Dyer 1984 Latin Psalters, Old Roman and Gregorian Chants
Dyer 1984 Latin Psalters, Old Roman and Gregorian Chants
Dyer 1984 Latin Psalters, Old Roman and Gregorian Chants
ORIAN CHANTS
.JOSEPf I DYER, BOSTON (USA)
The history of liturgical chant in the West has been probed with a wide variety of research tools,
but very little attention has been focused on the texts of the chants. Even studies'which concern
themselves with the choice of texts and their ordering within the liturgical year make only cur-
sory mention of the literary form of the biblical text. A thorough examination of these textual
matters represents an undertaking of considerable scope, for it requires the detailed comparison
of chant texts in the manuscripts with an extensive number of Psalter witnesses. In order to
make this initial survey more manageable I have decided to restrict the field of investigation to
the Mass chants of Lent in the Old Roman and Gregorian repertoires. The conclusions derived
from this body of material do not differ from those which proceed from comparisons I have
made covering the entire liturgical year. In the following pages I will review: {1) the Latin text
forms extant during the formative stages of the chant repertoire, (2) the dissemination 'and
geographical influence of these text forms, (3) the types of textual variants encountered in the
Mass chants, and (4) the way in which textual variants establish relationships within and be-
tween the two chant traditions.
The texts of the variable chants of the Mass have been studied previously by Bruno Stablein and
Thomas Connolly2. They both compared found in specific categories of Old Roman
and Gregorian chant with an array of Psalter manuscripts. Though the data on which they based
their studies was similar, it led them to strikingly dissimilar conclusions
3
Lack of either both
compared variant readings found in specific categories of Old Roman and Gregorian chants with
an array of Psalter manuscripts. Lack of a either common consensus or a uniform methodology
indicates that a fresh review of textual variants and the legitimate inferences which can be drawn
from them would be profitable.
Fundamental to any understanding ohhe psalmic foundation of the Mass chants is an inventory
of the text forms available to the early All of the Latin translations of the psalms,
Fur a brief introduction to the topic sec W. Ape], Gregorian Chant, Bloomington (Indiana) 1959,
p. 87-98 and E. Cardinc, Psautiers anciens ct chant grcgoricn, in: Richesses ct dCficiem:es des anciens
psauticrs latins ( = Collectanca Biblica Latina 13), Rome 1959, p. 249-58. The principal editions of texts
are: R.-J. Hcsbert, Antiphonalc Missarum Sextuplcx, Brussels 1935, and: Corpus Antiphonalium
Officii, 5 vuls., Rome 1963-1975. There is a modern text edition of all the Old Roman Mass chants in
P. Cutter, Musical Sources of the Old-Roman Mass ( = Musicological Studies and Documents 36), Rome
1979. This can be used only with the corrections provided by Th. Connolly in his review of the Cutter
edition, in: Early Music History 2, ed. I. Fenlon, Cambridge 1982, p. 363-69.
2 Karl Marbach's compendium, Carmina Scripturarum, Strassburg 1907, Hildesheim 1963, is useful for
locating biblical sources but does not consider the manner in which the chant texts differ from the
Vulgate used by Marbach. A comparison of chant texts with scriptural and patristic sources is P.
Peitschmann, Die nicht dem Psalter entnommenen Messgesangstucke auf ihre Textgestalt untersucht,
in: Jahrbuch fiir l.iturgiewissenschaft 12 (1932), p. 87-144. An illuminating study by H. Hucke (Die
Texte der Offertorien, in: Speculum Musicae Artis. Festgabe fur Heinrich Husmann, Munich 1970,
p. 193-203) concerns the choice and order of verses in the offertories.
3 B. Stablein, Nochmals zur angcblichen Entstehung des gregorianischen Chorals in Frankenreich, in: Ar-
chiv fiir Musikwissenschaft 27 (1970), p. 110-21; Th. H. Connolly, The Grad.uale of S. Cecilia in
Trastevere and the Old Roman Tradition, in: Journal of the American Musicological Society 28 (1975),
p. 413-58.
11
with the exception of Jerome's psalterium iuxta Hebraeos, were based on the Greek text which
embraced its own range of variants even before the Latin translation process began. It. was to
bring order to this textual confusion that Origen edited the greatest biblical research tool of anti-
quity. In parallel columns he arranged the Hebrew text, the same transliterated into Greek
characters, three of this newer Greek translations and his own edition of the Septuagint version.
Because of the six text forms it contained Origen's critical edition became known as the Hexapla.
Neither the date nor the provenance of the earliest Latin translations of the psalms is known.
There were many such translations, however. Augustine claimed that anyone with a smattering
of Greek would try his hand at a Latin version. Rome may have possessed a semi-official Latin
bible by the middle of the third century. Africa would have needed .a Latin Psalter at a very early
date as well; its existence is implied by Tertullian (ca. 160 -after 220). All of these earliest transla-
tions are known collectively as "Old Latin", a textual tradition which scholars generally divide
into a European and an African branch".
