Jump to ContentJump to Main Navigation
Inside the OffertoryAspects of Chronology and Transmission$

Rebecca Maloy

Print publication date: 2010

Print ISBN-13: 9780195315172

Published to Oxford Scholarship Online: May 2010

DOI: 10.1093/acprof:oso/9780195315172.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 21 February 2017

(p.248) Appendix 3 Commentary on Edition

(p.248) Appendix 3 Commentary on Edition

Source:
Inside the Offertory
Publisher:
Oxford University Press

(p.248) Appendix 3

Commentary on Edition

Appendix 3 Commentary on Edition

Ad te domine levavi (1)

The pitch‐level variants and probable nondiatonic pitches in this offertory have been demonstrated by Sidler, Pitman, Frasch, and Hankeln.1 In many sources, the respond and first verse are written at the affinal position, ending on a. Mo 159, for example, adopts a final of a and employs several pitches that cannot be written at the lower level in its notational system. As Sidler suggested, Mo's pitches D, F, and b‐flat would be Γ, low‐B‐flat, and E‐flat at the position of the regular final, D. Mo 159 employs a b‐flat at “erubescam” (phrase 2), the source of the parenthetical E‐flat in the base transcription of Ben 34. Mo 159 also implies an E‐flat in v. 2 at “custodi animam” (phrase 9).

These two cases of nondiatonic practice are variously treated in later MSS. The first, “erubescam,” is not reflected in many later MSS. Most peripheral sources notate the respond and first verse at the normal position, with a final of D and no nondiatonic indications. The nondiatonic practice, however, is preserved in Be 40078; the respond and first verse are written at the affinal position and employ the b‐flat at “erubescam.”

The opening of v. 2 is a point of variance in pitch level, probably reflecting the nondiatonic practice implied in Mo 159. The base version of Ben 34 presents a plagal reading of the verse, opening with the pitches A‐D, a version matched in Pa 780 and Lo 4951. In other core MSS, including Pa 776 and Pa 1121, the verse opening is written a fifth higher; the opening pitches, D‐a, are characteristic of mode 1. In Mo 159, the verse is designated as authentic by the tonary letters. The two versions join at “meam” (phrase 9), immediately after the problem spot at “custodi animam.”

Although both versions of the second verse are found across a broad spectrum of MSS, I consider the plagal reading of the verse the more probable unemended version, consistent with Sidler's hypothesis. The status of the authentic reading as an emendation is supported by the nondiatonic pitches that would result at the lower level of transposition and the relationship to the Roman version, which remains in the plagal range. Among later MSS, the nondiatonic practice suggested by Mo 159 is evident in Be 40078 and Gr 807, which are notated a fifth above Ben 34 and have corresponding b‐flats in v. 2 at “custodi animam.”2 The practice is further implied by a unique reading found in Pad 47, described below. Some early adiastematic MSS also suggest a performance at the plagal level. Ei 121, for example, has the letters “im” at the opening of v. 2, indicating that the verse begins below the final.

Pitch level of respond and v. 1 (in MSS that unequivocally indicate pitch):

  • = Ben 34: Pa 1235, Gr 807, Pad 47, Pst 120, RoV 52

  • (p.249)
  • 5↑ Ben 34 (= Mo 159): Be 40078

v. 2 “Respice…meam”:

  • = Ben 34: Pa 780, Lo 4951, Pa 1235, To 18

  • 5↑ Ben 34 (= Mo 159): Pa 776, Pa 1121, Pa 1132, Pa 1134–37, Be 40078, Cai 61, Pst 120, Gr 807

A unique reading in v. 2 in Pad 47 supports the hypothesis that an E‐flat was sung at “custodi.” The verse begins with the incipit D‐G, remaining a fourth above Ben 34 until “custodi.” On “custodi animam” the melodic segments where the hypothetical E‐flat occurs (DEF in Ben 34) are written a tone higher, EFG, suggesting that interval between the first two notes is a semitone.

The MSS otherwise show very few melodic variants.

Pad 47: v. 2 “[invocavi] TE”: the end the melisma differs from that of the majority version CF GF CD CF GFC FECD FFD

Roman sources: Vat 5319 is very similar in detail to Bod 74. Aside from differences in passing tones and the oriscus, it has the following small difference: v. 1 “et”: melisma starts CE.

  • RoS 22 (slashes indicate syllable breaks):

  • “Animam”: DC

  • “eruBESCAM”: FGaGaGFE/FGaGFE

  • “IRRIDEANT”: ED/FE/ED/DEFEDEDEDC

  • “NON”: DFEDC

  • “CON/FUNdentur”: DFEDC/EFGFGFEFD

Deus tu convertens (2)

This offertory presents few major variants among the core group of sources. The small variants within the core group involve the number of pitch repetitions or the pitches b and c, with a greater focus on b in the Beneventan sources. The b's in the opening, on “deus” and “vivificabis,” for example, are c's in Mo 159 and the Aquitanian sources.

More significant variants include the following: v.1 plebi/plebis tuae (phrase 7).

The reading of Ben 34 (the only Beneventan version of the verse) is anomalous here, closing with a deuterus cadence that imitates the verbally identical phrase of Benedixisti (3). Mo 159 and most Aquitanian sources have a shorter melisma and a verse ending on D, distributing the four syllables of “plebi tuae” as in Mo 159: aGaGFGa G FF FFFDED.

Most Aquitanian and German sources present a reading identical or very similar to that of Mo 159, as do all early adiastematic MSS in the sampling. The Italian sources, however, match Ben 34 in the cadential pattern on “tuae.” The melisma on “plebi/plebis” is a point of variance in some MSS. Although most Aquitanian sources match Mo 159, Pa 780 has a longer form of the melisma, aGaGFGaGaGF.

Among some peripheral sources, there are other points of variance in melismas, as shown in the sample readings below. Spaces between note groups indicate neume groupings, and boldface indicates specific variant points.

v. 1 “terram”: Cai 61 has bcd cdcb acedc acc ca G acba cb bc ded ec cb (Pa 1235 is identical in the bolded segment)

To 18 has a longer melisma, differing at the opening: b cd ede cb ac cc edc ac cc caG abcb abc cb bcd edec

(p.250) v. 2 “et”: Cai 61 acccaF acba cb bc ded ec cb

To 18: The last part of the melisma is repeated (perhaps an error): abcb bcd edec (repeat).

v. 2 “Caelo”: To 18 has a different ending: ac ca cccaG aGaG GGe Ga

Be 40078: G c ca bcaG c Ga c ca bcaG bbab db caaGa cc ca cc caaG aFG bcaG aca (similar in other German sources)

Many MSS, including Ben 34, have a prosula at the end of the second verse. In Ben 34, the repetendum cue follows the prosula.

Roman sources: This offertory is lacking in Bod 74. There are very few variants between RoS 22 and Vat 5319, mostly involving the filling in of intervals with passing tones.

Benedixisti domine (3)

There are few major variants within the core group of sources. Most involve passing tones and variants at the points E‐F and b‐c. “Peccata” in v. 1 is a good example of the latter kind of variant. Ben 34 has Gba/bGcbc. The same version is found in Pa 776, and a similar one occurs in Lo 4951 (Gba/bacbc). Pa 1121, however, has aca/cacbc.

The Beneventan MSS exhibit several small differences from the others:

Respond “[terram] tuAM”: all sources in the core group besides Ben 34 have DGFE

v. 2 “domiNE”: Ben 34 presents a slightly shortened melisma. Pa 776 is typical of the Aquitanian sources: cccbaccdccdaaG. A similar reading is found in most peripheral sources.

v. 2 “saluTARE TUUM”: The Beneventan MSS transmit an independent reading of the text underlay. Most MSS have a reading closer to that of Pa 776: aG/aGF/FGacacGE/FGFGFF ori. E

Individual MSS: v. 1 “operuisti”: Mo 159 adopts G rather than F as a reciting pitch at “operuisti” and notates the first three syllables a tone above all other sources examined.

Pst 120: “[salutare] tuUM”: aGaGFE

Pad 47:

  • v. 2 “ostende nobis do [mine]”: This verse opening is written at various pitch levels in relation to the majority version (slashes indicate syllable divisions): E/aa/GF/aFE/FGa/a/a.

  • v. 2 “domiNE”: cc cb a cc db cdaG

To 18:

  • respond, “benedixISti”: FGE GaG FGFE (shorter melisma)

  • v. 1 “Omnia”: bd edc dcb cc ca cdc (shorter melisma)

  • v. 1 “PECcata”: FaG

  • v. 1 “Iram”: aGaGE GFED Ga(different melisma)

  • v. 2 “NO/BIS”: cb aG/abc

  • v. 2 “domiNE”: cb abc dcb

  • v. “noBIS”: cc cG acb cdca acb cdc bca ccaG acaFGF GcccGEGG acaG

Pa 1235: v. 1 “domiNE”: cb ac c dc cda aG

The German MSS give a reading very close to that of the core group, with the expected raising of b to c.

Roman sources: There are no significant variants. RoS 22 has the following variant: “BenedixISti”: Fa abGffedFGFEFE.

(p.251) Confortamini (4)

The sources exhibit small variants in recitational style, mainly involving the pitches E and F or b and c.

Pitch‐level variants: Most sources are notated at the same pitch level as Ben 34 throughout. There are three exceptions, however, among the MSS in the sampling. Mo 159 notates the whole offertory at the affinal position, closing on b, probably because of the need for low B‐flat and a possible E‐flat. Two sources, Gr 807 and Cai 61, notate the final verse a fifth above the majority reading. The upward transposition of the final verse in Cai 61 may be attributable to the introduction of low B‐flat in the final verse. Gr 807, however, employs a notational sign for B‐flat in other places, so its move into the authentic range for the final verse may have been made simply for aesthetic reasons.

  1. 1. Respond, opening passage: At the affinal position, Mo 159 employs a b‐flat in the opening of the respond. In a contour that is duplicated in most Aquitanian sources, it opens with the pitches G a b‐flat c. As Sidler proposed, the b‐flats in the reading of Mo 159, written at the affinal position, suggest the use of the nondiatonic pitch E‐flat at the lower level.1 While this reading is implied in some Aquitanian sources, such as Pa 776 and Pa 780, it is not reflected in MSS from other regions. The Beneventan sources write a normal E. Most other MSS avoid E altogether. Because of the uncertainties surrounding this pitch, I have not included it in the transcription of Ben 34.

  2. 2. v. 2 “Praestat/praestatis certamen” (phrase 10): At the affinal position, Mo 159 employs an F here, implying the presence of low B‐flat in the normal range. The internal cadence on “certamen” is treated in various ways in the sources:

Mo 159:

a

GacaFGF

F F ori G

Pa 780:

D

CDFDCDB

B

Pa 1121:

D

CDFDCDB

B B (ori.) C

Pa 776:

DC (liq.)

CDFDCD

A A (ori.) B

Ben 34:

C (liq.)

CDFDCDB

AB

On the final syllable, Pa 1121 and Pa 780 present a version similar to that of Mo 159, in which the B's are presumably B‐flats. The other sources write the end of the cadential figure a tone lower.

Roman sources: Vat 5319 is the only source to transmit the verses of this offertory. Bod 74 and RoS 22 notate the offertory in the untransposed range, closing on E.

Respond, “tiMERE”:

  • RoS 22: FGaGabaGFEF/FGFE.

Respond, “NOS”:

  • Bod 74: GFGabaGFG

  • RoS 22: GFGabaGFGF

(p.252) Exsulta satis (5)

In most sources, this offertory closes on E, but in a few, such as Lo 4951, the respond ends on G.1

The second verse is a problem spot, with an extreme diversity among MSS. The modal ambiguity and pitch‐level variants of the verse are addressed in studies by Sidler, Frasch, and Hankeln.2 The peripheral sources I have incorporated are consistent with the diversity of the Aquitanian sources, demonstrated by Frasch and Hankeln, and unfortunately shed no new light on the reasons for the variants. Hankeln views Mo 159, chosen as a base version here, as representing the Aufführungslage and plausibly attributes the variants to a lack of modal clarity in the verse.3 Because indications of b‐flat are lacking in most Aquitanian MSS and in Beneventan MSS, it is not always possible to determine whether the various versions differ in intervallic structure.

The single Beneventan version, Ben 34, presents a reading not matched in any other source, with tritus characteristics (the other Beneventan MSS are incomplete, lacking this offertory). This version is probably best viewed simply as a different interpretation of the verse's modal properties.

v 2. “Quia ecce venio…omnipotens”: Sidler considers Mo 159 the lectio difficilior in this passage and thus the more probable reading.4 The interval of transposition between MSS changes several times, with the sources finally joining (temporarily) at “omnipotens.”5 The following summary disregards differences of less than three or four notes. With these small melodic variants in mind, a reading essentially matching that of Mo 159 is found in the geographically diverse sources Pa 776, Pa 1121, To 18, and Be 40078. Other sources may be summarized as follows.

a. “Quia ecce…medio tui”: The following sources begin 2↑ Mo 159 and join at “venio”: Pa 903, Pst 120, Pad 47 (a variant reading of “ecce”: d cdedc cb ab/ba); Gr 807 (melodically variant at “habitabo”). There are no intervallic differences that would have prompted the transposition.

Pa 780 is 2↑ Mo 159 for most of the passage. Pa 1132 begins as Mo 159 but is 2↑ Mo 159 beginning at “venio.” Lo 4951 has a variant opening and is alternately 2↑ or 3↑ beginning on the second syllable of “venio.” Ben 34 begins 2↓ Mo 159 and joins at “venio…. in medio” (with a variant reading of “venio”). Pa 1235 begins 2↑ Mo 159 and is alternately 2↑ or 3↑ beginning at “venio.”

b. “dicit dominus” (“dicit” is a point of melodic variance):

  • = Mo 159: Pa 1132, Pa 776, Be 40078, Pa 1235, To 18

  • 2↓Mo 159: Gr 807

  • 2↑Mo 159: Ben 34, Pa 1132 (beginning at “dominus.”); Pst 120 (at “dominus”)

c. “omnipotens”: all sources briefly join Mo 159.

2. “et confugient”:

  • = Mo 159 (with melodic variants): Pa 1235, Be 40078, Pad 47, Pst 120

  • 2↓ Mo 159: (with variants) Ben 34, Lo 4951

(p.253) 3. “ad te in illa die”:

  • = Mo 159: Pa 776, 780, 903, Lo 4951, Pad 47, Pst 120

  • 2↓ Mo 159: Gr 807, Be 40078

  • 2↑ Mo 159: Ben 34

  • 3↓ Mo 159 at “ad te in illa”: Pa 1121, 1132. The passage articulates the major third F‐a rather than the minor third a‐c.

  • Pa 1235: starts at Mo 159 level, 2↓ at “illa.”

  • To 18: starts at Mo 159 level, variant reading at “die”: aGFE FaG/G.

6. “omnes gentes”:

  • = Mo 159: all except Ben 34, which is 2↓.

  • To 18 lacks the melisma on “plebe.”

  • Roman MSS: RoS 22 has a slightly varied melisma on “syon”: GaG cdcdc bc abaGF GabaGaG.

Tollite portas (6)

In terms of pitch level, Tollite portas has three variant points: the respond, specifically its ending, the beginning of the second verse, and the final melisma of the second verse. The Beneventan version provided in the transcriptions is an unproblematic second‐mode melody with a final verse in the authentic range. A second version of Tollite is written at the affinal position, remaining a fifth above the Beneventan version until the end of the closing melisma on “gloriae,” which suddenly shifts downward by a tone to articulate the third G‐b‐flat. This G‐protus version is found in several key core sources, including Mo 159, Pa 776, and Pa 1121, as well as Be 40078 and Cai 61. The two different endings are shown in example 6.1.

Although this version is not reproduced in its entirety here, it may be studied in Hansen's transcription of Mo 159, as well as in Hankeln's edition (in the version of Pa 776).1

The D‐protus version of Ben 34 has been chosen for the base transcription because it is the more common reading among the sources examined and thus representative of a wider general practice. The version closing on G, however, clearly represents an authentic early tradition. Its authority is attested by its presence in early sources and later appearances in Cai 61 and Be 40078. Bomm viewed the G‐protus version as the “Originalfassung” and the D‐protus version as an emendation.2 Concurring with Bomm, Sidler noted that the b‐flats at the close of the respond would be E‐flats if the piece were written a fifth lower.3 The sudden change in modal character at the close of the G‐protus version is unusual, to say the least, and establishes the G‐version as the lectio difficilior. With two exceptions (Pa 1135 and Pa 1235, discussed in the notes below), the D‐protus versions present a nearly identical reading of the final melisma, which, coupled with its status as the majority reading, suggests that this version is an authentic alternative reading. The Roman version closes on D and has a (p.254)

Appendix 3 Commentary on Edition

Example AP 3 6.1

Appendix 3 Commentary on Edition

Example AP 3 6.2

pitch‐level profile similar to that of Ben 34. While the scribe of the tonary of Frutolf of Michelsberg opts for the D ending, his notation of the melisma indicates his awareness of it as a problem spot.

The differences at the close of the respond have implications for the pitch level of both verses. The versions that close on D are consistent in pitch level: the respond and first verse have typical second‐mode characteristics, with internal cadences on the final and D and F as the most prominent pitches. The final verse is in the authentic range, focused on a. All sources with a G final notate the respond and the first verse a fifth above the level of the Beneventan version. In the majority of these sources, the second verse begins a fourth above the level of the Beneventan version, as shown in example 6.2. Here the verse is a tetrardus authentic melody focused on D, consistent with the G ending of the respond. Cai 61 presents a unique pitch‐level profile. The respond, notated at the level of Mo 159, closes on G, whereas the two verses are in the same range as Ben 34. The repetendum is notated a fifth too low, implying a pitch‐level profile identical to that of Ben 34, with the low C‐protus ending.

Further variants in pitch level occur at the close of the second verse, as shown in example 6.3.

In most sources that notate the verse in the tetrardus authentic range, such as Mo 159, the interval of transposition relative to Ben 34 changes to a fifth at the end of the melisma. Sidler proposed that this apparent whole‐tone transposition indicates a use of nondiatonic pitches and thus argued for a “retransposition” of this segment (cdedeede, etc.) down a whole tone to derive a b‐natural, which would be F‐sharp at the level of Ben 34.4 Hankeln (p.255)

Appendix 3 Commentary on Edition

Example AP 3 6.3

has more plausibly attributed this transposition to the modal ambiguity of the melody as whole, and posited the written level of 776 and Mo 159 as the Aufführungslage. 5 The downward transposition of this reading facilitates the transition to the repetendum, with the protus characteristics of its internal phrases. Pa 1121, line 3 in example 6.3, remains a fifth above the reading of Ben 34, consistent with its notation of the whole verse at this level.

A summary follows of pitch‐level profile of the MSS relative to Ben 34. For the Aquitanian sources, I have read the base pitch level according to the “gloriae” melisma of the respond: those with the downward shift at the end of the melisma, matching the version of Mo 159, are interpreted as ending on G (rather than C, which is not a legitimate final), whereas those with a D‐ending are interpreted as being in the normal (rather than transposed) plagal range, like Ben 34. In summarizing the pitch‐level (p.256) profile as a whole for the Aquitanian sources, I have followed Hankeln's methodology, using the repetendum as a guide to reading the relative pitch level between the sections of an offertory. In most Aquitanian sources, the placement of the repetendum suggests that the intended pitch‐level profile of the respond and two verses is that of Ben 34, since the repetendum appears at the correct pitch level only if the pitch‐level profile of Ben 34 is assumed.6

  1. 1. Respond to “gloriae” melisma, and verse 1:

    • = Ben 34: [Lo 4951 Pa 780 Pa 1132] Gr 807, Pst 120, Pad 47, RoV 52, Mü 10086

    • 5↑ Ben 34: Mo 159, Pa 776, Pa 1121, Cai 61, Be 40078, Pia 65, To 18

  2. 2. End of “gloriae” melisma:

    • = Ben 34, ending on D: 4951, Pa 780, Pa 1132, Gr 807, Pst 120, Pad 47, RoV 52, Mad 20–4, Be 10086

    • 4↑ Ben 34, ending on G: Mo 159, Pa 776, Pa 1121, Cai 61, Be 40078, Pia 65

    • 5↑ Ben 34, ending on a: To 18

  3. 3. Verse 2 to the end of “eum” melisma:

    • = Ben 34: 4951, Pa 780, Pa 1132, Gr 807, Cai 61, Pst 120, Pad 47, RoV 52, Mad 20–4, Mü 10086, To 18.

    • 4↑ Ben 34: Mo 159, Pa 776, Be 40078, Pia 65

    • 5↑ Ben 34: Pa 1121

  4. 4. End of “eum” melisma:

    • = Ben 34: Pa 780, Pa 903, Lo 4951, RoV 52, Pad 47, Mü 10086, To 18

    • 5↑ Ben 34: Mo 159, Pa 776, Pa 1135, Be 40078, Pia 65

    • 2↑ Ben 34: Pa 1121, 1132,7 Pst 120, Cai 61

Beneventan distinctiveness in v. 2 at the “eum” melisma: Ben 34, the only Beneventan version of this offertory with the verse, has a form of the melisma that is unique among the sources examined, with an extra repetition of the opening material.

Individual MSS: Paris 1235 presents an unusual reading of the problematic end of the melisma, where the change in transpositional level occurs. Where Ben 34 begins with FGaGa, Pa 1235 has aFGaGaaGaGaGa.

The “gloriae” melisma” of respond: Pa 1135 presents a unique reading of the final melisma, FFFDFGFFFDGFDDCFEECD. Although it appears to end on D, it shares the change of reciting pitch with the G-protus sources. Pa 1235 has a shortened form of the melisma, in which the problematic ending is simply eliminated: FFFDFGFFFFDGFFD.

Roman MSS: The verses of this offertory are transmitted only Vat 5319. In the respond, the three sources exhibit no significant variants.

(p.257) Laetentur caeli (7)

Aside from variants around the pitches E and F, this piece shows very few differences between MSS on the Gregorian side. Small differences are found in the following MSS from the peripheral group.

Be 40078: respond, “terRA”: DGFE FGEFD (same in verse endings, terRA in v. 1, eIUS in v. 2).

Pad 47:

  • respond, “DO/MI/NI”: FGaG/GF/FGaG

  • v. 1 “cantiCUM”: GEG

  • v. 2 “[nomen] Eius”: GaGFG

  • v. 2 “benenunCIate”: EGaG

RoV 52:

  • respond “TERra”: EaGaF GaG (ori.) (same in v. 1, TERra and v. 2 Eius)

  • respond, “quoniam”: bottom note is E rather than D

  • v. 1 “benenunciA/TE”: acG.acG

Roman MSS: Vat 5319 and Bodmer 74 show more variants than usual.

Vat 5319:

  • respond “VEnit”: FaGaGF aG GF Ga(ori.)GFGF

  • v. 1 “domiNO” (first time): EFGaFE aGabaG

  • v. 1 “noVUM”: EG

  • v. 1 “CAN/TA/TE” (second time): FEF/EDFED/EFD

  • v. 2 “domiNO”: EFGaFE aGabaG (as in v. 1)

  • v. 2 “ET/BE/BE/DI/CI/TE”: Gc dc cb/a/G/F/F/F F(ori.)EDC FGaGabaG

RoS 22:

  • “exULtet”: EFG

  • “veNIT”: EFGF aGFEFE

Deus enim firmavit (8)

The MSS are inconsistent in their use of b‐flat. In the respond, Mo 159 indicates flats at several points where they are not indicated in many other MSS: “firMAvit,” “DEus” (phrase 2), “tunc,” and in the final cadence. Only the b‐flat on “firmavit” is witnessed in another MS, Pa 903 (through the use of the quilisma). The reading of Mo 159 thus has a G‐protus quality throughout the respond. In the base transcription of Ben 34, I have given parenthetical b‐flats only where they are found more widely.

There are a few significant variants between MSS of the core group. The opening passage, for example, is a point of variance among core MSS:

De‐us

Ben 34:

G FGF

Mo:

a aG

Pa 776:

G GF

(p.258) Pa 1121 and Lo 4951 match Pa 776, whereas Pa 780 matches Mo 159. Ben 34's reading is a common one in the peripheral group, found in MSS such as RoV 52 and Be 40078.

v. 1 “DOminus”(verse opening): Lo 4951 has a shorted melisma: DFFG GF GF F(ori.)D GccaFG Gcca cd cd ccaFGa c c(ori.).

v. 1 “decorEM [induit]”: Pa 1121 has a shorter version of the melisma, but gives more notes to the first syllable of “induit.”

de‐co‐

rem

in

[duit]

Pa 1121:

acb caG ba bGF

acca cdcG cdc c(ori.)a

v. 2 “domui tuae”: Most Gregorian MSS (with the exception of Pa 776) have “domum tuam.” Mo 159 and Pa 1121 have a different melodic reading on “doMUM,” as in Mo 159: cdcc GaG. The version of Ben 34, however, is matched in the early adiastematic MSS Cha 47, La 239, and Ei 121, as well as in most Aquitanian sources. Outside the core group, he version of Mo 159 is found only in Pa 1235 and Cai 61.

v. 2 “diErum”: The sources differ in the ways they close this melisma. Ben 34 (the only Beneventan MS to transmit this verse) gives a slightly shortened version of the melisma, with one fewer repetition of the figuration between D, G, a, and c. The same shortening is found in Lo 4951. Pa 776 gives a more typical reading (the boldface segment is missing in Ben 34): GE GabcccabaGb cccabaG aGFaaG ccG baGGEGaG FGaccaG G(ori.)DEFFFD GaccaG DGaccaFG DGaccaccaGb DGaccaccaGb ab b(ori.)ab.

The longer version of the melisma is found in most core MSS, including Mo, Pa 1121, and Pa 780, and in most sources in the peripheral group.

The sources also differ in their adoption of E, D, or F as the bottom note of the melisma. In contrast to Ben 34, Mo 159 adopts D rather than E as the bottom note in the first part of the melisma, whereas Pa 1121 adopts F as the bottom note in the second half of the melisma.

Peripheral MSS:

Pst 120:

opening, “DE/US E/NIM”: a/GaG/F/G

v. 2 “diErum” has a longer version of the melisma:

GD Gac cc ca baGccccabaG aGF aa ac ca baG GDG a aG DGaccaG GDF F FD GaccaG DGac ca ccaFG DGac ca ccaGa EGac ca ccaG cac c cac

RoV 52: v. 2 “DOmine” (phrase 11): melisma begins DG acGaGF.

Be 40078: v. 1 “DOminus” (verse opening): DEF G(ori.)F GEFDG G bcaFG Gc ca cdcG cdc dbcaF (etc.).

Roman MSS: In the verses there are several significant variants between Vat 5319 and Bod 74.

Vat 5319:

  • v. 1 “DOminus” (verse opening): Vat 5319 has a much longer melisma than Bod 74, with an internal repetition: c c(ori.)babaGFGaG Gac cbaGGFGaGGacb

  • v. 1 “domiNUS” (verse opening): acbaG (the more common cadential pattern)

  • v. 1 “[induit] doMInus”: dcbc(ori.)ba

  • v. 1 “FORtitudinem”: abaG Gc

  • v. 2 “testimonia tua”: Vat 5319 has just “testimoniam”

  • “testimoNI/AM”: c c(ori.)ba/aGa c c(ori.)ba aGa

(p.259) RoS 22:

  • “Enim”: bcba

  • “terRE”: GcbabaG cabaGF GabaGaba

  • “SE/DES”: a different text underlay is suggested by the neume groupings:

  • abaGcdc/cb

Tui sunt caeli (9)

The pitch‐level variants in Tui sunt caeli have been discussed by Hankeln, Frasch, and Sidler. The base version of Ben 34 represents both the majority version and Hankeln's Aufführungslage. Several Aquitanian MSS in the sampling, however, do exhibit variants in pitch level. In Pa 776, Pa 780, and Pa 1121, some or all of the verses are written a fifth below their level in Ben 34, but, as Hankeln argued, the repetendum, written a fifth too low, suggests a performance level of Ben 34.1 Outside of the Aquitanian tradition, these pitch‐level variants are not found in the sampling incorporated into this study.

Melodic variants: There are very few major variants among core MSS.

Peripheral MSS have the following variants.

Cai 61:

  • v. 1 “metuenDUS”: Gca baG aG

  • v. 1 “TUam”: CF FD CD DFF DCD F FD FGFGG GFE

Pa 1235:

  • Respond, “Eius”: FF FEFED

  • Respond, “pre/pa/ra/ti/o”: E/EF/Ga/G/Gab‐flatGF FEDED

  • v. 1 “metuenDUS”: aca baG aG

  • v. 2 “faciEM”: DCDF FDCD

  • v. 2 “PLAcito”: GaGa

  • v. 2 “CORnu”: melisma begins FFF FEGE aGF ba GaGE

  • v. 3 “SIcut”: DF FGF FED (liq.)

  • v. 3 “DEXtera”: melisma begins Ga cbc

Pad 47:

  • respond “Eius”: FF FD FF GE FF FE

  • v. 1 “[circuitu] Eius”: GaGFE FGaGFE FGaGFE

  • v. 2 “CORnu”: DF FF DGE FG aba GaGF acccaF GaFE FGEC (etc.)

Pst 120:

  • respond “TERra”: GFEaGF

  • v. 1 “metuenDUS”: acabaGaG

  • v. 1 “[circuitu] Eius”: GbaG FaGFE FaGF

  • v. 2 “CORnu”: DF FF FD GE FG aba GaGF accaF GaGE FG FGFDC (etc.)

(p.260) Be 40078:

  • Respond “IUstiticia”: DaG

  • v. 2 “DEXtera”: DF cbc c ca bcaG acGE G G bcaG acG etc.

  • Roman MSS: Bod 74 lacks the third verse. There are no significant melodic variants between Bod 74 and Vat 5319.

RoS 22:

  • “ceLI”: EDEFGFEF

  • “TERrarum”: FGaGbaGFGF

  • “iuSTI”: aGaFGaGFGaGF(liq.)

  • “preparatiO”: GaGFEDFE

Reges tharsis (10)

Reges tharsis exhibits pitch‐level variants that were discussed long ago by Sidler.1 Pa 776 is employed as the base reading because its scribe notates the problematic passages in their probable unemended range. This MS, however, is somewhat anomalous in lacking a repetendum cue after the first verse and giving a varied, written‐out repetendum at the end of the second verse. In many MSS, a repetendum to “omnes gentes” is indicated after the first and final verses. The scribe of Pa 776 does not maintain a consistent horizontal reference point between the first and second verses, but the custos at the end of the second verse and the written‐out repetendum at the end of the third verse confirm that the intended pitch level is that of Mo 159 and the other MSS.

Pa 776 has one small difference from other MSS that is probably a scribal error. In phrase 6, at “filio regis,” Pa 776 has the notes gfff on the last two syllables of “filio,” with an unclear text underlay. I have emended it to the more common reading gf/f, eliminating the extra f.

v. 1 “et iusticiam”—end of verse: At “iusticiam” (phrase 6), Pa 776 expands into the upper part of the range and remains there through the end of the verse. This reading is matched in two other key core MSS, Mo 159 and Pa 1121, and in Be 40078. In the majority of MSS, however, this passage is written a fourth below Pa 776 beginning at “iusticiam,” with a focal pitch of c. In Mo 159, moreover, the written‐out repetendum to “omnes gentes” is also written a fourth higher than its pitch level in other MSS. While this cue might imply that the performance level was a fourth below the written level, Mo 159 consistently employs b‐flat here, resulting in an intervallic structure identical to that of the majority version a fourth lower.

Perhaps the most likely reason for the transposition is the need for a nondiatonic pitch, eb, on the word “iusticiam,” which is indicated with a parenthetical e‐flat in the base transcription. The melodic gesture on “iusticiam,” an upward leap of a fifth followed by a half step, is a common one in the chant repertory, especially in first‐mode melodies. Typically, the leap is placed a fourth lower than it is here, from D to a, with a b‐flat. To replicate this intervallic structure at the level of Pa 776, an e‐flat is required. Pa 1235, written a fourth lower, does indicate a b‐flat here. Mo 159, however, writes this segment in the higher register and indicates an e‐natural. While the reading of Mo 159 might cast doubt on (p.261) the nondiatonic practice here, I am inclined to view Mo's reading as a simple ad hoc solution to the problem. When the same melodic gesture is written at the lower level in Mo 159, it consistently employs the expected b‐flat. (There are too many instances to list, but see the opening of the second verse of Dextera domini [13].) It seems improbable that the same gesture would be sung differently at the higher level of transposition, except as an emendation. In Be 40078, the e is raised to f, reflecting the German chant dialect.

The early adiastematic MSS reinforce the impression that the higher version of the melody found in Pa 776 is the preferred reading. In the MSS written at the lower level, the downward transposition begins on the word “et,” which starts on either C or D. The scribe of Ei 121, however, writes tractuli on the final note of “da” and on “et” (phrases 5 and 6) with an “e” (“equaliter”) between them, indicating that “et” should start on F, the final note of “da.” This version is consistent with Pa 776 and the other higher readings, but not the lower versions.

Two of the German MSS in the sampling, Gr 807 and Be 10086, exemplify a phenomenon I have observed in other offertories. These MSS represent a tradition in which a nondiatonic practice was probably suppressed, but the problematic segment of the melody is nevertheless written in the “emended” range, perhaps out of custom. In both MSS, the note corresponding to the b‐flat in Pa 1235 (and the e [flat] in Pa 776) is c, reflecting the German chant dialect. With this melodic change, the e‐flat at the higher level would no longer be needed, because the corresponding note at the higher level would be f (as it is in Be 40078).

Although Lo 4951 is written a fourth lower, its melodic reading differs here. “Et iusticiam” reads F/F/aGab(quil.)ca, probably a scribal emendation that facilitates the transition to the lower level.

Pitch level at “iusticiam”: =Pa 776: Mo 159, Pa 780 (joins the lower level at “filio”), Pa 1121, Pa 1132, Be 40078. A fourth lower: Lo 4951, Ben 34, Ben 35, Pa 1235, Cai 61, Gr 807, Mü 10086, RoV 52, Pad 47, Mod 7, Pst 120.

v. 3 “[suscipiant]MONtes” (phrase 9): This melisma varies at the point where Pa 776 recites on b (flat). The following sample readings begin with the first descent to D:

Pa 776:

D FG FFFFD

F babbbGbdc cccacdd (ori.) cdcd

Pa 1121:

D FGF FFF (ori).

D aba aa a(ori.)F bab bbb(ori.)G bccbcbc

Pa 780:

D FGF FFFD

F aG aa a(ori.)F a cb cc c(ori.)a cd d(ori.) c dcd

Lo 4951:

D FFGFFFFD

ac ccca acddcdcd (liq.)

Mo 159 (b's are flat):

D FGF FFFD

Gcb bbbG bdc ccca cddc dcd

Cai 61:

D FGF FF FD

FaG G GEF a aG a(liq.)

RoV 52:

D FGF FF FD

FaG GGGE Gcb c ca cd dcdcdc(liq.)

Pa 1235:

D FGF FF FD

FaG cc ca cc cac

Gr 807:

D FGF FFFD

cdc ccca cddc dcdc

The reciting pitch in the last part of the melisma is alternately c, b (flat), a, or even, in one case, G. Although there are intervallic differences between the versions, it is not possible to reconstruct an unemended reading. The variants may reflect a point of melodic instability rather than a non‐diatonic practice.

(p.262) v. 3 “Orietur…eius” (phrase 11): In Mo 159, this opening portion of the verse is written a fourth above its position in Pa 776. Mo 159 employs a b‐natural on “eius,” equivalent to f‐sharp at the level of Pa 776. The same upward transposition is found in other core MSS, including Ben 34, and in the peripheral MS Pad 47. That “eius” is the source of the transposition is suggested by the variant treatment of the verse opening, especially on “eius,” and the rejoining of most sources (in pitch level) beginning at “iusticia” (phrase 12). In some MSS, for example, the verse begins at the level of Pa 776 but is briefly written a fourth higher on “eius.” This solution is found in Lo 4951, the south German MS Mü 10086, and, with a somewhat different melodic reading, Pst 120.

Although the great majority of MSS join the level of Pa 776 at “iusticia,” two MSS included in the sampling, Pa 1235 and RoV 52, present reworked versions of the verse. Pa 1235 has a substantially different melody on “eius” and, beginning at “iusticia,” is a fifth below Pa 776, joining at “et dominabitur.” There are no intervallic differences that would suggest the use of a nondiatonic pitch. RoV 52 is a fifth below Pa 776 at “iusticia,” and in the following passage presents a different version of the melody that is alternately a third and a fourth below Pa 776, joining at “luna.”

Sample readings of “eius”:

Pa 776:

GEF(‐sharp)GaGabGGa

aGGEDDCD

Mo (b's are natural)

c ab cdcdeccd

bccaGGFG

Lo 4951:

cacd cdeccd

dcdccaGGF

Pa 1235:

GFFGFGaFFG

GFGF

RoV 52:

FDEFGFGaFFG

GFDCD

Pst 120:

cGabccdeccd

dcbaGa

Roman MSS:

Vat 5319:

  • v. 1 “iustiTIa”: dedc

  • v. 2 “PAcem”: dec

  • v. 2 verse ending, “iudiciO”: a longer melisma matching that of v. 1: FGaG c c c(ori.)baGFGaG c c c(ori.) baGcd

RoS 22 has no significant variants.

Iubilate deo omnis (11)

The Roman version has a longer text than the Gregorian does, lacking the text repetition in the respond. To facilitate comparison with the Gregorian version, I have adopted the phrase numbering of the Roman version.

This piece has a complex transmission. It is unlikely that any the MSS in the sampling notates the entire offertory at an unemended pitch level. Pitch‐level variants are found in both verses.

(p.263)

Appendix 3 Commentary on Edition

Example AP 3 11.1

The first variant point occurs in v. 1, phrase 5, at “non ipsi nos”:

non

ip‐ si

nos

nos

Pa 776:1

bcab

G GabaGa

aG

Gbdbdcdba

Ben 34:

bcGa

a abcbab

ba

FacacbcaG

Mo 159:

bcac

c bcdcbc

cb

acedecdca

Pa 1121:

bcac

c bcdcbc

cb

cdfdfefed

Lo 4951:

acGa

F FGaGFG

GF

FacbcaG

Pa 1235:

GaGFG

G Gacab

aG

Gacbdcdcdca

Cai 61:

bcab

b bcdcd

cb

acedecdca

Gr 807:

acGa

a accbab

ca

cdfdfefdc

Mü 10086:

bcac

c accbab

cb

cdfdfefdc

Be 40078:

acGa

a abcbab

ca

aceecdca

Frutolf tonary:

acGa

[c]a

acedecdca

Mod 7:

ccac

c acdcbc

cb

cdedecdcb

Pad 47:

ccac

c acdcbc

cb

bcedecdcb

Pst 120

abGa

a abcbab

ba

acdcdcdca

(p.264) Nearly all MSS join at “autem” or “populus” and remain at the same level throughout the verse. There are intervallic differences between the various versions at the cadential pattern on “ipsi nos” and the following passage on “nos.” On “ipsi nos,” Mo 159's cadence on b‐natural, matched in several disparate MSS, revolves around the notes b‐natural, c, and d, a semitone followed by a whole tone. The versions a tone lower, presumably sing with a b‐natural, have a tone followed by a semitone, a‐b‐natural‐c. The distinctive F-cadence in Lo 4951 has two whole tones, F-G-a. Intervallic differences are even more prominent at the second “nos,” where there are melodic variants as well as pitch‐level differences.

While I will not attempt to propose a single “original” version of this passage, the bewildering variety among the MSS suggests the presence of a nondiatonic pitch. In this rare case, an early adiastematic MS, Ei 121, offers enough information to help in a (very) hypothetical reconstruction. A few contrasting versions of the passage are presented in example 11.1. I have included the last few notes of the melisma on “nos” that precedes the problem spot. Since this melisma is uniform in the pitched sources, it is reasonable to suppose that the melisma ended on a in Ei 121. In Ei 121, the melisma is followed by a virga and the letter “s,” indicating a higher pitch on “et,” most likely b or c, followed by an “equaliter” sign indicating that “non” starts on the same pitch as “et.” “Ipsi,” in turn, begins on the same pitch as the closing note of “non.” There is an “e” between the two “nos's,” indicating that the second one starts on the closing pitch of the first one.

Each of the pitched readings in the sampling is contradicted by an indication in Ei 121 at some point. The most likely reading, as I have proposed with the solution in line 5 of example 11.1, is a version that is at the level of Mo 159 until the second “nos,” where, according to the “equaliter” sign (matched in only one pitched MS, Pad 47) it is a tone higher. This reading would produce the nondiatonic pitch f‐sharp. The variety of readings in the second “nos,” summarized above, supports the hypothesis that the second “nos” is the source of the problem. With the exception of Ben 34, the core MSS begin the second “nos” with the equivalent of a minor triad, placed in portions of the scale where one is possible: a‐c‐e, G‐b (flat?) d. If this minor third is a preferred reading, however, the Beneventan versions appear as reworkings of the passage, and Lo 4951 presents a similar reading. Because of the hypothetical status of this reconstruction, I have not incorporated it into the base transcription.

Verse 2: Although I have not been able to establish a definitive unemended reading of this verse, I can offer some hypotheses, building on previous work of Sidler, Pitman and Hankeln.2 There are three variant points: the beginning of the verse, “et usque (phrase 10), …” and “veritas” (phrase 10). The pitch level in the three variant points is summarized below, followed by a discussion of each passage.

  1. 1. “Laudate nomen eius…misericordia eius” (phrases 7, 8 and 9):

    • = Pa 776: Ben 34, Ben 35, Pa 780, Lo 4951 (?),3 Pa 1235, Pad 47, Pst, 120

    • 4↑ Pa 776: Mo (third? at “dominus”), Cai 61

    • 5↓ Pa 776: Be 40078

(p.265) Within this passage, “dominus” is a point of variance. The interval of transposition between Mo 159 and the other MSS changes to a third here, whereas Cai 61 remains mostly a fourth above the other MSS:

Mo (with b‐flats):

cdegffec

/dcc(ori.)bc

/cb

Cai 61:

cdegegfe

/feecd

/dc

Pa 776:

Gacdbdcb

/cbbGa

/aG

Pa 1121, Pa 780, Ben 34 are the same in pitch level as Pa 776. The majority of peripheral MSS match Pa 776, with the exception of Pa 1235, which cadences on a.

  1. 2. “et usque in saeculum saeculi” (phrase 10):

    • =Pa 776: Pa 780, Pa 1121

Pa 1235 = Pa 776 at “et usque” then is a fourth higher (joining Mo 159) at “saeculi.”

Pst 120 =Pa 776; interval of transposition changes to 2↑ at “saeculum saeculi.”

  • 4↑ Pa 776: Mo 159, Lo, 4951,Ben 34, Ben 35, Cai 61

  • 5↓ Pa 776: Be 40078 (2↓ at “[sae]culum saeculi)

  1. 3. “veritas,” final melisma (phrase 10):

    • =Pa 776: Pa 780, Pa 1121

    • 3↑ Pa 776: all others

Passage 1: Mo 159 begins the verse a fourth higher than most other MSS do, with a referential pitch of f rather than c. Sidler argued long ago that Mo 159's reading was a transposition made to represent the pitch F-sharp on [suavis] “est” where Mo 159's b‐natural corresponds to F-sharp a fourth lower. These intervallic differences suggest that level of Pa 776 is the preferred reading.

Passage 2: In the higher version of Mo 159, b‐natural's occur on the second syllable of “saeculum” and again at the cadence on “seculi.” Although most non‐Aquitanian MSS are a fourth above Pa 776 by this point, two Aquitanian MSS in the core group are at the level of Pa 776, requiring an F-sharp to duplicate the intervallic structure of Mo 159.4 The cadential figure on “saeculi,” which revolves around c, b, and a, is equivalent to G, F-sharp, and E at this lower level. Two MSS included in the sampling, Pa 1235 and Pst 120, begin this passage (“et usque”) at the level of Pa 776, moving to the higher register of Mo 159 at “saeculi.”

In previous editions of this piece, the passage is treated in various ways. Pitman presents a consensus Aquitanian reading that matches the pitch level of Pa 776. Hankeln, however, notes that the repetendum cue is notated a third too high in some MSS, including Pa 776. He thus proposes a discrepancy between the written and notated levels at the end of the verse. Beginning with “in saeculum,” he posits an Aufführungslage that is either at the level of Mo 159 or a tone lower, joining Mo 159 at “veritas.” As Hankeln's two alternative suggestions imply, no solution to this problem is fully satisfactory. While each of Hankeln's hypotheses is plausible, I have opted for an unemended reading that is at the level of Pa 776 (p.266) through variant point 2. Several factors suggest that the version of Pa 776 is the preferred reading of this passage, including the nondiatonic pitches implied by the variants, the choice of the same pitch level in unrelated MSS such as Pa 1235 and Pst 120, and the variety of the other versions. On the second syllable of “saeculum,” for example, where Mo 159 employs a b‐natural that suggests an F-sharp at the level of Pa 776, Pst 120 suddenly shifts down a whole tone. The last two syllables of “saeculum” are FE/FFF DED, a tone below Pa 776, allowing the irregular semitone G/F-sharp to be represented with F/E. This reading of “saeculum” is matched in Be 40078, which has been a fifth below Pa 776 until “saeculum,” where it is a whole tone lower. At the next problem spot, on “saeculi,” Pst 120 and Pa 1235 join the level of Mo 159, a fourth higher, with brief transitional sections, and Be 40078 remains a whole tone below Pa 776:

sae‐ cu‐

li

Pa 776:

EEDEGGG/EF(#)GFEF/F(#)E

Mo (b's are natural):

a aGa ccc/ abcbab

ba

Pa 1235:

GGAGccc/abcbab

ba

Pst 120:

FGaaGaccc/abcbaGa

aG

Be 40078:

DDCDFFF/DEFEDE

FD

This variety implies the presence of a nondiatonic pitch at this point, most likely at the level of Pa 776 and the other Aquitanian MSS.

Passage 3: At final melisma on “veritas,” the great majority of MSS are at the level of Mo 159, a fifth above Pa 776, but the three Aquitanian MSS write the passage a third lower, with some melodic variants. The preferred reading is not clear. I have given both the notated level of Pa 776 and the hypothetical performance level suggested by Hankeln, a third higher. Hankeln's suggestion is consistent with the repetendum cue and is the most probable solution. It is not clear, however, why this passage appears a third too low in a majority of Aquitanian MSS.

Other melodic variants: respond, first “TERra”: Some MSS in the core group have a few extra notes:

Pa 776:

Fac edc ac bbG

babc GFGF

Lo 4951:

Fac edc ac bbG bbbG

babc GFGF

Ben 34:

Fac edc acbbG bb bG

babc GFGF

v. 1 “iPSE”: Beneventan MSS present a distinctive and shorter reading of this melisma, as in Ben 34: cb decb dec c ca cc d dc dcd

Roman MSS: Vat 5319:

  • respond, “DOminus”: cccdedcb

  • v. 1 “[fecit] NOS”: cdcc(ori.)baGcd

  • v. 1 “[pascue] Eius”: GcaGaGaF(liq.)/GaGF (mostly written a tone below Bod 74).

(p.267) Iubilate deo universa (12)

The sources show some variants in the longer melismas, particularly among MSS outside the core group.

  1. 1. v. 1 final syllable of “dixerunt”: At the beginning of the melisma, some of the sources outside the core group differ in their use of the pitches a and c:

    • Ben 34: GFaaGccaccacGF (etc.)

    • Be 40078: GFaaGaaGccacGF (etc.)

    • Pst 120: aGaaGaaGccabGF (etc.)

    • Pad 47: baccaccaccacGF (etc.)

    These differences are replicated when the melisma returns in v. 2, on the final syllable of “medullata.”

  2. 2. End of v. 1 “labia” melisma: Three of the Italian sources in the sampling, Pad 47, Pst 120, and Mod 7, exhibit significant variants from the majority version in the first part of the melisma. Pst 120 transposes a segment of the melisma up a whole tone. Because this version is identical to the majority in intervallic structure, the reasons for the transpositions are unclear. A similar transposition, however, is found in Pa 1235; Justmann's transcriptions show a partially similar reading in Brussels 3823.1

    • Pad 47:ac cF abGF aG cba Gcba Gcbc ab-flatcbab GaGF (etc.)

    • Pst 120:ac cG abGF aG dc cb adccb adcd bcdcbc aba aG abc dcbaGaG

  3. 3. v. 2: Final melisma on “offeram”: Sources outside the core group show some small variants, often particular to individual MSS. Near the end of the melisma, Mo 159 and several other MSS reach several times to d.

    • Ben 34: GacccabaGaccc

    • Mo 159: Gadcdcccc(ori.)Gcdc

    • Pst120 begins the melisma slightly differently: caaGabGFGaGacaaGabGF…2

    • Pad 47: at the end of the melisma, the notes aGF are stated only twice (not three times, as in other sources).

