Peter Williams The Organ Music of J S Bach PDF

This page intentionally left blank
The Organ Music of J. S. Bach

Second edition
This is a completely revised edition of volumes I and II of The Organ

Music of J. S. Bach (1980), a bestselling title, which has subsequently
become a classic text. This new edition takes account of the Bach
scholarship of the last twenty-five years. Peter Williams’s
piece-by-piece commentary puts the musical sources of the organ
works in context, describing the form and content of each work and
relating them to other music, German and non-German. He
summarises the questions about the history, authenticity, chronology,
function and performance of each piece, and points out important
details of style and musical quality. The study follows the order of the
Bach catalogue (BWV), beginning with the sonatas, then the ‘free
works’, followed by chorales and ending with the doubtful works,
including the ‘newly discovered chorales’ of 1985.
Peter Williams is an internationally renowned Bach scholar and

performer. He was Professor of Performance Practice and the first
Director of the Russell Collection of Harpsichords at the University of
Edinburgh, Arts and Sciences Distinguished Professor at Duke
University, NC, and until recently John Bird Professor at the
University of Wales, Cardiff. He has written numerous books on the
organ, organ history and organ repertoire. The first edition of The
Organ Music of J. S. Bach was published in 1980 (vols. I and II) and
1984 (vol. III).
The Organ Music of
J. S. BACH
............
Peter Williams
Second Edition
  
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo
Cambridge University Press

The Edinburgh Building, Cambridge  , United Kingdom
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521814164
© Cambridge University Press 2003
This book is in copyright. Subject to statutory exception and to the provision of

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2003
- ---- eBook (EBL)

- --- eBook (EBL)
- ---- hardback

- --- hardback
- ---- paperback

- --- paperback
Cambridge University Press has no responsibility for the persistence or accuracy of

s for external or third-party internet websites referred to in this book, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
First published 1980 as The Organ Music of J. S. Bach, Volumes 1 and 2.

Second edition (in one volume) 2003
Contents
Preface [page vii]

List of abbreviations [ix]
BWV 131a Fugue in G minor [1]

BWV 525–530 Six Sonatas [2]
Preludes and Fugues (Praeludia) BWV 531–552 [37]
Eight Short Preludes and Fugues BWV 553–560 [141]
Miscellaneous pieces BWV 561–591 [145]
Concertos BWV 592–596 [201]
BWV 597 and 598 [225]
Orgelbüchlein BWV 599–644 [227]
Schübler Chorales BWV 645–650 [317]
Chorales formerly called ‘The Eighteen’ BWV 651–668 [336]
Chorales from Clavierübung III BWV 669–689 [387]
Chorales formerly called ‘The Kirnberger Collection’
BWV 690–713 [429]
Miscellaneous chorales BWV 714–765 [453]
Chorale variations (partitas) BWV 766–771 [499]
BWV 790 [528]
Four Duets from Clavierübung III BWV 802–805 [529]
BWV 943, BWV 957, BWV 1027a and 1039a, BWV 1029.iii,
BWV 1079.ii, BWV 1085 [536]
Chorales now called The Neumeister Collection
BWV 1090–1120 [541]
Further works, in part of uncertain origin [575]
[v]
vi Contents
Calendar [583]
Glossary [585]
Bibliography [591]
Index of names [608]
Index of BWV works cited [618]
Preface
The organ works of Bach never cease to arouse ideas, and a revision enables
me to express a few more. While the text is now largely new, its style and
method still work towards framing questions rather than defining answers,
aiming to give the performer and scholar some bearings on a unique reper-
tory, one about which there will always be more to say. In this connection,
I found particularly heartening the commendation of an early reviewer of
the first edition (G. M. Leonhardt), who discerned that I had more ideas
than I ‘wished to lay down in print’.
Since the early 1970s when work on this book originally began, the
findings of Bach research have been published at such a pace that it has
become necessary to add new material and delete some of the original. The
outlines of this revision are:
1. Volumes I and II are now combined, omitting duplication but
now including the chorales first published in 1985 (so-called ‘Neumeister
Collection’). The original volume III (A Background) needs a separate re-
vision, taking in the results of current thinking on historical performance
and how it might contribute to an understanding of the music.
2. The listing of sources for each piece, already selective in the first edition,
is revised and avoids duplicating fuller information now found in:
the Kritischer Bericht volumes accompanying NBA IV
the second edition of Schmieder’s BWV (including the ‘Little Edition’
1998)
the Bach-Compendium, planned Werkgruppen J, K
In the sources as now summarized, I use the word via to suggest who it was –
as MS-owner, copyist or teacher – through whom certain extant copies
derived.
3. I have kept in mind a newer approach to the whole notion of ‘The
Complete Organ Works of Bach’, recognizing that this repertory is not fixed
and that editions may be giving unfair privilege to one version (perhaps a
chance survival) above another, presenting a uniform appearance unknown
to the composer himself, and neglecting works, right through to the Art of
Fugue, that suit organ as one of several keyboard instruments. Doubtless
too, transcriptions played a bigger part than is suggested by the Schübler
Chorales and the five extant concertos.
Much help in rethinking questions of authenticity is given by the on-
[vii] going work of Dr Reinmar Emans and Dr Ulrich Bartels (Göttingen), who
viii Preface
generously shared with me their researches so far on ‘doubtful’ works at-

tributed at some time or other to J. S. Bach. If the ‘Neumeister Chorales’
are the work of J. S. Bach, so must many another piece be, and Bach’s work
must have been at first indistinguishable from that of his local predecessors.
It must also have gone through more versions or variants than are now
known.
4. For several reasons the book still resists dating this music. First, there
is a reasonably clear, broad chronology to most of it; secondly, greater pre-
cision is won only by speculating from inconclusive sources and putative
resemblances to other music (hence the frequent disagreements amongst
writers); and thirdly, with living and changing works of this kind there may
be a misleading, old-fashioned positivism in the whole notion of trying to
pinpoint a particular moment in their life.
5. I have been at pains to refer to other composers in relation to J. S.
Bach, not least since these are now better served by editions and studies
than they were in 1973. It is clear to anyone closely studying any keyboard
works of Bach that he knew a great deal of music, doubtless far more than is
listed in current literature, and responded to it in various ways: music not
only of major composers – those most often commented on ever since the
Obituary of 1754 – but also of minor.
6. I have selected only certain sources concerning the history of texts and
melodies, partly because Lutheran hymnology is a major study in itself with
limited relevance to Bach’s settings, partly in order to give due weight to the
work of C. S. Terry, who still gives the organist many a useful hint.
7. This is also the place, perhaps, to acknowledge again the contribution
made to the study of Bach’s organ works by some earlier writers, especially
Philipp Spitta and Hermann Keller. Though not always known to musicians
today, their musical aperçus are imaginative and useful, worthy of consid-
eration whatever factual shortcomings they reflect and however many new
territories have since been explored.
In revising this book I have received particular help from Ulrich Bartels,
Mark Bighley, Lucy Carolan, Reinmar Emans, John Druesedow, David
Humphreys, David Ponsford, Tushaar Power, Penny Souster and Tim Taylor,
for which I would like to thank them warmly. Planning a full-length mono-
graph for which one is entirely responsible helps one to develop an in-
terpretation of a subject, and accordingly I acknowledge gratefully three
early associates at Cambridge University Press for the opportunity they gave
me a quarter of a century ago: Michael Black, publisher; Eric Van Tassel,
copy-editor; and †Peter le Huray, originator (with †John Stevens) of the
Cambridge Studies in Music. Peter le Huray proposed this study originally,
and in affectionate and regretful memory of him I would like to offer this
revised edition.
Abbreviations
ABB the Andreas Bach Book (MS Lpz MB III.8.4)

AfMw Archiv für Musikwissenschaft
AM Acta musicologica
Am.B. MSS in Amalienbibliothek (SBB): Princess Anna Amalia’s library
AMBB Anna Magdalena Bach Books (1722, 1725)
AMZ Allgemeine musikalische Zeitung
Bä Bärenreiter edition
BG Gesamtausgabe der Bachgesellschaft, 46 vols., Leipzig, 1851–99
BJ Bach-Jahrbuch
BR MSS in Brussels, Bibliothèque Royale
BuxWV Georg Karstädt, Thematisch-systematisches Verzeichnis der musikalischen Werke von
Dietrich Buxtehude (Wiesbaden, 1974)
BWV Wolfgang Schmieder, Thematisch-systematisches Verzeichnis der musikalischen Werke
von Johann Sebastian Bach (Leipzig, 1950; 2nd edition, Wiesbaden, 1990)
BzBf Beiträge zur Bachforschung
CbWFB Clavierbüchlein für Wilhelm Friedemann Bach
cf. compare
c.f. cantus firmus
Cons MSS in Brussels, Bibliothèque du Conservatoire Royal de Musique
Darmstadt MSS in Hessische Landes- und Hochschulbibliothek
DDT Denkmäler der Deutschen Tonkunst
Dok I Bach-Dokumente, vol. I, ed. Werner Neumann and Hans-Joachim Schulze (Kassel
etc., 1963)
Dok II Bach-Dokumente, vol. II, ed. Werner Neumann and Hans-Joachim Schulze (Kassel
etc., 1969)
Dok III Bach-Dokumente, vol. III, ed. Hans-Joachim Schulze (Kassel etc., 1972)
DTÖ Deutsche Tonkunst in Oesterreich
EB Edition Breitkopf (Breitkopf & Härtel)
EF Editions Fuzeau
EKG Handbuch zum Evangelischen Kirchengesangbuch = R. Köhler, Die biblischen
Quellen der Lieder, vol. I.2 (Berlin, 1964)
EM Early Music
EP Edition Peters
[ix]
x List of abbreviations
Grönland MS in Copenhagen, Det Kongelige Bibliotek

Hamburg SUB MSS in Hamburg, Staats- und Universitätsbibliothek
HE Hänssler Edition
HJ Händel-Jahrbuch
HK Berlin Hochschule der Künste, Berlin (formerly Hochschule für Musik)
KB Kritischer Bericht (Critical Commentary to NBA), here referring to the relevant
NBA volume
LBL MSS in London, The British Library
lh left hand
LM MSS in Yale University Library (Lowell Mason Collection)
Lpz Go. S MSS in Lpz MB (Sammlung Manfred Gorke: Gorke Collection)
Lpz MB Leipziger Städtische Bibliotheken, Musikbibliothek
Mf Die Musikforschung
MGG Die Musik in Geschichte und Gegenwart, 1st edn, Kassel (1949–79)
Mö MS the Möller Manuscript (SBB MS 40644)
MQ Musical Quarterly
MT Musical Times
MuK Musik und Kirche
NBA Neue Bach-Ausgabe. Johann Sebastian Bach. Neue Ausgabe sämtlicher Werke (Leipzig,
Kassel, from 1954)
NBG Neue Bachgesellschaft
NZfM Neue Zeitschrift für Musik
Ob the Orgelbüchlein
Obituary the ‘Nekrolog’, in Dok III, pp. 80–93
P MS scores in SBB (Partitur)
Peters Peters edition, see EP
rh right hand
RV Peter Ryom, Verzeichnis der Werke Antonio Vivaldis: kleine Ausgabe (Leizpig, 1974)
SBB MSS in Berlin, Staatsbibliothek zu Berlin, Preussischer Kulturbesitz, Musikabteilung
Schmieder 1950 see BWV
SIMG Sämmelbände der Internationalen Musikgesellschaft
St MS parts in SBB (Stimmen)
Stuttgart WL MSS in Stuttgart, Württembergische Landesbibliothek
Vienna Cod MSS in Vienna, Österreichische Nationalbibliothek
Washington LC MSS in Washington, Library of Congress
WTC1 The Well-tempered Clavier, Book 1
WTC2 The Well-tempered Clavier, Book 2
BWV 131a Fugue in G minor
Copies via J. C. Kittel (P 320 etc.)
This is a transcription of the last forty-five bars of the final chorus of Cantata
131 (1707), whose opening and closing movements are, unusually, a prelude
and fugue, the latter a permutation fugue of three subjects (Example 1). This
conforms to the tradition of choral permutation fugues (Krüger 1970 p. 11),
as in other early works: Cantata 196, the Capriccio in B major. Perhaps the
model is Reinken’s sonatas and through them ultimately Frescobaldi’s Fiori
musicali. Unlike the Passacaglia fugue, BWV 131a has no interludes, and
its many tonic cadences are typical of such fugues. After Frescobaldi, one
line in a permutation fugue was often chromatic, with influential examples
in Kuhnau’s Clavierübung II (Leipzig, 1692) and also Pachelbel’s Magnificat
primi toni, v. 19 (1701–5?), which has a chromatic fourth subject and coun-
tersubject much like b. 3 of Example 1.
J. S. Bach is usually thought not to be the arranger (Spitta I p. 451),
and as with BWV 539, details make it unlikely to be authentic: the sources
(many, but from a common route), certain unidiomatic moments, omission
or alteration of fugal parts, and little in common with the authentic early
fugues BWV 531, 549a. Lines impossible for two hands are omitted and
the bass simplified. The succinct ending, though also vocal, need not be
Bach’s (as Bartels 2001 suggests), but could be the work of an arranger such
as Kittel. The cantata’s ending was surely the original, i.e. with a gradual
buildup from two to five parts.
Example 1
[1]
BWV 525–530 Six Sonatas
Autograph: a section of the MS P 271. No title-page (fol. 1r left blank,

BWV 525 begins fol. 1v); each sonata headed ‘Sonata 1.[etc.]’, perhaps only
subsequently. Three staves. At end: ‘Il Fine dei Sonate’. A title-page was
written by G. Poelchau (1773–1836): Sechs Orgel-Trios für zwei Manuale
mit dem obligaten Pedal (‘Six Organ Trios for two manuals with obbligato
pedal’).
Sources
The first section of P 271 gives the earliest complete set of the Sonatas (Kilian
1978 p. 65), a special compilation of c. 1730 (Dadelsen 1958 p. 104)
or, allowing for the date-range of the watermark, c. 1727–30 (Spitta II
pp. 692, 797). In this manuscript as now constituted, the Sonatas, the
chorales BWV 651–668 and the Canonic Variations all originally began
with a page left blank, each presumably for a full title?
Such a set of sonatas might have been compiled for publication, cor-
responding to the set of harpsichord partitas issued in 1731, matching the
progressive chamber music of the late 1720s for the Collegium musicum,
and even employing up-to-date notation (three staves, tempo marks, some
slurs and dots). Both Partitas and Sonatas use the treble G-clef, although
earlier versions of movements in both sets had used the soprano C-clef: a
change made perhaps for the sake of publication. P 271 has more conve-
nient page-turns than other copies and may have been intended as printer’s
fair copy to be used in the engraving process itself. (Was the Six Parti-
tas autograph lost because it was so used? The advertisement for No. 5,
in Dok II p. 202, spoke of a seventh partita, which would have made a
volume comparable to Kuhnau’s Clavierübung: were the organ sonatas to
have been the original Clavierübung II, replaced, perhaps because they were
too difficult, by the present Clavierübung which included the or a seventh
partita?)
The fascicle structure of P 271 – two bifolia, a gathering of five sheets,
a gathering of three, a bifolium, a gathering of three (see Goldhan 1987) –
need not mean that work on compiling/revising so many earlier movements
was still in progress at the time of writing, but it might. From the makeup it
seems that BWV 525, probably the last to be copied, was at one point meant
[2] to follow BWV 529, thus giving the order BWV 526, 527, 528, 529, 525, 530.
3 Six Sonatas
Another feasible order is BWV 526, 527, 528, 525, 530, 529. Makeup and
rastrum-types suggest that BWV 530 was a separate work, perhaps the
first to be written down in this form, with its own gathering and (like
BWV 525) a blank first side – on which the last section of BWV 529 was
copied in making up the set. The keys of the Six Sonatas do not compel
one order rather than another, and the composer seems not to have num-
bered them at first, either in P 271 or even when he wrote some headings in
P 272.
P 272 is a copy made by W. F. Bach as far as b. 15 of Sonata No. 4
(pp. 1–36 probably direct from P 271), and the rest much more spaciously
by Anna Magdalena Bach (pp. 37–86, certainly direct from P 271). To judge
from page-numbers, Anna Magdalena’s copy was complete but her first
forty-eight pages were replaced by Friedemann; why is not known (Emery
1957 p. 20). Watermarks are those of vocal works copied 1732–35, implying
that her pages had soon been ‘lost’ (KB pp. 23, 31). It seems the composer
participated in, supervised, revised or at least knew about this second copy:
the headings of Anna Magdalena’s Nos. 5 and 6 are autograph, as probably
are movement headings, Italian terms and – importantly – most ornaments
and articulation signs (Butt 1990). Perhaps P 271 was complete when W. F.
Bach entered the University of Leipzig as a law student (5 March 1729), and
P 272 when he moved to Dresden as organist of the Sophienkirche (summer
1733). Had Friedemann used his copy much it might show more signs of
use – damage, added slurs – but probably all such fair copies were re-copied
for practical purposes.
Perhaps tempo marks were entered in the autograph only after they were
in Anna Magdalena’s copy, leaving the first movement of No. 1 without a
tempo mark in either copy. Or all six first movements of the Sonatas in P 271
originally had no tempo-mark, thus joining the Italian Concerto and most of
the harpsichord transcriptions BWV 972–987 in consciously reflecting one
particular Italian usage. Another Italian detail would be the appearance of
movements in 2/4: a new time-signature found also in the contemporary Six
Partitas (but not in earlier harpsichord suites) for movements with Italian
names, Capriccio, Scherzo and Aria.
The compilation was not certainly copied again complete before the
composer’s death, even by students such as Kellner, Agricola, Kirnberger or
Kittel, the last of whom probably made at least partial copies (see KB p. 56).
Copies of individual movements, by J. G. Walther or J. T. Krebs, can be much
earlier than P 271. Later copies made directly or indirectly from P 271 include
Am.B.51 (for Princess Anna Amalia in Berlin); Vienna Cod. 15528 (J. C.
Oley, after 1762?); and Nägeli’s print (Zurich, 1827). Others appear to come
from P 272, partly through Forkel or Baron van Swieten (string trios ascribed
to Mozart, K 404a), somehow reaching London for the Wesley–Horn print
4 Six Sonatas
of 1809–10. Oley’s MS shows signs of revision, authorized or not, as if being

prepared for circulation or even printing (KB p. 95). Some copies made
in the decades around 1800 still preserve the early or variant versions of
movements in Nos. 1, 4 and 5.
Origin and purpose

Although the history of the set of six ‘begins only with the writing down
of P 271’ (KB p. 15), some movements exist in previous versions while
others may not be original organ works, judging by compass or tessitura.
From corrections in movements known to be adaptations of music from the
Weimar period, P 271 suggests that the composer was collecting or at least
revising them there and then. A general survey gives the following picture
(Eppstein 1969; Emery 1957; KB p. 66):
composed for composed previously

the compilation for organ as transcription uncertain later
525 ii i? iii? iii?
526 i? ii? i?, iii?
527 i? ii? i iii ii?
528 ii iii i
529 i iii? ii iii?
530 i ii iii
According to such surveys, no two originated in the same way, and only
No. 6 was composed throughout as an organ sonata. Several movements
show signs of being altered to fit the classic organ-compass CD–d –c (see
KB pp. 64–5). No significance in the present order of keys has yet been found
beyond a ‘tones-and-triads’ sequence: C minor, D minor, E minor, C major,
E major, G major (Kilian 1978 p. 66) or C minor, D minor, E minor, E
major, G major, C major (Butt 1988 p. 89). Comparing Bach’s ‘sets of six’
suggests that the idea of key-sequence gradually evolved: a few years earlier
the Sonatas for Violin and Harpsichord had no clear cycle of keys, while the
newer Partitas for Harpsichord did.
The Sonatas’ purpose and even period were clear to Forkel (1802 p. 60):
Bach hat sie für seinen ältesten Sohn, Wilh. Friedemann, aufgesetzt,
welcher sich damit zu dem grossen Orgelspieler vorbereiten musste, der er
nachher geworden ist . . . Sie sind in dem reifsten Alter des Verfassers
gemacht, und können als das Hauptwerk desselben in dieser Art angesehen
werden.
5 Six Sonatas
Bach drew them up for his eldest son Wilhelm Friedemann [b. 1710], who
must have prepared himself by this means to be the great organ-player he
later became . . . They were made during the composer’s most mature age
and can be looked upon as his chief work of this kind.
Perhaps Friedemann himself told Forkel this, having been involved in key-
board works that did get published, including the variations Forkel con-
fidently associated with J. G. Goldberg. Whether the Sonatas were more
than practice music can only be guessed: instrumental trios were played
during Communion in some northern churches (Riedel 1960 p. 180), but
organ trios are not reported. Nor was Mattheson thinking of them when he
wrote that preludes could take the form of ‘little sonatas or sonatinas’ (1739
p. 472). Similarly, nothing is known of organ trios said by Forkel to have
been composed by Handel while a boy (see Kinsky 1936 p. 160).
Though no doubt some organists practised on other instruments with
pedals, Forkel included the Sonatas as ‘Organ Pieces’, as did the Obituary,
and he did not say ‘composed’ for W. F. Bach but ‘set’ (‘aufgesetzt’).∗ Both
the words ‘Trio’ and ‘for organ’ were usual in references to them, as in the
Obituary, and though nineteenth-century commentators began to equate
‘Clavier’ with clavichord and speculate that the Sonatas and Passacaglia were
for domestic music-making (Peters I, 1844), 2 Clav. & Pedal did not denote
pedal clavichord or harpsichord. By c. 1730, a C–c compass implied organ
exclusively, as was not so in c. 1710.
One curious detail is that since neither hand goes below tenor c, the pieces
‘can be studied on organs of only one manual and pedal’, with 4 stop and lh
down an octave (Klotz 1975 p. 377). This is equally so for the chorale-trios
BWV 655a, 664a (earlier) and BWV 676 (later), and commonly for trios by
younger composers in the same tradition. (A 4 stop for left hand on its own
manual, played an octave lower than notated, is suggested several times in
Kauffmann’s Harmonische Seelenlust, Leipzig, 1733.) The two techniques –
tenor compass, octave-transposing left hand – may together reflect how
trios were often played.
Several references, such as this of c. 1777, are full of admiration:
so schön, so neu, erfindungsreich sind, dass sie nie veralten sondern alle
Moderevolutionen in der Musik überleben werden. (Dok III p. 313)
so beautiful, so new and rich in invention, they will never age but will
outlive all changes of fashion in music.
Pupils writing trios include Friedemann himself (on ‘Allein Gott’) and, in
the 1730s, H. N. Gerber. Though J. L. Krebs is not known to have made a
copy of the Six Sonatas, his own sonatas are the works most obviously based
∗ Forkel’s
word ‘aufgesetzt’ may have come from Friedemann and ‘obviously means “composed” ’
(KB p. 15). But Forkel’s usual words for ‘composed’ were ‘componirt’, ‘gemacht’, ‘ausgearbeitet’.
6 Six Sonatas
on them: all except his C major Fugue movement are under their influence,
and were even perhaps student assignments in writing both traditional and
more galant invertible counterpoint.
Trio types in organ music

While no ‘direct models for these Sonatas . . . have been discovered’ (Emery
1957 p. 204), their form and texture were known in the Weimar period.
Organ chorales à 3 are more feasible than organ fugues à 3, and are found
in different forms by c. 1700.
Parallel to German chorales were the trios, trios en dialogue and trios à
trois claviers of various ‘good old French organists’ admired by J. S. Bach
(Dok III p. 288). Most examples by Lebègue, Grigny, Raison, Boyvin and
Clérambault have two manual parts above a continuo pedal, sometimes
imitative, but with a lot of parallel thirds etc. The Six Sonatas’ binary and
ritornello forms are as good as unknown. Quite distinct from the baroque
tinkles fashionable in the twentieth century are the French registrations
based on three 8 lines: manual I with mutation (e.g. Cornet), manual II with
reed (e.g. Cromorne) or 8 +4 , pedal 8 Flûte, all of which were possible on
Friedemann’s Silbermann organ in Dresden. Sometimes the Sonatas seem
to confirm that pedal was at 16 (e.g. BWV 527.iii, bb. 61–6), as the basso
continuo had also probably been in the cantata movement transcribed as
BWV 528.i.
Formally, however, French trios cannot have contributed much to the
Six Sonatas. Much closer is the invertible counterpoint of Italian sonatas
for two violins, already turned to good use above a chorale cantus firmus by
Buxtehude, e.g. Vers 3 of ‘Nun lob, mein Seel’, a chorale known in Thuringia.
Here the imitation is only partial, as in Italian trio-sonatas. Meanwhile, the
chorale-trio technique of a modest composer of Central Germany such
as Andreas Armsdorff (1670–99) relied very much on parallel thirds and
sixths, seldom with much drama. A trio such as ‘Allein Gott’ BWV 664a is
one kind of successor to this, with a cantus firmus, a chorale paraphrase and
an independent bass, of nearly one hundred idiomatic bars.
Dating BWV 664a to the later Weimar years and the slow movement of
BWV 528 to the earlier gives some idea of how quickly Bach developed form.
(Also, BWV 664a shows a creative leap from Cantata 4.iv, one that cannot be
matched in the work of other composers.) The Sonata has a basso continuo
pedal and two alternating themes, with two-bar phrases of immense charm
but arbitrary continuity; BWV 664a has a thematic bass, a full ritornello
shape and episodes with broken chords. But of itself, the octave imitation
of BWV 528.ii is no more an ‘early’ sign than is the opening homophony of
7 Six Sonatas
No. 2. On the contrary, the non-fugal openings to Nos. 2, 5 and 6 are a later
kind of music than the fugal opening of others.
While it is generally true that the three movements are like those of a
concerto, and the three parts those of an instrumental sonata, the music
is clearly geared to manuals and pedals. Irrespective of compass, the upper
parts would rarely be mistaken for violin or even flute lines. Moreover, as
Emery observed (1957 p. 207), passages in the concertos that may resemble
some of those in the sonatas (compare Concerto BWV 594.i, bb. 93ff. with
Sonata BWV 530.i, bb. 37ff.) are typical of neither. If the organ concertos
had any influence on the sonatas it would be more in their form and types
of episode.
Trio types in instrumental sonatas

The closest parallel to the Six Sonatas is works for solo instrument and
obbligato harpsichord. But though they all contain at least one fugal Allegro,
the instrumental sonatas differ in important details. The organ’s compass –
rh f–c (mostly c –c ) and lh c–c (mostly c ) – is obviously planned for
the convenience of two hands, and, as any would-be arranger soon learns,
the lines are not easily adaptable to other instruments. The upper parts are
always in dialogue, whereas in the chamber sonatas the rh is sometimes like a
continuo accompaniment. At times the pedal-lines look like a basso continuo,
and indeed the distinction is not clear-cut. Whoever made the arrangement
BWV 1027a did not merely simplify the bass line of the Gamba Sonata BWV
1027; each version of the bass line has independent qualities. A common
point between organ and chamber sonata is that no movement begins with
the theme in the bass.
Though the variety makes a summary difficult, the organ sonatas’ first
movements have developed a more concerto-like shape than the violin
sonatas, while the violin sonatas tend to have a more active bass line, with
rhythmic complexities not expected in an organ sonata. Yet they do point
in the direction of the organ sonatas, and together, the two genres survey all
trio techniques, forms and textures:
slow first movements (not in organ sonatas)

changes of tempo and form within a movement (BWV 528, 1030)
ritornello movements of several lengths and sections, fast or slow
ABA-ritornello movements, fast or slow, with or without fugally answered subject,
with clear or disguised return to A2
binary slow and fast movements, with or without full reprise of first theme
ritornello subjects homophonic or imitative (at the octave or fifth), with or without
subject in bass
8 Six Sonatas
movements in four or more parts, the keyboard homophonic or contrapuntal

(not in organ sonatas)
the three parts in various areas of the compass (organ sonatas less varied)
bass line imitative, or with countersubjects, or ostinato, or thinly written (last two
not in organ sonatas)
simple proportions (e.g. 1 : 1 in BWV 525.iii and 3 : 4 : 3 in BWV 527.i)
The three-movement structure is not the obvious ancestor of any classical

sonata-type but rather, in Nos. 5, 2 and 6, like that of Bach concertos with
fugal finales. The most important parallel between the Six Sonatas and
classical Sonata Form itself is undoubtedly the development-like nature of
some middle sections, or the treatment given the subject of No. 2’s first
movement. Typical of the fast movements is the three-section plan in which
the middle section modulates and becomes ‘unstable’.
The comprehensive variety of the eighteen or (counting BWV 528.i as
two) nineteen movements seems to be planned to show the medium’s scope.
The Six Sonatas are very concise, clear in form, less diffuse in texture than the
instrumental sonatas. They are almost miniatures and yet take the principle
of equality of parts so far that the opening unisons of No. 6 are not a sign
of immaturity but the opposite: a concerto-like tutti, its unisons one more
trio effect.
Some further characteristics

Though without looking like organ music, Telemann’s Six Concerts et Six
Suites (c. 17I5–20?) do at times point towards BWV 525–530. J. L. Krebs’s
galant melody and simple harmony also bow to Telemann – as the throbbing
bass of Example 3 (Krebs’ Trio in B flat) suggests when compared with
Example 2. Any tendency for upper parts in Bach’s Sonatas to become a
duet above continuo, as at the beginning of No. 2, looks new and up-to-date
because simpler, indeed galant. Many turns of phrase in the Sonatas have
no part in the language of organ chorales or fugues; the slow movement of
No. 3 is quite at home in an arrangement from Mozart’s period, and all of
them make feasible duets for harpsichord (KB IV/7 p. 15).
In their short phrases and question-and-answer openings, Nos. 2 and
5 have an unmistakable chamber-like or concerto-like quality. Telemann’s
or Fasch’s chamber works can occasionally aspire to a similar idiom, as is
clear from the transcriptions BWV 586 and 585, where it is the working-
out and the sequences that betray their origin. Although occasionally, as
in the last movement of No. 6, lines resemble a chorale paraphrase, mostly
the chamber-like melody is sparkling, charming, either witty or plaintive,
9 Six Sonatas
Example 2
strangely free of the conventional associations there are between words

and themes in the organ-chorales. Some slow movements encouraged a
species of melancholy admired by the younger composers such as J. L. Krebs
(see Example 4, BWV Anh. 46). This was part of the idealized italianism
pervading the Six Sonatas, from their themes (Vivace = more energetic
than Allegro) to their actual terminology (‘Sonata’, ‘Il Fine dei Sonate’ –
compare the ‘Il fine’ at the end of the Italian Concerto, published 1735).
Example 3
Example 4
The Sonatas make a world of their own, as distinctive and accomplished

as the first movements of Leipzig cantatas or the preludes and fugues of
WTC1. The two hands are not merely imitative but so planned as to give
a curious satisfaction to the player, with phrases answering each other and
syncopations dancing from hand to hand, palpable in a way not quite known
even to two violinists. Melodies are bright or subdued, long or short, jolly or
plaintive, instantly recognizable for what they are, and so made (as the ear
soon senses) to be invertible. Probably the technical demands on the player
also contribute to their unique aura.
10 BWV 525
BWV 525 Sonata No. 1 in E major

Further sources: published by A. F. C. Kollmann in An Essay in Practical
Musical Composition (London, 1799), plates 58–67; first movement with
pedal only to c , in doubtful copies, e.g. P 597 (a copyist for C. P. E.
Bach?); St 345, arrangement in C major of movements i and iii, for strings
(c. 1750).
Headed in P 271 ‘J. J. Sonata 1. à 2 Clav: et Pedal’; second movement ‘Adagio’,

third ‘Allegro’. For ‘J. J’ (Jesu Juva, ‘Jesus help’) see BWV 651, also in P 271.
The likelihood that this originated as a chamber trio in B major (KB

p. 67) has led to a hypothesis that there were four versions: (a) a cham-
ber work in B, (b) an organ trio of one or more movements, also in B,
(c) a ‘Concerto’ or string trio version as in St 345 and (d) BWV 525, with new
middle movement (Hofmann 1999). Any preponderance of short phrases
in versions (a) and (b) implies that they were much earlier than (d). Despite
its title, the outer movements of (c) have the same bass lines as those in
P 271, which seem made for organ pedals; the scoring of violin, cello and
bass is surely an ad hoc arrangement, with added slurs (see KB p. 73).
The form of BWV 525.i – as if binary, with some recapitulation in the
second half – could mean that the movement is relatively late. In form
and figuration the outer movements are so contrasted, while their opening
harmony and melody are so similar, as to suggest that the composer carefully
paired them, perhaps for some didactic purpose. On the possibility that this
Sonata was a late addition to the set, see above, p. 2.
First movement
The form may be outlined as:
A 1–11 tonic, lh opens

B 11–22 to dominant, rh opens
A 22–36 to F minor, rh opens; inverts parts from A, extends to
15 bars (to include pedal entry b. 29)
B 36–51 to tonic, lh opens
A 51–8 pedal opens; b. 53(halfway)–b. 58(beginning) = bb. 6–11
The effect is that of a ritornello movement with a second half beginning

clearly at b. 22, and the final A ending like the first A. However, there is no
clear solo/tutti contrast in the movement, since motif a – Example 5 (i) –
runs through all sections inversus or extended or diminished, combining
both with scale (ii) and arpeggio figures (iii), the latter of which has the
11 BWV 525
Example 5
function of a second theme (B above). In bb. 29f. and 51f. the pedal has
its own version of the theme, changing its second bar apparently more for
reasons of three-part counterpoint than to make it easier. Thus, section B
makes play with three versions of the motif (see Example 6) while section
A has more scales, at least in one of the voices.
Example 6
Such emphasis on motif is rather more typical of Bach’s Two-part than

his Three-part Inventions. An ABABA shape can be seen in the Three-part
Invention in A major BWV 798, in which B is also a countersubject to a
line derived from A (bb. 9, 21). Moreover, some of the lines of this Inven-
tion are themselves rather like those of BWV 525.i in their triple coun-
terpoint: compare both movements at b. 27. But despite the similarities,
there are important differences. The triple counterpoint of the Inventions
can be more complete (the bass-line is not limited by pedal technique), the
Sonata’s forms are usually clearer, and as so often, each genre is tuneful in its
own way. Cadential pedal points, pauses or breaks before the final cadence
are unknown in the Six Sonatas where, except for the early Andante of
BWV 528, cadences are very succinct even when homophonic.
Although the final pedal bar quotes the opening motif, the composer is
not using motifs idly. For example, the pedal figure of b. 1 is heard again
12 BWV 525
only considerably later (b. 22), and the triadic motif constantly changes
shape. The way it is worked is known in concertos, and pedal lines derived
from a simple motif (as in bb. 6–8) recall the way the dactyl rhythm of
the Third Brandenburg Concerto creates long lines. Though much slighter
than the Brandenburgs, the Sonatas are comparable in two ways: melody
is spun out until it reaches a well-paced cadence, and the opening motif
counterpoints another theme. (The Third Brandenburg has examples of
both of these.) Also, the movement has a theme working both rectus and
inversus against two other subjects (bb. 11, 17), as does at least one of the
Three-part Inventions (E minor, bb. 14, 25).
Talk of motifs, however, does not reach the charm, pretty turns of phrase
and unusual feel of this movement, neatly phrased and executed. Curiously,
Cantata 140 (1731) also begins with a triadic theme in E followed by a
C minor Adagio.
Second movement
Binary (12, 16 bars); fugal first theme A, second theme developing

motifs from it, to dominant; second half beginning with theme
inversus, returning for quasi-recapitulation in b. 22; ends like first half,
upper parts exchanged.
Although this is a classic binary form, with partial recapitulation, the pat-
terns are developed to make it unusually continuous. There is much play
with the a motif, either as first heard (pedal from b. 6) or inversus (all
three parts from b. 13), or as bits of it are used. See Example 7. Thus the
movement is essentially monothematic, its patterns variously shaped but
still recognizable. In fact, the whole of b. 2 is open to inventive treatment
and is traceable in many semiquaver groups throughout. In the same way,
the lyrical fugue-subject informs much of the pedal-line.
Example 7
Probably the pedal quotation in b. 6 is not a subject entry but the point
at which a melodious bass sequence begins (Example 8). (There is a sim-
ilar sequence of incomplete bass entries in another trio slow movement:
that of the Sixth Brandenburg Concerto.) Even for Bach, the bass line is
unusually well motivated, almost as if the movement were written above
13 BWV 525
Example 8
a pre-composed bass. Not only are there five allusions to the theme in the
bass but it is part of the triple counterpoint: bb. 4–7, 6–10 and 10–12, all
reworked later. All trochaic/iambic figures seem to come from the opening
bar, just as all semiquaver groups do from the next.
The beginning of the second half, with its incomplete inversion of the
melody, is the least tense moment in the movement, particularly as the
section begins without pedal, uniquely in the Sonatas. The continuity tends
to disguise the fact that at key junctures, other phrases could follow than
those that actually do. The ‘recapitulation’ at b. 22 is not so much a tonic
return as a dominant answer to the entry of the previous bar, and in b. 23 it is
grafted on to a passage from the original b. 4, not b. 2 as might be expected.
The passage flows, but is less inevitable than appears at first.
The conciseness means fewer episodes than in the chamber sonatas (cf.
finale to the C minor Violin Sonata) and less distinction between ‘first and
second subject groups’ (cf. first Allegro of the D major Gamba Sonata).
Mature binary movements are often basically monothematic, as in the
Gavotta of the E minor Partita. All these movements have points in com-
mon with BWV 525.ii, particularly binary form with partial recapitulation
and two halves ending similarly. Inverted subjects opening the second half
are found in earlier 12/8 gigues. In addition, the melody keeps a plaintive
quality no matter what theme each part is playing. Remarkably little in the
movement is in the major – notably excepting the first three and a half bars
of the second half – and on these grounds alone the Adagio is a foil to the
finale.
Third movement
Binary (32 + 32 bars – cf. Goldberg Variations, Aria); second theme

develops motifs from fugue-subject, to dominant; second half begins
with inversion, closes like the first half, voices exchanged.
Though similar in form to the Adagio, this has no recapitulation before the
final pedal entry (b. 57, a tonic repeat of b. 25). Each half approaches its
closing key only by step, the two much alike, the second partly exchanging the
voices of the first. The subject’s inversion in the second half is accompanied
by an exact inversion of its countersubject, an ideal not often achieved
(cf. Gigue of the E minor Partita).
14 BWV 525–526
Example 9
While the main subject is only superficially like that of ‘Jesus Christus
unser Heiland’ BWV 688, its treatment is just as varied as the chorale’s: see
Example 9. So the subject is developed∗ in a manner not unlike the first
movement’s, and the opening quavers give rise to various other patterns.
The semiquavers of b. 3 are also responsible for many another line in the
movement, while the countersubject might have led to a later sequential
figure (compare b. 4 with b. 17). Such ‘derivation’ is of a different order
from the play with motifs in the first two movements; the ton of the sonata
has changed, and the gaiety is unmistakable.
For all its brio, the movement is not without subtlety. The second half
mirrors the first in several ways, literally (number of bars), contrapuntally
(upper parts exchanged) and thematically (inversus subject, countersubject
and episode), with contrary scales working cleverly back to the tonic. The
pedal theme is also more complete than appears, since the manuals take
over its semiquavers (bb. 25–7) in what is one of the most tightly organized
and self-referential of all J. S. Bach’s binary movements.
BWV 526 Sonata No. 2 in C minor

Further sources: early-nineteenth-century copies of string trio arrange-
ments (once said to be made by Mozart) of movements 2 and 3 as a pair.
Headed in P 271 ‘Sonata 2. à 2 Clav: & Pedal’; first movement ‘Vivace’, second
movement ‘Largo’, third movement ‘Allegro’ (P 298: ‘Moderato’).
While no movement of the Sonata is preserved in other versions, the cor-

rections in the autograph, and its provisions for organ compass, suggest
that it had an earlier version (KB p. 36), the second movement perhaps an
arrangement of a chamber trio (Eppstein 1969 p. 23). Neither contradicts
the idea that Sonatas Nos. 2, 5 and 6 form two groups of similarly conceived
first and last movements:
∗ The
bass of b. 41 is altered in the absence of pedal e ; the passage could not go down an octave
(Emery 1957 p. 135) because of spacing, etc.
15 BWV 526
first movements: concerto Allegro, beginning as if tutti (non-imitative),

then ‘solo’ episodes; pedal basso continuo; closes with opening
paragraph repeated.
finales: tutti fugue, ‘solo’ sections, fugal middle section, final ritornello;
pedal with fugal line. A type similar to the fugal Allegro of the violin
sonatas.
Such three-movement sonatas suggest less a chamber sonata than a very

succinct ‘concerto’, with tutti/solo first movement and fugal finale. Had the
set of sonatas started with No. 2, as suggested by the makeup of the MS (KB
p. 74), it would have established a genre: a neo-galant first movement, a
cantabile second, a fugato third.
First movement
A 1–8 tonic
B 8–16 tonic
A 17–22 relative major
B 22–31 to G minor
A 31–8 G minor
B 38–71 development section: gradually towards tonic
A 71–8 first 8 bars
That such ritornello movements sustain continuity is undeniable, but sec-

tions could follow each other in other orders. Thus the passage built on
sequential trills is followed on its first appearance by B (b. 22), and on its
second by A (bb. 70–1), both natural, the first slipping in ‘unnoticed’, the
second dramatic after a pedal lead-in calling attention to the reprise. Thus
in each case, between the sequential trills and what follows, the composer
has formed a link appropriate to the following material.
A is homophonic, B imitative; A begins on the beat with a conspicuous
pedal bass, B and the episode use patterns beginning off the beat. All of them
invite imitation and are alike enough for it to be possible to find this or that
semiquaver group derived from them. Samples are given in Example 10.
While in outline this movement resembles e.g. the B minor Flute Sonata first
movement (Keller 1948 pp. 102–3), details are different. The Flute Sonata,
though with a somewhat similar Affekt, has a much less clear ritornello form
and a more complex final section. Remarkable in the present Sonata is the
last-but-one section, a ‘Development’, very original in idea and perhaps an
addition made as the movement was being written out in P 271 (Butt 1988
p. 84). Its details seem prophetic:
16 BWV 525–526
Example 10
38–46 G minor pedal-point: repeats broken chords like a concerto;

then refers to A (in 3rds), then paired quaver semitones. (Slurs
wanted as at the pedal-points in Concertos BWV 1064.iii and
1063.ii?)
46–54 ditto, C minor, upper parts exchanged
55–60 new imitation above pedal line developing original quavers
61–2 from A (bb. 3–4)
62–5 developing the opening motif of B, including its pedal rhythm
66–70 developing the trills and countersubject of b. 20, over rising
chromatic fourths
Treatment of the main theme in b. 42 is less like the usual motif-play than
the development section of classical Sonata Form. The theme in outline is
both complete and easily recognizable; yet its intervals are altered and its
character is much less forthright than in b. 1. Also, the use to which the pedal
of bb. 55–60 puts one of the main motifs is different from the intensive play
in such mature chorales as BWV 678: in the Sonata it is used to spin out a
sequence and to be recognized as such.
The tonic–dominant–tonic strategy is clear. Clearly the opening pedal
point of the section beginning at b. 38 – serving at once as interlude, develop-
ment section and a kind of cadenza – contrasts with the shifting harmonies
and bass-line of section B. There seem to be many allusions to the various
themes. Rising semiquavers, for instance, seem to refer back to b. 4, and it
is striking how different the semiquavers are from those in the first Sonata.
The lines of No. 2 are clearly designed for keyboard, both in the broken-
chord figures and the sweeping lines (e.g. bb. 44–6 lh). Perhaps the fluid
semiquavers led to the sudden quoting of a passage from A in b. 61 and of
a passage from B in b. 62, though searching out thematic allusions in such
effortlessly spun lines is more than faintly pedantic.
17 BWV 526
Second movement
This is a unique movement:
1–8 subject (rh), countersubject (lh), codetta; with a

basso continuo
9–19 ditto, parts exchanged; episode on codetta theme
( = sequence 1)
20–6 two episodes or new themes ( = sequences 2, 3), latter with
pedal’s simplified version of opening subject ( = sequence 3)
27–9 sequence 4
29–35 subject G minor; pedal continues sequence, rh new
countersubject
35–8 sequence 2 in G minor, parts exchanged
39–45 subject and countersubject from 29, now in C minor
45–8 cadence in C minor, then half-close to finale
The key-plan, E to mediant, is unusual and suggests something specially

composed for P 271, i.e. ‘to link movements 1 and 3’ (Butt 1988 p. 86). More
traditional structures like the slow movement of the C minor Violin Sonata
or the organ Prelude in C minor BWV 537 close in their tonic before the
half-close.
The unusual key-plan is hardly evidence that this is a transcription or
shorter version of another movement (as Eppstein 1969 p. 21 suggests),
nor can one easily see it as ‘improvisation-like’ (Schrammek 1954). Despite
its simple shape (ABABAcoda) the movement again treats note-patterns
inventively, around statements of a main subject written in unusually long
notes. The movement’s characteristically fertile array of motifs is shown in
Example 11. As elsewhere in the Six Sonatas, the order the motifs appear
Example 11
in seems decided on the spot rather than by the ‘demands of form’, and
indeed, the shape of the movement is difficult to follow. At two points
(bb. 32–3, 42–3) subject and countersubject contrive to produce an off-beat
stretto, and – as often elsewhere – the composer picks up the final motif
18 BWV 526
for the coda. The pedal is a masterly bass-line: now a coherent continuo,
now détaché crotchets, now phrased quavers. The movement’s opening has
an apparent simplicity not borne out by the rest of it. It may begin like a
Telemann trio but by b. 5 is already developing complicated figuration and
turning the patterns upside down.
Third movement
This shows the type of ‘concerto fugue’ (as in Nos. 5 and 6) at its simplest:
A 1–58 Exposition, two episodes, two futher entries

B 58–82 new subject, then episode (b. 75); 4-bar link to:
A 86–102 unison stretto, answered at fifth below; to F minor
B 102–26 as bb. 58–82, parts exchanged; ditto the 4-bar link
A 130–72 stretto at fifth, then a further fifth below;
137–72 = 23–58
The form is clear and the details ingenious, chiefly in that the stretto potential
of the main subject allows the theme to be variously exposed. Moreover,
the quaver tail of the subject (Example 12) is developed as episode (from
Example 12
b. 18), as countersubject (from b. 30), as coda (from b. 51) and as the link
(bb. 82, 126). This unassuming quaver phrase is found in various guises in
other Bach works: see notes to the C minor Fugue BWV 546. Note how the
pedal’s rising semibreve 5ths anticipate the manual stretto that follows on
each occasion. Particularly interesting is running B2 into A3, for the form
then approaches a da capo fugue.
In view of such ingenuity, it becomes clear that the composer has carefully
distinguished the movement’s two fugue themes in style and application as
far as continuity, provided by the pedal, allows. The first theme is long-
phrased, like an alla breve (staid semibreves, dactyl rhythms, crotchet bass),
and is answered in the pedal, with correct middle entries and a classical
countersubject with suspensions. The second theme is short-breathed, dis-
tinctly stile moderno (rhythmic, repetitive, perky), with a basso continuo, a
lively countersubject vying with the subject, and a subsequent episode tend-
ing to galant simplicity. The first also modulates far less than the second,
and its entries slip in less conspicuously. The differences between two
19 BWV 526–527
fugue-styles are thus explored – but also dovetailed in a manner that suits
each.
So the three movements present three kinds of music: a concerto Vivace
with lively rhythms, a lyrical Largo (lines rise only to fall again), and a
chamber-music Allegro with old and new fugues. A passage like Example 13
may well have been heard by pupils as the newer idiom to imitate.
Example 13
BWV 527 Sonata No. 3 in D minor

Further sources: ‘early version’ in P 1096 (late eighteenth century) and Lpz
MB MS 1 (J. A. G. Wechmar, after 1740), both entitled ‘Sonata I’; ‘early
version’ of first movement only, in P 1089 and Lpz MB MS 7 (via J. N.
Mempell, before 1747); an ‘original manuscript’ owned by C. P. E. Bach
(BJ 79 p. 75); late copies of Adagio arranged for string trio (K 404a attrib.
Mozart, see Holschneider 1964); St 134, parts for a version of Adagio in the
Concerto BWV 1044.
Headed in P 271 ‘Sonata 3 a 2 Clav. et Pedal’; first movement ‘Andante’

(added after P 272 was made?), second movement ‘Adagio e dolce’ (‘dolce’
added? – KB p. 28; only ‘Adagio’ in P 1096), third movement ‘Vivace’.
‘It can be assumed that P 1089 and P 1096 are derived from a lost auto-
graph . . . written before 1730 . . . one of the sources from which P 271 was
compiled’ (Emery 1957 p. 90). Although the versions differ only in details,
the title ‘Sonata I’ might indicate an earlier plan to start the compilation
with it, and the impression it gives is of a work earlier than No. 2. That the
whole sonata ‘originated as a compilation or/and transcription’ (Eppstein
20 BWV 527
1969 p. 24) is suggested by the bass line (rewritten for pedals?) and by the
fact that in P 1089, the lines look as if they have been scored up from parts,
perhaps before 1727 (KB pp. 74–6).
P 271 shows the slow movement to have had its pedal in b. 4 altered
to avoid notes above d , but neither version of this movement seems to be
the source for the other. It is a model binary slow movement adding to the
variety surveyed by the Six Sonatas, while the organization of the first and
third movements is rather unusual.
First movement
Andante for a 2/4 movement could be a caveat (‘not allegro’), just as allegro
could be for the 2/4 finale of the Concerto in D minor for Three Harpsi-
chords, a movement more than faintly similar to this (‘not presto’). On 2/4
metre, see above, p. 3.
A 1–24 quasi-fugue above continuo bass, followed by coda

24–48 subsidiary material; 33–48 as 9–24, upper parts
exchanged
B 48–56 new theme in imitation; refers back (see b. 21 for 51, 55)
56–60 second sequence, using motif and bass from b. 1
61–4 third sequence, cf. 29
65–8 fourth sequence, cf. 21
68–72 fifth sequence, cf. 24
73–6 sixth sequence, cf. 16
76–88 opening section of B up a fourth, upper parts exchanged
89–92 pedal point, rh reference to motif from 4
92–6 seventh sequence, as 4 and 36 but in closer imitation
97–104 eighth sequence, corresponding to 17–24 and thus 41–8
104–8 ninth sequence; developed from 24 (cf. fifth sequence)
109–12 phrygian cadence decorated with previous motifs;
link to:
A 113–60 repeat of 1–48
Of particular interest is the middle or development section, which soon

turns almost exclusively to previous ideas, running from one to another
in an apparently arbitrary way through keys not fully represented in the
outer sections. While an ABA in such proportions (48 : 64 : 48 bars) may
be exceptional, and the work thought inferior to the others (Keller 1948
p. 105), its development section is full of significance, with its literal quo-
tation, series of themes, and display of motifs. Its technique is particularly
apt for organ trios, with their near-identity of upper parts.
21 BWV 527
Though the movement as a whole suggests little if any tutti/solo con-

trast, and certainly no dynamic changes of registration, its subtleties imply
that while simpler to play than some of the others, it is no early work.
Beginning both subjects on the mediant (b. 1, b. 48) is unusual; but more
significant is the constantly varying lengths of phrase, from the long opening
line down to the half-bar sequence of b. 29. Stretto within the first subject
(Example 14) is not so much a conventional fugal imitation as a device for
combining motifs.
Example 14
Also, the little anapaest from b. 1 crops up in very different contexts later.
The semiquavers’ potential for extension, sequence, and imitation from
b. 2 on is already familiar from early preludes and fugues. The pedal shows
a high degree of organization in depending on only a handful of ideas: the
detached quavers (b. 1 etc.), the short scale-like line (b. 8 etc.), the italianate
sequence (b. 24 etc.) and so on. The most interesting development is the
demisemiquavers, since from them come the subject codetta (b. 8), parts
of a countersubject (b. 12), and a kind of constant leitmotif. The fact that
A itself is ternary in multiples of eight – bb. 1–24, 24–32, 33–48 – gives the
movement a rounded form matched by its constant back-reference.
Second movement
Binary (8, 24 bars); contrasting themes (one in thirds, one more

imitative); second half with first theme, then new themes; reprise at
b. 21, followed at b. 23 by two previous bars ( = bb. 11–12); reprise of
first section.
So this binary form has elements of a ternary, a procedure not usually so

clear-cut in Bach, although both the E and G major sonatas have slow
movements of a similar cast (Schrammek 1954 p. 24). The ‘reprise’ is not
straightforward: two of the subjects appear with exchanged parts (b. 21 =
b. 1; b. 29 = b. 5), but between them is material from elsewhere, conforming
to the Six Sonatas’ technique of varying the order in which themes return.
Bar 26 is not a simple direct reprise of b. 3, since its rh line is an answer to
the lh; and the coincidence of pitch is of less moment than the chromatic
complexity of bb. 25–8.
22 BWV 527
The movement hangs on a succession of two-bar phrases, every one with

a new idea, at bb. 9, 11, 13, 15, 17 and 19. Of these, bb. 9, 13 and 19 have
been heard previously, as perhaps have bb. 11, 15 and 17. Is the descending
line of b. 3 to be heard decorated in b. 15 and b. 17? Is b. 1 as closely related
to bars 21 and 51 of the first movement as it seems?
The galant touch in the opening bars is rather belied by the rest, but the
rubric ‘e dolce’ seems to invite flute stops, while its thirds and appoggiaturas
could have seemed to whoever it was who copied out the ‘Aria’ BWV 587
(q.v.) to be the work of the same composer. As has been pointed out (Eppstein
1969 p. 24), the pedal line at b. 4 looks as if it started life elsewhere, since it
breaks the line and conforms less to b. 3 and b. 27 than might be expected.
But the original could have been an organ movement in a different key.
The version in the Triple Concerto BWV 1044 is more like a continuo bass
line at this and other points. It is ‘reasonable to suppose that the concerto
version is the later of the two’ not only because ‘it is more highly organized’
without repeats (Emery 1957 p. 122), but because its fourth part consists
of simple, easily added arpeggios. The binary–ternary form is typical of the
Six Sonatas’ advances in form.
Third movement
Like the finales of Nos. 2, 5 and 6, this has elements of the da capo fugue,
here with basso continuo rather than thematic pedal:
A 1–16 fugal exposition above continuo bass

17–25 subsidiary material, sequences
25–36 as 9–15, upper parts exchanged; short coda
B 37–60 six 4-bar phrases: invertibility, imitation, sequences
61–72 main subject as in 25–36, upper parts exchanged
73–96 six 4-bar phrases, motifs as before; refers to subject 73,
77 and countersubject in 81
97–108 main subject (decorated), as in 61–72, parts exchanged
108–44 nine phrases, mostly in 4 bars; motifs as before;
117–28 = 45–56; 133–40, see 37–44; new sequence;
141–4 = 57–60
A 145–80 Repeat
The parallels to the first movement are striking, though that is more concise.
Here there is scope for expanding the episode’s triplets. From the first episode
on, figure after figure follows, alike but varied and versatile: one almost
suspects the composer of seeking as many triplet-shapes as he can find.
Apart from the brief developments of the subject in b. 73 and b. 77, they are
absent from entries of the main subject, which therefore stands out rather in
23 BWV 527–528
the manner of a rondo. Instructive for the bar-by-bar process are the middle
entries at b. 61 and b. 97, as triplets spill over them.∗
It is characteristic of this movement that the ‘countersubjects’ to the
triplet figures are usually leaping quavers or tied notes (sometimes both): a
deliberate difference, underlining the old distinction between passus (steps)
and saltus (leaps). An unusual unifying factor is provided by the pedal,
particularly its repeated notes accompanying more than the fugue subject,
and the composer can introduce what figures he likes at any one point.
Since they are related, each triplet may be exchanged for another if compass,
spacing or harmony require it, and their shape can change. The lines become
reminiscent of Italian string sonatas whenever there is close imitation (e.g.
bb. 45, 108).
BWV 528 Sonata No. 4 in E minor

Further sources: Lpz MB MS 4 (J. A. G. Wechmar) and late copies; ‘early
version’ of first movement in Cantata 76; ‘early versions’ of the second
in Lpz Go. S. 311 (c. 1750?, in D minor) and later; in P 288, first thir-
teen bars of third movement appear at the end of the Fugue in G major,
BWV 541.
Headed in P 271 ‘Sonata 4 a 2 Clav: et Pedal’; first movement ‘Adagio’, and

‘Vivace’ in P 271 (not in P 272) and P 67 (Cantata 76); second movement
‘Andante’; third movement ‘Un poc’ allegro’ (also thus in P 272).
The sonata ‘is to all appearances a compilation of an instrumental sinfonia,

an early but rewritten organ piece and a later piece written for the Weimar
organ’ (Eppstein 1969 p. 24), and corrections throughout P 271 suggest the
rewriting to be still in progress. But though deriving from an earlier version
(KB pp. 41, 84), the last movement need not have been composed for the
Weimar organ – arguments from compass are inconclusive – or intended
by the composer to be part of the G major Praeludium. The last hangs on
the reliability of P 288 and has no other support.
The first movement is a scored-up version of the parts for oboe d’amore,
tenor viol and continuo of the Sinfonia in E minor opening Part II of
Cantata 76 (1723), though whether it derives from this directly or from a
transcription already made is uncertain. The autograph score of BWV 76
has enough corrections to suggest it to be the first form of the movement.
P 271 makes allowance for manual compass (left hand above tenor c), alters
∗ Since the triplets are significant for the conception of the movement, perhaps non-triplet rhythms
(e.g. b. 73, b. 77) should not be made to conform, despite the apparent ‘common sense’ of doing
this (Emery MT 1971 pp. 697–8).
24 BWV 528
some figuration, and gives the pedal a sonata-like basso continuo, here to
c only (e in the cantata). Both the slow introduction and the brevity of
the Vivace are exceptional in the Six Sonatas, and since the Vivace begins
uniquely with the left hand in a low tessitura, first impressions are unusual.
The middle movement exists in an early form in D minor, printed in
Peters I from a lost source and in Novello V from P 1115, and known in
yet a third version, none of whose copies dates from before 1750. It may
have been one of the ‘35 Organ Trios of J. S. Bach’ circulating as a set after
the composer’s death (see BWV 583). Its version in the autograph P 271,
whether or not made for this Sonata, is a unique contribution to the genre:
the short phrases are planned to be invertible (unlike the trio sections of
the early chorale BWV 739), and the chain of trills in bb. 36–7 is an early
anticipation of others in the C minor and D minor Sonatas∗ and even the
Musical Offering. Whether it was a trio composed specially for an organ with
pedal e , as often claimed, depends on whether the composer always kept
practical circumstances in mind. As in the Toccata in C major, the pretty
repeated Neapolitan sixths suggest an early date, and ‘c. 1708’ (Emery 1957
p. 102) is not implausible.
First movement
The form is unique (there is no double barline between the Adagio and
Vivace in either P 271 or P 272):
Adagio fugal exposition (modified bass, b. 3); b. 3 lh’s countersubject

not in cantata; accidentally (?) similar to subject of middle
movement
Vivace Imitative, in concise ritornello form: 5–13, 16–24, 31–9,
subject answered at 8ve 14–15, 25–30, 40–75, derived
episodes, final coda
Why this should be called a ‘French overture’ (Neumann 1967 p. 96) is

unclear.
The unusual form of the Vivace is as striking as its having an Adagio
prelude. Octave ‘fugal’ answers, which tend to continue through the move-
ment once they have begun, are not uncommon in the chamber sonatas’
slow movements (Sonatas in A major and G major for Violin and Harp-
sichord, etc.), and recur later in this Sonata’s second movement. The coda
from b. 61 looks as if in other circumstances it could become an imperfect
cadence, but here it ends brusquely with an italianate formula complete with
hemiola, rather simple for such a movement. Like the four-bar prelude, this
∗ In the present movement, the series of ornament signs are never elided (bb. 16, 73, 97, 128, 152) and
therefore ‘apparently do not mean chains of trills’ (KB p. 105) – a doubtful conclusion.
25 BWV 528
italianate cadence may have reached Cantata 76 via the sonatas prefacing
Buxtehude cantatas, rather than direct from Corelli.
All three lines of the Vivace – subject, countersubject, bass – have a
vivid melody and line rarely surpassed in the Six Sonatas. The characteristic
features of both subject and countersubject may well be seen as arising from
the special qualities of the viol: see Example 15, the second part of which
implies the crescendo natural to many passages in the Gamba Sonata in
D major. All three lines also have a high potential for generating motifs,
Example 15
as in bb. 13–15 (first bar of countersubject) and bb. 25–9 (plus first bar of
subject). These dominate the long episode from b. 40 onwards, including
the shortened entries at bb. 50 and 53. In the cantata version, the bass line
at b. 5 relies on crotchets, with the result that in the organ version crotchet
and quaver patterns are more systematically contrasted.
Using more notes in the pedal part than in the basso continuo of Cantata
76 suggests that the composer was compensating for the organ’s inability to
convey the natural tension of viol phrases. In the process, the pedal line gains
at least one important motif (b. 5), of which the composer makes curiously
little use: at the comparable point in b. 29, the autograph appears to show
an alteration. Nevertheless, bass and subject produce two-part counterpoint
typical of J. S. Bach, rich in accented passing-notes and appoggiaturas so that
the final notes of many bars are momentary discords. Such details render
the final cadence even more strikingly conventional, as too it often was in
some fugal movements of italianate sonatas by Handel. The final three bars
are very cramped in P 271, but follow the cantata parts.
Second movement
A1 1–11
subject a answered at unison, countersubject b; 2-bar
episode based on b; a plus octave answer, in dominant
B1 11–23 sequential imitation, lines derived from b?; cadence
for:
A2 24–8 as 7–11 in E minor
B2 28–38 as 11–23 but to G; continues as before, up a fourth
Coda 38–45 back to B minor (new material), then a in stretto
before final entry plus countersubject; interrupted
cadence
26 BWV 528
The little demisemiquaver slide of the countersubject can be heard in a

range of subsidiary themes. Equally striking and original is the main theme
itself, one of those early short melodies of Bach whose touching two-bar
phrasing would be tedious in a minor composer. It remains unaltered even
in imitation and stretto, so that the movement could be said to under-
line this phrasing throughout. One result of this is that harmonic devices
like the Neapolitan 6th become both predictable and wonderfully fresh:
see Example 16. (For a note on Bach’s early Neapolitan sixths, see also
Example 16
the Passacaglia.) The many perfect cadences might be ‘reminiscent of the

Legrenzi Fugue’ BWV 574 (Emery 1957 p. 101) but they also deliberately
emphasize the phraseology.
Descriptions of formal details cannot express how winsome this move-
ment is, though from the so-called early versions one sees how it evolved.
The figuration in the Peters I and Novello V appendices is simpler and
seems to show a maturing sense of melody: Example 17. Cadences and
Example 17
phraseology in the ‘early version’ are made less abrupt by some subtle
additions:
27 BWV 528
early version b. 5 becomes bb. 5+6 in Sonata version

b. 21 becomes bb. 22+23
b. 28 becomes bb. 30+31 (first half)
The ‘final version’ thus further underlines the two-bar phraseology. Its extra
passing-notes also render the melody more continuous. In the earlier ver-
sions the coda trills in b. 38 had been integrated with what had gone before,
and consequently, the effect now is more striking. But this final version has
also lost some invertibility: from b. 31 to the stretto in the coda, the parts
stand as they did before, but in the ‘early version’, B2 was not such an exact
repeat of B1. The left hand of P 1115 is unusually high, especially in the
(authentic?) key of D minor, with the two hands closer throughout than is
often the case in the Six Sonatas.
Third movement
I 1–28 exposition (subject A) complete with pedal subject

II 28–36 episode developing triplets
36–51 entry and answer in relative major; counterpoint as
in A
51–60 episode developing triplets, including one from A
(b. 16)
I 60–87 exposition; pedal subject, parts exchanged;
60–75 = 1–16
Coda 87–97 two 5-bar sections ( = episode bb. 28ff.), invertible
parts
As the left-hand column shows, the shape could be seen as ternary, the
outer sections similar to a concerto tutti (Eppstein 1969 p. 19). The extract
of it given with the Praeludium in G major in P 288 is not long enough to
show that this is a rondo fugue with regularly returning subject but without
second subject:
A 1–16 subject A, answered fugally

B 16–20 sequential episode
A 21–8 subject A, pedal
C 28–35 sequential episode
A 80–7 subject A, pedal
C 87–97 coda
28 BWV 528
The fugue-subject is of particular interest, being one of several Bach themes

in E minor, from the Toccata BWV 914 to the mature Fugue BWV 548,
that paraphrase the descending chromatic fourth (E D D C C B) in a
lively manner. The larger E minor Praeludium of Bruhns begins with a
flourish paraphrasing the same notes (see Williams 1997 pp. 95–8), which
also inform the theme of No. 6, middle movement. Here, the paraphrase
gives the impression of a minuet, indeed more dance-like than many another
chromatic minuet of the eighteenth century.
The triplet figures extend those already familiar in the finale of No. 3, now
also characterizing the subject entries. Some of the same melodic elements
can be seen in the organo obbligato part to the aria ‘Ich wünsche mir’ of
Cantata 35 (1726), although there the 3/8 is presumably slower than here.
The triplets are those of standard German variations – compare b. 9 with
No. 3 of Handel’s Variations in E major, HWV 430 – and their versatility can
be seen by comparing any two entries, where they accompany the subject
and become its countersubject, to an extent not common in the fugues of
WTC1. The aspect given the entry in b. 60 is new and unexpected, because
the triplets are dispersed between right hand and pedal. See Example 18.
Unlike most of the triplet figures in the finale of the D minor Sonata, several
of those here suit alternate-foot pedalling.
Example 18
The subject itself is without triplets save for b. 3. This probably suggests
that bb. 7, 15, 42, 50, 66 and 74 should remain paired semiquavers, while
apparently comparable moments at bb. 27, 86 should be played as triplets.
In P 272 the motif is dotted only in bb. 42, 50 and 74, but despite the claim
that such dots represent ‘not falsifications but rationalizations’ (Emery 1957
p. 75), the problems of inconsistency and ambiguity remain for this move-
ment (see KB p. 32). The most systematic answer would be to keep the
distinction between the two different patterns of b. 7 and b. 8, and to make
the dots of b. 25 etc conform to the triplets of the second of these. There
are in fact two different motifs in a continuous, unresting motion compa-
rable to the finales of some chamber sonatas, such as the Gamba Sonata in
G minor.
29 BWV 529
BWV 529 Sonata No. 5 in C major

Further sources: Lpz MB MS 1 (J. A. G. Wechmar, later eighteenth century)
entitled ‘Sonata 4’; ‘early version’ of second movement only, in Lpz Go.
S. 306 (J. T. Krebs c. 1725/6?), a Stockholm MS (J. C. Vogler – KB p. 53),
LM 4718 (J. G. Walther, from Vogler), P 286 (J. P. Kellner); this movement
associated by Walther, Vogler and Kellner with the Prelude and Fugue in
C major, BWV 545.
Headed in P 271 ‘Sonata 5. a 2 Clav: et Ped.’; first movement ‘Allegro’, second

‘Largo’, third ‘Allegro’.
The C major Sonata may have had its outer movements composed when the
set of Six Sonatas was compiled, while the middle movement seems to be an
earlier work, to judge by copies made by Weimar organists (Walther, Krebs,
Vogler). P 271 also has numerous corrections throughout, as if it were still
showing work in progress.
The sources imply better authority for an interlude in the C major Prelude
and Fugue than in the G major Praeludium (see BWV 528.iii). However,
since Walther and Vogler have the Largo only after the Fugue, one has to
(i) suppose a lost autograph of BWV 545 in which the composer cued
something somewhere (KB p. 86) and (ii) explain why J. T. Krebs copied the
movement separately.
First movement
A 1–17 tutti with question-and-answer phrases; scale sequences

17–32 parts exchanged; scale sequences altered to return to:
32–46 developed motifs from main tutti; pedal points; inverted
46–51 coda, scales from 12–17, upper parts exchanged
B 51–68 new fugue subject; answered at fourth, third and octave
68–84 alternating motifs from both main themes, then fuller
statement of first theme in F, then A minor
84–105 as 51–72 in A minor, parts exchanged; answer in 87
altered to produce D minor, then C (G in 55, F in 72)
A 105–55
B is continuous, and these bar numbers do not indicate distinct sections; it

begins fugally but becomes a development section. Throughout the move-
ment it is the main theme A which reappears to mark a new section.
The movement differs from the first of the C minor Sonata in that its outer
sections contain passages of ‘development’ – in particular, the pedal points
30 BWV 529
above which fragments of the main theme are heard. There are important
symmetries. Despite the ABA shape, the main theme returns conspicuously
almost halfway through, while A itself is symmetrical in subject matter if
not in bar numbers:
b. 1 statement b. 17 statement b. 33 statement

b. 9 pedal point b. 25 pedal point bb. 35, 42 pedal point
b. 12 scale sequence b. 28 scale sequence bb. 39, 46 scale sequence
B too is symmetrical, itself a kind of ABA.

Of all the Sonatas’ fast movements, this seems especially close to the
bright idiom of instrumental sonatas. If it were not for compass, the style
would suggest a sonata for two flutes and continuo. However, the spacing
and succinctness are typical of the Six Sonatas, and seldom outside the organ
works are motifs so developed – intricate despite the charming melody and
formal symmetries. The simple quavers marked a in Example 19 (i) not
only lead to direct derivations (ii), but can be heard in other figures (iii).
Clearly, the quavers also suit pedal, which is like both a continuo and a
derived counterpoint.
Example 19
The theme of the middle section seems related to the original semiquavers
of b. 1, though not as the result of arid calculation. Not the least memorable
moments are the pedal points, almost as if this was a galant movement
in classical sonata form. There is no over-use of motif, and even the scale
sequences are derivative only in general terms. But when two bars with the
same bass line are compared – e.g. b. 14 and b. 28 – it is clear that much
thought has gone into the motifs. The concentration of motifs in b. 32 is in
fact unusual in J. S. Bach and may have been intended more to create good
organ lines than to generate theoretically ingenious complexes.
Second movement
A 1–13 subject, chromatic countersubject (from b. 4 of subject);

sequences partly from both; a tonic cadence for:
B 13–21 second subject group, tonic; invertible counterpoint
(cf. first subject of first movement); sequential patterns
A 21–33 subject answered, upper parts exchanged; relative major
(avoids the chromatics); 29–32 = sequence 15–18
31 BWV 529
B 33–41 altered, in dominant of D minor, where upper parts then

exchanged (35–8 = 9–12); modulates back to:
A 41–54 1–12 repeated, plus countersubject 41–4; phrygian
cadence
Since the central sections alternate their components, the form is close to da
capo in which the middle begins independently but soon refers to previous
material. The whole contains elements of fugue, ritornello and da capo, all
achieved by means of two parts in dialogue above a basso continuo, and at
the same time conveying a distinctive and touching Affekt tending towards
the quasi-melancholy ‘sensitive style’ of younger composers. P 271 slurs
only the affettuoso appoggiaturas – thus the quavers of b. 1 etc but not the
theme’s opening gesture.
The typical sequence of b. 8 (and b. 47) involves a diminished fifth; cf.
similar moments in the Fantasia of Harpsichord Partita No. 3 (bb. 66–70 lh).
Meanwhile, the shapes taken by four demisemiquavers seem unlimited, each
an example of ‘varied figures’ taught by theorists (Walther 1708) from which
incomparably long lines are now generated. Different movements employ
different techniques: this Largo is an example of ‘generating cells’, while the
first movement of No. 1 has a single motif with a single shape bending to
different contexts. In both, the music is very complex at the note-by-note
level, more creative even than the Ob, where the chorale-melody governs
the direction taken. In this Largo, the theme itself is without motif-cells,
and its lyrical melody returns each time as a simple unmissable statement.
Third movement
As in Nos. 2, 4 and 6, the pedal participates in the fugue, though only the
opening notes of the theme are fit for pedal. As in No. 2, both subject and its
treatment are conventional, rather similar to the fugue in Corelli’s Sonata
Op. 5 No. 3 and also the A Fugue WTC2. In this way the movement contrasts
with the ‘modern’ first movement.
A 1–29 subject (in dominant, 9) with countersubject, above a

continuo bass; subject caput in bb. 21f. (sequence), 23f.,
25f.
B 29–59 new tonic subject, octave answer (again, 41); first subject
(A minor), countersubject; coda (51) combines both
capita
A 59–73 coda; stretto first subject, then episode from b. 13
A 73–119 development, minor keys; 73–89, first subject altered (73,
83); 89–97, entry with octave answer; episode; first subject
B 119–49 as 29–59 a fourth up, upper parts exchanged
A 149–63 coda as 59–73 (cadence altered), upper parts exchanged
32 BWV 529
This ingenious form serves as yet another example of modified binary

structure:
1–73 A, B, coda 1 (dominant)

73–163 A2, B2, coda 2 (tonic)
in which A2 is a development. Thus although it is as fugal as the finales of

Nos. 2 and 6, the movement is categorically different. The Sonata serves as
a complement to No. 2 in all three movements, in particular those with da
capo (C minor last, C major first) and those with developments (C minor
first, C major last).
Despite its conventional subject, the movement develops in a manner
quite typical of the Six Sonatas: bright, extrovert, tuneful, restless, intricate.
The pedal is especially instructive, the manual semiquaver figures espe-
cially inventive. The caput sequences of bb. 21–6 and 51–9 anticipate the
finale of No. 6 (bb. 8–13), and the same motif is taken effortlessly into a
longer line: Example 20. While the second theme appears rather sparingly,
special use is made of the opening notes of both themes, with the square
two/four-bar character of the subjects either emphasized (e.g. stretti begin-
ning in b. 83) or undermined (e.g. stretti beginning b. 59, six-bar cadence
bb. 67–73). The lively continuity is aided throughout by the tied notes and
suspensions typical of the first subject (though not the second) in all three
parts.
Example 20
The idea that this Sonata consciously emphasizes the natural hexachord
(CDEFGA – see Zacher 1993) has been overstated, perhaps, in seeking to
show that the slow movement has cadences on all these notes but out of
order. What other keys is a movement in A minor, or C major, likely to
modulate to? Also tenuous is the idea that its theme alludes to B A C H.
But as with other C major works of Bach, the player does feel a certain
elemental quality in this key, as if its basic musical figures (scales, broken
chords, triads, chromatics) have a distinct personality and every accidental
is telling. And there is undeniably a hexachordal flavour in a fugue-subject
that derives from six notes in C major, as there is too in the opening fugues
of both WTC1 and WTC2.
33 BWV 530
BWV 530 Sonata No. 6 in G major

Further sources: late copies only.
Headed ‘Sonata 6. à 2 Clav: e ped.’; first movement ‘Vivace’ in P 272 (prob-

ably autograph), not in P 271; second movement ‘Lente’ in P 271, third
‘Allegro’.
No. 6 may have had its three movements composed for the compilation,
including a middle movement with the binary structure of other middle
movements composed for the set, i.e. Nos. 3 and 1. An unusually high
number of corrections in P 271, especially in the first movement, suggests
that the composer was still working on it. (In the case of the Six Solos
for Violin, the last probably needed least altering during the compilation
process: see Eppstein 1969 p. 25.) No. 6 is therefore unique, placed last
perhaps because complete in itself. So the sonata with the biggest number
of up-to-date articulation signs was the last to be copied? – many of the
signs in P 272 for movements 2 and 3 may also be the composer’s.
First movement
The concerto-like arrangement with quasi-tutti and solo is at its clearest
in this movement. In structure, though not of course in manual changes,
it resembles the first movement of the Italian Concerto for harpsichord
(1735).
A 1–20 tutti; subject answered in dominant, as a fugue

B 20–57 solo; subject, answer, episode, broken chords; subject 53
A 57–72 tutti subject decorated; 60, episode from A
73–85 tutti subject developed in sequence
B 85–101 solo episode = 37–53 (motifs inverted, parts exchanged)
A 101–36 tutti subject decorated (101–9 = 53–6); episode from
bb. 8ff. developed (109–14 = 117–22); stretto
development of tutti
B 136–60 solo episode from 37/85, rectus and inversus combined;
solo from b. 21 developed in minor, dominant pedal
point
A 161–80 penultimate lh figure altered for final chord∗
However, this tutti/solo structure is no more than a framework invoked now

and then; the movement is not a concerto with clearly marked sections. In
∗ The NBA is surely correct to make b. 167 the same as b. 7 despite the reading in P 271 (KB pp. 33–4).
The unresolved fourth is a cadence à la Buxtehude.
34 BWV 530
concertos, the main theme is often hinted at in the solo episodes, but less
ambiguously than here in bb. 53–60 (ambiguous because of the invertible
counterpoint). If the ‘tutti’ begins in b. 57 not b. 53, it does so by force of
key rather than theme, and such ambiguities are typical of forms transferred
from one medium (concerto) to another (organ sonata).
The writing is concerto-like, particularly the unison theme – unique in
the Six Sonatas. Moreover, when the first solo passage appears in b. 20, it
is above a pedal point, as in the D minor Harpsichord Concerto and the
Fourth Brandenburg. Such a ritornello alludes to concertos, though here
with ideas typical of the Six Sonatas, for example the pedal point in b. 153
over which the first subject is developed, much as in Nos. 2 and 5. Also
characteristic is the minor chromaticism preparing a strong tonic entry
(bb. 153–61), and indeed, the main subject loses its ritornello feel if it is not
so prepared (as at the ambiguous G major of b. 125). Minor chromaticism
preparing a strong tonic entry is one of many details found in Vivaldi (see
the transcription BWV 973), one commentator even claiming BWV 530 to
be ‘a new piece generated from stuffs found in the work of Vivaldi’, using it
as a ‘database’ (Derr 1987).
The subjects are characterized by their own distinctive note-patterns
or figurae. Pedal lines are especially varied, with figures less difficult to play
than the semiquavers of No. 5’s finale, and lending tension to the stretti. One
particular motif serves as a link between phrases and subjects throughout
the movement – Example 21 (i) – and, taking various forms, it can be seen
operating in bb. 4, 8, 20, 28, 56, 60, 72, 84, 104, 108, 160 and elsewhere.
Bar 56 has a countersubject which appears three bars earlier – Example 21
(ii) – in which form it also appears in b. 104. Decorated versions of the tutti
subject tend to disguise its entry, for example at b. 101.
Example 21
Second movement
Like the slow movements of Nos. 1 and 2, this is a binary form whose second
part returns to the opening theme:
binary (16, 24 bars); first half develops motifs from one main theme;
second half with new theme (and new kind of bass); 25–40 = 1–16,
parts exchanged
35 BWV 530
Further details are familiar in slow movements: a bass below sequences (see
No. 2 b. 17, No. 5 b. 40), contrary-motion scales before the reprise (No. 5
b. 40), and pedal references to the subject. It all evolves so naturally that one
can miss how many thematic allusions there are. For instance, bb. 12–13 have
several in each part, while phrases can also be different and yet obviously
related – compare b. 2 (first theme) with b. 16 (second). The alien notes
introduced in bb. 21–4 produce a passage amongst the most skilful in the Six
Sonatas: strained, logical harmonies are worked above pedal motifs taken
from the subject, delaying an entry in a key already arrived at.
Though a binary movement, in its melodic style it is more like that of an
affecting aria with obbligato violin than a chamber sonata, where melodies
are usually less cut up. It is marked ‘slow’, thus not quite like a siciliano as
prescribed by Quantz:
muss sehr simpel und fast ohne Triller, auch nicht gar zu langsam gespielet
werden. (1752 p. 143)
must be played very simply, almost without ornaments yet not at all too
slow.
The movement conforms with this directive even less than do other cham-
ber works (Organ Sonata BWV 525, Violin Sonata BWV 1017, Gamba
Sonata BWV 1028, Harpsichord Concerto BWV 1063) and suggests the
‘Bach siciliano’ to be quite different from Quantz’s. The countersubject,
independent in rhythm and line, is conceived to be invertible: not a normal
feature in light dances but found elsewhere in the Six Sonatas. Less usual is
that the voices never join together and are united only at the cadences.
Third movement
This is another finale with a fugally treated theme in which the pedal also
joins. As in No. 5, it begins with a melody and counterpoint typical of the
Three-part Inventions, as does the second subject (b. 19); each, however,
soon passes to a simpler passage, almost galant at bb. 19–20. The form can
be outlined:
A 1–18 subject, answer, broken-chord episode; pedal entry b. 8

leads to sequence; coda, subject in stretto
B 19–31 second subject and answer; episode above bass from 22
A 31–41 stretto development of first subject, then derived episode
B 42–51 second subject answered in subdominant (after 4 bars);
B2 as B1 but filled in (bass between feet and hands)
A 52–77 return; extended, subdominant then second answer
(b. 59); 67–77 = 8–18 without change
36 BWV 530
An important detail is P 271’s dots at the beginning. Do they suggest that

otherwise one slurs 4–3s on the beat? As in the Vivace, a broken-chord
episode follows the initial subject and answer; as in the finale of No. 5,
the simplified subject in the pedal (b. 9) is then taken up in sequence; and
also as in No. 5, a tonic stretto at the coda helps to bring finality. In both
movements, simplified pedal themes can only with caution be regarded as
pedal entries/answers, since they are more like episodes, and the pedal is
not taking an equal part in the fugue, as it is in the finale of No. 2.
The whole movement fluctuates between the bright charm of a concerto
(jolly broken-chord figure of b. 3) and the sober counterpoint of an inven-
tion, and is both modern and traditional. The canonic imitation of bb. 14–18
leads to a somewhat circuitous harmonic sentence, while the tendency of
the second subject to be harmonized in sixths clearly suggests a proto-galant
style not far from Telemann’s trios. Similarly, the broken chords of bb. 4,
50, 60 are more pronounced than usual in the Sonatas’ fugal movements,
and surely aim at a more modern touch. Bars 48–52 have that descending
détaché bass known in many a concerto finale, such as the Concerto for Two
Harpsichords in C minor, BWV 1060.
The entry of the second subject is absorbed in a dazzling sequential
figure which drops to become a countersubject, alas not taken further
(Example 22). Several entries are further hidden by semiquavers. The pedal
Example 22
often has an ungainly look despite a wide variety of note-patterns; that may
be the reason why its line at bb. 21ff. becomes split between pedal and man-
ual in bb. 44ff., aiding the tension of the middle section. For a B-section
to modulate further and more often than the A-sections is a characteristic
of ABA form: cf. the first movement of No. 5, or the finale of the E major
Violin Sonata BWV 1016.
Note that in P 271, the last bar, unlike b. 18, is slurred, as if to suggest
that a marked articulation signals the end – as it does.
Preludes and Fugues (Praeludia) BWV 531–552
BWV 531 Praeludium (Prelude and Fugue) in C major

No Autograph MS; copies in Mö MS (J. C. Bach, before 1707?), P 274
(shorter fugue, J. P. Kellner 1724/5? Stinson 1989 p. 23); MS once thought
to be autograph (prelude, Washington LC, ML 96.B 186) copied by C. G.
Gerlach (c. 1720: Schulze 1984 p. 123); Stuttgart Cod. mus II.288 (prelude,
owned by W. H. Pachelbel c. 1740).
Two staves; title in Mö MS ‘Praeludium pedaliter’. Stuttgart has ‘Segue l’Fuga
un piu Largo’.
The fugue is already complete in Mö MS, so that since both it and the
carelessly written P 274 (Spitta I p. 400) derive directly or indirectly from
the same autograph, it seems that P 274 arbitrarily shortened the fugue.
The bass subject-entry of b. 36, during the omitted bars 26–54, is unlikely
to be for pedal and is thus no evidence that this section was a ‘later addition’
(Keller 1948 p. 50). Other copies, including Stutttgart, appear to have other
pedigrees, their different readings throughout reflecting problems in the
work’s transmission still evident in NBA.
It has become common to draw parallels between BWV 531 and the
praeludia of Georg Böhm, even dating the work to the ‘Böhmian’ years
before Bach travelled to hear Buxtehude (e.g. Schöneich 1947/8 p. 99).∗ Such
qualities as ‘the virtuoso brilliance of the closes and the freedom of the part-
writing’ also suggest the work to belong to an early period (Spitta I p. 401),
c. 1705. Resemblances to Böhm’s C major Praeludium are ‘unmistakable’
(Keller 1948 p. 50), but various works of Buxtehude suggest other similarities
to, and possible influences on, BWV 531, while in Lübeck’s Praeludium in C
the influence might be mutual. Clearly, the work is an early and imaginative
response to the music of established masters, with marked similarities in
figuration, texture, harmony and use of the organ, all of these implying a
common genre.
The Mö MS contains both the C major and D minor Praeludia of Böhm,
no Lübeck, and of Buxtehude only the less expansive A major Praeludium
and G major Toccata. Particularly apt parallels can be made with the D
minor Prelude and Fugue BWV 549a (also in Mö MS), almost as if they were
∗ That Böhm’s instrument in Lüneburg did not have independent pedal-chests until 1714 or so does
not mean that his major pedal-works need date only from then onwards (suggested in Wolff 1991
[37] p. 62).
38 BWV 531
conceived as complements: see notes to BWV 549/549a. But to call these

two works Bach’s ‘earliest surviving free organ compositions’ (Stauffer 1980
p. 129) would be to assume that early works without pedal solo, such as the
Fantasias 563 and 1121, are not ‘free organ compositions’, which may be
incorrect.
Prelude
As here, opening pedal solos based on alternate-foot pedalling tend to in-
clude dramatic rests or ‘rhetorical tmeses’ (BWV 549a and 564, Böhm in C
major, Buxtehude in C major), close sometimes with a pedal ornament –
more often than is notated? – and continue with a manual imitation of the
pedal, or vice-versa (e.g. Buxtehude in E minor, Bruhns in G minor). More
unusual is the pedal scale of b. 17, something perhaps that inspired the
virtuoso opening of the D major Praeludium?
The harmony of tonic–subdominant–dominant–subdominant–domi-
nant–tonic is more systematic than in the freer fantasies of earlier composers,
and there is an aura of sustained melody about the piece. The first eighteen
bars are almost entirely around a tonic pedal point, kept up longer than
was customary and filling the ears with the bright sound of C major, like
the opening bars of the WTC. Such bars as 17–18, though reminiscent of
early cantatas (BWV 106), are hard to match for the pleasure they give the
player. Other details can be found in other praeludia, such as the parallel
sixths in b. 22 (cf. BWV 568 or Lübeck’s Praeludium in C), while elsewhere
the material is wholly conventional. But the non-stop pedal points give the
movement a drive unknown in sectional toccatas such as BuxWV 165 in
the Mö MS.
The harmonic repetition of bb. 23–7 or bb. 30–2 suggests a new, original
version of the reiterated harmonies in Buxtehude’s Praeludium BuxWV 138
(bb. 10–14), where the repetition is simpler and winsomely obsessive. The
unequal interest of bb. 31 and 32 is probably a sign of immaturity, while
the climax of the final bars is out of proportion to the rest of the prelude,
even by the standards of Bruhns or Buxtehude. These composers are also
less likely to use both the top and bottom notes of the organ (C–c ) quite
so patently in the final bars of a first movement. So the Prelude mingles
the conventional and the unconventional, assembling various old prae-
ludium ideas expanded to a fully independent prelude of forty bars. Similar
points could be made about BWV 568, where the phraseology is more
regular.
Perhaps somewhere in the Praeludium’s transmission a tablature was
misread or an option misunderstood. Something is wrong in bb. 13–14:
should the top line read e g c g e g c g e , and notes 8 and 16 of
the bass-line be up an octave? Also, it seems unlikely that any missing pedal
note in b. 36 (if there is one) is d, as suggested in NBA; G was surely either
39 BWV 531
implied or restated, as in b. 24. And no doubt the demisemiquavers of the

final bars are distributed between the hands. Finally, the last chord is surely
too big and too long: did the left hand originally run down to a short, single
tenor C, with the Fugue following subito, senza pausa? Such readings both
suit this Prelude and complement the early D minor, BWV 549a.
Fugue
Such a perpetuum mobile fugue-subject is more characteristic of both the
smaller keyboard canzonetta and the variant fugue in a long praeludium,
such as Lübeck’s in C major; it is less characteristic of a self-contained organ
fugue, which from Scheidemann onwards tended to be ‘quieter’ in style. The
exposition is unusual:
four entries over three parts, resulting in a falling effect (g , c , g , c );
answers mostly subdominant (cf. the first fugue of the Capriccio in B), as
if the subject’s dominant notes are answered by tonics (cf. BWV 565).
The ‘falling effect’ is an early feature (Bullivant 1959 p. 344). Further devel-
opment of the subject produces a particular shape:
1 exposition, episode; 11 new material (in Pachelbel’s italianate

manner)
14 stretto use of subject caput in stretto; middle entry; more
‘Pachelbel’
24 middle entry (stretto with pedal version of caput); new episode,
relative
36 tonic entry in bass (pedal not cued in any source); derived
episode
41 4-part harmonization of entry; derived episode
49 dominant entry; episode; tonic entry
55 long coda, subject not heard again complete
The final bars are built on conventional flourishes – including a sudden

tonic minor (cf. Böhm’s Praeludium in C, and also BWV 549a) – and thus
recall old toccatas. But the fugue is better understood as:
A 1–27 beginning and ending in C major; no full pedal entry

B 28–55 ending in C major; a modified pedal entry
A 55–74 coda; pedal for point d’orgue
The free close is thus merely part of a longer coda. B depends on a passage
not given in P 274, which therefore has a version changing direction un-
expectedly (bb. 30–1, 34, 52–3); this passage also contains conventional
note-patterns found in A but now more ‘advanced’ (compare bb. 19–21
with 30–2). The harmonization of bb. 41–2 is both curiously original and,
40 BWV 531–532
surprisingly, taken no farther. The big chords against a pedal ‘entry’ in b. 23

are found in other early fugues, in particular BWV 549a and 533.
But what is the authentic form of BWV 531? That b. 25 ends with the
same eight notes in the right hand that begin b. 55 is open to various
interpretations. Perhaps for some fancied ‘improvement’ P 274 omitted the
section bb. 26–54 (leaving an unconvincing join), while bb. 26–54 in Mö MS
were original, without a bad join. One could also imagine further extension
of bb. 26–54, as for instance going on from b. 33 towards an entry in the
relative minor (b. 34, like b. 53, is rather abrupt). Although a bass entry
shortly after this in b. 36 might seem odd in view of the simplified version in
b. 23, it was surely intended for lh (Breig 1993 p. 48), and its pairing with the
countersubject recalls the Fugue in A minor BWV 551 b. 45. In the longer
version manual-changing becomes quite feasible:
b. 1 manual I, b. 14 manual II, 22 I, 26 entry II, 36 I, 45 II, 65 I
The pedal note in b. 70 is F according both to the sources and to the

old convention of making dramatic use of the dominant’s leading note (e.g.
Praeludium in D major BuxWV 139, bb. 89–94). But a conjecture that it
should be G, as in BG 15, is not inappropriate, especially if followed by
one long manual trill in bb. 70–1 (a trillo c –b is also in style with old
praeludia). The last few bars have reminded some of the ‘dark harmony’ of
minor chords in Bruhns and Buxtehude (Frotscher 1935 p. 866), and the
minor–major change gave Spitta the impression of ‘a spring storm at night
in March’ (I p. 401). But should the long eight-part final chord be short,
with e as the top note, whatever the sources say?
BWV 532 Praeludium (Prelude and Fugue) in D major

No Autograph MS; copies in e.g. Stuttgart Cod. 11. 288 (W. H. Pachelbel?
c. 1740), P 204 (1781? via C. P. E. Bach?); prelude, in Lpz MB MS 7 (J. N.
Mempell †1747) and P 287 (second half eighteenth century); fugue, in
P 595 (J. Ringk, after 1730?) and P 1095 (J. N. Mempell), in C major in
P 567 (J. F. Doles?).
Two staves; title in MS 7 ‘Praeludium’, in P 287 ‘Preludio – Claviecembalo’

and in P 204 ‘Piece d’Orgue’; ‘Praeludio Concertato’ in the Pachelbel MS,
where also at the end is written: ‘Nota Bey dieser Fuge muss man die Füsse
recht strampfeln lassen’ (‘note that in this fugue one must let the feet really
kick about’).
The title Pièce d’Orgue implies a festive character and sectional plan like a
Parisian organist’s Offertoire (Klotz 1962). But its authenticity is uncertain,
41 BWV 532
and neither idiom nor form is French. Rather, conventional northern toccata
sections, italianate sequences and a local fugue-subject are worked towards
a massive structure, each section more or less self-contained, the general
effect less capricious than earlier praeludia. Griepenkerl’s idea that the word
‘Concertato’ implied use outside church cannot be substantiated (Peters
IV); nor Spitta’s that it was ‘for an occasion, such as one of his artistic
travels’ (Spitta I p. 404); nor that it was played on the new organ of the
Liebfrauenkirche, Halle in 1716 (David 1951 p. 38).
Sources suggest the movements did not originate together (KB pp. 343,
715), but against the idea that the work began as the Fugue BWV 532a, was
then enlarged and given a prelude as well (Breig 1999 p. 659) are that the
Prelude is built up from various ‘building blocks’, too ‘early’ a sign for it to
be contemporary with the longer fugues. Either way, there is exaggeration
in seeing the fugue as ‘derived’ from the alla breve section (Dietrich 1931
p. 60) and that the ‘end of the fugue and the beginning of the prelude’ have
a ‘similar character’ (Keller 1948 p. 63). But one sees Keller’s point.
Prelude
First section
Though not very close to any extant work of Bruhns, the opening scales and
broken chords – all on a tonic pedal point – match his style. Closer still is
the start of the D major Harpsichord Toccata BWV 912, J. C. Bach’s copy
of which in Mö MS ties the notes of the broken chords, thus producing an
organ-like effect. The Toccata’s opening scale is now in the pedal, an original
gesture (but see BWV 531 above). Also in toccata tradition, both southern
and northern, are the dominant pedal point and manual figures in simple
stretto, and even the little figure of b. 9 is found in other organ works (BWV
566, 718). The rhetorical gestures are extreme.
Second section
Surprise chords are usually – as in recitative – first inversions, not root po-
sitions. But snapping rhythms, tremolo chords and quick scales give much
the same effect, the tremolo a new version of the northerners’ trilled thirds.
These first sixteen bars are those of a young, inventive composer ‘control-
ling’ the disparate elements of earlier praeludia, with an uncanny sense
of the drama of rests and the power of scales. The rhetoric is startlingly
accomplished, especially in the stormy B minor passage into and from
which the listener is thrust without warning.
And yet – the little section is very similar to one in the early D major
Sonata BWV 963 (known from a copy by Mempell), hardly a coincidence:
the F chord, the rhythms, the rhetoric are all virtually identical. Kuhnau-
like in so many respects, the Sonata too seems ideally to require pedal for
this very section.
42 BWV 532
Third section
The idea of a simple, sequential main theme with episodes is also to be
found in the Allegro of the Toccata in D. Although the rubric ‘allabreve’ is
not reliable (in the Pachelbel MS but not Mempell), its meaning is clear: the
new crotchet is twice as fast as the previous, whose opening scales are not
emptily virtuoso. Alla breve implies that none of the three sections is fast,
while the final ‘adagio’ sign (in the same sources) is slower still, to mean
free or ‘at ease’. Probably, such varied tempo was natural to organists of old
praeludia, and Italian terms were unnecessary.
The main theme of this alla breve embroiders a conventional chain
of suspensions which, depending on the inversion, can be described as
7–6, 5–6, 5–4, 2–3 or 9–8. In three parts, the sevenths would be 7/3, but
in four they require the fifth: 7/5/3. As sometimes in Buxtehude (the G
minor BuxWV 149, the F minor BuxWV 146), the result looks like a
model passage for the learner of figured bass, and would not be out of
place in the treatise Gründlicher Unterricht (1738), sometimes attributed
to J. S. Bach. The figuration itself (quaver lines, especially around b. 40)
is surely influenced by Buxtehude’s F minor Praeludium. The distinct
episodes could hardly be simpler: triads, repeated notes, repeated phrases,
all contrasting with the main material, which has none of these. The sim-
ple style can at times remind the listener of Cantata 4 (c. 1708) or perhaps
Corelli, as do other early keyboard works like the Aria Variata. Other déjà
vu italianisms include quaver lines of a kind found elsewhere, e.g. in the
overture to Handel’s Chandos Anthem HWV 247 and Harpsichord Suite
HWV 431.
The differences between theme and episodes suggest a second manual
for the latter, though it is not always quite clear where the episodes begin:
b. 31, then b. 62, b. 71 etc.? The many quasi-echos from b. 39 onwards also
suggest a second manual, as does the notation in the sources of bb. 62–3,
chords as simple as those in Böhm’s G minor Praeludium in the ABB. It
may seem out of character for the left hand to go alone to a second manual
in such bars as 37, 39, 41, 52 (Klotz 1975 p. 390), but the manner of writing
allows one to play with the keyboard(s) in various ways.
Fourth section
As in the Pièce d’Orgue BWV 572, an original interrupted cadence is pro-
vided by slipping to a diminished seventh. While the adagio harmonies are
certainly in the Buxtehude manner, closer comparison can be made with
the Grave of the C major Toccata, both for location (a short interlude,
a new tonic) and idiom (scales between the hands, diminished sevenths,
augmented triad, ninths, angular pedals). The part-writing of BWV 532 is
stricter, pedal might be doppio (not clear in any source), and harmonies
43 BWV 532
are calculated to mystify with dark, anxious, unexpected minors. Bach uses
the diminished seventh and Neapolitan sixth (cf. Bruhns, Praeludium in
G minor, b. 30) more systematically than any French composer. Once again
the section looks like an ‘enlargement’ of part of the Sonata in D major
BWV 963, where, curiously, one can also glimpse a far maturer work in
D major, the Fugue in WTC2.
Fugue
The extraordinary, rather violinistic subject and idiosyncratic counter-
subject have led commentators to search for similarities elsewhere, in
Buxtehude, Pachelbel, Reinken and especially the Thuringian tradition rep-
resented by a Fugue in G minor of J. H. Buttstedt, like J. C. Bach a pupil of
Pachelbel in Erfurt (Schäfertöns 2000).
Here is a distinct type of keyboard fugue-subject – long, with a conspicu-
ous opening, then a spun-out phrase, and finally a cadence – which proceeds
not to a toccata-postlude but to a virtuoso coda, mostly on a long final tonic.
Buttstedt’s Leipzig publication of 1713, Musicalische Clavier-Kunst, contains
several examples of long, ‘wild’ fugues and, in e.g. his D minor Capriccio,
more than a few similarities to BWV 532. But other works of Bach himself
are not so very distant – a pedal line in the G major Fugue BWV 550, a
bass theme in Cantata 71 of 1708 – and it is always possible that Buttstedt
(1666–1727), like Böhm, was himself influenced by J. S. Bach. Any organist
ever coming into contact with BWV 532 must have been startled by it.
However close to Pachelbel’s Fugue in D some of the motifs are, including
the opening bar (Wagner 1987 p. 26), and though it plays with ‘Thuringian
broken-chord counterpoint’, the Fugue is exceptional in melody and modu-
lations. Most extraordinary of all is that there is no true final cadence, either
perfect or plagal.
A 1–29exposition, two real answers; derived then free episode

30–53
middle entry (re-exposition tonic–dominant–tonic);
episode
B 53–64 entry, relative (first bar repeated); derived then free
episode
64–76 answer, dominant of relative; countersubject rhythm;
hovering in F minor at central axis (69)
77–96 caput on pedal; further answers, broken up, shortened, in
dominant of relative dominant; episode; ‘development’
C 96–124 final entries in dominant (then lengthy episode) and
tonic
124–37 coda: second half of subject, arpeggios from first codetta
(12); play of motifs, virtually a tonic pedal point
44 BWV 532–532a
The episodes could be differently described, for B is in fact a kind of

Development Section, in which this or that element is used here or there, in
different voices and keys, coherent but exceptional for a fugue. The cut-up
lines allow for changes of manual, though it is no more than an interesting
conjecture that the fugue was planned for ‘the four manuals of the great
Hamburg organs’ (Klotz 1975 p. 391).
The exuberant spaciousness of it all should not disguise its many inge-
nuities. After the first section, it is never clear whether the opening phrase
of the subject is going to herald a simple entry (b. 96), an episode (b. 77,
bb. 103ff.), or another voice (bb. 90–1), or be merely delayed (bb. 53–4).
The charming play with the trillo figure in bb. 69–71 might be a nod to
BuxWV 145 but is nonetheless unique even though its key of F minor is
prominent in the (older?) Prelude. The anchoring effect of the long domi-
nant preparation for the final entry (bb. 103–16) might be necessary but is
nonetheless contrived in a quite un-fugal manner.
All the tonics at the end of the Fugue could be seen as mirroring all the
tonics at the opening of the Prelude. And yet, despite the length of this final
section no other fugue in the literature actually ends so succinctly, with such
an exclamation, and (like the Missa solemnis, also in D major) without a
true cadence: an astonishing piece.
BWV 532a Fugue in D major

Peters IV (1845), from ‘a very good MS’.
Two staves; heading, ‘Fuga’.
This version differs most at the following points:
BWV 532a BWV 532.ii

28–9, 59–61 28–9, 59–64 different content
62–71 65–76 similar, but entry shorter in 532a
71–3 — episode in 532a
— 76–96 entries in further keys in 532
74–98 96–137 longer episodes in 532; cadence in 532a
BWV 532a is unlikely to be authentic, though often taken to be an early

version later expanded, or a later shortened version (Spitta I p. 405), or one
made (by whom?) for an organ unable to use such distant keys as the longer
version (Edler 1995). But the two Albinoni fugues BWV 951/951a and the
Reinken fugue BWV 954 are more reliable as models of reworked versions.
45 BWV 532a–533
The enlarged Fugue in A WTC2 does not offer a parallel to the putatively
enlarged D major (suggested in Breig 1993 p. 56), since a complete section
was added to the A, not interspersed.
Surely, whether it is genuine or not, few players find Spitta’s admiration
for BWV 532a ‘incomprehensible’ (Lohmann EB 6581 p. xi). In shape, it is
much more like the other early fugues than is BWV 532, and the cadence at
bb. 92–3 suggests a trained composer, as do the chromaticism in bb. 43–4,
the different version of bb. 27–8 and the way that BWV 532’s abruptness in
cutting off the stretto in bb. 58–9 is now avoided. Both final passages are
convincing, though it is easier to imagine the frenetic element of BWV 532
as material cut from a long fugue than as bars added to a short one.
BWV 533 Praeludium (Prelude and Fugue) in E minor

No Autograph MS; that once thought to be autograph (Lpz Bach-Archiv
Mus. MS 2, fugue only) copied by J. C. Vogler; copy by J. Ringk (P 425),
others probably via C. P. E. Bach (e.g. P 287) or J. C. Kittel (e.g. P 320);
copies of fugue only (P 804 eighteenth century) and prelude only (P 301).
Two staves; heading in P 287, ‘Praeludium et Fuga ped’; by Vogler, ‘Fuga

pedaliter’ (but no pedal cues).
Ringk’s is the only complete contemporary copy, which, if reliable, makes it

all the likelier that the Prelude is an early Böhmian work revised in Ringk’s
source. For a version of the Prelude without pedals (earlier? – KB p. 382), see
BWV 533a. Bar 18 is missing in some related copies, probably by mistake.
Neither version of the Fugue requires pedals, and no source asks for them in
b. 19 (KB p. 388), though for the final entry they are certainly appropriate.
Probably, the two ‘versions’ are of a work taking different forms, neither
demonstrably earlier or later than the other, or necessarily for or without
pedals, or always with two movements.
Spitta was full of admiration for the work, hearing certain expressive
qualities in both movements (‘gloomy pride . . . melancholy . . . magic’),
which he saw as ‘closely related, more so than usual’ (Spitta I p. 401). Yet
some copies provide good authority for the fugue circulating independently,
at least ‘for a time’ (KB p. 385). But whether or not BWV 533 really is the
first extant example of the fully separate prelude and fugue – too early to
get into Mö MS? (see Schulze 1984 p. 46) – it is true that the Prelude’s
toccata-like solo lines, free passages and durezze are absent from the Fugue,
which is strikingly free even of suspensions. Both movements are so un-
usual in conciseness, inherent melody, rhetorical gesture and bar-by-bar
46 BWV 533
detail, while differing in end-result, that they do look like typical Bachian
complements.
Prelude
The opening solo line resembles the Prelude of the Lute Suite BWV 996
(copied by J. G. Walther): ‘improvisations’ around a chord of E minor,
settling on a low tonic. BWV 996 is closer to the usual solo run-in of a
Buxtehude praeludium than is BWV 533, whose question-and-answer shape
is more regular and which begins more obviously in the tonic: Example 23.
Example 23
The freer passage beginning at b. 6 introduces vigorous ideas familiar in

‘northern’ praeludia (see Example 24), so that the gloomy weight familiar
in performances of it is perhaps an anachronism. All three ideas appear
in a further E minor work, the Toccata BWV 914, Adagio, about which
there is little very gloomy. (The figuration in b. 10 seems to be mis-written,
with redundant b . See the Prelude in A minor BWV 543 b. 23, and the
harpsichord toccatas.) Similarly, in the third section the harmonies are not
so much ‘atmospheric’ as an original way of handling keyboard mannerisms
of the day (cf. Bruhns’s ‘Nun komm’, b. 58). Such details as the final repeated
cadence recall the tonic re-affirmations in the early Cantatas 131, 106, and
71, certain works of Böhm, etc.
Example 24
An unusual feature is the many short phrases, resulting in a focus on the

most original section, the driving, pesante chords from b. 18. As in other
early works, the harmonic tension derives from simple diminished 7ths, here
47 BWV 533–533a
functioning as dominant minor 9ths. The harmony is not sophisticated but

the rhetoric is faultless.
Fugue
The first half of the fugue is taken up with five entries, one more than the
number of parts, as elsewhere in early fugues (BWV 531, 549a):
1–15 tonal answers, then real answer (not pedal?), cf. BWV 550
15–18 typical episode derived from codetta
19–27 entry en taille, soprano answer
27–36 entry, episode (countersubject’s dactyls); final entry bass, no
coda
Though brief, this is a classic fugue shape, all entries tonic or dominant.
The texture picks up on the Prelude (Fugue bb. 19 and 24, Prelude b. 15), as
perhaps does the melody (Fugue b. 18 alto, Prelude b. 29 – a coincidence?).
The second round of entries, an exposition corresponding to the first, begins
at the halfway point.
NBA’s policy on ties is not necessarily correct: on one hand, early copies
especially in or from tablature are often sparing in ties, whatever players
did in practice; on the other, early works might well make more of repeated
chords and notes as part of their style. The texture at the entry in bb. 24–5 is
found in chorale fantasias, while the harmony at the entry en taille and the
melody at various points are those of a future master. Though on a small
scale, the harmony and the melody can spin out lines – the pedal entry of
b. 33 could have appeared one and a half bars earlier – and produce episodes
less merely time-filling than those of BWV 549a. The final bars, simple and
undramatic, have a harmonic resonance typical of five-part writing in a
cantata sinfonia of Buxtehude or Bach (Cantata 4.i), and there is a touch of
the elegiac not rare in E minor (cf. the Three-part Invention).
BWV 533a Praeludium (Prelude and Fugue) in E minor

No Autograph MS; only copy, Lpz MB MS 7 (J. G. Preller).
Two staves; heading ‘Praeludium et Fuga’.
Pedals are neither specified nor required by the spacing, and the Prelude has
two extra bars, so that bb. 6–13 of BWV 533a are equivalent to bb. 6–11 of
BWV 533.
Though usually spoken of as an early version (KB pp. 382–3, 581), i.e. an
original pedal-free version, BWV 533a is demonstrably neither earlier nor
48 BWV 533a–534
even authentic in its detail. Some copies of BWV 533 (Ringk) correspond
in details to 533a, and one wonders if Preller was himself responsible for
passages in BWV 533a that are not there in 533 (Schulenberg 1992 p. 58).
Preller’s work probably dates to the 1740s, when there must have been a
MS source available in Leipzig (KB p. 382), though the fugue’s ornaments
(KB p.194) recall typical Walther sources.
The Preludes’ last five bars could imply that BWV 533a is either reduc-
tion or later simplification (Breig 1993 p. 48) of the organ version, which
alone has a recurrent motif (tremolo chords). Perhaps the composer began
to add harpsichord figures, omitted the unifying motif but extended the
Buxtehude-like idea of b. 6, going no farther than b. 13 with it. But since
differences in the harmonies from b. 20 were hardly due to carelessness,
as might be the case near the end of the Fugue, perhaps the organ version
is indeed a re-writing. The dominant minor ninth cadence in the Prelude
BWV 533a is probably a mistake, fine effect though it is.
The Fugue as it stands in BWV 533 is also playable by hands alone,
although this means that the real answer in b. 12 is not then so conspicuous.
BWV 534 Praeludium (Prelude and Fugue) in F minor

Only copies: Lpz MB III.8.21 (with BWV 544, 545, 548 – J. A. Dröbs, pupil
of Kittel) and a print of c. 1840/45 (G. W. Körner).
Two staves; heading, ‘Praeludium et Fuga ex F moll pedaliter’.
Since Körner used a further Kittel MS (KB p. 413), the work exists thanks to
a single source, from which the various infelicities it contains (see pp. 50, 51
below) might come. One can only guess whether its key-signature of three
flats means it copied a much earlier MS. Rather than indicating an early work,
its inconsistencies – e.g. careless counterpoint but mature harmony – have
led some to conclude that it was a new piece by Kittel himself, familiar with
the F minor WTC2 and producing ‘a pastiche of elements drawn’ from the
C minor Fugue BWV 546 and E minor Prelude BWV 548 (Humphreys 1985
p. 177). If the rich harmonies and melodies are not matched elsewhere in
Kittel, then perhaps he was helped by his teacher. However, there is nothing
unusual in a work of Bach being like none other, and the three other fugues
in Dröbs’s manuscript are above suspicion.
Qualities heard by Spitta made him see the work as one of those opening
up ‘new paths’ through its roundly shaped Prelude and more spacious Fugue
(Spitta I p. 581), although the Fugue’s ‘hesitation to leave the main key’ is
‘disproportionate to the ambitious length’ (Breig 1993 p. 53). Even if the
49 BWV 534
Fugue were an addition to an existing prelude, the poor source material

means that its errors (doubled leading-note b. 128, faulty suspension in
bb. 98–9) and the key itself need not be original. In G minor, the pedal
might have been able to play the low C two bars from the end.
Prelude
A 1–11 pedal point; 2 upper parts in canonic imitation

11–31 sequences (also pedal); 4 parts; hemiola cadence to:
B 32–43 as 1–11, dominant minor
43–67 sequences (also pedal); phrygian cadence; 3, 4 parts
67–76 pedal; big diminished 7th (4 parts in BWV 532 b. 96,
6 parts in BWV 572 b. 185).
The movement is thus a large binary form without much feel of ritornello.
Sequences are underlined by pedal (unusual), and the transitions produce
a figure that returns throughout (bb. 21, 26, 50), creating a sequence of
its own (bb. 64–6). Quaver figures all seem inter-related, as do semiquaver
figures. Almost all begin off the beat, hence the ambiguous metre when they
first do not (e.g. b. 17).
While G minor would be less ‘anxious’ than F minor, such a sarabande
doublée with continuous semiquavers, Neapolitan 6th and hemiola matches
other Bach sarabandes. To judge by the pedal-line at bb. 21, 26, 50, the
composer knew or was later to know the Aria of the Goldberg Variations,
as he surely knew the opening of the Toccata in E minor BWV 914. See
Example 25. The same underlying harmonies can be discerned at the be-
ginning of the E minor Prelude BWV 548. The binary form is unlike Italian
examples, being more like a toccata of the Pachelbel type: long tonic and
dominant pedal points, interspersed with and followed by other material,
as (on a bigger scale) in the Toccata in F major.
Example 25
It is not only in the last two and a half bars that the Prelude anticipates the
Fugue: its final eight bars trace in a freer, more prelude-like way the melodic
line of the final six bars of the Fugue. The melodious texture – closing up at
the halfway cadence, opening out for the close – results in a concentrated,
50 BWV 534
unusual movement, ‘bleak’ when widely spaced, ‘warm’ when congested.

The bars around the succinct tonic return (b. 50) have been said to lack ‘a
genuine sense of direction’ in an already static movement (Humphreys 1985
p. 180), but a distinctive melos sustains them. The pedal part resembles a
continuo bass more than it does a conventional pedal line of c. 1715, and
alone suggests a composer familiar with the E minor BWV 548.
Fugue
1–27 5-part exposition; counterpoint from subject (crotchets, 3);

then a ‘prolongation’ typical of ricercars
27–46 episode-entries, in three, four, two parts
47–72 entry, relative; episode to dominant and tonic entries;
episode to:
73–96 entry, relative; episode to tonic and dominant (pedal)
entries
96–119 entries, tonic, dominant, tonic; shorter episodes
120–38 entries, dominant (two); 130/131 implied tonic stretto
For the order of the exposition’s five voices (A S2 B T S1), compare the C
minor BWV 562 (A S2 S1 T B) and the Kyrie of the B minor Mass (T A S1
S2 B).
Despite many tonics and dominants, so distinctive a harmonic and
melodic character make it hard to believe that Bach had no hand in the
piece. The absence of a recurring episode, canon or stretto, when each was
possible, cannot prove it to be the work of a pupil, for one might expect
him to aim precisely at such imitable Bach hallmarks. Nor need Spitta’s
judgement that the countersubjects soon peter out and the subject ‘must
always look around for help’ (I p. 583) mean that for once, Bach could not
do such an unusual thing as to create a fugue whose subject and real answer
repeatedly enter on the same notes, in various voices, with various counter-
subjects, and at various intervals of time. It is true, however, that the fugues
BWV 535, 992, 579 (Corelli) and 951 (Albinoni), which all emphasize the
tonic, are early.
The countersubjects vary imaginatively from minims to crotchets to qua-
vers and in number of parts: that at b. 27 (Example 26) is rightly in two
parts not one. The parts countering the subject vary in texture from one
to four, as if intending to present it in various guises. There is comparable
variety between episodes: long crotchet lines, perhaps with suspensions
(bb. 20–6), truncated (b. 69) or repetitious (b. 113), a sequence free
(bb. 50–5) or derived (bb. 61–3), loose episodes (bb. 69, 93) contrasted
with the alla breve (bb. 105–9), and so on. The paraphrased fugue-subjects,
51 BWV 534–535
Example 26
outlined in Example 27, would be something new for Bach, but paraphrased
chorales were lingua franca. Granted the unusual character of BWV 534,
even that it has ‘awkward, clogged counterpoint and part-writing’, a ‘badly
thought-out tonal scheme’ and a ‘general absence of control’ (Humphreys
1985 p. 175), there is still a warmth to the harmony and melody hard to
attribute to any pupil. It may be a sign of immaturity to have two middle
entries in the relative major, but are not both too richly harmonized for a
Kittel or a Krebs?
Example 27
BWV 535 Praeludium (Prelude and Fugue) in G minor

No Autograph MS (see BWV 535a); copy in P 804 (prelude only, J. P. Kellner?)
etc; independent copies of final version (?) in Göttingen Bach-Institut (1st
half eighteenth century?), Lpz MB III.8.7 (c. 1740–50, with 2 bars copied by
J. S. Bach?), P 1097 (J. C. Oley? †1789), P 1098 (J. G. Preller †1786), and via
J. P. Kirnberger (Am.B.606) or Kittel (P 320 and derivatives).
52 BWV 535
Two staves; title in P 804 ‘Praeludium’ (no pedal cues), in P 1097 ‘Praeludium
et Fuga ex G moll con Pedale pro Organo pleno’. Fugue ‘allegro’ in most
copies.
Both sources and content suggest that BWV 535 is the ‘later version’ of a work
with an ‘early style’ toccata postlude. Evidently available to Leipzig pupils
working on its variants, the work must have originated in the composer’s
early twenties and was perhaps revised in the Weimar period. The Prelude
of P 804 has thirty-nine bars (usually forty-three), and shows no sign of a
Fugue. Since sources are inconclusive as to how many times 535a (see below,
p. 55) was revised, a question arises about the pedal-line in bb. 55–6 of the
Fugue and its striking similarity to passages in the mature Preludes in E
major (e.g. bb. 145ff.) and B minor: do all three passages belong to a late
phase?
Prelude
For the cello-like passage-work above an implied pedal point, see also the A
major Prelude BWV 536 and the opening of Preludes in E minor for Organ
and for Lautenwerk, BWV 533 and 996. The term passaggio, written in the
autograph BWV 535a, implies the alternating hands carefully specified in
Oley’s copy (KB p. 449) and such as one finds in e.g. Walther’s chorale ‘Wir
Christenleut’, v. 2.
Unlike BWV 535a, which is equally coherent, BWV 535 takes an idea
in b. 3 for the section preceding the expected dominant pedal point.
Though simple, the effect is strikingly like the chordal passages in the E
minor Prelude, and leads to Buxtehude-like repeated chords and – rather
puzzlingly – an apparent reference in the pedal to the head of the Fugue’s
subject. Curiously, this is also a phrase quoted by Mattheson (1739 p. 154)
as an ideal series of narrow and wide intervals: G A B G E D. As the
Prelude is merely passing from one pedal point to another à la Pachelbel, a
cross-reference to the Fugue is unlikely. But did Mattheson know it?
The next section, with opening and closing dominant pedal points, looks
like an afterthought to the version BWV 535a. So the series of scales and
diminished 7ths returns to where it began, and harmonies pick up where
they left off. Does this mean that some (all) of the passage is optional?
Sources transmit several versions of it (KB pp. 438–41), suggesting that the
original was merely a series of harmonies to be realized as broken chords,
either ad libitum or on a specified pattern, as in the early version of the
C major Prelude WTC1. Unlike the string of diminished 7ths in the first-
draft cadenza of the Fifth Brandenburg and the Gigue of the B Partita
(which also come back to their starting point), the progression in BWV 535
is 7–6.
53 BWV 535
Other patterns too have distant relatives elsewhere, such as b. 15 (see

BWV 571) and b. 33 (see BWV 543): all of these are ‘devices’ for improvising
preludes, all with a certain high seriousness. The Prelude closes with seven
bars looking like a realized version of the last six bars of BWV 535a. Note
that the pedal is obligatory now only in b. 37, if then, and that the five
parts appear to be manualiter. Runs of demisemiquavers are doubtless to
be distributed between the hands, and the repeats in the middle section are
optional echoes – for change of manual or stops?
Fugue
The fugue-subject alone is a mass of style-allusion: it has the repeated notes
of a ‘repercussion type’ (Buxtehude, Kerll), trillo semiquavers (Heidorn in
D minor and Böhm in D major, both in Mö MS), is both continuous and
broken up (cf. BWV 549a and 575), and has a ‘premature’ answer. Yet there
is also a new and distinct melodic shape to it, as it moves from crotchets to
quavers to semiquavers.
A premature answer, rare in Bach, has different consequences here each
time the subject runs its course. The overall shape is as clear as BWV 578’s:
1–25 long exposition, four entries but three parts; episode

25–46 entries in the three manual voices, each with episode
46–55 entry in relative, pedal; episode
55–70 entries, each with episode
70–7 coda: pedal solo, scales, Neapolitan 6th, pedal point; highest
and lowest notes of the fugue (C/c )
The fugue emphasizes the tonic, as do the (contemporary?) Capriccio in
B’s second fugue, the Albinoni Fugue BWV 951a and indeed Albinoni’s
original. Much in the subject’s semiquavers and working-out resembles the
keyboard version (BWV 965) of Reinken’s Sonata prima, a work of particular
influence on the young Bach.
The subject, first countersubject, episode and later countersubjects
present what looks like a catalogue of note-patterns, any of which can sud-
denly take off in an unexpected way (b. 69). Perhaps the composer intended
B A C H to be heard at the end of the pedal solo (b. 71). The coda’s scales
are more succinct than those of the Prelude, just as the Neapolitan sixth of
b. 72 – an ‘early’ sign for Bach – is slighter than the Passacaglia’s. A common
detail in early works is the part-writing’s awkward moments when parts
cancel each other out. Examples in BWV 535 occur in bb. 13, 14, 18, 19 –
perhaps the result of composing on paper or of writing tablature, where
such overlaps, grammatically correct, are less obvious?
There is a sense of drive and the counterpoint is well conceived, in
particular the four-part passage from b. 46 to b. 57. One sees why Spitta
54 BWV 535–535a
heard a ‘new and increasing liveliness’ of the counterpoint ‘each time the
theme enters’ (I p. 405). The countersubjects also become livelier: at b. 55
a canonic figure in contrary motion, at b. 64 wide-ranging arpeggios. Here
too at b. 55 is the reminder, already mentioned, of the Preludes in E and
B minor (a countersubject). In a sense, the postlude is unnecessary since
there have already been climactic moments, and it may be wrong to assume
that these dramatic, quickly modulating final ten bars were also there in the
missing pages of BWV 535a (see below): one can imagine a quite different
coda.
BWV 535a Praeludium (Prelude and Fugue) in G minor

Only source: Autograph MS, Mö MS c. 1705/6? (later known to Kellner? KB
p. 583).
Two staves; title and headings, ‘Praeludium cum Fuga ex G pedaliter’,

Prelude ‘Passaggio’, Fugue ‘Allegro’ (at least the last two inscriptions not
autograph?).
The Prelude is shorter (twenty-one bars), with a solo line above an implied
tonic pedal point, simpler figuration in the central section, and the last six
and a half bars similar to BWV 535. The Fugue now lacks the final twelve and
a half bars of the BWV 535 version, which is busier, more continuous and
inventive in its patterns. Probably BWV 535a was continued on a lost piece
of paper originally sewn into the MS (Hill 1991 p. xxiii), but there is no way
of knowing whether the completion was the same: see concluding remark,
BWV 535. Any ‘discrepancy’ felt today between the weight of the movements
may be anachronistic and over-encourage biographical speculation, as in
Breig 1999 p. 654.
How much earlier the work is than the copy in Mö MS is not known, but
perhaps considerably. Some detail, such as a pedal that enters in the fugue
only with the subject, unlike (it seems) BWV 533, suggests that BWV 535a is
not amongst the earliest, despite the Prelude’s simplicity. Or the Fugue was
composed independently, either expressly to be attached or simply ending
up attached to one or other ‘model’ prelude – hence the unusual title ‘cum
Fuga’?
Prelude
Although it might seem that the composer first ‘prefaced a predeliberated
fugue with an improvised prelude’ and then enlarged it ‘to produce a more
symmetrical plan’ (Stauffer 1980 pp. 39, 130), the two present Preludes
55 BWV 535a
need not have been the only versions of what is little more than a series
of formulaic harmonies and note-patterns. The durezza formulae of the
final bars, embroidered in BWV 535, are always open to figural decoration,
especially with such conventional motifs as those here – patterns found in
other keyboard works of c. 1700, such as Bruhns’s ‘Gelobet seist du’.
The term passaggio added above the first bar of the Suite in E minor
BWV 996 in J. G. Walther’s copy of it was defined by Walther himself as
Variatio . . . wenn an statt einer grossen und langen Note, allerhand
geschwinde Läufflein gemacht werden. (1708 p. 153)
A Variation . . . when instead of a large and long note, all kinds of quick
little runs are made.
His examples are not unlike the opening bars of BWV 535a, 996 or 533.
But in BWV 535a, does the word indicate that passage-work is already there
in the lute-like opening section or that the player is free to treat other bars
in this manner? The diminished sevenths in BWV 535 are a passaggio of a
more obvious kind: a ‘passage’ between two dominant pedal points, more
extensive than in BWV 535a.
In view of the distribution between hands in bb. 5–6 – necessary, with
no easy alternative – should bb. 1–5 also be divided? BWV 535 is more
helpful in this respect, and to specify a method when obligatory but not
when optional is common (cf. the Legrenzi Fugue BWV 574, bb. 105/112).
The slurs from b. 10 are unusual and probably indicate that the chords are
played sostenuto, as is implied by slurs in Raison’s table (1688) and more
clearly in Saint-Lambert’s Les Principes du Clavecin (1702).
Fugue
One can presumably trace the composer’s maturity in BWV 535’s greater
sense of climax in bb. 23–4 and the smoother continuity of bb. 35–8, com-
pared to 535a. The two versions of bb. 17–19 look as if the composer, pre-
occupied with little keyboard patterns, re-shuffled them for continuity and
imitation, without avoiding a certain aimlessness when episodes modulate
in several directions – including the dominant, where the entry in b. 32
is surprising. On the other hand, to criticize the modulations in bb. 52ff.
(Krüger 1970 pp. 48ff.) is to underestimate how the texture and stretto
produce a fresh and vigorous effect on the organ.
Particularly significant differences occur between the versions of the
passage bb. 46–65. BWV 535a was strikingly restrained at two points, and
possibly was so at a third, i.e. at b. 65 where the Mö MS’s incomplete bar has
tutti rests above the pedal. Entries in BWV 535a are treated en passant, as
in fugal sections of Buxtehude praeludia where the true climax is reserved
for the toccata postlude. This was presumably the case in BWV 535a. The
56 BWV 535a–536
observation, therefore, that the ‘later version’ rises in intensity and thus
follows ‘the famous rule that the first part of a fugue must be good, the
second better, but the third outstanding’ (Keller 1948 p. 62) seems to be an
anachronism.
On its position in the Bach oeuvre: similarities in theme (repeated notes),
modulation (limited), part-writing (crossing, to little effect) and contrapun-
tal detail (four-part harmonies, square phrases, quaver movement) will be
found between it, the early B Capriccio, and the B minor fugue in the
Sonata BWV 963, as well as here and there in early cantatas. Its move to-
wards a more reasoned form and careful figuration than found in BWV 963
is sometimes anticipated by Reinken, to whom several works in the Mö MS
might be considered a form of homage (Dirksen 1998 p. 135). But Reinken
seldom if ever matches Bach’s harmonic tension and melodic flair, and his
ultimate influence can be overestimated.
An important difference between BWV 535a and the harpsichord fugues
in BWV 992 and 963, or the Albinoni Fugues BWV 946 and 950, is that
its lines are more sustained, with more ties and fewer rests, as if carefully
conceived for organ.
BWV 536 Praeludium (Prelude and Fugue) in A major

No Autograph MS (see BWV 536a); copies in P 804 (prelude by J. P. Kellner
c. 1726/7?, fugue unknown copyist), P 837 (c. 1829, probably from another
source).
Two staves; title in P 804 ‘Praeludium in A. cum Pedale’. Mistakes in P 804
suggest that the Prelude’s source was tablature (KB pp. 474–5).
It used to be supposed that the ‘early version’, BWV 536a, was later
‘re-worked’ in Weimar where pedal e was available (Keller 1948 p. 81)
and remade with a ‘more lively organism’ (Spitta I p. 581), i.e. better use
of note-patterns. More likely, however, is that BWV 536a is neither an early
nor an authentic version, but rather a later arrangement made by L. Scholz
(see BWV 536a).
Because, as Spitta noted, its fugue-subject is somewhat like that of the
opening ‘Concerto’ of Cantata 152 (1714), BWV 536 used to be dated
1715–17 (e.g. Besseler 1955) – described as melodious, like a minuet or
forlana (Krey 1956 p. 191), and inspiring similar counterpoint. But the re-
semblances are too slight to indicate date (KB p. 473), nor need the date of
a vocal piece indicate the date of an instrumental. Similarly, pedal compass
57 BWV 536
with or without e and C is no reliable indication of time and place, since

one cannot know what the composer wrote or how literally in practice any
notation was followed. Here, the Prelude’s construction suggests an earlier
date (with the decorated chords, pedal points, modest length of a North
German toccata) than the Fugue’s (fully fledged ritornello), but this too is
inconclusive.
P 804’s having two copyists has led to the idea that BWV 536 contains
a Bach prelude copied by Kellner, with a fugue by someone else (Kellner
himself? – Humphreys 2000 p. 39). As with BWV 534, hypotheses are
based on identifying ‘weaknesses’ in harmony, counterpoint or modu-
lation. Perhaps the lightness and charm of both movements reflect its
composer’s familiarity with a certain toccata and passacaglia of Bernardo
Pasquini, associated in a lost MS with BWV 536 and once said to have
been copied by Bach (Beisswenger 1992 p. 57). A different conjecture
is that the Prelude once belonged to a Praeludium of four sections, like
BWV 566.
Prelude
Open broken chords were typical of keyboard preludes in major keys, from
Buxtehude’s Prelude in D major BuxWV 139 to mature works of Bach
(BWV 541). The opening ten bars have the conventional harmonies of a
pedal point spread over a large canvas (5/3, then 6/4, then 7/4/2 etc), and
as convention required in this bland spectrum, the first chromatic tone is
the dominant leading-note (b. 11). Pedal points frame the movement as in
BWV 534 or 535, with various keyboard patterns across bb. 15–27, in the
concentrated manner of J. S. Bach – for example, there seems the making
of a fugue over bb. 14–18.
There is a certain glowing, lyrical ton here, familiar in praeludia in bright
keys by Bach (E major) and Buxtehude (BuxWV 151, 141). The opening
arpeggio, which informs the piece from first bar to last, is of a kind found
in J. K. F. Fischer’s Blumenstrauss (Example 28), but more open to pleasing
development. Such figures go on appearing in chorales, as in BWV 651a. The
resulting feel of the prelude, with its wide tessitura, occasional playfulness
(bb. 5–10) and dance-like suspensions in bb. 15–27, is brighter than that of
BuxWV 139.
Example 28
58 BWV 536
A tablature origin would explain why the pedal-lines of the prelude in

both versions are unclear as to (a) when the pedal plays, (b) at which octave.
Perhaps players were given some licence in both respects?
Fugue
1–41 first dominant answer tonal, second real; countersubject

41–65 ‘false stretto’; tonal answer 49 answered en taille; ‘rocking’
figure
65–85 ‘false stretto’; tonal answer, answered in the bass
85–110 F minor, B minor, first with ‘rocking’ figure
110–36 entry and answer in D; episodes
136–53 closer 2-part stretti; tonic in b. 145
153–82 final entry (pedal); coda on scale pattern
The entry in (e.g.) b. 69 is disguised, and only gradually is it clear that this
is not merely an episode stretto. An overall shape is
A 1–45
B 45–153
C 153–82
in which B is characterized by pseudo-stretto, the last of which (from

b. 136) is at one bar not two bars. The original countersubject is hinted
at before it returns above the final entry, and the ‘rocking’ countersub-
ject is useful in the quasi-episode from b. 115. If the fugue-subject really
is derived from bb. 14–16 of the Prelude (bass) and its coda modelled
on the Prelude’s first half, then indeed one might claim that ‘virtually all
the thematically significant material in the prelude returns in the fugue’
(Humphreys 2000) – which would be unusual for the period, probably
unique.
This is an original fugal conception, with a smooth, effortless counter-
point treating the subject almost as an ostinato, an impression heightened
by the fugue’s rhythm and persistent eight-bar phrase. Although the work’s
invention has been called ‘minimal’, merely fourteen variations on a sub-
ject (Humphreys 2000 p. 33), many players agree with Spitta in hearing
a ‘wonderful intensity’ in the sustained three- and four-part counterpoint
(I p. 581), where entries have a more singing quality than even those of
BWV 535 or 578. An unusual effect overall is given by the constant series
of thirds and sixths, brought about in part by elementary stretti and pretty
dance-like cadences (bb. 76, 88, 114, 122, 181), more fluent than those of the
tight permutation fugue in Cantata 152. The particular flavour of such bars
as 60–70 is unusual and, like the non-stop quavers, rather like the moments
59 BWV 536–536a
between cantus firmus phrases in many an organ-chorale. The short final

chord suggests a strong rallentando.
Altogether, the A major Fugue is far more original than its unassuming
lyricism might at first suggest, and neither the canon at b. 136 nor the inner
thirds at bb. 146ff. would be out of place in the Ob. Of course, much of
this could result from a skilled pupil’s adoption of techniques learnt from
Bach works, and the argument for or against authenticity is difficult to take
further. For the player, a further question concerns manual-changing, which
is entirely practical here: the episodes are such that changing is effortless,
even to a third manual during one of them (b. 123).
BWV 536a Praeludium (Prelude and Fugue) in A major

Five copies in Scholz MSS, late eighteenth century (four of fugue only, one
in G major).
BWV 536a is different as follows:
Prelude
bb. 5–9, 12–13: in the inner voice, a single line of quavers only
bb. 10, 15, 16, 20, 25–7 lowest voice played by left hand
Fugue
notated in 3/8
bb. 33–41, 159, 160: pedal an octave lower
bb. 42–3, 89, 90: lowest voice played by left hand
bb. 182–4: three further bars, alluding to subject
BWV 536a would probably represent an early version if the lost source used
in Peters II really was autograph, but this is unlikely (KB p. 587).
On one hand, the difference in the notation of the inner voice from
b. 5 could mean that BWV 536 clarified what BWV 536a merely implied.
On the other, the differences alone between the two versions at bb. 20–1
and 25–7 are such as to suggest that BWV 536a is a typical simplifica-
tion by the Nuremberg organist Leonhard Scholz (1720–98). Differences
between Scholz’s copies probably mean not that he had more than one source
(KB p. 587) but that he had various shots at an arrangement, changing key,
dispensing with pedals, etc.
Irrespective of Scholz, the old idea that the sostenuto notation (held
notes) found in BWV 536a but not 536 is matched by the differing copies of
a rondeau by L.-C. Daquin, found plain in the AMBB (BWV Anh.III 183)
but sostenuto in Couperin’s Deuxième Livre (1717), is valueless: there is no
evidence that the Livre was AMBB’s source.
60 BWV 537
BWV 537 Fantasia and Fugue in C minor

No Autograph MS; source, P 803 (fantasia and bb. 1–89 of fugue copied by
J. T. Krebs, the rest by J. L. Krebs) and a lost copy perhaps once owned by
Kittel.
Two staves; heading, ‘Fantasia con Fuga pro Organo’, at end (J. L. Krebs)
‘Soli deo gloria d[en] 10 Januarii 1751’ (Zietz 1969 pp. 68, 98).
The copy made by J. T. and J. L. Krebs has added glamour through

Griepenkerl’s anecdote that the MS was almost used as waste paper
(Peters III, 1845). Some hear similarities between the Fantasia and the Fugue,
others that two themes per movement produce an overall shape ABABCDC
(Kloppers 1966 p. 22). In playing time, the movements are closer than
is often the case with prelude and fugue pairs for organ, and the details
are complementary: four parts, consistent (Fantasia) or varied (Fugue);
binary Fantasia, ternary Fugue; short imitative theme (Fantasia), long sub-
ject (Fugue). The Fantasia’s half-close is unique in the organ works, its de-
scending bass, hemiola and wandering semiquavers resembling half-closes
in chamber works (E major Violin and Harpsichord Sonata). Such linking
of movements, the first of which is no conventional prelude, could have
inspired the title ‘Fantasia’.
While both the economical style and the sources could point to a Leipzig
origin, as might the compass CD–c , similarities with Weimar chorales
suggest an earlier date; see below. Although the fugue-subject does allow
more complex treatment than it receives (e.g. ‘stretta inversa’), there is
no clear reason for thinking any ‘weaknesses’ in the last ninety bars due
somehow to J. L. Krebs (see below), although Krebs’s own F minor Fugue
does suggest the influence of BWV 537.
Fantasia
The binary form has a half-close or phrygian cadence:
A 1–12 pedal point, imitative upper parts; pedal begins B

B 12–21 imitative upper parts; hemiola close
A 21–31 virtual repeat of first ten bars, parts exchanged
B 31–47 denser development of B, including inversus and pedal;
41–6 = 15–20, partly decorated, parts exchanged
47–8 phrygian cadence, already anticipated in 9–10 and 29–30
This is more ‘cosmopolitan’ than any Italian binary movement. The opening
bars, with pedal point and imitative, wandering upper voices (rather like
61 BWV 537
obbligato wind parts), can remind one of subdued, yearning first movements
of Leipzig cantatas such as BWV 8 or 27; or of French en taille movements
in 6/4 (Grigny, 1699), minus the tenor solo; or of certain northern toccatas
(Buxtehude’s F major Toccata or G minor Praeludium BuxWV 150); or
even of the conventional Orgelpunkttokkata, now given a newly expressive
lease of life. The hemiola of b. 20 was surely known to the composer of the
F minor Prelude bb. 30–1.
Of all these, most like the Fantasia is a certain type of cantata first
movement. The lines, including the semiquavers accumulating towards the
end, are most like woodwind obbligati, despite the idiomatic organ style of
bb. 35–46. Similarly, despite its points d’orgue, the pedal is much like a fine
basso continuo line (e.g. bb. 12–21). And subject B (b. 12) sounds as if made
for a sung text. Since A has the typical leaping minor sixth exclamatio (a cry
of anguish, according to Walther’s Lexicon, p. 233) and B a very different
slurred figure (as if after an intake of breath), they are both ‘vocal’ – wordless
but contrasted and thus musically fruitful.
To the player, as remarkable as the fantasia’s mixed pedigree, careful tex-
ture and a form from which ‘all inorganic passage-work’ has been excluded
is its ‘noble, elegiac’ atmosphere (Spitta I p. 582), which it shares to some
extent with the C minor Prelude BWV 546. While no harmonies above the
pedal point are original, the phraseology is masterly: by b. 6 or b. 7 they
demand a turn to the dominant in b. 10 (like the opening paragraph of
the St John Passion), where the bass takes the opening motif. The result in
bb. 1–11 is an exceptionally well-conceived, natural and unforced state-
ment, in which technique is geared to expressiveness. The inverted theme
in b. 32 is introduced to be not merely ingenious but expressively beautiful,
as is not always the case with J. S. Bach; and although bb. 31–41 has the
theme in every bar, it is no more obliged to do so than the previous sec-
tion. Also remarkable is the almost complete absence of major keys, even as
cadences.
The very opening of the other C minor Fantasia, also somewhat ‘French’
and derived from the old pedal-point toccata, looks deliberately different:
BWV 537.i BWV 562

four parts, 6/4 five parts, 4/4
two subjects, binary form one subject, motivic development
pedal points: tonic, dominant tonic, dominant, subdominant,
relative
Fugue
The violin/organ fugue-subjects referred to by Mattheson (see BWV 539.ii)
imply that BWV 537 is a particular type, with a theme similar to another one
62 BWV 537
quoted by Mattheson, who drew attention to the striking semitone:

Example 29 (1739 p. 209, in G minor).∗ Such comparison is not to lessen
the ‘demonic power’ of Bach’s subject or Spitta’s admiration for it but to
suggest that it has regular, even textbook-like, features: a rising fifth (a run in
bb. 37, 45, as in Example 29), a repeated dominant note (hence a tonal an-
swer), a broken chord (diminished seventh), a tonic end, a four-bar phrase.
Every performer knows the exhilarating moment of the sequence in b. 18.
Example 29
The Fugue as it is in P 803 approaches the da capo perfected in BWV

548.ii:
A 1–28 exposition; episode tutti; tonic entry en taille, then no

pedal
28–57 episode; dominant, tonic, tonic entries; sudden
half-close
B 57–104 irregular exposition of two new fugue-subjects
A 104–28 = bb. 4–28 (en taille entry re-harmonized for pedal
point)
128–30 coda
‘Weaknesses’ observed in the last forty bars by O’Donnell 1989 include

a poor pedal line (?bb. 90–4), a static tonality (bb. 90–108), a banal alto
(b. 100) and poor part-writing (bb. 93, 127), which could all be attributed
to J. L. Krebs, i.e. if he was completing an incomplete fugue by introducing
a da capo, with or without the authority of the composer. Though it is
hardly a fault, starting the da capo with subject instead of answer, as in
BWV 548, leads one to wonder whether Bach reached b. 104 and then wrote
‘da capo’.
This return of A has long been found ‘meagre and unsatisfactory’
(Dickinson 1956 p. 22), as might also be the bass-line before the pedal
entry of b. 110 and the rather sudden pedal point of b. 124. Such ‘problems’
are dealt with in BWV 548 as follows: the da capo starts on a pedal point, the
pedal is absent before coming in with the subject, and A1 already includes
a dominant pedal point which therefore returns in due course. Putative
‘weaknesses’, therefore, might mean only that Bach had not yet perfected
the da capo conception for a fugue.
† If
Mattheson knew Bach’s subject and is quoting from memory, the crucial but forgotten tie in b. 3
implies, perhaps, something about his musicianship.
63 BWV 537
The counterpoint recalls the Weimar chorale ‘Nun komm der Heiden
Heiland’ BWV 661, though without a cantus firmus to compel and propel
its angular line. As is customary, pedal is not reserved for passages with
entries, and there is no marked end to the exposition, which runs across the
pedal paragraph. The ‘decorated suspensions’ style of counterpoint in the
first manual episode (bb. 29–37) is typical of a composer who seems to have
had an inexhaustible supply of it, unto the Art of Fugue itself.
Apart from its subject and its drive, the most striking features of the
Fugue are the da capo and the new fugal section in the middle. Although
the two new subjects of B are not combined with A, as might be expected by
analogy with the F major BWV 540, both have a pedigree. Rising chromatics,
already there in the Fantasia’s last bar, are as traditional in double fugues
as is a scalar theme in plain minims midway (cf. the C minor Prelude
BWV 546). And the quaver countersubject is not only ‘introduced in a
masterly fashion seven bars before’ the B section begins (Keller 1948 p. 83),
but has been gradually emerging throughout the first fifty-seven bars. Its
chief motif is in fact a countersubject to the original main theme from b. 24.
Therefore, although the three themes are not combined, one of them is
made from a motif that combines with the other two, and so adds a new
category to multi-subject fugues in works for organ (or harpsichord: see the
suites BWV 808, 830). This motif is one to appear in many guises: fugues
(BWV 546 or Art of Fugue), chorales (BWV 661), harpsichord works (Italian
Concerto, 1735). See Example 30. A similar motif but beginning on the beat
is also common, e.g. Violin and Harpsichord Sonata BWV 1016.ii b. 4 and
Prelude in B minor WTC2.
Example 30
As to section B: continuous quavers disguise the irregular entries of the

chromatic subject, which is treated imitatively rather than fugally, and since
the second bar of the subject is in effect a sequence to the first, the result is
a series of sequences. The phrase structure of the two sections is therefore
quite different. There is also the possibility that the layout at the beginning of
B and A2 allows stops to be changed or added without breaking continuity
64 BWV 537–538
too much, and (at least to modern ears) the more climactic A2, the more
convincing the da capo becomes. If Krebs was responsible for the manual
trills at b. 101 (reminiscent of the Passacaglia Fugue) and the half-close at
bb. 103–4 (very like that before section B, at b. 57), then he showed a grasp
seldom evident in his own works.
BWV 538 Toccata and Fugue in D minor

No Autograph MS; copies in P 803 (J. G. Walther, 1714–17?), P 1099 (J. G.
Preller), P 416 (later eighteenth century), and others probably via C. P. E.
Bach (e.g. P 290, P 277), J. C. Kittel (e.g. P 275) or J. P. Kellner (e.g. P 286);
separately in late derivative copies.
Two staves; title in P 803 (by whom?) ‘Toccata con Fuga’, in P 1099 (and
others) ‘Toccata ex D mol. per l’Organo â due Clavier et Pedal col la Fuga’;
in Forkel’s list (1802), ‘Prel.’. P 803 writes ‘O’, ‘Positif’ (b. 13 only) and ‘R’
(no other MSS use ‘R’).
The ‘Dorian Toccata and Fugue’ has no exclusive right to this name – already
there in 1845 (Peters III) – since sources also transmit BWV 549a and 588
without key-signature. More unusual is that except for the concertos, it is the
only work in which authentic manual changes are related to the structure.
This duologue-toccata has no exact parallel in or outside organ music and
is barely related to French dialogues.
Also unique is a claim on the copy by Kittel’s pupil Fischer that the
(whole?) work was ‘played at the examination of the large organ in Kassel by
S. Bach’ (‘bey der Probe der grossen Orgel in Cassel von S. Bach gespielt’),
a rebuilt organ in the Martinikirche. There was such an examination in
September 1732 (Dok II pp. 226–7) but neither stop-list nor manual-layout
is known, nor whether a public recital as such was involved. A Weimar
work could have been used on this occasion, revised or not, for as with so
many other pairs of preludes and fugues, MS variants imply more than one
original autograph, perhaps used in various connections. Walther’s copy
(which alone with P 416 gives all manual changes) may derive from the
earliest version, and Preller’s from one in which the fugue was notated in
4/2 time; sources associated with C. P. E. Bach are generous with ornaments
in both movements. One can only conjecture why Walther uses ‘R’ when the
Rückpositiv was rare in Thuringia, but he does in other MSS too. ‘Positif’
signifies any secondary manual.
The Toccata is a web of allusion to historical organ-music, and like BWV
562 virtually monothematic. Yet this is no fantasia woven from French
65 BWV 538
motifs over pedal points, but rather a concerto-like fantasia of basic ‘North
German’ figuration. In more than key it seems to recall the first couplet in
Buxtehude’s Magnificat primi toni, though the pedal point of bb. 78–9 makes
it certain that the composer also knew a further D minor work, the Vivaldi
Concerto BWV 596. In fact, these two works suggest that in BWV 538.i the
composer consciously combined the theme-type of earlier preludes such
as BWV 536 or 550 with a ritornello form learnt from up-to-date Italian
concertos, producing a unique amalgam of diverse forms and styles.
The Fugue may be older, to judge by certain separate copies and details of
notation (KB p. 365). Its counterpoint is exceptionally well reasoned, with
several countersubjects and idiosyncratic harmonies produced by stretti,
which turn out to be its spectacular achievement. Thus Prelude and Fugue
are complements: similar enough in length to form a more obvious pair
than the F major Toccata and Fugue, and closer in style than some other
supposed pairs.
Toccata
Resemblances are often found between the basic material of this movement
and other keyboard works in D minor or tonus primus, by Raison (Agnus
dei), Pachelbel (a Toccata and Praeludium), J. K. F. Fischer (a Praeludium),
or Buxtehude (Magnificat). Reinken’s Fugue in BWV 966 contains the basic
motif, in D minor and its relative; and something like it in F major also
appears in the course of the Toccata in F (b. 229).
Despite its original aura, the movement is close to other perpetuum
mobile toccatas with marked cadences but without clear returning theme
(Breig 1986a p. 33). The square motif (Example 31) seems to salute various
keyboard figures used but never so thoroughly explored in the praeludia
of Lübeck, Bruhns and others, or even in the G major BWV 550. Yet all
dialogue-types – Italian concerto, French mass, English double voluntary,
Spanish medio registro tiento – share characteristics: two manuals in alterna-
tion with the same melody, or one for bass and one for treble, accompanying
each other and joining together at the end. Using them antiphonally for the
sequences in bb. 43–5 or 73–7 looks like a more sophisticated working of
something in the Concerto BWV 595.i, bb. 3ff.
Example 31
To use the manuals in more or less simple alternation seems to be the

chief aim of the piece: there is no real récit or en taille as in French dialogues,
66 BWV 538
no fugal development as in tientos. Nor, despite the first change of manual in

b. 13, do the manuals belong only to main or secondary sections respectively;
rather, they appear in both. Themes and form are integrated. According
to simple principles of rhetoric outlined by Mattheson, the shape of the
movement can be expressed as:
Ow = Oberwerk, Pos = (Rück)positiv

Ow 1–13 Exordium by first speaker A, i.e. main theme, becoming:
Narratio, i.e. the theme develops; then
Propositio (5): further repetition, emphasis,
development, close in tonic (for the cadence,
cf. Contrapunctus III, Art of Fugue)
Pos 13–20 Confutatio, controversy: subject taken up by dialectic
partner B
Ow 20–5 Confirmatio, confirming main theme, further repetition
Pos 25–9 Confutatio, taking up 1–5 in dominant, parts exchanged
Ow 29–37 Confirmatio: A answers and develops, B interrupts with
antitheses
Pos 37–43 Confutatio: B variant theme (tenor 34), A’s antitheses
Ow 43–67 Confirmatio: new variant by A, answered at once by B,
further developed by A, who (47) re-introduces material
from 1–5
Pos 67–81 Confutatio, B interrupts when its material (37) is referred
to by A (66–7); B closes in tonic (73), A then with
material from 43; B answers twice, then speaks at the
same time (from 78); motifs repeated and accumulated
(congeries) for a climax (gradatio)
Ow 81–94 Confirmatio: A takes over before B has finished, refers
back (to 53), confirms the dominant (88) and produces
his own high point (90–4, now in contrary motion)
towards D major
94–9 Peroratio, exit, conclusion, coda
The outermost sections have no dialoguing.

If ‘such a complete approximation to speech’ is found in no other work
of J. S. Bach (Kloppers 1966 p. 90), nevertheless the rhetoric is purely mu-
sical: ‘approximation to speech’ is not what gives this movement its formal
perfection. No doubt, as J. A. Birnbaum claimed in 1739, Bach knew rules
and terms of rhetoric:
Die Theile und Vortheile, welche die Ausarbeitung eines musikalischen
Stücks mit der Rednerkunst gemein hat, kennet er so vollkommen, dass
man ihm nicht nur mit einem ersättigenden Vergnügen höret, wenn er
67 BWV 538
seine gründlichen Unterredungen auf die Aehnlichkeit und

Uebereinstimmung beyder lenket; sondern man bewundert auch die
geschickte Anwendung derselben, in seinen Arbeiten. (Dok II p. 352)
He understands so thoroughly the parts and benefits which the composing
of a piece of music has in common with oratory that not only does one
listen to him with a satisfying pleasure whenever he directs his profound
conversation to the similarity and correspondence between the two, but
one also admires the clever application of the same in his musical works.
But at most, BWV 538 merely illustrates the post facto descriptions of ars
rhetorica. Furthermore, so homogeneous is the material that the ritornello
structure is rarely clear:
1 A, 13 episode (a varied repeat of bb. 7–12)

20 A, 37 episode
47 A, 53 episode
58 A, 66 episode, corresponding to 37–46
81 A (but as b. 53), 94 coda
This could be seen as having three parts, the central one bb. 37–81. The
movement looks like an updated reworking of old German 4/4 semiquaver
motifs, the kind of thing found in Reinken’s Sonata reworked as BWV 966.
Some of the differences in detail in the MS sources could reflect later revision,
and perhaps note-patterns were even more uniform in the ‘first version’.
Bars 37–81 provide a striking example of the mature Bach organ
prelude, with returning phrases transposed but otherwise scarcely altered
(the sign of Italian concerto influence) and an overall symmetry (this is
the middle of three sections). The motifs, both quaver and semiquaver,
seem self-generating, different but unmistakable. As is clear from the homo-
geneity, this is no ordinary ritornello form: compare the passages from
bb. 7 (pedal), 15 (rh), 30 (lh), 53 and 81. Similarly, the main motif can be
used to create a pedal point (b. 86) or put above a pedal point in imitation
etc (b. 30). This kind of homogeneous music of a distinctive melos, one cast
in a complex ritornello form, is found again in the F major Toccata, but
clearly to different effect and much less economically.
Throughout, the rhythms are unusually square, to some extent coun-
teracted by phrase-lengths (e.g. six-bar phrase bb. 37–42) but produc-
ing remarkably few tied notes. The result is a highly unusual movement
characterized from first bar to last by little groups of four semiquavers.
Allied to this is a bland harmonic spectrum, with some conventional mo-
ments (compare bb. 8–9 with harmonizations of the D major fugue sub-
ject, BWV 532), and ‘interesting’ chords only at carefully timed intervals
(bb. 12, 35, 52, 65, 72, 93), three of them (bb. 51–2, 64–5, 93–4) functioning
68 BWV 538
as ritornelli. When there has been some rich harmony, the following passage
‘clears the air’ with a simple figure or sequence (e.g. bb. 35–7, 52–3). It is
difficult to see how any of this could have been applied again to another
composition: the toccata must remain an unicum.
Fugue
The Fugue, aeolian rather than dorian, is also an unusually complex move-
ment based on a curiously symmetrical theme that rises and falls an octave,
starts simply but runs into syncopations, preserves some alla breve ele-
ments (2/2 metre, suspensions, dactyl countersubject), and in some sources
is ornamented.
Unusual main features are that the episodes are canonic and that the sub-
ject has two countersubjects (b. 18), producing not so much a permutation
fugue as an overlapping counterpoint often confusing to the ear. Although
the pedal has three conspicuous tonic entries, they do not so much underline
a ternary canzona-fugue as imply a massive ostinato, not unlike the tonic
pedal entries in the Fugue in E.
I 1–36 exposition; two countersubjects (12, motif from the

toccata – see Example 31); from the codetta (15–17,
25–8) an imitative sequence leads to later development
36–42 episode, brief canon at the fifth in outer parts
43–56 entry, tonic, then episode, three parts canonic
57–63 entry, tonic, at first decorated
64–100 entries, dominant (71 = 18ff.), tonic (81); episode
sequences
II 101–66 entries in F (canonic, 101–2), C (115), G minor
(canonic, 130), B flat (146); episodes based on the
sequence
III 167–74 tonic entry in canon
174–202 episodes on the sequence; dominant entry (188)
203–11 tonic entry in canon (soprano entry decorated)
211–22 coda based on four-part version of x; final homophony
An alternative view is of four ‘sections’: 1–43, 43–101, 101–67, 167–end.