From the standpoint of the cham texts three Latin versions of the psalms are relevant: (1) the
Old Latin tradition, which antedates the activity of Jerome, (2) the related Roman Psalter, and
(3) the so-called "Gallican" Psalter. A fourth translation, produced by Jerome directly from the
Hebrew text, never found a place in the liturgy, though it surpassed in accuracy all previous
Other terms for these versions, like "Vulgate Psalter" or "!tala" are wanting in preci-
sion and should be discarded. Medieval Vulgate bibles did not always contain the same Psalter
version, and the term "Itala" derives from a single statement of Augustine which has resisted
all efforts at clarification
6
The portion of the Old Latin bible most strongly represented in the manuscript tradition is the
Psalter (161 witnesses in the Verzeichnis of the Beuron Latina). Many of these Psalters are
fragmentary, however, and the full extent of the tradition and its comple.te range of variants will
not become evident until.the Latina Psalter (five volumes) is published. In the meantime
research can be based on the Old Latin Psalter variants collated by Dam Weber in his exemplary
edition of the Roman Psalter and on the older studies of Arthur Allgeicr
7
The Old Latin
4 The distinction is based on the specific Latin words chosen to render the for cxampl(; doxa is
translated "claritas" by the African and "gloria" by the European text of the Psalter. Sec P. Capelle, Lc
textc du psautier latin en Afrique ( = Collectanea niblica Latina 4), Rome 1913, p. 30. T. A. Marazucla
(La Vetus Latina Hispana 1, p. 167-68) would abandon the traditional African-European division in favor
of distinction drawn on national lines, thus defining a Vetus Latina ltalica, Africana or Hispana.
5 The Roman Psalter can be compared with the Old Latin and "Gallican" versions in th.e edition of Dom
R. Weber, Le psautier romain et les aut res anciens psautiers latins ( = Collectanea Biblica Latina 10),
Rome 1953. The critical edition of the "Gallican" Psalter is published as vol. 10 of the Biblica Sacra
Iuxta Latinam Vulgatam Versionem ad Codicum Fidem, Rome 1953. The critical edition of the ittxt.t
Hcbraeos is H. de Sainte-Marie (Ed.), S. Hieronymi Psalterium iuxta Hebraeos ( = Collectanca Biblica
Latina 11), Rome 1953. For general surveys of the Latin bible in the Middle Ages the best sources arc:
B. Peebk-s, Bible, Latin Versions, in: The New Catholic Encyclopedia 2, p. 436-50; M. Tuya I.J.
Salaguero, lntroducci6n a Ia Biblia 1, Madrid 1967, p. 530-44; R. Loewe, The Medieval History of the
Latin Vulgate, in: Cambridge History of the Bible 2, ed. G. W. H. Lampe, Cambridge 1969, p. 102-154.
Very useful for the Hebrew background of the psalms and an extensive bibliography is Max Haas, Zur
Psalmodie der christlichen Frtihzeit, in: Schweizer Jahrbuch fUr Musikwisscnschaft, Neue Folge 2
(1982), p. 29-51. . .
6 "In ipsis autem interpretationibus [of Scripture] ltala ceteris praeferatur, nam est verborum renacior cum
perspicuitate sententiae." De doctrina christiana 2.15.22 (Patrologia Latina 34, col. 46; hereafter ab-
breviated PL). For an exhaustive discussion of possible meanings see Marazuela, La Vetus Latina Hispana
1, p. 157-72.
7 See note 5 above. A. Allgeier, Die altlateinischen Psalterien, Freiburg i. Br. 1928; idem, Die Psalmen der
Vulgata, Paderborn 1940; idem, Die Oberlieferung dcr alten lateinischen Psalmeniibersetzungen und
ihre kulturgeschichtliche Bedeutung, in: Publikationen der Freiburger Wissenschaftlichen Gesellschaft
20, Freiburg i. Br. 1931.
12
Psalter manuscripts, which stretch back to the sixth and seve.nth centuries, represent the oldest
version of the psalms in Latin. Only quotation!:i in the works of the Fathers, also to be included
in the vetus Latina, are more venerable.
The earliest Old Latin Psalter, a sixth or seventh-century manuscript (a) in the Chapter Library
at Verona, contains a text form which even at that early date had undergone several revisions.
Current opinion, while recognizing in it correspondences with the version used by Augustine
in his narrationes in psalmos (before 415), considers it a basically Italian text. Its
closest relative is the sixth-century Psalter of St. Gall
Another branch of the Old Latin tradition, the so-called "psautier gaulois", is represented by a
group of manuscript!:i headed by the sixth-century Psalterium Sangermanense (y). This recension
was known to Jerome, since he criticized a translation error in it. Other important witnesses of
the psautier gaulois are the eighth-century Corbie Psalter (o) and a sixth-century Psalter from the
South of France {e.)''. There is Gitical disagreement about the source of the peculiar gaulois
readings. Whatever their point of origin, however, this Psalter version held a dominant position
in Gaul until the time of Charlemagne.
The most important branch of the Old Latin Psalter tradition broadly defined is the "Roman"
Psalter, which the patristic scholar Alberto Vaccari sees as standing particularly close to the
"common root" of this tradition
1
c. As early as the fourth century the Roman Psalter was well
established in Italy. Its favored status at Rome engendered the name by which it is universally
known. Of special significance in the history of chant, the Roman Psalter l:ield sway when and
where the sung texts of the Mass were clothed with their melodies. The thousands of passages
collated for the present study bear this out, despite the frequency of contamination from other
Old Latin versions
11
For centuries it had been assumed that the Roman Psalter was the first of the three revisions of
the Psalter known to have been completed by Jerome. In the preface to his second Psalter revi-
sion, he referred to an earlier effort: "psalterium Romae dudum positus emendaram et iuxta Sep-
tuaginta interpretes, licet cursim, magna illud ex parte corrcxeram."
12
This cursory correction
of an older Latin Psalter was simply identified with the Roman Psalter until this attribution was
rejected by Dom Donatien de Bruyne. He claimed that, in spite of Jerome's disclaimer ("licet
cursim"), the Roman Psalter contains entire categories of textual flaws which the renowned
translator would never have tolerateu under any circumstances". Alberto did not
8 For complete citations sec r he List of s;mrccs. Concerning the Verona Psalter Weber maintains that
"dans son fonds, cc Psauticr n'cst pas africain, mais italicn" (Lc psauticr romain, p. x). Sec also D. de
Bruync, Saint Augustin: rcviscur de Ia lliblt, in: Miscellanea Agostiniana 2, Rome 1931, p. 544-75.
