Flute and Bach

J. S.
Bach's Compositions for Solo Flute:

A Reconsideration of their Authenticity
and Chronology
By ROBERT L. MARSHALL
Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

THE PUBLICATION OF THE "NEW CHRONOLOGY" of Bach's vocal
SINCE
music by Alfred Duirrand Georg von Dadelsen more than twenty
years ago,' there have been almost no important advances in Bach
chronology. Although we have now reached the point in Bach re-
search where, in effect, the last scribe and the last watermark have
been traced, catalogued and assimilated, we are left with countless
unanswered questions regarding the fundamentals: chronology and
even authenticity. We still know next to nothing, for example, except
in the very broadest terms, about when most of Bach's instrumental
compositions were written.2
The reasons for this impasse are not hard to find. Whereas for the
vocal works large numbers of composing-score autographs and origi-
I See Georg von Dadelsen, Beitrdgezur Chronologie der WerkeJohannSebastianBachs

(Tfibinger Bach-Studien, Heft 4/5 (Trossingen, 1958)); and Alfred Dfirr, "Zur
Chronologie der Leipziger Vokalwerke J. S. Bachs," Bach-Jahrbuch 1957,PP- 5-162.
A second, revised edition of Dfrr's chronology was published in 1976 in monograph
form as no. 26 of the series, Musikwissenschaftliche
Arbeiten,Herausgegeben vonder Gesell-
schaftfir Musikforschung(Kassel, 1976). An English-language summary of the chronol-
ogy appears in Gerhard Herz, ed., Bach CantataNo. z4o(Norton CriticalScores(New
York, 1972)), PP- 3-50.
2 Among the few significant contributions to the
chronology of the instrumental
compositions the following should be mentioned. (a) In reference to the organ and
keyboard music: Dadelsen, "Zur Entstehung des Bachschen Orgelbiichleins," Fest-
schrtftFriedrichBlumezum7o. Geburtstag,ed. Anna Amalie Abert and Wilhelm Pfann-
kuch (Kassel, 1963), pp. 74-9; Walter Emery, "The Compass of Bach's Organs as
Evidence of the Date of his Works," TheOrgan,XXXII (October, 1952), pp. 92-100;
Emery, "Some Speculations on the Development of Bach's Organ Style," TheMusical
Times, CVII (1966), pp. 596-603; Dietrich Kilian, "Studie uiber Bachs Fantasie und
Fuge c-moll," Hans Albrechtin Memoriam,ed. Wilfried Brennecke and Hans Haase
(Kassel, 1962), pp. 127-35; Robert L. Marshall, introduction to the facsimile edition,
JohannSebastianBach:Fantasiaper il Cembalo,BWV9o6 (Leipzig, 1976); Hans-Joachim
Schulze, "Johann Sebastian Bachs Konzertbearbeitungen nach Vivaldi und ander-
en-Studien- oder Auftragswerke?",DeutschesJahrbuch der Musikwissenschaft
fir 1973-
XVIII (1978), pp. 8o-Ioo. (b) Concertos: Heinrich Besseler, "Zur Chronologie
1977,
464 JOURNAL OF THE AMERICANMUSICOLOGICALSOCIETY
nal performingmaterialshavesurvived,therearevirtuallyno primary

sourcesfor the instrumentalcompositions.Most frequentlywe have
scribalcopies of greateror lesser authority, occasionallya later fair
copy in Bach's hand. This means that there is almost no unam-
biguous, hard, externalevidencethat'would permitus to date these
compositionswith certainty,or, in some cases, even to be sure that a
work is indeed by J. S. Bach.
BecauseBach'scompositionsfor solo flute are so few in number
and so typical in their source transmission,they can be taken as a

paradigmof the problemsof authenticityandchronologyposedby his
instrumentalmusicin general.They offeralsoa convenientarenafora
discussionof the conventionalmeansand for the testingof alternative
meansof dealingwith these problems.
The basicquestionsaboutthe flute sonatashavestill not been sat-

isfactorilyanswered.For example, how many compositionsfor solo
flute did Bach write, or moreprecisely, how many of the flute pieces
that have been attributedto him can be consideredauthentic?When
were they written? For whom? Were they originallyconceived as
flutesonatasor arethey arrangementsof worksfirstwrittenfor anoth-
er medium?
It is still widely believedthat Bach wroteeight worksfor the solo
flute. But the view that has prevailedamongBach specialistsfor the
pastfifteenyearsis thatthe rathersimilarSonatasfor fluteandharpsi-
chordobbligatoin G Minor(Bwv I020) and E Flat Major(Bwv
io31)
were probablycomposedby someoneelse-most likely Carl Philipp
Emanuel thatthe Sonatain C Majorfor fluteand continuo
(Bwv 1033) Bach--and
was written by one-or two-of Bach'sstudents(one of
them, again, might have been EmanuelBach), presumablyin part
under the composer'sactive supervisionand intervention.All three
works were excludedfrom the Neue Bach-Ausgabe;3 but they have
recentlyappeared in the Birenreiterseries, "Fl6tenmusik,"in an edi-
der Konzerte Joh. Seb. Bachs,"FestschriftMax SchneiderzumachtzigstenGeburtstage,ed.

Walther Vetter (Leipzig, 1955), PP-. 15-28; Martin Geck, "Gattungstraditionenund
Altersschichten in den Brandenburgischen Konzerten," Die Musikforschung,XXIII
(1970), pp. I39-52. (c) Music for solo instruments: Hans Eppstein, "Grundzuge in
J. S. Bachs Sonatenschaffen," Bach-Jahrbuch zp6p,pp. 5-31; Eppstein, "Chronologie-
probleme in Johann Sebastian Bachs Suiten fuirSoloinstrument," Bach-Jahrbuch z976,
PP. 35-57; Hans-Joachim Schulze, "Wer intavolierteJohann Sebastian Bachs Lauten-
kompositionen?", Die Musikforschung, XIX (1966), pp. 32-9.
3 Hans-Peter Schmitz, ed., Johann Sebastian Bach: Neue Ausgabe simtlicher
Werke, Serie VI, Band 3, WerkefiirFlite (Kassel and Leipzig, 1963), hereaftercited as
NBA, VI/3.
J. S. BACH'SCOMPOSITIONSFOR SOLO FLUTE 465
tion prepared by Alfred Diirr. The composer of the pieces is identi-
fied only as "Bach," with the tag "tiberliefertals WerkeJohann Sebas-
tian Bachs."4 The Neue Bach-Ausgabe has left us, then, with literally
a handful of authentic works for solo flute: the Unaccompanied Partita
in A Minor (Bsw I10 3); the Sonatas in E Minor and E Major for flute
and continuo (Bwv 1034 and 1035); and the Sonatas in B Minor and A
Major for flute and harpsichord obbligato (Bwv I030 and 1032).
It is no small irony that the Sonatas in G Minor, E Flat Major and
C Major were eliminated from the Neue Bach-Ausgabe--the veritable

symbol today of text-critical objectivity and responsibility--on the
basis of unstated stylistic criteria. The only reference to the omitted
compositions in the critical report appears under the list of abbrevi-
ations and reads in its entirety as follows:
Becauseof the strong doubts concerningtheir authenticity,the three-voice

Sonatas in G Minor (BWV1020), in E Flat (BWV 103I) and in G Major (BWV
xo38), as well as the two-voice Sonata in C Major (BWVo1033), have not been
includedin this volume;whichof them will appearin Volume4 of this series
[i.e., ChamberWorksof DubiousAuthenticity]hasyet to be determined.s
We learn from Hans Eppstein that at least the C-major sonata-

and by implication all of them--was excluded because Friedrich
Blume declared it, in passing and without any explanation, to be
"sicher nicht von Bach."6 In the meantime, Eppstein, in his Bach-
Jahrbucharticle and in an important monograph,7 has presented a
strong case against the authenticity of the three compositions on the
basis of a careful, specific, and altogether admirable stylistic analysis.
There is no need here to review his arguments except to say that he
successfully demonstrates that the compositions, at the least, are not
typical of J. S. Bach.
The source transmission of these three sonatas, as reported by
Eppstein and by Duirrin his new edition, is summarized in Table I.8
4 Alfred Dirr, ed., Bach, Sonate C-durfiir Fliite und BassocontinuoBWV 1o33, So-
naten Es-dur,g-mollfir Fliite und obligatesCembaloBWV zo3, ozo2o iiberliefertals Werke
JohannSebastianBachs(Kassel, 1975), hereafter cited as Durr 1975-
s NBA, VI/3, KritischeBerichte(hereafter KB), p. 6.
6 Hans Eppstein, "Uber
J. S. Bachs Flotensonaten mit Generalbap,"Bach-Jahrbuch
1972, pp. I2-23, especially p. I2. See Friedrich Blume, "Bach, Johann Sebastian,"
MGG, Vol. I, cols. 1013 and o1023.
7 Hans Eppstein, Studieniiber]. S. BachsSonatenfir ein Melodieinstrument und obli-
gates Cembalo(Uppsala, 1966). The discussion of Bwv o031 and i02o appears on pp.
176-81.
8 The table is based on Duirr 1975, PP. 39-40; also Eppstein, Studien, pp. 20 and
23, and Eppstein, "Fl6tensonaten," p. 12.
466 JOURNALOF THE AMERICANMUSICOLOGICALSOCIETY
TABLE I
THE PRINCIPAL
SOURCESFORTHECHALLENGED
FLUTESONATAS
Call No. Copyist Date Title

Sonata in C Majorfor Flute and BassoContinuo,swV 1o33
St 460 C. P. E. Bach c. i731 SONATA / a / Traversa / e /
Continuo / di / Joh. Seb. Bach
Sonata in E-Flat Majorfor Flute and HarpsichordObbligato,swv1 o30

a. P649 "Anonymous 4" mid-I8th Esd[ur] / Trio / Firs obligate Clavier
Title: C. P. E. century u. die Fl6te / Von / J. S. Bach.

Bach
b. P 10io56 C. F. Penzel c. 1755 Sonata / a / Flauto Travers. / ed /
(in part) Cembalo obligato /
di / I. S. Bach. /Poss. Penzel.
Sonata in G Minorfor "Flute"and HarpsichordObbligato,swv 1020

a. P io59 Unknown 2nd half SONATA. / del Signore Bach.
(Owner: Schicht) of i8th [lower right:] Schicht.
century
b. Vienna XI Michel 2nd half G. moll / SONATA. / Cembalo
36271 (Owner: Brahms) of 18th obligato. / con / Violino / Del
century Signore / C. P. E. Bach
c. P 471 A. Werner c. I840 Sonata in Gmo'/ per il / Clavicembalo
e Violino obligato comp: da / G.
Seb. Bach.
d. Breitkopf Catalogue of 1763 Sonata del Sigr. C. P. E. Bach, a Cl.
Music in Manuscript ob. c. V. [with incipit]
This tableis basedon informationin Dfirr1975(see n. 4). Unlessotherwisestatedthe MSS are
in the portionof the formerPreussischeStaatsbibliotheknow locatedin Berlin,Staatsbibliothek
PreussischerKulturbesitz,Musikabteilung. The lettersSt (Stimmen)andP (Partitur)are abbre-
viationsfor Mus.ms.BachSt. and Mus.ms.BachP respectively.
The principal source for the C-major sonata was written by C. P. E.