Already in 1777, Kirnberger was quoting excerpts from the Fugue to
demonstrate the composer’s use of sevenths and ninths (Dok III pp. 226–7),
as well he might. It is noticeable that neither of the countersubjects, first seen
together in b. 18, contains suspensions or tied notes; rather, the mainspring
of the movement comes from the canonic potential of the subject itself,
particularly in what seems to be a derived codetta (bb. 15–16) which yields
an exceptional series of imitative episodes throughout the fugue. From this
69 BWV 538
canonic seed grow imitations at all intervals except the third and seventh,
either at the bar or half bar, and all invertible.
Variety is achieved by avoiding simple repetition, creating canons at
different intervals, and varying the number of parts. The free parts vary –
chromatic from b. 156 – while passages of even freer quaver lines grow out
of the current and throw the canons into greater relief (bb. 64–7, 195–202).
One is often reminded here of later passages in the Art of Fugue, such as the
semiquaver counterpoint in the alla francese fugue and the quaver lines in
Contrapunctus III, all in D minor. The canon to which the subject itself is
susceptible produces parallel rhythms (as in the A minor Fugue WTC 1),
and clearly it is the episodes that give most variety. This variety may be
shown by comparing treatments of the same phrase, as in Example 32. Or
two different settings of the same bass line may also be compared, such as
Example 32
70 BWV 538–539
bb. 36–42 and 211–17. All of the suspensions produced in all of these bars
form a stark contrast to the style of the toccata, surely by design. The result
is a tour de force, so that a crucial passage from b. 125 has been said to ‘defy
harmonic analysis’ (Bullivant 1959 p. 539), a pardonable exaggeration in
the circumstances.
BWV 538 produces some of the most carefully argued four-part harmony
in the organ repertoire. In any pair of similar passages, two of the four voices
may well be identical; but the other two, without apparent contrivance, dis-
play a totally different harmonic character. For instance, compare bb. 43–50
with 115–22. The densest episode precedes the middle entries in the relative
major, producing a splendid inner line in minims which may or may not
refer to the head of the subject (alto b. 93, tenor bb. 95ff). A further effective
detail is that each middle entry is preceded by a strong perfect cadence. Al-
though an extra part appears immediately after the fugue’s loosest texture
so as to complete the canons in thirds and sixths (b. 164), the harmony
becomes richer as the coda gradually loses its quavers. The natural skill with
which the subject is re-harmonized, or the canonic interval made to vary,
or the countersubject’s quavers are effortlessly spun, is spectacular.
Most surprising of all are the last four bars of the fugue, harking back to
the toccata’s dialogue, dispelling any danger there might be of too didactic
a counterpoint, and offering an uplift to the spirit. In view of those last
four bars, and the obligatory (not optional) manual changes in the Toccata,
perhaps the Fugue is also a dialogue – now not obligatory but optional? There
is no great difficulty in playing all the themes and entries on Oberwerk, all
the codetta and episode canons on Positif (the first change in b. 15: see
Williams 2000). No other work of Bach allows this quite so patently.
BWV 539 Prelude and Fugue in D minor

No Autograph MS; movements paired in early nineteenth-century copies
(e.g. P 517, also Forkel, 1802); fugue only, second half eighteenth cen-
tury (Am.B.606, P 213) and later, copies probably all from one source (KB
p. 360).
Two staves (no indication of pedals in the Prelude); in P 213, one of six
fugues per il Clavicembalo, but with pedal cues.
Although it was once assumed that differences between this fugue and the
solo violin Fugue in G minor, Sonata BWV 1001.ii, were made by J. S. Bach in
the course of transcribing (Spitta I pp. 688–9), and that these say something
about his methods (e.g. Geiringer 1966 pp. 237–8), it is not known who
made the organ version or when. Readings suggest it was prepared from
71 BWV 539
P 268, Anna Magdalena’s copy of the violin sonatas made between 1725 and
c. 1733 (KB p. 354). Nor is it certain who composed the Prelude, whether it
was for organ, and who coupled the two movements in P 517. (This copyist
wrote out other transcribed works including the Concertos for Three and
Four Harpsichords.) So it is fruitless to speculate why Bach did not also
transcribe the violin’s ‘sublime and deeply passionate prelude’, substituting
for it ‘a little, insignificant praeambulum’ (Keller 1948 p. 99). The Prelude’s
authenticity, however, can ‘scarcely be doubted’ (Kilian 1961).
A separate history for the Fugue is implied by its separate sources, where
it is often transmitted with the Albinoni Fugues BWV 951 and 951a. The
transposition from G minor to D minor lowered the compass from f to c ,
also allowed some treble entries to be put up an octave, thereby extending
the range upwards as well as downwards (with new tenor or bass entries).
The pedal, which does not rise above tenor a, forces bb. 92–3 to be given
to the left hand, which it crosses at three of its four entries, and is reserved
largely for basso continuo – features quite untypical of Bach’s organ fugues.
The fugue was also transcribed into French lute tablature, probably be-
fore c. 1730 (Schulze 1966), by or for the lutenist J. C. Weyrauch (A. Burguéte
BJ 1977 p. 45). Whether the violin sonata was the (or only) original is un-
known, but both lute and organ versions appear to be made from it, not one
from the other, their additions appearing at different points in the work: or-
gan at bb. 5 and 28 of the violin version, lute bb. 2 and 5. Though more than
competent, the organ’s version of violin-writing is unlike that of authentic
arrangements, such as the violin concertos for harpsichord, and, though
perhaps quite typical of the time, spurious.
Prelude
This, whoever wrote it, may have been meant to resemble plein jeu or petit
plein jeu pieces in French organ masses, where the various quaver figures and
suspended chords such as the 9/7/5 in b. 20 could be found. Of all the organ
music in Schmieder’s BWV, this is the piece most plausibly played with
notes inégales for the conjunct quavers, especially in view of the harpsichord
idiom of the part-writing (bb. 3, 9, 19 etc.), whether or not organists of
Kirnberger’s period were intimate with French style.
A harpsichord piece similar in its suspensions to the Prelude is the
A minor Fantasia BWV 904.i, and both appear in one early-nineteenth-
century MS, though not together (Schulze 1977 p. 79). BWV 539 has a
miniature closed form:
7–12 = 1–6 in dominant, outer parts in inverted counterpoint

13–33 sequences towards half-close, then towards tonic return
34–9 = 1–6 in tonic
40–3 coda (41–3 = 22–4 in tonic)
72 BWV 539
This is near to a binary form, except that the ‘first half’ takes a long time to
cadence in the dominant and is longer than the second – details untypical of
Bach, as is the inconsistent texture. A simple, ‘French’ use is made of scales,
suspensions, and (from b. 24) sequences, all of which lead to striking har-
monies in more than half the bars. But plain cadences without suspensions
are not typical of French durezza styles, and the result is a prelude of mixed
genre, if charming and interesting.
Fugue
That all three fugues or fugue-subjects in the Six Solos for Violin (G minor,
C major, A minor) are archetypes – A minor a short theme of great potential,
C major a model for chromatic counterpoint – is suggested by Mattheson’s
quoting these two, the latter from an audition for organists (Dok II
pp. 294–5). The G minor Fugue represents the third archetype: a model can-
zona subject. Yet a fourth is found in the Albinoni Fugues in B minor, i.e. a
long melodious subject of the kind known in violin music from Frescobaldi
onwards.
Like the other violin-sonata fugues, BWV 539.ii has a ritornello structure
in which the subject (insistent, deliberate) contrasts with episodes (fluent,
fleeting). The subject has the repeated notes, and its countersubject the
implied suspensions, of countless canzonas, allowing easy invertibility and
even an extra entry (b. 5) in the irregular and almost Palestrinian exposition.
Just as bb. 5–7 are more than beginner’s work, so the accompaniments
added to episodes (bb. 8, 44, 66, 89) are no elementary block chords: an
intense, detached way of ‘placing’ them can achieve a remarkable intensity
in performance. Only a theoretical comparison with the violin version leads
to an opinion that the arrangement nowhere goes ‘beyond the scholastic’
(‘über das Schulmässige’: Ulrich Siegele, quoted in Kilian 1961).
1–7 irregular; two pairs of octave stretti; sixth part on the

mediant
7–15 episode, including reference to subject
15–30 a ‘second exposition’; stretti at 3rd and 4th; stretto
episode 25ff.
30–57 episode, first based on melodic extension of subject
57–60 stretto as at 25, subdominant, to relative
60–76 episode, first based on melodic extension of subject
76–81 partial entries, subject developed and followed by:
82–92 episodes and coda
The entries become less and less marked, although the change in texture
from episode (semiquavers, open texture) to entry (quavers, more closed)
73 BWV 539
makes them almost as clear to the listener as in the violin version, where the
entries are chordal.
A curious result of the many stretti is that the subject could be introduced
into the harmony more often than it is (e.g. in bb. 11–12 or 55). The tendency
for the fugue to go into five parts either for stretto entries or when the
harmony hangs fire, as in a string concerto (bb. 37–9, 85 – compare the
Vivaldi BWV 593.i b. 9), corresponds to the violin’s tendency to use four
strings when feasible. The ‘strain’ of double-stopping inspired the arranger
to find a comparable effect: see Example 33. Though creating two parts from
the original solo-episodes is not very systematic, it is not ‘unklavieristisch’
(as Kilian 1961 p. 327 claims).
Example 33
The lute version has a fairly regular exposition of tonic subjects and
dominant answers, the violin something less regular, the organ less regular
still (s = subdominant, m = mediant):
violin BWV 1001 bb. 1–5 dttd

lute BWV 1000 bb. 1–7 dtdtsdt
organ BWV 539 bb. 1–6 12 dttddm
The last version has many points of interest. The episodes produce new
organ textures, create possible echoes (from b. 49), and anticipate other
pieces (b. 66 – see BWV 565). The sudden springing up of fiery episodes
is in the Italian manner already perfected in Corelli’s Op. 5, and the final
cadenza is more like certain concerto ‘cadenzas’ (e.g. Triple Concerto in D
minor BWV 1063.ii) than those concluding organ praeludia. The energy
of the great organ fugues is replaced in BWV 539 by constantly re-worked
harmonies.
While the melodic inspiration of bb. 32–7 or 77–9 is difficult to attribute
to any composer but J. S. Bach, or at least a gifted pupil, the work is too
74 BWV 539–540
unlike authentic organ fugues (because less fluent) or known authentic

transcriptions (because more literal) for its authorship to be clear. As with
the lute version, perhaps friends or pupils were authorized (even super-
vised?) to widen the repertory by making such transcriptions.
BWV 540 Toccata and Fugue in F major

No Autograph MS; copies of both only in P 803 (Toccata copied by
J. T. Krebs c. 1714, Fugue by J. L. Krebs before 1731?), P 277 (from lost
Kirnberger/Agricola source? – KB p. 218), P 290 (via C. P. E. Bach?) and P 596
(eighteenth century), also a lost Kellner source; toccata only, in eighteenth-
century (P 1009 J. C. Kittel?, P 289) and nineteenth-century copies (e.g. Lpz
Poel 16 with an anon. fugue); fugue only, in eighteenth-century (P 287, Lpz
MB MS 3 J. A. G. Wechmar?, and a MS perhaps once owned by Christian
Bach: KB p. 171) and nineteenth-century copies.
Two staves; heading in P 803, ‘Toccata col pedale obligato’; first movement,
‘toccata’ in P 289 etc., but ‘preludio’ in P 277 and Forkel’s list (1802), etc.
Several conjectures are usually made about this work. The Toccata ‘dates
from a later, maturer stage of mastery’ than the Fugue (BG 15); or, on the
contrary, is some twenty years older; or it was connected with the Weissenfels
organ and its compass of pedal f and manual c , for/after a visit in 1712
(but the Weimar organ too may have had pedal f ); or the ‘Aria in F’ was
an interlude between them; or, with its distinct sections, this Toccata is
earlier than the D minor, BWV 538 (Zehnder 1995 p. 317). There is no clear
evidence for any of these conjectures. Most sources give the movements
separately, few as part of a regular collection of Bach works.
On the compass: both Toccata and Fugue use manual top c conspicu-
ously, but notes above were avoided. In P 803, the Toccata pedal part does
not go above c and is assumed to be a reduction (KB pp. 404–5), although
the organ of Buttstädt (1696) where J. L. Krebs became organist in 1721
had a pedal to f . Was the f -form written for him? Either way, the different
compass requirements serve as a reminder that works circulated in more
than one version, paired or not.
On the pairing: while different pedal compass does not prove that the
Toccata and Fugue originated at different times, it might suggest it. Nothing
in any of the copies’ title or cuing reliably indicates a pairing, and there
are further pointers to a different origin: transmission via J. C. Kittel seems
to have been of the Toccata only; the Fugue-only copies seem to be re-
lated; and the oldest extant pair is the work of copyists who, though related
75 BWV 540
and in Bach’s circle, were probably writing years apart, the younger per-
haps inserting the Fugue (KB p. 405). The pairing seems not to have been
obligatory or even expected.
Neverthless, since their difference in length, flow, shape and effect makes
the Toccata and Fugue complementary, their (optional) coupling was not
inappropriate, whenever it was first done. Drama is contrasted with con-
trapuntal ingenuity, and just as one is the composer’s longest extant organ
prelude, so the other is his only straightforward, integrated double organ
fugue.
Toccata
This gigantic movement couples a pedal toccata with a ritornello section in
a ratio of 2 : 3. The latter’s main theme is as Example 34 (i) and that of the
episodes as (ii), not vice-versa.
Example 34
The sections are continuous, and the overall shape can be described in
various ways:
Voigt BJ 1912 p. 36 A introduction, B ritornello, C coda

Sackmann 1985 3 sections: bars 1–176, 176–364, 365–438
Breig 1999 p. 697 2 sections: bars 1–176, 176–438, with ritornello
(176, 238, 290, 352, 382), interrupted cadence
(204, 318, 424), trio-episode (219, 271, 333)
Further details are:
A1 1–55 tonic pedal point below two-part near-canon∗

55–82 pedal solo, chief motif from 1; cadence figure 81
A2 83–137 dominant = 1–55 parts exchanged, modified
accordingly
137–76 pedal solo, as before but now to C minor, to prepare for:
B1 176–219 new related figure, imitative, four-bar sequence
(176–92), cadence figure; interrupted cadence;
Neapolitan; to relative
A3 219–38 opening material in three-part octave imitation,
D minor
∗ Hard to see as ‘modeled after’ Vivaldi’s BWV 596 (Wolff 2000 p. 126).
76 BWV 540
B2 238–70 as B1 in D minor but without interrupted-cadence

section
A4 270–90 as A3 in A minor, three parts exchanged
B3 290–332 as B1 in A minor (+ interrupted-cadence section),
sequence to:
A5 332–52 as A3 in G minor, three parts exchanged
B4 352–438 begins as B1 in G minor; last sequence (352–67)
changes direction, from B to F to C, for pedal-point
to end; as B1
No scheme, however, can convey the feeling of ‘endless song’ in the move-
ment, as if it were spinning out continuous melody to defy analytical labels,
gloriously massive.
While the main themes and episode are related – familiar keyboard
figures in canon or imitation – the toccata is by no means monothematic.
Its first two motifs (bb. 3–4) appear together less often than expected, and
recall other music in F whose second chord is a 4/2, such as Cantata 1, or
an aria in Cantata 208 that begins rather similarly (No. 13, c. 1712). Note
too the transposed B A C H references in bb. 204–7, 318–21 and 242–7. The
ritornello material modulating in regular steps while the episodes do not
modulate has suggested Torelli rather than Vivaldi as an influence (Zehnder
1991 pp. 90f.), though whether this means that BWV 540.i predates Bach’s
acquaintance with Vivaldi’s Op. 7 and 3 is doubtful. It might, however: the
Toccata in F is much like the Toccata in C writ large, like it developing the
principle of alternating themes and doing so in a more regular way than is
typical of the ritornello form of Italian concertos.
Very striking to the listener is the rhythm of the cadence figure, so much
that it becomes a kind of mini-rondo. The same figure leads to one of the
most startling interrupted cadences even in J. S. Bach’s peerless repertory
of them (Example 35). At the end (bb. 423–4) it is even more startling
in a major key. However, the same major cadence, enthused over by Felix
Mendelssohn in a letter of 3 September 1831 to his sister, occurs in the
Chromatic Fantasia BWV 903 (see bb. 54–5, 56–7). Its repeated effect in the
Toccata is without parallel, underlining amongst other things how necessary
the final dominant pedal-point is to the movement’s tonality.
As in other ritornello movements of Bach, material can be modified or
its order changed without any perceptible break. At some points, one cannot
foretell what the next section is to be. Yet it does not seem unnatural that
section B2 passes back to A without the interrupted cadence heard earlier, or
that the first B sequence (from b. 176) is a less complete circle of fifths than at
b. 352. Both main themes – each an octave canon on a subject used in various
77 BWV 540
Example 35
guises by various composers including J. S. Bach (e.g. Two-part Invention

in A minor) – are somewhat simplistic, throwing yet further weight on
interrupted cadences and novel ways of treating other progressions, such as
the Neapolitan sixth at b. 432. Similarly, while the cadence figure of b. 81
is not original (compare the final cadence of BWV 543.ii or the C major
Prelude WTC1), its extension and sudden minor turn in b. 169 are striking.
The figure was later taken up by J. L. Krebs in his Prelude in C major and
Toccata in E major.
The main melodic idea (octave imitation above a pedal point in 3/8
time) can be heard at the beginning of the later motet BWV 226, while the
main formal idea (ritornello, with its motifs heard in episodes) is found
in several of the English Suite preludes, which indeed have a broad fam-
ily likeness to the Toccata in F. In addition to length and thoroughness,
the Toccata’s contrapuntal handling, harmonic progressions and dramatic
pedal-points distinguish it, while it combines ideas current in other kinds
of toccata: tonic/dominant pedal points of ‘southern’ toccatas (Pachelbel,
Fischer, Kerll), pedal solos of ‘northern’ (Buxtehude, Bruhns). The three-
part invertibility at A3, A4 and A5 is not so patent elsewhere in contempo-
rary organ music. It could be that this invertibility, like the opening octave
canon, salutes traditional keyboard devices, as in Example 36 or in certain
Italian vocal music, e.g. Handel’s Dixit Dominus HWV 232.vi. More complex
counterpoint is reserved for the double subject of the fugue.
Example 36
To player and listener, the sustained energy of the toccata is incompara-

ble in its very reliance on simple elements. Despite the traditional tonic and
78 BWV 540
dominant pedal-points, the tonality is varied, and even the final cadence is
no platitude but almost a surprise. Obviously the motifs themselves mod-
ulate effortlessly. The second pedal solo is an interesting case, for if the
sources convey the composer’s intentions, its phrase-lengths change as the
line approaches the celebrated high pedal f :
bb. 137–68, 32 bars built up from two- and one-bar phrases:

2, 2, 1, 1, 1, 1, 1, 2, 2, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2
This hints at a textual crux, since the isolated one-bar phrase just before
halfway is at variance with the first pedal solo a hundred bars earlier. This
crux would still be there if bb. 152–65 or 156–9 were omitted in performance,
or were an addition by the composer, as one might freely conjecture –
though this is unlikely, for the second solo would then not balance the first,
as presumably it should.
The fact that the key motif of the movement is open to a two-bar or
one-bar interpretation (see added slurs in Example 37) is striking, and
recalls other examples of motifs in single- or double-length versions in
the Orgelbüchlein. The sheer number of variants this pattern gives rise to is
unique, leaving the impression that every group of six semiquavers is related.
The movement is ingenious in its use of the two basic motifs (Example 37
and the cadence figure), and plays with the obvious contrast between them.
They merge in the final dominant pedal point, which unites the rhythm of
one with the simple harmony of the other in a new kind of climax, insistent,
powerful, symphonic.
Example 37
Despite its obvious indebtedness, J. L. Krebs’s E major Toccata does not

offer a useful model for BWV 540 in its use of two manuals except in a
general way. That is, the possibility remains that the Weimar organists – J. T.
and J. L. Krebs, Walther and Bach – did change manuals in long ritornello
movements.
Fugue
While not unlike the D minor Fugue BWV 538 in rhythm, or the C minor
and E Fugues in its thematic combinations, this movement is a unique
example of the alla breve or ricercar fugue in which themes are separately
exposed and then combined:
79 BWV 540
A 1–23 exposition, consistent countersubject

23–70 tonic (30, 49, 56) and dominant (39) entries;
episodes from countersubject
B 70–93 irregular four-part exposition of new subject
(answer with subject-caput 75, further answer 81,
further subject 88)
93–128 further entries (without answers) in D minor, G
minor and C minor; episodes from countersubject
of B
A 128–33 return in tonic
A+B 134–70 entries of subject A in C, D minor, D minor, B, F
and F, accompanied by subject B, complete (134),
almost complete (142–3, 153–4, 158–9, 163–4),
incomplete (147–50)
The cumulative effect is therefore based on three levels: thematic (A, B, then
A + B), rhythmic (more and more quavers), and tonal (more key changes
towards the end). Probably for the second of these three, the composer
disguised most combinations of A and B by changing the first bar or so of B.
Its original caput would have held up the rhythm and harmony and drawn
too much attention to the combination.
The organ-writing is of a distinct style found elsewhere, e.g. in the Mag-
nificat BWV 733. Even for J. S. Bach, however, the counterpoint – a good
example of ‘cantabile polyphony’ (Besseler 1955) – seems effortless, espe-
cially in the last twenty bars: two subjects, spinning quavers, sure tonal grasp
(three entries in near-stretto), idiomatic texture (opening to its widest for the
final pedal entry), finally rounded off by three bars even more succinct than
similar closes elsewhere (e.g. BWV 537). The subject has the white notes,
incipient chromaticism, suspension and simple cadence of many such alla
breve themes (Pachelbel, J. C. Bach), and even the absence of codetta between
subject and answer is a common feature. Note the important crotchets, typ-
ical of the style (cf. E major Fugue WTC2), as are the contrary motion and
nota cambiata (bass, bb. 6–7). The countersubject crotchets produce fine
alla breve stretti in lower voices from b. 55 and recall other music: compare
bb. 61–2 with the A minor Fugue WTC1, in half-note values. And they can be
inversus (first in b. 37) or run across an entry (second subject in bb. 69–70).
Dactyl quavers also flavour the second fugue-subject, but differently, now
as a broken chord.
This second subject is a ‘character theme’, strong in rhythm, a bigger
contrast to the first than is the case in the Legrenzi Fugue. It produces a
quaver line as true to its tradition as the crotchet line was to its, taking
on various shapes and spun out right to the end. Quaver lines in Bach are
80 BWV 540
usually fertile, and as in the C minor Fugue WTC1, those here are fluent
and infinitely adaptable, though in principle merely built up from conven-
tional patterns. These patterns can appear in melody or bass – as in the
A Fugue WTC1 (there, semiquavers in 4/4) – and they can be twisted to
produce harmonic effects that herald the ‘clean’ subject (bb. 125–8). Even
if in bb. 125–8 Bach’s ‘diatonic sense failed him’ (Dalton 1966), the modu-
lation from C/F minor to D minor is presenting the same quavers in a new,
disturbed light. At other moments, the line is much like that elsewhere: see
Example 38.
Example 38
Further understanding of the composer’s methods is gained by com-

paring the bars after each complete subject entry, or by tracing how the
minor middle entries of B occur in order (bottom, middle, top). The Fugue
is working on several levels at once: style (alla breve elements), figuration
(quaver lines), fugal counterpoint (combining themes), key-structure (only
tonic and dominant for the first half), and texture (dense opening, wide
final entry), all more so even than the Toccata. This is far from the modest
examples of A B A + B form in Pachelbel’s Magnificat Fugues.
81 BWV 541
BWV 541 Praeludium (Prelude and Fugue) in G major

Autograph MS: SBB N. Mus. ms. 378 (c. 1733?); copies deriving from another
autograph, in P 288 (J. P. Kellner 1726/7?), P 595 (J. Ringk), Lpz MB MS 7
(J. G. Preller 1749), and perhaps via C. P. E. Bach (P 290 and P 597) or Kittel
(P 320, LM 4839) or other (Am.B.543, Kirnberger circle). BG 15 used a MS
‘with many corrections in the hand of the composer’; in P. 288, first thirteen
bars of the finale of Sonata BWV 528 added by J. C. Westphal (†1828) after
the Fugue.
Two staves; autograph heading, ‘Praeludium pro Organo con Pedal: obligat:’
and ‘Vivace’ (this also in Forkel, 1802); in P 288 (oldest extant copy?),
‘Praeludium con Fuga Pedalit: ex G’.
As Dürr observes (1984, plates 44, 45), the paper of the fair copy autograph
of 1733, once owned by W. F. Bach, is known only from letters written by
C. P. E. and J. S. Bach, including one connected with W. F. Bach’s applica-
tion at the Sophienkirche, Dresden in 1733. It is likely that this copy was
made by Sebastian specially for Friedemann’s audition on the Silbermann
organ – a work for his repertory or even the test-piece itself (Schulze 1984
p. 17).
Although Kellner’s copy is marked after the prelude ‘verte fuga’, ‘turn to
the fugue’, and after the fugue ‘Il fine’ (Kilian 1969 pp. 16–17), Westphal
might have seen an authorized copy with the trio between Prelude and
Fugue (KB pp. 428, 435). More likely, however, is that he added it on a
fancied parallel with BWV 545.
None of the extant MSS derives from the autograph, but Kellner, Ringk
and Preller have a common original (KB p. 429), and C. P. E. Bach may
have known a further original. To date the composition as early as 1712/14
because the Prelude has a hybrid form – opens with an ‘old’ passaggio and
continues with a ‘new’ ritornello – and because the Fugue’s subject seems
to recall Cantata 21 (Zehnder 1995 p. 337) is to exaggerate the amount in
common between different genres in Bach.
Prelude
Like BWV 538.i, this looks like a new, mature working of a traditional
idiom: an opening solo, repeated chords, and old note-patterns promoted
into an organized ritornello form. Perhaps it came as the composer worked
in various Italian concerto forms and thus well before the unique, succinct
ritornello of another prelude in G major, Partita No. 5.i (1729/30). But even
if it reflected ‘an older Italian concerto-type’ such as Albinoni’s (Wolff 2000
p. 126), which is doubtful, this would not mean that it was as early as the C
major Toccata.
82 BWV 541
Just as in its keyboard figures the C major Prelude BWV 545 can be
compared to other pieces in C, so the scales, broken chords and homophony
of the G major are comparable to those of the Toccata in G BWV 916
(Example 39). Another important influence must be the harpsichord
Example 39
transcriptions of Italian concertos, such as the Vivaldi Concerto BWV 972 –

compare, for example, bb. 20–3 of the Praeludium with bb. 35–7 of
Vivaldi’s first movement. An interesting detail of the Toccata in Example
39 is that it too is cast in elementary ritornello form, a form yet more
contracted in the organ prelude:
1–29 passaggio on tonic triad; thematic quaver chords and

semiquavers
29–46 allusion to passaggio in dominant; same figures developed
46–59 further development of quaver chords
59–82 further derivations; 74 return to opening toccata; 79 return to
cadence of 44–6
It may be a mistake to see this as a planned ritornello, since the main theme
returns less obviously than in concertos, and none of it is drawn out. Rather,
the Prelude suggests a working out of conventional toccata elements – tonic,
dominant and final pedal-point (b. 63) – into a tightly organized movement
for whose cohesion themes are re-used in the course of the movement,
though in what order and manner can not be predicted. Unity is ensured
in the four sections (described as ‘strophe-like’ in Breig 1986b p. 36) by
such details as the opening and closing bars being heard elsewhere in the
movement, at b. 29 and b. 45 respectively. It is possible to see it as having
both three main sections and two.
The Praeambulum of the G major Partita (1730) and first movement of
Cantata 192 (1730?) offer good parallels to the tightened ritornello shape
of BWV 541.i. The three have similar material, with a similar pulse, con-
centrated and free of time-filling episodes. There are other associations too:
for example, the ambiguous ‘threes’ of bb. 10–11 of the organ Praeludium
recall certain phrases in the Minuet of the same Partita, also ‘Partita VI’
83 BWV 541
of the Chorale Variations BWV 770, and the Sonata for Solo Violin
BWV 1001 (Presto, beginning of second half). With the quaver chords and
running bass typical of new concertos (compare b. 16 with BWV 593.iii
from b. 70)∗ are blended elements of German organ toccatas: a pedal part
(compare b. 12 with Bruhns’s ‘Nun komm’ bb. 102–3) and broken-chord
semiquavers (compare b. 18 with BuxWV 140).
The use that the passage bb. 12–16 is put to in bb. 32–8 would not be
found in either Buxtehude or Vivaldi, and even the opening passaggio is
transformed by its drive and ambiguous rhythms. The ‘Vivace’ direction
probably belongs to the autograph revision, and quite why it was added is
unclear: did the composer by c. 1730 have a livelier idea of such figuration
than earlier, or was Friedemann unlikely to understand it correctly? Did
it have some connection to the recent G major Organ Sonata, apparently
written for Friedemann and opening ‘Vivace’?
The movement works very much in one-bar units, including the ‘mini-
cadenza’ of b. 24, whose diminished seventh form in b. 76 is an updated
version of the Neapolitan sixth in earlier works like the Passacaglia. The
result is a restless, hectic work, kept up on a high level until the final cadence,
majestic in its unbroken swing.
Fugue
The subject sounds like a theme awaiting words. Spitta heard a resemblance
to the opening chorus of Cantata 21 (1714) and its rhythms in the Prelude
(II p. 689), as did Emery 1966 and Keller 1948. But the possibility – faint
and ambiguous – that the subject began originally with four quavers on the
beat (KB p. 430) marks it off both from BWV 21 and the Prelude. Besides,
repeated quavers and little dactyls have a quite different effect in the 3/4 of
the Prelude from what they have in the 4/4 of the Fugue.
Similar but shorter themes by G. F. Kaufmann (‘Vom Himmel hoch’,
Harmonische Seelenlust, 1733–6) or F. A. Maichelbeck (‘Fuga Octavi Toni’,
i.e. G major, Augsburg 1738) need not reflect J. S. Bach’s influence, since
the subject follows a norm, with its repeated notes on 4–3 and 7–6 suspen-
sions. Similar examples in Handel, Lotti, Pergolesi and others confirm its
origin in Italian rather than North German counterpoint. Thus the theme
in Cantata 21 is not far from the fugue of Vivaldi’s Concerto BWV 596, while
the opening of Cantata 77 (1723) makes something similar from material
derived from a cantus firmus. See Example 40. BWV 21 and 596 are in the
minor and exploit stretto from the beginning, unlike BWV 541 which has
this shape:
∗ To judge by a version in KB p. 679, the chords from b. 21 were at first more simply repeated, with
less implied inner counterpoint.
84 BWV 541
Example 40
1–17 exposition; first answer tonal, second real (new countersubject)

17–26 episode: quavers from subject, semiquavers from
countersubjects
26–35 tonic entry, then derived episode in relative
35–52 relative entry, then free episode (solo-like) towards:
52–63 dominant entry, codetta-episode, supertonic entry
63–71 derived episode, minor (G minor entry = dominant to
C minor)
72–83 stretto at 9th, then 5th; final entry 79 on a new fifth voice
(C major); final tonic pedal-point in soprano, then doubled
The modest modification of the subject in b. 66 or 72 looks ahead to BWV

547. A flagging turn to the minor before final entries is there too in the
Prelude: compare Prelude bb. 76–7 with Fugue bb. 71–2, now with ninths
in the harmony. So melodious a subject leads to singable quasi-entries in
the soprano of bb. 20–5 or bb. 30–2, then to trio-like passages crowned
with top soprano entries. Note that the fifth voice of b. 79 not only brings a
subdominant finality but is complete to its last note (c ).
There is a tendency in the first and last thirty bars or so for semiquaver
figures to spin around themselves, producing new patterns up to the last
couple of bars, whether open and vigorous (bb. 61–2) or closed and obses-
sive (from b. 72). The masterly semiquaver figuration produces harmony
more complex and mature than with other repeated-note themes, such
as the E major Toccata BWV 566 at bb. 34–8. A real contrast is provided
by the middle episode, the only passage without pedals or clear reference
to the subject, but with shifting harmonies. These broken chords corre-
spond to the scale passages in other fugues, e.g. those before the final stretto
of the D minor Fugue BWV 538, though more charming and dance-like.
The whole passage bb. 38–52 resembles episodes in the first movement of
the Fifth Brandenburg Concerto, subtly emphasizing main beats to counter
the subject.
85 BWV 541–542
The inverted pedal point at the end is unusual, though already hinted
at in an earlier work, the A minor Fugue, BWV 543 b. 95. Here, the effect
is much bigger, like a choir singing, though even Cantata 77 ends with a
conventional bass pedal point producing a less dense effect than this, which
is one of Bach’s most gripping closes. If BWV 542 suggests how Bach père
ended a competition fugue in Hamburg, BWV 541 suggests how Bach fils
did in Dresden.
BWV 542 Fantasia and Fugue in G minor

No Autograph MS; Fantasia alone, in late eighteenth-century copies (P 288,
Am.B.531 via Kirnberger?); Fugue alone, in P 803 (J. T. Krebs c. 1714?),
P 1100 (J. C. Oley), P 598 (J. F. Agricola c. 1740), P 288 (J. P. Kellner),
also Am.B.531 and derivations; Fugue in F minor, perhaps via C. P. E. Bach
(P 287, LM 4838) and others via J. C. Kittel (P 320); paired only in two late
copies, perhaps unintentionally (P 288 second copy c. 1800, also P 595 –
derived from Am.B.531, where Fantasia and Fugue are separate), reversed
in P 1071 (c. 1800).
Two staves; heading ‘Fuga’ (Krebs), ‘pro Organo pleno cum Pedal obligato’
(Kellner); in Am.B.531, ‘Fantasia’; in P 288 second copy, ‘Fantasia e Fuga in
G m: Per l’Organo pieno, col Pedale Obligato’.
In its counterpoint, texture and figuration, the Fugue may be no later

than the Passacaglia BWV 582, though doubtless still played in Weimar
by students, including Krebs whose copy already suggests some revision
(KB p. 462). Since the F minor version may come down partly via C. P.
E. Bach – from a source agreeing with Krebs’s readings in G minor – the
transposition probably belongs to a relatively early point (KB p. 458) and
was made to avoid pedal d . This version is ‘less fluent and natural’ (Peters
II) and not known to be authorized.
The Fantasia must be later, even post-Weimar (Spitta I p. 635). Only too
high a regard for written compass, or uncertain harmonic criteria, could
lead one to think the Fantasia older than the Fugue (as Stauffer 1980 p. 110
suggests). Since the Fantasia is not known in an F minor version, there were
at least two traditions for playing the Fugue as a separate piece. But though
no authentic pairing of the movements is known, their different language
and date would not put it out of the question, in view of some unlikely
pairings in the WTC.
Assumptions that the movements constitute a pair led to the idea that
it ‘belongs without doubt to the Cöthen period’ (BG 15), composed for
86 BWV 542
the visit to Hamburg in 1720 after the composer applied for the position
at the Jakobikirche (Dok II p. 77). This was the occasion – if the Obituary
is referring to this particular visit – on which Bach played to the elderly
Reinken, last representative of a revered organ school (Dok III p. 84). Spitta’s
idea that in the Fantasia Bach ‘wishes to surpass the Hamburg organists
on their own ground’ is a guess (I p. 635). Better evidence is Mattheson’s
report that the competition for a new organist in Hamburg Cathedral on
24 October 1725 included an extemporized fugue on the subject quoted in
Example 41 (i), complete with a countersubject (ii), this too as in BWV 542.
Example 41
Mattheson may be implying that he had seen a copy of the piece:

ich wuste wol, wo dieses Thema zu Hause gehörte, und wer es vormahls
künstlich zu Papier gebracht hatte; (1731 pp. 34f.)
I knew well where this theme originated and who brought it artfully to
paper;
But a simpler version of the theme had been published (‘brought to paper’?)
in the songbook Oude en nieuwe Hollantse Boerenliedjes, Amsterdam, 1700
(Dok I p. 219). That Bach knew either form, ‘touching it up later’ to make
his subject, is not proved though often supposed; but the earlier the date
assigned to the Fugue, the more it matches others based on existing themes,
such as the Passacaglia’s.
How significant the compass is is also unclear: of the notes C, E, A
and d in the Fantasia or the E and A in the Fugue, none was available at
Reinken’s Katharinenkirche, and almost none on the Jakobikirche organ as
Schnitger had left it (Fock 1974 pp. 63–4). Solutions to these questions – the
Fantasia shows enharmonic possibilities whether for Hamburg or not, the
Fugue is transmitted with an ‘ideal’ compass that organists had to realize as
best they could from organ to organ – remain conjectural.
Fantasia
The shape, unusually clear and suitable for two manuals, has been seen
as rhetorical (Kloppers 1966 pp. 76–7), though by analogy rather than by
Bach’s conscious planning:
87 BWV 542
A I 1–9 Propositio: ‘free’ main theme; tonic; dominant; pedal

point
B II 9–14 Confutatio: opposing statement; imitative; moving bass;
strict four parts
A I 14–25 Confirmatio: partial return (roulades etc., multiple
suspensions); more chromatics; enharmonic modulation
B II 25–31 Confutatio: as before a fifth lower, upper parts exchanged,
longer by one bar
A I 31–49 Confirmatio: further development of chromatic idea
Peroratio: return, 40; chromatics resolved in pedal solo;
cadence
The final chord appears variously. Am.B.531 has a natural while P 288, of
c. 1800, is without, presumably a slip – or were minor finals preferred by
then?
This shape seems to ask for two manuals, as do the section-ends: what are
the rests for if not to change manual? What such analogies with rhetoric do
not say is whether they are more than the stuff of any coherent and effective
utterance. Thus by analogy the key and the seventh and ninth chords may re-
mind one of the opening of the St John Passion; and one can find other analo-
gies for the shattering first chord (emphasis), the crying out (exclamatio),
the repetition (anaphora), the falling/rising lines (anabasis/katabasis), the
contrapuntal discussion of motifs (b. 9, declamatio), even the rests in the
penultimate bar (aposiopesis). But figures of speech need not be explicitly in
the mind of a composer. Such music naturally implies gradatio (rising to-
wards climax) and congeries (accumulated part-writing), and a passage like
bb. 31–4 depends on purely musical devices – major–minor change, chro-
matics (different from bb. 22–3), contrary motion and a quasi-crescendo.
The Fantasia is a regularized version of an earlier form, a systematic alter-
nation of the recitativo and arioso of old multi-sectional praeludia pedaliter.
Two ways of looking at it are:
Dietrich 1931 point d’orgue – interlude – point d’orgue –
interlude – improvisation – interlude –
improvisation
Zacher 1993a pp. 20f. seven sections:
7 end of tonic pedal-point
14 end of ‘intermezzo’, with A major
21 the ‘astonishing 6/4 chord in E minor’
28 the ‘intermezzo’ revised
35 the Fantasia’s ‘generative chord’
(a diminished 7th) broken off
42 the broadest layout for the ‘generative chord’
49 final triad in seven parts
88 BWV 542
Apparently, there are other operative sevens in the movement, such as seven
falling fifths from D to D over the pedal of bb. 31–4; also, versions of
B A C H (e.g. tenor bb. 43–4); also, ‘a secret scale’ running through the piece
(e.g. ABCDEFG over bb. 14–25).
The opening pedal-point harmonies are much like those elsewhere but
less extempore in style (e.g. BWV 546). Rarely will such a pair of diminished
sevenths be found as in the second and third chords here: they ‘threaten’,
as the sevenths opening the A minor Praeludium BWV 543 do not. This
diminished seventh is an old chord, newly thought out and taking many
guises here, despite regular returns to dominant and tonic. The device of
chords punctuating roulades can be found – in more whimsical and refined
form, perhaps – in the Violin Solos (fair copy 1720): Example 42. These may
Example 42
derive from the roulades added by violinists to sonata movements, to judge

by one edition of Corelli’s Sonatas Op. 5 or by Vivaldi’s recitative in the
Concerto BWV 594.ii, though this need not mean that the Fantasia is a
‘secular’ piece (as Hammerschlag 1950 suggests). The opening tonic and
dominant pedal points have something of the conventional Orgelpunkt-
tokkata, with chromatic harmonies of a durezza kind, and even the startling
penultimate bar adapts an old idea: see the same moment in the E minor
Prelude BWV 533. The solo line over bb. 6–7 is coherent because the im-
plied harmony is logical, and only in the next bar does the Fantasia start to
develop beyond its toccata-like opening.
The harmonies on shorter pedal points elsewhere (bb. 13 etc.) are rel-
atively conventional; it is other harmonic effects that give the movement
its power. By b. 49 an impression of immense complexity has been gained,
89 BWV 542
Example 43
with harmonies (Example 43) that can be put in several categories: pedal-
points, multiple suspensions, diminished sevenths treated enharmonically
(as in recitative), chromatics moving to unexpected minor chords, consecu-
tive diminished sevenths, and interrupted cadences. As well as conventional
Neapolitan sixths there is the distinctive chord 9/7/6/4 in bb. 19–20, an-
ticipated by Kuhnau in a Biblical Sonata of 1700, when the smitten Goliath
falls. Since a similar chord appears in the early Prelude BWV 921 b. 5 and
Fantasia BWV 1121 b. 42, the e here b. 19 is probably not a mere scribal
error for e, as some have suspected.
Effect is increased by the dramatic rests or tmeses (in particular bb. 15,
20, 35, 44), by the huge variety in the texture, and not least by the ‘ordinary’
passages that set the rest in relief (e.g. bb. 39–41). These last are unusual and
therefore interesting. Exploring the six harmonic devices of Example 43 re-
places more conventional kinds of development, and as in some Ob chorales,
this intensification of harmony does not exclude some inter-quotation (e.g.
bb. 15–17 in bb. 44–6).
Despite the closely reasoned detail suggested by any such description,
it could still be that, as in Schubert, the most startling chords are those
produced not by chromatics or diminished 7ths but by changes of direc-
tion. Thus, while 7ths and chromatics are certainly involved in bb. 23–4,
the most startling event is the close not in E minor (the key of the previ-
ous bar) but in F minor, only to change direction towards the G of the
next bar. It is as if the dominant chord at the beginning of b. 20 had
merely been delayed by a few bars; but the effect is unique in music. Minor
triads can never have been used to such effect, being behind the sudden
twists from B minor to C minor in b. 15, the abrupt change to E minor in
b. 36, the diversion to C minor in b. 39, the surprising F minor of b. 45.
So too with the harmonies above the descending scale of bb. 31–4: it is
not the slow chromaticism that is startling but the relentless logic of a
simple sequence taking listeners they know not where, from D major to
– G major?
90 BWV 542
Fugue
This subject too is unique, whether quoting a Dutch song in deference
to Reinken, or alluding to a northern (F major BuxWV 145) or local (a
Capriccio of F. W. Zachow) subject-type. It contains two sequences, one a
half-bar, the other a whole bar long: unique, a reason why the subject is
so memorable. (Note that in simplifying the subject, the pedal of b. 78 is
closer to Mattheson’s version of it.) Its unmistakable jollity prompts earnest
countersubjects, though one episode (b. 43) matches the subject in this
respect.
Both the copyist of P 287 – and C. P. E. Bach too? (KB p. 469) – thought
it ‘the best of all the pedal works of J. S. Bach’, but it has its slacker moments
that remind one of Reinken (Hortus musicus: Example 44).
Example 44
tonic 1–21 exposition, two countersubjects, then episode from

subject
tonic 21–36 entries, parts exchanged; episode from same motif (32)
36–65 entry in relative; long episode (entries in D minor);
answer in relative dominant (54); tonic; then a long
anticipation of:
tonic 65–72 entry; episode from subject-motif
72–93 entries, subdominant and its relative (79), long
episodes, the last (86) towards remoter keys before:
tonic 93–115 entries, episode (94–103 = 44–53); three-part entry
(103); old episode (106–10 = 32–6); final entry, no
countersubjects
An impression is given of the tutti/solo sections of a concerto in which the

tonic acts as point of reference (cf. E minor Fugue WTC2) and a long subject
91 BWV 542
stands out, slipping in with ease, spinning off into bubbling lines. Quavers
develop their own episodes (bb. 57, 68, 82), and final entries of subject and
countersubjects appear in the course of the ‘bubbling lines’ (bb. 103ff.), with
the pedal’s serving as coda.
While Mattheson seems to have known the first countersubject of the
fugue, he makes no reference to the second (Example 45), which is similar
to the B material of the Fantasia (bb. 9–10) and to moments in other fugues
(cf. B minor WTC1 b. 17, and BWV 544). Two countersubjects can be found
occasionally elsewhere (e.g. Bruhns’s E minor Fugue and BuxWV 155 b. 63),
and here they produce moments much like a permutation fugue, as in
another early G minor Fugue, BWV 578. Sources suggest that the composer
‘improved’ it over time (KB p. 462), as in b. 56.
Example 45
Great ingenuity is exercised in developing the opening motif of the sub-

ject, on whose melodiousness the episodes rely for their quasi perpetuum
mobile and from which a very unusual homophony is produced in bb. 61–3.
From b. 83 the motif even rises instead of falls. The repetitive episodes and
reiterated perfect cadence produce a fugue somewhat different from what
the first thirty bars imply, and changes of manual are neither more difficult
nor more disruptive than usual. Nowhere in all this is the harmony obscure,
and if Mattheson was criticizing this Fugue bb. 40–1 when he went on to
write
lieber was bekanntes und fliessendes genommen . . . darauf kömt es an,
und es gefällt dem Zuhörer besser, als ein chromatisches Gezerre.
(1731 pp. 34f.)
rather, something familiar and fluent [should be] taken . . . that is what
matters and the listener will like it better than some chromatic affectation.
then he cannot have known what ‘chromatisches Gezerre’ there are in the
Fantasia. Or, he did, and was showing his preference for the fluent Fugue.
The pairing of Fantasia and Fugue forms a complement not out of place at the
time, just as the sections in many a French ouverture do; and presumably
pairings were much less fixed when a whole church service could come
between prelude and postlude.
92 BWV 543
BWV 543 Prelude and Fugue in A minor

No Autograph MS; extant copies probably either via C. P. E. Bach (P 290,
Am.B.60 a Berlin copyist, after 1754) or J. C. Kittel (e.g. Lpz MB III.8.14,
J. A. Dröbs).
Two staves; title in Am.B.60 ‘Preludio e Fuga per l’Organo pieno’ (Italian
terms common in the Berlin school), in Dröbs ‘ . . . für die volle Orgel’.
Good extant sources suggest BWV 543.i to be a revised version of an earlier

Prelude BWV 543a paired with the present Fugue, the revised originating
after Kellner had already made a copy of BWV 543a (Breig 1999 p. 660). But
it would not be impossible for Kellner’s to be the revised version, despite
assumptions made about Bach’s ‘modifications’ being always in the direction
of greater complexity (Rienäcker 1995). In any case, it is hard to imagine
the Fugue being a Leipzig work, as is sometimes conjectured (Humphreys
1989 p. 85), whenever Kellner’s copy was made (see below, p. 95).
The Fugue has often been likened to the keyboard fugue BWV 944 in ABB
and claimed as some kind of version of it, as if it was only in organ fugues
that Bach was to ‘seek and find adequate expression’ (Oppel 1906 pp. 74ff.).
But resemblances – contours of subject and countersubject, a perpetuum
mobile element, a rather free close – are too slight to imply a history of
either, shared or not. While the subjects circumscribe similar harmonies,
these arise from conventional formulae not unlike an Italian ritornello’s;
and while both contain playful figures in a harpsichord-like style (Hering
1974 p. 49), the genres are quite distinct. The composer’s associations with
A minor can produce shared details.
Other resemblances have been found: between the subject’s outline and
that of the A minor Fugue BWV 559, or between the pedal figures in both
Preludes’ closing stages (Beechey MT 1973 p. 832). The outline has also
been traced in the Prelude’s opening rh figure, in a Corrente in Vivaldi’s
Op. 2 No. 1, of 1709, and in a Fugue in E minor by Pachelbel (Keller 1948
p. 84). Of course, minor-key subjects that first trace the triad and then run
into a sequential tail of some length are bound to sound similar. Such a
perpetuum mobile-like subject, however, is unusual for an organ fugue of
J. S. Bach and, like that in BWV 564, it breaks up towards the close.
Prelude
It is true, as Spitta pointed out, that the so-called early version of the pre-
lude shows ‘certain characteristics reminiscent of the Buxtehude School’
(II p. 689), but his instances of Buxtehude-like figures from bb. 22 and 33
are also found in the ‘later version’. Other characteristics of northern
93 BWV 543
praeludia are: an opening rh running solo; its latent counterpoint in two or

three parts; a pedal version of it some time later (rather than the dominant
pedal point that might be expected); and the kinds of note-pattern in bb.
1, 23, 30, 33, 36 (with pedal quavers), and 50–3. For the copyists’ notation
of b. 33, see a comment on BWV 549a below. Two further ‘errors’ may have
been transmitted by the copyists: should the pedal point begin in the second
half of b. 9, and should bb. 19, 21 continue the crotchet lines of bb. 11, 13,
15 and 17?
Traditional are the ‘latent counterpoint’ of the opening (Mattheson gave
a somewhat similar example in 1739 pp. 354–4) and its chromatic descent
(in fact two chromatic fourths A–E, E–B), the tonic pedal point (from
b. 10) followed briefly by dominant and then another tonic, and the running
figures isolated above other pedal points (b. 33 etc). More characteristic of
J. S. Bach, perhaps, are the regularity of phrase in the opening rh solo,
dramatic use of the tonic pedal point in b. 10 (a rise in tension), careful
reduction of note-values (semiquavers, triplets, demisemiquavers) to which
the trilled chord in b. 23 is a climax, the texture of bb. 31–3, and the systematic
pairing of pedal points and manual patterns in the second half. The trilled
figure of b. 23 may be found in Buxtehude, but less obviously as a climax than
here.
As a logical answer, the pedal solo of b. 25 would best begin in the
minor (i.e. with g), a detail perhaps missed by the various copyists. Other
conventions are explored, such as the little broken-chord or brisé effect in
b. 29 (Example 46). The pleasing keyboard idiom over bb. 36–46 derives
from the opening bar, now in the major and disguising the commonplace
harmonies – harmonies that have been improvised by countless organists,
on any registration from a single Open Diapason to plein jeu, depending
on local tradition. The final bars have something of a bariolage as found in
the (contemporary?) Passacaglia, and a pedal motif used very differently in
‘Alle Menschen müssen sterben’ in the Ob.
Example 46
The piece may reflect the composer’s interest in integrating different

prelude traditions. At the return to the tonic halfway through (b. 31), a
free-roving tenor melody in the lh keeps up the motion, in this Prelude
94 BWV 543
the only such figure but one of a type familiar elsewhere in early Bach.
(See Fugues BWV 535 at bb. 52ff. and BWV 578 at b. 51; also the Praeludium
BWV 566 at b. 85 and the chorale ‘Wie schön’ BWV 739 at b. 69.) ‘Northern’
are such details as the little obsessive g (bb. 10–14) and c (16–20), in effect
chromatic acciaccaturas colouring the buildup of a tonic pedal point in
preparation for a dominant answer.
Fugue
The subject’s head motif and lengthy sequential tail, which paraphrases the
A minor sequence at the beginning of the Vivaldi concerto BWV 593, are
broken chords suitable for pedals. They easily confuse the ear about the beat.
The codetta (bb. 11–14) already reduces tension, and the episodes (bb. 56,
66 etc.) rarely rise above a certain level of melodiousness against which the
subject is conspicuous. The shape is:
1–30 exposition, in regular four parts, of a subject 4 12 bars long;

consistent countersubject; two codettas
31–50 episode extending the exposition; pedal sequence; new
material; tonic entry (after hemiola cadence), head motif
in stretto
51–61 further hemiola cadence; entry in dominant (subject head
hidden); episode on a further circle-of-fifths sequence to:
61–95 relative major entry en taille; episode; answer; episode
(all episodes based on circle of fifths)
95–135 stretto entries 95/96, dominant 113/115; final 131; all
followed by derived episodes and short pedal points
(the last a trill?)
135–51 pedal point then solo; quasi-cadenza manual figures
The piece is a good instance of the growing interest in long-phrased fugues,

tight in neither counterpoint nor form. Entries appear as if delayed after
drawn-out episodes, effective and unusual, each time heightening the sense
of singable melody.
Like the Prelude’s opening solo, the Fugue’s final manual solo is not
free but regular, running straight into a cadence of great finality. It thus
resembles the C major Toccata’s Fugue, though the cadence itself and the
previous pedal solo remind one more of the Toccata in F. And for the pedal
of b. 145, see the first recitative of Cantata 161 (1715/16). Older features
include a profusion of circle-of-fifths sequences, rising but mostly falling,
as in the subject itself. Another ‘early’ sign is the array of Neapolitan sixths
(bb. 85, 111, 134), which like the brisé figures vaguely recall the Prelude
(Neapolitan 6th at b. 43). Such harmonic turns as the diminished sevenths
95 BWV 543–543a
over bb. 146–50 were highly inventive at the period and, like the dissonant
acciaccatura chord of Example 47, counteract the predictable sequences.
(The second part of Example 47, from the Concerto in D minor, shows a
simpler acciaccatura.) As in the C major Fugue BWV 564, the simple figures
sometimes turn into brief moments of complexity for the player (bb. 26–7
etc.).
Example 47
The Fugue is irrepressibly fluent: a singable, sequential subject whose

lively figures produce only two different harmonies per bar, hence the sig-
nificance of hemiolas early on and the cadenzas near the end. The metre
itself adds triplets and sextolets to the Prelude’s repertory of note-values.
(It could be anachronism to suppose that the final demisemiquaver sextolets
represent a ‘written-out rallentando’, to be played half as fast as written, as
suggested by Emery in MT 1967 pp. 32–4: succinct closes are in style with
these earlier Bach fugues.) Most semiquaver groups can be traced to the
way countersubjects spin off a tuneful subject, right to the end (bb. 132–4),
and the Fugue is free of mere scales until the last episode. If the ‘motoric’
subjects of Reinken, Buttstedt, Heidorn and others inspired this Fugue, its
sequences from bb. 28 or 132 were highly original at the time, almost as if
this were an essay in the art of writing them.
BWV 543a Prelude and Fugue in A minor

No Autograph MS; copies in P 803 (unknown copyist, perhaps contempo-
rary), P 288 (perhaps J. P. Kellner c. 1726/7?) and LM 4839g (via Kittel?).
Two staves; title in P 803, ‘Praeludium con Fuga’.
That the Kellner copy might have been collated with a MS of BWV 543 is
further support for the two versions being separate and distinct, though
with the same fugue – for which P 288 and P 803 probably drew on an
autograph (KB pp. 479, 590).
Differences between the two preludes in NBA IV/5 and IV/6 are as
follows:
96 BWV 543a–544
BWV 543 BWV 543a

1–9 1–6 different broken chords, the chromatic descent in 543a
more contracted; in 543, lh version inverts the rh figure
10–21 7–12 identical, but 543a appears to be notated twice too fast
22–5 13–16 identical, but 543a distributes the runs between the
hands
26–8 17–18 pedal of 543a again a shorter form of broken-chord
figure
29–53 19–43 almost identical
The demisemiquavers (b. 7) have led to an idea that the composer was
thinking ‘in the later version . . . on a larger scale’, preserving ‘a calmer
mood’, while the earlier invited the player to ‘feel free to improvise and
elaborate the score’ (Beechey MT 1973 p. 832). The hand-distribution in
bb. 13–14 of P 803 is either conjectural or implies a phrasing; bb. 23–5
(and bb. 33–5 of the other version) have no markings, nor does the solo at
the end of the Fugue. The longer of the two so-called cadenzas in the Fifth
Brandenburg Concerto also moves above a point d’orgue from semiquavers
into notes twice and then three times as fast, doing so systematically and
unambiguously.
The crucial differences between the versions – bb. 1–9 and 26–8 (first
and fourth sections above) – are generally taken to mean that BWV 543a is
the ‘earlier version’, but in fact the opening figure as it appears in BWV 543
is more conventional in its harmony, i.e. a series of prepared and resolved
7ths. Nevertheless, the extended and more developed triplets that follow in
b. 4 of BWV 543 do look like the result of revision, as does the alteration of
the opening figure when it passes to the left hand. The logical harmonies of
the second half of the prelude seem to have required no further revision.
BWV 544 Prelude and Fugue in B minor

Autograph MS (fair copy in private possession, c. 1727/31); copies (from
this?) via J. P. Kellner (P 891) or probably C. P. E. Bach (Am.B.60 Berlin
copyist after 1754, P 290, Am.B.54, P 276) or J. C. Kittel (Lpz MB III.8.21,
J. A. Dröbs).
Two staves; title in Autograph MS ‘Praeludium pro Organo cum pedale

obligato’.
Whether the autograph MS was based on a copy made in Weimar (Emery

1966) or one made early in Leipzig (KB p. 484) cannot be shown, although
97 BWV 544
the later copies probably derive from it. The work’s idiom has much in
common with B minor music in the St Matthew Passion and Cantata 198,
Funeral Ode for the Electress of Saxony, which was performed in the uni-
versity church in 1727 – with an organ prelude and postlude? The elegiac B
minor of the cantata’s opening chorus matches BWV 544 closely, and they
could well be contemporary. Perhaps the mature praeludia in B minor, C
minor and E minor were all associated with the university church’s organ.
The Prelude is an original contribution to new organ styles of the day,
aria-like and quite unlike the other mature preludes, with bold effects
achieved through appoggiatura harmonies, and matching Mattheson’s de-
scription of the Affekt of B minor as ‘unlustig und melancholisch’ (‘listless
and melancholy’: 1713 pp. 250–1). Spitta felt in it a ‘deeply elegiac note not
heard so intensively anywhere else in Bach’s organ works’ (II pp. 689–90).
More objectively, however, like the D minor Toccata BWV 538 it is confi-
dently new both in keyboard idiom and in its rounded form. By nature such
form is likely to express the Golden Section: see below.
Prelude
The concerto or ritornello shape can be outlined as:
A 1–17 two-part imitation; tonic then dominant pedal point

B 17–23 fugal exposition of a new theme
A 23–43 scale idea from A picked up, linked to a return in
dominant (27–33 = 1–7); sequences; second pedal point
(40–2 = 14–16)
B 43–9 fugal exposition (43–8 = 17–22)
A 50–73 thematic buildup: 50–6 motifs from A (50–2 = 11–13),
relative; 56–60 new theme (appoggiaturas) plus earlier
scales; 61, A (63–4 = 6–7); 65 beginning as 54; 69
sequence from theme of 56; scales
B 73–8 imitative exposition of B rectus and inversus
(77–8 = 49–50)
A 78–85 figures from A (79–80 = 38–9; 82–5 = 40–3 = 14–17;
81 new)
However, the limbs of the movement are not so distinct as they are in BWV
542, 546, 548 and 552, to whose general form-types it belongs, although they
are certainly clearer than those in BWV 538. A can be seen as returning not
at b. 23 but at b. 27 (lh second note), in which case there is no clear return
from the Positiv manual to the Hauptwerk; any return to the Hauptwerk in
b. 50 is also somewhat abrupt. But to conclude from this that manuals are
not to be changed is no more justified than it is elsewhere.
98 BWV 544
A 16-bar framework around three sections of 13 bars can be dis-

cerned: bb. 1–17, 17–43, 43–56, 56–69, 69–85 (Schmidt 1986). And there
are other symmetries: there are two halves (1–43, 43–85, tonic–dominant,
dominant–tonic); and there is a Golden Section both between ritornello
and first episode (16 : 26 bars, close to 3 : 5) and between the material over-
all (the episodes’ 32 bars to the ritornelli’s 52 = 8 : 13). Also, almost the
whole, and certainly the A section, can be seen as a succession of three- or
four-bar statements returning piecemeal:
for 1–4 see 27–30

4–7 30–3, 60–4
11–14 50–3
14–17 40–3, 81–5
37–40 78–81
53–6 64–9
Or one can see five entries of A (bb. 1, 27, 50, 61, 78 – Zahn 1985), of which
the fourth is less clear. In general, the ritornelli are stable, the episodes less so,
varying from being ‘non-thematic’ to having a new theme (b. 56). Perhaps
the episodes already contrast enough with the denser ritornelli for manual-
changing to be quite unnecessary, but to change requires only a tactful break
even over section bb. 50–73. The autograph notation is evidence neither for
nor against changing (KB pp. 38–9), for although the first note b. 17 was
re-written in the fair copy to make a continuous beaming, it could be the
nature of such a fair copy to rule out performing hints: to remain a ‘reference
document’ to be further copied as and when.
The elusive style of the B minor Prelude depends especially on appog-
giatura harmony, suspensions and accented passing-notes, so that virtually
every main beat of the whole opening ritornello has one or other of these
discordant effects. The opening bars, though based on the unoriginal idea
of invertible counterpoint imitated at the octave/unison (cf. the Two-part
Invention in E major BWV 777), explore an unusual tessitura characteristic
of Buxtehude openings, now with appoggiaturas. Once the dotted, swinging
rhythm begins, the louré effect combines with a plaintive melos to produce
a very distinctive movement. (It is this rhythm, presumably, that leads some
players to hear something French about it. See Krummacher 1985 p. 133.)
As with other mature preludes, there is a marked contrast between the two
main themes, i.e. the loure and the demisemiquavers, and the end-result is
unusual.
While in theory the pedal-point harmonies of bb. 14–15 reflect old toc-
catas (Orgelpunkttokkaten), in practice the five parts create a rich, lush
harmonic spectrum. The dotted rhythms are anything but siciliano-like;
99 BWV 544
however springingly played, what they supply is heaviness. Similarly, al-

though the lines in bb. 23ff. or 49–50 might look much like moments in,
say, the Corrente of the E minor Partita for Harpsichord (1725?), there is
nothing corrente-like in the Prelude’s tempo, texture or harmonic rhythm.
In comparison, the episodes are mostly without dotted rhythms and appog-
giatura harmonies, and have shorter phrases. Thus ritornelli and episodes
are reciprocal, and the new themes at b. 56 and b. 69 are a compromise, with
appoggiaturas from one and phraseology from the other.
From the scale of b. 8, all the scales of the movement could be claimed to
grow, both ascending and descending, but the point in the bar at which they
begin or at which they curl back on themselves varies. Some are like those
of the E minor Prelude BWV 548, whose opening bars are most curiously
hinted at (if seldom noticed) in b. 61. The ‘sighing thirds’ from b. 56 are
conspicuous for the listener, like BWV 537’s, reminiscent of woodwind
lines in a cantata movement. The cadence in b. 55 could have come from a
chamber sonata for flute or violin, however, or even from the Loure of the
G major French Suite. The fugue theme B produces a pretty sequence in
bb. 46–7 but is also put upside down in a didactic, some might think dry,
manner at bb. 74, 76, as if to match the C major Fugue’s equally gratuitous
inversion (BWV 547).
The extra bar slipped in between b. 80 and b. 82 is masterly, extending
the chromatic harmonies and forming the harmonic climax of the move-
ment. While section bb. 71–8 has to be there to satisfy the requirements
of superimposed form, the five-part harmonies of b. 81, particularly the D
major chord, are inspired.
Fugue
The fluent, restrained Fugue contrasts powerfully with the Prelude. Its lines,
moving largely by step throughout, are less like the driving subjects of earlier
organ-fugues than the Corellian bass-line from the last prelude of WTC1.
Its form as a tripartite fugue (i.e. with episode in the middle) is close to the
G major’s, BWV 541:
1–11 pedal is third, not last, to enter (cf. BWV 541); countersubject
12–17 episode from countersubject; tonic, subdominant entries;
short episode from subject
18–23 relative, answered in its dominant; short episode from subject
etc
24–37 entries, dominant twice (28 new countersubject), tonic,
subdominant, short episodes (32ff. from subject and second
countersubject)
37–49 episode from second countersubject (29); modulatory entries,
supertonic, dominant; episode from subject
100 BWV 544
49–59 quasi-entry in relative; episode from subject

59–67 tonic entry, two countersubjects (top one new), answered;
episode from subject and countersubjects
68–78 modulatory entry, supertonic, with new (third)
countersubject; episode to subdominant entry; further
episode (cf. b. 32)
79–84 chain of modulatory entries, pedal; use of earlier
countersubjects
85–8 final entry with two countersubjects
Three more or less equally large sections may be discerned: bb. 1–28, 28–59
(no pedal), 59–88 (combination of themes). The ‘walking quavers’ of the
subject are like those in the countersubject of the chorale BWV 698, just
as the new countersubject at b. 59 resembles a pedal-motif in the chorale
BWV 627, at v. 3 (‘ Christ ist erstanden’). It is possible to hear at b. 76 of the
Fugue a reference to the Prelude (bb. 14f., 81f.), but the similarity is slight,
and the passages’ functions differ, being more climactic in the Prelude.
A good deal of art has gone into this Fugue, its fine series of coun-
tersubjects and lines worked from a very few patterns. It is the patterns in
particular that produce the striking smoothness. As in BWV 543, the subject
has been glimpsed in the Prelude (penultimate bar, according to Stauffer
1980 pp. 130, 134), but perhaps only because the subject’s ambitus accords
with phrases in the Prelude, being founded on similar note-patterns. The
four-quaver groups in the subject are closer to such lines as the countersub-
ject to ‘Jesus Christus, unser Heiland’ BWV 689 (from b. 3), groups working
naturally well in diminution and producing the fugue’s persistent semiqua-
ver lines. Example 48 illustrates the kind of motivic derivation typical of a
fugue: compare that in the E Prelude, bb. 147–8.
Example 48
The countersubjects are carefully dissimilar: as first heard in b. 3, b. 28

and b. 59 they counter the theme by producing first angular lines, then
non-stop semiquaver scales (and broken chords, b. 29), then up-beat motifs.
Similarly, the final episode at bb. 73–7 concentrates on broken figures before
the final entries. Thus the Fugue is an extremely ingenious working of a basic
101 BWV 544–545
note-pattern, and one wonders why there is no full stretto, either of the
quaver subject or of its diminution. At times, stretto is approached, and
the entry of b. 24 even seems to be delayed for one. Of chief interest for
Bach was the unassuming but singable subject, with no attempt to use
countersubjects for some extravagant edifice. Constant re-harmonization
of the subject leads to happy results (e.g. the sixths of bb. 71–2), while less
colourful are the combinations (e.g. first and third countersubjects in b. 63)
and invertible sequences (bb. 32–4 or 44–7).
It is easy for a performer to miss the special flavour of this Fugue even at
its intense moment around b. 50, since its counterpoint is much like that in
‘quiet’ episodes found elsewhere, e.g. B minor Fugue WTC1. In the second
half it makes great play with the various motifs, so that (e.g.) b. 65 or b. 86
is a mass of allusions, some in diminution and producing textures difficult
to play. If changing manuals is an option, the return to the main manual
after the pedal-less middle section could be managed in more than one way,
leaving the density of the last dozen bars and its taut chain of bass entries
uninterrupted.
BWV 545 Prelude and Fugue in C major

In two movements: ‘Clauss MS’ (autograph fair copy?) now lost; other copies
known to Kittel circle (P 658, LM 4839c, Lpz MB 111.8.21 J. A. Dröbs) or
via C. P. E. Bach and Kirnberger (P 290, Prelude BWV 545a; also Am.B.60)
and later.
In three movements: ‘Moscheles MS’ once thought to be autograph but

copied c. 1729 by J. C. Vogler (Schulze 1984 p. 67); also J. G. Walther
(LM 4718, from Vogler’s?) and J. P. Kellner (P 286 after 1727? Stinson 1989
p. 24).
Two staves; title in Clauss MS ‘Praeludium pro Organo cum Pedale obligato’,
in Moscheles MS ‘Praeludium in Organo pleno, pedaliter’ (the composer’s
title?); in LM 4718, ‘Preludio con Fuga e Trio’ (NB order!), trio headed
‘Largo’.
Perhaps the several versions and forms of this work were less exceptional
amongst the major preludes and fugues than now appears, and others too
circulated like this:
a shorter Prelude with Fugue (BWV 545a)

longer versions of the Prelude and Fugue, including the ‘later’ BWV 545
(two-movement version)
102 BWV 545
the same with a trio, BWV 529.ii (forty bars placed before the Fugue by Vogler,
the rest after; entirely after, by Walther; between, by Kellner)
a version of the ‘early’ Prelude, in B major, made to avoid pedal d (?); plus a
version of a movement from a Gamba Sonata (BWV 1029.iii) before the Fugue;
plus two short interludes. See BWV 545b.
The main copyists probably worked from an autograph (KB IV/7 p. 86) or
autographs, and used an early version of BWV 529.ii (Emery 1957 p. 104).
To include one or other trio movement was surely because the prelude is so
brief – the reason too for its longer version, as the composer came to favour
such closed forms.
One possible order of composition for the work is as follows (see Emery
1959 and KB p. 299): (a) BWV 545a, before Weimar? (b) BWV 545b, at
Weimar? (Prelude with three extra ‘coda’ bars referring back to the opening);
(c) BWV 545 three-movement version (Prelude with a ‘coda’ also used as
preface; plus BWV 529.ii in an early version); (d) BWV 545 two-movement
version (slight variants throughout), and eventually a new fair copy with
revisions. Note that in its phrygian close, the trio of (c) suits both the C major
Fugue and the Sonata’s finale. Further doubts remain: was the Prelude of
BWV 545a in fact shortened (by whom?) from one or other longer version,
and can the idiom be as early as Weimar?
Prelude
The movement is organized as a pedal-point prelude:
Tonic durezze + broken chords above pedal point

1–3
4–11
main motif (Example 49), then interlude based
on it
Dominant 12–22 motif above pedal point; interlude (16–19 = 7–9)
22–6 motif above pedal points, dominant, tonic;
cadence
Tonic 28–31 similar to 1–3
This miniature da capo is unique, while still leaving clear the old
tonic–dominant–tonic pedal points. Dominated by a single motif, the pre-
lude is more like early WTC preludes than the organ works. The idea of a
framework is borne out by the number of parts: five or more at the begin-
ning and close, four at bb. 7 and 23, three at bb. 13 and 20, and four at b. 16
(the centre), thus a symmetry of 5–4–3–4–3–4–5. Also, bb. 1–3 and 28–31
(the additions) are both more sustained than the rest and have the key-
board’s top and bottom notes, the first bar alone covering C–c . Starting
at the top may be unique, although the alternate-foot pedalling and the
durezza element are traditional.
103 BWV 545
The Prelude’s original (?) opening in b. 4 represents a standard C major

prelude: Example 49. In both cases the prime motif is extracted and worked
into a contrapuntal texture, in the course of which it often changes shape
without losing identity. The first Prelude of WTC2 also exists in several
versions and, like the ‘organ version’, soon brings in a B over the opening
pedal point and an A diminished seventh at the end; but it develops its motif
more than does BWV 545. Together, they are subtly different examples of
idiomatic writing for the two different instruments: BWV 545.i has a much
more open texture, uses motifs more simply, and produces fine pedal lines.
Example 49
The splendidly expansive manual writing of both movements represents

a ‘standard C major sound’ (compare Fischer’s Praeludium 5 in Blumen-
strauss), and results in some similarities between them – e.g. the pedal in
the Prelude, b. 1 and the Fugue, b. 38. Much of the Prelude is based on
one-bar phrases, with at least two longer phrases (bb. 14–16, 24–6), and
one wonders why a bar like 21 was not treated in sequence. When the first
syncopated, suspended pedal phrase appears (b. 7), the motif in the right
hand goes off via an f beyond any usual ‘standard C major sound’.
The Fantasia in the AMBB, BWV 573, gives a third version of this prelude-
type, now in five parts, but in its sequences, bass-line and melos much
like BWV 545.i. Note that neither is fixed – one has variants, the other is
incomplete.
Fugue
The shape may be outlined:
1–19 pedal is third voice to enter; no constant countersubject

19–51 dominant and tonic (41) entries, episodes partly from
subject; countersubject, b. 45
52–72 entries, relative and its dominant, with episodes; 72,
suddenly to:
73–99 entries in dominant, tonic, subdominant and supertonic
100–11 final entries (106 above pedal point); cadence 108
(see bb. 81, 18)
104 BWV 545
For ideas similar to the tenor’s running quaver line at the end, see BWV 538
and 540. The possibility that ‘originally the piece ended shortly after b. 79’
(Breig 1993 p. 53), with the final tonic entry beginning in that bar, cannot
be ruled out. But no source suggests that the fugue was even more succinct
than now, and the quick succession of keys in the second half is typical –
surely not earlier than c. 1715, and probably later.
The Ob’s composer knew that the subject’s first notes can take many
forms: see Example 50. There was a tradition for ‘stepping’ themes of this
kind, to judge by a family likeness between it, the fugue-subject in the Prelude
BWV 546, the first subject of WTC1 and e.g. ‘Blessed be God’ in Handel’s
Cannons anthem HWV 256a (c. 1717). A result is that despite its jolly
broken chords and idiomatic sequences created on all possible occasions
(bb. 19, 31, 49, 65, 77, 96), the Fugue is calculating in its constant returns
to the tetrachord of Example 50. The tenor of b. 94 is surely an allusion.
Comparable points could be made about the B minor Fugue’s subject of
conjunct quavers.
Example 50
However similar in theory the subjects of BWV 544 and 545 are – narrow
compass, a scale-like line – the C major’s entries tend to slip in as if part of
the background (see bb. 28, 35, 52, 79, 84), which is not so in the B minor.
Similarly, in the C major, more entries go on into an extended discussion
of what the other voices were concerned with. A further distinction is that
while BWV 544 has three returning countersubjects, BWV 545 has at most
only one, although many of the lines accompanying the subject could have
become regular countersubjects (alto b. 73, bass b. 79, soprano in bb. 62
and 100). Nevertheless, even if the countersubject of b. 5 reappears only
once in the whole fugue (b. 45), its features – contrary motion, suspensions,
syncopations – colour the counterpoint throughout.
As often with the mature Bach, it is difficult to say whether the harmony
produces good contrapuntal lines or the counterpoint produces good har-
mony, e.g. the augmented chord in the relative-minor entry of b. 53. The
quaver patterns work ceaselessly to create the counterpoint, resulting in a
family resemblance between the last paragraph of this fugue and that of
the D minor, BWV 538. In the very block harmonies at the end, each voice
sings.
105 BWV 545a–545b
BWV 545a Prelude and Fugue in C major

No Autograph MS; copies second half of eighteenth century, perhaps via
C. P. E. Bach (P 290) or W. F. Bach (? Lpz Poel 12, Forkel’s thematic index
of 1802).
Two staves; title in P 290 ‘Praeludium Pedaliter’.
The chief differences between BWV 545a and 545 (NBA) are as follows:
BWV 545.i BWV 545a.i

1–3 —
4 1 (two further manual parts in 545)
5–26 2–23
27 24 (different in detail; dominant pedal point
in 545a)
— 25
28–31 —
The Fugue is different in minor details, e.g. no semiquavers in bb. 96–8.

Whether the Prelude BWV 545a is an abridgement is still uncertain.
NBA’s conjecture is that it is an early version, pre-Weimar (KB pp. 299,
568), but this hangs partly on assuming that Walther’s copy of BWV 545
is earlier than it is now dated (c. 1729). In comparison with the ‘later ver-
sions’, the Prelude of BWV 545a closes somewhat abruptly, thus suggesting
either that the composer came to feel the need for a coda restoring the
tonic–dominant–tonic shape of the whole, or that originally there had been
one but the composer or a copyist shortened it to avoid pedal d . There
may be other reasons why BWV 545a is shorter – the sources were poor, the
revision was not completed, etc. – but as it stands, BWV 545a opens more
like Book 2 of the WTC than does 545.
BWV 545b Prelude, Trio and Fugue in B major

Only source, LBL RCM MS 814 (copied by B. Cooke Jun. 1761–72 and B.
Cooke Sen. 1734–93).
Three staves; ‘Prelud[i]um pro: Organo Pedaliter’, ‘Adagio’, ‘Trio a 2 Clav:

e Pedal’, ‘tutti’, ‘Fuga pro Organo. Pedaliter’; at end, ‘By the late Mr. John
Robinson’.
Robinson was Cooke’s predecessor at Westminster Abbey, and it is possible

that with ‘by’ he was signifying not the supposed composer but the arranger
106 BWV 545b
(transcribed by), or the owner and/or copyist of the source (by courtesy of )
from which RCM 814 was made, or the route of its transmission (by the
agency of ). How it came to London is puzzling: through Handel, J. C. Smith
Sen. (†1763, his copyist), J. C. Pepusch (†1752), C. F. Abel (gamba-player,
visiting Leipzig in 1743, perhaps owning a copy of BWV 1029) or James
Hutton? The last visited Bach in 1749, brought back some music he called
autograph (see KB V/2 pp. 105–6): probably in fact an incomplete copy of the
Goldberg Variations printed in Hawkins’s A General History, London 1775.
Neither Robinson nor Cooke had more than a rudimentary pedalboard at
the Abbey (see Knight 2000), though a growing interest in such things could
be the raison d’être for making a copy whose date (at the latest, c. 1772), key,
shape and place of origin give a unique picture of the circulation of Bach
works.
The chief differences between BWV 545b and BWV 545 are as follows:
key, with the many octave displacements this entails

five movements, Praeludium, Adagio, Trio, Tutti, Fugue
prelude: BWV 545.i BWV 545b.i
1–3 —
4 1, with two further manual parts in 545
5–27 2–24
— 25–8, coda referring to opening bars
28–31 —
Whether the differences, including minor details, were there in the copy’s
source, or even all originated at the same time, cannot be known. The Trio
is a version of the movement now found as finale to the Sonata for Viola da
Gamba BWV 1029; both come from an earlier, unknown version. Perhaps
Abel had some hand in transmitting gamba pieces. (See also BWV 1029.iii
and 1027 below.) It is a curious coincidence that trios associated with BWV
541 and 545b are both fast movements and not, as might be expected, slow.
Someone, at some stage, seems to have thought of them almost as scherzos
in the later sense.
The Adagio and Tutti are connecting interludes added at some stage,
probably not for RCM 814 itself. Though brief, they evince a knowledge
of style (Adagio built on dotted figure, Tutti on a recitative line) and for
that reason alone are conceivably the work of J. T. Krebs, written already in
Weimar (KB p. 302).
Because its bass-line contains a few ‘improvements’ to BWV 545a ‘not
likely to have been made by anyone else’, one might agree that ‘the trans-
posed text [can be] best ascribed to Bach’ (Emery 1959 p. v), though not
necessarily the transposition to B major. Although chronology based on
107 BWV 545b–546
compass – e.g. a Mühlhausen work could be written for an organ with pedal
d and so need transposing later – is always speculative, the B Prelude’s coda
does look authentic (KB p. 300). Perhaps Cooke’s source was a German MS,
whose headings for the first, third and fifth movements are Bach’s own.
Furthermore, the clever and effective close to the Prelude, referring both
to the theme and to the concluding harmonies of the version BWV 545, is
typical of the composer of BWV 547 and 769.
BWV 546 Prelude and Fugue in C minor

No Autograph MS; copies by J. P. Kellner (P 286, from autograph? after
1730?), and perhaps via C. P. E. Bach (P 290, P 276, Am.B.60 J. P. Kirnberger)
or J. C. Kittel (e.g. P 320); fugue only, with Fantasia BWV 562, in P 1104
(J. C. Oley?).
Two staves; title for whole work in P 1104 ‘Praeludium Pro Organo cum
Pedal: Obligato’ (heading for first movement ‘Fantasia pro organo cum
pedali obligato’).
Two problems are: do the movements belong together? and are they con-
temporary? The discrepancy commonly felt between them has led to the
idea that the Fugue was written earlier, perhaps with the Fantasia BWV
562.i as prelude (Griepenkerl, Peters II 1844); this is attested by Oley’s
copy, which could well be based on a lost autograph (KB pp. 323–4). The
present Prelude, being in concerto form, was ‘completed in Leipzig’ (Spitta II
pp. 687–8) and added to an earlier Fugue much as the Toccata in F was, these
two fugues having ‘originated at the same time’ (I p. 581) – which, however,
could mean they were both Leipzig works. Less conjectural is that in the
complete copies of BWV 546, the Fugue shows signs of revision, as if made
when the Prelude was composed and the two coupled.
But it is not certain that the Fantasia BWV 562.i is earlier than the
Prelude BWV 546.i, and any ‘discrepancy’ between them might be no more
than the marked difference between complementary movements. After all,
at some point the composer doubtless did couple the massive ritornello
Prelude BWV 546.i with its present, much less dense Fugue. Similar points
may be made about BWV 537, and while BWV 546 may be less well matched
than the E minor BWV 548, as complementary prelude–fugue pairs they
are not dissimilar. The ending of the Fugue is similar enough to the ending
of the Prelude – richly scored, climactic, an important flat supertonic – to
suggest that the composer consciously paired them, whether before, during
or after the composition of the Fugue.
108 BWV 546
Prelude
The ritornello shape is of special interest, since A returns only in fragments
before the final reprise, as with the ‘sporadic recapitulation’ in the first move-
ment of the Vivaldi Concerto BWV 593. The idea is not so different from the
organization of certain organ-chorales and cantata first movements, where
chorale lines act as episodes of a sort.
A A1 1–5 homophonic dialogue between hands; tonic pedal

A2 5–13 distinct quaver motifs; then dominant pedal
A3 13–25 pedal motif from A1, pedal point, Neapolitan 6th,
triplet figure, perfect cadence
B 25–49 irregular exposition of new, derived figures; episode
A A1 49–53 dominant
B 53–70 more regular exposition (answer, codetta); episode
A A2 70–81 pedal point of 10, now (75) providing triplet motif
B 82–5 short statement
A A3 85–97 as 13–25 in subdominant, upper voices exchanged
B 97–120 two entries (97, 117) plus episode on motifs from
first codetta (31)
A 120–44 A1, A2, A3 as before; tierce de picardie
According to Meyer 1979, section A2 is bb. 70–8 and B bb. 78–85, but this
does not affect the symmetrical bar-numbers: 24, 24, 48 ( = bb. 49–97), 24,
24. Such symmetry is close to that in the harpsichord Fantasia BWV 904
(Stinson 1989 pp. 107f.).
The prelude is another example of rhetorical form (Kloppers 1966
pp. 74–5), clearer in its ABA shape than either BWV 538.i or 542.i:
A Propositio: main theme; contains essential features

(dialogue-chords, triplets, pedal points, scale-like bass); from
minims to semiquavers
B Confutatio and Confirmatio: spinning-out of triplet figures for the
‘high points’; A material restated in three extracts
A Peroratio: conclusion or exit
However, the definition of peroratio as counterpart to the exordium or intro-

duction does not quite fit the idea of da capo in music as usually understood.
Also, Kloppers understands the third B section to begin in b. 78, while Keller
(1948 p. 15) regards the whole passage bb. 70–96 as one section on the main
manual, which agrees better with the 24-bar plan of the movement. Either
way, manual-changing over the middle section is too awkward if the player
feels obliged to preserve continuity as written.
109 BWV 546
The opening dialogue chords recall the coda of the D minor Fugue
BWV 538 and, significantly, the close of the C major Concerto for Two
Harpsichords BWV 1061. That they do not necessarily require the massive
pleno customary today is clear from the lightly scored opening of Cantata
47 (1726). The startling contrast between the two themes (A in b. 1, B in
b. 25) has an opposite effect to the preludes of English Suites BWV 807
and 809, where the opening material is contrapuntal, the episode mate-
rial more homophonic. The opening tonic pedal point soon answered by
dominant is reminiscent of an Orgelpunkttokkata, now changed almost
beyond recognition but – like the E Prelude WTC1 – conveying an
impression that the movement is its own prelude and fugue. In b. 97,
the episode themes take over the pedal point and so unite material from
sections A and B.
As in many highly organized Bach movements, there seems at times
no particular reason why one theme or section rather than another
appears at certain moments, e.g. A3 at b. 85. Similar episodes at bb. 68–9
and 115–16 lead to different sections. Both at b. 85 and on other occasions
(in particular b. 120), there is a degree of abruptness not found in the
best seamless Brandenburg Concerto movements. Perhaps this is itself a
sign that no manual changing is required, since that abruptness scarcely
needs emphasizing by any additional change of timbre; perhaps it is also a
sign that the composer was governed by his twenty-four-bar structure.
After the opening exclamation, what follows in b. 6 is ‘conversational’:
a figure that springs naturally to mind when Bach requires contrast (see B
minor Prelude BWV 544.i, b. 56). Triplets add to the acceleration from min-
ims to semiquavers, and from b. 13 strict invertible four-part counterpoint
follows before a Neapolitan sixth (b. 19) makes yet another dramatic con-
tribution to a work in C minor. The insistent triplets leading to the cadence
(b. 21) correspond to the rhetoricians’ anaphora (‘repetition’), which like
the polyptoton (‘sequences’) in section B are natural to music. In its coun-
terpoint, this fugal section clearly picks up previous ideas (compare bb. 20
and 26), and triplet leaps produce unusual harmonic effects when inverted
and thus unresolved (bb. 109–11).
The triplets are the easiest figure of the movement to develop, leading
to little episodes like bb. 78–81 (where one pattern can be heard at least six
times) or to pretty sequences bursting out time and again (bb. 44, 102, 109).
Even so, it is unexpected that the triplets can be doubled in thirds against
a subject also doubled in thirds, which is what happens in b. 82. Another
concatenation occurs with the subdominant entry in b. 97, where subject
and countersubject are combined over the original pedal point, going on to
the only chromatic episode of the movement.
110 BWV 546
Fugue
‘Weaknesses’ heard in this Fugue – a listless subject, an unambitious coun-
tersubject, an out-of-style episode (b. 121) – have led some to attribute it to
another composer (Kellner), perhaps something ‘looked over’ by Bach or a
‘torso’ completed by another composer (Breig 1995 pp. 17f). Such doubts
come from later assumptions that a fugue has to be ‘bigger’ than its pre-
lude, and it is true that the first sixty bars suggest a fugue different from
what the quaver figuration gradually brings about. But in joining a five-part
exposition with imitative episodes exploring one of Bach’s base motifs, the
movement is of great interest: from section A a quaver motif emerges on
which a new section B is based, the two then combined. B is itself not fugal,
nor does it appear in A2 without much re-writing – a better reason to doubt
the authenticity?
A 1–45 exposition, five parts; episodes in alla breve

counterpoint
45–59 episode, quaver figures, tonic entry; mini coda (57–8)
B 59–86 invention-like development in three parts, of a
quaver figure (d in Example 51) found in every bar of
section B
A2 86–121 quaver figure in most bars, plus subject as a double
fugue; episode, double entry in relative 104, then
subdominant
(C) 121–39 free episode, quavers (derived?) embellishing the
crotchet figures heard earlier (e.g. pedal from 99)
(A3) 140–59 final double entry; coda 145, with ideas from A (pedal
theme 151), B (quavers) and C? (crotchets); cadence
as A1
Very puzzling is the free episode from b. 121. Spitta is right to see that
‘the most it has in common with the rest is the on-flowing quavers’
(I p. 583), but this says more than it appears to say, since on-flowing quavers
have characterized the fugue since the end of the exposition. See Example 51.
The quavers take over the Fugue, are adapted for B (often misleadingly called
a fugue or fugato), and at least some of the bars are ‘superfluous’ (Breig 1995
p. 17). One could simply omit bb. 121–37. Did someone add them? A long,
quasi-galant episode such as this is unlike any other in Bach and suggests
J. P. Kellner, except that just as light and quasi-galant is the echo theme of
the E Prelude BWV 552.
Elsewhere, the lines are in style. Such bars as 59–86 belong to the same
family as passages in BWV 540, 537, 661, 733 etc.; the counterpoint of
b. 73 or b. 98 is found note for note in the chorales BWV 694 and 646; and
111 BWV 546–547
Example 51
the countersubject of the A Fugue WTC1 can also be discerned here. The
closing bars and the quaver imitation running into them even anticipate
the Ricercar à 6 from the Musical Offering, and both the harmonic tension
in general and the Neapolitan D in particular (b. 151) are surely beyond a
Kellner, however versed he was in mature Bach works.
The way the quaver motifs wind in and out of the texture could lead to
unusually convenient manual-changes: Positiv with the left hand of b. 59,
Hauptwerk with the right hand of b. 86, Positiv with the left hand of b. 115,
Hauptwerk with the left-hand f of b. 140 (and with the right-hand g ).

No Autograph MS; copies by or via J. P. Kellner (P 274, after 1730?), C. P. E.
Bach (P 290) or Kirnberger (e.g. Am.B.60, P 276); good eighteenth-century
sources (Lpz Poel 32 from autograph?, Lpz MB MS 1), also via Kittel or
based on P 274.
112 BWV 547
Two staves; title in P 274 ‘Praeludium pro Organo pedal.’, in Lpz MB MS 1

‘Praeludium con Fuga ex C pro organo pleno’.
The sources and obvious maturity of musical detail, plus (in their dramatic
chords) as close a relationship between Prelude and Fugue as is ever demon-
strable in Bach, all point to a Leipzig origin. The dramatic chords towards
the close of both are complementary – dominant sevenths in the Prelude,
diminished sevenths in the Fugue – and both movements are built from
short, ‘neutral’ subjects looking at first hardly likely to lead to expansive,
original treatment. They were surely always coupled.
In the Prelude’s melody and the Fugue’s counterpoint the movements are
unlike any others, and both have a carefully planned finality. The Prelude is
spun out from its simple motif, almost at times ad hoc; the Fugue also has an
elemental subject open to wide, quasi-spontaneous development. The grand
pedal point of the Fugue ‘answers’ the succinct close of the Prelude, and the
final stages of both are derived from their respective themes. Presumably it
is its blend of the original and the traditional that has caused the work to
be dated variously, from c. 1719 (Stauffer 1980 pp. 57ff.) to even the 1740s
(Stinson 1990 p. 117).
There are similarities between several examples of five-part counter-
point in C major – the Fantasia BWV 573 (AMBB), the Prelude BWV 545a
and the present Fugue (bb. 54–5) – and comparable are the present fugue
bb. 66–72 with other final pedal points in C major, notably that of the
Canonic Variations. The similarity between the Fugue and the chorale BWV
677 is as puzzling as it is unique; see below. Since the dramatic diminished
7th chords also match those in another Clavierübung III chorale, BWV 681,
one might expect all three works to be roughly contemporary.
Prelude
Octave imitation at the start of a prelude or set of pieces is not rare (Inven-
tions Nos. 1–4, first Canonic Variation BWV 769, J. K. F. Fischer’s Ariadne
musica), but combining it with a pedal quasi-ostinato is more arresting. So
it is in ‘In dir ist Freude’ BWV 615, but in BWV 547 the theme is worked in
a more complex way.
The form is intricate, based throughout on at least three ideas, the second
much like a decorated version of the first: see Example 52. Each presents
a key rhythmic unit of compound time, and being simple, can be easily
inverted or converted into continuous semiquavers. There are two other
ideas: a countersubject (rh b. 2) and the detached pedal note, which comes
into its own in the dramatic chords near the end. Since the countersubject
rhythm is not the same as the pedal’s but its opposite (trochaic not iambic),
the latter need not ‘originate’ in the former (as Keller 1948 p. 117 suggests).
113 BWV 547
Example 52
Such motifs require particularly careful phrasing in BWV 547, not least
because too light and gigue-like a manner should be avoided, as with the
comparable Prelude to the English Suite in D minor BWV 811.
The motifs of Example 52 appear constantly throughout:
I 1–8 four rhythmic-melodic subjects, all in tonic

8–13 modulating episode derived from the motifs
II 13–20 as 1–8, dominant, parts exchanged, often plus extra part
20–31 modulating, derived episode
III 31–48 octave imitations A minor, D minor; episode
(39–43 = 25–9 a step higher); parts exchanged; last
bar in sequence to
IV 48–54 octave imitation in F; episode to:
54–60 octave imitation in C (54–7 = 48–51); episode 58–9,
cf. 22–3
60–79 octave imitation in G, chromatic; then in
C (60–7 = 31–8 down a tone); more chromatics
(68–72 = 25–9 in minor); dominant pedal point
V 80–8 tonic pedal point, motifs above; last reference to opening
subjects (83 = 5, 84–5 = 4–5), including the octaves
Depending on how one views the motifs, three sections can also be discerned:
bb. 1–24, 25–76, 77–88. Dating it as early as c. 1719 because of parallels to
the First Brandenburg Concerto’s ritornello (Stauffer 1980 p. 60) underrates
its complexity.
The nature of the triadic themes, including others not listed, allows them
to be easily combined, so that (e.g.) c can either follow a (bb. 56–7 etc.)
or be combined with it (b. 60 etc.), a can combine with b inversus and
d (b. 58) or d with b rectus (b. 63), and so on, as if there were just one
theme-complex. In view of the Fugue subject’s metamorphoses (see below),
an interesting quality in this music is how easily it modulates, leading to a
harmonic crescendo (b. 32, b. 37, b. 62 etc) resolved by Neapolitan sixth
(b. 29 had been a phrygian cadence), and so to the unique and startling
detached chords before the final pedal point. F minor and G minor are keys
not usually so well established in a C major prelude as here, and any formal
account of bb. 39–43 or 68–71 in relation to bb. 25–9 hardly hints at such
exceptional foreign tones countering all the sounds of C major.
114 BWV 547
A pair of expositions leading eventually to a final pedal point outlines

a shape more like traditional organ toccatas than Vivaldi concertos (Klein
1970 p. 77). The movement is a motivic fantasia with more internal repe-
tition than one might expect, and despite a concerto-like contrast between
static and non-static sections, the opening does not feel quite like a ritor-
nello statement. But sections alternate, and a glance will show how varied
is the harmonic rhythm. One curious consequence is that almost all first
beats have either a 5/3 or 7/3 chord, which only well-managed modulations
could save from monotony. Another is that the Prelude is based mainly
on one-bar phrases (see Example 52), between which are very few tied
notes of any kind. This relying on a few melodic ideas recalls the Toccata
BWV 538, and both works mould traditional keyboard patterns into confi-
dently handled quasi-ritornello forms, both of them original and unique.
Obviously, the repetitious 9/8 metre gives the Prelude its particular unity,
something not there in 6/8 versions of this theme also imitated at the octave,
such as in D. Scarlatti’s Sonata in B major, Kk 334. Related to but distinct
from this 9/8 are the horn motif and triads at the beginning of Cantata 65
(1724) – Example 53. Note the motif at ‘praise of the Lord’, for both this and
the bare octaves occur in the organ prelude. The performer who dislikes a
light, springing style for the Prelude would agree with Kirnberger’s remark
that 9/8 as distinct from 9/4 can ‘easily acquire the appearance of the light
and trifling’ (Die Kunst des reinen Satzes, 1774–9, II.i, p. 128), which he
illustrates with a theme in G major similar to Example 52 (a). Naturally, the
3 × 3 of compound triple time has been seen as ‘representing the Trinity’
(Siedentopf BJ 1974 p. 73), as presumably can all the triads.
Example 53
In two respects the pedal is used differently in the two movements:

without either main theme or tied notes (suspensions) in the Prelude, but
with both in the Fugue (see the pedal’s very first note!). Its chromatic basses
at the big dramatic chords in each are similar, however. On these chords:
both Cello Suites in C major and D major have something comparable at the
end of preludes, the latter built around triplets, suggesting either that they
are all roughly contemporary (early 1720s) or that dating different genres
from their similarities is unreliable.
115 BWV 547
Fugue
Ingenious counterpoint, lines derived from the subject, and a new shape (a
series of expositions) give the Fugue too a unique position in the repertory:
I 1–15 exposition (answers tonal 9, real 10, tonal 13); episode

II 15–27 second tonic exposition, new countersubject (semiquavers
against entries in other keys); imperfect cadence
III 27–34 irregular exposition, subject inversus, patterns rectus and
inversus; episode
IV 34–48 exposition of subject rectus + inversus on E, A and D; then
three entries inversus (on A, D, G) and three rectus either
tonal (C minor, G minor) or real (C minor); brisé link to:
V 48–72 mass-exposition of subject rectus, inversus and augmented
(pedal); from 56, subject twice transformed; pedal-point
coda, subject contracted in stretto and dismembered.
This is a particular kind of fugue in which the opening statement is a com-

plete fughetta of traditional type followed by a series of intricate expositions
showing four ways to handle a theme: rectus, inversus, in augmentatione and
cromatica. So BWV 546, 547 and 548 offer three different solutions to plan-
ning a fugue whose opening statement closes with a perfect cadence. Others
have no such clearcut section.
The new shape gives great power to the delayed pedal, more so than is the
case with delays in Buxtehude. Pedal has been busy in the Prelude but enters
now only for the last third of the piece, draws attention to the augmentation
and the piling-up of motifs above it, and contributes a fifth part. In the
C minor Fugue WTC2 too, an extra voice enters with the bass augmenta-
tion towards its close; perhaps the composer associated such devices with
C major/minor.
Probably in no other fugue of Bach does the subject appear so many
times (over fifty, according to Keller 1948 p. 118), and its type is familiar.
The opening motif incorporates the common little motif (y), while the
angular line z is also found elsewhere: see Example 54. Oddly, the tonal
Example 54
answer to this subject appears in the closing notes of the C major Prelude
as this was revised in order to open WTC2 (c. 1740). But most like it is
116 BWV 547
the exposition of the fughetta on ‘Allein Gott’ BWV 677 (published 1739).
See Example 55. It would be difficult to find two other keyboard works of
Example 55
J. S. Bach with quite such a correspondence. It is equally odd that the very
motif in the C major subject not found in the chorale’s first subject (i.e.
the opening figure y) can actually be found in its second (b. 7). In general,
the piling-up and inverting of thematically derived motifs in BWV 547,
even the strange harmony at e.g. b. 29, is very much of a piece with the
contrapuntal thinking in Clavierübung III.
More remarkable still is the astonishing metamorphosis of the subject in
b. 56 and its answer at the tritone: Example 56. Nor is this transformation
Example 56
merely the result of diminished sevenths such as appear in other C major

works and again later on here: the most remarkable progressions of
bb. 56–8 are not a diminished seventh but the augmented sixth resolved
in b. 57 and the melodic diminished third (tenor) in bb. 57–8. The fugue
117 BWV 547
has other examples of entries altered for the sake of modulation (bb. 9, 39)
and it is noticeable that of the two augmented entries in the pedal at b. 59
and b. 62 it is the latter, with its altered (diminished) interval in the second
half, that produces the better harmony.∗
Bach subjects are often transformed for harmonic effect – e.g. the
D major Fugue WTC2, shown in Example 57 – and usually produce in-
teresting harmonies rather than far-reaching modulation. In such respects
too, therefore, BWV 547.ii is unique.
Example 57
While in theory the episodes of bb. 6, 12, 23, 31, 46 and 53 are unimpor-
tant, most are characterized by a very melodious sequence probably derived
from the original semiquaver motif y. In fact, this motif colours the fugue
as a whole, and almost every bar contains it in one form or another. It exists
in two forms, single (four semiquavers) and double (eight), the longer of
which belongs to the same family as those listed under BWV 537 above.
Example 58 shows some instances, typical of the composer’s motivic com-
position at its densest.
From the prevailing y motif (up or down) spring subject, episodes, run-
ning semiquaver lines, the counterpoint above the pedal augmentation and
the final pedal point. The fugal techniques themselves, looking towards
the ingenuity of the Canonic Variations, are as follows: rectus/inversus lines,
contraction of subject, stretto, augmentation, transformation of the subject,
homophony, rhetorical rests, pedal point, diversions to the subdominant,
and valedictory reference to the subject (see tenor, penultimate bar). Some
of these are already unusual in organ fugues (e.g. augmentation and rhetor-
ical rests), while others achieve a new height: the preparatory chromaticism
before a final perfect cadence can never have been more richly employed
than it is here, over bb. 56–65. The accumulation of all these effects from the
modest start of the Fugue on middle c to the wide, five-part end previews
the Canonic Variations (whose motifs are similar) and contrasts with BWV
548, where by definition the ABA form is not cumulative in the same way.
For the detached chords in both Prelude and Fugue, see two other
fugues of c. 1736–40: the smaller Credo in Clavierübung III (BWV 681) and
No. 1 from The Art of Fugue. That the chord-progression in each of these
∗ This is subjective: while the harmonization of the pedal b at the beginning of b. 61 is ingenious and
imaginative, it can not be said to satisfy all ears.
118 BWV 547–548
Example 58
three fugues includes at least one diminished seventh while that of the
Prelude BWV 547.i does not, may suggest that the passage in the Prelude
was made to match the Fugue’s and not vice-versa.
BWV 548 Prelude and Fugue in E minor

Autograph MS P 274 (fair copy of Prelude and bb. 1–20 of Fugue; the rest by
J. P. Kellner?, c. 1727–32: Kobayashi 1989 pp. 128f.); MS based on this (Lpz
MB MS 1) and others in Kittel circle (e.g. J. Becker c. 1779, J. A. Dröbs);
119 BWV 548
others probably drawing on an earlier version, with da capo written out

(Anon 5 = Johann Schneider?), or via C. P. E. Bach (? P 290) or J. P.
Kirnberger or perhaps Kellner.
Two staves; autograph title ‘Praeludium pedaliter pro Organo’ in P 274,

where da capo not written out.
Whatever the reason for the change of hand in P 274, handwriting and
watermark are as for the fair copy of BWV 544. As with BWV 541, these
copies were no doubt made from older autographs, and were surely Leipzig
works (further in Kilian 1978 p. 62). That the pairing is original is also
suggested by their complementary form: an intricate concerto-ritornello
Prelude versus a clearcut ABA Fugue. Some inner relationships between
them can also be felt. On one level, both make much of scale motifs; on
another, the number of bars in the Fugue (231) relates to the total number
of bars in both (368) as 1 : 1.59, close to the Golden Section (1 : 1.618).
At least since Spitta recognized the ‘life energy’ of this ‘two-movement
symphony’, with ‘the longest amongst Bach’s organ fugues’ (II. p. 690), it
has encouraged warm words. Its riveting power is due partly to the easily
felt balance of two such movements, the first as logical-seeming as a mature
concerto (e.g. BWV 1043), the second an example of how to organize an
extensive fugue. If sources with the Fugue’s da capo written out go back
to an autograph earlier than P 274 (assumed in KB p. 391), then indeed a
literal ABA was for once intended (as one cannot be sure was the case with
BWV 537.ii), and any feeling one may have that A1 is shorter than expected
only makes this the likelier.
Prelude
This sectional ritornello shape is the most intricate amongst the organ
works:
A1 1–5 homophonic, pedal continuo; ‘instrumental rhetoric’

A2 5–7 more polyphonic
A3 7–19 sequences before tonic cadence (after Neapolitan 6th)
B1 19–24 new but continuous material, to dominant
B2 24–33 inner pedal point broken in 27–31 for material from A2
A1–3 33–51 dominant, parts exchanged except 46–8; 40–3 = 7–10
C 51–5 no pedal
B1 55–61 major; closes with reference to A1, now with new bass
C 61–5 as before, down a fifth, top line re-phrased to avoid d
B1 65–9 as 55–61, down a fifth, followed as before by:
A1 69–81 development over a new bass; freer episode (75–81),
same bass; 80 cadence as 69 before its interruption in 70
120 BWV 548
A1 81–90 development; subdominant; parts exchanged (see also

33–7)
C 90–4 an inversus form
B2 94–111 94–103 as 24–33, exchanged; episode (103–11 manual,
from A1), running bass
B1 111–15 as 55, 65 now in C major
C 115–21 C motifs rectus and inversus sequence, for key of:
B1 121–5 as 55, 65, 111 (i.e. B1); parts exchanged; dominant pedal:
A3 125–37 125–36 = 7–18 but re-written for the line to fall from b
down to C; final tierce de picardie
This is more succinct than a concerto Allegro, however, with too hectic a
continuity for that interrupted tonic return often found in concertos, when
the music shoots off in another direction to give the movement more space.
The similarity between the ritornelli of preludes BWV 544, 546, 548, 552 and
those of the Italian Concerto for harpsichord only emphasizes how totally
different they are in effect and Affekt.
Although the writing allows manual changes, they are not as inevitable as
elsewhere, including any concerto models there may have been. The texture
is surprisingly consistent, from three to five parts, with something of a
planned alternation between the two. Apart from four episodes, the pedals
continuously add to the tension, which is barely lightened by passages in
the major.
The first of the Prelude’s subjects is basically homophonic while
others are polyphonic, the opposite of the Fugue. Note that the opening
harmonies and bass are not unlike those of the C minor BWV 546, G minor
BWV 542, and even toccatas of Buxtehude that begin with a strong melodic
gesture above a pedal point. There is a focus on the sensitive soprano range
around e , which contributes to the intensity of a writing that ‘avoids strict
imitation’ (Frotscher 1935 p. 894). Particularly significant throughout are
sequences, spontaneous and inventive, constantly rising and falling. Sub-
jects are both re-introduced and developed, somewhat in the manner of the
Vivaldi partial ritornello. There are few cadences, and what there are usu-
ally rush into the next section, for the ritornello plan juxtaposes material
non-stop, and sections follow each other in almost random order. B1 is fol-
lowed on successive occasions by B2, C, A1, C, A3; and A3 gains finality by
alone quoting substantially from the original exposition – a ‘recapitulation’
typical of mature ritornello form.
Although any similarity fancied between the themes of A, B and C would
differ from Bach’s usual thematic allusion, certain resemblances can be
found: for example, between quaver patterns (bb. 14, 59 and 90). From
b. 1 on, there seems to be in the music either a question-and-answer or
121 BWV 548
a sequence, which is not true of preludes such as the C major BWV 547,
although scales running in sequence do appear in the harpsichord preludes
of the G minor English Suite and the G major Partita. The lines are no longer
traditional like BWV 545 or motivically single-minded like BWV 547 but
much more original, new to the corpus of organ music, and hardly imitable
despite their curious similarity to b. 61 of the B minor Prelude BWV 544.
The polonaise-like appoggiatura chords of bb. 2–3 belong with those of the
C minor Prelude BWV 546, though a ‘general E minor sound’ might remind
one of the opening of Cantata 125 (1725).
Fugue
The subject alludes both to the lament (a chromatic fourth) and toccata
(agitated virtuosity). The tradition for fugue-subjects in E minor to para-
phrase in some way the descending chromatic fourth is suggested by in-
stances in Example 59 and others in the finales of the E minor Organ Sonata
Example 59
and the E minor Harpsichord Toccata. Bruhns’s E minor Praeludium begins

with a comparable paraphrase, as does Kirnberger’s Fugue BWV Anh.III 181.
Similarly, the rocking figure of the first main episode (b. 60) is not unlike one
in Bruhns’s G major Praeludium but more complex: a broken chord with
acciaccatura, as in the Sarabande of the Harpsichord Partita in E minor. Its
harmony (Example 60) is not unlike fugal material elsewhere, such as the
finale of the Vivaldi Concerto BWV 596 (b. 4). But the sparse/rich harmony
of bb. 44–51 is unimaginable with any other composer.
Example 60
122 BWV 548
The movement brings together a fugue (regular exposition), concerto

(‘solo’ episodes), toccata (scales), and aria (da capo), resulting in a virtuoso
ritornello-fugue related to the Finale of the Fourth Brandenburg Concerto.
Against a concerto conception of the movement is the fact that manual-
changing, though feasible, is no simpler than usual, especially at the reprise
in b. 172, although as is clear in the B minor Ouverture BWV 831, a return
can go straight back to the forte manual. Much of the figuration in the
central episode – scales, broken chords, the patterns from b. 120 – recalls
the praeludia of northern composers, and may even be an allusion to them
(see below).
A 1–23 regular exposition (pedal last), constant countersubject,

from which a jumping motif (a) emerges in 9
23–38 episode from the quavers; sequence above a striding
bass; entry in 34 (pedal countersubject)
38–59 episode related to a (with suspensions); entry
subdominant, answer on pedal point; a developed
B 59–71 episode, manual only, new figure; truncated entry in
pedal
72–83 59–71 (modified ending); truncated dominant answer,
pedal
84–93 episode, scales; entry in D, with countersubject, pedal
point
93–112 episode in three sections; second based on
countersubject; entry on G as before, on pedal point
(106–12 = 87–93)
112–41 episodes: scales of two octaves; 116 episode as 25;
at 120 new figure (Buxtehude? Example 61); returns
(124 and 130 as 60 etc.); pedal entry supertonic, with
countersubject
141–60 episode, alla breve counterpoint (141–4 = 145–8);
sequence from 151 to pedal-point entry in C, with
countersubject
160–77 episode, scales as 93 but further; followed by pedal-point
entry en taille, with countersubject. Overlapping with:
A 172–231 da capo; entry at 172 = b. 1; 178ff. = from 6ff. except
for a presumed tierce de picardie (compare the Prelude
at bb. 19 and 137)
Hidden at first, the da capo in b. 172 has a double function (unique in the
organ works of J. S. Bach?) since it is also an entry closing the previous
episode. In fact, at b. 172 it is not at all clear that a full da capo is in process,
123 BWV 548
Example 61
for the pedal point is itself like codas at bb. 51 and 223. The symmetry in
bar-numbers means A + A = B, in which section A is already a complete
fugue with coda. The question what to do when A ends is thus given three
different replies in BWV 546, 547 and 548, and there is no reason to find
the ABA shape ‘inadmissible in fugal composition’ (Schreyer 1911).
Other da capo fugues appear in the spurious Lute Partita BWV 997
and Fugue BWV 998 (both c. 1740?) and a simple E minor Fughetta in
Telemann’s XX Kleine Fugen (Hamburg c. 1731). Perhaps the C minor Fugue
BWV 906 was intended to be da capo, like the semi-fugal finale to the Fifth
Brandenburg, while in the organ works, C minor fugues BWV 537 and 526
had approached it, with A2 modified in some way – shortened, or with
exchanged parts. Closer to BWV 548 are the fugues in the C major Violin
Sonata BWV 1005 (1720?) and the second movement of the Sonata in the
Musical Offering (1747) in which too the main theme returns at first against
further counterpoint. The ABA Duetto in F major in Clavierübung III is
part of the plan to present four specific fugues, and like BWV 548 refers to
A during B.
Close too are those fugues of certain ouvertures or suite-preludes, such
as the D major Ouverture BWV 1068, the English Suites in E minor and
D minor BWV 810 and 811 (ABA = c. 40–80–40 bars) and the B minor
Ouverture from Clavierübung II. In all of them, section B contains simpler
episodic material in which the subject from A1 appears shortened or isolated,
and in which A2 enters unobtrusively, without a break. In this respect,
the present Fugue is quite traditional, furthering an idea realized in the E
minor English Suite but now with new material, more drama and a greater
rhetoric including powerful pedal points. Its drive is spectacular and its
details ingenious.
Although neither subject nor countersubject yields other motifs used
much, the quaver figure of b. 9 is likely to occur anywhere, even inverted
(b. 57) or worked over several bars (bb. 22–31), in easy imitation and
invertible counterpoint (bb. 29–31). The pedal’s crotchets stride against
it, and a similar quaver figure occurs in another mature keyboard work,
the B minor Prelude WTC2 (b. 23). The manual scales present a whole
124 BWV 548–549
repertory, half-bar or whole-bar, ascending or descending, straight or

convoluted, sometimes producing bleak textures (compare bb. 86ff. with
bb. 71ff. of the B minor Prelude BWV 544), at other times weaving around
sequences, of which there are as many here as in the Prelude.
The final episodes juxtapose clearly different styles:
120ff., 132ff. a Buxtehudian figure (an allusion to bb. 74–6 of

Buxtehude’s Praeludium in F minor?)
141–9 alla breve style (traditional four parts)
150–5 Italian sonata style (invertible counterpoint above a bass)
160–4 a French rondeau progression, decorated with scales
(Ex. 62)
Example 62
In a composer so alert to style as J. S. Bach, such a ‘repertory’ is unlikely

to have come about by accident. For example, a progression in the ‘Grand
Dialogue’ of Louis Marchand’s MS ‘Troisième Livre’ bb. 15–20 (EF 90.400)
is close to Bach’s bb. 160–4, and both belong to the same family as a chaconne
en rondeau in the Deuxième Recréation Op. 8 of J.-M. Leclair (c. 1737), where
ninths and sevenths are typical.
The homophonic episode of bb. 120–35 is far better integrated than the
modish final episode of the C minor Fugue BWV 546. BWV 548’s episodes
never flag; sequence succeeds sequence (bb. 164–7, then bb. 168–70), and
the da capo is all the more striking. Not the least remarkable feature of the
fugue is that the truncated entries in the middle section quote the fugue
subject and do no more with it, though in the process placing subject-
entries on degrees of the E minor scale from E to D. As in other long fugues,
such as the Ricercar à 3 in the Musical Offering, the composer seems to be
deliberately walking a tightrope by interpolating new material and creating
his own version of the ritornello fugue.
BWV 549 Prelude and Fugue in C minor

No Autograph MS; copies from later eighteenth century, via C. P. E. Bach
(?P 287, 289, 319, LM 4838) or J. C. Kittel (?P 320, Lpz MB III.8.22); see
also BWV 549a.
125 BWV 549–549a
Two staves; title in P 287 ‘Praeludium pedaliter’.
Some sources also contain the Prelude and Fugue in E minor BWV 533,
which may imply that it was known in this form quite early. Nevertheless,
the oldest copy by far is of the D minor version, and ‘there was no reason for
transposing it up, but a very good reason for transposing it down’ (Emery
1959 p. iv), i.e. to avoid pedal d in the opening solo. Because, unlike other
works in D minor, BWV 549a happens not to use bottom C, the transposition
was straightforward.
The BWV order 549/549a arises because BG and/or Peters IV gave only
the first, not knowing the Mö MS. While the two versions are close enough
to imply that Bach need not have made the C minor version himself,
reliable copies could mean that he countenanced it during the Leipzig period
(KB p. 319).
BWV 549a Praeludium (Prelude and Fugue) in D minor

No Autograph MS; copies in SBB 40644 (Mö MS, J. C. Bach) and later
eighteenth-century source (P 218, shortened), also a lost copy by J. P. Kellner
(fugue only).
Two staves: title in Mö MS ‘Praeludium ô Fantasia. Pedaliter’.
That the only organ praeludia copied by J. C. Bach in Mö MS were BWV 531
and 549a underlines the complement each is to the other. Shared ‘Böhmian’
details are:
BWV 531, BWV 549a

A pedal solo; manual develops pedal motif; ambiguous pedal part
B fugue (four entries, but only two or three parts); pedal only at end
(as if it ‘arches back to the pedaliter prelude’ – Breig 1993 p. 49)
C coda, ‘growing’ out of patterns from the subject, and so integral
to fugue, but developing freer demisemiquavers
but differences are both consistent and conspicuous:
BWV 531 BWV 549a

major and longer minor and shorter
A pedal detached; thematic; textures pedal points only; consistently
broken in four or five parts
126 BWV 549a
B high, descending exposition; low, ascending exposition

tonal pedal entry integrated pedal entry, homophonic
C coda without pedal until the coda with pedal, thematic at
pedal point; final perfect first; no pedal point; plagal
cadence cadence
In the light of this, the sudden turn to the minor towards the end of BWV 531
begins to look like an equivalent to 549a’s tierce de picardie. In BWV 531, the
alternate-foot technique of the pedal solo leads to repeated figures, in BWV
549 to sequences; in BWV 531 the pedal’s final octave leap is followed by
rests, in BWV 549 by a pedal point. Such a catalogue of differences is possible
between other pairs of preludes and fugues, but here they are patent and
might even be meant to influence performance. For example, the continuous
demisemiquavers closing BWV 549a suggest a gradual rallentando, as the
close of BWV 531 does not.
Some problems arise in Mö MS probably because the or an original
was in tablature: the bass hiatus in bb. 14–15 (lh and pedal share the G,
or pedal keeps E?), the curious ornament in b. 45 (the tablature had a
wavy line?), uncertain distribution between the hands in bb. 56–7, etc. It
must be correct to hold the final pedal D of the Prelude (Bruggaier 1959
p. 177), though the C minor version suggests not, and the same with the
pedal’s other plagal cadence, in the Fugue. Perhaps b. 8 of the Prelude in
BWV 549a was altered in BWV 549 by copyists unfamiliar with pedal solos
that came to a close with their own perfect cadence (cf. Böhm’s C major
Praeludium).
Prelude
The pedal opening recalls extant praeludia of Böhm more than any other
composer, but the four-part counterpoint is a sustained version of what
happens in Buxtehude praeludia once the opening pedal or manual solo has
ended. Bars 9 to 18 – familiar from WTC1 (E Prelude) and elsewhere – are a
florid version of durezza suspensions, attempted too by J. K. F. Fischer. Also
Fischer-like is the homophony of bb. 20 and 24, an early idiom discarded by
the maturer composer though found in Buxtehude (Toccata in F) and in the
present Fugue (bb. 41ff.). If the motif-repetition in bb. 25–6 is Buxtehudian,
the chords are Bruhnsian, to judge by extant works.
Since, given the simple harmonies of the movement, the composer could
have employed the same motifs throughout, it seems that so far he had little
interest in such integration. A different unity is provided by the pedal points
of varying length, covering the diatonic steps between D and B.
127 BWV 549a–550
Fugue
The Fugue, whose long, unusual subject might derive from a motif in the
Prelude, consists chiefly of a series of entries, the first five of which rise in
tonic and dominant steps at regular four-bar intervals. Such regularity is
out of the question for genuine five-part expositions such as that in the C
minor Fugue WTC1. To a degree unusual in Bach, both Prelude and Fugue
centre on contrapuntally embellished tonics and dominants, in a manner
not unlike Buxtehude’s C major Fugue BuxWV 137, where these harmonies
eventually produce an ostinato. The late pedal entry on the keynote is a
precursor of the C major Fugue BWV 547, unlike whose subject, however,
BWV 549a’s has a folksy Thuringian quality one also hears in Buttstedt.
Though not those of a permutation fugue, the first countersubjects share
a rhythm: the little dactyl figure at b. 5 (cf. the E major Toccata BWV 566,
b. 40). Gradually, the two- and three-part counterpoint is overtaken by
semiquavers, spinning out as in some later fugues, and continuing over
the eventual pedal entry. This is a full entry and appears in elementary
stretto∗ before swirling away under toccata-like chords. Otherwise, this is
a manualiter fugue (Musch 1974), becoming at the pedal entry more like
a toccata. The coda from b. 46 develops previous motifs before the scales,
as does BWV 531. Bars 46–55 bear more than a passing similarity to the
closing section of the D minor Toccata BWV 538.i, as do bb. 52ff. to the
C minor Fugue BWV 575, and b. 58 to the G minor BWV 535a. The final
plagal cadence repeats the Prelude’s, while both cadences in BWV 531 are
perfect.
Two manuals are practical: II at b. 22, I at bb. 28 or 39 (right then left).
From bb. 47 to 52 the manuals can be alternated, first at each beat and then
at each half-bar.
BWV 550 Prelude and Fugue in G major

Autograph corrections on first 2 pages of P 1210 (Leipzig period?); Lpz
MB MS 7 (J. N. Mempell †1747), P 1090 (G. A. Homilius, a pupil c. 1740);
copies directly or indirectly via C. P. E. Bach (P 287), J. P. Kellner (P 642, 924)
and perhaps Kittel (?LM 4839a). Lost copy, perhaps by D. Nicolai (a pupil?
c. 1729).
Two staves; headed in P 1210 ‘Praeludium pedaliter’, fugue ‘alla breve e

staccato’.
∗ Spoilt, if one cuts four beats before the pedal entry, as suggested in BG 38.
128 BWV 550
The section bb. 46–62 is absent in P 642 and 924 (KB p. 421), and indeed the
Prelude could end in b. 46, a moment strangely like the G major BWV 541
at bb. 79–80. The version P 1210 is an instance of a copyist altering compass
(no pedal notes above d ), while P 287 is one of adding ornaments, as in
other MSS connected with C. P. E. Bach.
With its Bruhns–Buxtehude elements, the work seems to be another of
Bach’s early Weimar essays in writing long fugues, without postlude but with
a minimal interlude and a prelude that develops sustained sections. More
than BWV 549a, it explores a quaver pattern familiar in northern praeludia,
starting with manual, then pedal, then both (more or less) together. The old
‘sectional prelude’ of BWV 532 is now integrated by means of a persistent
motif, without the intense knitting together yet of related motifs as in BWV
541. In such respects, the closest work is the A major Prelude. Sectional
tempi are probably proportional: 3/2 minim = grave crotchet = alla breve
minim.
Prelude
As Spitta pointed out, Buxtehude also created a prelude from such material,
if less extensively (I p. 403): the A minor BuxWV 153 imitates a motif
taken up in a pedal solo, has derived counterpoint in four parts, and ends
with a tonic pedal point. But BWV 550 is three times as long, and original
in expanding a single idea over the old tonic–dominant–tonic plan. The
solo for pedal passes through its whole compass and has the clear, on-beat
harmony typical of Bruhns, driving up to the cadence of b. 46.
As elsewhere in Bach, the motif has shorter and longer forms, the first full
of gesture, the second more continuous: Example 63. The gesture is startling,
Example 63
as is its metre: does it begin in 2/2 or 3/2? The ambiguity contradicts the four-
bar phrases and the typical square motifs (cf. Vers III of Cantata 4, c. 1708),
and the metre continues to be handled dextrously, with unexpected hemiolas
(bb. 28–9, 43–4) and sequences of both two-bar and one-bar phrases. The
pedal solo produces the desired continuity, with little modulation until after
the point d’orgue, and the motif leads naturally to little harmonic ostinatos
à la Buxtehude (bb. 9–10, 38–9). The hemiola at bb. 43–4 supports the idea
that the Prelude was first meant to end at the cadence in bb. 45–6.
Perhaps originally the third beat in bb. 10 and 39 repeated the motif
unaltered, resulting in the unresolved fourth found not only in Buxtehude
129 BWV 550
(F minor Praeludium, b. 78) but in maturer Bach (G major Organ Sonata,
first movement, bb. 7 and 167). Familiarity with this effect is evident not
only in the Passacaglia but in the arrangements of Reinken’s Sonata prima
of 1687: see BWV 966 for examples. Ultimately, broken chords of persistent
harmony are a form of bariolage, q.v.
There are enough glancing similarities between this praeludium and
Bruhns’s in Mö MS to suggest that organists around 1700 had a ‘G major
vocabulary’, even if the dominating motif does not grow yet into a form as
complex as the D minor motifs in the Toccata BWV 538. One particular
sequence, in bb. 40–2, seems to belong to the same family as that of the
C major Toccata, bb. 67–70. As for date: the pedal e , integral on its two
appearances, has led some writers to seek an organ on which it could have
been played during the Weimar period, e.g. Weissenfels (Klotz IV/2 KB
p. 68), but other organs in the Weimar area were also possible (Kilian IV/5–6
KB p. 405).
‘Grave’
In theory related to the sustained interludes in Buxtehude’s praeludia and
chorales (‘Wie schön leuchtet’, bb. 74–6, noted in Keller 1948 p. 79), these
three bars have no more harmonic tension than similar preludes in Kuhnau’s
published suites (1689), despite two diminished 7ths and five parts, as at
the end of the D major Prelude. Very early or inauthentic?
Fugue
On tempo, see above. The direction ‘staccato’ could reflect either a copyist’s
ideas or a tradition for playing repeated-note subjects, broken triads and
chords, such as are found throughout the Fugue even at non-thematic mo-
ments. The unusual keyboard style is most like the Jig Fugue’s, particularly
at the close. The shape is also unusual:
62–95 five entries but three or four parts (cf. BWV 531, 549); first
answer tonal, second real; derived countersubjects
95–117 episode, first with pedal; two entries without (99, 107);
related countersubjects, partly repeated
117–44 two entries in relative, no answer; derived episode to a
series of:
144–202 quasi-stretti in dominant of relative minor, dominant,
supertonic minor, subdominant, tonic (two), all followed
by derived episodes
202–20 derived coda
Pedal entries are timed asymmetrically and material is developed with some
variety, despite an apparent sameness in the entries. Note that the very
130 BWV 550–551
striking en taille effect of the last entry (b. 192) has been prepared by the
tenor being silent for four bars.
Criticisms levelled at the piece seem not to recognize its distinct genre,
for at least its subject is related to others of the period in G major, such as
Handel’s HWV 571 (c. 1705). Criticism probably also underrates the way
the Fugue develops triadic figures as exhaustively as the Prelude develops
its motif. In the Bach conception, Prelude and Fugue are complements, not
using similar figures as such (despite claims to the contrary) but each work-
ing out its own. The episodes, though simple, weave triads to varying effect
(compare bb. 139ff. with 171ff.), and the subject is so easily transformed
that there is curiously little exact repetition. This is true of the counter-
subjects too, and if ‘one cannot speak here of counterpointing’ (Frotscher
1935 p. 866), ‘counterpoint’ is being defined too narrowly.
Similarly, though often threatening too much spinning out, the various
sequences are held in check, passing quickly to the next (as in the Prelude)
and preparing well for such entries as bb. 182–92 – a passage close to the D
major Prelude, as is much of the pedal writing. A concentration of chords
at the close is created by running further with both subject and various
countersubjects, which join in naturally since they use the same motifs.
The climax is more dramatic than the D major Fugue’s, with a close far
more succinct than was usual in the new long fugues of the early eighteenth
century, such as J. G. Walther’s Prelude and Fugue in C.

No Autograph MS; copies in P 595 (J. Ringk, after 1730?), Lpz MB MS 7
without first 11 14 bars (J. N. Mempell †1747, from Ringk’s?).
Two staves; title in P 595 ‘Praeludium con Fuga ex A Moll. pedaliter’.
Like the Toccata BWV 565, this now goes back to a copy by Johannes Ringk
(1717–78, pupil of Kellner), and is equally dubious, as its text is ‘unreliable
and full of mistakes’ (KB p. 566). If a Bach work, it shows signs of being
‘only an imitation . . . written before Buxtehude’s manner had been fully
understood and enlivened by the composer’ (Spitta I p. 316), i.e. before the
E/C major Toccata (Breig 1999 p. 648) and even before the Lübeck visit
(Keller 1948 p. 48). Insofar as the source can be trusted, another sign of
north German influence is the independence of the two fugue-subjects.
Buxtehude’s Praeludium in G minor BuxWV 148, with its toccata, fugue
and ostinato sections, was copied by J. C. Bach and possibly the young J. S.
Bach before the Lübeck visit (see Franklin 1991).
131 BWV 551
Despite their differences, some similarity can be discerned between the

two fugues, both of which have most entries in the tenor, and the symmetrical
plan might mean a common tempo:
1 prelude based partly on scale fragments (bb. 1–12)

2 fugue with chromatic subject and key semiquaver figures (12–28)
3 short sustained five-part section (29–38)
4 fugue (a section of 3 × 12 bars – Meyer 1979) with chromatic subject and semi-
quaver figure from 2 (39–74)
5 postlude based partly on scale fragments (75–89)
Certain parallels can be drawn with the five-section harpsichord toccatas,

BWV 910–915, all showing the ‘fluency with which Bach speaks Buxtehude’s
language’ (McLean 1993 p. 37), all symmetrical and thus unlike such prae-
ludia as the E minor BuxWV 143. A certain harmonic drive in the work as
a whole anticipates later work of J. S. Bach.
First section
Perhaps the section was too old-fashioned for Mempell to complete his
copy? Its tail-chasing figuration is not unlike that elsewhere (Buxtehude’s
G minor Praeludium BuxWV 149, Vincent Lübeck’s Praeambulum in C
minor), as are the three-part texture and a pedal point after semiqua-
vers. There is something one might hear as Bach-like in the insistence of
bb. 10–11, an insistence also found in certain ‘Neumeister Chorales’.
Second section
The chromaticism recalls many a seventeenth-century subject, for example
that of Buxtehude’s G minor BuxWV 176. There are two similar expositions,
the second moving to the relative; both are based on more answers than there
are parts (descending from e down to A), and both anticipate the exposition
of BWV 531. Like the simple imitation, such bars as b. 20 remind one of
South German styles. Spitta found the subject ‘melodically expressionless’
(I pp. 316–17), but it has three specific motifs: the trillo, the four-note
pattern, and the chromatics, all conspicuous.
Third section
The brief interlude is very much in the style of Bruhns or Böhm (gestures,
rests, caprice, durezze) or Buxtehude (BuxWV 149, 139, 142, 151), of whom
the f in bb. 29–32 is also characteristic (picked up from Frescobaldi?). The
durezza passage is no more chromatic than with the North Germans, its
progressions like those in Buxtehude’s A major Praeludium (from b. 64);
but the increase from four to five parts is typical of a Bach Grave (BWV 532,
132 BWV 551
549, 550, 564). To elaborate the passage with runs and other figurae ‘in the
Italian style’ is recommended in McLean 1993.
Fourth section
Sweelinck’s Fantasia in G has been claimed as influencing this section (Keller
1948 p. 49), but double subjects of which one was chromatic had long
been familiar, chiefly through Frescobaldi’s published fugues. The two-part
counterpoint (bb. 45–7, 51–9) is like that in similar work by J. C. Kerll and
others. Also, the angular ‘countersubject’, unconvincingly given to pedal in
most editions (bb. 44, 60 and a surely garbled b. 51), would not be out
of place in an Italian string trio. How far this subject is related to the first
fugue’s is not obvious, despite claims sometimes made, although all three
subjects do have a common quality: see Example 64.
Example 64
Although the two fugues exploit invertible counterpoint, with stretto and
spinning lines, there is no attempt at full permutation. The fugal writing
does not go much beyond three parts, yet there is variety of texture and
tessitura, and such a passage as bb. 65–73 contains both thematic cross-
reference (as if to both fugues) and Bach’s hallmark semiquavers. If it is
genuine, it represents an important step in the composer’s development.
But in an A minor fugue, the C minor passage at bb. 63–4 is as out of the
ordinary as the C minor entry in the D minor Toccata BWV 565, arousing
suspicions of Ringk’s MS.
Fifth section
The perfect cadence isolates a coda built on semiquaver figures familiar in the
genre but new here. The postlude can be seen as one long drawn-out plagal
cadence, finally breaking up the phrasing as in many northern praeludia and
using such common-property devices as the double trillo (BuxWV 149, 152,
155, 140 and BWV 533, 574, 543 and 532). The opportunities for dialoguing
between manuals are clear, particularly if bb. 83–4 are reversed, as perhaps
they should be; or the last beat of b. 4 put down a tone (NB uncertain alto
here).
133 BWV 551–552
The final bar with its d again recalls those cadences of Buxtehude in
which the subdominant is strong and/or the cadence is plagal (cf. BuxWV
153 in A minor, or mixolydian fantasias of Bull and Sweelinck). For the
manual sixths and the pedal figures, compare BWV 531. The question is:
do these stylistic allusions confirm it as a Bach work or, on the contrary,
something more likely to be the pastiche of a well-informed imitator?
BWV 552 Prelude and Fugue in E major

(Clavierübung III)
Published 1739: see BWV 669. No Autograph MS (? one referred to in 1774
by C. P. E. Bach, see Dok III p. 277); subsequent copies, only of the print.
Two staves; heading ‘Praeludium pro Organo pleno’, ‘Fuga à 5 con pedale
pro Organo pleno’.
Though united in key, number of parts (five) and themes (three), and un-
derstood as belonging together by such early writers as Forkel, the Prelude
and Fugue were printed apart in Clavierübung III, sometimes copied singly
during the eighteenth century and not always played together in the nine-
teenth. There may or may not be a significant proportion operating in and
between them: Prelude (205 bars) + Fugue (117) = 322, and 205 : 322 =
1 : 1.57, close to the Golden Section 1 : 1.618.
Since the first plan for Clavierübung III may not have included the open-
ing and closing pieces (see below, p. 388), perhaps E major was not their
original key? – D major is more likely for an ouverture or concerto, and
the Prelude’s E minor then becomes D minor. But transposition is not
demonstrable, and perhaps the composer knew both another E ouverture
(Couperin’s Quatrième Livre, printed 1730) and a remark of Mattheson
that this ‘beautiful and majestic key’ was not in the head and fingers of
most organists (1731 p. 244).∗ It is unknown how well E major suited the
Leipzig organs potentially associated with Clavierübung III (Thomaskirche,
Paulinerkirche), but both it and BWV 687’s F minor can be seen as modern
gestures.
Prelude
With BWV 540, this is the longest of the organ preludes:
∗ PerhapsMattheson’s treatment of double fugues (1739 pp. 440ff.), with examples from Handel,
encouraged the double fugues in WTC2?
134 BWV 552
A1 1–32 32 bars (2 × 16, cf. Aria of Goldberg Variations)

B1 32–50
A2 51–71 first part of A
C1 71–98
A3 98–111 second part of A
B2 111–29 as before, up a 4th; 129, 1 bar of A
C2 130–59
C3 159–73
A4 174–205 31/32 bars (overlaps C3, as the da capo in BWV 548
and 803)
Though A and B have an even number of bars, the sections are fluid and
could be further subdivided. The dotted figures dominating A can be spun
out, their lines inverted, or interchanged (compare bb. 17–18 with 1–2), or
decorated. On this last: compare the scales of bb. 54–7 with sections of the
E minor Fugue BWV 548. The second C section is not only extended but
begins and ends in an unexpected way: in bb. 129–30 a return to A is more
expected, and at b. 174 the key is C minor, not E major. A4 is the same as
A1 except that its return is disguised. The Preludes in B minor and C minor
also include a fugue after the previous section has come to a full close, but
as a second section, not the third of three sections as here. The Goldberg
Variations’ focus on 32 (32 movements, 32 bars in each, 32 pages) must be
roughly contemporary.
In Clavierübung II there had already appeared in print similar elements
of both the French ouverture (dotted rhythms, short runs, emphatic ap-
poggiatura chords)∗ and the Italian concerto (contrasting episodes, a devel-
oped ritornello form). But the E Prelude is unique, more continuous and
with fewer semiquaver runs than would be expected, so modified for organ
that to continue to describe it as a French Overture tout court (Horn 1986
p. 268) or even merely as ‘in the style of a French Ouverture’ (Breig 1999
p. 698) may be misleading.
The contrast between the three themes or sections is very striking, and
might be interpreted with respect to the Trinity (cf. Humphreys 1994):
A five-part contrapuntal harmonies based on two-bar phrases open

to extension and motivic development: the Father, majestic, severe
B staccato three-part chords, quasi-galant nature; one-bar phrases,
echoes, repeats, sequences; not further developed: the Son, the
‘kind Lord’
∗ Compare the opening chord of b. 2 with the same point in the Ouvertures of the Partitas in D major
and B minor.
135 BWV 552
C double fugue (three-part invention, modified countersubject),

built on semiquavers: the Holy Ghost, descending, flickering like
tongues of fire
As the piece proceeds, A remains much the same length while B becomes
shorter and C longer. None is typical of organ music of the 1730s and gone
are all toccata-like passages, though there are incidental reminiscences of
earlier ‘German’ works such as BWV 535.ii (pedal, b. 145), 544 (b. 147) or
739 (b. 163).
The three sections share a pulse but their styles are different, just as in
the fugue the three themes share a style but their written pulses are different.
The fugue theme is transformed for pedals in the usual way (compare the
E major Toccata, second section BWV 566.ii), requiring a conventional
alternate-foot technique (Bruggaier 1959 pp. 59–67). This transformation
also underlines the fact that the pedal does not take part in C1, and that it
is chiefly on its behalf that C2 is so much longer. Altogether, the pedal has
a different function in each section:
A a ‘modern’ bass, an instrumental basso continuo

B a pedal quasi-pizzicato bass, also ‘modern’
C absent at first, then an old-style pedal line (alternate-foot
pedalling)
In none has it kept its old role of providing pedal points at the beginning,
whether actual (BWV 546 etc.) or implied (BWV 548 etc.).
The double fugue subject C1 resembles that of BWV 546.i (b. 25) in
both the syncopations of the upper voice and the rising scale of the lower.
Again, this lower subject does not at first appear in the pedal though it
is a conventional fugue-subject – compare this subject with the C major
BWV 545, which has been exaggeratively claimed to be ‘borrowed’ for the
E Prelude. The change to minor at b. 161 is puzzling until it is seen as
various things: a contraction of C1, a change for variety and for a sense
of impending close (cf. minor at the end of both Prelude and Fugue in
C BWV 547 and in A WTC2), a reference to the previous minor (b. 144),
and a detail typical of Clavierübung III (see E minor Duetto, bb. 35–7).
Here, an Italian form absorbs a range of elements, therefore, through
a key-plan centring on E at crucial moments (bb. 32 and 130) but
with some unexpected modulations at bb. 91, 161 and 168. The contrast
between themes results not in a Vivaldian concerto form as such but in
an organ-like alternation, with both contrast and repetition. The result
used to be thought ‘monotonous’ here and there (Grace 1922 p. 226), but
its blending of the conventional and the new can now be better understood.
136 BWV 552
Thus the conventional two-part figuration in bb. 86 or 147 (compare the B

minor Fugue BWV 544) and the three-part in bb. 93 or 170 (compare the
Passacaglia) are planned as a marked contrast to the descant-like harmo-
nization of A in b. 100, which is a newer kind of organ music altogether.
The three themes share a little three-semiquaver motif: in b. 1, this is part
of a classical French ouverture figure; in b. 32, a galant Italian echo; in b. 71,
a typical German organ-fugue.
Although the movement is more continuous in texture and rhythm than
a true ouverture, the minor-key development of A does produce some obvi-
ously French progressions. See Example 65. Particularly French are the slurs
Example 65
in A, and the echoes and the turns to the minor in B. Echoes were famil-
iar to Bach from e.g. Kuhnau’s suites (Clavierübung 1689) and the Premier
Livre of Boyvin (1690) or Du Mage (1708), and were explored in the very
last piece he had published, the ‘Echo’ closing Clavierübung II. Yet because
it is neo-galant, one can view section B as Italian, like the ritornello
structure itself. Since theme C is close to traditional German organ-fugues,
one cannot fairly claim that the E Prelude is free of North German
elements (as Krummacher 1985 p. 129 suggests). Perhaps the very four-bar
phraseology is ‘German’, like Clavierübung III’s chorales in French or Italian
idioms later on. Part of any such ‘national agenda’ would be to add
articulation signs to the French and Italian themes (slurs, dots) but not
to a traditional German fugue-subject, which is a kind of music never
given slurs or dots.
Changing manual and/or stops is certainly feasible but, not being speci-
fied in even this carefully prepared publication, no more than optional. For
short piano echoes, stops can be pushed in, or even played up an octave,
according to Niedt 1721 p. 57. But the echoes have nothing to do with the
Prelude’s ritornello form or any manual-changes made for its sake, and a
case can be made for using three manuals:
section A: manual I (lh first in b. 51)

sections B and C: manual II
the short echoes: manual III (as implied in Du Mage, i.e. an Echo to the
Positif )
137 BWV 552
Why Clavierübung II carefully specifies manual-changes when Clavierübung

III does not is a puzzle: because German harpsichordists were only then
becoming familiar with two manuals and needed advice about using them,
while organists had long used them in alternation and did not?
Fugue
The Fugue continues to explore styles, now in part more antique. The old
idea that its three sections ‘represent’ the Persons of the Trinity is supported
by the three flats, the time-signature, the numbers of subject-entries (mul-
tiples of three) and the number of bars in the sections (all are multiples of
nine or 3 × 3 : 36 : 45 : 36). But sectional fugues using variants and/or com-
binations of a subject had long been admired, and Bach makes no attempt
to combine all three subjects, which would not be impossible if the aim were
to present Three-In-One – as music is peculiarly fitted to do. Furthermore,
over bb. 115–16 the second subject could be introduced but is not.
Yet there is an uncanny structure behind the Fugue: the number of bars
36, 45, 36 makes 72 : 45 or 1.6 : 1 (Golden Section), while the middle section
itself is divided at its midpoint, i.e. a conspicuous moment (b. 59) at which
the first theme modified enters against the second theme disguised. This
produces two further Golden Sections, 36 : 22.5 and 22.5 : 36 (see Power
2001), none of which gives any impression that the music has been forced
into a straitjacket. But if this were deliberate, it would represent a calculated
control of material quite as much as the late canons do.
Themes taking two or three forms (one for each section) were typical
of canzona or capriccio fugues of a lighter nature, as in Frescobaldi’s
Fiori musicali. Ricercar subjects like BWV 552’s do not usually change
metre, although they may be combined with different countersubjects. Two
previous Leipzigers working with sectional ricercars, counter-themes and
triple-time variants were N. A. Strunk (one of 1683 has a similar theme)
and F. W. Zachow (Fantasia in D major), and it is possible that the E Fugue
was conceived as alluding to local, learned tradition. The subject itself is
generic, an unambiguous salute to venerable tradition.
Certain stile antico elements found in the work of contemporaries are
discussed below (see BWV 669), and clearly fugue subjects of the kind
shown in Example 66 share with BWV 552.ii such details as the ‘quiet’ 4/2
Example 66
138 BWV 552
character, the rising fourths, suspensions, narrow compass (a minor sixth)

and invertibility. It is not typical of the North German school both to vary
the subject and to combine it with others, as here:
A 4/2 subject A, five voices, twelve entries, 36 bars

B 6/4 subject B, four voices, then A + B modified, fifteen
entries, 45 bars
C 12/8 subject C, five voices, then C + A, 36 bars
Three subjects are combined in the fugues in F minor WTC2 and Art of
Fugue Nos. 8 and 11, and the E subjects being in some degree related to each
other need not have forbidden this (compare Art of Fugue No. 6). Rather, the
subjects are complementary in various ways, such as their intervals: fourths
are prominent in A, seconds and thirds in B, and fifths in C. Stretti are
modest, easily produced in bb. 21–3, 26–8 with parallel thirds and sixths.
The stile antico subject sings through the counterpoint, emerging from
it each time like a melody – compare the accompanying parts in bb. 91–2
(which include subject A) with bb. 97–9 (which do not). It provides in-
tervals for development (e.g. rising fourths in bb. 21–3) and quasi-entries
(e.g. b. 54); but what is less to be expected, if the fugue were simply a contra-
puntal demonstration, is the way that the second subject B has to be altered
to fit the first in bb. 59–60. Moreover, the third subject passes to the first
(b. 88) before the two are combined; then it fits twice to A’s once.
As if alluding to Fux’s Gradus ad Parnassum (1725), the first subject
allows countersubjects of various ‘species’: the first of 4/2 crotchets, the
second of 6/4 quavers (subject diminished and syncopated), the third of
12/8 quavers and semiquavers (subject augmented and syncopated). The
three tempi appear related:
4/2 crotchet = 6/4 crotchet, while 6/4 minim = 12/8 dotted crotchet.
At each juncture, the player is helped to grasp the tempo: irrespective of
rallentando, the left hand in b. 36 runs into the new fugue subject, while b. 81
has a hemiola and thus provides the next beat (so minim = dotted crotchet).
The variations of the main subject in the second and third sections are
unique, producing ‘a degree of rhythmic complexity probably unparalleled
in fugue of any period’ (Bullivant 1959 p. 652).
Some further points:
A. The subject is so familiar in outline that many similarities have been

found, in chorales (end of Cantata 144), vocal/choral movements (Handel,
Krieger), older canzone à la francese (de Macque) and contemporary fugues
(J. G. Walther). Some thirty examples are listed by Lohmann in EB 6588,
who also finds the theme adding up to forty-one, ‘J. S. Bach’ (a = 1,
139 BWV 552
b = 2, etc.). But in principle the subject of the E major Fugue WTC2

is more closely related to BWV 552.ii than any subject that has some or all
of the same notes, such as Buxtehude’s E major Praeludium.
The discovery that the subject is very close to that of a Fugue in D major
by C. F. Hurlebusch published in a volume being retailed in Leipzig from
1735 by J. S. Bach himself (Compositioni musicali, c. 1734: see Beisswenger
1992 pp. 360f.) shed new light on the context. This work has been claimed
to be so similar in subject and treatment to the first section of the E Fugue
that one can speak of it as ‘Bach’s source’ and a commonplace modu-
lation in it as ‘borrowed verbatim’ by Bach (Butler 1983 pp. 206f.). But
Hurlebusch’s three-voice working is thin, entirely conventional, and more
like other fugues of the 1730s, seen at their best in Handel’s Six Fugues,
published in 1735. Resemblances may be natural when composers wrote
fugues true to type. Yet there has to remain the possibility that Bach was
responding to Hurlebusch and intending to blind players by science.
There are closer similarities to the E major Fugue WTC2 than the type
of subject. They both have a countersubject of passing crotchets, which
are a source of effortless counterpoint (in this respect the Credo of the B
minor Mass is also close to BWV 552) and their bass lines are more thematic
than those of the first two stile antico chorales of Clavierübung III. While
the E major Fugue WTC2 is the ‘strictest and most compressed of Bach’s
instrumental fugues’ (Wolff 1968 p. 99), the E too is a clear example of one
particular type, the fuga grave. A stretto following the first full exposition
in both of them (Fugue in E b. 21, Fugue in E major b. 9) and the parallel
thirds and sixths encouraged by such counterpoint are similar. In the case
of BWV 552, so vocal are the lines of a fuga grave that the subject may be
‘heard’ over bb. 35–6, dispersed between the lines and not obvious on paper.
B. Keller’s idea that the second fugue subject is ‘contained’ (einbezogen)

in the first is shared by many a listener, though were the quavers an actual
paraphrase of the alla breve theme, this would be easier to recognize. Typical
details – beginning off the beat, quavers running in 6/4 – are found in earlier
pieces with thematic metamorphoses, such as Heidorn’s Fugue in G minor
(Mö MS).
Whether subject B was altered in b. 59 to fit A, or whether the composer,
having found a countersubject to A, thought it needed to be changed for its
own exposition (b. 37), can only be guessed. The blending quality of A is
clear from hints of it in bb. 44–6 (alto, bass) and 54–5 (soprano); also, the
inversus form of B (from b. 47) is altered both to fit it and to run into it.
Just as the hemiola in b. 81 heralds the new section, so that in b. 58 leads to
the combination of themes as it cuts the 6/4 fugue into exactly equal halves.
Important, too, is that the top note of the fugue (c ) occurs in each section
shortly before the next (bb. 32, 57, 77, 105).
140 BWV 552
A ‘theological’ investigation of the fugue, in particular whether each

theme pictures a Person of the Trinity and if so in what order, depends on
whether B can be heard as containing within itself both A and C, which
some writers have persuaded themselves is so (e.g. Chailley 1974 p. 264).
C. Theme C seems to refer to theme B (compare notes 5–8 of C with notes

8–11 of B), as in its falling fifths and rising fourths it also does to A. When
A does appear it is both syncopated and accompanied by running semi-
quavers; it is not put into triple time, as in earlier canzonas, but syncopated
in compound time, a much more unusual idea, perhaps unique. A and
C are first combined only in bb. 91–3 and then somewhat obscurely, while
bb. 87–91 (top part) and 92–6 (pedal) run them together as a new composite
theme, C-plus-A. This is another unusual idea.
There are other important elements: the sequence in the subject, the
climactic combination at b. 114, semiquaver groups resembling the sec-
ond subject (e.g. bb. 105–6, rh), others reminiscent of other mature works
(compare b. 91 with b. 16 of BWV 547.i) and the increasing continuity.
The references to the first theme are various: hidden and circumstantial
(e.g. inner parts in bb. 103–4), quasi-stretto (bb. 108–11), extended (pedal
b. 110), even quasi-ostinato (there are four powerful pedal entries). This
quasi-ostinato effect recalls not only the first section’s pedal entries but
gives the last entry a thundering finality exceeding even that of the C major
BWV 545. Even so, the Fugue by no means fully exploits thematic combi-
nation. Rather, it is as if one were constantly hearing the subject singing out
in fine voice, in one or other part, especially in the last twenty bars or so.
By tradition a 12/8 section is the last of a composite fugue, here also the
last piece of a major collection, springing from a stile antico subject but with
a distinctly stile moderno sense of climax, particularly in the final bars, the
grandest ending to any fugue in music. Rather than imagining the com-
poser under pressure to complete the work, and doing so quasi-extempore
(Breig 1999 p. 700), one might see the 12/8 section as yet another way to
complete a fugue, at times thin but with a ‘singing, massed choirs’ effect
that in e.g. bb. 109–10 prefers a rising sequence to the mere stretto that
b. 108 suggests. There is more thematic combination than one is first aware
of, and there could have been more, as when B could have been introduced
in the final bars (see above). Finally, however plausible the Golden Section
created by Prelude and Fugue, there is little exaggeration in seeing them
as summing up the various resources of organ praeludia as current, super-
seded or anticipated during the composer’s lifetime, assembling styles and
techniques known from Palestrina to Haydn.
Eight Short Preludes and Fugues BWV 553–560
Complete copy P 281; a lost source used for Peters VIII (1852).
Two staves; P 281 headed ‘VIII Praeludia èd VIII Fugen di. J. S. Bach. (?)’.
P 281 was once thought to be a copy by J. C. G. Bach (†1814), and may

have belonged to J. C. Kittel. Its paper is known from three sections of the
MS P 803, including one written by J. L. Krebs (Dürr 1987 p. 34). A copy
of No. 2 in P 508 was made by F. A. Grasnick (†1877), who had access to
manuscripts transmitted through various Bach pupils. The MS used for
Peters VIII, either based on P 281 or sharing its source (Emery 1952 p. 5),
had belonged to Forkel.
P 281’s many errors make it unlikely to be a copy made by the composer,
whoever he was, and who deftly handles many styles: toccatas (No. 5), Italian
concertos (No. 1), neo-galant effects (No. 4), old durezze techniques (No. 3),
and ‘southern’ fugal styles (Nos. 1, 4, 6, 7). Errors like parallels in Preludes
No. 5 and 8 could reflect an unclear original. Some of these suggest a much
later date than the early non-thematic pedal fugal entry in No. 6. Though
frequently charming and melodious, they could hardly have been written
by J. S. Bach for his pupils since their ‘standard of counterpoint and general
musicianship’ does not fit the period in question, nor does the scarcity of
copies suggest they were much used (Emery 1952 p. 31), even as part of a
bigger compendium. Nevertheless, the pieces do amount to a fine book for
learners, teaching whether or how to add pedal, use a second manual, and
register according to so-called key characteristics (Vogel 1998).
Various details suggest various possible composers. Thus the compass –
to c in pedal, only to a in manual – is typical of J. L. Krebs, but nothing
here is very like known music of either J. T. or J. L. Krebs (Tittel 1966 p. 123).
BWV 560 in particular is said to show eccentricities typical of W. F. Bach
(Beechey MT 1973 p. 831), and there are many details rare or unknown
in his father’s music: differences between subject and answer; the incom-
plete second answer in No. 3 (Souchay 1927 p. 4); the many descending
SATB expositions. A tendency towards proportions between sections – 2 : 1
(No. 1), 1 : 1 (No. 2), 2 : 1 (No. 4), and 1 : 3 (No. 7) – implies a thoughtful com-
poser, and resemblances to certain music of F. A. Maichelbeck (Augsburg
1738) and J. C. Simon (Augsburg c. 1750) have been noticed.
Although ‘there seems no reason why they should not have been written
[141] about 1730–50 by some minor composer in central Germany, whether or no
142 BWV 553–555
he was a pupil of Bach’s’ (Emery 1952 p. 42), the eminence grise is more likely
to be a southern composer such as J. K. F. Fischer. Such modest and single-
minded preludes, modest fugues with exposition, episode and final entries, a
charming and coherent handling of the keys and cadences: these are closer to
Fischer’s idiom than to any northern repertories, and could reflect his wide
and lasting influence on organists of the time. Even in the longest Fugue,
No. 3, there is little modulation beyond what one finds in Fischer’s succinct
little essays, and any ‘updating’ of his idiom discerned in BWV 553–560 –
binary form, post-Vivaldian patterns, post-Bach melodies, further episodes
in some fugues, sometimes unclear handling of part-writing – could be that
of an admirer of his in 1750 or so.

Dietrich’s idea (1931) that the binary prelude resembles a Corelli allemande
has been adequately discounted (Emery 1952 p. 24), but its composer knew
Italian concertos, directly or indirectly, original or transcribed, as well as
traditional organ praeludia. The Fugue’s coupling of two basic motifs is
reminiscent of Fischer or Pachelbel, compact but more than a mere fughetta.
BWV 554 Prelude and Fugue in D minor

Such a miniature ABA shape as the Prelude’s, in which A is merely a frame-
work for a concertante middle section, would be unique in the organ works
of J. S. Bach, irrespective of harmony or melody. The Fugue’s closing bars
not only resemble the Prelude’s but both resemble the first and last lines of
the melody ‘Jesu, meine Freude’ – allusion of a kind unknown in J. S. Bach’s
free organ works. But J. L. Krebs published a praeambulum to two settings
of the same chorale in his Clavierübung of c. 1750/6, and the chorale melody
itself has an ABA framework.
BWV 555 Prelude and Fugue in E minor

The durezza style of the Prelude, though unmistakable, is not pronounced
and derives from organ versets of southern composers rather than string
trio sonatas. Sometimes the idiom also resembles passages in J. S. Bach,
e.g. bb. 12ff. recall the D major Prelude BWV 532, the Neapolitan 6th of
b. 23 that in BWV 535.ii, b. 72. The Fugue is stricter, the best-wrought of
the set, perhaps, with stretto, inversus, and a counterpoint typical of earlier
treatments of the descending chromatic fourth.
143 BWV 556–559
BWV 556 Prelude and Fugue in F major

Despite its patterns, the Prelude is hard to imagine being the work of the com-
poser of the faintly similar BWV 590.iii (the Pastorella’s third movement):
it looks like an exercise in simple rising sequences, with a basso-continuo
pedal part, the kind of italianate music produced by Soler’s generation rather
than D. Scarlatti’s. The Fugue’s motifs could be found in many northern
and southern fugues, including Magnificat versets of Pachelbel. Several bars
are much like those of vocal fugues.
BWV 557 Prelude and Fugue in G major

As a ‘miniature toccata’ (Frotscher 1935 p. 878), such a Prelude could be
improvised on the patterns demonstrated in Niedt–Mattheson 1721 or in
Kuhnau’s first suite (1689), especially by an organist acquainted with BWV
902 (Prelude in G major) or BWV 535a or the melodious cadences of a
Fischer. The Fugue’s syncopated subject has a potential for stretto more in
style with WTC, each entry leading to or following a neat modulation.
BWV 558 Prelude and Fugue in G minor

Only on paper could evidence be found for regarding the Prelude as an
‘Italian courante’ (Dietrich 1931); neither the form nor the figuration is
typical. The Fugue subject again supplies three distinct ideas, any one of
which can be found in other contexts, particularly canzona and ricercar sub-
jects. Modulation is neatly managed (Spitta admired bb. 68ff. in particular),
and perhaps the imaginative penultimate bar was inspired by J. S. Bach?

The Prelude’s demisemiquaver figures suggest the manual-play of a southern
toccata even though particular figures (e.g. b. 2) will be found in Buxtehude.
Other features again suggest certain organ traditions – compare the pedal
of bb. 12–15 with the close of the first section of the A minor Praeludium
BWV 543 (b. 24). The Fugue subject’s second half follows the ornate outline
of other A minor subjects (BWV 543 and 944) but is in no sense a sketch of
either, despite suggestions made by earlier commentators (Oppel 1906). It
is more like verset-fughettas in J. K. F. Fischer’s Blumenstrauss, such as the
F major No. 2.
144 BWV 560
BWV 560 Prelude and Fugue in B major

The Prelude’s keyboard style reflects the newer oboe concertos of the 1730s,
though specific elements are identical with those elsewhere in the Eight:
compare bb. 21–2 with bb. 16–17 of BWV 555. The varying texture com-
plements that of other preludes in the set. The Fugue subject is not likely
to have been written before c. 1740, and only then perhaps by someone
familiar with Handel’s Concerti Grossi.
Miscellaneous pieces BWV 561–591
BWV 561 Fantasia and Fugue in A minor

Later eighteenth- or nineteenth-century copies only (P 318, P 1066), and
Peters IX.
Two staves; headed in P 318 ‘Fantasia’, and by a later hand, ‘in A moll
(Preludio e Fuga per il Cembalo) compost: da Giovanne Sebast: Bach’.
One view is that this is an early work ‘composed for pedal harpsichord’
(‘Pedalflügel’: BG 38 p. xxii), like the A major fugues BWV 949 and 950.
Another is that whoever the composer was, he knew the Prelude and Fugue in
A minor BWV 543 (Keller 1937); perhaps it was Kittel, of whom the changes
of movement from semiquavers to demisemiquavers may be typical (Keller
1948 p. 57). Why the work could also be accredited to W. F. Bach (Frotscher
1935 p. 856) is unclear.
The figures of bb. 1 and 29 can be found (in the same key) in Buxtehude’s
D minor Toccata, and others suggest a familiarity with durezza conventions.
Details reminiscent of BWV 543 include such figures as the broken chords
above tonic pedal, the harmony at bb. 82–3 and the fugue-subject itself,
which is the most Bach-like thing in the whole. Like BWV 543, it consists of
an opening phrase followed by a sequence, a type known elsewhere amongst
contemporaries (e.g. Böhm’s C major Praeludium and BWV 948) or pupils
(J. P. Kellner’s Fugue Anh.III 180). A ‘style relationship’ with the Concerto
BWV 594 has also been heard (EB 6583 p. xiii).
The pedal points of BWV 551, 561, 949 and 950, and in some other early
or questionable works, are problematic. Were they meant to be adaptable
for organ or harpsichord, where the effect is ‘pale’ (according to Bartels
2001)? Only optionally held? Are pedals more than optional? Pulldowns or
independent? Could the notes merely be touched now and then, as in long
bass notes of a recitative? Or was there a convention for pedal points in A
major/minor, however practical (see A minor Fugue WTC1)? The last seems
to be the case, however the other questions are answered.
BWV 562 Fantasia and Fugue in C minor

Autograph MS P 490 (including Fugue fragment, see below); derived copies
[145] of Fantasia in P 286 (J. P. Kellner, 1727/40 – Stinson 1989 p. 24), P 533
146 BWV 562
(J. F. Agricola), Lpz MB MS 1 (J. A. G. Wechmar), and via C. P. E. Bach (e.g.