9 Complete citations will be found in the List of Sources. The Sangcrmancnse was edited by Dom Sabaticr
in: Bibliorum Sacrorum Latinac Vcrsioncs Antiquac scu Vetus ltalica 2, Rcims 1743.
10 As maintained in a review of the Weber edition in: Biblica 36 (1955), p. 100 ("a questo commune radice,
nessuno sta cosl vicino come il Romano"). Allgeier (Die Oberlieferung, p. !) considers the Augustine-
Veronese Psalter closer to this archetype.
II This combination is a normal situation in the history of psalmic textual transmission. for example,
Spanish influence has been claimed in the Verona Psalte;- by H. Schneider, Ocr altlateinische Palimpsest-
Psalter in Cod. Vat. lat. 5359, in: Biblica 19 (1938), p. 36182.
12 For a critical edition of the complete preface see Biblica Sacra luxta Latinam Vulgatam Versionem 10,
p. 304. .
13 D. de Bruync,. Le problcmc du psautier romain, ii1: Revue Benedictine 42 ( 1930), p. 101-126. De Bruync
enumerates these as: harmonizations, dependency on the "Western" Greek text usually rejected by
Jerome, misunderstanding of the Greek, irreconcilable differences in understanding of the Greek be-
tween the Roman Psalter and Jerome's known translations, deformation uf the text to arrive at an easy
solution to textual difficulties, stylistic crudities which Jerome would have improved. He points out,
furthermore, that Jerome's enemies never reproached him with the shortcomings of the Roman Psalter.
De Bruync held that traces of Jerome's (los.) first version arc preserved in the Commentarioli on the
psalms and in his letters written in the late 380s.
13
dispute the impressive array of evidence assembled by de Bruyne. He did maintain that, while
Jerome indeed recognized the deficiencies of his first attempt (the Roman Psalter, according to
Vaccari), he had neither the time nor the critical source material in Rome to make all the
necessary corrections
1
l.
In support of Jerome's authorship it can be observed that the patristic era tolerated a range of
imperfection in the sacred texts which would not be sanctioned by modern conceptions of tex-
tual accuracy. Jerome and his contemporaries believed that "singula verba scripturarum singula
sacramenta sunt; ... thesaurum sensum divinum habemus in verbis vilissimis"
1
'. In an exten-
sive study of Jerome's attitude toward contemporary biblical translations, G. Q. A. Meershoeck
concluded that Jerome accepted a certain degree of inaccuracy in deference to the text handed
down from previous generations: "aupres de la sublimite du contemi, toutes les imperfections ex-
terieures s'annihilent."
11
' Although de Bruyne's objections to Jerome's authorship of'the Roman
Psalter are forceful, they have not been received without substantial criticism by Vaccari and
other biblical scholars.
Even if the Roman Psalter cannot be identified with Jerome's first revision, there is no problem
in identifying his second revision: the well documented and widely disseminated "Gallican"
Psalter. After Jerome moved to Palestine in 385, he gained access to the authentic copy of
Origen's Hexapla. On the basis of the fifth column, Origen's own edition of the Septuagint,
Jerome framed his Psalter revision (ca. 389). He also referred to other columns on the page before
him and went directly to the Hebrew text for some rcadings
17
The resulting version was far
more exact and faithful to the Greek, but it was still not a fresh translation. Jerome felt constrain-
ed to make only the minimum number of corrections in the text of the accepted {Roman) Psalter.
He preserved an older vocabulary in making these corrections, but did not hesitate to impose
his own preferences in place of certain Old Latin and Roman choices
1
x. This "Hexaplaric"
Psalter- a name we will pr.efer to the more customary (and misleading) "Gallican"- was essen-
tially a scholarly edition. In it Jerome incorporated marks, derived from Origen, to indicate mat-
ter found in the Septuagint but not in the Hebrew (obelus -) or vice versa (asterisk*). In the
preface to this Psalter, composed as a letter to his Roman friends Paula and Eustochius, he does
not imply that this new revision was intended for the wide liturgical use it eventually enjoyed
in the West. At first the new recension encountered opposition, and its author had to justify its
existence to friend (Augustine) and foe (Rufinus) alike. In a well known letter to the Gothic
priests Sunnia and Fretala he acknowledged the shortcomings of the Hexaplaric Psalter and pro-
14 A. Vaccari, I saltcri di S. Girolamo e di S. Agostino, in: Vaccari, Scritti di crudizione e di filologia 1,
Rome 1952, p. 211-21. E. F. Sutcliffe (Cambridge History of the Bible 2, p. 84) claims that the arguments
of de Bruyne "have not been found convincing", a view shared by Allgeier (Die Psalmen der Vulgata,
p. 122-23). The chronology of Jerome's activity may be surveyed in Tuya and Salaguero, lntroducci6n
1, p. 530-37.
15 Tractatus de psalmo 90? (Corpus Christianorum: Series Latina 78, col. 130-131).
16 G. Q. A. Meershoeck, Le latin biblique d'apres Saint Jerome(= Latinitas Christianorum Primaeva 20),
i j m e g e ~ 1966, p. 241; see also pages 4-30 and 241-44. Jerome's methods have been studied by A.
Thibaut, La revision hexaplaire de Saint Jerome, in: Richesses et deficiences des anciens psautiers latins,
p. 107-149.
17 F. Stummer, Einfiihrung in die lateinische Bibel, Paderborn 1940, p. 87. The Greek text of this fifth
column is not preserved inde_pendentl_y.