For reasons that are unclear, To 18 has a completely different ending of the second verse, beginning with “medullata.” The melisma on “offeram” shares some material with the other readings, but begins differently and is truncated: (p.268)

  • a  c(liq.) cdede cdadcdc

  • me‐  dul‐  la   ta

  • edecbbcdedeccccaGabGFGaGacccacdccccaGacaGaddcddcccaGFaaGcccaGFbaGGacc

  • of-

  • edc  dca a

  • [of-] fe- ram

Roman MSS: Vat 5319 and RoS 22 show few significant differences from Bodmer 74, with one exception: In v. 2, Vat 5319 has the repeat of the words “locutus est os meum…” that are present in the Gregorian sources but lacking in Bodmer. The opening two notes of Vat 5319 are GG, contrasting with DD in Bodmer. In v. 2, Bodmer is missing the first note of “medullata” (or at least it is not visible on the facsimile). In Vat 5319, the corresponding note is F.

Dextera domini (13)

This offertory has variant repetenda. In base version of Lo 4951 and in most Aquitanian, Italian, and Beneventan MSS, the cue directs the performer to the beginning of the respond, leaving open the question of whether the initial phrase “dextera domini fecit virtutem,” constituted the repetendum or whether the whole respond was repeated. In Mo 159, the repetendum is to “non moriar.”

There are two problems of pitch level in this offertory. The first involves the notated level of the respond and first verse. In Mo 159 and in a few other MSS, the respond and first verse are notated at the affinal position, undoubtedly because of the need for low B-flat, which is represented by F at the higher level of transposition. In a majority of MSS, however, the respond closes on D, without emending the passages with low B. Although a low B-flat is clearly intended, it is indicated only in a few later MSS that employ a notational sign for low B‐flat, such as Gr 807.

A second problem arises in the final verse. In a great majority of MSS, the second verse is a first‐mode melody, as it appears in the base transcription. The scribe of Pa 776, however, notates the second verse a fourth lower, a position confirmed by the repetendum cue. To replicate the intervallic structure of the majority version at the position of Pa 776, an f‐sharp is required. A similar version is found in many MSS from St. Martial and in RoV 52.

With the principle that the lectio difficilior is the more powerful reading, the version of Pa 776 would emerge as the preferred pretheoretical version, and the majority version as the results of modal emendation. The presence of this reading in one Italian MS included in the sampling indicates that it was not limited to Aquitaine. The second verse is also modally irregular in the Roman tradition: melodic material typical of first‐mode Roman melodies is projected a fifth below its usual pitch level. Given the uncertainties surrounding the performance of this version, I am hesitant to posit it at the single preferred version. The version in the normal first‐mode range reflects the more common practice, and I have given it in the base transcription.

The lower version of the verse found in Pa 776 (Example 13.1) is modally irregular. In the second verse, melodic vocabulary typical of protus chants is projected a fourth below its usual position. The melodic activity of this verse is focused within two disjunct T‐S‐T tetrachords, A‐B‐C‐D and E‐F‐sharp‐G‐a, with C, the tritus pitch of the first tetrachord, (p.269)

Appendix 3 Commentary on Edition

Example 13.1A

(p.270)
Appendix 3 Commentary on Edition

Example 3 13.1B

serving a repercussive function. The parenthetical F-sharps in the transcription are derived from the b‐naturals that occur in the majority version a fourth higher, and are consistent with the typical melodic gestures of the chant repertory.

Pitch‐level profile: =Lo 4951: Pa 780, Ben 34 (lacks second verse), Cai 61, Pa 1235, Gr 807, Be 40078, Mü 10086, Pad 47, Pst 120, Mod 7.

Respond and first verse at affinal position, second verse = Lo 4951: Mo 159.

Second verse 4 ↓ Lo 4951: Pa 776, Pa 1134 (according to rep.),1 Pa 1136, Pa 1137 (no rep.), Pa 1135, Pa 1132 (according to rep.), RoV 52. (Pa 1121 lacks a notated repetendum.)

Individual MSS: In Pa 1121, the verse begins on a new line, with no repetendum cue. In Ben 35, the scribe alternates between F and c clefs, obscuring the intended pitch level. In Pa 780, the written level of the verse (relative to the first verse) matches that of Pa 776, but the repetendum cue begins on the next page, with no custos.

Other variants: There are very few significant variants among core or peripheral MSS. v. 2 “cadeREM” has a different reading in Beneventan MSS, as in Ben 34: FaGF FDF FDF FF.

Roman MSS: Respond, “narrabo opera”: The text underlay, indicated by the placement of the text and the neume groupings, is a point of variance in the three Roman MSS. RoS 22 and Bodmer present similar readings, with a different version in Vat 5319:

[narra] bo

o‐

pe‐

ra

Bod 74

DE

DFEDC

FGF

EFGFGFEF

RoS 22:

DE

DFEDCDF

GF

EFGFGFEF

Vat 5319:

DEDFEDC

FGF

[G]?

EFGFGFEF

In Landwehr‐Melnicki's transcription of Vat 5319, a G is given over “pe” that is not visible on the microfilm.2 Although the melody is not formulaic here, it does employ some standard melodic material, and Bod 74 and RoS 22 align the text and melody in a way that is (p.271) atypical. The notes DFEDC that occur on the first syllable of “opera” in Bod 74 and RoS 22 are more commonly used as a cadential figure (especially in eighth‐mode offertories, at the higher transposition), and the figure FGF would typically fall on the accented syllable, as it does in 5319. Despite its unusual alignment between melody and lyrics, however, the presence of a reading close to Bod 74's in the much later RoS 22 suggests that it is not a simple mistake.

Vat 5319:

  • respond, “moRI/AR”: GFE/EFEDCDC (liq.)

  • respond, “VIvam”: DEFEDCDCA

  • v. 1, “es”: DF FFEFEDCC(ori.)BACDE

  • v. 2 “MI/CHI”: BCDED/DCDEDCDC

  • v. 2 “saLU/TEM”: D FGFGFF(ori.)DCDEDC/ DC FFEFEDCC(ori.)BACDE

  • v. 2 “et dominus suscepit”: Vat 5319 lacks “et”

RoS 22:

  • “dexteRA [domini fecit]”: DCDEDEF FED

  • “virTUtem”: EEDCDEFED

  • “exalTAvit”: melisma begins FGaGFG (etc.)

Bonum est (14)

The Gregorian and Roman versions show a structural difference in the respond and repetenda. The Roman respond ends with a repeat of the opening, “bonum est confiteri,” that is lacking in the Gregorian, and this reprise is altered melodically to form a final cadence. It is probable that this repeat was sung as the repetendum in the Roman version, whereas the whole respond was probably sung as a repetendum in the Gregorian version.

Mo 159 indicates b‐flats in several places where they do not occur in other MSS, resulting in a G‐protus reading of the respond's opening phrases. In the respond, Mo 159 has b‐flats in the following places: respond, “boNUM”; respond, “noMIni”; respond, “TUo” (matched in Pa 903; indicated with a quilisma); v. 3 “corNU” (matched in Pad 47); v. 3 “MEum.”

Only two of these b‐flats, then, are confirmed in another MS. For the other passages, there are no flats indicated in the following MSS that often indicate b‐flat: Gr 807, Be 40078, Mü 10086,1 and Mod 7. In the case of the German MSS, however, several of the notes corresponding to b‐flat in Mo 159 are c's, reflecting the German chant dialect. Even though b‐flats were undoubtedly sung at some times and places, it is not clear how widespread this practice was, and I have decided not to give these parenthetical b‐flats in the transcription of Ben 34.

Aside from these differences in the use of b‐flat, there are very few melodic variants among core or peripheral MSS:

  • v. 3 “[cornu] ME/UM” and “[inimicos] ME/OS”

  • Mo 159: cdc b‐flat a/GaG

(p.272) Verse endings in Lo 4951:

  • v. 1 “tuE”: GGa Ga G baG

  • v. 2 “inquitaTEM”: GGa Ga G baG

  • v. 3 “tuA” (sic): cdccaG Ga Ga G baG

Among peripheral MSS, most differences involve variants in recitational style, with the following exception in the opening of the respond:

Bo‐

num

est

Pad 47:

FF

FaG

GE

Identical reading in Mod 7.

Roman MSS:

Vat 5319:

  • [Bonum] “est”: D EDEDCBDFEDFED (both times)

  • v. 2, “iniMIci”: EDCDE (whole tone above Bod. 74)

RoS 22:

  • [Bonum] “est”: DEDEDCB DFEDFED (both times).

Perfice gressus (15)

There is one point of melodic variance where the Beneventan MSS give an independent reading, on v. 2, at “ut pupillam oculi” (phrase 8). The following shows the passage in selected core MSS.

o‐

cu‐

li

Ben 34:

acbc

cabG

G

Mo 159:

b acbG aE Gbab

FG

G

Pa 776:

b acbG aFG Gbab

EG

G

Lo 4951:

b acbG GFE Gbab

FG

G

All peripheral MSS present a reading closer to Mo 159 and the Aquitanian MSS, though there are further variants in some of these sources:

o‐

cu‐

li

RoV 52:

a Gba GFGE Gc abc

G

G

Pad 47:

acba bG aGaG

Ga

a

Gr 807:

c acbG aFGcac

FG

G

Cai 61:

b ac bG aF acac

FG

G

Roman MSS: Bod 74 lacks the verse Custodi me and presents a shorter version of the verse Exaudi domine, missing the end of the verse, “auribus percipe orationem meam.” At the end of the verse Ego autem, Vat 5319 lacks notation for the words “gloria tua.” In the (p.273) transcription, these notes are supplied from Bodmer. There are very few melodic variants between MSS. Bod 74: v. 2 opening of the verse “Ego”: ab.

Benedictus es…in labiis (16)

The transpositional variants in the second verse of this offertory were described long ago by Sidler and further addressed by Hankeln.1 Based on the intervallic differences between MSS, both authors persuasively argue that the transpositions are attributable to the presence of the nondiatonic pitch high f‐sharp. Ben 34 was chosen as the base version because its scribe notates this verse in the probable unemended range, corresponding to Hankeln's Aufführungslage. In this respect, this piece is a departure from this scribe's usual tendency toward emendation.

The first hypothetical f‐sharp occurs at the repetition of the words “viam iniquitatis” (phrase 13) and the second at “de lege tua” (phrase 14), followed by a series of f‐sharps in the same phrase. Mo 159 is written a whole tone below Ben 34, with a b‐flat, so that the semitone e‐f in Ben 34 is equivalent to the whole tone d‐e in Mo. The third hypothetical f‐sharp occurs in the final melisma.

Although the intervallic differences between the versions, especially Mo 159, certainly suggest the use of f‐sharps in these passages, it is unclear how widespread and longstanding this practice was. Outside of Aquitaine and Benevento, most MSS place the verse a fifth below the reading of Ben 34, a transposition that would allow f‐sharp to be represented with b‐natural. Many sources, however, do not clearly distinguish between b‐flat and b‐natural. In Pa 1235 and Cai 61, the verse is notated a fifth below Ben 34. Where the first hypothetical f‐sharp occurs, on “viam iniquitatis,” the notes corresponding to the F-sharps are c's rather than b‐naturals (a reading also found in German MSS). At “lege,” the scribes of Pa 1235 and Cai 61 indicate a b‐flat, equivalent to f‐natural at the higher position. Although the b's that follow, on “tua” and “mei,” are not indicated as either flat or natural, it seems probable the flat was intended to apply to the whole passage. If so, however, the reasons for writing the verse in the lower register are not clear. In Cai 61, all the remaining b's are indicated as flat, even in the final melisma, where two hypothetical f‐sharps are suggested by Mo. In the final melisma, Pa 1235 indicates a b‐flat corresponding to the second hypothetical f‐sharp, but not the first. Cai 61 and Pa 1235, then, may represent a tradition in which the nondiatonic practice suggested by Mo 159 was suppressed but the verse was nevertheless written at the lower level of transposition. In other later MSS, such as Gr 807 and Be 40078, the b's on “lege,” “tua,” and “mei” can perhaps be assumed to be natural in the absence of a flat sign.

The question of how far‐reaching the nondiatonic practice was also arises in the case of Ben 34. Given this scribe's theoretical awareness and tendency to transpose problem spots, can the notation of the verse in Hankeln's Aufführungslage be seen as a suppression of nondiatonic practice? While the answer is not clear, the scribe of Ben 35 writes the verse a fifth lower, indicating that this practice was known at some Beneventan institutions.

(p.274) Another point of variance concerns the repetition of the words “iudicia tua” (phrase 16). This repetition is lacking in Ben 34, Ben 35, Compiègne, and the Roman MSS. While the circulation of the version without the repetition in these disparate sources may suggest that it is an alternative early reading, the text repetition is present in all other MSS, including the early notated sources in the sampling and Ben 38. In the transcription, this material is supplied from Ben 38.

As noted in Hankeln's summary of Aquitanian sources, the verse incipit is a point of variance. The following summarizes the pitch level at “domine” though the end of the verse:

  • = Ben 34: Pa 776 (according to repetendum); RoV 52

  • 2↓Ben 34: Mo

  • 5↓Ben 34: Lo 4951, Cai 61, Pa 1235, Gr 807, Be 40078, Pad 47

The pitch level of the second verse in some of the Aquitanian MSS is unclear. As noted in Hankeln's study, several Aquitanian sources appear to begin the verse on E, but the pitch level of the repetendum suggests that the scribe did not maintain a consistent axis between the two verses. With a reading of the pitch level derived from the repetendum, the verse incipit in Pa 776 is a tone above that of Ben 34, joining at “domine.” The reasons for this discrepancy are unclear. The repetendum is lacking in Pa 780. As summarized by Hankeln, the St. Martial sources (with the exception of Pa 1132) seem to be written a third below the level of Ben 34, but the repetendum cues give conflicting information about the intended pitch level.2

The sources are inconsistent in verse division and repetenda, and the designation of five verses (with no repetenda until the end) in the Beneventan MSS is unusual. A more common practice is to divide the material into three verses, Beati immaculati, In via testimoniorum, and Viam iniquitatis. In the Beneventan and several Aquitanian MSS, the last verse closes with a cue to “aufer a plebe” (phrase 8), suggesting that perhaps the performance concluded at the end of this verse, resulting in a migration of final from E to the affinal b.

There are few significant melodic variants, either among core or peripheral MSS.

Lo 4951:

  • respond, “omniA”: cbca

  • v. 1 “[exquirunt] eam”: ccc a cc edb ccc adcb

Cai 61: respond, “TUas” (2nd time): Gacc dec a cb cc cb

In verse 2, Ben 34 has a probable scribal error on the last syllable of “tua” (phrase 8), where it has a single e instead of e‐d, as found in the other MSS. I have emended the transcription here.

Roman MSS: The Roman verses have first‐mode characteristics, a trait that probably derives from the close verbal relationship between the first verse, Beati immaculati, and Confitebor tibi (39). In the Gregorian tradition, Confitebor tibi has a first verse, Beati immaculati, that is verbally identical to the first verse of Benedictus es. Although the Roman version of Confitebor tibi lacks this verse, the Roman verse of Benedictus es shows a compelling musical similarity to the Gregorian verse of Confitebor tibi. In Bod 74, there are several additional verse indications, at “aufer,” “viam veritatis,” and “viam mandatorum.” (p.275) Because these seem to mark sections rather than complete verses, I have not indicated them. Vat 5319 lacks notation beginning with the verse In via testimoniorum. There are few significant variants among the sources.

Vat 5319:

  • v. 1 “beAti”: a slightly shorter version of the melisma, lacking the last 4 notes of Bodmer's melisma

  • v. 1: Bodmer's reading “beati qui servantur” (Ps. 118:2) is probably a simple error. Vat 5319 has “scrutantur,” the reading of all known psalters.

  • v. 1 “AU/FER”: a/aG

  • v. 1 “obproBRIUM ET”: recitation on c

Exaltabo te (17)

The respond and first verse of this offertory are often written at the affinal position, undoubtedly because of the need for low B-flat in the respond. The transposition of the first two sections raises questions about the intended pitch level of the final verse, which, in most cases, is written in the normal first-mode range (in relation to the transposed plagal respond and first verse). The most likely preferred reading is that of Pa 776, given as the base version, where the respond and first verse are in the normal plagal range and the second verse is in the authentic range. The following summarizes the pitch-level profile of the sources.

There are very few significant variants among core or peripheral MSS. v. 2 “deCOri”:

  • Ben 34: c bca ccG aFa (etc.)

  • Mod 7: c bca cca aGa (etc.)

  • Ben 34: v. 1 “aniMAM” (5↑)

Roman MSS: In Vat 5319 the whole offertory is notated at the affinal position. There are no significant variants between MSS.

Domine vivifica (18)

This offertory exhibits few significant variants between MSS. The Beneventan tradition, however, does show some small differences from other core MSS, as evident in the summary below.

respond, “[eloquium] TU/UM”: This is a point of variance:1

  • Mo 159: FGEFDED/D

  • Pa 776: FGFD/FFFDED

  • Pa 780: FGF/FED

  • Pa 1121: FGFF/DED (same in Lo 4951, Pa 1235, Cai 61; Pad 47 FGF/DED)

  • Be 40078: FGF/DFD

  • Gr 807: FGFE/DFD

(p.276) respond, “testimoniA”: Pa 780: ccba bG abaGG

respond, “tuA”: Pa 780: EGEF/FE

v. 1 “DOmine”:

  • Mo 159: Gcca ccaG bdb

  • Pa 776: Gcca ccaG cdc (identical reading in Pa 780, Pa 1235, Cai 61, Pad 47)

  • Lo 4951: Gcca ccaG bdc Be 40078: Gc ca bcaG adc

  • Gr 807: G bca bcaG adc

  • Mod 7: Gc ca cccaG acb (identical to Ben 34)

  • v. 1 “[misericordiam] TUam”: Mo 159: bdbc (identical reading in Pa 780)

  • v. 2 “beneplaCIto”: Mo 159: bdbc (same in Pa 780)

  • Variants in individual MSS:

  • Lo 4951: v. 2 “domiNE”: melisma ends ccc a a(ori.) FGF (ori.) E

Pa 1235:

  • v. 2 “mandaTA”: cbcbced

  • v. 2 “benePLAcito”: ca dcdc ded

  • Be 40078: respond “domiNE”: ab(quil.)cdc bca aG adc

  • Mod 7: v. 2 “benePLAcito”: ca dcdc ded

  • Pad 47: v. 2 “benePLAcito”: cb cbcb ded

Roman MSS: Bod 74 lacks the second verse. Bod 74 and Vat 5319 exhibit a rare transpositional variant at the beginning of v. 1, where Bod 74 is written mostly a third lower than Vat, joining at “secundum.” At the level of Bod 74, the verse opening is nearly identical to that of two other verses, in Gressus meos (v. Cognovi domine) and Levabo oculos (v. Legem pone). It is not clear, however, that the reading of Vat 5319 is in error here, since passages with figural similarities are sometimes written at different pitch levels. With only two MSS transmitting the Roman version of this verse, there is not enough information to establish a preferred reading. I have selected Vat 5319 as the base version simply because it is the only MS to contain the second verse.

Scapulis suis (19)

In the Gregorian tradition, there are only minor variants between MSS, as the following sampling shows.

Verse 1: “tiMOre”: Most MSS begin the melisma with a, as in Mo 159: accc (ori.) ac; or Pa 776: acccabc.

Verse 2: a. In the Beneventan MSS, “UMquam” and “ofFENdas” are a literal repetition of “SUis” and “manDAvit.” While the sources in the core group are nearly uniform at “suis” and “mandavit,” they are variant when the material recurs, possibly because of the liquescent syllables.

“UMquam”: Mo 159, Pa 1121, and Lo 4951: Faca b (flat) a G (liq.); Pa 776: F acacaaG (liq.)

“ofFENdas”: Mo 159: Facab‐flat a a (ori.) F (liq.); Pa 1121 and Lo 4951: FacacaaG (liq.); Pa 776: FacacaaG (liq.)

(p.277) Verse ending, melisma on “TUam”: Mo 159: acccFGaGaG. Also a point of variance outside the core group, i.e. Pa 1235: caGaGacaG; Gr 807: bcaFGacbcaFGF.

Verse 3: melisma on “eum”: The following segments of the melisma take a number of slightly different forms:

  • Ben 34: f f (ori.) dd (ori.) cd fffdd (ori.)

  • Mo 159: ffefcdfff(ori)efcd

  • Pa 1121: efdd(ori.) cdfefdd (ori.) cd

  • Mod 7: ffdecdfffdecd

There are minor variants toward the end of the melisma, for example:

  • Ben 34: ccc da G Gac dc ac dcd

  • Pa 776: ccccaFGacdcabc (quil.) dcd (similar reading in the other Aquitanian sources)

  • Gr 807: cc dca abc dcG abc

Individual sources:

  • Pa 1235: v. 3 “speraVIT”: decdcbcb

  • Cai 61: v. 3 “tuUM”: cc ca cbG aG ca Ga

To 18 v. 2 “eum” melisma: some segments near the beginning of the melisma adopt e rather than f as the top note. The melisma begins: dacdfffdc dcd ff gfdc efed dc eedcd eeecbc e ecdca (etc.).

Roman MSS:

Vat 5319: “NOCturno”: bcecedcdc.

Verse 2 (major variant): Vat 5319 presents a much shorter version of the final melisma on the first syllable of “eum,” lacking the repetition scheme of Bod 74's version. The entire melisma reads ededed cbcd d fedc cdc.

RoS 22:

  • “tiBI”: GaGbaGFGF

  • “domiNUS”: FEGFEFE

Levabo oculos (20)

The parenthetical b‐flats are derived primarily from their use in Mo. In many later MSS, the b‐flats in passages such as “doceas” (phrase 2) are raised to c. The G‐protus opening of v. 1, however, is maintained in some later MSS, with b‐flats indicated in Gr 807 and Pa 1235. None of the sources examined undertakes a whole‐tone transposition here, as might be expected.

The few points of variance are as follows.

End of respond, “domine”: Mo 159 has “tua,” with the cadential pattern altered for a two‐syllable word: Ga(quil) cab/aG. Ben 34's version is found in La 239, Pa 776 and Pa (p.278) 780, whereas Mo's version, the more common one, is found in Ch 47, Ei 121, Pa 1121, Lo 4951, Gr 807, Cai 61, Pa 1235, Pad 47, RoV 52, and Mod 7.

v. 2 “tuAE”: The Aquitanian MSS match Ben 34 here, but Mo 159 has DFDD(ori.)C, a reading matched in most peripheral MSS.

Roman MSS:

Vat 5319:

  • v. 1 “Legem”: These notes are mistakenly written a third too low but corrected with the custos that follows.

  • v. 1 “exQUIram”: abcdcbab(ori.)aG

  • v. 2 “domiNE”: This melisma is written a third too low in Vat 5319, but again corrected with the custos that follows.

  • v. 2 “LEX”: ac dcd

  • v. 2 “meditatiO”: two extra notes are appended to the end of the melisma, aG

RoS 22:

  • “doceAS”: cdcba bcdcbabG

  • “[iustitia] tuA”: aGbaGaG

  • “intellecTUM”: aGFEFGFGbaGF

  • “MANdata”: aGbaGF

In te speravi (21)

There are no significant variants among the core group of sources and very few outside the core group. Individual sources:

  • To 18 v. 1 “TUam”: CD FGFDC FGD FF FGF

  • Pa 1235 v. 1 “SERvum:” DF F FD EFDC CDFF aG (liq.)

The Roman sources also lack significant variants.

Meditabor (22)

The parenthetical B‐flats indicated in the opening and at the final melisma on “dilexi” reflect a point of variance in the MS tradition and were probably not universally sung. In many MSS, this offertory is written at the affinal position and the corresponding notes are F's, clearly suggesting the use of low B‐flat. These MSS include Ben 34, RoV 52, and Pad 47. In Mo 159, however, it is written in the normal plagal range, with the low B's left as is, and the same version is found in Cai 61, Pa 1235, Mod 7, Pst 120, and Be 40078.

The base version of Ben 35 contains one certain scribal error and other anomalies that are most likely errors. The penultimate line of the offertory, stretching from “[se]cundum” through “pedes [meos],” is written a third too high in Ben 35. The status of this passage as a scribal error is clearly indicated by the custos at the end of the line, between “pedes” and “meos,” which serves as a correction of the cleffing. Another possible scribal error concerns “meAS” in phrase 2. The bracketed notes G‐F are lacking in Ben 35 but occur in all other (p.279) MSS, including the Beneventan sources. In v. 2, Ben 35 is missing four notes over “conVERti” (phrase 6) that occur in all other sources. These are indicated with brackets. A fourth variant is the last two syllables on “dominus” in verse 1 (phrase 3). The two F's in this passage are G's in most other MSS.

Distinctive readings in Beneventan MSS: In v. 1, the text “pars mea dominus” is matched in Ben 34, but reads “pars mea domine” in other sources. V. 1 “[vultum] TUum”: Ben 35's reading is matched in Ben 34, but most MSS have EFGFGFD.

Significant melodic variant: In some Aquitanian MSS, the second verse closes with a very long melisma on “conVERti.” Among the core MSS, this melisma is found only in Pa 776 and Lo 4951.1 The version of Pa 776 is aG aGFa DFDF FFFD aaaF GGEFG FG baGa FDE FF FF EG (quil.)aG aGFa DFDF FFFD aaaG GGE FG FG baG aFDE GaFE GaFE EFG GF GG EFGG Gc c(ori.)G aa Gc c(ori.)G aa Gaccccc caGF EFDC DE(quil.) FEF EFD EFG G(ori.) F FaGaG aFG(liq.).

Lo 4951:

  • v. 1 “meO”: DFDCD

  • v. 2 “miseREre” is missing a segment of the melisma. The melisma reads FGGFGFEF DE(quil.)FE CD FG FGabG

Mod 7: A lacuna on the opening melisma of v. 1 (“pars”).

Roman MSS: There are very few variants.

  • Vat 5319: v. 2 “TU/UM”: CDEDED/CBDFEDFE

  • RoS 22: “TUa”: FGFEGFEFDE

Benedic anima mea (23)

There are no significant variants in the respond and first verse. The second verse exhibits several pitch‐level variants. Pa 776 was chosen as the base version because it presents the verse in its most probable unemended range.

In many sources, including Mo 159, Cai 61, Lo 4951, and Be 40078, the second verse begins a fourth above its position in Pa 776. Mo 159, Lo 4951, and Cai 61 join the level of Pa 776 at “omnium” (phrase 9), whereas the scribe of Be 40078 continues at the higher level for the whole verse.

An upward transposition of a fourth is often employed to represent the nondiatonic pitch F‐sharp, equivalent to b‐natural at the higher position. It is probable that this pitch was employed on the last syllable of “caelo” (phrase 9). At this point, Mo 159 uses the special semitone sign at the position of b. A b‐natural at this level is equivalent to F‐sharp at the position of Pa 776, a fourth lower. Although Cai 61 lacks a natural sign here, it does not indicate a b‐flat, as it does for the rest of the b's in this verse. Other emendations of this passage support the hypothesis that an F‐sharp is intended at “caeLO.” Ben 34 and Gr 807, for example, start the verse at the level of Pa 776. Beginning at “caelo,” however, both versions are written a whole tone below Pa 776, rejoining the level of Pa 776 at “omnium.” Both sources thus have the half step F‐E on “caelo,” corresponding to the whole tone G‐F in Pa 776. In Pst 120, the verse likewise starts at the level of Pa 776, but beginning on the final (p.280) syllable of “caelo,” it is written a fourth higher, so that the whole tone G‐F in Pa 776 is the semitone c‐b in Pst 120. The hypothetical F‐sharp on “caelo” occurs just before the temporary change of recitation pitch from c to G and temporarily creates a tritus quality for G.

The emended readings Mo 159, Cai 61, Lo 4951, and Pst 120 are written a fourth above Pa 776 until “omnium,” thus lacking the shift of range at “omnium” that is found in Pa 776, Pa 1121, and other MSS. This shift of range, however, is suggested in the adiastematic MSS La 239 and Ei 121, which have the letter s (sursum) here, confirming the impression that Pa 776 represents the “unemended” pitch level at “in caelo…omnium.”

A summary of pitch level in the sources:

  • = Pa 776 throughout: Pa 1121, Pa 780, Pa 1235.

  • Verse begins a fourth above Pa 776, joins at “omnium”: Mo 159, Cai 61.

  • Entire verse a fourth above Pa 776: Be 40078, Mod 7.

Whole tone below Pa 776 beginning at “caelo,” joining Pa 776 at “omnium”: Ben 34 (transposition begins on the last two notes of “CAELo”), Gr 807 (transposition begins at “CAElo”).

Fourth transposition beginning at “CAElo,” joining Pa 776 at “omnium”: Lo 4951 (last syllable of “caelo” is b [flat?] a), Pst 120.

Fourth transposition beginning at “paravit,” joining Pa 776 at “omnium”: RoV 52

In RoV 52 the last syllable of “lo” is G‐F. This source shows no discernable evidence of nondiatonic practice in the verse, but the passage between “paravit” and “eius” is placed in a more common segment of the background scale.

Individual MSS:

Pad 47 presents a shortened version of the second verse, lacking “ut faciant ea…regnum eius.” The verse closes with the words “omnibus dominabitur.” The lengthy melisma on “omnibus” shares some material with that on “omnium” in the majority version but is not identical to it. The same shortening of the verse is not found in the other MS from Ravenna, Mod 7.

Mod 7: Differs in a segment of the final melisma on v. 2: “omniUM”: fgfc df fgfd fa gaga ff gfd fefdc ededc dd cd ff fcedcedc ec cdff gafe.

Roman MSS: There are no significant variants.

Domine deus salutis (24)

The Roman tradition lacks one of the three verses that circulate in Gregorian MSS. Here the phrases are given parallel numbers to facilitate comparison between the two versions.

This offertory shows two points of significant variance among the core group of MSS. The melisma on “mane” in the second verse (phrase 9) is shortened in the Beneventan MSS and in Lo 4951, lacking several of the internal notes that appear in most other MSS. The same melisma, with the same variant between MSS, occurs in the Easter Week offertory Benedictus qui venit (54). The early adiastematic MSS SG 339, Ei 121, La 239, and Cha 47 are consistent with the longer version. Some (p.281) sample readings of the melisma are given below, and more may be found in the notes to Benedictus qui venit.

Ben 34:

cdcaGFG

FaGaG

babGF

FGaca FGaGa

Lo 4951:

cdccaGFG

FGaGaG

babGF

GacaFGaGa

Pa 776:

cdccaGFG

FGaGa accca ccdcaaF FGaGaG

babGF

Gaca

FGaGa

Gr 807:

cdbcaFG

FGaGaGa cc ca cc dc aaF FGaG aG

cacGF

GacaF

FGaGa

The other significant variant concerns the pitch level in the opening of the third verse. In some MSS, “factus sum sicut homo” is written a fifth below its position in Ben 34. The pitch level of Ben 34 is matched in most core sources, including Mo 159 and all Aquitanian MSS except Pa 1121. The lower level is found in most MSS from St. Martial and a few unrelated sources, such as Gr 807.1 Since the two versions are identical in interval structure, the reason for the lower notation in some MSS is unclear. Perhaps the downward transposition is made by analogy to the second verse, whose incipit is otherwise melodically identical. In one MS in the sampling, Mü 10086, the whole third verse is written a fifth below the level of Ben 34, a reading also found in Pa 1133.2 In Mü 10086, moreover, the second verse is also transposed down a fifth, with the low B's indicated as flat, for unknown reasons.3

Pitch level of “factus sum sicut homo”:

  • = Ben 34: Mo 159, Pa 776, Pa 780, Lo 4951, Pa 1235, Be 40078, Cai 61, Pad 47, Pst 120, Mod 7, RoV 52, To 18

  • 5↓ Ben 34: Pa 1121, Pa 1133, Pa 1136, Pa 1137, Gr 807, Mü 10086

With the exceptions of Pa 1133 and Mü 10086, all MSS join Ben 34 at “sine.”

Transpositions in individual MSS:

Mü 10086: The second verse is written a fifth lower than the majority reading

Pa 1235 presents a unique version of the opening of the first verse, “inclina aurem tuam ad precem meam domine,” which is mostly a fourth below the majority reading, with the cadence on “domine” falling on D rather than G.

Other melodic variants: v. 2 “mea” (phrase 10): The Beneventan MSS transmit distinctive readings of this verse ending, with a more elaborate cadential pattern. The other sources in the core group have Ga/Ga.

Aside from the melisma on “mane,” most variants among core MSS involve differences in recitational style. The figure G‐b‐d‐c that occurs often in Ben 34, for example, is consistently G‐c‐d‐c in Pa 776 and in most other Aquitanian MSS. There are a few significant variants in peripheral MSS, as follows:

  • v. 2 “preVEniet” in RoV 52: GabGa ced

  • v. 3 “mortuOS”:

    (p.282)
    • Pa 1235: dcdbca aGa

    • Cai 61: dcda caaGa

    • Mod 7: cabaGa aG

  • v. 3 “[traditus] sum” in Be 40078: GaGF

  • v. 3 “egredieBAR”: the final melisma exhibits variants among several MSS, which are highlighted in boldface.

Mod 7: cd cc ca ccc dc ccc ca cac cc cacGF GG Fac GbaG acca cccaG abcb c ca ccca F abGFG FGaGa FGaGa GacccaG ccadcb bd eca cbGa

Cai 61: bd cc ca cc cd ccc ca cac c ca caF aa Ga ccc dcb ac cb cc aGa bcb c cb ccaF aca Ga (etc.)

RoV 52: cd cc ca ccc dc cc ca cac c ca caF F aa Ga cc cdcb (etc.)

Pst 120: cdc cc dc ccc dc cc ca cac cc cac GF FG Gac c cdcb ac ca ccaG Gabcb c ca cdca GbaGa FGaGa FGaGa (etc.)

To 18 is missing the final melisma.

Individual MSS: RoV 52 indicates a verse division within v. 3, at “traditus.”

Roman MSS:

Vat 5319:

  • respond: ME/E: The two “e” syllables were perhaps not distinguished in performance, but the neumes indicate the following: Gabcba/aGabaGaG

  • v. 1 diE: aG cdc cdcbdcbacba

  • v. 2 tradiTUS: a (ori.) GF

  • First verse lacks repetendum cue, second verse has a repetendum cue to “in.”

RoS 22: “ME/AE”: Gabcba/aGabaGaG

Benedicam dominum (25)

The pitch level variants in the second verse of this offertory have been demonstrated by Pitman and Hankeln. The pitch‐level profile of Ben 34, the base version, is matched in Mo 159 and in unrelated MSS such as Be 40078. In other MSS, however, the opening passage of v. 2, though “vias vitae” (phrase 7), is written either a whole tone higher or a fourth lower. Ben 34 has a rare notated b‐flat in the verse opening, and Mo 159 indicates b‐flats throughout the passage. The transpositions at levels a whole tone higher and a fourth lower, then, do not produce intervallic differences that would point to nondiatonic notes as the reason for the variants.

Some Aquitanian MSS, including Pa 776 and Pa 1121, have a written level for the second verse a fourth below Ben 34. In Pa 776, however, the repetendum cue suggests a performance level a fourth above the written level, at the position of Ben 34. Hankeln has accordingly posited the level of Mo (and Ben 34) as the Aufführungslage in Pa 776.1 The variety among the other readings, summarized below, reinforces the impression that they are emendations, or in some cases reworkings of the melody, and that the level of Ben 34 is (p.283) the preferred reading.2 One possible reason for the transpositions here is the G‐protus quality of the opening passage, particularly at “fecisti mihi.” The passage is notated a tone higher in the Italian MSS examined, consistent with the Guidonian recommendation for emending G‐protus passages. In many of the MSS that begin the verse a fourth below Ben 34, the melisma on the second syllable of “vias” is reworked to create a transition between the two transpositional levels, joining the level of Ben 34 within the melisma or at “vitae.”

Pitch level at “notas…vitae”:

= Ben 34: Mo 159, Pa 776 (according to repetendum), Pa 1121 (?),3 Pa 780, Cai 61, Be 40078, Gr 807, Mü 10086

2↑ Ben 34:

  • Pad 47 (joins Ben 34 during the melisma on “viAS”: c bcb aba GGF FGaGaca)

  • Pst 120 (joins Ben 34 during the melisma on “viAS”: c bcb aba GGE GaG aca a c [liq.])

  • Mod 7: same as Pad 47

  • RoV 52 (joins Ben 34 temporarily during the melisma on viAS, then again at “ad implebis”)

  • Frutolf tonary

4↓ Ben 34:

  • Lo 4951 is 4↓ Ben 34 until “VI/AS,” with a highly reworked version of the melisma that reads: DE(quil.)FGF/FGacaGaGFGFD FGaGaca aaa

  • Pa 1235 is 4↓ Ben 34 until the end of the melisma on “viAS,” where the interval of transposition changes to a third (DFD) and then a second (EGE): The melisma reads: CD GFF/EFEF DFD CCA C DFD EGE(liq.).

  • Most sources from St. Martial appear to notate the verse a fourth below the level of Ben 34, but lack repetendum cues to confirm this level.4

Melodic variants occur at the following points:

respond, “QUO/NI/AM”: Mod 7: DG/G abca /acb c FG

respond, “[est] miCHI”: Pa 776: ED

v. 1 “DOminus”:

  • Mo 159: melisma begins Fac

  • Pa 776: melisma begins CDa. Same reading in Be 40078

  • Pa 1121: melisma begins DFa

  • Gr 807: melisma begins Da

  • Mod 7: melisma begins FGa

v. 1 “meae” (verse ending). The Beneventan text underlay differs from that of other MSS. The majority reading places the last syllable on the last four notes.

(p.284) v. 2 “NOtas”: Ben 34's reading F‐c is anomalous. Mo 159, most Aquitanian MSS, and German MSS have a whole tone (B‐flat‐c, c‐d, or F‐G) here. Italian MSS are a mixture: Mod 7 and Pad 47 have G‐d.

v. 2 “viAS”: besides the emended versions described above, there are the following variants:

  • Pa 776: baba GaGF FG(quil.)aGacaa (no descent to D)

  • Be 40078 (b's are flat): b b ba GbG F FD FG(quil.)aG acG\a

  • Gr 807 (b's are flat): b(ori.)a GbF EFD FGaG aca

v. 2 “fiNEM”:

  • In the first two segments of the melisma, Mo 159 adopts c rather than d as the top note: a GbcbGa GbcbGa. Ben 34's version is matched in the Aquitanian MSS, but the version with c as the top note is found elsewhere:

  • Be 40078: melisma begins a Ga(quil.)c CGa Ga(quil.)c CGa

  • Gr 807: a GacbGa GacbGa

  • Mod 7: a GabcccbGa GabcccbGa (same in Pad 47 and Pist 120)

  • Gr 807 also differs in some other segments of the melisma following the Gacc segment: Gacc ca ca cG aa

Roman MSS:

Vat 5319:

  • v. 1 “E/GO”: aGacba/Ga(ori.)GFG (written a whole tone higher than in Bodmer 74)

RoS 22:

  • “SEMper”: GcaGaGaFG

  • “quoniAM”: acaGF GbaGaG

Miserere mihi (26)

This offertory shows only minor variants between sources, mainly around the pitches b and c. The melodic figure Gbdc that occurs several times in Ben 24 (see v. 1 “delictum” and “meum”) is realized in other sources from the core group, like Pa 776 and Pa 1121, as acdc.

  • v. 1 “me”: Ben 34 has a single a here. Most other sources have a aGa.

Individual MSS:

  • Pa 1235: Respond, “miseriCORdiam”: Pa 1235 has a more ornate alternative version, making it identical to the verbally parallel passage in Domine fac mecum: cccbacca(liq.)

  • Be 40078: Respond, “domine”: Gc (ori.) cba cG

  • Gr 807:

    • v. 2, “FEci”: f (ori.)eddefef

    • v. 2 “TUis”: bc deb cd cd eb cd cd ecbaGa

Roman version: Bod 74 has a shortened version of the first verse: “Tibi soli peccavi et malum coram te feci.”

(p.285) Domine in auxilium (27)

The Gregorian Domine in auxilium exhibits few significant melodic variants among the core group of sources. The Beneventan reading employed as the base version is representative of the group as a whole. Two issues, however, require discussion. The first is structural. Some sources close with a cue to repeat the opening of the offertory, “domine in auxilium meum respice,” which also serves as the repetendum cue in sources that indicate repetenda. Other MSS lack this cue, simply closing with “eam.” These differences are consistent with the remarks of John of Afflighem: “Many end this offertory badly by avoiding the heptaphone that is at the end because it seems to them ill‐sounding. Therefore in certain books they repeat at the end what occurs in its beginning.”1 The presence or absence of this reprise affects the offertory's modal assignment, an issue Bomm addresses.2 The internal phrases of the respond and first verse are characteristic of mode 6, sharing material with other sixth‐mode offertories. The sources with the reprise end unproblematically on F (or the affinal c) at “respice,” a practice also reflected in the Roman version. The sources without the refrain, however, close variously on C or F (sometimes G if the offertory is written at the affinal position). Although this chant is lacking in most tonaries, it is classified as a tritus chant by Bern of Reichenau, Frutolf, Ugolino, and the Tonary of Munich, Bayerische Staatsbibliothek clm. 9921, an eleventh‐century MS from Ottobeuren.3

Most sources in the core group lack a cue to the beginning of the offertory and close the respond on low C, as shown in Ben 34. Ben 38, however, does have a cue to return to the opening. It is doubtful that performances with verses actually ended on low C, which is not a legitimate final. Mo 159, for example, appears to close the respond on low C and lacks repetendum cues, but nonetheless classifies this piece among tritus plagal chants, perhaps implying that a repetendum was performed after the second verse. Most MSS with repetendum cues indicate “domine in auxilium,” the beginning of the offertory, as the repetendum. We may speculate that the repetendum consisted not of the whole respond but of “domine in auxilium meum respice,” thus coming to a conclusion on F. Ben 34 is unusual in indicating “confundantur” as the repetendum cue for the first verse and lacking a cue after the second verse.

To summarize the ending of the respond in the sources, a version matching the low C ending of Ben 34 is found in Mo 159, Pa, 780, Pa 776, Pa 1121, and Pa 1132. These sources lack a repeat of the opening at the end of the respond, but with the exception of Pa 776, they do have a repetendum cue at the end of the second verse (as do the other sources from St. Martial), suggesting that perhaps a performance with verses came to a conclusion on F, at “respice.” The respond closes on low C, without repetendum cues, in Be 40078. In Pa 1235, which is written at affinal position, it closes on G. In many Italian sources, including (p.286) Mod 7, Pad 47, RoV 52, To 18, Pst 120, and Pia 65, the respond concludes with a cue to return to the beginning and repeat the first phrase of the respond. With the cue observed in performance, the respond would conclude unproblematically on F. Finally, a few sources in the sampling emend the melisma on “meam” to close on F, D, or, at the affinal position, a, as in the following sources:

  • Tri 2254: CFDF

  • Lo 4951: FFGG(ori.)F

  • Frutolf tonary: CFCF

  • Mü 10086: DFD

A second variant point is the relative pitch level between the respond and second verse. Domine in auxilium presents a pitch‐level problem similar to most other sixth‐mode offertories. In the Beneventan version, the verse is an unproblematic tritus authentic melody in relation to the plagal respond. In some versions, however, the second verse is written a fifth below the level of Ben 34. In the second verse, the typical characteristics of mode 5 are projected a fifth below the final. Most of the melody remains within the fifth B (flat)‐F, and the cadence on “respexit me” is on low B (flat), a fifth below the final.

This latter version is the majority reading among Aquitanian sources and is also found (at the affinal position) in many Italian MSS. It is clear that the low B‐flats and probable E‐flats in this version of the final verse would have necessitated an upward transposition of the final verse. These problem spots, however, do not fully explain the differences in pitch‐level profile; this version could be transmitted in staff notation by simply writing the whole offertory, respond and both verses, at the affinal position, a solution adopted in many sources. Given the theoretical awareness of the Aquitanian scribes, it is probable that the affinal position was their intention as well. In the version of Ben 34, however, the respond and second verse appear in a relationship that is more consistent with contemporaneous understandings of mode, since the second verse appears as an unproblematic tritus authentic melody in relation to the plagal respond.

To some extent, the pitch‐level profile of the sources suggests a regional preference for one version or the other. In a few cases, however, variants in pitch‐level profile can be found among sources that are otherwise closely related. The repetendum cue in Pa 1137, for example, indicates a pitch‐level profile matching that of Ben 34, whereas the other sources from St. Martial match Pa 1121. These differences suggest that the pitch‐level profile was, to some extent, the purview of individual scribes.

A summary of sources follows: Pitch‐level profile of Ben 34: Mo 159, Pa 1137, Lo 4951, Pst 120, To 18, Pia 65, Tri 2254, Mü 10086, Be 40078, Gr 807. Final verse written a fifth below the level of Ben 34: Pa 776, Pa 780, Pa 1134, 1135, 1136, and 1132. At the affinal position: Pad 47, Mod 7, RoV 52, Madrid 20‐4. Pa 1235: Respond is at the affinal position; verse 1 is in the untransposed range.

Roman MSS:

Bod 74 and Vat 5319 have the verses in the following order: (1). Expectans; (2) Avertantur. There are no significant melodic variants.

(p.287) Illumina oculos (28)

The Beneventan MSS are unusual in one respect: “dormiam” in the respond (phrase 1) reads “obdormiam” in most other MSS (with only Compiègne matching Beneventan MSS). The melodic reading typically matches that of Mo 159: “OB/DORmiam”: E/GEGaG.

Other melodic variants:

  • respond, “eAM” Some MSS dip down to D:

  • Mo 159: FFF GaGD FFF aGF GF (quil.)E (same reading in Lo 4951 and Pa 776).

  • Pa 1121 and Pa 780 have the reading of Ben 34: FFF GaGF FFF aGF GFE.

  • Ben 34's reading is also found in the peripheral MSS RoV 52 and Be 40078.

  • Mo 159's reading is found in Cai 61, Pa 1235, Mod 7, Pad 47, and Pst 120.

  • respond, “IN [morte]”: Pa 776 and Pa 1121 have Da(liq.).

  • v. 1 “domiNE”: Pa 1121 and Lo 4951 are essentially identical to Ben 34, but Pa 776, Pa 780, and Mo 159 have a different reading:

  • Mo 159: cc cb(quil)a cc dc cdaa(ori.)G

  • Pa 776: b cccba ccd ccd aaG

  • Pa 780: cccba ccd ccd aaG

Ben 34's version is matched in Cai 61. Mo 159's version is matched in the following peripheral MSS: Pa 1235, RoV 52, Mod 7, Pad 47, Pst 120, and Be 40078.

Variants in individual MSS:

  • Pa 776: v. 2 “boNA”: bccaG aaG ccaaG ccGbaG (a truncated version of the melisma).

  • Pa 1121: v. 1 “CONsilium”: EG.

  • Pa 780: v. 2 “exAUdi”: cdc c(ori.)a cca ccaGa (longer version of the melisma).

  • Lo 4951: v. 1 “aniMA”: abaGa.

  • Cai 61:

    • respond, “MEos” (all b's are flat): aG ba bGF FGa baG ba

    • v. 2 “cantabo”: b-flats are indicated here, whereas they are not in other sources.

  • Be 40078: v. 1 “animam meam” (sic): indicates b-naturals rather than the b-flats found in most MSS that distinguish between b-natural and b-flat.

  • Gr 807:

  • respond, “morTE”: cccGF GF GaGF GFE

  • respond, “ADversus”: DFDF

Roman MSS: There are no significant variants. In phrase 6, the Roman reading concludes with a cadential pattern associated with the protus and tetrardus maneria, but written a tone lower. Vat 5319 employs a notated b-flat here. A similar cadence occurs in Immittet (69).

(p.288) Iustitiae domini (29)

This offertory exhibits significant variants at two points: the end of the respond—hence the modal assignment—and the relative pitch level between the respond and second verse.1

The deuterus ending of the respond in the base version of Pa 776 is the majority reading and is implied by the neumes in the early adiastematic MSS examined. In some MSS, however, the respond is emended to close on F. The alternative ending may be attributable to certain melodic characteristics this piece shares with the sixth‐mode offertories (discussed in chapter 3). The following MSS have alternative endings on F:

  • Mo 159: FFF(ori.) C FFF aGF GFE/F

  • Lo 4951: E FFF D FFF aGF aGFGF/F

  • Ben 34: FFF DFFF aGF aGFGF/F(liq.) (same in Ben 35)

In this respect, Ben 34 and Ben 35 differ from the earlier Beneventan MSS, Ben 38 and 40, which have the ending on E.