P 290) or J. C. Kittel (e.g. P 320); copy with Fugue BWV 546 (P 1104, owned
by J. C. Oley).
Two staves; headed in P 490 ‘Fantasia pro Organo. a. 5 Vocum, cum pedali
obligato’ (last phrase added later?).
All these sources but P 1104 are based directly or indirectly on P 490, which
begins as a fair copy presumably based on an earlier autograph (see KB
p. 28). A possible history of the work is as follows: (i) an ‘older’ version of
the Fantasia, with simpler close and without the penultimate bar of stage
(ii), which is a ‘newer’ version made in P 490 before c. 1738 and still being
amended in 1743/45 (? – see Kobayashi 1988 p. 59); (iii) a presumably new
Fugue added or begun, perhaps as late as August 1748 (Kobayashi ibid.). In
P 1104, the Fantasia is followed by the Fugue BWV 546.ii, an early pairing
(KB p. 336), with ‘early’ features: loose episodes in the Fugue, French idioms
in the Fantasia. But in the sources of BWV 546 itself, nothing suggests that
its prelude was paired with any other fugue (Kilian 1962).
Differences between the Fantasia’s final bars in P 1104 and P 490 suggest
a careful revision made during the 1740s: compare Example 67 with NBA
IV/5 p. 56. The later version’s reference to the opening theme at the end is
a ‘mature’ sign. As for the Fugue: in P 490 it takes the last of the four sides
of the MS, followed by directs to the next page, showing that the fugue was
either continued (KB p. 27) or planned. Not all incomplete works have a
full texture up to the breakoff point.
Example 67
Fantasia
While in its bleak C minor pedal-points the Fantasia resembles the C minor
Prelude BWV 537, its preoccupation with a single theme is unusual, more
147 BWV 562
so than in the Toccata BWV 538, whose theme-types are more convention-
ally German. Six pedal points are separated by bass entries. It is not quite
true that ‘the whole work is developed from a single theme’ (Keller 1948
p. 98), since the first twelve bars alone develop two ideas. There are various
countersubjects as well as stretto and doubling in sixths, and the motif is
heard against different harmonies as the piece proceeds, including cadence
(b. 37), sequence (b. 60) and episode (b. 68). New themes include the pedal
crotchets of bb. 57ff., and the whole becomes an idiosyncratic, contrapun-
tal tour de force, with that peculiar melancholy one often hears in French
baroque music.
It is no argument against the work’s Frenchness that this lies more in
appearance than in essence (a ‘rather superficial relationship’: KB p. 334),
for the opening motif is close to several melodies in Grigny’s Livre d’Orgue
known to Bach, of which the Gloria fugue for the petit plein jeu is typical
(Example 68). This style will include rising appoggiaturas and a five-part
texture spaced two parts rh, two parts lh, one part pedal. In this last re-
spect, Grigny’s Fugue à 5 is closer to it than the Gloria of Example 68: see
Example 69. But note that although Bach may have been ‘establishing a
Example 68
Example 69
French fabric so faithfully at the outset’ (Horn 1986 p. 263), it is not slav-
ishly observed. His monothematicism is rigorous, he does not invert the
subject (unlike the composer of Example 68), and his five parts cannot be
divided throughout between the hands quite as Grigny specified, i.e. each
148 BWV 562
hand on its own manual. (Nor can the Fugue’s: see b. 15. Paired manuals
in the chorales BWV 619 and 633/634 are clearer, since two of the parts are
canonic and the pieces are much shorter.)
Although Grigny is usually associated with this piece, there was some-
thing of a French tradition for a type of fugue in almost every bar of which a
short and decorated subject is carefully worked. Another example is a fugue
in Clérambault’s Livre d’Orgue (1710), a book dedicated to André Raison
and just possibly known to Bach. Had Grigny been the inspiration for such a
pedal-piece as BWV 562, one might expect its composer to have used three
staves or ended with an imperfect cadence (Clérambault’s has two staves
and a perfect cadence).
While the Fantasia’s key-plan recalls the South German toccata (pedal
points with fugal imitation above), its short, constantly reworked phrases
bring it within the French mode. Rising appoggiaturas are also charac-
teristic – not mere melodic ornaments but radical harmonic devices,
producing rich seconds, sevenths and ninths. Perhaps it was the appog-
giatura harmonies that attracted a later Leipziger, himself versed in such
techniques, to publish it in 1841 (Schumann in NZfM, Supplement to
No. 13).
Fugue
The Fantasia’s miscellaneous counterpoint is matched by the strict Fugue,
also in five parts, as the heading says. The subject and its hemiola would
not be out of place in a Livre d’Orgue, though any resemblance between
it and the Passacaglia’s French theme (see p. 183 below) upside-down is
superficial. The texture promises to be full, and one can easily believe such
bars as 13–18 to be contemporary with the chorale BWV 678. That a stretto
is already worked in b. 22 (i.e. after the first cadence) has suggested to some
that the composer had intended to proceed to a double fugue, with a new
subject (Keller 1948 p. 98); perhaps too the theme would have been inverted
later and a new section begun, as in BWV 547. Or, since the F minor from
b. 25 suggests a return to the tonic, perhaps the plan was to write another
da capo fugue like BWV 548, with a B section exploring various major keys
(Overholtzer 2001).
It is not the subject that is of greatest interest in these twenty-seven
bars but the quaver motif dominating the first section, producing a free
upper part of perhaps little conviction (bb. 10–11) but in theory open to
development of the kind seen in BWV 678, had there been a B section to
need it. Nevertheless, both theme and subsidiary motifs are short for a fully
developed five-part fugue; there is as yet no broad sweep, and one wonders
if it was ever taken very much farther.
149 BWV 563
BWV 563 Fantasia in B minor (‘Fantasia and Imitatio’)

No Autograph MS; copies in Lpz MB III.8.4 (J. C. Bach, ABB) from which
P 804 (partly by J. P. Kellner?) might derive, later MSS more certainly.
Two staves; headed ‘Fantasia’, the second section ‘Imitatio’ in ABB, which
may have been transcribing a tablature original (not autograph? – Hill 1990
p. 354).
Spitta thought the ‘light and minute character’ of the Fantasia did ‘not suit
the organ’ (I p. 432), while BG 38 included it in the organ works because of
its ‘organ-like nature’, the pedal necessary in bb. 15 and 20, and the crossed
parts at b. 129 of the Imitatio. Against this, the sources do not indicate
pedals; big pedal points do not always indicate organ (cf. A minor Fugue
WTC1); this Fantasia is no more ‘organ-like’ than that in A minor BWV
904 (also in P 804); and the Imitatio is neutral in style. Nevertheless, Bach’s
early method of composing-by-motifs, as here, can certainly be realized on
the organ as an instrument of instruction.
In principle a prelude and fugue, BWV 563 is unusually single-minded in
its exploitation of two kinds of motif: the little dactyl of the Fantasia (a ‘kind
of improvisation’ in the style of Pachelbel or Fischer – Breig 1999 p. 630)
and the stepwise 3/4 theme of the Imitatio. For these standard figurae, see
Example 70. The former produces a good – barely improvisable? – four-part
Example 70
texture with simple cadenza and pedal-points; the latter, a sectional fugue
with various derivative subjects, similar at several points to the Sonata in
D major BWV 963 or the C minor Fantasia BWV 1121. Although the full
subject of a fugue proper does not have to be heard complete after the
first section (cf. Three-part Invention in C minor BWV 788), the several
clearly related thematic groups of the Imitatio are more typical of the earlier
ricercar.
It is possible that the terms imitatio and fantasia were chosen (by whom?)
not least to enlarge the vocabulary used for titles in the ABB. Although
neither of the movements is doctrinaire in its use of motif, both are in
150 BWV 563–564
keeping with other pieces in the album that set out to exploit pedagogic
techniques, such as a chorale with canon. This Fantasia contrasts with the
next one found in the ABB, BWV 944 ‘pour le Clavessin’, while the previous
fantasia, BWV 570, is more like it in its dactyl motifs. Though these three
fantasias were copied by three different scribes, they amount to a survey of
the genre.
The Imitatio’s theme-type is also familiar from elsewhere, e.g. an
Offertoire in Grigny’s Livre d’Orgue and Sonata No. 3 from Kuhnau’s Frische
Clavier-Früchte of 1696, the latter surely known to J. S. Bach. Another similar
theme (also in rectus and inversus forms) is found in the ninth movement of
Cantata 21, and Georg Böhm has something like it in the chaconne of his F
minor harpsichord suite, found in the companion Mö MS. A similar theme
also appears as countersubject to the chorale ‘Christ lag in Todesbanden’
in the opening chorus of Handel’s Israel in Egypt – a sign, perhaps, that
he and Bach had been taught to work with similar material, in this case an
unassuming theme-type useful in many genres.
While some commentators doubt the work’s authenticity (Blume 1968)
or date it to early Arnstadt, its origin might be owed to an interest in stan-
dard note-patterns shared by Bach and Walther. Both movements have a
charming counterpoint, a genuine sense of melody and (as in the Fantasia’s
final pedal point) a striking grasp of harmony. The Imitatio handles tonal-
ity expertly: the final perfect cadence is fifteen bars from the end, the rest
a spacious coda referring to cadences already heard (bb. 46, 68, 98). Both
movements are as much models of three/four-part texture as certain bars
in the contemporary G minor Prelude BWV 535a are of five-part.
BWV 564 Toccata in C major

No Autograph MS; copies in P 803 (S. G. Heder c. 1719, based on lost auto-
graph?) and P 286 (partly J. P. Kellner, 1726/7?), others from an unknown
common source, including P 1101, P 1102 (fugue only), P 1103 (no middle
movement), and Brussels II.4093, all eighteenth century.
Two staves, headed ‘Toccata ped: ex C’ in P 803 and ‘Toccata ex C pedaliter’

in P 286, both heading the movements ‘Adagio’, ‘Grave’, ‘Fuga’.
The three-movement form was known to copyists who give no sign that
the Fugue is an earlier work, despite the fact that in bb. 84–5 it seems to
avoid manual d found in the Toccata (Emery 1966). Nor is the Adagio
known to be an addition, despite its absence in P 1103 (see Kobayashi 1973
p. 235. J. L. Krebs, imitating BWV 564 in his Prelude and Fugue in C, did
151 BWV 564
not keep the three-movement plan). To Spitta, the plan of quick–slow–quick

suggested an Italian concerto model (I p. 415), but like the Fantasia in G, it
could rather be seen as an updated multisectional praeludium. As happened
over time with concertos, sonatas and cantatas, traditional sections are now
crystallized into fully fledged movements, each in this instance strikingly
original.
Short phrases, rests, gaps and little repetitions characterize all the move-
ments except the Grave section of the middle movement, and each could be
aiming to use two manuals in its own way:
Toccata first for echoes in the opening solos, then for alternation in a
‘ritornello duologue’
Adagio for solo plus accompaniment (a melody over a realized
continuo)
Fugue for contrast (entries versus episodes)
Nowhere are two manuals obligatory, not even (surprisingly) for the Adagio,
and no sources suggest it. But the opportunities are clear: rests or phrasing
allow echoes in both opening manual and pedal solos, and manual-changes
in the ritornello; a solo line in the Adagio (played on Principal 8 ?) merges
into block harmonies at the Grave; and the Fugue’s episodes are clearcut.
Such variety might justify the guess that BWV 564 was composed for testing
an organ.
First movement
This seems to be a deliberate enlargement of an old prelude-type: manual
passaggio + pedal solo + motivic-contrapuntal section. The result is a join-
ing of toccata and quasi-concerto, its sections more distinct than in BWV
540. The join over the tonic of bb. 31–3 is logical and natural. The early
harpsichord Toccata in G BWV 916 is an essay in similar form, the organ
Prelude in G BWV 541 a later ‘tightening-up’ of it. In BWV 564 and 916
there are five statements (BWV 564: bb. 32 C, 38 G, 50 A minor, 61 E minor,
76 C), producing a short-breathed dialogue in a ritornello form distinct
from, and probably independent of, Vivaldi’s.
A manual and pedal solo introduction (the longest known in the

literature)
B a concerto-like dialogue
Example 71 suggests how traditional are the opening one-bar gestures, here
from the Reinken sonata transcriptions BWV 965.ii (see Toccata b. 33) and
BWV 966.iv (see Toccata b. 32). There is a touch of J. H. Buttstedt about the
152 BWV 564
Example 71
opening gesture, which is more arresting than one finds even in Buttstedt
praeludia, however. A rhetorical rest following a return to the tonic (bb. 2,
8, 10, 12) is conventional – see Lübeck’s C minor Praeambulum – as are
the three pedal Cs and their hint of Orgelpunkttoccata. Also typical are the
pedal’s opening motifs and its systematic phrase-structure, though not the
quasi-echoes and the array of motifs (triplets, dactyls, trills). The manual
demisemiquaver scales are in-turning, smooth, with potential echoes; the
pedal semiquavers are broken chords, varied, disjunct, with potential echoes
(bb. 14, 16, 17?, 18, 21–3, 28, and 30–1).
In modulating, the pedal solo enlarges on that in BWV 549a. The slurs
may well belong to the composer and are rare even in continuo bass-lines like
those of the Six Sonatas: do they indicate the use of heel for the demisemi-
quavers (right foot)?
Section B is marked less by ritornello episodes (bb. 55, 67) than by a
dialogue between two ideas, each of which could have its own manual:
see Example 72. Both are anticipated in the pedal’s solo (Spitta I p. 416),
Example 72
although Keller hears in the first the ‘energetic bowing’ of two violins (1948
p. 77), indeed as in Reinken’s string sonata in Example 71. The harpsichord
Toccata BWV 916 too has a ritornello movement based on short phrases
153 BWV 564
(and constantly moving to cadences in a similar way), of which the first

is scale-like, the second broken chords, as in Example 72. (ABB’s copy of
BWV 916 likewise does not specify two manuals, nor does Krebs for the
echoes in his C major Prelude and Fugue.)
The work’s ‘general cheerfulness’ and ‘less church-like’ mood need not
be reflecting the influence of Italian concertos (Hoffmann-Erbrecht 1972),
since Böhm’s C major Praeludium is equally cheerful. Nor need ritornello
elements be owed to concertos, since the returns here of complete material
are not characteristic of them, and there is no Vivaldian final reprise (Klein
1970 p. 26). The duologuing phrases, predominantly of six bars each, become
foreshortened towards the end, as can be clearly seen in the pedal part.
Passages such as bb. 67–70 are an original and charming slant on North
German praeludia, as is the turn to the minor before the final cadence –
compare the end of the first section of Böhm’s C major Praeludium.
Second movement
The Adagio is a short-breathed melody above a continuo (Schneider 1914)
realized simply in both harmony and rhythm. It has been compared with
Torelli’s Concerto in C major Op. 8 No. 1 (Zehnder 1991 p. 47), and one is
bound to wonder whether it originated as a movement for oboe solo, with
b. 13 up an octave.
While short phrases are characteristic of early Bach (e.g. Cantata 196,
c. 1708), more italianate are the quasi-pizzicato pedal, the Neapolitan sixths
and the petite reprise of bb. 20–1. Five Neapolitan sixths in one movement
is unusual, though there are more in the (earlier?) trio BWV 528.ii. Perhaps it
represents a new kind of organ music, one independent of Italian concertos
and created, like the Reinken arrangement BWV 965, in a spirit of invention.
The movement has no clear parallels even amongst the chorale preludes,
although short stretches of melody-plus-accompaniment by Bruhns and
others could have suggested the idea.
The Grave is equally distinct in idiom and like the Adagio of the D
major Prelude has its own kind of strained harmonies: diminished sevenths
suspended over the next chord (Example 73). These appear at least four
Example 73
times, adding French augmented fifths (as in Example 73) to typical chro-
matic durezze. This Grave, in its recitative link, thick chords, new harmonies
154 BWV 564
and ‘forbidden’ bass intervals, updates a passage in Buxtehude’s Praeludium

BuxWV 142, itself a development of links in the capriccios of Frescobaldi’s
Fiori musicali. Perhaps it puzzled the copyists, and the pedal should rise
a further diminished fourth two bars from the end, exchanging the usual
division between lh and pedal?
Third movement
Striking features are the length, the unique levity of theme, a countersubject
that dialogues with the subject (as in the D major Fugue), a long working-
out (middle entries answered at length), modest episodes, and an apparently
subdued close.
1–37 four-part exposition; countersubject typical of permutation

fugues
37–43 episode, pedal and manual motifs derived
43–123 middle entries, dominant (43), tonic (53), dominant (63),
episode (as before, parts exchanged), mediant (78) plus
answer (part stretto), episode, dominant of dominant (100),
long episode
123–32 final entry
132–41 coda (the longest episode), founded on various brisé figures
Much of the detail is unusual, including the demands made on the player
by quite conventional note-patterns. The rests and the dotted-note cadence
can both be found in Buxtehude, the length and figuration in Reinken and
Buttstedt, a similar motoric drive in BWV 532, and three-phrase subjects in
BWV 533 and 575. But nothing in these works approaches BWV 564. Entries
as far as the dominant of the mediant suggest a maturing stage in fugue-
writing, though whether the relative minor itself is ‘renounced’ because of
the middle movement (Breig 1993 p. 53) seems doubtful.
The block chords of simple counterpoint are typical of early fugues
and are part of the fun, as is the obsessive way the motif of Example 74 is
sometimes treated (b. 78). The episodes, which often include broken figures
Example 74
typical of harpsichord toccatas, are too brief for this to be considered a fully
worked-out ritornello-fugue. Like the subject, broken figures (as in b. 27)
return rondo-like throughout, as does a cadence-phrase much like one in
the early Cantatas 131, 71 and 4.
155 BWV 564–565
The final tonic pedal point is held, unlike the first movement’s which is
detached – a deliberate contrast? P 286 holds it through to the final chord,
which lasts a whole bar (KB p. 691). P 803’s short final chord suggests a strong
rallentando, as do all such short finals including the C major Fugue’s, BWV
547. How the last bar originally read (in tablature?) is not clear: perhaps
the apparently brusque and unassuming close alludes to North German
convention (cf. Buxtehude’s G minor Praeludium BuxWV 163), as does the
F slipped into the closing bars.
BWV 565 Toccata and Fugue in D minor

No Autograph MS; all known copies directly or indirectly from P 595
(J. Ringk 1717–78), which now also contains BWV 532.ii, 541.i and 551.
Two staves; heading in P 595, ‘Toccata Con Fuga: pedaliter ex d [sic] di

J. S: Bach: Scrips: Johannes Ringk’. For tempo indications, see below.
Ringk was a pupil of J. P. Kellner and, in a similar hand, copied keyboard

music by Böhm, Buttstedt, Buxtehude, Werckmeister, Pachelbel, Bruhns
and Handel, as well as the Wedding Cantata BWV 202. His attributions
are usually reliable, though P 595 contains important errors (KB p. 521).
Teacher and pupil seem not to overlap much in what of Bach they copied
(KB p. 203), implying collaboration between them. Typical of Ringk’s
calligraphy are the fermatas in the opening bars, whether intended for the
notes (NBA) or rests (BG) or as signa congruentiae to mark off the phrases.
Unlikely for non-Italian music copied before c. 1740, if then, are so many
tempo or section indications (ten in P 595) and staccato dots in bb. 12ff.
and 30f.
Being unique, the work is a puzzle:
Overall form
While the prelude–fugue–postlude is familiar from BWV 549a or 535a, the
cadenza-like writing of BWV 565’s three sections is more like that of the
interludes in a five-section praeludium. The pedal line of the Toccata keeps
to the familiar tonic–dominant–tonic framework, but about the Fugue there
remain many doubts because it is so simple in all respects (Bullivant 1959)
and exceptional in its subdominant answers, especially a unique flattened-
leading-note minor one (b. 86).
Detail of style
Spitta saw ‘traces of the northern schools in the detail’ (I p. 402), but
the ‘stretches of recitative’ and ‘fleeting, rolling passage-work’ are unique.
156 BWV 565
Parallels can be made with those praeludia of Böhm that have unique fea-
tures as if the genre itself was meant to produce many simple surprises
(G minor), often with flourishes (C major, A minor). Whether Buttstedt’s
wild idiom inspired the piece or was merely typical of the time and place –
would he write a C minor entry in a D minor fugue? – BWV 565 is un-
usually tuneful for a work of such free fantasy. Though in theory BWV 565
is comparable to early works such as BWV 531, 549a and even 578 (Claus
1995), such details as the opening octaves, spread chords, triadic harmony,
thirds, sixths, and solo pedal bear the hallmarks of the newer, simpler idioms
post-1730 or even post-1750.
Simplicity
Three simple diminished sevenths in the first twenty-seven bars produce a
patent rhetoric unknown in written-down organ music. Diminished sev-
enths in the G minor Prelude BWV 535 are not static in the same way, though
in both pieces the pedal picks up its last previous note (BWV 565 bb. 22/27
and BWV 535 bb. 14/32). If the falling line of bb. 16–20 is an old idea
(DCBA – cf. BuxWV 155 bb. 6–10), its repetition and simplicity are not.
Similar points could be made about the triplet sixths (cf. BuxWV 149) and
the decorated dominant seventh of the pedal solo before the cadence. Also,
a fugue without detailed imitative counterpoint, as here, is over-simple.
Patently rhetorical are several musical figures in the first thirty bars and
a whole catalogue of effects in the last seventeen (alternating hands, sus-
tained chords, pedal solo, change of tonal direction b. 133, simple chords
newly scored b. 137, a severely plain close). All of them are undeniably
effective.
Unusual organ textures

Though an isolated opening mordent is conventional, the octaves are
unknown in any toccata of Bach or any other composer. (But three transcrip-
tions in D minor – Harpsichord Concerto BWV 1052, its cantata version
BWV 146.i, and the Triple Concerto BWV 1063 – have such open octaves.)
Other unusual details are: the spread or built-up diminished seventh, the
characteristic rhythm of bb. 3ff. (the semiquaver pairs égal or inégal?), the
violinistic passage from b. 12 exploiting the open A string, the fourfold
phrase in bb. 16–20 interrupted by a scale, and the long broken diminished
seventh of bb. 22–7.
An amalgam of different idioms

The violinistic fugue-subject is also familiar in organ music: see Example 75.
The first of these is at the same pitch as BWV 565 in its arrangement for
organ ( = BWV 539 b. 66). Both the C major Fugue in CbWFB (BWV 953)
and the G major Prelude BWV 541 b. 19 have similar figuration, as do other
157 BWV 565
Example 75
works in G major such as the Prelude in WTC2. Even the unique pedal solo
entry recalls freer sections of northern praeludia, e.g. Bruhns in G major
b. 27.
Questionable harmonic details

To close a work with a minor plagal cadence is so unusual as to suggest (i)
a date after c. 1750, (ii) a Picardy third was originally written or intended
(see the chorale BWV 1098 for a likelier cadence), (iii) there was originally
no third, as was not uncommon in solo string-music. If the fugue-subject
can be glimpsed in the notes of the opening toccata flourish (Krey 1956),
this would be the result of a limited harmonic vocabulary rather than subtle
allusion, as would any supposed resemblances found to the melody of ‘Wir
glauben’ (Gwinner 1968).
Similarities to Bach works

The Fugue’s subdominant answer suggests a knowledgeable composer (see
also BWV 539 and 531), as do the first codetta (b. 34, cf. the Passacaglia
Fugue), the various hints of simple permutable counterpoint, certain
textures and motifs (compare bb. 87–90 with b. 77 of the G minor fugue
BWV 542) and not least the implied echoes (cf. BWV 539 again). To follow
each subject entry by striking material (bb. 41, 54, 62, 74, 90, 95, 111, 122),
and thus produce a sense of drive, certainly implies a skilled musician.
The final tonic entries (bb. 109, 124) anticipate those ritornellos of Bach
in which the main theme has a ‘false’ final appearance (e.g. the D minor
Harpsichord Concerto BWV 1052.i), as does the dramatic break-off in
b. 127 (e.g. C minor Harpsichord Toccata BWV 911).
Possible answers to these conundrums are:
Transcription?
A solo violin ‘original’, such as the final cadence suggests, could have been
in A minor up a fifth, the transposing made easier by Ringk’s soprano clef
had he been the one to do it (see Stinson 1990 p. 122). That there are ‘no
preserved North German violin works’ of this kind (Billeter 1997 p. 79) may
158 BWV 565
not be relevant, since one could as well argue that there are, in the form of
such transcriptions. Though many chords and figurations would have had to
be very different (e.g. bb. 86ff.), the harmonic spectrum is simple enough for
one to hypothesize on a string ‘original’: for some suggestions, see Williams
1981. For example, the violin’s final chord would be an open-string fifth,
without any third.
An alternative transcription
An ‘original’ for violoncello piccolo or five-string cello would be an octave
lower than for violin but otherwise much the same. The repertory for this
instrument being so small and ephemeral could explain why the work is not
known in this form. But the repertory certainly existed in manuscript form
at one time: in 1762, Breitkopf advertised thirteen little volumes of music
for the cello piccolo, none published and now all lost (see BJ 1998 p. 76).
One cannot know whether any such music took a form approximating the
present work, therefore. Furthermore, if its first string is tuned to d not e –
scordatura of the kind known in the Cello Suite BWV 1011 – idiomatic and
convenient figuration and chords would result (see Argent 2000).
For a harpsichord toccata to have been the original, with sections and
gestures familiar in Bach’s toccatas (cf. Billeter 1997), the arranger would
have had to add the octaves, the fermatas and the tempo-signs, and so much
interference is unusual.
BWV 565 merely imitates string music

The echo phrases in such violin music as BWV 1003 could always have been
imitated by the organ, as is also the case with many a ‘violinistic’ element in
this Toccata. The opening mordent itself – top e in a violin version – recalls
the opening of the E major Violin Partita. As the transcriber of the D minor
Fugue BWV 539 realized, an organ imitation of simple violin textures has
to be filled out with thirds and sixths: compare its bb. 13–14 with the Violin
Sonata BWV 1001.ii, bb. 12–13.
Of course, many works of J. S. Bach are unlike anything else, whether
or not they are imitating other genres. A work like the Sonata in D minor
BWV 964 would have seemed a perfectly self-contained, idiomatic work
for harpsichord – another lone masterpiece – had the solo violin sonatas
not survived. As it is, however, a comparison between it and the Violin
Sonata BWV 1003 offers many detailed parallels with the present work.
One assumption already is often made about the Sonata BWV 964: that
J. S. Bach transcribed it himself.
Bach was neither composer nor arranger

Perhaps Kellner inspired or acquired or even composed the ‘original’
(Humphreys 1982), for his circle was clearly interested in transcriptions – see
159 BWV 565–566
BWV 1039, below. Or perhaps an organist like Ringk, known for his fugal
improvisations and performance of Bach works (see Stinson 1990 p. 33),
could produce such a work himself and then ascribe it to a composer ad-
mired by the Berlin cognoscenti around him. Its ‘old’ features need not
mean that it was altogether an early work, as still so often claimed (Wolff
2000 pp. 72, 460), only that organists of Ringk’s generation were immersed
in earlier organ music and knew its more approachable characteristics –
could in fact fake them, even to deriving most of the themes from much
the same notes (a scale of D minor, up and down). The very simplicity of
so much harmonization in 3rds or 6ths argues for Friedemann’s generation
rather than his father’s, someone well read in keyboard styles as far afield as
‘Les Timbres’ in Couperin’s Troisième Livre, 1722.
BWV 566 ‘Toccata and Fugue in E major’

‘Toccata and Fugue in C major’
No Autograph MS; copies in C major in P 803 (J. T. Krebs), P 286 (J. P.
Kellner), P 203 (C. F. G. Schwenke, via C. P. E. Bach?), and via Kirnberger
(Am.B.59); copies in E major also via Kirnberger (Am.B.544) and a lost
Kittel MS (from the autograph?). First two movements only, in Am.B.59
(C major) and the lost Kittel MS (E major).
Two staves in P 803 etc.; headed in P 803 ‘Praeludium con Fuga’.
Various later titles show copyists becoming less familiar with multi-
sectioned organ works: ‘Praeludium’ is no doubt the original. Commonly
assumed now is that the original key was E major (NBA IV/6) and that the
C major version was made, perhaps by J. S. Bach (Peters III), perhaps by
J. T. Krebs (KB p. 302), to avoid the pedal D and/or pedal notes higher
than c (Emery 1958 p. iv), or even to simplify the first pedal solo (Keller
1948 p. 59). Yet from Example 76 one could argue either way; and from
Example 77 that neither (nor even a hypothetical D major) is obviously the
Example 76
160 BWV 566
Example 77
original: the first avoids C, d and e ; the second, AA and BB; the third, BB
and C. Perhaps this is another case in which equally authoritative variants
or versions circulated, in different keys with different details, the C major
version (or others now missing) already at an early period?
Also unclear is the reason for transposing from E to C and not, as with
concertos BWV1042 and 1054, from E to D. For the E major praeludia
of Vincent Lübeck or Buxtehude to have been a model, the first must be
older and the second a greater influence than other praeludia, neither of
which is certain. A problem with the E major version being the original is
the harmonies of bb. 16–17, impossible in any unequal temperament and
unusual in J. S. Bach, early or late. The progressions themselves, enharmon-
ically notated, are not advanced (doubled leading notes!), but the passage
of keys requires D major and E major to be equally sweet-tuned.
This has long been seen as ‘the only essay of Bach in the motivically
extended fugue form . . . of Buxtehude’ (Spitta I p. 322), or rather of
Frescobaldi, with two fugues, the subject of the second a variation on the
first. Both are more fully worked out than putative models, and there is
no postlude such as in the harpsichord Toccatas BWV 911, 912 and 915.
Formally, it resembles the Toccata BWV 913, which has four main sections,
the first with a solo bass line, the last a ‘variation’ on the second. BWV 566’s
sections are more distinct than often with Buxtehude, though the third has
not yet developed into the separate slow movements of BWV 564 or 913.
Buxtehude’s G minor Praeludium in the ABB shares certain details (such
as a lh opening plus pedal point) with BWV 566, which could well be an
Arnstadt work.
In C major copies the opening passagio is beamed to show hand-
distribution, presumed by KB p. 532 to be not the composer’s. But it is
idiomatic, and something similar is needed for the third section. Both fugues
161 BWV 566
give opportunity for changing manuals, and solo-like lines in the second
observe the French distinction between en dessus (bb. 204–14) and en basse
(bb. 227–end), especially with a tierce-registration.
The incomplete copies might be reflecting the growing tendency to pair a
single prelude and fugue (Krummacher 1985). Nevertheless, multi-sectional
praeludia did not have to be played complete, and in Buxtehude’s circle,
Frescobaldi’s advice to end ad libitum, ‘as you like’, might still have been
followed as a matter of course. When complete, however, the work, like
other ‘northern’ praeludia, has more than a passing resemblance to a four-
movement sonata da chiesa.
First section
As in many a northern toccata, the section progresses freely from a single-
line opening to a full final cadence; and as in many a southern, there are
full suspensions in organo pleno style. The pedal solo seems rather clumsy
in detail and to have a non sequitur in b. 9: through the copyist or the
composer?
Assuming they are authentic, the thick harmonies (up to ten parts)
create new, rich effects not always with a clear sense of direction, but vari-
ously phrased. The tendency to extract motifs and transform a near-banal
sequence into a miniature ostinato over bb. 24–32 is more marked and imag-
inative than was usual. In the richly harmonized passages certain infelicities
may be due to copyists (e.g. last beat of b. 7), but the harmonies, with or
without a Neapolitan 6th (b. 14), are ably spun out.
Second section
Repeated notes are typical of works with varied fugue-subjects (Example 78).
Such ‘characteristic repercussion themes’ (Apel 1967 p. 598) come from
canzone, the latter half’s sequences from a different tradition: compare the
Example 78
162 BWV 566
D major Fugue. Sequences are typical, resulting in a certain similarity be-

tween bb. 80f. and the close of the Prelude. Four parts are carefully worked,
the harmony at times even anticipating the G major Fugue BWV 541 (com-
pare b. 81 with BWV 541 b. 14). Entries are on tonic and dominant only,
except for one in the relative (b. 107), and countersubjects are so consistent
as to make it seem at times a permutation fugue (bb. 73–6, 101–4).
Length is achieved by means not only of somewhat pedantic sequences
but an unadventurous tonality, aimlessly wandering in and out of the dom-
inant, and pulled by gravity to the tonic. Nevertheless, the four-part texture
makes demands on the player, and one can imagine all these desiderata –
well-sustained length, better key-plan, astute counterpoint, playing profi-
ciency – gestating before fruition in Weimar.
Third section
Though short, this section includes the most obvious allusions to toccata
traditions: scales beginning off the beat, runs pitted against pedal mo-
tifs, simple overall harmonic progression (open to all kinds of figurative
treatment), pedal trillo, all rather more regular and less capricious than in
Buxtehude’s interludes. Nor do the northerners prepare a linking imperfect
cadence so dramatic as the one here.
Fourth section
Widor’s remark that the final section ‘begins as a fugue, becomes a chorale
and ends like a concerto’ (Keller 1948 p. 60) does not make it quite clear
that the final toccata flourishes are incorporated within the fugue itself.
As Example 78 shows, converting the head of a fugue-subject into triplet
time often produces dotted rhythms. The problem with this particular meta-
morphosis is that what one assumes to be the correct lively tempo at b. 134
cannot be kept up: there is far more diminution as the fugue proceeds than
is ever the case in Frescobaldi or Froberger. Did Bach, as later with Vene-
tian concertos perhaps, misjudge Italian tempi, thinking them slower than
Frescobaldi assumed in Fiori musicali?
Since only the caput is used, section 4 is not a ‘variation’ of section 2,
and is quite different: the last true entry is less than halfway through, after
which the subject makes a witty stretto (b. 181), or modulates (b. 206) or
is distantly paraphrased (bb. 218, 225). Textures at times resemble those
elsewhere (compare b. 209 with Var. 10 of the Passacaglia, b. 80), but
the loose fugal writing is more toccata-like and thus very different from
the more ‘correct’ fugue of the second section. Neither entries nor episodes
clearly grow out of the exposition, and the writing varies enough (and comes
back to the tonic often enough) to begin to sound like an ostinato.
163 BWV 567–569
BWV 567 Prelude in G major

Copy by J. L. Krebs in Brussels Fétis 7327, also later copies (unattributed).
While at least one passage shows a composer familiar with Bach keyboard
idioms (bb. 10–15), the tone of the penultimate bar is alien, as are the
harmonies in bb. 8 and 17–18. Such 3/4 preludes based on scales above
pedal points may have been a genre for improvisation, to judge by a similar
but monothematic movement in Fischer’s Ariadne Musica (c. 1702, No. 13).
The composer is now assumed to be the copyist (Kobayashi BJ 1978 p. 46),
but imitating a genre.
BWV 568 Prelude in G major

Copy in P 1107 (later eighteenth century) and derivatives; late copies via
another route.
Two staves; headed ‘Praeludium con Pedale’ in P 1107, where anonymous.
That in P 1107 the movement follows the ‘Harmonic Labyrinth’ BWV 591
(the only contents) does nothing to establish the authenticity of either.
Further questions concern the pedal: its lines at bb. 3, 8ff., 26, 32 etc. look
unreliable, the result of a copyist unclear what it plays outside its semiquaver
solos?
While part-writing, sequences and pedal points could suggest an early
work of Bach, the absence of thematic interest does not; nor do the galant
sounds in bb. 32–3 (parallel sixths, with acciaccatura and syncopation). If
its returning material is an example of ‘ritornello principle borrowed from
the pre-Vivaldi Italians’ (Stauffer 1980 p. 56), it surely was not borrowed by
J. S. Bach. Nevertheless, its composer was familiar with figures typical of
Böhm (scales, sixths) and Pachelbel (pedal points) and knew what was
useful to a practising organist. (Do differences between the notation of
pedal points in bb. 1 and 8 reflect poor sources?) If Bruhns’s Toccata in G
was a model (Geck 1968 p. 21), one might expect even more modulation.
BWV 569 Praeludium in A minor

Three copies perhaps from a lost Autograph: P 801 (J. G. Walther, 1714–17?),
Lpz MB MS 7 (J. G. Preller) and P 288 (J. P. Kellner); also a lost Kittel
source.
164 BWV 569–570
Two staves; title-page in P 801 (written by J. L. Krebs) ‘Praeludium pro

Organo pleno con Pedale’.
Since sources are good, BWV 569 is accepted as an early work. Spitta heard
in it ‘something monotonous’ (I p. 398), but its single-minded pursuit of a
little motif, prefaced and rounded off by faster lines, is something of a tour de
force, especially with part-writing so ‘flawless’ (Breig 1999 p. 631). Perhaps
the motif is typical of the South German praeludium, but its exploration
over some 150 bars conforms to Bach and Walther’s interest in note-patterns
c. 1708, and indeed in their interest in the continuo-player’s realization of
4/2 or seventh chords.
Several details suggest that the movement is not far from a chaconne
en rondeau: triple time, phrases of four or six bars; regular, simple episodes
(three parts as against the pedal tuttis); descending harmony for each phrase;
passacaglia patterns as in Muffat (Apparatus, 1690) or Pachelbel. For exam-
ple, the last twenty-four bars suggest an episode followed by three (four?)
chaconne variations, then a coda. Other moments are more ‘northern’ (har-
monic pedal points bb. 36, 80), or even anticipate mature Bach (compare
imitation at bb. 49ff. with the Gigue of Partita in G major). Schöneich
(1947/8) sees it as a movement in four sections (1–48, 49–85, 86–116,
117–52) based on a falling scale, with a partial ostinato theme close to
Buxtehude’s E minor Ciacona and not out of place in the improvisatory stylus
phantasticus.
BWV 570 Fantasia in C major

No Autograph MS; copy in Lpz MB III.8.4 (ABB, J. C. Bach) and later
derivations.
Two staves (no pedal cues); headed ‘Fantasia’, ‘di J. S. B’, no pedal cues.
Spitta thought the Fantasia perhaps originally connected with the Albinoni
Fugue, BWV 946, though J. C. Bach’s copy does not imply this. It must be
one of the earliest works: its non-thematic four parts give the impression
of a didactic piece, close to Pachelbel, encouraging ‘a very careful legato’
(Spitta I p. 398). As with the Imitatio BWV 563, Canzona BWV 588 and
Fantasia BWV 1121, pedal has been assumed for the bass line (Kilian 1982
p. 167) but was surely at most optional.
If ABB’s heading establishes authenticity, the young Bach was casting
his net wide in learning to compose with motifs, here dactyls like the B
minor Fantasia’s but treated differently. South German precedents for it can
165 BWV 570–571
be found (see Example 79), but the ultimate source may be Frescobaldi’s
toccatas or even variations. As a free fantasia, BWV 570 is a counterpart
to the ‘Neumeister Chorales’, not very different from some of them (BWV
1091, 1093, 1116). Melody, modulation, texture, motifs and continuity are
all promising, more so than in the various preludes of Krieger’s Clavier-
Übung of 1698, doubtless known to the young Bach.
Example 79
As the Prelude in G major BWV 902 shows, sustained four-part style

keeps a family resemblance wherever it appears, a kind of self-generating
plein jeu music familiar to organists far and wide. The motivic bass line of
BWV 570 distinguishes it from South German pedal parts (which however
are also often optional), and it is more organized, like BWV 571 in this
respect. Despite the various dominants, one cannot always anticipate in
what direction it will meander (bb. 7ff.).
BWV 571 Fantasia in G major

No Autograph; P 287 (J. P. Kellner, after 1727?) and later independent copies.
Two staves; headed ‘Fantasia’ in P 287, ‘Partita’ in Brussels Fétis 2960 (later
eighteenth century).
166 BWV 571–572
Textures suggest that the work is for organ, but composed by whom? There
are signs of a concerto shape: (i) ritornello-like theme, (ii) slow imitative
movement ending out of key (major not minor), its theme related to the
previous movement’s, (iii) Allegro ‘variations’ on the descending hexachord
or octave (in minims). The commonplace opening subject has been found in
Kuhnau, as has that of the middle movement. The working-out rarely rises
above either the motivic invention or the note-patterns of a J. G. Walther, and
even (ii) fails to achieve any harmonic tension. Spitta heard it as more mature
than BWV 551, with a thematic unity and hence under Buxtehude’s influence
(I pp. 318–19), whose C major Praeludium BuxWV 137 with ostinato was
included in the ABB. In (iii), the ostinato bass, key, modulations, imitation
and position after a prelude seem to bear more than a chance resemblance
to Corelli’s sonata da camera Op. 3 No. 12.
One can also find resemblances to a passage in Corelli’s Violin Sonatas
Op. 5 and even to mature works of Bach – for the third movement, see
the fugal finale of the Concerto for Three Harpsichords BWV 1064, whose
bass line had long been familiar as an ostinato. Altogether, the work sug-
gests an enthusiastic assimilator of various styles, perhaps the young Kellner
himself. Yet the sources are good (Bartels 2001), and much in the uneven
composition, such as the final pedal point, matches much in the ‘Neumeister
Collection’.
BWV 572 Pièce d’Orgue (‘Fantasia’) in G major

No Autograph MS; early version in P 801 (J. G. Walther c. 1714/17?) and
known to Kirnberger circle? (Am.B.54 and 541); copies of revised version
in P 1092 (J. Schneider, c. 1729?), P 288 (J. P. Kellner 1726/7?, perhaps from
an autograph, with ornamented second part), SBB Mus.MS 30380 (via
C. P. E. Bach?) and a lost contemporary MS perhaps by H. N. Gerber; also,
one known to J. C. Kittel.
Two staves; all copies roughly as in P 801, ‘Piece d’Orgue di Giov: Sebast:
Bach’ or P 1092, ‘Piece d’Orgue à 5. avec la Pedalle continu composée par
J. S. Bach’; a lost source for Peters IV evidently had ‘Fantasia’. Headings
in P 1092: ‘tres vistement’, ‘gravement’ and ‘lentement’, in P 801: ‘Piece
d’Orgue’, ‘gayement’, ‘Lentement’. Perhaps ‘gayement’ was authentic, as if
for a lively allabreve piece for harpsichord, with French title and headings
as for a presentation or dedication copy (Rampe 2002).
There must have been at least two autographs, one the source for P 801, one
revised and perhaps ornamented: another work known in more than one
form. Walther’s version preserves important hand-distribution in the first
167 BWV 572
section (see KB p. 208) but has no pedal-cue until it is necessary at b. 176 –

which suggests, but does not prove, that only the last section is pedaliter.
The dots in bb. 1, 5 and 17 imply staccato; would they perhaps not have
been found in the earliest copies?
Although the work draws on French idioms, Pièce d’Orgue is not as
common a term as one might assume, nor is there a similar movement in
Grigny’s Livre, Bach’s copy of which (c. 1709/12) may be contemporary with
BWV 572. Pièces appears on the title-page of two books probably known to
Walther and Bach (Du Mage 1708 and Raison 1688), and also in MS copies of
Marchand (c. 1700). Du Mage’s Livre begins somewhat like BWV 572: a free
prelude for petit plein jeu is followed by a denser contrapuntal movement
for grand plein jeu. But BWV 572’s second section also has features found in
French pleins jeux, such as suspended harmonies and a bass-line rather like a
purposeful cantus firmus (opening plein jeu of Boyvin’s Premier Livre, 1690,
probably known to Bach). While Walther’s term ‘gayement’ might just be a
misreading for ‘gravement’ – they are opposite terms in F. Couperin’s sonata
‘La Françoise’, 1690s – their tempi may not differ much (Gilbert 1993).
At the same time, both outer sections conform more to the tradition
for fresh, rather wild passage-work in preludes by e.g. Buttstedt (Clavier-
Kunst, 1713 and in ABB). The beginning reinterprets the northern toccata
with an original, repetitive figure demanding attention, such as was known
in France (the so-called perfidia: a repeated motif, ‘une affectation de faire
toujours la même chose’: Brossard 1703 p. 77); and the third section has a
form of passaggio, more thoroughly larded with acciaccaturas than any in
Buttstedt’s Clavier-Kunst. Both outer sections are unusual in the amount of
repetition on several levels, rather as if there were a quasi-French dialogue
in progress, though sources give no hint of an option for two manuals:
1–28 rh/lh broken chords, ‘pedal points’ in soprano or bass;

returns at b. 5 (early version) and 17 (partial); implied
tonic pedal
29–185 five-part alla breve harmonies; scales (rising semibreves,
falling crotchets); semibreve theme in G, D, B minor, G,
A minor, E minor, A minor, G minor, D minor, G; lastly
in 3rds (six parts)
186–202 rh/lh broken chords plus acciaccaturas; pedal falls
chromatically (but rises diatonically in second section);
dominant pedal point
In three different ways, each section works one distinct approach to har-
mony, and each has shifting harmonies which are linked by common notes
between the chords, either broken (outer sections) or sustained (inner),
168 BWV 572
and each is without disruptive cadences, the whole a unique tour de force in
harmonic manipulation. The ‘linking’ notes in the third section are often
the very non-harmony notes of the acciaccaturas.
Simple tripartite structure in e.g. J. Speth’s toccatas (Augsburg 1693),
though once thought an influence (Dietrich 1931 pp. 62–4), does not cor-
respond to BWV 572 except in the pedal points at begining and end. Only
a few details suggest parallels elsewhere, but perhaps the key of G major is
itself a French allusion to the petit plein jeu? Also, a common pulse may have
been intended: dotted crotchet = minim = quaver.
First section
For Reinken’s Toccata in G major (ABB) see Example 80. A prelude by
C. F. Witt (†1716) also has manual semiquavers followed by a durezza passage
with pedals, but no extant toccata approaches the catchy long-breathed
monody of BWV 572, one of the most original gestures even in Bach. Perhaps
fiddlers’ improvisations gave the idea for it, as they might have for the
Preludio of the E major Violin Partita?
The repetitions suggest echoes, as in the C major Toccata and the violin
solos, but here they are fully integrated in the regular swirl of notes. Is b. 24
too to be repeated (echo)? And a big question: since there is an implied tonic
from first to last, even avoiding a dominant in b. 24, is there any option to
add a pedal G throughout?
Example 80
Second section
An influence here might be J. Boyvin’s Livre d’Orgue, copied by Bach’s
Weimar pupil J. C. Vogler, where the phrase ‘plein jeu continü’ appears (cf.
the ‘Pedalle continu’ in P 1092) and where préludes tend towards sustained
four-part harmonies. Furthermore, the preface to Boyvin’s second book
reminds the organist how to play durezza harmonies on the organ. But it
has nothing as systematic as the descending semibreve bass of an earlier
piece much closer to BWV 572: the sixth verse of Weckmann’s ‘O lux beata
trinitas’, ‘à 5 im vollen Werck’.
Durezza harmonies often led to rising semibreve scales, as one sees
in Example 81. BWV 572 produces from them a tissue of ascending and
169 BWV 572
Example 81
descending lines, in all voices, now more systematically than in Weckmann.

It was an idiom to which Bach often turned in his maturity, in counterpoint
either stricter (Ricercar à 6, Musical Offering) or quicker, fluent and dra-
matic (Christmas Oratorio No. 21). The lines move predominantly by step,
leaping only to start again.
The section’s harmony is organized in an ‘ideal’ series of seventh and
ninth chords which, reduced, look like an equally ideal species-counterpoint:
see Example 82. It is too much to say, therefore, that ‘the Fantasia in G was
written completely in the French spirit for a French organ’ (Schrammek
Example 82
1975 p. 104). The Neapolitan sixths at bb. 57, 139 are perhaps ‘early’ signs,
but there are similarities with other works of Bach – compare the rising
harmonies of bb. 113–15 with the end of the Fugue in D minor WTC1.
The ornamentation transmitted by Kellner certainly strengthens the allusion
to a true French grave style, as he must have realized.
While the background for the section is clear, its length, non-fugal texture
and thoroughness of organization in what is essentially an improvisatory
style are found only here. In addition to parallel 3rds, a device necessary
170 BWV 572
in any five-part piece is contrary motion (bb. 113–15 etc.), and further
unity is provided by the periodic climaxes, the crotchet lines usually moving
by step, and most of all the pedal semibreves and their contrary motion,
appearing at regular points in different keys. No idle repetition results from
this technique, as can be seen by comparing two sections which begin with
the same progression (e.g. b. 76 and b. 118), and by the final two-octave
ascent (bb. 157–71), all of which are achieved without fugal imitation.
The puzzle of the low BB in b. 94 (an octave higher in Kellner’s copy)
has no clear answer. Perhaps it was an ‘ideal’ note; or a French allusion, to
pedal reeds below C, en ravalement; or written for harpsichord, with C
tuned down. In any case, it is not unique: it appears in copies of the C major
Toccata (bb. 138ff. in P 286, ‘not to be believed’: KB p. 492) and E major
Toccata (b. 18, version in C major).
Third section
The broken chords with acciaccaturas are kept up until the cadence:
Example 83. Of the various acciaccatura traditions a possible influence was
d’Anglebert’s Pièces de Clavecin of 1689, whose ornament table at least was
known to Bach and where the continuo player is recommended to play such
chords. More extravagant effects were suggested by F. Gasparini (L’armonico
Example 83
pratico, 1708) and by German writers he influenced (e.g. Heinichen 1711,

1728). It was not an effect recommended for organ: the chromatic notes
apparently ‘slipped in’ do produce strange combinations which only gradu-
ally – but quite noticeably – soften towards the end. Though this end, being
a kind of cadenza on a 6/4 followed by a trilled 5/3, anticipates a good deal
of later music, historians of musical form are unlikely to know it.
Thus, like the first two sections, the third single-mindedly exploits a par-
ticular musical device, pushing it beyond what was traditional. Moreover,
in its solo line and inner repetitions the third section is like the first, but
in its harmonic continuity more like the second. Together the three survey
the three main types of harmonic bass-line: an implied tonic pedal, a ris-
ing diatonic bass and a falling chromatic bass, and do so in proportional
tempi.
171 BWV 573–574
BWV 573 Fantasia in C major (fragment)

Autograph MS in P 224 ( = AMBB, 1722).
Two staves; headed ‘Fantasia pro Organo’, pedal line ‘ped’.
The Fantasia follows the French Suite No. 5, written down as it was being
composed (NBA V/4 KB)? It breaks off before the end of the page, after which
an empty side follows before the next piece, which is also incomplete (Air,
BWV 991). Were both meant to have been completed by family members?
A piece in four and five parts is exceptional in the two Anna Magdalena
Books (begun 1722 and 1725) and contributes to a repertory already very
wide and including a chorale. The order is BWV 812, 813 and 814 (both
incomplete), 815, 816, 573, 991, 728, 813 and 814 (their further movements),
841.
The pedal line, hardly suitable for a beginner, develops its own motifs.
The texture varies, developing parallel inner 3rds as in other five-part music,
and in idiom it is close to mature organ works – compare the last two bars
with the Fugue in C major BWV 547. (Both the 1725 AMBB and the CbWFB
have another ‘five-part prelude in C major’, i.e. the first prelude of WTC1.)
Melodious phrases such as the cadence at the end of b. 4 arise naturally, and
there are at least three promising sequences before the more conventional
close.
Since the thirteen bars do not suggest any particular shape before ending
on the mediant, the piece looks like an improviser’s prompt such as Bach
is said to have used (Dok II p. 397). The final full bar, modulating to E
minor, starts a new line in the MS. Until that point it looks as if the Fantasia
is going to cadence in the dominant, and it could have moved in any one
of several directions for a student to explore. This is more likely to be the
reason for such a fragment than that it was demonstrating the need to plan
page-layouts beforehand (NBA V/5 KB pp. 67f.) or that wife or son already
knew such pieces by heart (Schulenberg 1992 p. 130).
BWV 574 Fugue in C minor (‘on a Theme of Legrenzi’)

No Autograph MS (but with the Passacaglia in a so-called ‘Guhr autograph’,
see NBA IV/7 KB p. 129); copies in P 1093 (J. G. Preller), P 247 (c. 1730?),
Lpz MB MS 1 (without final section, c. 1740, via Kellner? Stinson 1989
p. 92).
Two staves; ‘Fuga’ (P 247), ‘Fuga ex C mol’ (P 1093).

172 BWV 574
What appear to be distinct versions of this confusing piece are best ex-
plained by supposing that the early text (BWV 574b) acquired several
reworkings, perhaps in more than one copy in the Bach household (KB
pp. 501–2), perhaps sometimes shortened without authority. Compare with
BWV 545. The reworked versions seem not to have mentioned Legrenzi,
as was also the case with some copies of the Albinoni fugue BWV 951,
headed by Walther ‘Fuga ò vero Thema Albinoninum. elaboratum et ad
Clavicembalum applicatum per Joa. Bast. Bachium’. The phrase ‘Cum sub-
jecto pedaliter’ for BWV 574b, which Spitta thought referred to the second
subject (I p. 421), probably indicates that the pedal is needed for the expo-
sitions, unlike BWV 575 or 549. (‘Cum subjecto’ means ‘with a persistent
countersubject’, as in BWV 579, and ‘cum subjectis’ indicates a permutation
fugue.)
To judge by J. C. Bach’s title for another piece in the Mö MS – ‘Fuga.
Thema Reinckianum à Domino Heydornio elaboratum’ – the verbal for-
mula belonged to a genre popular in c. 1700–10, not quite fairly described
as an ‘arrangement’ (KB p. 501). BWV 574 has a subject less melodious than
Italian string-fugue themes such as BWV 951’s, being more like keyboard
or vocal subjects with a common-property cadence – as in Example 84,
the Toccata BWV 914. Schöneich 1947/8 showed that the Benedictus from
Example 84
Palestrina’s Missa Pange lingua has a similar theme – the more similar it is,
the more original Bach’s second subject is made to appear – and Hill 1986
pointed to two themes in Legrenzi’s Sonata Op. 2 No. 11 (Venice, 1655). The
Sonata ‘La Cetra’ in Op. 10 (1673) also has a theme similar to the first, but
moreover with much the same notes as BWV 574’s second subject (Swale
1985). Though it is not improbable that Bach would extract his subject from
a complex of themes in a Legrenzi trio – as another C minor work, BWV
562, could have done from Grigny – Legrenzi himself might have been doing
no more than adopting common-property formulae.
Such themes could certainly inspire a long movement, even some per-
mutable counterpoint, as is hinted at in BWV 574 from time to time. Is it
possible that the Fugue in C minor on a Theme of Legrenzi and the per-
mutation Fugue in C minor on a Theme of Raison made a pair originally,
one with a toccata section added at the end, the other a long passacaglia at
173 BWV 574
the beginning? The Legrenzi Fugue followed the Passacaglia in the ‘Guhr
Autograph’, probably the copy by C. G. Meissner: two fugues in C minor on
foreign themes. It appears to be less dependent on Legrenzi than BWV 579
(see p. 180 below) is on Corelli:
1–37 exposition (one countersubject), episode, dominant, relative,

tonic
37–70 second theme with three- and four-part exposition; pedal
subject simplified; new countersubjects (53, 57)
70–104 themes combined seven times, invertible; coda pedal point
implied
105–18 toccata section thematically unrelated (including pedal
thirds?)
Moments such as bb. 77 and 89 imply that the ‘original’ was a trio, perhaps
for gamba and violin not two violins (see dialogue in b. 99), although a
similar impression given by the Concerto BWV 592 has been shown to
be misleading (see p. 206 below). Either way, Bach adds a fourth part,
converting it into organ music with Buxtehudian sixths (b. 100 etc.) and
section-breaks making it easy to add stops for a gradual build-up.
Spitta thought the cadences prefacing each subject entry gave a
‘disjointed and short-breathed’ effect (I p. 421), but this is counteracted
by having the subjects start off the beat. Nevertheless, so many perfect
cadences are a sign of early date, as in Sonata No. 4’s slow movement.
They also tend to be melodious (e.g. bb. 18, 23), as in other early works
such as the B Capriccio, and Frotscher had no evidence for think-
ing them Legrenzi’s cadences (1935 p. 860), although maybe the octave
imitation from b. 4 was his. Spitta too guessed in supposing that the open-
ing ‘goes back to Legrenzi’s original’, with ‘Bach’s real manner’ taking
over in b. 34. The gradual move from quavers to the semiquavers of the
second fugue, and the disintegration of these into toccata figuration, are
as Bach-like as the quite different continuous motion in the Albinoni
Fugue BWV 951.
The counterpoint may be Italian-inspired but the keyboard texture (in-
cluding quasi cross-references, bb. 67, 21) has little of the facile alla breve of
BWV 589. Although the coda’s broken chords resemble moments in Buxte-
hude, Bruhns, Lübeck and others, their prolongation over seven bars does
not; nor do the repetitive arpeggios of bb. 111–12 (not found in BWV 574b).
The close is uncertain: any tablature original might leave it unclear, even
optional, whether pedal C is taken off before the end of the whole piece and
whether the last two notes are manual.
174 BWV 574a–574b
BWV 574a Fugue in C minor (‘on a Theme of Legrenzi’)

No Autograph MS; copy in P 207 (late eighteenth century).
Two staves; title, ‘Fuga a 4. Voc’ only.
This has ‘often a more continuous and clever part-writing’ than BWV 574;
and in ‘leaving aside’ the last fourteen bars ‘points to a later, simplified
re-working’ (BG 38 p. xlix), an improvement that looks authentic (KB
p. 571). But not only does BG’s claim seem overstated – differences are
not enough to suggest chronology, each version is ‘more continuous’ than
the other at different points – P 207 may also be unreliable, insofar as other
music it contains, such as Handel suites, seems to have been ‘improved’ by
the copyist (Brockaw 1995). To be authentic, it must be so that the extra
fifth part in bb. 66ff. was not copyist’s work, that the ‘omitted’ part in
bb. 50–1 was not an error, and that the final pedal-point was the com-
poser’s, all of which are unproved. Either approach to the final cadence –
with pedal point but without final toccata, as here, or the opposite – is plau-
sible. A final pedal point instead of a toccata could reflect the later taste of
either the composer or an arranger.
The closer to the original Legrenzi string fugue this version without a
toccata coda is, the more it fits in with the style and contents of the MS itself,
where it follows part of WTC and its fugues of more than one subject but
without toccata flourishes.
BWV 574b Fugue in C minor on a Theme of Legrenzi

No Autograph MS; copy (?) in Lpz MB III.8.4 (ABB, J. C. Bach); indepen-
dently in P 805 (J. G. Walther, before 1714?), some others via J. C. Kittel
(? no final section).
Two staves; title by J. C. Bach, ‘Thema Legrenzianum. Elaboratum per Joan

Seb. Bach. cum subjecto. Pedaliter’; by Walther, ‘Fuga’.
BWV 574b has fewer continuous semiquavers in bb. 21, 34, 67, 77 and 86
than BWV 574, and a less clear fall and rise of arpeggios in bb. 111–13.
Sources suggest that BWV 574 is a later, revised version by the composer of
BWV 574b, but whether the differences are frequent or significant enough
to justify the term ‘version’ (either as something intended by the composer
or as reliably transmitted by sources) is questionable. The more continuous
semiquavers of BWV 574 would not be difficult for a musical copyist to
175 BWV 574b–575
incorporate, since no radical use of motif is involved. However, it is certainly

possible that such ‘early’ signs as the broken chords of b. 68 (Zehnder 1988
p. 103) would have been revised over the years. It seems that around 1705 the
composer was interested in making Italian contrapuntal harmonies thicker,
to judge by his figured copy of a cantata by Antonio Biffi (see Wollny 1997
p. 16), and such thickening can take various forms.
BWV 575 Fugue in C minor

No Autograph MS; two contemporary copies, P 247 (c. 1730?) and Lpz Go.
S. 310 (1740/50), and later, probably via other versions/copies, including
one by Kittel?
Two staves; headed in P 247, ‘Fuga di Bach’, ‘Adagio’ at b. 73 in Go. S. 310

and at b. 65 in the Clementi print (see below).
Sources support neither the attribution to C. P. E. Bach in Clementi’s English

edition of 1811 (KB p. 272) nor the assertion that it is for ‘Flügel mit
Pedalbass’ (BG 38), though they do specify pedal (C–c ) for the last twelve
bars, where it appears indispensable.
Although BWV 575 is probably an astute imitation by young Bach of old
canzonetta fugues, there are puzzles. Any similarity to the final fugue of the
E minor Toccata BWV 914 (Example 85) centres on the figuration (see also
BWV 549a, bb. 52–3), the breathless continuity, the simple accompaniments
Example 85
and a final toccata section. But is BWV 575 the final section of a lost toccata?
A subject starting on the submediant is not found in WTC, nor is its am-
biguous metre. These two details, combined with the dazzling harpsichord
figuration (e.g. bb. 23–34), justify the reliance on tonic and dominant for
the entries, which can appear as if out of the blue (b. 58).
A canzonetta subject produces a rondo-fugue in which the subject is
mostly accompanied by its countersubject, and episodes are brief inter-
ludes between entries. This one draws on other music: sequences from
176 BWV 575–577
Italian string music (see BWV 532 b. 32); Buxtehude–Bruhns–Böhm idioms

(bb. 41, 44, 67, 70, 74); the obsessive passage before the typically surprising
F (rather than f?) in b. 65; scales; alternate-foot pedalling. The result is
a fugue often admired, not least by Schumann who published it in 1839
(NZfM Supp. 5 Pt. 3).
The exact point at which the subject re-enters is often surprising and
stretto-like, and its tail is generally harmonized imaginatively (bb. 39–40,
54–5). Keeping to tonic and dominant entries does not preclude other keys
in the episodes (B minor, b. 45). The keyboard style and wide tessitura are
typical of the composer’s toccatas for harpsichord, although such details as
the broken chords are idiomatic to both instruments. Quick chord-changing
in such bars as 26 is unfamiliar in the maturer organ music, and most
indebted to tradition is the coda, including the new key at b. 65. Spitta
thought that without the final pedal solo ‘we would not otherwise believe’
that the fugue was at an end (I p. 250): the three final cadences, two perfect
and one plagal, are necessary because of the postlude’s new key at b. 65.
BWV 576 Fugue in G major

Copy formerly in possession of F. Hauser (Peters IX, 1881).
‘In view of their musical makeup, BWV 576 and 577 can scarcely go back
to Bach’ (KB p. 15). While the exposition may be authentic, the pedal entry
in b. 68 does not suggest J. S. Bach, any more than the long, unified shape
makes it likely to be work of a previous composer (Keller 1937). The ‘melodic
beauty and charm of the theme’ (Keller 1948 p. 51) are those of Italian string
fugues, including Handel’s or Corelli’s, and moments in it remind one of
concerto transcriptions.
As is usual in such fugues, most attention is directed to the subject,
but a few independent episodes are introduced, extending the movement
to almost 100 bars – a ‘German’ characteristic. The irregular entries and
answers and the minimal suspensions are those of a minor composer, one
familiar with alternate-foot pedalling in solo passages, perhaps a pupil of
Bach, one able to learn from his inventive sequences (see bb. 32–4). For
bb. 42–3, cf. the end of the D major Fugue WTC1.

Contemporary (?) copy formerly in possession of F. W. Rust (via Johann
Christian Bach? NBA IV/7 KB p. 124); later (?) copies include one by
L. Scholz.
177 BWV 577
On authenticity, see also BWV 576. Because of some effective moments,

especially in the final section, the composer was usually assumed to be J. S.
Bach until doubts were raised about the sources and the authority for the p
and f signs in Rust’s MS.
Spitta pointed out similar subjects in Buxtehude but heard here ‘a
bolder verve’ that precluded him, ‘who otherwise could well have written it’
(I p. 320). It might be the jig finale of a longer work – the variant of an
earlier fugal movement, as in Böhm’s Praeludium in D minor in the Mö
MS – but is already long. So too, however, is Buxtehude’s C major Canzona,
partly copied by J. C. Bach in the ABB and thought by Spitta also to be
part of a larger composition. See Example 86. Both there and in BWV
577 it would be possible to conjecture what an ‘original’ 4/4 version of
the theme was. The two works are similar, and the sudden move to the
dominant at the end is not particularly typical of J. S. Bach’s subjects, nor
are the persistently iambic chords. Whoever wrote it, BWV 577 is true to
genre.
Example 86
The simple sequences combined with a confident idiom make the piece
difficult to attribute. The confidence shows itself in such passages as bb.
26–7, where a four-part sequence exploits a well-spaced series of seventh
chords, provides an unusual but useful texture for practice, and is referred
to again only two bars later. Echoes within a subject do not suggest J. S.
Bach, but doubtless copyists could add the signs, and it is only surprising
how few appear in sources generally. Pedal seems necessary because of the
spacing, and the subject has been convincingly altered for its sake (b. 28
etc.). Large gaps in the pedal part are not out of character in early fugues,
and the cumulative effect of the whole last third of the piece reminds one of
the Fugue in D major.
Other details (here in italics) might cast doubt on its authenticity:
1–29 exposition, with long modulatory codetta after first answer,

and a shortened fourth part (pedal) merging into:
29–34 episode, keeping up exposition’s texture
35–40 entry (a) in mediant and (b) distributed over tenor and
soprano, settling on to the tenor and passing to:
178 BWV 577–578
40–7 episode, reducing the texture to one part

47–86 series of entries (sudden tonic return after mediant, 77), short
episodes
The movement is puzzling, for while the episodes contain motifs not found
in the subject, such a passage as bb. 78–86 is a thematic complex based on
bits of the subject, original and idiomatic. A similar motif can be found in
BuxWV 174, but not so exhaustively; nor does this contain regular entries
for the last third of the piece, or gravitate towards four parts like BWV 577.
The simple sequences do not argue against J. S. Bach’s authorship since they
throw the entries into relief, as if such jig fugues have room for the faux-
naif. And difficult though it is to imagine J. S. Bach writing such passages as
bb. 55–6, they might reflect a corrupt source.
BWV 578 Fugue in G minor

No Autograph MS; copies in Lpz MB III.8.4 (ABB, J. C. Bach), P 803 (J. L.
Krebs c. 1730) and derivatives from both; also SBB Mus. MS 11544 (J. C.
Vogler c. 1730); lost Kellner and Kittel copies known through derivatives
(P 288, P 320).
Two staves; headed ‘Fuga’ (ABB and P 803), ‘Fuga pro Organo Pleno’ in
P 320.
The many copies, including four prints by 1850, testify to the piece’s pop-
ularity, no doubt arising from its catchy violinistic subject (with open d
string). An early work, it seems to have existed in two versions in the Bach
portfolio (KB p. 538). Later on, a spurious prelude was associated with it
(Kobayashi 1973 p. 331).
1–22 exposition; codetta b. 5; real answer b. 6; constant

countersubject
22–30 episode; ‘false’ entry or quasi-stretto (tenor, then soprano),
tonic
30–45 episode; entry in relative (alto + codetta as at b. 11, then pedal)
45–55 episode as 22; entry on subdominant
55–68 episode; final entry, now in four parts; shortened for cadence
There are no learned effects (augmentation, stretto, etc.), only distinctive

motifs in a long theme of three phrases encouraging players to conjecture
various phrasings.
179 BWV 578–579
The subject belongs to a north German tradition of Spielthemen (idio-

matic, fun to play), but is more tuneful than most. Reinken’s G minor Fugue
shows similar semiquaver figures, a tendency towards broken chords, simple
sequences and a succinct close etc; but BWV 578 has clearer entries (always
well prepared and timed), more consistent counterpoint and a better tune.
Unlike another ‘fun to play’ fugue-subject – in the Concerto in C major
for Two Harpsichords BWV 1061a – this one remains within an octave.
Perhaps because it is so catchy, J. G. Schübler (a pupil, later engraver of the
Six Chorales) also wrote a fugue on this theme.
The counterpoint has been described as ‘mostly only one-part’ and thus
early (Spitta I p. 400); but in fact the three-part texture of bb. 17–21 is
that of a regular permutation fugue in which the counterpoint – including
the simple semibreves – returns in different keys and in different com-
binations. The three parts of bb. 27–30 are a complete inversion of
bb. 18–21, thus explaining why the pedal enters without theme in b. 26,
for by the next bar it takes up a role in the inverted three parts. The ‘some-
what facile sequences’ in episodes of BWV 578 have been aptly described
as ‘part of a successful emulation of Italian violin style’ (Schulenberg 1992
p. 83).
The countersubjects might be ‘derived from the second and third part
of the subject’ itself (Frotscher 1935 p. 878), and a certain pattern of semi-
quavers (from b. 5) is found in about half of the bars, rectus or inversus. The
alteration of both subject (b. 44) and countersubject (b. 51) argues that this
bass line was meant for pedal, though the sequences from b. 22 look more
like those of string trios. Typical of the composer is a fluency free from the
repetitive or motoric rhythms of fugues by Buttstedt, Vetter and others. Its
sources suggest an early fugue, while its simplicity implies that the composer
consciously gave it a shape different from the other early fugues BWV 574b,
944, 992, 531, 549a and 566.ii.
BWV 579 Fugue in B minor (‘on a Theme of Corelli’)

No Autograph MS; copies by W. F. Bach (? see Peters IV) now lost, and via
Kellner (? P 804 and Lpz MB MS 1) or Kittel (Lpz MB III.8.18).
Two staves; headed ‘Fuga’ in Lpz, ‘Thema con Suggeto Sigre. Correlli elabor.’
in P 804.
The subjects appear in the second movement (Vivace) of No. 4 of Corelli’s

Sonate da Chiesa a Tre Op. 3 (Rome, 1689). On the assumption – not
established! – that this print was the source, correspondences are:
180 BWV 579
Corelli BWV 579

1–3 1–3 octave lower
9–12 top part 6–9 top part
15 cadence to D ?10 cadence to B minor
16–19 B minor ?11–14 (F minor) or 23–4 (B minor) or
32–4 (B minor)
30–1 bass ?90–1 bass
Correspondences are slight and uncertain, though ‘elaborat’ and its cognates
usually implied a transcription: 39 bars have become 102, a fourth part is
added and pedal is required. All the ‘reworked’ themes by Corelli, Albinoni,
Reinken, Legrenzi and especially Raison (the Passacaglia) aim for length and
richer detail, and Corelli’s double subject also offered a model for tight part-
writing, thematic bass, exposition with tonic subjects, and a run of perfect
cadences as found in early Bach fugues. At the same time, however, there is
an energetic quality to Corelli’s fugue and a rich beauty of textured string
sound not obviously transferred to BWV 579, which must be slower.
1–24 subjects answered in tonic; 11/13, dominant answers; 22,

tonic
24–41 episode; new semiquaver figure; minims 25–34 from c
(see Example 87); entry + answer, each double; new
countersubject
41–58 episode, extending quavers; derived minims; tonic entries,
double
58–73 episode: derived minims; new motif (? 62); plus subject (67)
73–7 entry, with countersubject newly treated (D major, B minor)
78–90 episode, at first with material similar to previous
90–102 stretto final entries; ‘Italian’ adagio close
Example 87
The form is not clear, though sections are marked by the presence or absence
of pedal, and entries are more clearly distinguished from episodes than in
Corelli. (Schöneich 1947/8 saw the divisions as bb. 1–24, 25–34, 37–61,
181 BWV 579–580
62–71, 73–102.) Although there is little opportunity to change manual, a

‘strong sense of concerto style’ with episodes can be heard (Schulenberg
1992 p. 55), being fuller than Corelli’s trio yet not expanding much tonally.
To reserve semiquavers largely for episodes – a procedure familiar in Italian
string fugues – is untypical of the maturer Bach.
Apparently, ‘Corelli’s six theme-complexes have become ten’ (Braun
1972), and the double subjects are used differently. They appear together
only four times in Corelli but always in BWV 579, so b. 67 is no true entry,
and only the third stretto voice of b. 91 is strictly subject. Both fugues keep
to nearby keys, dominant in Bach, subdominant in Corelli. Corelli’s regular
stretti at both bar and half-bar do not appear in BWV 579, which reserves
stretto for the fourfold tonic–dominant climax in bb. 90–1, anticipated by
Corelli in three parts (bb. 35–6). Bach’s stretti here, rare in early fugues, will
resemble later stretti based on falling fifths or fourths, as in the B minor
Fugue, WTC1. Throughout BWV 579 the harmony is richer (but less deft?)
than Corelli’s, and is already developing a greater sense of urgency at
bb. 16f., 35f., 75f. and 93f. than anywhere in the E major Toccata.
The double theme is typical of italianate subjects adopted by Handel
(Concerto Op. 3 No. 2, or the Fugues in B and G minor from Six Fugues)
and individual themes of Bruhns (Praeludium in E minor) or Buxtehude (in
BuxWV 151), while the style behind such sections as bb. 79–90 is as Corellian
as the theme. This is so despite a typical German melody at bb. 82–3 and
various similarities to the Prelude and Fugue in D major. Nevertheless, even
the episodes from b. 25 or b. 65 are also not unlike passages in string fugues,
e.g. Corelli Op. 3 No. 12. Despite some play with the tied-crotchet motif of
bb. 54, 62 – which also may come from Corelli (see his b. 5, or the same
sonata’s previous movement) – the emphasis is on whole themes rather than
motifs. The final cadence is unusually modest, italianate as in Handel.
Evidently, the young Bach gained length by ‘spinning out’: of Corelli’s
39 bars, 24 have been counted as containing the theme, while of Bach’s 102
only 39 do (Tutino 1987 p. 69), leading to speculation about the Golden
Section (39 : 24). But was Corelli’s print-version the one used?
BWV 580 Fugue in D major

Later eighteenth-century Berlin copies only (Am.B.606, P 784).
The subject (‘of little worth’: Bartels 2001) is similar to the countersubject
of the Allabreve BWV 589, in notes, key and pitch, as if extracted by a less
than expert hand. The subject and some of its working out are not unlike
a Fugue No. 10 in G major in F. W. Marpurg’s Fughe e capricci (1777).
182 BWV 580–582
A further fugue in Am.B.606 is attributed to ‘Johann Christoph Bach’

( = BWV Anh.III 177), while P 784 also contains C. P. E. Bach’s Solfeggio
in C minor.

Copy in Lpz Poel 18 (c. 1790).
The MS Poel 18 (a single sheet) contains two three-part fugues compe-

tently composed on somewhat angular themes: BWV 581 and the chorale
‘Wir glauben all’ BWV Anh.II 70 (not attributed here to J. S. Bach). Perhaps
BWV 581 is also a chorale-fugue, though without any sign of being organ
music. Neither work has form, texture, figuration, invention or counter-
point characteristic of J. S. Bach at any period, although Anh.II 70 shows
familiarity with the old chromatic fourth in D minor.
BWV 582 Passacaglia in C minor

No Autograph MS, sources as follows (NBA KB IV/7): tablature-derived
copies by J. C. Bach in ABB and Lpz MB MS R 16, 9 (last 59 12 bars only),
further by J. C. Kittel (hence P 320); score-derived copies by J. T. Krebs
(P 803) and further copyists from Weimar (P 274) or Leipzig (e.g. P 286),
also probably via C. P. E. Bach (e.g. P 290) and C. G. Meissner (called the
‘Guhr autograph’ in Peters I).
Two staves in ABB and P 803 etc; headed in ABB, ‘Passacalja. ex C con Pedale’
and ‘Fuga cum Subjectis’ (for which see BWV 574); ‘Thema fugatum’ in
Meissner.
Evidence for a tablature original comes from the kind and number of errors
in some copies, such as octave displacements; and evidence for a revised
staff-score version from similarities in P 274 to Bach’s notation elsewhere
(KB p. 128). Whether the ABB copy, which is written in the book reversed,
dates from 1706/12 (Schulze 1984 p. 50) or c. 1708/13 (Hill 1991 p. xxii),
the tablature was earlier and perhaps made for or soon after the Lübeck
visit of 1705–6. Probably it had no pedal cues and left awkward playing mo-
ments, where parts collide or need juggling. Both movements were compo-
sitional essays leaving practical considerations secondary. As with the finale
of Capriccio BWV 992, perhaps the counterpoint was created on paper,
from conventional figurae (Passacaglia) or from permutable lines (Fugue).
183 BWV 582
Instrument and purpose

Even P 803 omits such phrases as ‘pro Organo’, but here is no authority for
BG 15’s rubric ‘Cembalo ossia Organo’ or Forkel’s phrase mehr für zwey
Claviere und Pedal als für die Orgel (‘more for double clavichord [?] with
pedal than for organ’: 1802 p. 60). Mattheson knew that organists wrote
ciacone (1739 p. 477), but he had in mind a different kind of dance, in a
church province with different traditions.
Such an essay in sustained form could have been prompted by Buxte-
hude’s ostinatos appearing – thanks to the Lübeck visit? – in the ABB; and its
handling of common-property motifs is surely earlier than the Ob’s, despite
claims to the contrary (Zehnder 1995 p. 334). The earlier it was composed,
the more it fitted in with the ABB’s survey of styles: a Toccata BWV 910,
an Ouverture BWV 820, a Passacaglia, a Fugue BWV 578, a chorale prelude
BWV 724, Variations BWV 989, the Legrenzi Fugue, three kinds of fantasia
BWV 570, 563, 944, and seven ostinatos, including unique copies of Buxte-
hude’s four (plus Pachelbel’s D minor Ciacona and Böhm’s Chaconne in D).
Assembling so many ostinato works – ‘according to French taste’ (Riedel
1960 p. 206) – was not at all common in Germany, and BWV 582 may have
been responding to all of them. It is also harder to play as it systematically
explores a series of common note-patterns from one to five parts, doing so
more thoroughly than a cantata ostinato like BWV 131 (1707).
Influences
The fugue’s main subject was found by Guilmant and Pirro, Archives des
Maı̂tres de l’Orgue II, 1899, in the Christe of the second mass of Raison’s
Premier Livre d’Orgue (Paris, 1688), subtitled ‘Trio en passacaille’:
Example 88. Whether either Bach or Raison, whose book was also copied
by J. G. Walther, knew that the subject resembles a Gregorian Communio
for the tenth Sunday after Whit is doubtful (see Radulescu 1979), but the
27-bar passacaille is not unique: in Raison’s sixth mass the Christe is an-
other ‘Trio en Chaconne’ with a four-bar bass very like the second half of
the Passacaglia theme – a curious coincidence, if that is what it is.
The possibility must be that BWV 582 began as a Fugue in C minor on
a Theme of Raison comparable to the Fugue in C minor on a Theme of
Legrenzi, and then used a second theme by Raison (as BWV 574 does by
Legrenzi?), rewriting it to make an eight-bar ostinato, longer than Buxte-
hude’s but like Krieger’s in Clavier-Übung, 1698. Ostinati are rare in the
keyboard music of the ‘old French masters’ whom Emanuel said his father
admired (Dok III p. 288), more so than the Chaconnes en Rondeaux such
as the one in Dandrieu’s suite copied by Walther in P 802. Perhaps the im-
itative opening of Example 88, unusual for a passacaille, stimulated Bach’s
interest? As for such dance-types in the liturgy: Raison directs that pieces in
184 BWV 582
Example 88
the style of ‘Sarabande, Gigue, Bourrée, Canaris, Passacaille and Chaconne’

are played more slowly ‘à cause de la Sainteté du Lieu’.
Another possibility is that the resemblance to Raison’s theme is coinci-
dence, ‘superficial’ (Buchmayer SIMG 1900–1 p. 270), no real ‘borrowing’
(KB p. 127). But the second half rather confirms the connection. And like
‘ciacona’ for chaconne, the ABB’s spelling ‘PASSACALJA’ looks like a quasi-
Italian form of a French word. The theme shares elements with all three of
Buxtehude’s themes and was less exceptional in Germany than Raison’s was
in France. To announce the theme first is unusual, though that too appears
elsewhere (Schmelzer, Violin Sonata in D, 1664), and one cannot be sure
that Buxtehude did not do likewise, whatever copies say.
The difference between passacaglia and chaconne was understood var-
iously from composer to composer. Since for Mattheson (1739 p. 233) the
passacaglia was a lively dance, ‘chaconne’ would have been a more suitable
title for BWV 582. But Walther’s Lexicon, following Brossard, describes it as
slower than a chaconne, in the minor, with a more refined ‘Melodie’ and a
less lively ‘Expression’. Specifically, it seems that for Raison and Buxtehude
passacaglias had a simple upbeat, chaconnes not, a distinction observed by
Bach in the organ Passacaglia and the violin Ciaccona.
In its sequence of note-patterns, Muffat’s Passacaglia in Apparatus (1690)
is similar: first quaver lines, then anapaests, semiquavers (rh, lh, together),
arpeggios, leaping semiquavers, and triplets. A miniature version of the
plan is also there in the F minor suite of Kuhnau (Clavierübung 1692).
Bach’s imitation from b. 24 looks like a more systematic version of a line
in Pachelbel’s F minor (b. 33), where a lighter dance still lurks. Pachelbel
too drops the bass theme at one point, dispersing its notes above, while his
D minor Chaconne anticipates not only Bach’s dactyl figures in imitation
(see Example 89) but the ‘modified repeat’ for Variation 2. Pachelbel’s
185 BWV 582
Example 89
dactyls decorate, Bach’s (in four parts) work towards a seventh and a minor
ninth.
BWV 582 is more systematic than any model, producing careful and
intensely contrapuntal four-part harmonies and avoiding the persistent
dominants of shorter ostinatos. No other composer is likely to write ten
7th-chords in his first two variations. Note-patterns are traditional, and in
some copies a slur at bb. 104ff. marks the first motif to appear on the beat.
(The reading of the slur in NBA IV/7 is surely incorrect: there is no sense
in its appearing off the beat, only on it, however ambiguous the source
might be. See KB p. 152.) Arpeggiation from b. 120 is more regular, with
two notes in each hand, than a similar one in Var. 5 of F. W. Zachow’s
‘Jesu, meine Freude’. Like the opening syncopation, the ‘obsessive figure’
from b. 153 appears more simply in Buxtehude’s Passacaglia, and both
ultimately derive from the seminal passacaglia, Frescobaldi’s ‘Cento partite’
(1615, see Example 90): it is a form of bariolage found too in an ostinato by
Weckmann (Silbiger 2001 p. 375). Like Buxtehude’s C minor Chaconne,
BWV 582 begins with a ‘painful longing’ (Spitta I p. 580), a deliberate and
typical C minor Affekt.
Example 90
186 BWV 582
Form of the Passacaglia

Kee 1992 finds the twenty-one sections (theme + twenty variations) sym-
bolic, but for b. 104 to mark the Golden Section (end of the thirteenth
section, 13 : 21) depends on there being a special break at this point. There
are others at bb. 88 and 128, and little in such number-counting is conclu-
sive. A similar proviso affects the fourteen supposed symmetrical groups
(1 + 2, 3, 4, 5, 6 + 7 + 8, 9, 10 + 11, 12, 13 + 14 + 15, 16, 17, 18, 19 + 20)
and any proposed connection to the Book of Revelation (Ouwerkerk
1995).
Two moments of tension are usually heard in the work: Var. 12, after
which there is an ‘intermezzo’ of three variations, somewhat like the trio-
intermezzo in Niedt’s model praeludium–chaconne (1720 pp. 122ff.), which
is itself like a French chaconne’s couplet; and then a rise towards the soprano
pedal point of the final two variations. Other breaks are heard at Var. 6 (first
unbroken semiquaver motion), at Var. 11 (theme leaves the bass), at Var. 16
(theme returns). Such moments as these might suggest a change of stops
or manuals, but nothing in the sources offers any hints about this, in either
Passacaglia or Fugue.
Methodical analysis of motifs, number of parts, use of pedal, position of
the theme, degree to which it is varied, tessitura, compass, possible manual
changes and other details cannot lead to one true interpretation of the form,
since the music is not dominated by any of them. Nor are there obvious
parallels between its shape and the symmetries of later works, whether the
shape is seen as axial (Wolff 1991 p. 312) or heard in internal tensions (Keller
1948 p. 96). Yet despite the work’s caprice and its resistance to dismembering,
the following schemes have been proposed:
Geiringer 1966 p. 228

1–2 3 4–5 6–7 8 9–10 11–12 13 14–15 16–17 18 19–20,
or 1–2–3–4–5 6–7–8–9–10 11–12–13–14–15 16–17–18–19–20
Vogelsänger 1972a
1–2 3–4–5 6–7–8 9–10–11–12 13–14–15 16–17–18 19–20
Klotz 1972
theme+1–2 3–4–5 6–7–8 9 10–11–12 13–14–15 16–17–18 19–20
Radulescu 1979
theme+1–2–3–4–5 6–7–8–9 10–11–12 13–14–15 16–17 18–19–20
Wolff 1991
1–2 3–4–5 6–7–8–9 10–11 12–13–14–15 16–17–18 19–20
187 BWV 582
But the only unambiguous principle of organization is the simplest: a

‘dynamic of development’, a shape formed by troughs and peaks, not a
‘symmetrical structure’ (Kobayashi 1995). ‘Pairing’ or repeating is part of
the tradition for variations; and since the second variation begins as if it is
going to be a repeat of the first, it is the sudden, beautiful seventh chord in
b. 17 that tells the player that it is not so. Buxtehude’s D minor Passacaglia
does something roughly similar, while Bach’s Violin Chaconne also begins
with repeated variations but moves on to subtler kinds of pairing.
Form of the Fugue

The ABB continues without double bar as if the Fugue were Var. 21 (see
Hill 1991 pp. 19–20), and no source authoritatively suggests a break here,
despite common assumptions. In closing b. 168 on a weak beat and rising to
the mediant, a composer of the period could not more clearly imply attacca,
senza pausa.
Using only half the theme as subject lessens its inflexible bass-like quality
and perfect cadence, both of which would be undesirable in a Fugue, whether
or not it was composed first. The new countersubject is immediately striking,
and there is a clear – one might say textbook-clear – difference between the
three subjects:
a minims-and-crotchets (from Raison)

b off-beat quavers related to the Passacaglia theme’s second half (?)
c perpetual semiquavers as in Pachelbel’s F minor or Buxtehude’s
E minor
Coupling a passacaglia with a fugue means presenting a theme in two guises,

long and short; one uses all the notes of the C harmonic minor scale, the
other just some of them as a short cantus firmus singing out from time to
time.
The three subjects a, b and c work in permutation:
169 174 181 186 192 198 209 221 234 246 256 272
S a b c c b c a a
A a b c a b a b c b b
T b c a b a c a b c
B a b c a b c a c
No permutations of themes and voices appear twice, and almost all possible
are there. Interludes and episodes are not independent, being based instead
on the countersubjects; but these episodes increase in length and complexity
188 BWV 582
as the fugue proceeds, creating a movement of broad sweep and unusually

tense continuity.
Insofar as the broad sweep has three sections – bb. 169–97, 197–250,
250–end – one might find in it the kind of tutti–solo alternation of a concerto.
But more noticeable is that the subject appears less and less often, as in
much maturer Bach fugues. Each time the subject does enter, it passes on
to different material, e.g. the countersubject’s harmonization bb. 201–3.
The theme’s rising fifth produces an initial imperfect cadence in each key,
resulting in an ideal key-plan:
168–9 tonic (with ‘fifth part’, b. 192)

197–8 relative, then its dominant (neither in the top voice – the
major 6th would jar?)
220–1 dominant–tonic–dominant
255–6 subdominant
271–2 tonic, then coda
The last twenty-two bars are amongst the most climactic in Bach: an entry
in the top voice, then a sustained sequence, a relentless pedal line, wide
texture (C-c in b. 187), a repetitive figure (b. 281), Neapolitan sixth, pause,
implied pedal point (last six bars), an added part and a ritardando (last
two). The final cadence is plagal, as it has never been in the Passacaglia
variations.
Just as the Passacaglia anticipates moments in the Ob (e.g. b. 97), so
the Fugue recalls old praeludia (bb. 217 or 237) and toccatas (compare
b. 264 with the Harpsichord Toccata in F minor, b. 67) or anticipates
later works: compare the whole coda section with the G major Prelude
BWV 541 or the semiquaver figure of b. 267 with the G minor Fugue
BWV 542. The composer of BWV 534 and 537, whoever he was, surely
remembered bb. 262 and 269–70. The Neapolitan sixth – which is no
occasion for an improvised cadenza! – is matched in BWV 532 and 535
and, complete with final six-bar pedal point, by the Fugue in A minor
WTC1.
If ever there was a work greater than the sum of its parts – a singable
theme, impeccable harmonic logic, clear pedigree, imaginative response
to other music, conscious manipulation of motifs, careful working-out of
permutation, calculated shape – it is the Passacaglia in C minor. Its ebb-
and-flow alone is hard to attribute to a young composer. So is its massive
structure, sustained by an archetypal theme matched only by two other,
much later, variation works, the Chaconne in D minor for violin and the
Goldberg Variations for harpsichord.
189 BWV 583
BWV 583 ‘Trio in D minor’

Copies in P 286 (C. P. E. Bach’s copyist Anon 300), P 1115 (? A. Kühnel
†1813); others via one line of transmission, Peters IV another (KB p. 115).
Three staves; headed in P 286 ‘Trio Adagio 2 Clav: Pedal’.
The Trio seems to belong in a miscellaneous collection of ‘35 Orgeltrio’s von

Sebastian Bach’ compiled in unknown circumstances, probably in Leipzig,
and containing questionable trios, genuine sonata movements (e.g. BWV
525.i, KB p. 58) and chorales, hence perhaps the titlepage ‘Choral Vorspiel’
in P 286 and an advertisement of 1780 (Dok III p. 296). While the form of
the Trio expressed as
A subject supplies a motif imitated in sequence; 13–17 =
1–19
1–5;
B 19–41 new subject, similar imitation, plus motif from A; two
sections (30–40 = 19–29 in dominant)
A 41–51 shortened reprise: 41–4 = 3–6, 45–51 = 7–13
Coda 51–3 inverted motif from A?
may appear to conform to genuine sonata shapes, the imitation through-

out is short-breathed and in this respect alone atypical. So are the non-
thematic opening bass and near-infelicities in the grammar (near parallel
5ths in bb. 12, 15–16 and unisons in 22, cross-relation in 48, etc.). All
themes are answered at the half-bar, even when the lines are extended (e.g.
bb. 26ff.). Such sequences and imitation above a moving bass line as those of
bb. 19ff. are found in the Six Sonatas only in secondary material, as in the
first movement of No. 3, bb. 24ff.
The short phrases resemble French trio-writing and are surely the work
of a composer familiar with the G minor Fugue BWV 542: compare b. 1
with its theme, b. 24 with its episode (b. 39), and b. 24 bass with its
pedal. Moments of trio-writing in this Fugue, at bb. 26, 37, 55, 73 or 103,
could also have been an inspiration for the Trio. The result is close to the
Six Sonatas, as comparisons show (e.g. bb. 39–40 with bb. 3–4 of Sonata
No. 2, first movement), and motifs are handled just as ingeniously, as when
the opening is decorated. The coda, which is not strictly necessary, could
equally well become an imperfect cadence, as in the Sonata No. 2, slow
movement. However, juxtaposing one subject with another is not so well
done (b. 41), nor is the linking effortless (b. 45). A further sign of the work’s
doubtful provenance is that though marked ‘Adagio’, the material would
equally well suit ‘Allegro’.
190 BWV 583–585
P 1115 also contains the trio on ‘Allein Gott’ BWV 664a, but while
the opening motif of BWV 583 appears in the hymn ‘Hier lief’ ich nun’
BWV 519 (twice in the first three bars), the Trio has no obvious chorale-
melody. Signs that perhaps a gifted pupil was responsible for it are the
counterpoint of such bars as 8, 12, 46 and 49–50, the sequences, the unusual
form of Neapolitan 6th in b. 52, and the ornaments (of C. P. E. Bach’s
period?) in a piece of mixed genre. Possibilities are that (i) it is a transcribed
chamber trio, or (ii) an embroidery of ideas prompted by the G minor
Fugue.
BWV 584 Trio in G minor

No Autograph MS; nineteenth-century copies only.
This is a version of the first section of a 78-bar ABA aria in Cantata 166
(1724):
right hand the oboe part

pedal basso continuo part
left hand some shared material with tenor part, but mostly
different
While it was once thought that the trio is the earlier of the two versions
(Oppel BJ 1909 pp. 27–40), more likely is that the original was neither of
these but rather a lost aria with two obbligato instruments. Since some
thematic references are missing in BWV 584, this was probably not made
by Bach himself (Dürr NBA I/12 KB pp. 18–20).
BWV 585 Trio in C minor

A lost ‘original MS’ of J. S. Bach? (see BJ 1993 p. 72); copies in Lpz MB
MS 7 (J. N. Mempell, c. 1730/40?), and a late Lüneburg MS also containing
BWV 587.
Title ‘Trio. ex. C mol. di Bach’ in MS 7, and the movements reversed.
Trios copied in Leipzig MS 7 – BWV 585, 586, 1027a – may have been part of
a bigger collection of chamber trios, including one in G major by Locatelli
(Schulze 1984 p. 78). It follows J. L. Krebs’s trio-plan of a pair of movements,
although if comparison with the Trio Anh.II 46 is justified (Keller 1948
191 BWV 585–586
p. 58), the composer would rather be J. T. Krebs (Tittel 1966 pp. 126–9). In
1973 H.-J. Schulze showed that it seems to be an arrangement of the first two
movements of a Sonata in C minor for two violins and continuo, preserved in
parts in a Dresden MS and attributed to J. F. Fasch (1688–1758), a competitor
for the Leipzig cantorate in 1722. Various grammatical faults in MS 7 do not
‘speak conclusively against Bach’s authorship of the transcription’ (Schulze
1974 p. 4), and the lost copy may have been closer to Fasch’s Dresden parts
than BWV 585 as now known.
BWV 585 and the Six Sonatas share a certain melos in the present
Allegro’s interplay of parts, here rather short-breathed. But the Adagio sub-
ject is long, the movement does not develop in proportion, the Allegro
subject has a unison answer, and the pedal plays an on-beat basso continuo,
none of which is typical of the Sonatas. Its neo-galant style implies a date
later than Fasch’s activities with the Leipzig collegium musicum in the years
up to 1710: perhaps J. L. Krebs and his teacher worked on it around the time
the Six Sonatas were being compiled?
BWV 586 Trio in G major

Copy in Lpz MB MS 7 (J. N. Mempell, c. 1730/40?); and Körner’s edition in
1850.
Headed in MS 7, ‘Trio. ex G.. 2. Clavier et Pedal. di J. S. Bach’.
Reported on by Seiffert in Peters Jahrbuch 1904, the movement was taken

into the 1904 edition of Peters IX. Later, in MuK 1942 pp. 47ff., K. Anton
claimed that it was a work of G. P. Telemann, ‘arranged by Bach’ from a
harpsichord piece or its theme (Siegele 1975 p. 76). The transcriber of the
trios BWV 585, 586, 1027a in MS 7 is thought to be J. N. Mempell (Schulze
1974), and various commentators have made attributions to possible Bach
pupils (see KB p. 90).
Not conforming in detail to the binary form familiar in Bach’s chamber
and organ sonatas, the movement plays with its themes, and works towards
cadences in various keys, in a manner typical of movements in Telemann’s
Musique de Table (1733). Perhaps BWV 586 was an entirely new compo-
sition – not by J. S. Bach – based on a theme of Telemann (Schulze 1973
pp. 150, 154), more sustained than an aria in Telemann’s Kleine Kammer-
musik of 1716, whose theme it resembles somewhat. In its simple imitation,
parallel thirds, basso continuo patterns, use of binary Allegro without con-
trast between subjects, it has more in common with BWV 587 – including
pedal above d – than with the Sonatas.
192 BWV 587–588
BWV 587 ‘Aria in F major’

Only source, a lost MS in Griepenkerl’s possession, used in Peters IX (1881)
and copied in a Lüneburg MS containing also BWV 585 (KB p. 79).
Headed ‘Aria’ (no known attribution to J. S. Bach).
This is an almost literal transcription, but without articulation signs and

some ornaments, of section 4 of ‘L’Impériale’, the first of ten movements
in François Couperin’s ‘Troisième Ordre’ for two violins and continuo
in Les Nations, sonades et suites (Paris, 1726), and headed not ‘Aria’ but
‘Légérement’. Bars 75–90 of BWV 587 do not appear in this print. Since, like
other sonatas in Les Nations, ‘L’Impériale’ had probably circulated in MSS
for as many as thirty years before publication in 1726, the source for and
date of the original transcription are as uncertain as its authorship. There is
no evidence that the movement was an interlude between the Toccata and
Fugue in F, suggested in Klotz 1950 p. 202 – all in F major!
The details of thematic development in such a well-constructed ABA
movement would have interested a player of BWV 527. However, Légérement
suggests a lively tempo far more in keeping with carefully articulated string
parts than with organ music. Curiously, this fourth section is the least con-
trapuntally imitative of Couperin’s original movement: a lively interlude
only.
BWV 588 Canzona in D minor

No Autograph MS; copies in BB 40644 (Mö MS, last sixteen bars only, J. C.
Bach 1705/6?) and derivatives (Lpz MB MS 7, J. G. Preller 1740s?, or via
J. C. Kittel, e.g. P 320); others from a revised autograph (?) probably via
C. P. E. Bach (P 204 C. F. G. Schwenke, and derivatives).
Two staves, no pedal cues; title in Kittel, ‘Canzona ex D a 4’ (first pages

missing in Mö MS). ‘Adagio’ for last two beats (?) in e.g. MS 7 and P 204
(but not Mö MS).
Why Peters IV and subsequent editions include BWV 588 among the organ
works, and why it is often played lugubriously, is not clear. The once-famous
‘opening pedal theme’ is not authorized by the sources; nor even at the
pedal-point and cadences is pedal necessary, though it appears to be so now
and then (bb. 54, 62?, 115?). The ornaments in MS 7 (KB p. 150) look like
copyist’s conjecture, contradicting the italianate counterpoint, its cantabile
193 BWV 588
and its tempo, and thus unlikely to be the result of lessons with Bach (as
suggested in KB p. 174). As harpsichord ornaments, they rather resem-
ble Gerber’s additions to the Inventions, also unjustifiably given the NBA
imprimatur.
Pirro heard a similar theme in a ‘Canzon dopo la pistola’ of Frescobaldi’s
Fiori musicali, 1635, Bach’s copy of which is dated by him 1714 (Dok I
p. 269). This is much too late for BWV 588, however, which belongs with
the Fantasias BWV 570 and 563 amongst the composer’s early genre-essays
in ABB and Mö MS, whose source for it was probably in tablature. Maturer
versions of the idiom can be heard in the D minor Fugue BWV 538, and
a similar subject appears in the opening movement of Cantata 25 of 1723
(see there b. 59). Repeated notes were typical of canzona themes, including
Frescobaldi’s double subject in Example 91. Repeated notes are also promi-
Example 91
nent in 3/2 sections, as well as in later German canzonas (e.g. Scheidemann’s

in G major). Such alla breve features as the dactyls and the continuity over
bb. 35–40 are vocal-melodic and more like ricercars than canzonas. For such
composers as Buxtehude ‘canzona’ always indicates a lively piece, and his
G minor Praeludium BuxWV 148, third section, has a somewhat similar
theme.
Thematic metamorphosis and combinations, i.e. italianate techniques
known to composers admired by J. S. Bach including the Leipziger N. A.
Strunk, are found in influential publications such as Krieger’s Clavier-Übung
of 1698. Krieger called his ‘ricercar’, so perhaps ‘canzona’ for BWV 588
comes from Frescobaldi sources circulating in c. 1700. Its plan is unusually
straightforward:
A 1–70 exposition, episode, exposition (octave answer); to

dominant
B 71–114 irregular exposition (octave answers); episode
114–40 second series of entries; episode, E minor to G minor
140–62 third series of entries; episode
162–9 final entry; ‘Italian’ Adagio close
No entries are in the relative or in any other major key. In B, where it is

convenient to add brighter stops, entries easily extend to episodes while A’s
194 BWV 588–589
entries are strict. Both have a chromatic countersubject as elsewhere in early

Bach: see Example 92. A falling chromatic fourth was associated with fugues
of the ricercar type, either as subject or countersubject (second ‘Christe’ of
‘Messa delli Apostoli’, Fiori musicali); at b. 111 it rises, as it does in Fiori
musicali but now with much greater tension.
Example 92
Full, new countersubjects to both this and the main subject are constantly
being produced, and in this respect B is rather less inventive than A, although
Spitta saw B’s part-writing as ‘bolder’ (I p. 420) – presumably because of the
episode that includes both a d and an e. The part-writing is so strict,
with each voice having the theme in turn, that the work could be laid
out in open score (Breig 1999 p. 636): perhaps it was composed less as a
keyboard piece than as an essay in the counterpoint of a particular italianate
genre.
Not only are ricercar elements mingled with canzona but the double-
subject section in 3/2 is like the third section of older canzonas, such as
Froberger’s Canzona II copied in Leipzig MB MS 51. It is possible to see
the piece as a lively canzona, with both cadences (particularly the link
between sections A and B) more dramatic than in the sectional canzonas of
Frescobaldi or even Buxtehude. Were it ever possible to show the ‘Adagio’
sign to be authentic, one could see the close – a drawn-out 5/4 chord, a long
trill and a long final, like a Sonata for Solo Violin – as specifically Italian in
style, more like endings in Frescobaldi, Corelli or Handel than Buxtehude
or Bach himself, which are almost always more succinct.
BWV 589 Allabreve in D major

No Autograph MS; copies in P 1106 (1740s?, but not a close copy of a Bach
autograph: KB p. 159) and ultimate derivatives.
Two staves; title in P 1106 ‘Allabreve con Pedale pro Organo pleno’.
195 BWV 589
The ‘ricercar-like and vocal-melodic’ nature of Italian alla breve counter-

point is even clearer here than in the Canzona in D minor. Some character-
istics of it are:
2/2 or 4/2 signature, mostly minims and crotchets

quasi-double subject
lines moving largely by step, but some conspicuous leaps (thirds,
fourths)
frequent minim suspensions (at least once every four beats)
characteristic stepwise crotchet lines (but no quaver dactyls)
a ‘singing style’ as in motets rather than cantatas
In such counterpoint the lines are not independent but only pretending to
be so, and planned to counter each other: as one rises the other falls, as one
moves the other is suspended, as one proceeds by step the other proceeds
by leap. The genre allows variety of tempo (slower in the ‘Gratias agimus’,
B minor Mass), figure (quaver dactyls in Goldberg Variation 22), subject
(longer in BWV 538), and consequently Affekt.
The uniquely high tessitura of the opening suggests string music, like
an Allegro in Corelli’s Concerto Op. 6 No. 1 (Keller 1948 p. 72), which
circulated long before its publication in 1712. But the idiom is not rare in
keyboard music, northern or southern. See Example 93.
Example 93
Spitta heard in it a ‘distant relation’ to the D major Prelude, alla breve

section (KB p. 161); Breig noticed a marked similarity between its final pedal
point and that of the first fugue of WTC1 (1999 p. 638); and in Graupner’s
cantata ‘Uns ist ein Kind geboren’ (1712), a similar theme in stretto produces
196 BWV 589–590
similar counterpoint below a cantus firmus of ‘Vom Himmel hoch’. Clearly,

there is a distinct type here.
1–37 tonic paragraph

37–90 entries and episodes towards relative minor
90–158/9 two sets of tonic entries, episodes
158/9–197 final tonic entries, stretto at one bar (174/5); chromatic
preparation for closing pedal point
One can view the divisions differently, since there are several episodes be-
fore each striking pedal entry. The subject not only can stretch across keys
(bb. 32–46) but allows stretto at one bar (fourth below), two bars (fourth
above or fifth below) and three (octave or third below), sometimes doubled.
The counterpoint flows effortlessly thanks to the simple diatonic steps of
the subject, and the original countersubject had already dropped out before
b. 37. Formulae include the falling chromatic fourth in D, which runs into
a Neapolitan sixth (bb. 180–5), and the upper theme, which paraphrases a
transposed natural hexachord (DEFGAB).
An effective entry each time is prepared by a rest that barely breaks the
work’s extraordinary continuity, a continuity typical of the ricercar-fugue
but far from Bach’s sectional, mature organ fugues. The non-structural use
of returning tonics and the array of subject-entries are also ‘early’ signs.
Yet the facility in manipulating motifs is already advanced (see bb. 57–9 or
bb. 111–18) and exceeds that of contemporaries, whose alla breve idiom was
never so on-driving as this. The many tonics work against the aimlessness
that easily arises in this idiom. As in the middle section of the (contempo-
rary?) Pièce d’Orgue, there seems no reason why this effortless counterpoint
should not go on and on.
BWV 590 Pastorella in F major

No Autograph MS; complete in P 287 (J. P. Kellner after 1727?), also via
C. P. E. Bach (P 290, P 277?, Am.B.59?) and lost MS used in Peters I; first
movement only in copies via Kittel (?).
Two staves; headed in P 287 ‘Pastorella pro Organo di Johann Sebastian

Bach’, in the Peters source probably ‘Pastorale’. No movement headings.
In plan and detail BWV 590 resembles no other organ work or keyboard
suite, and yet each movement can be shown to have features of one Bach
idiom or another, quite late in the case of the two middle movements. The
197 BWV 590
sequence of keys, unique in Bach, suggests a quasi-Italian sonata compiled

(by whom?) from movements of disparate origin (but genuine?). The main
sources transmit them together, even if in performance they are separated
like Magnificat versets (Keller 1948 p. 76).
It is possible that the whole work was composed/compiled for some
unknown occasion – but also that movements 2, 3, 4 have nothing to
do with the first (Spitta II p. 692), to which alone the title ‘Pastorella’
applies, whatever ‘ingenious synthesis’ the whole work might be said to
achieve (Stauffer 1983 p. 14) and however late its compilation (Stinson 1990
pp. 110ff.). The ‘Toccata sesta’ in F major in Muffat’s Apparatus (1690) has
a series of movements featuring toccata pedal points and finally a 12/8
fugue, and traditional organ pastorales encompassed several movements,
from Frescobaldi’s ‘Capriccio pastorale’ (Toccate, 1637) to Zipoli’s ‘Pastorale’
(A Third Collection, London, c. 1722), which has a shape A1BA2. Since there
is no early copy of movements 2–4 as a group, perhaps they were added to
a pastorale movement, ‘invited’ there by its mediant close, much as the
incomplete Fantasia BWV 573 may also have invited continuation.
Each movement subtly incorporates a pastoral drone: the second with
two held bass notes, the third a repeated bass, the fourth a fugue subject
circumscribing a tonic pedal point. And each has a dominant ‘answer’ to an
opening tonic phrase, bb. 11, 9, 25, 4 (and 25) respectively. (It is this domi-
nant answer that gives some other toccatas a superficial resemblance to the
Pastorale, e.g. Pachelbel’s Toccata in F.) Unified in one respect, movements
with and without pedal might well be grouped together, as in the Chorale
Variations BWV 768. If it could ever be shown that in its present form BWV
590 is authentic, it would be a unique imitation, contrapuntally worked, of
four Italian genres: pastorale, allemanda, aria, giga.
First movement
Kellner’s MS has empty staves for about twenty bars more before the next
movement (KB p. 180), leaving an open question whether it was completed
elsewhere or he thought it should be. While such Italian figures as b. 10
can be found in Handel’s Messiah pastorale (very likely inspired by arias of
Alessandro Scarlatti), the melos and modulations are surely Bach’s. Note
that the tonic could return in b. 27.
The dominant ‘answer’ is as in other pastorales (e.g. Corelli’s Concerto
Op. 6 No. 8) and continues with familiar motifs: compare bb. 25–6 lh with
the Ob’s ‘In dulci jubilo’. Compound-time figuration produces similarities,
so that b. 5 is not unlike b. 5 in the G major Prelude, WTC1. The chromatic
motif in b. 28 is also in keeping with Italian pastorales such as Zipoli’s,
where these tones allude to the dubious intonation of bagpipe-players
(e.g. in Zipoli’s Pastorale). The dominant seventh sequence of bb. 21ff.
198 BWV 590
is more typical of Bach: compare bb. 33ff. of the Pastoral Symphony in the
Christmas Oratorio. Like pastorales of Corelli, Locatelli (Op. 1) and others,
the movement lacks the dotted siciliano rhythm often found in latter-day
12/8 pastorales. (Locatelli’s Concerto in F minor Op. 1 No. 8 was known to
J. S. Bach probably by c. 1734/5 – see Beisswenger 1992 pp. 302f.) Smooth
12/8 figures on a pedal point produce quasi-pastoral idioms both in can-
tatas (for Jesus the Shepherd in Cantata 104.v) and in toccatas, especially in
F major (BuxWV 156).
To judge by such sonatas as the E major Violin or G major Gamba,
mediant cadences lead to further music a third down. So possibly a da capo
was intended, as in Corelli’s Pastorale, with a Fine somewhere around b. 20.
(In the F major Prelude BWV 556, a mediant close is followed by a da capo –
but whose work is it?) This mediant close may have been an italianate feature
in F major, one found again in Domenico Scarlatti’s Sonatas Kk 366 and 518,
Handel’s first ‘Piva’ for Messiah, and the opening Adagio of the Suite HWV
427. The same F–A mediant close in an allemande grave of Louis Couperin,
and in a textbook demonstration by Thomas Mace (Musick’s Monument,
1676, p. 143), suggests that it was an old idea, specific to these two notes of
F and A and to movements of gentle tempo – see also the chorale-fughetta
BWV 704.
Second movement
Although BG 38 likened this to an allemande, there is no up-beat, nor are
long-held bass notes usual. Yet the part-writing is allemande-like (cf. the G
major French Suite), and bb. 15–16 also resemble moments in suites or con-
certos. Since some early allemandes also have no upbeat (Chambonnières,
1670), knowledgeable copyists could have been uncertain quite what this
was, hence their time-signature of C rather than c. On the other hand, the
main cadences are as melodious as a violin solo in a cantata aria – something
like ‘Unerforschlich ist die Weise’ in Cantata 188 (1728). The question-
and-answer phraseology of bb. 19–20 is mature, while the broken chord
figures resemble some in manualiter settings of Clavierübung III.
Melody and counterpoint are Bach-like. While not many second halves
both begin like the first and include a shortened recapitulation in the tonic –
it is usually one or the other – a further example is the Sarabande of the
C minor French Suite.
Third movement
The shape broadly resembles such sonata movements as the Largo of the
F minor Violin Sonata, i.e. a melody rather improvisatory and expansive
in character is followed by a section leading to an imperfect (phrygian)
199 BWV 590–591
cadence. In general, the texture and melody again resemble an aria with
violin obbligato, or perhaps the middle movement of a harpsichord con-
certo, hard to ascribe to anyone but J. S. Bach. It is possible to discern the
notes of the second movement’s melody in the third’s, now in the minor
and extravagantly paraphrased. Two manuals are optional.
Fourth movement
The finale has more in common with older gigues – exposition, sequences
and entries, then inverted subject, modulations, final subject – than with
any other kind of movement, despite superficial resemblances elsewhere
(e.g. Third Brandenburg Concerto, finale). Even more than in the second
movement, the texture seems to call for harpsichord: compare the low tes-
situra opening the second half with the Gigue of the A minor Partita. It is
possible to see the triadic contours of the theme as related to the pastoral
motifs of the first movement. As with other links between the movements
already mentioned, had they been grouped together on these grounds by an
observant and musical copyist, why are the movements not found elsewhere?
BWV 591 Kleines harmonisches Labyrinth

Copies in P 1107 (later eighteenth century); several late MSS, including
Viennese.
Two staves; headed in P 1107 ‘Kleines harmonisches Labyrünth. Joh: Seb:

Bach’.
Only the word ‘Ped’ in P 1107 eight bars from the end – to denote a pedal
point? – justified Peters IX in including the piece amongst the organ works.
Since movements incorporating chromatic and enharmonic devices in-
terested such composers as Heinichen, Sorge and Kirnberger, BWV 591
has long been associated with one or another of these, in particular J. D.
Heinichen (Bartels 2001). The term ‘labyrinth’ appears also on the title-page
of Fischer’s Ariadne musica (c. 1702), though there was yet no question of
using all the keys. ‘Le Labyrinthe’ in Marin Marais’s Pièces de Viole Book IV
(Paris 1717) is a rondo in which the main theme returns in different keys,
beginning and ending in A major. Heinichen (1728 pp. 850ff.) gave several
examples of two-part pieces passing through twenty-two keys, while in the
same year and area of Germany as WTC1, Friedrich Suppig’s Labyrinthus
musicus (1722) contains a ‘Fantasia through all twenty-four keys’ which
‘could be played on the harpsichord without pedal or on the organ with’.
Suppig’s dedication refers to Kuhnau, Vetter and Buttstedt (see Rasch 1984)
200 BWV 591
and must indicate local interests. Locatelli’s ‘Laberinto armonico’ in L’arte

del violino, Op. 3 (1733), exploits no harmonic complexity but is an exer-
cise in violin technique, its motto ‘facilis aditus difficilis exitus’ a curious
reminder of the last section of BWV 591.
Mozart possessed a copy of BWV 591 also attributed to J. S. Bach
(Dok III pp. 512–13), the only name in the copies. And indeed the influence
of J. S. Bach can be glimpsed: the appoggiatura chords after the arpeggios
recall the Chromatic Fantasia; the fugue subject is somewhat like that of the
B minor fugue WTC1 and the doubtful B BWV 898; the part-writing in the
Exitus is Bach-like. But the programme – ouverture, lost direction, entry
into labyrinth, discovery of C major, exit beneath the ‘sun of clear harmony’
(Keller 1948 p. 57) – scarcely proves authorship, any more than a symme-
try in the bar-numbers does. Such competent harmonic progressions as
bb. 38–41 could result from familiarity with Bach keyboard idioms.
Despite the sources’ agreement on authorship (Bartels 2001), difficult
to attribute to Bach are such flaccid moments as the final pedal point, the
close, and the fugal working, which is little more than a set of harmo-
nized statements. Like the retrograde movement halfway through the
Fugue, the symmetry of prelude–fugue–postlude is simple and rather at
variance with the complex symmetry of e.g. the E Fugue in Clavierübung III.
The B A C H spelt out towards the end is, if anything, more a salute than
a sign of authorship, and in no way can fanciful, interdisciplinary explo-
rations of the labyrinth metaphor in and out of music ‘show that BWV 591
originated with Bach’ (Wright BJ 2000, p. 51).
Concertos BWV 592–596
No complete Autograph MS or copy.
Sources
It is not known whether the concertos were ever collected as a set, either
by the composer (like the ensemble harpsichord concerto transcriptions
BWV 1052–1059) or by a copyist (like groups of solo harpsichord concerto
transcriptions within BWV 972–987). Speaking against it are that individual
extant copies are varied, that harpsichord versions of two concertos appear
in separate MSS, and that when sources are very alike, as for BWV 593 and
594, they are still discrete copies.
The autograph MS of the D minor Organ Concerto BWV 596 is by
far the oldest extant copy of any concerto, and for that reason alone there
are likely to have once been more than the present five concertos, known
mostly from copies of the Leipzig period. Probably all five plus the harpsi-
chord concerto-transcriptions once existed in Kellner’s copies made before
any Leipzig revision, but it is only conjectural that all such were based on
earlier autographs or copies made in Weimar c. 1714. Since by 1709 Bach
knew of at least one Albinoni concerto (Beisswenger 1992 p. 226) and had
personal contact with German composers of concertos (e.g. Pisendel, in
Dok III p. 189), perhaps there had been a series of other transcriptions now
lost.
Transcriptions of Vivaldi’s Op. 3, including BWV 593 and 596, were
probably based on the Amsterdam print of 1711, while Opp. 4 and 7 were not
yet printed (KB pp. 13–14). The similarity of many details between Prince
Johann Ernst’s concerto BWV 592 (q.v.) and a Concerto in G from Vivaldi’s
Op. 7 (No. 8, RV 299) – ritornello, texture, figuration, bass-line, time-
signature, even a final scale-run – suggests that however Op. 7 was acquired,
it circulated among the Weimar musicians. Other German transcriptions
from Vivaldi’s Op. 3 are found in various sources, like Bach’s probably based
on the first edition. This was a set of eight parts, to score up which must
have been the first task of a transcriber.
Origin
[201] One explanation of the five extant concertos is given in Schulze 1972 p. 10:
202 Concertos
Despite the complex picture given by the sources, Bach’s organ and
harpsichord transcriptions BWV 592–596 and 972–987 belong to the year
July 1713 to July 1714, were made at the request of Prince Johann Ernst
von Sachsen-Weimar, and imply a definite connection with the concert
repertory played in Weimar and enlarged by the Prince’s recent purchases
of music. Since the court concerts gave Bach an opportunity to know the
works in their original form, the transcriptions are not so much
study-works as practical versions and virtuoso ‘commissioned’ music.
The young prince (1696–1715) was at Utrecht from February 1711 to July
1713, visited Amsterdam and sent Italian music back to Weimar, where the
organist at the town church, J. G. Walther, gave him lessons in composi-
tion. Walther also claimed later that he himself transcribed no fewer than
seventy-eight concertos (Schulze 1972 p. 12), many no doubt considerably
elaborated. In the first instance the point must have been to make a short
score on two staves (Klavierauszug), more easily playable than an open score
sufficient for study purposes.
Yet it is difficult to imagine all the transcriptions being made within a
twelve-month period. The taste did not suddenly appear in 1713 – perhaps
the prince knew some Vivaldi already, and had already played and worked
on concertos with Walther? – and is unlikely to have quickly disappeared.
While the prince’s departure in July 1714 and untimely death in August
1715 might have ended a particular call for transcriptions, those of his
own concertos could have been ‘in memoriam creations’ (KB p. 14); and
all of them could have much the same purposes as other virtuoso music
such as the D minor Toccata BWV 538. Perhaps the so-called harpsichord
transcriptions, being more ‘neutral’, were the first to be made and were then
adapted for organ and its more specific requirements (pedal, two manuals),
and perhaps more organ versions were made than are now extant or known
to have been made. Even if a newspaper report of Bach playing ‘diversen
Concerten’ in Dresden in September 1725 is unlikely to mean anything
as specific as transcriptions, much less ensemble works (pace Wolff 2000
p. 318), concertos need not have been associated so exclusively with Prince
Johann Ernst during 1713–14, and some may well belong around the time
of the Dresden visit of 1717 or to the years after Weimar.
That Vivaldi’s concertos made a huge impression on musicians of Saxony
and Thuringia in c. 1714 was confirmed later by Quantz (conversation with
him reported by Charles Burney, in Scholes 1959 II p. 185), and surely Forkel
was not entirely wrong to suppose them instructive in matters of form. From
them Bach learnt
dass Ordnung, Zusammenhang und Verhältniss in die Gedanken gebracht
werden müsse, und dass man zur Erreichung solcher Zwecke irgend eine
Art von Anleitung bedürfe. (1802 pp. 23–4)
203 Concertos
that order, continuity and proportion must be brought to bear on ideas,

and that to such an end some kind of guide [such as Vivaldi] was necessary.
Forkel has been criticized for oversimplifying the situation, and he only
guessed in saying that Bach had transcribed Vivaldi’s concertos ‘complete’.
His remarks suggest inspired conjecture, and they do not explain why Bach
transcribed the prince’s own works, or how, if ‘order, continuity and pro-
portion’ came to him only from Vivaldi, he could have produced the quasi-
ritornello of Cantata 196.iv by 1708 or so. Furthermore, a quasi-ritornello
was already familiar in fugues, a more straightforward ritornello form in-
deed than is found in the C major Concerto BWV 594. Also, instructive as
the details of form in BWV 593 were, so was the counterpoint itself in the
case of BWV 596.
The claim that the concertos are Communion music, on the analogy of
instrumental pieces at the Elevation, is conjectural; so too is the idea that
they were in some sense ‘commissioned’, though this ties in more closely
with what is known about musical life at the Weimar court. In April 1713 a
Bach pupil, P. D. Kräuter, asked his school board for further leave to study
in Weimar because the prince,
welcher . . . selbst eine unvergleichliche Violin spilen soll, nach Ostern
aus Holland nach Weimar kommen u. den Sommer über da verbleiben
wird, kunte also noch manche schöne Italienische und Frantzösische Music
hören, welches mir dann absonderlich in Componirung der Concerten u.
Ouverturen sehr profitabel seyn würde . . . Nun weiss ich auch, dass Hr.
Bach nach Verfertigung dieser neuen Orgel in Weimar absonderlich
anfänglich gwiss unvergleichliche Sachen darauf spilen wird . . .
(Dok III pp. 650)
who himself plays the violin incomparably, will return to Weimar from
Holland after Easter and spend the summer here; I could then hear much
fine Italian and French music, which would be particularly profitable to
me in composing concertos and ouvertures . . . I know too that when the
new organ in Weimar is ready Herr Bach will certainly play incomparable
things on it, especially at first . . .
The court organist’s study of styles and forms explains his interest in con-
certos, in which sources imply he kept up an interest throughout the Leipzig
period.
Style and influence

According to Forkel 1802 p. 24, more or less echoed by most later writers,
Bach learnt from such concertos of Vivaldi how to develop ideas (‘Führung
der Gedanken’) and how to think musically without waiting for ideas to
204 Concertos
come from the player’s fingers (‘auch musikalisch denken, so dass er . . .

nicht mehr von seinen Fingern zu erwarten brauchte’). But Forkel’s notion of
‘musical ideas’ belongs to a conception of the composer as poet rather than
creator of a ritornello form already explored in the Toccata in C. The main
theme of the Concerto in G’s first movement and the contrast between it and
the episodes sustain a movement of comparable length to the Toccata, and
the shape of both reflects the content. The short motifs of the Toccata should
not disguise their skilful development, so it is arguable which movement
has the ‘better-developed’ form. On the other hand, since the concertos for
organ as now known were more consistently up to date than those arranged
for harpsichord, perhaps it was indeed the newer ritornello shapes that were
of most interest.
Ritornello forms in J. S. Bach’s sonatas, preludes and fugues follow their
own line of development, seldom clearly based on, derived from, or even
paralleled by particular movements of Vivaldi. The concerto transcriptions
remain somewhat isolated. In this respect BWV 592 is interesting, since
it presents a (minor) German composer’s idea of Italian ritornello form:
simple, clear, less whimsical, more controlled than a Vivaldi first movement,
which stands or falls by the strengths of its caprice. From such a simple
ritornello idea as that of BWV 592.i – and not directly from Vivaldi? –
would develop the first movement of the G major Organ Sonata, despite
claims that this was composed from a ‘Vivaldi data-base’ (see p. 34 above).
Frequently mentioned concerto elements in the greater organ preludes,
such as new material after the opening exposition, are characteristic of many
kinds of music, too many for one to trace easily any direct influence of the
Vivaldi transcriptions. More instructive are the partial returns of the main
subject in the A minor Concerto, which somewhat resemble partial returns
in the C minor Prelude BWV 546. But in general, ritornello form seems
to arise naturally from certain material, and with the first movement of
Brandenburg Concertos Nos. 3 or 4 a highly intricate version had evolved,
perhaps prompted – but no more than that? – by Vivaldi’s ‘many-sided use
of motif ’ and ‘tendency to thematic contrast’ (Eller 1958).
Probably the transcriptions did introduce new figurations, which were
still surviving in the Goldberg Variations (1741): Example 94. Others in-
clude fast repeated pedal notes and ritornello octaves. Max Seiffert noted
Example 94
205 BWV 592
that Walther ‘remains true to the original’ (DDT 26/27 p. xxi), but both
composers produce textures uncharacteristic of their other organ music:
see Example 95. Conventional violin sequences are by nature alien to key-
board instruments, which have less of a natural vivacity to sustain interest.
Hence sequences in Italian violin-music can be more predictable than key-
board sequences of a Bruhns or Buxtehude. But Italian sequences were
certainly circulating amongst German organists at least by 1713, as is clear
from fugue-episodes in Buttstedt’s Clavier-Kunst (Leipzig).
Example 95
On the whole, the organ is used most originally in the episodes. Because
a solo or duo concerto is more likely than a concerto grosso to have such
finger-music developed at length in episodes, the Vivaldi concertos stand
out from the concerti grossi transcribed by Bach and Walther. It must be
for such passage-work as BWV 594’s that the concertos have often attracted
adverse criticism, particularly amongst older German editors (see Tagliavini
1986 p. 241) and their English followers (‘not much of musical value’: Grace
c. 1922 p. 248). But the opening paragraph of BWV 593 in particular would
have taught any transcriber a lot, its quasi-homophony good for strings but
rather clumsy for keyboard. Neither in Bach’s nor in Walther’s harpsichord
transcriptions is there another such paragraph.
As for the idiomatic use of two manuals: Vivaldi laid out schemes of forte,
piano and in particular pianissimo that do not appear in the organ transcrip-
tions. This is surprising since any pp in the A minor and D minor concertos
would not require outlandish ingenuity. The many and changing choruses
formed by string ensembles are suggested by mere blanket Oberwerk/Positiv
directions which, whether or not authentic, offer simple contrast, in texture
rather than dynamic.
BWV 592 Concerto in G major

No Autograph MS; copies in P 280 (plus BWV 972–982, J. B. Bach 1715 or
later: BJ 2000 p. 312), Lpz MB MS 11 (‘1739’), P 804 (Kellner, by 1725?);
206 BWV 592
later from a common source via J. C. Kittel (P 320) or C. P. E. Bach (? C. F.

G. Schwenke).
Headed in P 280 ‘Concerto à 2 Clav. et Ped.’, in Lpz MB MS 11 ‘Concerto. di

Giov. Ernest: appropriato. all’Organo. di Joh: Seb: Bach:’. Second movement
‘Grave’ in P 280, ‘Adagio’ in P 804; third, ‘Presto’ in P 280. In the string
version, ‘Allegro assai’, ‘Adagio’ and ‘Presto è staccato’. Manual indications
‘O’ and ‘R’ in P 280.
MS parts of the string version include a continuo part headed ‘Concerto

a 6 Violini e Violoncello col Basso per 1’organo’ (KB p. 64): the scoring
is principal violin, two obbligato violins, two ripieno violins, viola, cello,
figured bass. The paper used is also found in Bach works of 1714–16, and the
copyists worked on several Weimar cantatas (Schulze 1984 p. 166). Unlike
some other arrangements of Johann Ernst’s concertos, this is not one of
those published as a set by Telemann in 1718. ‘Appropriato’ is the term
used by Walther for his transcriptions, ‘accomodato’ by J. F. Agricola for the
four-harpsichord version of a Vivaldi concerto, BWV 1065.
The three-movement plan, with ritornello outer movements and a lyrical
slow middle, is the main type both of Italian concertos and of arrangements
by Walther and by Bach. Insofar as the extant string parts in KB pp. 105–22
do transmit the model Bach worked from, they show him ‘improving’ it
more than he did Vivaldi’s.
First movement
In texture, rhythm, manual-changes and key, the ritornello principle here
is more patent than in so many Italian concertos. The change of manual is
managed without inconvenience or disrupted phrases, the tutti/solo con-
trast is simple. The upper pedal part is not necessary to the harmony and is
mostly omitted in P 804.
Both subjects are open to development, and it says much for the quality
of the material that such sections (bb. 73ff., 121ff.) keep up interest through
repetition, sequence and many perfect cadences. Although the unusual tex-
ture of the sequences at bb. 5, 26, 74, 113, 121 and 123 might suggest a
concerto for two violins – with the second violin an octave lower? – the
parts show that this was not the case. In fact, Bach omits little imitations in
the tuttis (bb. 5ff.) or ignores other possible imitations (e.g. bb. 144–5). In
addition, the possibility of string crescendos in the final ritornello section is
lost, as are potential antiphonal effects in the sequence from b. 74 onwards.
The organ transcription therefore appears to lose much of the original.
Consequently, for a keyboard arrangement without much dynamic nu-
ance extra figures have been introduced, notably the semiquavers of bb. 38ff.
207 BWV 592
and the striding bass of bb. 48ff.; and a line depending originally on the vio-
lin’s lyricism has been made more ‘interesting’: Example 96. The bracketed
bar seems to be an addition, giving more momentum, as does a busier bass-
line in solo episodes. At the same time, the opening melody-with-harmony
and broken chords in the lh or both hands are new elements in organ mu-
sic, especially occurring so often in the course of one movement. The main
theme’s repeated notes are usually found in organ music only for fugue
subjects, while the Positiv episodes are atypically gigue-like and wide in
texture. Note that the third ritornello is made more climactic and the final
two bars are given klavieristisch scales, as in the G major harpsichord trans-
criptions BWV 986 (Johann Ernst?) and 973 (Vivaldi).
Example 96
Second movement
Again, the clear and simple shape – tutti piano framework around a solo –
is like a student’s essay in style. And again the Positiv parts suggest two violins,
with basses entering for the jeu en trio of b. 28; and perhaps manuals can
change more often than the copyists understood (see KB p. 71). But the
putative original is not so clearcut:
1 opening dotted-note theme accompanied by a simple continuo

6 solo with simple accompaniment, not canonic
18 original bass line has no repeated motif requiring change of
manual
25 BWV 592 melody more continuous; part-writing smoother;
five-part end
The contrast between framework and solo has become more stark, and the
solo’s cadence is now more of a climax. Though not unlike the chorale
BWV 654, the five-part passage has unusual scoring: two solo parts, two
accompaniment, one bass.
The dotted-note theme looks at first like an ostinato bass (cf. Cantata
31.iv), though such empty octave lines are known in Italian concertos, both
208 BWV 592
slow (Vivaldi in BWV 593) and fast (Handel Op. 6 No. 3). Octave imita-
tion for the solo theme is known widely, including the D minor Concerto
for Three Harpsichords, while the cast of the melody from b. 12 onwards
resembles Handel’s sequences derived from Corelli. The whole movement
is a web of Italian allusion, and rather touching.
Third movement
Much new figuration resulted from adapting the violin writing. While no
doubt the ‘third movement has gained most by the arrangement’ (Praetorius
1906 p. 100), it also lost some Venetian flavouring. So the original ritornello
bass line (Example 97) may lack poise and momentum but is far closer to a
bass line by Vivaldi. Bach seems to have been particularly free with Johann
Ernst’s original in this finale, substantially so in the section that moves to A
minor.
Example 97
Yet even as they replace something idiomatic the new bars are italianate
(bb. 81–6, cadence on the violins’ open g string), and their motifs crop up
in Bach’s own concertos, e.g. b. 47 in the E major Violin Concerto, first
movement. There are mostly only two parts, and string tuttis are indicated
by pedals and simultaneous semiquavers. Neither tutti nor solo figuration
is typical of organ music outside the obbligato parts in cantatas, though
Sonata No. 6 may owe something to this transcription. The perpetuum
mobile element is less typical of Italian concertos than might be thought,
certainly in the case of such ritornelli as these, whose shape is as textbook-
regular as the first movement’s.
Distribution between the manuals is ambiguous, and the changeover
of hands less clear than in the other movements. (Or at least its nota-
tion is not so clear: e.g. the first note of b. 13 lh could have double tails,
like b. 35 rh in the slow movement.) If episodes are solo (Positiv), do
the hands move to Oberwerk for the pedal sections? Where the first solo
begins is also uncertain, for to judge by the final bars, the scale in b. 12
is tutti not solo. Greater nimbleness than usual is required for manual-
changing across bb. 41–2, and perhaps the left hand remains on Ober-
werk throughout, with the right hand on Positiv in the episodes. Possible
reasons for not indicating manuals are (1) composer or copyist did not
distinguish tutti from solo, or wish to make it obligatory; (2) copyists
209 BWV 592–593
rejected and/or ignored indications; (3) they were more certain and/or
careful in the first two movements. Comparison with Walther’s transcrip-
tions rather suggests that the transcriber himself did not indicate manual
changes.
BWV 592a Concerto in G major

No Autograph MS; source Lpz Poel 39 (c. 1780?).
Headed ‘IV. Concerto per il Cembalo Solo del Sigr: Giov: Seb: Bach’.
To judge by its agreement with BWV 592 in those details in which J. S. Bach’s
arrangement differs from Johann Ernst’s original, BWW 592a is not an inde-
pendent transcription but (unlike the short-score or so-called harpsichord
transcriptions BWV 972–987) an arrangement of the organ transcription,
without pedal. Though not certainly authentic, it offers an interesting
comparison between organ and harpsichord transcription: the harpsichord
writing is usually thinner and leaps around more; a sense of tutti is given
in the ripieno sections both by bigger chords and much activity in the two
hands together; and no manual changes are indicated.
BWV 593 Concerto in A minor

No Autograph MS; copies in P 400b (J. F. Agricola 1738/9?), P 288 (c. 1780)
and probably lost MSS of J. P. Kellner and J. C. Kittel.
Partly three staves in P 400b, headed ‘Concerto del Sigre Ant. Vivaldi ac-
commodato per l’Organo a 2 Clav. e Ped. del Sigre Giovanni Sebastiano
Bach’; second movement ‘Adagio’, third ‘Allegro’ in P 288. Manual indica-
tions there, ‘O’ and ‘R’.
The concerto is a transcription of Vivaldi’s Concerto in A minor for Two

Violins, published as Op. 3 No. 8 (Amsterdam 1711, RV 522). Op. 3 is
likely to have originated between 1700 and 1710, with concertos whose first
solo entry has important thematic material perhaps the last to be com-
posed (Eller 1958). As Schering already suspected (1902 p. 236), such works
might well circulate with variant readings before being published, and while
for BWV 593 the Amsterdam print was probably the source (cf. VII/6 KB
p. 89), there is some uncertainty. Details in P 400b suggest that Bach himself
revised the pedal-line there (KB p. 36).
210 BWV 593
For points to make about the two-manual notation of the first movement,
see also BWV 592. Once again, in the transcribing of violin figuration for
organ new textures and figures appear, and particularly in the finale the
two manuals are used to distinguish both tutti from solo and violin solo I
from solo II. In its imaginative use of ritornello the work serves as a more
sophisticated model than Johann Ernst’s, while the middle movement too
shows a genuine art of combining themes.
First movement
The ritornello principle affects the five sections a–e of the main theme:
1–16 a (1–3), b (4–5), c (6–8), d (9–13), e (13–16)

22–5 e
39–42 c
52–4 a
62–5 d
68–71 a
78–86 b, c and e
90–3 e
There is some intricacy here: the episodes not only refer to each other but
use material from the main melodies; four of the last five episodes develop
an anapaest motif which comes from b; and Oberwerk during the second
episode furthers the merging of solo and tutti in the next section (in the
Amsterdam print, c in bb. 39–42 is shared between solo and tutti, unlike its
first appearance at b. 6).
The melodic material is very diverse, from the pleno chords of bb. 1–16
to slender two-part episodes, neither characteristic of organ music. The
episodes in two parts are clearly derived from violin lines, while held chords
in the tuttis have been filled in. The changes can be summarized:
tuttis with filled-in harmonies

imitation introduced in bb. 6–7, bb. 40–2, bb. 81–3
momentary gaps filled (bb. 19ff., 46, 47)
original bass in bb. 30–3 enlivened and rewritten
scales in bb. 42–4 originally more varied in scoring, including bass line
b. 44, originally no climax on c
octaves in bb. 51ff. originally a tutti in fuller octaves
bb. 71ff. pedal takes a viola line
Organo pleno in b. 51 is a puzzle: P 288 has ‘Obw:’ while Agricola has

‘O. plen.’ (and ‘pl. O’ at b. 62), perhaps a misreading. But Vivaldi’s bb. 51–4
211 BWV 593
are obviously climactic, so perhaps Bach or a copyist meant ‘add further

stops’ or ‘couple manuals’ or ‘do something’ to compensate for the thin
octaves, even if it breaks the continuity.
Not so much violin figuration in the episodes needed to be changed as
actually was, and string passages in bb. 55ff. and bb. 71ff. were less similar to
each other than the transcription suggests. Bach’s transference technique –
with its atypical pedal line – curbs the variety. But who was responsible
for the semiquavers of bb. 28–9 being down an octave, for the different
bass in bb. 30ff., and for omitting the harmonies of bb. 51ff.? Particu-
larly interesting are the filled-in gaps of b. 46 and bb. 19ff. (the latter in
Example 98) since this might suggest that Bach misunderstood Venetian
rhetoric. At b. 45, BWV 593 retains a violin figure that is not very idiomatic
on the organ, and indeed the whole passage bb. 43–7 illustrates the tran-
scriber’s priorities: string lines are simplified to suit organ but still need to
keep up tension.
Example 98
As to reducing the gaps: others in the finale are also filled in, and even
more extreme is Bach’s addition of a bass, in the 1740s, to unaccompanied
bars in another italianate work, an aria in Handel’s Brockes Passion (see
Beisswenger 1992 pp. 182ff.).
Second movement
Although the division into Oberwerk ostinato and Positiv solo (including
the solo duet for two violins from b. 14 onwards) is not specified in the
sources, analogy with BWV 592.ii suggests it. There is a strong and unusual
personality to the movement, due to the unusual spacing and tessitura and a
haunting melody for expressive violins, though compared to the Six Sonatas
the exchange of solo parts is elementary:
212 BWV 593
13–18 = 25–30
31 = 32
33–7 = 37–41
But exchange was a characteristic of the double concerto, and the sudden
return to the tonic in b. 24 seems to have been made for it. None of this
exchange of parts is in the 1711 Amsterdam edition. Characteristic of the
Italian duet tradition are the singing thirds, particularly after a passage of
imitative counterpoint, as at bb. 16–19. A theme in bare octaves with da capo
return is found in the Sinfonia of the Weimar cantata BWV 18, in triple time
and beginning with upbeat, like a French chaconne.
The transcription differs from the print as follows:
1 original heading ‘Larghetto e spiritoso’

9–12 violin II now down an octave
16–17 original imitative phrase altered to avoid d (see bb. 28–9)
26–31 violin 1 now down an octave, becoming the alto
31–41 two solos originally in thirds throughout (but exchanging
parts)
41 original ripieno marked ‘forte e spiritoso’, not ‘piano’
Third movement
Though the main theme is conspicuous in its bare scales, it is less versa-
tile than the first movement’s. The transcription differs from the print as
follows:
13ff. original bass line less active

42ff. string semiquavers altered (pattern varied, compass
narrowed)
51ff. left-hand line now an octave lower
59–63 pedal phrases to fill in original tutti rests
66–74 exploitation of a motif heard only in the original
bb. 69, 72
83ff., 115ff. original bare octaves now coloured by the same motif
86–113 repeated quavers originally on open strings in order: e ,
a , d , g. First two now dropped an octave, the order
disguised
104 d in melody avoided
118–27 simple alto sequence varied and put in pedal (an octave
lower)
128–31 Originally tutti
132ff. string semiquavers altered (same as bb. 42ff. in original)
142–4 octaves only, in print
213 BWV 593–594
The chief differences concern figuration (colourfully varied episodes in

BWV 593) and gaps filled in to avoid silences. The f/p marks as they appear in
the print are absent, and change is produced instead by different figuration
and manual-change.
Vivaldi’s concerto produced new effects by the interacting soloists (as
Spitta observed, I p. 414), and now the transcription does it with vari-
ous keyboard devices: two manuals for crossed lines or for antiphony or
for alternation or for melody-with-accompaniment. As in BWW 592.i, the
double pedal permits a richer harmony, whilst the repeated pedal e also
contributes motion – unusual in organ-music if not in string concertos.
Perhaps the biggest difference from the print concerns bb. 59–75: the pedal
not only fills in gaps (see Example 99) but does so with a motif convenient
for pedal and actually derived from a figure in Vivaldi’s original (b. 69).
The whole passage comes to concentrate on a motif that was given only en
passant in the original, and goes some way towards a ‘motivic unity’ rare in
(and of no interest to?) Vivaldi.
Example 99
Although the final entry of BWV 593 alone begins in thirds and
sixths, Vivaldi has supplied this material in another part of the movement
(bb. 3–4), causing one to question whether it was the print or a version
already including these harmonies that was Bach’s source. A more reliable
indication of Bach’s desire to add momentum to a big movement is the pedal
part made more active, presumably because pedals needed to do more than
string basses if they were to be as energized. Equally striking is that the spec-
tacular episodes of bb. 75ff. and 118ff. scarcely change the original notes,
simply scoring them between two hands.
BWV 594 Concerto in C major

No Autograph MS; copies in Lpz Inst. f. Musikwiss., Inv. 5138 (W. F. Bach
c. 1727, now incomplete) and Inv. 5137 (J. P. Kellner c. 1725), P 400c (J. F.
Agricola 1738–41?), further copies from Kellner or Agricola.
214 BWV 594
Heading by W. F. Bach ‘Concerto à 2 Clav: è Ped:’, in P 400c as for P 400b

(BWV 593); in Vivaldi’s autograph, first movement ‘Allegro’, second ‘Grave
Recitativo’. ‘O’ and ‘R’ most consistent in Agricola (not at b. 126 third
movement – unwanted?).
The original is Vivaldi’s Concerto in D major for Violin, in a version close to

MSS in Turin, Schwerin and Cividale (RV 208, see Tagliavini 1986 p. 242).
In another version it was published in Amsterdam, 1716–17, as Op. 7, Bk
2 No. 5 (RV 208a). BWV 594’s middle movement is neither this print’s nor
a Bach composition as once thought but resembles the Turin autograph’s,
while cadenzas in the outer movements resemble Schwerin’s. Vivaldi has no
cadenzas but directs ‘qui si ferma a piacimento’ (‘here one closes however
one wishes’), a wording he used elsewhere (Ryom 1977 p. 245).
Since therefore several versions circulated, one cannot say ‘in what bars
J. S. Bach transformed the musical text’ (Ryom 1966 p. 109), except that
having no concertino cello part for the episodes, he added a motivic bass
there. While it is possible that the concerto once existed in yet another form,
Spitta’s reasonable suggestion of solo viola da gamba (I p. 414) cannot now
be sustained, any more than it can be for BWV 592: the low-lying episodes
of bb. 26ff. are an octave higher in the published Op. 7. The transposition
to C major avoids notes above c .
As with BWV 596 and 593, there are details that suggest Bach to have
revised the transcription and Kellner to have shortened or omitted the ‘ca-
denzas’ for his copy (KB pp. 54, 50). Inconsistent indications suggest that
organists took manual-changing for granted.
First movement
Greater emphasis falls on solo episodes here than in the first Allegro of
BWV 593:
1–26 tutti, two particular motifs; preparatory chromaticism

(including Neapolitan 6th) before cadence
26–63 solo, non-thematic, gradually to dominant; tutti 58, opening
motif
63–93 solo, non-thematic, more modulatory; tutti 81, opening
motifs
93–117 solo, non-thematic, modulatory; tutti 111, opening motif
117–78 solo, mostly non accompagnato; tutti 174, opening motifs
cf. 25
Neither fourth nor fifth tutti is a reprise in the usual sense. The empha-
sis on the episodes seems to presuppose an ‘allegro vivace’ performance,
215 BWV 594
with a sharp-toned Positiv of the older kind. The organist does best by
carrying a memory of the original concerto, for the transcription’s busy
detail and thematic episodes seem more dependent on medium than
BWV 593’s:
3ff. original unison imitation of scales etc now at octave

5ff. original harmonies filled in
15–26 chords filled in (lh semiquavers); half-bar f/p contrasts
ignored
26ff. solos down an octave; lh parts added; new Ow contrasts,
rh only?
51ff., 64ff. bass lines absent in BWV 594 (as in Schwerin, not
Amsterdam)
77–80 pp marks in the string parts ignored
93ff., 118ff. bass lines different from Amsterdam print
105ff. lh figure replaces original basso continuo; lh scales added
118–20
137–73 modified version of Schwerin solo episode, like other
Vivaldi ‘cadenzas’; in Amsterdam, five bars for violin
alone link episode (ending b. 137) with final tutti
A long final solo episode having more than one form is found again in the
Fifth Brandenburg Concerto, with its two alternative so-called cadenzas.
In general, the transcription is more literal than in BWV 593 and realiza-
tions are straightforward: Example 100 is one of many. Perhaps lowering the
Rückpositiv part an octave suggests a 4 registration not 8 (Tagliavini 1986).
Such figures as those of bb. 65ff. and 93ff. are straight transcriptions, except
the left hand is down an octave and the implied staccato is now specified;
and lines are altered to use both bottom and top C of the organ.
Example 100
216 BWV 594
In general, the movement adds to the repertory of organ effects with

its unaccompanied solo line, the chords of bb. 65ff., the violin-like figures,
right-hand pedal-point effects, and quickly alternating hands. The final solo
episode’s pedal-point harmonies require ever more space to resolve, whereas
earlier returns to the tutti had been almost abrupt.
Second movement
The Grave of the Turin and Schwerin MSS is a 23-bar recitative with con-
tinuo, that of the Amsterdam edition a more conventional 11-bar melody
above repeated thirds in violins I and II. In the Schwerin MS the move-
ment is in score for violin and continuo, the chords notated as minims and
semibreves (Ryom 1977 p. 338).
The short chords in the accompaniment suggest what was played by
organists for whom Italian recitative was still a novelty. Such an idiom is
not only much less common in organ music than the Grave durezze of
the C major Toccata but is also unlike most actual recitative – in compass,
tessitura (octave lower than original), range (minims to fast runs), and quasi-
obbligato tenor line at the end. The melody is instrumental and, though it
includes harmonic progressions familiar in vocal recitative (bb. 5, 20 etc.),
is not far removed from a tierce en taille solo (bb. 15–19).
The movement is not only unique in the concerto corpus of Vivaldi
(Ryom 1966 p. 97) but no more than faintly resembles textures in other Bach
works, such as the opening of the G minor Fantasia. Though instrumental,
it is more vocally inspired and italianate than the solo lines in old organ
toccatas or even in the Pièce d’Orgue.
Third movement
The tutti ritornello has several limbs partially returning and making way
for solo entries more massive than the tutti returns.
1–64 tutti, quaver motif; then solo, new theme, to dominant

and back
64–112 tutti, contracted, quaver motif; solo at first less thematic,
towards:
112–64 tutti, dominant, to mediant; solo, new triplet figures, to
minor
164–79 tutti, beginning as second tutti, ending as first
180–283 solo, long sectional episodes
284–90 tutti, contraction, in octaves
BWV 594 differs from the other versions as follows:

217 BWV 594
1ff. unison imitation of motifs altered to octave imitation

24 etc. such bars filled in with scales
32ff. solo down an octave; busy lh runs etc replace original
continuo
81ff. original pp chords filled in and written short
90ff. new points of imitation attempted
106–11 violin’s abbreviated notation expanded; lh quavers replace
pedal point
126ff. further references to the quaver motif
180–283 not in Turin autograph; Amsterdam ends 179; Schwerin as
BWV 594
Like the finale of the A minor Concerto, the movement provides a greater
variety of textures than the original. Thus the first episode has a two-part
texture on Positiv, the second a lively line accompanied by Oberwerk chords,
the third with triplets, the fourth a solo line. The second episode is a rewriting
of a passage conceived in terms of the violin and not amenable to keyboard:
see Example 101.
Example 101
The final episode begins like a north German toccata, especially when it
changes to 4/4. But much of it is a transcription of violin figures as italianate
as the dissonances (bb. 247ff.) and the minor-key colouring, the latter being
found in other final episodes, e.g. in the Concerto for Three Harpsichords
BWV 1064.iii. From at least b. 210, the episode is unusually close to the
original – did Kellner not much care for Bach’s experiment with violinistic
keyboard writing? Positiv figuration generally is like that in a harpsichord
concerto, an idiom which the C major Concerto for Two Harpsichords
218 BWV 594–595
BWV 1061 shows to be typical of keyboard concertos rather than of tran-

scriptions as such. Or perhaps the style of bb. 32ff. and bb. 81ff. originated
in such transcriptions as this and then became associated with the keyboard
concerto.
BWV 595 Concerto in C major

No Autograph MS; copies in P 286 (eighteenth century, same copyist as for
BWV 594, and in P 288 for BWV 593); P 832 (from P 286 or both from
a common source), either or both directly or indirectly via a lost Kellner
copy?
Headed in P 286 ‘No: 2 Concerto del Illustriss: Prencipe Giov: Ernesto Duca
di Sassonia, appropriato all Organo à 2 Clavier: et Pedal’.
The attribution to Johann Ernst is based on the title in P 286 and on

J. N. Mempell’s contemporary copy of the harpsichord version, BWV 984.
Kellner’s own copy of BWV 984 does not mention him, and no original has
been found. BWV 595 (which consists of the first movement only) is fifteen
bars longer than BWV 984.i. If this was the result of ‘improvements’ by J. S.
Bach (Spitta I p. 413, and KB p. 76) it would confirm that he was less faithful
to the prince’s originals in organ transcriptions. But as likely is that in the
harpsichord version he shortened it by lessening its repetitiousness.
BWV 984 (harpsichord) BWV 595 (organ)

1–6 1–6
7–21 12 (2nd 12 )–27
22–34 35 (2nd 12 )–48
35–6 49
37–8 50–1
39–42 52–7
43–66 58–81
Perhaps the organ version has only the first movement because of a de-
fective copy, but the second movement would be problematic on organ
(inconsistent textures in F minor) and the third is harmonically meagre.
Although the apparently inescapable half-bar phraseology may justify the
usual opinion that Vivaldi and Johann Ernst had ‘widely separated talents’
(Schulze 1972 p. 6), the movement has a place in the repertory of italianate
concerto shapes. The opening theme is a classic ritornello, repetitive as if
the prince were imitating some model. It lacks the clean form of BWV 592.i,
despite the last section being like the first. Its theme is vaguely similar, but
219 BWV 595
the dangers of repetition are increased by a recurrent sequence that is part

both of the main theme and of the episodes. In the absence of the original,
it cannot be certain that the solo/tutti divisions in BWV 595 reflect those of
the string version, but they probably do.
Linking passages often suggest other organ works: for the figure in
b. 9 see the Dorian Toccata, for the cadences in b. 7 and b. 31 those in
the Concerto BWV 593.i. Johann Ernst had grasped the letter of Italian
concertos (see opening bass line) and at times its spirit (Neapolitan sixth of
b. 56). While the ‘soloist’ enters sooner than usual in Vivaldi, there are var-
ious Vivaldian passages including the non-modulating episode bb. 44–9.
Figuration is generally more organ-like than the scurrying semiquavers of
the harpsichord version, and if the half-bar and two-bar phraseology is more
naive than in BWV 538 (Example 102) the family likeness is still there in the
square phrases, the two manuals, and the semiquavers threading in and out.
Example 102
The static sequence in bb. 3–9 of the harpsichord version and the organ’s
more varied section do not allow one to judge which came first or which is
closer to Johann Ernst’s original. One could argue either way: either from
BWV 984 a new arrangement was made for organ (KB V/11 p. 122), and
was closer to the prince’s original; or a new arrangement was made for harp-
sichord from BWV 595, reducing the episodes because changing manuals
was unusual in harpsichord music.
In any case, commentators have found fault with the form of BWV
595. The organ version ‘repeats, perhaps unnecessarily’ and results at one
point in a ‘jarring juxtaposition’ of B major/G minor chords (Schulenberg
1992 p. 402) – though one might rather find this the highlight of the
220 BWV 595–596
movement. The short ritornelli, the limited modulation and the repeti-
tive second half make it unlikely that Bach added the extra fifteen bars,
though the manual-changing is likely to be his (Zehnder 1991 p. 87). The
last if verifiable would have implications for the composer’s habits, because
as it stands, BWV 595 has more manual-changes than any other Bach work,
and these are for simple phrases not unlike some of the D minor Toccata’s,
BWV 538.
BWV 596 Concerto in D minor

Autograph MS P 330 (1714/17: Dadelsen 1958 p. 79); later copies P 289
(2nd half of eighteenth century, from lost Kellner source?); lost copy of J. C.
Kittel.
Three staves in first movement, elsewhere two; headed in P 330 ‘Concerto a

2 Clav: e Pedale’ (autograph), and ‘di W. F. Bach manu mei Patris descript:’
added by W. F. Bach (c. 1770–80?). Second movement ‘Pleno. Grave’, third
‘Fuga’, fourth (also in Vivaldi print) ‘Largo e spiccato’. For the manuals,
see below.
The concerto is a transcription of Vivaldi’s Concerto in D minor for Two

Violins and Cello obbligato, Op. 3 No. 11 (Amsterdam [1711], RV 565),
evidently made straight from the printed parts. (The top stave shows signs
of original violin clef – see NBA VII/6 KB p. 89.) Until 1911, the work
was taken to be a concerto of W. F. Bach, as he claimed on P 330, and was
published as such by Griepenkerl in 1844, surprisingly so after C. F. Zelter’s
earlier suggestion that it was the work of W. F.’s father (KB pp. 28–9). The
watermark of P 330, known also from MSS of J. G. Walther, is found in
Weimar cantata parts performed in 1714 and 1715, i.e. at a period when
Friedemann was about five years old. As with the Concertos BWV 593 and
594, the composer probably returned to the work later.
The first movement has become celebrated for its autograph registra-
tions:
b. 1 rh ‘Octav: 4f.’ and ‘Oberw.’

lh ‘Octav: 4f ’ and ‘Brustpos.’
‘Princip. 8f ’ and ‘Pedale’
b. 21 rh ‘Brustw.’
lh ‘Obw. Princip. 8f et Octav. 4f.’
pedal ‘SubB: 32f.’
221 BWV 596
As with the so-called registrations in Ob and Schübler chorales, their main

point is to specify correct octave pitch. Whether directives or suggestions,
they establish that
1 manuals were not necessarily based on 8 , nor pedals on 16

2 in transcriptions, two manuals replaced various scorings, not only solo-and-
accompaniment
3 hands could exchange manuals in the course of a piece
4 stop(s) could be added to manual or pedal in the course of a piece
The last point is important, since the music provides no clear opportunity
for the organist himself to add stops to either manual or pedal without some
hiatus. Unlike the left hand in b. 21, the pedal has no break in its quavers;
perhaps registering 32 was an afterthought, just as the right hand first had
its chord higher (did it?– see KB p. 24). Or perhaps the lh break merely
reproduces the change from violin to cello in Vivaldi.
In the Grave, ‘Pleno’ is directed; in the Largo, ‘f ’ and ‘p’; in the finale,
‘R.’ or ‘Rückp.’ and ‘O.’ or ‘Obw.’. Since the title says ‘a 2 Clav:’, it seems
that whether called Brust or Rück – in the gallery-front, in the breast
of the organ, or to the side – only one secondary manual or Positiv is
meant. (Copyists might have interpreted Pos. as Rückpos., as in the Toccata
BWV 538.) Despite major rebuilds, the Weimar organ seems never to have
had a Rückpositiv. Perhaps Bach began a short score of Vivaldi’s concerto,
with violin I down an octave to avoid d , and added directions after-
wards. There seems no reason why each hand did not begin on the other
manual and so have avoided exchanging manuals in b. 21. The rh scale
at the end (not in Vivaldi’s original) was written after the lh part – an
afterthought?
First movement
In the print the Allegro begins as a duo for violins, followed by a duo for
cello and continuo. See Example 103. This is unusual in Vivaldi: dashing
fiddle sound in a 32-bar prelude more than half of which is a tonic pedal
point. The organ’s opening three-part texture is also unique in its unison
imitation, but its repeated bass quavers – found in concertos for organ
(A minor finale) and strings (Sixth Brandenburg) – are no substitute for the
lost rhetoric of strings.
Lowering the violins’ part an octave is not quite paralleled by the Sinfonia
to Cantata 146 (Klotz 1975 p. 385 and Tagliavini 1986), since there is no
registration there for 4 , and the organ part apparently avoids not only d
but even c .
222 BWV 596
Example 103
Second movement
Seven-part chords are rare, and Bach did not copy Vivaldi’s direction
‘Adagio e spiccato’. Note a new kind of Neapolitan sixth, becoming the mi-
nor third of an interpolated triad (C minor between E major and A major).
The Fugue differs from the print (where it is ‘Allegro’) in scoring and
layout:
Pedal takes a practicable line rather than the original bass (which
comprised both solo cello and basso continuo) and enters late, without
theme.
No distinction is made between tutti and solo (bb. 20–8, 45–52) –
because the fugue is too short? – but episodes could be played on the
Positiv
The parts are frequently exchanged, not always merely in order to
avoid d
Unusually, the Fugue develops four-part invertible counterpoint as if

Vivaldi were offering a distillation of Italian contrapuntal teaching, and
Bach’s changes (such as bb. 45–6, bb. 53–4) only underline the nature of the
223 BWV 596
counterpoint. But he also made his familiar additions, such as continuous

semiquavers above the closing pedal point, a taxing place for the player.
While one can imagine such an episode as bb. 21–4 influencing his later
writing, the pedal point is unusual for its rhythmic tonics and dominants.
While few Bach fugues of any period are so sectional, Vivaldi’s repetitious
dactyls at the beginning of almost all semiquaver groups now give way
to smoother continuity. No doubt the fugue’s strict invertibility was an
attraction for J. S. Bach, under whose name it was also known separately
(KB p. 26).
Third movement
Unlike the octaves of BWV 592 and 593, the tutti framework has an unmis-
takable siciliano character, and surrounds a sweet melody, one apparently
taken by Griepenkerl to represent Friedemann’s tenderness (KB p. 21). The
transcription differs from the print both in the spacing of the accompani-
ment (now for one hand) and at times in the harmony itself – ‘improvements’
by Bach or the sign of a different original? Neither homophonic tutti nor
lyrical solo has close parallels in Bach organ works, which is surprising in
view of their suitability for organ.
Fourth movement
Though basically a ritornello, the finale has unusual features: the soloists
provide not only the multi-limbed theme but also the episodes. Shape and
finality are given by a tutti passage with chromatic bass line appearing at
regular points (bb. 11, 27, 68), but the original form is blurred by the use of
two manuals:
Op. 3 No. 11 (‘Allegro’) BWV 596

A1 1–6 two violins Rp
7–11 solo cello, 11–14 tutti both Ow
A2 14–22 trio Rp
23–7 solo violin, 27–30 tutti both Ow
30–43 solo violins accompanied Ow, then Ow + Rp
B 43–6 tutti Ow
46–50 trio Rp
50–3 tutti with echoes Ow/Rp
A3 53–68 trio Rp, then Rp + Ow
69–73 tutti with echo Ow
Although the material yields some organ textures unusual outside the tran-
scriptions, it alludes a great deal to Italian string writing of a previous
generation:
224 BWV 596
clashing suspension style for two violins (1ff.)

paired quavers (4)
falling chromatic fourth, with Neapolitan 6th (44–6)
a version of the tutti tremolo effect (12)
characteristic solo cello figures (7)
tutti violin suspensions (12)
parallel thirds for two violins (14)
repeated-note figure for solo violin with accompaniment (35)
punctuating cadences (45–6, 50–3)
(For bb. 1ff., compare the opening subject of Cantata 21, sung on 17 June
1714, shortly before the sick prince left Weimar.)
Unusual organ textures result partly from finding equivalents for
idiomatic string music (b. 7, b. 59), partly from using it more or less unaltered
(b. 35, b. 44). The tremolo tuttis have been replaced by a busy line making a
fifth part (b. 11), and a few minor gaps have been filled in, though perhaps
fewer than usual. The left-hand accompaniment of bb. 59–67, assumed to
have been added by Bach (e.g. Schneider 1911), is by no means alien to
Vivaldi’s style, though his original rising line of bb. 63–4 has disappeared in
the need to avoid d .
A big impression is made by the falling chromatic fourths at the end,
giving a stirring ‘D minor finality’ such as marks the end of the Three-part
Invention in that key.
BWV 597 Concerto in E major
Only source: Lpz MB MS 7 (J. G. Preller, with BWV 585, 586 and 1027a).
Unique heading in the MS: ‘Concerto . . . A 2 Clavier con Pedal. di Mons:

Bach’.
Keller is no doubt correct to see in BWV 597 neither a concerto nor a

composition (or transcription) of J. S. Bach, but rather a trio sonata by a
composer of a later generation (1937 p. 66). The opening imitation suggests
two violins, the gap in the middle something missing, and the low harmonic
and melodic tension an inexperienced composer – but one who (to judge
by the charming cadences, sevenths and ninths in the Gigue) knew some
Telemann. Perhaps it was a student exercise in composing a pair of very
different movements based on or making use of similar material (Bartels
2001). The theme(s), the repetition and the decorative treatment resemble
those of no known work of J. S. Bach.
The term ‘Concerto’ recalls H. N. Gerber’s Concert-trios (1734: worklist
in E. L. Gerber’s Lexicon, 1790). Perhaps Leipzig pupils sometimes used
the term to distinguish such pieces from Trios based on chorales and from
Sonaten of two or (as there should be here?) more movements.
BWV 598 ‘Pedal-Exercitium’

Only source: P 491 (C. P. E. Bach, early).
Heading, ‘Pedal Exercitium Bach’ (written by C. A. Thieme: Schulze 1984

p. 126).
Several origins are possible for this piece: a fragment of a lost toccata (cf.
Lübeck’s Preludium in C); an independent pedal exercise, to be taken further
(brought back to the tonic); a prelude to a fugue, or a preamble to a written
prelude such as BWV 542; a paper exercise in composing for bass, whether
organ or (transposed) for cello; an étude by J. S. Bach to be completed
by C. P. E. Bach or C. A. Thieme (a Thomaner whose title-page of the
1738 figured-bass treatise attributed the latter to ‘Joh. Seb. Bach’), on the
analogy of the Allemande in CbWFB or the Fantasia in AMBB; an exercise
composed by C. P. E. Bach for whatever reason, and acquired by Thieme.
That the final bars cannot seem to escape the dominant, and thus imply
something of a compositional impasse, could be explained by any of these
possibilities.
[225]
226 BWV 598
Bars 19–23 read as much like a string-crossing exercise for cello as a

leaping exercise for pedal, and in either case imply counterpoint in two
parts; compare the Cello Suite in G major, Prélude. Hermann Keller heard
in it something ‘stormy and exuberant’ typical of the young Sebastian, but
the diminished fifth sequence of bb. 27–8 is unlikely to date before the Six
Sonatas. Its composer certainly seems to have been familiar with the violin
and cello suites as well as pedal-parts of J. S. Bach, and provides a repertory of
techniques for the advanced player – alternate-foot pedalling, leaps, the same
foot for adjacent notes, different feet for repeated notes, varied articulation,
perhaps off-beat slurs with heels, perhaps echo-registration for bb. 2 and 4.
Such a scope seems rather too well deliberated for the ‘hasty copy’ to have
been made from an improvisation by Emanuel’s father (as Dadelsen 1957
p. 39 suggests).
Orgelbüchlein BWV 599–644
Autograph MS P 283. Title-page of 1722 or 1723 (Dadelsen 1963 p. 77):
Orgel-Büchlein Worinne einem anfahenden Organisten Anleitung gegeben

wird, auff allerhand Arth einen Choral durchzuführen, anbey auch sich im
Pedal studio zu habilitiren, indem in solchen darinne befindlichen Choralen
das Pedal gantz obligat tractiret wird. Dem Höchsten Gott allein’ zu Ehren,
Dem Nechsten, draus sich zu belehren. Autore Joanne Sebast: Bach p. t.
Capellae Magistri S. P. R. Anhaltini-Cotheniensis.
Little Organ Book, in which guidance is given to an inquiring organist in
how to implement a chorale in all kinds of ways, and at the same time to
become practised in the study of pedalling, since in the chorales found
therein the pedal is treated completely obbligato.
For the highest God alone to Honour,

For my neighbour to instruct himself from it.
Composed by Johann Sebastian Bach, p.t. [pro tempore, ‘at present’? or

pleno titulo,‘with full title’?] Capellmeister to the Serene Reigning Prince of
Anhalt-Cöthen.
The album is now always known as ‘the Orgelbüchlein’, but its title-page,
written later than most of the contents, says nothing about what if anything
was originally intended. Its didacticism is more typical of title-pages of its
period, WTC1 of 1722 (or 1723) and the Inventions of 1723, when Friede-
mann was twelve or thirteen years old, and looks as if to match them. There
is no evidence of a previous title, and perhaps ‘p.t.’ implies it was written
pending the move to Leipzig in May 1723.
Büchlein was a common term: Gesangbüchlein (Weimar hymnbook),
Gebetbüchlein (Weimar book of prayers) and Clavier-Büchlein (1720).
‘Durchführen’ implies a model for composing or playing chorale-
harmonizations, used for the Inventions (P 610) and earlier chorale-books
(J. P. Treiber, Der accurate Organist, Arnstadt, 1704). Useful as exercises
though the pedal parts are, it has long been recognized that the album shows
no planned, progressive difficulty (Peters V 1846), and could hardly do so
even had it been completed. ‘Anfahend’, an old-fashioned term, appears
on the title-page of Ammerbach’s Orgel oder Instrument Tabulatur, Leipzig
1571, a book known to J. S. Bach (Dok I p. 269) and also subtitled Büchlein;
it too refers to young players (‘der Jugend’) and was the first keyboard music
[227] published by a holder of the cantorate to which Bach had recently been, or
228 Orgelbüchlein
was soon to be, appointed. ‘Anfahenden Organisten’ (‘learning organists’)

also feature in the dedication of Werckmeister’s book about a famous rebuilt
organ, Organum gruningense redivivum (Quedlinburg, 1705), a description
surely known to Bach.
The rhyming couplet salutes neither the author, as in Werckmeister’s
Orgelprobe, nor a dedicatee, as in Partita No. 1, but cites the Lutheran duty
‘to serve God and one’s neighbour’, as do BWV 639’s text and the Obituary,
this twice (Dok III pp. 85, 88). Pious allusion can be found in the album’s
handwriting (Schmögner 1995).
Sources
As interpreted in KB pp. 23ff. and Löhlein 1981, the contents of P 283
are:
I title-page
II blank
1 BWV 599 draft (Urschrift)
2–3 BWV 600 draft
4 BWV 601 careful fair copy (kalligraphische
Reinschrift)
5 BWV 602 draft
6+ BWV 603 draft (runs over to p. 7)
7 (one title, not set)
8 BWV 604 hasty fair copy (flüchtige Reinschrift)
9 BWV 605 careful fair copy (end in tablature)
10 BWV 606 careful fair copy
11 BWV 607 draft (last 2 23 bars on p. 10)
12–13 BWV 608 draft or revised fair copy
14 BWV 609 draft
16 BWV 611 draft
17 BWV 612 draft or revised hasty copy (end in
tablature)
20–1 BWV 615 careful fair copy
23 BWV 617 careful fair copy (? – end in tablature)
23a slip completing BWV 617
229 Orgelbüchlein
24a slip completing BWV 618

26 BWV 620 careful fair copy, revised (end in
tablature)
[26a lost slip completing revision of BW 620a?]
28–9 BWV 622 draft or revised fair copy
30a —
30b close of BWV 624 (later copy?)
32 (one title)
33 ‘O Traurigkeit’(fragment)
34–8 (four titles)
44 BWV 628 draft or revised fair copy
45 BWV 629 draft or revised fair copy
54 BWV 631 careful fair copy, revised
60 BWV 634 draft
62–72 (nine titles)
73 BWV 635 draft
74–7 (three titles)
79–88 (ten titles)
89 BWV 637 draft?
91–105 (thirteen titles)
106+ BWV 639 careful or hasty fair copy (runs
over to p.107)
107–12 (six titles)
113 BWV 640∗ careful or hasty fair copy
114 (one title)
115 BWV 641 careful or hasty fair copy
116–28 (twelve titles)
230 Orgelbüchlein

130–48 (seventeen titles)
149 BWV 643∗ careful or hasty fair copy
150–76 (twenty-seven titles)
178–82 (five titles)
∗
‘alio modo’, i.e. has same title as the previous (unset) entry
The distinctions between draft, careful fair copy and hasty fair copy are not
always clear, however; some pieces could have begun as one and become the
other.
Still unknown is whether, as in BWV 651–665, the script used for the
supplementary headings, ‘a 2 Clav. e Ped.’, is different because it was added
later or because Italian is written in a different script from German chorale-
titles. (This heading for BWV 605 was over-written by W. F. Bach, implying
that he used the album.) How many titles were written in before the music
is unclear – most of them, some in groups? Other uncertainties are whether
pieces in draft are newer than all those in fair copy, and whether coloratura
passages are written smaller in order to be clear or because they were added.
Most titles were given one page, a few two pages: for some settings, half-slips
and completions in tablature show that a page was not enough. Whether
alio modo for BWV 640 and 643 means ‘another setting of the same melody’
or ‘a setting of another melody to this text’ is also unclear: Frescobaldi’s Fiori
musicali, known to J. S. Bach at this period (Dok I p. 269), already used it
in both senses.
Although most extant copies go back directly or indirectly to the auto-
graph, no other group is complete or keeps its order. Probably by c. 1717,
J. T. Krebs had copied twenty-nine in P 801 and – judging by empty pages –
meant to copy more; six more appear in P 802 (grouped according to chorale-
type), where Walther also wrote one. Walther’s manuscript SBB 22541/1–3
has eleven, with other chorales on the same melodies. Krebs, knowing both
the revisions and ‘earlier versions’ (Dadelsen 1963), was surely close to the
composer at the time. Another copy, once thought to be autograph and con-
taining twenty-six chorales including BWV 620a, was written in c. 1727/30
by C. G. Meissner, a Leipzig pupil (KB p. 228), and later re-copied (Emans
2000 pp. 27f.). A third, containing seventeen by J. G. Müthel, is dated ‘1751’,
i.e. shortly after his intended study with J. S. Bach (KB p. 57).
Of the many copies, those by or associated with Kittel omit one chorale
(Lpz Poel 39) and Kirnberger (Brussels 12102, additions by Kellner) two
chorales. Others vary, such as Breitkopf ’s set copied for J. C. Oley (P 1160)
and C. F. Penzel (P1109), for C. P. E. Bach (?) in P 1110, or for J. N. Mempell
231 Orgelbüchlein
and J. G. Preller in Lpz MB MS 7. Why no copies follow the order of

P 283 is not to be explained by a missing source, or by liturgy, hymnol-
ogy, performing difficulty, or convenience of layout. During the Leipzig
years the composer doubtless kept this or another fair copy with his other
organ music, leading to further incomplete copies by pupils. Probably, P 283
came into C. P. E. Bach’s possession from his younger brother J. C. F. Bach,
who may have had it from their brother-in-law Altnickol (†1759: BJ 2001
p. 67).
Date
From such handwriting details as note-forms, clefs and staves, the following
table gives one possible chronology of the manuscript album (Dadelsen 1959
p. 80):
c. 1713/14: 599–609, 612 (later?), 616–619, 621 (later?), 622 (later?),
625–631a, 632, 635–639, 641–643
1714/16: 610–611, 614–615, 620a/620, 623–624, 633–634, 640, 644
(earlier?)
Leipzig (c. 1740): 613 and ‘O Traurigkeit, O Herzeleid’ (after 613?)
The paper of the MS is known from MSS made in 1714, and the handwriting
is like that of cantatas of 1714–15; but neither makes a start in Advent 1713
impossible. BWV 613 was written on the first entirely empty page in the
book (‘O Traurigkeit’ is almost the next), which suggests that in Leipzig the
composer set out to complete the album, at a time when he appears to have
had several projects for publication.
KB conjectures from appearances that BWV 603 and 601 were the first
to be written in (for the Third Sunday in Advent, 1713?), that BWV 599,
600 and 602 joined them only in the next church year, and that all of the
settings were probably composed during the relevant season:
Advent 1713 to Whit 1714: 601, 603–606, 608–610, 614, 621, 622, 625–627,
630, 631a, 637–644
Advent 1714 to Whit 1715: 599, 600, 602, 607, 612, 616–620a, 628, 629,
632–636
Christmas 1715: BWV 611
New Year 1716: BWV 615, and Passion 1716: BWV 623, 624
Later (Leipzig) entries: BWV 613 (New Year, c. 1740), ‘O Traurigkeit’
(Passion, c. 1740), 620 (revised after 1729) and 631 (revised after 1630)
This plan suggests that coloratura settings precede some canons, and that
skill in handling figurae gradually increased (Breig 1988). But the premiss
that Bach composed in the relevant season is doubtful, given so many non-
seasonal hymns.
232 Orgelbüchlein
In recognizing that the composer’s handwriting in his later twenties

barely changes and leaves few landmarks, a new chronology asks why com-
position, if not compilation, could not have begun shortly after the move
to Weimar (Stinson 1996):
1708–12? (as early as 1708 but no later than 1712): 601 (in ‘Neumeister’),
603–606, 608, 609, 621, 622, 630, 632, 635–638a
1709–13? (a ‘second phase’): 599, 600, 602, 607, 610, 612, 614, 625–629,
631a, 639 (also in ‘Neumeister’), 640–644
1715–1716?: 616–619
1716–1717?: 611, 615, 620a, 623, 624, 633, 634
after 1726: 613, 620, 631, ‘O Traurigkeit’
This dating implies that the (or an) album was begun (i) as Bach entered
on his new position at Weimar, (ii) for him to play in the Court Chapel. But
neither is demonstrable. It also begs the question of how quickly harmonic
style can mature, even for a Bach. In the case of ‘later’ groups too, the
reasoning is not obvious: chorales with unusual textures need not have
been entered only after Bach had made his copy of Grigny (as Stinson 1995
p. 65 suggests), since he doubtless knew several French Livres already.
Also questionable is whether the album ‘was planned as a more system-
atically organized collection of alio modo settings’ of chorales contained
in the ‘Neumeister Collection’ (Wolff 1991 p. 120), since this might imply
that Bach was still using ‘Neumeister’ in 1708 at Weimar, or even in 1714,
which is hard to believe, although the two collections do have complemen-
tary repertories. For the naive counterpoint of BWV 1108 to become the
polished and varied idiom of BWV 616, or for any part of BWV 1090 to lead
to BWV 612, a decade seems hardly enough. If ‘Neumeister’, authentic or
not, ‘paved the way towards concentrated and compact settings’ (Wolff 1991
pp. 302f.), so did many other chorales and variations of Central Germany.
BWV 601 compared with any variation in BWV 768 suggests either that
BWV 768 is much earlier than 1713, or that BWV 601 is much later than
1708, or both.
While some of the first settings to be entered probably originated earlier,
dating is vague and inconclusive. The Duke’s hymnbook of 1713, Geist-
reiches Gesang-Buch, might have inspired either composition or compila-
tion, though it was not the book actually followed. The chapel organ being
in and out of commission from June 1712 to May 1714 (Schrammek 1988)
could mean that e.g. some Advent and Christmas settings were older, or not
made for this organ. Dating the chorales from interior musical detail – e.g.
pedal quavers that end as each chorale-line ends (BWV 642) are earlier than
those that do not (BWV 611) – might neglect the sheer variety of technique.
More convincing is that work began with simple note-patterns (BWV 601)
233 Orgelbüchlein
and ripened into independent counterpoint (BWV 616), though this need
not mean that the ‘fantasia’ BWV 615 or the running tenors (BWV 617,
624) are late.
Purpose
The likely date when the compilation (i.e. as an album) began suggests that
Bach had in mind either the rebuilt Weimar organ or the larger new organ of
the Liebfrauenkirche, Halle, where he was invited to succeed F. W. Zachow
(Dok I pp. 23–4), in December 1713. This was some eight months after
work started on the Weimar organ.
Not only do the chorales’ immediate Affekte fit in with the pietism as-
sociated with Halle but they seem to conform to its contract-requirements
(Dok II p. 50):
langsam ohne sonderbahres coloriren mit vier und fünff Stimmen und
dem Principal andächtig einzuschlagen, und mit iedem versicul die andern
Stimmen iedesmahl abzuwechseln, auch zur qvintaden und Schnarr
wercke, das Gedackte, wie auch die syncopationes und Bindungen . . .
to play in a devotional manner, slowly without exceptional decoration in

four and five parts [voices? stops?] and with the Principal [alone], and at
each verse to alternate the other stops every time and also to apply the
Quintadena and reed-stops, the Gedackt, as too the syncopations and
suspensions . . .
Though unsure of the terms, what the committee wants is clear: discreet
registration, rich harmony and recognizable melody. It was in applying for
a job in the same Halle church in 1746 that J. G. Ziegler reported that Bach
had taught him to play ‘not merely indifferently but according to the Affekt
of the words’ (‘nicht nur so oben hin, sondern nach dem Affect der Wortte’ –
Dok II p. 423). Presumably, this was important to the appointing committee.
But not only Affekt: if the collection was begun with Halle in mind, its
special manner of harmonizing straight through without inter-line inter-
ludes could also reflect the town church’s style of hymn-singing. Inter-line
interludes are familiar both from hymn-settings presumed to be earlier
(such as the so-called Arnstadt Choräle – see BWV 715) and from those
known to be later (such as Kauffmann’s Harmonische Seelenlust, Leipzig
1733), and longer organ-chorales including fantasias likewise incorporate
inter-line interludes of a kind, though more integrated into the whole. But
the ‘Low Church’ convictions of Halle would require simpler or less dis-
tracting forms of chorale, replacing the formality of standard-hymns-with-
interludes with discrete, individual settings, simple in shape, expressive in
234 Orgelbüchlein
Affekt, and warmly registered on the organ. Hence could it be that the
Orgelbüchlein settings could be both solo pieces and (in most cases) accom-
paniments?
That the new Halle organ seems to have had chamber pitch (? see Dok II
p. 61) and a ‘tolerably good temperament’ (Dok I p. 150) could explain the
high pitch or distant keys of certain settings. Perhaps some were used when
Bach examined the completed new organ in 1716. At Weimar, appointment
as Konzertmeister on 2 March 1714 led to cantatas for the Duke’s chapel,
but the Ob can hardly have been ‘closely connected’ with this new work
(KB p. 88) – rather the opposite?
While Bach’s new duties as Konzertmeister need not have meant aban-
doning the compilation, finishing it would have been less urgent. Some such
reason for its being incomplete is likelier than that the unset chorales were
those ‘which do not lend themselves to musical description’ (Schweitzer 1905
p. 178), or that Bach had already used all possible note-patterns (Löhlein
1981 p. 12), or that after all, he was not ‘the man to set the chorale’ in 164
ways (Dürr 1988 p. 59). Settings could serve as teaching material, enabling
e.g. pedal-playing to progress from simple left/right alternation (BWV 612)
through partial alternation (BWV 615) to very little (BWV 622). But since
they could not have so served Wilhelm Friedemann in 1713–16, did the
title-page and its agenda belong only to when they could? Was pedal always
intended for every chorale, and two manuals for those now specifying them?
Or did P 283 contain two-stave harmonizations only later in need of per-
forming directions?
Hymnbook
Just as in cantatas Bach did not depend totally on Lutheran year-plans for his
choice of chorales (Gojowy 1972), so organ settings were not always associ-
ated exclusively with one day or season. Nevertheless, like J. H. Buttstedt’s
settings, the Ob was planned as a traditional Thuringian hymn repertory, if
not specifically for the Weimar hymnbooks of 1708 and 1713 as often said
(e.g. in EB 6587).
Recent hymns are not prominent: 147 of the 165 were in print before
1650, some 80 per cent are pre-1600 (Honders 1988), and the newer belong
mostly to the non-seasonal section. Practising organists knew many books,
as they still do, and while it is possible that the plan follows a Thuringian
hymnbook of c. 1675 (KB p. 104), that it did not is as likely – i.e. not Arnstadt
1666 and 1674 or Weimar 1666 (all without melody) but a general repertory
known to Johann Michael and Johann Christoph Bach (†1703), the titles
of whose Choräle zum Praeambulieren are also found in the Ob. Since it
235 Orgelbüchlein
includes no text by the Court secretary Salomo Franck or any Jesuslied texts
from the Weimar Gesangbuch of 1713, its connection with Weimar is not
obvious.
Not only does the order follow no known hymnbook, but no single tune-
book contains all the melodies used. The array of Advent and Christmas
settings implies that the album was to serve more than one church year,
while of the non-seasonal chorales listed or set, the largest groups are those
associated with penitence (11), Communion (9), time of trouble (7) and
death (16). Also included, though not as a group, are seven of Luther’s
Catechism hymns, and a text of his begins both parts, the seasonal (BWV
599) and the catechistic (BWV 635). Amongst those listed but unset are
three Trinity hymns and six metrical psalms, the last in biblical order.
Function
To start a collection with the main hymn of Advent was known since at least
August Nörmiger’s MS tablature book of 1598, prepared for a royal pupil in
Dresden, i.e. for devotional/practical purposes, not professional/liturgical.
If the original chorales later called Ob had a liturgical function, was it more
specific than Nörmiger’s? As preludes to a congregational hymn, preludes to
a choir hymn, interludes between verses, or voluntaries at other moments?
Each is possible.
Perhaps the Ob began with publication in mind, prompted by two re-
cent books. Daniel Vetter’s large, two-volume set of chorales, Musicalische
Kirch- und Haus-Ergötzlichkeit (Leipzig, 1709, 1713) begins as usual with
‘Nun komm’ – and in the less common key of A minor, like Bach’s – and was
evidently for church and home. In the publication of Walther’s variations,
the Musicalische Vorstellung of 1712, Bach may have been involved, as he was
with a later publication of Walther (see Dok II p. 377). That there was grow-
ing interest in collections of harmonized hymns is further suggested by the
ninety-seven figured chorale-variations in Musicalischer Vorrath (1716–19)
by J. S. Beyer, later cantor in Freiberg and closely associated with Silber-
mann organs. Whether P 283 was used by ‘Bach himself at the organ of the
Weimar court chapel’, as usually supposed (e.g. Stinson 1996 p. 28), is not
and cannot be known.
Chorales in the ‘Pachelbel manner’ compiled by Walther for the Weimar
town church were old-fashioned, and BWV 601 or 603–606 offered models
for the newer kind of harmony being developed in the Court Chapel. Many
of Walther’s extant chorales share two particular details with Ob: harmony
is realized in note-patterns (figurae); and a cantus can be set in canon, espe-
cially for certain seasons, sometimes with quasi-canonic accompaniments.
236 Orgelbüchlein
It would be no great step to see the Ob as reflecting interests in technique

shared by colleagues in the same town, especially in view of Vetter’s com-
petent but jejune treatments. Of course, it is the quality of its harmony
and melody, motifs and counterpoint, all developing techniques listed in
Walther’s Praecepta of 1708, that has led to greater attention being paid it
than to Vetter’s or Walther’s own settings.
There is a further possibility. If the settings had indeed been made for the
Weimar organ, and its pitch in 1713 was still high (chormässig: Schrammek
1988 p. 101), the yet higher keys of several chorales, including the first, would
make them even less suitable as preludes or interludes to a congregational
hymn. But a report of eight Weimar choristers singing chorales (Jauernig
1950 p. 71) could mean that they, rather than an aristocratic congregation,
sang the hymns, so benefiting from higher pitch: the upper limit of the
melodies varies from e (35 chorales) to f (7), to f (2) and to g (1). The
very location of the organ – in a ceiling gallery far above the chapel-floor –
speaks for a more direct relationship with professional singers nearby than
with the congregation below. But see remarks on Halle above.
Musical style
Characteristics can be listed, though there are important exceptions to each:
harmonizations of a cantus heard in the soprano

harmonies embroidered through figurae (often derived; treated imitatively)
in four parts, including cadences
without interludes between the lines
beginning with the melody, alone or accompanied
fermatas marking ends of lines (for articulation? a final pause?)
While other chorales are often described as ‘of the Ob type’, such as BWV 683,
727 or 730, various factors distinguish their form, harmony, texture or idiom
from most of the album. Similarly, if some chorale-variations anticipate the
style, as still often said (e.g. Breig 1988 p. 8), there is a perceptible gap: only
the last variation of ‘O Gott, du frommer Gott’ actually resembles an Ob
type, and then only superficially. The Ob has a level of inspiration simply
not found in the so-called chorale-partitas.
The principle of ‘melody chorale type’ is already there in the work of two
Halle composers, Scheidt’s ‘Mitten in dem Leben’ and Zachow’s ‘In dulci
jubilo’, as if a local speciality. Short settings by other accomplished com-
posers, such as ‘Jesus Christus, unser Heiland’ of Buxtehude, also hint in
this direction. The principle allows great variety, whether one motif runs
through all the parts (BWV 626) or through the middle parts (most chorales)
237 Orgelbüchlein
or individually to each part (605) or even outrunning the melody, having

produced its own impetus. Even so incomplete a MS juxtaposes settings so
different as to look like deliberately planned pairs, such as BWV 610 and 611
(both with tempo signs), 614 and 615, or 637 and 638. Canons are varied: at
the fifth (four, rare in organ music) or octave (five); in the cantus firmus only
(five examples); and sometimes in the other parts too, strictly or loosely –
though not in the accompaniment alone, as in Scheidt, Weckmann or
BWV 769.
Despite the attention given it, the ‘Ob style’ remains elusive. That in
it the figurae or note-patterns known to every composer generate excep-
tional harmonic tension is suggested by comparing any setting with one of
Walther’s or even of the young Bach. For while BWV 625 may be close to
the chorale in Cantata 4 (Kube 1999 p. 566), its harmonic tension is much
higher. The patterns themselves are found in many an earlier song-variation
(Example 104) but so imaginative a treatment of them as here was new. A
startlingly mature diatonicism is produced, and not simply because the
patterns are so concentrated; on the contrary, Steigleder’s ‘Vater unser im
Himmelreich’ (1627) already exploits a motif more single-mindedly than
the Ob’s inventiveness would have allowed.
Example 104
Figurae applied in the four-part chorale-variations of Pachelbel also ap-

pear in the Ob, and so do those illustrated in books of the time, such as
Niedt 1706 or Walther 1708. Niedt includes the very motif used in the early
BWV 601 (Sachs 1980 p. 143), as does other music of the period; but usu-
ally its effect is merely to decorate simple triads, not to generate so many
sevenths as in BWV 601. Clearly there was widespread interest in setting
chorales by using figurae, and it could be that ‘durchführen’ on Ob’s title-
page is acknowledging this abiding interest. Momentum towards cadences,
accented passing-notes and ties generated by the figurae give an impression
of a constantly propelled harmony. BWV 623 produces a series of original
accented passing-notes within a simple framework of four parts without
ever appearing to be as coolly calculated as J. G. Walther’s motifs in ‘Ach
Gott und Herr’.
It is significant that the chorale nearest to being doctrinaire in its figurae
is the most antique one, the three-verse ‘Christ ist erstanden’. By contrast,
‘O Traurigkeit, O Herzeleid’ (Example 105) has a later, more original tech-
nique. It was composed à 4, the melody not written in first: ‘molt’adagio’ and
slurs belong to the same operation, and the soprano passing-notes (tierces
238 BWV 599
Example 105
coulées) were soon added between original minims. Evidently Bach knew
the Affekt before he knew many of the notes, for the opening (including key)
already settles both mood and style, with new motifs easy to adapt to a com-
pelling harmony. Few Ob motifs are actually graphic – even the falling motif
of BWV 637 is metaphorical – but they have often been seen as ‘expressing’
the dogma or chief meaning of the hymn, especially if derived from the
melody, ‘contrapunctsweise zum gantzen Choral durch und durch geführt’,
as Praetorius said (‘contrapuntally developed through the whole chorale’:
Musae Sioniae, 1610). Canons invite symbolic interpretation, whether at
the octave or fifth, in close stretto or not.
A motif may emphasize a word in the text, as when the first notes of
the melody in BWV 632 are taken and used throughout the movement as if
repeating the opening vocative, ‘Herr Jesu Christ’. Coloratura settings suit
hymns concerned with prayer, complaint or trouble. Weimar cantatas too
use motifs to convey associations, e.g. with tumult in ‘Mit unsrer Macht’
BWV 80.ii or Advent in ‘Nun komm’ BWV 61.i. Less tangible or verifiable
is the significance of numbers: the multiples of 12 that seem to operate
(24 listed catechism texts, 60 seasonal hymns, etc.), the 158 notes in the
ostinato of ‘In dir ist Freude’ (158 = ‘Johann Sebastian Bach’), and so on.
BWV 599 Nun komm, der Heiden Heiland