18 Jerome preferred, for instance, "eruere" to the Old Latin-Roman "eripere" in psalms 24:17 and 20,30:3,
31:19, 38:9, 39:14, 85:13; "redimere" to the Old Latin-Roman "liberare" in psalms 43:26, 48:16, 54:19,
71:14, 77:42.
14
vided a list of
Jerome's third and final revision of the Psalter, the psalterium iuxta
Hebraeos, may be passed over with only brief mention. It formed the natural complement to his
other Old Testament translations directly from the Hebrew, but it never entered the liturgical
mainstream.
Since the relevance, or lack thereof, of the Hexaplaric Psalter has been a point of issue in previous
studies of the Mass chants and their texts, a short survey of its dissemination and possible in-
fluence is necessary. Despite its customary name ("Gallican") it did not hold a favored position
in Gaul until a relatively late date, about the beginning of the ninth century. (Amalar, for in-
stance, quotes regularly from the Hexaplaric Psalter.) Until the eighth century the psautier gaulois
seems to have been the Psalter used in public worship. It was also strong enough to influence
the St. Gall While it is true that six of the seven earliest manuscripts of the Hexaplaric
Psalter originated in Gaul, three of these are instruments of study, duplex and triplex Psalters,
not specifically liturgical manuscripts
10
The ascendency of the Hexaplaric Psalter began at the
time of Charlemagne or shortly before. Although the emperor did not give it official status, the
Hexaplaric Psalter came to be included in the splendid bibles produced by the scriptorium at
Tours under the supervision of Alcuin and his successors as abbots
11
Exactly how and by what stages the Hexaplaric Psalter ousted the psautier gaulois in the recita-
tion of the Divine Office has not been clarified, but a link with monastic reform movements
can be assumed. Since every monk probably learned the Psalter by heart from constant exposure
to its recitation in choir, manuscript evidence may be neither extensive nor reliable in tracing
the dissemination of a particular version in a liturgical context. Walafrid Strabo (ca. 808-849)
chronicled the change taking place during his own lifetime, while noting that the "Romani" still
clung to the Roman Psalter. He also claimed that the "Galliean" epithet derived from the legen-
dary introduction of the Hexaplaric Psalter into Gaul by Gregory of Tours (538/9-594y
2
In England the Roman Psalter began to yield to the Latin Hexapla in the Divine Office after
the middle of the tenth cenLury. The scriptorium at Winchester served as the principal agency
for its dissemination. Canterbury remained conservative: until the mid-eleventh century the
Roman Psalter persisted there. A manuscript in the British Library, Arundel115, a Roman Psalter
in use at Canterbury, was converted into a Hexaplaric Psalter at about this tirne
2
'. Continental
influence - the prestige of the Tours along with the monastic reforms initiated by con-
19 "Veterum. imerpretum consuetudinem mutare noluimus, ne nimia novitate lectoris studium ter-
reremus:' Ep. 106.12 (PL 22, col. !!43; Corpus Scriptorum Ecclesiae Latinorum 55, p. 255). A critical edi-
tion of the letter is published in the Vulgate edition cited in note 5 above, p. 8-42. De Bruyne was of
the opinion that the entire letter was an elaborate public-relations device: La lettre de Jerome a Sunnia
et Fretela sur le psautier, in: Zeitschrift fur die neutcstamentliche Wisscnschaft und die Kunde der
alteren Kirche 28 (1929), p. 1-13. Cf. also Jerome's letter to Augustine (Ep. 112, in PL. 22, col. 928;
Corpus Scriptorum Ecclesiae Latinorum 55, p. 389) and the Ep. adv. Rufinum 14 (PL 23, col. 468;
Corpus Christianorum: Series Latina 79, p. 86-87).
20 Rome, Vat. Reg., lat. 11, a duplex Psalter with the Hexaplaric text and the iuxta Hebraeos on facing pages;
Leningrad, Public Library, F. v. I. n. 5, sets side by sidethe iuxta Hebraeos, Hexaplaric and gaulois texts
of the psalms.
21 B. Fischer, Bibelausgaben des friihen Mittelalters, in: Settimanc di Studio del Centro Italiano sull'Alto
Medioevo 10, Spoleto 1963, p. 593.]. Gribomont, on the other hand, believed that Alcuin's bible was
"uniformement impose": L'Eglise et les versions bibliques, in: Maison-Dieu 62 (1960), p. 59.
22 De rebus eccl. 25 (PL 114, col. 957); seeS. Berger, Histoire de Ia Vulgate pendant les premiers siecles du
moyen age, Paris 1893, p. 4 and 185-258. The term "psautier hexaplaire" was suggested by Dom de
Bruyne, La reconstruction du psautier hexaplaire latin, in: Revue Benedictine 41 (1929), p. 298-99.
23 K. Wildhagen, Das Psalterium Gallicanum in England und seine altenglischen Glossierungen, in:
Englische Studien 54 (1920), p. 34-45. H. H. Glunz (History of the Vulgate in England from Alcuin
to Roger Bacon, Cambridge 1933, p. 59 ff.) maintained that the Latin Hexapla text was introduced into
England (via the Alcuinian revision) with the reforms of King Alfred (871-901).
15
gregations from across the channel - were significant in the changeover. Irish monks were far
ahead of the English in adopting the Hexaplaric Psalter, but their customs did not immediately
transfer to the continent
24
The rate of change in Upper Austria seems to have been comparable
to that observed in England, or even slightly slower. Conversion to the Hexaplaric version was
complete only in the twelfth century''. The e..trliest more or less thorough exposition of psalm
singing, the tenth-century Commemoratio Bn'Vis de torzis et psalmis modulandis, emanated from
a milieu in which the Psalterium Romanum was still in practical use'".