Bach around 1731, that is, when he was about seventeen years old and
still living in his father's house. The manuscript ascribes the work to
"Joh. Seb. Bach." As Duirr observes, few works attributed to J. S.
Bach have such clear and impeccable credentials.9 The main source
for the E-flat sonata dates from the mid eighteenth century and was
written out by an anonymous copyist who worked for both Sebastian
and Emanuel. The title page of the sonata, however, is in Emanuel's
mature (that is, not late) hand and identifies "J. S. Bach" as the com-
poser, as does the heading of the manuscript, which is in the hand of
the scribe. There is another mid-century copy of the E-flat sonata
which was once in the possession of Christian Friedrich Penzel
(1737-
9 "Wir waren froh, hatten wir fuirandere Werke eine ahnlich sichere Beglaubig-
ung!" (Dfirr 1975, p. 2). This point is made even more emphatically by Gloria Rose,
"Father and Son: Attributions to J. S. Bach by C. P. E. Bach," Studiesin Eighteenth-
CenturyMusic, ed. H. C. Robbins Landon (London, 1970), pp. 364-9-
J. S. BACH'SCOMPOSITIONSFOR SOLOFLUTE 467
I8o ). Penzel is one of the most importantand reliabletransmitters
and copyists of SebastianBach'smusic.10 This copy, partly in Pen-
zel's handwriting,also namesSebastianas the composer.For neither
the C-majornor the E-flatsonatais thereany conflictingattributionin
any source.
The situation is different for the G-minor sonata. There are three
known sources for this work. The first, from the second half of the
eighteenth century, has the non-committal title "Sonata del Signore
Bach." The title of a set of parts prepared by Emanuel's Hamburg

copyist Michel and now in the Vienna Gesellschaft der Musikfreunde
(from Johannes Brahms's estate) attributes the work to "Signore
C. P. E. Bach";" but a mid-nineteenth-century copy ascribes it to
"G. Seb. Bach." The work is listed also in Breitkopfs catalogue of
1763 as a composition of Emanuel's.12
I will take these cases in order. Eppstein points, among other
things, to the uneven quality of the C-major sonata. He considers the
first two movements to be quite mediocre and the last two rather fine,
observing that the first minuet resembles the first minuet of the Partita
in B Flat from the Klavieriibung,Part I. He concludes that the sonata
was likely written by two pupils of J. S. Bach, the second one superi-
or to the first and enjoying, moreover, Bach's active supervision and
assistance.13 Duirr is tempted to identify C. P. E. Bach as the com-
poser of the entire work.14 Both Diirr and Eppstein suggest that
Emanuel ascribed the work to Sebastian in recognition of his helping,
if limited, role in the composition.
10 See Hans-Joachim Schulze, "Die Bachuiberlieferung:Plidoyer fOr ein not-

XVII (1975), 45-57, esp. 48-52, and
wendiges Buch," Beitriigezur Musikwissenschaft,
Diirr, "Penzel,"MGG, Vol. X, col. 1o22. The most comprehensive discussion and as-
sessment of Penzel appears in Yoshitake Kobayashi, Franz Hauser und seine Bach-
handschriftensammlung (G6ttingen, 1973), pp. 109-12, 174-83.
11The
copyist, known only by his surname, and the approximate date of the
manuscript have recently been ascertained in Kobayashi, "Neuerkenntnissezu einigen
Bach-Quellen an Hand schriftkundlicher Untersuchungen," Bach-Jahrbuch 1978, pp.
43-60, esp. pp. 52-3-
12 Duirr 1975, p. 40, also refers to an additional source for BWV 10o20, said to be
part of the "Naumburg Church Archive," and attributing the work to J. S. Bach.
Dirr was not able to bring the copy to light.
13 Eppstein, "Flktensonaten," pp. 12-13.
14 Diirr 1975, p. 2. Diirr's suggestion that a correction in Philipp Emanuel's copy
of the sonata (St 460) indicates that the scribe was the composer is not convincing.
The correction, in m. 8 of the Adagio, affects the reading of an ornamental turn figure
(changed from d"c; "d"e" to d"e"d"c#") and could just as easily have resulted from mo-
mentary inattention on the part of the copyist as from the considered reflection of the
composer. The minor detail, in any case, has to be weighed against C. P. E. Bach's
ascription of the work to his father.
468 JOURNALOF THE AMERICANMUSICOLOGICAL
SOCIETY
Now it seems implausiblethat the young and, like most young

artists, surely proud C. P. E. Bach would have attributeda work
basicallyof his own compositionto his father, especially while he
lived at home. This surely stretchescredibility.On the other hand,
the compositionis strangely primitive, especially in the first two
movementswith theirrudimentarybass part.It is particularlyhardto
believethatJ. S. Bachcouldhave been responsiblefor that basspart.
We may indeed be dealinghere with a hybrid composition,but not
one in which a student (say, C. P. E. Bach)did the lion's shareand

the masterlent a helping hand only in the last two movements.The
fact alone that the first minuet has an obbligatoright-handharpsi-
chord part, while the rest of the sonatais for basso continuo, is suf-
ficient indicationof the hybrid natureof the piece.
Almostall the problemsof this compositiondisappear,it seems, as
soon as one entertainsthe possibilitythat the piece may in fact have
been writtenby J. S. Bach, but as a compositionfor unaccompanied
flute. The bassocontinuoin the first two movementsclearlydetracts
from rather than adds anything to them (see Exx. I and 2). The
flute line of the originalversionof the first minuet would have been
what is transmittedto us as the right-handline of the harpsichord
part, with occasionalborrowingsfrom the otherlines. The flute part
of the extantversionis obviouslya derivationof this right-handline
(see Ex. 3). The right-handpartneverdescendsbelow middled'-the
lowest note on the eighteenth-centuryflute-even thoughthe pieceis
in the key of C. It is even more significantthat the C-majorsonata,
like BWV10I3, and in contrastto the compositionsfor accompanied
flute, containspracticallyno restsfor the flute. Withthe exceptionof a
few quarter-restsin the firstpartof movementII the flute plays con-
stantly.15
The resemblanceof the minuet to that of the B-flat partitamay
even suggest a possible date of composition-or, more likely, com-
pilation-for the sonata. The partitawas first printedsingly in Sep-
tember 1726 with a dedicationto the newbornson of PrinceLeopold
of Cathen. It was reprinted as part of the Klavieriibungin 1731-the
time of Emanuel's copy of the C-major flute sonata. It seems altogeth-
er possible that J. S. Bach had assigned C. P. E. Bach the exercise of
fitting the sonata with a bass, or arranging it for flute and continuo or
harpsichord at that time. Hence the character of the surviving version
of the piece. The original--unaccompanied -sonata itself may have
'~ It is not really surprising that it took a flautist to be sensitive to the lack of rests
in the flute part: I am indebted to Mr. Jacob L. Berg, St. Louis, for this observation.
Examplei
J.S. Bach,Bwv 1033,firstmovement
(a) mm. I-5
Andante
Flauto traverso
Continuo 4 3

(b) mm. 10- 15
4
ii L•
,
-
t??i6
-61: I!IL
~
been composed around the same time as the Unaccompanied Partita

in A Minor, and conceived perhaps as a counterpart to it.
It is known, incidentally, that at about this time C. P. E. Bach
made arrangementsof other instrumental compositions of his father's,
namely of the lost Violin Concerto in D Minor and of the Sonata in
G Major for violin and continuo, BWV I021 . Emanuel's harpsichord
version of the violin concerto is differentfrom that known as BWV 1052.
Werner Breig suggests in passing that the transcription (Bwv I052a)
may well have been prepared at the request of Sebastian Bach.16
16
See Werner Breig, "Bach'sViolinkonzert d-Moll. Studien zu seiner Gestalt und
seiner Entstehungsgeschichte," Bach-Jahrbuch
1976, pp. 7-34. I am indebted to Joshua
Rifkin for communicating his findings corroborating the assumption of Breig and
others that C. P. E. Bach was the author, as well as the scribe, of the cembalo
arrangement.
470 JOURNALOF THE AMERICANMUSICOLOGICALSOCIETY
Example 2
BWV 1033, second movement, mm. I-6
Allegro
6 6 6 6 9 8
5 4 5

6 44
44
Example 3
BWV Io33, fourth movement, with suggested original version for flute alone
(a) mm. I-4
(Cembalo)
(b) mm. 9- o
'
At
J. S. BACH'SCOMPOSITIONSFOR SOLO FLUTE 47 I1
Breig's conjecture and mine concerning the Flute Sonata in C Major
would seem to support one another, and to shed light, moreover, on
J. S. Bach's method of teaching composition. Even more telling is the
case of the Trio Sonata in G Major for two flutes and continuo, Bwv
1038. Ulrich Siegele has convincingly argued that this composition,
which uses basically the same continuo part as that of J. S. Bach's
Sonata in G Major for violin and continuo, BWV 102 1, is actually a
work by C. P. E. Bach. In two respects this case represents the con-
verse of that of the Flute Sonata in C. In the trio sonata, J. S. Bach

evidently had Emanuel invent new upper parts to a given bass; in the
flute sonata, as we have seen, his assignment was the opposite. Sec-
ondly, whereas the C-major sonata is apparently a work essentially of
J. S. Bach's composition surviving in the hand of C. P. E. Bach, the
trio sonata is preserved in a manuscript penned by J. S. Bach.17
For the E-flat sonata, the most popular and indeed the most attrac-
tive of the three questionable sonatas, the argument against its authen-
ticity has been based first on its "un-Bachian"features (as Duirrputs
it: the thumping basses; the triadic, galant-sounding melodies; the
short-breathed phrases) and secondly on the fact that it has clear simi-
larities with the G-minor sonata.18 The latter, as we have seen, is
frequently ascribed to C. P. E. Bach in the early sources. Diirr's con-
clusion is that both the E-flat and the G-minor sonatas may have been
compositions of Emanuel's, that the E-flat was still written under his
father's supervision, and that the G-minor, while directly modelled on
the E-flat, was a sufficiently independent effort on C. P. E. Bach's
part, for him to justify his identifying himself this time as the com-
poser. But again, on the face of it, it seems unlikely that C. P. E. Bach
17 See Ulrich Siegele, Kompositionsweiseund Bearbeitungstechnik

in der Instrumental-
musik Johann Sebastian Bachs (Stuttgart, 1975), PP. 31-46. The manuscript of BWV
o1038, in the Germanisches Nationalmuseum, Nuremberg, mentions no composer; it
is datable, on the basis of its watermark-"MA, large form"-to the period 173 2-5. In
this connection, finally, there is a remarkable reference in the catalog of Emanuel's
musical estate to a "Trio fiir die Violine, Bratsche und BaP3,mit Johann Sebastian
Bach gemeinschaftlich verfertigt." See Heinrich Miesner, "Philipp Emanuel Bachs
musikalischer Nachla/3," Bach-Jahrbuch 1939, p. 86. I have not been able to identify
the composition; no work for this combination is included in Alfred Wotquenne,
ThematischesVerzeichnisder Werkevon Carl PhilippEmanuelBach (Leipzig, I905).
18 See DOrr 1975, Pp. 2-3. One wonders whether the phrase structure of the
sonata is accurately characterized as "short-breathed."In the first movement, for ex-
ample, the keyboard introduction, for all its use of "modern" redicta, proceeds for
eight rather long measures quite "breathlessly" over basically a single cadential pro-
gression. Similarly, the first "solo episode" after a curious six-measure double motto
(mm. 9-14) spins out for I2 full measures (mm. 14-26) before there is any strong
caesura.
472 OFTHEAMERICAN
JOURNAL MUSICOLOGICAL
SOCIETY
would wish to attributea workof his own entirelyto his father,even

if he enjoyedthe latter'sguidance,without at least providingan ex-
planationfor such a peculiaraction.19
Now the E-flatsonatais ascribedto J. S. Bachnot only by Eman-
uel but alsoby Penzel, whoseconnectionswith the Bachsourcesseem
to deriveprimarilyfrom WilhelmFriedemann,not EmanuelBach.20
In any case, Penzel's copy is demonstrablyindependentof Eman-
uel's.21Therefore,it seems likely that the two attributionsto Sebas-
tian arelikewiseindependent.This, in fact, is the cruxof the external