Further copies: by or via J. T. Krebs, J. C. Oley, C. F. Penzel, J. P. Kirnberger,
J. C. Kittel.
Two staves; second half of b. 7 corrected in tablature.
The TEXT is Luther’s translation of Ambrose’s Advent hymn Veni redemptor

gentium, Erfurt 1524. From at least c. 1600, chief hymn of the four Advent
Sundays, given in Latin and German in several Leipzig books (Vopelius
1682).
239 BWV 599
Nun komm, der Heiden Heiland, Come now, Saviour of the heathen,
der Jungfrauen Kind erkannt, acknowledged child of the Virgin,
des sich wundert alle Welt, at whom all the world marvels [that]
Gott solch Geburt ihm bestellt. God provided him with such a birth.
Four further verses concern the advent, the light in the darkness, and a
doxology.
The MELODY, published with the text, simplifies the Latin hymn (Example
106). Its form in BWV 659a, 660a and 661a is as in the Weissenfels hymnbook
of 1714 (NBA IV/2 KB p. 76) and it frequently opened hymnbooks. Set in
659, 660, 661, and 699, also in cantatas for Advent I: 36 (1731), 61 (1714,
1723) and 62 (1724 etc.). As in Buxtehude, the cantatas have the beat on
first and fifth notes (Example 106), Schein 1645 and Vopelius on the second
and fourth. Luther’s version (Babst, 1545) draws out the opening phrase
to produce a 2 12 -bar phrase, as in BWV 599. On the uncommon key of A
minor for this melody, see p. 235 above.
Example 106
The dotted pedal rhythms of BWV 599 have been seen as ouverture-like
(Luedtke 1918 p. 54), ‘a festive entrance-music for the King of Heaven’
(Arfken 1965 pp. 46ff.), as if recalling the opening of Cantata 61. But neither
tempo nor motif support this interpretation. More immediately striking is
the series of falling phrases (Keller 1948 p. 151), the ‘descente sur terre du
Sauveur’ (Chailley 1974 p. 196), falling figures being appropriate for both
Advent and the Incarnation (Meyer 1987). But the text does not say the
Saviour descends, and just as possible is that the main pattern is a so-called
‘talking figure’, i.e. it repeats ‘Now come, now come’.
The setting introduces various motifs heard again in the Ob. Not least
is the one used for texts referring to Life (the little anapaest), although not
once does it appear in as simple a form as in BWV 605. Two details are that
the motif could have been used more than it is, and the melody is much
less prominent or even recognizable than in BWV 659, 660 or 661. This
appears to be due as much to the density of motif affecting the melody, with
rhetorical rests in bb. 1, 8, as to the harmony, which is new even when a
previous passage could have been repeated (e.g. bb. 1–2 in bb. 8–9).
A more appropriate stylistic allusion could be the ‘French prelude’, asso-
ciated with lute or harpsichord and producing rich harmonies of the kind
240 BWV 599–600
found here. One typical way of breaking chords involved the same motif
as BWV 599, found both in Louis Marchand’s G minor Suite (1702) and
much earlier: see Example 107. Such chord-breaking was known both to
Example 107
the ‘old good French’ masters admired by J. S. Bach (Dok III p. 288), and to
German composers such as Froberger and Fischer (also admired) who left
performers to break the opening chords themselves. The D major Toccata
for Harpsichord uses it more boisterously. For as subdued an effect as here,
worked in five parts in awesome expectation of the Incarnation, one needs
to look at the ‘Et incarnatus’ from the Mass in B minor.
BWV 600 Gott, durch deine Güte / Gottes Sohn

ist kommen
Further copies: by or via J. T. Krebs, J. C. Oley, another contemporary
MS (SBB N. Mus.ms. 10117), J. P. Kirnberger, Mempell–Preller, J. C.
Kittel.
Two staves; in P 283, soprano ‘Man. Princip. 8 F.’, tenor ‘Ped. Tromp. 8 F’.
Canonic voices revised in bb. 1/2, 13/14.
J. Spangenberg’s TEXT was published in 1544 to the same melody as

‘Gottes Sohn ist kommen’, a hymn after the sermon. J. C. Olearius
(Jubilirende Liederfreude, Arnstadt 1717) calls it the old Thuringian Advent
hymn.
Gott, durch deine Güte, God, through your goodness,

wolst uns arme Leute [we beg you] us poor people
Herze, Sinn und Gemüte – heart, mind and soul –
für des Teufels Wüten against the raging of the devil
am Leben und im Todt in life and in death
gnädiglich behüten. graciously to preserve.
241 BWV 600
Three verses address the Persons of the Trinity in turn. The TEXT of ‘Gottes
Sohn ist kommen’ (1531) was also found in hymnbooks of the Bohemian
Brethren.
Gottes Sohn ist kommen God’s Son is come

uns allen zu Frommen to all of us believers
hie auf diese Erden here on this earth
in armen Gebärden, in lowly guise,
dass er uns von Sünde that he might free and release
freie und entbinde. us from sin.
Eight further verses describe the purpose of Advent, ending with a prayer
for faith.
The pre-Reformation MELODY, belonging to the hymn Ave ierarchia

celestis et pia (Terry 1921 p. 175) was published in 1544 to both texts in
different books. It is used in BWV 703 and 724 and harmonized in BWV 318
(Example 108).
Example 108
The ‘registration’ indicates that the canonic voices are to sound at the pitch
notated, differentiated flue/reed. Although these stops were on the Weimar
organ, this is no normal registration, for P 283 is a ‘short score’ in which
pedal could have taken either tenor or bass. Was the setting originally made
with no thought as to how it was to be realized? When the registration was
added is unknown, but if the Weimar pedal extended only to e it could have
taken either bass (cf. BWV 645 and 650) or the tenor an octave lower with
4 reed as in BWV 608, with which BWV 600 forms a pair. This is forbidden
neither by the compass nor by the ‘registration’. Since, as in BWV 608, the
heading ‘à 2 Clav’ is not authentic, the crossing in b. 22 suggests that nowhere
else in the Ob are two manuals obligatory either, even if indicated in P 283.
The left hand is unlikely to be separately registered with 16 (BG 25.ii), since
‘the right hand parts are braced together, and the brace was extended to
include the left hand as well’ (Novello 15) – i.e., the 8 registration serves
both hands, with crotchets more détaché than the quavers.
242 BWV 600–601
The 3/2 canon for a chorale melody found normally in duple time is
also hinted at in J. G. Walther’s F major setting of the same chorale (Vers 3),
and both composers knew canons in which the cantus has to be altered,
e.g. the ‘Veni sancte spiritus’ of G. G. Nivers’s Deuxième Livre, 1667. There
may also have been a tradition for falling motifs for a text speaking of
‘Gottes Sohn’, as in Buttstedt’s setting. Similarly, the almost doctrinaire
combination of three note-values (minims, crotchets, quavers) can be found
in a less strict form elsewhere, e.g. in Pachelbel’s ‘Nun lob mein’ Seel’, copied
in P 803. Perhaps the canon refers to v. 2, ‘He comes . . . to teach the
people’ (Chailley 1974 p. 124), but discussions of symbolism, as in Meyer
1987, forget how common it was to set Christmas and Passiontide melodies
canonically.
The bass line’s crotchets paraphrase the canonic melody at first, then
have a recurring shape (bb. 4, 8, 12, 21). The alto begins like that of
BWV 608, and remains within the ambit of the right hand by contrary-
motion figures, all derived from a little note-pattern of falling quavers. It is
this figure that produces the B A C H motif in b. 16 alto, but nothing further
indicates whether B A C H was deliberate, whether if it were deliberate its
position was calculated (b. 16 of 23 = Golden Section), and whether if it
were calculated it alludes somehow to the text (‘in lowly guise’).
Despite a masterly diatonic harmony, in which each problematic mo-
ment of the canon is ‘explained’ by accented passing-notes (see particularly
bb. 8–18), there is a strained feel to much of it, not to say unnecessary com-
plications (b. 22). But very mellifluous are the bars repeated in the second
half (bb. 1–4 = 18–21), and harmonizing the ninth produced by the canon
in b. 5 as a brief 6/4/2 is ingenious. One has the impression of a composer
pushing harmonic boundaries less for expressive than technical purposes,
though perhaps for him everything was ‘ad majorem gloriam dei’.
BWV 601 Herr Christ, der ein’ge Gottessohn / Herr Gott

nun sei gepreiset
Further copies: by or via J. T. Krebs, J. G. Walther, J. P. Kirnberger, J. C. Oley,
C. F. Penzel, J. C. Kittel (Lpz MB Poel 39: with figured-bass chorale Anh II
75); also ‘Neumeister Collection’ (C time).
Two staves; first title only in other MSS (in P 283 the second was added?).
The TEXT of E. Cruciger’s Christmas hymn was published in 1524, be-

coming the chief hymn for Third and Fourth Sunday in Advent in Weimar
hymnbooks 1708, 1713.
243 BWV 601
Herr Christ, der einig Gottes-Sohn Lord Christ, the only Son of God,
Vaters in Ewigkeit, of the Eternal Father,
aus sein’m Herzen entsprossen, sprouting from his heart,
gleichwie geschrieben steht, as is written:
er ist der Morgensterne, He is the morning star,
sein Glänzen streckt er ferne stretching his rays to the distance,
vor andern Sternen klar. brighter than other stars.
The five verses are a prayer and meditation on Christmas.
The second TEXT was published in Bapst’s hymnbook of 1553, being a grace
after meals, and sung to the melody below from at least 1609 (Terry 1921
p. 184).
Herr Gott, nun sei gepreiset, Lord God, now be glorified,

wir sagen frohen Dank, we give joyful thanks,
dass du uns Gnad’ erwiesen, that you have shown us grace,
gegeben Speis’ und Trank, given us food and drink
dein mildes Herz zu merken, to remember your liberal heart,
den Glauben uns zu stärken, to strengthen our faith,
dass du seist unser Gott. that you are our God.
Verse 3 gives a more symbolic aspect to meals: through Christ we avoid

hunger.
The MELODY, published with the first text, derived ultimately from the
song ‘Mein Freud möcht sich wohl mehren’ (Lochamer Liederbuch); its AAB
form is as in Example 109. Also in BWV 698 and Advent Cantatas 96, 164
and probably 132. As with BWV 603, 612, 632 and 633, Bach appears to
have added a repeat.
Example 109
The simple, straightforward technique supports the idea that this chorale-
setting served as the Ob’s basic model. In ‘Neumeister’, its form is AAB –
perhaps an earlier form of the movement, to judge by a few differences
between it and P 283 (Stinson 1993 pp. 473f.). Although BWV 601 uses
244 BWV 601–602
motifs heard elsewhere in the Ob but more simply, its simplicity should not
be overstated: not only is there an incipient canon in b. 1 (cf. BWV 599 b. 3)
but no other composer is likely to produce so many seventh, ninth and 6/5
chords on the beat, or extend a simple motif twice (pedal b. 1, pedal b. 3
into the cadence). The subtlety is hardly from the Arnstadt years (as Wolff
2000 p. 94 suggests).
Fanciful interpretations include Schweitzer’s (the pedal motif is a ‘motif
de la quiétude joyeuse’, as in the last variation of BWV 767: 1905 p. 349) and
Chailley’s (the motif is ‘almost visually’ a reference to the morning star: 1974
p. 129). Dietrich finds the bass motif often in variations of Buttstedt, Böhm
and Vetter (1929 pp. 44–5), and other examples can be found in Walther
and BWV 1115. If the motif was so common, BWV 601 must represent a
conscious attempt to create new language from it, for here it has two versions
(manual, pedal), with rich harmony, inversus forms, thorough imitation,
some contrasting scale motifs, and unification through repetition (each
half ends similarly, thus four times). There being so many broken chords
produces a sweetness of harmony highly contrasted with the settings either
side of it.
BWV 602 Lob sei dem allmächtigen Gott

Further copies: by or via J. T. Krebs, J. C. Oley, J. P. Kirnberger, and J. C.
Kittel.
Two staves.
M. Weisse’s Advent TEXT was published in 1531 for the Bohemian

Brethren.
Lob sei dem allmächtigen Gott, Praise be to Almighty God

der unser sich erbarmet hat, who has been merciful to us,
gesandt sein’n allerliebsten Sohn, has sent his well-beloved Son
aus ihm geborn in höchsten Thron. born of Him in the highest throne.
The following thirteen verses relate the purpose of Christmas and the danger
of ‘not hearing the voice of the Son’, and close with a doxology.
The MELODY (Example 110), published with the text, belonged to

‘Conditor (or Creator) alme siderum’, Vespers hymn for Advent I in the
Liber usualis. The melody of BWV 704 begins differently: the source for
BWV 602’s is unknown but shows no ambiguity in P 283 (written out first),
except for the last note; see below.
245 BWV 602–603
Example 110
As in BWV 599, the pedal and manual motifs are complementary but dis-
tinct, the manual’s perhaps derived from the melody (b. 5), the pedal’s built
on a pattern for alternate-foot pedalling. At times it brings the inner parts
with it, unlike most Ob chorales, creating new harmonies in b. 5. Perhaps the
falling thirds in the melody, less striking than in the Gregorian version, sug-
gested to Bach the various forms of the manual’s motif, just as the Gregorian
cadence suggested the close on A (cf. BWV 704). Bar 8 shows the manual
motif to be no idle decoration of chords but itself to motivate harmonic
progression. Twice the bass motif begins a sequence, is then drawn out
(bb. 3, 7), and a third time falls to the lowest note in the last bar.
Despite attempts to show otherwise, it is difficult to feel sure that
the motifs refer to any particular verse (‘leading to eternal light’ in v. 2:
Vogelsänger 1972b) or dogma – the ‘coming down of divine Majesty’ in the
falling bass (Keller 1948 p. 152) or the union between Father and Son in
the many thirds and sixths (Chailley 1974 p. 186). The pedal’s motif and its
tie appear in Walther’s Praecepta of 1708 as one of the ways to embellish a
simple progression (here F E D C), as does the little dactyl pattern, and one
can see both of them being worked here towards a new harmonic momen-
tum. The melody originally ended on the third beat of the penultimate bar
(crotchet complete with fermata in P 283), but the motifs, especially in
the pedal, take over, resulting in an extra bar, as if the long note a were a
Gregorian alleluia. This is like the long final g for ‘Kyrie’ in BWV 604 except
as that one sinks, so this one rises exultingly to the top note of the piece.
BWV 603 Puer natus in Bethlehem

Further copies: by or via J. T. Krebs, J. G. Walther, J. C. Oley, J. P. Kirnberger
and J. C. Kittel.
Two staves.
The Latin TEXT of the traditional Christmas hymn ‘Puer natus’ was pub-
lished by J. Klug in 1543 with a German translation; it became associated
also with Epiphany, in particular v. 4 with its reference to the Magi (Stiller
1970 p. 224).
246 BWV 603
Puer natus in Bethlehem, A boy is born in Bethlehem,

Bethlehem,
unde gaudet Jerusalem. wherefore Jerusalem rejoices.
Alleluia, alleluia.
The MELODY originated as the descant line to an early tenor melody

of which a later version is used in BWV 607 (Terry 1921 p. 287). Apart
from BWV 603, the descant melody is used in Cantata 65 (Epiphany 1724):
Example 111.
Example 111
The last bar of P 283 has two beats, the second with a fermata and passing
to a custos for B (flat, natural?); after this is a repeat mark, looking like an
afterthought. Whether ‘he meant the prelude to be repeated ad lib., and to
end eventually on the second beat of the bar’ (Novello 15), or simply played
twice, is unclear, but any such repetition reflects the repetitious text itself
(twelve short verses), as if taking further the repeated half of BWV 601. The
ending provided in some editions is less striking than the open, bare Gs the
composer apparently intended.
The accompaniment to BWV 603 is in the classic Ob manner: an active
and intimate motif between the two hands is underpinned by a developed,
almost ostinato descending motif in the pedal part, which is itself highly
idiomatic. Both motifs syncopate the harmony, as in a different way do
those of the preceding chorale, and both are persistent, making of every
bar an unrivalled piece of harmony. Naturally, the rocking quaver motion
(Example 112) has been credited with picturing the swaddling bands, and
Example 112
the pedal line the steps of the worshipping Magi (Schweitzer 1905 p. 349)
or even the Saviour’s descent to earth (Chailley 1974 p. 212). As the text
refers to no swaddling bands, reverential steps or descending Saviour, such
247 BWV 603–604
interpretations are conjectural, and the very importance of this text through-
out the Christmas season suggests that it is no mere accumulation of
Christmas images.
Despite the fall in each pedal phrase, the overall sense is of a rising,
intensive bass line. Every line of the chorale sees a rising sequence in the
bass below more and more imitative and therefore more and more tense
inner parts. The response to Christmas seems to be awe or fear rather than
jollity, and however one interprets the powerful lines in both pedal and
manual, their gesture is obviously very different from the pastoral canon in
Walther’s (contemporary?) setting of the same chorale.
BWV 604 Gelobet seist du, Jesu Christ

Further copies: by or via J. T. Krebs, J. G. Walther (with BWV 722), J. C.
Oley, C. F. Penzel, J. P. Kirnberger, J. G. Müthel, and J. C. Kittel.
Two staves; headed in P 283 (not in Krebs’s copy) ‘à 2 Clav. & Ped.’.
The TEXT of vv. 2–7 was derived in part by Luther from a Low German
version of Notker’s Christmas sequence ‘Grates nunc omnes reddamus’ and
became a main hymn of Christmas.
Gelobet seist du, Jesu Christ, Praised be you, Jesu Christ,

dass du Mensch geboren bist that you are born man
von einer Jungfrau, das ist wahr; of a Virgin, that is a truth;
des freuet sich der Engelschar. in this the angel host rejoices.
Kyrieleis. God have mercy.
Six further verses concern the light of the world, the Son ‘leading us from
the vale of misery’.
The MELODY was published with the text in 1524 and is ultimately derived
from the plainsong (Terry 1921 p. 169): Example 113. In addition to the
chorale BWV 314, it appears in BWV 697, 722, 722a and 723, in Cantatas 64
(1723 etc.) and 91 (1724), and in the Christmas Oratorio (First and Third
Days of Christmas).
Despite a conspicuous pedal motif, the accompaniment is less motivic than

elsewhere; nor is pedal needed for the bass-line. As in BWV 605, broken
harmonies make a continuous surround, but now incline to the ‘soft’ mixo-
lydian, and in both chorales there are several main beats without thirds.
Again, the melody inspired hidden allusions, as in bb. 1–2, alto (paraphrases
248 BWV 604–605
Example 113
line 2’s rise) and pedal (its fall). And again the pedal motif is typically
alternate-foot, ‘answering’ the rising inner voices, which then fall when it
rises (penultimate bar – the result of second thoughts in P 283?). The placing
of the pedal motif is neither repetitious nor predictable, but it runs into
cadences, including the final plagal, in a similar key-scheme to BWV 697’s.
The characteristic accompaniment leads to several en passant modu-
lations, with inner parts moving alternately, simply and by step, accented
passing-notes or short suspensions, seldom of more than one semiquaver
at once, and the accompaniment not as intricate as it could have been. The
melody, though its modest decorations consist of familiar patterns, is pre-
sented in a new guise, lyrical, even rapturous. In part, the sweetness comes
from the mixolydian harmonies (unlike those of the more diatonic Cantata
64.ii), with a tendency towards C major and, at the beginning, even F major.
Not the least striking effect is the bare fifth at the beginning of b. 8. But that
the mixolydian has more than one Affekt is clear from BWV 635, where it
is altogether more robust.
BWV 605 Der Tag, der ist so freudenreich

Further copies: by or via J. T. Krebs, J. C. Oley, J. P. Kirnberger, J. C. Kittel,
J. G. Müthel.
Two staves; headed in P 283 ‘à 2 Clav. et Ped.’, last four bars in tablature.
The TEXT of the first two verses, a pre-Reformation translation of the hymn
‘Dies est laetitiae’, had three further verses when published in 1525.
Der Tag, der ist so freudenreich This is the day so full of joy
aller Kreature; for all creatures;
denn Gottes Sohn vom Himmelreich because God’s Son from Heaven
über die Nature transcending nature
von einer Jungfrau ist geborn. is born of a Virgin.
249 BWV 605
Maria, du bist auserkorn, Mary, you are chosen

dass du Mutter wärest. to be the mother.
Was geschah so wundergleich? Was anything so miraculous?
Gottes Sohn vom Himmelreich God’s Son from Heaven
der ist Mensch geboren. he is born man.
The orthodox message appears in v. 2:
Wär uns das Kindlein nicht geborn, Had the child not been born to us,
So wärn wir allzumal verlorn. we would be altogether lost.
The MELODY, probably fifteenth-century, was published in 1529. Apart

from BWV 719, it appears only in the harmonization BWV 294
(Example 114). Only with difficulty does v. 1 fit the melody of BWV 605
(particularly in lines 2 and 4), which suggests either that a later verse was in
the composer’s mind or that the other text, ‘Ein Kindelein so löblich’ (see
BWV 719), was intended, its syllables a better fit. This text often appeared
as the second verse of ‘Der Tag’, e.g. in the Schemelli Gesangbuch, Leipzig
1736.
Example 114
As in BWV 604, the inner motif is dispersed between two parts, produc-
ing a continuous line. Early signs are the motif ’s simplicity, persistence
and even a notation whose differences (i.e. with or without tied note) are
not always obviously intended, as is also the case with the pedal phrase of
BWV 610. If the notation is followed, and rests taken as specified, many
chords are without the third, e.g. twice in the first two bars. (See also
BWV 604.) Other ‘early’ signs are that pedal begins and ends with the
melody’s lines, that these leave the middle parts with a void to fill, again un-
like BWV 604, that the bass has more falling-fifth cadences than usual, that
the left-hand rhythm barely changes, and that the harmony has few accented
passing-notes. The dissonance in bb. 3, 8, logical with the bass, suggests a
250 BWV 605–606
maturing harmony, however, as does the falling bass-line, and it could be

that the ‘joy’ of the hymn lies in its simple ‘rhythmic vitality’ (Stinson 1996
p. 83).
Again, there is a mixolydian flavour, with some dozen fs, making it
unlikely that the sudden f in b. 3 evokes the ‘coming of God’s Son as
a coming towards suffering’ (Arfken 1965 pp. 46ff.) or that the one in
b. 18 evokes the line ‘O, sweet Jesu Christ’ of v. 2 (Vogelsänger 1972). There
seems little agreement as to whether the left-hand motif explores the motif
de la joie dactyl (Schweitzer 1905 p. 352, where this is called an Easter
chorale), or pictures the rocking cradle (Keller 1948 p. 153), or symbolizes
the ‘super/contra-natural’ virgin birth (Arfken 1965). As in BWV 604, the
inner parts sometimes resemble the melody – see the alto of bb. 19–20 and
line 5 – while as in BWV 603, it is the scalar bass that gives momentum and
suggests a common tempo (crotchet there = quaver here).
BWV 606 Vom Himmel hoch, da komm ich her

Further copies: by or via J. T. Krebs, J. G. Walther, J. C. Oley, C. F. Penzel,
J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT of Luther’s hymn was published in 1539, v. 1 largely from the
song ‘Ich komm aus fremden Landen her’, and became associated with
the whole season (Gojowy 1972), especially accompanying the Christmas
manger-play.
Vom Himmel hoch da komm ich her, From Heaven on high I come,
ich bring euch gute neue Mär; bringing you good new tidings;
der guten Mär bring ich so viel, of good tidings I bring so much
davon ich sing’n und sagen will. of which I will sing and speak.
v. 15
Lob, Ehr sei Gott im höchsten Thron, Praise, honour be to God on the
highest throne,
der uns schenkt seinen eigen Sohn. who gives us his own son.
Des freuen sich der Engel Schar Thus the band of angels rejoices
und singen uns solch neues Jahr. And sings to us of such a new year.
The MELODY (one of three melodies with this text at first) was published
in 1539 (Terry 1921 p. 304), used in BWV 606, 700, 701, 738, 738a, 769
251 BWV 606–607
(five movements), Christmas Oratorio (three) and the Magnificat BWV 243a:
Example 115.
Example 115
While the off-beat semiquaver motif, a figura suspirans, produces runs typ-
ical of chorales concerned with angels (cf. BWV 607, 701, 769), no line of
BWV 606 is particularly scale-like. But the line derived from this motif –
up-, down-, in-turning – is a particularly telling example of the Ob’s figu-
ral technique. Superficially, the results are sometimes like those elsewhere,
such as Walther’s ‘Vom Himmel hoch’, but BWV 606 treats the motif more
freely, as required by the melody or the striding pedal (Schweitzer’s thème
de la démarche, familiar in earlier chorale variations). Perhaps the first and
last notes of each line are pulled out to minims to allow the semiquavers to
suggest the flurry of angels; compare BWV 700, 701, 738 and 769.
The syncopated final pedal phrase recalls another setting (BWV 738
b. 12) as well as ‘Da Jesus an dem Kreuze stund’ BWV 621, and writers have
discerned both here and in the semiquaver groups some cross-figures – see
Glossary (Meyer 1987 p. 28). Although the inner motif spills over into the
melody more than usual, except in old chorale-variations, its impetus finally
runs out towards not a full chord but bare Ds, just as in the Easter chorale
BWV 628. Despite a similar motif between these two D major chorales,
their treatment is quite different: BWV 606 is often harmonized in thirds,
BWV 628 more spare and on-driving. The bass-line’s shape is more or less
infinitely adaptable, and it is surely more than an ‘accompaniment to the
cantus firmus’ in the way that BWV 605 is (Stinson 1994), although the idea
that some Ob pieces form pairs of similar settings is certainly plausible.
BWV 607 Vom Himmel kam der Engel Schaar

Further copies: by or via J. T. Krebs, J. P. Kirnberger, J. C. Kittel.
Two staves; three in Brussels 12.102 (a Kirnberger copyist).
The TEXT of Luther’s last Christmas hymn (Stapel 1950 p. 142) was pub-
lished in 1543; its metre matches that of ‘Vom Himmel hoch’, to whose tune
it was often set.
252 BWV 607
Vom Himmel kam der Engel From Heaven came the host of angels,
Schaar
erschien den Hirten offenbar; appearing openly to the shepherds;
sie sagten ihn’n ‘Ein Kindlein zart they said to them, ‘A gentle child
das liegt dort in der Krippen hart.’ lies there in the hard crib.’
Five further verses centre on Luther’s message of Christmas, e.g. v. 4:
Was kann euch tun die Sünd und Tod? What can sin and death do to you?
Ihr habt mit euch den wahren Gott. You have with you the true God.
The MELODY was published in 1543 to ‘Puer natus in Bethlehem’; in

1553 it is found as the tenor to a soprano melody also associated with
that text and used for BWV 603 (Terry 1921 pp. 286, 309). The melodies
are closely related, and the BWV 607 form is not used elsewhere by J. S.
Bach.
BG 25.ii’s suggested two manuals separate alto and tenor unjustifiably; see in
particular b. 7, clearly written for one manual. (In BWV 617, the two upper
parts are more obviously paired.) The cramped handwriting of P 283 looks
as if the composer added the semiquaver runs to a harmonization already
on paper, one with more of G minor than it need have, turning Christmas
into an occasion for deep thought. The tempo must be slower than in
BWV 603, despite a comparable pedal part.
The descending scales for Christmas chorales, as in BWV 697 and 700,
are nowhere clearer than in the present movement, where they run at two
levels: a walking bass at quarter-speed follows the scurrying inner figures as
they rise and fall, emphasizing the beats, which exceptionally are without
syncopation, and marking each new line of the cantus by a rest. The resulting
harmony is full of rich, passing-note progressions in which most main beats
are simple concords. The scale line gradually widens, not only running
into the melody but eventually across it, twice right through three octaves,
when the pedal passes in contrary motion. In this way the motif is exploited
farther than in any other chorale, for example Buxtehude’s ‘Ich ruf zu dir’.
Note that the rushing angels supposedly represented by the scales (Spitta I
p. 602) are not referred to in the text itself.
The bass line’s first three phrases have four bars, the next phrase six, giving
an impetus towards the end even more striking than in BWV 612 or 626.
Similarly, while there is some back-reference, other potential repetitions are
varied (b. 7 = b. 3, b. 15 = b. 6). So developed has figural treatment become
in this setting that not only are the tenor and bass scales, in their different
way, pushed to a limit up and down, but the mood is elusive: a robust
flurrying or a subdued meditation?
253 BWV 608
BWV 608 In dulci jubilo

Further copies: by or via J. C. Oley, J. P. Kirnberger and J. C. Kittel.
Two staves; only direction in P 283: ‘Ped.’ by the opening note of the tenor
canon.
The TEXT of the pre-Reformation hymn appeared in an early Lutheran

hymnbook (Klug, 1535):
In dulci jubilo, In sweet joy

nun singet und seid froh! let us sing and rejoice!
Unsers Herzens Wonne The rapture of our heart
liegt in praesepio, lies in a manger,
und leuchtet als die Sonne and shines like the sun
matris in gremio: at his mother’s bosom
Alpha es et O, Alpha es et O. You are alpha and omega.
V. 3 begins
O patris charitas O love of the father,

O nati lenitas! O gentleness of the newborn!
Versions were known with one, three and four verses, with pure German
texts, with various dialect texts, and with the mariolatrous references
pruned.
The MELODY exists in variously embellished forms, e.g. BWV 368

(Example 116), and is used in BWV 729, 729a and 751.
Example 116
The notation of BWV 608 is that of a ‘short score’ on two staves. The
four parts enclose the canonic cantus as a tenor line at its required pitch,
254 BWV 608
beginning at a and rising to f . The tempo must be slower than in BWV 603
despite a comparable pedal part. With this kind of bass line, and because of
its compass, pedal plays (i) the tenor, at (ii) an octave lower than written, with
4 stop. P 283 thus notates the effect intended without further information
on how to achieve it – compare BWV 600.
Furthermore, like other old Christmas hymns, this is written in 3/2, now
divided not into quavers and semiquavers but into triplet quavers. It is often
assumed that the opening crotchets are to be played as triplets, although in
P 283 they are written as equally as possible, with only subsequent revision
of the alto at bb. 23–7 – a sign either of a change of mind or of a different
thematic pattern. There is an implied musette-drone A running throughout
the first twenty-four bars, right through to the very A of b. 25, and this
is best realized by equal repeated crotchets in bb. 3, 4, 7, 8, despite the
later triplets. (For another drone, see BWV 751.) Against triplet crotchets
there is a further argument: as in BWV 617, each voice subdivides the bar
differently, into minims, crotchets and triplet quavers, and since after all
the triplet quavers are ‘misnotated’ (they should be crotchets),∗ it seems the
composer meant a clear distinction between the patterns. Agricola’s remark
in 1769 that J. S. Bach distinguished between dotted and triplet quavers
unless ‘extremely fast’ is hardly relevant here (see also BWV 682), since there
are no dotted notes, and Agricola is not referring to this sort of music.
The canon’s similarity to J. G. Walther’s ‘In dulci jubilo’ is striking, but
which came first is unknown. Rather, a pastoral-canonic treatment of the
melody was already at least a couple of centuries old, as in Fridolin Sicher’s
Tablature Book (see Edler 1982 p. 229), and Johann Michael Bach had ten-
tatively used both canon and drone. Also striking is that the harmonization
BWV 368 decorates the melody with one of BWV 608’s motifs and develops
it towards the end, including a diminished version in bb. 31–2. The text
itself implies gentleness rather than brilliance.
The canon is strict except for bb. 14–15, and for the first twenty-four
bars the accompanying line is also treated canonically. Though this only
paraphrases what is a tonic drone, it is unique to the setting, despite a fitful
tradition for canonic accompaniment from Scheidt through Walther to
BWV 769. The motif, which is imaginatively explored, descends in the first
bar like that of BWV 600 and, also like it, runs through to the final cadence.
Again, it produces accented passing-notes typical of the album, and unusual
syncopations in the repeated passage (bb. 10–16 = 18–24). So naturally is
it developed that it appears to be neither contrived nor superimposed even
when heard in canon above the final pedal point.
∗ So written ‘to make the triplets more easily distinguishable’ (Peters V). Often in sonatas of D. Scarlatti,
triplets are similarly notated twice too fast.
255 BWV 608–609
Both this pedal point and the F major chord of b. 25 may serve to depict
the text, the former ‘Alpha es et O’, the latter ‘leuchtet als die Sonne’. An array
of A major chords embroidered in such a way as this, more than a merely
traditional canon and drone, conveys an unmistakable impression of both
dulci and jubilo.
BWV 609 Lobt Gott, ihr Christen, allzugleich

Further copies: by or via J. T. Krebs, J. G. Walther, J. C. Oley, Mempell–
Preller, C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT of N. Herman’s eight-verse hymn was published in 1560,

becoming a general Christmas hymn, for the Second and Third Days in
some books.
Lobt Gott, ihr Christen alle gleich, Praise God, you Christians all
together,
in seinem höchsten Thron, in his highest throne,
der heut schleusst auf sein Himmelreich who today opens up his Heaven
und schenkt uns seinen Sohn, and presents us with his son.
und schenkt uns seinen Sohn.
Seven further verses reiterate the praise, the ‘opening up’ and the gift of a
Son.
The MELODY was published with the hymn in 1580, having earlier had
another text (Terry 1921 p. 259). It appears in Cantatas 151 (Example 117)
and 195 (different text), harmonized in BWV 375 and 376 and set in
BWV 732, 732a.
Example 117
In P 283, it looks as if the melody was written in first, then the bass (complete
with its two great ascents), then the inner parts. A standard procedure?
256 BWV 609–610
Comparison with BWV 606 shows this to be less dominated by a single

motif despite the chorales’ similar motion, figuration and texture in the
inner parts. In view of the unusually few tied notes and rests in BWV 609,
its chief motif should be understood as on-beat semiquavers, BWV 606’s
as off-beat: such distinction between similar but different figurae is often
found in the Ob. The present chorale is unusually homogeneous, and its
secondary motif (the tenor’s second semiquaver group) is developed more
fully in another chorale, BWV 624.
The thrusting quavers of the pedal line (which looks in P 283 to have
been composed before the middle voices) rise and fall, by step and leap,
twice up and down from D to d , and offering less a motif than a vivid
counterpoint to the chorale-melody. It is not clear why the various motifs
are mostly absent from b. 3 – for variety? – but the clamour is if anything
increased as the line rises to ‘the highest throne’. One is bound to wonder
whether Bach was vying with Walther and his ‘Lobt Gott, ihr Christen’ to
produce Christmas exuberance or whether Walther was inspired by it to try
for himself. As with BWV 606, the very brevity adds to the exultation, for it
becomes a type of emphasis.
BWV 610 Jesu, meine Freude

Further copies: by or via J. T. Krebs, J. G. Walther, J. C. Oley, C. F. Penzel,
J. P. Kirnberger, J. C. Kittel, J. G. Müthel.
Two staves; headed ‘Largo’ in P 283 (an addition?), but not in Krebs.
The TEXT of J. Franck’s six-verse hymn of 1653 became a popular Jesuslied

(Stiller 1970 p. 234), used at Epiphany and (Weimar hymnbook, 1708)
Christmas. Modelled on the song ‘Flora meine Freude, meine Seelenweide’,
1641 (Terry 1917 p. 261).
Jesu, meine Freude, Jesu, my joy,

meines Herzens Weide, pasture of my heart,
Jesu, meine Zier: Jesu, my jewel:
ach wie lang, ach lange oh how long, how long
ist dem Herzen bange is my heart afraid,
und verlangt nach dir! and longs for you!
Gottes Lamm, mein Braütigam, Lamb of God, my bridegroom,
ausser dir soll mir auf Erden there shall be for me on earth
nichts sonst liebers werden. nothing dearer than you.
The MELODY by J. Crüger, published with the text, took varied forms
in Bach (Example 118): BWV 713, 753 and 1105, Cantatas 64 (1723), 81
257 BWV 610
Example 118
(1724), 87 (other text, Rogation Sunday 1725) and 12 (no text, 1714), motet
BWV 227 (four times as chorale, once as cantus firmus, once as paraphrase)
and harmonization BWV 358.
As a Jesuslied the chorale is relevant to Christmas, Epiphany and the ‘urging

of faith in adversity’ (Cantata 12), and there is no difficulty in hearing in
the setting a strangely ‘fervent longing’ (‘sehnsuchtsvolle Innigkeit’, Spitta
I p. 590). The low pitch, the strong opening minor triad in the centre of
the keyboard, the lowest note of the organ played four times, the con-
stant motif, the false relations, the ‘Largo’: all join to produce this dense
effect. Perhaps a parody-text based on the hymn was in the composer’s
mind (‘Jesu, meine Freude, wird gebohren heute’: see Honders 1988
p. 45), although its semi-doggerel is hardly matched by the music’s elevated
intensity.
As an instance of the Ob’s material – new semiquaver shapes weaving
around the basic harmony – see Example 119. As in BWV 602, 606 and 609,
the accompaniment achieves intensity when two of the parts are in simple
thirds or sixths – an unexpected by-product of this motivic technique. The
unusually shaped motif creates shifting harmonies in three dense semi-
quaver lines, far beyond the formulae-ridden variations on this melody by
J. G. Walther, published in 1712 and also in C minor.
Example 119
As elsewhere, the motifs are not applied to every conceivable progression,

despite their essential elasticity, nor is there repetition when the first line
returns (compare b. 18 with b. 1), only when the effect is somewhat hidden
(compare b. 15 with b. 3). Also important is the character of the pedal
phrase, ostinato-like and running across the end of one chorale line (b. 4)
to give continuity. The difference in its notation (tie or rest) cannot be very
significant. Naturally it is the motifs that produce the striking harmonies,
258 BWV 610–611
particularly the A-F-F complexes in bb. 4, 18, 19. Bar 19 becomes a kind
of richly coloured version of b. 2, and it is certainly possible to play the
setting in such a way as to reflect lines in v. 2:
Lass den Satan wettern, Let Satan thunder,
lass die Welt erzittern, let the earth tremble,
mir steht Jesus bei. Jesus stands by me.
That there is no first and second-time bar probably results from the repeat
marks being an afterthought in P 283. BWV 610 shows much less clear
repeat-marks than BWV 601, and as it stands, b. 6 runs into b. 7, not b. 1.
Three further questions are: since pedal is not necessary, is this one of the
chorales implying that the title-page’s agenda was not original? And, if this
is ‘Largo’, why not BWV 637, 643, 604? – because BWV 610 is ‘paired’ with
BWV 611? And was C minor chosen with respect to temperament and if so,
which one: less equal at Weimar (thus harsher), more equal at Halle (thus
sweeter)?
BWV 611 Christum wir sollen loben schon

Further copies: by or via J. T. Krebs, J. C. Oley, C. F. Penzel, J. P. Kirnberger,
J. C. Kittel.
Two staves; headed in P 283 ‘Adagio’, ‘Corale in Alto’ (both subsequently?).
The TEXT is Luther’s adaptation of the Christmas hymn ‘A solis ortus car-
dine’. In Leipzig, used as a Vespers hymn on the Second Day of Christmas
(Stiller 1970 p. 222).
Christum wir sollen loben schon, We should indeed praise Christ,

der reinen Magd Marien Sohn, son of the pure Virgin Mary,
soweit die liebe Sonne leucht’ as long as the dear sun shines
und an aller Welt Ende reicht. and reaches to the ends of all the earth.
The alternative title in BWV 696, ‘Was fürchtst du, Feind Herodes, sehr’,
refers to Luther’s adaptation of the second part of the same Latin hymn,
beginning ‘Hostis Herodes impie’ (Terry 1921 p. 129). The two texts shared
a doxology.
Was fürchtst du, Feind Herodes, sehr, Why are you so afraid, foe Herod,
dass uns geborn kommt Christ der that Christ the Lord comes born
Herr? to us?
Er sucht kein sterblich Königreich, He seeks no mortal kingdom,
der zu uns bringt sein Himmelreich. he who brings his own Heaven to us.
259 BWV 611
The MELODY is adapted from the Latin hymn, published in 1524. Its form
in Cantata 121 (1724) is Example 120. In Scheidt, Scheidemann, Walther
and Witt’s Hymnbook (1715), the melody takes various forms, and the
first line also appears in BWV 696. Walther’s ‘canone infinito gradato’,
a setting derived from this melody, is called ‘A solis ortus cardine’, like
Grigny’s.
Example 120
Although P 283 looks like a short score leaving the organist free to realize it
as best he may (e.g. 4 pedal cantus firmus, bass in the left hand), in fact the
spacing leaves no room for choice: compare BWV 600, 608. Only in b. 14 is
the layout ambiguous, perhaps reflecting a later emendation? – the bracketed
‘upper pedal part’ may be a lh part. P 283 suggests that the composer first
began with a minim d and then added the passing-note; moreover, for 10 12
bars the cantus firmus notes were written with stems up, so at first intended
for the top line. In this case, therefore, a setting evolved independently of
any idea how it should be played?
Except for the canonic BWV 618 and 633/634, this is the only alto cantus
firmus. After the unusually dense ‘Jesu, meine Freude’ (on the recto side
of the same folio), the spacing is very wide: the opening notes span almost
the whole keyboard, with pedal point and bare effect (no third at first) as
different from BWV 610 as possible. Is the contrasting texture a ‘reaching
to the ends of the earth’ of v. 1?
So unusual a setting has invited interpretation. The hidden cantus re-
flects a reference in v. 5 to Jesus in his mother’s womb (Clark 1984 p. 57);
the compass C–c in b. 6 alludes to the ‘ends of the whole world’, the
chromatic fourth of b. 5 to the ‘pure Virgin’ (both as in v. 1). The adagio
scales express not boisterous Christmas joy but a ‘mystical contemplation’,
an ‘exaltation joyeuse dans ce soprano’ (Schweitzer 1905 p. 353). Within
the web of ascending and descending scales the inner melody moves largely
by step, obtrusive only when its notes are longer than the counterpoint’s.
One hardly notices the double canon in bb. 11–12: cantus firmus and pedal
at a half bar, soprano and tenor at a half beat, the contrary motion facil-
itated by the scale-lines. Perhaps it was the opening stepwise melody that
260 BWV 611–612
suggested the scale patterns and their rhythm, hence the tenor’s quasi-stretto
in b. 1.
Four-part counterpoint of short scale-like motifs against this same chant-
melody, also in D minor, is found in G. G. Nivers’s Deuxième Livre d’Orgue
(Paris, 1667). Deriving such motifs from the melody is not so common
in the Ob, and results in a rather disguised cantus firmus. It also suggests
that by an inventive use of scale fragments of varying length, the style was
maturing. Leaps are found chiefly in the accompaniment, and are treated
imitatively in the usual way. Although there are many ties, the exceptions
are often at main beats (bb. 2, 4, 7, 12), and the chorale’s ‘fluidity’ does
not depend solely on the constant suspensions, despite the many tied pedal
rhythms.
The final setting of the chorale in Cantata 121 (Second Day of Christmas,
1724) is also lyrical and somewhat drawn-out, with a cadence compara-
ble to BWV 611’s: see Example 121. The modal cadence of the original
dorian chorale is preserved, as it is in the setting BWV 696. BWV 611’s
‘Adagio’ rubric may indicate ‘slow’ (‘langsam’ in Walther’s Praecepta, 1708)
or ‘at ease’ (Frescobaldi’s Fiori musicali) and ‘conveniently’ (‘commode-
ment’, Brossard’s Dictionaire, 1705). But as Brossard points out, to play thus
almost always means ‘lentement’.
Example 121
BWV 612 Wir Christenleut

Further copies: by or via J. T. Krebs, J. G. Walther, J. C. Oley, another
contemporary (? Lpz MB MS 1), C. F. Penzel, J. P. Kirnberger, and J. C.
Kittel.
Two staves; last 2 12 bars in tablature in P 283.
The TEXT of C. Fuger’s Christmas hymn ‘Wir Christenleut’ was published

by 1593.
261 BWV 612
Wir Christenleut’ We Christian people

habn jetzund Freud, now have joy
weil uns zu Trost Christus ist because Christ for our solace is
Mensch geboren, born man,
hat uns erlöst. and has redeemed us.
Wer sich des tröst’ Who trusts in this
und glaubet fest, soll nicht werden and believes firmly, shall not be lost.
verloren.
The remaining four verses concern the message of Christmas:
Die Sünd macht Leid; Sin causes sorrow;

Christus bringt Freud, Christ brings joy,
weil er zu uns in diese Welt ist kommen. for he is come to us in this world.
The MELODY (Example 122) was published with the text in 1593 but is
older. The versions differ in the repeat of line 1: see BWV 710, 1090.
Example 122
In P 283 it looks as if the composer wrote out the cantus firmus first (e.g.
third note of b. 3 was a minim, b. 10 was thoroughly revised: KB p. 38), and
various revisions show him searching for a tense harmonization realized
through note-patterns. The result is a miniature ritornello shape, pushing
the closing pedal-point into the margin. Dots between the stave-lines at the
beginning of b. 9 suggest that the section bb. 9–15 is repeated (as in NBA
IV/1 and BWV 632) but the chorale is not known to have a repeat here.
Perhaps on the contrary, bb. 9–15 were an optional omission: because the
melody is already repetitive, there is a lot of G minor (though no two similar
phrases have the same harmony), and b. 16 follows naturally on b. 8.
It is possible that the composer associated the ‘glauben’ of v. 1 with
such a firm, striding pedal line, as in the Credo setting in Clavierübung III.
This pedal phrase is of great interest, being related to the manual motif,
simplifying and accompanying it (Example 123) much as the pedal subject
of BWV 664 simplifies its manual subject above. (Compare BWV 664 at
b. 10 with BWV 612 at b. 1.) It is immensely pliable: the phrase-lengths
are varied, but b is found untransposed in several bars (bb. 1, 3, 8, 11, 14).
The longest bass phrase is the last, its motif driving on relentlessly, in effect
262 BWV 612–613
Example 123
embellishing a chorale’s ideal bass-line. (The Fourth Brandenburg Concerto

finale has a comparably driving bass line.) The absence of pedal for two and
a half bars gives the impression of an episode, especially as the upper parts
are repeating material.
Like the two chorales preceding it in the Ob, BWV 612 reaches new
heights in composing-by-patterns. Although the same semiquaver motif re-
turns in later 9/8 movements (Prelude BWV 547, Goldberg Variation No. 24),
it seems to spring from a phrase which occurs in the melody no fewer than
five times, DCBA. Perhaps deriving a theme in this way, and thus unifying
melody and motif, is a way of ‘confirming’ the text (‘We, we . . .’). Compar-
ing the first two bars and the last three shows how a pattern can appear in
different parts, in different keys and with different harmonies spun out to
only two per bar when the melody has repeated notes (bb. 11–13), and all
of it over a quasi-ostinato bass. The subdued chorale-melody, evidently as
apt for Christmas as rushing angels, has surely prompted an introvert
setting.
BWV 613 Helft mir Gottes Güte preisen

Further copies: by or via C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT of P. Eber’s six-verse hymn, an Advent hymn also sung in Leipzig
on Sunday after Christmas and/or New Year’s Day (Gojowy 1972), was
published in 1569.
Helft mir Gottes Güte preisen, Help me to glorify God’s goodness,

ihr Christen insgemein, you Christians all together,
mit Gsang und andern Weisen with song and other melodies
ihm allzeit dankbar sein, to be ever thankful to him,
vornehmlich zu der Zeit, especially at the time
da sich das Jahr tut enden, when the year draws to an end,
die Sonn sich zu uns wenden, the sun turns towards us,
das neu Jahr ist nicht weit. the new year is not far.
263 BWV 613
The MELODY by W. Figulus (?) is one of two similar tunes published with
this text, which was given the other melody in Freylinghausen (1741). BWV
613’s version appears in Cantatas 16 (1726?), 28 (1725) and 183 (1726), all
in A minor: Example 124.
Example 124
As in the fragment ‘O Traurigkeit’, BWV 613’s handwriting suggests that

the piece was written into P 283 ‘probably only after 1740’ (Dadelsen 1958
p. 80), or ‘at least after 1730’ (Dadelsen 1963). Whether it was composed then
is uncertain, though from the way the motif derives so explicitly from the
melody, and from the absence of earlier copies, a late date seems likely. In its
texture, complete and incomplete cadences, motifs and their combination,
and even its repetition, the technique is close to the others’, and yet the
two pedal scales and Corellian bass lines seem rather out of place – more
‘objective’, with a less immediate Affekt. Why B minor is used is not known,
but it matches the doubtful Anh.II 54 and Anh.II 68.
As in BWV 644, the scales draw attention to ‘passing time’ but now not
in every bar, and although the general impression is of a concentration of
motifs, there are moments free of them. Nor is the tempo languid. While line
1 certainly provides the head of the motif, line 2 might supply its downward
run (Example 125). Imitations built on a repeated-note motif are often seen
Example 125
as ‘speaking’ or ‘confirming’ the opening line of the text, as if in unceasing,

oft-repeated praise. There is some repetition in the chorale (end b. 10 to
middle b. 12 = end b. 12 to middle b. 14, written out only once and given
repeat signs in P 283) though not as much as in the melody itself (bb. 1–4 =
5–8; bb. 15–16 = 3–4). Surprising too is the number of dominant–tonic
progressions. In view of the following chorale, the alto’s chromatic line for
the text ‘da sich das Jahr tut enden . . .’ is conspicuous; but the chromatic
line in the pedal six bars earlier has no such reference in any verse. Does the
264 BWV 613–614
astute combining of disparate motifs throughout make it more ‘objective’

than the next chorale?
BWV 614 Das alte Jahr vergangen ist

Further copies; by or via J. G. Walther, C. F. Penzel, J. P. Kirnberger, J. C.
Kittel.
Two staves: headed ‘à 2 Clav. & Ped.’ in P 283.
The TEXT of the first two verses was published by C. Stephani in 1568,
vv. 3–6 in 1588 (J. Steurlein).
Das alte Jahr vergangen ist; The old year has gone by;
wir danken dir, Herr Jesu Christ, we thank you, Lord Jesu Christ,
dass du uns in so grosser Gefahr that in such great danger
so gnädiglich behüt dies Jahr. you preserved us this year so graciously.
From v. 3:
vor falscher Lehr, Abgötterei from false teaching and idolatry

behüt uns, Herr, und steh uns bei. preserve us, Lord, and stand by us.
Three other verses pray for the coming year, and the final verse is a doxology.
The MELODY, Example 126, was not at first associated with this text.
Its five phrases were made to produce various stanzas, of eight lines
(aabcdcde in Steurlein), four, or six (aabcde in BWV 288, 289). See also
BWV 1091.
Example 126
The supposed ‘chromatic grief motif ’ has caused much speculation, since
neither the text nor the aeolian melody seems to require what has been de-
scribed as ‘the greatest intensity’ (Spitta I p. 593), a ‘melancholy’ (Schweitzer
1905 p. 355), ‘a prayer, anxiety for the future’ (Arfken 1965), marking
265 BWV 614
the juncture between ‘the past and the future’ (Chailley 1974 p. 100).
For once, perhaps, a biographical speculation is justified: the Old Year
1713 saw the death of Bach’s infant twins. But there is no ‘Adagio’ or
‘Largo’, and the chromatics could as well imply supplication as sadness.
Nor, since the final major chord corresponds to various hymnbooks (Terry
1921 p. 140), does it necessarily imply ‘hope’, as was once supposed. A
recent idea that the six falling and six rising chromatic notes represent
the year’s twelve months raises a question why such figures would not in
other chorales.
The relationship of BWV 614 to the chorales on either side is clear:
the sequence forms a clear reference point in the church year, though one
not shown in the Weimar hymnbook of 1713, where hymns correspond-
ing to BWV 614 and 615 are respectively Nos. 39 and 29. In Freyling-
hausen’s hymnbooks, ‘Das alte Jahr’ is a New Year hymn, for 1 January
not 31 December. The texts of both BWV 614 and 615 are addressed
to Jesus; the first contains thanks and prayer, the second praise and joy,
both in their own way presenting Jesus as Saviour. Both exploit their key
motif fully, and as one of the few coloraturas in the Ob, BWV 614’s melody
also manages to include a clear reference to its chromatic motif (b. 5). The
chromatic fourth itself may therefore be derived from the melody’s decora-
tion, and its answer in inversion (bb. 1, 3, 5, 6, 7, 11) or in canonic stretto
(bb. 3–5) are skilful developments. This fourth is familiar as one form of
passus duriusculus, according to Schütz’s pupil Bernhard (Williams 1997
pp. 98–9).
Since P 283 is a fair copy, whether the coloratura decorations were added
cannot be known. Either way, unlike most Ob chorales, this has few other
places during the twelve bars in which more chromatics could be easily
introduced. They are already used in many ways, without regular stretto,
regular answer or even regular phrase-length. For example, bb. 3–4 are
neither a simple repeat nor an entirely new version of bb. 1–2. On the
other hand, several of the cadences are noticeably straightforward in the
pedal (bb. 2, 6, 8, 12) and give a firm anchor-effect under the extraordi-
nary rising ‘sighing motif ’ of the final cadence, where the melody is quite
lost.
Nevertheless, the problem remains: is the ‘melancholy’ heard in it by
organists over the last century or so justified by the ‘objective’ traditionalism
of its key motif? Is the little melisma in b. 2 more ‘subjective’ than in the so-
called Arnstadt Chorales, such as BWV 726? Note that the isolated a a at the
beginning and the appoggiaturas at the end anticipate respectively the two
settings of ‘Vater unser’ in Clavierübung III, a prayer ardent rather than
sad.
266 BWV 615
BWV 615 In dir ist Freude

Further copies: by or via J. T. Krebs, C. G. Meissner, J. C. Oley, J. P. Kirnberger,
J. C. Kittel.
Two staves; in P 283, no directions of any kind.
The TEXT was published in 1598 by J. Lindemann as a two-verse Christmas

hymn (Terry 1921 p. 217).
In dir ist Freude In you is joy

in allem Leide, in all suffering,
O du süsser Jesu Christ! O sweet Jesu Christ!
Durch dich wir haben Through you we have
himmlische Gaben, heavenly gifts,
der du wahre Heiland bist; you who are the true Saviour;
hilfest von Schanden, you help us from shame,
rettest von Banden; you save us from fetters.
wer dir vertrauet, he who puts trust in you
hat wohl gebauet, has built well
wird ewig bleiben, Halleluja. and will live for ever, Hallelujah.
Zu deiner Güte To your goodness
steht unser Gmüte, our spirit holds fast,
an dir wir kleben to you we cling
im Tod und Leben; in death and life;
nichts kann uns scheiden, Halleluja. nothing can separate us, Hallelujah.
The MELODY derives from G. G. Gastoldi’s balletto L’innamorato, published

in 1591, already a hymn-tune in D. Spaiser’s hymnbook of 1609, and asso-
ciated with ‘In dir ist Freude’ by 1646. Leipzig documents show Gastoldi’s
Balletti à 5 and tricinia available there by 1604 and 1607 (Wustmann 1926
pp. 172, 315). The melody of BWV 615 is also very like the form in Witt’s
hymnbook of 1715: Example 127.
Example 127
The greatest possible change is rung between this and the preceding chorale.
Alone in the collection, BWV 615’s melody is split up and used in a web
267 BWV 615
of thematic allusion, called the ‘Böhmian manner’ by Spitta (I p. 593), in

which the whole melody only gradually becomes audible. (In Böhm’s ‘Allein
Gott in der Höhe’, as in Buxtehude’s ‘Von Gott will ich nicht lassen’, both
copied by Walther, the melody passes from one voice to another, becoming
thus somewhat sectional and varied.) Quasi-ostinatos in chorale-settings
are also found from time to time, as in Walther’s ‘Dies sind die heilgen zehn
Gebot’. But BWV 615 is more than its parts: its varying but unified texture, its
momentum, its irrepressible gusto, even its repetitions, are found nowhere
else.
The cantus can be heard more or less continuously in three sections as
follows:
A text lines 1, 2 bb. 9–12, top part

3 bb. 13–16, alto, then top part
4, 5 bb. 26–9, top part
B 6 bb. 39–40, scattered through various parts
C 7–11 bb. 40–51, top part, middle lines decorated
12–16 bb. 52–end, ditto (12 bars)
Full repeats not written out in P 283 are: bb. 1–12 (18–29) and bb. 39–50
(51–62). Despite most commentaries, it is not quite correct to describe the
chorale as having interludes. Within the main sections, its compositional
technique – through-composition of a melody above motivic accompani-
ment and quasi-ostinato pedal – is typical of the album. Less typical are the
broken-up carillons of the opening, not only the ostinato but the manual
figures in bb. 3, 5 etc; these are matched by the lh figure in the second half
(bb. 40, 52). The ‘Freude’ of the text is breathless (bb. 8, 25: the only pedal
solos in the album) and clamorous (bb. 48, 50: rare pedal trills).
In addition to its carillonesque ostinato, the pedal has some melodic
phrases, the last two of which (bb. 48, 60) are decorated as in the rh,
and another of which quotes a line very like Gastoldi’s original (b. 34).
Nor is the pedal the only quasi-ostinato: the opening four notes of the
melody appear in each of the first eleven bars, and again on their repeat.
Only a melody with such short, repeated phrases could be treated in such
a manner, and the exceptional setting matches the text’s own short phrases
and repeated rhythms. Rather, therefore, than seeing it as ‘more akin to
Bach’s large organ chorales’ such as ‘The Eighteen’ (Stinson 1994) or won-
dering why it is in the Ob at all (Kube 1999 p. 569), one might consider
BWV 615 as a special evocation of a special text and melody, inspired by
them.
More traditional is the combination in bb. 48ff. of a cantus firmus phrase
with a decorated version of the preceding phrase. The quaver pattern is also
268 BWV 615–616
familiar from the (contemporary?) Weimar chorale ‘O Lamm Gottes’ BWV

656a, where however there is no thrusting bass to compel it onward in the
same way. Perhaps the turned trill evokes the ‘Hallelujah’ figure at the end of
‘Komm, heiliger Geist’ BWV 651, where again it leads to harmonies far more
conventional than the logical but at first puzzling bb. 48 and 60.∗ Despite a
claim in J. Krause, MuK 1967 p. 131, it is difficult to see that any ostinato
motif of the movement is related in shape (and thus in significance) to the
rising Kreuzstab motif of Cantata 56.
Krebs’s copy gives left/right (s/d) toe-pedalling for the ostinato motif in
b. 61:
A d F G A G A D
s d d s d s
BWV 616 Mit Fried und Freud ich fahr dahin

Further copies: by or via J. T. Krebs, C. G. Meissner, J. G. Walther, J. C. Oley,
J. P. Kirnberger and J. C. Kittel.
Two staves; last 3 bars in tablature in P 283.
The TEXT of Luther’s four-verse alliterative prayer of thanksgiving and

reconciliation with death is a version of the Nunc dimittis (Luke 2: 29–32),
associated with the Burial Service (Stapel 1950 pp. 222ff.). Hymnbooks used
it for the end of Epiphany, Purification, and less often Sixteenth Sunday after
Trinity.
Mit Fried und Freud ich fahr dahin With peace and joy I now depart
in Gottes Willen; in God’s will;
getrost ist mir mein Herz und Sinn, my heart and mind are consoled,
sanft und stille; soft and stilled;
wie Gott mir verheissen hat: as God has promised me,
der Tod ist mein Schlaf worden. death has become my sleep.
The MELODY was published with the text, and may be derived (by Luther?)
from an older melody. Used in Cantatas 83 and 125 (Purification 1724,
1725), 95 (1723), 106 (funeral, 1707?) and harmonized in BWV 382:
∗A ninth followed by seventh is found in the same key in the cadence of the Loure from the French
Suite in G, BWV 816.
269 BWV 616
Example 128
Example 128. Buxtehude’s published elegy on the death of his father in

1674 based a set of movements on it (see below, pp. 351, 390).
Of the three fully worked settings (BWV 95, 125, 616), the last is the least
‘fluid’. Since Schweitzer’s motif-list of 1905, the dactyl rhythm has been cred-
ited with symbolizing joy. But here, the dragging shape suggests something
much more restrained, Simeon’s dragging footsteps or some allusion to Lent
as following on Purification? That the rhythm itself, though so prominent,
is not of prime significance is shown by the pedal’s motif, which keeps the
shape but not the rhythm. Whether the manual’s version implies ‘joy’ and
the pedal’s simpler version ‘peace’ (Chailley 1974 p. 192) is a conjecture of
the kind inspired by the Ob.
The manual’s motif has two versions, one beat and two beats long,
and is developed both inversus (as is the pedal’s) and in stretto. Several
times it affects the melody, as is not uncommon when a motif is of arche-
typal simplicity (cf. BWV 606), though unlike the equally archetypal one in
BWV 642, it begins on a downbeat. Such distinctions are important in the
Ob – compare in this respect BWV 609 and 606 – and it seems unreasonable
to claim both versions to be motifs de la joie. The motif varies in another
way: the in-turning shape (b. 2, first beat) is essentially different from the
scale-like shape (b. 2, third beat), as both are from the broken-chord version
(b. 15, second half). Throughout, typical harmonic tension is realized by
varying the form the motif takes, never quite predictable and avoiding easy
repetition. The final bar’s diminished seventh under a tonic is a familiar
discord before peace: see BWV 727.
Also, as in BWV 612, 607 and elsewhere, the pedal phrases are carefully
graded towards the final cadence; each has a different length and begins with
a rest at each new chorale line. The cadences formed at the ends of the pedal
phrases are symmetrically arranged: plagal–perfect–plagal–perfect–plagal.
But the lines avoid simply alternating up and down forms of the motif, and
the harmonic complex is prompted by the motifs, far more interestingly
than when the same chorale in Cantata 106 is accompanied by patterns that
merely decorate the harmonies.
270 BWV 617
BWV 617 Herr Gott, nun schleuss den Himmel auf

Further copies: by or via J. T. Krebs, J. G. Walther, C. G. Meissner, C. P. E.
Bach (? P 603), J. P. Kirnberger, Mempell–Preller and J. C. Kittel.
Two staves; last seven bars on extra slip, last half-bar in tablature in P 283.
The TEXT of T. Kiel’s hymn was published in 1620, like the last hymn based
partly on the Nunc dimittis and becoming associated with Purification.
Herr Gott, nun schleuss den Himmel auf! Lord God, now unlock Heaven!
mein Zeit zu End sich neiget; my time inclines towards its close;
ich hab vollendet meinen Lauf, I have completed my course,
dess sich mein Seel sehr freuet; which much gladdens my soul;
hab gnug gelitten, I have suffered enough,
mich müd gestritten, am tired with struggling,
schick mich fein zu, send me carefully
zur ewign Ruh, to eternal rest,
lass fahren was auf Erden let him go who on earth
will lieber selig werden. would rather be blessed.
The last of the three verses alludes to the Nunc dimittis.
The MELODY was published with the text in a five-part setting (Novello
15 p. 52), from two voices of which a melody either gradually emerged
or was deliberately formed in early eighteenth-century hymnbooks. In
Freylinghausen (1741) it takes the form shown in Example 129. See also
Example 129
BWV 1092. It is possible that Bach gave the cantus firmus in BWV 617 a
unique two-voice form because the original melody itself only ‘emerges’
from two crossed parts. This doubling might justify ‘à 2 Clav. c Pedale’ in
BG 25.ii, although P 283 only brackets the two cantus firmus voices at the
beginning – as it does in the case of BWV 624, headed ‘à 2 Clav’. Something
like a ‘doubled cantus firmus’ had already been achieved more simply in
271 BWV 617–618
Cantata 106 (1707?), where ‘Ich hab’ mein Sach’ appears in two parts against
a fugue.
As an unusual kind of trio, BWV 617 has a cantus firmus, a running left
hand and a syncopated pedal, each with a strong character. Only if the alto
crotchets are taken literally does the lh need a separate manual in the second
half, and b. 18 alto suggests one manual only. A simple broken chord, first
used to lead back to the repeated section (bb. 7–11 = 1–5, not written out in
P 283), is particularly useful when the harmony suddenly changes (bb. 18,
22, 23). There is no reason for the ‘interlude’ rests in the right hand between
the chorale-lines, since the harmony does not change. But they do emphasize
the unstoppable accompaniment, for which the text supplies several images:
‘knocking on the gates of Heaven’ (Schweitzer 1905 p. 348), ‘the unease of
worldly life’ (Keller 1948 p. 157), ‘the faltering steps of the aged Simeon’
(Terry 1921 p. 190) and ‘the course of life’ running into lassitude (Chailley
1974 p. 136). Simeon’s feet might not be dragging as in BWV 616, but the
line wandering through the music is easily heard as ‘sad’ or ‘resigned’.
Pictorial or not, the accompaniment is immensely adaptable for harmo-
nizing a complex tune. It was surely added after the melody was written
in, hence 24/16 and 12/8 added to the original C signature? If so, P 283 is
hardly a fair copy. The astonishing harmonization of b. 19 is created by
doubled chromatics on a pedal point, and there is no grammatical need
to play the quavers as triplets (as proposed in BG), although P 283 itself is
not clear enough to prove that the lines are ‘completely independent rhyth-
mically’ (Finke-Hecklinger 1970), as in the equal quavers of the NBA. The
very ambiguity emphasizes how in the Ob, a singing line, harmonic drive,
continuous rhythm, original texture, chromatic turns, clear dominant end
and a strange but bewitching Affekt can all be unprecedented.
BWV 618 O Lamm Gottes, unschuldig

Further copies: by or via C. P. E. Bach (P 603), C. G. Meissner, J. C. Oley,
C. F. Penzel, J. P. Kirnberger and J. C. Kittel.
Two staves; headed in P 283 ‘Adagio’ and ‘Canone alla quinta’ (the latter
subsequently?); repeat marks for bb. 1–7.
The TEXT is N. Decius’s paraphrase of the Agnus dei (1542), sung partic-
ularly on Good Friday between sermon and Communion, and generally in
Passiontide.
272 BWV 618
O Lamm Gottes, unschuldig O Lamb of God, innocently

am Stamm des Kreuzes geschlachtet, slain on the stem of the cross,
allzeit funden geduldig, always found forbearing
Wiewohl du warest verachtet: however despised you were.
all Sünd hast du getragen, All sin have you borne,
sonst müssten wir verzagen. otherwise we should have despaired.
refrain vv. 1, 2
Erbarm dich unser, O Jesu. Have mercy on us, Jesu.
refrain v. 3
Gib uns dein’ Frieden, O Jesu. Give us your peace, Jesu.
The MELODY, at least whose first line resembles one Gregorian Agnus dei
(Liber usualis, Mass IX), was published with the text and took several forms;
see Example 130 (Terry 1921 p. 281). A simpler version is harmonized in
BWV 401 and used (with a different line 6) in the opening chorus of the
St Matthew Passion; also in BWV 656, 656a, 1085 and 1095.
Example 130
Like BWV 619, this does not begin with the cantus firmus; but its canon is
between the tenor and alto, BWV 619’s between second tenor and soprano.
To some extent, therefore, the two are complementary (text, key, form)
but contrasted (metre, length, disposition and number of voices). Canonic
treatment of at least some phrases had appeared earlier (Scheidt’s ‘O Lamm
Gottes’, Geistliches Konzert No. 2, 1634). Perhaps, to make the canon clear,
P 283 is a short score enabling various interpretations: (i) as usually played
or (ii), with double pedal, down an octave with 4 stop (cf. BWV 608) or
(iii), with three manuals above the pedal, in the French manner of quatuor
à quatre claviers (Schrammek 1975 p. 103). Either way, to make the canon
fit, the ends of the chorale’s phrases are frequently altered, in particular the
last line, where the resulting bass/alto phrase resembles the fugally altered
theme ‘Ein’ feste Burg’ in the first movement of Cantata 80.
Whether or not this canon can be regarded as symbolizing the ‘fol-
lower’ of Jesus referred to in associated texts (Arfken 1965) or the ‘following
out’ of God’s will (Keller 1948 p. 158) or the bearing of sin by Jesus the
273 BWV 618–619
Mediator in a middle part (Honders 1988 p. 31), it is clear that the slurred
semiquavers, rising or falling, have associations with both Passiontide
(St Matthew Passion No. 29) and Christmas (Christmas Oratorio No. 29).
Thus the slurred motif is more versatile than its usual associations suggest –
‘sobbing’, ‘sighing’, ‘bearing sins’ or ‘dragging the cross’ – and is useful
rising or falling when contrapuntal ingenuity is required for harmonizing
a canon (compare Goldberg Variations No. 15). Several lines it produces
are very like the obbligato melody of a cantata aria (see bb. 3, 7, 23) or
the Canonic Variations (see b. 6). The subsidiary motif (b. 1, third beat) is
also violinistic.
The chromatics at bb. 20–1 have been claimed to correspond to the word
‘verzagen’ or ‘despair’ (e.g. Stinson 1996 p. 128) as they have too in the longer
setting of the Agnus dei, BWV 656. But another claim, that BWV 618 and
619 have a ‘common key’ (ibid.), is not justified by their opening bars: the
‘mixolydian’ tendency in the first, with its es typical of Bach movements
in F major, contrasts with the quite different lydian cadence of the second.
Neither have much in the way of perfect cadence, BWV 618 only at the end
of some phrases, 619 not at all.
BWV 619 Christe, du Lamm Gottes

Further copies: by or via J. G. Walther, C. G. Meissner, C. P. E. Bach
(? P 778), J. C. Oley, C. F. Penzel, J. G. Müthel, J. P. Kirnberger and
J. C. Kittel.
Two staves; headed in P 283 ‘in Canone alla duodecima à 2 Clav. et ped.’.
The TEXT, another translation of the Agnus dei (see BWV 618), was
published in 1528, appearing with this melody in 1557.
vv. 1, 2
Christe, du Lamm Gottes, Christ, Lamb of God,
der du trägst die Sünd der Welt, who bears the sins of the world,
erbarm dich unser. have mercy on us.
v. 3
Christe, du Lamm Gottes, Christ, Lamb of God,
der du trägst die Sünd der Welt, who bears the sins of the world,
gib uns deinen Frieden. Amen. give us your peace. Amen.
The dorian MELODY (Example 131) may derive from a Gregorian tone
(e.g. Liber usualis, Mass IV). Used in Cantatas 23 and 127 (1723, 1725)
274 BWV 619
Example 131
and BWV 233/233a. In BWV 23.iv the melody is set in canon, in BWV 233
and 127.i it appears with other chorale melodies: both aim to counter the
melody’s brevity.
Both the five-part texture and three-bar introduction are unusual, more
so than the modal cadence (cf. BWV 611 and 620) and the opening pedal
point under imitative lines (cf. the Toccata in F). The texture of five parts
has been seen as ‘after the model’ of Grigny (Klotz 1969a) – two parts on
each manual, above pedal – though BWV 633 is more like Grigny in this
respect. The overlapping canonic lines increase the complexity, as do the two
major–minor progressions (bb. 8–9, 12–13) and the accented passing-notes
created by the scales. Apart from the soprano f in b. 10, the canon is per
giusti intervalli.
In its canonic scale motif, the opening few bars unexpectedly resemble
those of ‘Vom Himmel hoch’, BWV 769. This motif is present in every bar,
sometimes inversus, often rectus, and in the penultimate bar hints that it
originates in the ‘Amen’ of the Gregorian melody. As often, thirds between
inner voices are important. In particular, the contrary motion of bb. 5–7
and 10–11 produces new harmony not actually required to solve the canon
but arising from its inventive use of motifs; much the same can be said of
BWV 600. The three lines developing the crotchet scale motif can be played
by the hands, but a rescoring of the movement to enable the pedal to take
both canonic voices is not possible if P 283 shows the required octave pitch.
Six brackets have been written (subsequently?) in P 283 at various points,
to make it clear that the five lines on two staves are distributed rh A/S, lh
T1/T2, B, but these could equally signal that the original was a ‘short score’
open to various interpretation. A similar point could be made about the
(added?) direction for two manuals.
This brief canonic movement, in which harmony reaches new heights of
sophistication through accented passing-notes, and hovers for seven of its
sixteen bars around chords of A, is a peak in the Weimar canonic tradition
as glimpsed on a more prosaic level in BWV 714, 693 (Walther) and 744
(J. L. Krebs). It is possible that canons for Passion chorales imply a ‘closeness
to God’, but just as likely, perhaps, is that they are ‘musical offerings’, the
fruits of pious endeavour.
275 BWV 620
BWV 620 Christus, der uns selig macht

Further copies: by or via J. C. Oley, C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves; headed in P 283 ‘in Canone all’Ottava’.
The TEXT by M. Weisse deriving ultimately from the Good Friday

hymn ‘Patris sapientia, veritas divina’, was published in the first German
hymnbook of the Bohemian Brethren (1531), perhaps a translation from
Czech.
Christus, der uns selig macht, Christ, who makes us blessed,

kein Bös’s hat begangen, has committed no evil,
ward für uns zur Mitternacht was for us at midnight
als ein Dieb gefangen, taken like a thief,
geführt vor gottlose Leut led before godless people
und fälschlich verklaget, and falsely accused,
verlacht, verhöhnt und verspeit, ridiculed, jeered and spat on,
wie denn die Schrift saget. as the Scripture says.
Seven further verses tell the Passion story, meditating on ‘your death and its
cause’.
The MELODY adapts ‘Patris sapientia’, which was already metrical. Used in
BWV 283 and 747 (Example 132) and a later version in the St John Passion,
15, 37.
Example 132
The original version in P 283, BWV 620a, was revised at about the time
BWV 613 was added: bb. 1–19 ‘drastically’, after which ‘b. 20 ended in
complete illegibility’ (Novello 15 p. xxi). Secondary sources also transmit
the last bars in revised form, suggesting that the revisions were notated
276 BWV 620
on a separate sheet, now lost (KB p. 32). Greater rhythmic activity given
by the new syncopations and semiquavers make the work not only more
vivid but less bound to one quaver figure. Not only is the new syncopated
figura stronger and more emphatic but bits of it are quite like the chorale
melody (e.g. alto bb. 3–4, a truncated, chromaticized version of the opening
cantus).
As in BWV 629, the canon at the fifteenth (not octave) is in the
outer parts, and as in BWV 618 and 619, the other parts begin canoni-
cally and remain imitative. The melody needs to be altered – by entering
early (b. 6 etc.) or holding back (b. 15 etc.), both devices familiar in
stile antico imitation – and as a bass-line this cantus ensures a series
of clear diatonic progressions, without halting cadences. The chromatic
motif becomes increasingly prominent, perhaps in association with the
text: ‘kein Bös’s’ (end of b. 3), ‘als ein Dieb gefangen’ (b. 9), ‘verklaget’
(bb. 15–16), ‘verlacht, verhöhnt’ (b. 18), although the whole nature of the
hymn (its scopus) makes chromatics relevant throughout, whether formu-
laic or not. The harmonic maturity arises equally from chromatics and
from the need to ‘explain’ harmonic cruxes thrown up by the outer canon.
A certain similarity between this chorale and the middle section of the
finale to Cantata 63 (for Halle, Advent/Christmas 1714?) comes from com-
bining chromatic fourth and dactylic counterpoint. Note that having less
rhythmic energy, the earlier version’s chromatics are more like ordinary
formulae.
The fierce sentiments of the text justify the syncopation of this powerful
setting, an equivalent perhaps to the fierce voices of the chorale in the St John
Passion. Its combination of vigorous rhythms with wailing chromatics has
naturally led to poetic interpretation. Many harmonic details are original
(e.g. b. 15), but while the revised version allowed the false relation in b. 22
(c–c ), it seems that the composer altered the bass of the canon in b. 11
to avoid a similar but more obtrusive progression (f –F). The difference
between b. 11 and b. 22 is instructive: in b. 11, a pedal F would produce an
unlikely false relation when the tenor line is so diatonic; in b. 22, the fourth
quaver is yet more dissonant (F c c g ), but the dissonance is the result
of passing-notes and accepted by the ear as such.
Equally original is the progression over bb. 15–17. The lightening of
the harmony when a B minor chord rises to a clear G major, passes to
another brief B minor, then a C seventh and a highly chromatic turn to A
major/minor, then a B seventh: this passage deserves the closest examination.
Such harmonies are not at all obvious from the canonic cantus firmus, which
in itself need have led to no more than the mild triadism of a Walther canon.
As with b. 22, it is the two accompanying motifs that produce the inventive
harmonies, incited by the canon perhaps.
277 BWV 620–621
An especially characteristic passage, bb. 8–10, is largely repeated later,

bb. 19–21, including the unique low C. This C is no reliable evidence
for an organ with such a note, since the written-out canon makes it oblig-
atory. On the other hand, playing it up an octave (Arfken 1955 pp. 30–2)
seems rather drastic, unless this phrase alone used 16 reed, the rest pedal
Trompete 8 , as in BWV 600?
BWV 620a Christus, der uns selig macht

Written over in P 283; further copies by C. G. Meissner and late MSS.
Evidently some copyists knew the chorale before it was revised. While the
harmonies and the chromatic motif remain largely unchanged, clearly the
blander rhythms make for less pungent harmonies. But the original lines
should not be underestimated: Example 133 is a fine countersubject. The
‘sharpening’ of the rhythms anticipates that for the fugue alla francese in
the Art of Fugue, similarly revised after the composer wrote it in the score
P 200 – perhaps not very long after revising BWV 620a?
Example 133
BWV 621 Da Jesus an dem Kreuze stund

Further copies: by or via J. T. Krebs, C. G. Meissner, J. C. Oley, C. F. Penzel,
J. G. Müthel, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT of J. Böschenstein’s Passiontide hymn is based on the Seven Last

Words (cf. the hymn ‘Stabat ad lignum crucis’) and was sung on Good
Friday.
Da Jesus an dem Kreuze stund As Jesus hung upon the cross,

und ihm sein Leichnam war verwundt and his body was wounded
so gar mit bittern Schmerzen; with so much bitter pain;
die sieben Wort, die Jesus sprach, the Seven Words which Jesus spoke
betracht in deinem Herzen. consider in your heart.
278 BWV 621
Verses 2–8 relate the Seven Words, followed by an exhortation in v. 9.
The MELODY, from the Reformation period, is used for several texts and
is much like other melodies. Used as a fugue-subject by ‘southern’ com-
posers (J. E. Kindermann, J. Krieger, Pachelbel, J. K. F. Fischer), perhaps
during Lent, it appears in no known Bach cantatas. Krebs’s third cantus line
reads e a g , and is harmonized accordingly – presumably Bach’s original
(KB p. 73). But the form in Example 134 is usual.
Example 134
Since Spitta (I p. 593), the syncopated bass motif has been seen as either
symbolizing or picturing ‘a sinking body’ (Schweitzer 1905 p. 348), and cer-
tainly, if one set out to picture ‘dragging’ by conventional musical means, no
better bass line could be conceived than these masterly suspensions. Is the
similarity between the opening bars (from which the rest springs) and the
close of the Christmas chorale BWV 606 to be seen, therefore, as underlin-
ing the connection between Incarnation and Crucifixion? As in BWV 606,
the bass has its own motif while the middle voices produce some impor-
tant passages in thirds, more than faintly reminiscent of the Corelli fugue
BWV 579.
Density and intricacy in the chorale come from its constant reference to
motif, its compact harmony, and the total absence of rests (cf. BWV 602,
609). At the end of each chorale line the bass presses forward, never pausing
until the final cadence. Compared to the kind of stile antico treatment of this
melody by Fischer and others, BWV 621 does seem more ‘subjective’, inviting
one to see in the drooping bass a distinct cross figure (see Glossary). But the
text itself is mostly unconcerned with the actual incidents of the crucifixion,
only with it as the setting for the victim’s Seven Words.
Not only does each part have its own motif or prevailing rhythm, but
the tenor and bass motifs (each heard five or more times) consistently avoid
easy formulae or contrapuntal convenience. Moreover, the voices are paired;
soprano and bass work with or against each other, alto and tenor together.
Further concentration is given by the typical modified repetition (bb. 1 and
7, bb. 4 and 8).
279 BWV 622
BWV 622 O Mensch, bewein dein Sünde gross

Two staves, three in P 802 (Krebs); headed in P 283 ‘adagio assai: à 2 Clav.
& Ped.’ (written subsequently?); at end, ‘adagiissimo’.∗
The TEXT of S. Heyden’s Passion hymn was published in 1525.

O Mensch, bewein dein Sünde gross, O man, weep for your great sin,
darum Christus seins Vaters Schoss for which Christ left his father’s bosom
äussert und kam auf Erden; and came to earth;
von einer Jungfrau rein und zart of a Virgin pure and gentle
für uns er hie geboren ward; he was born here for us;
er wollt der Mittler werden. to become the mediator [for sins].
Den Toten er das Leben gab He gave life to the dead
und legt’ dabei all Krankheit ab, and banished all sickness,
bis sich die Zeit herdrange, until the time came on
dass er für uns geopfert würd, that he should be sacrificed,
trüg unsrer Sünden schwere Bürd bearing the heavy burden of our sins
wohl an dem Kreuze lange. long on the cross.
Twenty-two further verses alternate between the crucifixion and the ‘great
sin’.
The MELODY by M. Greitter was also published in 1525, later associated

with this text, harmonized in BWV 402 and used in the final chorus of
Part 1 of the St Matthew Passion (from the 1725 version of St John Passion):
Example 135. It is also sung to the Whit hymn ‘Jauchz, Erd und Himmel,
juble hell’ (1537).
Example 135
∗ The superlative of adagio was not always clear: Heinichen 1711 p. 179 wrote adagiosissimo, as probably
did Bach in the Capriccio BWV 992.
280 BWV 622
BG 25.ii surmised that ‘the melody was kept very simple’ at first, and the
arabesques were ‘added later’ – hence copied by Krebs with fewer orna-
ments? – but this is not clear from P 283, which began more as a fair copy
than it continued to be. Alterations were made in a hasty composing score
(KB p. 32), perhaps at several stages, and the totally rewritten b. 21 is the
only such instance in the album (see KB. p. 40). Whether two manuals were
(i) always the intention, (ii) necessary at all, is unclear; b. 22 suggests that it
was planned with only one in mind.
Though often likened to the coloratura found in Buxtehude, this cel-
ebrated setting’s ornamental melody is more original, less ‘instrumental’
than either BWV 659 or the ‘Adagio assai’ opening of Cantata 21 (1714?).
Most beats have the notes of the chorale – an old trait – and bb. 1, 2 and 5
would, at a much faster tempo, resemble a French ouverture. Many patterns
are conventional, others unique and mysteriously melodious (e.g. end of
b. 2, beginning of b. 20), perhaps later additions. At least one melodic pat-
tern was the result of second thoughts: the little rh demisemiquaver figure in
bb. 14 and 22 was originally simple pairs of semiquavers, and the lh probably
had fewer of the semiquaver patterns. Of course, the spectacular final line
has led to a search for allusions to the text (‘Kreuze’, ‘lange’), especially in
view of a key that is neutral only in equal temperament (E). The move-
ment gives the impression of inspired caprice and not a mere catalogue
of note-patterns, partly because in returning twice to simple crotchets the
melody is far beyond merely applying formulae. The invention appears
limitless.
The coloratura, sumptuously wide-ranging from b to b , disguises not
only the chorale melody but also the form of the hymn. Yet its four sets of
three lines each are strictly followed, and in particular, the rhyme-scheme
aab is mirrored in the two sets of dominant–dominant–tonic cadences of
the first six lines. The setting does not always reflect the repetitions in the
original chorale melody. Bar 8 can be seen as a kind of variation of b. 2,
whereas b. 7 is quite different from b. 1, despite the same chorale-melody,
for the accompaniment now moves into suspirans semiquavers. While the
coloratura too becomes more and more wide-ranging – something unusual
for such treatment – the two inner parts too are increasingly imitative, pro-
gressing gradually from crotchets to semiquavers and reaching a particular
intensity in b. 21 (‘bearing the heavy burden’), a bar revised and re-conceived
in P 283. This ‘peak’ appears after and before a chromatic bass. Generally,
these inner parts are freer but more active than those of BWV 659, whose
continuo-like pedal has much in common with BWV 622’s and sometimes
moves in a similar way.
Because the chorale melody is so long, changes in texture are desirable, as
are the varied reprise (bb. 1–6 = 6–10) and many touches of colour – the Ds
281 BWV 622–623
and the increasing chromatics, finally in the melody too. In view of the text’s
great length and the melody’s other association with Whit, perhaps BWV 622
relates more closely than usual to a particular verse (v. 1) and its key words,
though only special pleading can make close parallels, except for ‘lange’ at
the final melisma. Even ‘Kreuze’ does not coincide with the C chord, and
‘geopfert’ (bb. 19–20) is preceded, not accompanied, by bass chromatics.
Perhaps ‘Kreuze’ can be heard in the penultimate bar and its upbeat, but their
astonishing accented passing-notes transcend images, as does the sudden
simplicity of the melody when the bass twice rises chromatically.
One can look at the celebrated C triad a long time and not be quite sure
what it is – other than a preparatory chromaticism, i.e. E minor for the E
major cadence. A simpler final twist to the minor is found in Pachelbel’s
E Fantasia (copied by Walther) and often in Froberger and Buxtehude.
Another but lesser chromaticism colours the chorale in the St Matthew
Passion, where ‘Kreuze’ is less conspicuous, and in the St John version (in
E major) the chord is indeed C. Here in the Ob the C behaves rather as
a Neapolitan or augmented sixth, but is not exactly either, and is made the
more startling by the new spacing and sudden suspension of semiquavers.
A more closely related E–C progression is found for the text ‘deinen Leiden’
(‘your suffering’) in the second movement of Cantata 22 (1723).
BWV 623 Wir danken dir, Herr Jesu Christ, dass du für uns
gestorben bist
Further copies: by or via J. C. Oley, C. G. Meissner, C. F. Penzel, J. G. Müthel,
J. P. Kirnberger.
Two staves; last 1 12 bars in tablature in P 283, where second text-line added
later?
The TEXT of C. Fischer’s Passiontide hymn (different, after the first line,
from other texts beginning thus) was published in 1568.
Wir danken dir, Herr Jesu Christ, We thank you, Lord Jesu Christ,
dass du für uns gestorben bist, That you have died for us,
und hast uns durch dein teures Blut And through your precious blood
gemacht vor Gott gerecht und gut. Have made us righteous and good
before God.
In P 283, ‘du’ is written ‘DU’. The remaining three verses are a prayer for
‘assurance that you will not forsake us’.
282 BWV 623
The MELODY was sung to various texts, one form of it (1597) as in

Example 136 (Terry 1921 p. 334).
Example 136
Some have seen the pedal and accompanying rhythms as referring either
to ‘joyful thanksgiving’ or (in the bass) to an ‘expression of confidence’.
In its actual working-out, however, the note-pattern takes various lengths
and shapes. Such treatment suggests a different approach from that of (e.g.)
BWV 643, where a pattern is less often changed. Moreover, the ending of
chorale lines on dominant sevenths (bb. 4, 16) seems to undermine any
‘confidence’, as it does in other chorales using dominant sevenths in such
a way (e.g. ‘Mein teurer Heiland’, St John Passion). While the middle parts
are much like those in other chorales, the pedal motif is rather cello-like,
more so than a similar figure in the G major Prelude BWV 541. Particularly
good use is made of rhetorical rests and of displacement of the motif across
bar-lines, and in their use of a simple motif-cell all three lower parts show
an inventiveness that was unique to the Ob.
Only a three-note figure, the motif produces different patterns in each bar
yet leaves the chorale melody clear. Unlike the settings either side of it, BWV
622 and 624, the melody is as if merely harmonized and then decorated by
the dactyl figure between beats. As Marpurg pointed out in Abhandlung von
der Fuge (1753), ‘the two middle voices produce a mere counter-harmony’
(‘eine blosse Gegenharmonie’: Dok III p. 45), but this suggests he was not
fully aware of the harmonic nuances of the piece, or that a second motif
tends to emerge (bb. 7, 8, 11, 14, 15, 16).
The dancing metre, displaced rhythms and four-bar phraseology seem
typical of the polonaise, whose ‘popular’ character would then correspond
to the rather doggerel-like nature of the hymn, reminiscent of medieval
texts like ‘Mary’s joy of Six, Dancing on the Crucifix’. Was there an allusion
here to the melody’s known Polish connections (a Polish hymnbook of 1559:
Terry 1929 p. 149)? In any case, the Ob’s motivic harmonizations of chorales
achieve maturity in this movement, as well as in BWV 624. The 3/4 time-
signature is unique in the album, meant to be ‘modern’, perhaps, implying
283 BWV 623–624
a more pronounced dance character than the sarabandes BWV 652, 653
and 654. Does absence of fermatas and changes of beat for the pedal motif
suggest that the setting was meant to be unusually continuous?
BWV 624 Hilf, Gott, dass mir’s gelinge

Two staves; headed in P 283 ‘à 2 Clav et ped.’; two staves, pedal throughout
in tablature-letters (in place of a third stave).
The TEXT of H. Müller’s ‘Ballad of the Passion’ was published in 1527 and
appeared in early Lutheran hymnbooks.
Hilf, Gott, dass mirs gelinge, Help me, God, that I may succeed,
du edler Schöpfer mein, my precious creator,
die Silben reimweis zwinge, in forcing these syllables into rhyme
zu Lob und Ehren dein, to your praise and honour,
dass ich mag fröhlich heben an, that I may joyfully begin
von deinem Wort zu singen, to sing of your Word,
Herr, du wollst mir beistan. you will stand by me, Lord.
Twelve further verses recount the Passion and Ascension, referring to

scripture.
The MELODY draws on several versions associated with the text by

1545 (Terry 1921 p. 203). BWV 343 is similar to Freylinghausen 1741
(Example 137), and probably the difficulty of making a canon for the third
phrase occasioned the version in BWV 624. Walther uses a similar compos-
ite form for a canon. Note that the opening line does not need to go through
so many keys as in BWV 624, with its canon beginning at the tritone.
Example 137
284 BWV 624–625
Like BWV 618, the movement incorporates a canon at the fifth in adjacent
voices; for the fifth and sixth lines (bb. 9–13) it is a canon at the fourth. The
intervals of the canonic answer are not strict, and rhythms require changing
in b. 13, while the c is shortened to suit the accompaniment. There must
have been some fixed determination to make a canon here: in rivalry with
Walther?
The non-stop lh passage-work runs through the cantus even more
intensely (both higher and lower) than in BWV 617, though perhaps not
quite to so anguished an effect. Again, the Affekt is elusive: triplets make it
‘animated’ (Terry 1921 p. 204), a syncopated bass means ‘lassitude’
(Schweitzer 1905 p. 348), canon evokes the Creator ‘helping’ (Chailley 1974
p. 145) or pictures the effort of the ‘forced syllables’ in v. 1 (Clark 1984
p. 87). Also in common with BWV 617 are the repeat of the opening sec-
tion and the curious fact that without the lh the harmonies are already
complete, especially here. The final cadence, like those of BWV 616, 721,
727 and 659, carries the dissonant bass leading-note under a soprano pedal
point.
The lh line is as inventive as that of BWV 607 and 617, with distinct
patterns, some scale-like, some doubling back, according to requirements.
The particularly insistent triplets of the final 2 12 bars match the final driv-
ing scales of BWV 607. These three chorales present their obbligato lines
in three metres – semiquavers (BWV 607), sextolets (BWV 617), triplets
(BWV 624) – and have a compass of about three octaves from G (BWV 624
the largest, G–a ), and in all three the lh only gradually emerges through
and above the cantus firmus. The three bass lines, though equally motivic,
react in three different ways to such lh figures; BWV 624’s seems particu-
larly independent, not only because of the sophisticated passing-notes in all
voices, but because each line of the cantus ends on a weak beat. As in BWV
621, the bass syncopations invite a search for text-references (to the effort
implied in v. 1?), as do the left hand’s triplets (the persistence also implied
in it?).
The special aura of this unique setting – rather remote, subdued, strange
even – surrounds the listener, especially as the lh rises through the cantus.
Its uniqueness, owed to a harmony already rich even without the running
line, becomes clearer when compared to J. L. Krebs’s imitation of it in his
Clavierübung, ‘Christ lag’ (1752 – note the next Ob title).
BWV 625 Christ lag in Todesbanden

Further copies; by or via C. G. Meissner, J. C. Oley, J. G. Müthel, C. F. Penzel,
285 BWV 625
Two staves.
The TEXT is one of Luther’s two Easter hymns (see BWV 626): seven verses
built partly on the sequence ‘Victimae paschali laudes’, later the chief hymn
of Easter.
Christ lag in Todesbanden, Christ lay in the bonds of death

für unsre Sünd gegeben, given up for our sins,
der ist wieder erstanden he is risen again
und hat uns bracht das Leben. and has brought us life.
Des wir sollen fröhlich sein, Therefore we should be joyful,
Gott loben und dankbar sein, praising God and being thankful,
und singen Hallelujah, and singing Hallelujah,
Halleluja. Hallelujah.
v. 4 begins:
Es war ein wunderlich Krieg, It was a wonderful war,
da Tod und Leben rungen; as Death and Life wrestled;
das Leben behielt den Sieg, the victory went to Life,
es hat den Tod verschlungen. it has swallowed up Death.
The MELODY (Example 138) is from the older hymn ‘Christ ist erstanden’
(Terry 1921 p. 117), a variant or extract of the ‘Victimae paschali’ melody.
The sharpened second note, once rare, is prominent in Cantata 4, BWV
277–279, 625, 695, 718 and Bruhns’s ‘Hemmt eure Tränenflut’, but not in
Cantata 158. Both forms are found in Böhm and Scheidt, the latter within
one set of variations (Tabulatura nova, 1624).
Example 138
Like BWV 616, this uses a motif with both a one-beat and a two-beat version,
each developed throughout, joining finally in the last bar. As Example 139
shows, the motif is related to the cantus. Twice near the end the pedal also
has it in augmentation, enphasizing the perfect cadences. The motif suggests
to some ‘the bonds of death’ (Schweitzer 1905 p. 349), to others ‘the rolling
away of the stone’ (Keller 1948 p. 161), especially if played slow. Twice as
286 BWV 625–626
Example 139
fast, it would resemble the cello motif at ‘Gewalt’ (‘power’) in Versus III of
the early Cantata 4. As in BWV 718, its few suspensions have been seen as
‘the bonds of death’, though why just at these points is unclear, as is why
there are not more of them – the two penultimate bars could have supplied
the pattern for another whole setting.
Despite the possibility that this 4/4 is slow, the motif’s essential vigour
seems assured when it rises into the chorale melody at its highest point
(‘praising and thanking God’), aided by rising chromatics at that moment.
Although the movement begins as densely as ‘Jesu, meine Freude’, its tension
is less sustained (e.g. end of b. 8) and its motifs are sometimes neglected
(e.g. first half of b. 14). Nevertheless, there are vigour and intensity in the
setting, many bars of which have eight harmonies in quick succession, as if
disturbed and reflecting the image of war in v. 4, quoted above.
BWV 626 Jesus Christus, unser Heiland, der den

Tod überwand
Further copies: by or via J. T. Krebs, C. H. Meissner, J. C. Oley, J. G. Müthel,
C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT of Luther’s three-verse Easter hymn was published in 1524:
Jesus Christus, unser Heiland, Jesus Christ, our Saviour,

der den Tod überwand, who overcame death,
ist auferstanden, is risen,
die Sünd hat er gefangen. he has captured sin.
Kyrie eleison. Lord have mercy.
The MELODY appeared with the text in 1529; the version in BWV 626
and 364 closes with a different ‘Kyrie eleison’ phrase, first found in 1585
(Terry 1921 p. 229): see Example 140. The five melodic phrases have an
approximate form abcab.
287 BWV 626–627
Example 140
The same syncopated quaver motif runs through all three accompany-
ing parts, one after the other and sometimes together, thus appearing in
every half-bar. The weight of the inner parts seems to be characteristic
of Bach’s growing experience with the Ob conception; another example
is BWV 644. The syncopation can no doubt be seen as picturing the rise
from death, either symbolizing the triumph or giving a representation of
‘taking death prisoner’. However, by nature it resembles motifs often found
in compound-time variations of secular or chorale variations, such as the
gigues in Buxtehude’s ‘Auf meinen lieben Gott’ (copied by Walther) and
Bach’s ‘Sei gegrüsset’. Again thirds between the inner voices are important,
though not, as in other one-motif chorales (BWV 601, 623), between bass
and tenor.
As in BWV 620, there is a kind of embedded back-reference: b. 7 is much
like b. 2. However, although as a consequence of the abcab pattern the last
line is the same as the second, it is reharmonized, with new modulations,
despite each phrase actually beginning and ending much as before (bb. 3–4 E
minor to A minor, bb. 8–9 E minor to A minor, but with a b!). There seems
no end to how inventively short motifs can be explored, and the technique
never quite repeats itself. Although there is only marginally a greater use of
sevenths in b. 8 than in b. 3, the surprise b of b. 8 can be seen as crucial
in giving a colour unknown in b. 3. Reversing the bars would show how
naturally this unexpected note, appearing where it does in b. 8 (with its
hints of the Neapolitan sixth?), leads to the final cadence. It also gives a new
slant on the motif itself, whose second note otherwise is always diatonic.
The ‘monothematic’ accompaniment of BWV 601 and 626 is the reason
for their use as contrapuntal examples in the Abhandlung von der Fuge (1753)
of Marpurg, who fails to draw attention to the rare double time-signature.
BWV 627 Christ ist erstanden

Further copies: by or via J. T. Krebs, C. G. Meissner, J. C. Oley, J. G. Müthel,
Headed in P 283 ‘1 Vers.’, ‘Vers. 2’, ‘Vers. 3’.

288 BWV 627
The TEXT of the Easter carol was several centuries old when published in
1529. It came to be sung on all the days of Easter, on Sundays before the
sermon (Stiller 1970 p. 226), and at Ascension.
Christ ist erstanden Christ is risen

von der Marter alle; from all the torment;
des solln wir alle froh sein, therefore we should be joyful,
Christ will unser Trost sein. Christ will be our consolation.
Kyrieleis. Lord have mercy.
v. 2
Wär er nicht erstanden, If he had not risen
so wär die Welt vergangen; the world would be lost;
seit dass er erstanden ist, since he has risen,
so lobn wir den Vater Jesu Christ. we praise the Father of Jesus Christ.
Kyrieleis. Lord have mercy.
v. 3
Halleluja, Halleluja, Halleluia!
des solln wir alle froh sein, Therefore we should all be joyful;
Christ will unser Trost sein. Christ will be our consolation.
Kyrieleis. Lord have mercy
The MELODY was published with the text and may be as old. The three
verses have a melodic form AAB, but neither BWV 276 (three verses –
Example 141) nor the Easter Cantata 66 (v. 3 only) gives the melody in the
same form as BWV 627.
The three-verse form is unique in the album, and it was no doubt the three
different melodies that led J. C. F. Bach to count Ob’s contents as forty-eight
chorales, not forty-six, on the title-page of P 283 (Dok I p. 214). The three-
verse text ‘Christe, du Lamm Gottes’ BWV 619 has one melody; see also ‘O
Lamm Gottes’ BWV 656 and the two Kyrie groups in Clavierübung III. Each
Vers of BWV 627 develops its own motif, making a group similar to those
by Walther, except that it is not a set of variations but through-composed,
reflecting the different metres of the text. Its c.f. is like a cantus planus, in
minims such as are found otherwise in Ob only in BWV 635.
Each Vers has a pair of conventional motifs, similar but distinct, start-
ing with anapaests and dactyls. Thematic relationships can also be found,
as when the opening melody (especially with its sharpened leading-note)
traces the opening line of ‘Christ lag in Todesbanden’. Motifs derive from
the melody, and the common suspirans figure appears in v. 3, including a
quaver form in the pedal. Such relationships easily arise within Ob’s motivic
language and have the effect of integrating the three movements. The texture
flows more as the verses proceed, from the syncopations of b. 1 through the
289 BWV 627–628
Example 141
bar-long patterns in v. 2 to the cumulative final cadence, after the pedal has
explored its own ‘Hallelujah’ figure (b. 40) and in bb. 50ff. even anticipated
(as Clark 1984 p. 94 notes) the last six bars of the B minor Organ Fugue.
In BWV 627 a rigid cantus accompanied by busy but conventional note-
patterns, so worked as to produce a ‘standard 4/4 continuity’, leads to some-
thing closer to Pachelbel or Walther than J. S. Bach. There is a doctrinaire
feel to it especially in v. 1, owing chiefly to the common-property motifs.
Vers 3’s suspirans figure is clearly more conventional than in BWV 628 or
630, even when it affects the cantus in b. 49. The spinning around F major
in bb. 41–7 is difficult to imagine in a maturer chorale, and throughout,
harmonic progressions particularly at the cadences are straightforward and
orthodox. Not only does the striking fair copy in P 283 suggest that it was
older than some others but so does much of the musical content: various
moments in it sound like other chorales in the Ob, especially those in
D minor, rather as if it were a ‘dry run’ for them. Or to put it more positively,
perhaps the orthodoxy of the treatment is a means of celebrating a classic
hymn said to have been especially admired by Luther.
BWV 628 Erstanden ist der heil’ge Christ

Two staves.
290 BWV 628
The TEXT is a translation of the fourteenth-century carol ‘Surrexit Chris-

tus hodie’, published at Nuremberg in 1544 but of varying length in the
hymnbooks.
Erstanden ist der heilige Christ, The holy Christ is risen,
Halleluja, Halleluja, Hallelujah, Hallelujah,
der aller Welt ein Tröster ist, who is a comforter to all the world,
Halleluja, Halleluja. Hallelujah, Hallelujah.
In one version, nineteen verses (with ‘Hallelujah’ in every other line) narrate
the meeting of the Marys with the angel at the tomb.
The original MELODY of the folksong carol (Example 142) was published
by 1531; the version in BWV 628 follows later hymnbooks. BWV 306 har-
monizes in a similar way a melody published as a descant to it in 1555 (Terry
1921 p. 164).
Example 142
In their books on composition, German theorists such as Printz (1696)

and Walther (1708) compare and contrast such little note-patterns as the
suspirans and corta. Both are used in BWV 627, which is followed by three
triumphant Easter chorales that alternate them: suspirans (BWV 628), corta
(BWV 629), suspirans (BWV 630). In their different ways the rising lines
of all three surely refer to the Resurrection. Both alto and tenor in BWV
628 are graphic, while the pedal’s perfect cadences for most lines are more
in the way of an ‘affirmation of faith’. Moreover, in the latter half of the
movement both pedal and manual motifs fall as much as they rise, and the
final octave D recalls a similar effect at the close of one of the Christmas
chorales.
Characteristic of the Ob is the running line created between two manual
parts, supported by a constant and quite different pedal motif, which in this
case is unusually regular in its entries and tenuto only at the end of phrases.
The added passing-notes in the melody hint at the suspirans figure and may
be related to it, since crotchets are not unimportant in the movement. Either
way, the opening bars surge up as if to convey the shock felt by the three
Marys.
But lively surging lines based on this motif need not ‘picture’ resur-
rection: similar lines in the first movement of Cantata 66 (Second Day
of Easter, 1724) probably originated in a birthday cantata for Leopold of
291 BWV 628–629
Anhalt-Köthen (1718), to the text ‘may the sun shine’ (‘es strahle die Sonne’).
The demisemiquavers of the cantata’s violin lines rise and fall like a faster
version of the chorale’s inner lines, much as the violin scales of Cantata 26.i,
‘Ach wie flüchtig’ (1724), are a faster version of those in ‘Ach wie nichtig’,
BWV 644. Surging lines evoke an uplift of the spirits. But note: if resurrection
could only be invoked by dramatically rising lines, the previous resurrection
setting (BWV 627) would be anomalous because predominantly its lines fall
and lack drama. The text of both chorales refers to rejoicing at Easter but
BWV 627 follows with ‘Kyrie eleison’ and BWV 628 with ‘Hallelujah’: does
this explain the musical difference between them?
BWV 629 Erschienen ist der herrliche Tag

Further copies: by or via J. G. Walther, C. G. Meissner, J. P. Kirnberger, J. C.
Kittel and late sources.
Two staves; headed in P 283 ‘a 2 Clav. & Ped. in Canone’.
The TEXT of N. Herman’s Easter hymn was published in 1560.
Erschienen ist der herrliche Tag, The day of splendour has come,
dran sich niemand gnug freuen mag: at which none can rejoice enough:
Christ, unser Herr, heut triumphiert, Christ, our Lord, triumphs today,
all sein Feind er gefangen führt. he leads captive all his enemies.
Halleluja. Hallelujah.
Less ballad-like than BWV 628’s hymn, the present fourteen verses return
to the theme ‘Life triumphed and became Death’s master’.
The MELODY was published with the text but probably had a different ori-
gin, secular or sacred (e.g. the Easter antiphon ‘Ad monumentum venimus
gementes’). Used in Cantatas 67 (1724) and 145 (Third Day of Easter 1729?):
Example 143.
Although the dactyl rhythm of the accompanying motif is also found against
the words ‘et expecto resurrectionem’ in the B minor Mass, there appears
to be a further reason for it here: see BWV 628. The motif frequently en-
compasses a fifth, the ‘resurrection fifth’ found in the melody of BWV 629
and in the bass of 628. Though the canon is sometimes inexact and very like
Walther’s for the same melody – which came first is unknown – the accom-
paniment responds to the text far more energetically. As in the other canons,
the motif runs through to the end, ending more succinctly than BWV 608
292 BWV 629–630
Example 143
and affecting all three manual parts: they all rise. It is difficult to believe
that BWV 628 was composed without conscious reference to Buxtehude’s
‘Wir danken dir’ (copied by Walther), as too must have been the case for
Walther’s setting. The most thoroughly motivic treatment of the chorale is
here in BWV 629, whose dactyl seems to some ‘the Bach joy-figure’.
The octave canons (BWV 600, 608, 629) have a ‘joyful’ 3/2 metre clearly
different in mood from the canon of BWV 619, or from Walther’s similar
‘Puer natus in Bethlehem’. Two of Bach’s are particularly triadic, and the
triad is said to symbolize perfection (Krey 1956 p. 54ff.). As in BWV 620,
the lower canonic voice is also the bass of the harmony, and twice especially
it needs alteration to fit (bb. 8, 11–12), whereas in BWV 620 both voices
usually have to change. As elsewhere, the inner parts are imitative and at
times quasi-canonic. Their thirds and sixths are shown in a quite different
light from those of the Passion chorale BWV 624 (where they appear in the
canonic voices), particularly in the final upsurge, a canon sine pausa resulting
from the parallel motion that has been there right from the beginning.
Only towards the end are two manuals needed, to leave the canon unen-
cumbered. Otherwise, as in BWV 622, the inner parts give the impression of
being conceived to be played between the two hands, the alto only glancingly
interfering with the melody. Was ‘à 2 Clav’ originally intended or did the
idea occur only when (i) hands crossed in the final line as the composition
was completed, or (ii) clear instruction became part of the didactic ‘pro-
gramme’ for the album? (See a similar question below for BWV 639.) Was
the right hand expected to change manual at the end? If so, is it evidence
for further assumptions of this kind elsewhere?
BWV 630 Heut triumphieret Gottes Sohn

Further copies: by or via J. T. Krebs, J. G. Walther, C. G. Meissner, C. F.
Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves. (‘One of the earliest pieces entered in P 283’ – Kobayashi 1989
p. 38.)
293 BWV 630
The TEXT, first published by C. Stolzhagen in 1591, was included amongst

the Easter hymns in most later books, in Leipzig also for Ascension (Stiller
1970 p. 76).
Heut triumphiret Gottes Sohn, Today the Son of God triumphs,

der von dem Tod erstanden schon, having risen from the dead,
Halleluja, Halleluja, Hallelujah, Hallelujah,
mit grosser Pracht und Herrlichkeit, with great splendour and magnificence,
des dankn wir ihm in Ewigkeit. for which we thank him in eternity.
The following five verses continue the praise and Hallelujahs.
The MELODY (Example 144), published with the text in 1601, is closer
to BWV 630 than to BWV 342 (from a lost Easter cantata). The repeated
Hallelujah Ds at the end of BWV 630 either were added by the composer or
reflect local custom.
Example 144
For the suspirans motif in the accompaniment, see notes to BWV 628. While
the pedal motif looks like those of some other chorales, such as ‘O Trau-
rigkeit, O Herzeleid’ (see Example 105, p. 238 above) or the falling fifth of
BWV 628, properly it contains two ideas, one falling and one rising. They
always appear paired and at the same point, i.e. halfway through the first
bar of each line, and by way of climax are finally extended as often in the
Ob, here to make the Hallelujah. Terry (1921 p. 200) sees a resemblance
to an aria in Cantata 43 for Ascension Day, for which the chorale may be
intended.
The graphic pedal line below a seamless counterpoint encourages the
search for images: the ‘hero pressing down his enemies’ (Schweitzer 1905
p. 349) lying in the dust. In the spectacular final pedal phrase one can imagine
either the harrowing of Hell or a ‘Hallelujah!’, although in principle it is only
a decorated plagal cadence, a widely familiar way of breaking a chord as
already heard in BWV 599. See Example 145, from Buxtehude’s Praeludium
294 BWV 630–631
BuxWV 163. The final three bars of BWV 630 drop the scalar quavers for a
clearly articulated ‘Ha-lle-lu-jah’. The final chord of D major magnificently
prepares for the following chorale, but by accident: other settings before
Whit were to have come between.
Example 145
In its swinging 3/2 metre, its four-bar phraseology and almost-repeated

bass phrases – a line that could only be a bass-line! – the chorale is not far
from Buxtehude’s passacaglias. The suspirans quaver motif of b. 1 is also
familiar in chaconnes and in the Passacaglia itself, though in this chorale
it develops in classic Ob style, generating a harmonic progression (e.g. end
of b. 15) or embellishing one that is already clear. The melody requires the
motif to be constantly adapted (compare b. 11 with b. 3), but back-reference
is possible (b. 19 = b. 7 and 23; compare b. 9 with b. 1 or b. 21 with b. 5).
The unending quavers with their mellifluous thirds and sixths, in one hand
then the other then both, become a way of realizing a faultless four-part
chorale harmonization: were it a prelude to a hymn, one could then simply
pick out the main-beat harmonies for a ‘triumphant’ accompaniment.
BWV 630a Heut triumphieret Gottes Sohn

Copy: J. G. Walther.
Whether in being two bars long rather than nearly three the final
‘Hallelujah Ds’ amount to an ‘early version’ (KB p. 74) is doubtful: Walther’s
final bar has three beats (i.e. forgets or disregards the opening upbeat) and
seems short-breathed.
BWV 631 Komm, Gott Schöpfer, Heiliger Geist

Further late copies only (more copies of the longer version: see BWV 667).
Two staves.
295 BWV 631
The TEXT is Luther’s paraphrase (changing the verse-order?) of the ninth-

century Vespers hymn for Whitsunday, ‘Veni creator spiritus’, a stricter trans-
lation than Thomas Münzer’s (Stapel 1950 pp. 154ff.). The Whit cantatas
suggest that another hymn (‘Komm, heiliger Geist, Herre Gott’) was more
in use.
Komm, Gott Schöpfer, heiliger Geist, Come, God creator, Holy Ghost,
besuch das Herz der Menschen dein, visit the hearts of your mankind,
mit Gnaden sie füll, denn du weisst, fill them with grace, for you know
dass sie dein Geschöpfe sein. that they are your creatures.
Four further verses describe the Holy Ghost as comforter, the living fire, the
finger of God, and the Spirit directing faith. A fifth is the doxology.
The MELODY, which adapts the Gregorian melody (Example 146), was
published with the text. Used in BWV 667, 370 and 218 (a Telemann work).
Example 146
It is not certain that the shorter setting (first version BWV 631a) preceded the
longer (first version BWV 667a): a generalization that Bach always extended,
never shortened (KB p. 96), cannot amount to proof. Clearer from P 283 is
that the setting BWV 631a (originally a fair copy?) was revised much as BWV
620a was, by introducing a few more varied rhythmic groups of semiquavers.
The revision, BWV 631, is less uniform in figuration but richer in written
ornaments, corresponding to BWV 667 (where the revisions originated?:
KB p. 96) as 631a does to 667a. There are still some uncertainties in this
history, but from extant sources it seems that both Whit settings BWV 631
and 667 had a Weimar and a Leipzig version.
Perhaps Spitta exaggerates in saying the pedal has little to do (I p. 601),
but it is certainly not in the Ob style even if the setting as a whole is – melody
in soprano (as if being sung) with a standard motif in inner voices (often in
thirds) above a distinct pedal motif. Its startling gigue-like character makes
the search for images difficult. The middle parts are said to symbolize the
scattered tongues of fire (Steglich 1935 p. 122), and the compound time
expresses a Trinity of which the Third Person is heard in the pedal’s quaver,
the third of each beat (Arfken 1965). Two rests and a quaver do perhaps
amount to a figura of sorts. The 12/8 treatment seems an afterthought
296 BWV 631–632
in P 283, its signature placed after the C signature in which the melody
was first written; and as Terry noticed, the bass line is much like that of
the harmonization BWV 370. Clearly, the revision meant to build on the
tendency towards semiquaver sextolets and create a more cumulative effect
as the piece proceeds.
The autograph direction ‘organo pleno’ in BWV 667 is there perhaps
for cyclic reasons (see introduction to BWV 651–668) and because of a
pedal cantus firmus in the second half, neither of which is relevant to
BWV 631.
BWV 631a Komm, Gott Schöpfer, Heiliger Geist

Written over in P 283; further copies by or via C. G. Meissner, J. C. Oley,
C. F. Penzel, J. C. Kittel.
The smoothness of the revised version BWV 631 may reflect the composer’s
desire to soften too overt a gigue style in this earlier version, especially as the
piece moves to a climactic imperfect cadence. In its figuration, BWV 631a
corresponds to the first eight bars of BWV 667a/b. The sources made by
authoritative copyists who knew this version might imply that the revision
(BWV 631) was made late, indeed perhaps related in some respect to work
on BWV 667.
BWV 632 Herr Jesu Christ, dich zu uns wend

Two staves.
The TEXT, said to be by Duke Wilhelm II of Sachsen-Weimar, was published

in 1648 and sung every Sunday in many places as a prayer immediately before
the sermon, after the priest had entered the pulpit (Stiller 1970 p. 103).
Herr Jesu Christ, dich zu uns wend, Lord Jesu Christ turn to us,
dein’ Heilgen Geist du zu uns send, send your Holy Spirit to us,
mit Hilf und Gnad, Herr, uns regier rule us, Lord, with help and grace,
und uns den Weg zur Wahrheit führ. and show us the way to truth.
V. 3 speaks of ‘eternal joy and blissful light’, v. 4 is a doxology.

297 BWV 632
The MELODY was known from 1628 (Example 147), its metre much like
the Geneva Psalter’s (1562). So many organ settings (BWV 632, 655, 709,
726, 749) probably reflect the need for an interlude before the sermon. Also
in BWV 332.
Example 147
Both the manual motif and the pedal line are clearly derived from the cantus,
which, however, is somewhat disguised at first by being made more flow-
ing. The triads of the accompaniment and the interludes between phrases
are no empty broken chords but, in Ob fashion, press the harmony for-
ward and constantly surprise: see for instance bb. 3–4, a tissue of references
to the triadic cantus. The opening tenor motif is less developed than one
might anticipate, while the bass-line is more than a little similar to BWV
655’s. These two chorales have much in common in their penultimate line
(BWV 632 b. 12, BWV 655 b. 63) and elsewhere.
This bass is unusual, not a quasi-ostinato but a quasi-canon at the
fifteenth:
1–3 line 1
4 (last 3 notes)–6 line 2 (anticipated 12 bar earlier)
8, 13 line 3 (anticipated in b. 7)
9, 15 line 4 (fourth below)
Some diminutions are introduced, and with that of b. 1 or of b. 11 the pedal

seems actually to be avoiding a simple canon. Semiquaver triads derived
from the melody govern the inner parts rising around the chorale and are
sweetly melodious, triadic and consonant. Note that b. 14 does not imply
two manuals!
It is not known why the composer repeated (wrote in repeat-marks in
P 283 for) the second half of the melody, when hymnbooks give it without
repeat. But the resultant constant quoting of notes associated with the open-
ing syllables ‘Herr Jesu’ implies a prayer constantly being repeated: ‘turn to
us, turn to us’. In the repertory of broken-chord motifs accompanying the
melody in Böhm’s and Walther’s variations on ‘Herr Jesu Christ, dich zu uns
wend’, something faintly similar occurs: see Example 148. But the distance
in Weimar from the town church to the Court Chapel is only too clear. BWV
632 is not unlike a certain kind of allemande, even to the upbeat and the
final arpeggiated chord (cf. Buxtehude’s Suite in F major, BuxWV 238).
298 BWV 633
Example 148
BWV 633 Liebster Jesu, wir sind hier

Further copies: by or via J. T. Krebs, J. G. Walther, C. G. Meissner, J. G.
Müthel, J. C. Oley, C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Following BWV 634 in the autograph. Three staves, headed ‘distinctius’

(added), ‘Forte’ above top stave, ‘Pia’ between staves, ‘Ped’ above third
stave. Headed in Krebs’s copy, ‘alio modo distinctig [ = distinctius]’.
There are two TEXTS with this melody and first line, and it is not clear
which was meant (Leaver 1985 p. 232). Being used both before the sermon
and for Whitsuntide, BWV 633 is usually associated with T. Clausnitzer’s
hymn of 1663:
Liebster Jesu, wir sind hier, Dearest Jesu, we are here

dich und dein Wort anzuhören; to listen to you and your word;
lenke Sinnen und Begier direct our minds and desires
auf die süssen Himmelslehren, to the sweet teachings of heaven,
dass die Herzen von der Erden that our hearts be drawn from earth
ganz zu dir gezogen werden. wholly towards you.
The following two verses continue the prayer.
The MELODY, Example 149, was reshaped for this text in 1687 (Terry 1921
p. 251). Harmonized in BWV 373, and set in BWV 706 (twice), 730, 731
and 754. Krebs’s order in P 801 (BWW 706.i, 706.ii, 634, 633) presents a
pair of settings for each version of the melody – as in some original but lost
Bach manuscript?
In P 283, both BWV 633 and BWV 634 have repeat marks for their two
five-bar halves, as do Walther’s variations ‘Liebster Jesu’ (imitating BWV
633?). The added heading ‘for two manuals’ is in BWV 634 only, the rubrics
‘forte’ and ‘piano’ in 633 only. Not only are the five voices more spaciously
299 BWV 633–634
Example 149
written out on the three staves of BWV 633 but the opportunity was taken
to give the inner parts a little more activity at the beginning and end of each
chorale-line. Distinctius, ‘more distinctly’, must refer merely to the way the
music is written out on three staves, not to the way it is played, or to the
melody of b. 1 being ‘plainer’ (Keller 1948 p. 163). BWV 634 is on the left
page, 633 on the right: was the latter’s title originally there for an alio modo
setting?
On writing out BWV 633, the composer removed the uncanonic deco-
ration in b. 1 of 634 and put in four more references to the key motif of the
movement, in bb. 1 and 11. This motif is a group of four quavers, perhaps
derived from the first notes of the melody, taking various shapes in both
pedal and manual parts. In the first bar not only does the canon begin but
there are four versions of this quaver motif, with harmonies made com-
plex by accented passing-notes, which are especially noticeable whenever
the pedal has the motif. As a consequence, most bars have some unusual
or even dissonant harmonic progression, including two consecutive added
sixths which give an unusual tinge to the harmony (end of b. 4, beginning
of b. 5). The canon is complete, per giusti intervalli unlike BWV 619, and
keeps to the melody in the version found at the period. It may well symbolize
‘hearts drawn from earth wholly towards you’.
Although the motion of the chorale is quiet, the harmony is rich enough
to support the large amount of repetition there is in the setting. Its form
is a miniature ababcbcb, as is most clearly seen in the pedal line, which
has virtually the same five-bar phrase four times, ending with the same
semibreve A. By chance (?), the result is the most integrated chorale in the
collection. The two pairs of parts above a pedal reflect Grigny’s five-part
layout, unlike the canon of BWV 619, which has one canonic line in each
hand.

Further copies by or via J. T. Krebs, J. C. Kittel.
Before BWV 633 in the autograph; two staves, headed ‘in Canone alla Quinta’
and (later?) ‘à 2 Clav & Ped.’, with two brackets pairing off upper voices.
300 BWV 634–635
BWV 633 is neither a ‘variant’ (Schmieder BWV) nor an ‘alternative’ version

(EB 6589) in the usual sense of those terms. See BWV 633 above.
The distinction between ‘a 2 Clav.’ (BWV 634) and ‘forte/piano’ (BWV
633) may reflect the date: although the former phrase continued to be used
in Leipzig works (Clavierübung III, IV ), the latter was the more modern
(Clavierübung II, also BWV 552.i). Perhaps BWV 634 began as a single-
manual movement, not requiring two manuals as pressingly as BWV 624
(because of part writing) or 604 (because of solo colour); its two moments
of awkward part-crossings arise because the left hand is of thematic impor-
tance. Although its canonic distribution is different, similar points could be
made about BWV 619.
BWV 635 Dies sind die heilgen zehn Gebot

Further copies: by or via J. T. Krebs, C. G. Meissner, Mempell–Preller, J. C.
Oley, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT is Luther’s versification (with opening and closing stanzas) of the
Ten Commandments; a shorter version, beginning ‘Mensch, willst du leben
seliglich’, is one of the unset chorales in P 283.
Dies sind die heilgen zehn Gebot, These are the holy Ten Commandments,
die uns gab unser Herre Gott which our Lord God gave to us
durch Mosen, seiner Diener treu, through Moses, his true servant,
hoch auf dem Berge Sinai. high on Mount Sinai.
Kyrieleis! Lord have mercy.
The verses list the commandments, with a prayer for help from ‘our
mediator’.
The MELODY was published in 1524 with the text, probably an adaptation
of the pilgrim song ‘In Gottes Namen fahren wir’ (Terry 1921 p. 148),
although this might be vice-versa. Harmonized in BWV 298 (Example 150),
set in BWV 678, 679 and without text in Cantata 77. The cantata has ten
trumpet entries or separate phrases; BWV 635 contains ten entries of the
subject proper (i.e. the last two notes making a semitone); BWV 678 is the
tenth chorale in Clavierübung III; BWV 679 has ten fugal entries. The ten
entries of BWV 635 are not immediately recognizable, any more than the
ten semitones encompassed by g–f in the subject in BWV 679.
301 BWV 635
Example 150
Although the quaver motif is derived from the cantus firmus and is both
rectus and inversus as in BWV 632, the result is new, since the melody
now is in plain minims, the canonic imitation is between tenor and bass,
and the repeated notes produce a quite different Affekt. Also, the running
semiquavers paraphrase the first chorale-line (see Example 151), and are
instantly adaptable to the three other parts, sustaining a flow otherwise
endangered by so many repeated patterns. In the bass, they open out into
a shape typical of alternate-foot pedalling, with the lower notes referring
to the melody. This setting, therefore, is derived to an exceptional degree
from its melody, one way or another, with accented passing notes appearing
on two levels: both quavers and semiquavers. The plan of G-mixolydian
moving towards a minor plagal cadence is followed in the late setting of the
same chorale, BWV 678.
Example 151
The result is a striking chorale with which to open the Catechism section
of the album. Although the twofold use of repetition – repeated notes in the
motif, repeated use of the motif (twenty-five times) – can be seen as con-
stantly ‘confirming’ the text, whether there is actually a reference to Ten has
been disputed. Bach ‘was expressing the idea of insistence, order, dogma –
anything but statistics’ (Grace 1922 p. 123), and Schweitzer had to exercise
ingenuity in order to count only ten entries, for which he has been much
criticized. Nevertheless, there are indeed ten diatonic entries preserving the
exact intervals rectus (GGGGGABC); and if the final bar is read as a minim
(cf. BWV 621), there are exactly twenty bars in a movement whose cantus
firmus is notated in time-values twice as long as usual.
The use of a motif a whole bar long leads to one single harmony for many
a bar, something very unusual in the Ob. The main beats 1 and 3 usually
302 BWV 635–636
outline the harmonic progress, beat 2 frequently an accented passing-note,

beat 4 almost always so. To avoid too much repetition, four times the motif
enters halfway through the bar, but a setting with so many repeated notes
is easy to hear as drumming in the law. The plainest and most fluent bar is
the penultimate, perhaps originally intended as a cadence similar to BWV
727’s. (In P 283, pedal b. 19 might have begun the same as b. 14. That the
final bar should be a minim was perhaps forgotten in the hasty writing?)
BWV 636 Vater unser im Himmelreich

Further copies: by or via C. G. Meissner, an early anon copy, J. C. Oley, J. G.
Müthel, C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT is Luther’s versification of the Lord’s Prayer, a rather freer version
than previous German translations.
Vater unser im Himmelreich, Our Father in Heaven,

der du uns alle heissest gleich who bids us all to be equal
Brüder sein und dich rufen an brothers and to call to you,
und willst das Beten von uns han, and desires prayer from us:
gib, dass nicht bet allein der Mund, grant that our mouth alone does not pray,
hilf, dass es geh’ aus Herzensgrund. help, that it come from the bottom of our
hearts.
Verses 2–8 develop the first line of each section of the Paternoster, v. 9 the
Amen.
The MELODY, which may be based (by Luther himself) on an earlier song,
was published with the text in 1539 and remained unusually close to the
original: see Example 152. It is harmonized in BWV 461, set in BWV 682,
683, 737 and BWV 760–763, in the St John Passion and (to other texts) in
cantatas BWV 90, 101, 102.
Example 152
303 BWV 636–637
Typically of the Ob, the main motif appears as both a single and a double
cell (Example 153), one surely derived from the melody’s opening notes. It
may be the second that caused an unexpected bass line in b. 1 and an altered
melody in b. 3 (b instead of g ). The motif is unusually varied, particularly
in comparison with that of BWV 635, mingling rectus and inversus forms
freely, even arbitrarily. A description of such bass figures as ‘motifs de la
quiétude’ (Schweitzer 1905 p. 349) fits the slow harmonic rhythm already
Example 153
clear in b. 1. Keller’s demonstration that some of the counterpoint is ‘derived

from’ a simple four-part setting (1948 p. 150) begs questions about priority
but recognizes the pedigree of such chorales, especially their tendency to
close each line like a hymn, with a strong perfect cadence marking classic
key-progressions (tonic, relative, tonic, dominant, relative, tonic).
Keller’s impression must also be due in part to the motifs, since the
later setting BWV 683 has similar harmony and treats its motifs similarly.
Yet it is much farther from being a simple harmonization than BWV 636,
whose motifs are broken chords ‘circumscribing’ the harmony, sustain-
ing its tension and creating new effects (see last alto phrase). No bars are
repeated singly or otherwise, and the movement continues through the
cadences, with inner voices then dropping their suspensions. Naturally, part
of the special singing quality is owing to the melody itself, which, though
in some ways similar to BWV 637, is more warmly diatonic. In manner,
the setting reminds the player of BWV 625, for both rise in the melody
at one point and have three accompanying lines which draw on a single
motif. But their motifs are subtly different: on-beat in BWV 625, off-beat in
BWV 636.
BWV 637 Durch Adams Fall ist ganz verderbt

Further copies: by or via J. T. Krebs (two), C. G. Meissner, C. F. Penzel, J. P.
Kirnberger, J. C. Kittel.
Two staves.
The TEXT of L. Spengler’s hymn was published in 1524, associated in general

with penitential texts ‘of human misery and ruin’ (Freylinghausen 1741).
304 BWV 637
Durch Adams Fall ist ganz verderbt Through Adam’s fall is totally spoiled
menschlich Natur und Wesen, all human nature and being,
dasselb Gift ist auf uns geerbt, the same poison is bequeathed to us
dass wir nicht konnten gnesen, from which we cannot be delivered
ohn Gottes Trost, der uns erlöst without the solace of God, who has
hat von dem grossen Schaden, redeemed us from the great disgrace
darein die Schlang Eva bezwang, by which the serpent forced Eve
Gottes Zorn auf sich zu laden. to draw God’s anger down upon herself.
Eight further verses concern the need for the Saviour and for faith.
The MELODY, Example 154, was published with the words in 1535 but,
just as the text is Meistersinger-like, so the melody is of a Reformation battle
song (Pavia 1525). It appears in BWV 705, 1101 and Cantatas 18 (1713/14)
and 109 (1723).
Example 154
One of the most original of all settings, BWV 637’s expressiveness hangs
on the two standard chromatic ideas: step (passus) and leap (saltus). See
Example 155. So striking in Affekt and harmonic tension is it that one can
miss how ingeniously it uses its motifs: an Ob chorale par excellence. The
passus is constantly manipulated to produce unbroken semiquavers, the
saltus has all three kinds of 7th. (The first motif originally contained an
échappée: see KB p. 44.) A pedal leap signals each cantus line, and the last
Example 155
leap is delayed to pass straight into the cadence. In addition to the repeated
section (not written out in P 283) there is another important repetition
(line 6 = line 3 = line 1), which is not the case in the setting in Cantata 18.
At least twice the cantus is dissonant with the diminished 7ths, logical
but also part of an unease which is at its greatest when the pedal drops to
some new leading-note. Within six beats in bb. 13–14, the keys of D minor,
305 BWV 637–638
G minor, B major (?), G minor, E minor and G major are temporarily

established by this means. Another ‘strained’ effect is produced when the
chromatic line is inverted in the second half, its harmonies then even odder.
The final progression towards A major is simple, but expectation is built
up through diminished 7ths, both leaps and chords (four in the last five
beats). It easily escapes notice that the final cadence is very similar to that of
BWV 638 and that there is no ‘adagio’. Is it perhaps no slower, therefore,
than BWV 636?
In that it treats false relations and difficult intervals in its own way,
often unprepared or unresolved, BWV 637 offers many opportunities to
relate music to text – the discords with Original Sin (the breaking of rules
musical and moral), the falling pedal with the Fall of Adam, redeemed at
the end by a final major cadence of hope. For other suggestions, see Budday
1977. The broken bass is not only an allegory of ‘falling’ but also a good
example of tmesis – gaps or rests suspirantis animae, ‘for a sighing of the
spirit’, in Athanasius Kircher’s words (Schmitz 1970 p. 72). Against the
‘series of almost irremediable stumbles’ in the pedal (Terry 1921 p. 152), a
constant cantus expresses constant trust in Jesus (Arfken 1965). Perhaps the
major/minor twist, there from the very beginning, relates to the ‘spoiled’ of
v. 1 (Keller 1948 p. 164) or the ‘evil serpent’ of line 7 (Chailley 1974 p. 111).
Such detailed imagery goes against Spitta’s view that the whole text
is involved (I p. 593). In Buxtehude’s setting the imagery is more specific,
i.e. the falling bass accompanies line 1 only, the chromatic phrase line 3
only.
BWV 638 Es ist das Heil uns kommen her

Further copies: by or via J. C. Oley, C. G. Meissner, J. G. Müthel, C. F. Penzel,
Two staves.
The TEXT of P. Speratus’s hymn was published in 1523 and acquired various
associations. Fourteen verses proclaim a central doctrine of Protestantism.
Es ist das Heil uns kommen her It is salvation that comes to us

von Gnad und lauter Güte. through grace and pure goodness;
die Werk, die helfen nimmermehr, good works never help
sie mögen nicht behüten. and will not preserve us.
Der Glaub sieht Jesum Christum an, Faith looks to Jesus Christ
der hat gnug für uns all getan, who has done enough for us all;
er ist der Mittler worden. he has become the mediator.
306 BWV 638–638a
The MELODY is that of an Easter song, published with the text and used
in Cantatas for Weimar (155) and Leipzig (9, 86, 117, 155, 186). See
Example 156.
Example 156
As with other jubilant settings (BWV 606 and 609), the rhythmic scheme
is clear: crotchets for melody, running quavers for bass, running semi-
quavers (sometimes in sixths) for inner parts. Again, the common semi-
quaver suspirans figure (second note usually an accented passing-note) can
be derived from the cantus: see the last four notes of line 1 (CBAG) or the
last line (GFED). And again the bass quavers mark the structure by halting
at the end of each line with a ‘perfect cadence of affirmation’. The highest
phrase of the setting occurs at v. 1’s word ‘Glaub’ in b. 9, which is also the
point at which the bass, after beginning line 5 in the same way as lines 1 and 3,
immediately modulates.
Given that there is an overall conception common to both BWV 637
and 638 – i.e. in both a chorale-melody is accompanied by inner running
semiquavers and a strongly characterized leaping bass – the exceptional
contrast between them in mode, harmonic rhythm, flow, motif-shape
and presumably tempo, seems hardly an accident. Are they a pair, a de-
liberate presentation of the doctrines of sin followed by salvation, writ-
ten back-to-back in P 283 and each especially appropriate to a catechism
section?
BWV 638a Es ist das Heil uns kommen her

Copies: J. T. Krebs and J. G. Walther (two).
The suspirans at the end of the second line of BWV 638 (b. 4 third beat) is
found in P 283 but not in Krebs and Walther, which supports the idea that
the autograph version is revised from an earlier.
307 BWV 639
BWV 639 Ich ruf zu dir, Herr Jesu Christ

Further copies: by or via J. T. Krebs, J. G. Walther, anon early copy, C. G.
Meissner, J. G. Müthel, C. F. Penzel, J. P. Kirnberger, J. C. Kittel, J. A. G.
Wechmar; also LM 4708 (‘Neumeister Collection’, C time).
Two staves (three in P 802); headed ‘à 2 Clav. & Ped’ (not in ‘Neumeister’).
The TEXT of J. Agricola’s five-verse hymn was published in 1529 and became
associated with various Sundays after Trinity.
Ich ruf zu dir, Herr Jesu Christ, I call to you, Lord Jesu Christ,
ich bitt, erhör mein Klagen; I beg, hear my complaint;
verleih mir Gnad zu dieser Frist, grant me grace at this time,
lass mich doch nicht verzagen. let me not despair.
Den rechten Glauben, Herr, ich mein, The true faith, Lord, I aspire to,
den wollest du mir geben, which you wish to give me,
dir zu leben, [is] to live unto you,
mein’m Nächsten nütz zu sein, to be of use to my neighbour,
dein Wort zu halten eben. to keep your word.
Compare the reference to neighbour with the couplet on the Ob title-page

(p. 228).
The MELODY was published with the text, used in cantatas for 4th
Sunday after Trinity 177 (1732) and 185 (1715): Example 157. See also
BWV Anh.II 73.
Example 157
This, the only trio of the album, creates something new with the notion
of melody plus accompaniment plus pedal: a quiet melodious soprano, a
gently throbbing bass, a flowing accompaniment, each with a clear and
308 BWV 639
striking Affekt in a key with difficult thirds (a–c, d–f). The cello-like
obbligato line of the tenor has suggested to some that BWV 639 is a
transcription like BWV 649 (BG 25.ii), and there are certain parallels
in style and layout between it and the cantata movement BWV 180.iii
(1724). In P 283, the slurs appear largely where there is enough room,
not when there is not: but presumably they were meant throughout,
either indicating sostenuto (Butt 1990 p. 185) or, if lively enough, the viol-
like effect of Scheidt’s slurred groups in Tabulatura nova (1624, ‘imitatio
violistica’). Though without bowing marks, the lute obbligato in the St John
Passion No. 19 offers a parallel, with its broken chords against repeated bass
quavers.
BWV 639 has the most basso-continuo-like pedal part in the album, a
throbbing bass without accented passing-notes. Mostly the tenor line has
full broken chords that could easily have been turned into a regular two-part
accompaniment. Why the melody’s ornaments die out, like the tenor’s slurs,
is not clear but was probably the result of haste, and both might be supposed
to continue. (Ornaments added in J. T. Krebs’s copy – mordents in bb. 4,
14, 16, trills in b. 13 – are no more than suggestive and, for a movement
so supplicatory, quite unimaginative.) Pachelbel’s fugue and Buxtehude’s
fantasia on the same melody created no precedent for what is the least
traditional Ob setting.
Scarcely a better example can be found for some of the qualities Matthe-
son heard in F minor: ‘scheinet eine gelinde und gelassene wiewol dabey
tieffe und schwere . . . tödliche Herzens-Angst’ (‘seems to represent a mild,
calm, and at the same time a deep and heavy . . . fatal anxiety’ – 1713,
pp. 248–9), and though his ideas were based on vocal music, they were
well known. Meantone F minor makes three parts more feasible than four;
but this key, for a chorale listed by J. G. Walther as aeolian or in A minor
(1732 p. 414) and by Mattheson in D minor (1739 p. 162), has also been
seen as evidence that J. S. Bach knew the more modern temperaments (Eck
1981 pp. 154–61), perhaps on the new organ in Halle. The key was certainly
unusual. The chorale’s appearance in the ‘Neumeister Collection’ (a faulty
copy of a source other than P 283) does not prove it to have originated
before the Ob, much less to be an instance of ‘older material . . . absorbed’
in the Ob (Wolff 1991 p. 301). One could argue either way from a unique
trio layout such as would appeal to a late eighteenth-century compiler like
Neumeister.
As in some other Ob settings, only towards the end do the accompanying
semiquavers interfere with the cantus. Was the rubric for two manuals added
to the title because this was the original intention, because they turned out
to be desirable, or because the didactic purpose of the eventual titlepage
encouraged rubrics?
309 BWV 640
BWV 640 In dich hab ich gehoffet, Herr

Further copies: by or via J. T. Krebs, J. G. Walther, C.G. Meissner, J. C. Oley,
J. G. Müthel, J. P. Kirnberger, J. C. Kittel.
Two staves; headed ‘alio Modo’; in Krebs and Meissner, as in P 283, preceded
by unused staves.
The TEXT of A. Reusner’s hymn, based on Ps. 31, was published in 1533; it
was associated generally with ‘spiritual struggle and victory’ (Freylinghausen
1741).
In dich hab’ ich gehoffet, Herr; In you have I hoped, Lord;
hilf, dass ich nicht zuschanden werd help me that I be not disgraced
noch ewiglich zu Spotte. nor mocked in eternity.
Das bitt ich dich: This I pray you:
erhalte mich sustain me
in deiner Treu, mein Gotte. in your faithfulness, my God.
The opening alludes to the Te Deum. Six further verses are a prayer and
doxology.
The MELODY, Example 158, was printed in 1536 to the text ‘Christ ist
erstanden’ and known from the fourteenth-century hymn ‘Christus iam
resurrexit’. For a melody more commonly used by J. S. Bach and Walther,
see BWV 712 – perhaps the one intended for the first setting in P 283 (not
this, headed ‘alio modo’)?
Example 158
As in BWV 636, the accompanying motif can probably be derived from

the melody: bfg. A similar idea is used as the head motif in a lively aria in
E minor in Cantata 65 (1724). But typical of the Ob is that despite similarities,
the figuration of BWV 636 and 640 is different, even when the latter’s motif
is extended (e.g. tenor in bb. 1 and 6). The harmonies are much like a hymn-
setting’s, and the motif, being imitative by nature, gives many opportunities
for thirds and for a disjunctive pedal line of great independence, forcing
310 BWV 640–641
melodic suspensions in bb. 1–2 and 3–4. Although the serene kinds of
seventh chord in what is a series of simple harmonies – see for example last
two bars – are typical of the composer’s style by 1715, all the tonics and
dominants produce a homogeneous effect, not least the repeated section
bb. 7–8 (bb. 1–2) where the melody returns to its hymnbook form. Since
the original tie bb. 1–2 was an afterthought, however, perhaps its absence in
bb. 7–8 was unintended.
The motif’s angularity and a generally low, rich tessitura suggest no light
jubilation and certainly not the liveliness of such cantata movements as
BWV 65.iv. The dactyls have developed far from those found in simple
variations, for the semiquavers become continuous and, played in a certain
pesante manner, can be heard as allusion to the firm hope of the text.
BWV 641 Wenn wir in höchsten Nöten sein

Further copies: by or via J. C. Oley, anon. early copy, C. F. Penzel, J. P.
Kirnberger, J. C. Kittel.
Two staves; headed (subsequently?) ‘à 2 Clav & Ped’.
The TEXT of P. Eber’s seven-verse hymn was first printed in 1560, founded
on J. Camerarius’s ‘In tenebris nostrae’ (1546).
Wenn wir in höchsten Nöten sein Whenever we are in the greatest

distress
und wissen nicht, wo aus noch ein, and do not know where to turn,
und finden weder Hilf noch Rat, and find neither help nor advice,
ob wir gleich sorgen früh und spät, although we worry day and night,
so ist dies unser Trost allein, then is this our only comfort,
dass wir zusammen insgemein that all of us together
dich anrufen, O treuer Gott, call on you, O true God,
um Rettung aus der Angst und Not . . . for rescue from fear and distress . . .
The MELODY by Louis Bourgeois was published in 1543 (Ps. 140 or

Ten Commandments) and associated with this text in Ammerbach’s
Tabulaturbuch, 1571. According to Terry (1921 p. 316) the 1588 form is as
Example 159. It is harmonized in BWV 431 and 432, and set in BWV 668.
For the relationship to contemporary and later re-workings, see also

BWV 668 and 668a. Spitta already saw that the accompanying motif is
derived from the melody (I pp. 590-1) but the relationship is no more
obvious to the ear here than elsewhere in the Ob. It is clearer in the BWV 668
version because there the melody is less decorated, though there too
311 BWV 641–642
Example 159
imitations intersperse the lines and thus disguise the fact that the motif runs
through all its lines, rectus or inversus. (Most are inversus, as they are not in
BWV 640 – another difference between pairs of chorales?)
So highly decorated a melody suggests a tempo about half that for
BWV 668, where the fore-imitation and interludes also require more
momentum. Perhaps BWV 641 was already an intricate, more detailed
version of an earlier and simpler setting? A chromatics-free melody of this
kind hovering around g –b (compare the coloratura BWV 622) produces a
chorale of idiomatic beauty rather than rhetoric. The appoggiaturas in the
melody are ports de voix of the kind illustrated in the CbWFB (1720), intro-
duced here as if a certain frenchified elegance were slipped into a Sesquialtera
solo of the Buxtehude kind.
The many thirds and sixths, from first bar to last, help produce the air of
sweet gentleness, but so do the appoggiaturas on several second and fourth
beats. Again, the persistent motif in the accompaniment has the effect of
reiterating the opening words, as Schweitzer suggested (1905 p. 357), even
when buried as an appoggiatura in the tenor (aag in bb. 2 and 6).
The coloraturas, unlike most of those in BWV 614 and 622, centre around
turning phrases that lead to the next note of the cantus, which is placed where
it would be even if there were no decoration. This is a particular technique
that can be understood in two ways: these embellishments could be taken
out in order to produce BWV 668, or they could have been added in order
to produce BWV 641, where they are written in smaller notes in P 283.
Naturally, some of the patterns can be found elsewhere; the second half of
b. 1 in ‘O Mensch, bewein’, or the second beat of b. 2 in ‘Ich ruf zu dir’
Like BWV 639 and 643, it is a model for a particular kind of touching,
inexpressible expressiveness.
On the question of two manuals, see note under BWV 639.
BWV 642 Wer nur den lieben Gott läßt walten

Further copies: by or via J. T. Krebs (two), J. G. Müthel, J. P. Kirnberger,
J. C. Kittel.
Two staves.
312 BWV 642
The seven-verse TEXT by G. Neumark was published with its melody in

1641, often associated with the Fifth Sunday after Trinity (Stiller 1970
p. 229).
Wer nur den lieben Gott lässt walten He who allows dear God to rule him.
and hoffet auf ihn allezeit, and hopes in him at all times,
den wird er wunderbar erhalten will be wonderfully sustained by him
in aller Not und Traurigkeit. in all distress and sadness.
Wer Gott, dem Allerhöchsten, traut, He who trusts in God the most high
der hat auf keinen Sand gebaut. has not built on sand.
The MELODY has both duple and triple-time forms, versatile and much
used. See Example 160. Further used in chorales BWV 647, 690, 691, 691a;
Cantatas 21 (1714? ‘for all seasons’), 27, 93, 84, 88, 179 (1723–7), 166 (Fourth
after Easter 1724), 197 (wedding cantata, altered); and harmonized in
BWV 434. Cantatas 27, 84 and 166 use the melody with the text of the
funeral hymn ‘Wer weiss, wie nahe mir mein Ende’ (‘Who knows how near
is my end?’).
Example 160
Although to some this setting is ‘animated’ rather than ‘serene’, the dactyl
motif is surely heavier than in other instances and is not unlike that in
BWV 616. BWV 642 joins with BWV 602, 605, 615, 616, 618, 620, 621,
623, 627 (vv. 1 and 2), 629, 637 and 640 to complete a repertory of this
most adaptable of motifs, which in BWV 642 usually occurs in thirds (note
however that the pedal dactyls occur in isolation).
Holding back the last cantus line for a brief interlude corresponds with
BWV 690, where the harmony is similar and its motif (a simple suspirans)
equally fertile. Since BWV 642 and 643 may be two of the earliest in the album
(Dadelsen 1963), it is not surprising that they have in common such features
as harmonization by sequence (BWV 642 bb. 13–14, BWV 643 bb. 13–15), in
both cases in thirds or sixths. The two chorales are also more like a decorated
harmonization than (e.g.) BWV 644, with harmony changing on each beat,
many parallel thirds and sixths, and a motif reinforcing the 4/4 more than
the gliding scales of BWV 644 could. If BWV 642 and 643 were conceived as a
313 BWV 642–643
pair, separated by seventeen unset titles, such similarities in conception (and

tempo?) would once again serve to emphasize their difference in execution –
minor versus major, forcefulness versus resignation.
J. L. Krebs marks his chorale ‘Ich ruf zu dir, Herr Jesu Christ’, whose
figuration seems to be based on that of BWV 642, pro organo pleno.
BWV 643 Alle Menschen müssen sterben

Further copies: by or via J. T. Krebs, J. G. Walther, J. C. Oley, J. G. Müthel,
C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves; headed ‘Alio modo’.
The TEXT of J. G. Albinus’s hymn was written for a funeral in 1652, later
associated with texts ‘of Heaven and the heavenly Jerusalem’ (Freylinghausen
1741).
Alle Menschen müssen sterben, All mankind must die,

alles Fleisch ist gleich wie Heu; all flesh is as grass;
was da lebet, muss verderben, what lives must perish,
soll es anders werden neu. if it is to become somehow new.
Dieser Leib der muss verwesen, This body must wither away,
wenn er anders soll genesen if it is to be delivered
zu der grossen Herrlichkeit, to the great splendour
die den Frommen ist bereit. prepared for the righteous.
The remaining seven verses move towards the sentiment most clearly
summed up in v. 6:
O Jerusalem, du schöne, Jerusalem the fair,

ach wie helle glänzest du! O how brightly you shine!
The MELODY dates from c. 1660. BWV 643 takes the simplest of several
versions, as did pietist hymnbooks; Example 161 shows a 1687 form (Terry
1921 p. 93). Perhaps alio modo in P 283 meant an alternative melody (but
see BWV 1117) and the other was to have been the one used for the text’s
last verse in Cantata 162, 1715/16.
Whether the motif derives from the opening notes of the melody, or from
the final cadence, or from anywhere else, this rapturous setting provides a
good example of the single motif throughout a chorale, an example unique
in the Ob. The little pattern is found in partitas of Böhm and Vetter (1713),
314 BWV 643–644
Example 161
but as with the common dactyl, it has different aspects in different settings.
In BWV 643 (but not in BWV 602) the middle note of the three semi-
quavers creates a discord on most half-beats, and tossed between manual and
pedal the motif undergoes various changes. Heartless though it is to suggest
it, the special, rapt, bittersweet consonance of the setting hangs on this
discord.
The web of motivic allusion in the chorale is unbroken, with some variety
given by the overlapping between pedal and manual, a peculiar stretto. The
five bars which begin with two Bs in the melody (bb. 3, 4, 11, 14, 15) are
harmonized differently if similarly, and there are no duplicated bars or parts
of bars, despite the sequences from first bar to last. Harmonic tension occurs
exactly where it is most required, i.e. at the three-quarter point (b. 12).
Although all the thirds and sixths look earlier than the counterpoint of
‘O Traurigkeit’ (see p. 576 below), this celebrated setting is as sophisticated
as it is affecting, the very thirds and sixths often dissonant. Its secret seems
to be a judicious four-part harmony, immediately resolved discords, and a
plain, archetypal melody. Spitta must be right to find that not even such
moments of ‘indescribable expressiveness’ as the first beat of the last bar
(sudden modulation, false relation, melody inflected by the motif) are open
to particular imagery (I p. 590), though the ‘celestial happiness’ heard by
Schweitzer (1905 p. 350) is there in the last verse.
BWV 644 Ach wie nichtig, ach wie flüchtig

Further copies: by or via J. C. Oley, C. F. Penzel, J. P. Kirnberger, J. C. Kittel.
Two staves.
The TEXT of M. Franck’s eight-verse hymn, published in 1652, was not in

the regular list at Leipzig (Stiller 1970 p. 223). The verses alternately invert
the order of ‘flüchtig’ and ‘nichtig’ in the first line: BWV 644’s title is the
first line of v. 1 in Weimar 1681 but of v. 2 in Franck’s book. Cantata 26
follows Franck; Böhm’s setting (see below) is as Weimar.
315 BWV 644
Ach wie flüchtig, ach wie nichtig Ah how fleeting, ah how paltry
ist der Menschen Leben! is the life of mankind!
Wie ein Nebel bald entstehet As a mist soon rises
und auch wieder bald vergehet, and as soon disperses again,
so ist unser Leben, sehet! see! so is our life.
v. 8
Ach wie nichtig, ach wie flüchtig Ah how paltry, ah how fleeting
sind der Menschen Sachen! are the things of mankind!
Alles, alles was wir sehen, All, all that we see
das muss fallen und vergehen. must fall and decay.
Wer Gott fürcht’, wird ewig stehen. He who fears God will survive for ever.
The MELODY appeared with the text in 1652 and has various forms; that
of Cantata 26 and BWV 644 is simpler in outline than some others, as was
BWV 643’s, both deliberately so made? See Example 162.
Example 162
A classic example of motivic construction, BWV 644 is based on two motifs –

manual passus or step, pedal saltus or leap – used without a break from
beginning to end, and producing a texture far removed from an ordinary
harmonization, although the main beats could be extracted to provide
exactly that. Scales up or down being so easily adaptable, care has been taken
to vary them inventively. Thus the motif is basically two beats long, but half
of it often appears alone, and as well as appearing in contrary motion, it
alternates with similar motion in thirds (see examples of both in b. 9). The
bass begins as in many different works (e.g. Cantata 161, Organ Sonata
No. 2) but is now constant to the end (as in BWV 628), interrupted only to
avoid repetition in bb. 2 and 4. Like the scales, the octave drop often appears
in variations, e.g. Walther’s ‘Herr Gott, nun schleuss den Himmel auf ’, there
of course in simpler form.
That scale patterns also accompany the same chorale in the opening cho-
rus of Cantata 26 (1724) suggests an association for the composer between
music’s scales and life’s transience. Some have heard the rests in the quasi-
pizzicato bass as picturing ‘ach wie nichtig’, but being on weak beats these
are no true tmesis. Either way, the cantata movement is not simply a larger
version of the organ chorale, since it is about twice as fast: in BWV 644 the
316 BWV 644
contrary motion, the frequent accented passing-notes (Walther’s transitus

irregularis, 1708 p. 151) and the pacing bass all compel a slower tempo.
Faster, ‘fleeting’ semiquavers were one of the patterns for chorale vari-
ations, as in Partita 4 of Böhm’s ‘Ach wie nichtig, ach wie flüchtig’ (copied
by Walther – see Example 163), though whether this is older than BWV 644
is not known. Against Böhm and even Cantata 26, BWV 644 is thoughtful,
intimate, tempting the performer to hear in the contrary-motion scales the
‘dispersing mist’ of v. 1. Even the pedal dropping out at the end is original
and curiously final, more so than in BWV 628.
Example 163
Schübler Chorales BWV 645–650
Published 1748/9? Title-page:
Sechs Chorale von verschiedener Art auf einer Orgel mit 2 Clavieren und
Pedal vorzuspielen verfertiget von Johann Sebastian Bach Königl: Pohln:
und Chur-Saechs: Hoff-Compositeur Capellm: u: Direct: Chor: Mus: Lips:
In Verlegung Joh: Georg Schüblers zu Zella am Thüringer Walde. Sind zu
haben in Leipzig bey Herr Capellm: Bachen, bey dessen Herrn Söhnen in
Berlin und Halle, u: bey dem Verleger zu Zella.
Six Chorales of various kinds to be played as preludes on an organ with

two manuals and pedal, prepared by Johann Sebastian Bach, Royal Polish
and Electoral Saxon Court Composer, Capellmeister and Director of the
musical ensemble, Leipzig. Published by Johann Georg Schübler at Zella,
in the Thuringian Forest. To be had in Leipzig from Capellmeister Bach,
from his sons in Berlin and Halle, and from the publisher in Zella.
Origins
Five of the six are literal reductions in score of Leipzig cantata arias, three
of which are from so-called chorale-cantatas. Original keys are kept but not
the figures from the continuo parts (because full scores were used?); nor is
their harmony realized, or the string articulation in BWV 649 and 650 used.
Except for BWV 650, the titles are not the cantatas’ but the chorale’s first line.
Only BWV 646 has no known cantata version, and one could reason either
way: its idiomatic details might suggest an original organ piece, for some
reason otherwise unknown; or, since three of the others draw on cantata
full scores not surviving in autograph, so could this.
Who prepared the printer’s copy is unknown. The absence of autograph
scores of BWV 645, 647 and 650 leaves doubt about various details, despite
a few corrections made by the composer in a copy of the print, some time
between 1747 and 1750 (KB pp. 130–4, 155). As with Clavierübung III,
manuscript copies, though numerous, appear to derive directly or indirectly
from the edition, so there is little doubt that the chorales in this form
originated for it.
Though brusque, a remark by Walter Emery raises important questions:
The arrangements are much less effective than the originals, and it is hard
[317] to see why Bach published them. (in Abraham 1986 p. 677)
318 Schübler Chorales
The chorales are far more literal than Bach’s other transcriptions, such as
BWV 528.i. Neither the composer’s MS amendments to the musical text
as it is nor a few details in the print not found in the cantatas∗ prove
him to have made or even authorized the ‘transcriptions’: they could have
been made from a full score by any modestly competent pupil. Perhaps,
in connection with the Halle job, W. F. Bach or someone at his request
made the transcription, i.e. copied out the cantata scores as if for organ.
(That would mean that BWV 646 was from a lost cantata, but the argu-
ment then becomes circular.) Differences in slurring between print and
cantata-manuscripts could result from a pupil’s inexperience, although
in any case, instrumental articulation was not necessarily transferred to
organ.
Apart from the necessary corrections, certain details of notation such as
the order of staves for BWV 650 would be unusual for J. S. Bach, although if
BWV 646 were not a transcription, quite how it could be part of a collection
without the composer’s approval would be hard to see. As it is, the print
stands apart in two particular respects: this is the only publication of Bach
transcriptions, and of all his works, there is the biggest gap here between
date of composition and date of publication.
Date
W. F. Bach was appointed to the Liebfrauenkirche, Halle on 16 April 1746,
and several things point to a publication date some time after this. Schübler
worked on the engraving of the Art of Fugue and also the Musical Offering
(1747), whose quality of engraving looks earlier than the chorales’ (NBA
VIII/1 KB pp. 108–9). Although Schübler was paid promptly for work on
the Musical Offering, Bach owed him money on his death (Dok II p. 497),
just possibly for recent work on the chorales, though this has been doubted
(KB p. 154).
There are further hints of late publication, though less certainly relevant:
the musical style is so distinct from the Canonic Variations as to suggest
complementary publication, i.e. in 1748 or 1749; and there are sixty-five
words on the title-page (Bach was sixty-five on 21 March 1750). But either
way, the later the volume, the less likely that the composer was involved in
it before he scrutinized the print.
∗ e.
g. the appoggiatura in b. 20 of BWV 645, the slurs in BWV 645, the shortened pedal-point in
BWV 647, the change of title for BWV 650.
Order
The order suggests a purpose for the volume, since the texts describe a
conception of Christian life (Taesler 1969) and the music produces some
symmetry (Currie 1973):
645 E major, trio, c.f . in left hand

646 E minor, trio, c.f . in pedal
647 C minor, quartet, c.f . in pedal
648 D minor, quartet, c.f . in right hand
649 B major, trio, c.f . in right hand
650 G major, trio, c.f . in left hand (? – see BWV 650 below)
While the scoring of the last is problematic, the framing of the collection
by trio settings in major keys, with left-hand melody and string obbligato,
looks intentional. The first and last have fifty-four bars each, and BWV
645 has three pages of three systems, each of three staves. These and other
numbers involved – 14 pages, 14 lines on the title-page, a total of 256 bars
and 41 lines of music (14 = B+A+C+H; 41 = J+S+B+A+C+H) could be
accidental, or the work of an intimate.
Although only BWV 645, 648 and 650 have pronounced seasonal asso-
ciations, the texts present an order of events: BWV 645 Advent, 646 Trust,
647 Hope, 648 Rejoicing, 649 Steadfastness, 650 Incarnation. Another pos-
sibility emerges if the engraver had been meant to follow the reverse order,
from one Advent to the next:
BWV 650 Advent

BWV 649 1st or 2nd Day of Easter
BWV 648 Mariae Heimsuchung (Visitation, 2 July)
BWV 647 5th Sunday after Trinity
BWV 646 11th, 19th, 22nd and 23rd Sunday after Trinity
BWV 645 27th Sunday after Trinity
Drawing on all the chorale-verses and on Leipzig practice, Bighley 1991

proposes:
BWV 645 last Sunday before Advent: preparation

BWV 646 both texts related to Advent 1 through Collect and Introit
BWV 647 text related to Advent 2 through Collect and Epistle
BWV 648 text related to Advent 3 through Collect and Introit
BWV 649 text related to Advent 4 through Collect and Gradual
BWV 650 Christmas: incarnation, coming down to earth
It is hard to believe that such plans were ‘of no interest to the organist of
the time’ (Wolff 1991 p. 344), although links between music and texts do
remain intangible. As much the point, perhaps, is that these texts can be
understood personally: the life of any believer searching for Grace has an
Advent and an Evening, the Church’s seasons are themselves analogous. A
personal ‘cycle of faith’ could explain why the title of the last differs from its
cantata version. Or if Advent is taken literally, the work presages the Canonic
Variations, which are based on the Christmas hymn.
Few buyers knew the original cantatas or the implications of their text,
nor, to judge by their changes, was the order obvious to copyists. But this
is no evidence against the idea of (i) a symmetrical cycle in (ii) a particular
key-sequence.
Purpose
Perhaps the set was made for (not by?) W. F. Bach on his appointment to Halle
in April, 1746: tuneful, approachable settings matching other volumes partly
connected with him (Orgelbüchlein, Sonatas, Clavierübung, BWV 541). It
would be a strange irony if both Ob and Schübler originated for the Halle
Liebfrauenkirche, where Friedemann was to perform some of his father’s
cantatas.
Had No. 5 been in C major, the Schübler Chorales would have the same
keys as the Six Sonatas; already, like them, they outline a triad (E) and
consecutive minors (C, D, E) – odd, if the set was a merely diverse collection
of works in the same genre. Although the last two settings are particularly
demanding, giving each hand in turn a difficult and unmodified string
obbligato line, there is no rounded survey of organ arts: no attempt is made
to convey the dynamic variety implicit in the cantata versions, and even
the original echo effects and f /p changes are missing. Two manuals have to
be avoided for one chorale (BWV 647) because the unaltered cantata parts
make them impractical; the need for one hand to play on two manuals in
BWV 648 (b. 13 etc.) appears casual, not further developed; and several of
them, when played on the organ, seem to need a slower tempo than when
sung.
Perhaps, therefore, the publication was a hasty or delegated project cater-
ing for a taste in more popular organ music than could be satisfied by
Clavierübung III or the Canonic Variations. In his Sonatinen of c. 1744
dedicated to J. S. Bach, G. A. Sorge had spoken of ‘something to please’
music-lovers, and he was to attempt this later in his own simplistic chorales
(24 Vorspiele, 1754). But the Schüblers were not simplistic and, one imagines,
barely more popular, being technically too demanding for most organists
to use in services, however appropriate to the church year. But note: though
difficult, they are as geared to music’s practice as the Canonic Variations
are to its theory – a distinction made in the Obituary by Lorenz Mizler,
whose Society Bach and Sorge had joined in 1747 as fourteenth and fifteenth
members.
The ‘registrations’ are not, like Kauffmann’s in Harmonische Seelen-
lust (Leipzig, 1733), stop-selections for colour but are, in the way of the
Orgelbüchlein, aids for interpreting the score and octave pitches. Forkel
seems to have understood the indications literally, describing BWV 646
as showing how Bach ‘departed from the customary manner’ (‘von der
gewöhnlichen Art abging’, 1802 p. 51) – presumably pedal 4 cantus firmus
was rare by 1802, as indeed it was by 1750.
Musical style
The later eighteenth century did find things to admire in the Schübler (Dok
III pp. 313, 441). In particular, Nos. 1, 5 and 6 have a newness of idiom
unique in organ chorales until younger organists attempted it (J. L. Krebs,
Doles, Tag, Homilius). Its chief element is a melodious counterpoint, not
imitative, without Italian formulae, genuinely combining two themes rather
than pretending to do so. The counterpoint upon a cantus firmus now
achieves independence; and organizing the obbligato melody into periods
gives it a logic of its own, returning between lines of the chorale and ending
with ritornello codas (da capo in BWV 649, 650) even when less melodious
(BWV 646) or like an ostinato (BWV 648). Important ritornello codas are
occasionally found elsewhere (e.g. BWV 660) but the Schüblers have no
pedal point of the kind common in organ music (e.g. BWV 684, 658 etc).
With such arias for organ, the composer was indicating a trend, one easily
adaptable to the long-winded galant language of younger composers and
already to be seen in Kauffmann’s Harmonische Seelenlust. With Kauffmann,
this particular ‘trend’ meant certain forms and melodies (e.g. ‘Man lobt
dich in der Stille’), or pale, updated versions of cantus-firmus settings, with
awkward pedal-lines that look like basso continuo parts – some of which
are to be played by lh, with pedal playing cantus firmus. Kauffmann’s book
already included six chorale-settings for solo oboe and organ, and many of
the pieces throughout could be transcriptions, like Schübler.
Nor is the ‘Schübler style’ totally removed from earlier music: the
bicinium ‘Allein Gott’ BWV 711 points to BWV 649, though of course
is less richly worked, and other examples of the fully fledged counter-theme
appear in Clavierübung III (BWV 678, 684). The chorales of Kauffmann
and J. L. Krebs scored for organ and a solo wind instrument take the style
to greater lengths and (with Krebs) a more modern idiom. Earlier chorale-
settings developing long counter-themes, whether or not derived, were made
by composers familiar with Italian string music, such as Böhm (‘Freu dich
sehr’, Var. 12), Walther (‘Schmücke dich’) and Bach (Cantata 4, verse 3),
and while the Schüblers’ counter-melody has changed in style, the principle
is similar. In its sheer singable quality, the unique melody of ‘Wachet auf!’
is a step beyond that of ‘Ach bleib bei uns’, which begins as a paraphrase.
The ‘Schübler style’, seen at its clearest in BWV 645, 649, and
650, comprises a texture, a melodic-contrapuntal idiom, and an aria
form. (A further example is BWV Anh.II 55.) An aria-like pattern of
prelude–interludes–postlude is unusual in earlier organ music and belongs
more to cantatas, in which the instruments’ melody becomes self-contained.
Organ chorales of this kind make the cantus firmus even more prominent,
and the crucial postlude, though only four bars long in BWV 645, 646, 647
and 648, rounds off a movement in which the plain cantus has been quite
distinct. The Canonic Variations also use only plain cantus firmus, but each
movement closes with a pedal point on the last note of the melody, held to
the end. In this respect alone, therefore, the six Schübler Chorales provide a
complement to the five Canonic Variations.
Other potential ‘Schübler Chorales’?