Another variant concerns the relative pitch level of the respond and second verse. The reading of Pa 776 is the preferred version and the majority version among Aquitanian MSS. When the whole offertory is transcribed in the normal plagal range, with the final on E, several problems emerge in the second verse. This verse has the traits of the fifth mode, but projected a fifth below the normal fifth‐mode range. At “conspectu tuo” (phrase 8) a caesura occurs on low B‐flat, which is now heard as a temporary final, resulting in the need for a nondiatonic pitch, E‐flat, in the final melisma. In pitch‐specific MSS, these problems are addressed in two different ways. The group of MSS designated group 1 below preserves the pitch‐level profile of Pa 776. In pitch‐specific MSS, the entire offertory is simply transposed to the affinal position to close on b or c. This solution solves the notational problem of E‐flats in the final melisma (allowing this note to be represented by b‐flat) but preserves the modal irregularity of the second verse. Sources in group 2 indicate a shift in range between the two verses, creating a version more consistent with traditional conceptions of mode. In most cases, the respond and first verse are notated in the normal plagal range, closing on E or F, and the second verse is in the authentic range. A few MSS replicate this pitch‐level profile at the affinal position.

The indications of E‐flat and E‐natural in the transcription of Pa 776 derive from the indications in Mo 159, where the verse is written a fifth higher and alternates between b‐flat and b‐natural.

Group 1, pitch‐level profile of Pa 776: Pa 780, Pa 1121, Lo 4951, Ben 38, Pad 47 (affinal position), RoV 52 (affinal position), Mod 7 (affinal position)

Group 2, second verse 5↑ Pa 776: Ben 34 (at affinal position), Ben 35, Cai 61, Pa 1235, Pst 120, Gr 807, Be 40078

Melodic variants: The MSS differ in their indication of text underlay at “custodiet ea/eam.” Most MSS in the sampling follow the underlay indicated in Pa 776, but the following MSS vary:

Mo 159: “ET/E [am]”: F/FF F(ori.) C (etc.)

Ben 34: F/FFF (etc.)

v.2: Some MSS more consistently adopt low B‐flat (rather than C) as the bottom note (or, at the higher level of transposition, F rather than G). Mo 159, for example, has F's on “et [erunt]” and “ut [complaceant],” which would be low B‐flats at the level of Pa 776. Pa (p.289) 1121 adopts B (flat) as the bottom note at the end of the melisma on “complaceant,” corresponding to the final c of the melisma in Pa 776.

v. 2 “comPLAceant”:

  • Lo 4951: DFGFGFGFGa FFFDC FFG FG FGFGaFFFDC FDF FDC FD F(ORI.) DC FDB FDE(QUIL.)FG

  • Pa 1235: melisma ends cdaG caG aG

  • Cai 61: melisma ends daG caG caF cacd

v. 2 “SEMper”:

  • Pa 1121: FGFD ECB DF aGF FFF FabG F F(ori.) D FFD GF DD(ori.)B BDF (etc.)

  • Lo 4951: Ending of the melisma is FFF FG FG(quil.)a G FGF

  • Pa 1235: cdca baG ac edc ccc ce fdc caccc cdc ca ccadc aF Fac (etc.).

The ending of v. 2 (“semPER”) is a point of variance in a few MSS (though Pa 776 reflects the reading in most peripheral MSS examined):

  • Mo 159: cc cba ca FG

  • Pa 1121: FFFED EC BB(ori.) C

  • Lo 4951: F (following the altered version of the melisma on the first syllable)

  • Ben 34: cccbacG FG

  • RoV 52: cccbaca

Roman MSS:

  • Vat 5319:

  • respond, “RECte”: FGaGaGa(ori.) GF

  • respond, “cusTOdiet”: aGF FGaGbaG

  • v. 1 “preCEPtum”: baGaGFG FGF acGaGFGF EFE DFEDC DFFF GaGaGFGE FGaGaGF

  • v. 1 “VEra”: Vat 5319 has the longer version of the melisma's ending: FacGaG FGF EFEDFEDE(ori.)DC DFF GaGaGFGE GaGaGF

  • RoS 22:

  • “IUstitite”: aG

  • “dulciOra”: aFGaGaGF

  • “serVUS”: GEFEDC

Exaudi deus (30)

This offertory has few significant variants between sources.

1. Respond, “MEam”: In several sources outside the core group, notes 6, 7, and 8 of the melisma are Fga (rather than Fac): These include Gr 807 and the Italian sources.

Individual MSS:

Pa 1235: respond “exAUdi” [me]: ccbG aG acbG aba (liq.); v. 2. “retribuENdo”: d dc deda deda dedcba c cd ff gfgf fgf gag (liq.)

To 18: v. 2, “retribuENdo” (major variant): dedc dc cc ca bcbaG bcba cc dcca Ga cc dcdc de cc cad

(p.290) Roman sources: The only significant variant is at the end of the respond, “exaudi me.” Bod 74 has what appear to be two alternative endings. The second is nearly identical to that of Vat 5319, closing on a. The first is more ornate:

Ex‐ au‐ di‐ me

Gc dc c cbcbabaG acbaba G GF acbabaG (liq.) GacbabaGa GF

The ending of RoS 22 is similar to that of 5319 and the second ending of Bod 74, closing on a.

Domine fac mecum (31)

This fourth‐mode offertory is notated at the affinal position in most sources, undoubtedly because the low F in the respond would become B‐flat at the lower level of transposition, a note unavailable in most notational systems. Only three sources included in the sampling, Pa 903, Gr 807 (which has a sign for B‐flat), and Mo 159 notate the respond (along with the verses Deus laudem and Pro eo) in the untransposed range. In Pa 903 the passage is emended so that the B‐flat is a C, and in Mo 159, the note is written as a B‐natural.

Domine fac mecum exhibits a particular prominence of small variants around the pitches b and c. The Beneventan sources are the most consistently focused on b. Some change of tonal focus to c is evident already in the earliest pitch‐readable sources, Mo 159 and Pa 1121. On the final syllable of “peccatoris” in verse 1, for example, the Beneventan sources have babca, whereas Pa 1121 has cabcb. The change of focus to c seems to distinguish the sources chronologically. A marked emphasis on c is evident not only in late German sources such as Be 40078 and Tri 2254 but also in the thirteenth‐century Italian MS To 18. In most passages, these sources avoid b altogether, except at the final cadence; passages such as “deus laudem” in verse 1, for example, consist entirely of repetition of c rather than alternation between b and c.

Locuti sunt, the second verse in a majority of sources, contrasts modally and melodically with the rest of the offertory. The Italian sources included in the sampling present a distinctive reading of this verse, both in pitch level and melodic detail. These sources notate the beginning of the verse a fifth higher than it appears in Ben 34 and the other sources, joining at “circumdederunt me.” The Italian MSS, moreover, exhibit several other departures from the majority version, as shown in example 31.1. A very similar version of the verse may be found without pitched notation in RoA 123, and in Mod 7, Pad 47, Pst 120, and Pia 65. To 18 exhibits some, but not all, of these characteristics.

Appendix 3 Commentary on Edition

Example 31.1

(p.291) Roman sources: In Bod 74, the offertory is notated in the normal plagal range, ending on E. There are no major melodic variants.

Intende voci (32)

The parenthetical b‐flats at “laetentur” (phrase 8) are derived from their use in Mo 159, but are also indicated in Tri 2254.

In its hypothetical unemended version, written at the pitch level of Pa 1121, this offertory employs one nondiatonic pitch, at “viam meam” (phrase 7).1 In Mo 159 and most non‐Aquitanian MSS, “meam” is written a whole tone higher. Ben 34 is unique among the sources examined in notating the passage a fourth lower. A second pitch‐level variant occurs at “gloriabuntur” (phrase 9), described below.

Pitch level at “[viam] meam”:

  • = Pa 1121: Pa 776, Pa 780, Pa 1136, Lo 4951, Be 40078 (note corresponding to Pa 1121's E [flat] is F), Gr 807 (note corresponding to Pa 1121's E [flat] is F)

  • 2↑ Pa Pa 1121: Mo 159, Ben 35, Cai 61, Pa 1235, Pad 47, Pst 120, RoV 52, To 18, Tri 2254, Frutolf tonary

  • 4↓ Pa 1121: Ben 34

Pitch level at “gloriabuntur”: Many MSS notate this passage a whole tone above the level of Pa 1121. Mo 159 is a tone higher beginning on the second syllable of “aeternum,” whereas most other MSS are a tone higher beginning on the first syllable of “gloriabuntur.” All MSS join in the melisma on “gloriabuntur” or at “qui diligunt,” except for Pa 776. In Pa 776, the rest of the verse remains a tone above Pa 1121, including the repetendum to “quoniam,” which is notated a tone too high.

Although the reasons for the variants are not clear, the variety of versions suggests that an emendation has taken place. The passage is written at the level of Pa 1121 in several later MSS that distinguish between b natural and b‐flat. A b‐flat is indicated in Tri 2254, Cai 61, and Ben 34 (one of the few notated b‐flats in this MS). With a b‐flat, Pa 1121 has the same intervallic structure as the majority version a tone higher. Either the reading of Pa 776 is a scribal error or it represents a pretheoretical tradition in which the tonal focus of the verse migrated from c to d (with the e's in the majority reading sung as f‐sharp's) and the repetendum was also sung a tone higher. I have shown evidence for similar “modulations” in other offertories, such as Oravi deum and Ave Maria. In Intende, however, the evidence for this practice is limited to one MS included in the sampling and may well reflect an error. For this reason, I have opted for Pa 1121 as the base version.

Pitch level of “gloriabuntur”:

  • = Pa 1121: Ben 34, Ben 39, Pa 780, Lo 4951, Tri 2254, Cai 61

  • 2↑ Pa 1121: Pa 776, Pa 1136, Pa 1137, Pa 1132, Mo 159, Pst 120, Mod 7, Pad 47, To 18, RoV 52, Gr 807, Be 40078

On “gloriabuntur”:

Ben 34: One of the very few places where this MS employs a notated b‐flat: G/G/G/b‐flat G bbcba bcdc (liq.)/ed

(p.292) Pa 1121: G/G/G/bGbbcba bc(quil.)d/dc (very similar readings in Pa 780, Lo 4951, Cai 61, Trier 2254).

Melodic variants: Although there are very few major variants, small variants are found at several points around the pitches G and F. Melodic figures that revolve around the pitches c‐a‐G‐a occasionally appear as c‐a‐F‐a. These points of variance do not appear to be correlated with regional tradition or groups of MSS. One example is in the final melisma of the respond, on “DOmine.” Pa 776 is anomalous among core sources in having F as the lowest note. In most MSS G is the lowest note:

Pa 1121: FG(quil.)aGa ccc dcdc c(ori.) G acbac (same reading in Lo 4951)

Mo 159: FG(quil.)aGa ccc dc dcc(ori.)G Gcbac (same reading in Ben 34)

Pa 776's reading, however, is matched in some unrelated MSS such as Pad 47.

Another example occurs on “oRAbo” in the respond, where Ben 34 and Ben 35 start with Fac instead of Gac.

  • v. 2 “in conspeCTU”: Ben 34: GFa

  • v. 2 “aeTER/NUM”: Mo 159: cd/ded ede

Roman MSS:

Respond, “[orationis] meae”: The neume groupings seem to indicate a different text underlay in Bod 74, Vat 5319, and RoS 22, perhaps because the two “e” sounds were not distinguished in performance.

Vat 5319:

  • v. 1: “VER/BA”: deded(liq.)/dedc

RoS 22:

  • “inTENde”: babcbcba

  • “ME/E”: Gabcba/aGabaGaG

Gressus meos (33)

In both the Gregorian and Roman traditions, Gressus meos has an unusual structure in which a complete repetendum is written out after the second verse, with a long added melisma. As Kenneth Levy has shown, the melisma of the Gregorian version has several other functions in the chant repertory, including a use as the basis for an Italian sequence.1

Gressus meos exhibits several points of variance among the sources. The Beneventan reading presented as the base version is consistent with Mo 159 and Aquitanian MSS, except at the closing melisma, as described below. The major melodic and pitch‐level variants are as follows:

  1. 1. The opening pitches. The respond starts variously with D‐ED (Pa 776), D‐DF (RoV 52), E‐EG (Pia 65), EGF (Pa 780), or F‐FG (Mo 159, Beneventan MSS, Lo 4951, Pa 1121); and even G‐Ga (To 18).2

  2. (p.293)
    Appendix 3 Commentary on Edition

    Example 33.1

  3. 2. Respond, “secundum” melisma: Most Aquitanian sources present a reading similar to Ben 34, differing only in the number of repeated notes and the use of the oriscus. Mo 159, however, adopts a more literal repetition of the opening notes: cccaGG (ori.) FGacccaG G (ori.) FG (quil.) ab‐flat aG (liq.). A similar version is found in two of the Italian sources, To 18 and RoV 52.

  4. 3. Respond and repetendum, “ut non dominetur omnis”: The Italian sources have “ut non dominetur mei omnis.” The extra syllables are accommodated with notes from the melisma on “dominetur.”

  5. 4. Verse 1, “dat/dans”: The Italian sources present a shortened form of this melisma, omitting its last two notes.

  6. 5. Second verse, “cognovi domine”

In pitch level, Ben 34 represents the majority reading, found in Mo 159, Pa 776, several other Aquitanian sources, and the German MSS. The pitch level variants among a few of the Aquitanian sources in this passage are discussed by Hankeln, who notes that Pa 1133 begins a fourth below Mo 159 and several other sources.3 The passage is also variant among Italian sources. The scribe of Pad 47 begins “cognovi” a fourth above its usual position, joining with the majority version on “domine.” A very similar reading is found in two of the other Italian sources, Mod 7 and Pia 65. Pst 120 and To 18 begin like these MSS, a fourth above Ben 34, but subsequently place the melisma on “domine” a whole tone above its usual position. The reasons for the upward transposition of a fourth are unclear.

Two sources from German‐speaking areas have a slightly longer melisma on the final syllable of “domine,” as in Tri 2254 (with the b‐flat applying to the whole passage): Ga(quil.)b‐flat GbbabaGbG.

6. Second verse, “iustitia” melisma (ex. 33.1): The Beneventan version of this melisma, found in Ben 34, 35, and 38, differs from the others in its repetition scheme. The version of Mo 159, shown in example 33.1, is representative of a broad international tradition.

Notes on specific sources: v. 1,“dans” In Lo 4951 this melisma is lengthened to create a standard cadential pattern, with the following notes: Gccbab (quil.)cba.

Roman sources: The version of Vat 5319 is nearly identical to that of Bod 74 except for the “iustitia” melisma, where Vat 5319 omits a few passing notes toward the end of the melisma. The last seventeen notes of the melisma: FGaGFGFbacbabcbab. RoS 22 exhibits the following small differences from Bod 74:

  • “MEos”: Ga

  • “seCUNdum”: ccbabab

  • “oMNIS”: FEFGFEDFG

  • “IUstitias” (sic): aG

  • “IustitiAS” (sic): GaGacbcbabcbab

(p.294) Laudate dominum (34)

In most MSS, this offertory is notated at the affinal position, undoubtedly because of the low B-flat required at “populum” (phrase 10).

Aside from small differences in recitational style, there are very few variants among core and peripheral MSS.

The Beneventan MSS have the following very small differences from most other MSS:

  • respond: “EST,” and v. 2: “benediciTE”: Most MSS lack the second D in the Beneventan version, as in the following:

  • Mo 159 (at affinal position): Ga(quil.)cb(quil.)Da ca

  • Pa 1121: DE(quil.)FDCED

  • v. 1 “staTIS”: Most MSS have GFFD FGF

  • v. 2 “SE/CU/LA”: Most MSS have F/D/FD

  • Lo 4951: v. 3, “ierusaLEM” (final melisma): bottom note is B rather than A

  • RoV 52 (at affinal position): v. 2 “doMI/NUS”: cbGa/aG

Roman MSS:

  • Vat 5319: v. 2 “hierusaLEM”: This final melisma is transposed up a fifth, perhaps because of the need for low B-flat. There are no significant variants in RoS 22.

Exspectans expectavi (35)

As Pitman persuasively argued, this offertory has a variant point related to the use of the nondiatonic pitch high f–sharp in v. 2.1 In many sources, including Mo 159, the passage beginning “benenuntiavi” (phrase 8) is written a tone below Pa 776, remaining there until “magna” at the end of the verse; Mo 159 indicates a b–flat. The whole tone d–e in Mo 159, which occurs on “tuam” and in the melisma on “ecclesia,” is equivalent to e–f in Pa 776. To duplicate Mo 159's intervallic structure at the level of Pa 776, an f–sharp is required. The nondiatonic pitch is also suggested by Be 40078, where this verse is notated a fourth above Pa 776 and employs both b–flat and b–natural, equivalent to f and f–sharp at the level of Pa 776. Other sources, including Ben 34, Ben 35, Pa 1235 and most Italian MSS in the sampling, write the passage at the same level as Pa 776, evidently eliminating the nondiatonic pitch and changing the intervallic structure. Pa 776 was chosen as the base version because it places this passage in its “unemended” range.

Summary of pitch level of v 1:

  1. 1. At the level of Pa 776 throughout: Pa 1121, Ben 34, Ben 35, Pa 1235, RoV 52, Pst 120, Mod 7

  2. 2. The entire verse written a fourth above Pa 776: Be 40078, Cai 61

  3. 3. A whole tone below Pa 776 at “benenunciavi,” joining again at “magna”: Mo 159

  4. 4. A whole tone below Pa 776 beginning at “tuam” through the end of the verse: To 18

  5. 5. Individual solutions:

    (p.295)
    1. a. Tri 2254 is a tone below Pa 776 beginning on the third syllable of “iusticiam” and joins Pa 776 at “magna.”

    2. b. Lo 4951 is a fourth below Pa 776 beginning at “iusticiam” and a fifth below beginning on the last syllable of “iusticiam,” thereby representing the f–sharp with a b–natural.

    3. c. The Frutolf tonary indicates that “benenuntiavi” should begin on F, a fifth below the level of Pa 776, also allowing the f–sharp to be represented with b–natural.

Other significant variants:

v. 2 “ecCLESia” melisma: In addition to the whole–tone transposition, several sources in the core group differ from Pa 776 in one segment of this melisma:

Pa 776:…ggged egef(#)dc cdedecb…

Mo 159 (all b's are flat):…fffdc dedecbcdc cba bca…A reading similar to Mo 159 is found in Beneventan MSS and in many Italian and German sources.

Lo 4951 has a slightly shortened melisma (written a fifth below Pa 776): accaG cccaG abab GFGa cde c.

v. 3 “iusticiam tuam” (or “meam”) “non abscondi” (phrase 9): This passage is a point of variance, possibly because of the awkwardness created by the leap of a seventh, D–c, in Pa 776. There are variants among core and peripheral MSS as follows. All MSS except Pa 1121 join Pa 776 at some point in the passage that immediately follows, on “in corde.”

me/tu

am

non

ab‐

scon

Pa 776:

GaGE

EGEED

cd

c

cd(liq.) etc.

Mo 159:

GaGE

EGFFE

bc

b

bc(liq.)

Pa 1121:

FGFD

DFDDC

cd

d

cd2

Pa 1137:

abaF

GaGGF

cd

c

cd

Pa 1136:

GaGE

EGEED

bc

b

bc

Lo 4951:

GaGE

EGFFE

bc

b

bc

Ben 34:

cdca

acaaG

cd

c(liq.)

cd (liq.)3 (also in Ben 35)

T 2254:

FGFFD

DFDDC

c

c

c4 (also in Cai 61, Ber 40078, Frutolf)

Pa 1235:

ab–flat aG

GaGGF

cc

c

bc

RoV 52:

GaGF

FGFE

cc

c

cd (liq.)

Pst 120:

GaGE

FGEFD

Gc

c

bc(liq.)

Pad 47:

GaGE

FEFD

cd

c

cd

Mod 7:

GaGE

FGFE

cd

c

c

(p.296) The greatest point of variance between the MSS is the final syllable of “tuam/meam,” which ends variously on E, D, F, G, and C. While the intervallic differences between versions such as Mo 159 and Pa 776 on the last syllable of “meam/tuam” may suggest a nondiatonic pitch, other readings, such as Pa 1121 (matched in Trier 2254) are intervallically identical to Pa 776. I am inclined to attribute the different readings simply to the modal ambiguity inherent in the melody as a whole, which contains few references to the final, and to the awkwardness of the leap of a seventh (from “tuam” to “non”) in Pa 776. With this reasoning, Pa 776 emerges as the lectio difficilior.

Roman MSS: The Roman tradition lacks the first verse found widely in Gregorian MSS. Vat 5319 is written at the affinal position and shows the following variants:

  • respond, “CANticum”: ccdcbacbaba

  • v. 1: “mulTA”: ba

  • v. 2: “FE/CISti”: Ga/a

In the verses, Vat 5319 is more standard in its use of Formula A. Bod 74 alternates in its recitation between the standard torculus EFD and the single note F. All of the corresponding places are torculi in Vat 5319:

  • v. 1: “Deus”; “mirabiLIa”; “cogiTIoNIbus”; “QUIS”

  • v. 2: “MEo”; “VEriTAtem”; “MEus”

In Vat 5319, the start of a new verse is indicated at “benenuntiavi.”

RoS 22 shows the typical differences from Bod 74 and Vat 5319 in its use of Formula B. The last segment is dedcbcd (rather than dcbcd): “domiNUM”, “ME”, “meAM”

Other variants in F22:

  • “CANticum”: cdcbacbaba

  • “Ymnum”: bdcb

Benedicite gentes (36)

The respond and first two verses of Benedicite gentes exhibit no major variants, aside from minor differences in recitational style. In v. 3, however, the sources differ in pitch level, producing several different readings.1 The variants begin at “ad ipsum” (phrase 17). Pa 776 was chosen as base version because its reading of this passage is the most probable preferred reading; it is also the majority version among Aquitanian sources.

Example 36.1 presents several different versions of the passage. In Pa 776 (line 1), a common referential interval of mode 1, a‐c, is projected a tone lower, with the pitches G‐b‐ (p.297)

Appendix 3 Commentary on Edition

Example 36.1a

Appendix 3 Commentary on Edition

Example 36.1b

(p.298)
Appendix 3 Commentary on Edition

Example 36.1c

flat. This reading is found in Mo 159 and in several Aquitanian sources, including Pa 1121, Pa 780, and Lo 4951. In Ben 34 and Ben 35, lines 3 and 4, the passage is notated a tone higher, a range more consistent with the characteristics of the first mode. This reading of the passage is adopted in sources from a wide geographical spectrum, as summarized below.

The second variant point occurs between the melisma on “lingua” and the cadence on “mea.” Several of the G‐protus sources, including Mo 159, join the pitch level of Ben 34.

These sources move to the upper range at different places, but generally on the last syllable of “lingua.” Mo 159's reading of the melodic figure on the last syllable of “mea” thus differs intervallically from that of Pa 776. Without chromatic inflections, the semitone F‐E in Mo 159 is the whole tone E‐D in Pa 776. I would hypothesize that this upward transposition was made to represent the semitone D‐E‐flat at the level of Pa 776. This hypothesis is supported by the reading of Pad 47 (line 5), which is a fifth higher than Pa 776 and notates a b‐flat here.

It is doubtful that the nondiatonic pitch on “mea” is the only cause of the variant readings of this passage. The readings that match Ben 34 at the beginning of the passage articulate the third a‐c and are based in the diazeugmenon tetrachord; the lower version is based in the synemmenon tetrachord. Like several other passages in the offertories, this one exhibits the characteristic of modal “transformation” described by Guido, in which G and b‐flat, used in close proximity, create a G‐protus sound. The whole‐tone transposition in the other sources is recommended by Guido as a solution. It is worthy of note that Mo 159 adopts the ambiguous sign here that may indicate microtonal inflections, as indicated by the diamond in the transcription.

(p.299) Most of the sources included in the sampling join at or before “propterea.” The majority version of this passage, shown in five of the six sources in the example, is marked by a normal first‐mode structure centered on the pitch a. Two sources, Ben 35 (line 3) and Gr 807, however, remain a whole tone above the others. Gr 807 (not shown in the example) joins at “et intende” and Ben 35 in the final melisma of the verse. These sources present a different intervallic structure from the others at the cadential figure on “deus,” which is EFG in Ben 35 and DEF in the majority reading. It is unclear whether this transposition represents a nondiatonic pitch, such as E‐flat at the lower level.

  1. 1. “Ad ipsum”:

    • = Pa 776: Mo 159, Pa 903, Pa 4951, Pa 780, Be 40078, Pa 1121, Pa 1135, Pa 1136, Pa 1137, Pia 65

    • 2↑ Pa 776: Ben 34, Ben 35, Pa 1134, Pa 1132, Pa 1235, Cai 61, Gr 807, Pst 120, RoV 52

    • 5↑ Pa 776: Pad 47, Mod 7

  2. 2. Final syllable of “mea”:

    • 2↑ Pa 776: all except

    • = Pa 776: Pa 1121, Pa 1136, Pa 780, Lo 4951

    • 5↑ Pa 776: Pad 47, Mod 7

  3. 3. “propterea”:

    • = Pa 776: all except

    • 2↑Pa 776: Ben 35, Gr 807, Pa 1132

    • 5↑ Pa 776: Pad 47, Mod 7

  4. 4. “Et intende”:

    • 5↑ Pa 776: Pad 47, Mod 7

    • = Pa 776: all others except Pa 1132

Notes on specific sources:

Pa 776: In the final alleluia of the respond (phrase 7), the text has been erased.

Several Italian sources assume the same register profile as Paris 776, but notate the entire offertory a fifth higher.

  • Pst 120 lacks the third verse.

  • To 18 lacks notation in the third verse.

  • v. 3 cadence on “mea”: Pa 1235 has DDFDDC; Be 40078 has DFDC.

  • v. 3 melisma on “lingua” in Pia 65: abaGaGFFGaGFaGFFEFa/aaG.

Roman MSS: Bodmer C74 has the verses in the following order: (1) In multitudine; (2) Iubilate; (3) Venite. The first two verses in the transcriptions have been reversed to facilitate comparison with Pa 776.

There are very few significant variants.

  • Respond, “DOminus”: Vat 5319 and RoS 22 have FGFE

  • Melisma on final syllable of “alleluia”: RoS 22 has a more elaborate melisma: DFEDFEFDCD

  • v. 3 “Ipso”: Vat 5319 FGaGaF; et [exultavit]: G ori. FEaG liq.

(p.300) Populum humilem (37)

The sources exhibit variants in several of the long melismas in the second verse.

v. 2 melisma on “MEus”: Sources in the core group exhibit some small points of variance, as exemplified by the following (significant differences from Mo 159 are shown in bold): Pa 776: bbb GFGaFG bbb GFGa cdec a a (ori.) Ga ccacb cdcdcca cb c dcdcc aFGFa a (ori.) Ga cdccaG. The Beneventan sources have an abbreviated version, as in Ben 34: aa aGFGa FGaaaGFGa c dec aa(ori.)G a c cacb cdcc(ori)acb cdcdccaG.

v. 2 “exalTAbis”: Although sources in the core group show only small variants, some sources in the peripheral group have different versions of the melisma, particularly in Italian sources. The segments that reflect significant differences are shown in bold.

Ben 34 (minor differences from Mo 159): acc ccc abc dc Fac ccc abc dc ac dc ac e ed ce ec edc ac ca dcb GacccaGF acccaGF ac cc abc

Pad 47: accccc abc dcFa cc ccc abc dc ac dc ac d eed c e ecdcb a d dcedc bc e ecedc bceeecbacccaGFG ccccc abc d (similar version in Mod 7)

RoV 52: accccc abc dc Fga ccccc abc dc ac dc ac d e ed fe g gdfe bd db dcb bc eeecba bc eeecbaeee

To 18: an abbreviated melisma, acc ca abc dc ac d dc cf ecedc Gacc aGFacc abc (The melisma of Pst 120 is much closer to those of the core group.)

Pa 1235: ac cccc cac cdc FGa cccc ca cdc ac dc ac de ed df fcedc cd dc dcb Ga cbcaGF acbc aGF ca c cac

Variants in individual sources:

Gr 807: v. 2, “MEus”: differs in contour in the later part of the melisma, where the c‐d oscillation begins: acb db ca cb adc ebc (etc.)

Pa 1235: v. 2, “MEus”: differs in contour in the later part of the melisma, where the c‐d oscillation begins: adcdc ca cb adcdbc a (etc.)

Roman version: The beginning of the respond, through “populum humilem,” is written a whole tone lower both in Vat 5319 and RoS 22, presumably necessitating the use of b‐flat.

Factus est dominus (38)

Factus est dominus is one of the few offertories with substantial verbal differences between the Gregorian and Roman versions. The Roman respond begins as the Gregorian does, but continues differently. The lyrics of the Gregorian respond are found transmitted as the first verse of the Roman. The reason for the longer text of the Roman version is not clear. Since it is not found in any Gregorian MSS, it may reflect a change that took place after the transmission of the tradition to the Franks. In the base transcription, the two versions are aligned in the places where they have matching lyrics. To facilitate comparison of the two versions, the phrases are numbered according to the longer Roman version. Neither version has repetenda indicated.

The modal instability of this chant was discussed long ago by Bomm. The chant ends variously on F, E, or D.1 Mo 159 and the Aquitanian MSS close the chant on E, whereas the (p.301) Beneventan MSS close it on D. The endings are further discussed below, with other melodic variants.

Factus est dominus also presents a problem of pitch level and modal consistency at the end of the second verse, beginning with “inimicorum meorum” (phrase 14). The base reading of Pa 776 represents the majority Aquitanian version and the probable unemended pitch level. Similar readings are found in Pa 1121 and Pa 780. To some extent, the passage is G‐protus in quality: a b‐flat is employed in Mo 159 and implied by the sources that notate the passage, fully or partially, a whole tone higher, with c as the focal pitch. By the passage at “dedisti mihi” (phrase 15), most non‐Aquitanian MSS are a whole tone above Pa 776. On “mihi,” the interval F‐E in Pa 776 appears in these sources at the whole tone G‐F. To duplicate the intervallic structure of these versions, an E‐flat is required at the level of Pa 776. The MSS that are a tone higher move to this position at different points in the melody, a variety that suggests they are emendations.

  • = Pa 776: Pa 1121, Pa 780, Gr 807

  • 2↑ Pa 776 (Unless otherwise noted, the whole‐tone transposition continues through the end of the verse):

  • Mo 159 (beginning at “et inimicorum”)

  • Ben 34 and 35 (beginning on the second note of “dedisti,” returning to level of Pa 776 at “odientes”)

  • Pa 1235 (beginning in the melisma on “subplanTASti”:“subplanTAS/TI” is: cc adcd ceded fff/ded)

  • Cai 61: (beginning at “et inimicorum”)

  • Tri 2254: (beginning at “subplanTASti”)

  • Mod 7 (beginning at “inimicos meos”)

  • Pad 47 (beginning at “subplanTASti”)

  • Pst 120 (beginning at “et” [subplantasti])

  • RoV 52 (beginning at “et” [subplantasti])

Three MSS included in the sampling adopt different solutions to the problem. In Be 40078, the verse is written a fifth above its level in Pa 776, and the E‐flat at “mihi” at the level of Pa 776 is confirmed by a notated b‐flat in Be 40078, which also occurs on the first syllable of “dedisti.” The scribe of Lo 4951 notates the passage, beginning at “ad bellum,” a fourth below Pa 776 (and an octave below Be 40078), presumably representing the problematic E‐flat with a low B‐flat. Gr 807 is notated at the level of Pa 776, and the note corresponding to an E‐flat on “mihi” is a D.

Melodic variants:

respond, “eum”:

Mo 159 opens the melisma differently: a aGG(ori.)F EFDC DFF (etc.) respond, final cadence:

As mentioned, the closing three or four notes are a point of variance. The sources incorporated into this study, however, do not differ to the extent of those studied by Bomm; there is a clear majority ending on E:

  • closing notes GFE/E: Pa 776, Mo 159, Pa 1121, Pa 780, Cai 61, Pa 1235, Be 40078, Gr 807, Mod 7, Pst 120, Pad 47, RoV 52

  • Lo 4951: FGGF/D

  • Ben 34, 35: GFD/D

v. 1 “meOS”: Ben 34: DFDD(ori.)C (Ben 35 same without oriscus)

(p.302) v. 1 “COM/PREhendos” (small variant):

  • Mo 159 and Beneventan MSS: FGF/GaG

  • Pa 780: FGF/aGa

  • Pa 1121: FGF/aG (etc.)

  • Lo 4951: as in Pa 776

v. 1 “non” (small variant):

  • Mo 159: DG FG(quil.) a b‐flat

  • Ben 34, 35: DE DEab

  • Pa 780: FG FG(quil)ab

v. 1 “defiCIant”: Ben 34 (and other Ben MSS): FGaG

v. 2 “ME/US”: Ben 34: bbb/GaG (Ben 35 has the usual reading)

v. 2 “SUB/TUS”: Ben 34 and 35 Gb(liq.)/bcd

v. 2 “inimicorum”: Pa 776 is the majority Aquitanian version and found in Beneventan sources, Pa 1235, and the Italian MSS; Mo 159 and several other MSS have a different reading, departing from the literal repetition of the previous phrase that is found in Pa 776:

et in‐i‐ mi‐ co‐rum me‐o‐ rum

Mo 159 (2↑ Pa 776):

G/ac/[b]c/ba/bc/c/ c/c[b]cba/ba

Gr 807 (= Pa 776, b's are flat):

F/Gb/b /b /b/ b/b/ cbaG/bG

Roman MSS: As mentioned, the version of Bod 74 and Vat 5319 has a longer text than the Gregorian does. The Roman version transmits a different respond text, with three verses; the text of the first “verse” is identical to that of the Gregorian respond.2 Although this section is designated as a verse in both Bod 74 and Vat 5319, it closes with a cadence normally reserved for the ends of responds in sixth‐mode offertories. The verses, moreover, close with a repetendum to “et liberator,” which is in the middle of the first “verse.” This section, then, is atypical of a verse both in its closing cadence and in serving as the source for the repetendum. These features may suggest that the respond and the section designated as the first verse may be alternate responds, as Peter Jeffery has suggested.3 How these sections were performed is not clear. RoS 22, which lacks verses, transmits only the portion indicated as the respond in Bod 74 and Vat 5319.

Vat 5319 shows a significant difference from Bod 74 in one passage: the beginning of the third verse, Precinxisti, is written a tone above Bod 74 until the final syllable of “virtutem.” This melisma occurs numerous times in the repertory, and in all other cases it is a tone lower. This variant is probably best viewed as a simple error.

RoS 22 has the following variants:

(p.303) “DOminus”: aca GbaG

The alternate version of the final cadential segment in Formula B (common in RoS 22 but not found in the other MSS): GaGFEFG (rather than GFEFG) on “meUS,” “feCIT,” “potentiBUS.”

A different version of the final cadence, “hodeRUNT ME”: FGaGbaGFGa/GaGFEF

Confitebor tibi (39)

The Roman version of this offertory lacks two of the verses that circulate with it in the Gregorian tradition, Beati immaculati and Deprecatus sum. For purposes of comparing the two versions, the phrases in the transcriptions are labeled according to the longer Gregorian version.

The first verse of the Gregorian version, Beati immaculati, is verbally identical to that of the deuterus offertory Benedictus es…in labiis (16) but has a different melody. The Roman tradition lacks a corresponding verse in Confitebor tibi. In the Roman tradition, however, the first verse of Benedictus es…in labiis has a melody similar to that of the Gregorian version of the corresponding verse in Confitebor tibi, with first‐mode characteristics. The verse melody is clearly common to the two traditions, but at some point became associated with two different responds.

In both traditions, Confitebor tibi shares melodic material with other first‐mode offertories. The melisma that opens the Gregorian first verse also opens the first verse of Laetamini, also on the word “beati.” The lengthy melisma that opens the corresponding Roman verse, in Benedictus es…in labiis, is also found in Laetamini (79) and Gloria et honore (74).

In most Gregorian MSS, the verse that begins “deprecatus sum” (phrase 13) is transmitted as part of the second verse, but in Ben 34 and Be 40078 it marks the beginning of a new verse. In Lo 4951, “inclina” (phrase 10) marks the beginning of a new verse.

Beneventan distinctiveness: The Beneventan tradition has an independent reading of the cadence on “domini” in phrase 4. The other MSS have the traditional protus cadential pattern on the last two syllables of “domini”: DEFEDE/ED. V. 1 “eUM”: Beneventan MSS have a distinctive reading of this verse ending. The other MSS have DFDCD.

Other melodic variants: The melisma on “ioCUNda” (phrase 12) is a point of variance. Ben 34's reading is matched in the other core MSS. German MSS present a reading close to that of the core group. The Italian MSS, Pa 1235, and Cai 61, however, transmit variant readings:

  • RoV 52: FF GFD FFF GFD FaGaG c caca F FD FGFa GaG FGFD FF FD FFF

  • Pst 120: FFF GFF D FF FGFFD FaGaG ccacGFGF FGF a GaG FGFD FF FD FFF

  • Pad 47: FFF FGFD FF GFD FaG aG cca cG FD FaGa FaGF FFFDFFF

  • Pa 1235: FFF GF FD FFF GF FD FaGaG c caca F FD F FE FGE aGaG FGFD FFFD FFFD

  • Cai 61: FFF GF FC FFF GF FC FaGaG c cG ba G GE F FE FGFa GaG FGFD FF FD FF FE(liq.)

Individual MSS: (p.304)

  • Lo 4951: respond “VERbum”: FacaG aGFGFD

  • Pa 780: v. 3 “[legem] TUam”: a cca bcaG

Gr 807:

  • v. 1 “testimoNI/A”: G/GaG

  • v. 1 “Eius”: FG aG aGF (liq.)

  • v. 2 “ioCUNda”: the bottom note of the melisma is C (rather than D)

  • v. 3 “leGEM”: differs in the parts of the melisma in boldface: aGc cc cabaGa Gabcba Gabcba Gc cc ca bcaGa bcda bcd cd ec bca bcaG cca(liq.)

Be 40078:

  • v. 2 “ioCUNda”: the bottom note of the melisma is C (rather than D)

  • v. 3 “leGEM”, differs in the parts of the melisma in boldface: aGc cc cabaGa Gabcba Gabcba Gc cc ca bcaGa ccda ccd cd ec bca bcaG cca(liq.)

RoV 52:

  • v. 1 “[testimonia] Eius”: FGaGFG

  • v. 3 “leGEM”, missing the last few notes of the melisma: aGc cc ca cccaGa Gabcba Gabcba Ga c ca ccaGc cc ca cd cdec ca bcaG cc

Pa 1235: respond, “DOmine”: aGa bcba

Roman MSS:

Vat 5319:

  • respond “TIbi”: a ccc(ori.) ba

  • respond “ME”: FabcaG

  • v. 1 “leGEM”: ba(liq.)

  • v. 1 “cusTOdiam”: FGaba

  • v. 2 “viVI/FI/CA”: a (ori.) c/c (ori.)/GFGF FE

RoS 22:

  • “viVIfica”: ac cb

  • “seCUNdum”: FFEDEDEDC

Domine convertere (40)

This sixth‐mode offertory presents no major variants until the expansion into the authentic range in v. 2, at “sana me domine” (phrase 7), a point of variance in pitch level.1 There are at least five different readings of this passage among the sources examined. In the majority reading of the verse, found in Ben 34, Mo 159, Pa 776, and many peripheral MSS, the entire passage has a focal pitch of c. In these sources, the final melisma is written a tone higher than it is in Pa 1121, the base reading of the edition. In a second group, represented by Pa 1121, the passage is at the level of Mo 159 and Ben 34 until the final melisma on “ossa,” (p.305) which is a tone lower. In many of the MSS from St. Martial, the whole passage, beginning at “sana me,” is written a tone below the level of Mo 159, remaining there until the end of the verse.2 In two MSS, Pa 1235 and To 18, the respond, first, v. 1, and beginning of v. 2 are notated at the affinal position, with a final of c, but these MSS join the level of Mo 159 at “sana me domine.” Finally, in some sources, such as RoV 52, the pitch level of Pa 1235 is replicated a fifth lower.

Hankeln hypothesizes that the pitch level of Pa 1121, with its lower notation of the final melisma, is the Aufführungslage, and that a discrepancy exists between the notated and performance levels in Pa 776.3 The repetendum cue in Pa 776 is written a tone lower than the corresponding passage in the respond, suggesting a performance level a tone below the written level. A performance at the level of Pa 1121 requires the use of the nondiatonic pitch e‐flat, equivalent to b‐flat in the versions written a fourth lower. In the versions written a tone higher, the interval c‐e‐flat may be represented with d‐f. I find Hankeln's hypothesis plausible and accordingly I have adopted Pa 1121 as the base version and the most probable unemended reading. In this case, the preferred reading differs from the majority reading. As shown in the following summary, most MSS write the close of the second verse a tone above Pa 1121.

  1. 1. “Miserere”(beginning of v. 2):

    • = Pa 1121: all except

    • 5↑ Pa 1121: To 18, Pa 1235

  2. 2. “Sana me domine”:

    • = Pa 1121: all except

    • 2↓ Pa 1121: Pa 1132, 1133–36

  3. 3. “ossa mea”:

    • = Pa 1121: Pa 1132, 1133, 1134, 1135, 1136

    • 2↑ Pa 1121: Ben 34, Ben 35, Ben 39, Mo 159, Pa 776, Pa 1235, Cai 61, Be 40078, Gr 807, Pst 120, To 18

    • 4↓ Pa 1121: Lo 4951, RoV 52

Roman sources: In many passages, Vat 5319, Bod 74, and RoS 22 agree in striking detail in their use of Formula B. There are the following differences:

“Respond [misericordiam] TUam”: Vat 5319 has a more common and much less ornate form of this precadential segment, simply GaF, creating more literal repetition within the respond. Bod 74, by contrast, distinguishes the final cadence from the others with the special ornate form. RoS 22 is closer to Bod 74: FGaGaGF.

  • v. 1 “arguAS”: Vat 5319 has FgaGaGaGF (liq.) (more ornate than Bod 74)

  • v. 2 “miseREre”: Vat 5319 has a shorter form of the standard sixth‐mode melisma than Bod 74 does. In the first thirty notes, through the repeated F's, Vat 5319 is identical to Bod 74, then the melisma closes with GaGaGF.

Structural differences: RoS 22 closes with a repeat of the opening words, “domine convertere et eripe,” that is lacking in the other two Roman MSS, perhaps by analogy to (p.306) Domine in auxilium and Desiderium. Vat 5319 has a repetendum after the first verse that is lacking in Bod 74.

Sperent in te (41)

The repetenda in Sperent in te are a point of variance. The base reading of Ben 34 has “quoniam” as the repetendum, whereas other MSS, including Mo and the Aquitanian MSS, have “psallite,” matching the Roman reading.

This offertory shows very few variants between MSS.

  • Lo 4951 verse endings:

  • v. 1 “PAU/PE/RUM”: E/EGFE/E

  • v. 2 “deUS”: E GFGF F(ori.)E

  • Pad 47:

  • “iudiCAS”: aca aGa

  • v. 2 “domiNUS”: ba

  • v. 2 “iudiCI/A”: aG/GGGF Gaba

  • Pst 120: v. 2 “iudiciA”: melisma begins GGGE

  • Mo 159: v. 2 “iudiciA”: melisma begins GGGE

  • Be 40078: v. 2 “domiNUS”: ca

  • Gr 807: respond, “TU/UM”, a different syllable distribution: cadcb/b

  • Roman MSS:

  • Vat 5319: v. 1 “pauPErum”: GaGF

  • RoS 22 shows its standard differences in the use of Formula B. Where Vat 5319 and Bod 74 have GFEFG, RoS 22 has GaGFEFG on oMNES.

  • “doMIne”: GaGF

  • “DEreLINquis”: FGF/ GFEDEFGFEDEDC (same at oblLItus)

  • “querenTES/TE”: FGaGaF/GaGF

Eripe me…deus meus (42)

Ben 34, adopted as the base version, is missing four notes in a melisma on “FORtes” (phrase 5), a probable scribal error. These are shown in brackets in the transcription.

In the pretheoretical tradition, this offertory probably employed nondiatonic pitches at two places in the verse Quia factus es: the melisma on “meus” (phrase 6) and the final melisma of the verse, on “meae.” The scribe of Mo 159 writes the opening of the verse a fourth above the level of Ben 34, joining at “et refugium.” In the melisma on “meum,” Mo 159 indicates b‐naturals, equivalent to F‐sharp at the level of Ben 34. This nondiatonic practice is also reflected in Pa 1235 and in two sources written a fifth below Ben 34, Be (p.307)

Appendix 3 Commentary on Edition

Example 42.1a

(p.308)
Appendix 3 Commentary on Edition

Example 42.1b

40078, and Cai 61; the latter two MSS employ low B‐naturals, equivalent to F‐sharp at the level of Ben 34. It is unclear, however, how widespread this nondiatonic practice was. The pitch level of Ben 34 represents the majority reading, and the lack of transpositions in other MSS raises the possibility that the nondiatonic practice was suppressed in many traditions. This melisma is also a point of melodic variance, as described below. The pitch level at the beginning of the verse for the sources in the sampling can be summarized as follows (the transposition continues through phrase 7, unless otherwise indicated):
  • = Ben 34: Ben 38, Ben 39, Ben 35, Pa 776, Pa 780, Pa 1121, Lo 4951, Pad 47, Pst 120, RoV 52, To 18

  • 4↑ Ben 34: Mo 159 (joining at Ben 34 “et refugium”), Pa 1235 (joining Ben 34 at “in die”)

  • 5↓ Ben 34: Cai 61, Be 40078

The second problem spot is the end of v. 2, at the melisma on “meae.” Although the pitch level of Ben 34 is found in a great majority of other MSS, the nature of the variants among a few of the core and peripheral MSS suggests that the majority version may reflect an emendation. The pitch‐level profile of the MSS at “in die” and “meae” can be summarized as follows.

v. 2 “in die tribulationis”:

  • = Ben 34: All except Mo 159, Cai 61, and Be 40078

  • 2↓ Ben 34: Mo 159

  • 5↓ Ben 34: Cai 61, Be 40078

v. 2 meae:

  • = Ben 34: All except Pa 776, RoV 52, Cai 61, and Be 40078

  • 2↑ Ben 34: Pa 776, RoV 52, Gr 807

  • 4↓ Ben 34: Cai 61 (with melodic variants), Be 40078. In both MSS, the interval of transposition changes to a fifth at the end of the melisma.

Mo 159 presents a reading that is unique among the MSS examined: “in die tribulationis” is a tone below the level of Ben 34. It is doubtful that this whole‐tone transposition reflects a nondiatonic pitch at “in die tribulationibus”; Mo 159 consistently uses b‐flat, thereby retaining the same intervallic structure as Ben 34. The lower notation of Mo 159, however, does suggest a deliberate emendation on the part of the scribe, perhaps reflecting the need for a nondiatonic pitch further on. At “meae,” Mo 159 joins Ben 34 and the majority of other MSS. Pa 776, however, remains a tone above these MSS. Although Pa 776 is unique among the Aquitanian MSS examined, it is doubtful that its reading of this passage is a scribal error: it is matched in the unrelated MSS Gr 807 and RoV 52. Pa 776's pitch‐level profile is also confirmed by the two later MSS, Be 40078 and Cai 61. These MSS duplicate the pitch‐level profile of Pa 776 a fifth lower until the very end of the melisma, where the interval of transposition changes to a sixth.

Several different readings of the passage are presented in example 42.1. To replicate the intervallic structure of the majority version (Ben 34) at the level of Pa 776, a c‐sharp is required at the end of the passage. This point is precisely where the lower readings, Cai 61 and Be 40078, change in their intervallic relationship to the other MSS, presumably because an F‐sharp would be needed at the position a fifth below Pa 776. The half step b‐c in Ben 34 is equivalent to c (sharp)‐d in Pa 776 and E‐F in Be 40078. The nature of the relationship (p.309) between these MSS leads to a tentative hypothesis that Pa 776 represents the unemended version in pitch level and that the pretheoretical tradition employed a c‐sharp.

Melodic variants: The independence of the Beneventan tradition is evident in several brief passages: respond “MEis” (a point of slight variance):

  • Ben 34: ad ecc dcdcdc (same reading in Ben 35, Ben 39)

  • Mo 159: de cc dc dc dc (same reading in Pa 780, Lo 4951)

  • Pa 776: de cc dcdc

MSS in the peripheral group present a reading closer to that of Mo 159 and Pa 776.

v. Quia factus es, “MEus”: As mentioned, this melisma is a point of variance among core MSS, particularly at the segment in boldface. Pa 776 has a shortened version that is not found in other MSS included in the sampling.