This competition between Psalter versions which arose during the course of the late eighth and
ninth centuries in Gaul persisted in the medieval liturgy. Italy began to yield to the Hexaplaric
Psalter unJer German influence in the tenth and eleventh centuries. Pope Sixtus IV (1471-1484)
restricted the use of the Roman Psalter for chanting the psalms of the Divine Office to the area
delimited by the fortieth milestone from the city. (A special indult permitted its use at St. Mark's
in Venice until1808.) Pius V (1566-1572) restricted the Roman Psalter to the canons of St. Peter's
basilica, but Clement VIII in the bull Cum sanctissimum (7 July 1604) defended its use in the
liturgy against attempts to convert all chant texts to the reaJi'ngs of the Hexaplaric Psalter
27
With so many opportunities for the harmonization of one version with the other, it is
remarkable that so little of it took place in the liturgical manuscripts. The individuality of each
version was defended at length in the treatise De varia psalmorum atque cantuttm modulatione,
once attributed to Berno of Reichenau (d. 1048). The Roman Psalter, firmly entreched in the
chants of Mass and Office, maintained its place despite (in the words of Dom Cardine) "lc picgc
tendu par Ia dualite des
Thus far I have sketched briefly the textual conditions which prevailed as the chant repertoire
was being shaped in tlie fifth and sixth centuries. I have also added a few notes on the displace-
ment of all previous Psa1ter versions by the Hexaplaric Psalter, a historical development which
has yet to be fully clarified. Since I am likewise concerned with arriving at a secure methodology
for analyzing the texts of the Mass chants, I would like to offer a critique of conclusions reached
by Bruno Stiiblein and Thomas Connolly in their textual investigations footnote 3 above).
Stablein analyzed the gradual responsories of the Old Roman and Gregorian traditions, 88 pieces
in all. He catalogueJ passages in which the Hexaplaric Psalter differed markedly from the Roman
Psalter, the acknowledged basis of the chants he examined. He found absolutely no influence
which could be attributed to the Hexaplaric version. His useful study illustrated what might
have happened if this Psalter version had exercised its influence concurrently with a supposed
frankish revision of Roman chant, a revision which some scholars claim led to the creation of
what is now known as "Gregorian" chant. Despite its contribution, Stablein's intervention was
largely unnecessary. He may have made unconscious assumptions about the "Gallican" Psalter
and its position at the time of Charlemagne. As we have seen, it commanded no preferred status
during this critical perioJ in the history of chant.
In a study which appeard a few years after Stablein's, Thomas Connolly took the view that thl'
Hexaplaric Psalter had an impact on the texts of at least some Gregorian introits. He did not
24 K. Sisam, The Salisbury Psalter (Salisbury Cathedral, MS 150), ed. Celia and Kenneth Sisam ( = Early
English Text Society 242), Oxford 1959, p. 47-51.
25 Such are the conclusions of']. Marbock, Das Eindringen der Versio Gallicana des Psalteriums in die
Psalterien der Benediktinerkloster ( = Disscrtationen der Universitat .,Graz 5), Vienna
1970.
26 My analysis differs in this respect from that of T. Bailey in his edition of the Commcmoratio Brevis
(= Ottawa Mediaeval Texts and Studies 4), Ottawa 1979, p. 21-22.
27 Marbach, Carmina Scripturarum, p. 34*.
28 The "Berno" treatise is edited in M. Gerbert, Scriptorcs Ecclesiastici de Musica Sacra Postissimum I,
St. Blaise 1748, Hildcsheim 1963, p. 91-114. On the question of authenticity sec H. Oesch, Bcrno und
Hermann von Reichcnau als Musiktheoretiker, Bern 1961, p. 54-56.
16
offer a critique of Stablein's conclusions, though they confliqed with his own. He claimed to
have found in the Old Roman graduate of S. Cecilia in Trastevere textual adjustments to the
Gregorian sources. Old Latin Psalters were not collated with the chant texts, but the field of com-
parison included non-psalmic chants. Connolly dn::w up a list of nearly two hundred instances
in which the Old Roman introits depart from the readings of the Gregorian/Frankish sources,
or in which the three Old Roman gmdwdia differ among themselves. His purpose was threefold:
(1) to establish a characteristic Old Roman (and by contrast, Gregorian) text form for the introit
antiphons; (2) to discern possible Gregorian influences on the Old Roman text; and (3) to com-
pare Old Roman texts with manuscripts of the Roman Psalter in order to define. the scope of
"typically Italian psalm readings". In reviewing Connolly's ample documentation I have not
been able to confirm the first two of his theses; my own study of all the Mass chants in the Old
Roman manuscripts and in selected Gregorian witnesses (a portion of which will be presented
below) has led to quite different conclusions: that both Old Roman and Gregorian share a com-
mon textual basis - the Roman Psalter. Each has comparable variants which can be explained
without supposing the influence of one upon the other.
Before proceeding to a demonstration of this fact, 1 would like to suggest a few ways in which
Connolly's data might be otherwise interprete1.V'.
1. He cites 32 cases (footnotes 39-41) in which he sees a distinction between Old Roman and
Gregorian readings. In 8 of these cases the Beneventan (Gregorian) graduate agrees with the
Old Roman sources, and in 2 more (Nos. 54, 62) Northern gradualia (Mt. Blandin and Com-
piegne) also have the Old Roman reading. In a few instances (e. g., nos. 24, 113, 174) the Old
Roman tradition is itself divided, or else an introit is missing in the S. Cecilia graduale. Had
a wider range of Gregorian manuscripts, including the ones edited by Dom Hesbert in the
Antiphonale Missarum Sextuplex, been consulted, a unified Gregorian textual tradition would
have evaporated. In almost half of the introits cited the Sextuplex gradualia are at odds among
themselves.