argumentfor Bwv 1031; for both Penzel and C. P. E. Bach have at-
tributedto J. S. Bach elsewherecompositionswhose authenticityis
highly questionable,if not altogetherwithoutcredibility:in Penzel's
case the Christmascantata,Uns ist ein Kindgeboren,BWV 142; in the
caseof Emanuelthe Suite in A Majorfor violinandharpsichordobbli-
gato, BMV1025. Since the existenceof the latterattribution,in partic-
ular, has been adducedto buttressthe challengeto the credentialsof
the E-flatsonata(andthe C-major,too, for thatmatter),22it is impor-
tant to emphasizethat the attributionsof the cantataand the suite,
unlike that of BWV1031, are uncorroborated.23
19 Durr conjectures that Emanuel's faulty memory in old age could have accounted
for the "false" attribution of BWV 1031 to Sebastian (Durr 1975, PP. 2-3.). But the
attribution to "J. S. B." also appears, in the hand of the copyist Anonymous 4, in the
heading of P 649, and not only (in Emanuel's handwriting) on the title page. And,
again, this explanation leaves Penzel's independent and contemporaneous ascription
to "I. S. Bach" out of consideration.
20 See
Schulze, "Die Bachuberlieferung," pp. 49-50.
21 See
Diirr 1975, 39-
22 Diirr 1975, p. 2. p.Eppstein, Studien,
pp. 185-7, offers a cogent characterization
of the compositional failings of BWV 1025 (as distinguished from the presumed stylistic
anomalies he observes in BWV 1033 and 1031).
23 Penzel'sMS of BWV 142, (SPK) P 1042, is the earliestknownMS of the work
and, according to Dirr, StudieniiberdiefriihenKantatenJohannSebastianBachs,2nd ed.
(Weisbaden, 1977), PP- 57-8, is the source for all further copies. C. P. E. Bach's copy
of BWV 1025, (SPK) St 462, would seem to have served as the source for the two later
MSS that attribute the work to Sebastian: (SPK) St 442 and (SPK) St 442 is in
the hand of Michel; the wording of the heading of St 443, copied c. i8oo,443. St
is identical
to that of St 442. (See also Paul Kast, Die Bach-HandschriftenderBerlinerStaatsbibliothek,
Tibinger Bach-Studien, Heft 2/3 (Trossingen, 1958), pp. 87-8.). Another MS of the
violin suite, bound in the miscellany P 226, consists only of a harpsichord part; the
heading of the work mentions no composer. The very first system of the MS--and
only this much-is in the hand of J. S. Bach. The continuation, which disagrees
substantially from the version preserved in the C. P. E. Bach sources, is in the hand
of an anonymous scribe. In effect, the violin part has been transferred here to the right
hand of the harpsichord part and the piece rearrangedfor solo keyboard. This, pre-
sumably, represents yet another exercise by J. S. Bach the teacher of composition.
Having started the pupil off by entering the first system, Bach apparently expected
the pupil-scribe to prepare a transcription of the duo. It is conceivable that Bwv
J. S. BACH'SCOMPOSITIONS
FORSOLOFLUTE 473
The obviously galant style of the E-flat sonata should perhaps be
considered as a clue to its dating rather than a judgment of its authen-
ticity. As I have attempted to develop elsewhere, Bach seems to have
been increasingly influenced by the galant style during the 173os and
early i740s when he was interested in strengthening his connections
with the musical establishment in Dresden.24 That the E-flat sonata
could have been written by J. S. Bach in the early 1730s and that it
could have served as a model for C. P. E. Bach when he composed the
G-minor sonata-for violin25--at around the same time (that is, be-

fore Emanuel left Leipzig in 1734), is not only plausible but seems to
be the explanation that accords best with the external evidence. In
brief, the greater part of valor would seem to dictate that both the C-
major-as an unaccompanied sonata, perhaps-and the E-flat sonata,
as it has come down to us, should be restored to the Bach canon. They
should, at the least, be included in the NBA among the operadubia;
but even that would be perhaps more of a stigma than the evidence
warrants.26
The discussion so far has been rather agreeable; for the weapons
employed have been the conventional ones-external source evi-
dence-used in the widely accepted manner: namely, according to a
"strict constructionist" view of the objective data. In turning now to
the five authenticated compositions I shall be standing, quite a bit less
comfortably, on the other side of the fence. The following remarks
will be, frankly, quite speculative. The point of departure, however
(and this needs to be emphasized), will again be the sources.
The five flute compositions have reached us singly. It does not
io25--whoever may have been the actual composer, and for whatever medium the
piece was originally intended-was a favorite instruction piece of Sebastian's for as-
signments of this kind and was mistakenly taken by Emanuel later on to be an original
composition of his father's.
24 Robert L. Marshall, "Bach the Progressive: Observations on his Later Works,"
TheMusicalQuarterly, LXII (1976), pp. 313-57-
25 No instrument but the violin is mentioned in any source. The widespread con-
viction that a flute was intended is based on the fact that the lowest note called for in
the piece is d'. (See Eppstein, "Fl6tensonaten," p. 176, n. 2, for a summary of the
literature.) This would indeed be odd if the sonata were by J. S. Bach and intended
for the violin, since Bach normally exploits the entire available range of his solo in-
struments. But it is not at all certain that the young C. P. E. Bach would have felt
constrained to adopt this characteristic practice of his father.
26 This is Rose's conclusion, too. See "Father and Son," p. 368. Rose, though,
does not attempt to account for the stylistic peculiarities of Bwv 1031 and 1033. For
the E-flat sonata she seeks rather to deny them; and she avoids the issue altogether in
the case of the C-major sonata. As with the violin suite, BWV I025, she is content here
to accept the attribution on the strength of C. P. E. Bach's authority alone.
474 JOURNAL OF THE AMERICAN MUSICOLOGICAL SOCIETY
TABLE 2
THE PRINCIPAL SOURCES FOR THE AUTHENTICATED FLUTE SONATAS
Call No. Copyist Date Title

Partita in A Minor for UnaccompaniedFlute, 1
swv o13
P968 I) "Anonymous 5' c. 1722-3 Solo pour la Flute traversiere par J. S.
Bach
2) Anonymous
Sonata in E Minor for Flute and Basso Continuo,

swv 1o34
P 804 J. P. Kellner c. 1725-6 Sonata per la Flaute / Traversiere e Basso

di /J. S. Bach. /J. P. Kellner.
Sonata in E Major for Flute and Basso Continuo, Bwv 1035

P 622 Unknown early i9th SONATA / per / Traverso Solo e
century Continuo / del Sgre / Giov: Seb: Bach.
[Notation in the hand of Count Voss-
Buch:] Auf dem Exemplare, von welchem
diese Abschrift genommen worden ist,
steht die Bemerkung: "nach dem Auto-
grapho des Verfassers welches o. 17 da er
in Potsdam war, for den Geh. Kimmerir
Fredersdorfvon ihm angefertigt worden."
Sonata in B Minorfor Flute and HarpsichordObbligato,sBWV 1030
P 975 J. S. Bach c. 1736 Sonata a Cembalo obligato e Travers. solo
di J. S. Bach
The G-Minor Version of swv 1030

P ioo8 "Anonymous 300" 2nd half G. moll / Sonata / al / Cembalo obligato /
(C. P. E. Bach's of i 8th e / Flauto traverso / composta / da /
copyist) century Giov. Seb. Bach
Title: Voss-Buch
Sonata in A Majorfor Flute and HarpsichordObbligato,Bwv 1032
P 612 J. S. Bach c. 1736 Sonata a i Traversa e Cembalo bbligato di
J. S. Bach
The informationin this tableis basedon NBA, VI/3 (see n. 3). The MSS arelocatedin Berlin,
StaatsbibliothekPreussischerKulterbesitz,Musikabteilung, exceptfor P 975 andP 612, which
are in the DeutscheStaatsbibliothek,Berlin,GermanDemocraticRepublic.For the abbrevia-
tions P and St see the notesto Table i.
seem that they were ever gathered together into a set in any eight-
eenth-century source. Table 2 offers a brief description of the princi-
pal sources.27 We have a manuscript for BWV 10o3 dating from the late
Cothen period, about 1722-3; a manuscript for Baw 1034 probably
dating from 1725 or 1726; autograph score copies for BMV 10o30 and
1032, both dating from around 1736; and a nineteenth-century manu-
script for awv 1035 associating the piece with a visit by Bach to Pots-
dam, and thus dating it presumably 1741 or 1747. That is, of the five
27 The table is based on information supplied by Schmitz in NBA VI/3, KB, pp.
7-8, II,23, 28-30, 32, 43-4.
compositions for solo flute accepted by the Neue Bach-Ausgabe as
genuine, only one, the A-minor partita, Bwv 1013, survives in a
source evidently dating from the C6then period. Let us reflect on this.
It used to be thought that Bach's instrumental music (with the
exception of the organ works) was concentrated into a brief period: the
six years 1717-23, when Bach served as capellmeister in Cothen, and
that his church music was written gradually over the course of his life.
Can it be that just the opposite might be the case? It is now known

that Bach wrote the great majority of his church cantatas during the
concentrated period 1723-7, his first four years as Thomaskantor in
Leipzig. Perhaps, conversely, his instrumental music was written at
various times throughout his career. The manuscript evidence in the
case of the flute sonatas does not preclude that possibility. And we
know that the preparation and publication of the four parts of the
Klavieriibungand the composition of the Well-Tempered Clavier II as
well extended from at least the mid I720S to the early I740s.28
Even the dating of instrumental works apparently most securely
associated with the Cothen period cannot always be accepted without
qualification. In a reference to the sonatas for violin and obbligato
harpsichord, BWV o1014-19, Johann Nikolaus Forkel remarked in
chapter 9 of his biography that they were composed at C6then.29 It
must be noted, however, not only that the earliest surviving manu-
script of the final version of the set (in the hand of Johann Christoph
Altnikol) presumably dates from the last years of Bach's life (when
Altnikol was a pupil of Bach's),30but that a copy of what is apparently
the earliest of the three known versions of the sonatas would seem also
to have been prepared during the Leipzig period. For the principal
copyist of the keyboard part is Bach's "Hauptkopist C," an important
scribe who is represented in the manuscripts of Bach's vocal works
from April 1724 to February 1727.31
28 Earlyversionsof the A-minorand E-minor

partitas,Bwv 827 and 830, appear
in Bach's hand, as fair copies, at the beginning of the Klavierbiichlein
fiir Anna Magda-
lena Bach of 1725. See NBA, V/4, KB, pp. 40-6. The four parts of the Klavieriibung
were published in 1731, 1735, 1739 and c. 1742. Regarding the origins of the Well-
TemperedClavierII see Werner Breckoff, Zur Entstehungsgeschichtedeszweiten Wohltem-
periertenKlaviersvon Johann SebastianBach (Tiibingen, 1965), especially pp. 3, 8, 17-
24.
29 Hans T. David and Arthur Mendel, eds., TheBachReader(New York, 1966), p.
343-
30 According to Alfred
Dirr, "Zur Chronologie der Handschrift Johann Christoph
Altnickols und Johann Friedrich Agricolas," Bach-Jahrbuch1970, pp. 44-65 (esp.
pp. 46-8), P 229 can be dated no more precisely than "not before 1747."
31 The remainder of the keyboard part is in the hand of Bach himself. See the
description of St 162 in NBA, VI/I, KB, pp. I39-40., which, however, does not
This issue cannot be pursued here. Suffice it to say that it is of

course conceivable that Bach had drafted the first version of the so-
natas in Cothen and wished to make a copy of them after settling in
Leipzig where he eventually (sometime between the date of St 162,
i.e., c. 1724-5, and that of P 229, c. 1747-9?) composed the inter-
mediate and final versions. It is hardly plausible-at least it would be
most uncharacteristic of Bach-that the composer would have re-
copied in Leipzig the earliest version of a work that had already been
brought to a more advanced form in C6then. That is, the sonatas for