  • Ben 34: bcdedcbcaGccbdc bcaG bGGG FGaFG FGED GG FababaGa

  • Mo 159 (4↑): efgagfefdc fc ffcgfefd cecc bcdcc abaG cc bdededcd

  • Pa 776: bcded bcaG cGG FGaFG EFED GG GababaGa

  • Pa 1121: bcdedcbcaG caccbdc bcaG bGG GGaGaGG FGED GG GababaGa

  • Pa 780: bcdedcbcaG cG ccbdccdaG cGG FGaGG EFEED GG GababaGa

  • Lo 4951: bcdedc bcaG bG ccbdcbcaG bGG FGaFG EFED GababaGa

Sources outside the core group exhibit further variance, as the following sample readings show:

  • Gr 807: bc dedc bcaG cG ccb dc cccaG GEF FGEF DFDC FFE FGaGaGFG

  • RoV 52: bcdedc bcaG cac cb dc bcaG bG FGaGa GaGF GababaGa

  • Pst 120: bcdedcbcaG cac cdcbcaG GEF EFGF EFDC GG GabcabaGa

  • Mod 7: bcdedcbcaG bGa cb dcbcaG bGa bc dc bcaG aaG GabcabaGa

  • Cai 61 (5↓): EFGaGF EFDC FC F FE GF EFDC FCC BCDC ABAΓ CCB DEDEDCD

  • Pa 1235 (4↑): efg agf efdc fcffeg fefdc dcc ac dcc abaG ccb dededcd

Variants in individual MSS:

Verse endings in Lo 4951:

  • v. Quia factus es: cfffffdcdcbabaaG

  • v. Quia ecce: dffdcdcba ccdcb abaaG

  • v. Quia ecce, “FORtes”: Lo 4951 is missing four notes of the melisma, the figure cdcb.

Pa 1121: E (rather than F) is the top note in the melismas on “captaVErunt”, “FORtes”, and “[tribulationis] MEae”.

Roman MSS: Respond, “domiNE”: Vat 5319 and RoS 22 have a melisma on the final syllable of “domine” that is lacking in Bod 74. The melisma is varied repetition of the one that occurs on the first syllable of “domine”:

  • Vat 5319: Gb dedcb dedcb cbabG bacbaGa aG

  • RoS 22: Gb dedcb dedcbcbabG bacbaGaG

    • v. Quia factus, “QUI/A”: Vat 5319 has a less ornate opening: G/cb

    • v. Quia factus, “[refugium] MEum”: Vat 5319: c c(ori.)baG cbc

(p.310) Benedictus es…et non tradas (43)

The Roman version of this offertory has a text repetition in the respond that is lacking in the Gregorian, a difference probably attributable to the verbal and melodic similarities to Benedictus es…in labiis (16), which begins with identical words and music, and where the text repetition is present in both versions.

The opening of the second verse, “Adpropiaverunt confundantur,” exhibits complex variants in pitch level that are most likely attributable to nondiatonic practice. Because of the intervallic inconsistencies between readings and the lack of repetendum cues in some key Aquitanian MSS, however, I have not been able to establish an unemended version of the second verse with any degree of certainty. For this reason, I have given Mo 159, the majority reading, as the base version.

Example 43.1 presents five different readings of the verse opening. Because the opening pitches of the various versions do not necessarily correlate with their eventual pitch level, I will discuss pitch level beginning with “persequentes.” Here Mo 159 (line 1) represents the pitch level found in a majority of MSS, as summarized below. Ben 34 (line 2) is written a tone lower, a reading also found in Pa 1235. Pa 776 (line 3) is a fourth lower, a reading matched in Pa 1121 (line 4) and several other Aquitanian MSS. Be 40078 (line 5) is the only MS that presents the passage a fifth below Mo 159. Intervallic differences between these readings are evident at the ends of the short melisma on “me” and the melisma on “iniqui.” Mo 159 and Be 40078 have a semitone (c‐b‐natural and F‐E), Ben 34 has a‐b (flat?), and Pa 776 has G‐F. The semitone in the majority version of Mo 159 suggests that (p.311) perhaps an F‐sharp is intended at the level of Pa 776 and Pa 1121, both at “me” and “iniqui.” The same solution is suggested by Pitman.1

Appendix 3 Commentary on Edition

Example 43.1a

Appendix 3 Commentary on Edition

Example 43.1b

While the intervallic differences between the versions support a hypothesis that the level of Pa 776 and Pa 1121 is the unemended pitch level at “persequentes me iniqui,” it is unclear how a hypothetical unemended version of the chant would proceed at “iniqui confundantur” and through the rest of the verse. Nearly all non‐Aquitanian MSS, along with Pa 1121 and Lo 4951, join the level of Mo 159 here. The only exception is Be 40078, where the passage is positioned a whole tone below Mo 159 through the end of the verse. In the Aquitanian tradition, Pa 776, Pa 780, Pa 1136, and Pa 1135 remain a third below Mo 159 throughout the verse. These Aquitanian MSS, however, lack repetendum cues to confirm that the notated level is indeed the performance level (though Pa 1135 and 1136 do have a custos between the first and second verses).

Regarding the intervallic structure at “confundantur,” it is worth noting that the same melody and words occur elsewhere in the repertory, suggesting that the intervallic structure of the majority reading is the intended one.2 Because the reading of Pa 776 is found in several other Aquitanian MSS, however, I am hesitant to dismiss it as an error. If the level of Pa 776, a third below the majority reading, is considered the unemended version at “confundantur,” a c‐sharp would be required to replicate the intervallic structure of the (p.312) majority version. With a hypothesis that Pa 776 is the preferred reading at “persequentes” and the majority reading is the unemended reading at “confundantur,” however, the transition between the two sections is problematic. Assuming that an F‐sharp was sung as the final note of “iniqui,” the transition between the two sections would require the improbable leap of a diminished fifth, as suggested by Pa 1121.3

To summarize, the base reading of Mo 159 is probably emended at “persequentes me iniqui.” I would cautiously surmise that Pa 776, with F‐sharp and c‐sharp, represents the unemended version of the verse as a whole. In this version, the melody is displaced in relation to the diatonic background scale and never gets back “on track.” Because the evidence for this supposition is not conclusive, however, I have adopted the majority version of Mo 159 as the base reading.

There are few significant melodic variants between MSS. These are as follows:

  • v. 1 “faciES”:

    • Pa 776: cdcbcbaGaG

    • Pa 1121: cdbcaGa

  • v. 1 “IU/DI/CI/UM”:

    • Lo 4951: ccbab(quil.)cba/aG/Ga(quil.)bab/aG

    • On the syllable “ci” Gr 807 has the version of this common figure found throughout this MS: bdbc.

  • v. 2 “inIqui”: This is a point of variance; most Aquitanian MSS match Mo 159, but several sources have shorter melismas:

    • Pa 1235 has a shortened melisma: dedc dedc dcb

    • Ben 39: eded dcd edcdcb

  • “inIUste”: The Aquitanian and Beneventan MSS have a different reading:

  • Lo 4951: cccaG b(ori.)aGa

  • Pa 776: cccaGbaGa

  • Ben 39: cc ca Gca Ga

This reading is matched in MSS from German‐speaking areas, such as Gr 807: cc ca GcaGa; and in Italian MSS such as Mod 7.

  • v. 2 “IN ME”: Lo 4951 Ga(quil.)bab/aG

  • Roman MSS:

  • Vat 5319: “iustificatioNES”: (first time): c cbcabaGF ab(ori.)aGFGF

  • RoS 22: “iustificatioNES”: (first time): c cbca baGFGF

Improperium (44)

The MSS that indicate b‐flat are inconsistent in the places they do so. Mo 159 uses b‐flats in several places where they would not be automatically expected in the context and do not occur in other MSS: in the respond on “expectavit,” [cor] “meum,” and “aceto,” in v. 1 on “usque,” and in v. 2 on “in me.” I have indicated b‐flats only in the places where they are (p.313) both implied by the context and are consistently indicated in the MSS that distinguish between b‐flat and b‐natural.

In many MSS, including Ben 34, this offertory is an unproblematic eighth‐mode melody, with some passages expanding into the authentic range. Several sources, however, notate the final verse (beginning in phrase 14) a fifth below the regular eighth‐mode position, adopting F, rather than c, as the repercussive pitch. To some extent, the two approaches to notating the verse are regionally correlated: the normal eighth mode range is preferred in the Italian and Beneventan sources, whereas most sources from both eastern and western Frankish areas notate the verse a fifth lower. The Aquitanian sources, however, are a mixture: the lower level is implied in most of the St. Martial sources, whereas Pa 776, Pa 780, and Lo 4951 notate the verse in the normal eighth‐mode range.

In many of the MSS that begin a fifth below Ben 34, the interval of transposition changes to a fourth near the end of the verse, on the third syllable of “beneplaciti.” Although this reading is reflected in a majority of the MSS that notate the verse at this lower level, a few sources remain a fifth below the eighth‐mode readings.

The reasons for the lower notation of the second verse in some MSS are unclear. The four early adiastematic sources examined, La 239, Cha 47, SG 339, and Ei 121, begin v. 3 with upward motion, consistent with the higher, eighth‐mode reading of the verse. In the first part of the verse, Mo 159 consistently employs b‐flat, as do the other sources that distinguish between b‐natural and b‐flat, producing a reading that is identical in intervallic structure to Ben 34. At “beneplaciti,” where the interval of transposition changes to a fourth, Mo 159 adopts G as a repercussive pitch, showing an intervallic difference with Pa 1132, which remains a fifth below Ben 34. On “beneplacti” and the final syllable of “misericordiae,” Mo 159 employs F, the whole tone below the repercussive pitch, which would be an E‐flat at the level of Pa 1132 and a b‐flat at the level of Ben 34. While this difference in intervallic structure may explain the change in transpositional level at “beneplaciti,” however, the reasons for the initial downward transposition of a fifth are unclear.

Several of the Aquitanian sources, including Pa 1121, 1134, 1136, and 1137, lack notated repetendum cues at the end of the verse, making it difficult or impossible to establish the intended pitch level of the verse. In several of these cases, the notated level of the verse appears much lower than that of the second verse, suggesting that the level of Mo 159 is intended, but in view of the lack of consistent horizontal alignment between verses, this level is hypothetical. These cases are indicated with brackets in the summary below.

  1. 1. v. 2 “Ego vero”:

    • = Ben 34: Ben 38, Ben 39, Ben 35, Pa 776, Pa 780, Pa 1135, Pa 903, Pa 1235, Lo 4951, Mod 7, To 18, Pia 65, Pst 120, RoV 52

    • 5↓ Ben 34: Mo 159, 1132, [Pa 1121, 1134, Pa 1136, Pa 1137], Gr 807, Be 40078, Tri 2254,

  2. 2. [bene]placiti:

    • = Ben 34: Pa 776, Pa 780, Pa 1135, Pa 1235 (with melodic variants), Lo 4951, Mod 7, To 18, Pia 65, Pst 120, RoV 52

    • 4↓ Ben 34: Mo 159, Gr 807, Be 40078, Tri 2254, 1, [Pa 1121, Pa 1134]

    • 5↓ Ben 34: Pa 1132 [Pa 1136, Pa 1137], Mad 20‐4

Pitch level in Aquitanian sources:

Pa 1121: The third verse begins on a new line and has no repetendum cue.

(p.314)

Pa 1134: The third verse has no repetendum cue, but appears to be notated at the level of Mo 159.

Pa 1136: Relative to the second verse, the third verse appears to be a fourth below Ben, perhaps indicating that the level of Mo 159 is intended.

Pa 1135: The repetendum cue indicates that the intended level is that of Ben 34.

Pa 1137: Relative to the second verse, the third verse appears to be a fourth below Ben, perhaps indicating that the level of Mo 159 is intended.

Pa 1132: The third verse starts on a new line, but the repetendum cue indicates that the intended level is that of Mo 159.

Other melodic variants: There are few major variants among the core group of Aquitanian and Beneventan sources and Mo 159. These are as follows.

Many sources outside Benevento and Aquitaine begin the offertory on F rather than D, as in Mo 159: FGEFE. Beneventan sources are atypical at the cadential pattern on “miserIAM”: The Beneventan reading abcGG contrasts with a more elaborate cadential pattern found in most sources, as exemplified by the reading of Pa 776: ab(quil)cbaGa/aG. Mo 159 departs from the reading of the Aquitanian and Beneventan MSS in several places:

  • v. 1 “MEam”: cdc db caG G(ori.)F GF cc dcdb caG G(ori.)F GcaGaGFG (The Aquitanian MSS present a reading close to that of Ben 34.)

  • v. 3 “[ad te] DO/MI/NE”: Gba(quil)G/G (The Aquitanian MSS present a reading close to that of Ben 34.)

Lo 4951:

  • “quaesiVI”: aG

  • “potaveRUNT”: GFGababaaG liq.

  • v. 1“MEam”: shortened melisma: cdcdccaGGFGcaGaGFG

To 18:

  • v. 1 “Aquae”: shortened melisma: FabaGacaGabbcbccbabaGa

  • v. 2 “psallebant”: FFFEF/Ga/acccaG/Ga

  • v. 2 “VInum”: acdcaGbaGGFacaGaGFG

Pa 1235: v. 3 “oratiOnem”: edcbcaGcccaGccdfffdccaGacccedc Roman MSS: Bod 74 lacks the third verse and has the following variants:

  • “expecTA/VIT”: cdc c(ori.)ba/abaGaG

  • aceTO: GacbaGa aG.

RoS 22:

  • “NON”: dcdedfedcdc

  • “IN/I/STI ME/A”: A slightly varying melody, with a longer melisma on MEa: cba/cdeded/cd cdcbabcdcbabG/acbaG

  • “a/CE/TO”: cdcdcbcaGaG cdcdcbcba bcbabaG/ G acbaGa aG

(p.315) Eripe me…domine (45)

The three Holy Week offertories are characterized by an inconsistency of verse transmission. For Eripe me…domine, the Antiphonale Missarum Sextuplex MSS and the four early notated MSS included in the sampling (La 239, Cha 47, Ei 121, and SG 339) lack a second verse, which suggests that the early Gregorian tradition had only one verse, Exaudi me. A second verse is also lacking in all German and Italian MSS included in the sampling. Three different second verses are found in Mo 159, the Aquitanian MSS, and the Beneventan MSS. In factis serves as a second verse at both Rome and Benevento; for the purposes of comparing the Roman and “Gregorian” readings, a Beneventan version has been employed as the base version for the transcription. Many Aquitanian MSS have Expandi manus meas as a second verse, and Mo 159 has the second verse Velociter exaudi. The added Aquitanian verse, shown in example 45.1, is an individual melody, whereas Mo 159's second verse, Velociter exaudi, is melodically based on the first verse, Exaudi me.

Expandi manus and Velociter exaudi are based on Ps. 142:6 and 142:7. Both lyrics are consistent with the Roman and Gallican psalters, which are identical here.

Pitch‐level variants in the respond: In the respond, the pitch‐level profile of Ben 39 is the majority reading, matched in all MSS except the following.

1. Ben 34 is a fifth below the majority version beginning at “doce me” (phrase 3), so that it closes on the regular final E rather than b.

2. Pad 47: The opening of the respond, “eripe me,” is written a fifth above the majority version, joining at “domine.”

These readings of Ben 34 and Pad 47, unique among the sampling, may reflect attempts to correct a disparity of range between the beginning and end of the respond. The majority version of the respond, shown in the base transcription of Ben 39, is modally anomalous, beginning as a regular deuterus melody but migrating upward to close on b.

Pitch‐level variants in the verse Exaudi me: Most MSS match Ben 39 in pitch level throughout. The following are exceptional readings:

1. Mo 159 is briefly a whole tone above Ben 39 at “ne intres,” with b‐flats. Mo 159 joins Ben 39 at the fifth note of “inTRES”; inTRES begins b‐flat b‐flat (ori.) Ga caG (etc.).

2. Lo 4951 is a fifth above Ben 39 beginning at “in tua iusticia” and remains there through the end of the verse.

Appendix 3 Commentary on Edition

Example 45.1

(p.316) Pitch‐level variants in the Beneventan verse In factis: Ben 34 and Ben 38 are a fifth below Ben 39 beginning at “meditabor.” Ben 35 lacks this verse. The reasons for this variant are unclear.

Melodic variants: There is some inconsistency in the use of b‐flat between Mo 159 and later MSS. In the respond, Pa 1235 uses b‐flat on “domine,” whereas other MSS do not. I have placed parenthetical b‐flats only in the passages where they occur in the majority of sources.

The Beneventan MSS have an independent reading of the opening of v. 1, “exAUdi,” transmitting a much shorter version of this melisma than that found in the broader international tradition. The first twelve notes are omitted, as the following comparison shows:

  • Mo 159: Ga cc caG aaG aa Ga cc deed fed ec(quil.)b

  • Ben 39: Gac c de ed fedec c(ori.)b

Small melodic variants in core MSS: v. Exaudi, “IN [iudicio]”: Some MSS, including Pa 780 and Lo 4951, have DG.

Roman MSS: Vat 5319 has some missing notes at the beginning of the melisma on “exAUdi” (v. 1). The notes cannot be supplied from Bod 74, whose melodic reading differs here.

There are rare significant melodic differences among Roman MSS, including a transpositional variant in the respond. Bod 74 begins at the same pitch level as Vat 5319, has a different melodic reading on the last syllable of “voluntatem” (phrase 3), and is a fifth lower than Vat 5319 beginning at “tuam.” The latter part of the respond and the two verses, then, are notated a fifth below their position in Vat 5319, reflecting a type of variant also found among Gregorian sources. The reasons for these variants are unclear. RoS 22, which transmits only the respond, has the pitch‐level profile of Vat 5319. The reading of Bod 74 may be a correction to the perceived inconsistency of range between the beginning of the respond, in the deuterus plagal range, and end of the respond, on the affinal b. Vat 5319 was chosen as the base version in part because its reading is the majority among the three Roman MSS. Its overall range, beginning at the normal deuterus plagal position but closing on b, corresponds to that of the majority Gregorian version.

A second significant variant occurs at the beginning of the verse Exaudi me, where Bod 74 has a different melisma. In Bod 74, the opening of the melisma is similar to that of the verse In factis, whereas the closing of the melisma matches the one in Vat 5319:

  • Bod 74, “exAUdi” (at the lower transposition): aGaG (liq.) E (ori.) DC FGaGaGFGaGF GaGFEF(liq.)

  • Bod 74, “FACtis”: aGaG (liq.) E(ori.)DC FGaGaGF GaGFEFD

Other melodic variants:

Bod 74:

respond, “volunTAtem”: a(ori.) GFG

v. 1 “IUsticia”: EFGFaGFG

RoS 22: “DE”: DC

(p.317) Custodi me (46)

Like the other Holy Week offertories, Custodi me has an inconsistency of verse transmission. Roman MSS lack the first verse of the Gregorian version, Eripe me. It is not clear, however, that this verse is a Frankish addition, because the text matches the PsR against the PsG.1 The Roman MSS have a verse, Dominus virtus, that is lacking in Gregorian MSS. In the base transcription, the Roman verses are given in the order that they occur in Gregorian MSS in order to facilitate comparison between the two versions.

Although the Roman MSS have a corresponding verse, it appears that the verse Dixi domino (beginning in phrase 6) was not a consistent part of the early dissemination of the repertory in Francia. This verse is absent in most sources indexed in Antiphonale missarum sextuplex, with the exception of Silvanectensis, and in all early notated MSS incorporated

Appendix 3 Commentary on Edition

Example 46.1

(p.318) into the sampling, including La 239, SG 339, Cha 47, Ei 121, and RoA 123. The verse is also absent in the four MSS from German‐speaking areas included in the sampling (Gr 807, Be 40078, Rei 264, and Trier 2254). It is present, however, in some Aquitanian MSS (Pa 776, Lo 4951, and Pa 903), in Mo 159, in the Beneventan MSS, and in the five Italian MSS examined.

The instability of melodic transmission in Custodi me is illustrated in Justmann's transcriptions.2 The second and third verses, Qui cogitaverunt and Dixi domino, exhibit an unusual degree of melodic variance, with a completely different melody in Beneventan sources and yet a third melody in To 18. As Justmann's transcriptions show, To 18's version of Qui cogitaverunt is partially matched in Provins 12.3 Although Mo 159 and the Aquitanian sources have essentially the same melody for these verses, there are variants in the melisma that occurs on “cogitaVErunt” (v. 2) and “DOmino” (v. 3). The melisma has an AAB form in some MSS and no repetition in other sources. The part of the melisma shown in boldface below is lacking in Mo 159.

Mo 159: cdcb cdca FG(quil.)aGa aa(ori.)Ga

cdc dca b‐flat

Pa 776: cdcb cdca FG(quil.)aGa aa

Ga cdcb cdca FG(quil.)aGa

cdc dca b

The version of Mo 159, without the repetition, is implied in the early MSS, SG 339, Ei 121, La 239, and Cha 47. Mo 159 was chosen as a base version because it best represents this earlier tradition of the melody. The longer version of the melisma, with the repetition, is found in all Aquitanian MSS included in the sampling, except for Pa 1121 and the other sources from St. Martial. The longer version is also found in RoA 123 and most MSS in the peripheral group. The only exception, RoV 52, has the shorter version of the melisma in the verse Qui cogitaverunt and the longer version in the verse Dixi domino. 4

A Beneventan reading of the last two verses is provided as example 46.1. In the Beneventan tradition the third verse, Dixi domino, employs much of the same melodic material as the verse Qui cogitaverunt. There is one significant melodic variant among Beneventan MSS in the verse Dixi domino. Ben 39 and Ben 40 have a longer version of the melisma on DOmino, with an internal repetition:

Ben 34: GFD F GaF FEC DCB CDCD

FFF GFG (etc.)

Ben 39: GFD F GaFFFEC DCB CDCD FGFFDC DCB CDCD

FFF FGF (etc.)

Other melodic variants among core MSS: The last two verses exhibit more variance than usual among the core group of Aquitanian sources and Mo 159, as well as further variants in the peripheral MSS.

  • v. Qui cogitaverunt, “meOS”: Pa 776: DFDDC (same in Lo 4951)

  • v. Qui cogitaverunt, “absconDErunt”:

    • Pa 776: DFGa

    • Lo 4951: FGa

  • v. Dixi domino, “ES”: Pa 776: DFDCDED (same in Lo 4951)

(p.319) Significant variants in the peripheral group:

Qui cogitaverunt, “cogitaVErunt”: Gr 807: cdcb cdca FaG

v. Qui cogitaverunt, “meOS”:

  • Pst 120: DFDC

  • RoV 52 (at higher transposition): acaG

  • Cai 61: DFDCD

  • v. Qui cogitaverunt, “absconDErunt”: Italian MSS have FGa (matching Lo 4951 above), except for Pad 47, which has FGa FGa.

  • v. Qui cogitaverunt, “laQUEos”: Cai 61: DFG G FGa FF FD

  • v. Dixi domino, “DI/XI”: Cai 61: Da/a

  • v. Dixi domino, “DOmino”: Cai 61 is unique among the MSS examined in indicating b‐flats in this melisma.

v. Dixi domino, “VOcem”:

  • Cai 61: DF GG FGa FFD FGGG FGa FF FD CD EFEDE

  • Pst 120: DFG GFG FFFC DF FF FGa FFFDC CDEF FEDE

  • Mod 7: DF G GFG FFD DFFF GFE FFF GFE FD DEF FEDE

  • Pad 47: DF GGF GF FD DFFF FaccccaFaF Gab baGa (cadences on G)

Pitch‐level variants:

  • RoV 52: The entire offertory is written at the affinal position.

  • Be 40078: The verse Qui cogitaverunt is written at the affinal position.

Roman MSS: In the respond, the Roman sources show complex transpositional variants that are very rare in the repertory. Bod 74 begins a fifth above Vat 5319, but the interval of transposition changes several times before the two versions join on the second syllable of “hominibus.” Bod 74's relationship to Vat 5319 can be summarized as follows:

  • “Custodi me”: 5↑ Vat 5319

  • “domine de manu”: 2↑ Vat 5319 (melodically variant on DOmine: Gabacba bcaG bab; and “maNU”: cbcd)

  • “peccatoris”: 5↑ Vat 5319

  • “et ab ho [minibus]”: 2↑ Vat 5319 (“hoMInibus”: bG bcaGa)

  • “[ho] minibus”:= Vat 5319

Vat 5319's pitch level is matched in RoS 22. There is unfortunately too little information to posit one version or the other as a preferred reading. The differences between the two versions seem to reflect a confusion about the “modality” of the chant (in pretheoretical terms), a change in practice, or a difference in the melody as it was sung at different institutions. Although there is little melodic correspondence with the Gregorian version, the opening of Vat 5319 exhibits the protus characteristics of the Gregorian. In several cases, however, material in the melody occurs outside of its usual modal context. The verse Dixi domino, for example, begins with a gesture more typical of sixth‐mode verses but notated a tone higher.

(p.320) Domine exaudi (47)

Like the other two Holy Week offertories, Domine exaudi shows an inconsistency of verse transmission. Cha 47 and the Beneventan MSS have only one verse for this offertory, Ne avertas. The most complete Gregorian MSS have three verses, Ne avertas, Quia oblitus sum, and Tu exurgens. The order of the latter two verses is occasionally reversed. Quia oblitus sum is lacking in the Roman MSS, the sources indexed in Antiphonale Missarum Sextuplex, and, with the exception of SG 339 and Ei 121, the sampling of early notated MSS. While these factors may suggest that the verse Quia oblitus is a Frankish addition, its textual basis is the Roman psalter and not the Gallican psalter.1

The verse Tu exurgens might also be a later addition. The Gregorian Tu exurgens has the PsG reading “quia tempus miserendi eius quia venit tempus,” contrasting with the PsR‐based Gregorian tract sung on the same day. The Roman version has a text repetition that is also found in the tract for the same day, “quia venit tempus quia tempus venit miserendi eius.”

Among Gregorian MSS, there are two distinct melodic versions of Tu exurgens, a situation very unusual among core offertories. These patterns raise the possibility that the verse Tu exurgens was not a part of the core repertory disseminated in Francia but subsequently added on each side of the Alps. Because of their complexity of transmission, variants in these two verses are discussed separately. In the Roman MSS, Ne avertas is not designated as a verse.

Melodic variants in respond and first verse: [ad] “TE” [perveniat]: Pa 776 presents a unique reading here. The bracketed notes in the transcription do not appear other MSS. All core and peripheral MSS have shorter version of the melisma:

Mo 159: acaa(ori.)Ga

Appendix 3 Commentary on Edition

Example 47.1a

(p.321)

Appendix 3 Commentary on Edition

Example 47.1b

There is an identical reading in all Aquitanian MSS, a similar version in the Beneventan MSS (aca aGa), and the neumes in early adiastematic sources are also consistent with this reading.

Transpositional variants in v. Quia oblitus:

  • = Pa 776: Pad 47, Mod 7

  • 5↓ Pa 776: Mo 159, Pst 120, RoV 52

  • 2↓ Pa 776: Be 40078

Pitch level uncertain: Pa 1121, Pa 1137, Pa 1136

Melodic variants in v. Quia oblitus:

a. Core MSS and German sources: Mo 159, Aquitanian, and German sources present a unified melodic tradition, with the few significant variants as follows: “quiA”: Pa 1121 (assuming the lower level of transposition): FEFEDCDC. This reading is also matched in Gr 807.

Appendix 3 Commentary on Edition

Example 47.1c

(p.322)

Appendix 3 Commentary on Edition

Example 47.1d

b. Italian MSS: There are more significant melodic variants among the Italian sources that transmit this verse:

“manduCA/RE”:

  • Mod 7 and Pad 47: abaG/aG

  • Pst 120: FGFE/ED

RoV 52: EFED/ED

“PA/NEM”:

  • Mod 7 and Pad 47: Fa/c

  • Pst 120: DF/F

  • RoV 52: CD/ EGG

  • “ME/AM”:

  • Mod 7: dcdc cbab cc dca cbGa Ga bc dc cbaG/ acab

  • Pad 47: dc dcb ab cc dcaba FG Gabc dcba /acab

  • Pst 120: aGaGaGEF aa ba aF GFEF DEFGaGFE/DFDE

  • RoV 52: aGaGaGFEF aaaFGEFEDE DEFGaGFED/DFDE

Melodic variants in v. Tu exurgens: As mentioned above, there are two distinct versions of this melody in the core MSS. The melody of Pa 776, in line 1 of example 47.1, matches that of La 239, thus representing the earliest extant version of the melody. This version is found in most Aquitanian MSS and in one peripheral MS, Rei 264. The second version is best represented by Pa 780 in line 3 of example 47.1. This version also has early witnesses, including Ei 121 and SG 339, and is found in most peripheral MSS. Mo 159, transcribed in line 2, is more consistent with the version of Pa 780 but departs from it at “miserendi,” where it matches Pa 776, and especially at “eius,” where it has a reading not found in any of the other MSS examined.

The version of SG 339, Ei 121, and Pa 780 is found in all Italian MSS examined, in Gr 807, and in Be 40078. The four Italian MSS included in the sampling vary in the pitch level of the incipit: Mod 7 and Pad 47 begin a fourth above Pa 780, on c, and join Mo 159 at “misereberis.” Pst 120 and RoV 52 begin at the level of Pa 780, on F.

(p.323) Pa 776's version occurs only in one MS outside the Aquitanian sphere, Rei 264. The choice of melodic version is correlated with the pitch level—the versions that match Pa 780 are a fifth lower than the reading of Pa 776—except at the incipit of the verse, which is a point of variance.

Note on the transcription of Pa 776: In Pa 776, the pitch axis is consistently maintained throughout the respond and both verses. The third verse begins on a new line but is preceded by a custos, a departure from usual practice. The repetendum, however, appears to be notated a tone too high. In this case, I have given preference to the custodes rather than the repetendum as an indicator of pitch, thus deriving a reading similar to the other MSS that have this version of the verse Tu exurgens. To avoid confusion in the transcription, I have emended the repetendum cue at the end of v. 3 accordingly. Variants in Roman MSS: There are no significant variants (and few variants of any sort) between Bod 74 and Vat 5319.

Variants in RoS 22:

  • “exAUdi”: cbcabaGF acbabaG(liq.)

  • “TUam”: cdedfdcbc

  • “ME” (greatly shortened melisma): Gac cbca GaG EFGFEFE

Terra tremuit (48)

Although Terra tremuit shows few significant melodic differences between MSS, two issues require comment. The first is structural. The base version of Pa 776 is somewhat atypical in lacking repetendum cues after verses 1 and 2, though the same format is found in Lo 4951. In Pa 776, there are two alleluia melodies, one that closes the respond and another that follows the final verse. Although both melodies are found in all sources, the MSS differ in their indications of repetendum cues to these alleluias. In some MSS, the alleluia that is indicated at the close of v. 3 in Pa 776 occurs at the end of each verse, a format followed in Be 40078 and Pa 1235. In other sources, this “verse” alleluia occurs at the end of the first two verses, but the last verse closes with a cue to the alleluia that was heard in the respond. This structural format is found among several sources in the core group, including Pa 1121 and Ben 34. In other MSS, such as Mo 159, the repetenda are not notated.

The opening of the third verse is a problem spot. Several MSS, including Mo 159 and Cai 61, present a reading very similar to that of Pa 776. Mo 159 indicates a b‐natural on “ibi.” In other MSS, however, most of the passage is notated a tone higher. The passage variously begins with the fifth D‐a, the fourth E‐a, or the fifth E‐b. All MSS join with Pa 776 at the melisma on the final syllable of “confregit.” The following sampling of MSS at “ibi confregit” illustrates this variety.

I‐ bi con‐

fre‐

git

Pa 1121:

bd/decde/abG (liq.)

adcbcbb(ori.)

aGacc (etc.)

Lo 4951:

Da/acabc/Ga

acbabaaG

aGacc (etc.)

Ben 39:

Da/acabc/GaF (liq.)

acbabaa(ori)F (liq.)

GFacc (etc.)

Rei 264:

Gc/cdbcd/acb (liq.)

cdcbcbb (ori.)

aGacc (etc.)

Tri 2254:

Eb/bdbcd/ab(quil.)ca

bdcccb(liq.)

aGacc (etc.)

RoV 52:

Ea/abaab/GaG(liq.)

cdcbcbG(liq.)

aGacc (etc.)

(p.324) Lo 4951 and Ben 39 present a similar reading that is also matched in sources as diverse as Be 40078, Pa 1235, Pst 120, and Pad 47. Ben 34 erroneously begins the verse on D and notates the melisma on the last syllable of “confregit” a third below its usual position. That this reading is a scribal error is suggested by the custos at the end of the line, which is also written a third lower than the beginning of the next line, where Ben 34 joins the other MSS. The intention of the scribe at the beginning of the verse, however, is not clear.

The variety of different versions suggests a lack of compatibility between the aural tradition and the notational system at the opening of the melody. While it is likely that a nondiatonic pitch was employed, an unemended reading is difficult to construct. In the version of Ben 39 and Lo 4951, beginning with D‐a, the mode‐1 quality conforms to that of the other verses and the opening passage is placed in a more normal position on the background scale. The version of Pa 776 and Mo 159 reflects the tendency toward melodic repetition in this offertory, because the melodic material at “ibi” is echoed at “illuminans” (phrase 10).

Variants in individual sources: With the exception of the opening of v. 3, there are few significant variants among core sources. Most peripheral MSS present a reading very close to that of the core group, with a few exceptions:

Be 40078: The third verse is written a fourth above its level in the other MSS and employs b‐natural once, in the melisma on “confregit,” suggesting an f‐sharp at the lower level. This nondiatonic practice is not preserved in the other MSS examined.

Pa 1235:

  • v. 1 “IUdea”: D‐a (creating a more literal repetition of “notus”)

  • v. 1 “in” [israhel]: Cd GF

  • v. 1 “IsraHEL”: FGaG GaF FFF GaG

  • v. 2 “habitatiO”: GFaGE

  • v. 3 “miraBIliter”: FGFE FGFGFE FGE FGa cdc caG

  • v. 3 “eTERrnus”: FGE GbaG aG (liq.)

To 18: This MS is unusual in having tropes inserted between verses.

  • v. 1 “IsraHEL”: shortened melisma, FGaGE Ga aG

  • v. 2 “syON”: lacks the melisma

  • v. 3: “arCHUM”: shortened melisma, cccaG abc dccb

Roman MSS:

Vat 5319 varies slightly in the boldface segment of the melisma on “alLEluia”: aGaGF aGaGF aG GF Ga(ori.)GFG.

  • Vat 5319 dum: DF GF(liq.).

  • RoS 22:

  • “iudiciO”: Ga aaGF GabaGFGF

  • “alleluIA”: EFGFaGFEFE

(p.325) Angelus domini (49)

The repetenda in this offertory are a point of variance among core MSS. After verse 1, the base version of Ben 39 has a melodically varied, written‐out version of the repetendum “sicit dixit alleluia,” and this alleluia melody is also cued following the second verse. Although a similar version is found in Mo 159, many MSS, including Pa 776, have this more ornate version of the repetendum only at the end of v. 2, often following a prosula on the “videte” melisma. In these sources, v. 1 either closes without repetendum or is followed by a repetendum cue to the end of the respond.

Two of the Beneventan MSS, Ben 34 and 35, are anomalous in one passage at the end of v. 2. The end of the melisma on “videte” is written a fifth below its position in all other MSS: Ben 34 aGabaG cG abaG FG aGa c cacGG/G.

There is no discernable indication of a nondiatonic pitch here. I have chosen Ben 39 as the base version because it is representative of a broader international tradition in this passage.

There are two probable scribal errors in Ben 39. At the end of the respond (phrase 3), Ben 39 omits the words “sicut dixit.” Instead, these notes are included as part of the “alleluia” that follows. The presence of these words in all other MSS, coupled with the liquescence that should appear on the last syllable of “sicut” (which is in the middle of a melisma in Ben 39) suggests that this is a scribal error. I have provided the words “sicut dixit” in brackets and given the usual text underlay for “alleluia.” In the melisma on the second “alleluia,” at the end of v. 1 (phrase 4), Ben 39 omits two notes that appear in all other sources. In the transcription I have shown them in brackets.

Some Beneventan MSS have an additional verse, Surrexit dominus, that is not found in other traditions and is not given in the transcription. In Ben 39 it serves as the second verse, and in Ben 34 as the first verse. Its melodic material is derived from that of the respond and first verse.

Aquitanian MSS present a slightly different version of the varied repetendum, as in the version of Pa 776: v. 2 “alLEluia”: a dcdcbG ac bcaG G dcd G dcd cbG ac abG (similar version in the other Aquitanian MSS).

Individual MSS:

Pa 1121: In several passages, Pa 1121 has stepwise motion where other MSS have leaps and repeated notes:

  • respond, “desCENdit”: ca a(ori.)F GaGF FD EC CDEFGa caG G(ori.)FG(liq.)

  • v. 1 “Euntes” FG G bbbGF FD DEFGa (etc.)

  • v. 1 “alLEluia” (the varied repetendum): a dcdcbG ac bcaG G dcd G dcdcbF ac abG

Pa 1235:

  • v. 1 “Eius”: The first seven notes of this short melisma are mistakenly written a third too low but are corrected by the custos at the end of the line.

  • v. 1 “galiLEam”: cc cc ca bcd edec cb

Pad 47:

  • v. 1 “VI/DE/BItis”: acb/c acbc (etc.)

  • v. 1 “alLEluia” (second repetendum): d adcdcbG ac bc aG adcd adcdcbG ac bcG

Pst 120:

(p.326) The order of the two verses is reversed; the numbering below reflects their usual order.

  • v. 1 “E/UNtes”: CD/Da cdca bGF (etc.)

  • v. 1 VIdebitis: adc

  • v. 2, opening: The melisma is divided into three segments, each followed by a prosula. The melisma on “videte” is also prosulated.

RoV 52:

  • v. 1 “eUNtes”: c is the top note of the melisma.

  • v. 1 “Eius”: acb cdba Gb babG

  • v. 1 “VI/DEbitis”: acb c acbc (etc.)

  • v. 1 “alLEluia”: cdcdcbG Gb acba adcd cdcdcbG Gb acG

  • v. 2 “Vi/DEte”: Gb/bG (etc.)

Be 40078: The order of the verses is reversed.

Roman MSS:

respond, “CElo”: The three Roman MSS exhibit melodic variants at the end of the melisma and a transpositional variant in the middle. Much of the melodic figuration in this melisma is shared with the melisma common to sixth mode offertories. In Bod 74, the middle part of the melisma is written at the same position as it is in the sixth‐mode offertories, whereas in Vat 5319 and RoS 22 it is written a tone higher, a position more consistent with the eighth‐mode structure of this piece.

respond, “CElo”:

  • Bod 74: cbab(ori.)aGaG acba dcdcbcaFGF EFEDFEDEDCDFFF ababaGaF acbF

  • Vat 5319: cbabaGaG acba dcdcba GaG FGFEGFEFED EG GG(ori.) ababaGaF acbG

  • RoS 22: cbabaGaG acba dcdb cbaGbaGF GFEGFEFEDE GababG aGF acbG

A very similar melisma, with similar transpositional variants, occurs at the end of the respond on “alleluia.”

Vat 5319:

  • v. 1 “Eius”: Gaba babG(liq.)

  • v. 1 is followed by a repetendum to “sicut dixit,” lacking in Bodmer.

  • v. 2 On “STEtit,” a much shorter version of the melisma: babaGaG acba dcdc ababaGaF acbG

  • v. 2 “ET”: FaGFG

Intonuit de caelo (50)

In most MSS, the second verse of this deuterus plagal piece has an expansion to the authentic range, with a change of focal pitch from F to c, matching the base version of Ben 34. In a few sources, however, the whole offertory, respond and both verses, is notated in the transposed plagal range, with a final of b and a focal pitch of c; these MSS lack the contrast in range between respond and second verse.

Because the comparative study did not produce a clear unemended reading, a Beneventan MS representing the majority version was chosen for the base transcription. The transpositional variants, however, may attest to a nondiatonic tradition in the second verse. (p.327) Mo 159 employs a b‐flat on “eripies” (phrase 11), equal to E‐flat at the lower level of transposition. This pitch may be found in all sources that distinguish between b‐natural and b‐flat; it is indicated with a parenthetical b‐flat in the base reading of Ben 34. The F's that appear on “eripes,” moreover, would be low B‐flats if written a fifth lower. For the scribe of Mo 159, a notation of the whole offertory at the affinal position is not an option because the first verse uses a b‐natural on “liberator” (phrase 8), which would be an f‐sharp at the higher position.

Although a preferred reading of this offertory was not possible to establish with certainty, the move into the authentic range for the second verse may be an emendation made in response to the nondiatonic pitch at the end of the verse. If so, the plagal reading of the second verse would be the preferred one. In many diverse later MSS, such as Pa 1235 and RoV 52, the respond and both verses are written in the transposed plagal range. The Beneventan and Aquitanian sources, however, attest to the inconsistencies and perhaps ad hoc nature of the notation. In two early sources, Ben 38 and 40, the second verse appears to be written as the same pitch level as the first verse. In Ben 34 and Ben 39, the respond and first verse are in the untransposed plagal range, whereas the second verse is in the authentic range, a pitch‐level profile confirmed by the repetendum cue. The scribe of Ben 35, the latest of these sources, undertakes a unique emendation, in which the second verse begins in the plagal range but joins the majority reading before the problem spot at the end of the verse, with a transition in which the melody is briefly a third below the majority reading (on “[insuren] tibus me”) and joins on “exaltabis.” Although the unique reading of Ben 35 supports a hypothesis that the higher reading of the verse is an emendation, the transitional bridge attests to a concern for the actual sound of the melody, suggesting that the scribe's ad hoc solution to the notational problem in the second verse had implications for performance.

Among Aquitanian MSS, the repetendum cues are inconsistent and may attest to a variety of practice. In Lo 4951, Pa 780, Pa 1134 and Pa 1135, the cue suggests a performance in the authentic range, whereas in Pa 1132, a later MS from St. Martial, it points to a performance level in the plagal range. A few Aquitanian MSS hint at a further nondiatonic practice in the pretheoretical tradition, one that is not possible to reconstruct securely. If Pa 776 is interpreted as beginning the second verse in the authentic range, the repetendum cue appears to be notated a tone too high, beginning on G rather than F. For several other offertories, such as Oravi deum and Domine convertere, I and others have hypothesized that similar evidence in Pa 776 implies a performance of the verse a tone below its written level.1

The variety of versions points to the problems inherent in the assumption that there is always a single authentic version of the melody that can be reconstructed.

While a plagal reading of the verse emerges as the most probable unemended reading in the Beneventan sources, an authentic‐range performance of the verse is strongly implied in many Aquitanian MSS, including Lo 4951 and Pa 780. For these reasons, selecting a source to serve as a base reading for Intonuit de caelo is especially difficult. Given the inconsistencies among MSS from the same regional tradition, one wonders whether a higher or lower performance of the verse or its ending was the choice of individual cantors. A move into the authentic range for the final verse may have been undertaken not only in response to the need for emendation but also for aesthetic reasons, since it is stylistically consistent with the genre as a whole.

(p.328) Second verse in authentic range: Mo 159, Pa 780 (confirmed by repetendum), Lo 4951 (confirmed by repetendum), Pa 1134 (confirmed by repetendum), Ben 34, Ben 39, Mod 7, Pst 120, Pad 47, Cai 61, Be 40078, Tri 2254, Rei 264, Gr 807.

Second verse in plagal range (or whole offertory in transposed plagal range): Ben 38, Pa 1132 (confirmed by repetendum), Pa 1136 (written level; no repetendum), Pa 1121 (written level; repetendum starts on new line with no custos), Pa 1235, RoV 52.

Other melodic variants: Like other deuterus chants, Intonuit exhibits many variants around the pitches E and F, and in the second verse, b and c. The Beneventan tradition differs from the majority international reading in the following places.

  1. 1. respond, “alleluia”: In most non‐Beneventan MSS, the segments of the melisma in boldface below are notated a tone higher, as in the reading of Mo 159: “alLEluia”: E FaGFF(ori.) EG (quil.) aGa FG (quil.) aG aF GE aGG(ori.)F GaG.

  2. 2. v. 1 “[refugium] MEum”: most core MSS have a reading very similar to that of Mo 159: FF GF GFF DF(quil.)GF(quil.)E. Some peripheral MSS exhibit further variants, as indicated below.

  3. 3. “iraCUNdis”: This melisma is a point of variance. In most traditions, it is slightly shorter than in the Beneventan sources:

    • Mo 159: Gc dcc(ori.)G cdc db cba(liq.)

    • Pa 776: acdcca cdccba(liq.)

  4. 4. v. 2 “genTI/BUS”: Mo 159: a/aba

Individual MSS:

Lo 4951:

  • respond, “alLEluia”: shortened melisma. The second statement of EFGFG is omitted.

  • v, 2 “inIquo”: ca cb

Pa 1235 (respond and v. 1 are a fifth higher):

  • respond, “alTISsime”: ded de

  • respond, “alLEluia”: c bcedc cb cdede cdedec d dc ed dc ded

  • v. 1 “reFUgium”: c cc dfed db cdededcd

Cai 61: respond, “intonuit…dominus”: Cai 61 is unique among the MSS examined in notating the opening, “intonuit de ce[lo],” a fourth below the majority reading, and “[ce]lo dominus” a whole tone below. The reasons for this transposition are not clear. The whole‐tone transposition at “dominus” results in intervallic differences. In MSS that specify b‐flat, a b‐flat is indicated, resulting in a passage that articulates the minor third G‐b‐flat. In Cai 61, the referential interval is the major third F‐a. It is not clear whether this unique reading points to a nondiatonic practice.

  • respond, “alTISsimus”: GaG Ga

  • respond, “alLEluia”: E GaGF EG aGa FGaG aFGE aG GF GaG

  • v. 1 “reFUGium”: FF G b‐flat aG GE F GaG a

  • v. 2 “iraCUNdis”: ccc Gc dc cGc dccb cba(liq.)

Be 40078:

  • respond, “CElo”: GaG GEFED

  • respond, “alTISsimus”: GaGa

  • respond, “alLEluia”: FE FaGF FF FG(quil)aGa FG(quil)aGa aF GF aGGF aGaG

  • v. 1 [refugium] “MEum”: DF GF GFFD DGFEF

(p.329) Gr 807:

  • respond, “ET” [altissimus]: a

  • respond, “alLEluia”: FacaGFEF F FGaGa FGaGaF G (ori.)FE aGF aGa

  • v, 1 “reFUGium”: FF F(ori.)E GaGaGF

  • v. 1 [refugium] “MEum”: DG aG aE FD DF GFEF

  • v. 2 “GENtibus”: cd cdc c(ori.)bG bab cc ca ca c cdc(liq.)

In Italian MSS, G‐protus passages are often written a whole tone higher, consistent with the Guidonian recommendation for emending them. The sources, however, undertake this emendation in different places, as the variants below show.

RoV 52 (respond and v. 1 are written a fifth higher):

  • respond, “alLEluia”: bcedc ca Gacbc cdedecdc eddc ded

  • v. 1 “diLIgam” (a longer version of the melisma): bcbGa cc ca cccaG abcb abcb cdededcd

  • v. 2 “GEntibus”: melisma begins: cd cdcbcbaba (etc.)

Pad 47:

  • respond, “alTISsimus”: GaG Ga

  • respond, “alLEluia”: The segments equivalent to EFGFG EFGFG EFD in Ben 34 are a tone higher with a melodic variant at the end, FGaGa FGaG aFGF.

  • v. 1 “firmaMENtum”: G DGF

  • v. 1 [refugium] “MEum”: FF GF GF DEF GFE

  • v. 2 “ME/US”: melisma ends cdededcd/dc

  • v. 2 “GENtibus”: cdc bcb GaG a cc ca ba abc

  • v. 2 “iraCUNdis”: ac dc a cdc db cbG (liq.)

  • v. 2 “VIro”: a c b‐flat

  • v. 2 “E/RI/PI/ES”: aGaG/abcb/abcb/cG Gaba FaFa.

Mod 7 (has many of the same variants as Pad 47 above):

  • respond, “CElo”: GbabG FED

  • respond, “alTIS/SImus” [sic]: GaG Ga/FGF

  • respond “alLEluia”: F aba G GE FGaGa FGaG aFGF aGF GaG

  • v. 1 “domiNE”: FF FE DFEDE

  • v. 1 “firmaNENtum”: FF DGF

  • v. 1 [refugium] “MEum”: FF DF GFGF DEF GFE

  • v. 2 “ME/US”: melisma ends cdededcd/dc

  • v. 2 “VIro”: a c b‐flat

  • v. 2 “inIquo”: acbaG

Roman MSS: In v. 1, Bod 74 lacks the phrase “et refugium meum,” which is found in Vat 5319. The omission of this phrase is probably a simple error, and it is supplied from Vat 5319 in the transcription. In Vat 5319 and RoS 22, the whole offertory is notated at the affinal position. Following the second verse, Vat 5319 has a repetendum cue to “et,” which Bodmer lacks.