2. If the existence of a characteristic Old Roman textual preference ist doubtful, the influence
on it of a divided Gregorian tradition is even more dubious: in nearly half of the 13 cases of
supposed Gregorian influence on the Old Roman texts (footnotes 43-44), at least one of the
Sextuplex gradualia agrees with the purportedly isolated Old Roman sources (nos. 12, 52, 88,
138, 10, 182). Connolly wisely dismis5es as "the work of less careful scribes" approximately
SO additional apparent variants (footnote 45) which might give the wrong impression of a uni-
que Old Roman reading.
3. Connolly finds that the Old Roman introits tend to favor readings found in the Italian sources
of the Roman Psalter, while some Gregorian introits display readings derived from the older
(English) manuscripts of this Psalter. This is probably true, though a number of citations
brought forward in proof (footnote 47) must be considered of dubious validity due to the
presence of accidental textual corruption. Lack of agreement among the Sextuplex gradualia
blurs the picture yet further, and reference, to the Old I.atin Psalters would have revealed that
some of the "Roman" readings are also found in them.
4. The contenti<;>n that at times "one or more of the Gregorian sources agrees with the Gallican
Psalter" cannot be sustained.1c. Four of the five "Gallican" references (footnote 48: nos. 60,
24, 90, 157) can be found even in the restricted number of Old Latin Psalters collated by
Weber. The Old Roman and Gregorian introits thus confirm the absence of any influence
from the Hexaplaric Psalter. Despite the fact that some Old Latin Psalters were modernized
29 For the purpose of discussion all of his introit variants will be considered valid and non-psalmic texts
will be accepted. It is presumed that the reader has Connolly's generous documentation at hand.
30 The Graduate, p. 427.
17
in the direction of the Hexaplaric, I believe that the Beuron vetus Latina edition will show
that the chant texts have authentic Old Latin readings whenever they deviate from the main
Roman Psalter tradition.
I apologize for putting the reader's patience to the test by requiring him to relate these comments
to documentation published elsewhere, but 1 believe it is important to lay aside all questions of
the influence of one chant tradition on the other on the basis of textual comparisons. Both have
an identical textual basis, though they occasionally follow separate paths. These deviations are
not remarkable, since they parallel exactly the kinds of variants found in the Psalter manuscripts
themselves.
The sheer number of textual variants found by Connolly in the Mass chants might seem to
undermine this assertion of a common textual basis, but most of the- apparent "variants" do not
survive the application of a few simple principles of textual criticism. We have already noticed
the tolerant attitude toward variant readings in biblical manuscripts which antedates the earliest
gradualia and antiphonalia. Many of the "variants" proposed for the introits are merely the result
of common scribal errors; they cannot be placed on the same plane as possibly more significant
divergencies. Unless the contrary can be demonstrated, it must be assumed that the scribe inten-
ded to produce an exact copy of the text of his model. That he may have failed to do so resulted
from human frailty rather than from conscious design. In liturgical texts the mechanical copy
should be presumed, even if the same presumption does not hold true in the secular literary
tradition
11
A relevant guide to the kinds of scribal errors committed in the Middle Ages is the treatise De
Orthographia, compiled by Cassiodorus at the age of ninety-three for his monks at Vivarium.
It is an elementary guide to proper spelling. That such basic instruction was needed in a
monastery founded by acelebrated rhetorician schooled in the traditions of late antiquity is sug-
gestive of the problems encountered in other scriptoria-'
2
Despite daily exposure to the Psalter
and the chants derived from it, inattentive scribes could still iutroduce unintentional variants
without thereby creating a subsequent stcmma".
The common errors cited by Cassiodorus (incorrect cases, interchange of b and v, o and u, m
and n) are precisely the ones encountered in the texts of the Mass chants. Sometimes the context
itself invites scribal lapses, as when repeated letters are seen as one (haplography): fluctuum/fluc-
tum, cornua arcuum/cornu arcum, ad defensionem/a defensionem. Long or unfamiliar words
were another pitfall, though inadvertent omission of text (parablepsia) is unlikely to go uncor-
rected in a manuscript with musical notation. Errors like these, even if shared coincidentally by
several chant manuscripts, cannot he regarded as authentic variants, the tracing of which would
establish patterns of influence or the conventional stemma based on common errors. They can-
not be used to prove the influence of one chant tradition on another.
31 G. Pasquali argues convincingly for scribal intervention, bur the scribe would have no incentive to
clarify, embellish or adapt well known biblical texts sung annually -or even more frequently - in his
monastery; Storia della tradizione e critica del testo, Florence ~ 1 9 5 2 p. xvii. A non-psalmic intet-ven-
tion, significant for the history of the Old Roman Office is discussed by H. Barre, Corrections dans l'an-
tiphonaire de Saint-Pierre, iri: Revue Benedictine 76 (1966), p. 343-351.
32 De Orthqgraphia merely assembles relevant passages from earlier grammarians; it is reprinted in H. Keil,
Grammatici Latini 7/1, Leipzig 1878, p. '143-210. For a sensitive evaluation see James ]. O'Donnell,
Cassiodorus, Berkeley (California) 1979, p. 229-37.
33 For a sampling of variants in biblical manuscripts consult the listing "Orthographia tot ius Octateuchi"
in: Biblia Sacra luxta Latinam Vulgatam Versionem 4, p. 393-490. Pasquali's judgement is straightfor-
ward: "che peculiarita ortografiche non provano nulla, e risaputo ... ; Storia della tradizione, p. 17. Cf.
B. Stablein, Kann der gregorianische Choral in Frankenreich entstanden sein? in: Archiv fiir
Musikwissenschaft 24 (1967), p. 157.