violin and obbligato harpsichord, in their familiar, final form, are al-
most certainly a product not of C6then, but Leipzig.32
In brief, the assumption that Bach wrote all or almost all of his
instrumental chamber and ensemble music in Cothen is, in fact, noth-
ing but an assumption. The notion may have started with Spitta who
apparently expanded on the report in the obituary prepared by Eman-
uel Bach and Johann Friedrich Agricola, and repeated by Forkel, that
Bach composed most of his organ music in Weimar.33 Spitta elabo-
rated this into the admittedly very elegant conceit that Bach's artistic
career followed a rational plan: the bulk of the organ music composed
identify the first hand. A facsimile of a page from St 162 in the hand of Hauptkopist C
is reproduced as the frontispiece to Volume II of the Wiener Urtext Edition of the
sonatas, ed. Bernhard Stockmann and Hans-Christian Miiller (Vienna, 1973). For the
dates of this scribe's appearancein the Bach sources see DOrr, "Zur Chronologie," p.
I48. I should like to thank Dr. Dirr for having confirmed, in a private communica-
tion, the identification of the copyist and the continued validity of the dates of his
known association with Bach. Actually the dating of St 162 can be fixed more precise-
ly. The third and fifth movements of this version of Sonata No. 6, Bwv o1019, are early
versions of the Corrente and Tempo di Gavotte of the E-minor partita from the Kla-
vieriibung,Bwv 830-indeed earlier than the versions of the same movements in the
KlavierbiichleinfiirAnna MagdalenaBachof 1725. (See n. 28 above, and NBA, V/I, KB,
p. 53.) Since this portion of the Klavierbiichleinwas written in 1725 and Hauptkopist
C's association with Bach began in 1724, St 162 was almost certainly copied in 1724 or
1725. This conclusion is unaffected by the fact that these particular movements were
entered into St 162 by Bach himself (in their proper position), since it is hardly likely
that Bach would have entered an earlier version of them into St 162 after he had
revised them for the version in the Klavierbiichlein.
32 The general conviction, however, as articulated even by Eppstein, who has
convincingly established the relative chronology of the three versions of the sonatas, is
that the final version, too, is a product of Bach's C6then period. See Hans Eppstein,
"Zur Problematik von J. S. Bachs Sonate for Violine und Cembalo G-Dur (Bwv
o 19),"ArchivfiirMusikwissenschaft, XXI (1964), pp. 217-42, and Eppstein, Studien, p.
i64. In neither publication does Eppstein indicate that the principal scribe of St I62 is
Hauptkopist C nor does he discuss the absolute dating of the source except to suggest,
on the basis of its relationship to the Klavierbiichlein,
that it must be earlier than
See "Zur i725-
Problematik," pp. 223-4.
33 English translations of the obituary and of Forkel's biography are printed in The
Bach Reader, pp. 215-24 and 295-356, respectively.
in Weimar, the keyboard and chamber works in Cdthen, the church
music in Leipzig. I suspect that we may be laboring under the burden
of an unchallenged, comfortable myth that might have been keeping
generations of Bach scholars from noticing rather obvious clues in the
sources themselves.
To return to the flute music. There is no autograph for the A-
minor partita. The composition survives in a single manuscript that
was written out by two copyists, one of whom, known in the litera-

ture as Anonymous 5, was associated with Bach in both C6then and
Leipzig. The manuscript represents the earlier, C6then stage, in his
script, and dates from I722-3.34
In contrast to all of Bach's other flute music, the partita is consid-
ered unidiomatic by many flautists--especially the first movement,
where there are practically no natural places to take a breath, and the
second, where some extended arpeggio patterns are quite awkward.
Moreover, it is Bach's only flute composition to call for the high a"'-
the highest note on the baroque flute.35 For these reasons it is assumed
by some (including Schmitz) that the composition must have been
written originally for some other instrument.36 But the work obvious-
ly can be played on the flute; and it could have been conceived origi-
nally for the instrument-if Bach had reason to believe that there was
a virtuoso able to play it.
Now the modest technical demands of the surviving flute parts in
Bach's Cithen cantatas, the birthday cantata, Durchlauchtster Leopold,
Bwv 173a, and the (New Year's?) cantata, Bwv 184a (the text for which
is lost)37 suggest that Bach's regular flautists in Cothen, the Chamber
34 See NBA VI/3, KB, pp. 7-8.
35 Schmitz claims in fact that the partita is the only work for flute from the first
half of the i8th century that employs the a"'. See NBA, VI/3, KB, p. 8, n. io. But the
note does appear in the table of fingerings printed in Johann Joachim Quantz's trea-
tise. (See the English translation by Edward R. Reilly with the title On Playing the
Flute (New York, 1966), p. 42.) The table of ranges published in the appendix of
Charles Sanford Terry, Bach'sOrchestra(London, 1932), pp. 202-3, list the a'" as the
highest note in two cantatas, Bwv 8 and 145. In the latter instance this is a simple mis-
take: the highest note in the cantata isg"'. The case of Cantata 8 is quite complicated.
The Bach-Gesellschaft edition of the cantata contains a version of the fourth move-
ment, the aria, "Doch weichet, ihr tollen vergeblichen Sorgen," in which the obbligato
flute part contains several examples ofg#"' and a"'. It suffices to say here that none
of the surviving original sources of the cantata, including two versions of the flute
part, one of which is autograph, demand the a"'.
36 NBA, VI/3, KB,
p. 8.
37 Only the instrumental parts for Bwv I84a survive. They were reused in the
Leipzig parody version, Bwv I84, ErwiinschtesFreudenlicht.See the critical reports for
both cantatas, prepared in each case by Diirr: NBA, 1/35, KB, pp. 138 ff. (Bwv I84a),
and NBA, 1/14, KB, pp. 140 ff. (Bwv 184).
Musicians Johann Heinrich Freytag and Johann Gottlieb Wtirdig,

had rather limited abilities on the instrument.38(Even the flute part of
the Fifth Brandenburg Concerto, assuming for the moment that it was
intended for a performance by Bach's Cothen ensemble, is not partic-
ularly difficult.39 We may possess a clue, however, as to the identity
of the flautist Bach may have had in mind for the partita.
The heading of the A-minor partita, in the hand of the original
copyist, is in French. It reads: "Solo pour la flute traversiere par J. S.
Bach." This could be significant.40The greatest flute virtuoso in Ger-

many during Bach's lifetime was the Frenchman Pierre Gabriel Buf-
fardin (1689-1768). Buffardin was the principal flautist in Dresden
from 1715 to 1749.41 (Another flautist resident in Dresden during
much of that time, incidentally, was Buffardin's pupil, Johann Joa-
chim Quantz.)42
We have it on the testimony of C. P. E. Bach that J. S. Bach knew
Buffardin personally. In i774 Carl Philipp Emanuel sent along a copy
of the genealogy of the Bach family, a document prepared by his fa-
ther, to Johann Nikolaus Forkel, together with his own emendations.
Entry no. 23 referred to Johann Sebastian Bach's well-travelled elder
brother Johann Jacob Bach, for whom J. S. Bach had composed the
"Capriccio on the Absence of the Most Beloved Brother" (Bwv 992),
38Bach's Cothen flautists are identified name in Ernst

by K6nig, "Die Hofkapelle
des Fuirsten Leopold zu Anhalt-K6then," Bach-Jahrbuch t95p, pp. 160-7, especially
pp. i6o-i.
39The dedication of the concertos, of course, is to the Margrave of Brandenburg,
whose residence was in Berlin. Hans-Joachim Schulze, in his foreword to a facsimile
edition of the autograph performing parts of the Fifth Concerto (St I30), offers the
plausible argument that the dedication of the Brandenburg Concertos constituted a
barely disguised application for a position at the Margrave's court, made at a time
when the musical establishment at C6then was already in the midst of a substantial
decline. See Hans-Joachim Schulze, ed., JohannSebastianBach, Brandenburgisches Kon-
zert Nr. 5, D-Dur, BWV itoo. Faksimiledes Originalstimmensatzes (Leipzig, 1975), PP-
7-8.
40Frenchtitles or headings are extremely rare in the original--or early--sources of
Bach's instrumental music. Apart from the flute partita I am aware of only the follow-
ing examples: the autograph score (P Am. B. 78) of the Brandenburg Concertos; the
autographs of the lute compositions, BWV 995 and 998 (the first in the Biblioth~que
royale de Belgique, Brussels; the second in private possession in Japan); and Anna
Magdalena Bach's copy of the Suites for Unaccompanied Violoncello, Bwv I007-12
(SPK P 269). It is worth observing that both the score of the Brandenburg Concertos
and of the Suite for Lute, Bwv 995, are dedication copies; the heading of the latter
reads:Piecespour la Luth / a / MonsieurSchouster/ par /J. S. Bach. (Finally, for the sake
of completeness, it should be mentioned that it is not certain that the title page of
Anna Magdalena Bach's copy of the violoncello suites is also in her hand.)
41See Dieter Hartwig, "Buffardin,"MGG, Vol. XV, cols. 183-4-
42See Reilly, pp. xiii-xv. I
around i704. Emanuel added the following gloss to his father's ac-
count:
From [the Turkish town of] Bender [JohannJacob] journeyedto Con-

stantinople,and there had instructionon the flute from the famousflutist
Buffardin,who had traveledto Constantinoplewith a Frenchambassador.
This informationwas furnishedby Buffardinhimself when he once visited
J. S. Bach in Leipzig.43

It does not seem unreasonable to suppose that Bach had already
met Buffardin before he settled in Leipzig. The most likely occasion
for their first encounter would surely have been Bach's first docu-
mented visit to Dresden in the fall of I7i7 for his celebrated and well-
publicized recital--the remnant of the notorious, aborted contest with
Louis Marchand. This visit took place shortly before Bach was to
assume his new duties as court capellmeister to Prince Leopold of
Anhalt-C6then in December I717.44 Since the contest with Mar-
chand was arranged by the concertmeister of the Dresden Court
Chapel, Jean Baptiste Volumier (1670-1728), it is evident that Bach's
reputation had already reached the musicians of the Saxon capital.45
Indeed we know that Bach had connections with Dresden musicians
even earlier. He is known to have met the brilliant German violinist,
and Volumier's successor as concertmeister, Johann Georg Pisendel
(1687-1755), as early as March 1709, when the latter passed through
Weimar and took the trouble of introducing himself to Bach.46 It is
easy to imagine that it was Pisendel, a member of the Dresden Court
Chapel since 1712, who brought Bach to the attention of Volumier.
In 1717 Pisendel (or Volumier-it does not matter) could not have
failed to introduce Bach in his moment of triumph to his other Dres-
den colleagues, Buffardin among them. During the course of this
sojourn we can assume also that Bach had an opportunity to hear the
French flute virtuoso and to form, very likely for the first time, an
impression of the character and capabilities of the transverse flute in
the hands of a first-rate player. For, as far as is known, Bach had
43See The Bach Reader, pp. 207-8.

44See the obituary account, TheBachReader,p. 2 18. For evidence that the contest
must have taken place in the fall, no sooner than the end of September 1717, see
Werner Neumann and Hans-Joachim Schulze, eds., Bach-Dokumente, Band I:
von derHandJohannSebastianBachs(Kassel and Leipzig, I963), pp. 26-7-
Schriftstiicke
45TheBach Reader,p. 218.
46See Hans-Joachim Schulze, ed., Bach-Dokumente, BandIII: Dokumentezum Nach-
wirkenJohannSebastianBachs i75o-z8oo (Kassel and Leipzig, 1972), p. 189, and Hans
Rudolf Jung, "Pisendel," MGG, Vol. X, col. 1301.
never, up to that time, composed anything at all for the instrument.