The only other significant difference between Roman MSS is the final alleluia, a point of variance in several of the Easter Week offertories. Bod 74 employs a deuterus alleluia melodically identical to that of Bod 74's In die sollemnitatis, which also shares material with the closing melisma of Terra tremuit. Vat 5319 and RoS 22 close with a tetrardus (p.330) alleluia, which is similar but not identical to that of Benedictus qui venit. Vat 5319: “AL/LE/LU/IA”: cb(liq.)/dcdcdcc (ori.) baGaG cdcdc bcba bcbab/ba(liq.)/GacbaG aG

RoS 22: “AL/LE/LU/IA”: d/dcdcdbaGaG cdcdcbcba bcbab/bb(liq.)/GacbaGa aG. RoS 22 exhibits two other small variants:

  • “CElo”: cdcb

  • “deDIT”: cc1.

Portas caeli (51)

There is one point of Beneventan distinctiveness in v. 1. Most Gregorian sources have “in legem,” accommodating the extra syllable with one or two c's.

  • In v. 2 there are small differences in the melisma on “meum” among both core and peripheral sources, as shown in the sampling below (differences from Ben 34 in boldface):.

  • Pa 776: cdcb ac b dcaGa cba baGa/aG

  • Mo 159: cedc bcb dcba(quil.)Ga cb(quil.)a baGa/aGa

  • Pa 1121: cdcb a cb dc baGa cba baGa/aG

  • RoV 52: bdcb acb edcaG acbabaGa/aG

  • Mod 7: cdcb cdcdcbaGa cbabaGa/aG (very similar in To 18)

  • Gr 807: cdcb c(ori.)a CaG aGFG cba caGa/aG

In v. 2, there are minor differences in the “loquar” melisma:

  • Mo is nearly identical to Ben 34

  • Pa 776: EGbaG EGacaaG FGaccaccaG G(ori)F GacaGFG (quil.)abaGa/aG

  • Pa 1121 is nearly identical to Pa 776

  • Be 40078: DGcaG DGacaaG DGa c ca bcaF acba cG/G (shortened melisma)

Lo 4951 presents a different reading in many passages:

  • v. 1 Shortened melisma on “MEam”: GFGaaccd ccd aGGFG(quil.)a b

  • v. 1 “oris mei”: acd cdcc(ori.)/ ba/ GaG/G

  • v. 2 “paraBOLIS”: dfffdc cd(quil.)eddc

  • v. 2: [os] “MEum”: shorter melisma: cdcbacbcbaGa/aG

  • v. 2 On “loquar,” Lo 4951 has a shorter melisma that presents problems because the horizontal alignment is not maintained between the melisma and the previous passage. Using the repetendum as a guide, the melisma reads: EGbaG EGacaaG FgG (quil.)ab

Mod 7: “ORIS”: acdcdc/cb

Gr 807:

  • Respond, “HOMO”: a cdc db c (ori.) a cba caGa/aG

  • (p.331)
  • v. 1 “aPERiam”: melisma has top note of f: Ga bc dc dbc dfc dfd

  • Roman MSS: There are very few differences between Vat 5319 and Bod 74.

  • Vat 5319:

  • respond, “APparuit”: DG

  • v. 2 “secuLI”: cbcdcbc

  • RoS 22: “alLEluia”: cdededc ededc cbdcbacbabaGa ced cdcba dcdcbcbacbab

In die sollemnitatis (52)

The modal and pitch‐level variance in this offertory was illustrated in Justmann's dissertation.1 In an essay by Theodore Karp, In die is employed as an example of methodology in dealing with difficult chants.2 The south German theorist known as John of Afflighem describes this chant as “much disordered” (multi confusi), and suggests that it can be more easily emended at the affinal position.3 In addition to the core and peripheral MSS used in this study, this discussion of the piece draws on Justmann's fifty‐five transcriptions of the respond.

Although this chant has a variety of closing pitches, most versions are protus, ending either on D or at the affinal position, on a. Karp's study analyzes the many small melodic variants between the different readings, undoubtedly resulting from the modal ambiguity of the chant and the need for emendation. In larger terms, however, the sources can be classified into two groups, represented by the two readings of the respond given in the transcription. In terms of overall pitch‐level profile, Mo represents the lectio difficilior, and Pa 776, here transcribed with a final of D, represents the majority version (though it is more often written at the affinal position). After “vestrae dicit” (phrase 1), Pa 776 proceeds unproblematically as a first‐mode melody. Mo begins and ends the respond at the affinal position, a fifth above Pa 776. The internal part of the chant, however, is written a fourth above Pa 776 and consistently employs a b‐flat, beginning with “dicit dominus” (phrase 1). The interval of transposition returns to a fifth within the final melisma, on “alleluia” (on the twelfth note).

At the position of the normal protus range, a fifth lower, the b‐flats in Mo would be E‐flats. Although the reading of Mo is a minority version, it is found in several Frankish sources and geographically disparate MSS, including Be 40078 and the thirteenth‐century Italian MS Rossi 76. I would cautiously hypothesize that Mo's pitch‐level profile reflects that of the unemended reading, simply notated at the affinal position. In the majority version, matching the pitch‐level profile of Pa 776, there is little sign of the “great disorder” described by John except on the final syllable a “sollemnitatis,” a problem spot discussed below.

(p.332) The nondiatonic pitch E‐flat implied in the version of Mo points to the modal ambiguity of the melody, which reflects a mixture of protus and deuterus characteristics. Typical deuterus melodic gestures include the figure on the last two syllables of “dominus,” that on “vos,” and, to a lesser extent, on “in terram.” In Pa 776, these melodic figures are written at the normal deuterus position, employing the semitone between E and F. In Mo 159, they employ the semitone between b‐flat and a; the latter pitch also serves as the final of the protus chant. Mo 159, then, evidently reflects a tradition in which these deuterus melodic figures were sung a whole tone below their usual position, with a semitone between D and E‐flat.

In some MSS, final cadence is altered to close on E or, at the affinal position, G or b. These altered endings are probably attributable to the modal ambiguity of the melody. The deuterus characteristics of many internal phrases evidently led to an alteration of the final cadence to close on E or b, the most common alternate finals.

I would argue that the melodic variants in the respond, documented in Karp's and Justmann's studies, occur at specific places and are almost always related to the need for emendation. By way of illustration, I would like to consider the first problem spot, on the last syllable of “sollemnitatis.” As noted, Pa 776 employs a common solution to the overall “modal disorder” of the melody, writing the problematic melodic figures involving E‐flat and D a tone higher, in their normal deuterus position, beginning at “dicit dominus.” In these versions, however, one problem remains. On the last syllable of “sollemnitatis” (phrase 1), Mo has a b‐flat, equivalent to E‐flat at the position of Pa 776. The melodic variants that arise here are clearly attempts to circumvent the E‐flat of the unemended reading. Pa 776's reading, E (flat)‐F, is matched in Pa 780 and Lo 4951. Pa 1121 and 1132 simply have two F's here. The scribe of Pa 903 recomposes the passage: the previous syllable, “sollemniTAtis,” is transposed up a whole tone, ac abc, and “tis” is Ga. According to Justmann's transcriptions, a similar solution is taken in Avignon 181; Madrid 51 and Madrid 18 match Pa 776 at “ta” but have Ga, a transposition of a third, on “tis.” Following this passage, the Aquitanian MSS present a largely unified tradition. Some of the French MSS in Justmann's transcriptions adopt other solutions. Angers 96, Angers 97, and Leningrad 6, written at the affinal position, have a‐c on “tis,” equivalent to D‐F at the lower level. This solution is also found in many German MSS, including Gr 807 and others in Justmann's large sampling. Orleans 119 has FG on “tis,” a reading a whole tone above Pa 776, and remains a whole tone above Pa 776 through “vestrae,” joining on “dicit” (a version also found in Beneventan MSS and others summarized below).

The clear association of variants with nondiatonic pitches suggests that the varied transmission of this offertory is attributable to nondiatonic pitches rather than melodic instability per se. The scribes introduced small melodic variants by necessity, as ad hoc solutions to the problem of notating the chant within the diatonic framework. Among certain late German and Italian MSS in Justmann's large sampling, such as Munich 2541 and Salzburg 20, the final melisma on “alleluia” is also altered, followed by a reworked final cadence. These more extensive reworkings contrast with the consistent readings of the final melisma found in earlier MSS. In some sources, the melisma on “fluentem” is altered as a way of circumventing low B‐flat. In some Italian MSS, for example, the notes of the final melisma corresponding to Pa 776's F‐D‐C‐B(flat) are F‐E‐D‐C (Modena 0.1.3) or G‐F‐D‐C (Vercelli 162). In Beneventan MSS, by contrast, the b is simply left as is.

The pitch‐level profile of the sources is summarized below. Because the problem spots of this offertory are primarily in the respond rather than the verses, many more sources can be examined in a study of the problem spots. To give a broader picture beyond the MSS (p.333) normally used in this study, the following summary of pitch‐level in the respond incorporates selected MSS transcribed by Justmann, based on his transcriptions.

  1. 1. Pitch‐level profile of Pa 776, at the affinal position: Ben 34 (2↑ at [sollemnita] “tis vestrae”), Provins 12 (2↑ at [sollemnita] “tis vestrae”), RoV 52, Pst 120, Cai 61, Rei 264, Trier 2254 (2↓ briefly at “lac”), Pa 17310, Angers 96 and 97, Pst 119, Vercelli 161, Rossi 231, Wolfenbüttel 40 (2↑ at [sollemnita] “tis vestrae”), Namur 515.

  2. 2. Pitch‐level profile of Pa 776, at the normal plagal position (ending on D): [Pa 780, Pa 1121], Ben 39 (2↑ at [sollemnita] tis vestrae], Gr 807, Pa 903 (2↑ and 3↑ at [solemni]tatis), Brussels 3823, Orleans 116 (2↑ at [solelmnita]“tis vestrae”), Pa 1890, Modena 0.1.3, Vercelli 162, Milan S 74, Melk 109.

  3. 3. Pitch‐level profile of Pa 776, with an altered final cadence:

    1. a. Pa 1132 (affinal position): on G; last syllable of “alleluia” is changed to aG (also in Graz 1655).

    2. b. Lo 4951, Madrid 18, and Madrid 5: on E; written a whole tone above Pa 776 beginning with the ten notes of the melisma on “alLEuia”.

    3. c. Ben 39: On E; last syllable of “alleluia” is changed to ED.

    4. d. To 18 (affinal position): on G, with a substantially different melisma on “alleluia.”

    5. e. Pa 1669 (affinal position): on b; written a whole tone above the majority reading beginning with the last fifteen notes of the melisma on “alLEluia.”

    6. f. In Justmann's transcriptions, Munich 2541 and Salzburg 20 have similar deuterus endings, preceded by shortened and greatly varied melismas on “alleluia.”

  4. 4. Pitch‐level profile of Mo 159: Paris 904, Brussels 3824, Angers 96 and 97, Be 40078, Reims 224, Rossi 76.

  5. 5. Unique reading: Leningrad 6, at the level of Mo until the last syllable of “inducam,” presents a variant reading of “vos,” a whole tone higher.

Roman MSS: Bod 74 lacks the second verse for this offertory. The respond and first verse show no significant variants from Vat 5319. RoS 22:

  • “VEstre”: EGFEGFE

  • “TERram”: FEFEDC (etc.)

Erit vobis (53)

A first illustrated by Bomm, this offertory exhibits a great variety of modal characteristics.1 Its diversity of transmission is probably attributable to a mixture of factors, including nondiatonic pitches, modal ambiguity, and perhaps a resulting lack of clarity in the aural tradition. I have not been able to determine a single preferred reading or propose explanations for all the variants. The commentary that follows and the supplemental examples provide several possible solutions to the problems.

Because MSS in the core group show such diversity of pitch level in the respond, selecting a base version for the transcription proved especially difficult. The ending on G in the transcription of Pa 780 is matched in most pitched MSS. The Beneventan sources (p.334)

Appendix 3 Commentary on Edition

Example 53.1

present a highly emended tritus version of the respond that is matched in two German MSS, Gr 807 and Be 40078. In Mo 159, a lengthy passage of the respond is probably emended. As described below, two of the Aquitanian sources in the core group, Pa 776 and Pa 1121, exhibit too many inconsistencies with the other MSS to provide a reliable reading. Pa 780, chosen as the base version, presents most of the respond and first verse at the probable unemended pitch level, closing on G. As described below, the variants imply that F‐sharps were employed in phrase 3.

1. The opening passage, “erit vobis hic dies”: Four versions of the opening passage are shown in example 53.1. With the assumption that the intended starting pitch in the Aquitanian MSS is D or E, as in most pitched MSS, all Aquitanian MSS present a reading similar to that of Mo 159. I have used this majority reading as a guide to determining the opening pitches of Pa 780 in the transcription. Although similar readings of “erit vobis” are found in most MSS (except in Benevento), variants begin at “hic dies.” The alternate use of b‐natural and b‐flat in Mo 159 (line 1) exemplifies the problematic modality of this offertory. The b‐flats in the opening passage are matched in the other MSS that distinguish between b‐natural and b‐flat. The result is a mixture of G‐protus and tetrardus characteristics. In the Beneventan manuscripts, represented by Ben 34 (line 2), the passage is tritus in quality throughout, cadencing on F. Pa 1235 (line 3) presents a hybrid version, with an opening that matches Pa 780 and Mo 159, but an F‐tritus cadence on “dies.” A similar reading is found in Cai 61.

Since Mo 159 employs a b‐natural on “dies,” the tritus and tetrardus versions of the end of the passage are intervallically identical. The Beneventan version, however, differs intervallically from the majority reading on the opening “erit vobis,” articulating a major third, F‐a, rather than the minor third, G‐b‐flat, found in Mo and most MSS. It is possible that the difference in intervallic structure points to the use of a nondiatonic pitch, perhaps a‐flat at the level of Ben 34. The hybrid readings of the passage, such as that found in Pa 1235, would perhaps support this hypothesis. The variants, however, may also simply reflect the modal ambiguity of the piece as a whole. Since the tritus reading of the opening is confined to Beneventan MSS, perhaps the opening is simply an adaptation that brings the opening of the melody into conformity with the tritus ending in the Beneventan sources (p.335)

Appendix 3 Commentary on Edition

Example 53.2a

(see example 53.2). Be 40078 (line 4) gives an alternate reading of the opening passage that is matched in Rei 264, Tri 2254, To 18, and Mod 7.

2. Cadence on the first “alleluia” (phrase 1): This cadence is melodically variant. Pa 780's reading FGaGF/GF is the majority reading. Mo 159, Pa 1235, and Cai 61 have FGa/GF.

3. Cadence on “domino” (phrase 2): The deuterus cadence of Pa 780 is matched in all Aquitanian sources, all Beneventan sources, Pa 1235, Cai 61, Pst 120, and all German sources except for Be 40078. “Domino” concludes on F in Be 40078, Mod 7, RoV 52, and To 18.

4. “In progenies” (phrase 3) to the end of the respond: Here the sources show an extraordinary diversity in their choice of F or G as a focal pitch, beginning with “vestris/vestras.” A selection of various readings is given in example 53.2. The sources in the example (p.336)

Appendix 3 Commentary on Edition

Example 53.2b

were chosen to illustrate the very different approaches to emendation. Pa 780 adopts G as a focal pitch throughout, a reading also matched in Lo 4951. The Beneventan sources, by contrast, give a tritus reading, consistently adopting F as a focal pitch; this version of the ending is consistent with the tritus characteristics of the opening in Beneventan sources. This tritus reading is matched in some German sources, such as Rei 264. Most other MSS in the example alternate between F and G, moving between them at different points.

The pitch‐level variants are probably attributable, at least in part, to the use of a nondiatonic pitch. The tritus and tetrardus versions differ in the pitch below the final, which is the semitone in tritus and whole tone in tetrardus. In this passage, the semitone below the recitation pitch appears on “diem” and in the first and second “alleluia.” If the version on G is taken as the “unemended” pitch level, the downward transposition to F (p.337)

Appendix 3 Commentary on Edition

Example 53.2c

would suggest that an F‐sharp is required at this point, indicated parenthetically in the transcription of Pa 780. If the version on F is taken as the “correct” pitch level, an upward transposition to G would suggest that an E‐flat is intended at the lower level. The first alternative is more plausible for several reasons. The respond closes on G in the great majority of sources. With the exception of Be 40078 and Gr 807 (which end on F) and Rei 264 (which closes on E), most non‐Beneventan sources close on G. These sources move the referential pitch from G to F and back to G at various points. Mo (line 2), for example, moves to F at “legitimum,” whereas Cai 61 (line 5) moves on “sempiternum.” Nearly all MSS join on G in the last “alleluia,” where the hypothetical F‐sharp is no longer required. The general stylistic traits of the chant repertory further argue in favor of the version on G with an F‐sharp. This passage consists of melodic material that is typical of the chant repertory. Normally it is placed between the pitches F and D or a and c, with a semitone below the recitational pitch. The placement of this material a tone higher than usual in Erit (p.338) vobis, between the pitches E and G, suggests that an F‐sharp is required. The reading of Tri 2254 (line 6) is instructive: the focal pitch is G throughout, except for a downward transposition at “diem,” where the irregular semitone is required. In the first and second “alleluias,” where the irregular semitone occurs at the level of Pa 780, the pitch F (sharp) appears as G in Tri 2254, reflecting the German chant dialect. An emendation similar to that of Tri 2254 is suggested in the Frutolf tonary. While the need for F‐sharp is probably the primary reason for the diverse versions, the variants may also reflect inconsistencies in the aural tradition, attributable to a lack of modal clarity throughout the respond and verses.

Although the latter solution is reflected in the base transcription, I would like to explore yet a third possible solution to the problem. One key MSS in the core group, Pa 776, presents a reading not found in the other sources, with the possible exception of Pa 1121, which is difficult to decipher because of inconsistent heightening at the place in question.2 With the assumption that the starting pitch of the respond is D or E, these two versions close on a. Pa 776 is shown in the line 7 of example 53.1. The last “alleluia” is marked by a shift of recitation pitch from G to a. This reading of the passage follows the pitch‐level profile of Mo a whole tone higher, with the same upward shift of focal pitch for the last “alleluia.” Without chromatic alteration, however, an ending on the protus pitch a would result in an intervallic structure different from that of other MSS. The ending of nearly all MSS on G, without a b‐flat, argues against a protus quality for the closing passage and in favor of a tetrardus ending. Another possible solution, then, would be to assume a closing pitch of G and transcribe this version down a whole tone in toto. This reading would result in an opening passage that partly matches Ben 34 and partly Pa 1235. On balance, however, Pa 780 seems the more likely preferred reading.

  1. 1. Pitch level at “in progenies vestras”:

    • = Pa 780: Pa 776, Pa 1121 (?), Lo 4951, Pa 1235, Cai 61, Tri 2254, Rei 264, Gr 807

    • 2↓ Pa 780: Mo 159, Ben 34, Ben 35, Ben 39, Pa 1132, Be 40078, Pad 47, Mod 7, Pst 120, To 18, RoV 52

  2. 2. Pitch level at “legitimum”:

    • = Pa 780: Pa 776, Pa 1121 (?), Lo 4951, Cai 61, Gr 807 (starting on the third syllable)

    • 2↓ Pa 780: Mo 159, Ben 34, Ben 35, Ben 39, Be 40078, Tri 2254, Pad 47, Mod 7, Pst 120, To 18, RoV 52

  3. 3. Pitch level at “sempiternum diem”:

    • = Pa 780: Lo 4951, Tri 2254 (2↓ at “diem”),

    • 2↓ Pa 780: Mo 159, Ben 34, Ben 35, Ben 39, Be 40078, Rei 264, Cai 61, Pa 1235, Pad 47, Mod 7, Pst 120, To 18, RoV 52

  4. 4. Pitch level at “alleluia” 1 and 2:

    • = Pa 780: Pa 776, Pa 1121 (?), Lo 4951, Tri 2254

    • 2↓ Pa 780: Mo 159, Ben 34, Ben 35, Ben 39, Be 40078, Rei 264, Gr 807, Pa 1235, Cai 61, Pad 47, Pst 120, To 18, RoV 52

  5. (p.339)
  6. 5. Pitch level at “alleluia” 3:

    • = Pa 780: Mo 159, Lo 4951, Cai 61, Tri 2254, Rei 264, Pa 1235 (joins within the melisma on the second syllable of “alleluia,” when the melisma reaches c), Pad 47 (reading similar to Pa 1235), To 18 (reading similar to Pa 1235), Mod 7 (also joins within the melisma on the second syllable of “alleluia,” but earlier, on the eighth note), Pst 120 (same reading as Mod 7), RoV 52 (same as Pad 47).

    • 2↓ Pa 780: Ben 34, Ben 35, Ben 39, Be 40078

    • 2↑ Pa 780: Pa 776, Pa 1121 (?)

The verses present further problems in pitch level. The Aquitanian MSS differ in the order of the verses, and most include a repetendum cue only after the final verse. Although the Aquitanian MSS present no large‐scale partial transpositions in relation to Mo 159, the repetendum cues, when present, give conflicting information about the intended pitch level of the verses in relation to the respond. Outside the Aquitanian tradition, moreover, partial transpositions may be found in several sources.

In the transcription of Pa 780, I have derived the pitch level of the second verse, In mente, from the repetendum cue, resulting in a reading a fourth below the pitch level of Mo 159. (The written level is relation to the second verse is a fifth below Mo 159). Several F‐sharps are required to replicate the interval structure of Mo at this position. It is not clear, however, that this version is the preferred reading. In Ei 121, the scribe writes the letters “lm” at the verse opening, indicating a moderate heightening of pitch consistent with the majority reading of Mo but not with Pa 780. In the transcription, I have provided the version of Mo as an alternative. Because the verse Dixit dominus lacks a repetendum cue and because its relationship to the second verse is not always a reliable guide, its intended pitch level is unclear. I have tentatively transcribed it at the level of Mo 159, the majority reading. The summaries of pitch level below are provided in relation to Mo 159.

Beginning of v. Dixit Moyses: In the majority of MSS examined, the verse is notated at the same pitch level as Pa 780. These include the late German MSS, along with Cai 61 and Pa 1235. As mentioned, however, the repetendum cues in Aquitanian MSS, when present, give conflicting information about the intended pitch level of the verse. Three of the four Italian sources in the sampling, moreover, transmit different readings, joining the majority reading of Mo in pitch level at “estote.” These differences in the Italian MSS may be attributable simply to the modal ambiguity of the piece as a whole.

Pitch level of the verse Dixit dominus in relation to Mo/780:

  • = Mo 159: Ben 34, Cai 61, Pa 1235, Pst 120, Gr 807, Tri 2254, Be 40078

  • 4↓ Mo 159: Pa 1121 (according to repetendum)

Distinctive or unclear readings:

  • Mod 7: Verse begins 3↓ Mo 159, joins at “estote”; the words “bono animo” are missing.

  • RoV 52: Verse begins 3↓ Mo 159, presents a melodically variant reading of “animo estote,” joins at “veniet vobis.”

  • Pad 47: Verse begins 2↑ Mo 159, presents a melodically variant reading of “animo estote.” The passage that begins “et veniet vobis” is 3↑ Mo (like To 18 below), joining “deo.”

  • To 18: For reasons that are unclear, To 18 presents the passage that begins “et veniet vobis” 3↑ Mo 159, joining in the melisma on “deo.”

  • (p.340)
  • Pa 776: The written and performance levels are unclear because of the problematic ending of the respond and because verse lacks a repetendum cue.

  • Lo 4951: The written level throughout is a 2↑ Mo 159, but there is no repetendum cue to confirm the written level.

Position of In mente verse in relation to Mo 159 (with variants at “de terra et de domo”):

  • 2↓ Mo 159: Pa 776 (according to repetendum)

  • 4↓ Mo 159: Pa 1121, Pa 780 (3↓ at “de domo”)

  • 5↓ Mo 159: Pa 1235 (mostly); Mod 7 (joins Mo at “in quo existis”)

  • = Mo 159: Lo 4951, Ben 34 (2↓ Mo at “de terra”; joins at “manu”), Ben 39 (same as Ben 34) Cai 61, Gr 807, Re 264, Be 40078 Tri 2254

  • = Pa 780: Pa 1121, Ben 35, Pa 1235

  • 4↑ Pa 780: Ben 34, Mo 159, Pa 776, Lo 4951, Cai 61, Gr 807, Be 40078, Rei 264, Pst 120, To 18

Emended and distinctive versions: Tri 2254 is 6↑ Pa 780, until the melisma on the final syllable of “istum,” where the interval of transposition changes to a fifth. Mod 7 begins at the level of Pa 780 but is a fifth higher beginning at “in quo existis” (where the low B‐flat would be required). RoV 52 begins at the level of Pa 780 but departs from it at “in qua [sic] existis.” The pitch‐level relationship to Pa 780 is as follows: In qua 5↑; existis 3↑; egypti 4↑; in manu 3↑; potenti 4↑.

From after the initial phrase, these four sources diverge. Beginning with “in quo existis,” Tri 2254 joins at “in quo,” whereas Pa 1235 and Pa 780 remain a fifth below Mo 159 throughout.

Melodic variants: Melisma on “deo”: In sources from German‐speaking areas, the fourth note of the melisma is b‐flat rather than b‐natural, as in Mo 159.

Roman version: Bod 74 is the only Roman source to transmit the verses of Erit vobis. Vat 5319 and RoS 22, notated at the affinal position, are nearly identical to Bod 74 until the closing alleluias, where they show a rare significant difference. Vat 5319 and RoS 22 present a version more florid than Bod 74's. Their version is nearly identical to the closing “alleluia” melisma of Benedictus qui venit (54), thus closing on G rather than E. While the variants in the Roman tradition could reflect practices at the different institutions they represent, it is also possible that they attest to further melodic stereotyping in the tradition that took place between the copying of Bod 74 and Vat 5319.

RoS 22: “AL/LE/LU/IA”: G/aG/cd cdcdccbaGaG cdcdcb acbab/bb (liq.)/GacbaGaG. RoS 22 shows some variants from the others in the last cadential segment of Formula B. The segment GFEFG in 5319 and Bodmer is consistently GaGFEFG in RoS 22: on the first “alleluiA,” “celebrabiTIS,” and “domiNO.”

Benedictus qui venit (54)

In most MSS the order of the two verses is 1. Lapidem 2. Haec dies. Beneventan and Roman sources have the two verses in the opposite order.

b‐flats: The MSS that indicate b‐flats do not do so in the same places, perhaps indicating an inconsistency of practice. The parenthetical b‐flats in the base transcription are derived (p.341) from their use in Mo 159. Later MSS are not always in agreement with Mo or with one another. Be 40078 indicates a b‐flat at “rePRObaverunt” that is lacking in Mo 159. Cai 61 employs b‐flats at “haec dies” in v. 1. The later MSS, however, do not indicate b‐flat on “domiNI” in the respond.

Editorial emendation: In the melisma stated twice on “exulTEmus” and “laeTEmur,” a small segment of the melisma, indicated with brackets in the base transcription, is written incorrectly the first time in Ben 34. Beginning with the fifth note, the scribe writes babG instead of cbca. The status of this passage as a scribal error is evident from the other Beneventan sources and from the second statement of the melisma, where it is written correctly. I have emended this passage in the transcription.

Beneventan distinctiveness: respond, “vos”: Most MSS have “vobis” instead of “vos,” a Gallican psalter adaptation, with a melodically identical reading. The syllables of “vobis” are distributed FG/acaGa.

The Beneventan sources show small differences from the majority of other MSS in several of the melismas. In the following passages, Mo presents a more common reading. Some of the more significant differences are shown in boldface.

  • v. 1 “edifiCANntes”: Mo 159: eg de (quil.)f gd fff(ori.)d ff gdc de(quil.)f

  • v. 1 “anGUli”: Mo 159: de(Quil.)f ge fdc

  • v. 1 “exulTEmus” and “laeTEmur”:

Beneventan MSS present a shortened version of the melisma, which is also found in Lo 4951. The same melisma, with the same variant, occurs in the second verse of Domine deus salutis. All Beneventan sources omit a section of the melisma that is shown in boldface in the sample readings below. With the exception of Lo 4951, the Aquitanian, French, German, and Italian sources present very similar readings of the melisma, as shown in the sampling below (differences from Ben 34 are shown in boldface). The early adiastematic sources in the sampling also transmit the majority version. The reasons for the isolated presence of this shorter form of the melisma at Benevento and Toulouse (Lo 4951) are unclear.

  • Pa 776: cdca cbcaFG FG(quil.)aGaG a ccca ccdc aaF FG(quil.)aGaG babGF GacaF FG(quil.)aGa

  • Be 40078: cdcG cbcaFG FG(quil)aG aGa cc ca cc dca aF FG(quil.)aG aG ca caF GacaF FG(quil)aGa

  • Pa 1235: ac dca cbcaFG FGaGaG ac ccacbdc a aF FGaGaG babGF GacaGF GaGa

  • Mod 7: cdca ababGFG FGaGaG acc ca cc dc aF FGaGaG babGF Fa ca FGaGa

  • Cai 61: Gc db G cbcGGFG FGaGaG acc ca cc dc a aF FGaGaG babaF Ga caGF GaGa

As mentioned, the second verse in Mo 159 is written a fourth above its position in most MSS. Mo 159 employs a b‐natural in the second statement of the melisma, suggesting an F‐sharp at the lower level. This pitch, however, is indicated only in the second statement of the melisma. The two statements are otherwise identical. I am nevertheless hesitant to attribute the difference to an error, since the higher notation of the verse, not found in most sources, suggests the presence of a chromatic inflection.

Mo 159, “ExulTEmus”: (all b's are flat unless indicated): fgfc fe fdbc bc(quil.)dc dc dfff(ori.)d ff gfdd(ori.) bc(quil.)dc dc ed ecb cd fd (ori.) bc(quil.)dcd

“LaeTEmur”: fgfc fe fdbc bc(quil.)dc dc dfff(ori.)d ff gfdd(ori.) b‐natural! (quil.)dc dc ed ecb cd fdd(ori.) bc(quil.)dcd

(p.342) Individual MSS: Other than those described above, there are very few variants among core MSS. The Italian MSS present a reading close to that of the core group. Small variants in peripheral MSS include the following.

Be 40078:

  • v. 2 “rePRObaverunt”: b‐flat

  • v. 2 “Edificantes”: dc c adc ed cc aca

  • v. 2 “CAput”: d cdede

  • Gr 807: v. 1 “DIes”: FGaGaGF

Pa 1235:

  • respond, “domiNI”: no b‐flat indicated

  • v. 1 “DI/ES”: FGaGaGFG/GF

  • v. 1 “exulTEmus”: see above

  • v. 2 “edifiCANtes”: df fdf fd fffed f fe(liq.)

Cai 61:

  • respond, “domiNI”: no b‐flat indicated

  • v. 1 “ea”: melisma begins ac caG caG GF (etc.)

  • v. 2 “Edificantes”: dcc cdc edc cba ba

  • v. 2 “edifiCANtes”: fgd fgd ff fd c de fe(liq.)

Pst 120: v. 1 “reprobaveRUNT”: cfed

Roman sources: Vat 5319 presents a much shorter version of the respond text, lacking “benediximus vos de domo domini.” RoS 22 has a slightly less ornate version of the final melisma: “AL/LE/LU/IA”: aGcdcb/acbab/bb/GacbaGaGG.

Deus deus meus (55)

The Beneventan sources exhibit two small differences from the majority of other sources. In a segment of the final melisma of v. 2, the Beneventan version is missing the notes in boldface:

  • Mo 159: cca FG ab(quil.) cb(quil.) a FGbaG EF aGF etc.

  • Ben 34: cca FG baG EF aGF etc.

The Beneventan MSS also differ here: v. 2 “quia [factus es]”: Mo has DEF FGEF.

The final melisma is subject to small variants in some peripheral MSS, as listed below. There are otherwise very few significant variants.

Pa 1121: v. 2 “tuA/RUM”: acaGacaGba/b

Gr 807:

  • respond, “ET” [in nomine]: CFE

  • v. 1 “TUam”: FFFD CFDCD FFFD FGF GDEFD

  • v. 2 “QUIa”: DCFGaG

  • v. 2 “tuArum”: GaGF acaG ccc

(p.343) Pst 120:

  • v. 2 “exulTAbo”: ac ca aGaF Gac ca aGaF GabcaGa FaG c ca FGa baG GabaGEF aGFEF FGaGa (etc.)

Mod 7:

  • v. 2 opening, “INMa/TU”: Ga/F/G

  • v. 2 “exulTAbo”: ac ca aGaF Ga c ca aGaF abcbaGa FaG c ca aGF cba FG baG EF aGE F (etc.)

Pad 47:

  • v. 2 opening, “INMa/TU”: Ga/F/G

  • v. 2 “exulTAbo”: c ccaaFaF FG c cabGaF Ga caFa FaG cca FabaG EF aGF EF aG DF FGaGa caG (etc.)

To 18: This MS presents some variants not found among the other Italian sources in the sampling:

  • respond, “VI/GIlo”: DEDC/DEDC

  • respond, “alLEluia” : CFFGFGFCD FFFE FGFG

v. 2 “quia factus es adiutor”: This section migrates upward and is mostly a third above the majority reading of Ben 34, joining again at “meus.”

v. 2 “exulTAbor”: Shortened and varied melisma: acbcaG GF Ga cbcaFG FaG abGF ac c ca FGa cba FG baG EF aGF

Roman MSS: The Roman version of this offertory is unusual in corresponding most to the Gregorian version in its final verse. In both traditions, the final verse has the traits of mode 1. In typical fashion, however, the Roman melody as a whole is more uniform than the Gregorian. The two verses are nearly identical, beginning with a typical first‐mode gesture and continuing with repeated statements of the formula associated with mode 1. The first verse contrasts with the second‐mode characteristics of the Gregorian, lacking melodic resemblance.

Aside from the use of liquescences and the oriscus, there are very few variants between Vat 5319 and Bod 74. RoS 22 shows some minor variant readings and a substantial variant in the final melisma. In RoS 22, the ending of this melisma is very similar to that of the final alleluia in Terra Tremuit (48). In Bod 74 and Vat 5319, by contrast, the Deus deus melisma begins as the Terra tremuit melisma but continues differently. The greater similarity between the two melismas in the latest MS, RoS 22, may reflect further melodic stereotyping that has taken place in the tradition between the copying of the earlier sources and the copying of RoS 22.

Vat 5319: respond, “meAS”: FaGFEFF(ori.)ED

RoS 22:

  • “DEus”: EFD

  • “te”: ED

  • “luCE”: GF aGabaGFGF

  • “alLE/LU/IA”: GF FaGaGF aGaGF aGF GabaGFG/GG/EFGFaGFEFE

(p.344) Lauda anima (56)

This piece exhibits very few significant variants between MSS. In a pattern evident in most deuterus pieces, variants at the semitone E‐F are prevalent. The Beneventan reading employs E to a greater extent than other MSS do. In a similar sort of variant, the Beneventan base version departs from other MSS in one small variant in the opening syllable: “LAUda” is CDGE in Ben 34 and CEGE in most other MSS.

v. 2 “Tuus”: The text underlay indicated in Ben 34, though matched in other Beneventan MSS, differs from other core and peripheral MSS. The majority reading gives the first eight notes of the melisma to the first syllable.

v. 2 “Sion”: The Beneventan MSS have a different text underlay from other sources. In most MSS, only the first four notes are given to the first syllable.

Lo 4951:

  • v. 1 “FAciet”: CEGFG

  • v. 2 “saeCU/LI”: EGFE/EGFGFFE

Cai 61: The two verses have written‐out repetenda to “alleluia.”

Pa 1235:

  • respond, “AL/LEluia”: DF GF/FD FFF(etc.)

  • v. 1 “IN” [seculum] FGFGF FD GaG

Mod 7:

  • v. 1 “IN” [seculum]: FF GF aGF GaG

  • v. 1 “DAT”: FG aGa

RoV 52:

  • respond, “DEo”: FGFE

  • respond, “DIuero”: EFG aG GF

  • v. 1 “cuSTOdi”: EFG

  • v. 1 “DAT” GaGa

  • v. 2 “deus TU” (rather than “tuus”)

Pad 47

  • v.1 “IN” [saeculum] FF GF aGF GaG

  • v. 1 “DAT”: FG GFG

  • v. 2 “SAEculum”: A different text underlay, with melodic differences indicated in boldface: ca c ca ccaGb Ga c ca ccaGb acdc bc dcb ca c ca acaG Ga cbG a c cacba/c cac GFE/FGFE

Be 40078:

  • respond, “meA”

  • “alleLU/IA”: EGEF/FE

  • v. 2 “SAEculi”: melisma ends bcaG acGE GbcaG Gc ca cGE

Gr 807: (p.345)

  • respond, “alleLU/IA”: EGEF/FE (same at v. 1, “saeculum”)

  • v. 2 “SAEculi”: c c(ori.)a bcaG cGa bca bcaG c Gc dc ccedc cc cc ca bcaG acbG Ga cbG Ga bcacGE

Roman MSS:

In Vat 5319, the whole offertory is written at the affinal position. There are otherwise no significant variants in RoS 22 and Vat 5319.

Confitebor domino (57)

Although Confitebor domino has very few significant variants among core or peripheral MSS, there is one passage where the Beneventan MSS differ from other traditions: in the melisma on the final syllable of “letabitur” (phrase 11). This melisma, which echoes “pauperis” in the respond (phrase 3), is found in all Beneventan sources but not in other traditions, where the final syllable of “letabitur” is accommodated with a single note, G.

Variants in individual MSS:

  • Pa 1121: v. 2 “eam”: The end of the melisma appears to be written a tone below the majority version: Facbdc aa(ori.)Ga abaG bGE FGE FGF bbbGF Gb. This reading is not matched in the other MSS from St. Martial.

Pa 1235:

  • v. 2 “DOmine”: ac dcd

  • v. 2 “salvum me”: recitation note G

  • v. 2 “haec”: Ga cdc

Cai 61: v. 2 “SERvus”: top note of f rather than g: aG a cc de fe ec de fe (also in Be 40078)

  • Pad 47 v. 2 “SERvus”: aGaccdeged cde gege

  • To 18:

  • v. 1 “MEus”: cdedc cba dc

  • v. 1 “Eam”: Fga dcdcaG aGa acba caFG aF GaGaGFG cccaGa

Roman MSS:

Vat 5319:

  • v. 1 [manus] “TUa”: ad dedcdc

RoS 22:

  • “confiTEbor”: abG

  • “DO/MIno”: cbacba/GbaGFG

  • “meO”: aGbaGaG

  • “MEdio”: aGaba

  • “PAU/pe/RIS”: acbacba/Gacba cbaGF GbaGaG

  • “alLEluia”: aGaG cdcdcbaGaG cdcdcbcba bcbabaG

(p.346) Ascendit deus (58)

In addition to the base transcription from Ben 34, I have included an alternative reading of the ending of v. 1 (phrase 5), for reasons described below.

This offertory exhibits several variants in pitch level. The base version of Ben 34 represents the relative pitch level between sections found in the great majority of MSS.

In Mo 159, the second and third verses are notated twice. Version 1 in Mo matches the level of Ben 34 until the verse endings, which are written a whole tone lower than Ben 34 (at “super omnem” in v. 2 and “sub pedibus” in v. 3). In version 2, the second and third verses are notated a fifth below Ben 34 until the verse endings, where the pitch level joins that of Ben 34. There are small melodic differences between Mo 159's two versions, attributable in part to the need for pitches unavailable in Mo 159's notation system that would be required to write the verses at the lower level.1 In terms of pitch level, both versions that appear in Mo 159 are unique among the sources examined. While neither of Mo's versions is fully matched in any other MS, however, several MSS do correspond to Mo 159 in specific places. The pitch level at these variant points can be summarized as follows.

  1. 1. End of v. 1, “in voce exultationis”:= Ben 34: All MSS in the sampling except: 2↓ Ben 34: Mo 159, Pa 1121, Ben 38, Cai 61, Gr 807, Be 40078, Rei 264.

  2. 2. beginning of v. 2, “Quoniam dominus” through “magnus”:

    • = Ben 34: All except Mo 159, version 2

    • 5↓ Ben 34: Mo 159, version 2

  3. 3. End of v. 2, “super omnem terram”:

    • = Ben 34: all except Mo 159, version 1, and Rei 264

    • 2↓ Ben 34: Mo version 1, and Rei 264

  4. 4. Pitch level at beginning of v. 3, “subiecit” through “gentes”:= Ben 34: All except: 5↓ Ben 34: Mo 159, version 2, Pa 780 (written level), Pa 1121 (written level; repetendum not notated), and Pa 903

  5. 5. Pitch level at end of verse 3, “sub pedibus nostris”:= Ben 34: All except: 2↓ Ben 34: Mo 159, version 1, Rei 264

Pa 780 presents an unusual reading in which the written level of the third verse begins a fifth below Ben 34 but joins Ben 34 at “sub pedibus.” The repetendum, however, is written a fifth too high, and the scribe's intention is not clear. If the verse is read with the assumption that the repetendum closes on D, however, the result is a reading that is not matched in other MSS. A similar pitch‐level profile of the verse itself is found in Pa 1134, but this source lacks a notated repetendum cue to confirm the written level at the beginning of the verse. The early adiastematic MS Ei 121 shows a verse opening consistent with the higher reading, with a virga and the letter l (levate). All versions that begin at a confirmable level a fifth below Ben 34 join Ben 34 at “sub pedibus.” The reading of Ei 121 has the letter i here, consistent with the majority version of Ben 34 rather than that of Mo 159, version 2. Although these factors suggest that the majority reading of Ben 34 is the preferred and unemended version, the reasons for writing this section of the third verse a fifth lower are not clear. There are no intervallic differences in this passage that would suggest (p.347) the presence of a nondiatonic pitch. Perhaps the two versions simply represent performance alternatives.

As the foregoing summary shows, the verse endings are another point of variance. The three verses of Ascendit end with essentially the same melodic material, on “exultationis” (v. 1), “super omnem terram” (v. 2), and “[sub]pedibus nostris” (v. 3). In Mo 159, version 1, each statement of this material is written a whole tone lower than it is in most MSS, with a focal pitch of b‐flat rather than c, joining the majority reading for the last seven notes. For v. 1, Mo's reading is found in several unrelated MSS: Ben 38, Pa 1121, Cai 61, and the three German sources, Gr 807, Be 40078, and Rei 264. In the second and third verses, however, Mo's reading is matched only in Rei 264.

The intervallic differences between the two versions may point to a nondiatonic pitch, as the following comparisons show:

v. 1, last twelve notes:

Mo (b's are flat):

bbb FDF FGE [flat?]

FGF

Pa 776:

ccc GEG GaF [sharp?]

GaG

Ben 34:

ccc GDF GaE

EGa

Without accidentals, the two versions are different intervallically: The whole tone F‐G in Pa 776 occurs at the same position as the semitone E‐F in Mo 159. If Mo represents the unemended pitch level here, then the reading of Pa 776 (the majority version) could be an emendation to represent E‐flat. If Pa 776 is the unemended pitch level, an F‐sharp would be required to duplicate the intervallic structure of Mo 159. In other versions, however, the note corresponding to Mo's E is an F:

Pa 1121:

bbb FDF FGF FGF (identical readings in Gr 807, Be 40078, and Rei 264)

Lo 4951:

cc aFG GaF FGF F(ori.) DED

The problems in establishing a preferred version of this passage are complicated by the melodic similarities between the three verse endings. As mentioned, the scribes of Mo and Rei 264 write all three statements of this material a whole tone below Ben 34. In Pa 1121, Ben 38, Gr 807, and Be 40078, however, the ending of v. 1 is a tone lower than the corresponding passage in verses 2 and 3.

In v. 1, the passage immediately preceding this verse ending is also a point of variance. In Ben 34, the cadence on “deo” (phrase 4) is G‐a‐b(‐flat) G/G, a reading matched in Mo and Pa 776. In many other MSS, however, it is a tone higher: pitch level at v. 1, “deo”:

  • = Ben 34: Ben 38, Mo 159, Pa 776, Pa 1121, Cai 61, Gr 807, Be 40078, Rei 264, To 18

  • 2↑ Ben 34: Ben 39, Lo 4951, Pa 780, Pa 903, Pa 1235, RoV 52, Pst 120, Mod 7, Pad 47

To summarize, the lower reading of the verse ending in v. 2 and 3 is found only in Mo (version 1) and one other source, Rei 264. In verse 1, however, the lower reading of the verse ending is found in all three German MSS and the unrelated sources Ben 38, Pa 1121, and Cai 61. In verse 1, then, this reading cannot be dismissed as a peculiarity (p.348) of Mo 159. The majority reading of Ben 34 is certainly the more regular in tonal structure and the lower reading the lectio difficilior. I would cautiously surmise that the unemended reading of the ending of v. 1 is at the level of Mo and Ben 38, with a b‐flat and E‐flat, while the endings of verses 2 and 3 were sung a tone higher. In most MSS that adopt the lower pitch level of this passage, the hypothetical E‐flat is either left as is or written as an F (as in Pa 1121 above). With this scenario, the minority verse endings of verses 2 and 3 in Mo and Rei 264 can perhaps be explained as adaptations to make the second and third verse endings conform to that of the first verse. This hypothetical unemended reading, however, is preserved only in Ben 38 and Pa 1121. While I have not adopted Ben 38 as the base reading simply because of the uncertainties surrounding this passage, I have included it as an alternative reading in the transcription.

Transcription issues: In phrase 4, Ben 34 has a scribal error that is corrected by a custos at the end of the line. At [iubila]te deo (v. 1), an erroneous F clef occurs, corrected by the custos on D at the end of the line. I have emended the transcription accordingly. A similar error occurs at the end of v. 2 (phrase 7), where the close of the verse and repetendum are written a whole tone too high.

Beneventan independence: Beneventan MSS transmit a shortened and varied version of the melisma on “terribiLIS” (v. 2, phrase 6). The other MSS transmit a reading similar to that of Pa 776:

a Fac Fac Fac aGFG FFDDC D FFF C DFDC CDFF aGFG FF DGE FFF.

Other melodic variants: v. 3 “subjeCIT”: The melisma is a point of slight variance among core MSS. Benevenan MSS lack a threefold repetition of a that appears in Mo 159, Aquitanian MSS, and most peripheral MSS:

  • Mo (version 1): db dc dc c(ori.) a accc(ori.)a Gba aaa

  • Pa 1121: dc db ca a(ori.) Ga cc c(ori.)G aca aaa

  • Pa 776 gives shorter version of melisma: db dc caaG aca aaa

German MSS give a reading similar to that of Mo 159.

Italian MSS also have a variant reading of this melisma:

  • Pst 120: dcdc dcca acc ca cdc ccc

  • RoV 52: dc dc ca aG a c ca dec ccc

  • Mod 7 is closer to Mo 159: dcdc dc aaG a ccc ca Gba aaa

  • v. 3 is lacking in Pad 47

v. 3 “populos nobis et gentes”: Cai 61:The scribe appears to write this section of v. 3 a third too high, perhaps a scribal error.

Roman MSS:

Vat 5319:

  • respond, “DOminus”: shorter version of the melisma, FaGabGF

  • v. 1 “exultatioNIS”: FEFGFE (the less ornate version of this figure)

  • Vat 5319 lacks v. 2

RoS 22:

  • “iubilatiO/NE”: G ababaG/abaG

  • “alleluIA”: E FGFaGFEFE

(p.349) Emitte spiritum (59)

This offertory has very few significant variants between sources, especially in the core group. The minor differences between Ben 34 and the non‐Beneventan MSS are as follows:

  • respond “TUum”:

    • Mo 159: FGaaG

    • Pa 776: FGaG

    • (similar in other core and peripheral MSS)

  • respond “seCUla”: Mo 159: GacGG(ori.)FG (other sources in the core group are identical to Ben 34).

  • v. 1 “magnifiCAtus”: Most non‐Beneventan MSS have a reading identical to that of Mo 159: Gc ccc dc dcc.

  • v. 2 “veheMENter”: Mo 159: melisma opens G cdc abaG (similar reading in the Aquitanian MSS).

  • v. 2 “aMIC/TUS”: Mo 159: FaG/G

  • v. 2, “vestiMENtum”: Mo 159: melisma opens G cdc abaG (as in “vehementer”above)

  • v. 3 “QUI”: Mo 159: FG aG (same in other core sources)

  • v. 3 “Aquis”:

    • Mo 159: FG b‐flat GF b‐flat a

    • v. 3 QUI/POnis DF/FaG

Individual sources:

  • Lo 4951: v. 3 “aQUIS”: acc accc acc ab GFG

Cai 61:

  • respond “eMITte”: F

  • respond “creaBUNtur”: cccaGF GabaG(liq.)