18
Questions of Psalter versions and variants would hardly arise outside the context of cycles of
variable chants covering most or part of the liturgical year. Not all of the Mass chants were in-
troduced at the same time of course, but it can be regarded as certain that all the relevant Psalter
translations existed before the Mass cycles were created. Only the gradual may antedate the activi-
ty of Jerome, yet its earliest textual tradition will remain uncertain. The concept of a variable
cycle was surely established for th<.! introit and communion by the sixth century; perhaps the
other Mass chants had reached a similar stage of evolution at the same time'". The first men-
tion of a cantus anna/is, occurs more than a century later in an appendix to Ordo Romanus XIX.
The compiler of the Ordo credited several popes and Roman abbots with important contribu-
tions to sacred music. Pope Leo I (440-461) is the earliest figure associated with a yearly cycle of
chants, but the Ordo attributes similar achievements to a number of Leo's successors''. It seems
unlikely, however, that the compiler of the Ordo had much specific information about the
liturgical or musical activities of these popes.
Several considerations make the cycles of variable chants for the Lenten Masses a suitable founda-
tion for the present study. They form a coherent group of pieces within a liturgical season whose
main lines of evolution have been satisfactorily clarified. Furthermore, growth of the chant reper-
toire can be correlated at least in part with the various stages of development through which the
Lenten observance passed.
Although a fast of forty days preceding Easter existed at Rome in the late fourth century, not
every day was marked liturgically
3
'. Sundays had been observed with a Eucharistic liturgy since
apostolic times, but weekdays were not automatically provided with comparable services. By the
pontificate of Leo I (440-461), Wednesdays, Fridays and perhaps Mondays of Lent had been added
to the kalendar. All of the remaining Lenten feriae entered the liturgical kalendar between the
late fifth and early eighth centuries, though most of them were probably in place by the time
of Pope Hilary (461-468}'
7
There is some disagreement about the individual stages of this pro-
cess, but the late addition of the Lenten Thursdays under Gregory II (715-731) virtually closed
the liturgical development of the season. Since by that time compositional activity had ceased,
chants for these Masses had to be borrowed from other days.
The extent of the period defined as "Lent" also had an impact on liturgical organization. Sun-
days were never considered days of fast; hence, when it became imperative to have exactly forty
days of fasting before Easter, an additional..four days preceding the first Sunday of Lent were in-
corporated in the calculation. For the same reason Good Friday and Holy Saturday, heretofor
considered apart from Lent, were included. Thus came about the ambivalence between a forty-
day fast and the earlier custom of observing Quadragesima as six weeks of preparation for
34 This much can be assumed from the Libcr Pontificalis, which reported (ca. 530) that Pope Celestine I
(422-432) had introduced the practice of singing "psalmi David CL ante sacrificium" (1, 230); ed. L.
Duchesnt; in: Bibliothcquc des Ecolcs d'Athcncs ct de Rome, 2nd series, vol. 3. Sec also the
discussion of Chavasse's views on the Lenten communions below.
35 Ordo XIX, p. 36; M. Andrieu (Ed.), Les Ordincs du haut moyen age 3 ( = Spicilegium Sacrum
Lovaniense 24 ), Lou vain 1951, p. 223-24. Concerning the Frankish and monastic origin of the Ordo see
Andrieu 3, p. 9-15.
36 C. Callewaert, La duree et le caractere du Careme ancien dans l'Eglise latine, in: Sacris Erudiri, Steen-
brugge 1940, p. 449-506; A. Chavasse, Le Careme romain et les scrutins prcbaptismaux avant le IX" sic-
de, in: Recherches de Science Religieuse 35 ( 1948), p. 325-81. This seminal article and several of
Chavasse's later publications are the basis for: La structure du Careme et les lectures des messes
quadragesimales dans Ia liturgie romaine, in: Maison-Dieu 31 (1952), p. 76-119. A letter of Jerome to the
Roman virgin Marcella remarks on an strict fast in quadragesima (Ep. 24, 4; Corpus Scriptorum Ecclesiae
Latinorum 54, p. 216).
37 La duree, p. 500; Chavasse (La structure, p. 85) would place this development in the early
SIXth century.
19
pascha. Confusion about the real beginning of Lent remains embedded in the oldest layers of
the Gelasian sacramentary'". The Secreta prayer of Ash Wednesday contains the words "qui bus
ipsius venerabilis sacramentum venturum celebremus exordium: yet the collects of the day refer
to the (1) "inquoata ieiunia" and (2) "causas ... inquoatas". On the following Sunday, however,
a rubric refers to the day as "dominica in Quadragesima incoantis inicium"w.
A cycle of 26 psalmic communion chants was created sometime after this Wednesday began to
be regarded as the caput Quadragesimae in competition with the traditional commencement of
the season on Sunday.'