There are no transverse flutes in any of the Muhlhausen or Weimar
cantatas. The instrument makes its first appearanceno earlierthan the
year 1720 in one or the other of the two C6then cantatas mentioned
above.47
While we are speculating about Bach's experiences in Dresden in
the fall of 1717, we can assume further that Pisendel spoke with him
about his own recent personal experiences with one of the musicians
Bach most admired: Antonio Vivaldi. For not long before Bach's visit

to Dresden Pisendel himself had just returned there-in September-
after a year's study with Vivaldi in Venice.48 Pisendel may have
shown Bach on this occasion some of the most recent Vivaldi con-
certos he is known to have brought back to Dresden with him from
Venice.49 One or another may indeed have provided a direct impetus
to the composition of several of Bach's concertos. Pisendel may also
have shown Bach one of his own compositions, considered to be a
product of his Venetian sojourn: a sonata, in A minor, for unaccompa-
nied violin.50 This work, it has been claimed (quite plausibly), could
have served as the direct model for Bach's sonatas and partitas for
unaccompanied violin.51 (It does not seem ever to have been stressed
that, with the exception of Heinrich Biber's "passagaglia"[sic] from
the "Rosary Sonatas" (c. 1674), all known precursors of Bach's unac-
companied works are associated in one way or another with Dresden:
47 The precise dating, and even the relative chronology of

swv 173a and 184a, are
not clear. Owing to the strong stylistic similarities between the two works
(Diirr refers to them as "Schwesterwerke"), it has always been assumed that they
were composed at about the same time: the birthday cantata,
swv 173a, for io Decem-
ber and swv 184a for the following New Year's day. But the external evidence
contradicts this assumption. The watermark in the autograph composing score of
swv I73a (P 42) is not the same as that found in the surviving parts of swv 184a
(St 24). In his foreword to the facsimile edition of the parts for the Fifth Branden-
burg Concerto, Schulze mentions in passing (p. 8) that the watermark in the parts
OF BWv I84a suggeststhat the cantatawas composedfor New Year'sDay, i72
.
(See also Diirr in NBA, 1/14, KB, p. 145.) But, according to Dadelsen, Beitragezur
Chronologie,pp. 82 ff., Bach's handwriting in the score of swv 173a argues that that
work was composed in December 1722. See also NBA, 1/14, KB, pp. 144-5 (re
BWv
I84a); NBA, 1/35, KB, pp. 9-Io, 132-3 (re BWV173a) and 138 (re swv 184a).
48 See col.
Jung, "Pisendel,"
49 See Karl Heller, Die deutschei3.oi. derInstrumentalwerke Vivaldis(Leipzig,
Uberlieferung
1971), especially pp. i6-i8, 26-7, and Rudolf Eller, "Vivaldi-Dresden-Bach,"
Beitriigezur Musikwissenschaft, III (1961), pp. 31-48.
50 The sonata has been
assigned tentatively to the year 1716 by Jung, "Pisendel,"
col. 1301.
51 Such a claim is made, for example, by Giinter Hausswald in the foreword to
his edition of the Pisendel sonata, published in the Hortus musicus series, XCI
(Kassel, 1952).
J. S. BACH'SCOMPOSITIONSFOR SOLO FLUTE 48 I
Pisendel's sonata; a suite published in 1682 by the Dresden composer
Johann Paul Westhoff; and Francesco Geminiani's Sonata in B Flat a
violinosolosenzabasso,which survives only in a Dresden manuscript.)52
It is tempting to think, then, that it was Buffardin who stimulated
Bach's interest in the transverse flute, that the unaccompanied partita
was Bach's first composition for the instrument, that it was written
shortly after he had met the French virtuoso, and that it was inspired
by his having seen or heard, on the same occasion, Pisendel's unac-
companied sonata for violin in the same key. The unidiomatic charac-

ter of the partita, in that event, could be attributed to the composer's
inexperience in writing for the flute.53
After the unaccompanied partita, the earliest source for any of

Bach's flute compositions is a manuscript of the Sonata in E Minor,
BWV 1034. It was copied out, most likely in 1725 or 1726, by Johann
Peter Kellner (1705-72) and entered into his large collection of Bach
manuscripts.54 Now it is not clear whether the E-minor sonata was
52
See William S. Newman, The Sonata in the BaroqueEra (Chapel Hill, 1966),
especially pp. 65, 235, and 322.
53 To add to our conjecture: If the C-major sonata, as proposed earlier, was origi-
nally composed as an unaccompanied work-perhaps as a companion piece, or even
forerunner to the partita--then its deficiencies, too, after subtracting those caused by
the addition of the bass part, could be ascribed to the composer's inexperience: inex-
perience perhaps in composing unaccompanied works for melody instruments.
54 See NBA, VI/3, KB, p. ii. The miscellany, P 804, is described at length in
NBA, V/5, KB, pp. 24-35. Kellner knew Bach personally; but it is not known pre-
cisely when they met, nor how Kellner procured the sources from which he prepared
the manuscripts brought together (after his death) to form P 804. It is clear, however,
that many, but by no means all, of Kellner's copies must have been made directly
from autographs in Bach's possession. (See Ernest May, "J. G. Walther and the Lost
Weimar Autographs of Bach's Organ Works,"Studiesin Renaissance and BaroqueMusic
in Honorof ArthurMendel,ed. Robert L. Marshall (Kassel, 1974), pp. 264-82, especial-
ly pp. 268-70.) Three fascicles in P 804-none including the flute sonata-were dated
by Kellner: two with the year 1725, one with I726. The title page and also the final
page of the collection carry the date 1726. The miscellany, however, also contains a
copy of the A-minor partita from the Klavieriibung(Bwv 827), which not only repro-
duces the title page, including the date, 1727, of the original print, but also shares its
readings--readings, that is, that postdate those contained in the Klavierbiichleinof
1725. (See NBA, V/I, KB, p. 35.) This would seem to constitute conclusive evidence
that Kellner had access to Bach sources from the early Leipzig period and not only
from the Weimar and C6then periods. Since no systematic study of Kellner's script
has ever appeared, it is not yet possible to date these MSS precisely. As Dietrich
Kilian reports in NBA IV/5-6, KB, Teilband I, pp. 194-7, Kellner's handwriting
seems to have undergone several stages of development. My own examination of
Kellner samples from P 804 reveals that the script found in BWV 1034 shares features
with both the MSS of 1725 and that of 1726 and would seem to have been copied
within that two-year span.
composedin Leipzigor C6then. The consensusof opinion has auto-

maticallyassumed a Cathen origin.55But the fact that the earliest
manuscriptof the workdatesfromaroundi725 may not be altogether
coincidental.There is evidencethat Bachhad becomeintenselyinter-
ested in the flute shortlybeforethen, specificallyin the summerand
autumn of 1724. WilliamH. Scheide was the first to observe that
Bachwrote morethana dozenchurchcantatasbetweenthe monthsof
July and November 1724that containespeciallyelaborateand often
difficultsolos for the transverseflute.56(See Table 3.)

Wouldthese obbligatopartshavebeen intendedfor Bach'sregular
flute playerin Leipzig?First of all, it does not seem that there was a
regular,competent player of the flute availableto Bach in Leipzig
beforethis time at all."' Bachbeganthe weeklycompositionof church
cantatason 30 May 1723, the first Sunday after Trinity, with Die
Elenden sollenessen,Bwv 75. The transverseflute makesits firstappear-
ance almost eleven months (and aroundfifty cantatas)later, on 16
April 1724, with a modest ripieno part in Cantata 67: Halt im
GediichtnisJesumChrist.58 Bach did not call for the instrumentagain
until he preparedthe cantatasfor PentecostMondayand Tuesday, 29
and 30 May 1724. The compositionsperformedon those occasions
were, respectively,ErhihtesFleischundBlut, Bwv 173, andErwiinschtes
Freudenlicht,Bwv 184.59These are parodiesof the two secularC6then
cantatas,Bwv 173a and I84a, mentionedearlier. Like the originals,
both of the Pentecostcantatasemploy a pair of flutes. The parts, as
stated before, are quite simple.60
On ii June 1724, the first Sunday after Trinity, Bach beganthe
so-calledSecond LeipzigJahrgangwith the famous cycle of chorale
cantatas.There areno flute partsin any of them until the four-month
seriesof flute solos underdiscussionwas initiatedwith Cantata107 on
23 July.61 After the conclusionof that series with Cantata26 on 19
ss See, for example, Eppstein, "Fldtensonaten," p. 17.

56 Scheide is credited for this by Diirr in his Die KantatenvonJohannSebastianBach
(Kassel, 197'), I, p. 50, n. 24-
57 But see the discussion below of Bach's pupil, Friedrich Gottlieb Wild.
58 The flute
parts in Cantata 181 were only added after the first performance of
13 February 1724. See Diirr, "Zur Chronologie," p. 66.
19 Diirr, "Zur Chronologie," pp. 70-I.
60 The cantatas are published in NBA, I/14-
61 It is striking that between 6 August and 8 October the flute
appears almost
every week. (The cantatas for the 12th Sunday after Trinity, 27 August, and for the
Town Council Inauguration on 28 August are apparently lost. There are no flute
parts in the cantata for the I3th Sunday after Trinity, 3 September (Bwv 33).) It is
*justas striking that from 8 October to 19 November the weekly pattern is broken and
the flute appears at two-week intervals.
J. S. BACH'SCOMPOSITIONSFOR SOLOFLUTE 483
TABLE 3
THE SERIES OF CANTATAS FROMJULY TO NOVEMBER 1724 WITH EXTENDED PARTS FOR
SOLO OR OBBLIGATO FLUTE(S)
Date Occasion swv/movt. Description

23 July 7 p. Trin. 107/6 Aria, D major:flutes i and 2 all'unisono,
tenor, continuo
(30July 8 p. Trin. 178 No flute parts)
6 August 9 p. Trin. 94/i Chorus, D major:flute obbligato
94/4 Aria, E minor: flute, alto, continuo
13 August Io p. Trin. o101/2 Aria, G minor: flute

(replaced by violin before the first
performance), tenor, continuo
IoI/6 Duet, D minor: flute, oboe da caccia,
soprano, alto, continuo
20 August I I p. Trin. I13/5 Aria, D major:flute, tenor, continuo
(27 August 12 p. Trin. ? )
(28 August Inauguration ? )
(3 September I3 p. Trin. 33 No flute parts)
io September 14 p. Trin. 78/4 Aria, G minor: flute, tenor, continuo
17 September 15 p. Trin. 99/3 Aria, E minor: flute, tenor, continuo
99/5 Duet, B minor: flute, oboe d'amore,
soprano, alto, continuo
24 September i6 p. Trin. 8/1 Chorus, E major (later D): flauto piccolo
(replaced by flute before the first
performance)
8/4 Aria, A major (later G): flute, strings,
bass, continuo
29 September St. Michael 130/5 Aria, G major:flute, tenor, continuo
I October 17 p. Trin. II4/2 Aria, D minor: flute, tenor, continuo
8 October i8 p. Trin. 96/3 Aria, C major: flute, tenor, continuo
( 5 October 19 p. Trin. 5 No flute parts)
22 October 20 p. Trin. I80/2 Aria, C major:flute, tenor, continuo
(29 October 21 p. Trin. 38 No flute parts)
(31 October Reformation 8o? 76? No flute parts)
5 November 22 p. Trin. I 15/4 Aria, B minor: flute, violoncello piccolo,
soprano, continuo
(12 November 23 p. Trin. 139 No flute parts)
19 November 24 p. Trin. 26/2 Aria, C major:violin solo, flute
(partially in unison with violin), tenor,
continuo
November, the flute disappears once again until Epiphany, that is, 25
January 1725 (Bwv 123), and only turns up sporadically thereafter.
The sudden appearance and disappearance of such extensive flute
writing, much of it quite difficult, suggests that it was inspired by the
temporary and irregular presence in Leipzig of an unusually fine vis-
iting flautist. We can, however, identify one of Bach's regular flautists
in Leipzig. On i8 May 1727, Bach wrote a testimonial for Friedrich
Gottlieb Wild, a law student, and remarked:
Mons. Wild, during the four years [i.e., from the beginning of Bach's Can-
torate] that he has lived here at the University ... not only . .. helped adorn
the churchmusic with his well-learnedaccomplishmentson the Flautetra-

versi•re and Clavecin, but has also taken special instruction from me in the
Clavier,ThoroughBass and the fundamentalrules of composition.62
It is not difficult to imagine that Wild was entrusted with routine parts
on the flute or even, in time, with more challenging ones; but if he was
really a virtuoso of the first rank as early as the summer of I724,
capable of executing the rapid and ornate parts in Cantatas 94, 1 13, 99
or i14, then one is baffled by the chronological concentration and