  • respond “gloriA”: decc(ori.)Ga bcb

  • v. 1 “MEus”: Gcc ca ccaG acb a

  • v. 2 “PELlem”: G aca G GF Ga

Pa 1235:

  • respond “eMITte”: F

  • respond “TU/UM”: F GaG ac cc/c (different text underlay indicated)

  • respond “creaBUNtur”: cbcaGF GabaG(liq.) (similar to Cai 61)

  • v. 1 “MEus”: Gc ca bcaG ac aGca

  • v. 2 “conFES/SI/Onem”: ae/ed/dedb cdca ced

  • v. 3 “EX/TEN/DENS”: a/Ga cb ba(liq.)/baba aca aG

  • v. 3 “PELlem”: Gacba G GF GaG(liq.) (similar to Cai 61)

  • Pst 120 is very close to the core group.

RoV C 52:

  • respond “alLEluia”: cccc cabaGa ccbabcba

  • v. 2 “Amictus”: Fa

To 18: (p.350)

  • respond “TU/UM”: FGaGac/ccc

  • respond “FAciem”: Fa cdc

  • respond “seCUla”: abcbG

  • v. 1 “magniFIcatus”: aG

  • verses 2 and 3 not notated

Pad 47:

  • v. 1 “MEum”: slightly longer melisma that begins cdcacdcdc (etc.)

  • v. 3 “extenDENS”: aGa babG ababG

Gr 807:

  • v. 2 “EX/TENdens”: F/FG aGa

  • v. 2 “Aquis”: FGaGFa

  • v. 2 “POnis”: FaG

  • v. 2 “TUum”: cc ca cbcaG aGF G aca cGa

  • Be 40078 presents a reading very close to that of the core group.

Roman sources:

  • Vat 5319:

    • respond “spiriTUM”: aG (liq.)

    • respond “alLEluia”: a transpositional variant in the middle part of the melisma, in boldface: cdcbababa GaG ababa GaG ac dcdc bca GaGFGFE GFEFED EG GG abab (etc.)

  • RoS 2: “alLEluia”is similar, but not identical, to Vat 5319: cdccba baba GaG ababaGaG acba dcdcbca GaGFG FEGFEFEDF GaG baba (etc.)

Confirma hoc (60)

In terms of pitch level, Ben 34 presents an unusual reading of this offertory; a few of its peculiarities also occur in Ben 35. Ben 39, chosen as the base version, presents a reading consistent with other core sources, except in one passage, described under (3) below.

  1. 1. Opening passage, “confirma hoc”: For reasons that are unclear, the scribes of Ben 34 and Ben 35 begin the offertory a fourth below the majority reading represented by Ben 39. On the last syllable of “confirma,” there are further variants. The reasons for the variant readings in this problem spot are difficult to ascertain, as the following sampling shows.

    [Confir]

    ma

    hoc

    Ben 39:

    FGa

    abG babGaG (liq.)

    Mo (b's are natural):

    Gac

    b b(ori.)cG ba bGa

    Pa 1121:

    GGG

    GaF babGa

    Lo 4951:

    EFG

    GaF babGa

    Pa 780:

    GGa

    abG cbcab

    Pa 776:

    EFG

    GaF babGa

    Cai 61:

    FGa

    acG cbcab

    Be 40078:

    Ga

    acG ca cGa (same in Gr 807)

    Mod 7:

    Gac

    c ca cbbab (same in Pad 47)

    Pst 120:

    Gab

    abG cbcGa

    (p.351)

    Most of the sources join in the second part of the short melisma on “hoc.” On the last syllable of “confirma,” there are intervallic differences between MSS. In Mo and Ben 39, the neumes consist of two successive whole tones, perhaps suggesting that an E‐flat was required at the level of Pa 776. In Pa 776, the first five notes are mistakenly written a third too low and corrected with letter notation, which may attest to some uncertainty on the part of the scribe about the opening of the melody.

  2. 2. “DEO”: Ben 34 is a whole tone above the other MSS here, with a different intervallic structure. The reasons for the different reading are not clear.

  3. 3. v. 2 “dominum de fontibus”: The scribe of Ben 39 writes this passage a whole tone lower than it appears in all other MSS of the core group. The passage is indicated with brackets in the transcription. I consider Ben 39's reading an error and have provided the more common reading in the edition.

    The other sources included in the sampling match Mo here, joining with Ben 39 at “de.”

  4. 4. v. 2 “IsraHEL”: Where brackets are indicated in the base transcription, Ben 39 has a scribal error, writing a segment of the melisma a tone too low. The status of this reading as an error becomes evident with a comparison of the closing melisma of the final verse, on “orientem.” The two melismas differ in their opening notes but are otherwise nearly identical. In v. 3, Ben 39 matches the other sources in this segment. I have emended this segment in the base transcription.

  5. 5. v. 3 “cantate deo”:

  1. a. “cantate”: The low B on “cantate” was likely sung as flat, a pitch not theoretically available in most notation systems. In most MSS, however, this B is left as is, matching Ben 39, with the following exceptions:

    • Pa 1235, Pa 780, RoV 52: The last three notes of “cantate” are DDC.

    • Pad 47 and Mod 7: The beginning of the verse is notated at the affinal position, joining the majority version at “psallite.”

    • Tri 2254: The whole verse is written at the affinal position.

    • Cai 61 and Gr 807 have a notated low B‐flat.

    • In Ben 34, all of “cantate deo” is notated a tone above the majority version, producing some differences in intervallic structure.

  2. b. “DEo”: Some sources in the core group have FDF instead of CDCD (as in Ben 39).

    • CDCD: Mo 159, Ben 34, Ben 35, Pst 120

    • FDF Pa 776, Pa 780

    • DFDF: Pa 1121, Pa 1235, Cai 61, Be 40078

    • DFEF: Gr 807

    • DCD Lo 4951

    • DEDE: RoV 52

    • GaGa (at affinal position): Mod 7

(p.352) Pad 47 preserves a possible nondiatonic practice in this passage. The beginning of the passage is written at the affinal position, and “DEo” is written as G‐b‐flat G‐b‐flat, equivalent to E‐flat at the lower level.

6. v. 3 “PSALlite”: Beneventan MSS present a very small difference in this melisma, omitting the notes indicated in boldface, as in Mo 159: DG acaGF FFF aGF GFF DGF EFD (etc.).

Variants in individual MSS:

Verse endings in Lo 4951: The slightly altered verse endings create a more exact “melodic rhyme” between verses:

  • v. 1 “ilLI”: DF FFF FGFG FF(ori.)E

  • v. 2 “israHEL” and v. 3 oriENtem: melisma ends FGFGF F(ori.)E

Cai 61:

  • v. 2 “DOmino”: DFa

  • v. 2 “israHEL”: melisma has the additional notes FGFGEF at the end

Be 40078:

  • v. 2 “DEo”: FaGF FD GF GE FD CGFG F FD FGF aGF GFE

“DOmino”: DGa

  • Gr 807: v. 1 “asCENdit”: aG ca cGF (liq.)

Mod 7:

  • respond, “EST” GaGaFE

  • v. 1 “diciTE”: FaGF

  • v. 1 “NOmini”: FGF

  • v. 1 “DOminus”: FGaGa

  • v. 1 “NOmen”: FGaGa

  • v. 2 “DEo”: melisma ends baG aGF

  • v. 2 “DOminum”: EFGa

Pst 120: v. 1 “Eius”: FGFG

Roman MSS:

Vat 5319:

  • respond, “TUo”: aGa(ori.)GFEF

  • v. 1 “EST”: EFGFGa(ori.)GFGF

  • v. 2 “israHEL”: melisma begins abaGaGFEFE GaGFGFE (etc.)

  • v. 3 “oriENtem”: same variants as v. 2 above

RoS 22:

  • “operaTUS”: GGFEFD

  • “HIErusaLEM”: aaGFGabaGFG/EF

  • “muneRA”: FGaGbaG

(p.353) Precatus est Moyses (61)

In phrase 1, the base reading of Pa 1121 lacks a first statement of the words “et dixit,” which are found in a majority of MSS. The passage (which uses melodic material also found in Angelus domini) is supplied from Pa 776.

The problem spots in Precatus est Moyses have been addressed in studies by Frasch and Hankeln.1 As Hankeln discusses, there are probable scribal errors in several of the Aquitanian MSS, complicating the process of establishing a preferred reading. In Theinred of Dover's treatise, Precatus est is given as an example of an offertory that has three relocations of the semitones, implying that it uses low B‐flat, E‐flat, and F‐sharp.2

The problems in the respond begin with the passage “et placatus factus est” (phrase 6). I have provided the Pa 780 version as an alternative reading. Although Pa 1121's reading is matched in Mo 159 and in the majority of peripheral MSS, most Aquitanian MSS notate this passage a whole tone lower, at the position of Pa 780. The intervallic differences between the two versions suggest that one of two nondiatonic pitches was employed in the pretheoretical tradition. If Pa 1121 represents the intended interval structure of the passage, an e‐flat would be required to reproduce this structure at the level of Pa 780. In Pa 1121, f‐sharp would be required to replicate this intervallic structure of Pa 780. Either alternative is possible. Given the tendency of Pa 776 and other Aquitanian MSS to notate passages in their unemended range, however, the first alternative seems more plausible.

Pitch level at “et placatus factus est dominus”:

  • = Pa 1121: Mo 159, Ben 34, Ben 35, Gr 807, Be 40078, Tri 2254, RoV 52

  • 2↓ Pa 1121: Pa 776, Pa 780, Lo 4951, Pa 1235, Pad 47, Mod 7, Pst 120

Most MSS join Pa 1121 at “de malignitate.” Pa 776 continues a whole tone below Pa 1121 for the remainder of the verse, a probable error. Pa 1235 joins within “dominus” (with a slightly different melodic reading).

Pitch level of verse 1: The MSS exhibit many pitch‐level variants here, as summarized below. At “invenisti” (phrase 8), Mo 159 and most non‐Aquitanian MSS are written a fifth above the base version of Pa 1121, then a fourth above beginning at “et scio” (phrase 9) and continuing to the end of the verse. Pa 1121, however, represents the majority notated level among Aquitanian MSS.3 Hankeln sees a discrepancy between the written and performance levels of the verse in the Aquitanian MSS, beginning at “invenisti,” and proposes an Aufführungslage for the whole verse equivalent to that of Mo 159. Most Aquitanian MSS, however, lack a notated repetendum cue after this verse that would offer a confirmation of the performance level. Although Pa 1121 lacks a notated repetendum cue after the verse, there is a rare “equaliter” sign between the respond and first verse that confirms the intended pitch level of the verse.

I have opted for the lower version of the verse, found in Pa 1121, as the preferred reading for several reasons. First, Mo 159's shift of range between “Moysen” and “invenisti” is a minority reading, found only in two other sources examined, Pa 1235 and RoV 52. Mo 159's minority status here may suggest that its reading of this passage is an emendation. The (p.354) significative letters in Ei 121, moreover, are consistent with the reading of Pa 1121, but not with Mo or the versions a fourth higher. In Ei 121 the verse begins with an “equaliter” sign, suggesting that the opening pitch is G, inconsistent with the readings that begin on high c or d. Ei 121 has another “equaliter” between “Moysen” and “invenisti,” which contradicts the register shift of Mo's minority reading but is consistent with Pa 1121. Finally, the nondiatonic tradition for this offertory mentioned by Theinred of Dover is not discernable at the pitch level of Mo or the peripheral MSS. At the level of the majority Aquitanian reading of Pa 1121, however, F‐sharps are required to replicate the intervallic structure of Mo beginning in phrase 9. In this version, “invenisti gratiam” has protus characteristics (such as excursions into the lower tetrachord, G‐D, which are not uncommon in eighth‐mode melodies). In phrase 9, the tetrardus characteristics heard in the respond and the beginning of the second verse are projected a fourth lower, possibly to distinguish the voice of God from the voice of the narrator heard throughout the respond and the opening of the verse. The other readings place these passages in the traditionally tetrardus part of the background scale.

Pitch level of v. 1:

  1. 1. Verse opening, “dixit dominus”:

    • = Pa 1121: Mo 159, Pa 776, Pa 780, Pa 1134, Pa 1136, Lo 4951, Ben 34, Ben 35, Pa 1235 (cadence is on G), RoV 52, Gr 807, Frutolf tonary

    • 4↑ Pa 1121: Mod 7

    • 5↑ Pa 1121: Tri 2254

    • 5↓ Pa 1121: Pst 120

  2. 2. “ad Moysen”:

    • = Pa 1121: Mo 159, Pa 776, Pa 780, Pa 1134, Pa 1136, Lo 4951, Pa 1235 (cadence is on E), Gr 807

    • 5↑ Pa 1121: Tri 2254

    • 4↑ Pa 1121: Ben 34, Mod 7

    • 2↓ Pa 1121: Pst 120 (joins at “moySEN”)

  3. 3. “[Invenisti] gratiam in conspectu meo”: Most non‐Aquitanian MSS begin this passage at 4↑ Pa 1121 on “invenisti,” switching to 5↑ at “gratiam.” This difference is not reflected in the summary.

    • = Pa 1121: Pa 776, Pa 1134, Pa 1136, Pa 780, Gr 807, Frutolf tonary

    • 5↑ Pa 1121: Mo 159, Lo 4951, RoV 52, Mod 7, Pa 1235 (changes to 4↑ at “meo”)

    • Unique reading: Pst 120 (beginning= Pa 1121; grati [am] is 3↑; conspectu is 5↑)

  4. 4. “Et scio”: 4↑ Pa 1121: all except:= Pa 1121: Pa 776, Pa 780, Pa 1134, Pa 1136.

Melodic variants: The Beneventan reading is very similar to that of the core Aquitanian MSS. A few significant variants are found in Lo 4951 and in some peripheral MSS.

Pa 776: respond, “in conspectu” (both times): Pa 776 notates this passage a tone higher than Pa 1121. The reading of Pa 1121 is the majority reading among core MSS (the reading of Pa 776 is also found in several Italian MSS).

Lo 4951: (p.355)

  • respond, “et dixit” (both times): D/DFFDEFGaG/G (the first statement is shorter and written a fifth lower than the majority reading. The second statement is the same as the majority reading melodically but written a fifth lower).

  • respond, “aBRAM”: cd(quil.)edc

  • v. 1 “peccaTA”: shortened melisma: efffdcbcdecc ddcd(quil.) edccd fffdcbcdeccdd cd(quil.)edcd(quil.)ec

Pa 1235:

  • respond, [populo] TUo: GF abaGa

  • v. 1 “peccaTA”: small variants in final melisma: fefdcb cdeb cd dc ded c cd fefdcb cdeb cd dc dededc fgf fdc fefdc dfe (etc.)

  • v. 2 “acCEdite”: GE GaGF GF FFFECD DF GaG

Be 40078: v. 1 “Omnibus”: FacGE FDCD FF GaG cc caa(ori.)G aFG ac cc ccc

  • Gr 807: respond, “popuLO”: db ca cccbG aca caF Ga ccba bcba

Pad 47:

  • respond, [lac] “ET”: G abca ca (etc.)

  • respond, “PopuLO”: dca cc cc cG acacGF Ga cccbabcba (slightly shortened melisma)

  • v. 1 “moY/SEN”: dca cdedc/cbdcb

  • v. 1 “peccaTA”: fffdcb cde cc d dc c defd cdf fff dcb cde cc d dc cdefd dc f fgfdcd fffdc dfe fged ededc dca c defd fd fffd cca

Mod 7:

  • respond, “in conspectu” (both times): whole tone above Pa 1121

  • respond, [lac] “ET”: G abca ca (etc.)

  • respond, [placatus factus est] “DO/MI/NUS”: cded/dcdede/ed

  • v. 1 “peccaTA”: reading very similar to that of Pad 47

Pst 120:

  • respond, “[con] spectu do [mini]” (both times): Written a whole tone above Pa 1121

  • v. 1 “peccaTA”: ffd c cb cded ccd dc ded ccd f fd c ca cded cc d dc ded cdf fedf dc dff db dfe fgdc ededc dca cdfd fd fffd c ca

  • v. 2 “moySES (second time)”: melisma ends CDC

RoV 52:

  • respond, “in conspectu” (both times): Written a whole tone above Pa 1121

  • v. 1 “peccaTA”: cccaGF FGa FF G GF GaG FFG cccaGF FGa FF G GF GaG Fa c cdcaG accaG acb cdba babaGaGF Ga caca cc caG GE Ga (similar to Pst 120 above, but written a fourth lower)

  • lacks the second verse

Roman MSS: In the verse Dixit Moyses, the final cadence, on the last two syllables of “tempore,” is written a whole tone below its normal position. This apparent irregularity (which occurs a few other times in the repertory) is also present in Vat 5319.

Vat 5319:

  • respond “DIxit” (second time), a shorter and varied version of the melisma: Gac cbcba G GF acGF F(ori.)EDC FacbaG GF acbabG

  • (p.356)
  • respond, “FLUENTEM”: ba/G/ac

  • verse Dixit dominus, “TERra”: GababG(liq.)

  • verse Dixit dominus, “PEC/CA/ta”: neume groupings suggest a different text underlay: c/cc(ori.)b cba abcdcbaG

  • verse Dixit Moyses has “populorum israhel” instead of “filiorum israhel” (same melody)

RoS 22:

  • “DIxit” (first time): G ac cbcbaG acbG

  • “DIxit” (second time): GacbcaG aGaGF ac GGFED acbaGF acbabaG

  • “A/BRA/HAM”: defedefd/cb/ded

  • “FLU/ENtem”: ba/G

Immittet (62)

The base version of Ben 34 is atypical in a few very brief passages:

  • Respond, “viDEte”: In most MSS, this melisma is very similar to “TEMpore” in v. 1, as in the sample readings of Mo and Pa 776:

    • Pa 776: c dcd ccba ccbac

    • Mo: cdc db caG  ccb(quil.)ac

  • Beneventan MSS have a shorter version of the melisma, without the repeated notes and passing tone: cdcdccaG cac

  • v. 3 “eruBEScant”: Ben 34 has a slightly shortened version of this melisma, different from most core MSS in lacking the notes in boldface:

    • Mo: ccb (quil.) a Gb (quil.) c cba Gb (quil.) cb (quil.) a b (quil.) aGa

    • Pa 776: cc cbG ab(quil.)    ccbG ab(quil.) cba   baGa

The longer version of the melisma is found in Ben 38, Ben 35, Pa 1121 and Pa 780. Of the MSS in the core group, only Lo 4951 has a reading nearly identical to that of Ben 34: cc cbG ab(quil.)cba baGa. A similar reading is found in Pad 47: ccbG G abcba cbGa.

There are very few variants between MSS. Other melodic variants:

  • Pa 776: v. 2 “INvicem”, a shorter version of the melisma: c dcd ccbab(quil.)cba.

  • Peripheral MSS show a tendency to add or omit passing tones, as in Cai 61, respond, DO/MIni: ac cbaG/ac cbaG.

Other variants in peripheral MSS:

  • Cai 61: v. 3 “eruBEScant”: cb cbGa bcb c bGa bcbabaga.

A few MSS differ from the norm in their verse indications. Pa 780, for example, indicates only one verse, at “benedicam” (though it presents the full music and lyrics). Pst 120 has a new verse beginning at “iste pauper clamavi.”

Roman MSS: Bod 74 lacks the final verse of this offertory. One anomaly that occurs in both Vat 5319 and Bodmer is the cadence at the end of the second verse, on “invicem.” For reasons that are unclear, this typical tetrardus cadential pattern is written a tone below its (p.357) usual position, closing on F in both MSS. A similar cadence occurs in Illumina oculos (28) in phrase 6, with a notated b-flat in Vat 5319.

Bod 74: v. 2 “laudaBI/TUR”: cba/b (written a tone lower than Vat 5319).

RoS 22:

  • “IM/Mitet”: Gaba baGF

  • “Eum”: abaG

  • “DOmine”: melisma begins cdcdccaGF

Oravi deum meum (63)

Oravi deum has a complex and varied transmission, undoubtedly because nondiatonic pitches were employed in the pretheoretical tradition. I have adopted Pa 1121 as a base reading because it represents the most probable unemended version. Oravi begins as a deuterus melody and, in a great majority of sources, closes on E. In the pretheoretical

Appendix 3 Commentary on Edition

Example 63.1a

(p.358)
Appendix 3 Commentary on Edition

Example 63.1b

tradition, however, it most likely closed on D, with a deuterus E‐flat. Each problem spot is discussed separately.

1. Respond, opening “Oravi.” John of Afflighem mentions that Oravi deum requires emendation at two places in the respond, namely the beginning and at “super.”1 The opening passage is indeed a point of melodic variance. Several versions, including Pa 1121, incorporate a low B, on the second syllable of “oravi,” which was undoubtedly sung as B‐flat, a pitch that was theoretically nonexistent:

Pa 1121: EE(ori.) FDGFFFFD B(flat) DFD B(flat)DFDEF.

Rei 264, notated at the affinal position, employs F, equivalent to B‐flat at the lower level, and b‐flat, equivalent to E‐flat at the lower level: bc adc c cbG a b‐flat aG aca b‐natural c.

(p.359) In other sources, the b‐flat is avoided by means of melodic variants, as in the following: Mo 159: E E(ori.)F DGF FF FDC DFDC DFD EF (very similar readings in many other sources, including Pa 776 and Pa 1132).

2. Respond, “super sanctua[rium]” (phrase 3) to end: According to John, this passage requires an emendation so that “super” begins on c. This emendation is shown in Frutolf's tonary and found in the great majority of MSS. A few sources, however, give clues to the pretheoretical tradition. Example 63.1 shows three different versions. Beginning with “super,” Pa 1121 (line 1) adopts b‐flat as a focal pitch. The majority version, represented by Mo 159 (line 2), is written a tone higher through the end of the respond. In Pa 1121, typical melodic gestures of the chant tradition are positioned a tone below their usual position. At this lower position, several nondiatonic pitches are required to replicate the intervallic structure of Mo 159's majority reading, including low B‐flat, E‐flat, and a‐flat. The respond ends on D, with an E‐flat that results in a deuterus tonal structure. Singing “super” on b‐flat, then, evidently resulted in a modulation of sorts, in which the deuterus melodic characteristics heard in the first half of the melody were projected a tone lower; E‐deuterus becomes D‐deuterus.

Despite its irregularity, this reading cannot be dismissed as a scribal error; it is also found in Pa 776 and is preserved in three late geographically diverse sources, Cai 61, Rei 264, and Be 40078. These versions are notated a fifth higher, at the affinal position, and replicate the pitch‐level profile of Pa 1121. Be 40078 and Rei 264 begin as regular deuterus melodies notated at the affinal position but close on a, with a b‐flat. Cai 61 is similar to these versions but has a modified ending on G. The version of Rei 264 is shown in line 3 of example 63.1. Several of the b‐flats in Mo 159 would result in a‐flats at the level of Pa 776 and e‐flats at the level of Rei 264. Rei 264 avoids the latter pitch by either writing a regular e, as on” [sanctuarium] tuum”, or an f, as on “intende.”

Two sources in the core group, Lo 4951 and Pa 780, adopt distinctive emendations of the passage that begins at “super.” Both match Pa 1121 at “super.” Lo 4951, however, joins the level of Mo 159 on the first syllable of “sanctuarium,” and Pa 780 joins the level of Mo 159 on the third syllable of “propicius.” In Lo 4951, the starting pitch of each line in the respond is indicated with letters in a later hand, probably attesting to the problematic pitch level of the respond.

Near the beginning of this problem spot, on [sanctuarium] “tuum,” Mo 159, written a tone higher, has a b‐flat, the source of the parenthetical a‐flat in the transcription of Pa 1121. In other sources, a b‐flat is not specified.

= Pa 1121: Pa 776, Lo 4951 (2↑ beginning at “sanctuarium”), and Pa 780 (2 ↑ beginning at “propicius”)

5↑ Pa 1121: Be 40078, Rei 264, Cai 61 (Both sources begin the offertory at this level and remain there throughout).

2↑ Pa 1121: Mo 159, Ben 34, Ben 35, Ben 39, Pa 1132, Pa 1235, Pst 120, Pia 65, Tri 2254, Gr 807

The verses: In the pretheoretical tradition, the problematic ending of the respond may have had implications for the performance level of the verses. I have transcribed the verses of Pa 1121 both at their written level, which matches that of most MSS in the core group, and at their hypothetical performance level. The written level of Pa 1121 matches that of most MSS, including Mo 159, which consistently indicates b‐flat here. In Pa 1121, however, the repetendum cue after v. 1, to “super quem,” is notated a tone too high, beginning on C rather than B‐flat. Pa 776 presents a similar situation: if Pa 776 is read with the assumption that the verse begins at the level of the majority version, then the (p.360) repetendum cue to “super quem,” which follows the second verse, is notated a tone too high, beginning on C.

There are two possible explanations for this situation. The first, according to Hankeln's theory about the meaning of the “incorrect” repetendum cues, is that the cue attests to a performance tradition in which the verses were performed a tone below their written level, consistent with the lower ending of the respond. A performance level of the verses a whole tone below the level of Mo 159 would result in a nondiatonic melody. The opening of verse 1 would adopt ab as a focal pitch. With an emendation of the respond at “super,” so that it closes on E, the verses proceed unproblematically. A second possibility is that the verses were sung at their notated level, matching the majority reading, but that the repetendum was performed a tone above its original position, consistent with the emended version found in most MSS.

The inconsistency between the respond and repetendum is certainly related to the problematic ending of the respond. If the unemended version of the respond involves a downward shift of the deuterus structure by a whole tone, when did the chant get back “on track?” The notated level of the verses in Pa 1121 suggests a return to a diatonic position at the start of the first verse. This alternative, however, creates a problem in the transition to the repetendum. A performance of the respond at the notated level and the repetendum in its original, unemended range involves a rather sudden shift back to the D‐deuterus tonal area. If the repetendum is taken as a guide to the correct pitch level of the verses, the result is the hypothetical performance level shown in the transcription. In this version, the dislocation of the melody in the respond continues in both verses, and the chant never gets back on track.

In the two verses, most sources present a pitch‐level profile similar to that of Mo 159. Rei 264, Be 40078, and Cai 61, for example, join the level of Mo 159 at the beginning of v. 1 and remain there throughout. There are, however, two problem spots within the verses, both relating to the G‐protus characteristics of certain passages. At its written level, Pa 1121 alternately adopts b (flat) and c as focal pitches, a reading matched in Mo 159. The b‐flat is heard at the beginning of verse 1 (“Adhuc…orantes”) and at “Michael” in v. 2. In many other MSS, however, these places are written a whole tone higher, so that c is the focal pitch throughout. These readings are perhaps best viewed as emendations made in response to the G‐protus quality of these passages. I have shown that certain MSS exhibit a propensity for whole‐tone transposition of G‐protus passages, in other pieces such as Sanctificavit and Benedicite gentes.

  • v. 1: “Adhuc…orantes”:

    • = Pa 1121 (written level): Pa 776, Pa 780, Mo 159, Be 40078, Cai 61, Tri 2254, Rei 264

    • 2↑Pa 1121: Ben 34, Ben 35, Ben 39, Lo 4951, Pa 1235, Pst 120, Pia 65, Gr 807

  • v. 2: “Michael”:

    • = Pa 1121: Pa 780, Ben 34, Ben 35, Ben 39, Pa 1235, Cai 61, Be 40078 (without indicating a b‐flat), Rei 264, Tri 2254, Piacenza 65

    • 2↑ Pa 1121: Lo 4951, Pst 120, Ben 39, Gr 807

Other melodic variants: v. 2 “PO/PU/LI”: The version of Pa 1121 is matched in the Aquitanian and Beneventan MSS, as well as most peripheral sources. Mo 159 has a different reading and text underlay: G/Gab(quil.)cb(quil.)a/aG.

Roman MSS: Bod 74 lacks this offertory. RoS 22 has the following variants: (p.361)

  • “israHEL”: GF GaGabaGFGF

  • “doMIne”: GbaGFEF

  • “serVI”: GaGbaGFGF

  • “inTENde”: FGaGbaGFEF

Significant variant in the final cadence on “DE/US”: EFGFE/FE (lacks the long melisma of Vat 5319)

Sanctificavit (64)

In most Gregorian sources, the last verse ends with the words “tunc Moyses,” which provide a transition to the repetendum, “fecit.” These words, however, are probably a late addition, and they are a point of melodic variance. Some MSS have a more elaborate transition to the repetendum than the base version of Pa 776 does. Mo 159, for example, has a long melisma on “Moyses”:

  • Mo y‐ ses

  • Gab(quil.)cb(quil.)aG FE(quil.)D Gab(quil.)cb(quil.)aG FE(quil.)D FG Ga a(ori.)G

This melisma, however, is lacking in most Aquitanian MSS, in the Beneventan sources, and in most early notated MSS.1 The Roman version takes “in conspectu” as a repetendum.

Sanctificavit has an extraordinarily varied transmission in terms of pitch level, reflecting many instances of nondiatonic practice. Theinred of Dover, writing in the mid–twelfth century, describes this chant as having three relocations of the semitones, implying that it employs E‐flat, low B‐flat, and F‐sharp.2 Although the comparative transcriptions produced evidence for the use of all three pitches, as well as eb in the higher octave, the evidence for F‐sharp (at “illud” in the respond) is very tentative.

Pa 776 was chosen as the base reading because of its minimal tendency toward modal emendation. There are no less than twelve significant pitch‐level variants among the sources. Whole‐tone transpositions are particularly prominent. Although most sources fall into a few distinct groups in their treatment of the nondiatonic pitches, highly individual readings may be found in certain MSS such as Lo 4951.

1. Respond, “illud”: Mo 159 presents a minority reading of this passage that is notated a whole tone lower than Pa 776 and has a different intervallic structure:

  • Mo 159: aGFFFF (ori.) EFaG

  • Pa 776: aGGGG (ori)FGbaa

The version of Pa 776 is matched in the great majority of sources. Only one other MS in the sampling, Pst 120, notates the passage at the level of Mo 159.

Snyder sees the version of Mo 159 as the unemended pitch level and suggests that an E‐flat was sung here.3 Another option, however, is that Pa 776 represents the unemended pitch level of the passage and that the version of Mo is the transposed version. An F‐sharp would be required at the level of Pa 776 to replicate the intervallic structure of Mo 159. This passage, in fact, is the most plausible instance of the F# mentioned by Theinred. In some (p.362) sources at the level of Pa 776, the note corresponding to the hypothetical F‐sharp is written as an E, as exemplified in Cai 61: aGGGGEGbaa.

The evidence for nondiatonic practice here, however, is not conclusive. The variant may instead reflect a point of melodic instability. If the variant does reflect a nondiatonic practice, I see F‐sharp (rather than E‐flat) as the more likely alternative, given the tendency of Pa 776 to write passages in the unemended range, the tendency of Mo 159 to give the correct intervallic structure of problem spots at an “emended” pitch level, and the status of Pa 776's pitch level as a majority reading. In practice, however, an F‐sharp is admittedly aurally disconcerting. If an F‐sharp was indeed sung at one time, it may have been simply suppressed in most MSS. If performers choose not to sing the F‐sharp, a b‐flat (rather than b‐natural) should follow.

2. End of respond, “israhel”: Two German sources, Gr 807 and Be 40078, close the melody on E rather than F, as exemplified in the following version of the melisma on “Israhel”: Gr 807: GabcaGEFFDFGEGGE.

3. v. 1: “aSCENdit” melisma: Although Mo and the other Aquitanian sources present a reading very similar to that of Pa 776, Ben 34 notates the boldface parts of the melisma a tone higher:

  • Pa 776: cffcdedcdcb[flat]cb[flat]c(quil.)dcdcdfffdffgf

  • Ben 34: dffdefededcdcededefffdffgf

The other Beneventan sources present a reading closer to that of Pa 776. One source in the sampling, RoV 52, presents a unique reading of the melisma: ccccccabaGaGFaaGaaaGacdc.

4. v. 1: “Moyses procidens”: The variant starts on second syllable of “Moyses.” This problem spot, discussed in chapter 6, is best considered in the three parts “Moyses procidens,” “adoravit dicens,” and “obsecro…populi tui,” shown below. In “Moyses procidens,” the reading of Pa 776 matches Mo 159 and the other Aquitanian sources, but the passage is transposed up a whole tone in many other sources, including the Beneventan MSS. While the upward transposition is probably attributable in part to the G‐protus characteristics of the passage and the wish to avoid b‐flat in close proximity to G, the subsequent passages reflect the introduction of nondiatonic pitches.

  1. a. “Moyses procidens”:

    • = Pa 776: Ben 39, Mo 159, Lo 4951, Pa 780, Pa 1121, Pa 1132, Pa 903, Pa 1235, Gr 807, Rei 264

    • 2↑ Pa 776: Ben 34, Ben 35, Be 40078, Cai 61, Tri 2254, Mod 7, RoV 52, Pst 120, To 18

  2. b. “adoravit dicens”:

    • = Pa 776: Ben 39, Mo 159, Lo 4951, Pa 780, Pa 1121, Pa 1132, Pa 903, Pa 1235, Rei 264

    • 2↑ Pa 776: Ben 34, Ben 35, Be 40078, Cai 61, Pa 1235, Gr 807, Tri 2254, Mod 7, RoV 52, Pst 120, To 18

  3. c. “obsecro…populi tui”:

    • = Pa 776: Ben 39, Lo 4951, Pa 780, Pa 1121, Pa 1132, Pa 903

    • 2↑ Pa 776: Mo 159, Ben 34, Ben 35, Be 40078, Pa 1235, Cai 61, Pa 1235, Rei 264, Gr 807, Tr 2254, Mod 7, RoV 52, Pst 120, To 18

This passage is given as Example 6.2 in chapter 6. Although most non‐Aquitanian sources match the pitch level of Ben 34 at “obsecro,” the Aquitanian sources, represented by (p.363) Pa 776 in line 1, are united in writing the passage a tone below its pitch level in the other readings, a version also found in Ben 39. Without accidentals, the two readings differ in interval content: the Aquitanian sources articulate the major third c‐e, in contrast to the minor third d‐f found in Ben 34 and the other sources. These differences in intervallic structure suggest the use of a nondiatonic pitch. One possibility is that the major third of the Aquitanian sources is intended. If so, an f‐sharp would be required in the Ben 34 and Mo 159. The other, more likely, possibility is that the Aquitanian sources transmit the preferred pitch level, but that an e‐flat was sung.

The melodic characteristics shed some light on the question of which is the more likely alternative. The version of Ben 34 is unproblematic, adopting melodic vocabulary typical of the chant repertory in general: high f is the repercussive pitch, and the cadential pattern on “tui” is a typical deuterus cadential pattern. In the Aquitanian/Ben 39 version, these figures do not occur in their normal intervallic environment and sound foreign. With the Aquitanian reading as the lectio difficilior, it is probable that these sources preserve an authentic early reading, sung with an e‐flat, a nondiatonic note represented by whole‐tone transposition in the other versions.

5: end of v. 1: “[domi] us faciam secundum verbum tuum”: All MSS in the core group match the pitch level of Pa 776, as do most Italian MSS. The six German and French sources in the sampling, however, notate this passage a tone higher, so that the verse concludes on G: Gr 807, Be 40078, Pa 1235, Cai 61, Tri 2254, and Rei 264. The same reading is also found in the late Italian MS To 18. Because the passage is identical intervallically at the two transpositions, it is probable that this variant reflects the modal ambiguity of the verse rather than a nondiatonic pitch.

6. Beginning of v. 2 “Oravit moyses”:

  • = Pa 776: Ben 39, Lo 4951

  • 5↑ Pa 776: all others

Most MSS notate this passage a fifth above the level Pa 776 until “dum pertransiero” (with the interval changing to a fourth in some spots). Several factors, however, suggest that the lower level of Pa 776 is the unemended reading. In Ei 121, the verse begins with an i, consistent only with the lower, minority reading of the verse. The probable nondiatonic pitches discussed below, moreover, preclude the verse from being notated at the lower level without emendation. In all MSS, the verse closes at the level of Pa 776. The sources, however, join Pa 776 at different points. This variety reinforces the impression that Pa 776 is the preferred version.

7: “et dixit si” (phrase 15): Mo employs a b‐flat here, implying an E‐flat at the level of Pa 776. In one MS, Ben 34, the melodic segment in question is written a tone higher.

  • Mo (all b's are b‐flats): dcdcabcdcd/dc

  • Ben 34: dcdcacdede/dc

8. “si inveni gratiam in con [spectu]” (phrase 16):

  • = Pa 776: Lo 4951

  • 2↑ Pa 776: Ben 39

  • 5↑ Pa 776: Mo 159, Pa 1121, Pa 1235, Rei 264 (with melodic variants), Cai 61, Gr 807, Be 40078, To 18

  • 6↑ Pa 776: Ben 34, Pa 903

(p.364) Ben 34 and Pa 903 are a tone above the majority reading, a change reflected at the lower level in Ben 39, which is a whole tone above Pa 776 here. This version differs intervallically from the others on the final syllable of “gratiam,” ending with the semitone e‐f in Ben 34 and the whole tone in the other sources.

  • Pa 776: b(flat)GaG

  • Mo 159: fded

  • Ben 34: gefe

9. “ostende michi te ipsum” (phrase 17):

  • 5↑ Pa 776: all except Ben 39 (Lo 4951 joins the level of Mo and the other sources).

10: “manifeste” (phrase 17):

  • 5↑ Pa 776: all except:

  • = Pa 776: Ben 39

  • 4↑ Pa 776: Ben 34, Lo 4951

The majority of MSS are a fifth above Pa 776. Those that distinguish between b‐natural and b‐flat write b‐flat at the cadential passage on “ut videam te,” suggesting that an E‐flat is intended at the level of Pa 776 and Ben 39.

Ben 34 and Lo 4951 write most of the passage a tone below the other sources, joining (briefly) at “ut videam te.”

11. “et locutus est…lapidis” (phrases 18‐20):

  • = Pa 776: Ben 39, Lo 4951

  • 4↑ Pa 776: Mo 159, Pa 1121

  • 5↑ Pa 776: Ben 34, Pa 1134, Pa 903, Be 40078, Cai 61, Be 40078, Mod 7, Rei 264, Gr 807

Mo 159 is a fourth above Pa 776, adopting b‐flat as a temporary final, with an internal cadence on b‐flat at “potest.” This reading is matched in other MSS in the core group, but most sources outside the core group are a tone above Mo here, notating the passage with c as the cadential pitch. The reason for the change of transpositional level in Mo to a fourth above Pa 776 is not clear. Mo consistently employs b‐flats, resulting in an intervallic structure identical to that of Pa 776

12. “et protegat te” (phrase 20): 5↑ 776: all except Ben 39.

13. “donec pertranseam” (phrase 20): Here Ben 39, which has followed the level of Pa 776 until this point, is temporarily a tone above Pa 776, undoubtedly because the low B's in Pa 776 were sung as B‐flats. At the position a fifth higher, this pitch is represented as F. In Ben 39, these B‐flats are sung as C's, with no intervallic differences.

14. “dum pertransiero” (phrase 21):

  • = Pa 776: all except:

  • 4↑ Pa 776: Pa 1235 (joining with Pa 776 at “meam”)

  • 5↑ Pa 776: Ben 34, Mod 7

Most sources have a downward motion here, contradicting the readings of Pa 776 and Ben 39. Ei 121, however, has the letter l here, consistent with Pa 776 and Ben 39. (p.365) The joining of the other sources with Pa 776 here reinforces the status of Pa 776 as a preferred reading, since nondiatonic pitches were evidently no longer required at this point in the verse. Pa 1235 is a unique reading that remains a fourth above Pa 776 until “auferam” and then, after a transitional passage on “auferam manum,” joins with Pa 776 at “meam.” Pa 1235 is inconsistent in its use of a sign for b‐flat. Without a b‐flat, the intervallic structure of the passage differs considerably from that of the other readings. Ben 34 is unusual in remaining in the higher register, a fifth above Pa 776, through the end of the verse.

Roman MSS: Bod 74 lacks this offertory. RoS 22 has the following variants:

  • “SANCtificavit”: FGFEFGFEDFEDED

  • “OFferens”: aGacba

  • “offerRENS”: GaGFEFG

  • “ILlud”: GF aGa

  • “hoLOcaustum”: aGacaG

  • “vicTI/MAS”: The notes accommodating the final syllables are grouped differently and suggest a different text underlay: cbcd ededc/dedc.

Si ambulavero (65)

Although there are very few significant variants among the sources, the Beneventan MSS exhibit some small variants from the others.

Respond, “ME”: The Aquitanian MSS and most peripheral MSS adopt F as the lowest note (rather than the G in the Beneventan reading) as in Pa 776: c bcabGF bccbab (qui.) cba. Mo 159, however, matches Ben 34.

  • v. 1 “DIe”: The other MSS in the core group and the peripheral MSS have an extra 4 notes at the end of the melisma that are lacking in Beneventan MSS (the segment in boldface):

    • Pa 776: …      caF GaGF Ga(quil.)ba baGa

    • Lo 4951: melisma ends … caF Ga(quil.)b abaGa

  • v. 2 “DOmine”: There is a slight variant among MSS in the beginning of the melisma. Ben 34 has D as the bottom note, whereas other core MSS have F or E:

    • Pa 776: Melisma begins GccabGF accaGabGF …

    • Lo 4951: Melisma begins GccaGaFE accaGaGF

  • Verse endings in Lo 4951:

    • v. 1 “TU/AM”: cccacaaGaG/G

    • v. 2 “TU/AM”: ab(quil.)cbaGa/aG

Variants among peripheral MSS:

  • RoV 52: v. 2 “DOmine”: a major variant in the beginning of the melisma: bcaG GccaG aGF accaG abGFacbc acccaGF Gabcdcbabcba.

  • The other Italian MSS present readings consistent with the core group:

    (p.366)
    • v. 2 “TU/AM”: abcbaGa/aG (same as Lo 4951)

Variants in Roman MSS:

  • Vat 5319:

    • respond, “MEdio”: Gac cbc

    • respond, “meOrum”: melisma begins aGaGF …

  • RoS 22:

    • “tribulatiOnis”: Fac abaGFacbG

    • “ET”: GabaGaGaG

    • “meOrum”: aaGFGFGaGFEFED Fac abaGF acbG

    • “TUa”: ac dc bcabaGF acGaGF acbcba cabaGF acbabaga

Super flumina (66)

In v. 3, the Roman version has a longer text than the Gregorian does, reflecting either a Frankish abbreviation or an addition made in Rome subsequent to the repertory's transmission north.

In Gregorian MSS, the repetenda are a point of variance and confusion. In verses 2 and 3, the words “qui dixerunt” serve as an introduction to a melodically variant repetendum that brings the performance back to the beginning of the respond. As Pfisterer has suggested, it seems likely that the repetenda for the second and third verses, though written out, function as an introductory element, and that the whole respond was to be sung.1 One possible reason for the variant repetenda is the modal contrast between the first‐mode respond and the verses, which have the traits of mode 7.

The MSS differ in the pitch level of the second and third verses, as shown in the work of Justmann, Frasch, and Hankeln.2 Most Aquitanian MSS, including Pa 776, notate these verses a tone above the level of Mo 159, with a focal pitch of c. A few other sources place the verse a fifth above Mo 159, with a focal pitch of high f, or fourth below, with a focal pitch of low F. In all cases, the second and third verses are marked by a move away from the protus structure of the respond, creating a problematic transition to the repetendum. Pfisterer notes the resemblance of a passage in the third verse to the seventh‐mode alleluia Te decet and parts of the respond to two first‐mode graduals.3

In his edition of the Aquitanian version, Hankeln posits the level of Mo 159 as the preferred version, a hypothesis based partially on the pitch level of the repetenda.4 The second and third verses have melodically variant repetenda, and neither is melodically identical to the corresponding place in the respond, undoubtedly because of the need to provide a transition between the two modal areas. In Mo 159, the cadence (p.367) of the repetendum on “babylonis” is on F, matching the corresponding place in the respond. Assuming that the repetendum is an instruction to repeat the full respond, Mo's reading of the repetendum would allow for the performance to conclude on its original final, D. In Pa 776 and many other Aquitanian MSS, however, the repetendum is written a tone above its level in Mo 159, corresponding to the higher notation of the last two verses; at this level, a full repeat of the respond would result in an ending in a different mode, a tone higher. As a solution to this problem, Hankeln transcribes the second and third verses of Pa 776 at the level of Mo 159, a whole tone below their written level.5

While I cannot propose a definite a reason for the variants, I can offer some relevant observations. The pitch level of the majority Aquitanian version is matched in two of the Italian MSS included in the sampling, Pst 120 and RoV 52, and in two of the Beneventan MSS. The Italian MSS have the same repetendum as the Aquitanian sources do, and both notate the repetendum at the “correct” pitch level relative to the respond. The notation of these verses at the higher level, then, is not merely an irregularity of Aquitanian sources. There are intervallic differences between Mo and the version notated a whole tone higher, on “dextera mea” and “qui dixerunt,” where Mo has an E‐natural. At the higher level of Pa 776 and other Aquitanian MSS, an F‐sharp would be required to replicate the intervallic structure of Mo 159. At the level of Mo 159, an E‐flat would be needed to replicate the intervallic structure of Pa 776. Several of the MSS at the higher level periodically join Mo 159. In Lo 4951, for example, most of the verse is written at the higher level, but this version matches Mo at “dextera mea,” where the irregular semitone would be required, and again at both verse endings, so that the repetendum is placed at the normal position. Two other MSS, Pa 780 and To 18, join Mo 159 at “meminero.” The variety of the versions at the higher position perhaps suggests that they are emendations. Another possible reason for writing the verses at different positions in the background scale is a desire to avoid the G‐protus characteristics of Mo 159 throughout the last two verses. The tendency of some Aquitanian MSS, especially Pa 776, to notate G‐protus passages at their unemended pitch level, however, speaks against this hypothesis. Although the whole‐tone transposition is a common way to emend G‐protus passages, moreover, the verses of Super flumina also appear at other positions, as summarized below. This offertory illustrates well the sometimes ad hoc nature of scribal emendations and the possibility of fluidity in the performance tradition.

  • Pitch level of v. 2:

    • = Mo 159: Gr 807, Be 40078

    • 2↑ Mo 159: Pa 776, Pa 780 (joins Mo at “meminero”), Lo 4951, Ben 35, Ben 39, Pa 1235, Pst 120, RoV52, To 18 (joins Mo at “meminero”)

    • 5↑ Mo 159: Pad 47 (= Mo at “qui dixerunt”)

    • 4↓ Mo 159: Ben 34, Mod 7

  • Pitch level of v. 3:

    • = Mo 159: Pa 780, Gr 807, Be 40078

    • (p.368)
    • 2↑ Mo 159: Pa 776, Lo 4951 (joins Mo briefly at “Ierusalem”), Ben 38, Ben 39, Pa 1235, Pst 120, RoV 52

    • 5↑ Mo 159: Pad 47

    • 4↓Mo 159: Ben 34

Melodic variants in Beneventan MSS: The independence of the Beneventan MSS is especially apparent in the long melismas of Super flumina, which can be summarized as follows:

  • v. 1, “susPENdimus”: Beneventan MSS have a greatly shortened melisma, as in Ben 34: D CDF F Ga aGa FacGF FGD CD C DFF G abcba.

  • v. 1, “interogaVERrunt”: variant reading in Beneventan MSS, as in Ben 34: aFGE FGF FGDEDCD CDF F Ga aGa a Gac cccc dca.

  • v. 1, “alieNA”: Ben 34: D FF FD FGF FGDE D D(ori.)C (similar in other Ben MSS).

  • v. 2, “memineRO”: Near the middle of this melisma, corresponding to the last twenty‐seven notes of Mo's melisma, most sources have a repetition, as in Mo 159: Fa (quil.) b‐flat a (quil.) FG Fa (quil.) b‐flat a (quil.) FG. Ben 34 lacks this repetition, but it is present in the other Beneventan MSS.

  • v. 3, “hierusaLEM”: Beneventan sources present a different reading of the melisma, as in Ben 39 (written a whole tone above Mo 159): aE GF Ga aaG ab abaE a baG ccc abaa cdca c caG c caa c cGbaG GF G acaca cc caGG DEFG.

Other melodic variants:

Respond, “siON”: This melisma is a point of variance. The early adiastematic sources examined present a consistent reading, but Lo 4951 presents a shortened version of this melisma, omitting the internal repetition: FD FGaGa caGa FD FG FG FG FF DED. According to Justmann's transcriptions, some of the same notes are omitted not only in two other Aquitanian MSS, Ma 18 and Ma 51, but also in three Italian MSS.6 The three Italian MSS, in fact, present a reading of the melisma nearly identical to that of Lo 4951, suggesting that this variant entered the tradition at an early date.

v. 1, “interogaVErunt”: Aquitanian MSS begin this melisma slightly differently from Mo 159. The Aquitanian version is similar to the Beneventan reading at the start of the melisma, as in Pa 776:

  • Mo 159: aFGFDCD FFDCD

  • Pa 776: aF GE FG FF GDCD

  • Ben 34: aFGE FGF FGEFEDCD

Lo 4951 presents a longer reading of the end of this melisma, where the first part is repeated. The added portion of the melisma is indicated in boldface: aFG DFG FFG DCD CDFFG a a (ori.)G aa Gacc dcaFa GaG FGFD FECD CDFFG a a(ori.) G aaG accdca.