Although the pedal line of BWV 645, the left hand of BWV 649 and the
distribution of hands in BWV 647 are not fully characteristic of genuine
organ music of Bach, it could be that other suitable arias in the cantatas
would have given severer problems to the transcriber, whoever he was. The
choice of which movements to transcribe was limited, quite apart from
questions of text.
In addition to the three trio movements (Cantatas 6, 137 and 140) only
seven other surviving cantatas have movements in a suitable form, i.e. a vocal
cantus firmus and an instrumental obbligato melody, above a basso continuo
(4, 95, 113, 143, 166, 180 and 199). These movements are disqualified on
other grounds, however. BWV 4 and 199 are too early; the arias in BWV 95
and 180 are part of a longer movement, BWV 143 would have a compass
above c , while in BWV 166 neither cantus firmus (to g ) nor continuo
is suitable for pedal. The aria in BWV 113 would be suitable but is not
melodious in the preferred way.
The four-part BWV 647 and 648 are transcribed from a duet with basso
continuo and instrumental chorale melody. Only three further cantata
movements of this kind are known – in Nos. 163, 172, 185 – and all are
pre-Leipzig. Such arguments cannot prove that the composer had no choice
323 BWV 645
as to what he transcribed, or even that he resorted to including an original

composition (?BWV 646) because he had no suitable cantata movement.
But those that he did transcribe suggest that the requirements – a mature
Leipzig aria with cantus firmus, of suitable compass untransposed, with suit-
able figuration and spacing – limited the choice of both trios and quartets.
Had the transcriber been not J. S. Bach but someone else who felt obliged
to leave the key and spacing unaltered, that choice would indeed have been
limited and would help explain why certain cantata movements considered
suitable by some later writers (Dürr 1988 p. 59) were not used. That the
corpus of extant cantatas, therefore, offers little material for organ trans-
criptions comparable to the Schübler Chorales is not the least surprising
thing about them, and suggests a transcriber who knew the repertory very
well and was an intimate of the cantor’s library.
BWV 645 Wachet auf, ruft uns die Stimme (Schübler)

Three staves; headed ‘Wachet auf rufft uns die Stimme etc. à 2 Clav. et Pedal,
Canto Fermo in Tenore’; in the composer’s copy, ‘Dextra 8 Fuss’, ‘Sinistra 8
Fuss’, ‘Pedal 16 Fuss’. (Repeat written out in cantata score and parts.)
The TEXT of P. Nicolai’s hymn was published in 1599, later associated with
Twenty-Seventh Sunday after Trinity (Gojowy 1972), the close of the church
year.
Wachet auf! ruft uns die Stimme ‘Wake up’, there calls to us the voice
der Wächter sehr hoch auf der Zinne, of the watchmen high on the
battlements,
Wach auf, du Stadt Jerusalem! ‘Wake up, O city of Jerusalem!
Mitternacht heisst diese Stunde; The hour is midnight’;
sie rufen uns mit hellem Munde: they call to us in a clear voice,
Wo seid ihr klugen Jungfrauen? ‘Where are you, Wise Virgins?
Wohlauf, der Bräutigam kömmt, Arise, the bridegroom comes,
steht auf, die Lampen nehmt! get up, take your lamps!
Halleluja! Hallelujah!
Macht euch bereit zu der Hochzeit, Get ready for the wedding,
ihr müsset ihm entgegengehn! you must go out to meet him!’
v. 2 begins:
Zion hört die Wächter singen, Zion hears the watchmen singing,
das Herz tut ihr vor Freude springen . . . her heart does leap for joy . . .
The last verse is a hymn of praise.

324 BWV 645
The MELODY was published with the text but is probably older, its first
line resembling ‘O Lamm Gottes’ (Terry 1921 p. 315) and used only here:
Example 164.
Example 164
BWV 645 is transcribed from:

Cantata 140 ‘Wachet auf, ruft uns die Stimme’, 27th Sunday after Trinity
1731
Fourth (middle) of seven movements, ‘Zion hört die Wächter singen’,
called ‘Chorale’ in J. L. Krebs’s performing parts (NBA 1/27 KB p. 152)
Trio: obbligato melody (violin I + violin II + viola), cantus firmus (tenor),
basso continuo
The distribution of manuals and pedal in the composer’s MS rubric can

not be regarded as obligatory, even though the three-staff layout already
suggests it, as that of BWV 650 does not. A pedal part could be written on
an inner stave, as in BWV 769.iv, and the bass-line of BWV 645 is unlike
that of original organ pieces. In both BWV 645 and 650, therefore, the lh
can take either bass (16 ) or cantus (8 ). The composer’s added distribution
makes sense, of course, but organists may have welcomed the choice, given
them by the bare score, of where to place the melody. On pedal, it would
be down an octave and registered 4 , as in BWV 608, which too had no
indications.
Further characteristics of the transcription are that (a) ornaments in
the obbligato line are different (more generous but inconsistent); (b) the
chorale melody is more decorated; (c) the original figures in the basso part
(J. L. Krebs’s hand) are unrealized; and (d) the forte/piano signs are ignored,
both for echoes (bb. 15) and to indicate cantus entries. The extra grace-note
in b. 20 disguises the parallel unisons now exposed by empty harmony. The
325 BWV 645
new appoggiaturas in bb. 7, 8 recall those in Goldberg Variation No. 25 and

could belong to the composer or a transcriber replacing original trills and
modernizing other ornaments.
The achieving of a melody independent of the chorale is spectacular.
The right hand is developed to a half-close before the chorale melody
begins to combine with it, and its opening echo is even re-introduced
across the cantus firmus, as in Example 165. As the example shows, the
harmony is incomplete without continuo. So strong is the melody that
it suffers no sense of hiatus despite a halt in the interests of the cantus
firmus in bb. 54, 66 and despite so many tonic entries of the first or
second phrase. The obbligato melody has to be modified for the sake of the
chorale, and this process leads to a series of phrases which the ear accepts
as logical in their own terms (bb. 47–58). With the first section repeated,
the overall key-plan is tonic–tonic–relative/mediant–tonic, and this most
catchy of counter-melodies marshals the cantus into a reasoned ritornello
form.
Example 165
Presumably it is not only the chorale’s opening triad but the new melody
that resounds like the call of a street-watchman, complete with echoes (Keller
1948 p. 194). Perhaps its springy rhythms evoke the first two lines of v. 2,
an aria concerned with Zion’s enthusiastic reaction to the watchmen’s call
(Schmitz 1970 p. 65). Schweitzer heard in it the arrival of the bridegroom
(1905 p. 306), others an allemande with typically strong up- and down-beats
(Steglich 1962 p. 28).
With so dominating a melody one hardly notices how peculiarly discor-
dant the harmony often is, i.e. without the cantata’s continuo harmonies. In
Example 165, there are thirdless chords, échappées, accented passing-notes,
sevenths and unresolved appoggiaturas, all in quick succession, every beat
with something to strike the ear, hardly possible unless the melody is in
the middle, like a mediator. Meanwhile, the vocative, triadic hymn-tune is
harmonized conventionally. In fact, its phrases and their bass-line could be
extracted to make a satisfactory continuous chorale without interludes, as
if this best known of obbligato melodies were interrupting the hymn.
326 BWV 646
BWV 646 Wo soll ich fliehen hin (Schübler)

Three staves; headed ‘Wo soll ich fliehen hin etc. od: Auf meinen lieben Gott
etc. a 2 Clav. et Pedal’, also ‘1 Clav. 8 Fuss, 2 Clav. 16 Fuss, Ped. 4 Fuss’.
The TEXT of J. Heermann’s Busslied or penitential hymn was published in

1630, associated with various Sundays after Trinity in Leipzig (Stiller 1970
p. 231).
Wo soll ich fliehen hin, Whither should I flee,

weil ich beschweret bin since I am weighed down
mit viel und grossen Sünden? with sins many and great?
Wo soll ich Rettung finden? Where should I find salvation?
Wenn alle Welt herkäme If all the world were at my feet,
mein Angst sie nicht wegnähme. it would not take away my anxiety.
Ten further verses develop the theme of salvation for the sinner.
The MELODY, of secular origin, was first associated with the text ‘Auf meinen
lieben Gott’ from 1609 (Terry 1921 p. 344); both texts are listed but unset in
the Ob. The melody is as for BWV 694, used for various verses in Cantatas 5,
89, 136, 163 and 199, a penitential hymn for various Sundays before Advent
(Stiller 1970 p. 231).
The TEXT of ‘Auf meinen lieben Gott’ was published before 1603, becom-
ing associated with the Seventeenth and Twenty-First Sundays after Trinity
(Gojowy 1972).
Auf meinen lieben Gott in my dear God I trust

trau ich in Angst und Not; when in fear and misery;
der kann mich allzeit retten he can always save me
aus Trübsal, Angst und Nöten, from affliction, fear and need,
mein Unglück kann er wenden, he can turn away my misfortune,
steht alls in seinen Händen. all is in his hands.
The following five verses express faith and praise. The final verse later became
the last of ‘Wo soll ich fliehen hin’, and to pair these texts was something of
a tradition in Thuringia, to judge by J. M. Bach’s setting in the ‘Neumeister
Collection’.
A common view still is that BWV 646 comes from a lost cantata (KB
pp. 158–9): it is not known from any earlier MS of organ music, and one
can easily imagine a cantata scoring of basso continuo or bassoon for the
left hand, violin(s) or oboe da caccia for the right, and tenor for the cantus.
327 BWV 646–647
Whatever its relationship to the earlier chorale BWV 694, and however alike
their lines are, BWV 646 is surely more than simply ‘a much-altered new
version’ of it (Dürr 1956 p. 101). They need not even belong originally to
the same genre.
There is little problem in imagining BWV 646 as a cantata chorale.
A 16 registration makes the bass very like a basso continuo such as C. P. E.
Bach recommended for lh rather than feet (Versuch 1753 p. 245), when
accompanying cantatas. Also, the short-breathed cantus firmus is vocal
rather than instrumental, compared to BWV 651. And yet, the right-hand
part is not as expansive as some string obbligati, and the figuration in both
hands looks keyboard-like, more so than in the five other chorales – and
strikingly so, considering that this is the only one of uncertain origin.
It is, after all, similar to an earlier organ chorale, and one does not
need to conjecture that it was a solo organ piece inserted in a cantata
(BWV 188: Luedtke 1918 p. 68). All in all, arguments for and against tran-
scription are finely balanced and could be tipped either way by a new scrap
of evidence.
Like BWV 694, BWV 646 is a trio in which the left hand serves both as
bass line and as imitative second voice, the whole harmonized and motif-
based with an immense artistry that repays bar-by-bar examination. The
two hands do not cross parts, and the pedal has widely separated chorale
phrases. To liken such manual accompaniment to the Inventions (May 1986
p. 83) is fair, specifically the two-part in E minor. The main semiquaver motif
may be derived from the first line of the chorale melody (E E F G); it is
used inversus, and its segments create sequences. Often the inversus follows
immediately on the rectus in one or other hand, to create a running line. The
syncopated counter-rhythm is useful against the chorale’s crotchets, and at
times lh becomes bass-like. Except at the three cadences (bb. 6, 14, 24), the
motif is present in every half-bar of the movement, and yet the fleeing is not
as straightforward as in ‘Nun freut euch’ BWV 734. There is some unease
in it, borrowed from the words. Yet one early commentator heard in it ‘the
anxious seeking for peace’ (‘das ängstliche Suchen der Ruhe’: see AMZ 8,
1805, cols. 29–32).
The three cadences occur at similar points in BWV 694; but in length,
metre (4/4) and counterpoint, BWV 646 is tighter and more concentrated.
BWV 647 Wer nur den lieben Gott lässt walten (Schübler)
Three staves; rubric, ‘Pedal 4 Fuss’.
For TEXT and MELODY see BWV 642.

328 BWV 647
This is transcribed from:
Cantata 93, ‘Wer nur den lieben Gott 1ässt walten’, 5th Sunday after Trinity
1724
Fourth (middle) of seven movements, ‘Er kennt die rechten
Freudenstunden’
Quartet: cantus firmus (violin I and II, viola), vocal duet (soprano, alto),
basso continuo (parts copied c. 1732)
v. 4
Er kennt die rechten Freudenstunden, He knows the right time for joy,
er weiss wohl, wann es nützlich sei; he knows well when it is useful;
wenn er uns nur hat treu erfunden when he has found us to be true
und merket keine Heuchelei, and sees no hypocrisy, then
so kommt Gott, eh wirs uns versehn, God comes, before we are aware of it,
und lässet uns viel Guts geschehn. and lets great good befall us.
Since strings are not obvious for such a c.f . (an octave higher in the cantata,
hence Pedal 4 here), is the aria already an arrangement? Probably because
the bass figures are not realized in the transcription, the pedal point of
bb. 23–4 is avoided and the note shortened.
The cantata scoring implies two manuals, but the spacing only one: the
parts bump into each other and no attempt has been made to make them fit
the hands better except when the alto is slightly changed from the parts in
BWV 93 (bb. 39, 46), and again in b. 35 in the ‘composer’s corrected copy’.
Many moments are clumsy on keyboard (bb. 14, 19, 30 etc.), and much of
the texture is unlike traditional organ music, closer to the newer ways of a
Kauffmann (see bb. 10, 36).
The two fugally treated obbligato subjects are derived from the chorale-
lines, the first from line 1 but continuing to accompany line 2, the sec-
ond from line 3 but continuing to accompany line 4. The sophisticated
paraphrase technique is as typical of organ chorales as of cantata-arias-
with-chorale, if not more so: see Example 166. (The first was already noted by
F. W. Marpurg in 1753: Dok III p. 42.) The bass line manages to accompany
the two subjects with very similar material, and also incorporates thematic
Example 166
329 BWV 647–648
entries (bb. 8, 14, 45), contributing to the unity of the whole even though
the first subject reappears in the final bars (cf. BWV 695).
Considered as a ritornello section, the opening bars pervade the chorale,
moving continuously and not always coming to a cadence (e.g. before the
first c.f . entry). They are melodious in a way not familiar in organ music,
though precisely how is hard to say: one would think such lines more like
woodwind parts than vocal, and in any case the restless quavers are not
keyboard-like. The little dactyl figures seem a response to the first line of
the text concerned (‘right time for joy’), the aria’s serious Affekt a response
to the reticent anticipation of Advent.
BWV 648 Meine Seele erhebt den Herren (Schübler)

Three staves; headed ‘Meine Seele erhebt den Herren etc. a 2 Clav. et Pedal’;
‘sinistra’, ‘dextra forte’ and ‘Pedale’ added in the composer’s copy.
The TEXT is the German Magnificat (Luke 1: 46–55), used as the chief
hymn for Mariae Heimsuchung (Visitation) and sung after the sermon in
the regular Vespers, following a ‘praeambulo auf der Orgel’ (Stiller 1970
pp. 81, 22).
Meine Seele erhebt den Herren . . . My soul magnifies the Lord . . .

Er denket der Barmherzigkeit und He remembers his mercy and
hilft seinem Diener Israel auf. helps his servant Israel.
This is the only canticle to keep intact its original Gregorian MELODY: the
tonus peregrinus simplified in the harmonization BWV 324 (Example 167).
Example 167
Organ-playing at Vespers prompted many settings, especially alternatim

versets. Listed in the Orgelbüchlein, this melody is used in BWV 323, 324
and the Magnificat BWV 243/243a, called ‘Magnificat in the 9th mode’ in
Scheidt’s Tabulatura nova, 1624.
BWV 648 is transcribed from the full score (?) of:

Cantata 10 ‘Meine Seele erhebt den Herren’, Visitation 1724
Fifth of seven movements, Duetto ‘Er denket der Barmherzigkeit’
Quartet: cantus firmus (oboe I, oboe II, trumpet), vocal duet (alto, tenor),
basso continuo
330 BWV 648
The cantata layout suggests two manuals, as do the composer’s annotations;

but the left hand alone cannot play all of the middle staff, and the right hand
has to play on both manuals in b. 13 and perhaps 22 and 24 – only to reach
the notes, not for any truly idiomatic purpose. The rubric ‘Pedale’ for the
third staff suggests that without it, one would assume its bass-line to be on
manual 16 and c.f . on pedal 4 . But then, the right hand could not play the
vocal duet as is; and pedal would need to take the third staff. A minor point
is that the score’s slur in b. 2 appears as a tie in the print; but all this suggests
a literal transcription, done inauthoritatively, even inexpertly.
Though short, the movement has an intricate and unusual form:
A 1–5 a pedal-theme framework (slurred in the composer’s

copy, for heel-and-toe pedalling?) from which is derived:
B 5–9 inner framework of fugal imitation between inner parts
C 9–13 derived imitation à 3; two phrases of the melody
D 14–21 rising sequence derived, very new; F minor to A minor
C 22–8
B 27–31
A 31–5
The pedal theme, though paraphrasing a descending chromatic fourth, is

constantly modified and is no ostinato. (Originally in the cantata, this phrase
was less patently a chromatic fourth: see Marshall 1989 p. 93.) Such simple
symmetry is unusual, as, for an organ-chorale, are the silence in the inner
parts of bb. 9–10 and the barely idiomatic pedal.
The chromatic language and the appoggiaturas are generally associated
with texts concerning supplication or mercy, as in the aria ‘Achzen und
erbärmlich Weinen’ in Cantata 13, 1726. Other appoggiaturas, not chro-
matic but also in thirds, are used against the same melody for the same verse
in the choral Magnificat BWV 243. Details such as the unexpected change
to the minor in b. 13 are not uncommon (e.g. B Prelude WTC2, last four
bars) and need not be owed to older composers. It is also possible that
B A C H is to be heard in the course of the movement, e.g. in the tenor line
at the middle, bb. 17–20.
The degree to which the bass melody is constantly modified yet without
losing its melodic character is typical of the Schübler Chorales, as the skill
with which it harmonizes the two cantus phrases is of the Leipzig cantatas.
In both respects – development of motif, rich harmonic support – this bass-
line is inconceivable in the work of any other composer. One might think
that the original duet’s strong personality, concentrated, concise (two-bar
phrases!) and obviously complex, is rather less close to cantata arias than to
certain organ-genres.
331 BWV 649
BWV 649 Ach bleib bei uns, Herr Jesu Christ (Schübler)
Three staves; headed ‘Ach bleib bey uns Herr Jesu Christ’. Cantata 6 parts
headed ‘Allegro’ (Autograph MS) and ‘Allegro assai’ (late Autograph?).
The first verse of the TEXT is an early version of Melanchthon’s ‘Vespera iam
venit’ (1551), concerning the scene on the road to Emmaus (Luke 24: 29).
Verses 2–9 by N. Selnecker 1572 (see BC I p. 247) pray for Jesus’s help against
all dangers. It was used during the Reformation Jubilee in 1730 (Stiller 1970
p. 226).
Ach bleib bei uns, Herr Jesu Christ, Ah, stay with us, Lord Jesu Christ,
weil es nun Abend worden ist; because it is now become evening;
dein göttlich Wort, das helle Licht, your divine word – the bright light –
lass ja bei uns auslöschen nicht. may it not be extinguished in us.
v. 2
In dieser schwern betrübten Zeit At this sorely troubled time
verleih uns, Herr, Beständigkeit, grant us, Lord, steadfastness,
dass wir dein Wort und Sakrament that we your word and sacrament
behalten rein bis an das End. keep pure to the end
The MELODY is known in several versions, e.g. as alto to Calvisius’s ‘Danket

den Herrn’, 1594 (Terry 1921 p. 85): Example 168. Apart from a lost jubilee
cantata, the melody appeared only in Cantata 6 and the harmonizations
BWV 253, 414.
Example 168
BWV 649 is transcribed from:

Cantata 6 ‘Bleib bei uns, denn es will Abend werden’, 2nd Day of Easter
1725
Third of six movements, ‘Ach bleib bei uns, Herr Jesu Christ’, vv. 1 and 2;
called ‘Choral’ in continuo parts (Autograph?)
Trio: obbligato (violoncello piccolo), cantus firmus (soprano), basso
continuo
The transcription is shorter. Two verses in BWV 6 produce a shape of A

(ritornello or introduction), chorale v. 1, A (repeated), chorale v. 2, and A2
332 BWV 649–650
(first five bars replaced by one new linking bar). This is now simplified to
A–chorale–A, ignoring a potentially different Affekt for the aria’s second
verse (see above). As in BWV 648, the bass and obbligato lines would be
suitable for manuals and the c.f . for pedal, in the unlikely event of a 2 reed
being available. Untransposed, the obbligato melody is low for right hand
and suits the left; and as the bass is sufficiently continuo-like to suit pedal,
the cantata’s layout is convenient for the transcriber.
The obbligato melody’s character is unusual, doubtless a result of para-
phrasing the chorale melody for an agile string instrument: see Example 169.
However, its length not only gives it the weight of a full ritornello theme
(somewhat similar in form to BWV 645, 646 and 650) but allows for inge-
nious adaptation whenever the melody needs imaginative harmonization,
as in bb. 21–45. Each of the Schübler trios has an obbligato which combines
with the cantus either intact or modified, and in each, the melody reaches
clear cadences before two or more chorale entries. But the treatment varies,
and BWV 649 is unusually continuous – and is even more so in the cantata
version, with its two verses.
The length of the melody is alone enough to distinguish it from the usual
paraphrase-chorale, though in this respect BWV 660 is comparable. Four
conspicuously different motifs (see Example 169), together with the semi-
quaver patterns, ensure unity even when the melody is in fact much mod-
ified. Perhaps the length of the cantata version of the movement required
the Allegro heading, as too would the cello piccolo. On organ, part of the
unusual feel and difficulty must be due to its key of B major, otherwise rare
in the organ music, but there is also a strangely different sense of melody.
Not only was the cantata version presumably faster, but the melody has a
lightness and deft, string-like quality that one would not mistake for original
organ music.
Example 169
BWV 650 Kommst du nun, Jesu, vom Himmel

herunter (Schübler)
Three staves; headed ‘Kommst du nun Jesu vom Himmel herunter etc’;
cantus firmus on middle staff (beginning at g ), bass on lowest; in the
composer’s copy, ‘Dextra’ (top staff), ‘Sinistra’ (lowest), ‘Pedal 4 Fuss und
eine 8tav tiefer’ (middle, b. 13).
333 BWV 650
The TEXT of C. F. Nachtenhöfer’s hymn was published in 1667; Pietist

hymnbooks associated it with the Nativity and the Incarnation (Freyling-
hausen 1741).
Kommst du nun, Jesu, vom Are you coming now, Jesu, from
Himmel herunter auf Erden? Heaven down to earth?
Soll nun der Himmel und Erde Will now Heaven and earth
vereiniget werden? be united?
Ewiger Gott, Eternal God,
kann dich mein Jammer und Not can my misery and need
bringen zu Menschen Geberden? bring you to take human form?
The following four verses describe the need for the Incarnation.
The TEXT of J. Neander’s hymn ‘Lobe den Herren, den mächtigen König’
appeared in 1680; each of five verses begins ‘Lobe den Herren’, as in Cantata
137.
v. 2
Lobe den Herren, der alles so Praise to the Lord, who so
herrlich regieret, gloriously reigns over all,
der dich auf Adelers Fittichen who bears you safely on eagle’s wings,
sicher geführet,
der dich erhält who preserves you
wie es dir selber gefällt; as you yourself want;
Hast du nicht dieses verspüret? have you not felt a desire for this?
The MELODY appeared in 1665 (to the text ‘Hast du denn, Jesu, dein
Angesicht’): Example 170. Used in Cantata 57 (1725) to v. 6 of ‘Hast du
denn’, listed in the Ob, and with the text ‘Lobe den Herren’ in Cantatas 120a
(wedding, 1729?) and 137.
Example 170
The title of BWV 650 – which does not appear in Cantata 137 or anywhere
else in BWV – was perhaps chosen to conform to some overall Advent plan
(see p. 319 above), by the composer or someone else. The movement is
transcribed from:
334 BWV 650
Cantata 137 ‘Lobe den Herren, den mächtigen König der Ehren’, 12th
Sunday after Trinity 1725
Second of five movements, ‘Lobe den Herren, der alles so herrlich regieret’
Trio: obbligato melody (violin), cantus firmus (alto), basso continuo
The suggestion that the movement originated as an organ-chorale before

1725 (Grüss 1985 p. 144) cannot be substantiated, and the obbligato is
violinistic.
The ‘composer’s corrected copy’ and the original parts of Cantata 137
show variant readings in the obbligato melody (b. 2 is the same as bb. 15 and
25 in BWV 137 and in the original print of BWV 650) and in the ornaments
(c.f . trills in the corrected copy). Despite the print’s missing trills, the c.f . is
still decorated enough to qualify as ‘the composer’s only embellished cantus
firmus for pedal’ – if it really is a pedal part. The figuration of b. 2 is a crux
and differs between cantata and organ chorale, as printed, as ‘corrected’,
and as copied by J. C. Oley (KB pp. 132, 136, 149, 172). The literalness of
bb. 6, 8, 46 and elsewhere makes two manuals desirable. But this literalness
and these musical details make it unlikely that the composer had much if
anything to do with the transcription.
The print notates the movement in open score, with c.f . in the middle, as
in the original cantata. On the analogy of BWV 646, pedal is best, since the
continuo bass line is unidiomatic and reaches to e , higher than otherwise
required; but BWV 646 does not have its chorale on an inner stave, nor is
its bass line simply continuo, despite being 16 . (Three-staff notation always
puts the pedal on the lowest staff, including the printed Canonic Variation
BWV 769.v.) The closest parallel is BWV 645, and both have a sonata-like
melody and shape. Accordingly, symmetry is best served if the two have
their lines played in the same way, whichever way that is: cantus on pedal in
either both of them or in neither.
The chorale’s phrases have become virtually subservient to a short ri-
tornello sonata movement for solo violin, in whose theme may just be
made out the opening triad of the chorale melody. But if there is a para-
phrase, it is remoter than usual, enough to be an independent countersub-
ject (b. 14). As in ‘Wachet auf’, only perhaps less so, the different phrases
of this new melody occur in various orders, could occur in others, and
are patched together to form a seamless violinistic melody, as if this bar
or that could be moved around. Quite how these dancing figures, which
take over in b. 9 and bring a new pattern into organ music, relate to the
chorale’s new title (or vice versa) is unclear. The figuration is obviously
violinistic, but for organ to copy the violin’s articulation as in BWV 137
(as suggested in the Schmidt-Mannheim edition, 1965) is not obviously
appropriate.
335 BWV 650
It is usually assumed that the cantus firmus is to be read in 9/8 not 3/4
(Klotz 1969a). But unlike contrapuntal lines in movements elsewhere with
rhythmic ambiguities of this kind – C minor Praeludium BWV 546, Sonata
No. 4 finale, Gavotte of E minor Harpsichord Partita – a cantus firmus
is a discrete, pre-existing solo melody, with its own independent rhythm.
Another aria melody in 3/4 time against jig-like violin obbligato lines in 9/8
can be heard in Cantata 7.iv (1724).
Chorales formerly called ‘The Eighteen’
BWV 651–668
A section of the autograph MS P 271; no title.
Contents
There are either fewer or more than eighteen chorales in P 271:
pp. 1–56 the Six Sonatas, a self-contained MS; last page blank
p. 57 beginning of a further MS, ‘title-page’ left blank
pp. 58–99 fifteen chorales BWV 651–665, autograph; then
BWV 666–667, copied by J. C. Altnickol on blank
pages
pp. 100–6 Canonic Variations BWV 769a, autograph (begins verso
side of BWV 667, same fascicle; BWV 668 follows on at
end)
p. 106 BWV 668 (page ends at middle of b. 26), copied by
‘Anon Vr’
(pp. 107–8 missing?)
BWV 769a may have been copied while pp. 96–9 were still empty. Either
p. 99 was left for a new title-page, and pp. 96–8 or 96–7 were earmarked for
a further chorale; or BWV 769a was part of the same sequence, and all four
empty pages were to have been filled. The paper is as that for the preceding
Six Sonatas (c. 1727–31).
The present title on an extra page, old but not autograph, begins
‘Achtzehn . . .’, altered from ‘Siebzehn’. Though the MS dates from the
Leipzig period, the title ‘Leipzig Chorales’ is not much more appropriate
than ‘The Eighteen’ or ‘Seventeen’.
Sources
For chorales fair-copied in P 271, at least two versions exist and as many
as four. Secondary sources suggest that P 271 contains both details of the
(draft?) copies from which it was prepared and some revisions. The group
[336] was not copied as such, in any version, before Kittel and Kirnberger followed
337 ‘The Eighteen’
P 271, nor is the Trinity sequence (BWV 662–664) preserved elsewhere, or

any plan obviously followed.
While Walther’s copies of BWV 665a and 666a date perhaps as early as
c. 1708, and Krebs’s manuscripts contain all seventeen, there is no sign of
a grouping before P 271 or proof that all ‘early versions’ date from (were
copied during) the Weimar years. It is true that Walther was working with
similar material, but such work did not stop in 1717. His ‘Allein Gott’ Vers
4 resembles BWV 656 in part, and perhaps his publication of it in 1738
prompted Bach to make a publishable collection.
P 271 is a fair copy with alterations still being made in all but BWV 657,
661, 662 and 664. In particular, BWV 651 may have been revised and enlarged
at the point of copying (Stinson 2001 pp. 40ff.), though not necessarily
entirely in P 271.
Date of originals
Despite no evidence that such a group of chorales was conceived in Weimar,
their difference from Ob settings makes them complementary to it. Sources
for BWV 667a and 667b have been interpreted as showing chorales under-
going expansion already in Weimar, and if Bach was responding to chorales
published by Pachelbel in 1693, he was aiming at a yet greater scale. Some
of Pachelbel’s, such as ‘Wir glauben’, are quite extensive and can ‘be used for
preluding during the service’ (‘bey währendem Gottes Dienst zum praeam-
bulieren gebraucht werden können’). The long, meditative organ-chorale –
if not often as long as BWV 652a – was no stranger in Thuringia.
Even if copies of various chorales by Walther and Krebs belong to Weimar
1710–14 (Zietz 1969 p. 137), when most were originally composed is less
clear – mostly before the Ob, to judge by the music itself, its less con-
sistent part-writing, less extensive use of canon and less tense harmony.
‘O Lamm Gottes’ BWV 656 is surely earlier than BWV 618, just as the
three-verse BWV 656, an updated version of Pachelbel’s models, is earlier
than BWV 627. From comparing them with other music of Bach, some
such dating as the following has been proposed (Zehnder 1995 and Stinson
2001):
1707–8 BWV 665a, 666a, 652a, 656a

1709–17 667a
1711–13 662a, 659a
1712–14 654a, 653a, 655a, 664a, 663a
1714 657a, 651a, 661a, 658a
1715–16 660a
A ‘Böhmian’ fugal treatment of the cantus suggests BWV 652a, 665a and
666a to be early Weimar works; incidental similarities to Weimar cantatas
suggest a later date for BWV 657a (see ‘Jesu, dein Passion’ from BWV 182)
and BWV 655a; and motifs in the trios BWV 655a and 664a can be found
in Italian concertos circulating by then.
However, cross-influences from genre to genre in J. S. Bach are seldom
simple, and a cantata could treat chorales in a manner worked out long ago
in organ preludes, or (as in the Orgelbüchlein) draw on past and present
techniques. Works could circulate in more versions than are known, living
organisms not always enlarged. There may be anachronism in the idea of
dating such music.
Date of revision
The first thirteen chorales entered in P 271 have been dated to c. 1739–42,
the next two to 1746/7, and the Canonic Variations to c. 1747 – August
1748 (Kobayashi 1988 pp. 56ff.). Probably the original fifteen were copied
in order, but the date of Altnickol’s pair is uncertain: mid-1740s or after
Bach’s death (Wollny 1999 p. xvii) or early 1750s. The composer may have
intended a sixteenth chorale on a majestic scale to round off the collection,
occupying three or, leaving the verso blank, two pages; this sixteenth could
have been BWV 667 as we have it or a new composition never made (less
likely). Just as there is no known authority for Altnickol’s additions, so there
is not for his choice: BWV 735 would be as plausible an addition as BWV
666. Perhaps the copyist of the eighteenth, BWV 668, was Altnickol’s wife
Elisabeth née Bach, on whose authority is unknown (Kobayashi 2000 p. 1).
Since in early 1749, after completing his pages in P 271, Bach had some
problem with writing, two equal possibilities are that a group of fifteen was
already then complete and not intended to be taken any farther; and that
on the contrary, a bigger group was being planned, with BWV 666 and 667
or others, even to end with the Canonic Variations. The possibility that the
group was a true collection made for publication (Dürr 1984 plate 77) is not
much supported by P 271 itself: the fifteen are not clearly enough written to
serve for a facsimile etching like Clavierübung III, nor are there any of the
articulation signs often found in certain late prints.
Nature of ‘revision’
Some of the ‘Leipzig versions’ are longer by whole sections (BWV 651)
or by several bars (BWV 652, 653, 656); others were less systematically or
completely improved in individual motifs, ornaments, rhythm and part-

writing. Two had note-values doubled (BWV 656 v. 3, 661), and other
‘revisions’ could be merely notational: the sharper rhythms and ornaments
of BWV 653 may reflect only what was expected for BWV 653a. No such
differences need suggest conscious revision. Even the relationship between
BWV 651 and 651a is only assumed from the sources as they stand, and
it cannot be shown that here was one set of Weimar chorales ‘which Bach
arranged anew [only] in Leipzig and put together as a collection of seventeen’
(Zietz 1969 p. 10). Variants for BWV 655 suggest that it circulated in more
than two forms, and it would be surprising if this was the only one to
do so.
Shape of the collection

Are the chorales in P 271 (fifteen, seventeen or eighteen) an ordered collec-
tion or a miscellany?
In favour of the chorales’ being merely a miscellany of long, partly revised
settings is the absence of any concrete evidence otherwise, leaving one to
make inferences from such details as that chorales often follow on the same
page. But the sequence BWV 652–654 does not have the ring of a carefully
planned variety. Nor, although there is a common style – chorale-settings
in four parts, each line separated by interludes, all on a big scale – is it
consistent. In favour of there being a grand plan which only the composer’s
worsening health prevented from being completed are that (i) the first and
last pieces address the Holy Ghost; (ii) the first, last and a middle chorale
(BWV 661) are marked ‘organo pleno’ only in the Leipzig version, while
BWV 665a may once have been but was no longer; and (iii) two other
major collections (Ob, Clavierübung III) have a plan. Had a group of sixteen
been intended, one could speculate on two groups of eight, the second
beginning with the Advent settings – an otherwise strange position for them.
If BWV 661 were to have been central, four more settings would be required,
five if BWV 666 is there without authority.
The series does not follow the church year, the liturgy or a hymnological
agenda such as the Luther texts or the Gregorian emphasis of Clavierübung
III, shortly after the publication of which the composer began to work on
these revisions, i.e. in P 271. There is no ‘cycle’ or clear association with
Communion (as implied by Meyer 1987 p. 41). Yet some extra-musical
patterning can be discerned: within the Whit framework, texts evoke
Christian orthodoxy and the Central Mysteries of Communion, the Trinity
and Incarnation, as distinct from Catechism and Kyrie in Clavierübung III.
There are also several conspicuous threes:
three Communion hymns (BWV 654, 665, 666) and three Agnus dei
settings
three texts each to God the Father (BWV 657, 658, 662–664), the Son
as mediator (BWV 655, 656, 659–661), the Holy Ghost (three Whit
hymns)
three Trinity hymns and three Advent hymns
three sarabandes and three trios
Other texts are related to the Trinity and/or liturgical practice (psalms in
BWV 653 and 658, sermons in BWV 655 and 657) and textual allusions
can be found – the ‘Hallelujah’ closing BWV 652, the ornaments in ‘Adorn
yourself’, BWV 654.
The three sarabandes are perhaps the most surprising, whether grouped
by design or chance or on impulse. Many of their features had appeared
in less mature works such as Variation No. 10 of BWV 768 or settings by
other composers (Kirchhoff ’s ‘Ach Herr, mich armen Sünder’). Perhaps
Mattheson’s recent remarks on using dance-forms to set chorales also en-
couraged the group (on sarabandes, see Mattheson 1739 p. 162). In the case
of ‘Allein Gott’, each has a cadenza-like passage:
662 free right-hand, returning to the chord with which it began

663 free left-hand, embellishing notes from its starting chord
(C B G E)
664 an extra voice in a trio, above an exceptionally long pedal-point.
Sheer length and intricate melodic paraphrase distinguish the collection.

As is not so for the Ob or Clavierübung III, several candidates amongst the
miscellaneous chorales could have served in the collection: BWV 694 or 734,
BWV 735 or 712, which resemble BWV 665 and 666 in form. But note: the
last chorale to be fair-copied by the composer (BWV 665) has a particularly
conclusive ending, as if it were the end.
Purpose
While longer chorale-settings of this kind (or any settings of any kind) could
serve at Communion or other moments of prayer or meditation, no such
purpose explains the group’s musical variety and technical scope. There
is no evidence that it originated ‘out of Bach’s need for liturgical organ
music’ (Stinson 2001 p. 60) in Weimar and Leipzig, or, as is more likely,
Friedemann’s in Halle.
That there are at least six major examples of lines derived from the cantus
(BWV 651, 655, 656, 657, 664, 665) and six of decorated melody (BWV 653,
341 BWV 651
654, 659, 660, 662, 663) suggests a conscious survey of chorale technique,
indeed of the arts of paraphrase. The counterpoint as manual lines move
in suspension, while the pedal moves by step, evolves from simple sequence
(BWV 667 b. 13, BWV 665 bb. 39–40) to genuine counterpoint (BWV 655
b. 68) and on to complex imitation (BWV 658 b. 25). Commentators have
sometimes seen weaknesses in the collection, especially in the old-fashioned
‘objective’ settings BWV 657 and 666 (and BWV 652 in Meyer’s opinion,
1972). But other chorales seem to shine above tradition, as BWV 655 and 664
do above Pachelbel, or the chromatics in BWV 665 above any contemporary
music. Acknowledgments of Böhm (BWV 659), Buxtehude (BWV 652, 653,
654) and Pachelbel (BWV 655, 657, 658, 666) complement the antecedent-
less Ob.
If after all P 271 was compiled or planned (i) as a collection, (ii) with
publication in mind, one can suppose its aim was to be more approachable
than the recent Clavierübung III, joining the increasing number of organ
publications, Lutheran or otherwise (Kauffmann, Walther, Fischer, Gottlieb
Muffat, Quehl, Vogler). Quehl’s variations in Versuch (Nuremberg 1734) also
began with a version of ‘Komm, Heiliger Geist’. For the possibility that the
Ob too was a response to recent publications, see p. 235.
BWV 651 Komm, Heiliger Geist, Herre Gott

(‘Leipzig Chorales’)
Copies by or via C. F. Penzel, J. C. Oley, J. P. Kirnberger and J. C. Kittel.
Two staves; headed in P 271 ‘J. J. Fantasia super . . .’ (‘Jesu juva’, ‘Jesus,
help’), ‘canto fermo in Pedal’ (added later to title?), ‘in organo pleno’.
The TEXT has three verses, a pre-Reformation translation of the Whit an-
tiphon ‘Veni sancte spiritus’ with two verses added by Luther (Erfurt 1524),
and becoming the chief hymn of the three days of Whitsuntide.
Komm, heiliger Geist, Herre Gott, Come Holy Ghost, Lord God,
erfüll mit deiner Gnaden Gut fill with the goodness of your grace
deiner Gläubigen Herz, Muth und the heart, spirit and mind of your
Sinn, believers,
dein brünstig Lieb entzünd in ihn’. kindle your ardent love in them.
O Herr, durch deines Lichtes Glanz O Lord, through your light’s brilliance
zu dem Glauben versammlet hast you have gathered to the faith
das Volk aus aller Welt Zungen. people of every tongue on earth.
Das sei dir, Herr, zu Lob gesungen. Let this be sung to your praise, Lord.
342 BWV 651
V. 2 asks for protection from false teaching, v. 3 for ardour to sustain the
faithful.
The MELODY was published with the text but is also older, related perhaps
to ‘Adeste, sancte spiritus’; later versions have different ‘Hallelujahs’, one as
in BWV 651 and 652 (otherwise like BWV 226, Example 171). Used in Whit
cantatas 59 (1723 or 1724), 175 (1725) and 172 (1714, adapted); listed in
the Ob.
Example 171
This is the only certain appearance of the title ‘Fantasia super . . .’ by

J. S. Bach (it is unverified for BWV 695, 713 and 735) and recalls its use in
Scheidt’s Tabulatura nova, for long fantasias based on a chorale as opposed
to other themes. A huge continuous fantasia, musically and dogmatically as
grand an opening as the Prelude to Clavierübung III, this setting is easy to
see as a response to Pentecost:
And suddenly there came a sound from heaven, as of a rushing mighty

wind, and it filled all the house where they were sitting. (Acts 2, 2)
With its rushing theme paraphrasing the chorale on two levels

(Example 172), its internal repetition (bb. 55–86 = 12–43) and new subject
(‘Hallelujah’, b. 89), this is a masterly unified piece, indeed inflaming the
hearts of the faithful as its lines spin out the opening theme and invent a
new melody in b. 25.
343 BWV 651
Example 172
Of course, the opening pleno pedal point reminds the player of the
Toccata in F, not least since (as often in F major in J. S. Bach) they both
move towards e in the opening lines. No other setting with cantus firmus in
pedale begins with something else in the pedal. The point d’orgue’s ‘domi-
nant answer’ in b. 13 is brief, functioning as the last note of a cantus phrase.
The strong tonic/dominant pull of this cantus needs skilful handling at
bb. 17ff., 34ff. and their repeats, and there is something of a harmonic tour
de force here. The pedal’s final bars after the c.f. is complete share a coda-like
quality with those ending BWV 655 and 733.
The ebullient, ecstatic semiquavers never cease for a moment (unlike
BWV 655), and run right through into the final chord. The nonstop tech-
nique is there in BWV 651a, but in the longer version it naturally creates
more of a ‘rushing wind’, comparable to the other Whit setting, BWV 667
(second verse). ‘Every tongue on earth’ might be participating in this un-
ceasing cascade of sound, but there is a not dissimilar effect in another
mature work in F major, Prelude No. 11 WTC2.
Episodes provide variety of key and an important appoggiatura theme
(b. 25), and the end of line 4 on an A (b. 44) gives new opportunities for
modulation. Signs that the new material in BWV 651 probably belongs
to the mature Leipzig years are the similarity between its chromatics and
Contrapunctus IV bb. 61–80 in the Art of Fugue, the simple sequence and
thinning of parts in bb. 56–7 (see middle passages in BWV 544 and 547),
the ingenious use of motifs (b above the same phrase in bb. 87–8 and 102–3,
typical of Clavierübung III), and the final build-up (c.f. plus Hallelujah plus
motif from b. 26. Compare the Canonic Variations).
Sources of BWV 651 and 651a support the modern theory that ‘Bach
always lengthened, never shortened’. And yet, if it is true that the greater
length of BWV 651 produces a model ‘in which the structure of the cantus
firmus and the length of the work are appropriately proportioned to one
another’ (Breig 1986c p. 118), did the appropriate proportion not occur to
the composer earlier?
344 BWV 651a–652
BWV 651a Komm, Heiliger Geist, Herre Gott

No autograph; copies by J. T. Krebs (P 802) and J. G. Walther.
Two staves; headed in P 802 ‘Fantasia super . . .’.
The forty-eight bars amount to less than half of the version in P 271:
BWV 651: BWV 651a

l–43 (1st 12 ) = 1–43 (1st 12 )
43 (2nd 12 )–54 —
55–88 (repeat) = 12–45 (repeat)
89–103 —
104–6 = 46–8
Somewhat less than half the cantus firmus appears, but because the melody
includes repeated figures, one could as equally imagine BWV 651a telescop-
ing 651 as BWV 651 extending 651a. The Whit Cantata 172.v (1714) also
shortens it, as BWV 664 and 715 shorten their melody ‘Allein Gott’.
On the last cantus firmus notes in the pedal being shorter in BWV 651a
than in 651, see also BWV 769.ii, where the MS has longer notes than the
print. Partly from comparing it with Cantata 172, Werner Breig has suggested
that BWV 651a was composed for Pentecost 1714 (1986c p. 109), but other
comparisons could imply a later date – the appoggiatura figure of b. 26 plus
semiquaver accompaniment appears in both the A major Prelude WTC1
(c. 1720) and E minor Partita, Toccata (c. 1725).
BWV 651 should not obscure the originality and value of BWV 651a,
even if BWV 651a could ever be shown to be a shortened version made
by Krebs or Walther. An opening pedal point which rises to begin a cantus
firmus; the stretto; two pairs of cantus phrases separated by modulating
episode; three- and four-part fugal counterpoint drawing on the motifs,
never compromised by the bass theme; the glowing realization of a text –
all this is an achievement unparalleled in the period, whatever its pedigree.
Is the reverse B A C H in the penultimate bar intended?
BWV 652 Komm, Heiliger Geist, Herre Gott

Copies: as BWV 651.
Three staves; headed in P 271 ‘alio modo à 2 Clav. et Ped.’

345 BWV 652
For TEXT and MELODY see BWV 651. Perhaps BWV 652 refers to v. 3:
Du heilige Brunst, süsser Trost, O holy ardour, sweet comfort,

nun hilf uns, fröhlich und getrost now help us to remain constantly
in deinen Dienst beständig bleiben, joyful and confident in your service,
die Trübsal uns nicht abtreiben. do not let afflictions drive us away.
So astonishingly different from BWV 651’s is this setting of the melody, the
longest of Bach’s organ chorales, that it must be responding to something
different in the text: now the Holy Ghost is the ‘sweet comfort’ of v. 3 rather
than the ‘brilliant light’ of v. 1, though it too has a Hallelujah!
Sarabande-like features have long been heard in it (Dietrich 1929
p. 63), which as in BWV 768, 653 and 654 reflect familiarity with dance-
types in chorale-settings. However, none of these is a sarabande simpliciter,
which is in binary form, without upbeat but with feminine cadence, a four-
bar phraseology, etc. Buxtehude’s ‘Komm, heiliger Geist’ BuxWV 200 and
Böhm’s ‘Ach wie nichtig’ Var. 8 already offered smaller-scaled models. A fur-
ther major influence might well be the French textures learnt from Grigny,
Boyvin or Du Mage: so there is or could be a cornet de récit in BWV 652 and
a tierce en taille in BWV 653, paired as such in P 271.
Each cantus line is treated at leisurely length, as follows:
Fore-imitation: a derived fugue subject in tenor, answered (against
quasi-countersubject) by alto, then pedal (twice in the case of line 1)
further tonic ‘answer’ in soprano, ornamented and partly like a c.f. (in the
tonic), partly not (in the same note-lengths)
a few cadential bars before next tenor theme; and finally a coda
Like BWV 651 the setting sustains a flowing line, but one now gentler,
endlessly spinning anew: only line 3 is repeated (as line 7: bb. 42–66 =
124–48). Countersubject quavers derived from the subject help create flow
and merge into fore-imitations; in the final exposition b. 171, the alto’s entry
disguises the true answer, which begins only in b. 175. The overall Affekt is
uniform, mesmeric like no other setting.
The final paragraph from b. 171 contains two sections: (i) exposition de-
rived from the ‘Hallelujah’ of the cantus, related to BWV 651 (Example 173);
and (ii) a coda, not obviously derived from anything in the chorale. Codas
such as (ii) were known in northern repertories, Bruhns (‘Nun komm’, in
P 802), Buxtehude (BuxWV 200, new scale patterns), and Reinken’s ‘Was
kann uns kommen’ (also in P 802, rh wanders on solo manual). See also
BWV 671. It is not impossible to trace in this coda an extravagantly para-
phrased version of line 2, including its cadence.
A problem for the player is how lively or sarabande-like is the pulse
of a work almost 200 bars long with 37 c.f. phrases (9 × 4. Was the extra
346 BWV 652–652a
Example 173
pedal phrase in b. 19 meant as the first of a series?). The setting is neither a

simple organ motet nor a simple ornamental type. Its dotted first or second
beat gives a lilting rhythm, its ornamental lines a more unified texture than
BWV 652a. A somewhat doctrinaire feel to the counterpoint should not
hide its frequent charm (e.g. bb. 27ff.), which is too easily threatened by
a sluggish tempo. Whether the sudden coda was marked by adding stops
(which is practical) or a freer tempo, or both, can only be guessed: its
repetitious use of two motifs is imaginative, insistent and final.
BWV 652a Komm, Heiliger Geist, Herre Gott

Copies: as BWV 651a.
Three staves; headed in P 802 ‘a 2 Clav. è Ped.’.
Because of the cadences to lines 2–4 and 6–8, BWV 652a is shorter by six
bars:
BWV 652a becomes BWV 652

39 39–40
63 64–5
87 89–90
119 122–3
142 146–7
164 169–70
This suggests that any cadences considered perfunctory came to be length-

ened, in particular the first two and the last two.
Although the more highly ornamented style of the soprano melodies
follows the tradition for decorated right-hand solos on a second man-
ual, in fact each of them is the last entry in a series of complete four-part
fugal expositions. The ornamented line also helps to lead naturally into the
coda. The sources are reliable in respect of ornamentation (KB p. 66) and
347 BWV 652a–653
suggest that the different approach to ornaments in BWV 652 is not merely
notational.
BWV 653 An Wasserflüssen Babylon (‘Leipzig Chorales’)

Copies: as for BWV 651.
Three staves; headed in P 271 ‘a 2 Clav. et Pedal’.
The TEXT is a five-verse translation of Ps. 136/137 (‘Super flumina

Babylonis’, for Vespers), published in 1525. Not a regular liturgical hymn,
‘the complaint of Zion’ (Freylinghausen 1741) could be used analogously.
An Wasserflüssen Babylon By the waters of Babylon

da sassen wir mit Schmerzen; we sat down in sorrow;
als wir gedachten an Zion, when we thought of Zion
da weinten wir von Herzen. we wept from our hearts.
Wir hingen auf mit schwerem Muth Sorrowfully we hung up
die Orgeln und die Harfen gut our organs and harps
an ihre Bäum’ der Weiden, on their trees of willow,
die drinnen sind in ihrem Land; which are in their country;
da mussten wir viel there we had to suffer much
Schmach und Schand shame and disgrace
täglich von ihnen leiden. daily at their hands.
The MELODY, also sung to P. Gerhardt’s Good Friday text ‘Ein Lämmlein
geht und trägt die Schuld’, was published with the text (Example 174),
harmonized in BWV 267 and listed in the Ob.
Example 174
BWV 653 may be connected with Bach’s visit to Hamburg in 1720 as told
in the Obituary:
den Choral: An Wasserflüssen Babylon, welchen unser Bach, auf Verlangen
der Anwesenden, aus dem Stegreife, sehr weitläuftig, fast eine halbe Stunde
lang, auf verschiedene Art, so wie es ehedem die braven unter den
348 BWV 653
Hamburgischen Organisten in den Sonnabends Vespern gewohnt gewesen

waren, ausführte . . . Reinken . . . vor langen Jahren diesen Choral selbst,
auf die obengemeldete Weise gesetzet hatte. (Dok III p. 84)
At the request of those present [in Hamburg], Bach performed the chorale
‘An Wasserflüssen Babylon’ extempore, very amply for almost half an hour,
in a variety of ways, just as formerly the better amongst the Hamburg
organists had been accustomed to play during the Saturday Vespers . . .
Reinken himself had set the chorale many years previously in the manner
described.
Walther too tells an anecdote about Reinken’s setting this chorale and send-
ing it to ‘a great musician in Amsterdam’ (1732 pp. 547–8). Whether Bach
knew this setting, whether the Obituary authors had merely Walther’s story
in mind, and whether they or Walther were referring to the long fantasia
by Reinken still extant, cannot be established: the provenance, source and
implications of the copy made by Bach’s son-in-law Altnickol (or a pupil of
his – see Wollny BJ 2002 p. 42) are unknown.
Perhaps a version of BWV 653 was played on the Hamburg visit, though
the Obituary refers rather to a set of variations or a disjointed, extem-
pore fantasia. Only conjecturally can one see it as an ‘elaborate homage’
to Reinken as last representative of the Hamburg–Lübeck school (Wolff
2000 p. 64), though the Obituary story might be. BWV 653 is less elab-
orate and reflects more Bach’s and Walther’s interests as they were in
c. 1714, and could later have found a place in Vespers recitals such as
had become more widespread by c. 1740. Nothing is quite certain in this
picture.
BWV 653 is a ritornello chorale conceived as follows:
decorated c.f. phrase by phrase en taille

introduced by two upper voices, each derived from the first two phrases
of the c.f.; each tenor phrase accompanied by one or other theme
pedal continuo bass, often derived from the first line (bb. 1–2, 4–5, 16–17,
32, 61–2, 77–8) or second (bb. 27, 50), or perhaps others
(e.g. line 6 in b. 39?)
Its elusive character depends on several things: a sarabande style without

upbeat; homogeneous parts derived from the melody, with little free writing
but stretto (b. 1) and combinations (b. 4 pedal = line 1, b. 5 soprano =
line 2); soprano melody with unusual, ostinato-like returns; a striding pedal
which, when not derivative, has two points d’orgue below an en taille melody;
and a consistently elegant melos and stately rhythm. In melody, quavers, key,
metre, rhythmic figures and thematic derivation, it is similar to BWV 652;
but the crotchet chords are more sarabande-like.
349 BWV 653–653a
Being so constantly derivative, the lines, not least the pedal’s, have some-
thing ‘litany-like’ about them (Breig 1986c p. 111). Deriving inner parts
from the cantus occupied the composer in his later years – see the first Kyrie
in Clavierübung III – just as did constantly reworking the first two lines in
the upper voices, as in BWV 675 and 682, where however there is no obvious
reference to line 1 at the end as there is here. The last seven bars present
line 1 in the outer parts (imitation at the octave) and contract it in the alto
(bb. 79–80) whilst hinting at line 2 in the upper pedal part (bb. 81–2), all
around a long pedal point en taille that closes with a little descending run
at the end. Each little quaver phrase seems related to every other.
Though a detail, the final little run in the cantus part may well be saluting
an earlier northern fantasia such as Reinken’s, even if otherwise BWV
653’s continuity could scarcely be more different from Reinken’s one known
setting. The languor that commentators have felt implied by v. 1 of the
hymn, meditative rather than plaintive (Meyer 1987 p. 42), is plausibly
suggested by the ostinato elements in soprano and bass, by a smoothly
moulded accompaniment (the repeat of bb. 1–14 overlaps at b. 12) and, after
similar tonic cadences (bb. 13, 25, 36), by a long stretch before the next and
last. The many technical clevernesses have been geared to produce smooth
but now less flowing lines, with a gentle, harmonic rhythm in the major, as
mesmerizing in its way as BWV 652.
BWV 653a An Wasserflüssen Babylon

Three staves; headed in P 802 ‘Vers 2 à 4 con 2 Clav. è simp. ped.’, ‘alio modo
à 4’, lh ‘forte’, rh ‘piano’.
The last nine bars of BWV 653 were only three in BWV 653a and include two
pedal points, some further references to the theme, and a richer five-part
close. But 653a did already contain a pedal point, a five-part close and a
chromatic penultimate phrase (all as in BWV 653b), as well as the tierce en
taille idea. Perhaps this last dated from about the time of Bach’s copy of
Grigny’s Livre. Whether the more sharply defined rhythms of the ‘Leipzig
version’ show how it was usually played, or are there to make the imitation
clearer, is uncertain.
Whatever the relationship between BWV 653a and 653b, this four-part
setting is the more conventionally spaced of the two and preserves the har-
monic substance of the five parts above a single pedal line that now allows
16 registration. On whether this pedal part is a judicious reduction of the
350 BWV 653a–653b
two parts in BWV 653b, see below. With one exception, it keeps below d :
the e at the close (apparently authentic) is not there in BWV 653 or even
in BWV 654 (in E major).
BWV 653b An Wasserflüssen Babylon

Copies: J. G. Walther (P 802) and Mempell–Preller (via Walther?).
Three staves; headed in P 802 ‘Vers 1 à 5 con 2 Clav. è doppio pedale’.
P 802 may be a ‘later period’ in Walther’s handwriting, with many alterations

and erasures (Zietz 1969 pp. 89, 141). Both sources contain both the five-
part and four-part versions, and it can only be conjectured why P 802 calls
them Vers 1 and Vers 2. Perhaps Walther assumed they were variations like
his own sets but longer? Extant sources of these and other chorales seemingly
revised in P 271 surely give only part of the picture of variants, variously
transmitted versions, revisions circulating. There is no evidence for Spitta’s
suggestion (I p. 606) that BWV 653b was sent to Reincken and/or that it had
been adapted from the written-down extemporization in Hamburg.
Although contrapuntal motifs are not so effortlessly handled by J. L.
Krebs in BWV 740 as here in BWV 653b, two groups of pieces (four on
‘Wir glauben’ ascribed to Krebs, the three on ‘An Wasserflüssen Babylon’
to Bach) could have been somehow linked, vestiges of interests shared by
the Weimar organists. Whether BWV 653b is an arrangement of 653a (with
some harmonic infelicities in bb. 14 and 73ff.: Stinson 2001 p. 49) or vice
versa (the bass of 653a looks like a ‘compromise’ of 653b’s two: KB p. 67)
can only be argued from internal evidence.
The double pedal is different from three other notable examples: Scheidt’s
in Tabulatura nova (six parts, c.f. in alto), ‘Aus tiefer Noth’ in Clavierübung
III (six parts, c.f. in pedal), and the finale of the Concerto BWV 593.iii.
The closer the three chorales BWV 653 are to any tierce en taille models,
the likelier the double pedal of 653b is to be registered 8 only, not so much
because of spacing (Bruggaier 1959 p. 148) as because of French convention –
assuming that Bach or Walther understand pedalle de flutte in Du Mage and
others to mean 8 not 16 .
As it is scored in P 802, there is no particular reason why the cantus is an
octave higher than BWV 653a; nor is the single bass line of 653a ‘obviously
a compromise’, for there would be little difficulty in producing a fifth line
from and around it, using simple motifs and keeping up the motion. The
first two bars of ‘Schmücke dich’ should warn against regarding a disjunct
pedal line as the sign of a compromise. Furthermore, the spacing could
351 BWV 653b–654
imply that BWV 653b is not necessarily an organ piece, and not necessarily
by J. S. Bach: perhaps an exercise by Walther rather than a transcription, in
the tradition of Buxtehude’s ‘Mit Fried und Freud’ which Walther evidently
regarded as an organ piece. If the four-part version was only an ‘arrangement’
(Umarbeitung KB p. 67), why would this be the version, revised, to appear
in final form in P 271?
BWV 654 Schmücke dich, O liebe Seele

Copies: as BWV 651 (without Kittel?).
Three staves; headed in P 271 ‘a 2 Clav. et Pedal’.
The TEXT is J. Franck’s hymn for the Eucharist, published 1649; rare, it
appears for the Twentieth Sunday after Trinity in an Arnstadt book of 1666
(Gojowy 1972).
Schmücke dich, O liebe Seele, Adorn yourself, dear soul,

lass die dunkle Sündenhöhle, leave the dark cavern of sin,
komm ins helle Licht gegangen, come to the bright light,
fange herrlich an zu prangen! begin to shine in splendour!
Denn der Herr voll Heil und Gnaden For the Lord full of salvation and grace
will dich jetzt zu Gaste laden, wishes to invite you now as guest,
der den Himmel kann verwalten, he who rules over Heaven
will jetzt Herberg in dir halten. wishes now to make his dwelling in you.
The following six verses speak of the hunger and fear resolved in the
Eucharist.
The MELODY by J. Crüger (much like a Geneva Psalter tune) was published
with the text and used in Cantata 180 (1724): Example 175. Listed in the
Ob, and set in another form in BWV 759.
Example 175
352 BWV 654
Something of a ‘Jesus hymn’, this setting seems to many ‘as priceless, deep
and full of soul as any piece of music that ever sprang from a true artist’s
imagination’, according to Schumann (David and Mendel 1945 p. 372). Not
as contrapuntally tight as BWV 652, its lines have a similar lyricism; less
thematic than BWV 653, it nevertheless has a rather similar texture and a
common ritornello form. But with its homophonic opening and striding
continuo-like bass it is even more like a dance; its cantus is simpler, too,
though disguised by melismas. Again the melody is both ornamented as
cantus firmus and paraphrased in the counterpoint. See Example 176.
Example 176
As in BWV 655, motifs derived from the cantus create running lines:
triads in the former, smoother lines in the latter. See Example 177 for
examples typical of settings in P 271, as well as of the smaller chorales of
Clavierübung III. Using motifs in this way is more integrated than in the Ob,
Example 177
where they tend to be formulae. Thus line 2 is heard in the alto of bb. 5–9.
Quite apart from the continuity, its contrapuntal harmony is much later in
style than what its Buxtehude-like shape would suggest, not perhaps quite
as late as the Schübler Chorales (Kube 1999 p. 575) but with all the poise of
a spacious aria of the 1720s. However ingeniously the closing bars derive
from the motifs, as they do, their peaceful Affekt is unquestionable.
The length of time elapsing before each cantus, including the first, means
that the usual expectations of organ chorales are left behind. Particularly
353 BWV 654–655
unusual are the modulations in the interludes not called for by the cantus
(F minor b. 23, A major b. 99, B minor bb. 108ff.), calculated to help
produce the length desired for a setting which at other times manages to
hover around the basic keys, though without tedium. Its elegance obviously
matches that of the previous chorale.
Few would disagree with Spitta that the piece has a ‘strange, puzzling
magic’ (I p. 607), though whether the Eucharist was approached at Weimar
in c. 1715 with the solemn piety of the nineteenth century is doubtful. More
objectively, the chorale is remarkable for the ‘mini-recapitulation’ in b. 116
and for that familiar, sustained melos – a sense of effortless melody – in
the interludes and accompaniment. It shares key, metre, melodic and even
harmonic style with the Andante aria ‘Tief gebückt und voller Reue’ in
Cantata 199 (1713), ‘bowed down and full of remorse’, words which suggest
something more graphic than BWV 654.
BWV 654a Schmücke dich, O liebe Seele

Copies: as BWV 651a, with J. C. Kittel.
Three staves; headed ‘Fantasia super . . .’, ‘a 2 Clav. è Ped.’.
The differences are slighter than before, amounting to variant readings in

some rhythms (e.g. b. 5 more pointed in the ‘Leipzig version’), pedal phrases
(e.g. soprano pedal point in b. 105 duplicated in the bass in BWV 654a)
and ornamented c.f. (fewer ornaments in P 802). Sources probably reflect
variant readings in the copies rather than systematic alterations for the final
fair copy.
The unusual modulations in BWV 654 are already here, though whether
they suggest a later phase of composition than BWV 662a (Zehnder 1995
p. 338) or are a device selected in order to create length on this particular
occasion is uncertain.
BWV 655 Herr Jesu Christ, dich zu uns wend

Copies: derived from P 271.
Three staves; headed in P 271 ‘Trio super . . . a 2 Clav. et Pedal.’.

354 BWV 655
BWV 655 can be described as:
a trio for two manuals and pedal, material derived (entirely?) from the
cantus, within a ritornello framework
plus a complete cantus firmus in pedal, the last third of the piece, and final
ritornello statement.
The last three bars give a sense of unity to a heterogeneous genre (sonata
plus c.f.) but also mean that the pedal can have had no solo stop? BWV 655
resembles and complements BWV 654 in having lines closely and constantly
derived from the chorale, but now as a bright ritornello trio. For examples,
see Example 178. The pedal triad is heard more often than usual in a Bach
setting, as if it were an appeal (‘Lord Jesus! Lord Jesus!’). Also derived are
the scales (e.g. descending quavers of b. 40), from the second and fourth
lines of the cantus. Although the style is light, without true cantus until the
bass entry of b. 52, there could be as much motivic involvement here as in
Clavierübung III. On the logical key-order, see below.
Example 178
A trio derived from a chorale melody appears to be original: the tradition

was to give the c.f. to the pedal throughout, while Bach’s trios have it only
at the end (BWV 655, 664) or periodically (676). BWV 655 is both a new
genre (trio with integral pedal) and traditional (pedal c.f.), the latter forcing
a change of direction in the upper parts of b. 55. Weimar cantatas have
paraphrases in upper obbligato instruments (BWV 161, five parts) or in the
bass line (BWV 172.v, four) or in a solo instrument (BWV 199.vi, three),
so a pure trio seems a logical step. BWV 655 is most like the Six Sonatas in
its episodes, e.g. bb. 10ff., and as with BWV 664, the notation of its earlier
version announces a new genre: two G-clefs above a bass.
The triadic figure recalls (anticipates?) those in the Ob setting BWV 632
(see there b. 12), but are presumably lighter and gayer, having a simpler
texture and harmony, a brighter key and a livelier tempo. The opening bar
itself, both its motifs and the feel of a question-and-answer in each half bar,
strangely recalls (or anticipates) the opening of the A minor Praeludium
BWV 894, later arranged as a concerto. Also, the concerto-like length allows
a Vivaldian series of keys: G, D, E minor, B minor, D and so to G major.
355 BWV 655–655a
Luedtke (1918 p. 78) sees in this ‘jubilant’ trio a reference to v. 3 of the text
(see under BWV 632), less meditative than BWV 632 and 709, which have
more in common. Keller’s view that the movement is ‘completely in the
style of the Six Sonatas’ (1948 p. 184) needs modification, since the short
phrases, motivic compactness and use of pedal are not. In these respects,
BWV 664 resembles the Sonatas more closely than does BWV 655.
The one-bar phrases, the immediate inversions (e.g. bb. 8–9 or 18–19)
and the material returning in different keys (compare bb. 28 and 43) are
typical of the movement’s integration. Homogeneity of material is even
more pronounced in BWV 664 (where the cantus is shorter, thanks to the
nature of ‘Allein Gott’), but the short-phrased, question-and-answer tech-
nique of BWV 655 doubtless influenced Bach pupils, especially J. L. Krebs,
H. N. Gerber and W. F. Bach. The pretty charm anticipates galant taste in or-
gan trios, and the various arrangements by the Nuremberg organist Scholz
(Emans 1997 pp. 46–7) may represent only some of the versions, variants
and revisions to which the piece was subjected.
BWV 655a Herr Jesu Christ, dich zu uns wend

Three staves; headed ‘Trio super . . .’, ‘à 2 Clav. et Ped.’.
Apart from some minor differences (fewer ornaments in P 802), the so-
called ‘earlier version’ has more angular lines at bb. 54–5 (left hand) and,
most importantly, a different semiquaver countersubject. This gives not only
a different aspect to (and sometimes parallel fifths in) bb. 2–3, 8–9, 18–19,
28–9 and 43–4 but also a pair of final bars reminiscent of Buxtehude. Trio
treatment of this melody in this key with similar motifs is also found in one
of Walther’s variations.
In supposing that BWV 655a was composed by 1714 (Zehnder 1991
pp. 56ff.) one could consider it with the Toccata in C as a movement applying
concerto-like characteristics to another genre. However, the concentration
of motif in BWV 655a – such that its episodes do not have a distinct theme –
is quite at variance with the ritornello techniques of Torelli or Vivaldi. Sim-
ilarly, the idea that the chorale-statement in the pedal serves as a ‘final
ritornello’, as in a concerto, rather goes against the idea of ritornello form.
A possible answer to the conundrum is that BWV 655a is not an early ver-
sion but an arrangement (Emans 2001), either one circulating before Bach
made the final version for P 271 or an early independent form of it made
or arranged by Krebs or Walther. Quite why either musician would do that,
356 BWV 655a–656
however, is not clear, nor whether this would be an isolated case: if Anh.II
61 was Bach’s re-written version or ‘modernization’ of a Pachelbel chorale
(something not demonstrable, however), so could BWV 655 be of 655a.
BWV 655b Herr Jesu Christ, dich zu uns wend

P 285, Scholz MS and nineteenth-century sources only (KB p. 70).
This shortened version, presumably made by Scholz, is based on the pedal

c.f. section of BWV 655, the left hand an octave lower, the harmony often
‘banalisiert’ (KB), and the motifs largely suppressed in the final bars. Omit-
ting the opening section may be to avoid the much-repeated triadic motifs?
BWV 655c Herr Jesu Christ, dich zu uns wend

Copies: SBB Mus. ms. 30377 (2nd half eighteenth century), P 285, Scholz
MS.
The first MS contains much-altered versions of BWV 538.ii, 540.ii and 680;
its version of twenty-nine bars includes not only ornaments characteristic
of the ‘Berlin School’ (KB p. 72) but motifs still further removed from the
chorale melody, of which there is no full statement in the pedal. Whether
or not BWV 655b and 665c represent arrangements, authentic or not, of a
trio finalized only in P 271 and circulating earlier, their differences appear
complementary.

Copies: as BWV 651 (without Oley).
Two staves; headed ‘3 Versus’, then ‘1 Versus. manualiter’ (only); third verse
with ‘Pedal’ cue.
The first verse has a somewhat irregular fore-imitation based on a double

subject, both of which are derived from the cantus firmus (Example 179). The
357 BWV 656
Example 179
conspicuous motif in b. 2 looks rather like a certain formula recommended

by Mattheson (1731 p. 221) for decorating a c.f. so as to produce a new
melody when improvising. Although the movement seems subdued, quite
why ‘a deep Passion atmosphere’ has been heard in it is unclear (Leutert
1967); rather, its unbroken quavers are like those of the more standard
accompaniments in Walther’s variations.
The second verse contains the c.f. in alto; the figuration is livelier but
equally conventional and thus ‘early’, as in the sequences of b. 66. Less likely
to be found in Walther are the ornamented cantus from b. 94 (imitated in
b. 97, bass) and the rising line towards the cadence (compare v. 3). Both
the quaver pattern and the alto melody recall several ‘Neumeister’ chorales:
quavers in BWV 1108 and alto melody in BWV 1105, 1107, 1108, 1118.
Even the ‘wild’ way in which the pattern takes over, in bb. 44, 46ff., 59ff. etc.,
would not be out of place in ‘Neumeister’, though on a less spacious scale.
In any case, the three verses are closer in idiom to chorale variations than
to the harpsichord courantes suggested in Klotz 1962b.
In v. 3 the c.f. steps down to the bass, as in tripartite settings from
Sweelinck’s ‘Da pacem, domine’ to the Kyries of Clavierübung III. Bass
cantus suggests a bigger plenum, so each verse has more stops than the
previous? Perhaps the chorale-melody supplied the opening notes of the
new countersubject in b. 104, in whose up-and-down shape Keller hears a
cross motif (1948 p. 186). The first section-repeat is varied, unlike verses 1
and 2. Halfway through the verse, the five parts break off for a new theme
said to be derived from the cantus – (b. 122) – invoking the ‘bowed head
of the Saviour’ (Keller ibid.) or ‘illustrating the act of bearing’ sin referred
to in the Agnus dei text (Spitta I p. 602) or suggesting in its repetition ‘the
multitude of the sins of humanity’ (Schweitzer 1905 p. 357) or alluding in
its ten entries to the Commandments and subsequent need of the Lamm
Gottes (Leutert 1967)?
Less speculative is the graphic reference to the text heard in the penul-
timate line, ‘sonst müssten wir verzagen’ (‘otherwise we should have
despaired’), generating imitative chromatics (entries on C and F), cor-
rupting the cantus (pedal b. 136) and invoking other music, such as ‘quälen’
(‘torment’) in Cantata 63.vii. The chromatics over a pedal C – itself
358 BWV 656–657
rare – are richer than the pedal point with chromatics illustrated in C.
P. E. Bach’s Versuch (1762 p. 184). The St Matthew Passion setting uses a
different form of the melody at this point and does not become strikingly
chromatic any more than did vv. 1, 2 at the word ‘verzagen’, though the set-
ting BWV 618 has an incipient chromaticism in bb. 20–1. So does ‘sauren
Tritt’ (‘bitter step’) in the second aria of Cantata 71, to whose period (c.
1708) BWV 656 might belong.
Clearly the last section is less fraught, with simple major scales and long
final pedal point representing ‘give us peace’ – even, some think, a vision of
angels. The change is certainly dramatic and encourages some such response,
but it also has a purely musical function, serving as tonic return to the quaver
pattern from v. 1.
BWV 656a O Lamm Gottes, unschuldig

Copies: J. T. Krebs and Mempell–Preller (via J. G. Walther?).
Two staves (verses not originally marked?).
The differences between the versions are superficial, amounting to variant

readings. The ‘earlier’ has somewhat fewer ornaments, occasionally different
detail and less correct notation (triplet quavers instead of crotchets, compare
BWV 608). The repeat in v. 2 is shortened by omitting bb. 64–70, an omission
not perhaps authorized by the composer, and resulting in a hasty leap to the
next line of the cantus.
It is not certain from P 802 that Krebs understood the three sections
to be one continuous movement, since there is a pause at the end of v. 1,
and the pedal cantus firmus does not begin as the last note of v. 2 but is an
up-beat to v. 3 on a new page (see Zietz 1969 p. 145).
BWV 657 Nun danket alle Gott (‘Leipzig Chorales’)

Copies as BWV 651; ‘Weimar version’ by J. T. Krebs.
Two staves in P 271, where headed ‘a 2 Clav. et Ped. canto fermo in soprano’
and cue ‘Choral’ in b. 5.
The TEXT of M. Rinkart’s hymn of 1648 became associated with wed-

dings, Christmas/New Year, and Reformation Day, as the hymn after the
sermon.
359 BWV 657
Nun danket alle Gott Now let all thank God

mit Herzen, Mund und Händen, with hearts, mouth and hands,
der grosse Dinge tut who does great things
an uns und allen Enden, for us and for evermore,
der uns von Mutterleib who from our mother’s womb
und Kindesbeinen an and our first faltering steps
unzählig viel zugut has done us immeasurable good
und noch itzund getan. and still does today.
V. 2 refers to the peace and fortitude given by grace, and v. 3 returns to the
praise.
The MELODY is attributed to J. Crüger (1647) and was published with

the text. It is harmonized in BWV 386 and the ‘wedding chorale’ BWV
252, and used in Cantatas 79 (Reformation Festival 1725) and 192 (1730):
Example 180.
Example 180
While each line of the cantus in the right hand is anticipated fugally in the
familiar manner of Pachelbel and others (compare BWV 723), the piece has
many original elements, carefully worked to be continuous except at three
conspicuous points of stretto, and increasingly so in the final section. The
pedal and inner parts are all fully developed, rich in motifs, with good har-
mony contrapuntally managed, and with imitations worked out differently
each time:
stretto in b. 1, contraction and expansion in 11–13, dominant quasi-reprise
in 39, varied stretti in 47–8, chromatic alteration in 55, no final pedal entry
All parts have a wealth of motifs, indeed somewhat undisciplined and ‘early’.
The resulting harmony is usually very expert in a way not far removed from
the Ob, as in bb. 58–60; and the cleverly different 1st/2nd time bars are those
of an expert harmonist. But the varied motifs present a patchy appearance.
It is easy to believe that ‘the piece is probably very old, perhaps already
re-worked in Weimar’ (KB p. 73) – even the cue ‘Choral’ is an old sign –
and any similarity to the chorale movement ‘Jesu, deine Passion’ in Cantata
182 (1714) need not imply that they were contemporary, since the latter
360 BWV 657–658
is old-fashioned, especially for a cantata. One could as well argue that the
clear soprano c.f. reflects that of Cantata 192 (1730).
The ‘Weimar version’ of the work differs only in minor detail. Perhaps
the composer kept such a chorale intact in his late collection as an example
of modified chorale-fugue on old models, ‘modified’ by constantly having
newly thought-out detail. A fore-imitation cannot often have been chro-
matically altered as here in bb. 55–6 (see Example 181) and was perhaps an
inspiration for that later in BWV 656? If BWV 657 is earlier than most of
the set, its b. 13 (c.f. against a syncopated line plus semiquavers) became a
habit, recurring in BWV 644, 658, 665.
Example 181
BWV 658 Von Gott will ich nicht lassen

Copies: as BWV 651.
Three staves, lowest in P 271 cued ‘Ped.’ (in the Oley MS, P 1160, lowest
stave marked ‘Pedal 4 Fuss’); headed ‘canto fermo in pedal’.
The TEXT of L. Helmbold’s hymn was published in 1572, sometimes asso-

ciated with Advent and Epiphany.
Von Gott will ich nicht lassen, I will not forsake God,
denn er lässt nicht von mir, for he does not forsake me,
führt mich durch alle Strassen, leading me through all pathways,
sonst ging ich in der Irr. otherwise I should have gone astray.
Er reicht mir seine Hand; He reaches out his hand to me;
den Abend und den Morgen evening and morning he
tut er mich wohl versorgen, takes care of me,
wo ich auch sei im Land. wherever I am.
The following eight verses return to the ideas of support, faith, praise and
trust.
The MELODY may come from a secular song ‘Ich ging einmal spazieren’
(Terry 1921 p. 312) and resembles other melodies known both in Germany
361 BWV 658
(‘Helft mir Gottes Güte preisen’) and elsewhere (‘Une vierge pucelle’),∗
thereby acquiring Christmas associations. The melody is harmonized in
BWV 417–419, listed in the Ob and set in Cantatas 11 (Ascension 1735), 73
(Epiphany 1724), 107 and 186a (1724 and 1723): Example 182.
Example 182
BWV 73 (simplified)
This, the first of the settings in the minor, immediately changes the aspect
of the collection. Was F minor (otherwise unknown for this melody?) chosen
for BWV 658a merely to avoid d in G minor? Does certain spacing and a
lh part from C to a suggest that it was transcribed down from an aria and
re-written in the process? The difficult harmonies in unequal temperament
are in no way softened by careful part-writing, and while they gave less
trouble in 1745 than 1715, questions remain.
Vigour and continuity are created not only by alternating scale and
broken-chord note-patterns but by the division of the melody into only
three cantus-firmus phrases (four with repeat). While the lines derive from
the melody – see Example 183 – the most prominent feature from the be-
ginning is the countersubject motif a. The middle phrase of the cantus may
just be heard in the soprano in bb. 23–5, but a clearer reference in bb. 27–9
is coloured by this motif, as in Example 183 (ii). This amounts to a stretto
between soprano b. 27 and bass b. 29. Such paraphrase is quite distinct from
(e.g.) BWV 649’s, in which the new ritornello melody is much longer than
the chorale line it began by paraphrasing.
Example 183
∗ Marpurg
also noted this similarity in 1759 and compared Daquin’s canonic variations on it with
BWV 769, another Christmas chorale (Dok III p. 127).
362 BWV 658–658a
The bass line is largely made up of moving quavers below the crotchets
of the c.f., much as in chorales where manual and pedal parts are vice-versa.
The pedal’s ‘4 Fuss’ rubric in the Oley MS, which is not a copy of P 271,
might have come from another authentic source now unknown. Or it might
be an unjustified imitation (on Oley’s part) of the Schübler chorale BWV
646, for pedal at 8 supplies a tenor line otherwise absent, and for which, to
judge by bb. 5–8, the other parts make provision.
The rhythm of motif a has inspired much speculation. It is the ‘beatitude
rhythm’ of ‘Mit Fried und Freud’ in the Ob (Keller 1948 p. 187) or the
figura corta found with texts expressing an ‘awakening’ (Schmitz ‘Figuren’
MGG1). One might think its rhythmic insistence a reminder of the text’s
idea of persistence, and F minor chosen not for lugubrious effect but only
for the sake of a convenient compass – pedal to c , manual to c . The final
pedal point is very striking, its harmony long-spaced, rhythms new, motifs
more original than that at the end of BWV 656 or even the rhythmic coda
at the end of ‘Aus tiefer Noth’, BWV 686. Since the two penultimate bars are
not even necessary, it seems likely that they intend some special effect, such
as bells or even the text of v. 6 (Meyer 1972):
wir werden nach dem Tod after death we shall be

tief in die Erd begraben: buried deep in the earth:
wenn wir geschlafen haben, when we have slept
will uns erwecken Gott. God will wake us.
However, if seen as part of an Advent or even Christmas text, ‘awakening’

has another significance, something closer to ‘Wachet auf!’.
BWV 658a Von Gott will ich nicht lassen

Three staves; headed in P 802 ‘Fantasia super . . . ’, ‘à 2 Clav. et Ped.’.
The differences are more than notational or simple variants but do not
amount to total systematic revision. The opening right-hand paraphrase is
without ornament; and the bass line is often simpler and higher. The changes
of harmony in bb. 26 and 32 and of alto figuration in b. 35 suggest that
BWV 658’s revisions may have been made at the keyboard. What J. T. Krebs
can have meant by ‘Fantasia’ and, even more, by ‘à 2 Clav.’ is unclear: perhaps
it was the heading for another setting, here by mistake? How the piece could
have been notated or transmitted in another form, or understood as for two
manuals, is difficult to see.
363 BWV 659

Copies: as BWV 651.
Three staves; headed in P 271 ‘a 2 Clav. et Ped.’.
The chorale and its associations allow a whole range of approaches and
therefore Affekt, from rapt quietude to boisterous clamour. The Catechism
speaks of Jesus as beatifier, as the crucified, and as protector, and some
such ‘trilogy’ seems likely to be inspiring the three settings BWV 659–661.
It would not follow that the three were ‘obviously thought of as an inter-
dependent whole’ (Spitta I p. 607), since apart from there being no obvious
purpose in this, other settings than BWV 659a are found in Walther sources
and the three were not consecutive in J. T. Krebs’s copy.
The opening imitation of BWV 659, meditative, like two cello obbli-
gati, is derived from the cantus, as are lines throughout the chorale: see
Example 184. Using material this way under an ornamented cantus is often
described as ‘the Buxtehude manner’, but no extant chorale of Buxtehude
is quite so thorough. Nor in chorales with fore-imitation is there usually
such a stirring bass, here more like a violone continuo than a pedal part,
quasi-ostinato (bb. 1, 8, 9, 16–17, 24) and not far from the slow movement
of a concerto. A notable break in it occurs below the ornamented Neapolitan
sixth in bb. 22–3 (cf. the Prelude BWV 546 bb. 138–9).
Example 184
b. 8 (cf. bb. 16–17) b. 9 b. 24
D C F E D C B E DCD
As often in Bach, technical ingenuity is not an enemy of touching, ex-

pressive music. An exquisite melody spins out of and around the notes of
the cantus in the manner of Böhm, although the end must surely refer to
Buxtehude’s setting of the same melody, BuxWV 211. Florid treatments of
a cantus often took flight to an upper octave, but the beautiful expansion
of line 1 into the wide, melismatic melody of bb. 5–8 has no precedent.
Each line is treated in this way, beginning recognizably with the chorale but
giving free rein to bewitching sequences. The melismas arise particularly
at those points in the chorale melody that correspond to the second- or
364 BWV 659–659a
third-from-last syllables, something by no means common: it is not the

case with BWV 660, for example, though a Kauffmann might hint at it in
his simpler music. An earlier example of a spun-out melody in G minor,
Var. 1 of BWV 768, only underlines how beyond formulae is the present
melody.
The biggest melismas (bb. 14ff., 22ff., 32ff.) are inspired by the sequences
inspired by the cantus, in effect drawing out and colouring the penultimate
notes. Such bars as b. 23 naturally resemble free organ pieces (G minor
Fantasia, bb. 45–6). There is a puzzling relationship with another beautiful
melody probably from the Weimar period, the one re-used to open Cantata
156 (1729?) and found again in the F minor harpsichord concerto. Though
quite different in Affekt, the details of this melody’s melismas and turns of
phrase are very like ‘Nun komm’s’ – is one to suppose that Bach had a stock
of ideas he knew to be reliably expressive, in the minor in BWV 659 and
slower because of the accompaniment and the Advent text?
Although when sharpened the leading-note in the opening line is open
to fanciful interpretation (a ‘diminished fourth . . . significant of suffering’,
Terry 1921 pp. 18–19), the three main sources have it only for the return
in b. 28, and the derived line in b. 1 has no sharp. (Cantata 36 has it the
other way: first sharp (movement ii), then natural (vi, viii).) One could
argue that the modal melody at the beginning needed no f but the fully
diatonic accompaniment of b. 28 did. Also, it is not clear if b. 5 is meant to be
different from b. 29, though a reason could be conjectured if it were – b. 29
represents the composer’s last thoughts after the simple mordents of BWV
659a? Or simple mordents were customarily treated with some freedom?
BWV 659a Nun komm, der Heiden Heiland

Copies: as BWV 651a but no Walther.
Three staves; headed by J. T. Krebs ‘Fantasia super . . .’.
The difference lies chiefly in the ornamented cantus: P 802 gives few orna-
ments after the first line, and in the later version there are more melismas
in the third line, particularly around the Neapolitan sixth. It is possible that
when the piece was first written, its note-patterns were more conventional,
like those listed by J. G. Walther in 1708. (But by c. 1740, perhaps no-one
noticed that the pedal is constantly inventing different four-quaver patterns,
many like semiquaver patterns in BWV 680, all carefully varied.) A scale run,
such as in b. 15, was once a standard tirata figure, but so discreet here as to
seem original.
365 BWV 660

Copies: as BWV 651 (without Oley).
Three staves; headed in P 271 ‘a due Bassi è canto fermo’.
The double-length notes of the setting’s cantus have the effect of putting the
f on the beat, unlike BWV 599, 659 and 699. On the question of an ‘Advent
trilogy’, see BWV 659.
The kind of imitation in this extraordinary invention is not unknown
to Pachelbel, or its repetitive bass to Buxtehude, or its coloured melody to
Böhm; but two such bass lines, such a tight ritornello plan and such con-
stant diminished fourths produce something totally original. One might just
discern a Vivaldian concerto-form behind it, but there is nothing, anywhere,
as systematic as this:
1 stretto imitation
4 sequences
7 cantus firmus (partly over further stretti; its final cadence drawn
out)
11 as 4; spread ‘viol’ chord to end section
15 as 1
17 as 7 and 9–10 (except final cadence not drawn out)
20 as 1, to relative
24 as 17
26 as 1
30 as 4 (to C minor, as in b. 27 of BWV 659)
33 as 17
39 as 4 (inversion); spread ‘viol’ chord to end section; isolated
pedal note
The canonic imitation itself contains the whole first line of the melody,
dispersed (see Example 185). Note that the ornamented cantus firmus is
Example 185
366 BWV 660
closer than BWV 659 to the original hymn-tune. Its main notes (especially in
bb. 8–9) fall squarely on strong beats, and the basic line is as easily picked out
as in much more naive paraphrases. In detail, the ritornello sections are also
unusual: the sequences are simple (b. 4 etc.); each section cadences, having
modulated farther than the cantus demands; and tightness is achieved by
overlap (new stretti already during first cantus phrase). The most haunting
sound is the many diminished fourths, which become something of an idée
fixe, counteracted by the five patently conventional bars closing the first and
last ritornello statements.
There are various suggestions about how such a movement originated.
While the idea of a two-part invention accompanying a cantus is found
elsewhere (BWV 675, 688), as is some of the figuration (BWV 646), two
bass parts are unique. In theory like a cantata aria with cello obbligato and
continuo, and not unlike old trio-sonatas with gamba (Buxtehude, Marais),
in practice lower bass parts will rarely compete with the tenor in the same
way. Exceptions such as two bass parts in Legrenzi’s sonata ‘La Bevilaqua’
Op. 8, or even Frescobaldi’s canzoni a due bassi, may be relevant, imported
for a novel chorale-setting. But perhaps Example 185 explains it best: the
duo expresses the cantus heterophonically.
Not only the closing ritornello but the whole movement seems curiously
to anticipate the Schübler Chorales, with its sequential melody (compare
bb. 4–5 with BWV 649), the succinctness of the whole, and its air of being
a transcription of instrumental parts. Perhaps the aria for two obbligato
cellos in Cantata 163 (1715) offers the closest parallel. Yet the two basses are
certainly conceived in organ terms, whether registered 8 (not suggested by
any evidence) or both 16 , in the style of Kauffmann’s suggestions for adding
Fagott or Quintadena 16 to manual basses (Harmonische Seelenlust, 1733).
It depends very much on the organ whether either or both basses can have
a 16 stop.
Keller’s ingenious suggestion that the harrowing of hell in v. 3 prompted
this setting could certainly influence a player (1948 p. 188):
Sein Lauf kam vom Vater her . . . His course came from the father
fuhr hinunter zu der Höll. and led down to Hell.
But it cannot be assumed that the piece has so ‘rough’ an effect. Also fanci-
ful is the idea that the final short chord ‘shows God abandoning his son’
(Chailley 1974 p. 200), since in P 271, the last bar is so cramped that both
a natural sign and arpeggio symbol may have been perforce omitted. On
the other hand, the various roles of Jesus obviously include the crucified
Saviour – see BWV 661 – and the crossed lines of Example 185 are as likely
as any elsewhere to be allusive.
367 BWV 660a–660b
BWV 660a Nun komm, der Heiden Heiland

Autograph MS, three sides written between 1714 and 1717, added to P 271
in the nineteenth century (Dadelsen 1958 p. 79); other copies as BWV 651a,
plus J. P. Kirnberger.
Three staves; headed in P 271 ‘à 2 Clav. & Pedal’.
As is the case with BWV 659, the ‘early version’ – already revised in the
early autograph – has somewhat fewer ornaments (more cursorily written
in P 802 than P 271) and, with fewer small notes in the cantus, seems less
‘robust’. It is not clear why the left hand of BWV 660a b. 33 got re-written
(source had a tenor clef?), or whether the final major chord of BWV 660a
was really intended to be minor in BWV 660 (the dominant in b. 15 is major
in both versions) and played non arpeggio there (bb. 15, 42 arpeggiated in
BWV 660a). The opening canon between bass and tenor might suggest it
was contemporary with the D minor Concerto BWV 596, which begins
similarly.
The sudden arpeggio-chords in bb. 15, 42 might be remnants of an
‘original’ cantata version with viola da gamba, the lines suiting the compass
of gamba (D–g ) and cello (C–d ). But the first and perhaps second of these
chords in the autograph of BWV 660a – not of BWV 660 – looks like an
addition: the composer first wrote d alone. Nevertheless, since the decorated
cantus is rather like a soprano chorale in cantatas (BWV 80.ii, 1715), it is
possible that BWV 660a began as a transcription and was adapted somewhat
further for BWV 660 (more right-hand notes in the cantus) – the only such
instance, if this is so.
Either way, BWV 660a must be one of the composer’s first essays in
setting an organ-chorale (the cantus plus interludes) as a concise concerto
ritornello form (the cantus now as episode). For this new form, it seems he
imitated string instruments.
BWV 660b Nun komm, der Heiden Heiland

Copies: J. T. Krebs and J. G. Walther.
Three staves; headed in P 802 ‘a 2 Clav. e ped.’ (anon).
BWV 660b could be an arrangement by J. T. Krebs of BWV 660a already

copied into P 802 (KB p. 77). The two bass parts are in the right hand (up
an octave) and left hand respectively, the cantus firmus without ornaments
368 BWV 660b–661
in the pedal, an arrangement matching e.g. BWV 694. Or, in view of the
versions of BWV 655, some of these settings did circulate in more than one
form.

Copies: as BWV 651 (but no Oley).
Three staves; headed in P 271 ‘in Organo pleno – Canto fermo in Pedal’.
On the question of an ‘Advent trilogy’, see BWV 659. The ‘Full Organ’
character of the third setting is clear from the pedal c.f., the type of
counterpoint, and the rubric. Long ritornello sections comprise a series
of fugal expositions on a subject derived from a c.f. against which it is con-
stantly adapted rectus and inversus. The startling paraphrase (Example 186)
passes in the codetta bb. 7–13 to sequences typical of free fugues like the C
minor BWV 537. It also resembles a free fugue’s countersubject (cf. BWV
538), just as its angular motifs explore the figura messanza; episodes give
it its breadth, as in BWV 546; and the inversus in b. 45 anticipates fugues
such as BWV 547 that have no angular second subject (unlike BWV 540).
In short, it is much like a free fugue and as such is almost as unusual as
BWV 660.
Example 186
G G F B A G A (G)
The subject’s patterns are elastic enough for it to be relatively straight-

forward to combine them with cantus phrases (e.g. alto in bb. 24, 26; full
soprano entry in b. 28), even when inversus (e.g. tenor in b. 57; full entry in
b. 60) and despite those cantus phrases being of unequal length. The inverted
theme first appears just before the halfway point (alto b. 45), after which
much of the counterpoint returns inversus, including a complete three-voice
passage. Bars 48–53 are a close inversion of bb. 15–20. The result is a certain
‘remoteness’ in the counterpoint, with a bass that is only marginally success-
ful in bb. 48ff. But when the bass has the theme a grand, majestic celebration
369 BWV 661–661a
of Advent results: see Example 187. Combining cantus firmus and a fugue
theme derived from it occupied the composer in many ways (BWV 686, 695,
733), though the intervals of this melody make combination more awkward
than it was to be in the Art of Fugue BWV 1080.ix. The countersubject itself
(b. 4) seems at first to glimpse the chorale melody.
Example 187
Naturally, it is tempting to see three roles of the Saviour evoked in the

three Advent chorales. Both hymn and Catechisms speak of Jesus ‘the only
beatifier and Saviour’, Jesus who suffered crucifixion, and Jesus who ‘with his
power protects us against all enemies’. Perhaps all the inversion in the last of
the three settings was prompted by the hymn’s speaking of the Son ‘return-
ing to the Father’ (Meyer 1987 p. 44), although the number of things inversus
is supposed to denote is alarming. To have the cantus in the pedal was com-
mon for the last of three settings (BWV 656, 659–661, 662–664, 669–671),
and evidence from Kauffmann’s Harmonische Seelenlust and elsewhere sug-
gests pedal reeds for a powerful pleno. Here this could seem particularly
apt, as could the fact that the counterpoint throughout is not unlike the
Magnificat’s, BWV 733: the latter for Annunciation, the former for Advent?
BWV 661a Nun komm der Heiden Heiland

Two staves, no heading.
The ‘later version’ changes the earlier notation of 4/4 semiquavers to alla
breve quavers and made the final note pattern of the original countersubject
(b. 6) more angular and less ‘spun out’. The original time-signature produces
bars looking very like those of older praeludia and chorales in this repertory
(e.g. BWV 665) and thus, perhaps, a slower tempo than one would assume
for BWV 661.
370 BWV 662
BWV 662 Allein Gott in der Höh’ sei Ehr’

Copies: as BWV 651 (without Oley; Penzel’s MS with improved ornaments).
Three staves; headed in P 271 ‘a 2 Clav. et Ped. canto fermo in Sopr.’ (added?);
‘adagio’ below opening tenor.
The TEXT is an adaptation by N. Decius of the ‘Gloria in excelsis Deo’ (1522),

sung in Leipzig on each Sunday, four verses by choir and congregation after
the priest’s intonation from the altar (Stiller 1970 pp. 77–8, 103).
Allein Gott in der Höh’ sei Ehr’ Alone to God on high be honour
und Dank für seine Gnade, and thanks for his grace,
darum, dass nun und nimmermehr since that now and for ever
uns rühren kann kein Schade. no harm can touch us.
Ein Wohlgefall’n Gott an uns hat, God is well pleased with us;
nun ist gross Fried ohn Unterlass, now is great peace without intermission,
all Fehd hat nun ein Ende. all strife is now at an end.
The following verses address each Person of the Trinity in turn.
The MELODY derives from the plainsong Gloria (Liber usualis, Mass I
for Easter ‘Lux et origo’), particularly at ‘Et in terra pax hominibus’,
‘Benedicimus te’, and ‘Adoramus te’. Only the chorale repeats the open-
ing two lines: see Example 188. Listed in the Ob, harmonized in BWV 260
and set to other texts in Cantatas 85, 104, 112, 128, the melody is set more
often than any other (BWV 663, 664, 675, 676, 677, 711, 715, 716, 717) –
presumably because of being so often sung, not because it has simple two-bar
phrases (as Tusler 1968 p. 21 suggests).
Example 188
The ‘adagio’ direction in P 271 singles out the movement, and its treat-
ment of the Trinity hymn becomes a companion to BWV 659’s treatment of
the Advent hymn. The fugal subject is a distant paraphrase (Example 189),
with two important patterns, a and b, the first a pedal motif in b. 2, the
371 BWV 662
Example 189
second in the tenor b. 33, etc. Throughout the chorale, these patterns pro-
duce a most melodious line: to list their contrapuntal attributes does not
quite express their natural sweetness, which is curiously enhanced by the
many trills –
1 double subject, one derived from cantus, both supplying motifs

6–7 countersubject and subject appear (alto + tenor) in reversed
order
8–9 parts wait for the ornamented melody to begin
10ff. much motivic imitation between the parts, in the Pachelbel
style
33–4 decorative alto refers to cantus line 3 (clearer in pedal); main
theme re-written as fore-imitation of this line (cf. BWV 663,
717)
35–7 derived from bb. 5–7
41–4 line 4 against motifs a and b
– and so on, until all the spun-out, derived lines stop for free decoration
of the cantus’s last note a (bb. 49–53). This little ‘cadenza’ seems to be
anticipating or recalling – which? – those in the opening movements of
Cantatas 12 and 21 (1714).
Although the rh melody is one of the most ornate chorale-paraphrases
in the repertory, the cantus remains more recognizable than BWV 659’s
because its notes are there on the beats. In addition, the pedal is highly
derivative:
2ff. motif a spun out, with continuation (45–6) or without; often

inversus
6ff. motif b spun out, seven times with its continuation
It could be that the unusual ornament of motif b (the lombardic accent) is

left thus and not written out, so that it can be omitted in the pedal b. 6 etc.
This pedal begins more like a continuo part, with a cantus-derived phrase
that would be at home in the Canonic Variations.
372 BWV 662–663
Though highly decorated, the melody is more recognizable than that

of (e.g.) BWV 659, a developed version of the coloratura chorale known
elsewhere – even the unique cadenza (bb. 51–2) is more an extension of
the cantus-firmus tailpiece often found in Buxtehude (now around a dimin-
ished seventh) than a cadenza in any later sense. The melody’s flourishes
occasionally remind one of an obbligato string or wind aria in Weimar can-
tatas, ordering a decorative melody into fugal ritornello form. The extent to
which patterns are developed in the work may imply a new stage of devel-
opment: this melody suggests all kinds of motifs, and the lines of BWV 662
and 663 could hardly be more different. Similarities between BWV 662 and
656 (opening counterpoint) and between BWV 663 and 656 (final cadence)
arise because their chorale melodies begin alike.
The ornaments, unusual in themselves and in their frequency, seem to
suggest a languid mood, particularly in the Lombardic rhythm of motif b.
Various things may be read into this: the ‘bringing of Heaven down to earth’
(Keller 1948 p. 189) or the ‘condescension’ of the Trinity (Meyer 1972), all
as usual unverifiable.
BWV 662a Allein Gott in der Höh’ sei Ehr’

Three staves; headed in P 802 ‘a 2 Clav. e ped.’; in Gerber MS ‘forte’ (rh)

and ‘piano’ (lh: KB p. 80).
The chief difference is that BWV 662a has fewer ornaments, though the first
two accents of motif b are already there. Perhaps BWV 662 clarifies what was
earlier taken for granted – that such motifs at the end of b. 4 would have
ornaments, and that the trills would vary with context.

Copies: as BWV 651 (but no Oley).
Three staves; headed in P 271 ‘a 2 Clav. et Ped. canto fermo in Tenore’, and
below rh, ‘cantabile’.

373 BWV 663
The setting is a quaver perpetuum mobile in which the cantus is deco-

rated with long and short notes, as if a chorale was blended with a tierce
en taille. Both fugal theme and pedal part are derived from the melody:
Example 190. So is the harmony they produce. But whereas BWV 662 opens
with an important falling figure, BWV 663 and 664 rise, inviting a symbolic
interpretation of descent and ascent. The greater vigour and slow harmonic
rhythm of BWV 663 are striking, in their way original too. The final pedal
point invites another symbolic interpretation: a reference to the last word of
stanza 1 (‘Ende’), and this after the ‘adagio’ ( = rallentando then a tempo?)
has drawn attention to the phrase ‘without respite’ (‘ohn Unterlass’). But
note: the cadence finally is more succinct than that of the C major and
G major Fugues WTC1, which it resembles.
Example 190
Both motifs a and b are elastic, the first producing a nonstop running
line, the second an unusual bass, monothematic but unrepetitive. The tierce
en taille is complete with similar trills ending each phrase, a solo bar (b. 96),
an ‘adagio’ pause, and a division into two at b. 110 (holding up the cantus).
Tenor melodies whether manual or pedal are not usually so ornamented,
and before it enters, there is the impression of a pedal cantus firmus starting
some way into the piece (b. 9), as in BWV 651. Pedal keeps up the idea,
including stretti at bb. 69 and 73. Typical of the tierce en taille are the held
notes, scales and ornaments; less typical are the separate phrases, the divided
line, the long final, and above all the ‘Italian trio sonata Allegro’ style of the
movement as a whole (Ponsford 2000 p. 71).
So a much-used theme is now cast in a new contrapuntal ritornello
form, spacious and integrated, with fugato, fore-imitation, canon and col-
oratura as well as c.f., amalgamated to produce the length the composer
was clearly aiming for. The very rests are contrapuntal, and the running
decoration of the cantus is melodious and inventive. While details of the
ornamental melody are ‘Böhmian’ (rhythm of b. 28 – compare b. 9 of
BWV 662), the whole accords much more closely with paraphrase tech-
niques in BWV 651–665 as a whole, including the wish for sheer length.
Similarly, while the canon in bb. 69–79 resembles Walther’s in Var. 5
of his ‘Allein Gott’, the pedal’s work with motif b could only belong to the
composer of the Orgelbüchlein. This motif ’s ‘pure’ form is there in b. 1, but
it is also diminished (b. 2), doubly diminished (b. 1), paraphrased (tenor
374 BWV 663–664
bb. 16–17), inversus (b. 18), détaché (b. 14), détaché inversus (b. 15), doubled
(b. 124), coloured and détaché (b. 23), and may have yet further manifes-
tations. All this means that the setting is a fantasia on G A B and as such
anticipates Clavierübung III, as do the canons; compare bb. 69ff. here with
bb. 78f. in the trio setting BWV 676. The part-writing – a trio with two
or even three extra voices, but inconsistently – suggests a work predating
BWV 664, but the ingenuity is such that even apparently simple material,
such as the detached chords in bb. 103 and 105, is derived from motif b.
One wonders at times if there is any note in the whole setting that is not
derived from the cantus.
The order of Trinity settings, an order evidently specific to P 271, makes
the next chorale BWV 664 appear as yet another fantasia on the same Trin-
ity motif of a major third, all the little semiquaver patterns matching the
minims, crotchets and quavers it gave rise to in BWV 663.

Copies: J. T. Krebs (two).
Three staves; headed ‘à 2 Clav. è ped.’.
Apart from a few notational changes, the ‘later version’ has regularized the
cadenza and ‘adagio’, though it omits the ‘andante’ (which must mean a
tempo) of the following bar. ‘Andante’ suggests that the setting is not fast,
though ‘allegro’ for a 3/2 movement would in any case have been unlikely.
BWV 663a in P 802 is clearer than P 271 since it assumes that ‘adagio’ means
‘rallentando e più lento, e poi accelerando a tempo’ and that the solo tenor
line is free. Perhaps P 271 was made simpler for publication?

Copies: by and via J. P. Kirnberger and J. C. Kittel.
Three staves; headed in P 271 ‘Trio super. . . .’, ‘a 2 Clav et Ped’.
Entered into P 271 after a gap of three or four years, this was already thor-
oughly revised before being copied.
375 BWV 664
Spitta saw the trio as an inventive way of using fore-imitation of the Pachelbel
kind above pedal cantus firmus (I p. 604); here its motifs are developed
confidently, and with the invertible counterpoint of Italian string trios. See
Example 191. For this reason alone, it seems unnecessary to relate BWV 664
to Torelli’s Concerto Op 8 No. 2, as in Zehnder 1991: the counterpoint is
as like Corelli’s trios as anything else, and the scale of it is entirely Bach’s.
It is longer than the trio BWV 655, thanks chiefly to long episodes which,
in playing with thirds or triads (bb. 36, 49 etc.), might ultimately derive
from the cantus. In general character, as a three-part piece in A major, it has
more than a passing resemblance to the A major Prelude WTC2: was this
composed while the chorale was in the composer’s mind?
Example 191
All three lines are derived from the cantus, complementing the other
kind of tour de force in BWV 663: both rh and lh subjects (the second a
‘modified answer’) as well as pedal basso continuo bass from b. 1. Their lines
match the Six Sonatas and the later ‘Allein Gott’ BWV 676. The paraphrased
theme itself recalls chorales BWV 663 and 676, though the different metres
(3/2, 4/4, 6/8) give a different character right from the start. The ritornello
demands of trio-sonata form mean the subject enters at regular moments,
but the detail is unconventional:
1 double subject plus continuo; all three from cantus. Modified

cantus for pedal from end of 9; double subject in rh, end of 10
12 episode on second half of subject; modified theme, 16
25 cadence in A, the new and (for lh) unusually high entry in D
31 episode; motif a inversus and rectus; 35ff. broken chords as in
Sonata No. 6, first movement
43 short entries submediant minor and (64) supertonic minor;
episodes, a expanded (68 motif, as C minor Violin Sonata
BWV 1017.ii)
80 tonic entry with answer
85 cantus lines 1, 2 in pedal; motifs continue over final pedal point
(as in BWV 661); NB tenor b. 96
In the process, an unusual repetition occurs: 56–72 = bb. 35–51, parts

exchanged and up a fifth (down a fourth), the two sections ending identically.
376 BWV 664–664b
Consequently, the work is one of three sections: chorale trio + concerto or

chamber trio + chorale trio with c.f.
In this complex, the episodes account for more than half the music, with
sections returning in a different key (cf. BWV 655) and introducing passages
very like the Sonatas (e.g. bb. 35ff.) or even a cantata duet (b. 53). The final
c.f. section is only one seventh as long as the whole ‘trio sonata’, unlike
BWV 655 where the proportion is nearer two fifths. The chain of trills, if
that is what they are (bb. 39ff., 60ff.), also anticipates the Sonatas, though
the copyists of the so-called early version did not give them – were they
originally simple suspensions? The shortened c.f. in b. 85 leaves the balance
of the movement unimpaired, with finality achieved because the chorale’s
first and last two lines are similar (compare BWV 716). It also conforms
with other groups of three chorale settings in which the pedal takes the
final c.f., though this trio is brighter and gayer than either BWV 671 or
661.

In P 801 three staves, two treble clefs; headed ‘Trio super’, and at end ‘SDG’
(‘Soli deo gloria’).
The differences do not amount to a systematic revision, the earlier ver-
sion showing fewer ornaments throughout (originally left to the player? –
e.g. from b. 39), a simpler rhythm in bb. 1–2 etc., and occasional minor
difference in a line. One may surely doubt whether this version is many
years older than the Six Sonatas as finalized, so close to them in idiom
is it.
BWV 664b Allein Gott in der Höh’ sei Ehr’

Copies: by or via J. C. Oley, J. P. Kirnberger and other Leipzig sources.
Two of the more important differences are that the pedal is occasionally
an octave lower (most of bb. 40–3) or less smooth, and the theme begins
without the échappée on note 10. The theme on subsequent appearances is
as in BWV 664a and 664, and one cannot know whether the copyists made
a mistake or the composer changed it to the other, vastly superior phrase.
But together, BWV 664a and 664b suggest an originally simpler paraphrase
of the melody than BWV 664’s.
377 BWV 665
BWV 665 Jesus Christus, unser Heiland

Copies: only via J. C. Kittel?
Two staves; headed in P 271 ‘sub communione.’ and ‘pedaliter’ (added?).
The TEXT is Luther’s free translation of the hymn ‘Jesus Christus nos-
tra salus’, said to have been written by John Hus. It served as a doctrinal
hymn before Communion, during which it was sung and played alternatim
(Luedtke 1918 p. 87). Schein (1645) and Vopelius (1682) give it as a hymn
for Maundy Thursday.
Jesus Christus, unser Heiland, Jesus Christ, our Saviour,

der von uns den Gottes Zorn wandt, who turned God’s anger away from us,
durch das bitter Leiden sein through his bitter suffering
half er uns aus der Höllen Pein. helped us out of the torment of Hell.
Nine further verses discuss the sacrament and the love of one’s neighbour.
The MELODY, perhaps late Gregorian, was published with the text in
1524 (Example 192), beginning like that of ‘Wir glauben’. It is used in
Clavierübung III, listed in the Ob and harmonized in BWV 363.
Example 192
The form is regular and thus old-fashioned:
line 1 derived theme in tenor, countersubject in manual bass, alto

answer, then pedal (plain notes; manual bass drops out),
then soprano (each with countersubject); freely derived
four-part coda
line 2 as line 1, new countersubject begun in upper part; coda
ditto
lines 3, 4 as line 2 but pedal with held finals; parts added at end
To give each voice the cantus in the same note-lengths is something found
in Böhm; and moving the bass from manual to pedal appears in BWV 549a.
378 BWV 665–665a
(The two staves in NBA stress this antique element more than P 271 does,
where fewer settings have three staves.) The breaks between sections are
clearer than in more recent works: in BWV 666, the breaks are caused by
interludes before the following line, rather than by the previous line worked
to a climax as in BWV 665. The harmony is masterly (e.g. bb. 5–6) beyond
the conventional figuration, and each line’s freely derived coda brings out
clearly the allemande-like character of the texture, e.g. bb. 11–13.
The countersubjects have long been seen as giving the chief interest to
the movement, not because they can be traced to the chorale melody –
they cannot – but because they impart different Affekte to each chorale line.
Spitta, who admired the piece (I pp. 602–4), found reasons for the ‘sub
communione’ heading, and Schweitzer saw representations of key words
from v. 1: God’s anger (bass, bb. 14–15), bitter suffering (chromatics b. 27),
and resurrection from ‘the pain of Hell’ (rising demisemiquaver motif b. 38).
The first line and its opening countersubject may be less easily labelled –
‘carrying of the Cross’, perhaps (Grace c. 1922 p. 279). The setting was
retained by the composer and put in the Leipzig autograph to give a ‘glimpse
of his compositional development’ (according to Meyer 1979b pp. 40ff.).
The motifs thought to express the various images work gradually towards
the final pedal point: the biggest close so far in the whole collection, a clear
attempt to give shape to the disparate elements, perhaps an expression of
‘escaping from the torment of Hell’, and even more rhetorical than the com-
parable endings of Weimar cantatas such as BWV 161. While the chromatic
line (b. 37 etc.) is no doubt meant to be evocative, note that the harmony
is quite simple, with repeated Gs on the main beats and Ds in the bars be-
tween (bb. 30–2). There is no good pictorial reason for the line to fall in the
particular way it does, but for a not dissimilar Affekt or effect, see BWV 656.
BWV 665a Jesus Christus, unser Heiland

Copies: J. T. Krebs, and others via J. C. Kittel?
Two staves; headed in P 802 ‘in pleno Organo’.
While the differences do not amount to a radical revision, sources imply

that the composer made two alterations in the motifs: the demisemiquaver
dactyl was added in bb. 28, 31, 34 and 36; and in bb. 49–50 a plain figure
became chromatic. The authority for ‘pleno organo’ is questionable, since
this registration is usually given for continuous, non-sectional movements;
perhaps Krebs added it because of the pedal, which being last with each line
looks rather like a cantus firmus.
379 BWV 667–667b

No Autograph; copy by J. C. Altnickol (P 271); other Leipzig sources includ-
ing copies via J. C. Kittel.
Two staves; in P 271 ‘alio modo’, melody-phrases cued ‘Choral’, final pedal
‘Ped.’.
Altnickol’s source may well have been revised well before c. 1740, and
some of the later copies also include BWV 665a instead (KB p. 84). The
technique is motif-imitation rather than fugal, though one can trace the
melody in the opening alto line. The chorale melody produces a thematic cell
(E B A B), but less patently than elsewhere. BWV 666 also resembles 665 in
that the soprano cantus for lines 2, 3 and 4 is anticipated in the tenor. Alt-
nickol might also have seen resemblances between the semiquaver figuration
of the two settings, and thus paired them.
As in BWV 665, each line has a different countersubject, well developed,
and leading to a fuller exposition of the last line, plus pedal point above
which a manual ‘cadenza’ is built upon the final countersubject, skilful but
somewhat distancing. As Spitta pointed out (I p. 602), the first semiquavers
in b. 10 look like an interlude-run between lines of a congregational hymn,
and are then developed further after the next interlude, even looking towards
the Canonic Variations (b. 18).
The impression is of a series of countersubjects, of quavers and semi-
quavers in different forms, rather as if four variations had been contracted
into one setting. In this, and the working towards running semiquavers and
final pedal point, the setting is much like BWV 712, though the actual semi-
quaver sextolets do not have the same shape, deliberately so perhaps, and
BWV 712 has a more original air to it. The second of the countersubjects
(b. 11) resembles many by Böhm or Walther and thus suggests an early and
more ‘objective’ treatment than BWV 665, one complete with an inversus
(bass, b. 12) but no obvious imagery?
BWV 666a Jesus Christus, unser Heiland

Copy by J. G. Walther.
Two staves; headed ‘alio modo’.

380 BWV 666a–667
Walther includes two extra directions to the player: to slur the opening
figure and to alternate hands in the cadenza figuration of b. 35. There is
also a different form to a motif in bb. 26, 28, 31 and 33, suggesting that it
was altered with hindsight after the cadenza was written. Or the differences
reflect a more complex situation of circulating versions.
BWV 667 Komm, Gott Schöpfer, heiliger Geist

No Autograph; copy by J. C. Altnickol (P 271); other copies via J. P.
Kirnberger.
Two staves; headed in P 271 ‘in Organo pleno con Pedale ob[bl]igato’.
The two sections do not amount to variations, such as Walther’s three-verse

partitas, nor do they patently correspond to vv. 1, 2 of the text, as Spitta
saw (I p. 601). There are examples of double chorales (fugal, then pedal
cantus) in Pachelbel and in the ‘Neumeister Chorale’ version of BWV 714,
but neither quite prefigures BWV 667:
1–8 = BWV 631. Changes in P 283 imply that BWV 631 came
first.
8–12 interlude, picking up semiquaver figures and rising to c
13–26 cantus firmus in four phrases on pedal, below loosely
imitative lines
One can only guess why an Ob chorale was expanded in this way, if it was,
and whether other chorales were ever so treated. Did Altnickol copy it here
after 1751 on the analogy of another expanded Ob setting since published
with the Art of Fugue (BWV 668a)? Walther’s copy of the longer version
has a change of handwriting for the second section, perhaps because he was
reflecting a change in his source, or because J. T. Krebs took over (? NBA
IV/5–6 KB p. 189), or of course both.
The accompaniment’s startling offbeat rhythm is twofold – lh then
pedal – both without known precedent except for faint precursors
in compound-time variations. The result is unforgettable, the stir of
Whitsuntide unmissable. It sends many organists in search of symbols,
though its insertion in P 271 is most striking for being out of style with the
381 BWV 667–667b
rest and therefore unexpected. The harmonic aura is consistently modal,

G-mixolydian with F and plagal cadences, diatonic but not quite regular,
with almost-parallel fifths in b. 2 and a second verse which uses the canto
in basso as a means of modulating further. Whether or not ‘tongues of fire’
are painted in the second verse (particularly at bb. 10, 26?), the continu-
ously rising semiquavers certainly return to the pentecostal clamour of the
opening chorale BWV 651, suggesting why either the composer or Altnickol
might want it there.
The second section does not pick up quite the same semiquaver figures
as the first but still gives the impression of an integrated work, with its many
semiquaver patterns of a kind associated generally with compound time, to
judge by the C major Prelude BWV 547. They begin in b. 9 as if improvised,
not unlike Pachelbel’s but rising to top C, then spilling over into various
shapes and settling on simple scale-figures which when imitated (bb. 20–1
etc.) resemble ‘Vater unser’ BWV 683 – or would if a ‘cantus firmus in organo
pleno’ did not ensure a different world. In Altnickol’s copy pedal is cued
only when its c.f. begins (b. 13), but it is needed earlier; perhaps the rests
and the cue mean that a pedal reed is to be drawn there.
The theme being in the bass, its harmony differs from BWV 631 and 370,
tending, as in other big works in the major, towards diminished sevenths
at cadences, here handled with great originality (bb. 18–19). Whilst the
hemiola in the last bar is unexpected, the final harmonies anticipate another
chorale in the G-mixolydian, BWV 678. (The Fugue BWV 541 has similar
associations for G major, including a final top and bottom tonic pedal point
plus a diminished seventh.) Whether the altos in the final bar consciously
enunciate B A C H is hard to know, but this second section does seem a
response to the first, as if the Holy Ghost were accepting the invitation, and
this were the end of the set.
BWV 667a Komm, Gott Schöpfer, heiliger Geist

Copy: J. G. Walther (SBB Mus. ms. 22541/3).
BWV 667b Komm, Gott Schöpfer, heiliger Geist

Copy: J. T. Krebs (P 801, fragment, formerly thought to be autograph), also
J. G. Walther (P 802, improved by J. L. Krebs after 1731).
Walther included BWV 667a as the first of six settings of the melody, by
Bach, Pachelbel, Zachow and himself, including in it some ‘improvements’
382 BWV 667b–668
found also in BWV 667 (not autograph). Whether minor differences be-
tween it and BWV 667b amount to a different version or are merely further
evidence for circulating variants is unclear (NBA IV/1 p. 95).
In its first part BWV 667b differs from both BWV 631 and 667 in having
a few unlikely semitone clashes in bb. 12 and 18 (miscopies?), a falling figure
at the end of b. 21, and parallel octaves in b. 13, the last so clear in P 801
(see NBA IV/2 p. vi) as almost to suggest that the cantus firmus was to have
been in the alto. At this point, P 271 gives the alto a crotchet rest: it could
be an ‘improvement’ or a sign that another copy was misread.
BWV 668 Vor deinen Thron tret’ ich (‘Leipzig Chorales’)

No Autograph; fair copy by ‘Anon Vr’ (Anon 12 = Altnickol’s wife Elisabeth,
née Bach?) in P 271: 25 12 bars only, on the lower part of the last page, after
and below BWV 769. No known copies.
Two staves; heading in P 271 ‘Vor deinen Thron tret ich etc’. No pedal cues.
The TEXT by von Hodenberg was published in 1646 as a hymn for ‘Morning,
Noon and Evening’.
Für deinen Thron tret ich hiermit Before your throne I now appear,
O Gott, und dich demütig bitt O God, and beg you humbly
wend dein genädig Angesicht turn not your gracious face
von mir, dem armen Sünder, nicht. from me, a poor sinner.
v. 15
Ein selig End mir bescher Confer on me a blessed end,
am jüngsten Tag erwecke mich Herr, on the last day waken me Lord,
dass ich dich schau ewiglich: that I may see you eternally:
Amen, amen, erhöre mich. Amen, amen, hear me.
The intervening verses contain prayers suitable for the dying.
The MELODY is that usually associated with ‘Wenn wir in höchsten Nöthen
sein’: Example 193 (see also BWV 641). Only BWV 668 gives this melody the
text ‘Vor deinen Thron’, which in Freylinghausen is associated with other
melodies. In what follows, BWV 668 and 668a are discussed together.
History
The history of this work has been conjectured as follows (Wolff 1991
pp. 282–94). ‘Wenn wir in höchsten Nöthen sein’ BWV 641 was written
383 BWV 668
Example 193
in the Orgelbüchlein in c. 1714, with a coloratura melody above three parts

developing a derived motif, nine bars long, no interlude. BWV 668a con-
tains BWV 641 but with its melody stripped of its coloratura and the whole
enlarged to forty-five bars by means of fore-imitations and interludes. This
was published posthumously at the end of the Art of Fugue as a compensa-
tion to the buyer for the incomplete final fugue, said in the preface to have
been ‘dictated extempore by the deceased man in his blindness to one of his
friends’.
In P 271 a somewhat ‘improved’ version – BWV 668 – was copied by
Anon Vr (a scribe known from MSS of 1742 onwards) on the blank staves
following BWV 769a, only six of which were drawn by J. S. Bach, probably
for BWV 769a. The manuscript now ends after 25 12 bars, at the bottom of
a page. Since there are directs for the next chord, and the last fascicle has
three sheets not four, there is probably a lost page on which the piece was
completed.
But all these steps are conjectural. Thus:
(i) if the ‘enlarged’ form of BWV 641 is the work of Bach, made either c. 1715 or
some thirty-five years later, it is unique. But sources do not prove which came
first, short or long version, or why one was made from the other, least of all
whether Bach made the enlarged version.
(ii) since nothing shows BWV 668 to be earlier than the final Leipzig years (KB
p. 96), perhaps it entered P 271 on the analogy of BWV 667/631, another ‘enlarged’
composition. Or vice-versa. In style it is close to another published chorale (BWV
687), and may well not have entered P 271 during Bach’s lifetime or on his
authority. Nor is it demonstrably his work: there seems little reason why a
competent pupil, if familiar with both BWV 641 and the Clavierübung III chorale
BWV 687, could not have concocted it. See p. 424 below.
(iii) the copy of the chorale said conjecturally by Forkel to be ‘dictated a few days
before his death to Altnickol’ (p. 53) is unknown and perhaps never existed.
Altnickol did copy the MS’s last chorale (BWV 667) and did write the title-page
(and title?) of the autograph MS of the Art of Fugue (P 200); so perhaps C. P. E.
Bach, knowing all this, drew conclusions he transmitted to Forkel, who had read
the Art of Fugue’s story about BWV 668a. But BWV 668a, had it been a deathbed
work, would surely have had the other title?
384 BWV 668
(iv) It is only conjecture that Anon Vr’s copy in P 271 derives from an ‘original
dictation copy’ (according to Kobayashi 1988 p. 64), or that it was added to
P 271 before the Art of Fugue appeared in print. Furthermore, the differences
between print and MS versions are hardly enough to speak of thorough-going
‘improvements’ in the latter, much less ‘emendations that elevate [this] final
version’ (Wolff 2000 p. 451) so as to give us an idea of Bach’s final, indeed dying,
pieties.
From a musical point of view, the story of the dictation (published in

1752) is doubtful: it looks like a biographical legend matching the ‘moon-
light’ anecdote of Bach’s infancy (published in 1754). In view of BWV 641,
how can the composer have dictated BWV 668 ‘on the spur of the moment’?
Conceivably, he could give directions for de-embellishing the melody, but
composing the new sections by dictation seems out of the question, despite
their being so like another chorale (BWV 687). The most I can imagine
is that the ailing, sight-impaired Bach had a chorale played over (by his
daughter?) and suggested a few changes, perhaps but not necessarily in
readiness for a copy to be inserted into P 271.
The music
The work’s special associations have made realistic appraisal of it difficult.
In most references, ‘in his blindness’ has become ‘on his deathbed’, which
the original anecdote need not have meant. Already in 1754, the work was
invoked to do battle with the ‘champions of materialism’ as an instance
of miraculous human endeavour (Dok III p. 73). Forkel heard it as ‘the
expression of pious resignation and devotion’; and more recent enthusiasts
find deep mystical or numerological references (see Smend 1969 p. 173).
But as an enlargement of an Ob nucleus, it has problems.
In some respects simpler than BWV 641, it is old-fashioned in form
and its counterpoint tends to the commonplace. Like BWV 687, it has been
composed so as to dispense with pedal – not indicated in P 271 – and cast
in old-fashioned form:
cantus in soprano; fore-imitation and interludes based on motifs derived
from each phrase of the cantus in turn, imitated inversus
each phrase comes to a complete close (unlike e.g. BWV 652–654)
the final episode augments and inverts its theme
Much of this is found also in BWV 687, but in being plainer BWV 668 seems
less mature. The rhythmic interest of its lines is weaker than BWV 687’s, just
as the harmony of its interludes is less original than BWV 641’s; compare,
for example, bb. 21–7 of BWV 668a with b. 5 of BWV 641. The ends of
the cantus phrases, particularly bb. 22 and 31–2, are more run-of-the-mill
than the preceding bars, rather as if subsequently added. There are also
385 BWV 668–668a
inconsistencies, such that the suspensions and accented passing-notes of

b. 17 seem maturer than the following bar, with its simple cadence.
The similarities to BWV 687 in the texture, form, inverted fugal answers,
plain cantus firmus in soprano, clearcut opening of sections, long-held note
and details of the cadence, are matched by the differences: BWV 687 has
a more modern key and metre (2/4). In BWV 668, it is difficult to believe
that b. 14 or b. 37 was written (drafted? revised? dictated?) nearly half a
century after Pachelbel’s death, even as a farewell salute. Despite this, one
must recognize that the last thirteen bars in particular are not only a web of
thematic allusion but have a touching euphony and are harmonically astute
(b. 40), with a rich myxolydian or plagal cadence hard to attribute to anyone
but J. S. Bach. But at what age?
BWV 668a Wenn wir in höchsten Nöten sein

(Die Kunst der Fuge)
Published c. 1751, no Autograph MS; copies derive from print.
Four staves (open score, four different clefs, fifth part at end on lowest stave);
headed in the Art of Fugue ‘canto fermo in canto’; described in preface as
‘worked-out church chorale in four parts’ (‘vierstimmig ausgearbeiteten
Kirchenchorals’).
BWV 668a differs in its title, new (as for the final chorale of another recent
set, the Schübler); in its notation (now in open score as commended in
Marpurg’s preface to the 1752 edition); in being complete (so leaving it
uncertain whether BWV 668 was to have had the same forty-five bars); and
in certain details –
9 tenor a written as two tied quavers in BWV 668a, untied in b. 41, but a
crotchet in BWV 668. (This implies that the original untied quavers in
the Ob b. 2 were not recognized as derived from the theme.)
26 rh quavers dotted in BWV 668 (like first beat in bass in b. 9)
7 imitative semiquavers in tenor in BWV 668
10 interrupted cadence in BWV 668
These differences have been interpreted as the composer’s final improve-

ments (Wolff 1991 p. 292), meaning that the printed chorale represents an
earlier, less polished version than a manuscript benefiting from dictated
386 BWV 668a
revisions. But these differences are minor, perhaps creeping in as the

open score was engraved, perhaps (in the case of the cadence in b. 10) in
error.
Nothing in the Art of Fugue indicates that this is an organ piece, nor is
pedal needed. It has surely been made to be playable by hands alone? – see
the last three bars.
Chorales from Clavierübung III BWV 669–689
Published 1739. Title-page:

Dritter Theil der Clavier Übung bestehend in verschiedenen Vorspielen
über die Catechismus- und andere Gesaenge, vor die Orgel: Denen
Liebhabern, und besonders denen Kennern von dergleichen Arbeit, zur
Gemüths Ergezung verfertiget von Johann Sebastian Bach, Koenigl.
Pohlnischen, und Churfürstl. Saechss. Hoff-Compositeur, Capellmeister,
und Directore Chori Musici in Leipzig. In Verlegung des Authoris.
Third Part of the Keyboard Practice, consisting of various preludes on the
Catechism and other hymns for the organ. Prepared for music-lovers and
particularly for connoisseurs of such work, for the recreation of the spirit,
by Johann Sebastian Bach, Royal Polish and Electoral Saxon Court
Composer, Capellmeister and Director of the chorus musicus, Leipzig.
Published by the Author.
Two groups of engravers worked on the volume: one in Nuremberg (a single

engraver, paper made in Nuremberg) for thirty-five pages including the
title-page, one in Leipzig (three engravers, Leipzig paper) for forty-three
pages including pp. 1–18, i.e. those once thought to have been engraved by
Bach. The engraving process probably used at least in part the autograph
manuscript itself, through which to trace the image on to the plates, an
operation damaging the paper beyond recall.
Two pulls or identical editions can be inferred (Butler 1990 p. 79), sold
for 3 Reichsthaler. For comparison, a new clavichord in 1745 might cost
only 10 (Dähnert 1962 p. 230). In 1740, Mizler’s translation of Fux’s Gradus
sold at 2, in 1751 the first edition of the Art of Fugue at 5, the second at 4.
The copy in SBB has minor corrections by the composer, and copies
in London and Vienna have been called ‘control copies from which Bach
compiled a list of corrections’ (Butler 1990 p. 129), though not exhaustively.
All major MSS are direct or indirect copies of the print, complete or incom-
plete, some still being made in the nineteenth century. Some copies have
a different order for the last four chorales, and several include corrections
that may derive from a lost autograph.
The period
The volume appeared towards Michaelmas 1739 (29 September), although
[387] J. E. Bach had thought it might be ready for the Easter Fair (Dok II
388 Clavierübung III
p. 335).∗ The year 1739 saw three Reformation festivals in Leipzig: 25 May
(bicentenary of Luther’s sermon in St Thomas), 12 August (bicentenary
of Augsburg Confession) and 31 October (Reformation Day). Perhaps the
composer played some or all pieces on his visit to the new organ in Altenburg
Castle in September 1739 (? Dok II p. 368).
Erased page-numbers and other details on the engraved plates sug-
gest that the plan evolved, beginning with the Kyrie–Gloria and larger
catechism settings (first BWV 676?), then the Prelude and Fugue in E
(already composed? added with the manualiter settings in 1738?), and
finally the Duets in mid-1739. Except for the trio ‘Allein Gott’ BWV
676a (but q.v.) all the pieces seem to be new, but only hypothetically
did work begin so soon after Clavierübung II (1735). The plan of the
work and its publication were surely prompted by other publications:
Kauffmann’s Harmonische Seelenlust 1733–6 (in which Walther too was
involved), C. F. Hurlebusch’s Compositioni musicali 1734–5 (see BWV 552),
H. F. Quehl’s two chorales 1734, Walther’s Allein Gott in der Höh’ sei Ehr’
1736, J. C. Vogler’s Vermischte Choral-Gedanken 1737, and even old French
Livres.
The title-page’s note on connoisseurs does not appear for Clavierübung
II, though the two title-pages otherwise correspond and are typical of their
time. Kauffmann too promised ‘delight for high and lowly lovers of music’
(‘allen hohen und niedern Liebhabern . . . Vergnügen’) as well as useful
service music. Quehl noted that his were in part fugal, in part for two
manuals and pedal on three staves.
The title of Saxon court composer in late 1736 made it appropriate for
Bach to compile some elevated organ music equivalent to his recent compi-
lation of elevated vocal music – Kyrie and Gloria for organ, matching those in
the B minor Mass – especially since W. F. Bach was then organist in Dresden.
Perhaps some of it originated for an organ recital in the Frauenkirche on
1 December 1736, or for Friedemann’s repertory at the Sophienkirche,
although both those Silbermann organs would have made the book’s
remoter keys problematic. To conjecture further: perhaps it was on some
such occasion, and with such music, that Bach found Silbermann’s tuning
not to suit ‘today’s practice’ (Dok II p. 450).
Context
Clavierübung III was Bach’s first publication for organ, respectfully received
by younger contemporaries such as Lorenz Mizler:
∗ Finalized while the composer was fifty-three? – the titlepage has fifty-three words.
Der Herr Verfasser hat hier ein neues Exempel gegeben, dass er in dieser
Gattung der Composition vor vielen andern vortrefflich geübet und
glücklich sey . . . Dieses Werk ist ein kräfftige Widerlegung derer, die sich
unterstanden des Herrn Hof Compositeurs Composition zu critisiren.
(Dok II p. 387)
The author has given here new proof that in this kind of composition he
excels many others in experience and skill . . . This work is a powerful
refutation of those who took it upon themselves to criticize the Court
Composer’s music.
The last remark must refer to the attack made on Bach by J. A. Scheibe in 1737
(see below), although Scheibe had not specified organ music and it is hard
to see how such complex music could be ‘Bach’s rebuttal to Scheibe’s barb’
(Butler 1990 p. 17) – rather the contrary. Perhaps Mattheson’s remarks in
1739 on the limits of modern organ music prompted a monumental survey
(Butler 1983), though this may over-estimate Mattheson’s influence as well.
More likely is that really fine music such as Frescobaldi’s Fiori musicali
wielded lifelong influence on Bach and would inspire him to produce his
own Kyrie settings. (In the same way, Fiori musicali’s Bergamasca was surely
to influence the Goldberg Variations quodlibet, 1741.)
The title too saluted tradition:
J. Kuhnau, Neue Clavier Übung I (Leipzig, 1689), II (Leipzig, 1692)
J. Krieger, Anmuthige Clavier-Übung bestehend in unterschiedlichen Ricer-
carien . . . (Nuremberg, 1698)
Clavir Übung Anno 1709, MS album of J. C. Bach (Gehren, 1673–1727)
V. Lübeck, Clavier Übung (Hamburg, 1728)
G. A. Sorge, Clavierübung . . . sowohl auf der Orgel, als auf dem Clavicymbel und
Clavicordio mit Vergnügen zu hören (Nuremberg, c. 1739)
Sperontes singende Muse . . . Clavier-Übung und Gemüths-Ergötzung, I–IV
(Leipzig, 1736–46)
Much of Bach’s wording, as on his other Clavierübung title-pages, is close to

Kuhnau’s, the first of which appeared in Leipzig exactly fifty years earlier, to
the day perhaps. Moreover, Kuhnau’s second volume distinguishes between
beginners and those knowledgeable enough to find in its fugues material
for further contemplation. The term Clavierübung was probably coined by
him as a quasi-translation of musica prattica in earlier seventeenth-century
Italian publications.
While Clavierübung III is clearly not merely a miscellaneous album,
its nature has been in some dispute, whether it is a ‘closely knit group of
pieces’ or actually in one way or another a ‘cycle’. That the volume was being
expanded in the course of being engraved would not necessarily explain why
the Prelude and Fugue are separated, why the Duets were included, or why
the title-page mentions neither.
Overall plans for published collections reflected practical needs in the

Mass (twenty-one pieces in Couperin’s Messe c. 1690) and Office (Kerll’s
Magnificat versets in Modulatio organica 1686), or demonstrated learned
counterpoint (Buxtehude’s Hinfarth 1674, known to Walther). Parisian
Livres often included Vespers movements, as in Grigny’s book mentioned by
J. A. Birnbaum in 1736 when defending Bach against the Scheibe criticism
(Dok II pp. 304f.). Clavierübung III’s French, Italian and German music
harks back to Ammerbach’s Tabulaturbuch, a Thomaskantor’s publication
promising German, Latin, Italian and French pieces.
Since the engraver Krügner also worked on Kauffmann’s Harmonische
Seelenlust, it is likely that Bach was responding to such local chorale-settings,
using their styles to new ends. And since the E Fugue shares minor details
with a fugue by Hurlebusch, perhaps it was composed in response to it, not
originally for Clavierübung III? On its key, see notes to BWV 552.
Textual plan
Though perhaps only by chance do the twenty-one chorales recall twenty-
one movements of a French Mass, the collecting together of Mass and
Catechism settings represents the two main religious observances on a
Leipzig Sunday (Humphreys 1994 p. 48): the Main Service and the after-
noon Catechism. In the Leipzig hymnbook of G. Vopelius, the Missa or
Kyrie plus Gloria is in the section ‘of the Holy Trinity’, and Clavierübung III
has many threes. Hymns sung every Sunday such as ‘Allein Gott’ or ‘Wir
glauben’ gave the organist opportunity to make use of different keys, as Ad-
lung noted (1758 p. 726); he reports the Gloria hymn being played in the keys
of E, F, F, G, G, A and B, three of which are found in Clavierübung III.
Since the Leipzig Catechism Examination itself did not use organ (Stiller
1970 p. 242), the settings must have served other purposes, not least as
a personal gesture of orthodoxy, something set against a background of
penitence. Luther’s reformed liturgy included Kyrie, Christe and Gloria
just as his reformed doctrine centred on Ten Commandments, Credo,
Prayer, Baptism, Penitence and Eucharist. Both Catechisms consisted of
a series of questions and answers outlining the principles of faith, and
from these could be drawn six headings, introduced by the German Kyrie
and Gloria. Perhaps the Penitence hymn, BWV 686, belonged to an early
phase when the stile antico settings of the Kyrie took shape (Butler 1990
p. 16).
The six principal sections of such ‘evangelical song catechisms’ were used
for morning assembly in Thuringian schools (Trautmann 1984), the seventh
day being Sunday, with Kyrie and Gloria. All six hymns are in Luther’s
hymnbooks, and the melodies of Nos. 1–4, 6 can even be combined in a
quodlibet (Hilgenfeldt 1850). As representing six pillars of orthodoxy, they
were important not only in the Jubilee Year 1739 but – since half the hymn
melodies were of Gregorian origin – as an answer to the Saxon Consistory’s
directive of 1730 that ‘new hymns . . . shall not be used in public divine
services’ without permission (David and Mendel 1945 p. 119). In such
respects, Clavierübung III is more a tribute to Saxon Lutheranism in Leipzig
than to school-catechisms in Thuringia.
The early reformers offered the Bible, the hymnbook and the Catechism,
and Bach, after setting Bible texts and collaborating in Schemelli’s hymn-
book, was now to supply the Catechism. Perhaps such texts were a reaction
to the pietist flavour of Schemelli’s hymnbook? Ergötzung implied a pious
‘recreation’ of the spirit and was common on title-pages, including Vetter’s
Leipzig collection thirty years earlier.
Musical plan
One aspect of ‘practical music’ is that each lesser setting develops a particular
kind of fugue, and yet only the last (BWV 689) resembles anything in WTC2,
then being or about to be assembled. There are musical schemes here beyond
mere ‘esoteric brooding’ (Albrecht 1969 p. 46), for the volume is a careful
compendium, with a systematic musical variety and cyclic elements clear to
the reader if not player:
552.i Praeludium pro organo pleno E

669 Kyrie, Gott Vater c.f. in soprano G
670 Christe, aller Welt Trost c.f. in tenor C (G?)
671 Kyrie, Gott heiliger Geist c.f. in pedal (pleno) G
672 Kyrie, Gott Vater 3/4 manualiter E
673 Christe, aller Welt Trost 6/4 manualiter E
674 Kyrie, Gott heiliger Geist 9/8 manualiter E
675 Allein Gott in der Höh’ trio, manualiter F
676 Allein Gott in der Höh’ trio, pedaliter G
677 Allein Gott in der Höh’ trio, manualiter A
678 Diess sind die heil’gen zehn Gebot’ c.f. in canon G
679 Diess sind die heil’gen zehn Gebot’ fugue, manualiter G
680 Wir glauben all’ an einen Gott à 4, in organo pleno D
681 Wir glauben all’ an einen Gott fugue, manualiter E
682 Vater unser im Himmelreich trio + c.f. in canon E
683 Vater unser im Himmelreich non-fugal, manualiter D
684 Christ, unser Herr, zum Jordan kam à 4 c.f. in pedal C

685 Christ, unser Herr, zum Jordan kam fuga inversa, manualiter D
686 Aus tiefer Noth schrei’ ich zu dir à 6, in organo pleno E
687 Aus tiefer Noth schrei’ ich zu dir motet, manualiter F
688 Jesus Christus, unser Heiland trio, c.f. in pedal D
689 Jesus Christus, unser Heiland fugue, manualiter F
802 Duetto I 3/8, minor E
803 Duetto II 2/4, major F
804 Duetto Ill 12/8, major G
805 Duetto IV 2/2, minor A
552.ii Fuga pro organo pleno E
Note the organo pleno framework, the three inner groups (Mass, Catechism,
Duets), three genres for the Mass (three polyphonic, three manualiter, three
trio), pairs for the Catechism (two canonic c.f., two pedal c.f., two pleno, and
pedaliter/manualiter pairs), and total variety in the Duetti. The opening and
the close are both somewhat French, almost ballet-like: an entrée and a gigue.
The Praeludium passes to the Kyrie quite as aptly as it does to the eventual
Fuga, for both of these begin on b.
Significant threes abound: settings of the Trinity hymn (all in three parts,
the keys forming a major third F G A), themes in the opening Prelude,
sections in the closing Fugue; three flats, parallel thirds, Clavierübung ‘Third
Part’. Other allusions are more conceptual than perceptual: number of Mass
chorales (3 × 3), total number of pieces (3 × 3 × 3, like the twenty-seven
books of the New Testament or entries in the Fugue), progressive triple time
in the manual Kyries. Authors find numerological reference to religious
belief (see Lohmann, EB 6588), in the proportion between sections of the
Kyries with c.f. and without (Humphreys 1994 pp. 43f.) and in patterns
created by playing with the number 27, sub-groupings, cross motifs, and
Lutheran texts (Clement 1999 passim). Three is bound to be significant: the
Ob had already used a digit 3 for the word drei in the unset title ‘Der du
bist drei in Einigkeit’. The dogma of the Trinity would have been one of the
things Bach was examined in when taking up the Leipzig cantorate (see BJ
1998 p. 29).
But there are also purely musical significances. Though a fughetta, the
central piece of the collection (BWV 681) has the typical rhythms of a French
Overture, as does the central piece in all other parts of the Clavierübung:
the opening of Partita No. 4, the first movement of the B minor Ouverture,
and Goldberg Variation No. 16. That the four are in the related keys of D,
B minor, E minor, G, suggests a level of organization musical rather than
symbolic. This may also be the case with the Duetti in E F G A, the very
notes of Walther’s tetrachordum excellentium (1732 p. 600).
Musical idiom
It is always possible that the composer intended the chorales as service pieces
for Lutheran organists. But as with the Canonic Variations, Musical Offering
and Art of Fugue, the technical demands put it out of the way of most players,
and both musical idiom and organization evince more the private labours
of a pious composer.
There is an unconventional, even strange, quality about the counter-
point, whether modal or diatonic. Already in the 1770s Kirnberger noted
that only the Trinity trios were firmly in major keys (Dok III pp. 221, 583), a
consequence of their hymn-melody, perhaps, though in practice BWV 677
is hardly more diatonic than BWV 674. Both are ambiguous in their first
bar: if BWV 674 is in G, why does it begin on the mediant? If in E minor, why
a supertonic answer? If B minor, why the C? If modal, how is there such
a diatonic modulation as bb. 18–22? The so-called modality lies in a kind
of diatonic ambiguity suggested by the key-signature and expressed in the
cadence. Because of the key-signatures there are no accidentals in the print
for any of the volume’s cantus firmi except for an occasional leading note,
surely not an accident. In BWV 677, the subject is tonally uncertain (unlike
its other manifestation in the C major Fugue BWV 547, which is unam-
biguously diatonic), as are the mediant harmony and mediant entries in
bb. 7–8. In general, mode is far more pronounced than in Telemann’s XX
Kleine Fugen of c. 1730, described as ‘composed according to particular
modes’, but without such a mediant progression as to make the key tem-
porarily ambiguous, as in the Kyrie BWV 670 at bb. 55–6.
Techniques are systematically surveyed: fugue, paraphrase, canon, ritor-
nello, motif development, invertible counterpoint, cantus firmus. Thus the
three settings of ‘Allein Gott’ are a manual trio with inner cantus, a trio-
sonata-like movement with partial cantus, and a fughetta based on the first
two lines without cantus. None of them, however, could offer a template to
other composers such as Pachelbel’s preludes had done, nor is the volume
a compendium of all up-to-date treatments. Indeed, it could be that the
Schübler Chorales were meant to make up for this deficiency by offering
more tuneful models than BWV 678, 686 or 688.
The intention to develop distinct styles is clear in the five stile antico
pieces. Contemporary musicians on whom Palestrina’s influence is most
direct, notably Fux, Caldara and Zelenka, were said to be admired by J. S.
Bach (Dok III p. 289), who seems to have acquired his own copy of Fux’s
Gradus soon after it was published in 1725 (Wolff 1968 p. 28). Also, his
pupil Mizler translated it in 1742 (‘very well’ according to Schering 1941
p. 202), lecturing on it in the university. While the clearest sign of the
stile antico is the larger note-values, the style meant a stricter polyphony
394 BWV 669
than usual in alla breve keyboard music, one suiting the ambiguous modes.
That ‘antico’ is at least partly a question of notation does not lessen its
significance, since much of the pattern-making in the late works is indeed
notational.
But even Clavierübung III’s pieces in stile antico are not strict textbook
demonstrations. While BWV 686 may tend towards a more polyphonic tex-
ture than Cantata 38’s treatment of the same melody, the rhythms, compass
and intervals are not much more Palestrinian. Neither in Palestrina nor
in Fux is one likely to find sequences such as one does in BWV 671, and
yet the opening of the E Fugue is more like species counterpoint than
e.g. the Fugue in F BWV 540 or older Italianate works like the Canzona and
Allabreve. A relationship with Frescobaldi is suggested by the Fiori musicali’s
stated purpose (‘mainly to assist organists’ in Mass and Vespers), shape
(a free piece before and after liturgical movements), details of polyphony
(stile antico counterpoint, many cantus firmi, new countersubjects, pedal
points), and other technical details (mutation and combination of themes,
quasi-ostinato bass).
But can more than a few connoisseurs have had their spirits refreshed by
the volume? G. A. Sorge’s Vorspiele (Nuremberg, c. 1750) provided simple
three-part settings because such chorales as Clavierübung III were ‘so diffi-
cult and almost unusable by beginners’. The Choräle by the Weimar pupil
J. C. Vogler (1737) were composed ‘principally for those who have to play
in country’ churches. J. L. Krebs’s Klavierübung II (1741) was made to be
playable ‘by a lady, without much trouble’. In view of all this, those who
agreed with Scheibe in 1737 that Bach
seinen Stücken durch ein schwülstiges und verworrenes Wesen das

Natürliche entzöge, und ihre Schönheit durch allzugrosse Kunst
verdunkelte. (Dok III p. 280)
deprived his pieces of all that is natural by giving them a bombastic and
confused character, and eclipsed their beauty by too much art
could have found examples in this volume, for the very mastery has a for-
bidding air.
BWV 669 Kyrie, Gott Vater in Ewigkeit

Two staves; headed ‘Canto fermo in Soprano’, ‘a 2 Clav. et Ped.’
The TEXT is one of three sections published in early Lutheran hymnbooks

as a version of the troped ‘Kyrie summum bonum: Kyrie fons bonitatis’
395 BWV 669
(Liber usualis, Mass II, Feasts of the 1st Class, I). Each Sunday in Leipzig,
the German or Latin text was sung after an organ prelude (Stiller 1970
p. 103). Strictly, the ‘Kyrie summum’ was sung from Trinity to Christmas,
the similar ‘Kyrie paschale’ from Easter to Trinity.
Kyrie, Gott Vater in Ewigkeit, O Lord – God Father in eternity,

gross ist dein Barmherzigkeit; great is your mercy;
aller Ding’ ein Schöpfer und Regierer, sole creator and ruler of all things
eleison! – have mercy!
The MELODY adapts the plainsong (Terry 1921 p. 250), its three sections
sharing a second half: Example 194. Bach’s five c.f. paragraphs are as in
hymnbooks (cf. BWV 371). The melody is used only in BWV 672 and
BWV 233, and organ settings are rare; Scheidt’s Tabulatura nova III (1624)
uses another melody.
Example 194
The three massive 4/2 Kyrie preludes are both unique and keyboard-like
though related to vocal works, in motif, inversion techniques, c.f. style and
the tripartite plan (Mass in F, BWV 233). The form can be described as
‘organ motet’, irregular in being monothematic; fugal theme from the first
two lines of the cantus, which is augmented, line by line, in the top part
(‘God the Father’); all regular entries dependent on the cantus
and its style as
three-part alla breve counterpoint plus cantus, strictly antico; modal

(G-phrygian), with ambiguities (e.g. B/E major bb. 29–35).
The three settings refer back to such works as the versets of Frescobaldi’s
Fiori musicali or to those they influenced (e.g. ricercari of J. K. F. Fischer)
396 BWV 669
rather than to the usual German motet-chorales. Though generally similar

to the ‘Confiteor’ from the B minor Mass, the latter’s motifs are livelier. The
c.f., moving entirely by step, gives the piece a characteristic smoothness by
no means out of place in a movement that follows (and could be paired
with) the Prelude in E.
Stile antico features are:
4/2; modal cantus (opening imitation ‘unrelated’ to the final cadence);

constant inversion and stretto (so a ‘tight’ fugue); antique subjects; many
suspensions, dactyls, crotchet lines moving by step; canon sine pausa in the
final bars.
None of the latter features are exclusive to this style. But although the pedal
is often frankly bass-like (e.g. bb. 17–18), the parts are unusually strict:
free phrases like the quavers of b. 36 are more in character with other
movements (e.g. E Prelude, b. 70). There are fourteen entries of the theme
(with two partial entries) and seven inversions; the seven stretti include
a rectus/inversus stretto (bb. 19–20). Subsidiary ideas are developed very
largely from implied suspensions, the dactyls and rising crotchets from the
theme. The last are strikingly ‘effortless’ in the working-out, and sometimes
amount to sub-themes (b. 32).
The whole is developed below a c.f., whose last note each time could be
held longer than notated. The moulding of c.f. into fugue-subject involved
little paraphrase, since plainchants naturally served as ricercare subjects.
While the cantus phrases of both BWV 669 and 670 are played on a separate
manual, the counterpoint is more complete than is often the case with such
movements: so plein jeu for the accompaniment, jeu de tierce combination
for the solo? Reserving reeds for the cantus in pedale suited organs of the
period with no strong manual reed.
Fux’s Gradus and Mizler’s comments on it suggest that the intention
behind stile antico was to present music ‘grounded on the unchangeable
rules of harmony’:
BWV 669 monothematic, ricercare-like, vocal polyphony

BWV 670 cantus firmus en taille, given freer treatment
BWV 671 several subjects combining in turn with the c.f.
Whether it was meant to evoke more – the strength of faith, confessional

orthodoxy – is uncertain, though in being so adaptable the solidity of the
style does suggest such things. Common to all three movements is a certain
seamless motion that rarely leads to full cadences or sequential repetition,
both of which would be more diatonic than suits the desired transcendental
style.
397 BWV 670
BWV 670 Christe, aller Welt Trost (Clavierübung III)

Published 1739, no Autograph MS.
Two staves; headed ‘Canto fermo in Tenore’, ‘a 2 Clav. et Pedal.’
The TEXT is the second section of ‘Kyrie, Gott Vater in Ewigkeit’.
Christe, aller Welt Trost, Christ – consolation of all the world,

uns Sünder allein du hast erlöst; you alone have redeemed us sinners;
Jesu, Gottes Sohn, Jesus, Son of God,
unser Mittler bist in dem höchsten you are our mediator at the highest
Thron; throne;
zu dir schreien wir in Herzens to you we cry in our heart’s desire, – have
Begier, – eleison! mercy!
The MELODY is adapted from the plainsong; its eight paragraphs are as
in hymnbooks, except that the second is divided into two (bb. 14–16 and
20–2), on the analogy of bb. 33–5 and 39–42 (a traditional division). From
b. 39 to the end, the c.f. is virtually the same as that of BWV 669, including
the ornaments.
The form, style and features of the stile antico are those of BWV 669, with
cantus in the tenor (‘God the Son’, middle Person of the Trinity). Twenty-two
entries of a theme derived from the first two lines of the cantus are coun-
tered by only one inversion (b. 43), perhaps because the theme’s angularity
is more obtrusive inversus. There is the same ease of counterpoint based
on smooth transitions, suspensions and counter-rhythms, while phrygian
features result again in some ambiguity of key, especially in the first twenty
bars. B major at the opening makes it appear that the subject enters on
the submediant, and only when the music moves elsewhere (bb. 19–22) is
there a clear perfect cadence; even the diatonic versions in soprano and bass
bb. 52–4 (compare with bb. 14–16, 20–2, etc) give no firmer sense of key
or prepare the final cadence.
The entries are variously disguised (e.g. alto bb. 6–7, soprano b. 28,
soprano b. 52 with alto stretto), and again, much is made of the dactyl motif
taken from the theme, with some passages serving almost as a model of
italianate counterpoint (bb. 17–18), in spacing and tessitura far removed
from pure stile antico (bb. 23–4). The double entry in sixths (b. 32) is a classic
canon sine pausa, interpreted as concords expressing the idea of ‘mediator’
(Chailley 1974 p. 178). The motifs countering the themes in such bars as
46–51 are typical of stile antico (Example 195): the figura corta (i), tirata
mezza (ii), and circolo mezzo (iii), as well as Third Species crotchets (iv). The
crotchets produce a pedal line of wider ambitus than the other voices.
398 BWV 670–671
Example 195
The spacing is noticeably different from the preceding Kyrie’s because

the soprano is now independent of the c.f. en taille: the wide spacing of
bb. 31–2 has no parallel in BWV 669. Nevertheless, there are bound to be
such similarities as the crotchet motif in BWV 669 (b. 32, tenor plus inversus),
BWV 670 (b. 43, alto and bass rectus) and BWV 671 (bb. 32–3, alto sequence).
The pedal part of BWV 669 and 670 is especially ‘modern’, requiring a
confident technique as much as the trio or canons of Clavierübung III,
where it can afford to be more consistently détaché than here.
BWV 671 Kyrie, Gott heiliger Geist (Clavierübung III)

Two staves; headed ‘à 5’, ‘Canto fermo in Basso’, ‘Cum Organo pleno’.
The TEXT is the third section of ‘Kyrie, Gott Vater in Ewigkeit’.
Kyrie, Gott heiliger Geist, O Lord – God, Holy Ghost,

tröst’, stärk’ uns im Glauben comfort and strengthen us in faith
allermeist dass wir am letzten End’ most of all that at the final end we may
fröhlich abscheiden aus diesem Elend, depart joyfully out of this misery
– eleison! – have mercy.
The ‘Kyrie paschale’ version of the third section (see BWV 669) lists the
attributes of the third Person of the Trinity: wisdom, faith, love, justice.
The MELODY adapts the plainsong, the six paragraphs corresponding to

the divisions in hymnbooks. From b. 34 to the end, the c.f. is virtually the
same as that of BWV 669, but without ornaments and now played with
pedal reed.
The form, style and features of the stile antico are like those of BWV 669,
but with a new kind of texture from the five parts: two sopranos (unusual)
and c.f. in the bass (‘God the Holy Spirit’ – see also BWV 651 and 667).
The subject of the fugue (now in four parts) comes again from the first two
lines of cantus, and is answered in octave stretto by its own inversus which
accompanies it on all its entries – a stile antico idea. This pairing of rectus
and inversus is such that neither theme alone accompanies the c.f., though
either could (e.g. in b. 24).
399 BWV 671–672
The note-patterns developed in the episodes (which the c.f. has now
become) are less like Fux’s than those in BWV 670, and there are se-
quences like some of those in WTC2 (Fugue in B major). As in the first
Kyrie, a simple stepwise theme pervades the whole texture, as do classic
syncopations derived from it and its answer. Continuous quavers from
b. 37 to b. 54 both incorporate syncopation and move by step: each
quaver idea leads to the others, with a fresh beginning at b. 50 and an
abrupt end at b. 54. Though again G-phrygian, it all seems more diatonic
because of the sequences, full harmony and bass cantus: the penultimate
cantus phrase (bb. 43–5) opens more firmly in a key than in BWV 670 (bb.
46–8). But any anchoring effect the final phrase’s B has is countered
by the chromatics.
These chromatics are a marvellous surprise. Since themes combined
with their inversus recall Frescobaldi (Kyrie 3 from ‘Messa della Madonna’
in Fiori musicali), perhaps the Italian chromatic toccata was also behind this
coda, with its durezza suspensions and rising/falling semitones. Chromatics
for ‘eleison’ (see Frescobaldi’s Kyrie 3) are known in other stile antico music
(B minor Mass, Kyrie 2) and generate the final phrygian cadence A/G. The
strict five parts develop semitone drops and diminished thirds, beginning
in double stretto and resulting in a unique progression over bar-line 59–60:
a Neapolitan sixth changed beyond recognition.
Although the passage seems to anticipate moments in the Ricercar à 6
from The Musical Offering, the five-part writing here is quite different –
chromatically slipping lines, few thirds, more incidental dissonance and a
less clear tonality. It looks like a fuller version of the sudden chromatic end
to the last verse, in five parts with pedal c.f., of the ‘O lux beata’ of Matthias
Weckmann (†1674, whose son had been organist of St Thomas, Leipzig).
The effect is a repeated cry of ‘have mercy’, its twelve chromatic steps more
‘vocal’ than e.g. falling chromatics at the end of the Chromatic Fantasia for
Harpsichord.
BWV 672 Kyrie, Gott Vater in Ewigkeit

Two staves; headed ‘alio modo’, ‘manualiter’.
Although the three lesser Kyrie preludes are sometimes called fughettas –
not by J. S. Bach! – their form and character are unconventional; nor could
they be mistaken for ‘South German versets’ to which they have been likened
400 BWV 672–673
(Kube 1999 p. 591), for again an original idiom, in part modal, produces a
new, transcendental quality.
The cantus supplies not only the E-phrygian tonality but the cadences in
bb. 6–7 and 28–9 and much of the material (see Example 196), including the
falling motif in b. 8 (top part, an inversus?). From these motifs are woven four
smooth contrapuntal parts shot through with many and subtler allusions to
the theme’s notes than one expects even of the mature Bach: any two rising
minims or even quavers are probably allusive. Four parts create clear keys
(G, A minor, D minor, A minor, E-phrygian) and there is a curious, remote
sweetness from so many thirds duplicating the subjects.
Example 196
The doubled F in b. 5 was noted by Kirnberger as an example of how ‘this

great man departs from the rule in order to sustain good part-writing’ (Dok
III p. 217). Such passages, slight in themselves, certainly give character to
the movement, as do similar progressions in the B Fugue WTC2 (bb. 64–6
etc.), and as does the liquefying effect of the triple time itself. Chromatic
semitones in bb. 8, 9, 25 and 31 add further smoothness and are quite
without the shock effect of the previous chorale, despite the parallel thirds
they have in common. For all three smaller settings to begin in the major
and end in the minor is a modal gesture, with the final E approached in
three different ways (F–E, D–E, A–E).
BWV 673 Christe, aller Welt Trost (Clavierübung III)

That the lesser Kyrie preludes are a group is suggested by their ‘progressive’
time-signatures, common cadences on E, and unconventional fugal form;
also, BWV 673 and 674 lack separate headings. Again the cantus supplies
material in a complex of allusions (see Example 197), inspiring a movement
of immense subtlety. Section a serves also as first countersubject, motifs
b, c and d can be often discerned, and the little semiquaver figure that
becomes more and more prominent originates in falling phrases of the
401 BWV 673–674
cantus, to which however there is no simple, obvious reference even in the

closing bars. Nor does the composer allow motivic development to govern
the movement, as in BWV 672: these thirty bars have an original, almost
capricious shape quite different from the previous setting.
Example 197
Thematic allusions include the irregular bass entry of bb. 2–3 and the
rising fourths in b. 23, and much again depends on parallel thirds and
sixths, now in a freer texture than BWV 672. The entries are very original,
on all degrees of the scale but G, in stretto (bb. 18–19, disguised in NBA
by direction of stems), canon cum pausa (bb. 11–13) and canon sine pausa
(bb. 20–1, 24–5). The lilt given the piece especially by motif d is matched by
the semiquaver figure; both are second nature in 6/8 time, and b. 22 looks
as if it might develop a familiar sequence. As with the two Kyrie fughettas,
close inspection will often reveal paraphrase; see Example 198. The quaver
in BWV 672 and in BWV 673 must be the same?
Example 198
BWV 674 Kyrie, Gott heiliger Geist (Clavierübung III)

For TEXT and MELODY, see BWV 669.
In addition to its E-phrygian tonality, the cantus supplies themes: see

Example 199. Though most phrases of the plainsong melodies can be traced
in BWV 672, 673 and 674, the result is not a gratuitous intricacy but a seam-
less texture exploiting motifs as they naturally develop. Thus motif a is
developed more than the opening ‘fugal’ theme itself, appearing in almost
every bar, often in thirds and sixths though never inversus, and once in
combination with the opening (b. 17).
402 BWV 674–675
Example 199
A complement to the other lesser Kyries, BWV 674 is equally original,

smooth, compact and formula-free. The turning quaver lines may suggest
the final ‘eleison’ phrase of the plainsong (though there is no clear quota-
tion), and their motion is constant, suggesting that BWV 672 crotchet =
BWV 674 dotted crotchet, and thus that the settings are in proportion,
the bars the same in each metre. The final cadence A/E resembles that of
BWV 669 C/G, and suggests a pattern of cadences over the six Kyries:
BWV 669 C/G – BWV 674 A/E

BWV 670 F/G∗ – BWV 673 D/E
BWV 671 A/G – BWV 672 F/E
A further pattern emerges when the manualiter settings, all ending on E,

are followed by the Trinity settings on F, G and A. These are the four keys
of the Duetti BWV 802–805 and are prefaced now (as the Duets are later
followed) by a piece in three flats. Why this should be is obscure, but the
patterning seems to be there.
BWV 675 Allein Gott in der Höh’ sey Ehr’

Two staves; headed ‘à 3’, ‘Canto fermo in Alto’.
The trio is a two-part invention between whose lines, busy with many note-
patterns of wide compass and spacing, appears the cantus firmus, plain
and smooth, in almost entirely stepwise motion. Since there is no heading
manualiter, perhaps a choice was intended for the alto melody, between rh
and pedal 4 . But the c.f. does fit the hand, and the two previous settings
were not marked manualiter either.
∗ I.e. as in bb. 59–60. The final cadence seems out of line with BWV 669 and 671 and unlike them
disagrees with the harmonization BWV 371.
403 BWV 675
The subject contains several important motifs of its own and also refers.
to the first and fifth lines of the cantus (bb. 5–9, 26–30): see Example 200
Example 200
More is made of the melody’s upbeat than was customary, and perhaps the
prominent opening three notes (F G A) are a reference to the major third
formed by the keys of the three settings. Unusually, the paraphrase theme
incorporates the same notes and length of phrase as the following cantus
line (compare soprano bb. 1–5 with alto bb. 5–9), and although sometimes
called ‘not quite worthy’ of its place (Keller 1948 p. 202), the piece clearly
suits Clavierübung III in its motivic ingenuity and reference to Trinity. The
motifs are constantly adapted: d can change beat, and also combine with
b (b. 18); the countersubject motifs prove particularly versatile, combining
with others (b. 14), perhaps inverted (b. 36) or extended (b. 15), and run-
ning into another motif (bb. 15–16, bass). The key-plan, with its major–
minor alternations, largely follows the harmonizations found in Leipzig
cantatas.
With three motifs, b. 14 is typical of Clavierübung III: Example 201. In
not compromising this complexity with looser passages of galant melody,
the piece is a grander version of the kind of ‘tight’ chorale-settings attempted
by Kauffmann 1733 but more systematically exploring its motifs, running
into simpler phrases at the end of sections and (when compared with the
Example 201
404 BWV 675–676
paraphrases of this melody in Cantata 112, 1731) clearly conceived for key-
board. All of this contrasts with the last chorale, which is based on a single
motif.
The sudden dropping of complexity near a rather commonplace final
cadence is typical of the late works and Clavierübung III in particular. While
the fugal imitation is far more straightforward than in the lesser Kyries,
only a rigid view of Clavierübung III’s ‘musical concept’ (as in Kube 1999
p. 591) could find it out of place. Some of the triplet lines might have the
quality of a textbook exercise, but the underlining harmony is in classic Bach
chorale-style.
BWV 676 Allein Gott in der Höh’ sey Ehr’

Three staves; headed ‘a 2 Clav. et Pedal.’.
This is a strict trio with dominant answers as in the Six Sonatas, incor-
porating the chorale melody both as intermittent cantus firmus and as
paraphrase: Example 202. As the smaller notes show, the paraphrase also
combines canonically with the c.f. (see bb. 12ff.). An astounding integra-
tion is achieved not only in this way – more completely than in the trio
BWV 664 – but also in the harmony, which a melody of so elemental a
nature (GABCDCBAB) can truly pervade. Many possible references to this
melody can be found, even during the final pedal-point (e.g. bar 125 lh),
and all melodic lines, it sometimes seems, derive from the hymn.
Example 202
Though difficult to play, BWV 676 is one of the more approachable

settings. Such sonata-like features as the patterns and syncopations of
bb. 70ff. or the cello-like pedal of bb. 18ff. are charming and almost galant.
405 BWV 676–676a
But the logical if unconventional cadence in b. 99 is unknown in the Sonatas,

as too is the c.f. type of ritornello form:
l–33 trio exposition, lh answering rh; then c.f. 12 lh

33–66 trio exposition repeated but inverted; c.f. 45 rh
66–78 trio episode modulating but returning to G major for –
78–92 next two lines of cantus in canon, pedal and each hand in
turn, countersubject from subject (for the canon here, see
BWV 663)
92–9 trio episode
99–end last line of cantus lh (avoiding tonic) answered by rh; short
episode; pedal/rh canon; pedal point
The last section raises acutely the problem of a trio with c.f., since Bach
seems to have found it difficult to end the work with the sense of final-
ity easier in plainer ritornello trios. This finality eludes BWV 676, despite
the repeated last cantus line (heard four times from b. 99), the return of
bb. 30ff. at 119ff., and the farewell return of the trio theme in the last four
bars.
Although the work’s ‘integration’ is often recognized (Breig 1987), the
paraphrase-theme is so fluidly developed that an inversus is easy to miss:
bb. 30–3 soprano, 63–6 and 119–22 alto. It is not difficult to imagine that
in BWV 676 the composer was consciously ‘summing up’ his chorale-trios,
creating a ritornello form with cantus firmus, pedal melodies, invertible
counterpoint, paraphrased subject, all over a spacious 126 bars. This ritor-
nello principle is underlined by BWV 676 using an older form of the melody
in which the last line is the same as the second, unlike the previous setting
(Jacob 1997 p. 230). The overall length reflects the length of the phrases, very
different from the trios BWV 655, 660 and 664. Other than in the canonic
section, it is difficult to agree that it bears any similarity to Vers 5 of Walther’s
Acht Vorspielen on the same hymn, as is sometimes suggested, even if this
was known to Bach by now (see Dok II p. 265).
BWV 676a Allein Gott in der Höh sey Ehr’

Late sources only.
Though sometimes called ‘an early Weimar work’ serving as the kernel from
which BWV 676 was developed, BWV 676a is unlikely to be authentic. To
give the rh only the c.f. but have the free coda as in BWV 676 (KB p. 34),
to have ornaments in bb. 10, 33 and 40 not used elsewhere by J. S. Bach,
406 BWV 676a–677
and to have so many reiterated Gs in the pedal, suggest that the piece was
extracted to provide a shorter prelude. There seems to be some command
of idiom (e.g. the repeat of line 2 in bb. 39ff.).