isolation of these parts within a single four-month period. But the
possibility that these parts were indeed written for Wild or some other
regular member of Bach's Leipzig ensemble can by no means be elimi-
nated.
The name of the player of the obbligato flute parts is unimpor-
tant.63 What is significant here is Bach's unusual, concentrated inter-
est in the transverse flute during the second half of the year 1724-
shortly before the presumed date of origin of the earliest surviving
copy of the E-minor sonata. At the least, it is altogether conceivable
that the sonata was composed at about the same time as the series of
obbligato flute parts in the cantatas.
It is worth observing, too, that in a significant respect the texture
of the E-minor sonata, scored as it is for basso continuo, is more close-
ly related to the "impure" texture of the continuo-accompanied arias
with obbligato flute than to the "pure," "clean"texture of Bach's ma-
ture "duo" sonatas from the Cothen period: the sonatas for melody
instrument and obbligato keyboard.64
The probable (approximate) date of the earliest manuscript, then,
along with a constellation of biographical and stylistic data (the con-
centrated series of flute solos in the cantatas, as well as the "reversion"
in the sonata to the texture of the continuo-accompanied solo sonata),
suggests that the Sonata in E Minor was composed in Leipzig some-
time in the late summer or fall of
i724-
The B-minor sonata, Bwv io3o, survives also in a G-minor ver-

sion. (At least the harpsichord part does.)65 While the G-minor ver-
62 TheBachReader,pp. 11xx-12.
63 That of Buffardin, who "once visited J. S. Bach at Leipzig" (see above), also
suggests itself.
64 See Eppstein, Studien,
pp. 29-30.
65 See Table 2 and NBA, VI/3, KB, pp. 32-3. Eppstein has suggested that the
sonata may also have existed in one or more preliminary forms in a different scoring,
now lost. See Eppstein, Studien, pp. 75-90.
J. S. BACH'S COMPOSITIONS FOR SOLO FLUTE 485
sion is definitelyearlierthan the B-minor,66it survivesonly in a post-
humouscopy (P ioo8), so that it is impossibleto determineits date.
But the vocally inspiredopening theme of the sonata bears an un-
canny resemblanceto the beginningof the openingchorusof Cantata
117, Sei Lob und Ehr' demhbchstenGut, a movement in G major, com-
posed some time between 1728 and 1731 and scored with flutes.67
More striking is the remarkablecorrespondence between the idiomatic
right-hand accompanimental figure in the sonata and the equally char-
acteristic, but highly unusual, obbligato cello accompaniment in the

cantata movement (see Exx. 4 and 5). Since it was in 1729 that Bach
took up the directorship of the collegium musicum and turned his
attention to the performance-and composition -of instrumental mu-
sic, it does not seem too audacious to entertain the possibility that the
G-minor version of the sonata, BWV 1030, may have been prepared
between 1729 and i731, i.e., within the chronological limits estab-
lished by DOrrand Dadelsen for Cantata 117.68 When Bach prepared
Example 4
Bwv 1030/I, mm. 1-2
Andante
fik-U
66The fair-copyautographof the B-minorversion,P

975, containsa numberof
transposition corrections that reveal that it was prepared from a source in G minor.
See the facsimile edition of P 975, edited with an afterword by Werner Neumann,
published in the series Faksimile-ReiheBachscherWerke und Schriftstocke,IV
(Leipzig, n.d. [I961?]).
67 Forthe datingof Cantata
I 17 see Diirr, "ZurChronologie,"2nd ed., p. 0o6.It
is evidentfromthe incipitscompiledby May DeForestMcAll, in herMelodic Indexto
theWorks ofJohannSebastian Bach(New York,n.d.), pp. 55-6, that no otherthemeby
Bacheven approximatelysharesthis particularrhythmic-melodicconfiguration.
68 Alfred Diirr, in his review of
Eppstein,Studien,in Die Musikforschung,XXI
(1968),pp. 332-40, also suggeststhe possibilitythat one of the versionsof Bwv 1030
with keyboardobbligatowas preparedin connectionwith Bach'sactivitieswith the
collegiummusicum.(See, especially,p. 340.) It is perhapsnot entirelyirrelevantto
mentionherethat a numberof passagesin the sonatacall to mind, in a generalway,
certainnumbersin the St. MatthewPassion:sharedrhythmicor melodiccontours;
Example5
BWV 1 17/I, mm. 1-4
Fl.trav.II
Oboe II o
ViolinoII
Viola _

the fair copy of BWV 1030o(P 975) around 1736, he decided on the new
key, B minor-a key, among other things, much better suited to the
flute than G minor.69
the use of certain syncopation patterns (for example, tied and decorated suspensions
and appoggiaturas approached and/or left by leap); florid, ornamental, but strangely
unflowing "jerky" melodies punctuated with erratic, start-stop rhythms that feature
short-long anapestic patterns; and so on. Compare, for example, the aria "Erbarme
dich, mein Gott," and the "second theme" (mm. 20 ff.) of Bwv I03o/I, or the gigue
from BWV I030/3 and "Mache dich, mein Herze, rein." But it would surely be pre-
mature to speculate as to whether the G-minor version of the sonata was composed
perhaps at about the time of the second performance of the St. Matthew Passion in
1729.
69 The watermarkof the autograph ("NM + Heraldisches Wappen von Zedwitz")
can only be dated to the second half of the 1730s. (See DOrr, "Zur Chronologie," p.
143.) But it appears in the parts prepared for the 1736 performance of the St. Mat-
thew Passion (DUrr, "Zur Chronologie," p. 115). The title page of P Ioo8, in-
cidentally, with its description of the G-minor version as a flute composition, seems to
have been written by the 19th-century owner of the manuscript (see NBA, VI/3, p.
32). Accordingly, it is not inconceivable that the G-minor version was intended for
some instrument other than the flute. Indeed, the work is often performed in G minor
on the oboe. Eppstein, however, argues on the basis of the compass of all three parts
(obbligato instrument and each of the two hands of the harpsichord part) that the solo
instrument was the flute. (He also posits yet an earlier version for two flutes and
continuo, likewise in G minor. Eppstein, Studien, pp. 75-8.)
The B-minorsonatais exceptionallydemanding.The prestofugue
with the breakneckgigue conclusion,in particular,is a tourdeforceof
technicalvirtuosity.The inordinatedifficultyof this movement,along
with the fact (touched upon earlier)that Bach's connectionswith
Dresdenhad increasedconsiderablyduringthe I73os and indeedcul-
minatedin November1736with his receivingthe title of Composerto
the Royal CourtChapel,suggestthat Bachpreparedthe finalversion
of his greatestand mostdifficultflutecomposition,too, for the master
flautistBuffardin.As in the whirlwindbadinerie that concludesthe B-

minorsuite for flute and strings, BWVIo67, Bach seems to have had
Buffardin in mind in this treacherousgigue. Buffardin'sgreatest
strength, after all, was, accordingto Quantz, playing rapid pieces
(geschwinde
Sachen).70
The A-major sonata, BWV 1032, also survives in an autograph, a

fair or revision copy dating from around I736.71 The piece, then, was
definitely composed earlier. Indeed, the retention of the same tonic
for all three movements of the sonata (A major-A minor-A major) is
totally unknown otherwise in Bach's compositions in this genre and
lends support to Eppstein's suggestion that the outer movements were
originally in C major. Eppstein, however, has argued in addition that
the work must have existed in one or more earlier versions--possibly
as a trio sonata for flute, violin, and continuo (and/or as a flute con-
certo)-that were substantially different from the surviving version.
70 Quantz's description of his teacher's

playing appeared in "Herrn Johann Joa-
chim Quantzens Lebenslauf, von ihm selbst entworfen," printed in F. W. Marpurg,
Historisch-KritischeBeytraigezur Aufnahmeder Musik, I (Berlin, 1755; reprint Hildes-
heim, 1970), pp. 197-250, especially p. 209. Martin Bernstein was the first to doubt
that the B-minor suite was written at C6then and to connect it with Dresden and with
the 1730s. See his remarks in the symposium "Bach Problems," Reportof the Eighth
Congressof the InternationalMusicologicalSocietyNew Yorkp161, ed. Jan LaRue (Kassel,
196I), II, p. 127.
71 See the facsimile edition, Faksimile-Reihe Bachscher Werke und
Schriftstiicke,
XV (Leipzig, 1979), ed. Hans-Joachim Schulze, and the discussion below. P 2 is
6i
one of the handful of manuscripts, and the only Bach manuscript, from the Prussian
State Library that had been missing since the second world war and was returned in
1977 to the Deutsche Staatsbibliothek. Its recovery has made it possible to examine
the watermark and to confirm the dating c. 1736, that had been suggested by Philipp
Spitta. The autograph is something of a curiosity in that it contains two works en-
tered one above the other. The score of the Concerto in C Minor for two harpsi-
chords, Bwv io62, is notated on folios ir to 15v above the score of the A-major sonata
which occupies the bottom three staves of those leaves and then continues on fol. I6r.
Another peculiarity of P 6 2 is that from fol. 9r to 14v the staves for the first move-
ment of the sonata, beginning after m. 62, have been completely removed from the
score. That is, over 40 percent of the movement is irretrievably lost.
Moreover, he seems to be asserting that all of these versions must have

originated in C6then.72
Actually, Eppstein is quite vague-even ambivalent--about the
chronology of this sonata. Since the autograph was not available to
him, he was obviously greatly handicapped in dealing with the prob-
lem. On the one hand he argues, among other things on the basis
of the restricted compass of the harpsichord part (from C# to b"),
that even the surviving version (but not of course the autograph

itself) must have been one of Bach's first chamber compositions-or
transcriptions--with obbligato harpsichord (p. 169). It was written,
he thinks, along with the B-minor sonata and the sonatas for viola da
gamba and harpsichord, Bwv I027-29, before the violin sonatas, Bwv
1014-i9. These, in his view, as we have seen, were not only com-
posed but given their final form during the C6then period (p. 159)-
Since Eppstein dates the violin sonatas during the "later Cothen peri-
od," i.e., 1720-3, he concludes that the flute sonatas, BWV 103o and
1032, must have been written by 1720 (p. I6o).
Eppstein is extremely critical of the musical quality of the A-major
sonata, specifically the first movement, in its present and only known
form. He points to the lack of "polyphonic intensity," the thin tex-
ture, the "careless compositional technique" (especially the redundant
E-major cadences in measures 33 and 35), and the "stylistic uneven-
ness" (pp. 100-2). Indeed, one suspects that Eppstein would have
been pleased to strike Bwv 1032 from the canon of authentic works,
were it not for the known existence of the autograph.7 His solution is
72 See
Eppstein, Studien, pp. 9o-102, I6I-9. In the following discussion page
references to specific points in Eppstein's argument will be cited in the text in paren-
theses.
73 It is of interest to compare Eppstein's characterization of the A-major sonata,
BWV 1032/I, or at least the first movement, with his arguments against the authentici-
ty of the E-flat sonata, Bwv 103I. (a) Re I032/I: "it completely lacks . . . the intensive
three-part texture which otherwise characterizes all the polyphonic movements of the
sonatas for melody instrument and harpsichord. If one disregards the passages in
parallel motion (flute, mm. I 1-12, 18-19, etc.) and the barely melodic filler parts ...
then the movement is basically for two voices" (pp. 99-Ioo). Re i03 1: ". . . a loose-
ness of polyphonic writing [Schlaffheitder polyphonen Arbeit] which finds expression
for example in the constant parallel motion between the two upper parts (third move-
ment, mm. 2, 4, 6-8, o10, 12)"(p. I79). (b): Re ro32/I: "In m. 25-37, where the 'tutti'
theme of the beginning is repeated by the flute in the dominant, it is curiously divided
between the flute and the right hand of the harpsichord in the manner of 'durchbro-
chene Arbeit.' This technique is hardly a stylistic feature of Bach's epoch" (p.
"Bach draws on this technique again . . . mm. 55 ff. . 93).
... The reason for it in this
movement, a movement so lacking in polyphony ... is probably again Bach's desire
for quasi-polyphonic activity in the two parts"(p. 93, n.). Re 1o31: "Justas unusual is
the obvious preference for small-scale voice exchange. . . . This phenomenon is re-
to conjecture--again without having been able to see the autograph-
that the version preserved in it was probably not really the final, i.e.,
"finished," version of the sonata. He accepts Spitta's dating of c.
1736-a dating now known to be correct-but he argues that it may
have been some kind of sketch or draft which in some sections was still
so unsatisfactory that Bach himself perhaps cut out the missing leaves
of the manuscript.74That is, while Eppstein seems to argue explicitly
that wv o1032 was written early in the C6then period, he is also hint-
ing at the same time that the present state of the sonata is really a