To 18 is missing a section of this melisma, corresponding to notes 51–60 of Mo 159.

  • v. 1 “cantiCOrum”: Lo 4951: aFG DEDCD

  • v. 2 “alieNA”: To 18 D FF GDCD

  • v. 2 “memineRO”: To 18 is missing the last 23 notes

  • v. 3 “IerusaLEM”: Lo 4951 leaves out one segment of the melisma: G(ori.) D Ga FGFD GaGFacccaba [missing segment here] acdca ccaG (etc.)

(p.369) Roman MSS:

Vat 5319:

respond, “organa nostra”: This passage appears to be written a third too high, but a custos after “nostra” corrects the error.

  • v. 1 “cantiCUM”: FGFEFD

  • RoS 22:

    • “tuI”: G

    • “syON”: DC DFEF aGaGEFGF GaGFED FFEDFED

Vir erat (67)

In most MSS, Vir erat is notated in the transposed second‐mode range, with a final of a, reflecting the use of nondiatonic pitches in the verses. Three sources in the sampling, Ben 34, Ben 39, and Rei 264, begin the respond in the normal plagal range and later move to the

Appendix 3 Commentary on Edition

Example 67.1a

(p.370)
Appendix 3 Commentary on Edition

Example 67.1b

transposed range (Ben 34 and 39 at the beginning of v. 1 and Rei at the beginning of v. 2). In Ben 39 and Rei, however, the repetendum cue begins at the affinal position, a fifth above the level of the corresponding passage in the respond. This discrepancy may suggest that the respond and first three verses were sung at the same pitch level, reflecting the notated level of other MSS, and that the move to the affinal position does not reflect the pitch level in performance.

Vir erat presents two problem spots: the close of v. 3 and the nine repetitions of “ut videat bona” in v. 4. In the transcriptions, Pa 776 serves as the base Gregorian reading and is transcribed in the normal second‐mode range, with a final of D. This transcription facilitates comparison to the Roman version, which is also notated with a D final. In the supplementary examples below, however, Pa 776 is transcribed at the affinal position to facilitate comparison with other Gregorian sources.

Verse 3: [lapidum est] “fortitudo” (phrase 15) to the end of the verse: Several readings of this problem spot are shown in example 67.1 In many MSS, including Mo 159, most of this passage is written a fourth below the level of Pa 776, but the interval of transposition is briefly a third at “fortitudo mea.” As shown in example 67.1, this passage is also a point of

(p.371)

Appendix 3 Commentary on EditionAppendix 3 Commentary on Edition

Example 67.2A

(p.372) Appendix 3 Commentary on Edition

Appendix 3 Commentary on Edition

(p.373) Appendix 3 Commentary on Edition

Appendix 3 Commentary on Edition

(p.374) Appendix 3 Commentary on Edition

melodic variance, particularly on “mea.” Pa 1121 (line 5) presents an alternative reading of the melisma that differs from that of Pa 776, Mo 159, and Ben 34, but ultimately joins the pitch level of Pa 776. Pa 1235 (line 6) gives a reading of the melisma on “mea” similar to that of Pa 1121 but written a tone lower, with a different intervallic structure.

The reasons for the variants are unclear. The various versions differ intervallically on “fortitudo mea,” as is evident in a comparison of Pa 776 (line 1) and Mo 159 (line 3). On “mea,” there are two intervallic differences. On the final syllable, semitone f‐e in Pa 776 appears as the whole tone d‐c in Mo and Ben 34, and as the whole tone e‐d in Pa 1235. The melodic differences between Pa 776 and Mo 159 also result in an intervallic difference at the melisma on the first syllable. In Pa 776 the last four notes of the melisma on mea are defe, whereas Mo has b (natural) cdc, implying the use of a nondiatonic pitch. If the reading of Mo and Ben 34 is taken as the preferred intervallic structure, then an eb would be required to replicate this structure at the level of Pa 776. In several MSS, however, the passage is notated at the level of Pa 776 and simply written with an e‐natural, as shown in the reading of Be 40078 (line 2). The hypothetical e‐flat in Pa 776, however, is equivalent to an a‐flat in the normal plagal range, a nondiatonic pitch that is implied infrequently in the chant repertory.

With the next passage, “aut caro mea…” the two versions are a fourth apart and remain so through the end of the verse. Be 40078 employs a b‐flat at “enea,” which is given as a parenthetical b‐flat in Pa 776 and a parenthetical E‐flat in the base transcription.

pitch level= Mo (with melodic variants): Pa 903, Lo 4951, Ben 34, Ben 34, Ben 38, Pst 120, RoV 52, Mod 7

(p.375) pitch level= Pa 776 (with melodic variants): Pa 1121, Pa 780, Be 40078, Rei 264, Pa 1235

Verse 4, “ut videat/am bona” (phrase 19) to the end: The passage that consists of nine statements of “ut videat/am bona” is among the most variant spots in the repertory, with several distinct versions. Transcriptions of v. 4 are provided in examples 67.2a and b. Example 67.2a shows variants among some of the core sources and additional Aquitanian MSS. Example 67.2b shows variants among selected peripheral sources. In terms of pitch level, many of the core sources match either Pa 776 or Ben 34. Pa 1121, for example, presents a reading similar to that of Pa 776. The pitch‐level profile of Ben 34 is matched in the other Beneventan sources and, in a shortened version, Mo 159. As described in chapter 3, Pa 776 and the Roman version are very similar in their pitch‐level profiles.

Before considering the reasons for the variants, some transcription problems warrant consideration. Some Aquitanian sources present a problem I have not been able to fully solve. In the transcription of Pa 776, I have determined the starting pitch level of v. 4 based on the end of v. 3. The resulting transcription matches the pitch‐level profile of the Old Roman version and other diverse sources, such as Pa 1121 and Gr 807. Pa 776, however, exemplifies a persistent problem in the offertories: an inconsistency between the notated level of verse and repetendum cue. In most Aquitanian sources, the final verse is followed by a cue to “vir erat,” the beginning of the respond. In Pa 776, the repetendum cue is notated an octave above the level of the corresponding place in the respond, as shown in line 1 of example 67.2a. When Pa 776 is transcribed at the affinal position, it starts on low E. The repetendum, however, starts on high e. The same inconsistency is found in other MSS, including Pa 1121, Paris 909, and Pa 1137.

I see two possible explanations for the incompatible repetendum cue. One is that the notated level of the fourth verse represents the intended pitch level and that the repeat of the respond was really sung an octave higher than its original pitch level. If so, perhaps the higher notation is attributable to the dramatic expansion of range in the final verse. The disparity of range between the opening of the respond and the close of the v. 4 may have created a practical problem for the singers in the transition to the repetendum. If the repetendum was sung at its original pitch level, it would have required a sudden downward shift of more than an octave. As a solution, the repetendum may have been sung an octave above its original pitch level. A second possibility, consistent with Hankeln's theory about the “incorrect” repetendum cues, is that the verse was sung an octave below its original level, and that the repetendum cue is correctly notated.

I consider the first alternative more likely in this case. A performance of the final verse an octave below its written level seems counterintuitive, simply because it would be so anticlimactic and incongruous with the range of the piece as a whole. The Roman version, discussed below, attests to a tradition in which the repetendum was performed at a position a fifth higher than the corresponding passage in the respond, with the performance concluding on the affinal a. The notation of the repetendum cue in Pa 776 and 1121 suggests that a similar tradition prevailed in Aquitaine, with the repetendum sung an octave higher. While I find Hankeln's theory generally persuasive, then, I think it more likely in this case that the repetendum was simply sung an octave higher and that the performance concluded on high d, in keeping with the climactic effect of the final verse. The ending on high d, which has no theoretical grounding, is reminiscent of some early sequences. In the edition, I have transcribed the verse and repetendum in Pa 776 at their written level.

Two other Aquitanian sources, Pa 1134 and Lo 4951, have been included in example 67.2a to illustrate the variants in the Aquitanian tradition. In the case of Pa 1134, the pitch (p.376)

Appendix 3 Commentary on EditionAppendix 3 Commentary on Edition

Example 67.2B

(p.377) Appendix 3 Commentary on Edition Appendix 3 Commentary on Edition (p.378) Appendix 3 Commentary on Edition level again presents problems I have not been able to fully resolve, and the transcription is in part hypothetical. I have included it simply to illustrate the problems inherent in the tradition. The verse begins on a new line. I have transcribed it with the premise that the repetendum was sung at the same pitch level as beginning of the respond and that the starting notes of the repetendum (at the affinal position) are thus E‐G‐a‐b. The resulting transcription produces a reading that is a fourth below Pa 776 until the seventh statement of “ut videat…” where it is an octave lower. In performance, this reading would require several nondiatonic pitches to derive the correct interval structure. On the third statement of “ut videat bona,” for example, the D‐F interval on the last syllable of “videat” is equivalent to G‐b (natural) in 776 and Ben 34, requiring an F‐sharp. The corresponding passage in the sixth statement has the notes a‐c‐a, corresponding to G‐b‐natural‐G in Mo 159. If Pa 1134 is transcribed with the assumption that the repetendum was sung an octave higher, however, the resulting version of the verse would be an octave lower than it is written here, an even less probable reading. Similarly surprising versions, however, are found among the sources transcribed in example 67.2b, which are unambiguous with respect to pitch.

These shifts in pitch level most likely reflect the need for nondiatonic pitches. The transcription of Pa 776 at two different levels, the normal plagal range in the base (p.379) Appendix 3 Commentary on Edition transcription and the affinal position in the example 67.2a, illustrates the problems that arise at both positions. As I have shown, the end of the third verse, if notated in the normal plagal range, requires an E‐flat. The final verse, moreover, requires low B‐flat, represented as F at the higher level. For these reasons, the whole offertory is typically written at the affinal position. At the affinal position, however, new problems arise. The reading of Pa 776, transcribed at the affinal position in line 1 of example 67.2a, differs intervallically from the others in the sixth statement of “ut videat bona,” where it has the pitches d‐f‐d‐e. In Mo 159 (line 2) and many other sources, this passage is written a fifth lower, with the pitches G‐b‐natural‐G‐F. The use of b‐natural rather than b‐flat in Mo implies that an F‐sharp would be required at the level of Pa 776, here and at the subsequent cadence on “bona.” This impression is reinforced by the Italian source Pst 120 (example 67.2.b, line 5), which matches Pa 776 until this point. The passage in question, however, is a whole tone below Pa 776, with the pitches c‐e‐c‐d, again implying an f‐sharp at the level of Pa 776. Pst 120 remains a tone below Pa 776 for the rest of the verse. The difficulties of writing the offertory at both the normal and affinal position are undoubtedly the primary reason for the diversity of versions.

An additional reason for the variants may be the previously mentioned transition to the repetendum. In many sources, the final verse closes in a range much higher than that of the (p.380) Appendix 3 Commentary on Edition respond. If the repetendum was sung at the same pitch level as the respond, the transition may have posed a problem for singers. In Pst 120 (example 67.2b, line 5), for example, the respond is notated at the affinal position, beginning with the notes E‐G‐a‐b. The final verse, however, closes in the higher octave, with the third d‐e‐f. The transition to the repetendum, then, would have required a leap of a ninth. Although Pst 120 has a repetendum, the cue has no notation, and it is unclear how the singers would have resolved the problem. Several other readings of the verse, however, conclude in a range and tonal area more congruous with that of the respond. Mo 159, for example, closes the verse with the third a‐b‐c, consistent with the transposed second‐mode range of the respond.

A third reason for the variety of versions may lie in a simple desire to normalize the melody's extraordinary range. The diversity of range among the sources is indeed striking. In Pa 776 the verse spans nearly two octaves. Ben 34 and Mo 159, by contrast, have a range from F to g, just over an octave (lines 2 and 3 of example 67.2a). Perhaps the most interesting emendation of range is that of To 18 (example 67.2b, line 6), which begins in the range of Ben 34 and Mo but closes a fifth lower, perhaps to facilitate the transition to the (p.381) repetendum. This reading lacks all traces of the climactic expansion of range found in the other versions.

With these observations, we may hypothesize that the striking diversity of readings in the sources results from a combination of factors: pitches that were unavailable on the gamut, an irregular modal structure, and an extraordinary range and resulting problematic transition to the repetendum. While the extent to which the different versions reflect actual performance is difficult to ascertain, the problematic transition to the repetendum probably did result in a diversity of performance traditions, perhaps dependent on the preferences and voice ranges of individual soloists. Some of the transcriptions provided in the examples, however, present problems whose solutions are elusive. For reasons that are unclear, the scribe of Be 40078 (example 67.2b, line 2) sets much of the verse in the G‐protus tonal area. A few sources depart from the norm in their repeat structure. Mo 159, for example, has only seven statements of “ut videat bona.” In the seventh and eighth statements of “ut videat bona,” To 18 presents a reading that is unique among the sources examined: both statements are set to the more elaborate melody of the sixth statement. Lo 4951 (example 67.2a, line 5) differs from the other sources in the sixth statement of “ut videat bona”: most sources repeat the melodic material of the third statement at a higher pitch level, but Lo 4951 has different material here.

The Roman version: Bodmer 74 lacks this offertory. Several features of the version in Vat 5319 warrant comment. First, the letter v, normally used to mark verse divisions, occurs at most points of text repetition in the verses (i.e. in phrase 18: V. Quoniam V. Quoniam V. Quoniam). These seem to simply serve as a visual marker of text repetitions. I have not included these in the transcription, but rather followed the verse division of the Gregorian reading. The offertory is written in the normal plagal range. It is likely that the low B's at the end of verse 2 on “agam” were sung as B‐flats, just as they clearly were in the Gregorian version. The final verse has a pitch‐level profile similar to that of Pa 776. The repetendum is written a fifth above the level of the corresponding place in the respond. The melody of the repetendum, moreover, differs from its initial statement, appearing in a much more florid version and coming to a close on the affinal a. The melodic differences suggest a special performance practice, with a melodically varied repetendum that lacks an equivalent in the Gregorian reading.

The horizontal alignment is briefly lost at several points, especially at the ends of lines. At “carnem quoque eius” (phrase 6), the cadential figure on “eius” is incorrectly written a tone too high. The mistake is evident both from the melodic figuration (the cadential pattern, on C, occurs many times in Vir erat and, in many other offertories, on G) and in the custos that follows “eius,” a D that is indicated at the correct level and not the written level. The alignment is corrected at the beginning of the following line. I have transcribed the passage a tone lower than written, and the transcription differs from that of MMMA. A very similar situation is found at the first “quae est enim” (phrase 12) and the first syllable of “numquid” (phrase 15), which is written a third too low. Each time the mistake is corrected by the custos that follows, and I have used that custos as the guide in transcription.

The final verse presents two different readings of the text, “videam” (common in German sources) and “videat” (the majority reading elsewhere in Gregorian sources). The use of the b‐flat sign in the verses on “calamitas” (verse 1) and the parallel passage on “est” (verse 2) is very rare in the Roman MSS.

(p.382) Recordare mei (68)

This offertory is the only one with two distinct melodic and textual versions, which have been discussed by Justmann, Maiani, and Steiner.1 The version given in the base transcription, which I shall call version 1, is found in the Roman, Beneventan, Aquitanian, and many French MSS; it also circulates as a Gregorian and Roman responsory. A second version, found in most German and Italian MSS, as well as some French sources, is melodically distinct from the first and shows some textual differences. Mo 159 is unique among MSS examined (and also among Justmann's larger sampling) in transmitting both versions.

The Roman version of the offertory has no verses. In each Gregorian version, there are differences in verse transmission. Version 1 has Recordare quod steterim as a first verse in some Aquitanian and French MSS and Memento domine et ostende as a second verse in some Aquitanian MSS. Yet another verse, Memento nostri domine, circulates in some Beneventan MSS. In version 2, the part of version 1 that begins “everte cor eius” is transmitted as a verse, and one MS cited by Justmann has Recordare quod steterim as a second verse.2 The status of these verses as later additions is evident from the regional variation and their absence in the Roman tradition.3

As Justmann, Steiner, and Maiani have argued, version 1 appears to be earlier than version 2; version 1 is found not only in the Roman tradition but also in Mont‐Blandin and Corbie. Steiner has argued persuasively that the second version of Recordare mei was created after the piece was adopted as a responsory, in order to distinguish the two liturgical uses.4 Another factor consistent with the chronological priority of version 1 is its melodic correspondence to the Roman version. Although version 2 appears to be based melodically on version 1, it has a long melisma at the end of the section designated as a respond.

Melodic variants among core MSS in version 1:

  • opening, “RE/COR/DA/RE”: Pa 776, Lo 4951: F/aG (liq.)/a/aG

Although the rest of the respond is at the same level as Ben 34 among core MSS, Hankeln notes that the part beginning “nos autem” is written a fifth lower in Pa 903, Pa 1135, and Mo.5

Final cadence: Mo 159: [recordare] “ME/I”: G/G (a less ornate version than the majority reading).

Mo 159 classifies this chant as a tetrardus melody with a final cadence on G. The final syllable of “aeternum” is simply aG rather than ba.

Roman MSS: Bodmer 74 lacks this offertory. There few variants between RoS 22 and Vat 5319.

RoS 22:

  • [hos] “MEum”: cdcba bcdcbabG

  • “eTERnum”: end of melisma: cbcbabG cbcd

(p.383) De profundis (69)

Although the Roman MSS lack verses for this offertory, they are present in the earliest Gregorian MSS. Mont‐Blandin, however, has a different second verse indicated, “Quia apud te qui propiciato est” (Ps. 129:4).

Mo presents a reading of De profundis that is unique among the MSS included in the sampling. It is notated a fourth above the reading of Ben 34, with a final of G. Despite the G final, the scribe of Mo includes De profundis with the protus chants. Although most passages are protus in quality, with a b‐flat, Mo does employ a b‐natural twice: on the second syllable of “clamavi” in the opening and at the ending of the first verse, where the b on the second syllable of “tui” is a b‐natural. These notes are equivalent to F‐sharp at the majority lower level of Ben 34. If Mo's intervallic structure were replicated at the level of Ben 34, “claMAvi” would read C‐D‐E‐F‐sharp‐E, and “tuI” (phrase 3) would be D‐F‐sharp‐D. The second b‐natural in Mo is not matched when the same melodic material returns at the end of the second verse, on “sustineBIT” (phrase 5). Here all b's are flat. Because this possible nondiatonic practice is witnessed in only one MS included in the sampling, Mo 159, I have not shown it in the base transcription of Ben 34.

Because of the close proximity to F, it is likely that the low B in the opening of Ben 34 was sung as a B‐flat. In most MSS, the first note is A rather than B.

Variants among core MSS: There are very few melodic variants in either core or peripheral MSS. The Aquitanian MSS and Mo 159 differ from other MSS in v. 1, at the cadence on “tue.” Most MSS in the core group have the standard cadential pattern DEFEF/ED. Italian and German MSS have DEFEFEDE/ED, a reading closer to Ben 34, and this version is also found in Pa 1121). Ben 34 exhibits another small difference from the majority of sources in the respond at “ad” (phrase 1). Most MSS have the reading of Pa 776, FEFD.

  • Mo 159:

    • respond, “AD”: b‐flat a b‐flat G

    • respond, “MEam” (all b's are flat): Gbb FaG Gba bcaG aG aGF GF(quil.)D

  • Pa 776:

    • respond, “meAM”: melisma begins DFFFDED D FEFG ED

  • Pa 1121:

    • respond, “meAM”: melisma begins DFF F(oei.)DCED

Peripheral MSS:

Pa 1235:

  • v. 1 “TU/E”: DEFEFEDE/ED

  • v. 1 “intenDENtes”: CF GFCD F FED F FD(liq.)

  • v. 1 “TUi” (significant variant): D DCDA DF EFD DCDA DFEFDCD FGa FD EFDC FFFED.

When this melodic material returns in v. 2, Pa 1235 has some of the same variants, but not all: v. 2 “domiNE”: DDCDA DFE FDCD FGaFD EFDC FFFED.

RoV 52: v. 2 “TU/AM”: DEFEFEDE/ED

Pst 120:v. 2 “TU/AM”: D EFEFEDE/ED

(p.384) Be 40078:

  • respond, “MEam”: melisma begins FF FC FD DFE

  • v. 1 “TU/AM”: DE(quil.)FEFEDE/FD

  • v. 1 “intenDENtes”: FGFFCD FF FED F F FD(liq.)

  • v. 2 “oratiOnem”: DED FF FED FEF

Roman MSS: Bod 74 74 lacks this offertory. RoS 22 has the following small variants:

  • “exAUdi”: c cbca bagF acbabaG

  • “oratiOnem”: c

  • “meAM”: GcaGbaGFGa

In virtute (70)

This offertory exhibits several pitch‐level variants. The differences in its overall pitch‐level profile are similar to those of Desiderium (89), Domine in auxilium (27), and Iustitiae (29): the sources differ in their indication of the relative pitch level between the respond and second verse. In Ben 34, the respond and first verse are notated at the affinal position, with a final of c, and second verse is in the tritus authentic range. In a few MSS, however, the respond and first verse are notated in the normal plagal range a fifth below Ben 34, closing on F, and the second verse is written at the level of Ben 34.

The variants in pitch level arise from problematic pitches in at least three places. In the opening phrase of the respond, Mo 159, notated at the affinal position, employs b‐flats (given as a parenthetical b‐flat in the transcription of Ben 34). A b‐flat is also found in most other sources that employ a sign for it, such as Be 40078 and Rei 264. At the lower level of transposition, this pitch would be E‐flat. Mo 159 also employs some passing b‐flats in the first verse; in Ben 34, these pitches are simply c's.1 In the final verse, Mo uses b‐flat and F, which would be E‐flat and low B‐flat at the lower level. In Ben 34, a corresponding b (flat) occurs before the cadence on “impones.”

In its indication of the relative pitch level between the respond and two verses, Ben 34 is clearly the preferred reading, as suggested by its presence in the majority of sources and the nondiatonic pitches that would result if the piece were written in the untransposed plagal range. With the preferred status of Ben 34 in mind, other possible nondiatonic pitches occur in the final melisma of the second verse, which is a point of variance among the sources. Several different versions are shown in example 70.1. In Mo 159 (line 2), most of the final melisma is written a whole tone below that of Ben 34 (line 1) and employs both b‐natural and b‐flat. If the reading of Ben 34 is considered to be a fifth above the normal plagal range, as seems likely, then Mo's reading, a whole‐tone below Ben 34, represents a shift in the transpositional level to a fourth above the untransposed range. A downward transposition of Mo's final melisma by a fourth would result in an F‐sharp at the beginning of the melisma. In view of the proximity of f and the b‐flats that immediately follow, however, this b‐natural is dubious and may be a scribal error.2 A more likely candidate for a nondiatonic pitch is the (p.385)

Appendix 3 Commentary on Edition

Example 70.1

interval between the seventh and eighth notes of the melisma. The whole‐tone g‐f in Ben 34 and Pa 776 appears as the semitone f‐e in Mo 159 and many other sources.

Most sources in the core group, including Pa 776, present a reading of the final melisma very similar to that of Ben 34. In Pa 1121, however, this melisma is written a whole tone lower, matching the level of Mo 159, with some melodic variants. Outside the core group of sources, the final melisma exhibits melodic variants that result in brief shifts of transpositional level, as shown in the readings of Pad 47 (line 4) and Rei 264 (line 5). Although these readings mostly match the level of Ben 34, certain segments deviate from it by a whole tone.

In some sources included in the sampling, including Gr 807 and Mü 10086, the modal structure of the offertory as a whole is normalized: the respond and first verse are written in the untransposed plagal range, a fifth below Ben 34, and the second verse is written at the level of Ben 34, in the authentic range. These sources, however, require an adjustment to circumvent the nondiatonic pitch E‐flat at the beginning of the offertory, the notes equivalent to the b‐flats in Mo 159 (given parenthetically in the transcription of Ben 34). (p.386) In Pst 120, the whole opening passage, “in virtute tua domine,” is written a tone above its normal position, allowing the problematic interval C‐E‐flat to be written as D‐F. In Mü 10086, it is simply written as a normal E. The downward transposition of the respond and second verse regularizes the modal structure of the offertory: the second verse appears unproblematically as a tritus authentic melody in relation to the plagal respond. These sources, however, present an additional problem: the new disparity of range between the respond and second verse and the transition to the repetendum. In all three sources, the final melisma of the second verse is altered, presumably either to facilitate the transition to the repetendum or, in performances without a repetendum, to bring the offertory to a close on the final F. The scribe of Pst 120 (line 6 of example 70.1) extensively alters the melisma so that it concludes on F. In Mü 10086 and Gr 807, the melisma is simply transposed down a fifth, as shown in Gr 807 (line 7), producing an ending on F.

Respond and v. 1= Ben 34 (final of c): Mo 159, Pa 776, Pa 1121, Pa 1132, Pa 780, Pad 47, Mod 7, RoV 52, Pia 65, To 18, Rei 264, Be 40078, Tri 2254.

Respond and v. 1 5↓ Ben 34 (final of F): Gr 807, Mü 10086, Pst 120.

Verse 2, final melisma (beginning with “super”):

  • = Ben 34 (sometimes with melodic variants): Pa 776, Pa 1136, Pa 1137, Lo 4951, Pa 1132, Pad, 47, RoV 52, Mod 7, Pia 65, Rei 264, Be 40078, Tri 2254

  • 2↓ Ben 34 (with melodic variants): Mo 159, Pa 1121

  • 5↓ Ben 34: Mü 10086, Gr 807, Cai 61.

Notes on individual sources:

Pa 780: final melisma of second verse: The starting pitch of the final melisma is not visible on the microfilm.

Cai 61: The respond is written in the transposed plagal range, v. 1 in the untransposed range.

Roman version: In the Roman tradition, this offertory lacks the second verse that circulates in Gregorian MSS. Vat 5319 shows some significant variants from Bodmer, as follows. Differences are indicated in boldface.

  • Vat 5319 is written in the normal plagal range, closing on F, a fifth below Bod 74.

  • Respond, “super salutARE”: accc(ori)ba/Gccc(ori.)baG.

In the verse, Vat 5319 has a shorter form of the melisma on “diErum”:

baGaGFGFEFEDFEDEDCDFFFGaGaGFGFEFGaGF.

melisma on “SEculi”: In Vat 5319, the last thirteen notes of the melisma are GaGaGFGFEGaGaG. The boldface segment shows a change in the interval of transposition from a fifth to a sixth. In transposed plagal range, the passage would be written a tone lower than the corresponding passage in Bod 74.

RoS 22:

  • “saluTA/RE”: egfe/dgfed

  • “vehemenTER”: dedcbcd

(p.387) Iustus ut palma (71)

Following the second verse, many Gregorian MSS have a melodically varied repetendum to “sicut caedrus” that is lacking in the base reading of Ben 34 and in the Roman version.

Ben 34 is the only Beneventan source in the sampling to transmit this offertory with its verses. This MS, selected as the base version, presents a shortened version of the final melisma in the second verse. The reading of Mo 159 is more typical (the notes in boldface are lacking in Ben 34):

  • Mo 159, v. 2 “floreBIT”: aa(ori.)G cccaa(ori.)F GFE FG(quil.)aG aFD FF EG(quil.)aG aa(ori.)G ccaa(ori.)F GFE FG(quil.)aG aFD FC EG FGF GF GFE FGFEDC DE(quil.)FE FGFEDCD FGFGFEF.

Two other very minor points of Beneventan independence occur in v. 2, on “domiNI,” where most MSS have the reading FGFE rather than EGE of Ben 34, and on “atriis,” where a more typical reading of the first two syllables is that of Mo 159: GaGa/aaGa.

This offertory otherwise shows few significant melodic variants between MSS. Most MSS in the core group present readings very similar to that of Ben 34.

  • Mo 159:

    • v. 1 “maNE”: DFDF

    • v. 2 “in atriIS”: caG aFE FGFF(ori.)E

    • v. 2 “miseriCORdiam”: G FFFG FFF(ori.)D GaFF(ori.)EF

  • Pa 1121: v. 2 “miseriCORdiam”: G FF FG FF F(ori.)D Ga FF(ori.)EF

  • Pa 780: v. 3 “atriIS”: aaGa caG aF EF(quil.)GFE

  • RoV 52: v. 2 “miseriCORdiam”: GFFGFF GFD GaFEF

The varied repetendum at the end of v. 2 differs from Mo 159 and the Aquitanian MSS at the end of the melisma on “CAEdrus”: abc dca GabaGa cccGbaG aca ccc cdededcbaGa.

  • Pad 47:

    • v. 2 “maNE”: FGFDED

    • v. 1 “boNUM”: GFGF FD FF FD FGFE

    • v. 2 “miseriCORdiam”: GFFFG FF FD GEG

    • v. 2 “Et veritatem” (words and music) is missing

    • v. 3 “doMI/ni”: GEG

    • v. 3 “atTRIIS”: a different text underlay is indicated, with the longer melisma on the second syllable.

    • v. 3 “floreBIT”: a aG ccca FE GFGFE FGaGaFEF FGaGa aG GGEG GFE

  • FGaGaFE FD EG FGF GFGFE (etc.)

  • Pst 120:

    • v. 1 “maNE”: FGFF CDC

    • v. 2 “miseriCORdiam”: G FFG FF FG GaF FEF

    • (p.388)
    • v. 3 “atriIS”: caG aFE FGFE

  • Pa 1235: v. 1 “boNUM”: GFGF FD FF FD FGFE

  • Cai 61:

    • v. 1 “boNUM”: GE GF FD FF FD FGFE

    • v. 3 “plantatus”: indicates a different text underlay, with the long melisma on the last syllable.

  • Be 40078:

    • v. 1 “tuO”: FGF F(ori.)

    • v. 2 “miseriCORdiam”: GFF FG FF FD GaF FF(ori.)

Roman MSS: In Vat 5319, the order of verses 2 and 3 is reversed. There is one significant variant between Bod 74 and Vat 5319, in the final melisma of the verse Plantatus. Vat 5319 presents a much longer (and slightly varied) version of the melisma. The notes in boldface are lacking in Bod 74.

  • “floreBIT”: aGaGbaG baGaGFGFF(ori.)EDF aGaG EFDEDC FGFGFDE GFGFEDCDCD FGFEDaGbaG baG aGF GFEDF aGa GEF DEDC FGFG FDE GFGFEDCDCD F FEDaGaGbaGF F(ori.)EDC DCD FGF F(ori.) GaGFEDEDFF

  • Other variants in Vat 5319:

    • respond, “multipliCAbitur”: end of melisma goes to a instead of b: GF Ga(ori.)GFG

    • v. 2 “BOnum”: melisma ends GFGabaGFGF

  • RoS 22:

    • “floRE/BIT”: GaFGF/DGFD

    • “QUAE”: GaGaGFFEDC

    • “LI/BAno”: GaGaGFFEDC FGa/aGaG c cbcbaGa

    • “multipliCAbitur”: Fa ba GaGFGFE aGaGF GabaGFG

Anima nostra (72)

This offertory exhibits variants in its notated pitch level and thus in the tonal structure of the piece as a whole.

Pa 776, transcribed with the final on D, represents the probable unemended version of the chant. The end of the second verse (phrase 8) employs two pitches that are unavailable on the gamut, low B‐flat and the F below Γ. Phrase 8 is irregular in modal structure. Melodic material associated with the first and second modes is projected a fifth below its usual pitch level. The gesture on “nos,” for example, often marks a caesura in first‐mode offertories, and the figure on “in” is a typical opening phrase for first‐mode pieces.

Despite the irregularity of this version, the MS evidence suggests that it is the preferred reading in the pretheoretical tradition. It is the majority reading among Aquitanian MSS, and most other MSS replicate the pitch‐level profile of Pa 776 a fifth higher, at the affinal position. In a common minority version, found in Ben 34 and other MSS, the respond and first verse are written in the normal plagal range, with a final of D, and the second verse is (p.389) written a fifth higher. A few MSS included in the sampling adopt distinctive emendations, moving to the higher register within the second verse, as summarized below.

The repetendum in Pa 776 is written a fifth too low. With other instances of this phenomenon, these “incorrect” repetendum cues seem to indicate a discrepancy between the written and notated levels of the verse, as Hankeln has suggested. In Anima nostra, however, I consider it more likely that the repetendum was actually sung at this lower level in the pretheoretical tradition. Because the closing of the second verse has moved into a new tonal area, singing the repetendum at its normal position (starting on F) would require a leap of a seventh between verse and repetendum.1 The significative letters in Ei 121 support the hypothesis that the pitch‐level profile of Pa 776 is the preferred version. The scribe places an i (inferius) at the beginning of the second verse, consistent with the majority version of Pa 776 but not with the emended version of Ben 34, where the second verse is written a fifth higher. This emendation corrects both the irregular tonal structure of the melody and the problematic transition to the repetendum.

Another notable feature of Anima nostra is the melodically variant repetendum, with an added melisma on the final syllable of “laqueus” and some other small variants. The varied repetendum is found in most other MSS from the core group.

Summary of pitch‐level profile:

  • = Pa 776: Pa 780, Lo 4951, Pa 1121, Cai 61, Be 40078 (at affinal position), Mo 159, Pist 120 (at affinal position),

  • Second verse 5↑ Pa 776: Ben 34, Pa 1235, Gr 807, Rei 264, Mü 10086, Pad 47, Mod 7

Unique emendations: RoV 52 is written a fifth above Pa 776 beginning at “qui non dedit.” Pa 903 is written a whole tone above Pa 776 on the last syllable of “captionem” and a fifth above Pa 776 on “eorum,” with a transitional section at “dentibus.”

Melodic variants:

  • v. 1, “doMInus”: Most core and peripheral MSS have FGF.

  • “eO/RUM”: This final melisma of the second verse is a point of variance, perhaps because of the theoretically nonexistent pitch F below Γ and the problematic transition to the repetendum. The following are sample readings:

  • Pa 1121 (similar to Pa 776): DFD BBΓ CDCB CBF ΓBABA FΓ /FΓ

  • Mo (at affinal position): G ca FF(ori.)D GaGF GFD FGF FE(quil.)DE/DF

  • Ben 34 (at affinal position): G Gca FF(ori.)D FaGF GEC DFE FECD/CD

  • Be 40078 (at affinal position): GcaF FF(ori.)D FaGF GFD FGEFEDE/DDE

  • Pad 47 (at affinal position): Gba GGF FaG FGFD FGFE CD/CD

Roman MSS: There are no significant variants.

Inveni (73)

In both the Gregorian and Roman traditions, Inveni has an inconsistency of verse transmission. Gregorian MSS fall into two groups. In the first group, represented by Pa 776 in the transcription, the second verse begins with “veritas mea cum ipso in nomine meo exaltabitur (p.390) cornu eius.” The melody and text are the same as that of the offertory respond Veritas mea (76) but notated a fifth higher to correspond to the tetrardus modality of Inveni. The verse continues “et ponam in saeculum saeculi…” The second group transmits a truncated version of the verse, which begins with “et ponam.” The two traditions have the same melody for the section beginning at “et ponam.” The MSS may be summarized as follows.

Longer version of the verse, beginning with Veritas mea: Mt. Blandin, Silvanectensis, Lo 4951, Pa 776, Pa 780, Pa 1121, Ei 121, SG 339, Cha 47, Gr 807 (with many melodic variants.)1

Shorter version of the verse, beginning with Et ponam: Mo 159, Beneventan MSS, Compiègne, Cai 61, Pa 1235, Rei 264, Be 40078, Pad 47, Pst 120, RoV 52, Pad 47, Mod 7.

As the summary shows, both traditions are found in Antiphonale Missarum Sextuplex and early adiastematic MSS. Most later MSS, however, have the shorter version of the verse. Among pitch‐readable MSS, Gr 807 is the only non‐Aquitanian MS in the sampling to include the “veritas mea” material, and its melody differs substantially from that found in Aquitanian sources (and from the corresponding offertory respond Veritas mea). In the Roman tradition, Vat 5319 has a corresponding verse that begins with “veritas mea,” using the same melody and text as the Roman version of the respond Veritas mea (76), notated in the second‐mode range. The portion of the verse that begins “et ponam” is indicated as a separate, third verse and is not notated. Bod 74 presents only the material corresponding to the Gregorian first verse, Potens es, and divides it into two verses at “posui adiutorium.”

Pa 776 and Vat 5319 were chosen as base versions because they present the fuller version of the second verse, thereby allowing for the most extensive comparisons between the Gregorian and Roman versions. Although the reason for the widely circulated shorter version is not clear, it is notable that Mt. Blandin and Vat 5319 give the same text for both verses, perhaps suggesting that the longer version should be regarded as the preferred reading and the shorter version as an abbreviation. The differences between Vat 5319 and Bod 74, however, may also reflect the practices of the different institutions they represent.

Melodic variants:

  • respond, “brachium” (phrase 4): The text underlay Pa 776 is not the majority reading, though it is also found in Pa 1121. In Mo 159, Lo 4951, Beneventan MSS, and most peripheral sources, the ascent to c is placed on the second syllable of “brachium,” as in Mo 159:

  • “bra‐chi‐um:” G/Gab(quil.)cb/ c

  • respond, “oleo sancto” (phrase 2): Pa 1235 has “sancto meo,” accommodating the extra syllables with a repeat of the material on “sancto.”

  • v. 1 “circuito TUo” (phrase 5): Ben 34: aGFG cbabG

  • v. 2 final melisma on “Eius” (variant portion in boldface): Pa 1235: ac dcdc cacc deca cd cdec cd cdbcaF Fa acbc aGF ac c cbcaGF Fac cd(liq.)

Pa 1121 lacks the second verse.

Gr 807: (p.391)

  • v. 1 “SUper”: Ga GaF Gcac

  • v. 2. As mentioned, Gr 807 is the only late non‐Aquitanian MS to transmit the “veritas mea” part of v. 2. The melody appears to be related to the Aquitanian version at “misericordia mea” and “cornu,” where it adopts standard recitational vocabulary, but differs from it substantially in the other passages.

  • “VE/RI/Tas”: cc caG c ccc/aG/FG

  • “ME/A”: GaFG cc ca cG/F

  • “ET” [misericordia]: Ga

  • “ET/IN/NO/MI/NE”: cd/cdc/c/c/c

  • “ME/O”: ca ccc/abc

  • “EX/AL/TA/BI/TUR”: G/a/cd/dc/c/

  • “[cornu] E/IUS”: acaG Gabca ca(liq.)/GF

Roman MSS: As previously mentioned, Bod 74 transmits only the first verse, divided into two verses at “posui.” Vat 5319 lacks notation for a small portion of v. 1, “tu dixisiti” (phrase 6), and the notes are supplied from Bodmer in the transcription. There are otherwise no significant variants between Bod 74 and Vat 5319.

Offerentur (minor) (74)

Like several other offertories based on Psalm 44, the Roman and Gregorian versions of Offerentu r present complex variants. In the Gregorian tradition, there are two Offerentur offertories, traditionally referred to as “minor” and “maior.” The texts may be compared in the following table.

Offerentur maior, Gregorian

Offerentur minor, Gregorian

Offerentur, Roman

Offerentur regi virgines proxime eius offerentur tibi adducentur in laeticia et exultacione adducentur in templo regi

Offerentur regi virgines post eam proxime eius offerentur tibi

Offerentur regi virgines post eam proxime eius offerentur tibi adducentur in laeticia et exultatione adducentur in templum regi offerentur regi virgines

V. 1 Eructavit cor meum verbum bonum eructavit cor meum verbum bonum dico ego opera mea regi lingua mea calamo scribe velociter scribentis adducentur

Eructavit cor meum verbum bonum dico ego opera mea regi

Eructavit cor meum verbum bonum dico ego opera mea regi

v. 2 Diffusa est gratia in labiis tuis propterea benedixit te deus in aeternum

Adducentur in laeticia et exultatione adducentur in templum regi

Specie tua et pulchritudine tua intende prospere procede et regna

Although the text of the Roman respond might appear on first glance to correspond more closely to Offerentur maior, the melodic relationships suggest that the Gregorian Offerentur minor is cognate with the Roman version. The lyrics shown in the table, however, (p.392) are presented in a different order in the Gregorian and Roman readings. Adducentur in laeticia consistently serves as the second verse in the Gregorian Offerentur minor, whereas in the Roman version this material is simply incorporated into the respond. In the base transcription, the Gregorian verses are presented in reverse order (verse 2 followed by verse 1) to facilitate comparison to the Roman version.

In the Roman tradition, Offerentur has the same two verses, in text and music, as Filiae regum (78) and Diffusa est (80), Eructavit and Specie tua. Because the verse Specie is lacking in the Gregorian Offerentur but present in the Gregorian Diffusa, the Roman version is transcribed from Diffusa est (80).

The Roman version has a varied, written‐out repetendum that is lacking in the Gregorian tradition.

Melodic variants in core MSS: There are only a few variants.

v. 2 “adducentur” [in templum] second time:

Mo 159: melisma begins aFaaGEDC FF …

v. 2 “REgi”:

Mo 159 is missing notes 23–25 of the melisma.

In Pa 776, Pa 780, Lo 4951, E rather than D as the bottom note in the latter part of the melisma (corresponding to the second and third D's of Ben 34).

Variants in peripheral MSS: There are no specific points of instability in Offerentur. Variants entered the melodic tradition to varying degrees. Cai 61, Be 40078, and Gr 807, for example, present readings very close to that of the core group, whereas Pa 1235 and some of the Italian MSS reflect the filtering of small variants into the tradition.

Pa 1235:

  • respond, “tiBI”: DCDF F FD FF FED ED

  • v. 1 “reGI”: EF DEF FDFED EDF

  • v. 2 “adduCENtur” [in laeticia]: DF GFDF CD Da GabGF Gaca DF FFDC (etc.)

  • v. 2 “LAEticia”: aFa GaFE GFGF EDCD FF

  • v. 2 “adduCENtur” [regi]: aFaGF FED FEFDCBCD (liq.)

RoV 52

  • v. 2 “adduCENtur” [in leticia]: DF GFDF DF aa GabGF Gaca DFFFDCFGG FGa cc aGFaGF GF Ga

  • v. 2 “adduCENtur” [in templum]: aF aGF DC FFF DCB CD

Pad 47:

  • v. 1 “BO/NUM”: FGaF/F

  • v. 2 “reGIS”: The entire end of the verse is notated a third lower than the other versions, possibly a scribal error.

Roman MSS: Vat 5319: The entire offertory is written at the affinal position. There are no significant melodic variants. RoS 22: “reGIS”: DFDFED.

(p.393) Gloria et honore (75)

The Gregorian and Roman versions of this offertory share melodic material with others of the first mode. The opening melisma of the first verse in the Gregorian tradition also occurs in Repleti sumus (82). The corresponding Roman melisma is found in Laetamini (79) and the deuterus Benedictus es…in labiis (16), a verse that probably originally circulated with Confitebor tibi (39).

There are some inconsistencies between Mo 159 and later MSS in their use of b‐flat, perhaps reflecting a change in practice. The parenthetical b‐flats in the base transcription are found in Mo 159 and later MSS; some later MSS such as Gr 807 and Be 40078 also indicate flats at “noster” (phrase 3).

Beneventan MSS show a small difference from others at the end of v. 1. In Mo 159 and Aquitanian MSS, the notes over the two syllables of “caelos” are distributed as follows: Pa 776: aG baF FGaGa GbaFGF/F.

Variants among core MSS:

  • v. 1 “NOster”: Aquitanian MSS have acc cba Gba

  • v. 2 “HOmo”: There are small variants in this melisma:

  • Ben 34: FGa ccc cdcdaG acc cGF aG abc abca (same reading in Ben 35, Lo 4951)

  • Pa 776: FGac ccc cdc caG a accGF aG ab(quil.)c ab(quil.) ca (same reading in other Aquitanian MSS)

  • Mo 159: FGa ccc deb caG acc caG aG ab(quil.)ca b(quil)ca

Variants in individual MSS:

  • Pa 776: v. 2 “hoMInis”: cdcd ccaGGF Gab(quil.) ca (slightly shortened melisma)

  • Pa 780: respond, “oPEra”: DE(quil.)FEDCD

Lo 4951:

  • v. 1 “admiRAbile”: cdcaG ccca cdca

  • v. 1 “NOmen”: aGG (ori.) F accc dcab

  • v. 2 “Eum”: slightly shortened melisma:

  • a a(ori.)Ga Faca aGa FacaG abaF GaGE FGFD FFDEC

Pa 1235:

  • respond, “Eum”: GaF FD

  • v. 1 “QUOniam”: Fac

  • v. 2 “HOmo”: version of Mo 159 above

  • v. 2 “Eius”: FGaGaG acaca G GF aG abca ac ccc dcdc(liq.)

  • v. 2 “hoMInis”: cdcdb ac G GF FGa caca

Cai 61:

  • respond “Eum” (second time): Ga bcbaG GaFF(ori.)D

  • respond “domiNE”: F FGFED

  • v. 1 “NOster”: acccba

  • v. 2 “HOmo”: FGa ccc cdc da Ga cc caG aG aca bca

(p.394) Gr 807:

  • v. 1 “NOster”: acccb acccba Gca

  • v. 1 “admiRAbile”: cdcaG acaG

  • v. 2 “HOmo”: FGaccc deb caG (etc.)

  • v. 2 “QUOniam”: D abca

  • v. 2 “VIsitas”: acaaG

  • v. 2 “Eius” (b's are flat): aaGa Faca aGa Fb baGb baF GaGE (etc.)

Be 40078:

  • v. 1 “NOster”: a cccba Gca

  • v. 2 “Eius”: a cdc da cdb caGGF aF FGaGa ccc cdc dc(liq.)

  • v. 2 “VIsitas”: acaaG

RoV 52:

  • Respond, “DOmine”: GaGGF FaGF FDF FFF GFGFG

  • v. 1 “CAE/LOS”: bGbaF FGaGa GbaFGF/FG

  • v. 2 “HOmo”: FGa ccc decaG acc daF GaG aca bca

  • v. 2 “Eius”: acdcda cdcdcdccaGGF Gaca a ccc cdcd

Pst 120:

  • v. 1 “NOmen”: aGaFG Ga dca

  • v. 1 “uniVERsa”: acba

  • v. 2 “HOmo”: FGa cc decaG GcccaF aG aca abca

  • v. 2 “Eius”: acdcca cdc aGF aG abca ccc cdcd

  • v. 2 “hoMInis”: cdcca cbcaGF Ga cabca

Roman MSS: There are no significant variants between Bod 74, Vat 5319, and RoS 22.

Veritas mea (76)

In the Gregorian tradition, there are very few significant melodic variants between sources. In Mo 159, the respond and first verse are written in the transposed plagal range, ending on a, and the second verse in the normal authentic range. Mo 159 employs several pitches that cannot be written at the lower level, including b‐flat (at the quilisma on “miseriCORdia” in the respond), equivalent to E‐flat a fifth lower, and D on “EXaltabitur,” equivalent to Γat the lower level. In its notation at the affinal position and the b‐flat on “misericordia,” Mo 159 differs from most pitched sources, which are written in the normal plagal range and use an E‐natural.

Small variants are found in the following MSS.

Gr 807:

  • v. 1 “poTENtem”: DFFEDE

  • v. 1 “pleBE”: DFDF

Pad 47:

  • v. 1 “POtentem”: DGFE

  • (p.395)
  • v. 2 “MEo”: a different opening of the final melisma: FG aGF aGF Fa aF EG abb aaFa aGaF abaE EGaF (etc.)

Mod 7: v. 2 “MEo”: a different opening of the final melisma: FabaF GFCD FFDF FD F CF Ga aa aGaaGaF abGE EG

RoV 52:

  • v. 2 “DIS/PERgam”: FGa/ GaFG

  • v. 2 “MEam”: in the middle of the melisma, a (rather than b‐flat) is the top note: FabaF GFD DF FD DF Ga a a aGa aGaF

Pa 1235:

  • respond, meO”: ACA CD

  • v. 1 “pleBE”: DFDF

  • v. 1 “MEa”: CDF FGFE FGFG

  • v. 2 “MEo”: FabaF GFD CF FD CD Da a Ga Ga bGaF GaGE FGaF (etc.)

Roman sources: Because Bod 74 lacks the verse Posui, the more complete reading of Vat 5319 is given as the base version. At the end of v. 2, however, the repetendum cue in Vat 5319 is notated a fifth too high, a problem found often in Gregorian MSS but very rarely in the Roman tradition. Bod 74 presents a substantially different reading of the final melisma that ends in the lower part of the range and is followed by a repetendum cue at the “correct” level. The melisma begins a fourth lower (starting on “conSPECtu”), is longer, and is melodically variant.

Other variants in Bod 74:

  • respond [misericordia] “ME/A”: ED/DEFEDEFE

  • v. 2 “miseriCOR/DIa”: FGaG(liq.)/ac c(ori.)ba

Constitues eos (77)

There are small variants among the core group of sources at the following places. Peripheral sources generally exhibit the same range of variance, and for these passages, only sample readings are provided from the peripheral group.