Two staves; headed ‘Fugetta super . . .’, ‘manualiter’.
The work is not a simple but a double fughetta, completed in twenty

bars:
1–7 subject (and countersubject?) based on first line; stretto

answer; second answer, 5–6. Motif a (see Example 203),
appears inversus
Example 203
7–16 second exposition, also begins and ends in tonic, based on

opening of second line (Example 204); motif a from the first
16–20 themes combined (bb. 17–18); refers to motif a
Example 204
407 BWV 677–678
The two subjects are clearly contrasted. Most semiquavers can be traced
to motif a, which is present in fifteen bars of the twenty, and again the
cantus imbues the entire texture in a novel way. Subjects and treatment
are puzzlingly similar to the opening bars of the C major Fugue BWV 547
(see p. 116).
To the player, an important detail is the contrast between smooth semi-
quavers and detached quavers, including moments when each is extended
(quavers bb. 6–7, semiquavers bb. 15–16). One would expect such disjunct
quavers as the bottom line of Example 203 to be detached in performance
even without the dots: was it necessary by 1739 to mark them, and if so
because ‘counterpoint by articulation’ was an art becoming lost? Or, since
a not dissimilar détaché theme was also the subject of a fughetta in Kauff-
mann’s Harmonische Seelenlust, perhaps BWV 677 represents a conventional
subgenre of the day.
The scale figures in bb. 15–16 and at the close may, like runs in BWV
675 and the closing bars of BWV 676, represent the heavenly host singing
‘Gloria in excelsis Deo’ (Terry 1921 p. 98), very discreetly. The open thirds in
the theme, like the emphases on mediant harmonies and entries, probably
allude to the Trinity hymn, while the chromatic touches are there to create a
little tension (b. 18 – see also bb. 30–1 of BWV 675). But note in Example 203
that as might be expected of Clavierübung III, this paraphrase theme refers
to the melody on its main beats much less regularly than does a simpler
paraphrase of it in BWV 717.
BWV 678 Dies sind die heil’gen zehn Gebot

Three staves; headed ‘a 2 Clav. et Ped:’, ‘Canto fermo in Canone’.
For TEXT and MELODY (and a note on number symbolism) see BWV 635.
The big Ten Commandments setting attracts many kinds of attention. In

1757, Padre Martini quoted the opening as an example of imitation at the
octave (Dok III p. 117), although it could also have reminded Italian organ-
ists of organ pastorals. In c. 1776, Kirnberger found it a typical G-mixolydian
work (Dok III p. 301), in which the dominant is minor (b. 40) and the final
cadence is not perfect. Schweitzer saw it as representing order (the canon)
and disorder (upper voices wandering ‘without rhythm, without plan’: 1905
p. 346); Dietrich heard a pre-Fall quietness in its opening, then sinful devi-
ousness before final salvation; and Schering counted the five phrases of
408 BWV 678
the melody as producing ten when duplicated canonically (Keller 1948

p. 203). That there are strictly six and not five phrases in the cantus firmus
(as in BWV 298) still leaves ten sections in the sixty bars. More objectively,
the pun that Law = Canon implies that the New Lawgiver follows the Old,
as too does the text.
Canonic treatment of this melody is already there in the Ob setting and
Cantata 77, where a quasi-diminished canon between trumpet and continuo
‘summarizes’ the Law. In BWV 678, the canon has two unusual details: it
enters after episodes of differing lengths, and it changes order (alto or tenor
first). So intricately conceived are the note-patterns in the upper parts that
BWV 678 becomes a quite exceptional fantasia on motifs, rectus or inversus.
See Example 205 and the following list:
Example 205
bars
motif a 1, 15, 21–3, 37, 38
motif b 1–3, 6, 7, 12, 13, 17, 18, 25, 29–31, 34, 35, 44, 45, 49, 53, 54,
56, 59
motif c 2, 3, 4, 8, 9, 11, 30, 31, 40, 51, 52, 58
motif d 5–7, 9, 11, 12, 14, 16–18, 21, 22, 26, 34, 35, 43, 44, 48–50, 53,
55, 59, 60
motif e 6, 7, 10, 17, 18, 22, 23, 25, 27, 35–8, 44, 45, 54, 56, 57
motif f 13, 15, 19, 20, 24, 27, 28, 32, 33, 34?, 37–9, 46–8, 51, 52
motif g 4, 7–11, 18, 35, 36, 40, 41, 42, 51–7, 59
motif h 5, 10, 16, 25, 36, 43, 55–7
motif i 5, 16, 27?, 36, 38, 43, 45, 47, 48, 58, 59
motif j 6, 7, 11, 12, 17, 20, 21?, 22, 23, 26, 32–4, 44, 49, 50
motif k 7, 9, 14, 18, 23
The motifs are not always as clear as mere pedantry could have made them,
nor are they literally exhaustive: motif h could easily have slipped into the
last bar.
409 BWV 678–679
The two upper voices are like obbligato instruments in a cantata aria,
and are not based on the cantus. However, the opening pedal point is a
counterpart to all the repeated Gs and their harmonies as they appear in the
following fughetta BWV 679, in whose subject the same motif g reappears.
Furthermore, the melody of the chorale so governs the first four bars of
BWV 678 that it could actually be sung against them – a kind of unspoken
allusion – and the next two lines too could be anticipated in this rather
unusual way, though less convincingly (bb. 16–19).
The final c.f. phrase seems to express the ‘Kyrie eleison’ of the text by a
chromatic fall in the upper canonic voice – an effect less striking than in
BWV 671 but more so than at the end of ‘Allein Gott’ BWV 663, which
is simply a preparatory chromaticism. Other chromatic motifs, in bb. 5–6
and elsewhere, emphasize the ‘purity’ of the c.f., particularly when it falls
towards the G minor/B major of bb. 51ff. Reaching a relatively remote key
at this distance from the end is known elsewhere in Bach and is occasioned
here by the use of a B in the cantus, rare in the hymnbooks but there in
BWV 298. Throughout, each canon has begun in one key and ended in
another.
The five parts might be laid out ‘after the model’ of Grigny (Klotz 1969a),
with four manual parts paired off and registered accordingly; but Grigny
taxed his right hand a good deal less than is the case here, where very busy
parts are ingeniously laid out.
BWV 679 Dies sind die heil’gen zehn Gebot

Two staves; headed ‘Fugetta super . . .’, ‘manualiter’.
For TEXT and MELODY (with a note on number symbolism) see

BWV 635.
The theme of the fughetta, a complete fugue, is derived as in Example 206.

The jolly subject not only paraphrases the chorale but produces important
motifs, unlike the other fughettas with their stretto answers. The phrase
GABC ( = motif b inversus) will also be a reference to lines 2 and 4 of the
melody.
Example 206
410 BWV 679–680
Despite appearances, the work has much in common with BWV 678,
notably the G-mixolydian tonality, the opening Gs, references to number
ten (ten entries, four inversus) and the final diminished seventh and plagal
cadence. Most striking is the motivic ingenuity: as well as inverting the
subject as a whole (from b. 12), BWV 679 extends and inverts the two cells
of motif a and the countersubject’s jig-figure. Motif a occupies nearly half
the composition, obviating the need for more than ten entries, dividing
every bar (and the movement as a whole) into two, and creating a restless
three-part invention before the final entries in canon. Dissonances created
by accented passing notes contrast sharply with the triads, and a real unity
is given by all the repeated 12/8 rhythms, unlike those of BWV 712 which
change and vary.
So conspicuously lively – and strange – a jig as this may well be explained
by Luther’s words in the Lesser Catechism: ‘we should . . . cheerfully do
what he has commanded’ (Leaver 1975), rather than by Bach’s having in
mind a certain pilgrim song to the same melody (Steglich 1962 p. 32). The
psalms too speak of ‘delight in thy statutes’ (Ps. 119) and of rejoicing in
the Law (Pss. 19, 119). The ‘rejoicing’ is clear in the long final episode – a
structural detail familiar ever since the Passacaglia – and the result here is
a fugue far in advance of the usual chorale-fughetta, another unicum in a
book of unica.
BWV 680 Wir glauben all’ an einen Gott

Two staves; headed ‘in Organo pleno con Pedale.’.
The TEXT is Luther’s version of the Nicene Creed, placed in the hymnbooks
as a Trinity hymn, sung after the Gospel on each Sunday by the whole
congregation.
Wir glauben all an einen Gott, We all believe in one God,

Schöpfer Himmels und der Erden, creator of Heaven and Earth,
der sich zum Vater geben hat, who gave himself to be the Father
dass wir seine Kinder werden. that we might be his children.
Er will uns allzeit ernähren, He will always feed us,
Leib und Seel auch wohl bewahren, and keep us safe in body and soul,
allem Unfall will er wehren, he will ward off all misfortune;
kein Leid soll uns widerfahren. no harm shall befall us.
Er sorget für uns, hüt and wacht, He cares for, guards, watches over us;
es steht alles in seiner Macht. all stands in his power.
411 BWV 680
Verse 2 concerns chiefly ‘Jesus Christ, seinen Sohn’, v. 3 ‘den Heilgen Geist’.
The MELODY, based on the Credo cardinale (free paraphrase of Credo IV

in the Liber usualis), was popular before the Reformation; Example 207.
Example 207
Harmonized in BWV 437 and partly used in BWV 681, 765 and 1098. ‘Wir
glauben all’ an einen Gott’ listed in the Ob refers either to this chorale or
that of BWV 740.
As Example 208 shows, the fugue-subject paraphrases the first cantus
line, its countersubject the second, its answer the third. (For an earlier
Example 208
412 BWV 680
Thuringian melody made up of subject and answer cf. J. R. Ahle’s ‘Jesus

Christus, unser Heiland’ in Krummacher 1978 p. 469.) BWV 680 is the only
larger chorale in Clavierübung III to have no cantus firmus, but the melody,
which includes repeats of lines 1 and 2, is impractically long and rarely set
whole in one movement (there is an example by Walther). Another ‘partial
setting’, similarly fugal, is found in BWV 1098, where too there is a complete
statement of the line at the end. Was this final statement conventional in
Thuringia? Thus line 2 is countersubject not only to line 1 (both of which are
subtly paraphrased in the alto line) but to the bass ostinato theme as well.
If not as a firmus, the cantus pervades the whole ritornello structure,
through its characteristic contours against a dorian background. Although
no direct reference is certain, the outline of the chorale melody is there
in various note-patterns, as inspection would reveal, and not for the only
time in Clavierübung III, the setting is shot through with the cantus. The
resulting impression is one of strength and, each time the subject enters, of
voices singing ‘Wir glauben, wir glauben’.There seems no question but that
the quasi-ostinato motif in the pedal corresponds to ‘firm faith’.
But quite apart from the paraphrasing technique, the cantus is difficult
to recognize in the fugue-subject because the third note is sharpened, as the
same note in BWV 681 is not. The simplest reference in the whole piece is in
the tenor near the end (Example 209), a line that is both the second and last
of the hymn. It appears with the final pedal phrases, again after increasingly
longer episodes.
Example 209
The general style of the piece is Italian. This kind of fugue is as close
to a ritornello shape as the trio BWV 655 – hardly a sign of pre-Vivaldian
influence (according to Zehnder 1991 p. 93)? – and even the appearance of
the chorale-line at the end has much the same effect as the final statement
in BWV 655. There are further Italian elements. A striding ostinato bass
line without pauses is also there in a Credo section added by Bach to G. B.
Bassani’s Mass in F (Wolff 1968 pp. 202–3) and the influence of Frescobaldi
is likely, though nowhere else is the alternate-foot pedal idiom so well inte-
grated as here. The lh takeover of the ostinato phrase leads to several bars
of good, traditional counterpoint (bb. 76–82), with italianate suspensions
and a style obviously distinct from BWV 679.
In addition, the semiquaver groups throughout are much like quavers in
stile antico (see BWV 670), almost as if the piece were providing a catalogue
413 BWV 680–681
of them. All the figures in Example 210 are used, many of them constantly
in the course of the piece, as a few sample passages show (ii–iii). Most are
used imaginatively, with much less repetition than in BWV 684. Any passing
Example 210
similarities to the 2/4 counterpoint of the Echo in Clavierübung II (1735) –

and there are several – serve to show how three-part counterpoint can be
turned to Italian or French effect in keyboard music as in any other. And there
are many original touches in BWV 680. For example, each ostinato passage
ends with a two-bar transition in invertible counterpoint (bb. 8–9, 19–20,
31–2, 44–5, 64–5), moving towards the new key and giving coherence. Or the
recurring chromatic line (b. 44) incorporates the falling chromatic fourth
of antiquity. Or the final tenor reference to the melody is like a simplified
version of the countersubject as it occurs in b. 40, both of them with a singing
quality that permeates the texture.
Observations on the piece have often led to fanciful conjecture, such as
that the fourteen subject-entries allude to B A C H. More pertinent, perhaps,
are the dominant to relative-major sequence in the pedal entries (d, a, F, C,
g, d), the increasing gap between entries, and the fact that the opening D
minor and A minor paragraphs recall the tonic–dominant answers of the
old Italian trio sonata.
BWV 681 Wir glauben all’ an einen Gott (Clavierübung III)

Two staves; headed ‘Fugetta super . . .’, ‘manualit:’.

414 BWV 681
Curiously, the central piece of Clavierübung III is the shortest. But rather
than reflecting a kind of inverse emphasis (as if ‘marking it in a negative
way’: Jacob 1997 p. 53), its brevity is only notational, and the second ‘Vater
unser’ is no longer. The too-short final bar is probably the result of crammed
space in the engraving.
The paraphrase is clearest at the second answer: Example 211. Although
the key-signature led Kirnberger and others to speak of E-dorian (Dok III
p. 302), the key is firmly E minor, complete with sharpened leading-note
in the tonal answers of bb. 1 and 5. The hymn’s first line is implied when
Example 211
the opening motif returns, as in the alto b. 7, and furthermore, its second
line can be heard through the penultimate phrase, shown in Example 212 –
a thematic paraphrase typical of the volume. The dramatic, broken-off di-
minished seventh chords foresee later music more, perhaps, than those of
BWV 547.ii – because of the following runs?
Example 212
Misconceptions often arise with this piece: it is not a French ouverture,

which has a quite different shape, but a fugue using ouverture rhythms for
unusual effect. (On ouvertures in the middle of all four Clavierübung vol-
umes, see p. 392.) Thus the setting is a complement to the preceding ‘Italian’
treatment, and because of this, the setting is no more harpsichord music
than the other’s is string music. Yet although lh runs and final appoggiatura
need not imply harpsichord, the work most like it is the Gigue from the
D minor French Suite, with dotted-note subject, three parts, sequences,
similar final bars. The conventions of the genre produce in one case a binary
dance, in the other a fughetta whose subject is a chorale-paraphrase.
Both BWV 680 and 681 are fugal, based on the first two lines of a cantus
but different in as many parameters as possible, glossing Luther’s German
Creed with French and Italian accents. It is not a wish to portray majesty
415 BWV 681–682
but rather the contrast between Italian and French styles (cf. Clavierübung
II) that suggested the ouverture rhythms.
BWV 682 Vater unser in Himmelreich (Clavierübung III)

Three staves; headed ‘à 2 Clav. et Pedal è Canto fermo in Canone’.
Perhaps the most complex of all organ-chorales, for both composer and
performer, this is a ritornello trio sonata with distinct patterns (detached
triplets, slurred snaps) above a restless pedal part, plus c.f. sung in long
notes in octave canon between two further voices. (Such a compound form
had already been perfected in Cantata 78.i, which combines c.f., ritornello,
chaconne en rondeau, lamento bass, and four-part chorale.) A trio theme
again paraphrases the cantus (Example 213). Triple metre is unusual with
this melody, and its rhythms have been altered here and there, not always
for an obvious reason (e.g. the harmony does not require the longer notes
of bb. 67–70). Canon may allude to the Law, the keeping of which Luther,
in his commentary on the Lord’s Prayer, saw as an aim of constant prayer
(Leaver 1975).
Example 213
The musical language is as intricate as the form, for the articulation signs
alone make a ‘counterpoint of articulation’ obligatory: the sighing Affekt of
the trio (as in modern flute-and-violin sonatas) meets with the impervious
c.f. in canone. The tonality is E minor rather than the E-dorian implied by
the melody and its key signature, producing the following pattern:
BWV 680 D-dorian, 681 E-dorian, 682 E-dorian, 683 D-dorian.
416 BWV 682
The cantus is treated with a coloratura more elaborate than one in a setting
of ‘Vater unser’ by Georg Böhm, and it could be that we are being invited
to interpret it as wandering like ‘unsaved man’, sighing, in need of the
Lord’s protection (Weismann 1949–50). Whether the chromatic lines or
appoggiaturas allude specifically to the text is no more certain here than
elsewhere, though the astute musical effort involved in it may well convey a
sense of the strain of prolonged supplication.
If both ‘patience’ and ‘suffering’ can be heard in b. 41, and 41 = J. S.
Bach, so the setting’s ninety-one bars can be seen as the product of 13 (sin)
and 7 (prayer), and the full cadence in b. 56 as coming at the point of Golden
Section (1.62:1). The music’s ideas and the special significance of the Lord’s
Prayer make it inevitable that close connections are heard between them,
especially when cantatas use similar motifs, as they often do – when, for
example, Cantata 131’s chromatic motif descends for sin but rises for hope.
To speculate further – for example that BWV 682 alludes in some way to
Bach’s quarrel with Rektor Ernesti (Scheide 1999 p. 94) – is to attribute to
Bach particular ideas of what music is and does or should do.
A list of motifs on the lines of those indexed for the Ten Command-
ments setting is also possible for this piece (see BWV 678), but note
that now the harmony is predominantly minor, thus very different from
BWV 678’s. The upper parts of the trio sonata are founded on line 1 of the
cantus (bb. 1 and 5, 19 and 23, 56 and 60) and include three crucially dif-
ferent figurae or note-patterns (Example 214) in the codetta (a), counter-
subject (b) and continuation (c). The triplet figure c is capable of great
variety, a good instance of J. F. Agricola’s point in 1769 (Dok III p. 206)
that J. S. Bach taught players to distinguish between dotted figures and
triplets. All three are typical of galant flute music and suggest conscious
allusion to chamber trios, as do lombardic rhythms on rising lines elsewhere,
e.g. Cantata 114 (1724), ‘Where is the refuge for my spirit in this vale of
misery?’
Example 214
This main trio-sonata theme preserves the simple repeated note

(‘Va-ter’) of the c.f., which otherwise contains mostly longer notes than
the other parts, and is thus closer to BWV 684 than to 686. The opening
ritornello section is long and rich in thematic detail, and is not literally
repeated but, rather, each of the six later ritornelli introduces motifs from
it in order, one at a time. As a counter to the chromatic b, motif a could be
417 BWV 682–683
interpreted as ‘hopeful’ or ‘trusting’. The lombardic rhythm has been traced

as a device found in works of the early 1730s, which may or may not imply
a date for the composition of BWV 682 (Herz 1974 p. 96).
The continuo-like pedal has its own motifs that emerge clearly during
episodes and in the sequences following the prevailing E minor of the first
thirty bars. Its more or less consistent quavers supply the much-needed
impetus. A few ideas are introduced as the piece proceeds – syncopated
motif around b. 52, lh in b. 62 – otherwise, the ninety-one bars seem to be
a fantasia on constantly recycled motifs. Despite the dominant passage in
the middle the work repeatedly returns to the tonic, using the motifs with
a variety of incomplete cadences to avoid over-strong tonics. The last nine
bars are effectively a coda, highly reminiscent of the Six Sonatas, and the
final cadence (with the longest bass notes of the movement) bears more than
a slight resemblance to the close of the E minor Fugue WTC2. The coda’s
sense of finality is clear from comparing the bass of bb. 83–8 with that of
bb. 7–12.
For a note on the five parts, see BWV 678. Again, the canon alternates
the starting voice (upper or lower), but unlike BWV 678, now between the
hands.

Two staves; headed ‘alio modo manualiter’.
As in so many Ob chorales, the melody is in the right hand without inter-

ludes, accompanied by counterpoint made from motifs. In this case one
of them may be derived from a line of the chorale: Example 215. In their
chords and progressions this and the Ob’s ‘Vater unser’ are surprisingly
alike in their sweetly melodious harmony. The contrast in all respects with
BWV 682 could hardly be greater, the smaller setting more old-fashioned
and as sweet as the other is awesome.
Again, however, there is great ingenuity. The running motif a is also
inverted, allowing it to appear on every half-bar but bb. 3 and 23, and the
second motif b appears both with and without its tie. Clearly, the mood
is supplicatory and toned down, as conveyed by the isolated upbeat at the
beginning (the repeated a opening BWV 601 and 636 is accompanied); by
the thinning of parts at the beginning of each line; by the text of line 4
invoked in the motif (see Example 215); and by the low and apparently
418 BWV 683–684
Example 215
subdued close. Though like the Ob in conception, and just as usable in the
service as a hymn-prelude, a more integrated texture results from being
manualiter.
BWV 683a Vater unser im Himmelreich

Late sources only.
Like the longer version of ‘Ich ruf zu dir’, this lengthened version is likely to
be an arrangement, characteristic of posthumous Bach reception (NBA IV/1
KB p. 97). The sequences, inconsequential voice-leading and inconsistent
provision of interludes cannot be authentic (NBA IV/4 KB p. 34) and thus
cannot represent, any more than BWV 691a does, ‘Bach’s first attempt at
using the concertato principle in chorale preludes’ (Eickhoff 1967).
BWV 684 Christ, unser Herr, zum Jordan kam

Two staves; headed ‘a 2 Clav. è Canto fermo in Pedal.’.
The TEXT is Luther’s Baptism hymn, generally associated in the hymnbooks

with St John the Baptist’s Day (Stiller 1970 p. 323).
Christ, unser Herr, zum Jordan kam Christ, our Lord, came to the Jordan
nach seines Vaters Willen, according to his father’s will,
von Sanct Johann die Taufe nahm, and was baptized by St John
sein Werk und Amt zu ’rfüllen, to fulfil his work and office.
Da wollt er stiften uns ein Bad, There he ordained for us water
zu waschen uns von Sünden, to wash us of our sins,
419 BWV 684
ersäufen auch den bittern Tod and to drown bitter death

durch sein selbst Blut und Wunden; through his own blood and wounds;
es galt ein neues Leben. it was a matter of new life.
One of the six other verses speaks of the symbolic nature of water, which
ist vor ihm ein rote Flut, is for [the faithful] a red flood
von Christi Blut gefärbet. coloured with Christ’s blood.
The MELODY (by J. Walther?) was published in 1524, only later associated
with this text; used in Cantatas 7 (St John, 1724) and 176, harmonized in
BWV 280 and listed in the Ob. Example 216.
Example 216
Though similar in form and technique to other pieces in Clavierübung III,

BWV 684 is distinctive. The three manual parts together constitute the
ritornello, often avoiding firm cadences and spinning out the patterned ac-
companiment to the c.f. Below the imitative parts, lh runs with a semiquaver
figure derived, as in BWV 677 and 681, from the melody at two different
phases: a kind of double paraphrase, shown in Example 217. It is possible
Example 217
to find other allusions too, as when rh seems to take in the second line at
two different levels, long and short (Example 218). So the cantus pervades
the idiom – but quite naturally, for had the aim been to allude endlessly to
it, more was possible with the rising fourth common to lines 1, 2, 5 and 6.
Also, the soprano of bb. 42ff. could have made a clear reference to the fifth
chorale line, and bb. 50–1 to the sixth line.
Resemblances between BWV 684 (top part bb. 77–9) and 687 (final bars)
are coincidences suggesting that they may have been composed within a
short space of time. This alone may be enough to still the player’s first
420 BWV 684–685
Example 218
suspicion that the movement was either transcribed from a (lost) aria or
transposed from a draft version in another key (D minor). The tonality is
C minor rather than the C-dorian implied by the key-signature, which –
as mostly throughout Clavierübung III – allows the modal c.f. to be written
originally without accidentals.
The running, swirling semiquavers are usually interpreted as picturing
the flowing Jordan, rather more convincingly, perhaps, than the ‘sound of
a rushing, mighty wind’ felt at the presence of the Holy Spirit at baptism
(Leaver 1975). Others have seen a connection with the C minor sections of
the E Prelude (Trumpff 1963 p. 470), which may again say something about
the dates. Also open to speculation, in this setting of a melody whose text is
itself peculiarly symbolic, is the opening rh motif of quavers: a typical cross
figure or sign, as at the opening of the Order of Baptism itself. Furthermore,
the melody of this ‘Jesus chorale’ appears in the tenor as middle voice or
mediator, second Person of the Trinity.
Not the least interesting detail is the rubric ‘a 2 Clav’. Presumably this
means that the running bass has its own manual and registration, the latter
of which (if it conforms to Kauffmann’s Harmonische Seelenlust) imitates a
violone bass played by a 16 manual stop. It also means that care needs to be
taken with such bars as 7 and 14, where the lh is also needed on the other
manual. Such a ‘continuo bass registration’ for the lh is a development of
the 1730s?
BWV 685 Christ, unser Herr, zum Jordan kam

Two staves; headed ‘alio modo’. ‘manualiter’.
The work is not a simple fughetta, and its twenty-seven bars are amongst
the most closely reasoned of the whole collection. Both subject and counter-
subject are derived from the chorale melody (Example 219) and the shape
is as follows:
421 BWV 685
Example 219
A 1 subject and countersubject rectus

4 subject and countersubject inversus, with free middle part
8 episode from countersubject (answered in bass 8–9)
B 10 subject (alto), countersubject (bass 12), rectus; episode
continues against alto entry
14 subject, countersubject (bass 16), inversus, alto derived
18 episode from countersubject heard intact in soprano
C 20 subject (bass), countersubject (soprano), rectus, alto
derived
23 subject (alto), countersubject (soprano 24), inversus,
bass derived
The combination of constant inversion, derived motifs and modal progres-

sions (bb. 1–3, 15–16) results in a highly original composition, with unusual
harmony and a capriciousness about the number of parts, repetition and
direction not suggested by the shape. Note the antique, Scheidt-like nature
of the sequences in b. 9. The taut feel of it all has been likened to the short,
‘somewhat indigestible’ settings in Kauffmann 1733 (Butt 1995 p. 50): a dis-
tinct subgenre contrasting both with new aria-like settings (in Kauffmann)
and with modest chorales for modest players (in Sorge’s collection). The
result is a dense fugal style as ‘remote’ as many a canon.
There have been many attempts to ‘explain’ the movement. The turning
motif a gives the ‘visual appearance of a wave’ (Schweitzer 1905 p. 345); the
three inversus entries represent the threefold immersion in baptism (Keller
1948 p. 207), by playing on inversus = immersus (Leaver 1975); the three
rectus entries, passing from soprano to bass, suggest a reference to the Trinity;
and the two subjects correspond to the Old and New Adam (Smend 1969
p. 166). Other suggestions could be made for the little falling scales from
b. 4 onwards. If having the melody in triple time refers to Baptism as ‘la
manifestation par excellence de la Sainte Trinité’ (Chailley 1974 p. 90), then
so should ‘Vater unser’ BWV 682. But more likely is that a simple contrast
was desired between longer and shorter settings, particularly in view of the
four-square perpetuum mobile of BWV 684. There is no evidence to support
or disqualify these and other speculations about a text recounting an act
which is itself symbolic.
422 BWV 686
BWV 686 Aus tiefer Not schrei ich zu dir

Three staves, lowest ‘Ped: dopp:’; headed ‘a 6’, ‘in Organo pleno con Pedale
doppio’.
The TEXT is Luther’s free and highly personal translation of Ps. 130 (‘De
profundis clamavi’), used (as Ps. 129) in the Roman Burial Service and
Office for the Dead.
Aus tiefer Noth schrei ich zu dir, From deep distress I cry to you,
Herr Gott, erhör mein Rufen. Lord God, hear my call.
Dein gnädig Ohr neig her mir Incline your gracious ear to me
und meiner Bitt sie öffne! and open them to my entreaty;
denn so du willst das sehen an, for if you will take notice of
was Sünd and Unrecht ist getan, what sin and wrong is done,
wer kann, Herr, vor dir bleiben? who may abide you, Lord?
V. 3 turns to hope:
Darum auf Gott will hoffen ich, On God therefore will I place my hope,
auf mein Verdienst nicht bauen. and not on my deserts.
By 1525 this was already both a Communion and Burial hymn (Stapel 1950
p. 176), to which a doxology was added. In Schein and Vopelius a Palm
Sunday hymn (1645, 1682), it became associated with Trinity 21 and 22 in
Dresden and Leipzig.
The MELODY as used in Cantata 38 (1724) preserves the phrygian character

(Example 220) and the first five notes are common to several sixteenth-
century themes or theme-types. Listed in the Ob; see also BWV 1099.
This is the grand climax of the so-called organ motet, one of the few
six-part pieces in the organ repertory and the only known example by
Bach, unless Ricercar à 6 in the Musical Offering (1747) is counted. Double
pedal parts in Buxtehude, Reinken, Bruhns and others are less reliable than
modern editions imply, though tablature or staff-scores (open and key-
board) could often be interpreted in this way: the extant tablature score of
Weckmann’s ‘O lux beata trinitas’ notes that it can be so read that pedal
plays both tenor c.f. and the bass (the latter read down an octave: see
W. Breig in Bä 6211). Earlier models in Scheidt’s Tabulatura nova (1624) are
organ motets for organo pleno, with 16 pedal reed. Different articulation for
the two pedal parts of BWV 686 makes 16 pedal registration on appropriate
organs practicable, for what is almost an augmented canon for pedals.
423 BWV 686
Example 220
BWV 686 is the composer’s strictest motet chorale, just as the other set-
tings in Clavierübung III are models of other techniques. As such, it has more
parts, its polyphony is more continuous, there are more countersubjects,
the expositions are less stereotyped and the final section more keyboard-like
than the choral motets it resembles, such as the opening of Cantata 38, which
is closer to note-against-note. Consequently, the organ setting has been seen
as involving the whole of the chorale’s text, while Cantata 38 involves only
the first verse (Meyer 1985 p. 74). The stile antico of BWV 686 is purer, more
‘objective’, though the lively dactyl figures – common motifs in the contra-
punctus floridus style, of which this is a model (Wolff 1968 p. 69) – invite the
listener to hear the note of penitence as moving towards a positive outcome.
The shape can be outlined:
1line 1/3, fugal, all voices; stretto at octave (3) and fifth (9), this
against augmented c.f.; syncopated (cf. BWV 687) + crotchet
countersubjects.
13 line 2/4, rising caput (minims or crotchets) in all voices; c.f. in
pedal; crotchet countersubject inversus, most parts moving by
step
22 line 5 in all voices but manual bass; countersubjects are
syncopated by step (22), or small leaps (32), or quaver anapaests
(42)
31 line 6 in three voices only; motif from 32 (cf. bb. 57ff. of
BWV 687) leading to more broken texture
41 line 7 in all voices, paraphrased, partly inverted; motif from 42
leading to coda of lively, more ‘modern’ figuration
The lower pedal part systematically passes on to the countersubject after

each subject, and certain rhythmic or melodic shapes look as if they were
derived. The careful variety in the texture and spacing, which is increas-
ingly varied in the second section, leads to good lines. (Note the top part
throughout.) The massive opening is only one facet of a six-part texture that
is constantly varied, leading to the chord of widest extent exactly halfway
424 BWV 686–687
through, b. 27. The lines are so constantly allusive that a sample bar (b. 14)
will contain a motet subject derived from the cantus, the same answered, and
a crotchet countersubject line like the cantus in diminution. Other moments
will remind one of other chorales, e.g. BWV 629 in b. 69.
The setting is altogether less chromatic than the Ricercar à 6, its fleet-
ing augmented chords barely counteracting the repetitious tonic-and-
dominant at the ends of each major section. The Ricercar has some of
the same motifs but more diminished sevenths, 6/4/2 chords, and suspen-
sions. However, certain passages – bb. 53–62 of BWV 686, bb. 90–4 of the
Ricercar – are more similar to than different from each other. The Ricercar’s
suave chromaticism serves a modern royal chamber-music theme, just as
the organ-chorale’s diatonicism does an archaic church hymn-tune.
BWV 687 Aus tiefer Not schrei ich zu dir

Two staves; headed ‘a 4 alio modo manualiter’.
Like the last, this is a chorale-motet with augmented c.f. and contrapuntal
lines constantly referring to the subjects. But it has its own technique:
line by line, derived fugue subject answered in inversion; only in

bb. 11–14 do the lower parts have no fugue theme, rectus or inversus
or both
each exposition very like the others (5 bars fore-imitation, 8 bars c.f.,
1 bar cadence), with a subject answered in stretto
c.f. in top part in minims (double beat), fugue subjects in quavers
(half beat)
each c.f. section gradually works towards an increasing motion
Manualiter motets with c.f. are uncommon, and so is such strict coun-
terpoint, even denser in its way than BWV 686’s. But tempering this stile
quasi-antico are two distinctly modern elements: the galant key and the
2/4 metre.
Melodic cross-references emerge naturally in such textures with such
a theme – e.g. the first line inverted is not unlike the last line – and the
part-writing is geared to produce unusually restless semiquavers. Constant
little syncopations keep up motion just as, in a different way, do the sus-
pensions of the larger setting BWV 686. While constant inversion invites
425 BWV 687–688
symbolic interpretation, such as that atonement = conversion = melodic

inversion (Meyer 1985 p. 73), much of the detail is like the ‘dictated chorale’
BWV 668. This implies that the response to the text’s distress was subdued, as
does the whole setting. Objectively, it makes an obvious pair with BWV 686,
as is the case elsewhere in the collection, where two settings are conceptually
similar but perceptually quite different.
F minor is a more feasible transposition of E-phrygian than F minor
would be, and some modal character is kept, as at the two phrygian ca-
dences bb. 14–15 and 28–9. The same phrygian colour is heard in J. K.
F. Fischer’s little eight-bar fughetta in Ariadne musica on a similar theme
(presumably known to Bach), which in turn resembles certain traditional
canzonas found as late as 1722 in Zipoli’s Collection of Toccates. While
BWV 687 is founded on a chorale melody, it also shows one of the possible
developments of a theme widely known in various forms and contexts, none
of which necessarily alludes to any other. A theme countered by a diminu-
tion of itself is an old convention, and there is a family likeness between
BWV 687 and, for example, J. C. Kerll’s Canzona in G taken over in Handel’s
Messiah for ‘Let all the angels of God’ (1741), as well as ‘Vor deinen Thron’
BWV 668.
BWV 688 Jesus Christus, unser Heiland, der von uns den
Zorn Gottes wandt (Clavierübung III)
Two staves; headed ‘a 2 Clav. e Canto fermo in Pedal.’.
By Clavierübung III standards, BWV 688 is straightforward: a fugal,

through-composed monothematic trio on two manuals, with pedal c.f. line
by line, in long notes, modal and almost ‘Gregorian’. But the motif de-
velopment is subtle and arcane, the trio theme no simple paraphrase. See
Example 221. Then the fugal codetta from b. 6 supplies an idea that will
recur rectus and inversus for the rest of the piece, as does the countersubject
(bb. 7–9). See Example 222. As (ii) shows, the theme is used in inversion,
mirror image or retrograde, mirror-image inversion, syncopation (bb. 20f.),
and syncopated mirror-image inversion: a whole catalogue of metamor-
phoses in which versions of the theme alternate with each other. In one
form or another, it appears some seventy-two or seventy-three times, and
the non-stop semiquavers (tending more and more towards scales) are built
up from motifs in alternation and in various combinations, at least twice in
brief canon (bb. 13ff., 29ff.).
426 BWV 688
Example 221
Example 222
How far all this intricacy bears on the hymn’s text or function has led
to much conjecture. Spitta hears in it the ‘life-strengthening beliefs’ of
v. 5 (II p. 694), others a separation and coming together of God and Man
in the leaping subject (Dietrich 1929), or the ‘lively exertions’ implied in
v. 6 (Steglich 1935 p. 123), or ‘the anger of the Father deflected by Jesus’ in
Communion (Chailley 1974 p. 163), or the treading of the winepress (Isaiah
43, 2–3) symbolizing victory over the cross, whose motif opens the piece
(Leaver 1975). The inversion beginning in b. 47 may reflect the text of line 2,
‘Who turned God’s anger away from us’ (Jacob 1997 p. 213). Naturally, the
number of times the main motif appears has Trinity associations (72 = 1
× 23 × 32 ), while the wedge-shaped theme and its inversion seem to trace
iota-chi, jc, Jesus Christus (Krause 1965).
However, motivic intricacy – a kind of restless self-reference – may well
be not only the nature of the piece but also its purpose: it is a further step
towards ‘self-generating composition’. Fanciful interpretations at least draw
attention to the originality of detail, the one-bar phrases, the inversions, the
constant appearance of the subject. The result is quite different from other
works with a similar theme and countersubject (E Sonata’s Finale) or faintly
similar figuration (Cantata 72, opening) or arias with a c.f. sung between
two instrumental lines. The tone throughout is original, without precedent,
and even the coda is unusual for such a ritornello movement: here, a pedal
point is drawn from the last note of the c.f. (cf. BWV 684) followed by a
final ritornello (cf. BWV 675 and two Schübler Chorales). Ending without
pedal is unusual and heightens the dissonant effect of the two upper parts
and final syncopations, which imply a written-out rallentando.
427 BWV 689

Two staves; headed ‘Fuga super . . .’, ‘a 4 manualiter’.
The melody supplies the theme for a regular fugue, longer and with clearer
entries than the previous fughettas: note the title, ‘Fuga’. The sharpened
fourth is as found in Vopelius’s hymnbook (Leipzig 1682) and helps to turn
the modal, long-note cantus of BWV 688 into a diatonic fugue subject in 4/4.
The countersubject provides much of the running quaver material through-
out (Example 223), including the inversus from b. 19 – was it suggested by
the last line of the cantus? (If it was, the connection could have been made
clearer, e.g. in the last two bars.)
Example 223
While such a fugue could serve at Communion, its musical purpose is to

explore the chorale melody in a clearly defined genre different from the one
before – then a c.f. trio, now a complete fugue à 4 – and mastered beyond any
distant model. As in BWV 680 and elsewhere, the first answer is in stretto;
this sets the pattern for the movement almost as much as it does in the C
major Fugue WTC1, for stretti then occur at varying intervals of time and
in various textures:
1–2 middle voices after six beats (beginning on a down-beat)

10 upper voices after one beat
16 middle voices after two beats
23–4 lower voices after four beats
36–7 upper voices after five beats
37–8 soprano and tenor after six beats (beginning on an up-beat)
57 middle voices in stretto of augmentation: simultaneous
The stretti exploit six distances, much as those in the (contemporary?) B

minor Fugue WTC2 exploit different harmonies. It is particularly convenient
for an organ chorale that the augmented stretto in b. 57 brings in the melody
as a kind of rounding-off cantus firmus en taille, closing the work except for
a deceptively simple coda derived, as in BWV 686, from the counter-motifs.
428 BWV 689
(This coda shares a family likeness with that to the chorales BWV 687 and
658, the latter also in F minor.)
Note too that the augmented stretto appears shortly after an imitative
episode has used earlier material (compare bb. 30–3 and bb. 53–4), con-
tracting it and producing a sequence reminiscent of Pachelbel and others.
Further imitation and stretti concern the quaver countersubject and the sub-
ject’s little dactyl (b. 3 – see for example bb. 41ff.), and semiquavers become
prominent before disappearing towards the end. In the process, the theme
is constantly reharmonized, which must be one of the purposes of fugues:
to present the subject in varied but always intelligible, singable harmony.
The same is true of organ chorales.
Chorales formerly called ‘The Kirnberger Collection’
BWV 690–713
Misunderstood source-material gave this group of chorales its name dur-

ing the nineteenth century, all of them found (in a different order) in
the MS copy Am.B.72a bought by J. P. Kirnberger from the Leipzig pub-
lisher Breitkopf in 1777 – a professional copy similar to copies owned by
other late Bach pupils, C. F. Penzel’s P 1109 and J. C. Oley’s P 1160 (see
May 1996 pp. 24f.). Unclear is whether Breitkopf worked from a Sammel-
mappe, a collected portfolio belonging to the late composer, revised by
him here and there and called ‘Variirte und fugirte Choräle’ (‘decorated
and fugued chorales’); or whether Breitkopf acquired (bought? commis-
sioned?) such a portfolio only after 1750, advertising copies of it on sale in
1764 (May 1974a p. 100). Kittel and others also had access to some of the
pieces.
Nothing suggests these twenty-four pieces to have been a set. On the con-
trary, perhaps BWV 700, 718 and 741 were already too old to be included
in the albums copied by Walther and J. T. Krebs, who knew at least twelve
other BWV numbers between 714 and 762. BWV 690, 694, 712, 713, 741 were
copied by J. L. Krebs, but as individual pieces. There is often little to choose
in style between BWV 700, 705, 707, 716 and 724 and various anonymous
settings in C. H. Rinck’s late album LM 4843 (Krumbach 1985), making it
hard to discern a collection as such or to be confident of authorship, both
here and by analogy in the ‘Neumeister Collection’. Breitkopf seems not to
have known the ‘Arnstadt Chorales’ as a group, and in any case, any prop-
erly ordered or complete portfolio of chorales would surely have included
BWV 718, 733, 734 and 741.
The most promising group is the seven Advent and Christmas fughettas,
perhaps intended or originating as a separate set. They were copied in the
order 696, 697, 699, 698, 703, 704, 701 (then plus BWV 702) by one of
Breitkopf ’s copyists, c. 1760 (Brussels II 3919). In them well-wrought
counterpoint provides motion and rich harmony for melodies also set in
the Ob, and the variety both in the way themes were derived from the cantus
and in their countersubjects – compare BWV 703 and 704 – must be de-
liberate, part of a bigger plan. Possible models can be guessed: if the stiff
fughettas in the 44 Choraele attributed to J. C. Bach (†1703, but see BJ 2001
pp. 185–9), where a second line can also be a subject, led to the ‘Neumeister’
fughettas BWV 1098 and 1103, then perhaps the mature Seven Fughettas
[429]
430 BWV 690
were responding to something less provincial in answering the need for

chorale-fughetta preludes.
One model was surely the ‘Canzon dopo l’epistola’ in Frescobaldi’s
Fiori musicali, a volume whose lasting influence on Bach has not yet been
fully traced. Another would be the high-quality counterpoint of some of
Pachelbel’s Magnificat versets, though these too are likely to be following
Frescobaldi. Bach’s look like Leipzig works, mature in detail, harmony,
derived countersubjects, modal elements, dense and abstract counterpoint
exhaustively using subject and countersubject in the course of a short move-
ment, more ‘approachable’ and just earlier than the manual settings in
Clavierübung III. As such, they were as much ‘demonstrations’ as Telemann’s
four-part XX Kleine Fugen (Hamburg, 1730), though based on chorales and
less formulaic, didactic or whimsical than these. Since there was a tradition
for fughettas with short three- or four-part fugal exposition, episode, one
or two final entries, and short pedal point (Fischer’s Blumen-Strauss, c. 1732
or Muffat’s seventy-two Versetl, 1726), perhaps they were a direct response
to some such publication.
The thematic complexity of BWV 698 or 701 is not known elsewhere –
not, for instance, in Pachelbel, as often claimed, and it is difficult to agree
that BWV 698 or any of the ‘Seven’ lies close to Buxtehude (see Burba 1994
p. 94). If the stretto structure of BWV 697 can also be found in Fischer,
its pervasive countersubject, its chorale basis and its unusual modal frame-
work cannot. This ‘modality’ is more a question of ambiguity of diatonic
key than modalism in any antique sense, as is also the case with some
Clavierübung III fughettas. This is not the least reason to think them Leipzig
works.
BWV 690 Wer nur den lieben Gott lässt walten

Further copies by J. L. Krebs (P 1117), via C. F. Penzel (P 1109) and later.
Like BWV 683, this contains the melody as a soprano c.f. above a run-
ning motif in three and four parts below. The suspirans, from which the
scales are developed, is prominent when each cantus final note is length-
ened (e.g. bb. 4–5), as at interludes in the earlier hymns BWV 722 and 729.
The movement is much like a partita variation, e.g. Böhm’s ‘Wer nur den
lieben Gott’ (Nos. 2, 4, 6 and 7) and the double of Buxtehude’s ‘Auf meinen
lieben Gott’ (BuxWV 179), except for more suspensions in the harmony.
Another point in common with (some of) Böhm’s variations as known to
431 BWV 690–691
Walther is that both halves are repeated, unlike BWV 642 and 691. Were
repeats more characteristic of harpsichord and organ chorale-variations
(cf. BWV 767), on the analogy of the allemandes and courantes they
resemble?
Spacing, unlike BWV 683’s, suggests domestic keyboard instrument,
which would not make the following simple harmonization inappropri-
ate despite the change of metre. Krebs’s copy of BWV 734 also has a fig-
ured chorale, but whether any such harmonizations go back to J. S. Bach is
unknown. The harmony underlying the counterpoint of BWV 690 is no-
ticeably more sophisticated than the chorale’s, which also has no repeat
marks for the second half. That there was some need for simple harmoniza-
tions or figured chorales is clear – for a public service in church (one after
each hymn in Kauffmann’s Harmonische Seelenlust) and for domestic devo-
tions or ‘youth studying music’ (one after each of ninety-seven settings in
J. S. Beyer’s Musikalischer Vorrath, 1716–19).
BWV 691 Wer nur den lieben Gott lässt walten

Autograph MS: CbWFB (early 1720); also in AMBB (after 1725), and by or
via J. G. Walther and J. C. Oley.
Like BWV 753, BWV 691 was probably composed for CbWFB, where it
appears on fol. 5v. The decorated right-hand melody is without interludes
between chorale lines; and the two left-hand parts are not derived from the
melody, nor do they develop any one motif. An unusual genre: presumably
it had domestic uses?
The eight bars contain many (but not all) of the ornaments in the
Explication or Table of Ornaments placed two leaves earlier in CbWFB,
incorporating many cantabile or written-out ornamental figures, and thus
giving a succinct model of both French and Italian embellishments. The
copy in the AMBB is less exact in its ornaments. The order of pieces in
the older album (BWV 994, 924, 691, 926, 753, 836 . . .) suggests that the
chorales supply examples of decorative effects contrasting with the simpler
figuration of the surrounding pieces: miniature models of technique. Thus
the first note of BWV 691 is plain; the next figure is a much-used one (e.g.
BWV 656 b. 2); the short rest in the melody in b. 3 is a good example of
the tmesis (see also Walther’s BWV 692); and the whole is as unusual as it is
touching.
For a note on the slurs, see BWV 728.
432 BWV 691a–694
BWV 691a Wer nur den lieben Gott 1ässt walten

Late sources only (P 285).
Such bars as 12–13 hardly date from the earlier eighteenth century,
and the change in style recalls that of other unauthorized enlargements,
BWV 639a and 683a. Perhaps the relationship between the Orgelbüchlein’s
BWV 641 and the Art of Fugue’s BWV 668a led Bach admirers to try their
hand at expanding a short chorale.
BWV 692 Ach Gott und Herr

Copies by J. G. Walther, J. T. Krebs (BWV 692), and others via a supposed
‘Portfolio’.
Relying on certain sources, M. Seiffert in DDT 26 included these as Vers 4

and Vers 3 of a seven-movement partita of J. G. Walther, who signed one
copy of BWV 692. Sixteen sources are known for BWV 692 and BWV 693,
five for the set of seven verses (Emans 2000), and from their other contents
some imply Bach as composer of one or other or both, without naming him.
Much depends on how likely it is that two composers would collaborate on
a set of variations, or one of them use work by the other in compiling
a set.
Such bland harmony and doctrinaire working of motifs is not char-
acteristic of J. S. Bach, not even, as far as is known, by way of demon-
stration or exercise. At most, the cantus decorations of BWV 692 and
the simple accented passing notes of BWV 693, both of which could be
instructive, suggest common interests amongst Weimar organists at the
time.
BWV 694 Wo soll ich fliehen hin

Copies: via the ‘Portfolio’ only.
Two staves; headed ‘à 2 claviers et pédale’ (? NBA IV/3).
For TEXT and MELODY see BWV 646. The melodic form in BWV 694 is
as in Cantata 5: Example 224.
433 BWV 694
Example 224
Like BWV 646, this is a trio in which the two hands do not cross parts (thus
playable on one manual), the left hand imitates the right, and the pedal has
separate c.f. phrases. Despite their relative lengths, the ritornello shape of
BWV 646 is clearer because it ends on manuals alone; but its registration
suits BWV 694 equally well. This is much longer and has a slower-moving
cantus and – unlike maturer chorales in this form (BWV 646, 684, 688) –
contains an extra bar before certain pedal phrases enter. It also exploits
inversion much less than BWV 646, the motif appearing only in rectus
form for the first forty bars, wth one exception (b. 12). From b. 40, how-
ever, there is an alternation of rectus and inversus comparable to BWV 646.
Perhaps all along the motif was derived from the cantus: the inversus form of
Example 225 (iii) occurs at the point reached by line 3.
Example 225
Despite the length and occasional harmonic infelicity, the movement

attempts a well-knit development of motif, with syncopated countersubject
(b. 1), all hard to think of as pre-Weimar. The amount of repetition is itself
a graphic description of the text – a ‘fleeing’ to no avail – more so than is
the tightly controlled setting BWV 646. Some have heard in the sustained,
twisting lines a reference to v. 7 (Luedtke 1918 p. 68): ‘with your blood I will
overcome death and sin’.
434 BWV 695–695a

Further copies via C. F. Penzel (P 1109) and later.
Two staves; headed ‘Fantasia super . . .’ (authentic?).
For TEXT and MELODY see BWV 625, including a note on the melody.
This is a ritornello setting with first two lines repeated, each section be-
ginning with a two-part fugue (long fore-imitation), on lines 1 and 5 of
the chorale: see Example 226. The first theme returns in the coda but there
Example 226
is no combination of subjects as in a double fugue or some harpsichord

gigues. Rather, the movement is highly organized in its patterns, with
quavers and semiquavers drawn either from the subject (x in b. 62, etc.)
or developing from standard 3/8 patterns, in a typical way. The patterns
go on to produce an original and imaginative coda after the melody is
finished – a mature characteristic (cf. BWV 646, 675), in this instance on to
a cadence that corresponds to the hymn’s ‘Hallelujah!’ The texture is varied
and accomplished, much like the composer’s harpsichord music, particu-
larly in the section from b. 84 – compare the Gigue from the G major Partita –
and carries conviction in its striking subject, harmony and counterpoint.
An early Weimar work?
In addition to these conventional details, there are several unusual fea-
tures: one is the opening dominant ‘answer’ and consequently ambiguous
tonality (what is the key?); another is the first and second-time bars to the
first section (why are so many different?); a third is the mere figured bass
for the final harmonization (the MS took it from another source?). On a 4/4
chorale to a 3/8 setting, see BWV 690.
BWV 695a Christ lag in Todesbanden

Late sources only (e.g. Scholz).
A few 6/4 chords, unconvincing points d’orgue and congestion in

bb. 110ff. make it likely that this version (EB 6589) is an inauthentic ar-
rangement. But transferring the alto to the bass works well enough – and
435 BWV 695a–696
such new details as the line across bb. 104–7 are convincing enough – for
one to wonder whether such c.f. settings circulated in two versions, even in
Clavierübung III (BWV 675). One would need better sources to conclude
that J. S. Bach was the author of BWV 695a, however: such details as the 7–6
progressions of bb. 126–7 do not ring true.
Perhaps in practice an inner cantus firmus was often ‘scored up’ for pedal,
though in the case of J. L. Krebs’s setting of the same chorale, the original
c.f. could well have been for pedal en taille, rescored without pedal for the
printed version in his Clavier-Übung I (Emans KB).
BWV 696 Christum wir sollen loben schon / Was fürchtst

du Feind, Herodes, sehr
Further copies include J. C. Kittel sources.
For TEXTS and MELODY see BWV 611.
The first line of the chorale supplies the subject (see Example 227) including
both the important motif a and the tied or dotted fifth note that gives the
movement its flow and a certain similarity to the D major Fugue WTC2.
From a, the second part of the subject, derives the countersubject, as it does
in the C major Fugue WTC2.
Example 227
Though only twenty bars long, BWV 696 is subtle in its harmony,
counterpoint (subject + countersubject each entry), motifs (compare the
sequence in bb. 9–10 with bb. 2–3 of ‘Da Jesus an dem Kreuze stund’ BWV
621), and allusions (the final long phrygian e and its cadence come from
the chorale). The form is modest: a single exposition, episode and final
entry, the last with octave answer and a subject. This last, as elsewhere in
the ‘Seven’, is longer than in some fughettas, and rather Grigny-like when
supplied with ornaments, as here in P 1119 (late, unauthorized?).
Despite the final cadence, the ‘modality’ lies in the ambiguity of key, as
when the answer implies no clear dominant such as appears in the harmoni-
zation BWV 121. Moreover, the finely shaped a motif introduces unex-
pected chromaticisms and modulations and at other times moves in thirds
and sixths to create mellifluous false relations (bb. 4, 10, 11, 15, 18). The
436 BWV 696–698
harmony and texture have a richness and originality easily missed, though
not on a fine harpsichord.

Copies, similar to BWV 696.
The subject is the cantus’s first line, diminished (eight quavers) and livelier
than the ‘canzona’ BWV 723 or settings by Buxtehude and Böhm. However,
the same motif is inventively used again in a soprano recitative in Cantata
91 (‘Gelobet seist du’, 1724), and it seems unlikely that BWV 697 is earlier
than the cantata. Though quite different from BWV 696, this fughetta is
likewise original in all respects, with the quaver subject in most bars, and
semiquavers running down in all but the last bar.
An unusual feature is found in each exposition: the fourth entry in
b. 5 is at the fifteenth to the previous and follows a codetta bar, and the two
other complete expositions also have irregular entries (bb. 6–9 and 10–14).
The countersubject supplies the running semiquavers associated with the
Angel Throng of other Christmas chorales, particularly BWV 607 and 701.
These runs persist against the many entries of the theme (twelve times in
fourteen bars), as if expressing the repeated ‘Gelobet’ given out by the angels.
Christmas hymns were particularly appropriate to such Affekte as fluttering
angels’ wings, to judge by settings of ‘Gottes Sohn ist kommen’ and ‘Vom
Himmel kam der Engel Schaar’ by (e.g.) J. H. Buttstedt. Subtleties here
are a subject tending to supertonic or relative, a countersubject developing
inversus, and off-beat phrases aiding cohesion.
Although the cantus’s last line is not there, its mixolydian cadence is.
Tonics are generally avoided, and the sequences are rather disguised, both
idiom and contrapuntal complexity surely later than those of the Ob? As
a fughetta, it is the peak of a tradition, both as it had been in music for
organ (Pachelbel’s Magnificats) and harpsichord (Kuhnau’s Clavierübung,
1689).
BWV 698 Herr Christ, der ein’ge Gottes Sohn

Further copy via J. C. Oley.

437 BWV 698
Continuing the Seven Fughettas’ survey of fugal types, this has a subject from
line 1 of the chorale, with a countersubject of three motifs used throughout
(Example 228). The working-out of the countersubject is particularly in-
ventive – in sequence, in the bass, in the soprano – and one is hardly aware
of all the tonics (bb. 1, 7, 11, 17). The melodious lines are more conventional
than in the smaller chorales of Clavierübung III but work to similar ends,
with ingenious combinations.
Example 228
Thus in bb. 11–12, line 2 of the melody appears in the bass against both
subject and countersubject, and the following bass quavers look like a
diminution of this line. In bb. 15–16 the top part has a paraphrased version
of the cantus (Example 229); this means that bb. 15–16 see a paraphrased
line 1 plus original countersubject plus a diminished line 2, and in bb. 17–18
the second (= last) line of the melody can be heard paraphrased in the top
part against line 1 in the tenor. A fresh bass-line prevents all this adding up
to a merely empty quodlibet, and one which no doubt incorporates other
ingenious thematic diminutions.
Example 229
Also derived are the running lines against the familiar G-soprano pedal
point at the end (cf. BWV 541, 657, 668), just as every figura corta of the piece
probably comes from motif a in Example 228. Such complexity in a twenty-
bar miniature is reaching out to a new kind of expression, ‘unworldly’,
rather private, hard to put earlier than 1725, to judge by the chromatic
transformation of the cantus in b. 18 lh. The ingenuity results in a genuinely
new and newly expressive music.
438 BWV 699–700

‘Portfolio’ copies only.
With its subject from line 1 of the chorale, this matches BWV 696 but with
voices in reversed order. On paper the form is little more than exposition,
further tonic entry and coda, as in BWV 696. But comparison with the
coloratura setting BWV 659 suggests that the countersubject in b. 4 is derived
from the subject, circumscribing the notes D D C F and in the course of
sixteen bars achieving independent development with its diminished fourth.
See Example 230.
Example 230
(Such free-ranging countersubjects are found in the fugal sections of

chorale fantasias, e.g. Buxtehude’s ‘Nun freut euch’.) The result is a move-
ment as rapt as BWV 686 or 704, as original in demeanour, and as equally
suited to harpsichord: a beautiful piece.
A further ‘old’ sign is the broken figuration of bb. 10–11, but the subject
and its countersubject are strikingly like the double theme of the C minor
Fugue WTC1. The countersubject looks progressively more like a paraphrase
of the chorale melody, whose five entries in bb. 12–15 serve as reminders,
and it keeps a mystifyingly sad-winsome quality throughout.
BWV 700 Vom Himmel hoch, da komm’ ich her

‘Portfolio’ copies, also via C. F. Penzel and J. C. Kittel.
Two staves; headed by Penzel ‘Fuga sopra . . .’.
In part because of the pedal doubling, BWV 700 is usually described (with
BWV 741) as ‘very early’, though improved in slight detail during the com-
poser’s last decade (KB p. 11). Pedal doubling is also found in a setting
439 BWV 700–701
by Pachelbel (DTB 58), and BWV 700’s narrow pedal compass (C–c) may
reflect some such influence.
The form of the movement is unusual, early like the square 4 /4 rhythms
or Corellian moments (bb. 31–3), but soon with tell-tale signs of Bach’s
thematic allusion. Bars 1–23 are a conventional monothematic fughetta,
the subject line 1 of the chorale. But in b. 23 the subject is diminished and
answered by a new subject, line 2 now in diminution; this in turn has two
dominant answers before the c.f. appears against its diminished form in the
tenor. In b. 37 a theme derived from line 3 becomes the stretto subject of
the next section, and other motifs could be derived from other lines – e.g.
does the little dactyl come from the melody’s first four notes? In b. 47, line 4
supplies another subject partly in diminution, and a last pedal c.f. draws out
the end of the line (as in four variations in BWV 769). This final pedal entry
is striking enough to appear again at the end of the Canonic Variations –
not, presumably, borrowed from it by a copyist.
Such fugal treatment of more than line 1 is represented by examples
in the ‘Neumeister Collection’, and is later developed on a larger scale in
Weimar. Perhaps BWV 700 belongs to the ‘early layer of Böhm-imitation’ in
J. S. Bach’s chorale work (Zehnder 1988 p. 106) and perhaps it was part of the
hymn-plan for his first church-year in Arnstadt, along with BWV 705, 766,
739 and 724 (Krumbach 1985 II.14). Such speculation on the early biography
is almost limitless, and as likely is that in being a kind of miscellany-in-
itself, the setting is a ‘contribution’ to this chorale’s voluminous tradition,
from Scheidemann onwards. The tapestry of ordinary rhythms puts one in
mind of J. M. Bach, the phased fore-imitations Pachelbel, the pedal cantus-
crotchets Böhm, etc. The diminished reference to line 1 in the closing bars,
where a B counters the preceding dominant passage, could be a detail
picked up from J. M. Bach – if genuine.

Copies as BWV 699.
Whether or not the scales can be seen as a reference to the angels of

vv. 1–2 and/or the bells of v. 15 – did Lutheran bells peal in this way? –
this is a fluent, very musical exercise in the techniques of tonal answer
(bb. 3, 12, 20), countersubject scales (segments up and down), subject
diminution (line 1 semiquavers GFEF, line 2 quavers b. 10, etc), and the-
matic combination. See Example 231 for themes. There is motif b in stretto
440 BWV 701
Example 231
(bb. 10–11), c in stretto (bb. 16–18, 21), b plus a (bb. 12, 14, 20), c plus
a (bb. 15, 24). The scales run effortlessly, changing to other motifs only
when themes are combined (e.g. b. 15). The significance of other moments
is easy to miss – e.g. the syncopated alto of b. 9 appears again later, and the
alto of b. 10 is derived from the theme. C major scales as counterpoint to
‘Vom Himmel hoch’ were also attempted by the composer of Anh. II 64,
attributed to Bach in P 285.
Thus, although Christmas scales dominate the movement, there are more
than a few hints of the intricacies of the Canonic Variations, with a different
tone between simple sequences (bb. 6f., 22ff. – both like moments in WTC)
and thematic combinations (bb. 15–16, 19–20). The strict part-writing and
derived counterpoint suggest that this was the most mature fughetta (Burba
1994 p. 94). Like BWV 698, it is a web of allusion difficult to unravel, and
this itself becomes a species of musical language which, like the canonic
techniques on the same chorale in BWV 769, is sometimes ambiguous in
tonality. The key to Example 231 is as follows:
a line 1 (here functioning in turn as fugue answer, subject and entry)

b line 2, in diminution
c line 3, in diminution (subject and answer in succession)
x countersubject inversus
y line 4, in double diminution
The impression left behind is one of brilliance, indeed a seamless, ringing

bell-sound. The last four notes of the piece are surely singing ‘Hal-le-lu-ja’?
441 BWV 702
BWV 702 Das Jesulein soll doch mein Trost

B&H’s ‘portfolio’ copies including AM.B.72a (see May 1974a, Emans 1997
p. 31).
The TEXT of B. Helder’s New Year hymn was published in 1636.
Das Jesulein soll doch mein Trost, The infant Jesus should be and remain
mein Heiland sein und bleiben, my consolation, my Saviour,
der mich geliebet und erlöst; who loves and redeems me;
kein G’walt sol mich abtreiben. no power shall drive me from him.
Ihm tu’ ich mich ganz williglich To him I shall devote myself willingly,
von Hertzensgrund ergeben, from the bottom of my heart,
es mag mir sein weh oder fein, whether it goes well or ill for me,
mag sterben oder leben. whether I die or live.
The MELODY was published with the text: Example 232. It appears only
in BWV 702, whose unusually high cadence – unusual enough for some
to have doubted its authenticity – may invoke the last verse (Terry 1921
p. 142), particularly the line
Zum Leben fein zu gehen ein. To enter upon a pure life.
Example 232
Bad opinions of the part-writing and the pedal line (Keller 1937) are based on
the BG edition, which is unlikely to reflect the wishes of copyist or composer.
The combination of fugue-subjects drawn from the first two lines of the
cantus is not untypical of J. S. Bach, especially as the resulting counterpoint
is italianate. See Example 233. The upper parts in bb. 17–18 are also derived
from the subject, which enters at the end of b. 18. Varied, well-regulated
stretti (bb. 3–4, 7ff., 12ff., 19ff.) are more in line with Bach’s handling of
fugue-themes than Walther’s or even Pachelbel’s, as is the treatment of keys
and cadences, harmony, semiquavers, and rhythmic variety. Compare with
it BWV 693 on all five counts. The ‘conventionality’ of the episode bb. 17–18
442 BWV 702–703
Example 233
(according to Emans KB) is not egregious, and the very irregularity of such
a fugue might be characteristic of J. S. Bach:
1 subject (tenor), countersubject (alto) in dominant

3 subject (soprano), countersubject (bass) in tonic
4 countersubject (soprano) in octave stretto
5, 6 subject (tenor), countersubject (alto) in tonic to subject in tonic
(cf. b. 1)
7 subject (bass), answered by double stretto, etc.
The bass semiquavers of bb. 16–19 should presumably be up an octave in

the left hand – perhaps the original source was tablature? Like the final
tenor semiquavers, these lines are possible to believe as early work of the
Ob’s composer, though quite untypical are the spacing and tessitura in the
sources and the wandering keys of bb. 6–12 and 16–17. An early work of
J. L. Krebs aping his master?
BWV 703 Gottes Sohn ist kommen

‘Portfolio’ copies only.
Like others of the ‘Seven Fughettas’, this is little more than the exposition
of a subject derived from line 1 of the chorale, followed by an episode
developing the countersubject, and closing with a tonic entry, the whole
in twenty-two bars of rapt, ‘mystical’ counterpoint. The tonal answer is
irregular (as in BWV 701), and the codetta in bb. 7–9 (like some in WTC)
develops material used later in the movement – in this case, semiquavers
derived from the countersubject in b. 4 and then running through every bar
of the movement. So semiquavers lacked by the subject are supplied by the
countersubject, which itself might be derived. Though the end-result looks
slight, the relative complexity of all this means an unusual, remote musical
language, as with others of the group.
443 BWV 703–704
Though like the other Advent/Christmas countersubjects (BWV 701),

the semiquaver phrase appears on various degrees of the scale and so makes
a distinct contribution to the harmony as it fluctuates between E and
E – the only accidental of the piece, constantly returning and colouring
the lines throughout, surely not by accident. The tendency towards E in
F major pieces preserves something of the old Fifth Tone, and is found
in both sacred music (Pachelbel’s Magnificat quinti toni) and secular (here
and there in the F major ordre of Couperin’s Premier Livre). The chorales
BWV 618 and 619 have the same tendency, as does the Pastorale BWV 590
(where E is the first accidental, as too in the F major Toccata). It could
well be that the F major of BWV 703 with its E was deliberately contrasted
with the F major of BWV 704 with its lydian B, producing a new musical
dialect.
BWV 704 Lob sei dem allmächt’gen Gott

Copies as BWV 701.
For TEXT and MELODY see BWV 602. The opening of the melody in both
BWV 602 and 704 differs from the traditional form (see Example 7).
The fughetta, based on line 1, matches others of the ‘Seven’: exposition

(bb. 1–10), episode (bb. 10–11), tonic entry (b. 12), quasi-supertonic
(b. 15), quasi-submediant (b. 19) leading to close on A. The countersubject
(bb. 4–6) is developed in most of the remaining bars, many of which
combine its various motifs (e.g. a+b+subject in b. 16, a+c+subject in b.
17): Example 234. The little motif of (ii) is also important, as it is in other
fughettas (e.g. BWV 703), especially when spun out – compare bb. 20–1
with BWV 697. Such bars as 12–13 are like those in other 3/2 settings in
three sections with c.f., BWV 656 or 663.
Example 234
Although the cadence on A is the original plainsong’s, the movement has

had modal tendencies from the start, with a subject beginning on a nominal
mediant and including the lydian b : a somewhat ‘remote’ version of the
chorale melody. The firmest key is G minor in the middle, and even the final
entry is at first harmonized as F, not D minor. The conception may be not
444 BWV 704–706
so much an ‘adaptation of fugal plan to the anti-tonal modality of certain

chorales’ (Chailley 1974 p. 187) as a carefully detailed fugue brought to a
close on its first middle entry.
On the lydian element, see a remark at the end of BWV 703. Presumably
the absence or presence of ties in the subject is intended: absent from the
exposition but present in a sequence (bb. 15–19 alto) and final entry (bb.
20–1 bass). Again, technical complexity results in a musical dialect one hears
seldom outside the Seven Fughettas.

‘Portfolio’ copies, including P 1160 (for J. C. Oley); two staves.
Heading in MSS: ‘Fuga’.
BWV 705 deserves attention as a typical ‘organ motet’, so similar in style

and form to choral movements of Bach (e.g. in Cantata 2) as to look like a
transcription, with or without pedal. It was published as a motet in NBG 26
and has a c.f. in the same note-values as the stretto fore-imitations, the last
answer in each case the ‘true’ chorale-line. The result is a diatonically treated
modal melody, whose original fifth line (here from b. 73) was described by
Mattheson as beginning in the minor and ending in the major subdominant
(1739 p. 384).
The form follows the hymn’s (ababcade), as the close follows its cadence.
Occasional dorian turns of phrase (e.g. bb. 20 or 42) are more evident than
in the cantata settings of the melody, and there is virtually no development
of motif compared to BWV 737’s. On the other hand, Oley’s copyist also
provided BWV 664b, 698, 712 and 713, indisputably authentic works. As
often in classical ricercars and capriccios, a repeated-note subject automat-
ically leads to stretti, and the counterpoint, though simple and anonymous,
hardly puts a foot wrong.

Further copies by J. T. Krebs and via J. C. Oley.
Two staves; harmonization headed ‘alio modo’ in P 801.

445 BWV 706–707
The usefulness of this text and melody in connection with the Sunday ser-
mon (cf. ‘Herr Jesu Christ, dich zu uns wend’) may explain the array of
settings which, though varied in technique, leave the melody immediately
recognizable. In key alone, BWV 706.i, 706.ii, 634, 633 appear to belong
together (this order in P 801). But since they are not an obvious set of vari-
ations, perhaps BWV 706 originated as an exercise for Krebs c. 1710, and
perhaps even developed later into the canon BWV 634 complete with Ob
motifs (Zietz 1969 p. 130, also KB p. 11).
Whilst the second setting here looks like a simple vocal chorale of the
kind called ‘un-Bachisch’ by Spitta (I p. 588), the first is a more idiomatic
harmonization for keyboard, with the makings of true motivic lines –
elementary but, in the final scales, promising.
BWV 707 Ich hab’ mein’ Sach’ Gott heimgestellt

‘Portfolio’ copies, two staves.
The TEXT of J. Leon’s hymn was published in 1589 and became associated
with ‘death and resurrection’ (Freylinghausen 1741 etc).
Ich hab mein Sach Gott heimgestellt, I have placed my cause in God,
er machs mit mir, wie’s ihm gefällt. he does with me as he pleases.
Soll ich allhier noch länger lebn, Should I live on earth longer,
ohn Widerstrebn without resistance
sein’m Willen tu ich mich ergebn. I will give myself to his will.
v. 3
Es ist allhier ein Jammertal, Everywhere here is a vale of tears,
Angst, Not und Trübsal überall . . . all anxiety, distress and trouble . . .
Fifteen further verses trace the soul’s conversion from misery to hope and
praise.
The MELODY of Cantata 106 is the tenor of a song published in 1589, ‘Ich
weiss mir ein Röslein hübsch und fein’ (parody of an earlier song?); that of
BWV 707, 708 and 1113 (Example 235) is the soprano.
Thought by Spitta to be by Walther because of its canonic technique
(bb. 15–16, 72, 97–9: I p. 820), this resembles BWV 705 and 737, also in
the matter of the pedal. The direction manualiter in one copy (P 1160)
may reflect uncertainty: the pedal plays the lowest voice all through,
despite some unidiomatic moments, or it does not play at all. Keller (1937)
sees as more Bach-like the chromaticism of bb. 52ff., but both this and the
446 BWV 707–709
Example 235
other expositions suggest at best an acquisitive pupil, one capable both of

infelicities (alto b. 8, bass bb. 56–8 etc.) and of rich and imaginative har-
mony. The counterpoint of bb. 89ff., countersubject of bb. 28ff. and 108ff.,
and harmonization of the unpromising final line (bb. 128ff.) could be those
of a gifted learner – perhaps even the young Bach – as could the consis-
tent appearance of the figura corta and the countersubject note-patterns:
crotchets or quavers, scales or in-turning figures, diatonic or chromatic,
syncopated or unsyncopated, etc.
Perhaps the closing harmonization is authentic, showing many hall-
marks of the four-part Bach chorale and lending authority to the ‘organ
motet’.
BWV 708 and 708a Ich hab’ mein’ Sach’ Gott heimgestellt
‘Portfolio’ copies for BWV 708, later for BWV 708a; both in Brussels Fétis
3237 C Mus, complete with BWV 760 and 761, all attributed to J. S. Bach
(C. G. Gerlach, c. 1730).
Despite the interesting sevenths, neither harmonization is obviously an

‘organ chorale’, presenting the melody in duple and triple time but showing
no sign of authentic handling – less so, therefore, than the simple chorale
closing BWV 707.
BWV 709 Herr Jesu Christ, dich zu uns wend’

Further copies by or via J. C. Oley, C. F. Penzel and J. C. Kittel.
Two staves in the older copies; headed in Lpz III.8.10 (contemporary?),

‘à 2. Clav. e Ped.’; ornaments mostly in Penzel.

447 BWV 709–710
Like the Ob setting BWV 632, the complete melody is in the soprano without
interludes,∗ and with an accompaniment exploring several motifs in all three
parts. In the decorated melody (the notes of which occur on the main beats)
and the quality of accompaniment, BWV 709 is almost a match for the
Ob, although the long notes beginning each line suggest somewhat earlier
origins, as does the final.
Several little patterns, a, b and c (Example 236), draw attention away
from and affect the already disguised melody, which includes several patterns
familiar in Ob chorales BWV 622, 641 etc. Such a bar as 18 incorporates
a, b and c in one part or another, perhaps a little too single-mindedly.
Example 236
There might be more imitation between the inner voices than is customary
in ornamented chorales, but b. 16 (with its thirds and pedal motif) is a
rehearsal for other Ob work, its harmony almost as advanced. Similarly, the
cadences of bb. 4–5 and 14–15 seem to look ahead to a longer chorale like
‘Nun komm, der Heiden Heiland’ BWV 659a, as does the (unnecessarily?)
long final.
BWV 710 Wir Christenleut’ hab’n jetzund Freud

Further copies via J. C. Oley, J. L. Krebs (SBB Mus. MS 12012/6) and
J. C. Kittel.
Two staves; headed by Krebs ‘für 2 Clav. u. Pedal’.
For TEXT and MELODY see BWV 612. The melody used by Krebs (EB
6589) includes a repeat of line 1, as do the versions in Cantatas BWV 40,
110, 142, 248.iii, and BWV 612 (Ob). One of the hymnbooks which the
composer seems to have known (Darmstadt 1699) also omitted the phrase
(Luedtke 1918 p. 48).
A tradition that J. L. Krebs was the composer – probably because his

name appears at the top of the copy – is now discounted (Tittel 1966
∗ The rests at the ends of the first three lines could be ignored and the note sustained, as is not possible
at the corresponding points in BWV 632.
448 BWV 710–711
pp. 133–4). As in BWV 694, the two-part invention above pedal c.f. is such
that the hands do not cross; and as in BWV 695, the subject is clearly derived
from the chorale melody (Example 237). Although the hands tend to follow
the contours of the pedal, rising and falling with it, the theme and its motif
seem to develop independently, especially a, whose many appearances lead
one to hear ‘Wir, wir’ being repeated.
Example 237
There is much internal repetition, as there is in the original chorale,

where the same notes are heard during lines 3, 5, 6 and partially in 4 and
7. Motifs and phrases are reharmonized or parts exchanged, so that the
counterpoint of bb. 28–31 is immediately inverted when the cantus firmus
repeats its phrase.
When a is inverted in b. 7, it paraphrases line 2 (DCBA) just before the
pedal has it, a striking coincidence with the inverted motif in BWV 694. In
one form or another this motif comes to dominate the movement, even in
thirds, and as often with settings of melodies which move largely by step,
thematic allusion is not difficult to find in the upper parts – here to DCBA
or ABCD. Moments of chromaticism in the middle, an uncertain direction
of key around b. 39, or an ending surprisingly succinct for a pedal point,
leave behind an unusual impression, like the Ob setting less jolly than the
text might lead one to expect.

‘Portfolio’ copies.
Headed ‘Bicinium’ (authentic?).
Although the lh of BWV 711 is more like that in known bicinia of Johann
Bernhard Bach (e.g. ‘Nun freut euch’) than Böhm (‘Auf meinen lieben Gott’)
or Walther (‘Durch Adams Fall’) or even J. S. Bach (BWV 718), there is no
particular reason to ascribe it to him (KB p. 41). The melodic character
looks ahead to one of the Schübler Chorales as the final ritornello does to
449 BWV 711–712
another (BWV 649 and 646), and like so many Bach settings is specific to
its type: here, in the bass, a vigorous melody.
The piece may have been intended as a partita movement, by itself or
with BWV 716 and 717 or other settings. As in J. B. Bach’s ‘Nun freut euch’,
the first line is paraphrased in the opening (and closing) left-hand theme:
Example 238. Cohesion in BWV 711 comes partly from the first and penulti-
Example 238
mate lines of the cantus being similar, the ritornello shape giving almost the
impression of an ostinato. Telemann’s bicinia (e.g. ‘Allein Gott in der Höh’)
also involve broken-chord figures, some paraphrasing the chorale, but with
a less rounded shape. BWV 711’s left-hand line may paraphrase the opening
of the chorale but soon passes without break or change of direction into its
own sequences, hiding the tune by integrating with it. Its rhythmic drive
and harmonic tension are unknown to Telemann or Kauffmann but typical
of Bach’s Weimar works.
BWV 712 In dich hab’ ich gehoffet, Herr

Further copy via J. C. Oley.
For the TEXT, see BWV 640.
The MELODY is one of two used by Bach for this text, published by Seth
Calvisius in 1581. See Example 239. The simplified version in BWV 712 is
as in Cantatas 52 (1726), 106 (funeral c. 1707), the St Matthew Passion, and
the Christmas Oratorio.
Example 239
If BWV 666 really belongs to ‘The Eighteen’, so might BWV 712 have done.
From the chorale lines as they appear in BWV 52 etc., fugue-subjects are
derived:
450 BWV 712–713
Example 240
A 1–5 1, clearest in soprano and bass (canon, b. 2): Example 240

B 5–10 2, clearest in soprano
C 10–15 3, clearest in soprano; subject similar to A
D 15–19 4, clear in all voices
E 19–22 5, clear in all voices; subject similar to D
F 22–end 6, clearest in bass (b. 30)
The increasing tendency towards running figures in the final twelve bars –
though not the increasing chromaticism – may reflect the melisma at the
end of Calvisius’ original melody, surely known to his Leipzig successor and
relating to v. 1 of the text. Since there is no c.f. as such, simple or ornamented,
the lines that emerge give the appearance of growing out of the fugue’s
subjects, rather than vice-versa. The effect is of an organic, original series
of fughettas.
The character of the ‘fughettas’ varies. A and B are regular stretto ex-
positions, C has a series of answers on three different notes plus stretti,
each of D’s answers appears a fifth up, E is similar to D but a step higher,
F is irregular. The lifting of section D up a step is highly unusual and only
hinted at in the melody itself. All sections become smoother as the chorale-
line emerges, the texture as a whole idiomatic to the harpsichord. Answers
other than tonic and dominant remind one of stretto fugues in WTC1
(C major) and WTC2 (D major), and at least the first theme resembles the
C major Fugue WTC2, as do the movement’s increasingly faster figures.
Bars 20–30 are particularly harpsichord-like, though all the patterns are
typical of compound-time chorale fugues (compare with BWV 673 and
674, even 679) or similar sections in chorale fantasias (e.g. Buxtehude’s
‘Wie schön leuchtet’). The end-result has the conviction of a self-contained
genre, the ‘jig chorale’, versatile in its application.
BWV 713 Jesu, meine Freude

Further copies by or via C. F. Penzel and J. C. Oley.
Headed ‘Fantasia super . . .’ (authentic?).

451 BWV 713
The shape is unusual:
A 1–52, a continuous two- and three-part fugue on a subject which,

modified, serves as a countersubject to each of the first six lines of a
c.f. (S T B T B S), prefaced and interspersed with regular fugal
entries
B dolce 3/8, an imitative three- and four-part section paraphrasing
(or following the harmonies of) the final four lines of the cantus,
whose text begins ‘Lamb of God, my Bridegroom’
Both sections are skilfully kept manualiter, and the form of the cantus makes
it clear that AB is the correct form, not BA. (On comparison with the
‘Neumeister’ version of BWV 714, one might expect the stricter section to
be second rather than first.)
The fugues are essentially in two voices, with a third added between the
cantus phrases, when the subject usually appears. Their development is re-
stricted by the cantus in two ways: the key returns constantly to E minor; and,
presumably because a ten-line hymn is long, there is little interlude devel-
opment (e.g. no sequential treatment of b. 46, such as one might expect). As
in BWV 710, invertible counterpoint plays an important part (bb. 26–39 =
bb. 1–14), though chorale line 6 (b. 49) is newly harmonized. And through-
out, motifs are intensively explored, particularly semiquaver fragments.
The change in metre at b. 53, and a new way to use the cantus, is more
striking than in the motet ‘Jesu, meine Freude’ BWV 227 (verse 3), where
at the same point in the stanza a similar paraphrase emerges, but without
the organ-chorale’s change of direction. Here the cantus becomes extensile,
pulled out over a long phrase; see Example 241. Spitta described the chorale-
melody here as ‘varied freely in the Böhmian manner’ (I p. 601), but there
seem to be no antecedents. In the motet,∗ the thirds and the dolce motifs are
much less foreign. In the organ chorale, the section becomes a movement in
Example 241
∗ Sincethe cantus version used in BWV 713 is closer to that of BWV 610 than BWV 227, probably
the organ chorale preceded the motet (Meyer 1974 p. 85). However, the Choral which follows
BWV 713 (NBA IV/3 p. 57) has the passing-note of Bach’s ‘Leipzig arrangements’ of the melody, so
added later?
452 BWV 713–713a
itself, with its own developed motifs and keyboard-like cadences (cf. BWV
695). So subtly do the last three chorale lines pervade the section that several
suggestions can be made as to where they fit. Keller hears the final line in
the last ten bars (1948 p. 180), but it fits bb. 77–82 equally well.
In bar-numbers the two sections are similar in length; but since a propor-
tional tempo was probably intended (minim = dotted crotchet), the sym-
metry is only visual, and the playing time reflects the chorale’s structure of
six lines to four. The last bar of the 4/4 section has a ritardando so that its
last four lh semiquavers match exactly the first rh four of the 3/8 section.
As with BWV 691 and 695, it is uncertain whether the simple chorale at
the end goes back to Bach or a copyist, or if it speaks for a common practice.
See also a remark on BWV 957.
BWV 713a Jesu, meine Freude

Late copies only (L. Scholz: source as for BWV 638a, 691a).
This differs chiefly in two respects: the key (D minor), and the c.f. given
to pedal in the bass octave, as are non-thematic phrases. Obviously an
arrangement.
Miscellaneous chorales BWV 714–765

Copy in ‘Neumeister Collection’; second part, also in J. T. Krebs and Walther
copies.
Two staves; headed in P 802 ‘per Canonem’.
The TEXT of J. Major’s Lenten hymn of 1613 has six verses; four more were
added by 1625, moving the association to Passion.
Ach Gott und Herr, Ah God and Lord,

wie gross und schwer how great and heavy
sind mein begangne Sünden! are the sins I have committed!
Da ist niemand There is no one
der helfen kann who can help
in dieser Welt zu finden. to be found in this world.
The MELODY appears in the minor (J. Crüger 1640, BWV 714) and major
(Cantata 48, BWV 255, 692, 693): Example 242. Listed in the Ob.
The prelude to the canon is a 37-bar fantasia based on the rising and falling
lines of the chorale-melody, in a natural, adept and reasoned counterpoint,
with incipient motifs (soprano from b. 1 and alto b. 38; soprano of bb. 28–9
and bass of bb. 49–50). Partly with durezze, partly imitating a string prelude
to an early cantata, this quasi-improvised section might well represent an
early model prelude, as in turn the canon’s separate cantus phrases in the
soprano could be the hymn as sung, with inter-line interludes. The canon
is less strict than the Ob’s ingenuity would make it (e.g. b. 17); it must be
earlier, or sources corrupt. (The fifth tenor phrase could have appeared in
the soprano a bar earlier if the alto allowed for it.)
The distribution of parts in both sections, especially the bass, is uncertain,
and the two staves suggest a score playable in various ways. Pedal seems
necessary in the prelude, though the C might imply that the score is an
‘ideal’, adapted in practice. Thus the four parts can be manualiter; or pedal
plays bass; or soprano canon is on a separate manual; or pedal plays the tenor
canon at 8 . Such ‘adaptability’ is common to such canons, and marking each
cantus line ‘Choral’ encourages a solo registration.
As in BWV 693, the theme and its derived countersubject use note-
[453] patterns (scale-fragments) open to development up or down or diminished,
454 BWV 714–715
Example 242
so that the canon becomes as much a ‘fantasia on a theme’ as the prelude

has been. For a comparison with J. L. Krebs, see BWV 744 or Krebs’s ‘Herr
Gott, dich loben alle wir’, where the counterpoint is more natural and less
doctrinaire than e.g. BWV 693. (He, rather than Walther, was taught by J. S.
Bach?) That there was interest in canonic settings amongst the Thuringian
organists is clear also from works of Andreas Armsdorff (†1699), known
sometimes from copies made in the Walther circle.

Copy by J. P. Kellner (after 1727?).
Such harmonizations with interludes between lines were called Arnstädter

Gemeindechoräle in Keller 1948, i.e. settings supposedly used by the young
Bach to accompany the congregation in Arnstadt, and partly anticipated in
the ‘Neumeister Collection’. See also BWV 722, 726, 729, 732, 738. Copies
made in the Weimar years (two-part ‘skeletons’ of BWV 722, 729, 732, 738
in P 802) and later (written-out harmonies in Mempell–Preller MS) say
nothing about when they were composed. Spitta’s idea that the settings’
bold harmonies date from after the visit to Lübeck was a guess, no more
(I pp. 309f.), as was Keller’s dating to Arnstadt.
Whether the six settings are evidence for the Arnstadt Consistory Court’s
reprimand of Bach in 1706 for playing ‘many curious variations in the hymn’
and ‘mingling a wandering key in it’ (‘in dem Choral viele wunderliche
variationes, einen tonum peregrinum mit einbringen wolte’ – Dok II p. 20) is
not established by the sources. They could be later works, solo organ-settings
or models for Weimar pupils. Nor is it clear what ‘tonum peregrinum’
means: a musically knowledgeable clergyman’s term for ‘strange tones’ (as
often translated), a modal Magnificat melody of that name (unlikely), notes
strange in certain temperaments (as for Treiber’s chorales in Der accurate
455 BWV 715–716
Organist, 1704), or unexpected harmonies in final verses (as still common

in English cathedrals)? How rare such complaints were is unknown, and
clergy could have said such things of many kinds of Bach setting, including
some of the Ob and Clavierübung III.
Chromatic changes in the melody, uncertainty now and then when the
next line is to be sung, and the length of final cadences are reasons to think
the settings unsuitable for accompanying a congregation (Sackmann 1998),
but against this is sheer custom: these hymns were deeply familiar and were
doubtless sung in a rougher way than cantatas. Also, interludes can be played
with rallentandi as to imply the next line to be sung, as was the case in other
Protestant traditions of the time. It is true that in BWV 715 the inter-line
runs do not prepare the following chord, and at one point (b. 2) agree with
neither the preceding nor the following harmony; but there is no insuperable
problem here, especially if chorales were sung slowly and ‘lining out’ was
practised, i.e. the next line’s words were read out for those who could not
read.
Similarly, when in 1770 J. F. Agricola reported that his former teacher
Bach had no regard for inter-line interludes (see Czubatynski 1993), the
implication is at least that they were still well known, even perhaps that
he himself had once added them. As in England in the eighteenth century,
such complaints about ‘objective’ inter-line interludes were part of the move
towards a more ‘subjective’ hymn-playing. Signs that these settings are rep-
resenting the text’s Affekt in some way need not mean they were for ‘purely
solo performance’ (Sackmann 1998 p. 249), nor despite some superficial
resemblances in the roulades themselves have such inter-line interludes any-
thing to do with scrabbling embellishments added, on uncertain authority,
to Corelli’s Sonatas Op. 5 (Amsterdam 1710).
Why such organ chorales are represented in J. P. Kellner’s voluminous
fund by only some of them (BWV 715, 722, 726, 732: Stinson 1989 p. 52) is
unknown, nor on whose authority four were grouped as ‘Four Christmas
Chorales’ in Mempell–Preller’s Lpz MB MS 7 (Vier Weynachts Chorale). But
the other two, BWV 715 and 726, could have been played on each Sunday,
and are thus models for less common hymns. Also unclear is whether in
these hymn-settings the pedal is more than optional for a bass-line of some
complexity: see heading for BWV 722.

Copies in P 1160 (owned by J. C. Oley) and late MSS (P 285, 311).

456 BWV 716–717
This has a mixed form: regular three-part fugue (without pedal?) on line 1,
paraphrased subject for line 2 (bb. 29ff., third voice for pedal?), new fugue
on line 1 (end b. 56, alto), with final pedal cantus firmus of two lines and
final soprano answer. The crotchet line begins for the line 2 paraphrase and
is then sustained. The use of partial c.f. at the close makes sense because the
chorale melody itself ends as it begins – for which see also BWV 664 and
(especially) 724.
Although the copies probably all derive from the same source and can
not therefore confirm the attribution (Emans KB), the general competence
of the lines would not rule out the young Bach, composer of BWV 724,
nor would the setting’s unusual (experimental?) form. The likelier it is
that Bach wrote (most of) the ‘Neumeister Collection’, the more it is that
BWV 716, 723 and 724 need to be considered along with them as part of
the same repertory.

Copies by or via J. T. Krebs and Mempell–Preller.
Like the first part of BWV 713, this is a two- and three-part manualiter
fugue built predominantly on a derived subject against whose motifs
(rather than the subject itself) all the lines of the chorale appear as cantus
firmi. See Example 243. Unlike BWV 713, it keeps the cantus in the
soprano.
Example 243
Motifs a and b remain important in both rectus and inversus forms,

changing the direction of the second subject (b. 35), which looks as if it means
to paraphrase line 5 of the melody (b. 39) but becomes more fancy-free. The
subject fits both line 6 (b. 47) and the final pedal point, with motifs producing
sequences and imitation. The result is very like harpsichord music, such as
the fugal gigues of the English Suites (e.g. bb. 50–6). A sign of the subtle
pervasiveness of the subject, already admired by Spitta (I pp. 597–8), is
that the motif b usually appears on weak beats, and weak beats are usually
characterized by motif b. Such pervasiveness clearly surpasses the simple
paraphrases of a Scheidt or Pachelbel, with the chorale’s notes on the fugue-
subject’s main beats.
457 BWV 718

Copies by J. L. Krebs and via J. C. Kittel.
Two staves; headed by Krebs ‘à 2 Claviers et Pédale’. For registration, see
below.
Sources suggest that the movement circulated in versions – perhaps revised

in part by the composer – differing in such details as ornaments in either
hand, in the ‘forte’ marks, in the ‘Allegro’ sign, and in the manual-changes:
Krebs Ow and Uw or P (P = Positiv, Unterwerk = forte?)

Others Ob and R (Oberwerk, Rückpositiv) or piano forte (piano = Ob
and forte = Rp)
Variants in ornaments and manual-changes could also reflect late copyists’

uncertainty as to old practices in chorale fantasias. The implications are that
lines in a bicinium were ornamented, and that two manuals were to be used
for echoes or melody-with-accompaniment. On such small-scale fantasias,
see also BWV 739.
Since at least the time of Spitta (I pp. 210–12), the influence of Böhm has
been seen in BWV 718: the opening bicinium and the tonic–dominant
alternations (b. 43) are ‘unifying Böhmian characteristics’ taken from
various pieces (Zehnder 1988 p. 92), though Böhm could also have been
influenced by Bach. Spitta also heard elements of Pachelbel in the interludes
(e.g. bb. 13, 24) and guessed it was ‘conceived for pedal harpsichord’, perhaps
because of the last note, which, however, would suit a Rp Sesquialtera stop
and its major third. Parallels can be drawn between the opening bass and
continuo arias in early cantatas, including quasi-ostinatos in BWV 4, 71 and
106. Though called the composer’s only chorale-fantasia (Dietrich 1929
p. 7), it is true to type only in its echo passages; the modest length and
clearcut sections suggest rather an attempt to survey some of the variation
techniques known on a bigger scale in the chorale partitas. Thus the short-
phrase imitation from bb. 24ff. is not far from Scheidt’s setting of the same
chorale melody.
Long by Thuringian standards, it applies a series of ‘northern’ techniques:
A 1–13 lines 1 and 2, bicinium. For motifs, ornaments, melody

and quasi-ostinato, see Böhm’s ‘Vater unser’
B 13–33 lines 3, 4 and 5, cantus anticipated but no fore-imitation;
lines increase in length (line 5’s semiquavers: cf. BWV 4.v
or 766.iv)
458 BWV 718
C 33–42 line 6, partial paraphrase in triplets, sequences as in

Pachelbel etc.
D 42–61 line 7, echoes (both octave and manual-change) in
sequence, after halfway point (cf. Buxtehude’s ‘Gelobet
seist du’)
E 61–73 line 8, three minim statements (last on pedal,
cf. BWV 656), with derived line first in rh, secondly lh,
thirdly both; leading to
F 73–end coda; rh runs on after cadence, as in Reinken etc.
(cf. BWV 720).
The sections are articulated by clear tonic or dominant cadences.

The old-fashioned square echo figure from line 7 and the simple para-
phrasing of line 8 show a composer very familiar with these devices
(Example 244). Such patterns running against a minim cantus are typical
Example 244
of northern composers. The echo section, being repetitious, pictures the

words at this point (‘and singing Hallelujah’), as in effect does giving a
simple c.f. to each voice in bb. 63, 68 and 71 – ‘Hallelujah’ is the last word in
each verse. The ‘objective’ nature of a bicinium suits the subdued nature of
line 1 (as in Böhm’s ‘Vater unser’), while the jig of line 6 could as well suit
the Affekt of line 5 (‘joyful’). In general, the harmonies are still standard,
decorated rather than generated by the motifs.
Two performance problems arise: the tempo, and the distribution of
manuals for section C. Having so many ornaments implies a slower tempo
for sections A and B, hence ‘Allegro’ at b. 24 in a source giving the ornaments
(J. A. G. Wechmar). Also, the sections of longer North German fantasias
often seem to require changes of tempo, as too they do changes of registra-
tion. The choice of manual for section C is unclear: both the spacing in b. 35
and the piano in b. 41 may suggest both hands on the louder manual up to
the second half of b. 41 – in which case both hands may also have played on
the louder manual from b. 27 onwards. Pedal is necessary only for the final
two bars, and even these may have been governed by the same conventions
as in (e.g.) the Fantasia BWV 561.
459 BWV 719–720
BWV 719 Der Tag, der ist so freudenreich

Copies: HK Berlin Sp 1491 (with the 44 Choraele attributed to J. C. Bach
†1703) and Yale LM 4708 (‘Neumeister’, subtitled ‘oder Ein Kindelein so
löbelich’).
Spitta, who owned the first MS, describes it as written about 1700, but it is
now thought to be later than 1719 (Wolff 1997 p. 159). Why BG 40 arranged
it on three staves, and what version the melody from b. 34 had been based
on, are both uncertain. The two copies show signs of going back directly
or indirectly to a common source; in ‘Neumeister’, the chorale is the first
to be attributed to J. S. Bach, after other uneventful Advent and Christmas
chorales by J. M. Bach and Zachow. LM 4708’s alternative or subtitle text is
v. 2 of ‘Der Tag, der ist so freudenreich’, also found as an independent hymn.
The first two lines of the melody are treated fugally at some length,
separated by an ‘interlude flourish’: the first with seven entries, the second
in diminution until the last entry. Pedal is not necessary, though optional
for the two cantus phrases (first eight notes of bb. 5–7 only, and bb. 22–4)?
At first glance rather primitive, the counterpoint is not unskilful, and the
length and the flourishes at the Golden Section (bar 24 of 38 = 1 : 1.58) and
the close itself are all credibly early work of J. S. Bach.
BWV 720 Ein’ feste Burg ist unser Gott

Copies by J. G. Walther, J. T. Krebs, and in the Plauener Orgelbuch (before
1710? from Walther?); and later.
Two staves, three for trio sections; headed by Plauener ‘à 3 Clav. e ped.’, by
Krebs ‘a 2 Clav: e Ped:’; for Walther’s registration, see below.
The TEXT of Luther’s hymn, a free paraphase of Ps. 46, became associated
with the Third Sunday in Lent and with Reformation Day (Stiller 1970
p. 226).
Ein feste Burg ist unser Gott, A firm stronghold is our God,
ein gute Wehr und Waffen. a good defence and weapon.
Er hilft uns frei aus aller Not, He helps us out of all distress
die uns jetzt hat betroffen. that has come upon us now.
Der alte böse Feind The old wicked enemy
460 BWV 720
mit Ernst er’s jetzt meint; means it now in earnest;

gross Macht und viel List great power and much cunning
sein grausam Rüstung ist, are his fearful armament,
auf Erd ist nicht seinsgleichen. there is no equal to him on earth.
Three further verses combine defiance and faith.
The MELODY, said to be adapted by Luther from a plainsong, was published

with the text in 1531: Example 245. Harmonized in BWV 302 and 303, listed
in the Ob and used in Cantata 80 (Reformation 1724) and 80a (Weimar,
Lent).
Example 245
Although sometimes likened to much longer North German chorale fan-

tasias, this setting, like BWV 718, is a ‘catalogue’ of different treatments:
A 1–20 lines 1, 2 (3, 4) in dialogue, one or more times in each

hand, decorated or paraphrased
B 20–4 line 5, paraphrase, two-part imitation above basso continuo
C 24–33 lines 6, 7 c.f. in pedal (cf. words of v. 1, lines 6, 7)
D 35–9 line 8 paraphrase, two-part imitation above pedal
E 39–end line 9 as a bicinium; restated in four parts (plus previous
motif)
As well as the simple decorations of line 1 in bb. 1, 4 and 12, and of line 2
in bb. 8 and 16, there is a tapestry of paraphrases (Example 246). As is
fitting for a motif derived from a line that occurs three times (lines 2, 4, 9), a
appears constantly throughout the piece. Line 5 (which emerges through the
lh phrase bb. 20–4) and line 8 (ditto, rh bb. 35–7) are spun out and disguised,
appearing between simpler chorale lines. Every line of the chorale melody
is there as a strand in the tapestry.
The registration, layout and supposed origin are interconnected. From
at least Spitta onwards (I pp. 394–7) the setting has been linked with a
461 BWV 720
Example 246
putative opening of the rebuilt three-manual organ at the Divi-Blasii Church

in Mühlhausen in 1709, composed to show its colourful possibilities by the
organist who had advised on the instrument. As with the ‘Arnstadt Chorales’
(see BWV 715), this link is over-simple:
(i) The musical style suggests a date earlier than 1709: the square rhythms,
texture, parallel motion of such passages as bb. 50–3, conventional motifs.
Also conventional are the running passages leading to the next chorale line
and rounding off the pedal point. Yet though repetitive and partita-like, the
motifs are treated inventively, e.g. the two suspirans in bb. 35 and 51.
(ii) There is no evidence of an opening dedicatory recital on the re-

built organ, and though the headings and ‘Fagotto/Sesquialtera’ registra-
tion in Walther and Plauen are plausible, they cannot be authenticated.
Mühlhausen’s three manuals had these stops; but the Fagotto compass may
have been only C–c or C–c (J. Adlung, Musica mechanica organoedi, 1768
pp. 92, 260) and thus unsuitable. A dialogue of Sesquialtera and Fagotto was
described by later writers such as Adlung and was common, even generic.
Pedal reed for the c.f. in bb. 25–32 might reflect the Gravität of the new
Posaunen Bass as desired by the composer (Dok I pp. 152–5); but pedal
reeds of a suitable kind were more common in the area, both then and later,
than the Rückpositiv required for the section beginning b. 20.
(iii) There is nevertheless a clear opportunity for organ effects: dialogue,

manual and pedal c.f., and solo, duo, trio and quartet passages. (Compare
462 BWV 720
J. N. Hanff’s ‘Ein’ feste Burg’, copied by Walther.) It is unusual to find both

registration and manual indication so explicit in a source, but in the Plauener
Orgelbuch at least one movement by Walther (‘Hilf, Gott, dass mir’s gelinge’)
was registered even more fully than BWV 720 (Seiffert 1920 p. 373):
Rp (c.f.) forte, Prinzipal 4 , Sesquialtera

Ow (accompaniment) piano, Viola da Gamba 8
Ped (c.f. in canon) Cornet 2
All four stops were not only at Mühlhausen (a typical if large Thuringian
organ of the kind described by Werckmeister) but also in Walther’s church
at Weimar.
(iv) Nothing proves or disproves the Mühlhausen association. The (wide-

scaled?) Fagotto, complete or not, would probably have been at 16 ;∗ also
uncertain is whether ‘Sesquialtera’ meant a stop or a registration. In b. 20,
Rückpositiv seems to apply to both hands (specified for left hand in P 802),
presumably with a plenum; this implies that the Sesquialtera was a Brustwerk
stop (or registration) and the Fagotto was on the Oberwerk. (Rests occur so
conveniently as to make it possible for the Fagotto to belong to any manual –
hardly an accident?) In b. 24, the lh Oberwerk suggests a plenum of some
kind, the right hand still on Positiv.
From b. 25, the two Walther sources have the pedal c.f. an octave lower ‘to
obtain the effect of the 32 Untersatz at Mühlhausen’ (Klotz 1975 p. 386),
producing a strange three-part spacing. (To obtain the 32 effect or to imitate
it?) The free lh in b. 34 enables a pedal reed to be taken off, and Oberwerk
in b. 39 suggests that in b. 35, lh plays Rückpositiv, which seems desirable.
Whether Oberwerk applies also to the rh in b. 41 is unclear; from there to
the end, it could play either, but probably Rückpositiv. Perhaps from b. 50
the lh plays on the Rückpositiv too, since that would probably be the louder
manual unless an untypical registration change had taken place near b. 39
or b. 50.
As in the longer works of Bruhns and others, the changing texture of
such pieces does allow licence in the use of manuals, and was no doubt
meant to. Perhaps Walther was merely adding his own suggestions, as he
did elsewhere? – though they would have been close to Bach’s too, no
doubt.
∗ This would not make it unreasonable for the opening phrase: the downward run across two octaves
would demonstrate any reed, and no organist in 1700 was so dominated by the 8 norm as his
descendants.
463 BWV 721
BWV 721 Erbarm’ dich mein, O Herre Gott

Only copy, by J. G. Walther.
Two staves.
The TEXT of E. Hegenwalt’s hymn of 1524 is a translation of Ps. 51, as-

sociated with ‘penitence and conversion’ (Freylinghausen 1741), and with
the Third, Eleventh, Fourteenth and Twenty-second Sundays after Trinity
(Vopelius 1682).
Erbarm dich mein, O Herre Gott, Be merciful to me, O Lord God;

nach deiner gross’ Barmherzigkeit according to your great mercy
wasch ab, mach rein mein Missethat, wash away, make clean my misdoing;
ich kenn mein Sünd and ist mir leid. I know my sin and am sorry for it.
Allein ich dir gesündigt hab, Only against you have I sinned,
das ist wider mich stetiglich; that is constantly against me;
das Bös vor dir nicht mag bestahn, evil cannot exist before you,
du bleibst gerecht, ob man urtheile you remain just, however one judges
dich. you.
The MELODY accompanied the text and, like most hymns in J. Walther’s
books, was later simplified: Example 247. Listed in the Ob and harmonized
in BWV 305. The sharpened third note occurs also in Busbetzky (see below).
Example 247
Authenticity has often been doubted because of the texture: repeated three-
and four-part chords, without break from first bar to last, below a soprano c.f.
with interludes. No other example is known by J. S. Bach or contemporaries,
even for texts as austere as this. In P 802, Walther specifies neither pedal nor
two manuals, and it is playable by two hands on one keyboard – surely not by
chance in what is at times a five-part piece? There are no real paraphrases,
though the harmony expresses the modal cadence of the melody, and at
times the bass does seem to anticipate the chorale line.
464 BWV 721–722
Suggestions as to where the style originated include a ‘Vater unser’ of

Böhm (Spitta I p. 212), but the chords there are less constant, and there is a
clear pedal line – radical differences. Terry (1921 p. 158) and Keller (1937)
point to the movement ‘Il tremore degl’Israeliti’ in Kuhnau’s first Biblical
Sonata (1700), the passage headed ‘their prayer to God’ in which repeated
chords accompany the melody ‘Aus tiefer Noth’. This is much likelier, for
Thomascantor Kuhnau held a commanding influence in Thuringia–Saxony.
Mahrenholz (quoted by Klotz KB p. 47) claims that the composer was ‘stylis-
tically dependent’ on the cantata setting ‘Erbarme dich mein, O Herre Gott’,
now known to be by Buxtehude’s pupil Ludwig Busbetzky (M. Geck Mf 1973
p. 175), and where the string writing is virtually identical. This, or some-
thing very like it, was surely known to the composer of BWV 721. So it
might have been to Handel, whose setting of the same text in the Chandos
Anthem HWV 248.ii (but in English) uses similar repeated quavers ‘adagio
e staccato’.
The organ-builder Esaias Compenius had noted earlier that a suitable
rate for the organ Tremulant was eight times per bar (Blume 1936), which
either strings or keyboard instruments could have imitated. Examples given
in Krummacher 1978 include Busbetzky (Ex. 35a), Kuhnau (35b and 79a)
and Boxberg (87), though in comparison with keyboard techniques from
Scheidt to J. K. F. Fischer, or string accompaniment in the Sonatina of Cantata
106, BWV 721 is exceptionally uniform. Scheidt’s ‘imitatio tremula organi’
(Tabulatura nova) is a lively bicinium, and its repeated notes, superficially
resembling BWV 721, do not accompany a chorale. The wealth of seventh
and ninth chords in BWV 721 makes it likely that its composer was familiar
with Italian continuo harmonies of c. 1700 and earlier: a possible, strong
influence is the Adagio in the first of Corelli’s Sonate da camera Op. 4 (1694).
For a note on the penultimate chord, a particular leading-note seventh
approached here most skilfully, see also BWV 727.

Copies by J. G. Walther, J. G. Preller and J. C. Kittel.
Headed by Walther ‘man’ (manualiter).
On this type of movement, see BWV 715. The Mempell–Preller copy collects
BWV 722, 738, 729 and 732 as ‘Vier Weynachts Chorale’, in the same order as
the simple figured versions in Krebs’s P 802: see BWV 722a below. Whether
465 BWV 722–723
the two versions had a different origin or purpose – the figured versions
a model for students, the fully composed versions free organ-chorales – is
unknown, as too is how distinct the group of four was from the similar
settings BWV 715 and 726.
In the spacing of the hands, in the ornaments, and in the incipient
motifs (bb. 5, 10), BWV 722 approaches an independent organ-chorale.
The last four bars correspond to the ‘Kyrie eleison’ part of the text (see also
BWV 604 at this point) and follow on the previous line immediately. They
also hint at other G major cadences, such as that in BWV 698. The flourish
after each line first keeps the harmony of the previous chord, then anticipates
the next. Simpler flourishes preparing the following chord (Zu-Lenckungen,
‘linkings’) were provided by H. F. Quehl for the two chorales in Der musi-
calische Versuch (1734).
Although quick modulatory harmonies are found (b. 2), those of
bb. 8–9 suggest a language approaching the Ob, where too Advent/Christmas
settings were the first to be grouped together.
BWV 722a Gelobet seist du, Jesu Christ

Only copy, by J. T. Krebs.
The harmonies of the figured bass are much like those of BWV 722, and
it is doubtful that the bass note is to be held through the second and third
inter-line flourishes, as suggested in Zietz 1969 p. 166. The chief difference
is that the unbarred BWV 722a gives no sign of the half-speed of the last
four bars as they are in BWV 722. But neither version can be presumed to be
the earlier: in the case of all four figured hymns in P 802 (BWV 722a, 738a,
729a, 732a), either could be drawn from the other. The figured ‘sketches’
may be reductions rather than drafts, despite what is usually claimed, and
despite their extant source being earlier.
Two further examples of such chorales occur without attribution in
Krebs’s MS. Again the chorale in minims is interspersed with four sets of
runs, alternating treble and bass. Perhaps all six are the student work of J. T.
Krebs and not J. S. Bach’s teaching models, as is often supposed from Spitta
onwards (I p. 586).

Copies via J. C.Kittel (?), also LM 4708 (‘Neumeister Collection’).
Two staves. In LM 4708, ‘J. M. Bach’.

466 BWV 723–724
There is more than a passing similarity here to another (very early?)

Christmas setting in G major, BWV 719, but the derivative fore-imitations
are closer to Pachelbel, as in his ‘Komm Gott, Schöpfer, heiliger Geist’. There
is a G-mixolydian key ambiguity to the opening fugue, and the whole move-
ment wavers between C and G major, as too does BWV 697.
Kittel’s probable access to Bach’s own copies may support the attribution
to J. S. Bach (KB pp. 14, 17) but the four parts – playable by hands on one
manual, managed more than competently, and without too many formulae –
look more like Johann Michael’s work. Whether or not the clear perfect
cadences at the end of each line affirm the phrase ‘that is true’ from v. 1
(Chailley 1974 p. 123), they evoke no particular composer.
BWV 724 Gott, durch deine Güte

Only copy ABB (J. C. Bach), in tablature.
For TEXT and MELODY see BWV 600. A later hand has added ‘Gottes
[Sohn] ist kommen’ (cf. BWV 600).
Such contrapuntal accompaniment ‘with neither sustained motif nor

thematic interludes’ has long been recognized as ‘the most primitive
form of organ chorale’ (Spitta I p. 595). It has also been claimed to
‘correspond exactly’ with chorale-types in the ‘Neumeister Collection’
(Wolff 1992 p. 244), though there is little there in the texture or counterpoint
that is similar, except certain ‘weaknesses’ common to both. Its absence from
Weimar sources may be due to its being very early, or never being transcribed
from tablature, or its genre having become permanently outmoded.
But the careful four- and five-part texture has points of interest, not
only in the tentative canons of bb. 7–8, 12–13, 17–21, 33–4, 38–9, but the
easy lines and effortless counterpoint of the whole, at first predominantly
in minims, then in crotchets. This must be the result of conscious planning.
While according to NBA IV/3 the pedal enters in b. 9 for the fourth (partly
free) part, the only cue in the tablature is at b. 34, where it is given for the
fifth part, thus for two phrases in canon (see Hill 1991 p. xxvii). Bar 28
suggests short octave in the manual, b. 41 a pedal with F. This, if missing,
could go up an octave (BG 40). As in BWV 716, the final pedal entry serves
as a kind of coda, since the opening and closing lines of the cantus are rather
alike, and the result has little in common with Pachelbel’s treatments.
Despite attempts to show that the canonic imitation somehow refers to
the text, the composer seems rather to be experimenting with the technique
467 BWV 724–725
itself, as with BWV 716 (opening canon at the fourth) and as eventually
perfected in BWV 600. There is also a careful handling of harmony from
two to five parts: an ‘exercise’ well accomplished for a setting that also keeps
a certain Christmas-pastoral feel to it.
BWV 725 Herr Gott, dich loben wir

Only source, Peters VI (copy by Forkel).
The TEXT is Luther’s version in rhyming couplets of the Te Deum, associated

with Mattins and Vespers, later a hymn for New Year’s Day (Gojowy 1972)
and general ‘praise and thanksgiving’ (Stiller 1970 pp. 223, 232). The first
four of fifty-three lines are:
Herr Gott, dich loben wir. Lord God, we praise you.

Herr Gott, wir danken dir. Lord God, we thank you.
Dich, Vater in Ewigkeit, You, Father in eternity,
ehrt die Welt weit und breit. The world honours far and wide.
The simplified MELODY (Liber usualis, appendix, Hymn for Thanksgiving)

appears intact in BWV 725. Listed in the Ob, used in part in Cantatas 16
and 190 (New Year 1726, 1724), 119 (Council Election 1723), 120 (wedding,
c. 1729), 190a and 120b (anniversary of Augsburg Confession, 1730). BWV
725 often agrees with BWV 328 when it does not re-harmonize for a new
verse; bb. 188–202 have a different melody. It is not certain that Forkel’s MS
included the unique text incipits.
The setting appears to be a written-out accompaniment for a repetitious

chant (Spitta I p. 588), preserving the phrygian harmonies and admitting
contrapuntal motifs particularly for repeated lines, so that the livelier verses
bb. 18–37, 183–212, etc. look rather like alternatim interludes. However, it
seems not to be laid out as verse and response in the way that the Te Deum
was still sung in the Leipzig Nikolaikirche at Mattins (Spitta II p. 109). This
‘accompaniment’ could be for organ alone or (as in cantata chorales) for
instruments playing colla parte.
This is fine, five-part harmony with some unambiguous references to
the text: ‘angels’ in the scale of b. 18, ‘the incarnation and its purpose’ in
bb. 123ff., ‘divine power’ in the pedal of bb. 143ff., ‘appeal for help’ in
the chromatic lines of bb. 163ff., and ‘the vigour of belief and praise’ in
bb. 193ff. For passages serving two sets of words the treatment, as too in the
cantata settings, is more neutral: bb. 143–52 (223–32), bb. 153–7 (213–17)
and bb. 163–72 (233–42). The repeated passages include some of the setting’s
less conventional harmonies.
468 BWV 725–727
The modified repeats, particularly bb. 158–9 as altered in 218–19, may

be for continuity rather than word-painting. But the verse structure and
the many perfect cadences make it less continuous than usual – a major
difference between it and Ob chorales which use the same kind of motif as
that in b. 153 or 188. Nevertheless, the harmony has striking progressions
from time to time (e.g. bb. 133–56) and there is little to say whether, its style
being so straightforward, it was written before or after Clavierübung III. If
one looks for a suitable occasion for a festive performance – a celebration
in 1730, or the Thanksgiving on 9 January 1746 after a recent Prussian
win (Butler 1992) – more than one can be found, including much earlier
occasions around the time of the Ob.

Source as BWV 715.
On this type of movement, see BWV 715, which immediately precedes

it in P 804. As in other examples of the type, BWV 726’s interludes grow
progressively more arpeggiato; unlike others, it remains strictly in four parts,
of which only the bass is unvocal. The chromaticism of the penultimate bar
is unusual in G major and its style seems earlier, more elementary, than e.g.
Cantata 131’s (1707).
BWV 727 Herzlich tut mich verlangen

Copies by J. G. Walther and J. T. Krebs.
Two staves; headed in P 802 only, ‘2 Clav. e ped.’ (J. L. Krebs?).
The TEXT of C. Knoll’s hymn was published in 1605, becoming associated

with funerals and Resurrection (Freylinghausen 1741).
Herzlich thut mich verlangen From my heart I am longing

nach einem selgen End, for a blessed end,
weil ich hie bin umfangen, for here I am surrounded
mit Trübsal und Elend. by trouble and misery.
Ich hab Lust abzuscheiden I have a desire to take leave
von dieser argen Welt, of this wicked world,
sehn mich nach ewgen Freuden, longing for eternal joys,
O Jesu, komm nur bald. O Jesu, only come soon.
469 BWV 727
Eleven verses alternate earthly misery and heavenly joy, as in v. 4 (Cantata

161) and v. 5, which is ‘obviously represented’ in this setting (? – Sackmann
1998 p. 249):
Ob mich die Welt auch reizet . . . Although the world charms me . . .

das Himmlisch’ ich betrachte . . . I contemplate heavenly things . . .
The MELODY belonged to Hassler’s love-song ‘Mein G’müt ist mir ver-
wirret von einer Jungfrau zart’ (published 1601), attached to this text from
1613 (Example 248) but associated with many texts: ‘Ach Herr, mich armen
Sünder’ (Cantata 135, 1724), ‘O Haupt voll Blut und Wunden’ (St Matthew
Passion 1727, Cantata 159, 1720), ‘Befiehl du deine Wege’ (St Matthew Pas-
sion, Cantata 153, 1724), ‘Wie soll ich dich empfangen’ and ‘Ihr Christen
auserkoren’ (Christmas Oratorio), as well as ‘Herzlich thut mich verlangen’
(Cantata 161, 1715); without text in Cantatas 25 (1723), 127 (1725) and
161 (first movement). Listed in the Ob as ‘Ach Herr, mich armen Sünder’.
Mattheson (1739 p. 473) claimed that there were twenty-four different
hymns sung to this melody, and asked how such a ‘moving death-hymn’
(‘bewegliche Todten-lied’: 1731 p. 71) supported the Greek view of the
phrygian mode as inflammatory.
Example 248
The quality of the music, the suspirans figures in the decoration (already b.
2), the tmeses or gaps in b. 3 etc., the relationship felt between the music and
‘longing’ (‘verlangen’), and its overall conception – accompanied melody
without interludes – have all suggested to players that BWV 727 is very like
an Ob prelude. However, in the virtual absence of motivic development or
imitation, BWV 727 could properly be compared only with (at most) ‘O
Mensch, bewein’, minus the coloratura melody. While BWV 727 differs from
the CbWFB settings BWV 691 and 728, its suspirans is found in the former;
and while it differs from the Tremulant setting BWV 721 (also in P 802),
the spacing and final cadence are similar.
The movement lies somewhere between the simple four-part harmo-
nizations and the Ob, though nearer the latter in both style and date, late
enough to be known to Krebs and Walther. As a harmonization, its rapt
mood is immediate, continuing through telling motifs (b. 5 etc.) and a final
470 BWV 727–728
cadence whose special Affekt might change from text to text. Obviously, this
particular melody and its natural harmonies create an unmistakable aura
of ‘longing’, of whatever kind; the unnecessarily long final pedal point alone
plays a part in this.
The penultimate chord, a particular combination of a held tonic and
(implied) diminished seventh on the leading note, was something of a
fingerprint in early Bach works: see BWV 721 and ‘Neumeister Chorales’
BWV 1095 and 1105. (A decorated version of it, originally closing the
Commandments setting in the Ob, was altered to the present reading. See
BWV 635.) It is clear from its appearance at the end of the early Fantasia in
C minor BWV 1121 that the chord need not be associated with any partic-
ular text; but drawn out in a ritardando, in such a chorale-setting as this, it
compels a response from the touched listener.
BWV 728 Jesus, meine Zuversicht

Autograph: AMBB; a later copy via C. P. E. Bach.
The anonymous TEXT of the Easter hymn was published in 1653.
Jesus, meine Zuversicht Jesus, my trust

und mein Heiland, ist im Leben. and my Saviour, lives again.
Dieses weiss ich; soll ich nicht This I know; should I not
darum mich zufrieden geben, therefore be content
was die lange Todesnacht with whatever thoughts
mir auch für Gedanken macht? the long night of death gives me?
The nine verses that follow contrast death with the afterlife.
The MELODY became attributed to J. Crüger (Terry 1921 p. 238); for its
version in BWV 365, see Example 249. A varied form appears in Cantata
145, and BWV 728’s differences do not all appear to be the result of embel-
lishment.
Example 249
Although in AMBB the piece looks like a finished fair copy, perhaps it was
written out at the moment of creation, in 1722 or 1723 (NBA V/4 KB
pp. 11, 20). In form and type it relates to AMBB as BWV 691 does to the
471 BWV 728–729
CbWFB. Whether in either the composer intended the decorated melody

to convey a subdued mood of death is less certain than that he was incor-
porating ornaments and figures already familiar from BWV 691 and the
Ornament Table in CfWFB, and so providing practice in florid melodies
(Spitta I p. 585) and ornamentation. For the order in AMBB, see BWV 573.
The cadences of both BWV 691 and 728 are their most conventional
moments, and here the second half of b. 5 is very like Walther’s cadences.
On the other hand, the melody of BWV 728 tends to rely more on groups
of small notes and varied decorations of the dotted quaver motif. As with
BWV 691 only more so, the decorations include important slurs: slurring
was a manière quite as much as adding ornaments. They are all used to
mark the main notes of the original melody on the beat, thus showing both
how to paraphrase a melody with beautiful patterns and how they are to be
played cantabile.

Copies by J. G. Preller and by or via J. C. Kittel.
Two staves.
On this type of movement, see BWW 715. The interludes now are not
only better integrated but more obviously related to the text, even perhaps
to the last verse with its bells and choir of angels (Spitta I p. 587). How
suitable it is for accompanying a congregation or even choir is doubtful: the
breaking of the melody in bb. 19–21, the irregularly prolonged cadence at
bb. 22–5, the unequal lengths of the interludes (2, 2, 2, 3, 3, 2 bars), the break
necessary for the congregation to pick up the new line at the end of b. 32,
the length of the coda – these suggest an independent organ work, a model
improvisation, perhaps a (rather messy) postlude reaching for effect, and
something impossible to sing with. Curiously, the first two cantus phrases
also avoid the obvious cadence: a young man’s originality?
Pedal, though nowhere specified, seems to be more necessary than in
other ‘Arnstadt Chorales’, and most clearly at the end, as often in early
works. The broken scales of the first three interludes suggest the angel throng
(cf. BWV 607) or the purpose of Christmas (cf. BWV 600); the broken chords
of the fourth interlude are like those of a bicinium bass; the bass lines of the
fifth resemble those of BWV 600 and 607; and the arpeggio of b. 46 matches
those in Handel’s harpsichord preludes pre-1710. The high full chords of
472 BWV 729–731
the close are rare outside doubtful or early chorales (cf. BWV 702, 716), but
much of the rest looks ahead to maturer work.
BWV 729a In dulci jubilo

Only copy, by J. T. Krebs (no composer’s name).
The figures are closely observed in BWV 729; on which came first, how-
ever, see BWV 722. Three major differences between the two versions are:
(i) like BWV 722a, BWV 729a has no bar-lines, possibly suggesting more
freely played interludes than BWV 729; (ii) BWV 729a has no interlude
between lines 4 and 5 of the chorale; (iii) BWV 729a gives no hint of the
coda.

No Autograph MS; copy in Lpz Poel 39 (Kittel) and late sources.
BWV 731, two staves, headed ‘à 2 claviers et pédale’ in NBA IV/3.
BWV 730’s four or five-part harmonization of the melody without inter-

ludes points towards the Ob, as does the diversionary 6/4/2 of b. 4 and
especially the tenor and bass of the third cantus line (compare BWV 638).
The harmonies are of a higher order than the diminished sevenths of BWV
715, with dominant and major sevenths and ninths (bb. 4, 13). The rising
pedal of b. 13 has suggested to some the words ‘von der Erden ganz zu dir’
of v. 1 (Chailley 1974 p. 185).
Only at times does BWV 731 read like a variation, chiefly in b. 1, where
tenor and bass seem to anticipate the Ob setting BWV 633/634. Harmoni-
cally and melodically, bb. 11–12 of BWV 730 could replace the corresponding
bars of BWV 731 if an organist wanted an ornamented chorale based on this
version of the melody. The juxtaposition of the two settings in the source
allows this.
The inner parts, the moving bass line and the ornamental melody of
BWV 731 seem to be ‘mere variations’ around basic four-part harmony,
arising from an imaginative treatment of it. Yet just as the accented passing-
notes of b. 3 already hint at the sophistication of the Ob, so too does the
473 BWV 731–732a
melody, especially as it takes flight in bb. 4 and 13, in a manner well beyond
the ‘Neumeister Chorales’. Perhaps too early to be known by Walther and
Krebs, the settings suggest an intermediate step towards the motivic integrity
and intensity of BWV 601.
BWV 732 Lobt Gott, ihr Christen, allzugleich

Copies as BWV 729.
On this type of movement, see BWV 715. Like the other settings of this type,
this one modifies the conception somewhat:
1 typical interlude figure integrated into the harmonization

2–3 semiquaver lines more in Orgelbüchlein or partita style
6–8 cantus sinks below soaring top line
Perhaps line 3 pictures ‘today opens his Heaven’ (Spitta I p. 586), as

b. 8’s downward-running interlude does Jesus’ descent to earth (‘and gives
us his son’). Whether it is ‘a significant forward step’ in the evolution of
chorales (Keller 1948 p. 142) is as uncertain as the chronology, but it is
possible that strict four-part harmonization in BWV 706 came after the
full-chord technique of BWV 732, rather than vice-versa. Much here could
be improvised upon further: the short-lived motif development of b. 1, the
change to quavers in b. 4, the free runs, the drawn-out cadences.
Replacing the hymn’s opening upbeat with the demisemiquaver motif
need not have confused any congregation it was accompanying (as Sack-
mann 1998 p. 234 asserts): performance of hymns was surely far from being
polished and exact, and the harmony is clear enough. For similar reasons,
the burying of the cantus in the alto of b. 8 and its following ‘disintegration’
do not mean BWV 732 could not accompany a congregation.
BWV 732a Lobt Gott, ihr Christen allzugleich

Copy, as BWV 729a (no composer’s name).
The figures of BWV 732a are observed in BWV 732; for which came first,
see BWV 722a. There is no indication of the long close to lines 1, 2, 3 and
5 in BWV 732 except the fermatas, but the tie to the first note of the first
474 BWV 732a–733
interlude may well suggest it was taken from BWV 732, with interludes of a
kind improvised at the period.
BWV 733 Meine Seele erhebt den Herren (Fuga sopra

il Magnificat)
Copies by or via J. C. Oley, J. P. Kirnberger and J. L. Krebs (two).
Two staves; headed by Oley and Kirnberger, ‘Fuga sopra il Magnificat’; also
in Oley ‘Meine Seele erhebet den Herren – pro organo pleno con pedale’.
One long fore-imitation based on a melody that is heard briefly at the end is
an idea associated with Pachelbel, as it is too with the simpler BWV 716 and
Magnificats of other composers, both older (Scheidt’s ‘Noni toni’, also on
the tonus peregrinus) and younger (‘Meine Seele erhebet den Herren’ by the
Leipziger J. C. Schiefferdecker). In particular, the first verse of Buxtehude’s
Magnificat BuxWV 205 may have influenced BWV 733. But the setting
is greater than its pedigree and beyond possible antecedents, better seen
perhaps as a fugal equivalent to the trio ‘Allein Gott’ BWV 664, i.e. a densely
contrapuntal setting based both on a subject derived from the first cantus
line and on its countersubject, the whole crowned by two phrases of pedal
c.f. at the close. A further parallel is BWV 661, which also develops quaver
patterns, or even BWV 651, where a c.f. is introduced by four rising notes,
just as BWV 733’s is rounded off by them. Perhaps like BWV 651 it was
entitled in the autograph ‘Fantasia’ or ‘Fant’, the latter misread by an early
copyist (J. L. Krebs?) as ‘Fuge’?
It is not obvious in what sense BWV 733 is a fugue, or how (or even
whether) fore-imitation is involved. The cantus appears not as a fugue-
subject but as a series of intonation-like entries, most of them signalling a
new part in the texture:
1 tonic (two parts)

10 tonic (three parts)
30 dominant (four parts, including countersubject)
55–6, 75–6 two stretti
98, 119 augmentation, lines 1 and 2 (five parts)
Though the form is different, the general effect is not unlike (e.g.) Contra-
punctus IX from the Art of Fugue, in which too the subject is pervasive. The
opening seems fugal, owing to the conventional nature of a countersubject
475 BWV 733
which incorporates useful ideas. See Example 250. It is the motifs that are
picked out for development, and the line is neither answered as a fugue-
subject nor treated as a regular countersubject. The counterpoint generally
is italianate, first two-part, then in three, four, and for the two pedal c.f.
sections, five parts.
Example 250
Discrete motifs are again open to expansion, inversion, sequence and

other kinds of development, and all of them in Example 250 were common
in standard alla breve counterpoint. Thus f is used against the c.f. in one of
Buxtehude’s settings (‘Noni toni’ BuxWV 205), e and d in another (BuxWV
204), d in the opening movement of Bach’s Cantata 10, and so on. More even
than moments in Pachelbel’s Magnificat cycles, BWV 733 is remarkable for
its systematic use of particular motifs. Strictly speaking, e, f , g and h are not
countersubject motifs, being accompanied themselves by a crotchet line of
a kind also familiar in stile antico (b. 6).
The sheer profusion of usable motifs seems to force the composer to rely
chiefly on one of them (a), and apart from such moments as the cadences
in b. 43 and b. 65, the others are not much used; those that are (c, e) are not
particularly melodic or striking. But a is more or less continuous, and its
inversion is either complete or partial, thus a good example of a particular
technique: Example 251. It fits in anywhere, in any voice, against any line,
Example 251
sustaining a long movement so effortlessly (from b. 2 to its inversion in

the penultimate bar) that it can allow a new counter-theme to enter as if
it were a genuine second subject – though quite where it starts is not clear
(b. 42 or 43? and b. 63 or 65?). Similar fluent quaver themes are used in
mature Leipzig works, and partly because of this quaver theme, there is a
family likeness between this Fugue and the C minor BWV 546, the finale
to the Italian Concerto BWV 971, and the Ricercare à 6 from the Musical
Offering.
This strong, vivid movement is a stark contrast to the sung settings of
the same chant-melody, ‘Suscepit Israel’ in the Magnificat BWV 243 and
476 BWV 733–734
Cantata 10. There, the same theme lends a plaintive colour quite alien to
the Magnificat tradition in organ music. Here, the writing is closer to the
big organ fugues – compare its close with BWV 540 or 545 – and the two
five-part sections are as well conceived as anything in mature Bach. The
e of bb. 105–6, the counterpoint written to ‘justify’ the repeated Gs of
bb. 121–5, and the pairing of such bars as 128–9, are especially fine, ingenious
solutions to a taxing cantus firmus. A standard type of counterpoint has been
handled very imaginatively, distracting the ear from the peculiarity of a piece
that opens in an ambiguous key and continues to fluctuate between a minor
key and its relative major. Other Magnificat fugues too had ended with a
major chord prepared in the penultimate bar (Pachelbel), but not after so
sustained a fantasia as this.
BWV 734 Nun freut euch, lieben Christen g’mein / Es ist

gewisslich an der Zeit
Copies by or via J. C. Oley, J. L. Krebs and probably J. C. Kittel.
Two staves; headed by Krebs ‘Choral in Tenore’, and in Oley’s ‘manualiter’.

Second title from the (late) copies of BWV 734a.
The TEXT of Luther’s ten-verse Advent hymn is a ‘ballad on Christ’s Incar-

nation’ (Stapel 1950 pp. 203ff.), later associated with Ascension and Sundays
after Trinity:
Nun freut euch, lieben Christen gmein, Now rejoice, dear Christians all,
und lasst uns fröhlich springen, and let us leap with joy,
dass wir getrost und all in ein that we, confident and united,
mit Lust und Liebe singen, sing with pleasure and love
was Gott an uns gewendet hat of what God has given for us,
und seine süsse Wundertat; and his sweet miracle;
gar teur hat ers erworben. very dearly has he bought it.
The TEXT of Ringwaldt’s Advent hymn was published in 1582:
Es ist gewisslich an der Zeit It is certainly time

dass Gottes Sohn wird kommen that God’s Son will come
in seiner grossen Herrlichkeit, in his awful splendour,
zu richten Bös und Fromme. to judge the wicked and the righteous.
Da wird das Lachen werden teur, Then jeering will cost dear
wenn alles wird vergehn im Feuer, when everything perishes in the fire,
wie Petrus davon schreibet. as St Peter writes of it.
477 BWV 734
The next six verses, ultimately based on the Dies irae sequence, concern Jesus
the intercessor. The double title is also used by J. M. Bach in the ‘Neumeister
Collection’.
The MELODY is said to be derived by Luther from a song ‘Wach auf, wach
auf du schöne’ (Terry 1921 p. 270), associated with both hymns. In the
Christmas Oratorio it is set to an Epiphany text. See Example 252. Listed
in the Ob, set in BWV 755, and used without text in Cantata 70 (Sunday
before Advent 1723).
Example 252
As in bb. 36–8, the three voices are surely written to be played manualiter,
though pedal c.f. is always an option: see remarks on BWV 695. Both the
non-stop semiquavers and the continuo-like bass line are unusual. The
rh (Example 253) is a paraphrase rather than a variation in the manner
of Partita IV of BWV 767, and is built on the turning motifs of Scheidt’s
imitatio violistica. The left hand occasionally augments and inverts the same
figures (b. 4), and similar motifs, pursued less single-mindedly, can be found
above pedal c.f. sections in music of Pachelbel (‘Nun freut euch’) and of those
he taught.
Example 253
The semiquavers seem to gloss more than one line of the cantus, whose
lines 2, 4 and 7 are in any case the same. Moreover, it appears sometimes as
a ritornello: b. 3 dominant, 9 tonic, 30 relative, 36 dominant, 40 tonic, 45
tonic. The closes of each half are also similar. Characteristic of the melody’s
organized cell-construction is that it consists chiefly of two motifs: the in-
turning x and the scalar y in Example 253. Both are hugely adaptable, as a
comparison of bb. 3 and 35 shows. The result is a masterly version of the
common perpetuum mobile movement included in sets of variations.
478 BWV 734–735
That the setting somewhat resembles the final chorale of Cantata 22 is

probably because their melodies are rather alike, not because of a connec-
tion between Advent and Lent. Nevertheless, one might see in the chorale’s
exuberance an allusion to the second line. The figured chorale which fol-
lows has different harmony; see also BWV 690. It is nowhere clear whether
BWV 734 is a prelude to the figured chorale, whether this (with its keyboard-
like bass line) was intended by Bach or anyone else to serve organists
without hymnbook melodies, or neither. But whoever is responsible for
any particular instance of this pairing, a plain chorale following an organ-
chorale does give the impression of a prelude leading naturally to a following
hymn.
BWV 734a Es ist gewisslich an der Zeit

Late sources only, including P 285 (‘a 2 Clav. et Pedal’), without the chorale.
In the cantus of bb. 33–5 and 39, BWV 734a differs from 734, taking a
form associated with its different text, though not based on any of the
extant MSS of BWV 734 (Emans KB). Although the other parts (rewritten
as a consequence?) are not unmusical, the lines show occasional infelicities,
surely not authentic: uncharacteristic lh tenths b. 33, new (?) rh bb. 32–5,
lapse in rh sequence b. 39. Such infelicities have been described as not
so marked that the version must be inauthentic; on the contrary, perhaps
BWV 734a is the earlier version, ‘developed’ later by Bach into BWV 734
(Emans KB). But is there a precedent for ‘development’ from something less
grammatical to something more?
BWV 735 Valet will ich dir geben

Reputed Autograph MS ‘from Guhr Collection’ used in Peters VII; late copies
only.
Heading ‘Fantasia super’ in BG 40 not authenticated.
The TEXT of V. Herberger’s hymn was published in 1614.
Valet will ich dir geben, I shall say farewell to you,

du arge, falsche Welt; O wicked, false world;
dein sündlich böses Leben your sinfully evil life
durchaus mir nicht gefällt. I detest through and through.
Im Himmel ist gut wohnen, To live in Heaven is good,
479 BWV 735
hinauf steht mein Begier, and on this rests my desire;

da wird Gott herrlich lohnen there will God reward well
dem, der ihm dient allhier. him who serves him here.
The following four verses look to the saviour of the soul.
M. Teschner’s MELODY, which varied, was published with the text

(Example 254). It seems to derive from a melody in the Geneva Psalter
(Ps. 3). Listed in the Ob, found in BWV 415, Cantata 95 (1723) and the
St John Passion.
Example 254
The authenticity of BWV 735 is not certain enough to be sure that Bach
was picturing ‘the soul rising to peace’ at the end (Keller 1948 p. 175), as
he appears to do in BWV 656: it is suspiciously like BWV 702. But in shape
BWV 735 is similar to ‘Jesus Christus, unser Heiland’ BWV 665:
irregular exposition of cantus lines, each finally in crotchets in the pedal;
gap between lines except 6–7 (see BWV 735a) and 7–8; each (except line 7,
42ff.) uses the whole cantus line; few motifs, mostly announced in the first
bars.
Although BWV 735 and 665 have details in common – see BWV 735 b. 29
etc. and opening of BWV 665 – the former appears to be more ‘objective’,
without textual allusion but with constant reference to one or two semi-
quaver patterns, familiar in the partitas but here spinning out a decorated
chorale line recalling Buxtehude or Böhm. Moments in the final pedal point
are more than faintly reminiscent of the end of the C major Toccata, while
b. 51 seems to return in more exacting form before the final stretto entries
in another fugue (D minor, BWV 538).
The semiquavers from b. 1 appear rectus or inversus against every line of
the chorale (bb. 2, 10, 16, 22, 29, 37, 43, 52) as well as on the pedal point and
the ‘old’ texture at bars 43–7, so called because of the broken-chord figure
and the incomplete chorale line. Despite frequent tonics, such economy
of means combined with well-harmonized chorale lines (especially in the
bass) is credibly the work of Bach. So are the mastered ‘tricks of the trade’
480 BWV 735–736
(e.g. the a when the final c.f. note is reached) – but all these are already in
the form shown by BWV 735a, which is without the final rise.
BWV 735a Valet will ich dir geben

Copies by J. G. Walther and in Plauener Orgelbuch (J. A. Lorbeer, before
1710).
Two staves; headed ‘con Pedale’ in Plauen.
The main differences are: certain details in the imitation (BWV 735a
b. 34, BWV 735 bb. 29ff.) and in accompanying figures (e.g. two upper
parts transposed an octave in bb. 39f.); and old patterns in bb. 51ff. which
become more contrapuntal in BWV 735, whose final pedal point is also
‘bigger’. But 735a already had accomplished fugues on each line, and whether
BWV 735a/735 correspond to the early/late versions of ‘The Eighteen
Chorales’ as often suggested (KB p. 11) hangs on whether the changes are
reliably attributed to Bach. They become more radical as the piece pro-
gresses, unlike those of ‘The Eighteen’, even in the case of BWV 651a/651.
For a remark on the ‘new’ coda, see BWV 735 above. Clearly, the final
three bars of BWV 735a are older, as is the broken-chord figure from b. 51.
This last has precedents in several ‘Neumeister’ chorales (BWV 1092, 1106,
1107, 1117), and much in BWV 735a resembles Georg Böhm’s pedal c.f.
setting of ‘Vater unser’. All such details could well have been consciously
rejected/improved in BWV 735, but by whom? The simple sections of
BWV 735a, where each line had a distinct beginning, are disguised in
BWV 735 by the motif-extension in bb. 42 and 52. Other ‘improvements’ are
also not quite happy, producing repetition (bb. 29–32 and bb. 34ff., also in
b. 53), interfering with sequences, and leading to a thin texture at the end.
Some of these are changes in conception quite untypical of Bach’s revisions
as known. A recent view that BWV 735 may be a nineteenth-century
arrangement (EB 6589 p. vii) is plausible, since neither the sources nor the
musical details – modelled on ‘The Eighteen’ revisions? – are conclusively
against it.
BWV 736 Valet will ich dir geben

Copies via J. C. Kittel and possibly other contemporary sources.
Two staves.
481 BWV 736
For TEXT and MELODY see BWV 735. For the figured chorale added in
some late copies (KB p. 59), see BWV 734. That BWV 736, a very ac-
complished work, is found in a Kittel source may support the idea that
it was intended for the set of ‘Leipzig Chorales’ as assembled in P 271, after
BWV 735a had been rejected.
Highly original in general effect, BWV 736 also combines subtle motif and
paraphrase techniques. Its form is:
A 1–12 fore-imitation, paraphrase of lines 1 and 2; accompaniment

to pedal cantus based almost entirely on a (see Example
255)
B 24–36 fore-imitation, paraphrase of line 5 via a (modified); ditto
C 36–43 fore-imitation, based on theme from B (inversus, rectus);
ditto
D 43–52 fore-imitation, paraphrase of line 7 via a further derivative
of a; accompaniment to cantus partly based on a
E 52–9 fore-imitation, paraphrase of line 8 by means of a
(inversus)?
Quick-moving triplets above a slow c.f. may ultimately derive from Scheidt
(Dietrich 1929 p. 62), but some idea of the motivic subtlety of the movement
can be grasped from the final cadence, where – to look only at the inner
parts – the last three bars incorporate a, a inverted, a as modified in section
B, and a as modified in section D, the whole cross-referring to the previous
pedal-points: see Example 255. The c.f. paraphrases become progressively
Example 255
less obvious during B, D and E, and only the plan makes it likely that the
composer did in fact have line 8 in mind for E. If the main notes of the cantus
were off the beat, the paraphrases would be less apparent, since the triplets
sweep up everything, breathless to the end.
482 BWV 736–737
A further example of the work’s unity-within-variety is the pedal points.

The final cadence refers back but now with brief, ‘finalizing’ diminished
sevenths; the middle cadences (b. 35, b. 42) are quite distinct, despite the
similar motifs; and the cantus, being in the pedal (unlike the figured chorale),
leads to punctuating cadences on D, D, A, A, F/B and D. The gigue-like motif
a suggests that the composer had in mind the second half of v. 1 but also that
the setting dates from the Leipzig period: Kauffmann used a similar pattern
in his ‘Komm, heiliger Geist’ (1733), to which BWV 736 could be a response,
but now with organo pleno replacing Kauffmann’s discreet registration of
Vox humana + Salicional 8 + Spillpfeife 4 .
So majestic and exuberant a setting of words that speak of resignation
might be explained by a ‘striving up towards heaven’ (Meyer 1987 p. 21), or
it may be a response to the chorale-melody itself, which has little resigned
about it. Either way, the note-patterns are familiar in other compound-time
preludes, such as BWV 712.

Copies by J. G. Walther and in Yale LM 4708 (subtitled ‘Nimm von uns,
Herr, du treuer Gott’).
For TEXT and MELODY see BWV 636. Alternative texts are not uncommon
in ‘Neumeister’: ‘Nimm von uns’ (M. Möller, 1584) is another prayer, against
retribution for sins. In some hymnbooks, the two texts share a verse.
Like the doubtful BWV 705 and 707 (both headed manualiter in P 1160),
the parts are playable by hands, and Walther gives no pedal cue for what
is an undistinctive bass. Spitta saw it, like BWV 724, as an example of ‘the
most primitive form of organ chorale . . . with neither sustained motif nor
thematic interludes’ (I p. 595), which would suit its position in ‘Neumeister’.
The patterns are typical of alla breve counterpoint in 4/2, and more than
once the soprano c.f. is anticipated in an inner part (e.g. bb. 19–20). The
opening fore-imitation of b. 1 with bass countersubject leads the player
to expect a stricter organ-motet of Scheidt’s monothematic type, as do
the inconclusive cadences (typical of Pachelbel) and the lengthening of the
cantus as the piece proceeds. Though looser than his, its lines are typical
of Scheidt, with consistent but undeveloped motifs – an early work of
J. S. Bach.
Effective cohesion is given by the dactyl/anapaest figure, and there is
good harmonic tension between the modal moments (b. 3) and the chro-
matic/diatonic (b. 28). The distinctly modal details in NBA reflect Walther’s
483 BWV 737–738a
copy, resulting in a by no means inappropriate archaism from time to time.

Whether Neumeister’s several added accidentals (bb. 2, 5, 16) make his ver-
sion ‘superior’ (Wolff 1997 p. 160), or merely reflect the diatonic interference
of a copyist later and less authoritative than Walther, is not known.

Copies, similar to BWV 729.
On this type of movement, see BWV 715; on running figures in Christmas

chorales, see BWV 607, 697 and 701. While the semiquavers resemble those
of other 12/8 chorales – interludes in Vars. 2–7 of the Partita BWV 766,
accompaniments in BWV 666 and 667 – there is little attempt at developing
a single motif à la Orgelbüchlein. Many such patterns fall naturally to hand in
extrovert music in compound time, as is suggested by the similar opening of
the A major Suite for Harpsichord, BWV 806, which presumably the chorale
predates.
Interludes suggest congregational accompaniment, which the differing
length of first notes – two beats for the first two lines only – would not
unduly complicate, despite claims to the contrary (Sackmann 1998 p. 237).
However, without the interludes, the movement would be a useful pointer to
the Ob conception. At least one of the patterns was used in other connections
with the same chorale melody, in BWV 769. Although given in the Preller MS
with other Christmas hymns (BWV 722, 729 and 732), the setting is clearly
developed beyond the ‘Arnstadt Chorale’ type, chiefly by means of non-stop
semiquaver patterns, and is nearer than most ‘Neumeister Chorales’ to the
Orgelbüchlein.
BWV 738a Vom Himmel hoch, da komm’ ich her

Only copy, by J. T. Krebs (no composer’s name).
The figures of BWV 738a give the harmony of BWV 738; for a remark on
the relationship, see BWV 722a. That BWV 738a could be a reduction from,
not a draft for, BWV 738 is suggested by BWV 738a having those motifs in
BWV 738 unlikely to be readily improvised (b. 3, b. 4) but not those that
can be (b. 10, b. 20).
484 BWV 739
BWV 739 Wie schön leuchtet der Morgenstern

Autograph MS P 488 (see also BWV 764); copies in Mö MS (J. C. Bach,
similar paper to P 488, a MS he ‘slavishly’ copied: Hill 1991 p. xxv),
Plauener Orgelbuch (before 1710, from P 488 before its minor revisions);
later derivatives.
Two staves; headed in P 488 (a fair copy) ‘a 2 Clav. Ped.’.
The TEXT of P. Nicolai’s seven-verse hymn was published in 1599, associated

variously with Advent, Whit, Annunciation and Sundays after Trinity.
Wie schön leuchtet der Morgenstern How beautiful shines the morning-star,
voll Gnad und Wahrheit von dem full of the Lord’s grace and truth,
Herrn,
die süsse Wurzel Jesse. sweet root of Jesse.
Du Sohn Davids aus Jakobs Stamm, Son of David from the lineage of Jacob,
mein König und mein Bräutigam, my king and my bridegroom,
hast mir mein Herz besessen; you have taken possession of my heart;
lieblich, freundlich, lovely, kind,
schön und herrlich, fair and splendid,
gross und ehrlich, great and faithful,
reich von Gaben, rich in gifts,
hoch und sehr prächtig erhaben. exalted to great magnificence.
The MELODY, in part from older material, was published with the text,
and varied in detail. See Example 256. Listed in the Ob, set in BWV 763,
harmonized in BWV 436 and used in Cantatas 1, 36, 37, 49, 61 (1714), and
172 (1714).
Example 256
J. S. Bach’s hand in P 488 no more proves that he was the composer than
does the Plauener Orgelbuch copy. Nevertheless, the setting has hallmarks
of the young Bach, perhaps writing in preparation for the Lübeck visit in
1705, even for an audition there (Dürr 1984, plate 1). Either way, BWV 739
and 764 now represent ‘the oldest handwriting of Bach’ (Kobayashi 1989
pp. 25, 16), from about 1705, or already 1703/4; the paper dates between 1703
485 BWV 739
and 1709. P 488 could be a fair copy made from tablature, revised (Stinson
1985 p. 236), now with all pedal-entries clearly marked, lh/rh distribution
marked when needed (bb. 38, 44), and a bracket to show left-hand notes
(b. 55).
It would not be out of the question that some notes are incorrect, and
that the distribution in bb. 63ff. (four notes left, four right) does not convey
what happens in practice. The direction ‘a 2 Clav Ped’ (added when?) may
be an ideal.
The form and style are Thuringian rather than purely North German.
The note-patterns, the manual-changes and the sectional treatment of each
cantus line are reminiscent of Buxtehude and others, but like Böhm’s ‘Christ
lag in Todesbanden’, the setting is succinct; and like Buxtehude’s ‘Ich ruf zu
dir’ it remains continuous-but-varied by changing not metre but texture,
from two to four parts, including a trio with pedal as c.f. The details it shares
with the two extant and much longer chorale-fantasias of Reinken (echoes,
a certain square figuration) are a little too common-property for it to be
unarguably ‘indebted to a Reinken model’ (Wolff 1991 p. 63), although
Mö MS does contain various pieces looking like a ‘homage to Reincken’
(Dirksen 1998 pp. 133ff.). See also the Praeludium BWV 535a.
One can certainly find parallels between BWV 739, the incomplete
BWV 764 and a setting of ‘Erhalt uns Herr bei deinem Wort’ in Rinck’s
LM 4843 (Krumbach 1985, 5 p. 12). How typical of local Thuringian styles
in 1700 the conciseness is, or whether Bach had to learn from Buxtehude
(BuxWV 212, 196) about a cantus migrating from voice to voice, is un-
clear. Such contraction and migration seem characteristic of J. S. Bach
(BWV 739, 718, 720, 767.ix and 770.x), and in many respects, the closest
parallel to BWV 739 is BWV 720. The form is:
1–4 line 1, fore-imitation, cantus; unrelated interlude then line 2;

much use of a motif from b. 2
14–19 line 3, ditto; similar motif
21–35 lines 4–6, pedal; dialogue (lh Rp, rh Ow); Pachelbel
semiquavers
36–40 echoes, two manuals: harmonies drawn from next cantus line
40–54 lines 7–9 pedal; 8–9 anticipated by their harmony (bb. 44–5)
or melody (bb. 46–7); alternating manuals (b. 43 = Rp?);
broken chords
55–64 line 10 anticipated; scale lines derived from motif from b. 2?
65–end line 10 in pedal with free parts above; pedal point ditto
To vary the treatment is a way of responding to what is a somewhat

intractable melody, one which occasions very charming paraphrases in
486 BWV 739–740
Cantatas 1 and 49. The scales of the last two sections are typical of fan-
tasias, while other figures are of toccatas and preludes. Those of b. 42 and
b. 46 can be found in Buxtehude’s Toccata in D minor, other bar-by-bar
detail elsewhere: for example, the counterpoint at bb. 22 and 57 of Pachel-
bel’s printed ‘Wie schön leuchtet’ (Acht Choräle, 1693) has an ‘astonishing
relationship’ with BWV 739 and already served Buttstedt as model (Kube
1999 p. 582). But both the tenor of b. 69 and the arpeggio close, each a little
wild, are original Bach fingerprints.
Two particular questions about the manual changes are: why are there
directions for the Ow in both bb. 42 and 44? and do the signs suggest the
original was tablature? On the first: perhaps the right-hand O was a mistake
in b. 42; or R was omitted (for either hand or both hands) in b. 43; or the O
signs in b. 44 are merely cautionary, a warning against changing to R (Mö
MS has the sign only in b. 42). There is no compelling reason to take one
rather than another in simple dialogues; nor are two manuals necessary in
the way that they are for trios.
On the second: it is possible that the copy was made for the trip to Lübeck
(where Rückpositiv was common), and would have been less suitable for
auditioning or for testing a new organ in Thuringia, where chair organs
were scarce. Rather, O and R recall the practice in earlier chorale-fantasias
as copied in tablature: they could be typical or conventional terms meaning
primo and secondo manuals, neither of them truly piano. With comparable
plena between the manuals, it would not be mistreating the work to interpret
it as one wished, playing any section on either manual, rh or lh, above a
pedal reed.
BWV 740 Wir glauben all’ an einen Gott, Vater

Late copies only.
The TEXT of T. Clausnitzer’s Trinity hymn was published in 1668.
Wir glauben all’ an einen Gott, We all believe in one God,

Vater, Sohn und heilign Geist, Father, Son and Holy Ghost,
den der Cherubinen Rott whom the band of Cherubim
und die Schaar der Engel preist, and host of angels glorify,
den durch seine grosse Krafft, who through his great power
alles würcket, that und schafft. accomplishes, does and creates all.
As in Luther’s Nicene Creed (see BWV 680), three verses address the Persons
of the Trinity.
487 BWV 740–741
The MELODY was published in 1699 (Terry 1921 p. 339): Example 257.
The title’s first six words are listed in the Ob – for this text or Luther’s?
Example 257
A nineteenth-century editor attributes the two versions of this to

J. L. Krebs (see Emans 1997 p. 82), who may well be the composer, though BG
40 thought it an arrangement of a ‘piece by his great teacher’, perhaps for an
organ whose pedal went up only to a much-used middle c (Bruggaier 1959
p. 149). Probably both the four- and five-part settings, each in two rather
distinct versions, are the work of Krebs, including b. 7, the final melisma,
the double pedal, the frequent returns to tonic Fs in the bass, and the well-
digested fore-imitation, even the attempt at unity at the close of each half.
There is some general resemblance to BWV 653b, also known to the young
Krebs. But the four pedal cantus phrases are inconsistent (one is missing,
one is in both parts) and the harmony a mixture of good and bad (sequence
in b. 34). Or perhaps it is optionally laid out to be a four-part organ piece
with obbligato c.f. for violin, in the Kauffmann style developed further by
Krebs at Altenburg? Alternatively, there seems good circumstantial evidence
that the five-part version is an arrangement made in the nineteenth century,
by J. N. Schelble (Stinson BJ 2002, p. 131).
BWV 741 Ach Gott, vom Himmel sieh’ darein

Copies by or via J. P. Kirnberger, J. C. Oley and later.
Two staves; headed in the first ‘in organo pleno’.
The TEXT is Luther’s version of Ps. 12, later associated with various Sundays
after Trinity. Like others by Luther, the versification is ‘popular’, almost
folksong-like.
Ach Gott vom Himmel sieh darein O God look down from Heaven
und lass dich des erbarmen, and have pity on
wie wenig sind der Heilgen dein, how few your saints are;
verlassen sind wir Armen. we wretches are abandoned.
Dein Wort man lässt nicht haben wahr, Your word is not held to be true,
der Glaub ist auch verloschen gar faith is quite extinguished
bei allen Menschenkindern. amongst all the children of men.
488 BWV 741–742
The MELODY was published with the text in 1524: Example 258. Listed in
the Ob and set in Cantata 2 (1724) and (to other texts) 77 and 153.
Example 258
Despite problematic sources, the work is boldly said to be ‘obviously one

of the rare youthful works of Bach . . . revised between 1739/40 and 1750’
(KB p. 40), or directly related to the ‘Neumeister Chorales’ (Kube 1999
p. 581), or even ‘Bach’s first chorale-setting known to us’ (Meyer 1979b
p. 39). Zachow and J. M. Bach are also possible composers (Zehnder 1988
p. 100) amongst those able to stamp some individuality on a mixed genre of
fore-imitation, bass cantus in same note-lengths, sustained counterpoint,
inventive harmony and final double-pedal stretto.
But details in the harmony do not ring true – the major seventh in bb. 8
and 15, unusual resolutions (no third end b. 1, minor ninth b. 2, end b. 46
etc.), strange progressions (second half b. 56), spacing (bb. 45–6), novel
chromatics, an ending out of key (with its own perfect cadence), and many
others, all signs of a less skilful composer filling out canonic counterpoint
in such a way as to sound almost convincing. The form is also unusual:
seven cantus lines in the pedal, each preceded by fore-imitation, sometimes
in stretto, sometimes harmonized but incorporating a cantus motif (e.g.
soprano crotchets b. 13). The free chromatic tenor to line 5 of the chorale
melody (bb. 29–36) suggests a reference to the words of v. 1, as possibly does
the harmony of line 6 (bb. 45–7). But only at certain moments, such as the
fore-imitation of bb. 29–37, does one hear a composer well in control of the
harmony.
It seems unlikely that the pedal should play in the first six bars, de-
spite cues in some copies and despite what does seem to be a final double
pedal.
BWV 742 Ach Herr, mich armen Sünder

‘Neumeister Collection’ (headed ‘oder Herzlich thut mich verlangen J. S.
Bach’) and a late copy (C. Sasse).
489 BWV 742–743
For the MELODY, see ‘Herzlich tut mich verlangen’, BWV 727.
Sasse’s copy, perhaps via Kittel, does not justify the attribution to J. S. Bach
any more than for BWV 752 and 763, which it also contains: the open-
ing melodic gesture is uncharacteristic, as are some of the sequences and
the direction ‘poco adagio’. However, the three-part working and a rather
hectically varied figuration are not untypical of ‘Neumeister’ and fit in
with its intermittent air of experimentation. (Only Sasse gives the manual
indications.)
The Sesquialtera-like bravura part suits the picture of a young Bach
inspired either by Böhm (as at the opening – Seiffert 1904) and/or by the
violinists Westhoff and J. J. Walther (as at the close). Imaginative paraphrase
technique could also have been learnt from certain French livres or even from
Pachelbel, who, as here, would have brought in most notes of the cantus on
the strong beats. The free paraphrase is skilful and imaginative, ‘explaining’
the repetitive rh from b. 9, and wittily varying the line with slurred figurae
in bb. 18–19 and a dashing four-bar introduction.
But a good question remains: how young would Bach need to have been
to leave the implied parallels over bars 9–10 (Dürr 1986)?
BWV 743 Ach, was ist doch unser Leben

Copy in Lpz MB MS 7 (Mempell–Preller, from J. G. Walther?) and MS
R 24.
Like BWV 691a, BWV 743 contains a chorale (harmonization with inter-
ludes), prelude and postlude; it is likely to be no more authentic than BWV
691a and 683a, though for different reasons, i.e. a shorter authenticated
version is unknown. The figuration is characteristic of e.g. Armsdorff and
Kirchhoff, though on an unusual scale, and J. S. Bach’s influence may be
heard not least in the tone-painting of the victorious end (Luedtke 1918
p. 13). The shape resembles a small chorale-partita, i.e. variations framing a
harmonization complete with interludes and (in Mempell–Preller) making
use of two manuals. This manual-division is much like that in BWV 739
and as probably intended in BWV 764.
The strong close of the third section is characteristic of many chorales
in the ‘Neumeister Collection’, as is the anonymous two-verse form in
MS R 24 (i.e. without the central harmonization, attributed to Bach by
Mendelssohn). There is always the possibility that such a piece belongs to
the Arnstadt repertory (Krumbach 1985, 5 p. 17), and the stronger the
490 BWV 743–745
case for ‘Neumeister’, the likelier that MS 7’s attribution di J. S. B. is also

trustworthy.
BWV 744 Auf meinen lieben Gott

Copy by J. L. Krebs (P 802, anon) and a late source, P 311.
Two staves; headed ‘per Canonem’, followed by a second piece (see below),
‘alio modo per Canonem’; anon in P 802.
The facing pages in P 802 containing the two movements were insertions
made before 1731, originally entitled only ‘Chorale, per Canonem’ (Zietz
1969 pp. 93, 171). Perhaps Krebs was imitating BWV 714 previously copied
in P 802, first in canon at the octave and then at the fifth (fourth below), and
doing so very competently as part of studies that included a canonic setting
of ‘Herr Gott, dich loben alle wir’ (Tittel 1966)? Or the early editions inde-
pendent of P 802 and P 311 were following a reliable source in attributing
it to J. S. Bach (Emans KB).
The accompanying motif in the first canon (treated both in imitation
and sine pausa) is like that of BWV 714, less rigidly applied than in ‘Ach Gott
und Herr’ BWV 693. Only b. 13 seems unsuccessful, and b. 5 in the second
canon, both of which could result from study-sessions in canon-writing
by teacher and/or pupil. Krebs’s authorship is still in doubt (Weinberger
1986), and indeed ‘neutral’ canons may well have few distinguishing or
attributable characteristics. Perhaps they were not even organ-chorales in
any usual sense of the words.
BWV 745 Aus der Tiefe rufe ich

Late sources only, including P 285 and a MS of L. Scholz.
BWV 745 is a harmonization in full chords (implying pedal), somewhat

similar to but less convincingly handled than BWV 766’s or 770’s, and fol-
lowed by a fantasia incorporating chorale lines as soprano c.f. The harmonies
(e.g. augmented sixth in b. 11), cadences (b. 12), melodic detail (bb. 14–16),
form and obbligato-like texture of the whole make no part of it likely to
be the work of a composer working before 1750–75. It is largely identical
to the allemande of C. P. E. Bach’s Suite in E minor Wq 62/12 (1751, pub-
lished 1761), slightly modified as if optionally for organ. If this allemande
was made from a pre-existing chorale-setting by interpolating two-part pas-
sages in it – see Leisinger and Wollny BJ 1993 pp. 139–40 – then Philipp
491 BWV 745–748
Emanuel was adapting a work (by whom?) for a different medium, much
as Johann Christian did the (authentic) Praeludium from the Harpsichord
Partita in B.
No copies mention C. P. E. Bach but some late ones do J. S. Bach.
Like strict canons, full chorale-harmonizations have too few distinguish-
ing marks to give reliable evidence for authorship, though such a piece does
suggest the way chorales may have been treated, especially for keyboard
instruments other than organ.
BWV 746 Christ ist erstanden

Copy by J. G. Walther (attributed to Fischer), anon in later copies (P 311).
This was published in 1702 (or at least by c. 1715) as ‘Ricercar pro festis
paschalibus’ (J. K. F. Fischer, Ariadne musica), one of five such settings;
Fischer’s ‘Da Jesu an dem Kreuze stund’ is also in P 311, implying Bach as
composer (Emans KB). Since all are monothematic fugal treatments of a
chorale line without c.f. or pedal cues, BWV 746 is unlike chorales which it
is often said to resemble (BWV 707, 737 etc.).
BWV 747 Christus, der uns selig macht

Copy only in Lpz MB MS 7 (Mempell–Preller).
The confusion of idiom and form between one part of this chorale and
another – modern obbligato-melody opening, antique chorale-fantasia
close – makes it most likely to be the work of a young composer c. 1750.
(The usual dating of MS 7 ‘before 1747’ is not firm evidence against this,
nor is the fact that this MS may have drawn on Walther–Krebs copies for
some genuine Bach.) The texture in the middle of the work and the pedal
line throughout suggest no intimate knowledge of the organ, just as the rep-
etition suggests no creative gifts. Two possibilities, not mutually exclusive,
are that it is a transcription from an ensemble work, at least for the first half
(v. 1), and that an organ composer of mid-century, knowing such works as
BWV 718, cobbled together what he could.
BWV 748 Gott der Vater wohn’ uns bei

Copies include an album of Walther, to whom it is attributed (Emans 1997
p. 39).
492 BWV 748–751
The figuration and its working are similar to Pachelbel’s setting of the same
chorale, its partial canonic technique typical of Walther. P 285 and a Scholz
MS add five introductory bars ( = BWV 748a).