product of the 173os and not quite finished even then.
It turns out that the autograph for the first two movements is a
neatly penned fair copy, showing absolutely no traces of the sub-
stantial rewriting of parts from a pre-existent model that Eppstein has
argued must have taken place in the process of transcription and trans-
position and which he has attempted in part to reconstruct (pp. 193-
7).75 The third movement, however, can best be described as a revi-
sion copy. There are corrections of detail in all parts (especially the
bass), although none, again, is of major structural importance. More
pertinent to our discussion is the presence of a few corrections which
indicate that Bach may have been copying from a C-major source.76
The condition of the manuscript, then, reveals only that the sonata
had achieved its present compositional state some time before it was
lated to another: two-voice episodes are often simply repeated with the voices ex-
changed" (p. i79). (c): Re I032/I: "The flute ascends to d"', the highest note in the
movement, only once; for the remainder Bach apparently set c#"' as the upper limit"
(p. ioo). "Neither the highest available notes of the flute nor the harpsichord are
exploited" (p. 92). Re I031: "Incidentally, they [Bwv o031 and 102o] differ from all of
Bach's authentic compositions for flute . . in that they hardly exploit the high range
of the flute: neither exceeds d""' (p. I8o).
74 "One wonders whether this movement in its present fragmentary state can be
considered a finished composition at all, or whether it might not be the sketch of a
transcription (arrangement). Could it have been Bach himself, dissatisfied with the
result, who cut out the missing portion?" (p. ioi).
7s This confirms the conjecture expressed in NBA, VI/3, KB, p. 44, that the
reading crossed out after m. 21 in the first movement of the score resulted from a
copying error and was not a rejected compositional sketch. The latter had been sug-
gested by Georg Schiinemann in his "Bachs Verbesserungen und Entwiirfe," Bach-
Jahrbuch1935, p. 4. The commentary in Robert L. Marshall, TheCompositional Process
ofJ. S. Bach (Princeton, 1972), II, sketch no. 169, should be emended accordingly. P
6 2 does, however, contain a few interesting corrections, such as the addition of the
characteristic written-out 32nd-note mordents in mm. 3 and 4 of the theme-but not
thereafter.
76 Flute part, mm. 122-3; bass part, mm. 144, first note, m. 194, first note: all
originally a third higher. It is also conceivable, but by no means certain, that Bach
entered the right-hand stave of the harpsichord part for the entire sonata in the so-
prano clef in order to facilitate the transposition down a third (in the outer move-
copied. The correctionsof a third suggestthat, as in the case of Bwv

103o, Bach may havedecidedon a new key for the sonataat the time
he pennedthe manuscriptand transposedthe outer movements(or, to
be more cautious, at least the final movement)on the spot from C to
A. Moreover,the calligraphyindicatesthat Bachapparentlyintended
this to be the finishedversionof the work.
Whetherin fact the composition,with the likely exceptionof the
slow movement," everexistedin a substantiallydifferent"Gestalt"is
not at all clear, and certainlynot evident from the autograph.The

very fact that Eppstein is ultimately unable to decide whether the
originalform of the first movementwas a concertoor a trio sonataor
both (pp. 98, ioo) raisesthe suspicionthat there may have been no
earlierform-that Bachhad conceivedthis movement,and by exten-
sion perhapsthe finalmovementas well, from the beginningas a so-
nata for flute and obbligatoharpsichord.78
As for the suggestion(p. 162) that the characterof the opening
theme of Bwv I032--in Heinrich Besseler'sterminologya Laufthema
(Ex. 6)--argues for a quite early dating,79it shouldbe mentionedthat
the theme also bearsa strongresemblanceto the openingof the aria,
"Halleluia,Stirk' und Macht,"fromCantata29, Wirdanken dir, Gott,
a workcomposedfor 27 August 1731, i.e., at just aboutthe date sug-
gestedabovefor the challengedSonatain E FlatMajor,Bwv 103I (Ex.
7)s80
It is not necessaryto entertainhere(aloud)any naive speculations
as to whetherthis thematicsimilarityhas any bearingon the datingof
ments) from a treble-clef source in C major. It should also be mentioned that in mm.
34-5 of the second movement, Bach had at first entered the right-hand part a third too
low, which suggests that his source here was written, as presumed, in the same key (A
minor) as the current version, but, again, in the treble clef.
77 The largo survives in a source, (SPK) St 345, dating from the mid i8th century
and scored for violin, cello, and "bass," which presumably has no connection with
Bach but which seems to be derived from an (earlier?)version of Bwv 03 2/2 that was
not identical with that in P 612. See Eppstein, Studien,pp. 24, 91-2, and NBA, VI/3,
KB, pp. 43 ff. and 55 ff.; also Dtirr's review, p. 336.
78 Eppstein is quite convinced that the third movement was originally a trio. But
the fact that it could have been does not, of course, mean that it must have been. (Cf.
Diirr's expression of caution, Review, p. 335.) It is curious, moreover, that in present-
ing his argument for the trio origins of the third movement-as opposed to the uncer-
tain (trio or concerto) pre-history of the first movement--Eppstein points out its simi-
larity to the finale of the Harpsichord Concerto in E Major, BWV 1053 (p. 99).
79 One infers from the context in which this observation is made that Eppstein
would be tempted to suggest a Weimar origin for the presumed earliest version of
Bwv
1032, were it not for the fact that Bach did not begin to compose for the flute until he
settled in C6then.
80 See Marshall, The
Compositional Process,sketch no. 19.
Example6
Bwv 1032/I,mm. i-3
Vivace

Example 7
Bwv 29/3, mm. 20-3
Hal -- le - lu - jah, Stark' und Macht
piano
the pre-autograph version of the A-major sonata. But considered in

the light of the severity of Eppstein's criticism of the work's artistic
value, and, moreover, the similarity of his arguments on that point to
those adduced in his challenge to the authenticity of Bwv I031, one is
inclined to whisper one's suspicion that the A-major sonata, too, may
have been composed--as a work indeed in C major, but in its present
scoring for flute and obbligato harpsichord--in the early 173os, at
about the same time that Bach composed the E-flat sonata. Both
works, along with the Sonata in B Minor, were presumably written in
conjunction with Bach's activities with the collegium musicum;8' the
A-major and E-flat sonatas being slightly later than the B-minor and
reflecting, moreover, Bach's flirtation at the time with a lighter, sim-
pler style.
With the fifth of the "authentic" flute pieces, the Sonata in E Ma-
jor for flute and continuo, BWV 10o35, we would seem, at first glance,
to be on firmer ground again. An old archival notice on the title page
of a nineteenth-century copy declares that Bach prepared the score of
81 See also Diirr, Review, p.

340o.
the work "when he was in Potsdam in 17 [sic] for the royal valet
Fredersdorf."82 Bach is known to have gone to Berlin in 1741 and
1747. is, nonetheless,rathercurious that the identity of the com-
It
poser of the E-majorsonata has apparentlynever been called into
question.The sourcesfor the sonataarenot only all posthumous,but
they all date from the nineteenthcentury. Furthermore,the sonata
(one is inclinedto add, "asusual")is not at all "typical"of J. S. Bach.
As Eppstein observes, the composition, first of all, dispenses with
imitativepolyphonyalmostentirely;in addition,it is Bach'sonly en-

semble sonatato adopt the form of the sonatada camera.83
The biographicalcircumstancesallegedby the nineteenth-century
annotatorsto have surroundedthe workseem to have been sufficient,
however, to forestallany doubts about its authenticity.The lighter
textureand the freerformof the sonataleadEppstein,for example,to
conjecturethat Bach not only deliveredthe sonatain Potsdamin the
I740s but that he may well have composedit at that time with a view
to pleasingthe galanttaste that prevailedin Frederick'scourt.84It is
certainlynot my intentionto challengethe authenticityof the E-major
sonatahere. On the contrary,in light of Bach'sprogressivetendencies
in the I730S and early I740s and the increasinglyarchaicand con-
trapuntalnatureof his musicafterabout 1745, I am inclinedto accept
Eppstein'ssuggestionthat the sonata was composed specificallyfor
Bach'sfirst known visit to the Prussiancapitalin 1741.85
But it must be reportedthat almostnothingis knownaboutBach's
1741 visit to Berlin except that it took place some time between late
July and mid-August, and that it began before 5 August and con-
cluded after 9 August-those being the dates of two letters sent to the
composerin Berlinby his cousin,JohannElias (who was living with
the Bach household in Leipzig at the time).86 There is no evidence that
Bachwent to Potsdamduringthis sojourn.Eliasmakesno allusionor
82 A somewhatearliercopy of the sonata, P 62 (c.

I8oo), also refersto Fred-
ersdorf. See NBA, VI/3, KB, p. 22-3. MichaelGabrielFredersdorf(1708-58), an
amateurflute player, was the valet, privatesecretary,and personalconfidanteof
Frederickthe Great, as well as his flute partner(NBA, VI/3, KB, p. 24)-
83 See Eppstein,"Fl6tensonaten," pp. 14-15, 18-19-
84 Eppstein, "Fldtensonaten," pp. I4-15, 18-19-
85
Eppstein, "Fl6tensonaten," p. 19. See also Marshall, "Bach the Progressive,"
PP. 354-5-
86 The letters were in answer to two letters written
by J. S. Bach from Berlin.
Both of Sebastian's letters are lost. See Werner Neumann and Hans-Joachim Schulze,
eds., Bach-Dokumente,Band II: Fremdschriftliche undgedruckteDokumentezur Lebensge-
schichteJohann SebastianBachs, 1685-175o (Kassel and Leipzig, 1969), PP. 391-2, Bach-
Dokumente I, p. 112, and The Bach Reader, pp. 168-9.
response to any important or exciting business such as an appearance
at court that might have been mentioned in J. S. Bach's letters. In
fact, he apologizes in his letter of 5 August for disturbing the com-
poser's "present peace and contentment"--a formulation that suggests
that the Berlin visit was private-even recreational-in nature.
Moreover, it seems that Frederick was away from Berlin on a mili-
tary campaign in connection with the First Silesian War for much of
1741.87 And it would be odd that there should be absolutely no docu-
mentation of any contact between Bach and Potsdam in 1741 when

the famous visit of 1747 not only found its way into the eighteenth-
century anecdotal literature and, of course, into Forkel's biography,
but was even covered at the time by the Berlin newspapers.88 Finally,
as Wolff has demonstrated, Bach was able, even within the stylistic
framework of the Musical Offering, to make a gesture toward the mu-
sical taste of Potsdam.89 In short, Bach could conceivably have com-
posed BWV 1035 some time shortly before his celebrated journey to
Potsdam in May 1747 as a homage to be presented to the king on that
occasion. Eppstein's objection to this notion90on the ground that such
a gesture would have been unnecessary in view of the fact that Bach
had been invited to appear before the court is not compelling. Even
invited guests, after all, sometimes bring gifts. On the other hand, if
Bach had "prepared"the E-major sonata in connection with the royal
invitation of 1747, then he surely would have prepared the score not
for Fredersdorf but for the king himself. In any case the work would
surely have been mentioned somewhere in the contemporary com-
mentaries on the famous visit.
The case for i741, then, can not be discounted by any means.
Again, the musical evidence of the E-major sonata itself favors this
date. In addition, it was just about this time, shortly after his acces-
sion to the throne on 31 May 1740, that Frederick began the elaborate
expansion of his musical establishment. Within two months the king
sent Carl Heinrich Graun to Italy to recruit singers for the new opera
house whose construction he had just ordered. In the same year, 1740,
C. P. E. Bach was called to Potsdam to serve as principal cembalist.91
It is easy to imagine thatJ. S. Bach would have sought to establish an
early contact of his own with the new, musically ambitious monarch
87 See Heinz Becker, "Friedrich II," MGG, Vol. IV, col. 956.
88Bach-Dokumente II, pp. 434-5, and TheBachReader,p. 176.
89 Christoph Wolff, "New Research on Bach'sMusical
Offering,"TheMusicalQuar-
terly, LVII (1971), pp. 401-3-
90 Eppstein, p. 19.
"Fl6tensonaten,"
91 See Becker, col. 956.
by offering (through the agency of Philipp Emanuel, perhaps) a com-

position -BWV 1035 -to the king's flute partner, Fredersdorf. And it
is most plausible to think that such an approach specifically to Fred-
ersdorf would have been made before, rather than after, the arrival in
Potsdam in December 1741 of a new, most highly favored, flute part-
ner for Frederick: Johann Joachim Quantz.92 But the sonata, of
course, would have been prepared for Fredersdorf even after Quantz's
arrival-indeed at almost any time, in connection with some journey
by Bach to Potsdam of which we have no specific information at all.93