0
This cycle could hardly be the result of haphazard evolution. Derived
from psalms 1-26 in ascending numerical oder, the series covers most of the weekdays between
Ash Wednesday and Friday of the fifth week of Lent. At the time of its introduction the
Eucharist was celebrated on most of the Lenten Jeriae except the still a liturgical Thursdays. A
distinction existed, moreover, between the period of penance and the week commemorating
Christ's
25:6
mozmcdHex
LXX: lxuxAwa01
multiplicatac sunt
GR. Tribulationes,
Fer. iv, Hcb. I)
et siwt owli ancille
ecce oculi ancille
(TR. Ad te levavi, Dom. III)
adversum me
.tt!V(TSUS me
(OF. vs. Domine fac mecum,
Fer. iv, Hcb. Ill
ad ipsum
ab ipso
(OF. vs. Renedieite gentes
Fer. iv, Hcb. IV)
in testamento tuo
in testamentum tuum
(GR. Respicc, dominc,
Fer. v, Hcb. IV)
et circuibo a/tare
et eircumdabo altare
(CO. Lavabo, Fer. iv, Heb.V)
B. Relatively even division among the gradualia surveyed
33:8
mozrnedHex
KTD*V
LXX:
IXntAoc; xup(ou
25:12
t moz
immittet a11gelum dominus
immittet angelus domini
immittet angelus dominus
(OF., Fer. v, Heb. I)
bencdicam dominum
benedicam domino
(INT. Redime me,
foer. ii, ric b. II)
28
MonComCorSglaAbEcTrlrPr
Bl
MonCom CorSgMcEcT r Lr Pr
Mn
BILaSgEclr
Alifr
Com LaSgA h F.cTr
Lr
BllrPr
MonComCorSillaSgAbMcEcTr
CorLaSgAbEcTrl.rPr
RhComMc
BIComLaEcTrlr
CorSilSgMc(immitit)Pr
Ab
ComC:::orlaAbEcTrlrPr
McMnSg
15:8 ne commovear
nee commovear CorlaMcAbEclrPr
(OF. Benedicam dominum,
fer. ii, Heb. II)
18:8 sapientiam CorlaSgEcTrPr
a mozU sapicntia ComAbMcLr
(INT. Lex domini, Sabb., Heb. II)
53:3 Iibera me AbMnTrlrPr
IXOE K Hex iudica me ComCorMcLaSgEc
LXX: xp1'116v f.I.E (INT. Deus in nomine
Fer. ii, Heb. IV)
C. Influence of the Italian branch (PQRUVX) of the Roman Psalter
94:4 yo mozmedHex sub pinnis eius Bl
N!MKTBCDPQRUVX sub pennis eius RhComCorAbMcMnSgLaEcTrlrPr
69:6
OE CPQR
2
UV]X
39:15
mozPQ
2
RUVX
85:17
IX medH
2
PQRUVX
ot
2
M BCD
PUVX
34:26
N*T
2
PQRUV
medHcx
142:2
IX mozT
2
PQRUVX
(OE Scapulis suis, Dom. I)
es tu
esto
(GR. Adiutor meus,
Fer. ii, Heb. II)
re-vereantu1 simul qui
revereantur lnlllliCI mei qui
revereantur qui
(OF. Domine in auxilium,
Fer. vi, Heb. II)
in bono
in bonum
(INT. Fac mecum,
Fer. vi, Hcb. III)
qui magna !oquuntur
qui maligna loquuntur
(CO. Erubescant,
Fer. ii, Heb. VI).
adversum me
adversus me
super me (CO. Erubescant)
non mtres in iudicio
ne intres in iudicio
(OF. vs. Eripe ... Domine.
Fer. ii, Heb. VI)
Mon
BIComCorAbMcMnLaSgEcTrLrPr
TrlrPr
ComCorAbMcMni.aSgEc
ComCorSi!LaSgEc
RhBlAbMcMnTrLrPr
BlComCorAbMcLaSgEcTrl.rPr
CorAbMcLaSgEc
TrlrPr
BlCom
ComAbLaSg(in iudicium)EcTrlr
D. Influence of the Old Latin Psalter independently of the Roman Psalter
118:47 quae di!exi nimis Mn
IX'"( quae dilexi valdc ComCorAbMcla(dilexit)SgEcTrlrPr
118:57
IXO
118:58
oc:precatus
oty med: vultum tuum
43:26
(OF. Meditabor, Fer. iv, Heb. I) Meditabar)
portio mea
pars mea (OF. vs. Meditabor)
deprecatus sum faciem tuam
precatus sum vultum tuum
(OF. vs. Meditabor)
adiuva nos
29
AbLaSgEcTrLr
ComlaSgEc
depr. sum v. tuum: AbTrlr
o
2
:fers o. nobis
ro*:fer
fer opem
(GR. Exsurge domine,
Fer. iii, Heb. IV)
MonBIComCorSiiAbMcLaSgEcTrLrPr
E. Chant text unattested/rare in the Psalters collated (excluding ccntonization)
30:17
118:18
LXX:
118:77
118:15
B mozHexHMN
2
ST
102:3
medHexBC
37:22
moz
33:6
Hex
73:22
Mo7.: obprobriis
servo rum
118:121
Hex
in tua misericordia
propter misericordiam tuam
(OE In te spcravi,
Fer. iii, Heb. I)
revela omlos
levabo oculos
(OF., Fer. ii, Hcb. I)
mihi
super me
(OF. vs. Levabo)
in mandatis tuis me exercebo
in prcceptis tuis me exercebor
et a preeeptis . 0 0
(OF. vso Levabo)
qui propitius fit
qui propitiatur
(OF. vs. Benedic anima,
Fer. v, Heb. I)
0
ne discesseris
0
ne discedas
(INT. Ne dcrelinquas,
Fer. iv, Hcbo II)
vultus vestri mm erubescent
facies vcstrae non confundcntur
(GR. Venitc filii,
Fer. iv, Hebo lV)
inproperiorum tuorum
obprobrii servorum turorum
obprobrium servorum tuorum
(GR. Respice domine,
Fer. v, Hebo IV)
persequentibus me
{;alum(p)niantibus me
(OF. Benedictus ... nc tradas,
Fer. vi, Heb. V)
30
ComAbLaSgEcTr Lr
McSg
RhBIComCorSi!AbMnLaEcTrLrPr
BILaTrPr
ComSilAbSgEc
Tri.r
ComAbLaSg
ComSi!Ab LaSgEcT rlr
RhComCorAbMcMnLaSgEcTrLrPr
BIComCorLaSgEc(AbMC: f. vcstras)
(This verse, absenr from the OR
chant msso, is Frankish.)
ComCorAbMcLaSgEclorPr
Lr
ComCorAbMcLaSgEcTrLr