Late sources only (e.g. P 285).
With BWV 750 and 756, this is a fughetta somewhat in the manner of
the 44 Choräle attributed to J. C. Bach (†1703). Keller hears the youthful
J. S. Bach in its ‘suppleness’ (1948 p. 144), but in the nineteenth century it
or something like it was attributed to Telemann (Kobayashi 1973 p. 338). As
yet, there is no known way of tracing the young J. S. Bach’s language so as
to confirm that BWV 749, 750 and 756 are early, imitative works pre-dating
even the ‘Neumeister Collection’ (suggested in Wolff 1992 p. 249).
BWV 750 Herr Jesu Christ, mein’s Lebens Licht

Late sources only (e.g. P 285).
See BWV 749: the language of both is merely common property.

Copies in Yale LM 4708 (‘Neumeister’) and Lpz MB MS 7 (Mempell–Preller).
Two staves; headed ‘di Bach’ in MS 7, ‘J. M. Bach’ in LM 4708.
The Latin title is matched by an unusual italianate setting: the pastoral

pedal point accompanies two ‘verses’, charming in the hint they give of a
tonic–dominant musette and in the canonic ‘trahe me post te’ (‘draw me
after you’). But the piece is too undeveloped and simple to be likely work of
J. S. Bach. Keller already suggested a composer under Pachelbel’s influence,
and although ‘Neumeister’ is likely to be right, there is nothing comparable
in J. M. Bach’s extant music. Frotscher thinks the carillon figure in the third
and fourth lines deserves a Glockenspiel stop (1935 p. 934), such as was
known in Thuringia.
493 BWV 752–754
BWV 752 Jesu, der du meine Seele

Late copy, as for BWV 742.
Although the attribution to J. S. Bach is questioned usually because of the

weak canon (perhaps the result of an incomplete source – Emans KB), its
weakest points are the harmony and inept keyboard technique.
BWV 753 Jesu, meine Freude (fragment)

Autograph MS: CbWFB (probably early 1720).
Two staves.
The 8 34 bars are not known to have been completed, and the fragment
appears to be a first writing-down rather than a fair copy (NBA V/5 KB
p. 77). In CbWFB the decorated chorales appear as isolated movements. For
the order, see BWV 691; for the style, BWV 728; for a remark on incompletes,
BWV 573; and for a suggested completion of this one, see Schulenberg 1992
p. 135.
It is possible that BWV 691, 728 and 753 were meant as demonstra-
tions. BWV 753 has long runs of semiquavers, including a fine example of
the figura messanza in bb. 13 and 14: Example 259. Its inner parts are a
model, particularly the long rise in the tenor bb. 2–4, followed by a falling
bass.
Example 259

Copy in Lpz MB MS 7 (by J. N. Mempell, attributed to ‘Bach’).
Luedtke (1918 p. 20) thought the trio texture and bass quavers suggested
J. G. Walther, but the square phrases, continuo-like pedal, melodic detail
494 BWV 754–757
and simplistic harmony could well be those of a later generation, perhaps the
composer of Anh.II 55 (J. L Krebs?) or a Bach son attempting a proto-galant
trio on the model of certain moments in BWV 655. See also a remark on
BWV 759. The trio technique is elementary, the bass-line not very idiomatic
for pedal, the key-plan too dominated by tonic, and the cantus so thoroughly
paraphrased as to be barely recognizable.
Mempell writes a sharp by an E instead of a natural, presumably like his
source (Emans KB).
BWV 755 Es ist gewisslich an der Zeit

Late volumes (e.g. P 285, P1119) naming J. S. Bach only in overall title.
The c.f. with fore-imitation resembles Pachelbel’s treatment of the same

melody, and several composers of his ‘school’ could have composed it,
including the two Bach brothers Friedemann and Emanuel. Seiffert, who
knew a source with the title ‘Nun freut euch, lieben Christen’ (1904) –
hence BWV’s original title – thought it possibly a very early work of
J. S. Bach, with a ‘smoothness in the lower parts and certain turns of
harmony’ not typical of Pachelbel. In this respect the setting matches some
in the ‘Neumeister Collection’.
BWV 756 Nun ruhen alle Wälder

Late copies only (e.g. P 285).
See BWV 749, 750, 755. Unlike that of BWV 755, the fore-imitation answer
(bb. 2–4) does not contain grammatical errors, but can hardly be credited
to J. S. Bach, despite a certain ‘charm’ (Emans KB). And yet again, on the
reliability of P 285 hangs the possibility that this might be a very early
chorale-fughetta, a genre whose purpose is by no means clear.
BWV 757 O Herre Gott, dein göttlich’s Wort

Copies in Lpz MB MS 7 and P 409 (both second half eighteenth century).
For TEXT and MELODY, see BWV 1110. Both MSS sandwich the setting
between genuine Bach works: BWV 600, 757, 609 in the first, BWV 993, 651,
723, 736, 737, 540.ii in the second. As in BWV 755, the short imitations,
495 BWV 764–765
bass cantus firmus and semiquaver figures are characteristic of the Pachelbel
‘school’ – which could, after all, be taken as including the teenage Sebastian
Bach. Having fugal fore-imitations, the piece is unlikely to be a partita
movement (as Dietrich 1929 p. 66 asserts), and there is little to distinguish
it from countless other pieces of the kind.
BWV 758 O Vater, allmächtiger Gott

Only source, P 291 (late eighteenth century?).
Although the source is careful with its attributions and contains genuine
Bach works, BWV 758 is unlikely to be one of them. The techniques dis-
played in the four verses can be found in works of J. G. Walther, and if the
static quality of the first is uncommon after 1650, its archaic harmonies
are not, belonging as they do to a conventional genre. The pedal-point
(bb. 49–52) looks eighteenth-century but hardly a ‘very early work of Bach’
(BG 40).
The argument that despite ‘not having its own face’ (Keller 1937 p. 73)
a piece such as this may still be work of Bach (Emans KB), cannot easily be
developed further.
BWV 759 Schmücke dich, O liebe Seele

Copies in P 1115 (by G. A. Homilius) and by J. L. Krebs, etc.
The attribution to Homilius in a destroyed Hauser MS (Kobayashi 1973

pp. 76, 162), already noted in BG 40, is supported by the identification of
the copyist in P 1115 (Kast 1958 p. 62) and by its presence in the chief sources
of Homilius’s music (see EB 8541). The bass line, cadences and figuration,
plus the general competence, suggest a galant composer to whom several
spurious settings attributed to J. S. Bach might belong.

Three copies by J. G. Walther (attributed to Böhm), one copy each by
C. G. Gerlach c. 1730 (see BWV 708), and later copies via J. P. Kirnberger
and J. C. Kittel.
496 BWV 761–764
The coupling in P 802 of a bicinium with a fore-imitation setting (not

coupled in other Walther copies) accords more with Böhm’s chorales with
several verses than with full partitas. Though the later copies name J. S. or
J. C. Bach (†1703), details conform with other Böhm pieces (see EB 8087
p. 122). Kirnberger may have thought it a Bach work.

Copies by J. T. Krebs, and e.g. in Lpz MB MS 7 (Mempell–Preller, via
Walther?).
Two staves; headed by Krebs ‘a 2 Clav. e ped:’; anon (therefore JTK?).
The fore-imitation to each decorated cantus, whose notes are on the main
beats, is ‘good school-work but no more’ (Keller 1937), and the many returns
to the tonic are difficult to ascribe to J. S. Bach at any age. On the other hand,
many of the figures in the accompaniment (b. 7, b. 22 etc.) suggest that their
composer was acquainted with Bach melodies of the Weimar period; and
the pedal line is that of an accomplished player. The later copies have more
ornaments. If this part of P 802 is dated 1710–14 (Zietz 1969 p. 100) and
perhaps earlier still, the composer may be J. T. Krebs; a later date would
suggest other pupils.
BWV 763 Wie schön leuchtet der Morgenstern

See BWV 752 for a note on the source and the canon technique.
BWV 764 Wie schön leuchtet der Morgenstern (fragment)

Autograph MS: P 488 (c. 1705 or 1703/4: paper similar to Mö MS).
Two staves; headed ‘à 4’ (only).

497 BWV 764–765
In the autograph, BWV 764 follows immediately on BWV 739, is with-

out title or attribution, does not complete its second page and is followed
by empty staves. No completed copy is known, though in situ alterations
in note-values might suggest that P 488 was being made from another
copy (in tablature?), and this could have been complete (Emans KB).
The fantasia-like form of BWV 739 makes it unlikely that BWV 764 was
versus 2 in a pair of settings, on the analogy of BWV 760 and 761 in P
802, though this is possible. So consistent a figural counterpoint, at least as
it is so far in the fragment, is found also in several ‘Neumeister’ chorales
(see BWV 1104).
This might well be an early work of Bach himself, despite repeated
rhythms, a square motif partially shaped by the phraseology, and a sound
much like Walther’s. Its harmonic grasp is firm, there are signs that a new
motif is to develop – the first is surely exhausted – and changes in note-
value in P 488 give an impression of four parts being moved along. (This is
speculative, as is the idea in Stinson 1985 p. 236 that P 488 is the composing
score.) The main motif accompanying the first three chorale-lines is not
idly repetitious in either its rectus or inversus forms, and seems to be derived
from the opening chorale line, circumscribing the notes G D B.
BWV 765 Wir glauben all’ an einen Gott, Schöpfer

Copies by J. T. Krebs (four lines) and Lpz MB MS 7 (Mempell–Preller, two
lines only of the chorale, bb. 1–53).
Two staves; headed by Krebs ‘à 4 di [–]’, in Mempell–Preller ‘di J. S. Bach’.
As in BWV 705, 707, 737 and 746, the figures are typical of Scheidt’s
polyphony and later alla breve styles, but now developed farther. On grounds
of its competent and by no means simple handling it is frequently thought
authentic (Zietz 1969 p. 136). In Krebs’s P 802, Ob chorales too were nor-
mally anonymous.
The form is unusual. From each of the first two cantus lines a regular fore-
imitation is derived in stretto (cf. BWV 680), and the result is counterpoint
in an accomplished alla breve style (e.g. bb. 8–12). The expositions are not
regular, and cohesion is given more by the counterpoint which, doubtless,
includes references to the cantus. For example, the second line (cantus b. 20)
seems to be there in the alto from b. 10. In bb. 49 and 65, quicker, shorter
498 BWV 765
alla breve themes emerge and are answered rectus, inversus and in stretto, as
if someone was consciously running through conventional techniques; but
then they become themes more in their own right than is usual in the organ
motet. The final little flourish, more typical of Buxtehude’s ornamented
chorales than organ-motets, is not the least original touch.
Only four of the eleven lines of Luther’s chorale appear, a contraction
possible because of the virtual identity of lines 4 and 11.
Chorale variations (partitas) BWV 766–771
Whether J. S. Bach ever called such works ‘Partita’ is not known: ‘variatio’
or Veraenderung are the authentic titles in the Canonic Variations and the
Goldberg Variations. Secondary copies call the chorale-variations variat.,
variatio, partie and partite indiscriminately, and Walther used Partita for
BWV 770 but variatio for separate movements in Böhm’s ‘Wer nur den lieben
Gott’. The theme itself is often labelled variatio, whether a song (Buxtehude’s
‘More Palatino’, in tablature) or a chorale (Böhm’s ‘Jesu, du bist allzu schöne’
in the Mö MS).
What purpose chorale-variations have is uncertain, but presumably
they could have been used at home, in church (voluntaries, especially for
Communion? Kube 1999 p. 550), as interludes between congregational
verses, as models for independent chorale-preludes, or as exercises in dif-
ferent genres or composing by note-patterns. The common plan – a playing
over of the hymn, then a bicinium, then figural variations, various dance-
types, a final plenum chorale – suggests some of these uses more than others.
So do other sets of variations on chorales set by Bach, such as Buttstedt’s on
‘O Gott, du frommer Gott’ and ‘Sei gegrüsset’, which with Walther’s imply
that Thuringian organists had a common interest in such pieces, whether
for church or home. This interest would have been there irrespective of
what the young Bach learnt in Böhm’s Lüneburg, to which time and place
one can still sometimes find his chorale-variations being attributed. Except
for the Canonic Variations, they presumably date from the time up to the
Weimar appointment, BWV 768 perhaps beyond, but in any case never as
a group. Although there is no single theme, as there is with Pachelbel’s four
sets in his Musicalische Sterbens-Gedanken of 1683 – all of which deal with
death and eternity – BWV 766, 767 and 770 are comparably concerned with
Evening, Lent and eternity. Perhaps this suggests some domestic use for such
music.
BWV 766 Christ, der du bist der helle Tag (‘Partita’)

Copies in Darmstadt Mus. 73 (first half eighteenth century), Lpz MB
MS 4 (middle eighteenth century, no Partita VI), by or via F. Hauser and
J. C. H. Rinck (1770–1846, no Partita VI); reputed autograph (Kobayashi
1973 p. 255), probably a copy.
[499]
500 BWV 766
Two staves; headed in Darmstadt ‘Partite diverse. . . . manualiter’, each

movement ‘Partita’, No. II ‘Largo’.
The TEXT of E. Alberus’s hymn ‘Christe, du bist der helle Tag’ or ‘Christe,
der du bist der helle Tag’ is a translation of ‘Christe, qui lux es et dies’ (Lent),
published in 1556 and later used as an evening hymn:
Christ, der du bist der helle Tag, Christ, you who are the bright day,
vor dir die Nacht nicht bleiben mag. before you the night may not endure.
Du leuchtest uns vom Vater her You illumine us from the Father
und bist des Lichtes Prediger, and are the preacher of light.
und bist des Lichtes Prediger.
The six verses that follow are a prayer for safety.
The MELODY was published with the text in 1568, differing from version
to version: Example 260. Harmonized in BWV 273 and listed in the Ob.
Example 260
The seven movements seem not to ‘describe’ the seven verses but follow the
plan of chorale-variations c. 1700, for domestic music-making. Although
the variation closest to harpsichord idiom is omitted in some copies
(VI, rejected by later tastes?), the key of F minor makes harpsichord like-
lier than organ to be instrument of first choice. Pedal seems to have been
added to the last (see below), as if the original was very early, though not
necessarily from the Lüneburg period (see Spitta I p. 207). A new element
is the inclusion of interludes between the chorale-phrases, less conspicuous
than in BWV 715 or even 738, but not merely in the bicinium, as had be-
come conventional. All the variations belong to types long known, and the
comparisons given below are selective.
Partita I (Chorale)
The harmonies of the chorale are fuller and more varied than in ‘Sei
gegrüsset’ or in Böhm and Pachelbel partitas. The four- to seven-part har-
monies seem also to be conceived less obviously for the organ than those of
e.g. BWV 715.
501 BWV 766
Partita II (Var. 1)
This is a traditional bicinium without opening bass statement. But it does
have the ‘anticipatory’ first phrase before the melody proper enters in
b. 2 (cf. BWV 767 and 768); also, the ends of phrases are repeated (cf.
J. G. Walther’s ‘Schmücke dich’, Vers 3) and marked piano. Perhaps late
sources are untrustworthy and there was and/or should be an opening bass
statement.
In such ritornello movements the bass line both accompanies and adds
its own melody, as in the continuo arias of early cantatas (e.g. BWV 106,
‘In deine Hände’), cello-like and slurred somewhat like the bass line in
the sixth movement of Cantata 71 (1708). Halfway through the variation
(bb. 15–21), the melody is spun out in imitation, as in versus 3 of Böhm’s
‘Auf meinen lieben Gott’.
The slurs, less consistent in the sources than in modern collated editions,
are puzzling. If they are meant to last a beat (KB p. 193), is the line broken
up? Are they the composer’s, or a copyist’s? – one who knew BWV 639
(also in F minor), where an unbroken cantabile line is more plausible than
here?
Partita III (Var. 2)

The cantus is accompanied, embroidered and separated off by a little motif
rectus or inversus in every bar, i.e. with a single-mindedness unknown except
occasionally to Pachelbel or Zachow, who however do not demand such
finger-dexterity. Other motifs add variety, as does the altered form in the
penultimate bar (soprano). A c.f. line disguised by its own counter-motif is
common in keyboard partitas, e.g. Böhm’s ‘Wer nur den lieben Gott’ Partita
4, and Pachelbel’s ‘Christus, der ist mein Leben’ Partita 1. Example 261 can
stand for many.
Example 261
Partita IV (Var. 3)
The rh perpetuum mobile variation is traditional (examples in Pachelbel
and Böhm); less usual is that the rh line explores one or two motifs and
takes in interludes between the cantus phrases. While the manner of dis-
guising the theme is also typical of Böhm, the opening passage is moving
towards Bach’s maturer paraphrases: the notes FFGAFABC can still be
502 BWV 766
made out on the beat and yet the line is an independent melody of strong
character.
Partita V (Var. 4)
Placing the chorale melody in the middle part (en taille), with interludes
between its phrases, is familiar from larger-scaled sets of chorale-verses (e.g.
Buxtehude’s ‘Ach Gott und Herr’). In keyboard partitas, en taille variations
usually have no interludes (e.g. Pachelbel’s ‘Herzlich tut mich verlangen’).
Less common is the density with which a complicated motif is developed,
one announced by the rh and appearing several times in each bar, without
inversion or disguise. As in Var. 2, the midpoint of the movement sees the
fourth line of the melody anticipated and developed before its full entry in
b. 13. It is difficult to see such treatment as justifying Terry’s view that ‘they
are not in the ordinary sense Variations at all, but movements in Fantasia
form’ (Terry 1921 p. 112), except that he meant to praise them as more than
mere formulaic variations.
Partita VI (Var. 5)
The influence of such chorale-suites as Buxtehude’s ‘Auf meinen lieben
Gott’ on this gigue movement is clear, as it is in others using much the
same motifs spilling across the cantus firmus (e.g. Partita 9 of Böhm’s ‘Freu
dich sehr’). While the phrase structure of Var. 5 is curiously bicinium-like,
the melody itself is completely integrated in a texture even more motif-
ridden than Var. 4. The main motif is fragmentary and versatile, open to
constant alteration, compelling the cantus to change octave, at one point
interrupted by repetitious tonic–dominant harmonies (between b. 9 and
b. 12).
Partita VII (Var. 6)

The pedal rubric ‘con pedale se piace’ (‘pedal ad lib.’) is from the Hauser
MSS, and the part may derive only from copyists. While the rhythm of
the accompanying figure can be found in (e.g.) Böhm’s ‘Wer nur den
lieben Gott’ Partita 5, the idea of a final bass c.f. is familiar in sets by
older composers (Buxtehude’s ‘Nun lob, mein Seel’) and by Bach himself
(BWV 656). Apparent heterophonic doubling of pedal and left-hand bass
parts is not uncommon in Pachelbel chorales and surely means no more
than an optional alternative, for the piece is carefully conceived for two
hands.
The startling treatment of the melody from b. 10 onwards, with broken
chords in the Böhm tradition (G minor Praeludium, and also BWV 535),
gives the movement an effect greater than expected. Perhaps refers to the
doxology of the final verse?
503 BWV 767
BWV 767 O Gott, du frommer Gott (‘Partita’)

Copy by J. T. Krebs; Peters V used ‘a very old copy’ from Forkel’s Collection.
Two staves throughout; called ‘Partite diverse über O Gott du frommer Gott’
in a catalogue of 1781 (Dok III p. 269); movements headed ‘Partita’ in Krebs.
The TEXT of J. Heermann’s hymn was published in 1630; various verses

became associated with Sundays after Trinity in Weimar and elsewhere
(Gojowy 1972).
O Gott, du frommer Gott, O God, righteous God,
du Brunnquell guter Gaben, fount of good gifts,
ohn dem nichts ist, was ist, without whom nothing is that is,
von dem wir alles haben: from whom we have everything;
gesunden Leib gib mir grant me a healthy body,
und dass in solchem Leib and that in such a body
ein unverletzte Seel an unviolated soul
und rein Gewissen bleib. and a pure conscience remain.
Various verses that follow pray for safety in all danger, including death:
lass hören deine Stimm let your voice be heard
und meinen Leib weck auf, and waken my body,
und führ ihn schön verklärt and lead it, beautifully transformed,
zum auserwählten Hauf. to the chosen throng.
A doxology was added in some versions. Listed in the Ob.
The MELODY was published with the text in 1646; the form here does not
appear in Zahn: Example 262. It is one of three with this text: (i) BWV 767;
(ii) Cantatas 24, 71, 164; (iii) Cantata 45 (Nos. 64, 94, 128, 129, 197a to
another text).
Example 262
As with BWV 766, the absence of pedal does not indicate date. Conjectural
too is Schweitzer’s interpretation of the last three variations in relation to
504 BWV 767
the hymn (1905 pp. 65–6): the falling line of VII expresses death and burial
in v. 7; the chromatics of VIII, a ‘sad wait for the signal of resurrection’ in
v. 8; and IX, the ‘animation’ of a doxology. Keller heard in the unexpected
andante–presto passages of IX the last two lines of v. 8 (1948 p. 137). This
last, which is less fantasy-like than the finale of BWV 770, has also been
likened to the French dialogues aux grands jeux (Klotz 1975), though these
have no cantus and are rare for actual finales.
Since such variations work with conventional figurae, and are so shaped
as to increase in complexity, it is unclear how much more they ‘express’ than
similar but less expert sets by Walther or other Thuringers (cf. Ziller 1935
p. 46). Perhaps the copy in MS P 802 is the product of instruction Krebs took
with Bach in Weimar (KB p. 196). For comparison with works by Böhm
and Buxtehude, see BWV 766.
Partita I (Chorale)
For such harmonies (with strong up-beats), see Var. 1 of BWV 766. The
shape of the melody as it appears here (A1 A1 B A2) gives a rounded form
to each variation.
Partita II (Var. 1)
With a lh introduction, an anticipated rh first phrase (bb. 2–3), repeats in the
melody, and an ostinato-ritornello theme in the bass, the movement has all
the hallmarks of the traditional bicinium and can be compared with Böhm’s
melodic fragmentation or with the bicinium of BWV 768. One individual
detail is that the melody is rather more cut up than usual (see BWV 711 and
718 first part): parallels can be made with early continuo arias (such as the
ritornello-like movement 4 from Cantata 131, 1707), but a comparison of
bb. 2–7 with 31–42 – sections based on similar chorale lines – shows how
the idea is developed in an organ bicinium.
Partita III (Var. 2)

The suspirans (first four rh notes), a standard figure in German variation-
types, is found in other partita movements (e.g. here Var. 5), the Ob and
other chorales that by this means look rather like a partita movement (e.g.
BWV 690). It sustains motion, suggests imitations rectus and inversus, and
can imply a thematic reference, since it passes through the opening fourth
(g–c ) of the cantus.
Partita IV (Var. 3)
For perpetuum mobile variations, see as well as ‘Neumeister’ BWV 1106,
‘Nun freut’ BWV 734 and BWV 768 and 766. Note that in detail – number
of parts, a continuous or disjointed lh, the violinistic character of rh – the
505 BWV 767
corresponding movements in the three partitas show three distinct treat-

ments. Another appears in cantatas, e.g. BWV 4.iii. (For the bass motif of
Partita IV, see also BWV 644: since this too has continuous rh semiquavers,
was it deliberately modifying a type?)
Partita V (Var. 4)
Now the suspirans is extended into longer scale sections, as too in partitas
of other composers (e.g. Böhm’s ‘Ach wie nichtig’ Partita 4), complete with
broken-chord cadences. Less usual is the resulting octave displacement of
the chorale melody, migrating down through two (or even three) octaves in
the course of bb. 9–12.
Partita VI (Var. 5)
The unusual bass part, like a cello obbligato, may owe its origin to the stan-
dard lh divisio (i.e. variation with lh passage-work) of the partita tradition,
as in Böhm’s ‘Gelobet seist du’. A bass aria written for keyboard, it para-
phrases the harmony more prosaically than Bach’s Cello Suites do, but at
times anticipates them.
Partita VII (Var. 6)

The triple-time variation seems (like Var. 9 of BWV 768) an equivalent
to certain movements in Pachelbel’s partitas or earlier composers such as
Froberger, who included 3/4 dance-types in their sets. The scale motif of
Var. 6 might be a more original touch, while the style of the middle sec-
tion is not unlike that of the Courante in Buxtehude’s ‘Auf meinen lieben
Gott’.
Partita VIII (Var. 7)

Whatever this movement may be ‘expressing’, chromatic variations near
the close of a work were long familiar: Froberger’s in ‘Auf die Mayerin’
was already within the tradition. The main motif is the bass’s chromatic
fourth, as found in the last variation of Scheidt’s ‘Da Jesu an dem Kreuze
stund’ (the ‘choralis in cantu per semitonia’, in Tabulatura nova 1624),
where as here it appears both rectus and inversus. There are also at least
ten appearances of the tenor’s opening chromatic motif. (One wonders if
BWV 767 is a transcription from D minor, the most common chromatic
key?)
The unusual harmonies produced by such chromaticism (e.g. b. 19,
second beat) were traditionally associated with chromatic countersubjects.
The coda, particularly bb. 17–18, resembles cadences in the instrumental
Sonatina from Cantata 106 (1707?).
506 BWV 767–768
Partita IX (Var. 8)
The form of this fantasia-like finale, unique to Bach variations, can be
expressed:
1–3 line 1/3 of the chorale melody; last phrase extended

8–9 line 2/4
11–14, 19–22 lines 5 and 6, both with echoes in and after the phrase
26–7, 35–6 line 7 (Andante in P 802) and line 8 (Presto in P 802)
As in French dialogues, the short phrases imply that the whole work is
nearing its close, a musical doxology perhaps. They are carefully varied in
length, for example four quavers (b. 16), then eight (bb. 16–17), then twelve
(bb. 17–19).
The second manual allows complete or partial echoes at the same octave,
the octave below or the octave above. Its use from b. 26 on is not clear from
Krebs’s copy (KB p. 199): it seems that each hand in turn plays a solo phrase
accompanied by the other, but the intention could be that a forte phrase
in one hand is then echoed piano by the two hands together. The dynamic
markings are too few to show either of these conclusively, though the second
is plausible, since each forte statement contains an echo-like repeat. Clearly,
both hands are forte from the middle of b. 31 though not necessarily earlier,
and both are piano from the beginning of b. 33.
The phrase for the andante section in b. 26 is like some of Walther’s,
and the echoes recall the Suite in B BWV 821, Finale. As Keller suggests
(1948 p. 137), perhaps the andante corresponds to the line of v. 8 ‘and lead
[my body] beautifully transformed’, just as the adagio section of Partita 7
of Böhm’s ‘Wer nur den lieben Gott’ (known to Walther) may refer to the
line, ‘For who puts his confidence in God’. Tempo can act as a reminder of
the text and need not imply a change of manual, while extant registrations
in BWV 720 are a warning that copyists might sometimes be adding their
own suggestions.
BWV 768 Sei gegrüsset, Jesu gütig (‘Partita’)

No autograph MS; contemporary copies by J. T. Krebs (Vars. 1, 2, 4, 10) and
three others: Carpentras MS 1086 (copyist known from Plauener Orgelbuch,
Vars. 1, 2, 3, 4, 5, 6, 11, 9, 7, 8, 2, 4, 10, 3), Lpz MB MS 7 (J. G. Preller, Vars.
1–6, 10, 7, 9, 8, 12, 11) and Lpz MB III.8.17 (another Leipzig student? Vars.
1–5, 7, 6, 9–12); also a former Königsberg MS (contemporary? Vars. 1, 2, 4,
10, 3, 5, 7, 11, 9, 6, 8); others via J. P. Kirnberger and later, ultimately from
same source as III.8.17?
507 BWV 768
Two staves; heading in Krebs ‘cum 4 Variat’, in Carpentras ‘à 11. Part’, in
Preller ‘Partite diverse’ (but movements ‘Variatio’), III.8.17 ‘per il Organo’,
and in MSS via Kirnberger ‘Variationen’. Also, ‘à 2 Clav’ in Carpentras for
Vars. 3, 7, 9, in Preller for Vars. 3, 10, and in III.8.17 for Vars. 5, 7, 8, 9, 10,
where Var. 11 is also marked ‘in Organo pleno’.
The TEXT of C. Keimann’s hymn was published in 1663; several hymnbooks

relate it to the prayer ‘Salve [or Ave] Jesu, summe bonum’.
Sei gegrüsset, Jesu gütig, Hail to you, kind Jesus,

über alles Mass sanftmütig, Beyond all measure gentle,
ach! wie bist du doch zerschmissen, O, how you are dashed in pieces,
und dein gantzer Leib zerrissen! your whole body torn to bits!
(R) Lass mich deine Lieb ererben, (R) Let me inherit your love
und darinnen selig sterben. and die happy in it!
The hymn exists in versions of five or seven stanzas, the first (as in the
Weimar Gesangbuch 1713) with a common refrain (R); vv. 6 and 7 have a
different refrain:
Singen immer Heilig, heilig: Sing always Holy, Holy,

alsdenn bin ich ewig selig. then I shall be ever blessed.
J. Böttiger’s ‘O Jesu, du edle Gabe’, a Jesus-song for Communion, has the

refrain:
Dein Blut mich von Sünden wäschet Your blood washes me from sin
und der Höllen Glut auslöschet. and extinguishes the fires of Hell.
and is throughout less meditative than ‘Sei gegrüsset’. On the grounds of

supposed text–music correspondences, it is the ‘more correct’ title (Clement
1993 p. 193).
The MELODY appears with the text in 1682, as Example 263. (Its history is
unclear: Grimm 1969 p. 172.) Both texts are listed in the Ob. Harmonized
Example 263
508 BWV 768
in BWV 410 and 499, and used for Walther’s chaconne ‘O Jesu, du edle
Gabe’.
Schweitzer saw that BWV 768, because the number and order of variations
differ, would not serve to relate text and setting (1905 p. 66); but he accepted
Spitta’s conclusion that the variations belong to different periods (I p. 594),
as have many later authors: a Weimar set revised later. Luedtke (1918
pp. 47ff.) tried to show that Vars. 8, 9 and 10 referred to a text ‘added’ to
the ‘earlier’ variations, Zehnder (1995 p. 335) that both longer and shorter
version belong to 1711/13, though the cited comparisons to the Ob and
Böhm would equally suggest half a decade earlier. Ulrich Meyer has argued
that of the various orders, ‘the latest and most convincing’ is 1, 2, 3, 4, 5,
7, 6, 8, 9, 10, 11 (1973 p. 478), given in EB 6590 but not NBA IV/1, and
producing a clear pattern: opening chorale, four semiquaver settings, two
demisemiquaver, two compound time, two triple, then final chorale. On the
other hand, 7, 8, 9, 10, 11 keep the pedal variations together.
Since the point of chorale-variations is to explore conventional genres
one by one, only subjectively can they be dated or ordered chronologically.
It could be that too much is still being argued from extant sources as to
the ‘correct’ order (Kube 1999 pp. 556ff.) and the way it relates to the text,
arguments ruled by later ideas on ‘organicism in music’. While some sources
do no doubt transmit minor revisions, there is nothing concrete to say that
Krebs’s first four variations were the ‘original’ or ‘earliest’ to which others
were added gradually (or at once), or to conclude that there was then a
‘second version’ with a new, ‘cyclic order’ (KB p. 206). It is certainly possible
that the text is relevant and that this is, in one or other version, the first
fully patterned organ cycle. But equally, the text may be less relevant than
technical details like number of parts or the presence of pedal or even (for
an over-arching conception) the particular sequence of time-signatures.
Chorale
In its four fluent parts, this not only is more organ-like than opening chorales
in BWV 766 and 767 but ‘opposes’ the finale’s five parts to make a framework
with it.
Var. 1
Unlike BWV 711, the bicinium is organized into three paragraphs, with a
motif found in the fugue-subject of BWV 578.
1–12 lines 1 and 2, to relative major

12–22 lines 3 and 4, to subdominant
22–37 lines 5 and 6; with final ritornello (unlike BWV 766, 767)
509 BWV 768
This shape can also be expressed as a regular alternation between an ostinato

phrase of three bars and chorale lines of more than three. The result–
3 bars – then 8 12 – 3 – 6 12 – 3 – 9 – 3
– is much closer to new Italian ritornello-arias than to any early bicinium of
Scheidt, and was surely influenced by them. This particular bass is very like
that of an aria by Marcello later used by Mattheson as a model; see Williams
1984 p. 113.
Line 1 is anticipated in the rh of b. 4, as in Böhm’s ‘Vater unser’ and
certain arias in Weimar cantatas, and the melody matches the ritornello-
like bass in having a common close to lines 2 and 6 (Meyer 1979). In its
spinning-out of cantus lines well beyond their two-bar phrases, this melody
is theoretically like that of ‘Nun komm, der Heiden Heiland’ BWV 659,
but its tempo must be livelier, and the ornamental technique, as in the last
phrase from b. 28, is not yet so touching and original. The bass contains
two important motifs (Example 264) both of which can be found in J. B.
Bach’s ‘Jesus, Jesus, nichts als Jesus’. Though not unreasonably likened to
ostinatos, the development and clear returns of these motifs in different
keys are merely characteristic of ritornello themes in general.
Example 264
Preller’s decorations (NBA IV/1 pp. 155–7) make it yet more like the
work of Böhm, and the movement looks later or more developed than such
chorales as BWV 718. Flowery melodic lines in Cantatas 12 and 21 suggest
a date for it near theirs, thus later than BWV 766 and 767; but dating across
genres can be misleading.
Var. 2
In a three- and four-part texture around new harmonies, a pattern first
heard in the tenor is exploited rectus and inversus in the inner parts, three
times passing into the melody (i.e. more often than in e.g. BWV 653 or 644).
It is present in every bar, but the beat on which it enters, its direction and
its intervals are all carefully varied.
Var. 3
A two-part perpetuum mobile like many another – Walther’s ‘Meinen Jesum
lass ich nicht’, published 1713, or Pachelbel’s ‘Herzlich tut mich verlangen’
510 BWV 768
Var. 7 – this is joined by non-stop lh quavers. The semiquavers develop in

their own way, and become less dependent on the cantus, particularly after
the ‘interlude’ bar (b. 5), where a variant of the basic motif arises.
Var. 4
A texture otherwise resembling BWV 644 (G minor scales) is varied by a
second semiquaver motif (tenor, b. 2). When this appears, the number of
parts always increases to four. Though only a three-note figure, it is more
than a mere harmonic decoration as in Var. 4 of Pachelbel’s ‘Alle Menschen
müssen sterben’, also known to Walther. Its harmony varies from beat to
beat – each of the three notes may or may not be part of the chord – and is
insistent, like a bariolage.
Var. 5
Presumably ‘à 2 Clav’ in III.8.17 is a mistake. The three ideas are cantus
firmus, bass divisio (much the same line as Var. 4) and smooth inner parts,
fuller in the second half. Some attempt is made to mitigate the squareness
of the bass motif, especially towards the end. Such motifs had remained
undeveloped and repetitious in earlier keyboard partitas (e.g. Partita 5 of
Pachelbel’s ‘Christus der ist mein Leben’), as they had too in some French
basse de trompette pieces.
Var. 6
The opening bass motif of the 12/8 variation occurs widely in music, both
for organ (Walther’s ‘Jesu Leiden, Pein und Tod’) and elsewhere (Finales of
Bach’s F minor and G major Violin Sonatas), but is perhaps nowhere so
fully exploited as here. Unlike a similar motif in BWV 626, it also affects the
melody. Variety of treatment is achieved not by replacing the motif but by
breaking it up into sub-motifs.
To group the pedal-variations together (Vars. 7–11) might seem reason-
able, but it is not certain that Var. 6 should precede Var. 7.
Var. 7
This is similar to a trio in Böhm’s ‘Christe, der du bist Tag und Licht’ (also
copied in P 802) which, after several verses, introduces pedal c.f. below
imitative upper parts, at last running together in sixths. Both composers
use the little demisemiquaver suspirans figure, but the present variation is
more continuous, evidently avoiding (like BWV 694 and 710) crossed hands.
Something similar is found in Pachelbel’s partitas, but the dramatic leap of
lines here is unusual.
511 BWV 768
Var. 8
Once again a motif, here the circulatio or familiar curling figure of compound
time, is absorbed into the chorale-melody itself, having originated as a
paraphrase in b. 1. The motif is presented along with its own inversion,
which becomes increasingly prominent, particularly above (or below) the
penultimate chorale line (b. 10) and the final pedal point. The 24/16 time-
signature is presumably intended to show that this is a different pattern
from the one in 12/8 in Var. 6.
The tendency from Var. 4 on for the melody’s last note to be drawn out
is explored further in Vars. 7, 8 and 9, as if they were deliberately demon-
strating three different figural embroideries for a final pedal point. (Perhaps
they are.)
Var. 9
Here too, as in Vars. 7 and 10, a second manual is not necessary from the
point of view of crossed parts. As in several of Pachelbel’s partitas, the c.f.
sings in the tenor between two highly imitative parts. As in BWV 688, the
duo-like manual parts begin to incorporate inversions of both their note-
patterns (compare bb. 1 and 15). And as in BWV 661, the final bars include
rectus and inversus together, again in G minor.
Var. 10
Unlike the Preller and other sources, P 802 and Carpentras have no heading
‘à 2 Clav e Ped’ but mark the rh cantus sections ‘Choral’. The NBA solution,
with rh and lh together except for the solo cantus, rather simplifies the
scoring: perhaps the MS leaves it to the player (or his circumstances) to
decide how it could be laid out.
Similarities between Var. 10 and other organ works are striking. In its
sarabande-like air it is clearly comparable to three settings in ‘The Eighteen’
(BWV 652, 653 and 654), with an accompaniment especially like BWV 654’s;
and in a more general way, the ostinato-like pedal and chaconne-like motion
resemble Walther’s chaconne on the same melody.
The form is unusual:
plain cantus firmus, each line after an ornamented version on the same
notes, lasting six bars; from the first phrase a figure is derived
(Example 265) which becomes ostinato-like both in pedal and in
ornamented soprano
Canonic treatment of a similar figure can be heard in the Sarabande of the

B minor Ouverture BWV 1067.
512 BWV 768–769
Example 265
If the rh is played solo, so presumably are the last two lines of the melody
when it splits into two parts (b. 75), marked ‘forte a 2 voci’ in Kirnberger
sources (i.e. two parts on one solo manual). Alternatively, the plain rh c.f.
phrases could be played on the lh’s softer manual, the decorated phrases on
the louder (a Sesquialtera stop?), though the spacing of the hands suggests
the opposite. Perhaps piano and forte mean a change either of registration
or of manual, with a third manual if available playing the echo, as in the
later E Prelude.
Mostly, the inner parts accompany simply, with surprisingly little imi-
tation, and they never make use of the ostinato figure that appears nearly
forty times in the pedal, not even in the coda, where the top part is coloured
by it. The last twenty-five bars in particular have the very mellifluous har-
monies, part-writing, motivic bass and melodic flair of an accomplished
organ-chorale.
Var. 11
The ‘organo pleno’ heading matches that for BWV 667: a final tutti hymn
without interludes, a richly harmonized finale such as was already estab-
lished in J. C. Bach’s Aria Eberliana, 1690. Unlike superficially similar move-
ments in the Ob (e.g. ‘Jesu, meine Freude’), the five parts do not much
develop a motif, either in the pedal or (despite the continuity of the woven
lines) in the manual. As in Var. 10, these five parts are more simple and
homogeneous than those of Ob’s ‘Liebster Jesu’, not paired off but each
vigorous and active like a full-throated choir.
BWV 769 / 769a Canonic Variations, Vom Himmel

hoch, da komm’ ich her
Stichfassung, print version
Published 1747 (Kobayashi 1988 p. 60). Title-page:
Einige canonische Veraenderungen über das Weynacht-Lied: Vom Himmel

hoch da komm ich her. vor die Orgel Mit 2. Clavieren und dem Pedal von
Johann Sebastian Bach Königl: Pohl: und Chur Saechss: Hoff Compositeur
513 BWV 769
Capellm. u. Direct. Chor. Mus. Lips. Nürnberg in Verlegung Balth:

Schmids.
Some Canonic Variations on the Christmas hymn Vom Himmel hoch, da
komm ich her. For organ with two manuals and pedal, by Johann
Sebastian Bach, Royal Polish and Electoral Saxon Court Composer,
Kapellmeister and Director of the Musical Ensemble, Leipzig. Nuremberg,
published by Balthasar Schmid.
Autograph manuscript version

A section of MS P 271 (before August 1748: Kobayashi ibid.), headed ‘Vom
Himmel hoch, da komm ich her. per Canones. à 2 Clav: et Pedal’.
Origin
In the Obituary of 1754, Lorenz Mizler added a note about Bach’s member-
ship of the Leipzig ‘Society for the Musical Sciences’ founded by Mizler in
1738:
In die Societät der musikalischen Wissenschaften ist er im Jahr 1747
Junius . . . getreten . . . Zur Societät hat er den Choral geliefert: Vom
Himmel hoch da komm’ ich her, vollständig ausgearbeitet, der hernach in
Kupfer gestochen worden. Er hat auch den Tab. iv. f. 16 abgestochenen
Canon, solcher gleichfalls vorgeleget . . . (Dok III pp. 88–9)
In June 1747 he entered the Society for the Musical Sciences . . . He
presented to the Society the chorale Vom Himmel hoch da komm’ ich her
completely worked out, and this was afterwards engraved on copper; in
much the same way he also presented the canon printed in Plate IV
Fig. 16 . . . [ = BWV 1076]
The canon BWV 1076, written at least a year earlier (Kobayashi 1988 p. 55),
not only shows the interest that Mizler’s Society took in such music but has
a bass much like the melody of ‘Vom Himmel hoch’ and the opening bass
of the Goldberg Variations (1741), in a copy of which the composer wrote
down this canon plus thirteen others (the Fourteen Canons, BWV 1087).
See Example 266.∗ Wolff 1991 p. 177 thinks that ‘actually, the thematic
relationship [between BWV 769 and 1087] suggests that the idea to celebrate
∗ The eight-note theme is also found in other guises: the opening toccata-like semiquavers of
Buxtehude’s C major Praeludium BuxWV 138 or the theme of the Basse de trompette (septième ton)
in Jacques Boyvin’s Second Livre d’Orgue (Paris, 1700).
514 BWV 769
Example 266
the Christmas cantus firmus originated as an afterthought in connection

with the fourteen canons’. But as BWV 700 and 701 show, Bach had long
explored various contrapuntal ideas with this melody.
Mizler shows no awareness of the possible significance of Bach’s being
the fourteenth member of his Society (B + A + C + H = 14).
Origin of the Canonic Variations

Researches on extant copies, their publisher and engravers conclude that the
print was probably engraved and published in 1746 or for the New Year’s
Fair 1747 (Butler 1990 pp. 91ff.). Perhaps the canons were begun, or at least
Nos. i–iii completed, for some special occasion at Advent or Christmas 1745.
Some five stages have been suggested (KB pp. 11ff.):
A composition of the Canon at the Octave, Canon at the Fifth,

Canon at the Seventh, the Augmentation Canon
B composition of the largest Canon with Inversions, perhaps with
Mizler’s Society in mind
C preparation of copy for engraving ‘in learned showpiece notation’
D new amended copy of A (known from three copies), based on
Mizler’s presentation copy and probably becoming part of
Breitkopf ’s collection
E fair copy (P 271)
A is a generalization, B conjectural; C, D and E can be discerned from dif-

ferences in the order of movements and discrepancies in the text (including
ornaments); each stage also includes both ‘early’ and ‘late’ readings. Recent
re-examination of the print-copies and the autograph fair copy suggests that
there was no fixed or ‘final’ version (Butler 2000 p. 34), as was probably the
case too with many a chorale and free work. The print version of BWV 769.i,
ii and iii seems to precede that in P 271, vice-versa for movement iv, and
for movement v the two were probably contemporary. Also, presumably the
canons were originally worked out on paper.
The alto line of the Canon at the Seventh is more florid in E than in C,
with details seen now as characteristic of Bach revisions. However, other
details suggest that D and E give an earlier reading than C; E also contains
515 BWV 769
errors suggesting it was not checked. Any date proposed for E and C does
not affect the putative sequence, since for C this is the date of publication.
Whether or not A and B are discernible stages is open to question, but the
answer could have a bearing on the order of pieces (see below); so could
a yet different interpretation – that D was the last version, with a simpler
alto coloratura made for the Canon at the Seventh when engraved (Emery
1963). From Kobayashi 1989 one could conclude the following:
a original version of i, ii, iii and then v (engraved before summer

1746?)
b iv later, to complete cycle and for copy to Mizler (June 1747)
c copy of title-page etc. for engraving (printed by Michaelmas
Fair, 1747)
d revised text, as copied in P 271 (full open score, as throughout
this MS)
e composer’s copy of the print, with revisions
f new print pulled in 1751
Two conclusions from this, in which c and d could be reversed, are that first,
composition took over a year and went through at least two changes; and
secondly, the fair copy version (P 271) represents the composer’s desired
order, while the other came about for one or other mundane reason (see
below). However, both are speculative, if plausible. The verses of the chorale’s
text might be read as corresponding to the engraving’s order (Clement 1989),
but there are too many unknowns – which is the definitive version of the
augmentation canon? – for such speculation to clarify much.
Order
The two major differences between BWV 769 and 769a are the order of
movements:
769 i ii iii iv v
769a i ii v iii iv
and the notation:
769 i 2 staves, canonic answer not written after the first five notes
ii 2 staves, ditto after three
iii 2 staves, ditto after eleven
iv 4 staves, open score, four different clefs
v 3 staves
769a all movements on three staves, all canons written out
516 BWV 769
The exceptional nature both of the sources and the music makes the cus-
tomary distinction between early and late versions, authorized and non-
authorized, by no means as certain as often assumed. Canonic chorales in
which the canons are not realized demand work ‘with the pen at home’ (as
Marpurg said in 1759 – Dok III p. 127), and the published BWV 769 could
not, any more than the Fourteen Canons, serve as a playing copy or indeed
enforce any particular order.
It is no more certain that Bach prepared the version for engraving (KB
p. 94) than he did the almost contemporary open-score Schübler Chorales.
The idea that Var. v must be the last to be written because the most cumulative
rather ignores the ingenuity already displayed in the earlier setting BWV 701.
Nor because it ends supposedly with a B A C H reference (KB p. 91) does
it mean it comes last: neither the notation (there is no b in the print) nor
part-writing (the notes are dispersed between voices) suggests B A C H.
Its forte close and combination of chorale lines, however, are comparable
to the final variation of the Goldberg Variations, and clearly, it is a suitable
end-piece, its final ingenuity, alas, clearer to the eye than the ear, as D. G.
Türk pointed out in 1787 (Dok III p. 432).
The engraving’s order may also be notational: it is arranged so that
no turn-over is required, and each two-page opening has a different stave
system. It cannot be out of the question that the engraver Schmid was
responsible for the whole of the appearance, saving space and eliminating
page-turns. After all, it is Var. iv that is visually a climax, being laid out in
open score, and the one with the indisputable reference to B A C H near its
end.
The order of movements in the autograph, though independent of layout,
is equally logical, perhaps more to the composer’s taste than a modern
performer’s:
i canon (possibly derived from melody) above c.f.
ii canon (certainly derived from melody) above c.f.
v various invertible canons from the cantus, final stretti and
diminution
iii canon (derived from melody) + free part, c.f. in soprano
iv canon + free part, pedal c.f. (en taille)
Each order – (a) progressive and (b) symmetrical – appears to be authentic
and reasoned; and whichever came first, the composer seems not to have felt
bound to it. For different reasons, the order of movements in the Musical
Offering and the Art of Fugue is also confused, suggesting respectively a
progressive and a symmetrical arrangement. The Goldberg Variations can
be seen to be both, to some extent rising progressively towards the end, to
some extent ‘rotating’ around Variation 16.
517 BWV 769
The music
Earlier settings of the melody in BWV 606, 700, 701 and 738 had already
produced motifs open to imitation or combination above or below the
chorale-melody. The early BWV 700 not only ends with the same pedal point
but, like 701, combines cantus lines 1 and 2 and imitation per diminutionem,
features found in BWV 769.
At other moments, the canonic idiom recalls the Goldberg Variations:
compare Canonic Variation i with Goldberg Variation 3, including the bass
line. Melodic lines are occasionally similar (e.g. some phrases of Canonic
Variation iv and Goldberg Variation 13), as are several turns of phrase or
actual motifs (e.g. Canonic Variation iii b. 13 and Goldberg Variation 15
b. 1). Although the resemblance is subtler than a few shared motifs suggest,
the ruling difference between the two works – one is based on a melody, the
other on a bass – prevents them from being too alike. Clearly, with the strict
canons of the Musical Offering and two fugal canons in the Art of Fugue,
Bach’s canonic composition was ranging wide at this period.
The Canonic Variations offer ground for speculations. Smend counted
49 chorale lines throughout the work (7 × 7), producing 441 notes
(7 × 3 × 7 × 3) – though whether this includes ties is unclear (Smend 1969
p. 169). As in BWV 606, 701 and 738, the various scale passages can be
seen ‘not only [to] represent the ascending and descending angels, but [to]
sound joyous peals from many belfries ringing in the Saviour’s birth’ (Terry
1921 p. 307), if we assume bellringers of Saxony rang in diatonic scales. The
so-called dragging motif in Var. iii recalls that of the Ob’s ‘O Lamm Gottes’,
thus relating Christmas and Passion (Klotz 1973 p. 14), but it also appears
in the Goldberg. Naturally, the falling lines of Var. i (particularly beginning
and end) can be seen as the Saviour’s descent from heaven to earth, and the
rising lines closing ii and iv as departing angels, the soul’s elevation, etc.
Canons were not uncommon for Christmas chorales, and Schein’s Opella
nova I (Leipzig 1618) had included a setting of ‘Vom Himmel hoch’ with
canonic phrases and intermittent c.f. At least one of the Fourteen Canons
was explicitly symbolic: BWV 1087.xi, ‘Symbolum. Christus Coronabit
Crucigeros’, ‘Symbol: Christ will crown the cross-bearers’ (referring to the
chromatic sharps). Furthermore, despite its jejune harmony, Kauffmann’s
Harmonische Seelenlust could well have been a stimulus of some kind, with
its Christmas canon above pedal c.f., a descending motif in C major (as in
BWV 769.i), a walking bass in ‘Vom Himmel hoch’ (as in BWV 769.v), and
various syncopations. Perhaps the Canonic Variations, the Schübler Chorales
and Clavierübung III were all a response to Kauffmann’s volume.
In speculating about rhetoric, Zacher 1981 analyses the variations to
show that melodic patterns allude to particular words. The first four verses
518 BWV 769
underlie the first four variations, so that in Var. ii when the elect Virgin is
‘picked out’ (‘auserkorn’), the alto elects to rise to a dissonant f and its line
is thus picked out. However, f could be a misprint (a in the autograph),
and there is no reason why dissonance should express ‘standing out’, or
why so important a word as ‘auserkorn’ is marked so obscurely (a single
semiquaver off the beat). As often with such interpretations, Zacher draws
other unsupported conclusions: BWV 769a was prepared ‘certainly for a
later publication’, with particular personae associated with particular figurae
(angel in the canon at the octave, shepherds at the fifth, child at the sixth),
etc.
Neither the ‘some variations’ of BWV 769’s title-page nor the ‘several’ of
the Goldberg’s are variations in the familiar sense (the type Bach elsewhere
called ‘double’), despite the print calling each movement ‘variatio’. Nor is
BWV 769 a chorale partita in the sense of BWV 768. While particularly
Var. i, ii or iv could serve as prelude to the sung hymn, clearly the work has
more scientific aims, as perhaps is implied when the title-page says ‘Lied’
(‘melody’), as distinct from ‘Chorale’ for the Schübler. In their form, the
nature and ingenuity of their counterpoint, their complexity of notation,
the association of their texts with Advent and Christmas, and above all their
musical language, the Canonic Variations are an obvious contrast to the
Schübler Chorales. Perhaps Schübler had originally been intended for the
same publisher?
The melody, canonically generated harmony and keyboard idiom are as
unlike Schübler and other organ music as the Goldberg Variations are other
harpsichord music: at times strangely new, at others very approachable,
the style is elusive enough to prompt admirers to search outside music for
suitable expressive metaphor. There is certainly nothing quite the same in
the cantatas. Wherever one looks, something unusual happens: the stamping
bass below short phrases in Var. v changes to a smooth cantus firmus; the
final C major quodlibet quickly summarizes the proceedings, more so than
at the end of BWV 547; the ingenious free lines in Vars. iv and v ‘explain’
the harmony and counter the other lines. But more than this, there are
moments, especially in Vars. iii and iv, that touch the listener, showing how
at their best canons create harmonies, melodies and progressions not only
otherwise unheard but strangely rapt and intense.
Variation i
BWV 769 ‘Variatio 1’, ‘in Canone all’ottava à 2 clav. et pedal’

BWV 769a ‘Canone all’ottava’, ‘à 2 Clav: et Pedal.’ (added?)
The canon is a two-part fore-imitation to the pedal’s c.f., its subject perhaps
paraphrasing both the first and the last lines (cf. BWV 651). Both subject
519 BWV 769
and counterpoint recall the opening of ‘Christe, du Lamm Gottes’ in the Ob,
and the running figure itself is scattered around the literature, from Georg
Muffat’s Toccata No. 12 (Apparatus, 1690) to various sonatas of Domenico
Scarlatti.
The canon seems particularly successful when the phrases are short or
when the c.f. is supplying a clear bass line (e.g. bb. 10–12, 16–18), although
compound time enables the composer to deal fluently with the chords as
they arise (e.g. b. 3).
Pedal 8 is called for, since the style does not demand 16 tone (as for
plenum) nor the spacing 4 (as for a solo). Subjectively speaking, BWV 769a
supplies the better text across the barline 7–8, neutralizing the chromatic
motif of BWV 769.
Variation ii
BWV 769 ‘Variatio 2’, ‘Alio Modo in Canone alla Quinta à 2 Clav.
et Pedal’
BWV 769a ‘Canone alla quinta’, ‘canto fermo in Pedal’
The canon takes the form of a two-part fore-imitation before the pedal c.f.,
based on the first line (bb. 1, 16) and the second (b. 10). Like the Canon at
the Second in the Goldberg Variations, a canon at the fifth produces a more
natural line – with sequences – than a canon at the unison or octave. Bar
12 resembles a sequential figure in ‘Wir glauben’ BWV 680 (bb. 72ff.), and
the imitative use of scales (b. 5 etc.) and leaps (b. 20 etc.) is characteristic of
chamber trios. In b. 16 the theme returns in the tonic, recapitulation-like
(though here syncopated), as in b. 13 of Var. i.
Pedal registration is as in Var. i. Now the pedal point (which by itself
makes both 16 and 4 unlikely) has ascending figuration above in the manual
parts, which leave the canon incomplete at the close. The versions do not
differ in ways significant to the performer; but note that the print (because
the canons are not written out) does not direct left-hand sharps in b. 3.
Variation iii
BWV 769 ‘Variatio 3 Canone alla Settima’, ‘cantabile’

BWV 769a ‘Canone alla settima’, ‘cantabile’
The canon takes the form of a pair of lower voices running as a kind of
ostinato with ‘interludes’, against a free melody in the alto and c.f. in soprano.
The canonic parts begin with the first line of the melody and continue
to allude to it. The interludes become somewhat ostinato-like, occurring
approximately when the c.f. lines appear in the top part, rather as episodes
in a ritornello chorale (compare BWV 662 etc.).
520 BWV 769
The alto melody, though not formally an aria, sounds at times as if

it were – see, for example, several motifs introduced in the first few bars
(Example 267). Those labelled a b c d e f and g can all be found in other highly
Example 267
embellished pieces, such as the slow movement of the F minor Harpsichord

Concerto BWV 1056, while many cantabile cantata and keyboard move-
ments contain the motifs labelled x, y and z. Some, like the ‘dragging motif ’
c, may have textual connotations; and the varied appoggiaturas increase
the aria-like effect, adding gratuitous clashes (e.g. bb. 10, 25, 27) as if to
‘explain’ those that arise through the canon. At other moments, the free alto
line hints at the chorale melody (e.g. b. 10), as do the inner parts in the
final bars. Neither version – nor any other reading produced by juggling
with accidentals – softens the effect of b. 19, which is both logical and very
striking, the more so as it precedes a simpler passage leading to the lovely
cantabile close.
The chief difference between the versions is that the autograph has a
rather more ornate melody and more motif-exploitation as the movement
proceeds, and a little more apparent freedom of line (e.g. b. 6) made easier
by being on three staves – a revision or the original? While a diatonic
interpretation of the unwritten canon in the print is no doubt correct,
some notes are more unexpected than others (e.g. f in b. 8, c in b. 23).
The arioso alto makes a 16 +8 pedal possible, and ‘Canon at the Seventh’
does not mean that it cannot be a ‘Canon at the Fourteenth’.
Variation iv
BWV 769 ‘Variatio 4 à 2 Clav. et Pedal per augmentation. in Canone

all’ottava’
BWV 769a ‘Canon per augmentationem’, ‘à 2 Clav: et Pedal’
The canon takes the form of a new, long, melismatic subject in the soprano
followed in doubled note-values in the manual bass, against a free alto
521 BWV 769
and the tenor c.f. in pedal (which came first?). There are no significant
differences between versions except placement. Using three manuals, if the
keys are short, is not impossible.
The variation’s four parts cross more than is usual in BWV 769, and
while the free line often imitates one or other of the canonic parts (as do
free lines in the Goldberg), all three refer to the theme from time to time.
Example 268 gives a few instances; those in the soprano are later augmented
in the bass, and the whole variation becomes a dense tissue of allusions.
Those meandering in the soprano part are inconspicuous and mostly en
passant, becoming clearer when they appear in augmentation in the bass.
Example 268
The weight of reference is hardly oppressive: references to themes are woven

in without seeming repetitive or contrived, and lines 2 and 3 of the melody
never appear in undecorated form because their leaps of a fourth could not
so easily have been integrated.
While therefore in Vars. i, ii, iii it is the harmonic implications and in Var.
v the canonic potential of the original melody that occupied the composer, in
Var. iv it is the melody’s simple cells. Significantly, lines 1 and 4 are similar
to each other in their scale fragments, lines 2 and 3 in their fourths and
repeated notes. Moreover, if lines 2 and 3 are expressed in a more ‘fluid’
form with passing-notes, they begin to resemble lines 1 and 4. However
conscious the composer was of this, its results can be seen throughout the
unusual soprano part. Every scale passage, every fourth (e.g. b. 32) invites
522 BWV 769
the listener to hear the chorale melody, even to ponder and admire the
manner in which a given theme can be spun into a long melody.
In some respects this is more like Goldberg Var. 13 than Canonic Variation
iii – i.e. the right hand is more in the style of Bach’s mature harpsichord
music than organ. The soprano ranges widely from c to c , and the whole
of its second half (from the latter part of b. 21) is free, working towards a
rich close with special spacing for the last chord, full of thematic allusion.
As in Vars. i and ii, there is a quasi-ritornello return of the opening melody
towards the end (b. 34, top line, in diminution). The final bars, whether
derived or not from the cantus, are in the coloratura tradition, a developed
form of the final pedal point known from (e.g.) Buxtehude’s ‘Durch Adams
Fall ist ganz verderbt’. On the other hand, the dragging motif in bb. 38f. (c in
Example 267) and the fact that the pedal point begins in the ‘wrong’ key are
unlikely elsewhere. The one dry moment of the melodic line (b. 14) produces
a very good bass line in bb. 27–8, implying that the composer’s technique
in writing canons was to find the bass-line first, or to give it priority when
problems arose.
The final B A C H might be twofold (Example 269, both parts) and comes
in the last bars to be written by Bach himself in the MS P 271. Especially
these bars show how immense motivic ingenuity need not dehydrate the
music, for in b. 39 are heard the ‘dragging’ motif, two or three motifs a,
the c.f., the augmented canon, and a melodic line in b. 20 so managed as to
allow the canonic bass part of b. 39 to combine with B A C H (or vice versa).
But the end-result is one of the best bars in the whole of P 271.
Example 269
Variation v
BWV 769 ‘Variatio 5’, ‘L’altra Sorte del’Canone all’rovercio

[rovescio], 1) alla Sesta, 2) alla Terza, 3) alla Seconda è 4)
alla Nona’
BWV 769a ‘Canto fermo in Canone’, ‘alla Sesta è all’ roverscio [sic]’,
then ‘alla Terza’, ‘alla Seconda’, ‘alla Nona’
Now the canon passes to the chorale-melody itself, but a canon of another
kind in that its answers are per giusti intervalli:
523 BWV 769
1 cantus in inverted Canon at the Sixth, line by line; pedal continuo

bass
14 cantus in inverted Canon at the Third (thus like the preceding
canon – line 3 also drawn out to a four-bar phrase); pedal
continuo bass
27 cantus in inverted Canon at the Second, bass and tenor;
occasional free part in alto, quasi-imitative; free part in soprano
40 cantus in inverted Canon at the Ninth between outer parts
(therefore like the preceding canon); free part formerly in
soprano, now in tenor
52 cantus line 4 in pedal, plus diminutio recta and inversa of line 1;
pedal point plus stretto of lines 1, 2, 3 and 4, with diminutio as
before
The giusti intervalli produce the f in the canon of b. 32.

The build-up of Var. v is unmissable, but this description does not do
justice to all the coda’s musical subtleties. The pedal statement of line 4
acts as a confirmation of the previous soprano phrase (compare BWV 664);
the variation’s pedal point produces the lowest note of the five movements
(compare it with BWV 547), and of the organ; the chromatics are slight
but telling (Example 270); and there are finally six parts, the number having
risen from one/two/three (bb. 1–2) to four (b. 28) and five (b. 53). Moreover,
all six parts are thematic, i.e. including the second soprano running in thirds
(canon sine pausa) and the first tenor altering line 3 in the final bar or so.
Example 270
A further detail is that just as the five movements become gradually

longer, so the chorale melody changes: the c.f. is almost the same in the
first two canons and identical in the last two. The free parts may ultimately
derive from melodic cells (e.g. line 1 in b. 36, now minor), but there is a clear
change in character between the free semiquavers of b. 51 and those of the
diminutio in b. 52. The free opening pedal-line is less conventional than the
derived bass in Var. iii, and looks like a more thorough version of the kind
of walking pedal-part sometimes found elsewhere, e.g. Kauffmann’s ‘Vom
Himmel hoch’ already mentioned. Very striking is that the two-bar phrase
structure of the chorale melody is emphasized throughout by the canonic
524 BWV 769–770
answer appearing at the same point in each line, i.e. halfway; to counter any
unwanted dryness, line 3 is expanded into three bars, and later on non-stop
semiquavers and syncopations are introduced. Mizler’s phrase ‘completely
worked out’ is no empty description.
The versions do not differ significantly, but the forte signs are problem-
atic. Of versions C, D and E, only E (P 271) has forte at b. 27; in b. 39, all
versions have forte, but the print puts it between staves, and not below the
left hand as in NBA IV/2. Whether forte means a louder manual or extra
stops, it is difficult to see why it should appear in b. 27, for any or all the
parts; at the end of b. 39, it would apply to both manuals. If, however, it
suggests merely a second manual, it would be especially relevant to the sec-
tion bb. 27–39, and either way, the hands must be on the same manual at
the beginning of the diminutio.
BWV 770 Ach, was soll ich Sünder machen (‘Partita’)

Copies in P 802 (J. G. Walther, movements 1 and 2 now missing), P 489 (first
half of eighteenth century), Brussels Cons. XY 15.137 (nineteenth century).
Two staves; ‘Partita terza’ etc. in P 802; in P 489 ‘Partite diverse’. Manuals
‘Organ:’ (Ow) and ‘R’ or ‘Rückpos.’ in P 802.
The TEXT is J. Flittner’s seven-verse Jesus-hymn of 1661:
Ach! was soll ich Sünder machen? Oh, what should I, sinner, do?
ach! was soll ich fangen an, Oh, where should I begin?
mein Gewissen klagt mich an, my conscience accuses me,
es beginnet aufzuwachen; it begins to awake.
dies ist meine Zuversicht, This is my confidence:
meinen Jesum lass ich nicht. I do not forsake my Jesus.
The MELODY was published with the text. Harmonized in BWV 259.
The attribution in P 802 may be the work of J. L. Krebs (Zietz 1969
p. 101) for what is possibly the earliest of the authentic chorale partitas,
an Arnstadt work (? KB p. 177). However, in being very like harpsichord
variations of the late seventeenth century, this work is not quite like the
others. The bottom C in Partita IX might suggest a later date were it not
that in the ABB the Fantasia BWV 563 also has this note, and the passage
following the C recalls the early keyboard Sonata in D major BWV 963. At
any given date the harpsichord compass is likely to be less narrow than the
organ’s, and the full harmonies of the opening hymn, in the Böhm manner,
certainly look like harpsichord music.
525 BWV 770
P 489 comprises only BWV 770, the Brussels MS also BWV 739. The
copy P 802, though more weighted towards organ music than P 801, con-
tains Pachelbel’s similar variations ‘Was Gott thut, das ist wohlgethan’, just
as Walther’s MS Kö 15839 contains both BWV 768 and seven of Böhm’s
and Pachelbel’s (harpsichord?) partitas. The running bass of Partita V does
not appear in Bach’s organ variations, though it is common in Pachelbel,
while the running soprano of Partita VI (‘reminiscent of J. K. F. Fischer’:
Löffler 1923) is not much like comparable movements in BWV 766, 767 and
768. However, the extant copies, apparently independent of each other (KB
p. 176), transmit at least the possibility of reading Partita X as organ music,
with appropriate manual-change.
Whether the forte–piano directions in Partita IX mean the same as
manual-changes, or originated in the same source, or are reliable/authentic
and more than optional, is unknown. The Adagio implies something
approaching a Böhmian sarabande. Apart from echo passages (bb. 17ff.,
25ff.), the two manuals are used for a question-and-answer that is known
in later harpsichord music (e.g. B minor Ouverture BWV 831, Echo) and in
principle different from that of the last movement of the Variations BWV
767. The technique might be derived from French dialogues, as in Grigny’s
Livre, and such a passage as bb. 44–9 is characteristic of the French style as
transmitted most significantly by Johann Kuhnau.
Doubt has also been expressed about the manual directions in Partita
X, which were probably added later in P 802 (Emery 1970 p. 168); if so, the
intention may have been to make the movement conform to short fantasias
such as BWV 718, which make use of various organ-chorale techniques,
commonly supposed to be in response to the Affekt of each line of text.
The fantasy of the movement, however, again seems rather to suggest a
harpsichord piece, borrowing or adapting treatments from the ‘northern’
chorale-fantasia in its own terms.
Partitas II and III are so like corresponding variations in Böhm’s ‘Ach
wie nichtig’ as to ‘seem to copy Böhm directly’ (Zehnder 1988 p. 92), specif-
ically his harpsichord style. It is striking that there is no bicinium, as in the
organ variations BWV 767 and 768. The planning of the set, however, is
surely more thoughtful than in regular variations of the period, as the detail
of contrapuntal or motivic working results in richer harmony than usual.
Particularly the final movement looks farther afield than conventional sets.
It is certainly possible to consider BWV 770 one of the composer’s first
masterworks and one as versatile in genre as other early works, such as the
Sonata BWV 963. To be versatile in this sense means to be both domestic/for
harpsichord and liturgical/for organ.
There are many details of musical interest in the fine keyboard effects
of BWV 770, filling out the picture of a young Bach as it was sketched in
526 BWV 770–771
BWV 771 and in the ‘Neumeister Collection’, where the settings differ from
chorale variations in having inter-line interludes. Thus the ‘harpsichord
harmonization’ in Partita I, rather reminiscent of the Aria Variata BWV
989, is followed in Partita II by the kind of Corellian bass that colours
many a two-part variation, as in Var. 4 of the Aria Variata. The rh part in
Partita III could be the violin obbligato in a chorale-based cantata c. 1690,
while there are analogies to Partitas IV and V in many other harpsichord
variations, particularly those influenced by Froberger. Just as from time to
time the Sonata BWV 963 anticipates much maturer pieces in D major (see
Prelude BWV 532), so bb. 2–3 of Partita VI are like the opening of the G
major Prelude BWV 541 – a violin solo making much of crossed strings. It
is difficult to believe that the present Partita VI, the ‘Neumeister Chorale’
BWV 1106, and even the G major Toccata BWV 916 (cf. the Adagio), were
written many years apart.
The gigue idiom of Partita VII is familiar both in suite-finales of Bux-
tehude, Böhm or others and in works of Handel (see HWV 429, 432, 438,
443, 450), which draw on similar traditions, probably in much the same
period. The principle behind the motivic consistency of Partita VIII might
have been picked up from Buxtehude (e.g. BuxWV 250). But like the early
Sonata BWV 963 and Fantasia BWV 563, Partita IX suggests more cosmo-
politan influences perhaps via Kuhnau, such as the question-and-answer
phraseology of early concertos (bb. 1ff.), the frenchified paraphrase and imi-
tation of bb. 29ff. ( = bb. 7–8 of the theme), the chromatic fourth (bb. 37ff.),
the common-property theme of bb. 44ff., and so on.
The last Partita expands freely on phrases of the melody, which is heard
more or less straightforwardly at certain points (bb. 1, 9 end, 17, 25, 33 end,
40/62), and the resulting fantasia is one that aims at looking afresh at the
conventions. Though shorter, ‘Neumeister’ chorales BWV 1092, 742, 1102,
1114, 1115 are still similar enough to seem even earlier in the composer’s
development. It is clear that the composer of Partitas III, VI, X and perhaps
VIII (compare this with the Adagio of the Sonata in F minor for Violin, BWV
1018) was very familiar with pre-Vivaldian violin idioms, while the opening
paragraph of Partita IX would not be out of place in Handel’s Concerti grossi
Op. 6 or any other concertos imitating Corelli.
BWV 771 Allein Gott in der Höh’ sei Ehr’ (‘Partita’)

Copy in P 1143 and Brussels Cons. XY 16.142 (eighteenth century?).
Though under Bach’s name in P 1143, this is more likely to be the work of
Nikolaus Vetter (1666–1734) to whom Vars. 3 and 8 are attributed in SBB
527 BWV 771
Mus. MS 40035. But these movements show a more primitive use of motif
than do others, for which Spitta proposed young Bach under the influence of
Pachelbel (I pp. 250–1) – as shown in the number of parts and the treatment
of the c.f. in Vars. 2 and 11.
But weaknesses are evident: Vars. 1, 2, 11 and 13 are less developed
than trio or melodic treatments in (e.g.) Böhm’s ‘Herr Jesu Christ, dich zu
uns wend’. There is a reliance on simple, undeveloped motif in Vars. 3, 4
(cf. a figure in the anonymous BWV 743), 5, 6, 7, 9, 13, 14 and 17, while the
Pachelbel influence on Var. 12 seems to have led to nothing special when
compared with BWV 768, Partita iii. Yet while the fughetta of Var. 8 has
none of the hallmarks of shape, motif and harmony of BWV 716, despite
the latter’s simpler texture, it does have a partial c.f. in the bass at the end,
rather suggesting that examples of this very feature, found in mature Bach
settings of ‘Allein Gott’, were traditional.
BWV 790 Trio in B minor
Copy, Lpz MB MS 1 (via Kellner?).
Headed ‘Trio ex H mol di J: S: Bach’.

A transposed version of the Sinfonia in D minor BWV 790, perhaps this
was made from J. P. Kellner’s copy of the Sinfonias in P 804 (see Stinson 1990
p. 47). Along with other trios copied or owned by organists in Kellner’s circle
(BWV 1027 or 1039), it typifies activities of a Bach pupil and/or his pupils,
as perhaps does the arrangement BWV 131a, known from more copies.
[528]
Four Duets from Clavierübung III BWV 802–805
Published 1739, no Autograph MS. Later copies only.
‘Duetto’
Some uncertainty as to a suitable name for imitative two-part pieces is
suggested by the term ‘Praeambulum’ for works in the CbWFB later called
‘Invention’ in the fair copy of 1723. Theorists understood ‘duetto’ as (i) a
‘petit duo’ above a bass (Walther’s Lexicon 1732, after Brossard’s Dictionaire
1703), (ii) a dialogue aria with an ‘opportunity to introduce and develop
two subjecta opposita’ (Mattheson, Critica musica II, 1725, p. 28), (iii) an
instrumental or vocal piece above a bass, ‘skilfully fugued’ (ibid. I, 1722,
p. 131), and (iv) a two-part piece incorporating more than mere ‘imitation
at the unison and octave’ (ibid. pp. 305, 360). Although the last comes closest
to BWV 802–805, it refers less to technique than to a form.
Only in general terms do the Four Duets allude to French organ duos,
which in Grigny, Raison, Boyvin and Du Mage have no more than a loosely
organized counterpoint. Similarly, in length and idiom they go far beyond
the two-part fugal verses in Pachelbel’s Magnificats. In Cantatas 140 and 110,
‘duetto’ is a dialogue or duet plus bass. In the search for wider significances
in Clavierübung III, Lutheran devotional dialogues have been invoked to ex-
plain the term: that between pastor and pupil in the catechism (Humphreys
1994 p. 48), or between the Soul, the Word, Cross, Death and Heaven in the
influential Geistliche Erquick-Stunden, 1672 (Clement 1999 p. 320).
The instrument
J. E. Bach remarked that Clavierübung III was ‘mainly’ for organists
(‘hauptsächlich’: Dok II p. 335), and much in the Duets looks quite unlike
organ music: spacing and countersubject in No. 1, spacing and figuration in
No. 2, the opening bass of No. 3, answers at the twelfth in Nos. 2 and 4, etc.
The two-part sections in the E Fugue are much more conventional organ
textures, as are the Two-part Inventions which, despite frequently made
claims, the Duets do not much resemble in any way – in texture, form,
themes or length. Duet to No. 1 is only superficially similar to the Courante
[529] of the E minor Partita for Harpsichord, and of the four, No. 4 most resembles
530 Four Duets
the mature harpsichord music. But none feels as idiomatic as two-part music
in Clavierübung II and IV (B minor Gigue, Goldberg Variation No. 1).
Nevertheless, as also for the canons of the Art of Fugue, a harpsichord or
(as then new) a fortepiano is a good contrapuntal worktool for such music.
Although the compass CDE–c looks more organ-like (as in the Canonic
Variations) than GG–d (as in the Goldberg Variations), it is also an ‘ideal’
counterpoint compass.
As part of Clavierübung III

Although in performance the Duets can intersperse the chorales (Albrecht
1969), there is no evidence for them as ‘Communion music’ (Keller 1948
p. 198), nor is it certain that Clavierübung III was as liturgical in practice
as on paper. Even less certainly do the Duets allude to the book’s chorale-
melodies (Ehricht 1949–50). Is it really significant that like the fughetta
BWV 679, No. 3 is in G major and 12/8 time, or that selective sieving can
find ‘Allein Gott’ in it? See Example 271.
Example 271
Extra-musical speculation has been popular partly because the Duets

plus the six manualiter catechism settings make ten pieces, and Vopelius’s
Gesangbuch has the same six catechism hymns (Commandments, Credo,
Lord’s Prayer, Baptism, Confession, Communion) plus four prayers
(morning, evening, before and after eating). However, any idea that the
Duets somehow represent these or the four table prayers in Luther’s Small
Catechism (Leaver 1975) or refer to the cross (Clement 1993 pp. 209ff.)
can never be pursued with concrete evidence. A more musical question
is: are the changes of key – E minor after the last chorale’s F minor, then
A minor before the Fugue’s opening b – so extreme in order to mark them
off?
Purpose
Perhaps the Duets were slipped into Clavierübung III for the convenience of
the book-maker (to complete the page-gatherings) or to make the number
of pieces twenty-seven (3 × 3 × 3). The first seems unlikely, since the duets
531 Four Duets
take up exactly eight pages, four openings left and right, with space before
the next (pp. 63–70); the second is possible, but organists would have found
more useful other pairs of organ chorales on Luther texts, such as ‘Christ lag
in Todesbanden’ or ‘Ach Gott, vom Himmel sieh darein’. That each takes two
sides of paper (left and right), neither of them filled to capacity, does suggest
some practical reason for their inclusion, however, especially as they were
probably added late, shortly before publication (Butler 1990 pp. 19–20).
The many imaginative and imaginary explanations found for the Duets do
not take into enough account such mundane circumstances.
The kinds of symbolism often proposed seldom if ever match the more
verifiable allusions in Clavierübung III, such as that canons represent divine
Law. The significance of other well-known fours – Gospels, Elements, stages
in the life of Jesus (Birk 1976) – is far less certain, nor can it ever be more than
a guess that the Duets mark the dualities of Bread/Wine and Saviour/Grace
at Communion (Chailley 1974 p. 267). More likely is that the composer, in
adding to a big collection, was offering musical skill ‘ad majorem gloriam
dei’, and alluding to music’s history. As to the first, any Golden Section in
No. 3 could be alluding to its traditional name, the Divine Proportion. As
to the second, E F G A are the top four notes in the tetrachord-scale of
classical theory as illustrated in Walther’s Lexicon of 1732 p. 601, and
they may appear more often than is now recognized. The first four keys
in Krieger’s Clavier-Übung of 1699 are E minor, F major, G major and
A minor.
Techniques
Techniques included are regular fugue, double fugue, ABA fugue, fugue
with bass; strict invertibility, canon, inversion, stretto; motivic derivation
(scales, triads, chromatics); and different modes (major/minor), rhythms
and metres (duple, triple, compound). It is possible that as with the Italian
Concerto in Clavierübung II, there is a common pulse: 3/8 quaver = 2/4
crotchet = 12/8 dotted crotchet = 2/2 minim.
Invertibility is part of the composer’s interest in melodic counterpoint
free of species formulae, and such details as the falling augmented octaves
in No. 1 and the effect of an augmented triad in No. 2 – both of them
quite logical in situ – contrast with the stile antico of the movement follow-
ing in Clavierübung III, i.e. the fugue in E. They may also hint that here
the composer was investigating ‘the consequences of equal temperament’
(Eck 1981 pp. 21–5), though that could equally be said of any advanced
music of the 1730s. Throughout, as with the Canonic Variations, the strict-
ness of counterpoint results in a musical language recognized immediately
by the listener to be both distinct – like nothing else – and ‘other-worldly’.
532 BWV 802–803
BWV 802 Duetto I (Clavierübung III)

double fugue, 73 bars; all material invertible; chromatic, scale-like,
angular and syncopated elements; one-bar phrases; tonic–tonic–
relative–dominant–tonic
The subject modulates in order to be answered (‘rare’ – Souchay 1927),

but only in b. 6, which is not strictly part of the subject. The row of par-
allel major thirds produced by the stretto scales has an effect comparable
to the ‘augmented triad’ in No. 2: the false relation between minor and
major sixths comes from the two forms of the melodic minor scale. The
strange, consonant harmonizing in two chromatic parts resembles that of
the (contemporary?) A minor Prelude WTC2.
The two subjects are very different. The lower has a paraphrased descen-
ding chromatic fourth (cf. E minor Fugue BWV 548), the upper a more
modern chromatic appoggiatura. The top subject’s extension in b. 6 is devel-
oped in a highly original way in bb. 12ff. and 40ff., and equally unusual are
the final tonic entries (bb. 61, 66): the first in the course of the overlapping
scales, the second in place of the original b. 6. Their invertibility is exact,
unlike pairs of final tonic entries elsewhere (e.g. Two-part Invention in E),
and the counterpoint is new, free of old formulae.
BWV 803 Duetto II (Clavierübung III)

ABA fugue, 149 bars (37, 75, 37):
A regular exposition; most material invertible; triadic and broken
chords; little syncopation; various phrase-lengths
B canonic second theme; stretto treatment of first theme; inversus of
it (74); shifting tonality, minor, chromatic; syncopations, slurs
A2 returns in one hand across a running line in another, as it does in another

ABA Fugue, the (contemporary?) BWV 548.
B differs in its mode, stretti, chromatics and phrase-lengths, a ‘learned’
or ‘hard’ section contrasting with the lighter A and built up of thirty-one,
thirteen and thirty-one bars. Even its slurs across the beat are distinc-
tive, clarifying the canon (especially from b. 52) and contrasting with A,
where such slurs would be inappropriate. Very like Bach’s late canons is the
right-hand of bb. 57–60 and its many returns. From the shifting tonality
of the second theme a new, chromatic countersubject emerges (bb. 69–72),
casting new light on A and propelling it to an inversion in the minor. The
middle section is extensive, twice as long as the outer sections and produc-
ing symmetries: its opening and closing stretti begin with the right hand,
533 BWV 803–804
its middle stretto with the left, and the inverted minor version of A appears
almost exactly halfway (b. 74).
Quite the most startling effect is the augmented triad progression in one
particular stretto, bb. 62–3, 90–1 and 105–6, giving a new aspect to a theme
that was brightly triadic in its original version. Throughout, there is a tension
peculiar to this piece between its quasi-naiveties (four-bar opening theme,
triads, broken chords, the last bar) and the calculated counterpoint. The
listener senses a theoretical idea being played out, especially in the ‘strain’
of B’s counterpoint: its canonic subject and stretti, the inversion per giusti
intervalli of both theme and countersubject at b. 74, the steady ‘denials’ of
basic four- and eight-bar phrases. A’s sunny major triads contrast with B’s
dark augmented triads.
Perhaps one also senses the proportions: the 1 : 2 of sections A : B and
the five- and eight-bar division of the middle section (5, 8, 13: see Glossary
under Fibonacci). When Siegele 1992 observes, however, that the sections
subdividing the 149 bars (37 + 31 + 13 + 31 + 37, all prime numbers) can
be juggled so as to produce 81 bars (37 + 31 + 13) or 75 (31 + 13 + 31)
which ‘equate’ with the words ‘mediator’ and ‘passio’; or that the notes on
which A’s fugal imitation appear (f , c, c , F) produce the cross figure; or that
the main theme appears fourteen times (counting a stretto as once), thus
making B A C H; or that in the number-alphabet JSB is anyway the same as
‘SDG’, ‘soli deo gloria’ – then one passes from the actual-perceptual to the
conjectural-conceptual.
BWV 804 Duetto III (Clavierübung III)

invention fugue, 39 bars; non-chromatic, non-modulatory subject;
rolling sequential figures; staccato elements; non-invertible bass
Although in form, melody and harmony (see the tonic across bb. 1–3)
Duetto III is the simplest of the four, it is no more conventional than the
others. The detached bass is not a theme, is not found in the right hand, and
does not lead to much in the way of development. The fugal answer plus
countersubject in bb. 3–4 could conceivably be accompanied by a third part,
though this is no trio manqué. A piece similar to it is the B minor Two-part
Invention, which in turn resembles some Three-part Sinfonias more closely
than it does the other Two-part Inventions.
As b. 3 compared with b. 9 shows, one version of the invertible counter-
point is more successful than another; but where the piece is in its element
is in developing the rolling semiquavers of 12/8 time in a clear diatonic
G major, taking in the regular fugal entries and a seemingly natural stretto.
The middle section, bb. 12–21, brings minor more to the fore, by way of
534 BWV 804–805
contrast. Broken-chord sequences will resemble those elsewhere in Bach,

such as the G major Sonata BWV 530, first movement or even (from b. 13)
the Third Brandenburg Concerto. Like Nos. 4 and 2, No. 3 has an early
tonal answer (b. 3); but here, a tonal answer is not strictly necessary and
results in a much-altered subject. The rather tiresome amount of tonic over
bb. 28–31 arises because the long tonic subject is duplicated in stretto.
BWV 805 Duetto IV (Clavierübung III)

fugue, 108 bars ( = 33 × 22 ); subject with minim head and quaver tail;
chromatic element throughout; almost total invertibility
Like those of Nos. 2 and 3, No. 4’s answer is tonal, which is not obviously
necessary. But between them, the Duets are also surveying the ways to treat
dux and comes, as if anticipating, even participating in, the new didactic
interests signalled by Mizler’s translation of Fux and by Marpurg’s book on
fugue. Two modern touches given to this interest in strict counterpoint are
the slow harmonic rhythm of the long subject’s second half and the notes’
penchant to sink a semitone, in subject (b. 4) and episode (b. 18). The
result is a movement whose visage says ‘constant invertible counterpoint’
but which reminds one at times of a bourrée in Clavierübung II.
Although sometimes called a regular fugue, No. 4 explores several
unusual intricacies in its two-, four- and eight-bar phrases. The lines are
constantly exchanged; all entries are either tonic or dominant; a stretto
at bb. 31–5 returns inverted at bb. 94–7; and various symmetries can
be discerned:
1 A1 (lh), A1 (rh), A2 (lh), to dominant

17 B1 (rh), B2 (lh), to relative
26 C1 (rh), C2 (lh), to tonic
31 (end) A1 (stretto); then b. 33 A1, b. 35 A2, b. 41 A1, b. 43 A2, to
dominant
49 B2 (rh), B1 (lh), to relative (or G major)
58 C2 (rh), C1 (lh), relative (or G major) to dominant
64 A3 (rh), A4 (lh), to tonic
70 A1 (rh), b. 72 A2 (lh), section identical to bb. 9–17
78 B2 (rh), B1 (lh), to relative; 79–85 as 18–24 (partial
inversion)
86 A4 (rh), A3 (lh), to subdominant
93 link
94 (end) A1 contracted (for stretto), b. 96 A1 (lh), b. 98 A2 (rh)
104 coda
535 BWV 805
The subject appears on tonic and dominant only, countered by a chro-

maticism that takes the forms of flattened notes in the melody, including
Neapolitan sixth (b. 105), and raised semitones in the passages that modu-
late. Thus old and new elements are contrasted. Only the bourrée-like coda
is not inverted, and even that, with some modification, could be made to
invert.
BWV 943 Prelude in C major
Copy: P 804 (W. N. Mey, by 1727? – Stinson 1990 p. 33).
Mey probably worked for Kellner, though perhaps later than proposed, and
some of their joint copies are of authenticated works (BWV 535.i, BWV 827).
Because of the (simple) final pedal point the work was published for organ
by Peters in 1852, but neither counterpoint nor harmony have authentic
hallmarks. Nevertheless, its composer was familiar with authentic keyboard
works (WTC, Inventions, possibly suites), to judge especially from bb. 9ff.,
16ff., 30ff., 40ff. (F major WTC1?), the hemiolas at cadences and the dimin-
ished seventh of b. 56. Also, the subject looks curiously like a contraction of
the first 11/4 bars of the Canonic Variations – is that how it originated?
BWV 957 Machs mit mir, Gott, nach deiner Güt

Copy in LM 4708 (‘Neumeister’) with chorale (b. 25); without it in other
late copies.
The TEXT, evidently by J. H. Schein, Leipzig Thomascantor, was published

as a five-verse Hymn for the Dying in 1628, probably based on psalms
(EKG p. 477).
Machs mit mir, Gott, nach deiner Do with me, O God, according to your
Güt, goodness,
hilf mir in meinem Leiden, help me in my suffering.
was ich dich bitt, versag mir nicht, What I ask you do not deny me
wenn meine Seel will scheiden, when my soul wishes to depart,
so nimm sie, Herr, in deine Händ’, so take it, Lord, in your hands.
ist alles gut, wenn gut das End. All is good if the end is good.
Schein’s MELODY (Example 272) appears in the St John Passion and

Cantatas 139 and 156 (triple time); also BWV 377 (D major) and listed
in the Ob.
Although the piece was one of ‘140 variirte Choräle’ in the late Schelble–
Gleichauf MS collection (EB 4322), and is not regularly fugal, Schmieder
1950 called it ‘Fugue in G major’ with reason, since the chorale melody is
not obvious in those first twenty-five bars, and Scholze’s MS calls it ‘fuga’
(Emans 1997 p. 59). The ‘Neumeister’ version is without certain infelicities
[536] found in this version.
537 957–1027a and 1039a
Example 272
Paraphrased, the cantus first line supplies a running theme in the style of
old canzonettas, with octave imitation like an invention, dominant answers
and long derived episode before final entry of b. 23. Perhaps the crotch-
ets of bb. 16–19 or the rh descent in 24–5 allude to the last cantus lines,
though it is the opening paraphrase G A B C D D C B A that seems to
operate at different levels, even perhaps twice (in two phases) in the subject
itself. To a well-articulated harmonization which includes a canon at the
fifth for the last line, the canzonetta serves as an apt prelude, raising the
question whether other fughettas could/should be followed by four-part
chorales.
In theory, the semiquaver scale-runs anticipate those of the Ob’s Hymn
for the Dying, but the passagework is closer to various types of fugue, espe-
cially in G major (e.g. Buxtehude Canzonetta, Pachelbel ‘Magnificat octavi
toni’). For the insistent little figure of bb. 9–10, see a minor version in
BWV 1114, b. 7. The texture at bb. 15, 23f. is not quite conventional, per-
haps suggesting an articulation more common than most notation implies.
The mixolydian flavour of the last bass entry (bb. 23–5) and final cadence
(b. 33) is found in a much later G major fughetta, BWV 679.
BWV 1027a and 1039a Trio in G major

Movement 1: copy in P 804 (J. P Kellner) = BWV 1039.i
Movement 2: in P 288 (2nd half eighteenth century) = BWV 1039.ii
Movement 4: in Lpz MB MS 7 (1730s, for J. N. Mempell?) = BWV 1027a
The movements are versions for two manuals and pedal of the Adagio,
Allegro ma non tanto and Allegro moderato known in the Sonata in G for
Viola da Gamba and Harpsichord BWV 1027 or for Two Flutes and Con-
tinuo BWV 1039. Both BWV 1027 and 1039 may derive from a version
for two violins or flutes, all of them from the Leipzig period (Wolff 1991
p. 234). The organ version of at least the last movement seems to have been
538 BWV 1027a and 1039a–1029.iii
made from one or other sonata (Eppstein 1965), probably the flute version
(Siegele 1975 p. 69). Whether the third movement was ever transcribed for
organ is not known, but as a transcription in Stinson 1992 shows, it could
be played directly on the organ from a chamber score.
Of all the miscellaneous trio movements, BWV 1027a is closest in melody,
harmony and form to the Six Sonatas. As well as a length of phrase and a
melodic charm typical of J. S. Bach’s sonata work of c. 1730, the pedal
goes some way towards variety of motif, and the fugal ritornello introduces
simple episode material very much huius generis. Compare bb. 26ff. with the
first movement of the G major Organ Sonata bb. 37ff. On the other hand,
there is not the same careful interchange of parts, the pedal is not always
well handled (though more organ-like than continuo-like), and some bars
are omitted.
Good reasons, based on sources and details of the transcription, have
been exhaustively set out for concluding that perhaps J. P. Kellner, but in
any case not the composer, was the transcriber (Stinson 1990 pp. 75–99 and
Bartels 2001).
BWV 1029.iii (Trio in G minor)

Only source, London RCM MS 814 (Benjamin Cooke, c. 1770).
Headed ‘Trio a 2 Clav: e Pedal’, second subject ‘cantabile’.
The work is a version of a movement known as the finale of the Sonata in

G minor for Viola da Gamba and Harpsichord BWV 1029, itself possibly
a version of a three-movement concerto for unknown forces (Siegele 1975
p. 9). For its position in ‘BWV 545b’, see above: it cannot be certain that the
composer had nothing to do with the form and key of this organ version
(Emery 1959 p. viii).
The versions differ in figuration, in length (only the gamba version has
two coda bars) and other details; two-part passages imply that the organ trio
is ‘an arrangement of a string trio of some kind’ (Emery 1959 p. iii) and that
its source is not the gamba version BWV 1029 but another (note the chord in
b. 93, and KB pp. 144, 302). The awkward spacing and missed contrapuntal
detail look inauthentic rather than an early attempt at trio writing, though
this has been done with some thought as to its medium. When motion
is required, the alternate-foot motif for pedal is applied indiscriminately,
unlike its carefully placed appearances in the development sections of the
first movement of the G major Sonata BWV 530.
539 BWV 1029.iii–1085
BWV 1079.ii Ricercar à 6 from the Musical Offering

Organ score: copies in P 667 (? J. F. Agricola), P 289 and P 565 (eighteenth
century), etc.
Title in P 667 ‘Ricercata a 6 Voci . . . sonabile sull’Organo col Pedale

obligato’.
The similarities between its alla breve counterpoint and the middle
section of the Fantasia in G major, and the tradition of open-score organ
music, were such as to encourage Bach students to arrange this for organ
with pedals, perhaps with the composer’s tacit approval (EB 6584). Although
the engraved chorales BWV 645–650 and 769 suggest that open score could
be read in various ways to include pedal, in the case of the Art of Fugue the
open score is conceived for manual alone. Adding pedals (and presumably
16 tone) to the bass line of BWV 1079.v distorts the nature of the original
counterpoint, and moreover, Agricola’s simplified bass-line omits quavers
that are part of the motivic counterpoint.

Copies by J. G. Walther (first movement) and by J. C. Bach of Gehren
(LM 4983).
Two staves; for headings, see below.
Walther attributes it to ‘J.S.S.’, LM 4983 to ‘Giovan. Sebastin. Bach’, where

the two movements occupy different pages but ‘obviously belong together’
(see Kobayashi 1983 pp. 170f.). Walther has no second movement (‘Choral’),
but some empty staves follow; each copy omits ornaments given by the
other.
The two pieces treat the chorale melody in distinct ways, somewhat like
BWV 760 and 761, also copied in P 802. In the first movement, lines 1–3 are
ornamented cantus answers to fore-imitations; line 4 is without interlude;
line 5 follows without fore-imitation and dissolves into a quaver paraphrase,
bb. 46–9, all in three parts. The second is no chorale in the sense of the
figured hymn accompanying BWV 695 etc. but a four-part harmonization
with ornamented melody, again ‘dissolving’ towards the end. This is not
uncommon in Böhm (‘Christ lag in Todesbanden’, known to Walther) and
may, somehow, allude to ‘have mercy on us’ or ‘give us peace’.
540 BWV 1085
The harmony of the first movement is more sophisticated than Walther’s,

whose style it otherwise resembles. Questions are whether it is the first of
three settings written or intended as an ‘Agnus dei partita’, like BWV 656a;
whether it might be even a fore-study for BWV 656a (Meyer 1987 p. 14); and
how inauthentic the dissolved final line can be (Meyer 1974 p. 82), when its
cadence is only a major version of those in BWV 721 and 727. For a remark
on the many Es in the last twelve bars, see BWV 703 (p. 443 above), also
in F major.
Chorales now called The Neumeister Collection
BWV 1090–1120
‘Neumeister Collection’ is the name given by NBA, BWV and Wolff 1985
to a set of chorales in Yale University, MS LM 4708, called Das Arnstadter
Orgelbuch when they were first publicized as ‘Bach works’ in 1984 or earlier
(Krumbach 1985), and published in connection with the tercentenary year
1985. Twenty-seven other chorales in the manuscripts LM 4843 (‘Rinck
Collection’) and Leipzig MB MS R 24 were also attributed to J. S. Bach by
Krumbach, but these MSS have no comparable likely lineage: see below,
works listed under ‘[no BWV]’.
Neumeister’s MS
Johann Gottfried Neumeister (1756–1840) was a pupil of G. A. Sorge in
Lobenstein, probably making his collection while assistant organist to a
Herrnhuter congregation in Hessen, from 1790 (Krumbach 1985). The ul-
timate source could well have been an early collection of chorales corre-
sponding to collections of free keyboard pieces made by Johann Christoph
Bach (ABB, Mö MS) or by Johann Michael Bach, father of Maria Barbara
(Krumbach 1985 p. 7). Thirty-eight of eighty-two pieces are attributed to
J. S. Bach, including six known from elsewhere; and since various others
of its attributions agree with other sources (J. M. Bach, Pachelbel, Zachow
and Sorge), there is a prima facie case for thinking it reliable. It is not cer-
tain whether some extra accidentals in Neumeister’s version of BWV 737
mean that he copied reliably (assumed in Wolff 1997) or updated wilfully,
but nothing argues against Neumeister himself being the copyist, for most
chorales compiling or selecting from a single source (Sackmann 1991 p. 168),
and doing so over a brief period in the 1790s.
Probably Neumeister imitated the handwriting of both music and titles
from his source, and preserved its order though not, presumably, any blank
pages it may have contained. There is nothing to say as to whether the
ultimate source contained many more pieces than Neumeister knew, but
it may have been in tablature, partly (Wolff 1992) or entirely (Krumbach
1985). Neumeister’s archaisms such as occasional flats cancelling a sharp
suggest that what he copied was already an early transcription. Sorge’s five
chorales are from his Vorspiele (Nuremberg c. 1750), but whether the two
[541]
542 Neumeister Collection
from the Ob (BWV 601, 639) were already in Neumeister’s main source
cannot be known, nor, therefore, can they suggest anything reliable on the
Ob’s dating. The version of BWV 601 may be earlier than the autograph’s,
but musically BWV 639 has always seemed one of the most modern of the
Ob settings. Both are freer of errors in ‘Neumeister’ than are many other
pieces.
Quite why such a collection was made in the 1790s is an interesting ques-
tion. Perhaps Neumeister was a modest player needing his own selection –
particularly for funerals? – but the period was also one in which many sects
required simple, concise music, as is clear from J. C. Kittel’s Der angehende
praktische Organist (Erfurt, 1801). No doubt most organists would find
Neumeister’s selection more practicable than later Bach collections, which
must often have been copied as ideals or optimum models, and so preserved
more carefully than simpler settings. The presence of Sorge’s pieces suggests
too that Neumeister needed easy music: see above, concerning Clavierübung
III, p. 394. But since not all pieces are easy to play, perhaps his aim was also
to include older styles and the authority they brought.
J. S. Bach
Musically, few if any of these pieces are more obviously authentic than those
in other collections, such as MS R 24 (mentioned above) or P 285 (Anh.II 78
etc.). The very idea that they ‘equal or surpass’ such masterpieces as BWV
697 and 703 (Wolff 1985 p. vii) shows how treacherous subjective responses
can be, and one could just as well claim that few of the settings give any sign
of the enormous gifts of, say, the young Handel. On the other hand, if one
were to postulate how the teenage Bach treated organ-chorales in the light
of what he had learnt from local organists of Thuringia, the result might
well resemble several ‘Neumeister’ settings.
One can make various suggestions about problematic pieces of an ‘early’
nature and apply them to many works listed from BWV 690 onwards. One
is that settings in which the actual chorale is well harmonized in four parts
but its inter-line interludes are less expert were arrangements by someone
else of a Bach ‘chorale nucleus’. Another is vice-versa: J. S. Bach could have
arranged or made versions of works circulating under other composers’
names, particularly Pachelbel (Emans 2001).
A certain roughness, even eccentricity, allows some of the chorales to
stand out from the music of a J. M. Bach or from other doubtful works in the
Bach canon, such as the partita ‘Herr Christ, der einig’ Gottessohn’ Anh.II
77. Such characteristics – changing textures, a tendency to end dramatically,
some spinning-out of phraseology, adventurous or infelicitous harmony or

turns of phrase far from the ‘norm’ – could well be the work of an immature
but promising composer aiming for originality. An attempt to contract
the normal long-winded chorale variations of the day by bringing into a
single movement their changing metres, wandering melody and various
paraphrases would produce unconventional music.
The very variety of pieces attributed to J. S. Bach speaks for their
authenticity, for they have an air of experimentation in an idiom that is
rarely complex. The final chords of BWV 1113 make a simple coda, but what
an original touch! Here surely is a young composer’s reaction against the
anodyne style of Pachelbel and those he influenced in the Bach family, such as
J. M. Bach, whose cadences are uniformly calm and formulaic. The various
surprises are close to the ‘wildness’ Hermann Keller heard in early Bach
and which one can also glimpse in Buttstedt’s works, where, however, ec-
centric originality seldom results in genuine harmonic tension. Some of
the 44 Choraele attributed to J. C. Bach (†1703) are also whimsical at their
cadences, with some unexpected touches at the final pedal point, suggesting
that there was a local tradition for treating this moment with an original
touch.
One negative sign of youthfulness in ‘Neumeister’ is faulty grammar –
parallels in BWV 1109, 1111, 1113 and 1117, empty fifths and doubled thirds
in other chorales – though open fifths are still quite common in the early
Ob settings BWV 604 and 605. Nevertheless, I agree that ‘if authentic they
must be very much earlier’ than the Ob (Dürr 1986), very few of which show
signs of being as early as 1708 (as proposed in Wolff 1991 p. 427). Since so
spectacular a maturing is inconceivable in less than ten years, ‘Neumeister’
would have to be dated to about 1700. It is no argument against a very early
date that from time to time some of them resemble settings by J. G. Walther
(see remarks on BWV 714).
One positive sign of youthfulness is the obvious enthusiasm for a
certain discord described as a ‘diminished seventh chord on the seventh
degree of the scale’ (Stinson 1993 p. 464) but whose real distinction is that
a tonic in the treble is held over the leading-note diminished seventh. A held
tonic in the bass is not so piquant, as one sees by comparing the ends of
BWV 1105 and BWV 1113. Such ‘enthusiasm’ is less known to staid talents,
and the same chord’s return at the end of chorales BWV 721 and 727 cer-
tainly argues for J. S. Bach’s authorship. It seems that the Neapolitan sixth
was not yet so beloved by Bach as it became in the Weimar years.
Clearly, if authentic, ‘Neumeister’ offers many insights into influences
and the maturing of style. The idea that it includes music played by Bach
in Arnstadt is reasonable (Krumbach 1985) but cannot be substantiated.
Assuming the order is original, some of the sequences in it, such as from
BWV 1095 to 1099, suggest that the composer was engaged in a ‘learning
process’, working with imitation, fugato and then canon (Sackmann 1991
p. 170). In accepting that the chorales are all authentic, Breig 1991 points out
how traditional are the correspondences between texts and music, especially
in the use of two particular figurae: the chromatic fourth in BWV 1113, 1114
and 1093 and the ‘sighing motif ’ in BWV 1099. Both motifs anticipate the
Ob and therefore suggest a line of development.
Musical context
A few differences between Neumeister’s copy and the (later?) autograph ver-
sion of two Ob chorales suggest that as his style matured, Bach thought out
the note-patterns and the accented passing-note harmony more carefully.
The original structures and gestures in earlier settings give way in the Ob to a
reliance on harmony so subtle as to convey alone, without structural exper-
imentation, an original and unique Affekt, even with more-or-less standard
cadences.
Generally, the chorales attributed to J. S. Bach have counterpoint more
carefully conceived and sustained, and chorale-melodies more integrated
(same note-values), than those by Johann Michael Bach, whose treatment
of harmony and motifs is closer to that of BWV 1100. In such a setting,
decorations embellish the harmony but do not move it on in new directions.
Details typical of Böhm, including his phraseology, have been recognized in
BWV 1120, which, like some others, gives a vocal impression as if imitating
(older) choral music, including work of Bach-family members (Sackmann
1991 p. 169). But priority is not always certain – are Böhm settings, or the
relevant choral works, always earlier than ‘Neumeister’?
Some of the chorales attributed to J. S. Bach seem to anticipate moments
in the Chorale ‘Partitas’, as if the latter were a stepping-stone on the way
to the Ob, as perhaps they were. But comparisons with first-class music
might not always be appropriate: the idea that some ‘Neumeister’ chorales
borrow elements from characteristic North German chorales – changes of
metre, migration of c.f. – misses the likelihood that broken-up settings of a
simple kind were known throughout Lutheran provinces, easily improvised
or composed by minor organists, polished somewhat by a J. M. Bach but
given their grandest form only in the work of the northern masters, who
are therefore not quite representative. Just as the modest if voluminous
collection of chorales by Daniel Vetter of Leipzig (1709, 1713) is likely to
have prompted the Ob project as much if not more than finer music by
Buxtehude (see p. 235), so local hymn-collecting in Thuringia would have
545 BWV 1090
been the context for Neumeister’s source, as for other collections like the
Plauener Orgelbuch (c. 1708).
Texts, layout
The texts of all chorales uniquely set here, except BWV 1094, 1104, 1106,
and 1119, are in G. C. Schemelli’s Gesang-Buch (Leipzig, 1736), edited in
Zahn 1889–93, in Bighley 1986 (with further bibliography) and in the EKG.
Most of the thirty-eight settings attributed to J. S. Bach are of hymns listed
but not set in the Ob, whose presumed plan makes clear their associations,
which are cited below in brackets. Although there are only three settings of
hymns by Luther, ‘Neumeister’ includes sectarian Lutheran texts otherwise
rare (‘Erhalt uns Herr’, ‘O Herre Gott’), and it is possible that in Thuringia
such hymns as ‘Du Friedenfürst, Herr Jesu Christ’ and ‘Gott ist mein Heil’
were still sung regularly in memory of the Thirty Years War. Whether so
many settings related to death and dying reflect the nature of Neumeister’s
source or his own needs as an assistant organist is unknown.
In the following remarks, no further note is made of the layout of
the chorales in LM 4708 (two staves, rh soprano clef) or of the form of
attribution (‘J. Seb: Bach’, ‘Joh. Sebast. Bach’ and, mostly, ‘J. S. Bach’),
although the ambiguity of names that occurs so often in problematic
sources – Bach/Pach/Pachelbel – might now and then bear on the question
of authorship.
BWV 1090 Wir Christenleut (‘Neumeister’)

While the first half’s three-part harmonization does little with its semi-
quaver figure that could not be matched in chorale-variations by several
composers, or its subsequent triplets and dotted-note fughetta by music
of Zachow or Kuhnau, the overall conception is striking. The triplets and
fughetta are based on repeated lines of the cantus, whose final line rounds off
the setting with a kind of Amen cadence. The shape results from contract-
ing two variations of a conventional chorale-partita towards a half close,∗
followed by an irregular fugue complete with a sequence like that in another
dotted-theme fugue (compare b. 29 and D major Fugue WTC1). In princi-
ple, the setting is a miniature equivalent of BWV 718 or 720.
∗ BA 5181 p. 5 (b. 22) fills in the chord as D minor, but if it is to have a third (uncertain), it should
be f?
546 BWV 1091–1092
BWV 1091 Das alte Jahr vergangen ist (‘Neumeister’)

To begin with the melody is to anticipate an Ob characteristic, indeed a

conspicuous detail in BWV 614 itself. After this opening, the setting is an
unexpected mixture of harmony (via note-patterns in a consistent contra-
puntal style) and fore-imitation of the soprano melody (c.f. in b. 5 bass and
b. 15 tenor), all preserving the modal or mixed tonality of the melody. The
little dactyl figure is rather indiscriminate, but the cadence is drawn out
again as a long Amen, with yet another apparently much-relished E major
chord at the end (compare BWV 4 and 718). The well-conceived four-part
harmony anticipates the Ob, and the interludes between lines are not very
different from those of very much later chorale-settings.
BWV 1092 Herr Gott, nun schleuss den Himmel

auf (‘Neumeister’)
BWV 1092 differs from what seems to have been a particular interest of the
young Bach – the ‘miniature chorale fantasia’ (Stinson 1993 p. 459) – in
having the hymn-melody only in the top part (unlike BWV 739, 718, 720,
767.ix and 770.x) and in being on a yet more modest scale. But in growing
out of a hymn-like start it is more like other ‘Neumeister’ chorales than it
is BWV 739 or 764: less a miniature fantasia than a hymn with interludes,
echoes and coda. Its originality lies in freely using whatever traditional
means of harmonization appear: ‘northern’ octave-echoes (bb. 1, 19–22),
suspirans patterns (b. 10 etc.), broken chords (b. 5), four-part chorales
(bb. 8, 13), long final melisma (in effect repeating the cadence – cf.
BWV 617), all rather disjointed and yet intense.
The form is not much like BWV 739 but its texture and figuration have
a lot in common with it, so that bars 17 or 10ff. might appear in BWV 739
or 720. The sustained, rolling continuity of BWV 617 looks like a reaction
to this setting, and a comparison of the two modal endings is instructive:
both play with the final alternating es and ds, one broken up and dramatic
(with two manuals?) and yet simple, the other continuous and much more
sophisticated in harmony, canonic cantus, rhythm, motion, pedal-bass and
texture.
547 BWV 1093
BWV 1093 Herzliebster Jesu, was hast du verbrochen

(‘Neumeister’)
The TEXT is a Passion hymn by J. Heermann, published in 1630 and
formerly said, in error, to be based on Chapter 7 of the Meditations of
St Augustine:
Herzliebster Jesu, Heart’s dearest Jesus,

was hast du verbrochen, what crime have you committed
dass man ein solch scharf that such a severe
Urtheil hat gesprochen? judgement has been spoken?
Was ist die Schuld? What is the guilt?
In was für Missethaten What kind of trouble
bist du gerathen? have you fallen into?
Fifteen verses move towards the hope inspired by the Passion story.
The MELODY, published by J. Crüger in 1640, is found in both the St

John and the St Matthew Passions (first chorale) and listed in the Ob:
Example 273.
Example 273
After a potentially solemn opening, the most striking musical gesture is

the conversion of the last line of the hymn-melody to a descending chro-
matic fourth momentarily in D minor, thus a passus duriusculus ex tono
primo (see also BWV 614). Although the whole text concerns suffering
there is nothing to compel a subdued Affekt at this point, which is defi-
ant rather than melancholy, so perhaps it alludes to the last line of v. 1.
As often in ‘Neumeister’, the chorale’s close is ‘conceived as a final mu-
sical climax’ (Breig 1991 p. 297) – rhythmically and harmonically, hav-
ing progressed from crotchets to quavers to dactyls and finally continuous
semiquavers. The conventional chromatic fourth is anticipated by equally
conventional rising chromatics (b. 30), and the accompanying figuration’s
548 BWV 1093–1094
semiquavers could belong to the early seventeenth century (Scheidt,

Sweelinck, Bull).
BWV 1094 O Jesu, wie ist dein Gestalt (‘Neumeister’)

The TEXT of the anonymous Jesus-hymn, headed ‘for Lent’, was published
in 1627 with music by Melchior Franck:
O Jesu! wie ist dein Gestalt O Jesus, how is your physical form,
In Marter hoch und mannigfalt, in great and manifold torment,
Mit Wunden tief versehret, injured deeply with wounds,
Von Heiligkeit der Leib so gross, this body so great with holiness
Am Creutz ist ausgespannet bloss, is stretched out naked on the cross,
Hat seinen Glantz verzehret, has exhausted its splendour.
Hertzlich, schmertzlich Heartfelt, painful
Ist dein Liebe, is your love,
Heiss und trübe, ardent and troubled,
Reich von Gaben, rich with gifts,
Die dich an das Holtz erhaben. which raised you on the cross.
The following nine verses contemplate Jesus’ wounds. Listed in the Ob.
The syllabic MELODY had the form in 1627/8 of Example 274.
Example 274
The harmonization of the cantus resembles others in ‘Neumeister’ by being

freely created by an obviously able musician, hinting at the melody from
time to time and moving from quaver figures to dactyls. Only on its repeat
is the first cantus line introduced with fore-imitation (bb. 12ff.), or rather,
obscurely anticipated; these fore-imitations are irregular, not at all on the
Pachelbel model. The falling and finally rising lines of the melody might
be discerned in the counterpoint, and the four parts remain well-wrought,
549 BWV 1094–1096
with moments not so far from the Ob (bb. 21–2). The final cadence is both
phrygian and of a particular kind described above (see p. 543), therefore
doubly dissonant.
BWV 1095 O Lamm Gottes unschuldig (‘Neumeister’)

For TEXT and MELODY, see BWV 617. On the triple time, see also
BWV 1109.
Particular similarities with BWV 1085 (both parts) include key and metre,
harmony replete with Es, and the tendency to increasingly active note-
patterns (more quavers). But the fluent four-part writing is found in neither
part of BWV 1085, and since the melody seems to be slightly different in
line 3, it is unlikely that they belong together as three ‘Agnus dei’ verses on
the analogy of BWV 656a.
On the long seven-bar interlude heard only when the first line is re-
peated, see also BWV 1094. In no obvious way could this interlude have
also originally been an introduction now missing, i.e. one wilfully cut out
by Neumeister: to begin with the first note of the hymn, to go on to in-
clude clear interludes, and to close with a long final all suggest a hymn as
actually sung. Although in theory the accompanying lines, neither derived
from the cantus nor exploiting a clear motif, could be those of many another
composer c. 1700, there is a noticeable sureness in the harmonic direction,
especially in the second half.
BWV 1096 Christe, der du bist Tag und Licht / Wir danken
dir, Herr Jesu Christ (‘Neumeister’)
One copy of the first ten bars and another of the first twenty-nine (this by
J. G. Walther) ascribed to Pachelbel (see Emans 1997 p. 28).
The TEXT of the anonymous Lutheran hymn (Erfurt, 1526) is based on Ps.
91 and the sixth-century hymn ‘Christe, qui lux es et dies’ (as are BWV 1120
and 766).
Christe, der du bist Tag und Licht Christ, you who are day and light,
für dir ist, Herr, verborgen nicht, before you, Lord, is nothing hid,
du väterliches Lichtes Glanz, you, lustre of the Father’s light,
lehr uns den Weg der Wahrheit ganz. teach us fully the path of truth.
550 BWV 1096
The hymn is a prayer for evening (vv. 2–6) with doxology (v. 7). Listed in
the Ob (Evening). Perhaps the second TEXT ‘Wir danken dir’ was not the
one used for BWV 623 but another: ‘Wir danken dir, Herr Jesu Christ, dass
du das Lämmlein worden bist’, a title also listed in the Ob (Evening).
For all three hymn-texts, the MELODY (1535) would require melismas and
to be less simple than in BWV 274. Example 275 is as Bapst’s hymnbook
of 1545.
Example 275
On the strength of Walther’s attribution (a later addition by him?), the first

half of BWV 1096 was included in the Pachelbel edition DTB IV/1 (1903)
as No. 12. Hartmann 1986 supports this attribution for a piece in which
nothing can be identified as proof of authorship; but it could be that the
Walther fragment is simply corrupt (see Wolff Facsimile, 1986 p. 9). The
idea of a fugal fore-imitation before the melody could equally indicate a
Thuringian imitator of Pachelbel such as J. M. or J. S. Bach, as too might
the coupling of different treatments:
1–26 (in Walther) fughetta on line 1, seven entries; episodes; mostly

three parts
28 further dominant answer, as fore-imitation for:
31 the four cantus lines in minims (soprano), with interludes,
long final; mostly four parts
The counterpoint accompanying ‘section 2’ is apparently not derived but

explores patterns familiar elsewhere in ‘Neumeister’. Such bars as 47–9 or
14–15 are conventional and would go on appearing for many decades.
The possibility that one Thuringian organist added a section to a pre-
existing fughetta composed by another, so as to give the hymn’s full cantus,
is plausible, and perhaps something met again in the case of a setting ‘Ach
Herr, mich armen Sünder’ (see Emans 1997 p. 13 and KB). But neither the
musical style nor the sources of either longer piece makes this more than
conjecture. The double versions BWV 631/667 and probably BWV 714 are
not comparable, being presumably work of the same composer.
551 BWV 1097
BWV 1097 Ehre sei dir, Christe, der du leidest

Not (‘Neumeister’)
The TEXT (1560) is based on ‘Laus tibi, Christe qui pateris’ and is the
last verse of ‘O wir armen Sünder’ of H. Bonnus. Listed in the Ob
(Passion).
Ehre sey dir, Christe, Honour to you, Christ,

der du littest Noth you who suffered distress
an dem Stamm des Kreuzes on the trunk of the cross,
für uns den bittern Tod, bitter death for us,
und herrschest mit dem Vater and reigns with the Father
dort in der Ewigkeit, there in eternity,
hilf uns armen Sündern help us poor sinners
bald zu der Seligkeit. soon to blessedness.
Kyrie eleison! Christe eleison! Kyrie eleison!
Both the MELODY (1527) and its treatment in BWV 407 (‘Ach, wir armen
Sünder’) are similar to the cantus sections of BWV 1097, with yet richer
melismas for the ‘Kyrie’ section: Example 276.
Example 276
The rubric ‘Choral’ by the first full line of the hymn – a clear cantus at the
top of four-part counterpoint – raises the question whether such settings
are congregational hymns, with the organ playing prelude and interludes
between the lines. Although the ‘prelude’ (fugal with five entries) and the
‘interludes’ (canonic bb. 25, 31, 37, 47) are contrapuntally worked, with
further imitation or stretto, the chorale’s harmonization line by line is quite
552 BWV 1097–1099
distinct and could even be extracted and played separately. A comparison

with BWV 407 underlines this.
The four-part harmony is straightforward and the exploration of motifs
modest, thus raising a further question whether the composer was J. M.
Bach, as for the next but one chorale in the MS as a whole (‘Jesus Christus,
unser Heiland’). There seems little to choose between BWV 1097 and a
setting of ‘Gott hat das Evangelium’ attributed by Neumeister to J. M. Bach
(see HE 30.650), except perhaps the penultimate bar.
BWV 1098 Wir glauben all an einen Gott (‘Neumeister’)

The final phrase has two particular points of interest, both of which reappear
in Clavierübung III: a cadence with diminished seventh (though the chord
is more dramatic at the end of BWV 1097) and a final clear statement of
part of the cantus, as at the end of BWV 680. The ‘Lesser’ Commandments
chorale BWV 679 also has a diminished seventh cadence, and the ‘Greater’
chorale BWV 680 is a fugue whose final entry is a more complete statement
of the subject than has been heard clearly before. This is the case too in
BWV 1098, where the subject–answer of bb. 16–21 and the first cadence
of bb. 47–51 positively disguise the hymn-line before it is eventually made
quite clear in b. 52. G. A. Sorge too incorporated the final chorale line in the
closing bars of his ‘Wir glauben all’ (see Heinemann 2000 pp. 13–14).
Despite a wavering tonality hard to attribute to c. 1700, the four-part
harmony at the first complete cantus-line (bb. 16–27) is accomplished, as are
the allabreve details (quaver dactyls, suspensions), varied stretti (crotchets
b. 27, in augmentation bb. 52/53), episodes lengthening as the piece
proceeds, and a promising quaver countersubject in the final bars. The fugal
answers beginning in bars 16 and 29 paraphrase a longer stretch of the
chorale-melody. There is a harmonically weak passage in bb. 31–5, but the
plan of a fugal exposition rounded off by c.f. either established a prece-
dent for Bach or was copied from him by another composer. The section
beginning in b. 38 might even recall, though coincidentally, what Parisian
composers had produced in a similar style (F. Couperin, Messe des Paroisses,
first Kyrie fugue).
BWV 1099 Aus tiefer Not schrei ich zu dir (‘Neumeister’)

For the TEXT, see BWV 686. The MELODY is not that found in Cantata
38 or BWV 686 and 687 but a hymn-tune of 1525 (Zahn Nr. 4438), the
553 BWV 1099
only organ-setting by Bach: Example 277. Perhaps this was intended for the
unset Ob chorale (Atonement).
Example 277
The technique, found elsewhere in ‘Neumeister’, of treating the hymn-

melody with changing note-values is taken to great lengths here:
1 crotchets, harmonized (soprano-bass = inversus)

3 minims, canon
10 crotchets, quasi-canon
14 minims, canon, up to five parts
24 dotted crotchets, after fore-imitated paraphrase
31 mixed quavers and crotchets, a form of paraphrase
coda semiquaver figura, no hymn-melody unless the top line
bb. 33–end is a protracted paraphrase of D B C A G F
GAG
The canon produces unusual but logical harmony, such as the 6/4 in the
contraction of b. 17 (deemed a fault in Wolff 1985 p. vii). The final triplets
are presumably a written-in rallentando, as in the second movement of
Cantata 106, and it is possible that ‘adagio’ (an early use, if authentic)
implies ‘freely, at one’s ease’. Certain similarities between this chorale and
settings by J. G. Walther – metre-change, the choral harmonies, the motifs –
seem more marked than usual, suggesting either common precedents or
Walther’s acquaintance with such pieces.
As with BWV 1093, the setting becomes more intense as it proceeds,
moving away from the expected cadence in b. 35, not towards durezza
chromatics but towards a perfect cadence decorated at length by the ‘sighing
motif ’ (cf. BWV 622 and 619). The appoggiaturas do not produce expressive
harmony, and it looks as if they bring Affekt simply by being slurred and
played accordingly. Presumably, the motif transforming the last line of the
melody (bb. 31–4) had some conventional Affekt associated with it, suffi-
ciently telling for the melody to become quite lost. But the disparate treat-
ments of the melodic lines look like miniature sections of a chorale-fantasia
554 BWV 1099–1100
and may be just as ‘objective’ in the use they make of conventional note-
patterns.
BWV 1100 Allein zu dir, Herr Jesu Christ

(‘Neumeister’)
K. Hubert’s TEXT is an early Lutheran catechism hymn, published in 1540.
In Bapst’s hymnbook of 1545 it is headed ‘a general confession’.
Allein zu dir, Herr Jesu Christ, Only to you, Lord Jesus Christ,
mein Hoffnung steht auf Erden, does my hope on earth turn.
ich weiss, dass du mein Tröster bist, I know that you are my comforter,
kein Trost mag mir sonst werden. no other solace may be mine.
Vom Anbeginn ist nichts erkorn, There is no Elect from the beginning,
auf Erden war kein Mensch geborn, there is no earth-born person
der mir aus Nöthen helfen kan, who can help me in my need,
ich ruf dich an, I call to you,
zu dem ich mein vertrauen han. to whom I have entrusted myself.
V. 4 (the last) is a doxology. Listed in the Ob (Atonement).
The MELODY reprinted in Bapst’s hymnbook (1545) is simpler by the time

of BWV 1100, 261 and Cantata 33 (C major). Example 278 follows Calvisius,
1597.
Example 278
A canzonetta subject, typically beginning off the beat and paraphrasing the
cantus, introduces a three-part setting that seems little different from some
of Walther’s. The various note-patterns, the harmony with only a few turns,
conventional syncopations, a constant semiquaver motion, might equally
be the work of J. H. Buttstedt or a Pachelbel student. Only the final figura
(bass b. 28) seems possibly derived from the cantus (last line), though not
555 BWV 1100–1102
much ingenuity would have been necessary to unify all the motifs and/or
derive all of them from the chorale.

(‘Neumeister’)
As elsewhere in the collection, a complete four-part fughetta on a derived

theme prefaces the c.f.; the fore-imitations are based on the following chorale
lines, but less patently than usual – e.g. b. 21 is derived from the cantus of
bb. 24–6. Other themes or patterns treated imitatively – chromatics b. 15,
repeated notes b. 25, broken chord b. 36 – are either new or yet more
remotely derived. (Line 3 is diminished to produce a fugue subject, then
further diminished to produce a counter-motif.) The patterns themselves
are not yet original, the dactyls and suspirans being common property.
Accordingly, the music has no obvious Affekt.
But there is a great variety in the note-patterns, a familiar sense of ac-
cumulation towards the end, and a well-wrought four-part harmony that
conveys an impression of the hymn as a distinct whole, including its modal
cadence. Unusual touches (bb. 20–1, 26, 41) make it credibly attributed
to J. S. Bach, though certainly early (see bb. 36, 16, 20a). Pedal is not
required.
BWV 1102 Du Friedefürst, Herr Jesu Christ

(‘Neumeister’)
The TEXT is J. Ebert’s prayer for peace in time of war, published in 1601:
Du Friedefürst, Herr Jesu Christ, You Prince of Peace, Lord Jesus Christ,
wahr Mensch und wahrer Gott, real man and real God.
ein starker Nothhelfer du bist you are a strong helper in time of need
im Leben und im Tod. in life and in death
Drum wir allein, Therefore we alone
im Namen dein, in your name,
zu deinem Vater schreyen. cry out to the Father.
Verse 7, the last, begins:
Erleucht auch unser Sinn und Herz, Enlighten also our mind and heart,
durch den Geist deiner Gnad, through the spirit of your grace,
556 BWV 1102–1103
The MELODY, by B. Gesius, 1601, appeared often in the hymnbooks, and

in three cantatas (67, 116, 143): Example 279. Listed in the Ob (In Time of
War).
Example 279
The setting has two verses, the first a bicinium, the second a further two-
part paraphrase, shorter (briefer interludes), lively (allegro) and more like a
variation. Comparable pairs are found elsewhere in ‘Neumeister’. Particular
Böhmian characteristics are the ritornello elements in the first verse (rec-
ognizable returns of material), particular motifs, the luthé style (bb. 29f.),
excerpts of the chorale-melody, and certain repetition (Sackmann 1991
p. 169). If the first section is what it seems, a conscious imitation of contem-
porary aria-like bicinia, then not the least unusual detail is the ambiguity of
its opening beat – 3/4 or 4/4?
In the second section, all three patterns will be found in other early Bach
work: the luthé (bb. 29f.), the triplets, and b. 37. In particular, the luthé is
like that in BWV 1092, 1106, 1107 and also 735a, except that here there is
no cantus but instead a soprano paraphrase of the melody (D B E D F E E D
etc.). The augmented octave in b. 32 is not a fault (Wolff 1985 p. vii) since
there is a phrase-break between them.
The setting’s version of the cantus is close to that of BWV 67.vii but seems
to have paraphrased it in both verses with more licence than usual, with some
fantasy-like development, various references to the various phrases, and a
new plagal cadence at the end, as if depicting ‘Amen’.
There is nothing to say that a change of manual is likely in b. 29, from
Sesquialtera to a plenum, though the word ‘allegro’ might imply it, since a
change of actual tempo does not appear necessary.
BWV 1103 Erhalt uns, Herr, bei deinem Wort

(‘Neumeister’)
The TEXT’s first three verses by Luther were a Children’s Hymn against the
pope and Muslims (EKG p. 245); also a closing hymn in services (Bighley
1986 pp. 283f.).
557 BWV 1103–1104
Erhalt uns Herr, bey deinem Wort, Sustain us through your word, Lord,
und steur des Pabsts und Türken and check the murder by pope and
Mord, Turk,
die Jesum Christum deinen Sohn who would hurl Jesus Christ your Son
stürzen wollen von seinem Thron. from his throne.
All seven verses appear in Cantata 126 (1725).
The MELODY, published by J. Klug in 1543, appears in Cantatas 6 and 126:

see Example 280 for Bapst’s version (A minor), 1545. Listed in the Ob (‘The
Church’).
Example 280
A three-part fughetta on the first line opens out to four parts for the last line
at the end, as a c.f. (b. 24) and then codetta (last three bars), between which
is the final entry (b. 27). The last line of the cantus is paraphrased and stated
simply over the course of the last seven or eight bars of the soprano part.
(For a fugue written in such a way as this to imply snatches of the melody,
see BWV 1098 as well as BWV 681.) While the unimaginative harmony
matches J. M. Bach’s, and the formulaic suspensions, the easy quaver lines
and the conservative fugal technique match Pachelbel’s, the conception is
once again unusual.
Pedal is not required, though presumably optional for the final tonic
entry. Perhaps the final minor chord is a copyist’s error?
BWV 1104 Wenn dich Unglück tut greifen an

(‘Neumeister’)
The TEXT is an anonymous hymn of 1609, for times of trouble:
Wenn dich Unglück thut greiffen an, When misfortune has you in its grip
Und Unfall will sein Willen han, and mishap has its way of you,
So ruf zu Gott im Glauben fest, then call to God in firm faith,
In keiner Noth er dich verlässt. in no distress will he leave you.
558 BWV 1104–1105
The MELODY is one of several tunes for the hymn, Example 281 being as
published by Vulpius, 1609 and Schein, 1627. Listed in Ob (Persecution).
Example 281
The preface to the c.f. of b. 3 is unusual – perhaps some bars are lost, and
these twenty-six bars are the final cantus statement of a longer setting whose
key was clearer than it is in bb. 1–2 here. Comparable questions arise in
connection with the chorales BWV 957, 714, 1096 and the anonymous
three-part ‘Ich ruf zu dir’ in ‘Neumeister’: has this or its source preserved
only part of a setting? Although this ‘Ich ruf’ has also been proposed as the
work of Bach (Stinson 1993 p. 456), the four-part BWV 1104 is harmonically
richer and more inventive.
A harmonization which gradually admits figuration in lower voices
seems to have been familiar to J. M. Bach (cf. ‘Warum betrübst du mich,
mein Herz’), and to him probably via Pachelbel. The note-patterns emerg-
ing in bb. 4–5 and 7 were common property, and yet the form the first of
these takes in bb. 20–1 produces something very like the opening of the early
autograph chorale BWV 764.
BWV 1105 Jesu, meine Freude (‘Neumeister’)

This stopping-and-starting chorale is rhetorically effective in performance

and more than a theoretical working out of ideas, which include a cantus
that wanders (soprano, alto, bass, as in ‘Jesu, meine Freude’ BWV 713),
imitative figurae (bb. 3, 5), echoes, tmeses (breaks), chorus-like beginnings
and ends, and Ob-like phrases (b. 11). The general effect is cut up but
undeniably striking, as again is the penultimate chord, for which see BWV
1113, 1097 and various harpsichord allemandes (suites BWV 806–808,
811).
It is difficult to agree that the last five bars are simply an out-of-
proportion ‘final cadence’ (Wolff 1992 pp. 246f.): rather, they draw out and
paraphrase the final line of the hymn in a manner comparable to bb. 7–10’s
treatment of the corresponding line. As the movement becomes increasingly
559 BWV 1105–1106
drawn out in this way, however, one can certainly picture a young com-
poser striving to make something new, even fervidly responding to the
text.
Since pedal seems indispensable for one or perhaps two of the cantus
lines, presumably it is optionally used throughout. There is also a clear
potential for a second manual, in the echoes of bb. 7–8 and 14–16. Neither
is indicated in the source, any more than for BWV 742 (q.v.).
BWV 1106 Gott ist mein Heil, mein Hilf und

Trost (‘Neumeister’)
The TEXT was published 1592 as an anonymous ‘song of solace in the
cross’:
Gott ist mein Heil, mein Hülff und God is my salvation, my help and
Trost solace
mein Hoffnung und mein Vertrauen, my hope and my trust,
Der mich mit sein’m Blut hat erlöst, Who has redeemed me with his blood,
auf ihn will ich fest bauen. on him will I firmly build.
Denn ich hab all mein Zuversicht For I have placed all my reliance
zum lieben Gott gericht, on the dear God,
denn er verlest die Seinen nicht. for he does not abandon his own.
Four verses share the last line.
The MELODY was published by B. Gesius in 1605 and is found in major

hymnbooks of the seventeenth century (Example 282). Listed in the Ob
(Persecution).
Example 282
Patterns of semiquavers run in the soprano or middle voice or bass as the

case may be, and create no more than three parts until the full cadence – a
detail typical of this early repertory. But there is also an ingenious play with
the chorale:
560 BWV 1106–1107
1, 4, 7, 9 tenor line 1, tenor line 2, soprano line 1, soprano line 2

11, end soprano, line 3 broken up with unison echoes
15, end line 4 between soprano and alto, then octave echo
18, end soprano line 5, answered in 22, tenor in 27
The melody is handled inventively, and at moments (e.g. bb. 7–9) gives an
impression of stretto. A second manual is possible for the echoes, though
unspecified.
The prevailing feel of G-mixolydian means that the two strong dominants
are important (in b. 16 and again the answer in b. 22); otherwise the harmony
is not yet very sophisticated (e.g. last six bars). The figuration of bb. 7–8 is
not distant from moments in the chorale BWV 735a, although the fluency
of BWV 1106 might be thought superior. Both the echoes and final cadence
suggest a vivid, imaginative response to setting a chorale.
BWV 1107 Jesu, meines Lebens Leben (‘Neumeister’)

The TEXT by E. C. Homburg (1659) is a typical Jesus-song – hence is
not listed in the Ob – see p. 235 above? Eight verses describe the Saviour’s
suffering, each ending with thanks:
Jesu, meines Lebens Leben, Jesu, life of my life,

Jesu meines Todes Tod, Jesu, death of my death,
der du dich für mich gegeben, you who gave yourself for me,
in die tiefste Seelennoth, into the deepest distress of soul,
in das äusserste Verderben, in the most extreme undoing,
nur dass ich nicht möchte sterben. merely that I might not die.
Tausend, tausendmal sey dir, A thousand, thousand times,
liebster Jesu, dank dafür. dearest Jesu, thanks be given you for it.
The MELODY (W. Wessnitzer, 1661) is one of ten known for this text.
BWV 1107 corresponds only to lines 1, 2, (3, 4) and 6 of Zahn No. 6795:
Example 283.
Example 283
561 BWV 1107–1108
The repeat of lines 1 and 2 is unusual: a three- or four-part harmoniza-

tion of the melody without interlude is broken off for a kind of variation
before two further phrases in the soprano are harmonized. The result is a
kind of miniature fantasia, followed in turn by a 12/8 continuation for the
other lines, first in a fugal paraphrase, then the two last in the bass, then a
final Amen pedal point under wild and freely composed triplets. The final
result is unique, although theoretically the opening section is much like
some of the settings by J. M. Bach. The sectionality recalls other chorales
(BWV 1090, 1099, 1111, 1115, 1118), as do the triplets (BWV 717, 1099,
1110). Throughout, the note-patterns are conventional, unlike the turn to
G minor near the end.
BWV 1108 Als Jesus Christus in der Nacht (‘Neumeister’)

The TEXT of J. Heermann’s Communion hymn was published in 1636:
Als Jesus Christus in der Nacht, When Jesus Christ, in the night
darin er ward verrathen, that he was betrayed,
auf unser Heil war ganz bedacht, was wholly intent on our salvation,
dasselb uns zu erstatten. in order to retrieve it for us,
v. 2:
Da nahm er in die Hand das Brodt, as he took bread in his hand,
und brachs mit seinen Fingern . . . and broke it with his fingers . . .
Eight verses draw on the four eucharist texts (Matthew 26, Mark 14, Luke
22 and I Corinthians 11), followed by doxology.
The MELODY, published by J. Crüger in 1649, is virtually identical to that

of BWV 265, a later simplified form (Zahn No. 258): Example 284.
Example 284
The two verses offer an interesting view of attempts to control an idiom

(four-part with quavers, three with semiquavers) and give it free rein. The
562 BWV 1108–1109
quaver motif is not merely imitative but often extended, and the harmonic
progressions are sure, though the conception is not much of an advance on
J. M. Bach. The alternation of crotchet and minim c.f. in the ‘Variatio’ is
unusual and suggests a reaction to Pachelbel’s uneventful continuity, as does
the accumulation of semiquavers in the last two bars. While the sequence
of bb. 38ff. is not one of J. S. Bach’s (?) best, the two sections’ very different
cadences are original.
Perhaps Neumeister or his source selected two variations from a longer
work, hence the unusual heading ‘Variatio’ for the second setting.
BWV 1109 Ach Gott, tu dich erbarmen (‘Neumeister’)

M. Müntzer’s dramatic TEXT was published in c. 1550:
Ach Gott! thu dich erbarmen, Oh God, have mercy
durch Christum deinen Sohn, through Christ your Son,
Über reich und über armen, on rich and on poor,
hilf, dass wir busse thun, help us to repent
und sich ein jedr erkennen thut, and for each to recognize himself.
ich fürcht, Gott hab gebundn ein Ruth, I fear that God has bound a whip
er will uns damit strafen, with which he will punish us,
den Hirten mit den Schafen, the shepherd with the sheep,
es wird ihm Keinr entlaufen. no one will escape him.
Listed in the Ob (‘Day of Judgement’).
The MELODY was published by S. Calvisius in 1597 (Example 285).

Example 285
How far BWV 1109’s composer was responsible for simplifying the melody
is not known, or whether he put it into triple time to give it ‘weight and
563 BWV 1109–1110
staidness’ (Breig 1991 p. 295), for it was already so in Scheidt’s Görlitzer

Tabulaturbuch, 1650. Marking the chorale lines Choral, as elsewhere in the
volume, might be to clarify structure after long fore-imitations, but it could
also indicate the hymn-lines as they were sung by the congregation, i.e.
usually with simpler harmony accompanying their unison melody.
The four-part harmony of these lines is mostly well done, especially
for the last line. ‘Blemishes’ in the form of parallels in bars 48 and 70–1
have made it difficult to accept as the work of young Bach (Dürr 1986
p. 310), though neither parallel is ordinary or difficult to avoid. Similarly,
evidence that the copy is garbled is not obvious from the passage beginning
b. 7 (suggested in Wolff 1985 p. vii) since in fact the four-part harmonization
of what is anyway not a straightforwardly diatonic tune is rather more
successful than the opening two-part. One can hear, especially in the Choral
sections, an attempt to control a somewhat intractable melody, partly with
a crotchet line typical of 3/2 chorales (BWV 1095).
BWV 1110 O Herre Gott, dein göttlich Wort

(‘Neumeister’)
The TEXT, published in 1531, was written with reference to Luther’s sermon
‘On the freedom of a Christian’ (EKG pp. 211f.):
O Herre Gott, dein göttlichs Wort O Lord God, your divine word
ist lang verdunkelt blieben, long remained obscured
bis durch dein Gnad uns ist gesagt, until by your grace we were told
was Paulus hat geschrieben, what Paul has written,
und andere Apostel mehr, as too other apostles,
aus deinem göttlichen Munde; from your divine mouth,
des dank ich dir for this I thank you
mit Fleiss, dass wir assiduously, that we
erlebet habn die Stunde. have witnessed the hour.
Listed in the Ob as the last of the de tempore hymns (for the Saints).
The MELODY is as for BWV 757 and Cantata 184. Its triple-time form
may be owed to Praetorius (see Zahn 5690); Bapst’s 1545 version as in
Example 286.
As in other settings such as BWV 1117, subsequent figuration suggests that it

begins more slowly than appears at first. Only in three parts, like comparable
chorale-settings of Pachelbel or J. G. Walther, the setting shows neither an
advanced working with motifs nor unusual harmony, except for the final
564 BWV 1110–1111
Example 286
cadence. The opening paraphrase keeps the chorale’s notes on main beats
(an ‘early’ sign) but moves on to new themes after the first line is repeated.
Since this line appears three times in the chorale, there are three different
harmonizations for it (see bb. 2, 21 and 42), with motifs shaped accordingly.
A ‘defiant’ performance (see the text) could be appropriate.
BWV 1111 Nun lasset uns den Leib begraben

(‘Neumeister’)
The TEXT is a seven-verse hymn by M. Weisse, sung at the place of burial
and known also in a version with alternate verses, i.e. as responses.
Nun lasst uns den Leib begraben, Now let us bury the body,
daran wir keinen Zweifel haben, of this we have no doubt but
er wird am jüngsten Tag aufstehn, that it will rise up on Judgement Day
und unverwesslich herfürgehn. and go forth incorruptible.
Luther’s added v. 8 is a prayer to Christ. Listed in the Ob (Burial).
The MELODY in a hymnbook of 1544 (Zahn 352) is as Example 287.
Example 287
The form of BWV 1111 is puzzling. Line 1 is set fugally (bb. 1–20, complete
with coda); line 2 has a new countersubject (bb. 20–30); line 3 is a new fugue
in triple time and therefore with change of Affekt (from b. 30). The fugues
565 BWV 1111–1112
gradually become shorter, so that line 4 (A B A G E G A G) is probably to be

heard in the soprano from b. 35 and again in the last few bars. The gauche
and uninteresting final cadence might suggest that something is missing.
The music is rarely distinguished by much melodic flair and could be
the work of J. M. Bach, though details both formal (the unpredictability)
and figural (the common dactyls, then running semiquavers) conform with
other settings attributed to J. S. Bach.
BWV 1112 Christus, der ist mein Leben (‘Neumeister’)

The TEXT, an anonymous hymn published before 1608, is a prayer of the
dying:
Christus der ist mein Leben, Christ is my life,

sterben ist mein Gewinn, dying is my gain,
dem thu ich mich ergeben, to him I make over myself,
mit Freud fahr ich dahin. with joy I depart.
For line 4, see the Nunc dimittis (BWV 616). Listed in the Ob.
The MELODY is by M. Vulpius 1609, found in BWV 281 (similar to BWV

1112) and Cantata 95 (cf. BWV 282, in triple time): Example 288 is as
in 1666.
Example 288
A plain fore-imitation in two parts is followed by the first chorale-line and

then the other lines, each of which is harmonized in four parts. The brief
interludes are not derived but worked from a simple figuration typical of
Pachelbel or J. M. Bach, including the omnipresent dactyls, except that
halfway through, the c.f. is anticipated (end b. 16). The formula is simple,
but the well-handled harmonizations, the sequences developing in unex-
pected ways, and an interesting final cadence, all suggest a composer working
beyond formulae.
566 BWV 1113–1114
BWV 1113 Ich hab mein Sach Gott heimgestellt

(‘Neumeister’)
The echoes faintly recall those in J. T. Krebs’s copy of the variations BWV 767,
last movement, while the chordal phrases are like those in other ‘Neumeister’
chorales, such as J. C. Bach’s ‘An wasserflüssen Babylon’ or, better, J. M.
Bach’s ‘Allein Gott in der Höh’’, registered Rückpositiv/Oberwerk in a copy by
J. G. Walther (HE 30.650 p. 36). According to the f and p marks, BWV 1113’s
hymn-lines and introductory passages are played on the main manual, only
the echoes on the second.
As it stands, BWV 1113 offers the clearest layout in ‘Neumeister’ for
a chorale as performed in church: the sung lines are harmonized simply,
being prefaced and followed by organ interludes appropriate to the chorale’s
phrases but not actually derived from them. The effect is another kind
of fantasia, like BWV 714. Unspecified ‘compositional weaknesses’ in the
chorale caused Breig 1991 to regard it as ‘very early’ work, but given that
its plan mirrors one way of performing chorales, there are many felicitous
touches – in the harmony of each cantus line, in the chromatic fourth bb.
29–30 (to the text’s ‘ohn Widerstreben’?), and in the way that the usual
build-up towards the end (cf. BWV 1099) stops for the sake of simple final
statements, including a climactic codetta. The concept of the whole has
character and originality: a fresh way of setting a chorale.
BWV 1114 Herr Jesu Christ, du höchstes

Gut (‘Neumeister’)
The TEXT is B. Ringwaldt’s hymn ‘for the forgiveness of sins’, 1588:
O Jesu Christ! du höchstes Gut, O Jesu Christ, you highest good,

du Brunnquell aller Gnaden, you source of all grace,
sieh doch, wie ich in meinem Muth see how in my spirit I
mit Schmerzen bin beladen, am weighed down with pains
und in mir hab der Pfeile viel, and have many arrows in me,
die im Gewissen ohne Ziel which oppress me without purpose,
mich armen Sünder drücken. poor sinner, in my conscience.
Eight more verses hope for forgiveness.
The MELODY was published in 1593 and appears in Cantatas 131, 166
(minim c.f., G minor, C minor), 48, 113 and 168 (final chorale, G minor,
567 BWV 1114–1115
B minor, B minor); also BWV 334, see Example 289 (BWV 48.vii). Listed in
the Ob (Atonement).
Example 289
The two halves offer another example of the ‘paired setting’ (cf. BWV 1102,
1107 and 1116), again so written as to give a sense of progression and
build-up. The first half has fugal fore-imitation either derived from (b. 1)
or independent of (b. 10) a melody which is decorated and, probably, to be
played ‘un poc’ adagio’. The second half has a new and more brilliant fugal
imitation, far less restricted in compass, against which the melody is more
like a c.f.
There are many points of interest. The theme, which in the first half is
surely so written to have its own manual, is at first almost as decorated as
those in the albums for Wilhelm Friedemann and Anna Magdalena. It stops
and starts like an aria, its notes (F F E F G A G F) only gradually emerging,
which they do against a rising chromatic fourth in the bass of b. 8. It cannot
be as lively as conventional c-time chorale-fugues, though whether the sec-
ond half is faster is unclear. One could view that half ’s angular figuration
as alluding in some way to the sharp arrows of conscience, though musi-
cally their interest lies in being so different from the first half ’s. The cantus
sings out (bb. 15, end, b. 20) before being decorated once again for the last
line, paraphrased with tmeses and therefore conveying a sense of hesitation.
Like the harmonized versions, BWV 1114 ends in the major, as if the
Affekt of F minor (sin) moves at the last to the major (atonement). That
F minor is gloomy is suggested by comparing any bar in BWV 1114 with
any bar in BWV 1115.
BWV 1115 Herzlich lieb hab ich dich, O Herr

(‘Neumeister’)
The TEXT of M. Schalling’s three-verse hymn was published in 1569:
568 BWV 1115
Herzlich lieb hab ich dich, O Herr, I love you with all my heart, O Lord,
Ich bitt, du wollst seyn von mir nicht I ask that you will not be far from me
fern
mit deiner Hülf und Gnaden, with your help and grace.
die ganze Welt erfreut mich nicht, The whole world delights me not,
nach Himmel und Erden frag ich nicht I do not ask after heaven and earth
wenn ich dich nur kan haben. if only I can have you.
Und wenn mir gleich mein Herz And were my heart even to break
zubricht,
so bist du doch mein Zuversicht, you are still my confidence,
mein Heil und meines Herzens Trost, my salvation and my heart’s solace,
der mich durch sein Blut hat erlöst. who has redeemed me through his
blood.
Herr Jesu Christ, mein Gott und Herr, Lord Jesu Christ, my God and Lord,
in Schanden lass mich nimmermehr. let me never be shamed.
The MELODY was published by B. Schmid in 1577. It appears in

Cantatas 19 (c.f., in G), 149, 174 and the St John Passion (final chorales,
C, D and E); also BWV 340 (C), as Example 290. Listed in the Ob (Dying,
Burial).
Example 290
The setting exemplifies ideas current c. 1700. Beginning like a bicinium, the
first line uses a bass figure similar to BWV 601’s (Ob), also found elsewhere in
‘Neumeister’. This is a broken-chord figure familiar to J. M. Bach, Pachelbel,
Böhm and others, merely decorating, without propelling, an unadventurous
harmony. (Bar 11 is also less than successful.) For such reasons it is hard
to see BWV 1115 as being more than superficially similar to Buxtehude’s
569 BWV 1115–1116
Praeludium BuxWV 137 (Krumbach 1985, Wolff 1986), since just as likely
an influence is non-organ music like the new continuo aria-basses, as seen
too in Walther’s chorale ‘Herzlich lieb hab ich’. Perhaps BWV 601 entered
the collection, in ‘Neumeister’ or its source, because its figuration is similar
on paper to BWV 1115’s.
Although different sections explore different techniques, they do not
make it feasible to change registrations. The opening two parts become
three, suggesting that the composer did not yet know the organ bicinia that
imitated solo arias; the gigue would suit a petit plein jeu but is short; manual-
changes are possible in bb. 21, 28, 32 but would break the line. Perhaps such
breaking was more acceptable than now assumed, as is suggested too by
‘Wie schön leuchtet’ BWV 739, a larger version of the fantasia-miscellany
here. In other respects, the second half shows the weakness and strength
of one particular ‘Neumeister’ type: although each section is short, yet this
helps produce a typical touch of wild originality (e.g. the obsessive thirds
over bb. 28–32).
While the junction at b. 21 can be rhetorically handled by the performer,
b. 18 is hard to accept as it stands – better evidence than bb. 16f. that the copy
is garbled? The gigue section might be recalling Buxtehude’s ‘Gelobet seist
du’ but only briefly. More striking, perhaps, is that the in-turning, equally
‘obsessive’ soprano line at the end has something of the positive outlook
one hears in the same chorale at the end of the St John Passion. Even the
awkward moment at b. 18 corresponds to the cantus, and indeed the whole
chorale-melody of an unusually long hymn-verse can be found faithfully
observed from beginning to end.
BWV 1116 Was Gott tut, das ist wohlgetan

(‘Neumeister’)
The TEXT of S. Rodigast’s hymn was published in 1675.
Was Gott thut, das ist wohl gethan, What God does, that is done well,
es bleibt gerecht sein Wille, his will remains just,
wie er fängt meine Sachen an, as he originates my affairs,
will ich ihm halten stille; I will hold quietly to him.
er ist mein Gott, der in der Noth He is my God, who in my need
mich wohl weis zu erhalten, knows well how to support me,
drum lass ich ihn nur walten. Therefore I let him alone hold sway.
Six verses begin with the same line. Listed in the Ob (Persecution).
570 BWV 1116–1117
The anon MELODY, 1690 (Example 291), appears in BWV 250 (Wed-
ding), Cantatas 12 and 98 (B), 69a, 99, 144, 75 and 100 (G), the last in six
movements.
Example 291
Though superficially with conventional fore-imitations, the setting is rather

a free-for-all contrapuntal tapestry in which lines of the cantus appear in
a variety of ways. So an octave imitation at the beginning is joined by a
stretto (b. 3) and then after a codetta by the pedal, which is answered with
the rest of the line by the soprano (b. 8); stretto diminutions introduce the
conspicuous repeated notes of the next two lines (bb. 11ff., 16), the second
of which is answered by the soprano; and the last line is paraphrased with
dactyl figures, imitated in their own right. Again, there is more activity
towards the close, which is marked by an original cadence.
BWV 1117 Alle Menschen müssen sterben

(‘Neumeister’)
For TEXT and MELODY, see BWV 643 (alio modo).
Despite the array of charming ideas, from semiquaver accompaniments

through fast decorations to big adagio chords at the end, this setting follows
the cantus as strictly as – perhaps more strictly than – any other ‘Neumeister
Chorale’. The notes of the melody can be discerned, without interludes and
usually on the beat, as follows:
1, 3 lines 1 and 2, tenor

5 line 1, dispersed
7, 9, 11 lines 2, 3 (decorated and dispersed) and 4, soprano
13 lines 5 and 6, soprano, decorated, ending an octave above
17, 19 lines 3 (decorated) and 4, soprano
21 lines 5 and 6, soprano, partly decorated
25 coda incorporating the last line
571 BWV 1117–1118
The setting is therefore like a chorale-partita contracted into one movement,

in principle not unlike the longer ‘contracted chorale-fantasias’ BWV 720
and 739.
Yet despite the undoubted charm and originality of BWV 1117, the
texture, largely in two or three parts, is anything but intricate. Nor do the
simple motifs have a clear Affekt: whether the new figure at b. 17 is alluding
to the ‘withering away’ in text-line 5, and the scales of b. 21 are alluding to
the ‘great splendour’ of text-line 7, is uncertain. But it is hard to see them
as meaningless figuration.
The final three bars resemble the coda to the finale of the Suite in B BWV
821, also marked ‘adagio’ and effective in performance; they are also not
unlike the end of another early work in B major, the Capriccio BWV 992.
The parallel fifths of b. 25 are puzzling, since they could be easily avoided:
nothing is obviously miscopied, and the progression is acceptable as it is,
thanks to the doubled chords.
BWV 1118 Werde munter, mein Gemüte (‘Neumeister’)

The TEXT of J. Rist’s hymn ‘seeking the protection of the Almighty’ (1642)
is:
Werde munter, mein Gemüthe, Be cheerful, my soul,

und ihr Sinnen, geht herfür, and you senses, go forth
dass ihr preiset Gottes Güte. so that you praise God’s goodness
die er hat gethan an mir, which he has done me;
da er mich den ganzen Tag since the whole day,
für so mancher Sorg und Plag against so many cares and afflictions,
hat erhalten und beschützet, he has sustained and protected me
dass mich Satan nicht beschmützet. so that Satan does not defile me.
Twelve verses, like BWV 1096 and 1120, are related to Luther’s Evening
Blessing.
The MELODY was published by J. Schop in 1642 (Example 292) and used
in Cantatas 55, 146, 147 (triple time, G), 154 and the St Matthew Passion,
also BWV 359 and 360 (other texts). Listed in the Ob (Evening).
This is another more or less continuous setting of a chorale-melody, with

different note-patterns every other line, a keyboard piece offering little op-
portunity either for introducing pedal or for accompanying a congregation.
The opening four bars prepare for the Ob, even in the fermata of b. 2; the
572 BWV 1118–1119
Example 292
next four, evidently an exhortation to be cheerful, shape triplets like those

in Handel’s early gigues except now with a tenor c.f. Each line is set as if for
the sake of the Affekte implied in v. 1, and the second half is drawn out and
developed as follows:
9, 11 lines 5 and 6, tenor, soprano

13, 15 lines 1 and 2, dominant, alto
9–17 repeated
18, 20 line 7 and 8, tonic, soprano
BA 5181 has several doubtful readings: in b. 8, the triplets surely fade away,
requiring no ties at the end; and b. 15 appears to be garbled. The problem
is due to too far a modulation, which the following passage (b. 17) does not
entirely solve. Hence, perhaps, the tonic triads of b. 18?
Two details look forward to other chorales: b. 17, a texture found in
BWV 735 and 665; and a final G-mixolydian cadence, as in BWV 719 and
723.
BWV 1119 Wie nach einer Wasserquelle (‘Neumeister’)

The TEXT is uncertain. The title surely alludes to Ps. 42, on which several
texts were based, including ‘Ach wann werd ich dahin kommen’, v. 2 of
which is:
Wie nach einem Wasser-Brunnen As for a spring of water

Ein Hirsch schreyet mit Begier, a hart cries with desire,
Also auch mit meiner Zungen so too with my tongue
Lechs’ ich, O Herr Gott, zu dir. I pant with thirst for you, O Lord God.
The title is the first line of other texts too long for the present melody (Bighley
1986 p. 328). Listed also in the Ob (‘Christian Church’).
573 BWV 1119–1120
The MELODY’s origin is not known and perhaps had no single or fixed
form; one or other version was used by six further texts in eight Leipzig
cantatas. A simpler version appears in Witt’s hymnbook of 1715 for the text
‘Ach wann werd ich dahin kommen’ (Zahn 1294): Example 293.
Example 293
The trochaic phrase-ends of the cantus are unusual, but for a remark on
triple-time versions of melodies, see BWV 1109. Imaginative harmony and
a very competent handling of motif suggest that this setting belongs to
the same composer as most of the others in ‘Neumeister’, though it is in
three parts only. The main motif (a dactyl figure) is familiar from chorales
composed in Weimar, not only the Ob but settings by J. G. Walther, D minor
having particular associations with it. The coda and final cadence must be
amongst the most original passages in any three-part chorale, though there
are other moments too of adroit figural harmony.
BWV 1120 Christ, der du bist der helle Tag

(‘Neumeister’)
For TEXT and MELODY, see BWV 766. Listed in Ob (Evening).
If some bars look like J. G. Walther’s handling of a cantus, the disjointed

references to the hymn-melody have an effective rhetoric and pleasing melos
suggesting Georg Böhm. Amongst the elements in common between this
setting and such other early works as BWV 533, 700 and 741 are the chords
and a sectionality which might be corresponding to the text’s Affekt, as too
perhaps was BWV 1117.
Though not much more than a playing-through of the chorale, BWV
1120 in fact surveys many different treatments: chords and potential echoes
of a kind found in choral music, with fore-imitation (lines 1 and 2), brief
fughetta (line 3 in b.12), pedal c.f. (lines 4 and 5), and a striking pedal-
point cadence. The idea of a ‘miniature survey’ alone matches others in
‘Neumeister’, so that one could see the fantasias BWV 720 or 739 as expanded
574 BWV 1120
versions of BWV 1115, 1117 or 1120 rather than as contracted versions of

long northern fantasias.
The piece seems to require pedal (a c.f. reed) and a second manual for
echoes, if not perhaps as many as suggested in BA 5181 (but is b. 1 all piano?).
Neither pedal nor echo is cued, but as with the longer fantasias BWV 720
etc., both are feasible.
Further works, in part of uncertain origin
BWV 1121 Fantasia in C minor

Autograph MS: in ABB (c. 1706? Kobayashi 1989 p. 17; December 1708? Hill
1987; possibly as late as 1710: Schulze 1984 p. 50).
Tablature (only complete extant example of J. S. Bach); headed ‘Fantasia ex

C dis’ and ‘adagio’.
Although its being a tablature of J. S. Bach does not prove this to be a
work of his, he would surely have attributed it otherwise in ABB had this
been the case (Kilian 1983 p. 166). Seiffert’s edition in 1925 already guessed
the composer to be ‘one of the great masters’ (meaning Buxtehude?) and
Keller thought its counterpoint close enough to the Canzona in D minor to
date it to Bach’s Weimar period (1948 p. 71). Nothing makes it specifically
for organ, although the Prelude in C minor BWV 921 that precedes it in the
ABB (copied by J. C. Bach, also anon, but last three bars autograph) is less
contrapuntal and more exclusively for harpsichord. These two pieces, being
notated differently, do not obviously belong together, yet as a pair they do
anticipate the Fantasia and Imitation in B minor BWV 563, also in ABB.
Since the bass of the C minor Fantasia enters as a contrapuntal voice, pedal
might be supposed, although as in the Canzona, everything can be played
by hands.
Some points made about BWV 563 can also apply here, including com-
parison with Böhm. But the character of melody at bb. 20–2 is close to that
of the Sonata in D major BWV 963.i, just as the imitation at bb. 37ff. also
recalls BWV 563. The less than completely imitative counterpoint is unusual
and at first glance not more than competent, hardly in an obvious ‘Italian
contrapuntal style’ despite recent references to it as such. But with BWV 563,
the chorale BWV 724 (also in tablature in ABB), the G minor Praeludium
BWV 535a and the Passacaglia, the Fantasia makes its own contribution to
the composer’s survey of contrapuntal working, as if now ‘combining vari-
ous musical ideas in one genial movement’ (Kilian 1983). Certain thematic
resemblances to the much later unfinished Fugue in C minor BWV 562.ii
are not entirely superficial.
The chromatic elements noted by Kellner are interestingly different from
those in the Canzona, which has a conventional chromatic fourth. Even if the
composer had much to learn about where to place perfect cadences – here
still rather arbitrary – there are nevertheless moments of harmonic interest:
[575]
576 BWV 1121–Anh.II 46
the interrupted cadence at b. 43 (recalling Böhm’s G minor Praeludium?), a

certain harmonic drive at bb. 46f., a final cadence using a Bach ‘fingerprint’
(leading note diminished seventh plus treble tonic, see BWV 727). One
would have expected the final chord to be major, as in the preceding Prelude
in C minor.
Anh.I 200 O Traurigkeit, O Herzeleid

(fragment, Orgelbüchlein)
Two staves in P 283, headed ‘molt’adagio’.
See above, p. 238. The fragment, probably added in the late Leipzig period,
suggests a different technique from the rest of the album: all parts begin in
one operation, the melody is soon modified as the Affekt of key and tempo
inevitably takes over. The passing-notes added to the cantus recall those
instances in cantatas in which Bach ‘composed to some extent in half-note
values [minims], “dissolving” the individual lines subsequently’ (Marshall
1989 p. 120).
Anh.II 42 Fugue in F major

Copy in P 817 (c. 1800).
The subject has a post-Bach ring to it such as one also finds in BWV 580 or
in the subject said by A. F. C. Kollmann to have been given his brother by
C. P. E. Bach at an audition in Hamburg (quoted in An Essay on Practical
Musical Composition, 1799, p. 35). Perhaps it owes something to Handel’s
Six Fugues (1735), as many keyboard fugues of the mid-eighteenth century
appear to do: a general italianate contrapuntal style, somewhat as in Johann
Christian Bach’s Fugue on B A C H also in P 817, competently handled but
without any obvious sign of J. S. Bach’s working.
Anh.II 46 Trio in C minor

Copy in SBB Mus. ms. 12011 (J. L. Krebs) and later.
Headed by ‘Trio a 2 Clav: e Ped: di J. T. K.’.
Although P 833 ascribes the movement to J. S. Bach and more recent au-
thors to J. L. Krebs (Keller 1948 p. 58), the title in BB 12011 seems to point
577 Anh.II 46–Anh.II 55
conclusively to Johann Tobias Krebs (Tittel 1962). The supposed resem-

blance to the subject of BWV 585.i, the plaintive quality of the melody and
the curious avoidance of strong cadences are all characteristic of trio-writing
in the wake of the Six Sonatas.
Anh.II 49 Ein feste Burg ist unser Gott

Anh.II 50 Erhalt uns, Herr, bei deinem Wort
Copies in P 285, P 1119 (via Breitkopf) and later.
An attribution to Bach comes only from the MSS’s general titles for a col-
lection of miscellaneous chorales. Though the technique in neither setting
would be out of place for Walther, and any infelicities in the second may be
due only to faulty sources (Emans KB), nothing in the music points clearly
to authorship.
Anh.II 52 Freu dich sehr, O meine Seele

Late sources only (P 285).
As often with P 285, this has little in common with authenticated works,
especially here the weak opening. But ‘on second look’ there are a charm,
drive and keyboard idiom such as to remind one of Bach, and perhaps the
first three bars were added by someone (Emans KB)? The questions thus
raised are typical of the doubtful works.
Anh.II 54, 68 Helft mir Gottes Güte preisen / Wer nur den
lieben Gott lässt walten
Copy in P 285 and later MS.
For another B minor version of the melody, see BWV 613. Despite moments
of more than competence in both settings, P 285 is not authority enough to
establish Bach as composer.
Anh.II 55 Herr Christ, der einig Gottes Sohn

Copy by J. T. Krebs (anon).
Two staves; headed ‘à 2 Clav. è ped.’.