Prudence dictates, however, that speculation be confined to Bach's
two documented visits to Berlin (or Potsdam), and, on balance, the
case for a 1741 origin of the Sonata in E Major is rather more per-
suasive.
It is difficult at this point to suppress the suspicion that the prob-
lems of style, idiom and authenticity (not to mention the questions of
chronology) posed variously by the flute sonatas in C, E flat, A and E
and by the A-minor partita derive not so much from the shortcomings
or idiosyncrasies perceived in them as from our own still underdevel-
oped knowledge of the stages of Bach's stylistic development and also,
perhaps, from our underestimation of the breadth of his style-per-
haps even our refusal to acknowledge the extraordinary extent of its
range. Just as the traditional notion of the general outlines of Bach's
artistic career is apparently an oversimplification resting on largely
unsubstantiated preconceptions, it seems that our notions of the Bach
style itself have been unjustifiably restricted and marked by our in-
ability to imagine that the greatest master of the arts of counterpoint,
the "fifth evangelist," could, at certain times and on certain appropri-
ate occasions, also deliberately cultivate a lighter, more popular
idiom.
A proper understanding of the music has also been hindered by
one further assumption: that Bach's music is in some particularly pro-
found way abstract, "pure" music-musica mundana. Bach, in this
view, was, in the most literal sense, "profoundly" unconcerned about
92 See Reilly, p. xxii.

93 Could it be, for example, that Bach went to Berlin at the beginning of 1744 to
attend the wedding of Emanuel and met one or another of his son's musical col-
leagues, including Fredersdorf, on that occasion? We are particularly ill-informed
about Bach's activities and movements in the early I740s. For example: with the
exception of a number of receipts and routine letters of recommendation, no letters--
in the proper sense of the term-survive in his hand from the period fromJuly 1741 to
July 1748. See Kalendariumzur Lebensgeschichte Johann SebastianBachs, edited by the
Bach-Archiv, Leipzig (Leipzig, 1970), especially pp. 5i-6.
the actual sounding, acoustical realization of his creations, uncon-
cerned about their technical difficulties or their idiomatic effective-
ness, and "sublimely" unaware of the abilities of the musicians ac-
tually available to him.94 But perhaps Bach was quite concerned about
all these things. As noted above, the flute parts in the works known to
have been written for C6then are quite simple; and in the permutation
fugue Bach developed a choral form that was relatively easy for
young, nonprofessional singers to learn; finally, the opening choruses
of the chorale cantatas make use characteristically of a simple soprano

cantus firmus that obviously takes cognizance of the limitations of
young children. Conversely, the trumpet parts of the Leipzig cantatas
surely testify to the specific, perhaps unique, abilities of the brilliant
virtuoso, Gottfried Reiche, Bach's regular trumpeter. And there is, in
the last analysis, no reason not to assume that Bach-who was pre-
sumably able to play all of his own keyboard music-had personal
knowledge of, say, a violinist who would have been able to play the
unaccompanied sonatas and partitas (Pisendel? whose own sonata for
unaccompanied violin is in some ways technically more difficult than
Bach's), or a flute player-who shall remain nameless-who would
have been able to play the opening Allemande of the A-minor partita
or the concluding gigue of the B-minor sonata.
And there is every objective indication that it may in fact have
mattered more to Bach than to most of his contemporaries whether,
say, an ensemble sonata was played by a particular solo instrument.
None of the original sources, definitely none of the surviving auto-
graphs, suggest that any sonata for melody instrument and keyboard
could be played, for example, alternatively on "violino o flauto traver-
so." Could it be that Bach had a quite definite conception of the prop-
er style of music for the flute that was clearly different from music for
solo violin, that music for the one instrument should not necessarily
be played (or even playable) or be effective on the other, that is, that
he was "profoundly" concerned about the "instrumental, sounding,
and other performance-related realities" (Eppstein p. 32)?
When Bach made the acquaintance of the transverse flute it was
pre-eminently a French instrument nourished by a French reper-
toire-a musical tradition characterized by light textures, clearly ar-
ticulated phrases, and dance-inspired rhythms and forms. Should it
be unimaginable or even particularly surprising that in his own cham-
ber music for the instrument Bach would frequently wish to draw on
94 This view is eloquently propounded throughout Eppstein, Studien. See, in par-

ticular, pp. 32-3.
this fresh style and not invariably impress upon it all the composition-
al rigor and sophistication at his command?
To summarize briefly: A review of the sources, in conjunction

with a consideration of some neglected biographical information and a
number of stylistic observations, has suggested the outlines of a chro-
nology of Bach's music for solo flute.
(i) The first of Bach's flute compositions seems to have been the

unaccompanied partita, BWV 101o3, a work presumably composed
around 1718 for Pierre Gabriel Buffardin, whom Bach would have
met in Dresden in the fall of I717.
(2) The C-major sonata, BWV 10o33,too, may have been composed
originally as an unaccompanied work at around the same time--or
even slightly earlier.
(3) The earliest, lost, versions--if such indeed existed -of the two
sonatas for flute and harpsichord obbligato, BWV 1030 and 1032, may
have been composed, in different instrumentations, in C6then.
(4) The E-minor sonata, BWV 1034, may have been composed in
the late summer or fall of 1724 in Leipzig (that is, shortly before the
probable date of the earliest surviving manuscript) at a time when
Bach was writing an elaborate series of obbligato flute parts in his
church cantatas-in themselves an indication that he then enjoyed the
services of a virtuoso of the first rank.
(5) The G-minor version of Bwv I03o, now scored for flute and
harpsichord obbligato, may have been prepared (or composed) around
1729, perhaps in connection with Bach's having assumed the director-
ship of the collegium musicum.
(6) Sometime in the early I73os, and before i734, Bach probably
composed the E-flat sonata for flute and harpsichord obbligato (Bwv
i031). The composition then evidently served as a model for Philipp
Emanuel Bach's Sonata in G Minor for violin and harpsichord (Bwv
1020).
(7) At about the same time, perhaps in the year I731, Bach may
have prepared (or composed) the C-major version of BWV 1032.
(8) Fair copies of both aBWV103o and 1032, the former transposed
on this occasion to B minor, the latter to A major, were prepared
about 1736, perhaps, again, with a view towards a performance with
the collegium musicum in Leipzig-or, conceivably by Buffardin in
Dresden.
(9) BawV I35 was probably composed at the time of Bach's first
known visit to Potsdam, in 1741, for Frederick the Great's flute part-
ner, Fredersdorf.
Bach's involvement with the flute as a solo chamber instrument,
then, apparently was not restricted to his five-and-a-half year sojourn
in C*then but rather spanned close to a quarter-century of his matur-
ity, extending from around 1718 to around 1741 .
Since the second world war Bach research has been dominated,
appropriately, by textual criticism.9s But the text-critical techniques
that have achieved such sensational results in investigations of Bach's

vocal music seem to have reached--or are rapidly approaching--their
limits in dealing with his instrumental music, since (as mentioned at
the outset) virtually no primary sources for these compositions sur-
vive, and it is not very likely that many new ones will turn up. Faced
with such a situation we can obviously either abandon the effort to
solve the remaining problems, or we can attempt to approach them
anew. We should, at the least, attempt to augment the meager, frag-
mentary, inadequate or contradictory source evidence we might have
with whatever potentially enlightening information we are able to ob-
tain from other domains.
Two important areas of Bach research that have stood in the shad-
ow of textual criticism over the past few decades--biographical stud-
ies and stylistic analysis-have in fact registered substantial achieve-
ments during that time as well.96 But until now these three disciplines
have largely been pursued independently of one another, indeed by
different specialists in each case. I suspect that in the future the most
important advances in Bach scholarship are likely to be achieved only
by uniting these disciplines.
Of course, it is possible, indeed probable, that even if we make
this effort, we shall never be able to obtain the quality of proof and
verifiability we have become accustomed to. But do we really have
much choice?
University of Chicago
This study was originallydeliveredas a paperat the MidwestChaptermeet-

ing of the AmericanMusicologicalSociety in November, i977; at Brandeis
University; and at the Bachfest of the Neue Bachgesellschaft,Marburg,
95 The "new chronology" and the Neue Bach-Ausgabe, of course, epitomize this.
96 As a symbol for the former the three volumes of Bacb-Dokumente, ed. Werner
Neumann and Hans-Joachim Schulze, of the Bach-Archiv, Leipzig (Leipzig and
Kassel, 1963, 1969, 1972) may serve admirably; Eppstein's monograph on the
sonatas for melody instrument and obbligato keyboard, mentioned frequently in the
course of this study, serves just as well as a symbol for the latter.
1978. The presentrevisedand expandedversionhas greatlybenefitedfrom

comments made during the discussions following those presentations.I
should like in particularto express my gratitudeto Joshua Rifkin, Hans-
JoachimSchulzeand AlfredDtirrfor havingcalledmy attentionto a number
of items that had escapedmy notice.

Flute and Bach

Uploaded by

Copyright:

Available Formats

Flute and Bach

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

Flute and Bach

Uploaded by

Copyright:

Available Formats

J. S.

Bach's Compositions for Solo Flute:

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

I See Georg von Dadelsen, Beitrdgezur Chronologie der WerkeJohannSebastianBachs

nal performingmaterialshavesurvived,therearevirtuallyno primary

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

The basicquestionsaboutthe flute sonatashavestill not been sat-

der Konzerte Joh. Seb. Bachs,"FestschriftMax SchneiderzumachtzigstenGeburtstage,ed.

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Becauseof the strong doubts concerningtheir authenticity,the three-voice

We learn from Hans Eppstein that at least the C-major sonata-

Call No. Copyist Date Title

Sonata in E-Flat Majorfor Flute and HarpsichordObbligato,swv1 o30

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Sonata in G Minorfor "Flute"and HarpsichordObbligato,swv 1020

The principal source for the C-major sonata was written by C. P. E.

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

10 See Hans-Joachim Schulze, "Die Bachuiberlieferung:Plidoyer fOr ein not-

Now it seems implausiblethat the young and, like most young

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

been composed around the same time as the Unaccompanied Partita

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

17 See Ulrich Siegele, Kompositionsweiseund Bearbeitungstechnik

would wish to attributea workof his own entirelyto his father,even

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Call No. Copyist Date Title

Sonata in E Minor for Flute and Basso Continuo,

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Sonata in E Major for Flute and Basso Continuo, Bwv 1035

The G-Minor Version of swv 1030

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

28 Earlyversionsof the A-minorand E-minor

This issue cannot be pursued here. Suffice it to say that it is of

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

Musicians Johann Heinrich Freytag and Johann Gottlieb Wtirdig,

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

38Bach's Cothen flautists are identified name in Ernst

From [the Turkish town of] Bender [JohannJacob] journeyedto Con-

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

43See The Bach Reader, pp. 207-8.

never, up to that time, composed anything at all for the instrument.

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

47 The precise dating, and even the relative chronology of

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

After the unaccompanied partita, the earliest source for any of

composedin Leipzigor C6then. The consensusof opinion has auto-

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

ss See, for example, Eppstein, "Fldtensonaten," p. 17.

Date Occasion swv/movt. Description

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

the churchmusic with his well-learnedaccomplishmentson the Flautetra-

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

The B-minor sonata, Bwv io3o, survives also in a G-minor ver-

Downloaded from http://online.ucpress.edu/jams/article-pdf/32/3/463/152127/831251.pdf by guest on 07 June 2020

66The fair-copyautographof the B-minorversion,P