Though anonymous in P 801, where it follows the partita BWV Anh. II 77,
this has been claimed as a Bach work (EB 6589) on the grounds that the
incomplete source Lpz MB MS 1 was to have included a chorale on this
text, ascribed to J. S. Bach and coupled with BWV 765, which is also in
P 801.
The concept is unusual:
cantus in the tenor of a trio, with an obbligato melody derived from it

and accommodating it throughout; all above a continuo bass
Given its kind of melody, the result is more like some of the Schübler Chorales
(BWV 645, 649) than a trio such as ‘Allein Gott’ BWV 664a. The melodious,
angular line may well suggest J. L. Krebs, not least as the musical flow appears
to be interrupted for the sake of the repeat in b. 9, as in Krebs’s ‘Von Gott
will ich nicht lassen’.
Anh.II 59 Jesu, meine Freude

Late copies only (including P 285).
The setting’s free-ranging paraphrase and (partly) detached chords result in

a setting scarcely more obviously inauthentic than BWV 1105. The question
must remain, therefore, whether P 285 is less reliable as a Bach source than
‘Neumeister’.
Anh.II 61 O Mensch, bewein dein Sünde gross

Late copies (e.g. P 285).
A variant setting is transmitted under Pachelbel’s name, less flowing and

seemingly earlier (Emans KB). P 285 and a further MS by Scholz are not
good evidence for Bach’s authorship, though a ‘modernization’ of an earlier
work, made for musical reasons by a later organist or copyist, is plausible
whether by J. S Bach or someone else. (Bach would have been exceptional
amongst organists of the time if he never altered other composers’ music.)
Alternatively, the ‘Pachelbel version’ could as well be a simplification of a
setting that survived in later copies, a detached quaver passage (with dubious
staccato dots) replacing harder-to-play two-part counterpoint.
Anh.II 70 Wir glauben all’ an einen Gott (Fuga)

See BWV 581.
Anh.II 73 Ich ruf’ zu dir, Herr Jesu Christ

Copies in Lpz MB MS R 25 (second half eighteenth century) and later in
P 1149.
Headed in MS R 25 ‘dell Sig. C. P. E. Bach’ (‘C. P. E.’ added? Emans 1997

p. 52).
Like BWV 683a, this seems to be the result of a not ungifted later-eighteenth-
century composer adding introductory bars and other material to an organ
chorale conceivably regarded as too short for church use at that period, in
this case BWV 639 (a setting also found in ‘Neumeister’).
Anh.II 77 Herr Christ, der einig Gottes Sohn

Copies by J. T. Krebs (c. 1714?), and in P 826 (derived from P 801?), both
anon.
Spitta thought these variations to be the work of Bach, like BWV 766 and
767 (I p. 207); but Krebs, who wrote it in P 801 between Anh.II 65 and
Anh.II 55, gave no composer. His heading for the first variation, ‘la prima
alla maniera’, suggests he meant to write ‘alla maniera italiana’ as for the
Aria variata BWV 989 in the ABB, which became ‘all’manual – italiana’ in
Krebs’s copy of it in P 801 (Zietz 1969 p. 54).
The final variation (v. 7) suggests why BWV Anh.II 77 is worth con-
sideration: its texture is much like the opening of the fourth movement of
Cantata 106 (1707?), implying perhaps either that J. S. Bach composed it or
that Krebs, a pupil not long after, copied the idea. Similarly, although the
work throughout lacks melodic drive or harmonic tension, other variations
recall note-patterns in the ‘Neumeister Collection’, which had no (room for)
chorale-variations as such. But the opening of v. 1 shares ideas with BWV
1115: for the parallel thirds in the left hand see bb. 28ff., for the right hand
paraphrase in semiquavers see bb. 33ff. or moments in another chorale,
BWV 1102.
One is reminded of various moments in early works: Cantata 106, the
Variations BWV 989, and the Capriccio BWV 992. The paraphrasing in vv. 2
580 Anh. II 77 – Erhalt uns, Herr
(main notes on the beat) and 7, or alto treatments in vv. 4 and 5, certainly
suggest a competent composer with new ideas and an ability to take them
far, while v. 6 anticipates the rich, allemande-like Ob chorale in F major,
BWV 632. Whether by Krebs, Walther or early Bach, the work offers a model
for details often found elsewhere.
Anh.II 90 Fugue in C major

Copy in Mylau Tabulaturbuch (c. 1700?, lost), P 804 (anon, c. 1727?) and
later.
In P 804 ‘di Bach’ has been added after ‘Fugue in C’ (Stinson 1990 p. 136),
thus weakening its authority. Although the entries are tonic and dominant,
as in early authenticated fugues, there is little in the way of episode, not
much likelihood that pedal was required except for the familiar final point
d’orgue, and nothing to associate J. S. Bach with such a canzonetta subject,
its harmony or its treatment. Nothing suggests that ‘Bach’ means C. P. E.
Bach. The subject’s length, repeated notes and static harmony are more in
the style of the Erfurt composers J. H. Buttstedt and A. N. Vetter or their
pupils (‘watered down Böhm’ – Keller 1937 p. 64), as is not the case with the
semiquaver subjects BWV 575 or BWV 914. The blatantly triadic harmony,
the joie de vivre, the continuity, and the Neumeister-like originality of one
cadence (b. 33), are not unattractive.
[no BWV] Erhalt uns, Herr, bei deinem Wort

Copy in LM 4843 (C. H. Rinck, c. 1800).
This fantasia of 117 bars (Emans 1997 No. 13, Wollny and Zehnder No. 12)
gains its length by treating the cantus line by line in four ways: fugal fore-
imitation, an embellished c.f., a c.f. in soprano, finally c.f. in bass, all in
an idiomatic keyboard style believable as the work of the composer of the
‘Neumeister Chorales’. (Do the thirty-one bars of BWV 1103, part fugal part
harmonization, imply that Neumeister chose only shorter settings than this
one?) Each line/section ends with a spacious cadence bar, having explored
conventional patterns with something more than competence. Rather than
suggesting the influence of the ‘North German chorale fantasia’ (accord-
ing to Emans KB), the conception could be that of an imaginative young
581 Aus tiefer Not, etc.
organist creating a spacious setting while systematically surveying current

local techniques.
[no BWV] Freu dich sehr, O meine Seele

Late copies, as Anh.II 52.
This is apparently a manualiter version of a setting with pedal which in eight

MSS is attributed to J. L. Krebs (Emans KB). Because its motivic conception
imitates the Ob’s, because some word-painting can be discerned (‘joy’ in
the demisemiquaver motif), and because the pedal version appears to have
been made from it, claims for Bach’s authorship of the ‘original manual
version’ are sometimes made. But the motif is treated to little more than
empty repetition and the harmony is immature (b. 7), suggesting only a
student’s pale imitation of the master.
[no BWV] Aus tiefer Not schrei ich zu dir

Da Jesu an dem Kreuze stund
Erhalt uns Herr bei deinem Wort
Ich ruf’ zu dir, Herr Jesu Christ
Komm, heiliger Geist, erfülle die Herzen
Vom Gott will ich nicht lassen / Auf meinen lieben Gott
Herr Christ, der einig Gottes Sohn
Only copy: Yale LM 4843 (by or via J. C. H. Rinck).
In the wake of the publication of ‘Neumeister Chorales’, also formerly part

of Rinck’s library, some of these have been offered as possible Bach works
(see Wollny and Zehnder 1998 and Emans KB), not yet developing specific
keyboard idioms and thus earlier than BWV 766, which is also found in part
in the MS. The MS leaves many works anon but may imply J. S. Bach from
the volume-title (Krumbach 1985). Rinck’s teacher was Bach’s pupil Kittel.
The lines of the hymns are introduced and interspersed with short
chordal passages not based on the cantus, and in Nos. 4, 6 and 7 there
is a relatively lengthy coda. In No. 5, inter-line interludes anticipate those
of BWV 715 but are much simpler, and as with other doubtful works in this
MS, quite without harmonic tension. If they were actual accompaniments,
582 Was mein Gott will
the organist would have needed to signal the congregation’s entries, perhaps
with rallentando. If they are indeed works of the teenage Bach, he had not
developed harmony as much as he had a certain imaginative touch with
conventional motifs at the codas of such pieces, and the counterpoint (e.g.
in No. 1) has poor moments. It cannot be ruled out that Bach enlarged or
helped enlarge a Pachelbel setting in the same MS of ‘Ach Herr, mich armen
Sünder’, or that e.g. Nos. 1 and 2 are Arnstadt settings, or that No. 3 suggests
acquaintance with North German chorale-fantasias (Krumbach 1985 and
above, p. 580). But there seems seldom if ever sign of the individuality that
often appears in the ‘Neumeister Collection’.
[no BWV] Was mein Gott will, das g’scheh allzeit

Copy in LM 4708 (anon).
A four-part treatment, with undeveloped motifs and simple interludes be-

tween lines, the movement is sufficiently like other harmonizations of the
chorale by J. S. Bach in Cantata 144 and the St Matthew Passion as ‘not to
exclude him from authorship’ (Krumbach 1985, 2 p. 10). Since compara-
ble old-fashioned settings in ‘Neumeister’ (BWV 1109, 1116, 1119) are also
attributed to J. S. Bach, the case is strong, although there is ‘no decisive
evidence to tip the balance’ (Wolff 1986 p. 11) between J. S. or J. M. Bach.
Calendar
Phrases in quotation marks are taken from the Obituary or from contem-
porary documents, all to be found in Dok I–III.
1685–1700 (i) Eisenach. Possibly taught organ by Johann

Christoph Bach (first cousin once removed),
organist of the Georgenkirche.
(ii) Ohrdruf. Possibly taught by elder brother
Johann Christoph Bach (a pupil of Pachelbel).
Mar. 1700 Lüneburg, chorister of St Michaeliskirche; possibly
organ lessons there or in the Nikolaikirche or
Johanniskirche (where G. Böhm organist). While in
Lüneburg, said to have travelled ‘occasionally’ to
Hamburg and to have heard Reinken there.
c. 1700 Perhaps learnt ‘French taste’ in Lüneburg
(orchestra of the Duke of Braunschweig-Lüneburg,
from Celle).
1702–3 Applied for post of organist at the Jakobikirche,
Sangerhausen.
1703 Few months at Weimar. May have studied Italian
string music there. Commission to test organ in the
New Church, Arnstadt (Bonifatiuskirche, organ by
F. Wender).
9 Aug. 1703 to 29 June 1707 Organist at Bonifatiuskirche, Arnstadt. Criticized
for long interludes in chorales and for too bold and
chromatic harmonization. At Arnstadt ‘revealed the
first fruits of his industriousness in the art of
organ-playing and composition’.
1705–6 Winter journey to hear Buxtehude, i.e. probably the
special Abend-musiken performances (Dec. 1705).
June 1707 to 25 June 1708 Organist at Divi Blasii, Mühlhausen (organ by
F. Wender, new proposal by Bach, February 1708;
? tested by him, Reformation Day 1709?).
July 1708 to Dec.1717 Organist to the court of Weimar, a position
enabling Bach to perform ‘well-ordered church
music’; ‘here too he wrote most of his organ works’.
13 Dec. 1713 Elected organist at Liebfrauenkirche, Halle;
position finally not taken up.
2 Mar. 1714 Promoted at Weimar to Konzertmeister.
1 May 1714 With Kuhnau and C. F. Rolle, reported on new
[583] organ of the Liebfrauenkirche, Halle, by C. Cuncius.
584 Calendar
Aug. 1717 On payroll of Prince Leopold of Anhalt-Cöthen as

Capellmeister; allowed to leave Weimar after 2
December 1717.
Sep.? 1717 Visit to Dresden.
Extempore (harpsichord?) competition with Louis
Marchand called off.
17 Dec. 1717 Reported on the rebuilt organ of the Paulinerkirche
(University Church), Leipzig, by J. Scheibe.
1717–23 Capellmeister to the court of Cöthen.
Oct.–Nov. 1720 Played to Reinken at the Katharinenkirche,
Hamburg; 23 Nov. 1720, leaves Hamburg after
candidature at the Jakobikirche (organ by Arp
Schnitger), withdraws his name.
15 May 1723 First payment of salary at the Thomaskirche,
Leipzig.
1 June, ‘entered upon the cantorate’ at the Thomas
School.
2 Nov. 1723 Inaugurates small new organ at Störmthal
(by Z. Hildebrandt, extant).
25 June 1724 New organ at Johanniskirche, Gera, tested and
dedicated by the ‘famous Cantor and Capellmeister
Bach’ (organ by J. G. Finke).
Sep. 1725 Plays organ of Sophienkirche, Dresden
(by G. Silbermann).
14 Sep. 1731 Plays organ of Sophienkirche, Dresden, where eldest
son (W. F. Bach) appointed organist 23 June 1733.
Sep. 1732 Said to have examined rebuilt organ of the
Martinikirche, Kassel (by H. Scherer, rebuilt by
N. Becker).
1 Dec. 1736 Plays large new organ in the Frauenkirche, Dresden
(by Silbermann), for two hours in the presence of
‘many persons of rank’.
Michaelmas 1739 Clavierübung III published by the author.
1739 Visits the large new organ of Altenburg
Schlosskapelle (by G. H. Trost).
26 Sep. 1746 With Silbermann, examines the large new organ of
the Wenzelskirche, Naumburg (by Z. Hildebrandt,
partly extant).
1746/47? Six Chorales published by J. G. Schübler (Zella).
c. 1748 Canonic Variations on Vom Himmel hoch published
by B. Schmid (Nuremberg).
8 May 1747 Plays organ of the Heiligegeistkirche, Potsdam
(by J. J. Wagner).
28 July 1750 Dies in Leipzig, ‘mourned by all true connoisseurs
of music’.
1751 Art of Fugue published.
Glossary
acciaccatura: a ‘delicate and admirable secret’ Example 294

(F. Geminiani, A Treatise of Good Taste, London
1749 p. 4) enabling a keyboard player to enrich
the harmony by adding (short) notes outside the
chord itself. See Example 294, also BWV 543,
572.
Affekt: a term used especially by modern writers
to denote the mood of a piece of music or of
feelings aroused in the listener, particularly as
invoked by such details as key and figurae
(see below). The original term affetto, used by
composers of the new expressive style c. 1600, is
found some decades earlier in Zarlino, who also obviously playable in this manner may not have
spoke of effetti, ‘effects’. For Bach’s pupil J. G. been intended as pedal parts (e.g. BWV 531.ii
Ziegler’s remark on being taught to play b. 36 as compared with b. 23), however later
‘according to the Affekt of the words’ – though understood.
in what respects, or to what level of alternatim: ‘alternately’ – specifically in the
sophistication, is unclear – see p. 233. present repertory, performing a hymn, canticle
alio modo: ‘in another manner’, usually of a or psalm so that choir or congregation sings a
chorale, either a setting of a certain melody verse, then the organ alone either replaces or
different from the previous one in a collection prefaces the next with its own ‘verse’ based on
or a setting of a different melody with the same the melody. Whether or how this was a practice
title. in Arnstadt (small town church), Weimar (court
alla breve: strictly, a term from late medieval chapel – likelier?) or Leipzig and Halle (city
theory denoting music in which the beat is a churches), and whether or how any settings
breve, not (as usual) a semibreve – in effect, (including chorale-variations) are related to it, is
quick duple time (2/2 not 4/4). By c. 1700 the still uncertain.
term had come to denote a certain style of ambitus: the total compass or range of pitches
counterpoint, vocal or instrumental, in a vocal or instrumental part, as distinct from
characterized by a lively minim pulse, certain the prevailing tessitura, q.v.
rhythms (quaver dactyls, crotchet lines) and anapaest: see figura corta
suspensions, all derived ultimately from appoggiatura: a dissonant note ‘leaning’ down
late-sixteenth-century counterpoint celebrated on, or up to, the following consonant note.
by Fux in Gradus ad Parnassum (Vienna, 1725). While writers since c. 1675 have demonstrated
alla stretta: in ‘narrow’ imitation, as found this ornament in melodies, its significance in
especially at climaxes in music of this period. In harmony and harmonic evolution is more
modern usage, stretto denotes a theme imitated crucial. The frenchified melodic appoggiaturas
before it is finished (so overlapping, but the of BWV 562 already lead to modern four- and
imitation not necessarily exact), or a close but five-part harmonies, while the appoggiatura
incomplete imitation of one and the same theme chords of BWV 546.i or 552.i are a chief feature
(BWV 529.iii), a pseudo-canon (e.g. subject at of the main themes. ‘Appoggiatura harmony’
end of BWV 541.ii), or canonic imitation of may be an appropriate term for such harmonies
brief motifs (e.g. BWV 538.ii). At the close of as those of BWV 562, as distinct from the
BWV 769.v, ‘alla stretta’ signals four overlapping ‘accented-passing-note harmony’ developed in
themes and at least three derived motifs. the Ob.
alternate-foot pedalling: a phrase denoting the augmentatio: ‘enlargement’ of a theme’s or
technique (old-fashioned by 1750) in which the motif’s note-values, usually to twice as long
feet alternate in a characteristic figuration (diminutio, ‘reduction’: half as long). Per
(e.g. BWV 531). In such early pieces, this augmentationem (diminutionem) signifies
pedalling suggests three particular things: counterpoint, particularly a canon, in which a
(i) the heel is not used; (ii) the bass line is made theme is heard combining with itself in enlarged
[585] specially to suit the technique; (iii) bass lines not (reduced) note-values.
586 Glossary
Example 295
bariolage: a ‘variegated’ sound as between the conspicuous. In late Italian theory (e.g. G. B.
rapidly alternating open and stopped strings on Martini, Esemplare, Bologna 1774–5), the
a violin, usually a lively, scrabbling figuration of livelier sections of a fugue subject were called
an obsessive character around an implied the andamento as distinct from both the attacco
soprano pedal-point. See notes to BWV 550 and (a subsidiary motif for imitation) and the
582. soggetto (‘subject’, denoting the ‘head’ of the
basso continuo: see continuo theme as distinct from the livelier coda or
bicinium: a two-part piece going back to sung ‘tail’).
repertories of the early Lutherans; during the circle of fifths: a common sequence, gradually
seventeenth century it became specifically an regularized during the seventeenth century, and
organ-chorale in which a melody derived from based on harmonies in a succession of
the chorale is accompanied by a lively bass, dominant–tonic or tonic–dominant
ostinato-like and marked by certain figures relationships: see Example 295. Of the two, the
(broken chords, octave drops, rests). falling sequence (in which each pair of chords is
brisé: a (modern?) term particularly denoting dominant–tonic) is the more common, but the
the seventeenth-century manner of ‘breaking’ rising tends to be more interesting.
chords on harpsichord or organ in imitation of colla parte: in ensemble music such as the
the gentle arpeggiation (plus non-harmonic chorale of a cantata, a directive for a performer
notes) practised by lutenists and guitarists to to play the part of another; e.g. violin I, oboe I
sustain their harmony. See BWV 599 or the and cornett play what the trebles sing. See
sectional cadence points in the D minor Toccata BWV 725.
BWV 538, which supply there an ‘antique’ continuo or basso continuo: a bass line giving a
flavour. straightforward accompanying bass foundation
broken chord: a term to denote triadic figures below more imitative or motivic lines played
such as open the A major Prelude BWV 536, in or sung above. In organ music it is usually
distinction to arpeggiated or spread chords, all possible to distinguish between a thematic
of which were called ‘arpeggio’ by Heinichen pedal line and a ‘continuo pedal line’, though
(Der Generalbass, Dresden, 1728) and others. the Six Sonatas and late works (e.g. Canonic
Brustwerk (Bw): the small organ chest (usually Variation BWV 769.v) might hover between
played by its own manual) above the keyboards the two.
and below the Hauptwerk, ‘in the breast’ of the cornet de récit: the kind of rh melody played
organ. Although with a penetrating solo stop or with the Cornet from c , sometimes on its own
two, it was always the smallest department when manual, in a movement from such French livres
present, and often used for continuo, being so d’orgue as those by Grigny, Boyvin and Du
near the performers in the gallery. Walther’s Mage.
Lexicon (‘Brust’) pointed out that this chest the cross motif: an angular group of four notes
could be placed above the Hauptwerk. of which the first and last are around the same
canon sine pausa: ‘canon without a pause’, i.e. a pitch, the second and third respectively higher
subject worked simultaneously at two or more and lower (or lower and higher), so that two
levels, resulting in parallel thirds or sixths. lines drawn between 1 and 4, 2 and 3 would
cantus: used here as synonym for ‘the chorale or cross halfway.
hymn melody’. dactyl: see figura corta
cantus firmus: a pre-existing melody on which a diminutio: see augmentatio
contrapuntal movement is based by creating divisio or division: in such music as this,
new lines around it. Although in the organ particularly a bass-line producing a lively
music of Bach there is normally a clear variation (running bass, etc.) from the main
distinction between pieces including a cantus notes of the harmony it is supporting.
firmus and those not, newly composed themes dorian: a minor key so called will have
in augmentation (e.g. Art of Fugue) can give a prominent major sixths (particularly B in D
similar impression. minor), flattening them only in descending. Its
caput: the ‘head’ of a fugue subject, a term final chord is quite likely to be – to use diatonic
coined for its opening motif, which is often terms – a ‘half close on the dominant’.
587 Glossary
Example 296
durezza: a term long established in Italy (e.g. figura corta: one of the most important patterns
G. M. Trabaci, Ricercate . . . durezze, ligature, for little notes, described by Walther (Lexicon,
Naples, 1603) to denote keyboard music 1732) as having two forms – dactyl
incorporating drawn-out and slowly resolving (long–short–short) and anapaest
suspensions or ‘ties’ (ligature), in which seconds, (short–short–long). The Ob uses such figures
ninths, diminished or augmented intervals, and systematically and with great variety: anapaest
chromatic ‘hardnesses’ (durezze) play a big part. on the beat (BWV 610), dactyl on the beat
In later organ music the style became less ‘hard’. (BWV 616), dactyl phrased after not on the beat
See however BWV 564, Grave, also Example 296. (BWV 629) etc., each with its own articulation.
échappée: an ‘escaped’ note, a (rising) galant: a term belonging to the eighteenth
dissonance off the beat as if a passing note did century but applied today – apparently with a
not pass but went off in the wrong direction, for more specific meaning – to light, elegant music
pleasing melodic effect. of the middle of that century, reflecting a turn
en ravalement: an eighteenth-century term towards new kinds of public music mostly
which now denotes the ‘extending’ of an organ outside the Church, for which even C. P. E. Bach
compass below C, usually for the pedal reeds of a continued to employ old styles.
French classical organ (e.g. to FF at Du Mage’s at G-mixolydian: a G major with prominent Fs,
St-Quentin, 1697). It is not always clear in organ particularly at the beginning (which therefore
contracts whether it means simply ‘the complete has something of the flavour of C major) and
bass octave’ (i.e. with C, D etc.) or ‘further end (in which a plagal cadence is likely). See
notes below C’ (BB, AA etc.). See BWV 572. BWV 604, 635.
In L’Art de toucher le clavecin (1716), Golden Section: a proportion between two
Couperin also speaks of a treble compass above unequal parts, a greater and lesser, equal to that
c as ravalement, suggesting that a certain between the whole and the greater. The notes
passage be put down an octave if the player does B A C H at b. 16 of the 26-bar chorale BWV 600
not have the notes. Was this a widespread appear at this point of division (1 : 1.62).
custom? grand jeu: a characteristic French registration
en taille: phrase denoting a melody ‘in the waist’ consisting of Prestant 4 , chorus reeds, Cornet
or tenor. In classical French organ music the and Tierce, the last three not used in the plein
most characteristic tenor lh solos are played jeu, which was made up of the Diapason chorus
with a Tierce or Cromorne registration including Mixtures. While theorists from
(N. Lebègue, 1676). Typical tierce en taille Mersenne (c. 1625) to Bedos (c. 1775) were
figuration (scales, ornaments, lively runs rarely unanimous in their description of the
to the cadence etc.) is to be found in BWV 663. grands pleins jeux, they agreed in general on the
Fibonacci series: a series of numbers in which kinds of music for these registrations: fugues on
the next is the sum of the previous two the grand jeu, but massive homophony (often
(so 0.1.1.2.3.5.8.13.21 or 0.3.3.6.9.15.24.39. with durezze e ligature) on the plein jeu,
63.102 . . .). including the petit plein jeu (Positif ). See
figura: a ‘figure’ or distinctive note-pattern in a BWW 572 and 532.
particular shape, catalogued by such theorists as Hauptwerk (Hw): the main chest and manual of
J. G. Walther (Praecepta, 1708) and patently the an organ, as distinct from one or other Positiv.
basis of many works of J. S. Bach. Even so Many eighteenth-century German sources
familiar a figure as the eight-note pattern of continued to call it the Oberwerk: this is the
Example 30, woven into many movements meaning in BWV 720, 596 and other
(see C minor Fugues BWV 537 and 546) and registrations of J. S. Bach and/or his copyists. In
almost a signature of J. S. Bach, had a name of North German organ music, this Oberwerk was
its own: the minuta. echo to the smaller Rückpositiv and not vice
588 Glossary
versa, presumably because it was more distant tending constantly towards its dominant, e.g. in
from the listeners in the body of the church. In F major with prominent Bs.
Saxony by c. 1710 (G. Silbermann etc.), melisma: a group of notes sung to one syllable,
Oberwerk meant the secondary manual, hence a discrete melodious phrase in
replacing the out-of-fashion Rückpositiv. instrumental music, increasingly often slurred
hemiola: in such music as this, a characteristic during the eighteenth century.
rhythm in triple metre in which two notes of melos: the general melodic character of a
three beats (dotted minims in 6/4) are replaced particular melody or melodious passage.
by three notes of two beats (as if plain minims messanza: a term still found in Walther 1732
in 3/2). (p. 401) for a group of four fast or small notes
heterophony: the simultaneous sounding of two mixing leaps and steps and thus generally both
versions of a melody or, in the present repertory, angular in shape and distinct from other
of the same melody in two different, overlapping note-patterns. See BWV 661, 753.
phases. mixolydian: melody or harmony in a major key
hexachord: a scale-like progression of six tending constantly towards its ‘soft’
diatonic notes, the so-called ‘natural hexachord’ subdominant; see G-mixolydian.
with the semitone between the middle pair motoric subject: a term sometimes used for
(c–d–e–f–g–a). Eighteenth-century (long) fugue-subjects built up of lively
keyboardists, not yet practising octave and semiquaver figures (repeated notes, broken
multi-octave scales, would be the more likely to chords, sequential arpeggios, etc.), developed
perceive hexachord allusions in (e.g.) the from the old canzonetta. See BWV 532 and 575.
opening bar of BWV 769. Such subjects – ‘perhaps introduced by
inversus: a term for a melody or motif whose Weckmann’ (Apel 1967 p. 599) – often took a
original intervals (‘right’, rectus) have been more galant form in chamber music of c. 1725
inverted, producing a new shape, rising where (see also BWV 585).
the rectus fell and falling where the rectus rose. Neapolitan sixth: a progression associated with
The coincidence in English between ‘inversion’ Neapolitan composers from c. 1675 onwards
of intervals in a melody (inversus) and and immediately discernible by its flattened
‘inversion’ or voice-exchange in counterpoint supertonic (a in Example 297). No doubt
(invertible) has led to a confusion not entirely originating in passages developing traditional
obviated by using – as some dictionaries do – passus duriusculus harmonies for Affekt (q.v.), it
the terms ‘melodic inversion’ for the first and appears often in Bach’s earlier keyboard music, a
‘harmonic inversion’ for the second. habit perhaps picked up from Georg Böhm
invertible counterpoint: counterpoint whose (see BWV 564.ii).
intervals allow each of two parts to be or
become the bass, e.g. without producing
unprepared fourths. The voice-exchange this Example 297
allows is usually achieved by one part being
transposed an octave.
lombardic: ‘Lombard rhythm’, a manner of
interpreting a slurred pair of small notes
(often semiquavers in 4/4, quavers in 2/2) as
short–long, not long–short or equal. Familiar in
England long before Bach chorales, the manner
became typical of the new proto-galant Italian
music from the 1720s on, associated by Quantz
1752 p. 309 with composers from ‘Lombardy’, nota cambiata: a ‘changed note’, an unaccented
i.e. Northern Italy including Venice. passing-note passing to the next by leap (saltus)
loure: a slower or heavier kind of gigue, rather than step (passus); conventionally, the
typically in compound time, in which many leap is down by a third.
main beats are dotted and many phrases begin notes inégales: an old and ‘unequal’ way of
on an upbeat to the upbeat. playing equally notated quavers (or semiquavers
luthé: ‘like a lute’ (Walther, Lexicon), a term in a leisurely movement), an articulation or
sometimes used today to denote simple broken playing manière associated with classical French
chords, as at certain moments in BWV 535.i or music. The degree to which certain movements
even BWV 665, less sophisticated than the brisé of J. S. Bach might have been treated in this
technique of BWV 599. manner is controversial, but in any case notes
lydian: melody or harmony in a major key inégales probably arose from a manière or
589 Glossary
way of playing instruments, rather than from a perfidia: a ‘treachery’ or figurative passage, so
desire for jerky rhythms as such. See also called occasionally in seventeenth-century
BWV 539.i. Italian violin music (examples by Torelli) and a
Oberwerk (Ow): see Hauptwerk few later writings (e.g. Brossard’s Dictionaire,
Orgelpunkttokkata: a modern term to denote re-used in Walther’s Lexicon) to denote fast
those organ toccatas associated in particular ostinato-like passage-work, ‘treacherous’ for the
with Central and South German composers player. See also bariolage. The first and third
(probably based ultimately on the toccata sopra i movements of the Pièce d’orgue BWV 572
pedali in Frescobaldi’s Secondo Libro, Rome might represent a new approach to such perfidia
1627), in which the hands weave motifs above a figurations as that in Reinken’s Toccata in G
series of pedal points whose pattern of (see Example 80).
harmonies (tonic–dominant–subdominant– permutation fugue: a fugue in which subject
tonic, etc) are much simpler than Frescobaldi’s. and two or more countersubjects reappear
ostinato: an ‘obstinate’ phrase, usually but not together on all subsequent entries, combined in
necessarily in the bass, which recurs throughout different vertical orders, each part able to
the piece of music concerned, hence a piece so function as the bass line. The form became less
called. See BWV 582. ‘Harmonic ostinato’ strict through the inclusion of episodes. See the
signifies a harmony or group of harmonies Passacaglia and the Concerto in D minor.
repeated ‘obstinately’; if the group is repeated on perpetuum mobile: a nineteenth-century phrase
another degree of the scale (e.g. closing section denoting a piece of music with non-stop motion
of BWV 544.ii), a sequence is the result, in a lively tempo (e.g. unbroken semiquavers
harmonic rather than melodic. allegro or presto), clearly suiting fugues for
paraphrase: a modern term referring to the harpsichord (BWV 855.ii and 944.ii) more than
ornate treatment of a theme or cantus firmus, so for organ.
as to produce a new melody incorporating the phrygian cadence: a modern term for the
original. While examples by Böhm, Pachelbel or cadence in which the bass falls a semitone,
Buxtehude tend to leave the notes of the original analogous to the church mode whose lowest
theme on the main beats as if merely notes are E and F. Though by 1700 common
embellished, those in Clavierübung III and for the half close before an Allegro in Italian
elsewhere are more independent; see sonatas (see BWV 537.i), it appears as a final
BWV 675–677 or 622. cadence in Clavierübung III chorales
passaggio: a figuration (usually semiquavers) (BWV 671, 672), i.e. is more ‘modal’. A
shared between the hands, e.g. broken chords in ‘phrygian tendency’ would be the prominent
regular motion. For Walther’s definition in use of the semitone above a tonic, in melody
1708, see BWV 535a; for a later example, BWV or bass.
541. In Monteverdi’s Combattimento (1624), Picardy third: probably deriving from ‘sharp
passeggio indicates ‘walking by step’; third’ (tierce picarde), and denoting the major
perhaps passaggio implies ‘passing through tonic at the end of a piece in the minor. Since it
chords’. was so firmly established by 1700, particular
passus duriusculus: a term (‘a somewhat interest attaches to where it might not be meant
hard step’) probably coined by Christoph (see Fantasia BWV 542 and Toccata BWV 565)
Bernhard, c. 1650 for a passage hard to sing or might be varied in a da capo piece (see Fugue
(? – see Williams 1997 pp. 62–3), such as the BWV 548).
chromatic fourth, i.e. the semitone steps down plein jeu: see grand jeu
from tonic to dominant. See BWV 131a, 537.ii, point d’orgue: by usage, a long-held note below
588, 596, 614; paraphrased in BWV 528.iii, shifting harmonies, generally tonic or
548.ii, 648. dominant, sometimes in the soprano (see
per augmentationem, diminutionem: see BWV 541.ii) but usually in the bass, especially
augmentatio with a pedal (hence ‘pedal point’).
per giusti intervalli: strictly a term applied only Positiv (Pos): the department of an organ
to a canon accomplished ‘through exact resembling a smaller organ, but not as small as
intervals’, e.g. when the rectus rises by a minor the Portativ. Strictly it applies to any lesser
third the inversus falls by a minor third, not a manual of the organ (Brust, Rück, Echo, Unter,
minor or major depending on circumstances Seiten), and composers/copyists would mean by
(BWV 769.v bb. 1–27). The inversion is usually ‘Pos.’ simply the second manual.
less exact, as in BWV 547.ii bb. 34–8 (did this quodlibet: ‘what you please’, a work or passage
inversion suggest the chromatic metamorphosis combining various known themes either
later in the fugue?). successively or simultaneously. Though
590 Glossary
humorous in the final variation of the Goldberg length of subject in BWV 550 or 578 makes each
Variations, combined themes have a more a Spielthema.
serious Affekt in the final variation of the stile antico: ‘old style’, a term found in the
Canonic Variations. seventeenth century for a range of church or
rectus: see inversus learned music, now applied specifically to
ritornello: the ‘little return’ or recurring tutti counterpoint of the ‘Palestrina style’, with
section/theme in an Italian concerto’s first its rules for melodic lines and imitation,
movement, separated by or interspersed with restrained modulation and rhythms, and
solo sections; also, the form that results. The convention for suspensions or discords. An early
returns are often partial, with perhaps only the work such as BWV 588 might well be less pure in
last being complete. its stile antico than a mature such as BWV 669.
Rückpositiv (Rp): the ‘back positive’ or little suspirans: one of the most common of all
organ behind the organist, usually in the figurae, beginning with a rest or ‘sigh’ (catch of
gallery-front from which it speaks to a breath), as in Example 298 (i). Although
congregation directly. By 1700 or so, the theorists are not always clear, it seems that the
Rückpositiv was found in few new organs of suspirans and the corta are essentially different
central Germany, and an upper pipechest took from the tirata (a run between notes apart),
over its functions as the solo, continuo or circolo (a curling pattern), tremolo and groppo
contrasting manual. (quick note-repetitions or alternations), or
siciliano: a term now used for a supposed messanza (an in-turning pattern with a striking
‘Sicilian dance’ of a quiet pastoral nature, leap – see BWV 661 and 753).
generally in 12/8 (or 6/8) with a conspicuous tessitura: the music’s predominant range
dotted rhythm; a slow movement as in BWV 596 (‘texture’) within the total compass of an
(Vivaldi) or BWV 525. instrument or voice – wide or narrow, high or
signum congruentiae: ‘sign of agreement’, a low, thin or thick.
small mark indicating some relationship tierce en taille: see en taille
between the various parts (e.g. another voice tirata: another common figura so called because
enters in canon or drops out); by extension, a as a run of little notes it is ‘drawn’ between two
sign to warn the performer that ‘something is notes on the beat, a fifth or sometimes much
happening here’. more apart.
species counterpoint: a didactic counterpoint tmesis: in grammar, the separation of a word’s
in which lines are combined according to a syllables by another in between, hence a gap or
step-by-step schedule, note-against-note (First rest for rhetorical emphasis in a melodic or bass
Species), two notes against one (Second), four line.To be musically meaningful, the rest needs
against one (Third), two with the second tied to be on the beat, not off it.
over (Fourth), and a succession of these trillo: commonly used to denote rapid
(Fifth). alternation of notes in keyboard music, an
Spielthema: modern term for a ‘playful subject’, equivalent to the Italian vocal trillo of repeated
a recognizable type of fugue-theme, particularly notes. In the earlier music of Bach, found in
in the hands of North German composers. The subjects (BWV 535.ii), countersubjects (532.ii)
broken chords, lively rhythms, and spacious or at cadences (BW 533.i).
Example 298
Bibliography
Works cited, plus selected titles

Adlung, J. 1758. Anleitung zu der musikalischen Gelahrtheit (Erfurt; 2nd edn,
1783)
Albrecht, C. 1969. ‘J. S. Bachs Clavierübung Dritter Theil: Versuch einer Deutung’,
BJ 55, pp. 46–66
Albrecht, T. 1980. ‘Musical rhetoric in J. S. Bach’s organ Toccata BWV 565’, The
Organ Yearbook 11, pp. 84–94
Apel, W. 1967. Geschichte der Orgel- und Klaviermusik bis 1700 (Kassel)
Arfken, E. 1965. ‘Das Weimarer Orgelbüchlein Johann Sebastian Bachs’
(unpublished diss., Göttingen)
1966. ‘Zur Entstehungsgeschichte des Orgelbüchleins’, BJ 52, pp. 41–58
Argent, M. 2000. ‘Stringing along’, Musical Times 141 (Autumn), pp. 16–23
Bach, C. P. E. 1753, 1762. Versuch über die wahre Art das Clavier zu spielen, 2 vols.
(Berlin)
Bartels, U. NBA IV/xi KB [in preparation]
Beisswenger, K. 1992. Johann Sebastian Bachs Notenbibliothek = Catalogus musicus
13 (Kassel)
Benary, see J. G. Walther
Bernhard = J. Müller-Blattau, Die Kompositionslehre Heinrich Schützens in der
Fassung seines Schülers Christoph Bernhard (Kassel, 2nd edn 1963); English
trans. W. Hilse, ‘The treatises of Christoph Bernhard’, Music Forum 3 (1973),
pp. 1–196
Besseler, H. 1955. ‘J. S. Bach als Wegbereiter’, AfMw 12, pp. 1–39
Besseler, H. and G. Kraft (eds.). 1950. Johann Sebastian Bach in Thüringen
(Weimar)
Bighley, M. 1986. The Lutheran chorales in the organ works of J. S. Bach (St. Louis)
1991. ‘The Schübler Chorales as cycle: a liturgical and theological perspective’,
The Organ Yearbook 22, pp. 97–118
Billeter, B. 1997. ‘Bachs Toccata und Fuge d-moll für Orgel BWV 565 – ein
Cembalowerk?’, Mf 50, pp. 77–80
Birk, R. 1976. ‘Die Bedeutung der Vier Duette in Bachs Klavierübung III’, MuK 46,
pp. 63–9
Blankenburg, W. (ed.). 1970. Johann Sebastian Bach (Darmstadt)
Blechschmidt, E. R. 1965. Die Amalien-Bibliothek: Musikbibliothek der Prinzessin
Anna Amalia von Preussen (1723–1787) (Berlin)
Blume, F. 1936. Michael Praetorius und Esaias Compenius Orgeln Verdingnis
(Wolfenbüttel/Berlin)
1968. ‘J. S. Bach’s youth’, MQ 54, pp. 1–30
1973. Syntagma Musicum [collected writings] (Kassel)
[591]
592 Bibliography
Braun, H. 1972. ‘Eine Gegenüberstellung von Original und Bearbeitung,

dargestellt an der Entlehnung eines Corellischen Fugenthemas durch J. S.
Bach’, BJ 58, pp. 5–11
Braun, W. 1999. ‘Ein unbekanntes Orgelbau-Attestat von Johann Sebastian Bach’,
BJ 85, pp. 19–33
Breig, W. 1986a. ‘Bachs freie Orgelmusik unter dem Einfluss der italienischen
Konzertform’, Bach-Studien 9, pp. 29–43
1986b. ‘J. S. Bachs Orgeltoccata BWV 538 und ihre Entstehungsgeschichte’, in
W. Hirschmann et al. (eds.), Festchrift Martin Ruhnke zum 65. Geburtstag
(Stuttgart), pp. 56–67
1986c. ‘The “Great Eighteen” Chorales: Bach’s revisional process and the
genesis of the work’, in Stauffer and May 1986, pp. 102–20
1987. ‘Bachs Orgelchoral und die italienische Instrumentalmusik’, in W. Osthoff
and R. Wiesend (eds.), Bach und die italienische Musik (Venice), pp. 91–108
1988. ‘Zum geschichtlichen Hintergrund und zur Kompositionsgeschichte von
Bachs Orgel-Büchlein’, in Brouwer 1988, pp. 7–20
1990. ‘Textbezug und Werkidee in Johann Sebastian Bachs frühen
Orgelchorälen’, in P. Petersen (ed.), Musikkulturgeschichte. Festschrift für
Constantin Floros (Wiesbaden), pp. 167–82
1993. ‘Form problems in Bach’s early organ fugues’, in P. Brainard and
R. Robinson (eds.), A Bach tribute: essays in honor of William H. Scheide
(Chapel Hill), pp. 45–56
1995. ‘Versuch einer Theorie der Bachschen Orgelfuge’, Mf 48, pp. 14–52
1999. ‘Freie Orgelwerke’, in K. Küster (ed.), Bach Handbuch (Kassel),
pp. 613–712
Brokaw, J. A. 1995. ‘The perfectibility of J. S. Bach, or did Bach compose the
Fugue on a Theme by Legrenzi, BWV 574a?’, in R. Stinson (ed.), Bach
Perspectives 1 (Lincoln, NE), pp. 163–80
Brossard, S. de. 1703. Dictionaire de musique (Paris)
Brouwer, F. (ed.). 1988. Bachs Orgel-Büchlein in nieuw perspectief (Utrecht)
Bruggaier, E. 1959. Studien zur Geschichte des Orgelpedalspiels in Deutschland bis zur
Zeit Johann Sebastian Bachs (Frankfurt)
Bruggaier, R. 1987. ‘Das Urbild von Johann Sebastian Bachs Choralbearbeitung
“Nun komm, der Heiden Heiland” (BWV 660) – eine Komposition mit Viola
da Gamba?’, BJ 73, pp. 165–9
Budday, W. 1977. ‘Musikalische Figuren als satztechnische Freiheiten in Bachs
Orgelchoral Durch Adams Fall ist ganz verderbt’, BJ 63, pp. 139–59
Bullivant, R. 1959–60. ‘The fugal technique of J. S. Bach’ (unpublished diss.,
Oxford)
Burba, O.-J. 1994. ‘Bachs Choralfughetten aus der Kirnbergerschen Sammlung’,
MuK 64, pp. 87–94
Busch, H. J. 1985. ‘Unbekannte Registrierungsanweisungen aus dem Umkreis
Johann Sebastian Bachs’, Ars organi 33, pp. 122–5
Butler, G. G. 1981. ‘Borrowings in J. S. Bach’s Klavierübung III’, Canadian
University Music Review 4, pp. 204–17
593 Bibliography
1983. ‘Der vollkommene Capellmeister as a stimulus to J. S. Bach’s late fugal

writing’, in G. J. Buelow and H. J. Marx (eds.), New Mattheson Studies
(Cambridge), pp. 293–305
1990. Bach’s Clavier-Übung III: the making of a print, with a companion study of
the Canonic Variations on “Vom Himmel hoch” BWV 769 (Durham, NC)
1992. ‘Johann Sebastian Bachs Gloria in excelsis Deo BWV 191: Musik für ein
Leipziger Dankfest’, BJ 78, pp. 65–71
2000. ‘J. S. Bachs Kanonische Veränderungen über “Vom Himmel hoch”
(BWV 769). Ein Schluss-strich unter die Debatte um die Frage der “Fassung
letzter Hand” ’, BJ 86, pp. 9–34
Butt, J. 1988. ‘Bach’s Organ Sonatas BWV 525–530: compilation and
recomposition’, The Organ Yearbook 19, pp. 80–90
1990. Bach interpretation. Articulation marks in primary sources of J. S. Bach
(Cambridge)
1995. ‘J. S. Bach and G. F. Kauffmann: reflections on Bach’s later style’, in
Melamed 1995, pp. 47–61
Chailley, J. 1974. Les chorals pour orgue de J.-S. Bach (Paris)
Clark, R. and J. D. Peterson (eds.). 1984. Johann Sebastian Bach: Orgelbüchlein
(St. Louis)
Claus, R.-D. 1995. Zur Echtheit von Toccata und Fuge d-moll BWV 565 (Cologne,
2/1998)
Clement, A. 1989. ‘ “O Jesu, du edle Gabe”. Studien zum Verhältnis von Text und
Musik in den Choralpartiten und den Kanonischen Veränderungen von
Johann Sebastian Bach’ (unpublished diss., Utrecht)
1990. ‘Zur rhythmischen Ausführung des dritten Satzes der Sonata e-moll
BWV 528 von J. S. Bach’, Ars organi 38, pp. 75–83
1993. ‘Die ideale Grundlage der vier Duette BWV 802–805?’, in R. Steiger (ed.),
Die seelsorgliche Bedeutung Johann Sebastian Bachs. Kantaten zum Thema Tod
und Sterben (Heidelberg), pp. 193–226
1995. ‘Musik, Text und Theologie im Orgelwerk J. S. Bachs’, in Wege zu
Bach II = Jahresgabe der Internationalen Bach-Gesellschaft Schaffhausen
(Stuttgart)
1999. Der Dritte Teil der Clavierübung von Johann Sebastian Bach (Middelburg)
Currie, R. N. 1973. ‘Notizbuch für Studenten – cyclic unity in Bach’s Sechs
Choräle: a new look at the Schüblers’, BACH (Riemenschneider Institute) 4:1,
pp. 26–38 and 4:5, pp. 25–39
Czubatynski, U. 1993. ‘Choralvorspiel und Choralbegleitung im Urteil J. S. Bachs’,
BJ 79, p. 223
Dadelsen, G. von. 1957. Bemerkungen zur Handschrift Johann Sebastian Bachs,
seiner Familie und seines Kreises = T übinger Bach-Studien 1 (Trossingen)
1958. Beiträge zur Chronologie der Werke Johann Sebastian Bachs = Tübinger
Bach-Studien 4/5 (Trossingen)
1963. ‘Zur Entstehung des Bachschen Orgelbüchleins’, in A. A. Abert and
W. Pfannkuch (eds.), Festschrift Friedrich Blume zum 70. Geburtstag (Kassel),
pp. 74–9
594 Bibliography
Dähnert, U. 1962. Der Orgel- und Instrumentenbauer Zacharias Hildebrandt

(Leipzig)
Dalton, J. 1966. ‘Bach interpretation’, MT 107, pp. 341, 440, 536ff.
David, H. T. and A. Mendel. 1945. The Bach Reader (New York; revised edn with
supplement, 1966; revised and enlarged by C. Wolff, 1998)
David, W. 1951. Johann Sebastian Bachs Orgeln (Berlin)
Daw, S. 1976. ‘Copies of J. S. Bach by Walther and Krebs: P 801, P 802, P 803’,
Derr, E. 1987. ‘Bach’s Sonata in G, BWV 530.i and its Vivaldi data-base’, The
Diapason 78 (September), pp. 14–16
Dickinson, A. E. F. 1956. Bach’s fugal works: with an account of fugue before and after
Bach (London)
Dietrich, F. 1929. ‘J. S. Bachs Orgelchoral und seine geschichtlichen Wurzeln’,
BJ 26, pp. 1–89
1931. ‘Analogieformen in Bachs Tokkaten und Präludien für die Orgel’, BJ 28,
pp. 51–71
Dijk, P. van. 1988. ‘Bachs registratie-principes, toegespitst op het Orgel-Büchlein’,
in Brouwer 1988, pp. 157–90
Dirksen, P. 1992. ‘The background to Bach’s Fifth Brandenburg Concerto’, in
Dirksen 1992, pp. 157–85
1998. ‘Zur Frage des Autors der A-dur-Toccata BWV Anh. 178’, BJ 84,
pp. 121–35
1999. ‘Bachs koraalpartita Ach, was soll ich Sünder machen’, Het Orgel 95,
pp. 15–24
2002. ‘Bach’s “Acht Choralfughetten”. Ein unbeachtetes Leipziger Sammelwerk’,
in U. Leisinger, ed., Bach in Leipzig – Bach und Leipzig. Konferenzbericht Leipzig
2000 (Hildesheim), pp. 155–82
Dirksen, P. (ed.). 1992. The harpsichord and its repertoire: proceedings of the
International Harpsichord Symposium Utrecht 1990 (Utrecht)
Dufourcq, N. 1948. J. S. Bach, Le maı̂tre de l’orgue (Paris)
Dürr, A. 1951. Studien über die frühen Kantaten J. S. Bachs (Leipzig)
1954. ‘Neues über die Möllersche Handschrift’, BJ 41, pp. 75–9
1956. ‘Gedanken über J. S. Bachs Umarbeitungen eigener Werke’, BJ 43,
pp. 93–111
1970. ‘Zur Chronologie der Handschrift Johann Christoph Altnickols und
Johann Friedrich Agricolas’, BJ 56, pp. 44–63
1977. Studien über die frühen Kantaten J. S. Bachs (2nd, enlarged edn, Wiesbaden)
1984. Johann Sebastian Bach. Seine Handschrift: Abbild seines Schaffens
(Wiesbaden)
1986. ‘Kein Meister fällt vom Himmel’, Musica 40, pp. 309–12
1988. ‘Gedanken zu den späten Kantaten Bachs’, in C. Wolff, ed., Johann
Sebastian Bachs Spätwerk und dessen Umfeld. Perspektiven und Probleme (Kassel),
pp. 58–63
Dürr, A. (ed.). 1987. J. S. Bach, formerly ascribed to, Eight little Preludes and Fugues
BWV 553–560 = BA 6497 (Kassel)
Eck, C. L. van Panthaleon van. 1981. J. S. Bach’s critique of pure music (Buren)
595 Bibliography
Edler, A. 1982. Der nordelbische Organist. Studien zu Sozialstatus, Funktion und

kompositorischer Produktion eines Musikerberufs von der Reformation bis zum 20.
Jahrhundert (Kassel)
1995. ‘Thematik und Figuration in der Tastenmusik des jungen Bach’, in Heller
and Schulze 1995, pp. 87–110
Ehricht, K. 1949–50. ‘Die zyklische Gestalt und die Aufführungsmöglichkeit
des III. Teiles der Klavierübung von Joh. Seb. Bach’, BJ 51, pp. 40–56
Eickhoff, H. J. 1967. ‘Bach’s chorale-ritornello forms’, Music Review 28,
pp. 257–76
Eller, R. 1957. ‘Zur Frage Bach-Vivaldi’, in Kongressbericht der Gesellschaft für
Musikforschung, Hamburg 1956 (Kassel), pp. 80–5
1958. ‘Geschichtliche Stellung und Wandlung der Vivaldischen Konzertform’, in
Kongress-Bericht Wien 1956 (Graz), pp. 150–5
Emans, R. 1997. Johann Sebastian Bach. Orgelchoräle zweifelhafter Echtheit.
Thematischer Katalog (Göttingen)
2000. ‘Zu Un(r)echt?’, Organ 2000, Heft 4, pp. 25–31
NBA IV/x KB [in preparation]
Emery, W. 1952. Notes on Bach’s organ works, 1: Eight Short Preludes and Fugues
(London)
1957. Notes on Bach’s organ works, 4: Six Sonatas for Two Manuals and Pedal
(London)
1957–8. ‘The dating of Bach’s organ works’, The Organ 37, pp. 181–7
1963. ‘A note on the history of Bach’s Canonic Variations’, MT 104, pp. 32–3
1966. ‘Some speculations on the development of Bach’s organ style’, MT 107,
pp. 596–603
1970. ‘Der Klaviaturumfang von Bachs Orgeln als Beweismittel für die
Datierung seiner Werke’, in Blankenburg 1970, pp. 162–77; translation of ‘The
compass of Bach’s organs as evidence of the date of his works’, The Organ 32
(1952), pp. 92–100
1974. ‘Cadence and chronology’, in Marshall 1974, pp. 156–64
Emery, W. (ed.). 1959. J. S. Bach: Prelude, Trio and Fugue in B flat = Novello Early
Organ Music 12 (London 1959)
Eppstein, H. 1965. ‘J. S. Bachs Triosonate G dur (BWV 1039) und ihre
Beziehungen zur Sonate für Gambe und Cembalo G dur (BWV 1027)’, Mf 18,
pp. 126–37
1966. Studien über J. S. Bachs Sonaten für ein Melodieinstrument und obligates
Cembalo (Uppsala)
1969. ‘Grundzüge in J. S. Bachs Sonatenschaffen’, BJ 55, pp. 5–30
1976. ‘Chronologieprobleme in Johann Sebastian Bachs Suiten für
Soloinstrument’, BJ 61, pp. 35–57
Ernst, H. P. 1987. ‘Joh. Seb. Bachs Wirken am ehemaligen Mühlhäuser
Augustinerinnenkloster und das Schicksal seiner Wender-Orgel’, BJ 73,
pp. 75–83
FaIck, M. 1913. Wilhelm Friedemann Bach (Leipzig)
Finke-Hecklinger, D. 1970. Tanzcharaktere in Johann Sebastian Bachs Vokalmusik =
Tübinger Bach-Studien 6 (Trossingen)
596 Bibliography
Fock, G. 1974. Arp Schnitger und seine Schule (Kassel)

Forkel, J. N. 1802. Über Johann Sebastian Bachs Leben, Kunst und Kunstwerke
(Leipzig)
Forst, I. 1990. ‘Johann Sebastian Bachs Orgelkonzert d-moll: Vorlage und
Endform’, in R. Emans and M. Wendt (eds.), Beiträge zur Geschichte des
Konzerts. Festschrift Siegfried Kross zum 60. Geburtstag (Bonn)
Franklin, D. O. 1991. ‘The Carnegie Manuscript and J. S. Bach’, BACH
[Riemenschneider Institute] 22, pp. 5–15
Freylinghausen, J. A. 1741. Geistreiches Gesangbuch (Halle; 1st edn, 1706)
Friedrich, F. 1983. ‘Johann Sebastian Bach und die Trost-Orgel zu Altenburg.
Bemerkungen zur Problematik der “Bach-Orgel” ’, BJ 69, pp. 101–7
1991. ‘Aufführungspraktische Beobachtungen am Orgelschaffen von Johann
Ludwig Krebs’, Ars organi 39, pp. 135–50
1998. ‘Johann Sebastian Bach und der thüringische Orgelbau’, Ars organi 46,
pp. 130–6
Frotscher, G. 1934–5. Geschichte des Orgelspiels und der Orgelkomposition, 2 vols.
(Berlin)
Fuchs, I. (ed.). 1992. Johann Sebastian Bach. Beiträge zur Wirkungsgeschichte
(Vienna)
Geck, M. 1968. Nicolaus Bruhns: Leben und Werk (Cologne)
Geck, M. (ed.). 1969. Bach-Interpretationen (Göttingen)
Geiringer, K. 1966. Johann Sebastian Bach: the culmination of an era (London)
Gernhardt, K. 1986. ‘Versuch einer Rekonstruktion der Orgel in der Schlosskirche
zu Weissenfels’, Bach-Studien 9, pp. 51–5
1988. ‘Über das “Gedackt zur Musik” in Orgeln des 17. und 18. Jahrhunderts’,
in Hoffmann 1988, pp. 133–6
Gilbert, K. (ed.) 1993. Johann Sebastian Bach. Pièce d’Orgue BWV 572 (Monaco)
Gojowy, D. 1972. ‘Lied und Sonntag in Gesangbüchern der Bach-Zeit: Zur Frage
des De tempore bei Chorälen in Bachs Kantaten’, BJ 58, pp. 24–60
Goldhan, W. 1987. Johann Sebastian Bach Sechs Sonaten für Orgel / Six Sonatas for
Organ (BWV 525–530), Faksimile des Autographs = Documenta musicologica
2.XIX (Kassel), preface
Grace, H. c. 1922. The organ works of Bach (London)
Grimm, J. 1969. Das Neu Leipziger Gesangbuch des Gottfried Vopelius (Leipzig, 1692)
(Berlin)
Grüss, H. 1985. ‘Über die Tradition des Cantus-firmus-Kanons. Eine Ergänzung
zur Thema: Bach und das Mittelalter’, BJ 71, pp. 134–46
Gwinner, V. 1968. ‘Bachs d-moll Tokkata als Credo-Vertonung’, MuK 38,
pp. 240–2
Hackel, W. 1987. ‘Johann Sebastian Bachs Ruf als Orgelsachverständiger’, BzBf 6,
pp. 92–5
Hammerschlag, J. 1950. ‘Der weltliche Charakter in Bachs Orgelwerken’,
Bach-Probleme. Festschrift (Leipzig), pp. 17–37
Hartmann, G. 1986. ‘Authentischer Bach-Elbel. Marginalie zu einem der
angeblichen Bach-Choräle der Neumeister-Sammlung’, NZfM 147/4, pp. 4–6
597 Bibliography
Heinemann, M. 2000. ‘Neumeisters Choräle. Zu einer Sammelhandschrift

“früher” Bach-Werke’, The Organ Yearbook 29, pp. 7–18
Heinichen, J. D. 1711. Neu erfundene und gründliche Anweisung zu vollkommener
Erlernung des General-Basses (Hamburg)
Heller, K. 1980. Bearbeitungen fremder Werke = NBA IV/8 KB (Kassel)
Heller, K. and H.-J. Schulze (eds.). 1995. Das Frühwerk Johann Sebastian Bachs
(Cologne)
Henkel, H. 1986. ‘Zur Geschichte der Scheibe-Orgel in der Leipziger
Johanniskirche’, Bach-Studien 9, pp. 44–50
Hering, H. 1974. ‘Spielerische Elemente in J. S. Bachs Klaviermusik’, BJ 60,
pp. 44–69
Herz, G. 1974. ‘Der lombardische Rhythmus im “Domine Deus” der
h-Moll-Messe J. S. Bachs’, BJ 60, pp. 90–7
1984. Bach-Quellen in Amerika / Bach sources in America (Kassel)
Hilgenfeldt, C. L. 1850. Johann Sebastian Bachs Leben, Wirken und Werke (Leipzig)
Hill, R. 1985. ‘ “Echtheit angezweifelt”: style and authenticity in two suites
attributed to Bach’, EM 13, pp. 248–55
1987. ‘The Möller Manuscript and the Andreas Bach Book: two keyboard
manuscripts from the circle of the young Johann Sebastian Bach’
(unpublished diss., Harvard)
1990. ‘Tablature versus staff notation: or, why did the young J. S. Bach compose
in tablature?’, in P. Walker (ed.), Church, stage and studio. Music and its contexts
in seventeenth-century Germany (Ann Arbor), pp. 349–59
Hill, R. (ed.). 1991. Keyboard music from the Andreas Bach Book and the Möller
manuscript (Cambridge, MA)
Hobohm, W. 1987. ‘Ein unbekanntes Urteil Johann Matthesons über Joh. Seb.
Bach. Anmerkungen zu einem Literatenstreit um 1730’, BJ 73, pp. 19–27
1988. ‘Überlegungen zu Bachs Beziehungen zu dem halleschen Orgelbauer
H. A. Contius (Cuncius)’, in Hoffmann 1988, pp. 125–31
Hoffmann, W. and A. Schneiderheinze (eds.). 1988. Bericht über die
wissenschaftliche Konferenz zum V. Internationalen Bachfest der DDR in
Verbindung mit dem 60. Bachfest der Neuen Bachgesellschaft, Leipzig, 25 bis 27
März 1985 (Leipzig)
Hoffmann-Erbrecht, L. 1972. ‘J. S. Bach als Schöpfer des Klavierkonzerts’,
in K. Dorfmüller (ed.), Quellenstudien zur Musik (Frankfurt), pp. 69–77
Hofmann, K. 1999. ‘Ein verschollenes Kammermusikwerk Johann Sebastian
Bachs. Zur Fassungsgeschichte der Orgelsonate Es-dur (BWV 525) und der
Sonate A-dur für Flöte und Cembalo (BWV 1032)’, BJ 85, pp. 67–79
Hollmann, G. 1990. ‘Eine unbekannte Fingersatztabelle des Bach-Schülers Johann
Gotthilf Ziegler’, in W. Schlepphorst (ed.), Orgelkunst und Orgelforschung.
Gedenkschrift Rudolf Reuter (Kassel), pp. 59–68
Holschneider, A. 1964. ‘Zu Mozarts Bearbeitung Bachscher Fugen’, Mf 17,
pp. 51–6
Honders, C. 1988. ‘Het Orgel-Büchlein gehoord vanuit de teksten’, in Brouwer
1988, pp. 21–51
598 Bibliography
Horn, V. 1986. ‘French influence in Bach’s organ works’, in Stauffer and May 1986,
pp. 256–73
Huggler, H. E. 1935. Johann Sebastian Bachs Orgelbüchlein (Bern)
Humphreys, D. 1982. ‘The D minor Toccata BWV 565’, Early Music 10, pp. 216–17
1983. The esoteric structure of Bach’s Clavierübung III (Cardiff)
1985. ‘Did J. S. Bach compose the F minor Prelude and Fugue BWV 534?’, in
Williams 1985, pp. 173–84
1994. ‘ “Und besonders für Kenner von dergleichen Arbeit” – a study of the
symbolic aspects of Bach’s “Dritter Theil der Clavierübung” ’, The Organ
Yearbook 24, pp. 41–64
2000. ‘J. S. Bach, J. P. Kellner and the Prelude and Fugue in A major, BWV 536’,
The Organ Yearbook 29, pp. 27–44
Jacob, A. 1997. Studien zu Kompositionsart und Kompositionsbegriff in Bachs
Klavierübungen = Beihefte zum Archiv für Musikwissenschaft 40 (Stuttgart)
Jauernig, R. 1950. ‘Johann Sebastian Bach in Weimar’, in Besseler and Kraft 1950,
pp. 49–105
Joelson-Strohbach, H. 1987. ‘Nachricht von verschiedenen verloren geglaubten
Handschriften mit barocker Tastenmusik’, AfMw 44, pp. 91–140
Kast, P. 1958. Die Bach-Handschriften der Berliner Staatsbibliothek = Tübinger
Bach-Studien 2/3 (Trossingen)
Kaufmann, M. G. 2000. ‘Die innere Ordnung in Johann Sebastian Bachs
Orgelbüchlein’, in K. Johannsen et al. (eds.), Musicus Doctus. Festschrift für
Hans Musch zum 65. Geburtstag (Freiburg), pp. 33–43
Kee, P. 1992. ‘Die Passacaglia C-moll BWV 582 von Johann Sebastian Bach und
die Passacaglia von Dietrich Buxtehude’, in Fuchs 1992, pp. 251–64
Keller, H. 1948. Die Orgelwerke Bachs (Leipzig)
1973. ‘Unechte Orgelwerke Bachs’, BJ 34, pp. 59–82
Kent, C. 1992. ‘J. S. Bach and the “Livre d’Orgue” of Nicolas de Grigny’, in
R. Judd (ed.), Aspects of keyboard music: essays in honour of Susi Jeans (Oxford),
pp. 45–59
Kilian, D. 1961. ‘J. S. Bach, Praeludium und Fuge d-moll, BWV 539’, Mf 14,
pp. 323–8
1962. ‘Studie über J. S. Bachs Fantasie und Fuge c-moll (BWV 562)’, in Hans
Albrecht in Memoriam (Kassel), pp. 127–35
1969. ‘Dreisätzige Fassungen Bachscher Orgelwerke’, in Geck 1969, pp. 12–21
1978. ‘Über einige neue Aspekte zur Quellenüberlieferung von Klavier- und
Orgelwerken Johann Sebastian Bachs’, BJ 64, pp. 61–72
1978–9. Präludien, Toccaten, Fantasien and Fugen für Orgel = NBA IV/5–6 KB,
3 vols. (Kassel)
1983. ‘Zu einem Bachschen Tabulaturautograph’, in W. Rehm (ed.), Bachiana
et alia musicologica (Kassel), pp. 161–7
1988. Sechs Sonaten und verschiedene Einzelwerke = NBA IV/7 KB (Kassel)
Kinsky, G. 1936. ‘Pedalklavier oder Orgel bei Bach?’, AM 8, pp. 158–61
1937. Die Originalausgaben der Werke Johann Sebastian Bachs (Vienna)
Klein, H.-G. 1970. Der Einfluss der Vivaldischen Konzertform im Instrumentalwerk
Johann Sebastian Bachs (Strasbourg/Baden-Baden)
599 Bibliography
Kloppers, J. 1966. Die Interpretation und Wiedergabe der Orgelwerke Bachs

(Frankfurt)
Klotz, H. 1950. ‘Bachs Orgeln und seine Orgelmusik’, Mf 3, pp. 189–203
1962a. ‘J. S. Bach und die Orgel’, MuK 32, pp. 49–55
1962b. Die Orgelchoräle aus der Leipziger Originalhandschrift = NBA IV/2 KB
(Kassel)
1962c. Die einzeln überlieferten Orgelchoräle = NBA IV/3 KB (Kassel)
1966. ‘Über J. S. Bachs Kanonwerk Vom Himmel hoch, da komm’ ich her’,
Mf 19, pp. 295–304
1969a. ‘Originale Spielanweisungen in Bachs Orgelwerken und ihre
Konsequenzen für die Interpretation’, in Geck 1969, pp. 112–18
1969b. ‘Les Critères de l’interprétation française sont-ils applicables à la
musique d’orgue de J-S.Bach?’, in L’interprétation de la musique française
aux XVIIe et XVIIIe Siècles, Colloques Internationaux du CNRS (Paris),
pp. 155–72
1973. ‘Die Kanonische Veränderungen in Entwurf, Reinschrift und Druck’, in
Wörthmöller 1973, pp. 11–14
1975. Über die Orgelkunst der Gotik, der Renaissance und des Barock (2nd edn,
Kassel)
1984. Die Ornamentik der Klavier- und Orgelwerke von Johann Sebastian Bach
(Kassel)
Kobayashi, Y. 1973. Franz Hauser und seine Bach-Handschriften-Sammlung
(dissertation, Göttingen)
1983. ‘Die Gehrener Kantor Johann Christoph Bach (1673–1727) und
seine Sammelbände mit Musik für Tasteninstrumente’, in Rehm 1983,
pp. 168–77
1988. ‘Zur Chronologie der Spätwerke Johann Sebastian Bachs. Kompositions-
und Aufführungstätigkeit von 1736 bis 1750’, BJ 74, pp. 7–72
1989. Die Notenschrift Johann Sebastian Bachs. Dokumentation ihrer
Entwicklung = NBA IX/2 (Kassel)
1995. ‘The variation principle in J. S. Bach’s Passacaglia in C minor BWV 582’,
in Melamed 1995, pp. 62–9
2000. ‘Probleme zu Bachs “18 Chorälen” ’, Organ-Kenkyu 28, pp. 1–14
Kollmann, A. F. C. 1799. An essay on practical musical composition (London)
Krams, P. 1974. Wechselwirkungen zwischen Orgelkomposition und Pedalspieltechnik
(Wiesbaden)
Krause, J. 1956. ‘Die grosse Bearbeitung von Jesus Christus unser Heiland aus
Clavierübung III von J. S. Bach’, MuK 35, pp. 117–26
Krause, P. 1964. Handschriften der Werke Johann Sebastian Bachs in der
Musikbibliothek der Stadt Leipzig (Leipzig)
Kremer, J. 1993. ‘Die Organistenstelle an St. Jakobi in Hamburg: eine “convenable
station” für Johann Sebastian Bach?’, BJ 79, pp. 217–22
Krey, J. 1956. ‘Bachs Orgelmusik in der Weimarer Zeit’ (dissertation, Jena)
Kröhner, C. 1995. ‘Johann Sebastian Bach und Johann Friedrich Bach als
Orgelexaminatoren im Gebiet der freien Reichsstadt Mühlhausen nach 1708’,
BJ 81, pp. 83–91
600 Bibliography
Krüger, E. 1970. Stilistische Untersuchungen zu ausgewählten frühen Klavierfugen

Johann Sebastian Bachs = Hamburger Beiträge zur Musikwissenschaft 2
(Hamburg)
Krumbach, W. 1985. ‘Sechzig unbekannte Orgelwerke von Johann Sebastian Bach.
Ein vorläufiger Fundbericht’, NZfM 146/3, pp. 4–12, 146/5, pp. 6–18
Krummacher, F. 1978. Die Choralbegleitung in der protestantischen Figuralmusik
zwischen Praetorius und Bach (Kassel)
1985. ‘Bach und die norddeutsche Orgeltoccata: Fragen und Überlegungen’,
BJ 71, pp. 119–34
Kube, M. 1999. ‘Choralgebundene Orgelmusik’, in K. Küster (ed.), Bach Handbuch
(Kassel), pp. 545–612
Leaver, R. A. 1975. ‘Bach’s Clavierübung III: some historical and theological
considerations’, The Organ Yearbook 6, pp. 17–32
1985. ‘Bach and hymnody: the evidence of the Orgelbüchlein’, EM 13, pp. 227–36
le Huray, P. and J. Butt. 1985. ‘In search of Bach the organist’, in Williams 1985,
pp. 185–206
Leutert, H. 1967. ‘Betrachtungen über Bachs Choral-triptychon “O Lamm Gottes,
unschuldig” ’, Musik und Gottesdienst 21, pp. 21–5
Löffler, H. 1923. ‘Die Choralpartita “Ach, was soll ich Sünder machen” ’, BJ 20,
pp. 31–56
1940–8. ‘Johann Tobias Krebs und Matthias Sojka, zwei Schüler Johann
Sebastian Bachs’, BJ 37, pp. 136–48
Löhlein, H.-H. 1987. Orgelbüchlein. Sechs Choräle von verschiedener Art
(Schübler-Choräle), Orgelpartiten = NBA IV/1 KB (Kassel)
Löhlein, H.-H. (ed.). 1981. Orgelbüchlein BWV 599–644: Faksimile des Autographs
(Leipzig)
Lohmann, H. 1974. ‘Bemerkungen zur Zahlensymbolik in der Klavier-Übung III
von Joh. Seb. Bach’, Der Kirchenmusiker 25, pp. 55–8
Luedtke, H. 1918. ‘Sebastian Bachs Choralvorspiele’, BJ 15 (1918), pp. 1–96
McLean, H. J. 1993. ‘BWV 551, a Bachian orphan’, BACH [Riemenscheider
Institute] 24, pp. 35–42
Mann, A. 1964–5. ‘Eine Kompositionslehre von Händel’, HJ 10/11, pp. 35ff.
Marpurg, F. W. 1753–4. Abhandlung von der Fuge, 2 vols. (Berlin)
Marshall, R. L. 1986. ‘Organ or Klavier? Instrumental prescriptions in the sources
of Bach’s keyboard works’, in Stauffer and May 1986, pp. 212–39
1989. The music of Johann Sebastian Bach. The sources, the style, the significance
(New York)
Marshall, R. L. (ed.). 1974. Studies in renaissance and baroque music in honor of
Arthur Mendel (Kassel)
Marx, H.-J. 1983. ‘Johann Matthesons Nachlass. Zum Schicksal der
Musiksammlung der alten Stadtbibliothek Hamburg’, AM 55, pp. 108–24
Mattheson, J. 1713. Das neu eröffnete Orchestre (Hamburg)
1721. ‘Sammlung von Orgeldispositionen’, in F. E. Niedt, Musicalische
Handleitung anderer Teil (Hamburg), pp. 156–204
1731. Grosse General-Bass-Schule oder exemplarische Organistenprobe (Hamburg)
1739. Der vollkommene Capellmeister (Hamburg)
601 Bibliography
May, E. D. 1974a. ‘Breitkopf’s role in the transmission of J. S. Bach’s organ

chorales’ (unpublished diss., Princeton, 1974)
1974b. ‘J. G. Walther and the lost Weimar autographs of Bach’s organ works’, in
Marshall 1974, pp. 264–82
1974c. ‘Eine neue Quelle für J. S. Bachs Einzeln Überlieferte Orgelchoräle’,
BJ 60 (1974), pp. 98–103
1986. ‘The types, uses and historical position of Bach’s organ chorales’, in
Stauffer and May 1986, pp. 81–101
1996. ‘Connections between Breitkopf and J. S. Bach’, in G. B. Stauffer (ed.),
Bach Perspectives 2 (Lincoln, NE), pp. 11–26
Melamed, D. R. (ed.). 1995. Bach Studies 2 (Cambridge)
Meyer, U. 1972. ‘Zur Frage der inneren Einheit von Bachs Siebzehn Chorälen
(BWV 651–667)’, BJ 58, pp. 61–75
1974. ‘Zur Einordnung von J. S. Bachs einzelnüberlieferten Orgelchorälen’,
BJ 60, pp. 75–89
1979a. ‘Zum Problem der Zahlen in Johann Sebastian Bachs Werk’, MuK 49,
pp. 58–71
1979b. J. S. Bachs Musik als theonome Kunst (Wiesbaden, 1979)
1985. ‘Aus tiefer Not – ein Lobgesang. Versuch über J. S. Bach’s Orgelchoral
BWV 686’, MuK 55, pp. 73–80
1987. ‘Über J. S. Bachs Orgelchoralkunst’, Beiträge zur theologischen
Bachforschung 4 (1987), pp. 7–46
1988. ‘In profundis – in excelsis. Zu drei Orgelchorälen J. S. Bachs über “Allein
Gott in der Höh sei Ehr” ’, MuK 58, pp. 12–16
Musch, H. 1981. ‘Eine Spiel- und Registriermöglichkeit für das mitteldeutsche
Orgeltrio des 18. Jahrhunderts’, Ars organi 29, pp. 177–80
Neumann, W. 1956. ‘Zur Frage der Gesangbücher J. S. Bachs’, BJ 43, pp. 112–23
1967. Handbuch der Kantaten Johann Sebastian Bachs (3rd edn, Leipzig)
Niedt, F. E. 1700, 1706, 1717. Musicalische Handleitung I (Hamburg), II
Handleitung zur Variation (Hamburg; 2nd, enlarged edn by J. Mattheson,
1721), III Musicalische Handleitung, dritter und letzter Theil (ed. J. Mattheson)
O’Donnell, J. 1989. ‘Mattheson, Bach, Krebs and the Fantasia and Fugue in C
minor BWV 537’, The Organ Yearbook 20, pp. 88–94
Oppel, R. 1906. ‘Die grosse A-moll-Fuge für Orgel und ihre Vorlage’, BJ 3,
pp. 74–8
1910. ‘Über Joh. Kasp. Ferd. Fischers Einfluss auf Joh. Seb. Bach’, BJ 7,
pp. 63–9
Osthoff, W. and R. Wiesend (eds.). 1987. Bach und die italienische Musik (Venice)
Ottenberg, H.-G. 1986. ‘Zur Frage der Authentizität der Choralbearbeitung Aus
der tiefe rufe ich (BWV 745)’, BJ 72, pp. 127–30
Ouwerkerk, P. 1995. ‘La Passacaglia BWV 582: una nuova interpretazione’,
Informazione organistica VII/1, pp. 10–18
Overholtzer, Harry, Private communication
Pauly, H.-J. 1964. Die Fuge in den Orgelwerken Dietrich Buxtehudes (Regensburg)
Pirro, A. 1906. J.-S. Bach (Paris)
Plantinga, L. B. 1967. Schumann as critic (New Haven, 1967)
602 Bibliography
Ponsford, D. 2000. ‘J. S. Bach and the nature of French “influence” ’, The Organ
Yearbook 29, pp. 59–74
Power, T. 2001. ‘J. S. Bach and the Divine Proportion’ (unpublished diss., Duke
University NC, 2001)
Praetorius, E. 1906–7. ‘Neues zur Bach-Forschung’, Sammelbände der
internationalen Musikgesellschaft 8, pp. 95–101
Praetorius, M. 1610. Musae sioniae vol. IX (Wolfenbüttel)
Printz, W. C. 1696. Phrynidis Mytilenaei oder des satyrischen Componisten anderer
Theil (2nd edn, Leipzig/Dresden)
Quantz, J. J. 1752. Versuch einer Anweisung die Flöte traversiere zu spielen (Berlin)
Radulescu, M. 1980. ‘On the form of Johann Sebastian Bach’s Passacaglia in C
minor’, The Organ Yearbook 11, pp. 95–103
Rampe, S. Forthcoming. ‘Bachs Piece d’Orgue G-dur BWV 572: Gedanken zu ihrer
Konzeption’, in M. Geck (ed.), Dortmunder Bach-Symposium 2002 (Witten)
Rasch, R. 1984. ‘Friedrich Suppig and his Labyrinthus Musicus of 1722’, The
Ratte, F. J. 1989. ‘Die Temperatur als Mittel der musikalischen Rhetorik am
Beispiel des Orgelbüchleins von Johann Sebastian Bach’, in J. P. Fricke (ed.),
Die Sprache der Musik. Festschrift Klaus Wolfgang Niemöller (Regensburg),
pp. 401–17
Rehm, W. (ed.). 1983. Bachiana et alia musicologica: Festschrift Alfred Dürr zum 65.
Geburtstag (Kassel)
Riedel, F. W. 1960. Quellenkundliche Beiträge zur Geschichte der Musik für
Tasteninstrumente in der 2. Hälfte des 17. Jahrhunderts (Kassel)
1968. ‘Der Einfluss der italienischen Klaviermusik des 17. Jahrhunderts auf die
Entwicklung der Musik für Tasteninstrumente in Deutschland während der
ersten Hälfte des 18. Jahrhunderts’, Analecta Musicologica 5, pp. 18–33
Rienäcker, G. 1995. ‘Beobachtungen zu den Präludien BWV 543.i und 543a.i’, in
Heller and Schulze 1995, pp. 116–29
Ryom, P. 1966–7. ‘La Comparaison entre les versions différentes d’un concerto
d’Antonio Vivaldi transcrit par J. S. Bach’, Dansk Aarbog for Musikforskning,
pp. 92–111
1973. Antonio Vivaldi: table des concordances des oeuvres (Copenhagen)
1977. Les Manuscrits de Vivaldi (Copenhagen)
Sachs, K.-J. 1980. ‘Die “Anleitung . . . auf allerhand Arth einen Choral
durchzuführen”, als Paradigma der Lehre und der Satzkunst Johann Sebastian
Bachs’, AfMw 37, pp. 135–54
Sackmann, D. 1988. ‘Toccata F-dur (BWV 540) – eine analytische Studie’, in
Hoffmann 1988, pp. 351–60
1991. ‘Der “Yaler” Bach. Beobachtungen zur Handschrift US-NH (Yale) LM
4708 und deren Umfeld’, BzBf 9/10, pp. 165–72
1998. ‘Johann Sebastian Bachs vermeintliche “Arnstädter Gemeindechoräle” ’,
Schweizer Jahrbuch für Musikwissenschaft 18, pp. 229–50
Schäfertöns, R. 2000. ‘Ein Vergleich zwischen Johann Sebastian Bachs Fuge D Dur
(BWV 532) und Johann Heinrich Buttstedts Fuga in g’, The Organ Yearbook 29,
pp. 19–26
603 Bibliography
Scheide, W. H. 1999. ‘Theologische Überlegungen zu Bachs Orgelchoral “Vater

unser im Himmelreich” BWV 682’, BJ 85, pp. 81–104
Schering, A. 1902–3. ‘Zur Bach-Forschung’, Sammelbände der internationalen
Musikgesellschaft 4, pp. 234–43
1941. J. S. Bach und das Musikleben Leipzigs im 18. Jahrhundert (Leipzig)
Schiffner, M. 1991. ‘Werk – Sammlung – Zyklus: Bachs Klavierübung Teil III’, BzBf
9/10 (1991), pp. 77–84
Schmidt, C. M. 1986. ‘Zur Entwicklung der Form im emphatischen Sinne. Bachs
Orgelpräludium h-Moll BWV 544’, Musiktheorie 1, pp. 195–204
Schmidt-Mannheim, H. (ed.). 1965. J. S. Bach, Sechs Choräle für Orgel (Berlin)
Schmitz, A. 1952. ‘Die Figurenlehre in den theoretischen Werken Johann
Gottfried Walthers’, AfMw 9, pp. 79–100
1970. ‘Die oratorische Kunst J. S. Bachs: Grundfragen und Grundlagen (1950)’,
in Blankenburg 1970, pp. 61–84
Schmögner, T. 1995. ‘Neue Wege zu Johann Sebastian Bach? Kalligraphische und
graphologische Untersuchungen zu Bachs Orgelbüchlein’, Österreichisches
Orgelforum, no. 2, pp. 3–22
Schneider, M. 1911. ‘Das sogenannte Orgelkonzert d-moll von Wilhelm
Friedemann Bach’, BJ 8, pp. 23–36
1915. ‘Der Generalbass J. S. Bachs’, Jahrbuch der Musikbibliothek Peters 1914
(Leipzig), pp. 27–42
Schneiderheinze, A. 1987. ‘Zu den aufführungspraktischen Bedingungen in der
Thomaskirche zur Amtszeit Bachs’, BzBf 6, pp. 82–91
Scholes, P. 1948. The Great Dr Burney (London)
Scholes, P. (ed.). 1959. Dr Burney’s musical tours in Europe, 2 vols. (London)
Schöneich, F. 1947/8. ‘Untersuchungen zur Form der Orgelpräludien und Fugen
des jungen Bach’ (dissertation, Göttingen)
Schrammek, W. 1954. ‘Die Musikgeschichtliche Stellung der Orgeltriosonaten von
Joh. Seb. Bach’, BJ 41, pp. 7–28
1975. ‘Johann Sebastian Bach, Gottfried Silbermann, und die französische
Orgelkunst’, Bach-Studien 5 (Leipzig), pp. 93–107
1982. ‘Versuch über Johann Sebastian Bachs Vorstellung von Orgelbau,
Orgeldisposition und Orgelregistrierung’, Bach-Studien 7, pp. 192–211
1983. ‘Zur Geschichte der Orgel in der Thomaskirche zu Leipzig von 1601 bis
1885’, BzBf 2, pp. 46–55
1985. ‘Johann Sebastian Bachs Stellung zu Orgelpedalregistern im 32-Fuß-Ton’,
BJ 71, pp. 147–54
1988. ‘Orgel, Positiv, Clavicymbel und Glocken der Schlosskirche zu Weimar
1658 bis 1774’, in Hoffmann 1988, pp. 99–111
Schreyer, J. 1911–13. Beiträge zur Bach-Kritik, 2 vols. (Leipzig), II pp. 34–6
Schulenberg, D. 1992. The keyboard music of J. S. Bach (New York)
Schulze, H.-J. 1966. ‘Wer intavoliert Johann Sebastian Bachs
Lautenkompositionen?’, Mf 19, pp. 32–9
1968. ‘Johann Sebastian Bach und Christian Gottlob Meissner’, BJ 54, pp. 80–8
1970. ‘Beiträge zur Bach-Quellenforschung’, in Bericht über den internationalen
musikwissenschaftlichen Kongress Leipzig 1966, pp. 269–75
604 Bibliography
1971. ‘Das c-moll-Trio Bwv 585 – eine Orgeltranskription Johann Sebastian

Bachs?’, Deutsches Jahrbuch der Musikwissenschaft 16, pp. 150–5
1972. ‘J. S. Bach’s concerto-arrangements for organ – studies or commissioned
works?’, The Organ Yearbook 3 (1972), pp. 4–13
1974. ‘Wie entstand die Bach-Sammlung Mempell Preller?’, BJ 60, pp. 104–22
1978. ‘Johann Sebastian Bachs Konzertbearbeitungen nach Vivaldi und
Anderen – Studien oder Auftragswerke?’, Deutsches Jahrbuch der
Musikwissenschaft für 1973–1977 18 (Leipzig), pp. 80–100
1984. Studien zur Bach-Überlieferung im 18. Jahrhundert (Leipzig/Dresden)
1985. ‘Johann Christoph Bach (1671 bis 1721), “Organist und Schul Collega in
Ohrdruf ”, Johann Sebastian Bachs erster Lehrer’, BJ 71, pp. 55–81
1991. ‘Bach und Buxtehude. Eine wenig beachtete Quelle in der Carnegie
Library zu Pittsburgh/PA’, BJ 77, pp. 177–81
Schulze, H.-J. (ed.). 1974. J. F. Fasch: Sonate C-moll für zwei Violinen und
Cembalo = EP 9041 (Leipzig)
1977. Katalog der Sammlung Manfred Gorke (Leipzig)
Schweitzer, A. 1905. J. S. Bach le Musicien-Poète (Leipzig)
Seiffert, M. 1905. ‘Neue Bach-Funde’, Jahrbuch der Musikbibliothek Peters 1904
(Leipzig), pp. 15–25
1907. Letter to editor, BJ 4, p. 180
1920. ‘Das Plauener Orgelbuch von 1708’, AfMw 2, pp. 371–93
1921. ‘Seb. Bach’s Bewerbung um die Organistenstelle an St. Jakobi in Hamburg
1720’, AfMw 3, pp. 123–7
Senn, K.-W. 1964. ‘Über die musikalischen Beziehungen zwischen Johann
Gottfried Walther und Johann Sebastian Bach’, MuK 34, pp. 8–18
Siegele, U. 1975. Kompositionsweise und Bearbeitungstechnik in der
Instrumentalmusik Johann Sebastian Bachs (Stuttgart)
1992. ‘Bach’s theological concept of form and the F major Duet’, Music Analysis
11/2–3, pp. 245–78
Sietz, R. 1935. ‘Die Orgelkompositionen des Schülerkreises um Johann Sebastian
Bach’, BJ 32, pp. 33–96
Silbiger, A. 2001. ‘Bach and the chaconne’, The Journal of Musicology 17, pp. 358–85
Smend, F. 1969. Bach-Studien: Gesammelte Reden und Aufsätze (Kassel)
Snyder, K. J. 1987. Dietrich Buxtehude: organist in Lübeck (New York)
Souchay, M.-A. 1927. ‘Das Thema in der Fuge Bachs’, BJ 24, pp. 1–102
Spitta, P. 1873, 1880. Johann Sebastian Bach, 2 vols. (Leipzig)
Stapel, W. 1950. Luthers Lieder und Gedichte (Stuttgart)
Stauffer, G. B. 1980. The Organ Preludes of Johann Sebastian Bach (Ann Arbor)
1983. ‘Bach’s Pastorale in F: a closer look at a maligned work’, The Organ
Yearbook 14 (1983), pp. 44–60
Stauffer, G. B. and E. D. May (eds.). 1986. J. S. Bach as organist: his instruments,
music and performance practice (Bloomington)
Steglich, R. 1935. Johann Sebastian Bach (Potsdam)
1962. Tanzrhythmen in der Musik Johann Sebastian Bachs (Wolfenbüttel/Zurich)
Stiehl, H. 1984. Das Innere der Thomaskirche zur Amtszeit Johann Sebastian
Bachs = BzBf 3
605 Bibliography
Stiller, G. 1970. Johann Sebastian Bach und das Leipziger gottesdienstliche Leben
seiner Zeit (Berlin)
Stinson, R. 1985. ‘Bach’s earliest autograph’, MQ 71, pp. 235–63
1989. ‘Three organ-trio transcriptions from the Bach circle: keys to a lost Bach
chamber work’, in D. Franklin (ed.), Bach Studies (Cambridge), pp. 125–59
1990. The Bach manuscripts of Johann Peter Kellner and his circle. A case study in
reception history (Durham, NC)
1992. Keyboard transcriptions from the Bach circle (Madison)
1993. ‘Some thoughts on Bach’s Neumeister Chorales’, Journal of Musicology 11,
pp. 455–77
1995. ‘The compositional history of Bach’s Orgelbüchlein reconsidered’, in
Stinson (ed.), Bach Perspectives 1 (Lincoln, NE), pp. 43–78
1996. Bach: the Orgelbüchlein (New York)
2001. J. S. Bach’s Great Eighteen Organ Chorales (New York)
Swale, D. 1985. ‘Bach’s Fugue after Legrenzi’, MT 1713, pp. 687–9
Taesler, W. M. 1969. ‘Von Zusammenhang in einigen zyklischen Orgelwerken
Johann Sebastian Bachs – Beobachtungen eines Orgel-spielers’, MuK 39,
pp. 184–7
Tagliavini, L. F. 1969. ‘Johann Gottfried Walther transcrittore’, Analecta
Musicologica 7, pp. 112–19
1986. ‘Bach’s organ transcription of Vivaldi’s “Grosse Mogul” Concerto’, in
Tatlow, R. 1991. Bach and the riddle of the number alphabet (Cambridge)
Terry, C. S. 1917. Bach’s chorals, II: the hymns and hymn melodies of the cantatas and
motets (Cambridge)
1921. Bach’s chorals, III: the hymns and hymn melodies of the organ works
(Cambridge)
Tessmer, M. 1969. ‘Von Zusammenhang in einigen zyklischen Orgelwerken Bachs’,
MuK 39, pp. 184ff.
1974. Dritter Teil der Klavierübung = NBA IV/4 KB (Kassel)
Theobald, H. W. 1986. ‘Zur Geschichte der 1746 von Johann Sebastian Bach
geprüften Johann-Scheibe-Orgel in Zschortau bei Leipzig’, BJ 72, pp. 81–91
Tittel, K. 1966. ‘Welche unter J. S. Bachs Namen geführten Orgel-werke sind
Johann Tobias bzw. Johann Ludwig Krebs zuzuschreiben?’, BJ 52, pp. 102–37
Tittel, K. (ed.). 1962. J. L. Krebs: Ausgewählte Orgelwerke 1 (Cologne)
Trautmann, C. 1984. ‘Bach’s Clavierübung III’, BACH [Riemenschneider
Institute] 15.3, pp. 3–11
Trost, J. C. 1677. Ausführliche Beschreibung des neuen Orgelwercks auf der
Augustus-Burg zu Weissenfels (Nuremberg)
Trumpf, G. A. 1963. ‘Der Rahmen zu Bachs Dritten Teil der Klavierübung’, NZfM
124, pp. 466–70
Tutino, L. 1987. ‘La fuga per organo in si minore “über ein Thema von Corelli”.
Modalità di lettura di un modello italiano’, in Osthoff and Wiesend 1987,
pp. 61–89
Vier, P. 1990. ‘Die Stimmweise der alten Meister’, Acta organologica 21,
pp. 103–20
606 Bibliography
Vogel, H. 1998. ‘ “Acht kleine Präludien und Fugen” von Johann Sebastian Bach’,
MuK 68, pp. 274–5
Vogelsänger, S. 1972a. ‘Zur Architektonik der Passacaglia J. S. Bachs’, Mf 25,
pp. 40–50
1972b. ‘Zur Herkunft der kontrapunktischen Motive in J. S. Bachs
Orgelbüchlein (BWV 599–644)’, BJ 58, pp. 118–31
Vopelius, G. 1682. Neu Leipziger Gesangbuch (Leipzig)
Wagner, G. 1987. ‘Praeludium und Fugue in D-dur von Johann Sebastian Bach
(BWV 532)’, MuK 57, pp. 24–30
Walker, P. 1989. ‘Die Entstehung der Permutationsfuge’, BJ 75, pp. 21–41
Walther, J. G. 1955. Praecepta der musicalischen Composition (1708), ed. P. Benary
(Leipzig)
Weinberger, G. (ed.). 1986. Johann Ludwig Krebs Sämtliche Orgelwerke
(Wiesbaden), preface
Weismann, W. 1949–50. ‘Das grosse Vater-unser-Vorspiel in Bachs dritten Teil der
Klavierübung’, BJ 38, pp. 57–64
Wenke, W. 1988. ‘Die Orgel Johann Sebastian Bachs in Arnstadt – Erkenntnisse
anlässlich der Restaurierung und Rekonstruktion des Spieltisches der
Wender-Orgel von 1703’, in Hoffmann 1988, pp. 93–7
Williams, P. 1968. ‘The harpsichord acciaccatura: theory and practice in harmony,
1650–1750’, MQ 54, pp. 503–23
1972. Bach organ music (London)
1981a. ‘The musical aims of J. S. Bach’s Clavierübung III’, in I. Bent (ed.), Source
materials and the interpretation of music: a memorial volume to Thurston Dart
(London), pp. 259–78
1981b. ‘A Toccata and Fugue in D minor for organ by J. S. Bach?’, EM 9,
pp. 330–7
1982, 1986. ‘J. S. Bach – Orgelsachverständiger unter dem Einfluss Andreas
Werckmeisters?’, BJ 68, pp. 131–42, and BJ 72, pp. 123–5
1984. The organ music of J. S. Bach, III A background (Cambridge)
1986. ‘The snares and delusions of notation: Bach’s early organ works’, in
1997. The chromatic fourth during four centuries of music (Oxford)
1999. ‘Some thoughts on Italian elements in certain music of Johann Sebastian
Bach’, Recercare 11, pp. 185–200
2000. ‘Further on changing manuals in works of J. S. Bach’, The Organ Yearbook
29, pp. 59–74
2001. ‘Witting and unwitting allusion in certain keyboard music of J. S. Bach’,
MQ 84, pp. 757–85
Williams, P. (ed.). 1985. Bach, Handel, Scarlatti. Tercentenary essays (Cambridge)
Witt, C. F. 1715. Psalmodia sacra . . . Neues Cantional (Leipzig/Gotha)
Wolff, C. 1968. Der Stile antico in der Musik Johann Sebastian Bachs (Wiesbaden)
1991. Bach: essays on his life and work [collected writings] (Cambridge, MA)
1992. ‘Aspekte und Probleme des Bachschen Frühwerkes unter besonderer
Berücksichtigung der Orgelchoräle der Neumeister-Sammlung’, in Fuchs 1992,
pp. 237–49
607 Bibliography
1997. ‘Zum Quellenwert der Neumeister-Sammlung: Bachs Orgelchoral

“Der Tag der ist so freudenreich” BWV 719’, BJ 83, pp. 155–67
2000. Johann Sebastian Bach: the learned musician (Oxford/New York)
Wolff, C. (ed.). 1985. Johann Sebastian Bach, Orgelchoräle der
Neumeister-Sammlung (Yale University, MS LM 4708) (Kassel)
1986. The Neumeister Collection of chorale preludes from the Bach circle. Facsimile
edition of the Yale Manuscript LM 4708 (New Haven)
Wollny, P. 1994. ‘Bachs Bewerbung um die Organistenstelle an der Marienkirche
zu Halle und ihr Kontext’, BJ 80, pp. 25–40
1995. ‘Wilhelm Friedemann Bach’s Halle performances of cantatas by his
father’, in Melamed 1995, pp. 202–28
1997. ‘Neue Bach-Funde’, BJ 83, pp. 7–50
Wollny, P. (ed.). 1999. Johann Sebastian Bach. Die Achtzehn grossen Orgelchoräle
BWV 651–668 und Canonische Veränderungen BWV 769. Facsimile der
Originalhandschrift (Laaber)
Wollny, P. and J.-C. Zehnder (eds.). 1998. Aus dem Umkreis des jungen Johann
Sebastian Bach. Neunzehn Orgelchoräle (Stuttgart)
Wörthmüller, W. (ed.). 1973. Die Nürnberger Musikverleger und die Familie Bach
(Nuremberg)
Wustmann, R. 1911. ‘Tonartensymbolik zu Bachs Zeit’, BJ 8, pp. 60–74
1926. Musikgeschichte Leipzigs in drei Bänden. Erster Band: bis zur Mitte des 17.
Jahrhunderts (Leipzig, 1926)
Zacher, G. 1981. ‘Canonische Veränderungen BWV 769 und 769a’, Musik-Konzepte
17/18, pp. 3–19
1993a. ‘Die Form der g-moll-Fantasie (BWV 542a [ = 542.i]) für Orgel’,
Musik-Konzepte 79/80, pp. 20–39
1993b. ‘Bach gegen seine Interpreten verteidigt’, Musik-Konzepte 79/80,
pp. 85–105
Zahn, D. 1985. ‘J. S. Bachs Präludium und Fuge in h-moll (BWV 544)’, MuK 55,
pp. 63–73
Zahn, J. 1889–93. Die Melodien der deutschen evangelischen Kirchenlieder aus den
Quellen geschöpft und mitgeteilt, 6 vols. (Gütersloh)
Zehnder, J.-C. 1987. ‘Die Weimarer Orgelmusik Johann Sebastian Bachs im
Spiegel seiner Kantaten’, Musik und Gottesdienst 41, pp. 149–62
1988. ‘Georg Böhm und Johann Sebastian Bach. Zur Chronologie der
Bachschen Stilentwicklung’, BJ 74, pp. 73–110
1991. ‘Giuseppe Torelli und Johann Sebastian Bach. Zu Bachs Weimarer
Konzertform’, BJ 77, pp. 33–95
1995. ‘Zu Bachs Stilentwicklung in der Mühlhäuser und Weimarer Zeit’, in
Heller & Schulze 1995, pp. 311–38
1998. ‘J. A. L. – ein Organist im Umkreis des jungen Bach’, Basler Jahrbuch für
historische Musikpraxis 22 (1998), pp. 127–55
Zietz, H. 1969. Quellenkritische Untersuchungen an den Bach Handschriften P 801,
P 802 und P 803 (Hamburg)
Ziller, E. 1935. Der Erfurter Organist Johann Heinrich Buttstedt (1666–1727) =
Beiträge zur Musikforschung 3 (Halle)
Index of names
An organist’s appointments (‘appts’) include his last major post. ‘Pupil of J. S. Bach’ indicates those
owing their musical education largely or partly to the composer, including choristers of St Thomas,
Leipzig (Thomaner). ‘Author’ indicates a twentieth-century author.
Abel, C. F., 1723–87, son of a Cöthen gambist, as composer, 10, 175, 182, 190, 225–6, 490,
from 1759 in London, 106 579, 580
Abraham, G., author, 317 as a source of copies etc., 19, 40, 45, 64, 74,
Adlung, J., 1699–1762, from 1728 organist at 81, 85, 92, 96, 101, 105, 107, 111, 124, 127,
Erfurt Predigerkirche, 390, 461 146, 159, 166, 182, 189, 192, 196, 206, 230,
Agricola, J., 1492–1566, hymn-writer, 307 470–1
Agricola, J. F., 1720–74, pupil of J. S. Bach Bach, E. J. F., 1726–81, fourth daughter of
1738–41, court composer at Berlin from J. S., in 1749 married J. C. Altnickol, 338,
1751, 254, 416, 455 382
as copyist, 3, 74, 85, 146, 206, 209, 213 Bach, J. Andreas, 1713–79, fifth son of J. C. Bach
Ahle, J. R., 1625–73, hymn-compiler, 412 of Ohrdruf and brother of J. B. Bach
Alberus, E., ?–1553, hymn-writer, 500 (below)
Albinoni, T., 1671–1750, Venetian composer, ABB, 92, 149, 153, 160, 164, 168, 174, 177,
81, 180, 201 178, 182–4, 186–7, 466, 579
see also BWV 946, 950, 951, 951a Bach, J. B., 1670–1749, organist from 1703 in
Albinus, J. G., 1624–79, hymn-writer, 313 Eisenach, 205, 449, 509
Albrecht, C., author, 391, 530 Bach, J. C., 1673–1727, ‘Bach of Gehren’, 389,
Altni(c)kol, J. C., 1719–59, pupil (from 1744) 539
and son-in-law (from 1749) of J. S. Bach, Bach, J. C., 1642–1703, cousin of J. S.’s father,
231, 336, 338, 348, 379, 380–1, 383 organist from 1665 at Eisenach
Ambrose, Saint, 238 Georgenkirche, 79, 234, 429, 459, 492, 496,
Ammerbach, E. N., c. 1530–97, organist from 512, 566, 583
1561 at the Thomaskirche, 227, 310, 390 Bach, J. C., 1671–1721, elder brother of J. S.
d’Andrieu, see Dandrieu (who lived with him from 1695), pupil of
d’Anglebert, J. H., 1635?–91, Parisian composer, Pachelbel, organist from 1690 in Ohrdruf,
170 583
Anna Amalia, 1723–87, Princess of Prussia and as copyist, 37, 41, 125, 130, 149, 164, 172,
sister of Frederick II, 3 174, 177, 178, 182, 192, 466, 484, 541,
Anton, K., author, 191 575
Apel, W., author, 161, 588 Bach, J. C., 1735–82, son of J. S., from 1762 in
Arfken, E., author, 239, 250, 264, 272, 277, 295, London, 74, 176, 491, 576
305 Bach, J. C. F., 1732–95, son of J. S., 1758
Argent, M., author, 158 capellmeister at Bückeburg, 231, 288
Armsdorff, A., 1670–99, organist in Erfurt, 6, Bach, J. C. G., 1747–1814, organist in Ohrdruf?,
454 141
Augustine, Saint, 547 Bach, J. E., 1705–55, grandson of J. S.’s uncle,
1736–42 secretary to J. S., 387, 529
Bach, A. M., 1701–60, second wife of J. S. Bach, J. M., 1648–1694, J. S.’s first father-in-law,
(1721), 171 organist in Gehren, 234, 326, 439, 459, 466,
AMBB, 59, 103, 112, 171, 225, 431, 470–1, 477, 488, 492, 541–3, 582
541–3, 567, 575 Bach, M. B., 1684–1720, first wife of J. S., 541
as copyist, 3, 71 Bach, W. F., 1710–84, second child of J. S.,
Bach, C. P. E., 1714–88, fifth child of J. S., 1740 organist at Dresden Sophienkirche 1733,
musician to Frederick II at Potsdam, 1767 Halle Liebfrauenkirche 1746, from 1774 in
succeeded Telemann at the Hamburg Berlin, 3, 4, 5, 6, 83, 85, 171, 220, 223, 227,
Johanneum, 90, 133, 183, 231, 317, 327, 234, 317, 318, 319, 355, 388, 584
[608] 358, 383, 576, 587 as composer, 5, 141, 145, 159
609 Index of names
as copyist, 81, 105, 179, 213, 230, 288 Brockaw, J. A., author, 174
CfWFB, 156, 171, 225, 431, 469, 471, 493, Brossard, S. de, c. 1654–1730, appts in Paris
529, 567 and Strasbourg, 167, 260, 529, 589
Bapst (Babst), V., hymn-compiler (pub. 1553), Bruggaier, E., author, 135, 350
239, 243, 550, 554, 557, 563 Bruhns, N., 1665–97, pupil of Buxtehude
Bartels, U., author and NBA editor, 1, 145, 166, (1681), organist at Husum Cathedral
181, 199, 200, 225 from 1689
Bassani, G. B., c. 1657–1716, maestro di chorales, 46, 55, 83, 285, 345, 462
cappella at Ferrara, 412 free works, 28, 38–40, 43, 65, 121, 157, 161,
Becker, J., 1726–1804, copyist in Kittel circle, 163, 181
118 influence of, 41, 77, 126, 128–9, 131, 173,
Becker, N., organ-builder in Mühlhausen 205, 422
(first half of eighteenth century), 584 Buchmayer, R., author, 184
Bedos de Celles, Dom J. F., 1706–79, French Budday, W., author, 305
organ-builder, 587 Bull, J., 1562/3–1628, 133, 548
Beechey, G., author, 92, 96, 141 Bullivant, R., author, 39, 70, 138, 155
Beethoven, L. van, 44 Burba, O.-J., author, 430, 440
Beisswenger, K., author, 57, 139, 198, 201, 211 Burguéte, A., author, 71
Bernhard, C., 1627–92, ‘pupil’ of Schütz, Burney, C., 1726–1814, 202
from 1681 capellmeister at Dresden, 265, Busbetzky, L., ?–1699, pupil of Buxtehude and
589 Flor, from 1687 organist in Narva
Besseler, H., author, 56, 79 (Estonia), 463–4
Beyer, J. S., 1668–1744, cantor of Freiberg Butler, G. G., author, 139, 387, 389, 390, 468,
Cathedral from 1718, 235 514, 531
Biffi, A., 1666/7–1733, Venetian composer, 175 Butt, J., author, 3, 4, 15, 17, 308, 421
Bighley, M., author, 319, 545, 556, 572 Buttstedt (Buttstett), J. H., 1666–1727, pupil of
Billeter, B., author, 157–8 Buxtehude, from 1684 organist at various
Birnbaum, J. A., 1702–48, from 1721 teacher Erfurt churches, 199
(Dozent) of Rhetoric in Leipzig, 66, 390 chorales, 234, 242, 244, 436, 499, 544, 554
Blume, F., author free works, 43, 95, 149, 151, 155, 167, 179,
Böhm, G., 1661–1733, from 1698 organist of 205, 580
Lüneburg Johanniskirche, 155, 573, 583 influence of, 127, 154, 156
chorales, 244, 267, 285, 314, 316, 322, 324, Buxtehude, D., 1637–1707, in 1668 succeeded
345, 363, 377, 416, 418, 421, 430, 436, Tunder as organist of Lübeck
457–8, 464, 480, 485, 495, 499–527, 540, Marienkirche, 33, 122, 366, 464, 575,
589 583
free works, 37, 38, 42, 53, 126, 145, 150, 153, chorales, 6, 65, 236, 237, 252, 267, 269, 279,
177, 183, 575–6, 580 287, 292, 305, 308, 311, 345, 351, 363, 390,
influence of, 43, 46, 131, 156, 163, 341, 365, 430, 436, 438, 450, 458, 474, 485, 498,
439, 457, 489, 496, 544, 568, 588 499–526, 544, 569
Böschenstein, J., 1472–1530, hymnwriter, 277 free works, 25, 37–40, 42, 46, 47, 57, 61, 91,
Böttiger, J., 1613–72, hymn-writer, 507 93, 115, 120, 124, 126–7, 139, 145, 154, 155,
Bonnus, H., hymn-writer, 551 156, 160, 161, 162, 164, 181, 183, 184, 185,
Bourgeois, L., c. 1510–c. 1561, hymn-writer, 310 186–7, 193, 198, 260–95, 486, 513, 537,
Boyvin, J., c. 1653–1706, from 1674 organist at 568
Rouen, 6, 136, 167, 168, 345, 513, 529, influence of, 42, 43, 47, 77, 83, 92, 128–33,
586 143, 154, 155, 161, 166, 173, 177, 194, 205,
Boxberg, C. L., 1670–1729, student of N. A. 341, 355, 365, 422, 485
Strunk in Leipzig, organist at Görlitz from
1702, 464 Caldara, A., c. 1670–1736, Venetian composer,
Braun, H., author, 181 1716 court appt in Vienna, 393
Breig, W., author, 40, 41, 45, 48, 54, 65, 75, 82, Calvisius, S., 1556–1615, cantor of St Thomas,
92, 104, 110, 125, 130, 134, 140, 149, 154, Leipzig from 1594, 331, 449–50, 554,
164, 194, 195, 231, 343, 344, 349, 405, 422, 562
544, 547, 563, 566 Camerarius, J., hymn-writer (pub. 1546), 310
Breitkopf, Leipzig publisher, 158, 230, 429, 514, Chailley, J., author, 140, 239, 242, 244, 245, 246,
577 265, 269, 271, 284, 305, 366, 397, 421, 426,
see also BWV 690–713 444, 466, 472, 531
610 Index of names
Chambonnières, J. C. de, 1601/2–72, 198 Dürr, A., author and NBA editor, 141, 190, 234,
Christiane Eberhardine, 1671–1727, Electress of 323, 327, 338, 484, 489, 543, 563
Saxony, 97
Clark, R., author, 259, 284, 289 Eber, P., 1511–69, hymn-writer, 262, 310
Claus, R.-D., author, 156 Ebert, J., hymn-writer, 555
Clausnitzer, T., 1618–84, hymn-writer, 298, 486 Eck, C. L. van, author, 308, 531
Clement, A., author, 392, 507, 515, 529, 530 Edler, A., author, 44, 254
Clementi, M., 1752–1832, 175 Ehricht, K., author, 530
Clérambault, L.-N., 1676–1749, Parisian Eickhoff, H. J., author, 418
organist, 6, 148 Eller, R., author, 204, 209
Compenius, E., c. 1600, organ-builder in Emans, R., author and NBA editor, 230, 355,
Magdeburg, 464 432, 435, 441, 442, 456, 478, 487, 490, 491,
Cooke, B., Sr, 1734–93, pupil of Pepusch, 1757 493, 494, 495, 497, 537, 542, 549, 550, 577,
appt Westminster Abbey, 105, 106, 107, 578, 580–1
538 Emery, W., author and Novello editor, 3, 4,
Cooke, B., Jr, 1761–72, son of preceding, 105 6, 7, 14, 19, 22, 23, 24, 26, 28, 83, 95,
Corelli, A., 1653–1713 102, 106, 125, 141, 142, 150, 317, 515,
works, 31, 73, 88, 166, 176, 179–81, 195, 525, 538
197–8, 455, 464 Eppstein, H., author, 4, 14, 17, 20, 22, 23, 27, 33,
influence of, 25, 42, 173, 194, 375, 439, 526 538
see also BWV 579 Ernesti, J. A., 1707–81, from 1734 Rektor
Couperin, F. (‘le Grand’), 1668–1733, 59, 133, Thomasschule, 416
159, 167, 192, 390, 443, 552, 587
see also BWV 587 Fasch, J. F., 1688–1758, pupil of Kuhnau, from
Couperin, L., c. 1626–61, late uncle of 1722 capellmeister in Zerbst, 8, 191
F. Couperin, 198, 240 see also BWV 585
Cruciger, E., ?–1535, hymn-writer, 242, 547 Figulus, W., c. 1520–91, hymn-writer, 263
Crüger, J., 1598–1662, from 1662 cantor of Finke, J. G., organ-builder in Saalfeld (first half
Berlin Nikolaikirche, 256, 351, 359, 453, of eighteenth century), 584
470, 561 Finke-Hecklinger, D., author, 271
Cuncius, C., 1676–1722, organ-builder, 583 Fischer, C., c. 1520–97, hymn-writer, 281
Currie, R. N., author, 319 Fischer, J. K. F., c. 1670–1746, court musician in
Czubatynski, U., author, 455 Baden, appts from 1692 or earlier, 57, 65,
77, 103, 112, 126, 142, 143, 149, 163, 169,
Dadelsen, G. von, author and NBA editor, 2, 195, 199, 240, 278, 341, 395, 425, 430, 464,
220, 226, 227, 230, 231, 263, 312, 367 491, 525
Dähnert, U., author, 387 Fischer, M. G., 1773–1829, pupil of Kittel, appts
Dalton, J., author, 80 in Erfurt, 64
Dandrieu (d’Andrieu), J.-F., 1682–1738, Flittner, J., 1644–78, hymn-writer, 524
Parisian composer, 183 Fock, G., author, 86
Daquin (d’Aquin), L.-C., 1694–1772, Parisian Forkel, J. N., 1749–1818, university organist at
composer, 59, 361 Göttingen from 1770, 3, 4, 5, 70, 81, 105,
David, H. T., author, 352, 391 133, 141, 183, 202, 203, 321, 383, 384, 467,
David, W., author, 41 503
Decius, N., ?–1541, hymn-writer, 271, 370 Franck, J., 1618–77, hymn-writer, 256,
Derr, E., author, 34 351
Dickinson, A. E. F., author, 62 Franck, M., 1609–67, hymn-writer, 314
Dietrich, F., author, 41, 87, 142, 143, 168, 345, Franck, Melchior, c. 1579–1639, 548
407, 426, 457, 481, 495 Franck, S., 1659–1725, Secretary Weimar Court
Dirksen, P., author, 56, 485 from 1701, 235
Doles, J. F., 1715–97, pupil of J. S. Bach, from Franklin, D., author, 130
1755 cantor of St Thomas, Leipzig, 40, 321 Frescobaldi, G., 1583–1643, organist at
Dröbs, J. A., 1784–1825, pupil of Kittel, music St Peter’s, Rome from 1608, 162, 185,
teacher in Leipzig from 1808, 48, 92, 96, 366
101, 119 influence of, 72, 131, 132, 161, 165, 193, 194,
Du Mage, P., 1674–1751, 1703–10 organist in 197, 589
St-Quentin, 136, 167, 345, 350, 529, 586, Fiori musicali, 1, 137, 154, 193, 194, 230, 260,
587 389, 394, 395, 399, 430
611 Index of names
Freylinghausen, J. A., 1670–1739, hymn-writer, Guilmant, F.-A., 1837–1911, organist of

pastor of Halle Ulrichskirche, 265, 270, the Trinité, Paris, 183
283, 303, 309, 313, 333, 347, 445, 463, 468, Gwinner, V., author, 157
509
Froberger, J. J., 1616–67, son of a cantor in Halle, Hammerschlag, J., author, 88
pupil of Frescobaldi, appts in Vienna, Handel, G. F., 1685–1759, pupil of Zachow, 106,
London, etc., 162, 194, 240, 281, 505 138, 211, 542
Frotscher, G., author, 40, 120, 130, 143, 145, works, 5, 25, 28, 42, 77, 83, 104, 130, 133,
173, 179, 492 139, 150, 155, 174, 176, 194, 197, 198, 208,
Fuger, C., (father and son?), hymn-writers 425, 526, 576
(pub. 1586?, 1593), 260 Hanff, J. N., 1665–1711, Thuringian organist
Fux, J. J., 1660–1741, court composer at with appts in Hamburg etc., 509
Vienna from 1698, 138, 387, 393, 396, 534, Hartmann, G., author, 550
585 Hassler, H. L., 1564–1612, appt in Dresden,
469
Gasparini, F., 1668–1727, pupil of Corelli, appts Hauser, F., 1794–1870, director of Munich
in Venice, 170 Conservatory, 176, 499
Gastoldi, G. G., ?–1622, Mantuan composer, Hawkins, Sir John, 1719–89, 106
266, 267 Haydn, J., 140
Geck, M., author, 163, 464 Heder, S. G., 1713–?, copyist in P 803, 150
Geiringer, K., author, 186–7 Heermann, J., 1585–1647, hymn-writer, 326,
Geminiani, F., 1687–1762, pupil of Corelli, 503, 547, 561
violinist in London, 585 Hegenwal(d)t, J., hymn-writer (pub. 1524), 463
Gerber, E. L., 1746–1819, son of H. N. Gerber, Heidorn (Heydorn), R., pupil of Reinken?,
court organist at Sondershausen from appts in or near Hamburg, 53, 95, 139,
1775, author of Lexicon, 225 172
Gerber, H. N., 1702–75, at University of Leipzig Heinemann, M., author, 552
from 1724, pupil of J. S. Bach, appts at Heinichen, J. D., 1683–1729, pupil of Kuhnau,
Sondershausen from 1731, 5, 166, 193, 1717 court capellmeister at Dresden, 170,
225, 355, 372 199, 281, 586
Gerhardt, P., 1607–76, hymn-writer, 347 Helder, B., ?–1635, hymn-writer, 441
Gerlach, C. G., 1704–61, pupil of Kuhnau, 1729 Helmbold, L., 1532–98, hymn-writer, 360
organist at Leipzig Nikolaikirche, 37, Herberger, V., 1562–1627, hymn-writer, 479
446 Hering, H., author, 92
Gesius, B., 1555?–1613/14, hymn-writer, 556 Herman, N., ?–1561, hymn-writer, 255, 291
Gilbert, K., author, 559 Herz, G., author, 417
Gojowy, D., author, 234, 250, 262, 323, 326, 351, Heyden, S., 1491–1561, hymn-writer, 279
467, 503 Hildebrandt, Z., 1688–1757, organ-builder in
Goldberg, J. G., 1727–56, pupil of J. S. and/or Dresden, 584
W. F. Bach, 5 Hilgenfeldt, C. L., nineteenth-century author,
Goldhan, W., author, 2 391
Grace, H., author, 135, 205, 301, 378 Hill, R., author, 54, 149, 172, 186–7, 466, 575
Grasnick, F. A., 1798–1877, copyist in Berlin, Hodenberg, B. von, hymn-writer, 382
141 Hoffmann-Erbrecht, L., author, 153
Graupner, C., 1683–1760, a Leipziger, appts in Hofmann, K., author and NBA editor, 10
Darmstadt from 1709, 195 Holschneider, A., author, 19
Greitter, M., ?–c. 1550, hymn-writer, 279 Homburg, E. C., 1605–81, hymn-writer, 560
Griepenkerl, F. C., 1782–1849, pupil of Homilius, G. A., 1714–85, pupil of J. S. Bach,
Forkel, Peters editor, 41, 60, 107, 192, 220, from 1742 appts in Dresden, 127, 321, 495
223 Honders, C., author, 234, 257, 273
Grigny, N. de, 1672–1702, from 1695 organist of Horn, C. F., 1762–1830, Saxon composer,
Rheims Cathedral, 1, 61, 147, 150, 167, from 1782 in London, 3
172, 232, 259, 274, 345, 349, 390, 409, 435, Horn, V., author, 134, 147
525, 529, 586 Hubert, K., hymn-writer, 554
Grimm, J., author, 507 Humphreys, D., author, 48, 50, 57, 58, 92, 134,
Grüss, H., author, 334 158, 390, 392, 529
Guhr, C. W. F., 1787–1848, capellmeister in Hurlebusch, C. F., 1696–1765, appts from 1743
Frankfurt am Main, 171, 182, 478 in Amsterdam, 139, 388, 390
612 Index of names
Hus, J., 1371–1415, 377 Kittel, J. C., 1732–1809, pupil of Bach 1748–50,
Hutton, J., 1715–95, English visitor to Leipzig, organist at various Erfurt churches from
106 1756, 48, 51, 64, 141, 542, 581
as copyist etc., 1, 3, 45, 51, 64, 74, 81, 85, 92,
Jacob, A., author, 405, 414, 426 95, 96, 101, 107, 111, 118, 124, 127, 145,
Jauernig, R., author, 236 146, 159, 163, 166, 174, 178, 179, 182, 192,
Johann Ernst, 1696–1715, Prince of 196, 206, 209, 220, 230–314, 336–79,
Sachsen-Weimar, 202–3, 206–8, 209–10, 429–47, 457, 464, 465, 470–2, 476, 480,
218–19 489, 495
see also BWV 592, 595, 982–7 Klein, K.-G., author, 114, 153
Kloppers, J., author, 60, 66, 86, 108, 122
Kast, P., author, 495 Klotz, H., author and NBA editor, 5, 40, 42, 44,
Kauffmann, G. F., 1679–1735, pupil of Buttstedt, 129, 186–7, 221, 274, 335, 357, 409, 462,
court organist at Merseburg, 5, 83, 321, 464, 504, 517
328, 341, 364, 366, 369, 388, 390, 403, Klug, J., hymn-writer, 245, 253, 557
407, 420, 421, 431, 449, 482, 487, 517, Knight, D., author, 106
523 Knoll, C., 1563–1650, hymn-writer, 468
Kee, C., author, 186 Kobayashi, Y., author, 118, 146, 150, 163, 178,
Keimann, C., 1607–62, hymn-writer, 507 186–7, 292, 338, 484, 492, 495, 499,
Keller, H., author, 15, 20, 37, 41, 56, 63, 71, 83, 512–13, 515, 539, 575
92, 108, 112, 115, 129, 130, 132, 139, 145, Kollmann, A. F. C., 1756–1829, organist in
147, 148, 152, 159, 162, 176, 186–7, 190, Lüneburg, from 1782 in London, 10,
195, 197, 200, 225, 226, 239, 245, 250, 271, 575
272, 285, 299, 303, 305, 325, 355, 357, 362, Körner, G. W., 1809–65, publisher in Erfurt,
366, 372, 403, 408, 421, 441, 445, 449, 452, author of organ tutors, 48, 191
464, 473, 479, 492, 496, 504, 506, 530, 543, Krause, J., author, 268, 426
580 Kräuter, P. D., 1690–1741, pupil of J. S. Bach,
Kellner, J. P., 1705–72, ?pupil of J. S. Bach and from 1713 appts in Augsburg, 203
of H. F. Quehl (of Suh1), from 1727 Krebs, J. L., 1713–80, son of J. T. Krebs, pupil
appts in Gräfenroda, 130, 155, 158, 166, of J. S. Bach, from 1756 organist in
217–18 Altenburg, 74, 141, 191, 321, 355
as composer etc., 57, 110–11, 145 as composer, 8, 9, 60, 77, 78, 142, 150, 153,
as copyist etc., 3, 29, 37, 51, 54, 56, 64, 74, 81, 190, 284, 313, 350, 394, 435, 442, 447, 454,
85, 92, 95, 96, 101, 107, 111, 118–19, 125, 487, 490, 581
127, 145, 149, 150, 159, 163, 165, 166, 169, as copyist, 5, 60, 62, 64, 74, 141, 163, 164,
170, 171, 179, 196, 197, 205, 209, 214, 220, 178, 324, 381, 429, 430, 457, 468, 474, 476,
230, 454, 455, 528, 537–8 495, 524, 576
Kerl(l), J. C., 1627–93, 1677–84 organist of Krebs, J. T., 1690–1762, father of J. L. Krebs,
St Stephen, Vienna, 53, 77, 132, 157, 390, pupil of J. G. Walther and J. S. Bach
425 (Weimar), from 1721 organist in
Kiel, T., 1584–1626, hymn-writer, 270 Buttstädt, 51, 78, 141, 191,
Kilian, D., author and NBA editor, 2, 4, 71, 72, 576
73, 81, 119, 129, 146, 575 as copyist, 3, 29, 60, 74, 85, 159, 182,
Kindermann, J. E., 1616–55, appts in 230–313, 337–81, 429–44, 453–96, 497,
Nuremberg, 278 503–8, 566, 577, 579
Kinsky, G., author, 5 Krey, J., author, 56, 157, 292
Kircher, A., 1602–80, Jesuit polymath in Rome, Krieger, J., 1652–1735, from 1681 music
305 director at Zittau, 138, 165, 183, 193, 278,
Kirchhoff, G., 1685–1746, succeeded his teacher 389, 531
Zachow at Halle Liebfrauenkirche in 1714, Krüger, E., author, 1, 55
340 Krügner, J. G., c. 1684–1769, Leipzig engraver,
Kirnberger, J. P., 1721–83, pupil of J. S. Bach 390
and J. P. Kellner, 1758 musician to Princess Krumbach, W., author, 429, 439, 485, 489, 541,
Anna Amalia in Berlin, 71, 199, 429 569, 581–2
as copyist etc., 3, 51, 74, 81, 85, 101, 107, 111, Krummacher, F., author, 98, 136, 161, 412,
119, 159, 166, 230–314, 336–80, 474, 487, 464
495, 506, 512 Kube, M., author, 237, 267, 352, 400, 404, 435,
as theorist etc., 68, 114, 393, 400, 407, 414 488, 499, 508
613 Index of names
Kuhnau, J., 1660–1722, predecessor of J. S. Bach Marshall, R. L., author, 330, 575
at St Thomas, Leipzig (organist 1684, Martini, G. B., 1706–84, Bolognese priest,
cantor 1701), 1, 41, 89, 129, 136, 143, 150, composer and theorist, 407, 586
166, 184, 199, 389, 436, 464, 525, 526, 545, Mattheson, J., 1681–1764, cantor of Hamburg
583 Cathedral from 1715, etc., 5, 52, 57, 61, 66,
Kühnel, A., c. 1770–1813, violinist and organist 72, 86, 90, 91, 93, 97, 133, 183, 185, 308,
in Leipzig, 189 340, 357, 389, 444, 469, 509, 529
May, E. D., author, 327, 429, 441
Leaver, R., author, 298, 415, 421, 426, 427, 530 McLean, H., author, 131, 132
Lebègue, N. A., 1630–1702, from 1678 Parisian Meissner, C. G., 1707–60, at Leipzig
court composer, 6, 587 Thomasschule 1719–29, from 1731 cantor
Leclair, J.-M., 1697–1764, Parisian violinist, 124 at Geithain, 173, 182, 230–309
Legrenzi, G., 1626–90, Venetian composer, Melanchthon, P., 1497–1560, 331
172–3, 180, 366 Mempell, J. N., 1713–47, Pupil of J. P. Kellner?,
see also BWV 574, 574a, 574b cantor in Apolda, 19, 40, 41, 127, 130, 131,
Leisinger, U., author, 490 190, 191, 218, 230–300, 358, 454–97, 537
Leon, J., ?–1597, hymn-writer, 445 Mendel, A., author, 352, 391
Leopold, Prince of Anhalt-Köthen, 1694–1729, Mendelssohn-Bartholdy, F., 1809–47, 76, 489
293, 584 Mersenne, M., 1588–1648, Parisian Minorite
Leutert, H., author, 357 and theorist, 587
Lindemann, J., c. 1550–1634, hymn-writer, 266 Mey, W. N., eighteenth-century copyist, pupil of
Locatelli, P., 1695–1764, pupil of Corelli, Kellner?, 536
violinist in Amsterdam etc., 190, 198, 200 Meyer, U., author, 108, 131, 239, 242, 251, 339,
Löffler, H., author, 525 341, 349, 362, 369, 372, 378, 425, 451, 482,
Löhlein, H.-H., author and NBA editor, 234 488, 508, 509, 540
Lohmann, H., author and B&H editor, 45, 138, Mizler, L. C., 1711–78, pupil of J. S. Bach,
392 founded Societät der musikalischen
Lo(h)rbeer, J. A., copyist c. 1700, 480 Wissenschaften 1738, editor of
Lotti, A., c. 1667–1740, Venetian pupil of Musikalische Bibliothek 1736–54, 321, 387,
Legrenzi, from 1717 appts in Dresden, 83 388, 393, 396, 513, 514–15, 524, 534
Lübeck, V., 1654–1740, organist in Stade, then Möller, M., hymn-writer, 488
from 1702 Hamburg Nikolaikirche, 37–9, Monteverdi, C., 589
65, 131, 152, 173, 225 Mozart, W. A., 1756–91, 3, 14, 19, 200
Luedtke, H., author, 239, 327, 355, 377, 433, Muffat, Georg, 1653–1704, studied in Paris,
447, 489, 508 appts in Salzburg, Passau, etc., 164, 184,
Luther, M., 1483–1546, 238, 239, 247, 250, 251, 197, 519
258, 268, 285, 286, 289, 295, 300, 302, 341, Muffat, Gottlieb, 1690–1770, son of the
377, 410, 415, 418, 422, 459, 467, 476, 486, preceding, pupil of Fux, appts in Vienna,
487, 498, 530, 556, 563, 571 121, 137, 341, 430
Müller, H., 1631–75, hymn-writer, 283
Mace, T., c. 1612 – c. 1706, appts in Cambridge, Münzer, M., 16th-cent. hymn-writer, 562
198 Mün(t)zer, T., c. 1489–1525, hymn-writer and
Macque, J. de, 1551–1614, Flemish composer early reformer, 295
working in Rome and Naples, 138 Musch, H., author, 127
Mahrenholz, C., author, 464 Müthel, J. G., 1728–88, student-boarder with
Maichelbeck, F. A., 1702–50, priest and Bach, appts in Schwerin, from 1753 Riga,
composer in Freiburg im Breisgau, 83, 230–313
141
Major, J., hymn-writer, 453 Nachtenhöfer, C. F., 1624–85, hymn-writer, 333
Marais, M., 1656–1728, Parisian composer and Nägeli, H. G., 1773–1836, music dealer in
viol-player, 199, 366 Zurich, 3
Marcello, B., 1686–1739, Venetian composer, Neander, J., 1650–80, hymn-writer, 333
509 Neumark, G., 1621–81, hymn-writer, 312
Marchand, L., 1669–1732, Parisian court Neumann, W., author and NBA editor, 24
composer, 124, 167, 240, 584 Neumeister, J. G., 1756–1840, 541
Marpurg, F. W., 1718–95, secretary in Paris and see also BWV 1090–1120
Hamburg, from 1763 court appt in Berlin, Nicolai, D., 1702–64, from 1730 organist in
181, 282, 287, 328, 361, 385, 516, 534 Görlitz, 127
614 Index of names
Nicolai, P., 1556–1608, hymn-writer, 323, 484 Quantz, J. J., 1697–1773, flautist in Dresden
Niedt, F. E., 1674–1708, lawyer in Jena, appts in from 1718, from 1741 Berlin, 35, 202, 588
Copenhagen from 1704 or earlier, 136, Quehl (Kehl), H. F., organist in Suhl? 1730s?,
143, 186, 237 341, 387, 465
Nivers, G.-G., 1617–1714, Parisian court
musician from 1667 or earlier, 242, 260 Radulescu, M., author, 183, 186–7
Nörmiger, A., c. 1560–1613, from 1581 court Raison, A., ?–1719, from 1666 organist in Paris,
organist in Dresden, 235 6, 55, 65, 148, 167, 172, 180, 183, 184, 187,
Notker, sequence author, 247 529
Rampe, S., author, 166
O’Donnell, J., author, 62 Rein(c)ken, J. A., 1643–1722, in 1663 succeessor
Olearius, J. C., 1611–84, hymn-writer, 240 to his teacher Scheidemann at Hamburg
Oley, J. C., 1738–89, perhaps a pupil of J. S. Katharinenkirche, 86, 90, 348–9, 583,
Bach, organist in Aschersleben from 1762, 584
3, 4, 51, 52, 85, 107, 146, 230–314, 334, as composer, 56, 65, 67, 90, 95, 152, 168, 172,
341–76, 429–50, 455, 474, 476, 487 179, 180, 345, 422
Oppel, R., author, 92, 143, 190 influence of, 1, 43, 154, 485
Ouverkwerk, P., author, 186 see also BWV 965, 966
Overholtzer, H., author, 148 Reusner (Reissner), A., 1496 – c. 1575,
hymn-writer, 309
Pachelbel, J., 1653–1706, pupil of Kerll in Riedel, F. W., author, 5, 183
Vienna, successively organist in Eisenach, Rienäcker, G., author, 92
Erfurt, Stuttgart, Gotha and Nuremberg, Rinck, (J.) C. H., 1770–1856, pupil of Kittel and
385, 541–68 Forkel, from 1805 appts in Darmstadt, 429,
chorales, 237, 242, 278, 289, 308, 337, 341, 485, 499, 580, 581
356, 393, 430, 436, 439, 443, 466, 476, 477, Ringk, J., 1717–78, pupil of J. P. Kellner, from
492, 494–5, 499, 500–27, 529, 537, 578, 1755 organist of the Marienkirche, Berlin,
582, 589 40, 45, 48, 81, 130, 132, 155, 159
free works, 1, 43, 49, 65, 80, 92, 142, 143, 155, Ringwaldt, B., 1530–99, hymn-writer, 476,
164, 183, 184, 185, 187, 197, 281, 441 566
influence of, 43, 77, 79, 149, 163, 164, 337, Rinkart, M., 1586–1649, hymn-writer, 358
341, 365, 369, 375, 380, 430, 439, 456, 457, Rist, J., 1607–67, hymn-writer, 571
482, 485–6, 489 Robinson, J., 1682–1762, pupil of John Blow,
Pachelbel, W. H., 1686–1764, son of preceding, from 1727 organist of Westminster Abbey,
friend of J. G. Walther, appts in Nuremberg 105, 106
from 1706, 37, 40, 42 Rodigast, S., 1649–1700, hymn-writer, 569
Palestrina, G. P. da, c. 1525–94, 72, 140, 172, 393 Rolle, C. E., 1681–1751, organist in
Pasquini, B., 1637–1710, 57 Quedlinburg 1709, cantor of Magdeburg
Penzel, C. E., 1737–1801, Leipzig Thomasschule 1721, 583
from 1751, from 1765 appts at Merseburg, Rust, F. W., 1739–96, music director in Dessau
230–314, 341–70, 429–50 from 1775, 176
Pepusch, J. C., 1667–1752, Berlin composer by Ryom, P., author, 214, 216
c. 1700 settling in London, 106
Pergolesi, G. B., 1710–36, 3, 83 Sachs, H.-J., author, 237
Pirro, A., 1869–1943, author, 183 Sackmann, D., author, 75, 455, 469, 473, 483,
Pisendel, J. G., 1687–1755, from 1711 appts in 541, 544, 556
Dresden, 201 Saint-Lambert, M. de, 55
Poelchau, G., 1773–1836, from 1813 appts in Sasse, C., 1721–94, possibly pupil of J. C. Kittel,
Berlin, 2 organist in Halle, 488
Ponsford, D., author, 373 Scarlatti, A., 1660–1725, 197
Praetorius, E., author, 208 Scarlatti, D., 1685–1757, 114, 143, 198, 254, 519
Praetorius, M., 1571–1621, 238, 563 Schäfertöns, R., author, 43
Preller, J. G., 1727–86, pupil of J. T. Krebs?, from Schalling, M., 1532–1608, hymn-writer, 567
1753 cantor in Dortmund, 47–8, 51, 64, Scheibe, J., ?–1748, organ-builder in Leipzig, 584
81, 163, 171, 192, 225, 231, 358, 506, 509 Scheibe, J. A., 1708–76, pupil of J. S. Bach (?),
see also J. N. Mempell appts Hamburg 1736, Copenhagen, 389,
Printz, W. G., 1641–1717, from 1665 cantor in 394
Sorau, 290 Scheide, W. H., author, 416
615 Index of names
Scheidemann, H., 1596–1663, pupil of Schwen(c)ke, C. F. G., 1767–1822, from 1789

Sweelinck, succeeded father at Hamburg director of music in Hamburg, 159, 192,
Katharinenkirche, 39, 193, 259, 439 206
Scheidt, S., 1587–1654, pupil of Sweelinck, Seiffert, M., 1868–1948, author, 191, 205, 432,
appts in Halle, 236, 237, 254, 259, 272, 285, 462, 489, 575
308, 329, 342, 350, 395, 421, 422, 456, 457, Selnecker, N., 1532–92, hymn-writer, 331
464, 474, 477, 481–2, 505, 548, 563 Sicher, F., 1490–1546, Swiss pupil of Paul
Schein, J. H., 1586–1630, succeeded Calvisius in Hofhaimer, 254
1615 as cantor of St Thomas, Leipzig, 239, Siedentopf, H., author, 114
377, 422, 517, 536, 558 Siegele, U., author, 72, 191, 533, 538
Schelble, J. N., 1789–1837, collector, 536 Silbermann, G., 1683–1753, organ-builder and
Schemelli, G. C., c. 1680–1762, Thomaner, piano-maker in Freiberg, 6, 81, 235, 388,
hymn-compiler, 1727 cantor in Zeitz, 249, 587, 588
391, 545 Silbiger, A., author, 185
Scherer, H. (‘the Younger’), organ-builder in Simon, J. C., c. 1705 – c. 1750, appts in
Hamburg 1590s–1631, 584 Nördlingen, 141
Schering, A., author, 209, 393 Smend, F., author, 384, 421, 516–22
Schiefferdecker, J. C., 1679–1732, Thomasschule Smith, J. C., 1683–1763, Handel’s copyist, 106
1692, 1707 Buxtehude’s successor in Soler, A., 1729–83, Catalonian composer, 143
Lübeck, 474 Sorge, G. A., 1703–78, from 1721 organist in
Schmelzer, J. H., c. 1623–80, violinist, court Lobenstein, 199, 320, 389, 394, 421, 541,
appts at Vienna from 1649, 184 552
Schmid(t), B., sixteenth-century hymn-writer, Souchay, M.-A., author, 141, 532
568 Spaiser, D., hymn-writer (pub. 1609), 266
Schmid(t), B., 1705–49, organist and publisher Spangenberg, J., 1484–1550, hymn-writer, 240
in Nuremberg, 584 Spengler, L., 1479–1534, hymn-writer, 303
Schmidt, C. M., author, 98 Speratus, P., 1484–1551, hymn-writer, 305
Schmidt-Mannheim, H., editor, 334 Sperontes ( = J. S. Scholze), 1705–50,
Schmieder, W., author and compiler of songbook-compiler, 389
BWV, 536 Speth, J., 1664–1719?, from 1692 appts in
Schmitz, A., author, 305, 325, 362 Augsburg, 165, 168
Schmögner, T., author, 228 Spitta, P., 1841–94, 1, 2, 37, 40, 41, 44, 45, 48,
Schneider, J., 1702–88, copyist (perhaps 50, 53, 56, 62, 70, 83, 85, 92, 97, 107, 110,
Anon. 5), from 1729 organist Leipzig 119, 128, 130, 131, 143, 151, 152, 155, 160,
Nikolaikirche, 119, 166 164, 166, 173, 176, 177, 179, 185, 194, 195,
Schneider, M., author, 153, 224 197, 213, 214, 218, 252, 257, 264, 267, 278,
Schnitger, A., 1648–1719, organ-builder in 295, 305, 310, 314, 353, 357, 362, 378, 380,
Hamburg, 29, 584 445, 451, 454, 456, 457, 459, 460, 464, 465,
Scholes, P., author, 202 466, 467, 471, 473, 482, 500, 508, 527
Scholz, L., 1720–98, organist in Nuremberg, 56, Stapel, W., author, 251, 268, 295, 422, 476
59, 176, 355, 356, 434, 452, 490, 492, 578 Stauffer, G., author, 38, 54, 85, 100, 112, 113,
Schöneich, F., author, 37, 164, 172, 180 162, 197
Schop, J., ?–1664/5, hymn-tune composer, 571 Steglich, R., author, 295, 325, 410, 426
Schrammek, W., author, 21, 35, 169, 236, 272 Stephani, C., hymn-writer (pub. 1568), 264
Schreyer, J., author, 123 Steurlein, J., 1546–1613, hymn-writer, 264
Schubert, F. P., 1797–1828, 89 Stiller, G., author, 245, 256, 258, 288, 293, 296,
Schübler, J. G., c. 1725-?, music-engraver in 312, 314, 326, 329, 331, 370, 390, 395, 418,
Zella, 179, 317, 318, 584 459, 467
Schulenberg, D., author, 48, 171, 179, 181, 219, Stinson, R., author, 37, 101, 108, 112, 157, 159,
493 171, 197, 232, 235, 243, 250, 251, 267, 273,
Schulze, H.-J., author, 37, 45, 71, 81, 101, 182, 337, 340, 350, 455, 485, 497, 528, 536, 543,
190–1, 201, 202, 206, 218, 225, 575 546, 558, 580
Schumann, R., 1810–56, 148, 176, 352 Stolzhagen, C., hymn-writer c. 1590, 293
Schütz, H., 265 Strun(g)k, N. A., 1640–1700, appts in Hanover,
Schweitzer, A., 1875–1965, 234, 244, 246, 250, Dresden, Leipzig opera etc., 137, 193
251, 259, 264, 269, 271, 278, 284, 285, 293, Suppig, F., first half eighteenth century, organist
301, 303, 311, 314, 325, 357, 378, 407, 503, in Dresden?, 199
508 Swale, D., author, 172
616 Index of names
Sweelinck, J. P., 1562–1621, 132, 133, 357, 548 Wagner, G., author, 43
Swieten, G. B. Baron von, 1733–1803, 3 Wagner, J. J., 1690–1749, organ-builder in
Berlin, 584
Taesler, M., author, 319 Walther, J. G., 1684–1748, from 1707 town
Tag, C. G., 1735–1811, pupil of Homilius, organist of Weimar, 78, 167, 202, 205, 388,
321 390
Tagliavini, L. F., author, 205, 214, 215, 221 as copyist/transcriber, 52, 64, 138, 166, 205,
Telemann, G. P., 1681–1767, organist in Leipzig 209, 235, 247, 322, 341, 388, 405, 412, 441,
1704 (where founded the collegium 445, 508, 543, 553, 554, 573, 577, 580
musicum), from 1721 appts Hamburg, 8, as theorist, 31, 61, 150, 164, 184, 236, 392,
18, 36, 123, 191, 206, 225, 295, 393, 430, 529, 531, 586, 587, 588, 589
449, 492 as copyist, 3, 29, 46, 48, 55, 64, 101, 105, 163,
Terry, C. S., 1864–1936, author, 241, 243, 250, 166, 172, 174, 183, 206, 220, 230–316,
255, 256, 265, 266, 271, 272, 281, 282, 283, 337–81, 429, 453–95, 499–511, 524–5, 539,
284, 285, 286, 290, 293, 300, 305, 310, 313, 540, 549, 566
324, 326, 331, 360, 364, 395, 407, 441, 464, Walther, J. J., c. 1650–1717, violinist in Dresden,
470, 477, 487, 502, 517 489
Teschner, M., hymn-writer (pub. 1614), 479 Wechmar, J. A. G., 1727–99, son of Weimar
Thieme, C. A., 1721–95, Thomasschule, 1767 organist, copyist in Kellner circle?, 19, 23,
Konrektor, 1938 29, 74, 146, 307, 458
Tittel, K., author, 141, 191, 447, 490 Weckmann, M., 1619–74, pupil of Schütz, court
Torelli, G., 1658–1709, 76, 153, 355, 375, 589 organist at Dresden 1640, 1655 Hamburg
Trabaci, G. M., c. 1575–1647, Neapolitan Jakobikirche, 168, 185, 237, 399, 422, 588
organist, 587 Weinberger, G., author and editor of J. L. Krebs,
Trautmann, C., author, 390 490
Treiber, J. P., 1675–1727, professor of law Weismann, W., author, 416
in Jena, 227, 454 Weisse, M., c. 1480–1534, hymn-writer, 244,
Trost, G. H., c. 1673–1759, organ-builder 275, 564
in Altenburg, 584 Wender, J. F., 1655–1729, organ-builder in
Trumpff, G. A., author, 420 Mühlhausen, 583
Türk, D. G., 1756–1813, pupil of Homilius, Werckmeister, A., 1645–1706, successively
from 1776 appts in Halle, 516 organist in Brunswick, Quedlinburg and
Tusler, R. L., author, 370 Halberstadt, 155, 228, 462
Tutino, L. author, 181 Wesley, S., 1766–1837, organist in London and
editor, 3
Vetter, A. N., 1666–1734, pupil of Pachelbel, Wessnitzer, W., c. 1615–1680/90, hymn-tune
from 1691 organist in Rudolstadt, 244, composer, 560
526, 580 Westhoff, J. P. von, 1656–1705, violinist in
Vetter, D., ?–1721, from 1679 organist of Dresden, 489
Leipzig Nikolaikirche, 179, 199, 235, 313, Westphal, J. C., the Younger, 1773–1828, pupil
391, 544 of Kittel, 81
Vivaldi, A., 1678–1741, 218–24, 355 Weyrauch, J. G., 1694–1771, lawyer and
as composer, 34, 76, 83, 88, 92, 94, 201, lutenist, 71
203–5 Widor, C. M., 1845–1937, organist of
as arranged by Bach, 206, 209–15 St-Sulpice, Paris, 162
see also BWV 593, 594, 596, 972–980 Wilhelm II, Duke of Saxe-Weimar, 300
Vogel, H., author, 141 Williams, P., author, 28, 70, 158, 265, 509,
Vogelsänger, S., author, 186–7, 245, 250 589
Vogler, J. G., 1696–1763, pupil of J. S. Bach, Witt(e), C. F., 1660–1716, hymn-writer and
appts in Weimar from 1721 (incl. capellmeister at Gotha, Altenburg, 259,
burgomaster 1736), 29, 45, 101, 168, 178, 266, 573
341, 388, 394 Wolff, C., author and NBA editor, 37, 75, 81,
Voigt, W., author, 75 139, 159, 186, 202, 232, 244, 308, 320, 348,
Vopelius, G., 1635–1715, from 1675 cantor of 382, 384, 385, 393, 412, 423, 459, 466, 483,
Leipzig Nikolaikirche, 238, 377, 390, 422, 485, 492, 513, 538, 541, 582
427, 463, 530 Wollny, P., author, 175, 338, 348, 490, 580–1
Vulpius, M., seventeenth-century Wright, C., author, 200
hymn-composer, 558, 565 Wustmann, R., author, 266
617 Index of names
Zacher, G., author, 32, 87, 517–18 Zelenka, J. D., 1679–1745, appts at Dresden
Zachow, F. W., 1663–1712, from 1684 organist court from 1710, 393
at Halle Liebfrauenkirche, 90, 137, 185, Zelter, C. F., 1758–1832, director of Berlin
233, 236, 338, 459, 488, 501, 541, 545 Singakademie from 1800, 220
Zahn, D., author, 98 Ziegler, J. G., 1688–1747, pupil of J. S. Bach,
Zahn, J., 1817–95, compiler of hymns, 503, 545, from 1718 appts in Halle, 233, 585
552, 560, 561, 563, 564, 573 Zietz, H., author, 60, 337, 339, 350, 358, 445,
Zarlino, G., 1517–90, 585 465, 490, 496, 497, 524, 579
Zehnder, J.-C., author, 74, 76, 81, 153, 175, 183, Ziller, E., author, 504
220, 337, 353, 355, 375, 412, 439, 457, 488, Zipoli, D., 1688–1726, organist at the Gesù,
508, 525, 580–1 Rome from 1669, 191, 197, 425
Index of BWV works cited
This index does not list the main reference to a work (or other references to it in its group if it has
one) or any mere cross-reference. The number can refer to a work cited only by title in the text.
1 76, 484, 486 70 477

2 444, 488 71 43, 46, 154, 358, 501, 503
4 6, 42, 47, 128, 154, 237, 285, 286, 322, 457, 72 426
505, 546 73 361
5 326 75 570
6 322, 331 76 23–5
7 335, 552 77 83, 85, 408, 489
8 61 78 415
9 306 79 359
10 329, 475–6 80 238, 272, 367, 460
11 361 80a 460
12 257, 371, 509, 570 81 256
13 330 83 268
16 263, 467 84 312
18 212, 304 85 370
19 568 86 306
21 83, 84, 150, 224, 280, 312, 371, 509 87 257
22 478 88 312
23 273 89 326
24 503 90 302
25 193, 469 91 247, 436
26 314, 315–16 93 312
27 61, 312 94 503
28 263 95 268, 322, 479, 565
31 207 96 243
33 554 98 570
35 571 99 570
36 239, 364, 484 100 570
37 484 101 302
38 422, 423, 526 102 302
40 447 104 198, 370
43 293 106 1, 38, 46, 268, 271, 445, 449, 464, 501,
45 503 505, 553, 579
47 109 107 361
48 453, 566 109 304
49 484, 486 110 447, 529
52 449 112 370, 404
56 268 113 322, 566
57 333 114 416
59 342 117 306
61 238, 239, 484 119 467
62 239 120 467
63 276, 357 120a 333
64 247, 248, 256, 503 120b 467
65 114, 246, 309–10 121 259, 260, 435
66 288, 290 125 121, 268
67 291, 556 126 557
[618] 69a 570 127 273
619 Index of BWV works cited
128 370, 503 232 50, 195, 291, 388

129 503 233 274, 395
131 46, 154, 183, 416, 468, 566 243 329, 330, 475
131a 528 243a 251
132 387 244 97, 272, 273, 279, 281, 358, 449, 469, 547,
135 469 571, 582
136 326 245 61, 87, 275, 279, 281, 282, 302, 308, 479,
137 322, 333–4 536, 547, 568, 569
139 536 248 247, 251, 273, 447, 449, 469, 477
140 12, 322, 324, 529 250 570
142 447 252 359
143 322 253 331
144 570, 582 255 453
145 291, 470 259 524
146 123, 156, 221, 571 260 370
147 571 261 554
149 568 265 561
151 255 267 347
152 56, 58 273 500
153 488 274 550
154 571 276 288
155 306 277 285
156 364, 536 278 285
158 285 279 285
159 469 280 419
161 94, 315, 354, 378, 469 281 565
162 313 282 565
163 322, 326, 366 283 275
164 243, 503 288 264
166 190, 312, 322, 566 289 264
168 566 294 249
172 322, 344, 484 298 408, 409
174 568 302 460
175 342 303 460
176 419 305 463
177 307 306 290
179 312 314 247
180 308, 322 318 241
182 338, 359 323 329
183 263 324 329
184 563 328 467
185 307, 322 332 297
186 306 334 567
186a 361 340 568
188 198, 327 342 293
190 467 343 283
190a 467 358 257
192 359, 360 359 571
195 255 360 571
196 153, 203 363 377
197 312 364 286
197a 503 365 470
198 97 368 253, 254
199 322, 326, 353, 354 370 295, 381
208 76 371 395, 402
218 295 373 298
226 77, 342 375 255
227 257, 451 376 255
377 391 547 84, 99, 107, 121, 123, 127, 135, 140,
382 268 148, 155, 171, 262, 343, 368, 381,
386 359 393, 407, 414, 518, 523, 589
401 272 548 28, 48, 49–50, 62, 97, 99, 107, 115,
402 279 117, 134, 135, 148, 532
407 551 549 129, 132, 172
410 508 549a 37–40, 47, 53, 64, 93, 152, 155–6,
414 331 179
415 479 550 43, 65, 68, 590
417 361 551 40, 166
418 361 552 52, 78, 97, 100, 110, 120, 200, 300,
419 361 396, 512, 529, 585
431 310 556 198
432 310 559 458
434 312 561 458
436 484 562 50, 61, 107, 172, 575, 585
437 163, 411 563 38, 164, 183, 193, 524, 575
461 302 564 24, 42, 81, 92, 94, 95, 129, 132, 160,
499 508 170, 204, 479
519 190 565 73, 132
525–530 189, 191, 211, 226, 320, 336, 354–5, 566 41, 57, 84, 127, 135, 170, 180
375–6, 404, 417, 538, 577, 586 568 38
525 189, 426, 590 570 150, 183, 193
526 123, 189, 315 571 53, 165
527 9, 192 572 42, 151, 216, 539, 589
528 121, 153, 173, 318, 335 573 103, 112, 171, 197, 493
529 102, 585 574 26, 55, 79, 132, 183
530 83, 208, 375, 533, 538 574b 179
531 41, 47, 125 575 53, 127, 154, 172, 580
532 67, 128, 130, 131, 132, 142, 154, 176, 577 129
177, 181, 195, 526, 590 578 53, 58, 94, 156, 183, 508, 590
533 40, 52, 54–5, 88, 125, 132, 154, 573, 579 50, 172, 173, 278
590 582 5, 26, 53, 64, 86, 93, 136, 162, 171,
534 57, 188 410, 575, 589
535 50, 57, 58, 94, 135, 142, 156, 502, 588, 583 24
590 585 8, 190, 577
535a 127, 143, 150, 155, 485, 575 586 8, 190
536 52, 65, 586 587 22
537 17, 79, 99, 107, 110, 117, 123, 146, 588 64, 164, 195, 394, 590
188, 368, 587 589 173, 181, 394
538 78, 84, 97, 104, 114, 127, 129, 147, 590 143, 443
155, 193, 195, 202, 219–20, 221, 591 163, 411
368, 479, 585, 586 592 173, 204, 211
539 156, 158 593 73, 83, 94, 108, 204, 214, 221, 350
540 49, 63, 65, 94, 104, 107, 110, 133, 151, 594 7, 88, 145
192, 274, 343, 368, 394, 476 595 65
541 23, 27, 29, 57, 99, 106, 119, 128, 151, 596 65, 75, 83, 84, 121, 367, 587, 589, 590
156, 162, 188, 282, 320, 381, 437, 599–644 59, 78, 89, 183, 188, 221, 320, 337–8,
526, 585 339, 340, 341, 359, 373, 384, 392,
542 85, 97, 108, 120, 157, 188, 189, 190, 408, 417–18, 442, 445, 447, 455,
216, 225, 364 465, 468, 469, 472, 483, 504, 573,
543 53, 77, 85, 88, 100, 132, 143, 145 585
544 52, 91, 104, 119, 120, 121, 124, 134, chorales listed but not set 333, 342, 347,
135, 136, 289, 343, 589 351–2, 361, 370, 377, 411, 419, 422, 453,
545 29, 82, 121, 135, 140, 476 460, 463, 467, 477, 479, 484, 487, 488, 500,
545a 112 507, 536
546 18, 48, 61, 63, 78, 88, 94, 97, 104, 108, see also BWV 1090–1120
120–1, 123, 134, 135, 146, 204, 599 365, 588
335, 363, 368, 475, 585, 587 600 467, 471, 568, 587
601 417, 542, 568 664b 444

602 443 665 479
603 587 666 449, 483
604 465, 543 667 398, 483, 550
605 543 668 425, 437
606 517 668a 432
607 436, 471 669–689 116, 133–9, 198, 300, 320, 338,
608 324, 358 339–40, 341, 342, 343, 352, 354,
610 451, 512, 587 374, 377, 430, 437, 455, 468, 517,
612 447 584, 589
614 546, 547 669–674 288
615 112 669 137, 349, 590
616 362, 579, 587 671 345, 376, 589
617 546 672 589
618 337, 358, 443 673 450
619 148, 443, 519, 553 674 450
621 435 675 349, 366, 434, 435
622 553 676 5, 354, 374, 375
623 550 677 63, 112, 116
626 510 678 16, 148, 300–1, 321, 381
627 100 679 300, 450, 537, 552
629 424, 587 680 261, 497, 519, 550
631 380 681 112, 117, 557
632 354, 447, 580 682 254, 265, 349
633/4 148, 445, 472, 486 683 236, 265, 303, 381, 431
635 470 683a 432, 489, 579
636 417 684 321, 433
639 501, 542 686 350, 362, 369, 438
639a 426, 432 687 133, 383, 384
641 432 688 14, 366, 433, 511
642 431 689 100
644 360, 505, 509–10, 537 690 312, 478, 504, 559
645–650 221, 352, 366, 385, 393, 426, 516, 691 432, 489
517–18, 539, 578, 584 691a 418, 489
645 241, 578 693 274, 453, 490
646 110, 366, 433, 434, 449 694 110, 340, 476, 510
649 308, 361, 366, 449, 578 695 329, 342, 369, 477, 540
650 241 696 260
651–668 2, 230, 296, 480 697 248, 252, 466, 542
651 268, 327, 474, 518, 529 698 100
651a 57, 480 699 365
652 283, 384, 511 700 252, 514, 517, 573
653 283, 384, 511, 514 701 436, 514, 517
654 207, 283, 384, 511 702 472, 479
655 297, 405, 412, 494 703 540, 542
655a 5 704 244
656 273, 288, 443, 445, 479, 502 705
656a 268, 540, 549 706 473
656b 431 707 491
657 437 709 355
658 428 710 510, 516
659 280, 284, 438 711 321, 504, 508
659a 447 712 309, 340, 379, 410
660 321, 332, 405 713 342, 456, 558
661 63, 110, 474, 511, 590 714 274, 380, 451, 550, 558, 566
662 520 715 344, 500, 581
663 405, 409, 443 716 376, 429, 449, 467, 527
664 261, 404, 405, 456, 474 717 371, 449, 561
664a 5, 6, 190, 578 718 41, 286, 429, 448, 504, 509, 545–6
719 249, 572 808 121, 558

720 506, 545, 546, 571, 573–4, 809 109
587 810 123
721 284, 540, 543 811 113, 123, 558
722 540 812 312, 414
723 359, 436, 572 813 171, 198
724 183, 429, 439, 575 814 171
726 265 815 171
727 236, 269, 284, 302, 543, 576 816 99, 171, 198, 268
728 171 820 183
730 236 821 506, 571
733 79, 110, 343, 369, 429 825–830 2, 3, 320
734 327, 340, 429, 431, 504 825 52, 228, 491
735 338, 340, 342, 572 827 31, 199
735a 556, 560 828 134, 392
737 444, 445, 541 829 81, 82, 121, 434
738 500, 517 830 13, 63, 99, 121, 335, 344, 529
739 24, 94, 135, 439, 525, 546, 569, 571, 831 2, 122, 123, 134, 136, 137, 392, 413,
573 525, 589
740 350, 411 841 171
741 429, 438, 573 846–869 9, 28, 32, 102, 143, 175, 199, 227,
742 526, 559 440
743 527 846 38, 52, 104, 171, 195, 373, 427,
744 274 450
751 254 848 77
753 590 849 80, 127
755 494 850 176
757 563 852 109, 126
760 446, 539 853 169
761 446, 539 854 536
764 546, 558 855 121, 589
765 578 856 536
766 439, 457, 483, 549, 579 860 197, 373
767 244, 431, 477, 485, 546, 566, 579 862 80, 111
768 197, 232, 236, 287, 345, 364 864 344
769/769a 2, 107, 112, 117, 237, 254, 273, 274, 865 69, 79, 145, 149, 188
318, 320, 322, 324, 334, 343, 344, 867 181
361, 371, 379, 393, 439, 440, 499, 869 91, 99, 104, 200
530, 531, 536, 539, 584, 585, 586, 870–893 32, 133, 174, 391, 442
588, 589, 590 870/870a 103, 105, 115, 435
769a 336, 338, 382 871 115
770 83, 485, 546 872 450
772–786 112, 227, 529, 536 874 43, 116, 117, 435, 450
776 532 878 79, 139
777 98 879 90, 417
778 327 880 343
784 77 881 48
786 533 883 138
787–801 12, 35 886 31, 135
788 149 889 532
793 47 890 330, 400
798 104 891 63, 427
802–805 402 892 115, 399
802 135 893 123
803 56, 123 898 200
806–811 77, 456 902 143
806 483, 558 903 76, 200, 399
807 109, 558 904 71, 108, 149
906 123 1028 13, 25, 35

910–916 131 1029 102, 106, 538
911 157, 160 1030 7
912 41, 160 1039 528, 538
914 46, 49, 121, 172, 175, 580 1042 159, 208
915 160 1043 119
916 82, 151, 153, 526 1044 19, 22
921 89, 575 1046–1051 12, 109
944 143, 150, 179, 183, 589 1046 113
946 56 1048 12, 199, 204, 534
949 145 1049 34, 122, 204, 262
950 56, 145 1050 52, 84, 96, 123
951 44, 50, 71, 172, 173 1051 12, 221
951a 44, 53, 71 1052–1059 201
953 156 1052 34, 95, 156, 157
954 44 1054 159
957 452, 558 1056 364, 520
963 41, 43, 56, 149, 525–6, 1060 36
575 1061 109, 179, 218
964 158 1063 16, 19, 35, 71, 73, 156, 208
965 53, 151, 153 1064 16, 71, 166, 217
966 67, 77, 129, 151 1065 71, 206
971 9, 33, 63, 120, 134, 137, 300, 475, 1067 511
531 1068 123
972–987 3, 201, 209 1076 513
972 82 1077 see BWV 1087
973 34 1079 111, 123, 124, 318, 393, 399, 422,
984 218 424, 475, 516–17, 539
986 207 1080 63, 66, 69, 117, 138, 318, 343, 369,
988 13, 49, 134, 195, 204, 262, 273, 300, 380, 383, 385–6, 387, 393, 474,
325, 392, 499, 513, 516–22, 530, 516–17, 539, 584
590 1085 549
989 42, 183, 188, 526, 579 1087 513, 516, 517
991 171 1090–1120 165, 166, 232, 243, 307, 308, 456,
992 1, 50, 53, 56, 173, 179, 182, 571, 473, 477, 483, 488, 489, 492,
579 494, 526
996 36, 46, 52, 55 1090 232
997 123 1091 165
998 123 1092 480, 526
1000 73 1093 165
1001–1006 33, 72, 88 1095 470
1001 70, 73, 83, 88, 157, 158 1098 157, 412, 429
1003 158 1102 526, 579
1004 184, 187, 188 1103 429
1005 88, 123 1104 497
1006 158 1105 357, 470
1007 226 1106 480, 504, 526
1009 114 1107 357, 480
1011 158 1108 232, 357
1012 114 1109 582
1014–1019 4 1113 445
1015 24 1114 526, 537
1016 36, 60, 63, 198 1115 244, 526, 579
1017 13, 35, 375 1116 165, 582
1018 198, 510, 526 1117 313
1019 24, 510 1118 357
1027 7, 106, 198, 528, 538 1119 582
1027a 7, 190 1121 38, 89, 149, 164, 470
Anh.II 54 263 Anh.II 70 182

Anh.II 55 322 Anh.II 77 542
Anh.II 61 356 Anh.II 78 542
Anh.II 64 440 ‘Gründlicher Unterricht’ (1738) 40
Anh.II 68 263 ‘O Traurigkeit, O Herzeleid’, see Anh.I 200

Peter Williams The Organ Music of J S Bach PDF

Uploaded by

Copyright:

Available Formats

Peter Williams The Organ Music of J S Bach PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Peter Williams The Organ Music of J S Bach PDF

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What does Peter Williams discuss in his commentary?

What does Peter Williams discuss in his commentary?

This page intentionally left blank

The Organ Music of J. S. Bach

This is a completely revised edition of volumes I and II of The Organ

Peter Williams is an internationally renowned Bach scholar and

Cambridge University Press

© Cambridge University Press 2003

This book is in copyright. Subject to statutory exception and to the provision of

First published in print format 2003

- ---- eBook (EBL)

- ---- hardback

- ---- paperback

Cambridge University Press has no responsibility for the persistence or accuracy of

First published 1980 as The Organ Music of J. S. Bach, Volumes 1 and 2.

Preface [page vii]

BWV 131a Fugue in G minor [1]

generously shared with me their researches so far on ‘doubtful’ works at-

ABB the Andreas Bach Book (MS Lpz MB III.8.4)

Grönland MS in Copenhagen, Det Kongelige Bibliotek

Autograph: a section of the MS P 271. No title-page (fol. 1r left blank,

of 1809–10. Oley’s MS shows signs of revision, authorized or not, as if being

Origin and purpose

composed for composed previously

Trio types in organ music

Trio types in instrumental sonatas

slow ﬁrst movements (not in organ sonatas)

movements in four or more parts, the keyboard homophonic or contrapuntal

The three-movement structure is not the obvious ancestor of any classical

Some further characteristics

strangely free of the conventional associations there are between words

The Sonatas make a world of their own, as distinctive and accomplished

BWV 525 Sonata No. 1 in E major

Headed in P 271 ‘J. J. Sonata 1. à 2 Clav: et Pedal’; second movement ‘Adagio’,

The likelihood that this originated as a chamber trio in B major (KB

A 1–11 tonic, lh opens

The effect is that of a ritornello movement with a second half beginning

Such emphasis on motif is rather more typical of Bach’s Two-part than

Binary (12, 16 bars); fugal ﬁrst theme A, second theme developing

Binary (32 + 32 bars – cf. Goldberg Variations, Aria); second theme

BWV 526 Sonata No. 2 in C minor

While no movement of the Sonata is preserved in other versions, the cor-

ﬁrst movements: concerto Allegro, beginning as if tutti (non-imitative),

Such three-movement sonatas suggest less a chamber sonata than a very

That such ritornello movements sustain continuity is undeniable, but sec-

38–46 G minor pedal-point: repeats broken chords like a concerto;

1–8 subject (rh), countersubject (lh), codetta; with a

The key-plan, E to mediant, is unusual and suggests something specially

A 1–58 Exposition, two episodes, two futher entries

BWV 527 Sonata No. 3 in D minor

Headed in P 271 ‘Sonata 3 a 2 Clav. et Pedal’; ﬁrst movement ‘Andante’

A 1–24 quasi-fugue above continuo bass, followed by coda

Of particular interest is the middle or development section, which soon

Though the movement as a whole suggests little if any tutti/solo con-

Binary (8, 24 bars); contrasting themes (one in thirds, one more

So this binary form has elements of a ternary, a procedure not usually so

The movement hangs on a succession of two-bar phrases, every one with