Respond, “MEmores”:

  • Pa 776: FF FF(ori.)D F aG

  • Lo 4951: FF FFF FaG

  • Pa 1121: FFFFD Fba

respond, “PRO/GEnie”: Mo 159:DG/Gc (also in Mod 7)

respond, “generatioNE”: The bracketed notes in the base transcription (phrase 3) are omitted from Ben 34, a probable scribal error. They occur in all other MSS. In the base transcription they are supplied from Ben 39.

v. 1 “eructavit”: The Beneventan version matches the Roman verbally, but most non‐Beneventan MSS included in the sampling have “eructuavit,” accommodating the extra syllable with a single G. The exceptions are the Italian sources Mod 7, Pad 47, and RoV 52.

v. 1 “DIco” (like “memores” above): (p.396)

  • Pa 776: FF FF(ori.)D F aG

  • Pa 1121: FFFFD Fba

  • v. 1 “eGO/Opera”: some sources indicate a different text underlay, as in Mo 159: FE/GabGa.

  • v. 2 “scriBE” is a point of variance:

  • Ben 39: a aGaGFGD GaGaD GaG

  • Pa 1121: same notes as Ben 39

  • Mo 159: melisma begins a a(ori.)G aGF GD (cont. like Ben 34)

  • Lo 4951: same notes as Ben 39

  • Pa 776: aaG aGFG C GaGa C GaG

  • Pa 1235: a aGaGF GE FGF FD FaG

  • Mod 7: a aGa GFGD GbabF GaG

  • RoV 52: a aGaGFGF GaGaF GaG

  • Gr 807: a aGaGF GD GaGGE GaG

v. 2 “SCRIbentis”: Mo 159: D

v. 2 “speCI/Osus”: Most Aquitanian, Italian, and German MSS match Ben 34 here, with a leap of a sixth on the third syllable (a b‐flat is indicated in Gr 807). Mo 159 has an alternative reading, two successive fourth leaps:

  • Mo 159: DG/GcaG (etc.)

  • Pa 1235: DG/GbaG

v. 2 “fiLI/IS”: The Beneventan reading is distinctive. Most MSS have a text underlay similar to that of Mo 159: DFED/D.

v. 3 “eTERnum”: Mo 159: melisma begins GF aGF GF(quil.)E.

Significant variants in peripheral MSS (there are very few):

v. 2 “tuIS”: Variants within small segments of the melisma occur in some Italian sources:

  • Mod 7: cda cbcaG cb bdeca cdca cbcaG (etc.)

  • RoV 52: cda cbcaG cb cdeca c ca cc aG cb bab (etc.)

The tonal contrast of the final verse, with its repercussive pitch on F, is found in all MSS. Pa 776 has a written level a fifth higher, but the repetendum suggests that the performance level is that of Ben 34 and the other MSS.

Roman MSS: This offertory is lacking in Bod 74. Vat 5319, the only MS to transmit the verses of this offertory, presents a shorter version of the first verse, lacking an equivalent of the Gregorian phrase 5.

RoS 22 has the following variants:

  • “TERram”: FGaGbaGFG

  • “TUi”: dedcbcbabaG

  • “generatioNE”: aaGF FGaG baGFEGFEFE (closes on E rather than F)

(p.397) Filiae regum (78)

The verses of Filiae regum and two other offertories based on Psalm 44, Diffusa est (80) and Offerentur (74), present a complex picture of transmission between Rome and Francia. In the Roman tradition, the three offertories have the same verses, in text and melody: Eructavit and Specie tua. In Bod 74, these verses are notated once, for Offerentur (f. 3v), and cued in the other two cases. Vat 5319 has the same cues in Diffusa est but lacks verse cues for Filiae regum. In most Gregorian MSS, these three offertories have different sets of verses. Filiae regum normally circulates with the verses Eructavit and Virga recta est, the latter verse lacking an equivalent in Roman MSS. In Mont‐Blandin, however, Filiae regum circulates with the same two verses as in the Roman graduals, Eructavit and Specie (with only incipits provided), suggesting that perhaps the verse Virga recta is a Frankish addition. Specie, moreover, is also the second verse in two peripheral MSS included in the sampling, Mod 7 and Pad 47, with the same melody as the Specie verse for Diffusa est. The other Italian MSS have Virga recta as the second verse. If Virga recta is a Frankish addition, it dates from a early period, since it is found in Compiègne, the later Silvanectensis, as well as Cha 47 and Ei 121.1 Textually, the verse corresponds to the PsR rather than the PsG.2

There are several small melodic variants in this offertory, including the final cadence of the respond and resulting modal assignment. The ending on F in the Beneventan version is a minority reading. In many other MSS, the final cadence is on E, as the following sampling shows:

  • Mo 159: G GaG aGE

  • Pa 776: GFaGE

  • Mod 7: GFaGE

  • Pad 47: FaGE

  • Pst 120: GFaGFE

  • Cai 61: GFaG G (ori.)E

  • Be 40078: G GaGE

Cadences on F:

  • Lo 4951: GFaGF

  • Pa 1235: GFaG GF

  • RoV 52: G GaGF

Other melodic variants:

  • respond “reGUM”: The Beneventan reading aca at the beginning of the melisma is distinctive. In most sources the melisma begins with cdc and continues as in Ben 34.

  • v. 1 “cor meUM”: Unlike Ben 34, most MSS have cccaGc

  • v. 1 “BOnum”: The Beneventan version is anomalous in notes 8–14, and this passage is also a point of variance between Mo 159 and the Aquitanian MSS:

  • Mo 159: melisma begins G acaGa Fb‐flat‐aG a‐b‐flat‐aG b‐natural dc

  • Pa 776: melisma begins GabaGa F aGF abaG cdc (etc.)

  • Lo 4951: GacaGa FaGF acaG cdc (etc.)

  • (p.398)
  • The readings of Pa 776 and Lo 4961 are also found in the following: Pa 780, Pa 1235.

  • Cai 61: G aca Ga F b‐flat GF acaG

  • Mod 7: FacaGa FGFE aacaG (same in Pad 47)

  • RoV 52 GacaGa GaGFacaG

v. 1 “BOnum”: The Beneventan reading differs from others in the last four or five notes of the melisma. The other sources have a reading closer to Mo 159 and Pa 776.

  • Ben 34: FGaGa abaG

  • Mo 159: FG(qui.)aGa cb(qui.)aG

  • Pa 776: FG (qui.) aG ab(qui.) cba

v. 2 [deus] “TU/US”: Mo 159: ad ecba/c

v. 2 verse ending “tuIS”:

  • Lo 4951: GaF

  • Pa 1235: cdcba/acac

Among peripheral MSS, there are differences in recitational style. In the passage of v. 2, “unxit te deus deus,” for example, Pa 1235 adopts c rather than d as the highest note.

Variants in individual MSS:

Cai 61:

  • respond, [dextris] “TU/IS”: cdcba/abca a(ori.)G

  • v. 1 “operA MEa”: ccc edc cbaG/GcaGa

Mod 7: v. 1 “verBUM”: acaGa

RoV 52:

  • v. 1 “verBUM”: acbac

  • v. 2 “iniquitaTEM”: acaG

Be 40078: v1. “reGI”: G

Roman MSS:

Vat 5319: This offertory has no verses cued. The respond is written at the affinal position, with the following small variants:

  • “tuO”: cbcd

  • “regNA”: aG

  • “A” [dextris]: cde

RoS 22:

  • “tuO”: FEFGFE

  • “tuIS”: GaGFEFG

  • “inVESti”: GaG

  • “circumDA/TA”: FaGFFED/DFDEDCB DFED FGFF

Laetamini (79)

This offertory exhibits few variants. Core MSS: (p.399)

  • v. 2 “temporRE”: Most MSS have a slightly different reading, as in Pa 776: GFG FF DED.

Lo 4951: respond, “CORde”: FG FFGFED(liq.)

Pa 780:

  • v. 2 “TE”: DE(quil.)F GF F(ori.)D GFGFEC DFD DDD

  • v. 2 “SANCtus”: FG

  • v. 2 “muTArum”: acGF acaG ccccc

Most peripheral MSS present a reading very close to that of the core group, with the following exceptions.

  • Be 40078: respond, “CORde”: F FG EFG EFGFE

  • Gr 807: v. 1 “peccaTA”: DEFEDCD

  • RoV 52: respond, “IUsti”: DF FFF GaGaFED

  • Cai 61:

  • respond, “ET”: Da baG

    • v. 2 “SANCtus”: FG

    • v. 2 “diluVIo”: dcbcbaG

Roman MSS:

Vat 5319:

  • v. 1 “beaTI”: FG

  • v. 2 “verumTAmen”: FGababa

  • v. 2 “diLU/VIo”: c cdcdc dc c(ori.) aGa/baG

RoS 22: “corDE”: FFED FFEDFED (shortened melisma)

Diffusa est gratia (80)

In the Roman tradition, this offertory has the same verses, in text and melody, as Offerentur (74) and Filiae regum (78). In Vat 5319 and Bod 74, the verses Eructavit and Specie tua are written out for Offerentur and merely cued in the other two offertories. In early Gregorian MSS, Diffusa circulates with the same two verses. The first, Eructavit, is cued in most sources, with neumes that indicate the melody of the corresponding verse in Filiae regum (not that of Offerentur, which uses a different melody). In most later Gregorian MSS, however, Diffusa circulates with only one verse, Specie tua. In Ben 34 and in the base transcription, only Specie tua is provided. For the Roman verse Eructavit, see the transcriptions of Offerentur and Filiae regum.

Melodic variants:

respond, “difFUSa”: This syllable is a point of variance within the core MSS. Most peripheral MSS match the Aquitanian readings. (p.400)

  • Ben 34: FGa b bab bab

  • Pa 1121: FG a a(ori.) G a a(ori.) G a

  • Pa 776: EFGa a(ori.) G aaGa (same reading in Cai 61)

  • Pa 780: FGaaG aaGa (same reading in Be 40078, RoV 52, Pad 47)

  • Mo 159: Ga b c c (ori.)b c c (ori.) b c

  • Mod 7: FGa abaa aaaa

respond, “IN” [aeternum]: Most MSS have a reading identical to Ben 34. Pa 776 has cccGa cdec.

v. 1 “REGna”: In Mo 159, the latter part of this melisma is notated a tone higher than it is in other MSS: cdecc a cca ccaa(ori.)FG FG(quil) aGa Ga(quil.)bab ab dd dbaG bdd dbaG bdba bdba ab/G. A different reading at the opening of the melisma is found in several of the Italian MSS, as in RoV 52: cdecd cdeca c caGFG FGaGa (etc.).

Individual MSS:

RoV 52:

  • v. 1 “SPEcie”: DF G aG baG GF Ga cdcb dcdca cc ca cc dec

  • v. 1 “inTENde”: This melisma is written a tone below the majority reading of Ben 34 and most MSS. That this is a scribal error, however, is suggested by the custos at the end of the line, followed by the final segment of the melisma notated at the correct level.

Cai 61: respond, “IN” [aeternun]: cc cG ac dec (like Pa 776 above)

Mod 7

  • The verse Specie is cued in Diffusa and written out as the second verse of Filiae regum

  • v. 1 “TUa”: acaaG

  • v. “REGna”: melisma begins cd ecd cd eca cccaG GFG (similar to RoV 52 above)

Pad 47:

  • The verse Specie is cued in Diffusa and written out as the second verse of Filiae regum

  • v. 1 v. 1 “TUa”: acaG

  • v. “REGna”: The beginning of the melisma has the reading found in several of the other Italian sources; it is also variant in the second segment in boldface: cd ecd cd eca cccaG GFG aGa FGaGa FGaGa etc.

Roman MSS: There are only a few small variants.

Vat 5319:

  • respond, “SEculi”: cdcdc bcba(ori.)GF GFGFGaG acbcba bcaG GFacbaG

  • The verse Specie, written out in Offerentur and merely cued in Diffusa, is written at the affinal position, like the respond and first verse of Offerentur.

RoS 22: “SEculum”: c dcdc bcbaGF GFGFGaG Fac cbcba bcbabaGF acbaG.

Ave Maria (81)

The pitch level variants in this piece have been illustrated and discussed by Hankeln and Sidler.1 Sidler proposed that Ave Maria was originally a protus melody, based partially on the first‐mode characteristics of the final verse. Hankeln, however, notes the consistent tetrardus (p.401)

Appendix 3 Commentary on Edition

Example 81.1a

character of this chant in the Aquitanian MSS, and the sources included in the sampling present a similar picture. Although the melody does exhibit a mixture of modal traits, there is little evidence that it was ever consistently protus.

The passage “spiritus domini superveniet in te” (phrase 8) is a problem spot. Based on the transpositional variants, Hankeln persuasively posits an Aufführungslage that is equivalent to Pa 776 here.2 Three different versions of the passage are shown in example 81.1. In (p.402)

Appendix 3 Commentary on Edition

Example 81.1b

Mo 159 (line 1), this passage is marked by a move into the higher register, with high f as the repercussive pitch at “superveniet.” A similar reading is found in several other MSS, as summarized below. In Pa 776 (line 3), however, the passage is written a fifth lower. In this version, “superveniet in te” is marked by a modulation of sorts, in which typical tetrardus melodic characteristics are projected a tone below their usual position: b [flat] (equivalent to high f in Mo 159) serves as the recitational pitch, and a hypothetical E‐flat (equivalent to b‐flat in Mo 159) is introduced at the end of the melisma on “te.”

(p.403)

Appendix 3 Commentary on Edition

Example 81.1c

Hankeln's hypothesis that Pa 776 is the preferred reading of this passage receives support from a group of sources that notate the passage a whole tone above the level of Pa 776. Ben 34 (line 2), for example, writes the passage a tone higher than Pa 776 does, beginning with the melisma on “te.” This transposition places the passage in the normal eighth‐mode segment of the background scale and allows the irregular semitone at the end of the melisma, E‐flat‐D, to be represented with F‐E. Although a similar transposition is found in Lo 4951, Gr 807, and in most Italian MSS in the sampling, these sources begin the transposition at different places, as summarized below. The variety of versions among this group suggests that the transpositions are emendations made in response to the irregularity of the passage and the need for E‐flat. With this hypothesis, Pa 776 and Pa 780 are the only MSS in the sampling that notate this passage in its unemended range.3

The final passage of the verse, “et virtus altissimi obumbravit tibi” (phrase 9), presents a problem I have not been able to fully solve. As shown in example 81.1, Pa 776 continues at an interval a fifth below Mo 159 and a whole tone below Ben 34, a reading unique among the MSS examined. Hankeln considers the written level of Pa 776 at the end of the verse a scribal error, and in his edition he has placed it a whole tone higher. One reason for positing a scribal error lies in the two prosulas that follow the first verse in Pa 776, based on the “quomodo” melisma. After the move into the new tonal area at “super veniet,” the version of Pa 776 never rejoins the level of Ben 34 or Mo 159; the two prosulas are thus written a whole tone lower than the melisma itself. With the premises that the prosulas were sung at the same level as the source melisma and that the pitch level between verse and prosula is accurately indicated in the MS, one must posit a scribal error somewhere earlier on. The problem of the written level in Pa 776 continues with the second verse and repetendum: the second verse is written a tone below its usual level, and the repetendum is a tone lower than the corresponding place in the respond.

This problem spot in Ave Maria is reminiscent of a similar passage in Oravi deum. In respond of Oravi deum, Pa 776 and 1121 present a reading that contains a modulation of (p.404) sorts. Beginning in the middle of the respond, deuterus melodic passages are projected a tone below their usual position. In the case of Oravi deum, this reading is confirmed in some unrelated MSS that notate the passage at the affinal position: the chant starts as a regular b‐deuterus chant and ends as an a‐deuterus chant (with b‐flats). The same questions are raised about the continuation of the chant in the verses: when, if ever, did it get back “on track?” In the case of Oravi deum, I have cautiously hypothesized that in Pa 1121 and Pa 776, the answer is never. In Oravi deum, this hypothesis is supported by a repetendum cue in Pa 1121. In Ave Maria, however, the repetendum is written a tone below its position in the respond, perhaps suggesting an Aufführungslage a tone higher. Hankeln's proposed solution to the problem, that “et virtus” is a scribal error and should be written a tone higher, is found in only one MS, Pa 780. This MS matches Pa 776 at “spiritus domini” but is a tone higher beginning at “et virtus,” thereby returning to the normal eighth‐mode range. The relative pitch level between the two sections (“spiritus domini…” and “et virtus…”), as it is notated in Pa 776, is duplicated in a great majority of sources: the interval between the end of the melisma on “te” and “et [virtus]” is a third, not a fourth, as it is in Pa 780 and in Hankeln's proposed correction of Pa 776. The relative pitch level between these two sections as they appear in Pa 776 is found in all MSS except Pa 780: at a position a fifth higher in Mo 159 and a whole tone higher in Ben 34, Gr 807, and most Italian sources. Either all sources except Pa 780 are emended versions in this particular passage or the level of Pa 776 represents the unemended pitch level, as it does in the earlier passage.

I consider the second possibility more likely and am thus hesitant to regard Pa 776's reading of the end of the verse as a scribal error. It is more probable that the whole‐tone downward “modulation” that occurs in Pa 776 was maintained throughout the verse. The variety of the other versions, summarized below, reinforces the impression that they are emendations. The other versions move to a level a whole tone above Pa 776 at different points. In the edition, I have transcribed this passage in Pa 776 at its written level. At this level, several nondiatonic pitches, including low b‐flat, E‐flat, and a‐flat, are required to duplicate the intervallic structure of the melody as it appears in Mo 159 and Ben 34. The three versions may be compared by examining the last system of example 81.1. With this hypothesis, the intended pitch level of the prosulas and second verse remains an unresolved problem: it is unclear when, if ever, the chant got back “on track.” If such an adjustment was made, it was likely done at the beginning of the prosula or at the beginning of the second verse. In the transcription, I have given the end of v. 1 at its notated level and the second verse, hypothetically, at the level of the majority reading, a tone above its written level.

A few MSS notate v. 2 a fourth above the majority reading: Mo 159 and Pa 1121 and the other MSS from St. Martial. All other MSS match the majority reading given in the transcription, with the first‐mode incipit C‐D‐D‐a. Mo 159 employs a b‐natural on the last syllable of “ideoque,” which would be equivalent to F‐sharp at the lower level. This nondiatonic practice, however, is not preserved in other traditions. In Pa 776 and many other MSS, the corresponding note is a. Ben 34 is the only Beneventan MS with verses to transmit this piece (because of lacunae in the other MSS), and it lacks the second verse.

Pitch level at “spiritus domine superveniet in te”:

  • 5↑ Pa 776: Mo 159, Cai 61, Pa 1121, Be 40078

  • 2↑ Pa 776: Lo 4951 (at “superveniet”), Ben 34 (at “te”), Pad 47 (at “spiritus”), To 18 (at “superveniet”), Pia 65 (at “spiritus”), Gr 807 (on the third syllable of “superveniet”: the (p.405) first two syllables are G and a), Frutolf tonary (on the third syllable of “superveniet”—same as Gr 807), Pa 780 (at “et virtus”)

Editorial emendation: Pa 776 presents one small transcription problem. In the respond, “mulieribus” is the last word on the page, followed by a custos on pitch a. The a, however, is missing on the next page, which is probably a scribal error, since it is present in all other MSS. I have supplied it in the transcription and indicated it with brackets.

Individual MSS: Pa 780 is a fourth below Pa 776 at “quae virum,” joining at “spiritus domini.” Gr 807 moves to the lower register during the melisma on “virum”” ccb cGG(ori.)G EF Ga FFD(liq.).

Melodic variants: Ave Maria has an unusual number of melodic variants, even among the core MSS. In the earliest MSS, this offertory lacks the words “dominus tecum” (phrase 3), which, as Hesbert suggested, appear to be a later addition, made separately in the Gregorian and Roman traditions.4 The passage is lacking in Mont‐Blandin, Compiègne, Corbie, and Ei 121. These added words are accommodated with familiar melodic material: the melisma on “dominus” also occurs in Angelus domini.

Given the later addition of this material, the melodic variants here are not surprising. La 269 indicates a shorter melisma on “dominus” (without the repetition of material), a reading also found in Lo 4951 and Pa 780. Other sources, such as Pa 1235, lack the melisma on “dominus” altogether.

Some sources present a variant reading of “tecum,” with a short melisma:

te‐

cum

Pa 1121:

G GGG aGF GD

acab/aG

Ben 34:

GaGFGD Gbab

/aG

Gr 807:

GG aGF GD Gcac

/aG

Be 40078:

GG GED EGaGF

/G

Pst 120:

GGG FaGFGD

Gcab/aG

Both versions, with and without the short melisma on “tecum,” are found across a broad regional and chronological spectrum. The melisma is lacking in Mo 159, Pa 780, Lo 4951, Pa 1235, Pad 47.

v. 1 “VIrum”:

  • Pa 1121: c c(ori.) bc G G (ori.)FG ccde

  • Be 40078: c cb cG G(ori.) FG cc de

  • Pst 120: cc cbca aFG Gc de

v. 1 “DOmini”:

  • Gr 807: DCD EF Ga caG FG acaGF c

  • Pst 120: aGa cc de ged efg cde g edefedef

v. 2 “NA/SCE/TUR”: The reading of Pa 776 is found in the Aquitanian MSS, with exception of Pa 780, and also matched in Pa 1235 and Cai 61. Mo 159 and many other MSS have a different reading: (p.406)

  • Mo 159 (at higher pitch level): dfed/dc/cedc

  • Pa 780: acba/aG/GbaG

  • Gr 807: acba/aG/G b‐flat aG

  • Be 40078: acba/aG/G c aG (similar reading in Pst 120 and Pad 47)

v. 2 “DEi”: Pa 776 is the Aquitanian version, also found in Pa 1235, but many MSS have a more elaborate version of this melisma, as found in Mo 159:

  • Mo 159 (at higher level of transposition, and all b's are flat): ccc(ori.) cdbaG Gabc cc(ori.) cdbaG Gabc dd(ori.)c

  • Gr 807: GaFED D EF Ga FED DEF Ga(ori.)G

  • Pst 120: GaFED DEFGa FED DEFGa aG G

Versions very similar to Gr 807 and Pst 120 found in Cai 61, Be 40078, and Pad 47.

Roman MSS: In the Roman tradition, v. 1 consists largely of material from the respond; v. 2 has protus characteristics, a trait that ties it to the Gregorian version. There is one significant variant between Bodmer and Vat 5319 in the melisma that occurs at the end of v. 1. The part in boldface is written a whole tone lower in Vat 5319: aGaG baG GF aGaG ababaGaG.

Other variants:

Vat 5319: v. 1 “naSCEtur”: ba

RoS 22:

  • “maRIa”: c cbcabaGF acbG

  • “gratiA”: GFG FG

  • “beneDICta”: cba

  • “beneDICtus”: cba bcdcbabG

Confitebuntur caeli (82)

In the opening passage, on “confitebuntur,” Pa 776 differs from all other MSS. On the last two syllables, the scribe writes b/bdc. In all other MSS, these neumes are reversed: bdc/b. This is certainly a scribal error, and I have undertaken an editorial emendation of it.

There are also some differences between MSS in repetendum structure. In most MSS, the second verse closes with a repetendum to the first “alleluia” of the respond, indicating that the both alleluias were sung and that the performance came to a conclusion on G. In Pa 776, the verse closes with a written out repetendum of this “alleluia,” with two tiny variants. Although it is not clear from the MS whether this repetendum is an instruction to close the melody here, on F, or to proceed to the final alleluia, it is likely that the performance was concluded on the original final G, requiring the singers to sing the second alleluia. In Mo 159, the repetendum cue is to the final “alleluia” of the respond.

Pitch‐level variants in v. 2, Quoniam: Pa 776 was chosen as the base reading because it presents the most probable unemended version of the second verse. This verse presents complex transpositional variants, with several distinct versions. In most MSS, the verse begins a fourth below the reading of Pa 776. With the exception of Mo 159, which remains a fourth below Pa 776 throughout, all of these MSS join the level of Pa 776 before the end of the verse. The transition to the higher level, however, occurs at different times and in different ways, as summarized below.

(p.407) While it was not possible to determine a preferred reading with certainty, various circumstantial factors point to the pitch‐level profile of Pa 776 as a preferred reading: the ending of the verse at this position in a majority of MSS, the diversity of readings in the sources that begin a fourth lower, and the tendency evident in Pa 776 to notate problematic passages in their “unemended” range. A downward transposition of a fourth would typically be employed to represent the nondiatonic pitch high eb, with a b‐flat. A b‐flat is indicated only once in Mo 159, on the penultimate note of “erit” (phrase 10). I have indicated this spot with a parenthetical e‐flat at the corresponding place in the transcription of Pa 776. It is doubtful, however, that this hypothetical e‐flat is the sole cause of the different readings. Several sources included in the sampling move to the position of Pa 776 before the problem spot on “erit” (phrase 10), hinting at a possible nondiatonic pitch earlier on in the verse. The following intervallic differences with Pa 776 are evident in individual MSS, suggesting nondiatonic pitches.

1. “aequabitur” (phrase 9): A possible nondiatonic pitch is suggested here in some sources. Ben 34, 35, and 39 begin the verse at the level of Mo 159, a fourth below Pa 776. On the first syllable of “aequitabitur” (phrase 9), some MSS are a tone below Pa 776. The notes fgfg in Pa 776 appear as a semitone, efef, in the Beneventan sources, possibly attesting to a nondiatonic practice at this point. Pa 1132 also has this reading, but it is not found in other MSS from St. Martial. At the level of Pa 776, an f‐sharp would be required to duplicate this intervallic structure.

  • Pa 776: fgefgfg

  • Ben 34: dfdefef

  • Pa 1132: dfdefef

The other sources match either Pa 776 here or are written a fifth lower, with no trace of a nondiatonic pitch. Because the problem seems particular to a small group of sources, I have not included accidentals in the transcriptions.

2. “filiOS” (phrase 10): This passage follows the problem spot on “erit deo.” On the last syllable of “filios,” Mo 159 is briefly written a fifth below Pa 776 rather than a fourth below. In Ben 34, this passage is part of a lengthy section that is written a whole tone below the level of Pa 776. The three versions may be compared as follows:

  • Pa 776: defed

  • Mo 159: Gabcb

  • Ben 34: cdededc

The minor third d‐f in Pa 776 is the major third G‐b‐natural in Mo 159 and the major third c‐e in Ben 34. An f‐sharp would be required to replicate this intervallic structure at the level of Pa 776. On “dei,” the interval of transposition between Mo 159 and Pa 776 changes back to a fourth, and Mo 159's motion around the pitches a, b, and c is intervallically identical to the motion around d, e, and f in Pa 776. Ben 34, however, places the passage on “dei” between c and e, equivalent to f‐sharp at the level of Pa 776.

In most MSS there is no trace of this hypothetical nondiatonic practice. Many MSS, including Lo 4951, Tri 2254, Cai 61, and the MSS from St. Martial, for example, are identical in pitch level to Pa 776 here. Because of the inconsistency of transmission of this passage, I have not indicated any nondiatonic notes in the transcription.

Pitch level in individual MSS:

  1. 1. Pa 780 and Cai 61 are at the level of Pa 776 throughout.

  2. (p.408)
  3. 2. Tri 2254 is at the level of Pa 776 throughout, except for “aut quis si [milis],” where it is a tone below. In this MS, the e‐flat on “erit” is circumvented by simply reworking the melisma: df gf fff dcdfff.

  4. 3. The following MSS begin 4↓ Pa 776 and join it at various points: Pa 1121, Pa 1235, and Pst 120 join Pa 776 on “aequabitur.” Pa 1132 joins on the second syllable of “aequabitur.” Lo 4951 joins Pa 776 at “deo,” with a transition during the melisma on “erit.” Mod 7 and Be 40078 join Pa 776 at “deus.” Pad 47 joins Pa 776 at “dei.” Ben 39 joins on the third syllable of “aequitabitur,” and is a second below Pa 776 on the last syllable of “filios.” Ben 34 presents a highly emended reading of the verse that differs from the other sources in several passages. It is variously a fourth, a third, and a whole tone below Pa 776.

    • At verse opening: 4↓ 776

    • “aequabitur domino”: 2↓ 776

    • “aut quis similis erit”: at various invervals of transposition, with a different melodic readings.

Other significant variants: v. 1 “TU/AS” (phrase 5) is a point of variance among core and peripheral sources, as the following sample readings show:

  • Lo 4951: identical to Pa 776

  • Mo 159: dg ag aged/de(quil.)f ge

  • Pa 1235: dg agagd/dgf fd ced

  • Ben 34: df gfgfd/fgff (ori.)d ced

  • Mod 7: dg afed/decacded

v. 2 “DEus” (phrase 11): variant reading in Beneventan MSS: Ben 34: d ec c(ori.)d edec c(ori)d ff.

Roman MSS:

Vat 5319: respond, final “alleluia”: The second half of the melisma on “le” and the final two syllables of “alleluia” are not notated. The melisma on “le” reads as follows:

  • cdededc ededc c (ori.) cbdcbacbaGaGF

  • v. 1 “in”: bacbG

  • v. 1 “meO”: cd

  • v. 1 concludes with a slightly varied written‐out repetendum to the first “alleluia,” with a modified cadence:

    • al‐ le‐ lu‐ia.

    • dc (liq.)/ dcd cdc c(ori.)baGaG cdcdcbcbabcbab ba(liq.)/ GacbaGa/ aG

v. 2, “deus” (signifcant variant): For reasons that are unclear, most of this melisma is notated a fifth below that of Bod 74:

  • De‐             us

  • GGG aG FEFG GG aG FEFG G cbaG/cdc cbcd

For the last four notes of the melisma, in boldface, the interval of transposition changes to a third.

Vat 5319 indicates no repetendum at the end of v. 2.

RoS 22:

  • “conFItebuntur”: GabdcG

  • (p.409)
  • “sancTOrum”: edc

  • first “alLEluia”: RoS 22 has a slightly shorter version of the melisma:

  • dcdcdccbaGaG cdcdcbabcbab

Repleti sumus (83)

The Roman version has a longer text than the Gregorian does, as well as a written-out repetendum that is lacking in Gregorian MSS.

This piece exhibits almost no significant variants among the MSS included in the sampling.

  • Lo 4951: v. 1 “reFUgium”: aca Gba

  • RoV 52: respond, “alLEluia”: abaGFF GaGF GF FD FGF GbaG GF GFGF

Roman MSS:

  • Vat 5319: respond, “noSTRIS”: EGFF (ori) GFEDEDF FE (same at “progeniE” in v. 1)

  • RoS 22 is written at the affinal position and has the following variants:

    • “oMNI/BUS”: bcba/bc

    • “noSTRIS”: bdcdcbababcb

Mirabilis (84)

Mirabilis exhibits one variant in pitch level. In some MSS, including Mo 159, the beginning of v. 2 is written a fifth above its level in Ben 34, joining at “aepulentur.” The reading of Ben 34 is found among all core Aquitanian MSS, Pa 1235, and Cai 61, whereas Mo 159's version is matched in all Italian and German MSS. Because the two versions are intervallically identical, it is doubtful that the variants are attributable to nondiatonic pitches. The two versions may simply represent performance alternatives.

Pitch level a v. 2, “Pereant peccatores…iusti”:

  • = Ben 34: Ben 35, Pa 776, Pa 780, Pa 1121, Lo 4951, Cai 61, Pa 1235

  • 5↑ Ben 34: Mo 159, Mod 7, Pad 47, Pst 120, RoV 52, Be 40078, Gr 807

Other melodic variants:

  • respond, “PLEbi”: Most MSS have the reading of Mo 159: aG cdc

  • verse endings:

  • v. 1 “eIUS”

  • v. 2 “leticiA”

Some MSS have a slightly shorter version of this melisma, as in Mo 159: GaGF acb cca FG; or Pa 1121: GaGF acb bcaFG. Pa 780, however, presents a reading similar to Ben 34: GaGF acb cca caFG.

Pa 776: “disSIpentur”: acdc

Lo 4951: v. 2: missing “qui” [epulentur]

(p.410) Pa 780:

  • respond, “alleluIA”: Gca aGGF

  • v. 2: missing “qui” [epulentur]

Pa 1235:

  • respond ends at “deus,” omitting the “alleluia.”

  • respond, “SUis”: ac dcdcba

  • respond, “DEus”: melisma begins cdcdbca (etc.)

  • v. 1 “EXurgat”: DG

  • v. 1 “dissiPENtur”: cc adc

  • v. 1 “ET” [fugiant]: DG

  • v. 1 “QUI”: Faca

  • v. 2 [facie] “DEi”: FGFE

  • v. 2 “EXultent”: DG

  • v. 2 “IN”: Faca

Cai 61:

  • respond ends at “deus,” omitting the “alleluia.”

  • respond, “SUis”: ac dc dcba

  • v. 1 “EXurgat”: DF

  • v. 1 “dissiPENtur”: c ca cdc

Mod 7:

  • respond, “israHEL”: FGa baG

  • respond, “SANCtis”: ac dcbaG

  • respond, “alleluIA”: melisma ends abaGF

  • v. 2: missing “qui” [epulentur]

  • v. 2 ”epulentur”: ac ccc dcd

Pad 47:

  • respond, “israHEL”: Fa baG

  • respond, “VIRtutem”: abcb(liq.)

  • respond, “SANCtis”: ac dcbaG

  • respond, “alleluIA”: melisma ends abaGF

  • v. 1 “inImici”: acb

  • v. 1 “QUI”: abca

  • v. 1 [facie] “Eius”: melisma begins EFGF

  • v. 2: missing “qui” [epulentur]

  • v. 2 “epulentur”: ac ccc dcd

Pst 120:

  • respond, “israHEL”: Fa ba aG

  • respond, “VIRtutem”: abcb(liq.)

  • v. 2: missing “qui” [epulentur]

RoV 52:

  • respond, “israHEL”: Fa baaG

  • respond, “VIRtutem”: abcb(liq.)

  • (p.411)
  • v. 2: missing “qui” [epulentur]

Be 40078:

  • v. 2: missing “qui” [epulentur]

Roman MSS:

Vat. 5319: v. 2 “qui epulentur”: Vat 5319, like many Gregorian MSS, lacks the “qui.” As a result, the text underlay in this passage is different: “E/PU/LEN”: aG/aGa/a(ori.)GF Gaba babaG(liq.).

RoS 22: “DEus”: last 12 notes are bcabaGF acbaG.

Gloriabuntur in te (85)

In the Gregorian tradition, the first verse of this offertory is identical, verbally and musically, to that of Intende voci (23). In most MSS, the verse is simply cued. The Roman melody, however, differs from that of the corresponding verse in Intende voci, instead adopting Formula B. To facilitate comparison between the Gregorian and Roman versions, I have transcribed the first verse from Ben 34 as it appears in Intende.

While Ben 34's normal tritus version is found among a great majority of MSS, two sources in the core group, Pa 776 and Pa 1121, hint at possible modal irregularities in the pretheoretical tradition. In three of the core Aquitanian MSS (Pa 776, Pa 780, and Pa 1121) the written level of the cue to the first verse, Verba mea, as well as the entire second verse, is a fifth below their position in Ben 34 and the great majority of other MSS. In Pa 780, the repetendum cue after the second verse, to “domine,” implies a performance level for the verse equal to that of Ben 34. In Pa 776, however, the repetendum cue suggests a performance level a fourth below the level of Ben 34. Although Pa 776 is the only pitched source in the sampling to imply a lower performance level for the verse, this reading is consistent with Ei 121, which has the letter i (inferius) at the beginning of the verse, suggesting that it begins below the final. A similar pitch‐level profile is found in two geographically diverse later versions, Tri 2254 and Pad 47. In both MSS, the respond is written at the affinal position, ending on c, and the second verse is in the normal authentic range, lacking a repetendum cue after the second verse. In the case of Tri 2254, this transposition of the respond may have been undertaken to represent low B‐flat, as described below. In Ben 34 and most core sources, the corresponding note is c.

Pa 1121 and most MSS from St. Martial lack repetendum cues. For the second verse, a performance level identical to that of Ben 34 is implied by the repetendum cues in Pa 1134 and 1136. In Pa 1121, however, the end of the verse is notated a whole tone lower than it is in other MSS, beginning with “domine.” To replicate the intervallic structure of the other versions at this level would require a high eb, corresponding to the high f in the other MSS.

Although Pa 776 and Pa 1121 suggest possible nondiatonic practices, their readings are inconsistent with one another. The other pitched MSS give too little information to posit either of these distinctive readings as a preferred version. For this reason, I have adopted the majority reading of Ben 34 as the base version.

There are very few melodic variants. The Aquitanian MSS provide a distinctive reading of one passage, in the second occurrence of “quoniam” in v. 2, as described below.

(p.412) Mo 159:

  • v. 2 “oraBO” (second time): Mo 159 employs a b‐flat here, whereas most MSS that distinguish between b‐flat and b‐natural indicate a natural.

  • v. 2 “TE” (second time), a different opening of the melisma: dcc(ori.)a cb adcc(ori.)a cb Ga cb(quil.)a (etc.).

Pa 776: v. 2 “quoNIam”(second time): cdc ab(quil.)c abcb

Pa 780:

  • v. 2 “quoNIam” (second time): cdc ab(quil.)c abcb

Lo 4951:

  • respond “diLI/GUNT”: DEFEDE/ED

  • v. 1, Verba mea, is written out

  • v. 2 “quoNIam” (second time): cdc ab(quil.)cbcdc

Cai 61:

  • respond, “gloriaBUNtur”: BD c cdc c cdc

  • v. 2: Cai 61 presents a shorter version of this verse, with the text repetition omitted.

RoV 52: v. 2 “MA/NE”: FGaGFG/GF (replaced with a more common cadential pattern)

Mod 7: has a lacuna here

Pst 120:

  • v. 1, Verba mea, is written out

  • v. 2 “TE” (second time): dcdcca GdcdcG acba

  • v. 2: The word “mane” and its music are omitted

  • v. 2 “MEam”: cac cac ca ccc ac cac ca ccc dc ca c ca ccc dfffffdc dfffffdcdc dedc dcdc (etc.)

Pad 47:

  • v. 2 “TE” (second time): dcca cb deccacb acba

  • v. 2 “MEam”: cac cac cca ac cac cca ccc dca cca ccc d fff fffdc ff ff ffffdc dc deca cb (etc.)

Be 40078: transmits the second verse only. The respond and verse are notated at unusual positions on the background scale, the respond an octave above its position in Ben 34 and the second verse a fourth higher. In the case of the respond, the reason may lie in a melodic variant that requires a b‐flat on the third syllable of “gloriabuntur,” b‐flat d ff gf ff gf. The scribe may have written the respond an octave higher because the low B‐flat was not part of the notational system. The same melodic reading of “gloriabuntur” is found at the normal position in Cai 61 above, but without an accidental indicated, and in Tri 2254, where the whole respond is written at the affinal position.

Roman MSS:

Vat 5319:

  • respond, “TE”: FGaGbaGF

  • respond, “doMIne”: Ga(ori.)GF

  • verses: Vat 5319 has no verse division indicated between the first and second verses, and lacks the text repetition of “quoniam ad te orabo.”

  • (p.413)
  • v. [2]: “doMIne”: Ga(ori.)GF

RoS 22 has the cadential variant in Formula B is that is used consistently in this MS: GaGFEFG rather than GFEFG (as in Bod 74 and Vat 5319): This variant occurs on “omNES” and “tuE.”

  • “doMIne”: GFED

  • “IUsto” [sic]: FGaGaGF

  • “coronaSTI”: F GaGbaGFGa

Mihi autem (86)

This piece exhibits numerous melodic variants and, in the final verse, pitch‐level variants. The base version of Pa 776 presents a pitch‐level profile of the third verse that reflects the most probable unemended version. Pa 776, however, is anomalous among core sources in two passages, namely the melisma on “meam” in v. 1 (phrase 4) and its repetition in v. 2 (phrase 7). Although this melisma is a point of variance, especially in peripheral MSS, most sources have a reading closer to that of Mo 159: DF GFF (ori) EDCD FFD CGE GF GE FGDC D FF F(ori.) D FGF GF GF(quil.) E.

For this melisma, the version of Mo 159 is implied in the neumes of the three early adiastematic MSS examined, including Cha 47, Ei 121, and La 239. Several MSS in the peripheral group, however, present a version closer to that of Pa 776, including Gr 807, which has the same internal descent to D and c: Gr 807: CFGFDCD EFD FGF GE FD CFDCD FFD FGF GFGFE. This latter reading is also implied by the neumes in SG 339, suggesting that an alternative reading of this passage was a part of the early melodic tradition.

The parenthetical b‐flats in the base transcription are derived from the indications in Mo 159. In a great majority of cases, the b‐flats in the three verses are matched in later MSS that employ a sign for b‐flat. Pa 1235 and Cai 61, however, also indicate a b‐flat in the opening of the respond, on the second syllable of “mihi,” and at “sunt” and “eorum” in the respond. At “eorum” a b‐flat is also indicated in Be 40078.

Pitch‐level variants: In Mo 159 and most Aquitanian MSS, the final verse (phrase 8) is followed by a varied, written‐out repetendum to “nimis confortatus est,” with a different melody for “nimis confortatus.” In the base reading Pa 776, the repetendum is notated through the first two syllables of “principatus,” indicating that the full repetendum was to be sung following the melodically variant “nimis confortatus est.” In Pa 776, the cue to the repetendum is notated a fifth above the expected level; at the notated level, the ending of the second verse and repetendum extend well beyond the deuterus range of the respond and first two verses; a full repetendum at the written level would conclude on the affinal b rather than the original final, E. The pitch‐level profile of Pa 776, with the same inconsistency of repetendum, is found in other Aquitanian MSS, including Pa 1121, Pa 1134, and Pa 1137.1 In most MSS, however, the verse is written a fifth below Pa 776 beginning at “tu formasti.” In the majority reading, the second verse ends a fifth below the written level of Pa 776 and the repetendum concludes on E.

There are two possible ways to interpret the notation in Pa 776. The first is that the “incorrect” repetendum indicates a discrepancy between the notated and performance levels of the verse and that the whole verse should be sung a fifth lower. This hypothesis (p.414) rests on the assumption that the melodic material of the repetendum was sung at the same pitch level each time it occurs. The second possibility is that the notated level reflects actual performance, and the expansion into the higher part of the range at the end of the verse resulted in a migration of the final from E to b. I have provided both versions in the transcription, labeled “written level” and “hypothetical performance level.”

I would cautiously hypothesize that the lower reading, the hypothetical performance level, is the intended one. Among the adiastematic MSS examined, only Ei 121 gives a clue to the performance level. The third verse, which immediately follows the second verse (with no repetendum), begins with an “im,” indicating that the verse begins at a pitch lower than the end of the second verse. This reading is consistent only with Pa 776 at its hypothetical performance level. This evidence, coupled with the usual meaning of the “incorrect” repetendum cues, suggests that the lower reading is the preferred one.

The evidence for this hypothesis is strengthened by the variant readings in other MSS. At the end of the verse, beginning with “et posuisti” (phrase 10), most versions are a fifth below the written level of Pa 776, with notated b‐flats in Mo 159. The various versions, however, move to this lower level at different points. The variety of readings suggests that an emendation has taken place in response to the presence of nondiatonic pitches. At the beginning of the verse, the nondiatonic pitches low‐B‐flat and E‐flat would be required to reproduce the intervallic structure of the majority reading at the hypothetical performance level suggested by Pa 776 and others like it. By the close of the verse, most MSS have moved to the hypothetical performance level, a fifth below the written level of Pa 776, at a point when nondiatonic pitches are no longer required.

The hypothetical a‐flat at “formasti” (phrase 9) is suggested by the intervallic relationship to Mo 159. Just before this point, at “tu,” Mo 159 moves from the written level of Pa 776 to position a whole tone lower, a reading unique among the MSS in the sampling. At “formasti,” Mo 159 is briefly a fourth below the written level of Pa 776 and employs a b‐flat, equivalent to an e‐flat at the written level of Pa 776 and, by extension, an a‐flat at the hypothetical performance level. This hypothetical practice, however, is not reflected in most MSS. Most sources notate this passage a fifth below the written level of Pa 776 with no partial transpositions. The pitch in question is written as a regular a, suggesting that this nondiatonic practice was either confined to Dijon or suppressed in other areas.

In most MSS, the transpositional variants begin at “tu formasti.”2 In relation to the written level of Pa 776, these variants may be summarized as follows:

  1. 1. Pitch level at verse opening, “ecce”:= Pa 776: all except the Beneventan MSS, which have partial transpositions here. Ben 34 and 38 are 2↓ Pa 776, with a notated b‐flat. Both join Pa 776, on the final syllable of “domine.” Ben 35 begins as Pa 776, but is 2↑ Pa 776 on “[domi]ne cognovisti,” rejoining at “omnia.”

  2. 2. Pitch level at “tu” (phrase 9):

    • = Pa 776: Pa 780, Pa 1121, Pa 1137, Pa 1134, Pa 1137, Pa 1235

    • 2↓ Pa 776: Mo 159

    • (p.415)
    • 5↓ Pa 776: Lo 4951, Ben 34, Ben 35, Cai 61, Gr 807, Be 40078, Rei 264, Mod 7, RoV 52, Pst 120 (beginning within the melisma; the melisma reads DG acaG aFE FFFD etc.)

  3. 3. Pitch level at “formasti” (phrase 9):

    • = Pa 776: Pa 780, Pa 1121, Pa 1134, Pa 1137, Pa 1235

    • 5↓ Pa 776: Mo 159 (interval of transposition changes briefly to a fourth on the second syllable of “formasti”: aG b‐flat‐a‐b‐flat), Ben 34, Ben 35, Pa 1235, Lo 4951, Cai 61, Rms 264, Gr 807, Be 40078, Mod 7, RoV 52, Pst 120.

  4. 4. Pitch level at “et posuisti”:

    • = Pa 776: Pa 1121, Pa 1134, Pa 1137

    • 5↓ Pa 776: Pa 780, Mo 159, Ben 34, Ben 35, Pa 1235, Lo 4951, Cai 61, Rei 264, Gr 807, Be 40078, Mod 7, Rov 52, Pst 120

Melodic variants: There are several points of variance in this offertory, among both core and peripheral MSS. The following summary gives the distinctive Beneventan variants first, followed by a listing of variant points, then a list of variants particular to individual peripheral MSS.

Beneventan distinctiveness: The following gives the readings of Ben 34. There are very similar readings in Ben 38 and Ben 35.

Ben 34: v. 2 “LONge,” shortened melisma: aG G(ori.)F a baG aGE G aGF GFD FF FD FGaGaG babG (liq.)

v. 3 “Omnia”: FGFD FGaGaG a cc ca a (liq. ori.) (Lo 4951 has similar shortened melisma)

Beneventan MSS lack the melodically varied repetendum that appears in Mo 159 and the Aquitanian MSS.

Other melodic variants in core MSS:

  • respond, “AUtem”: Lo 4951: DDG

  • v. 1 [cognovisti] “ME”: Lo 4951: CD FF aGFF(ori.)D GF DD(ori.) C GFGFF(ori.)D CFDC CDFF aGFG FF(ori.)D CGEFFF

  • v. 1 “cognoVI/STI” [sessionem]: Lo 4951: abGF/FED

  • v. 2 “investiGAsti”: At the beginning of the melisma, E (rather than D) is the bottom note of the melisma in several core and peripheral MSS, including Pa 1121 and Pa 1235.

  • v. 3 “cognoVI/STI”: Mo 159: Da/G b‐flat AG

There is an identical reading in Lo 4951 and several peripheral MSS, including Pst 120 and Be 40078.

v. 3 “Omnia” has several smaller variants in the first half of the melisma:

  • Mo 159: GaGE FG(quil.)aGa G b‐flat aE FG(quil.)aGa etc.

  • Pa 1121: GaFE FG(quil.)aGa Ga FE FG(quil.)aGa

  • Pa 780: identical to Mo 159

Lo 4951 presents a shorted form. The entire melisma reads: GaGE FG(quil.) aGaG ccc a aG(liq.). The text scribe, however, has left room for a longer melisma.

Ben 34 and Ben 35 have a shortened form very similar to that of Lo 4951.

Variants in peripheral MSS:

Pa 1235:

  • v. 1 “DOmine”: FF FaGE FGacbc ac c cGaF babGa

  • (p.416)
  • v. 1 [cognovisti] “ME”: a shorter form of the melisma with melodic variants: CDFF aGF D FEDC GFGF FD DEDC CDFF FaGF GFF DGF

  • v. 1 [sessionem] “MEam” (identical at v. 2 [directionem] Meam): DF GF EDCD F FD FGFGE FD DEDC CDF FD FGFGF GFE

  • v. 2“noVISsima”: FGa b‐flat a aG cccac

  • v. 3 “Omnia”: FGFD FGaGa FGFD FGaGaG ccac

  • v. 2 “TU”: melisma begins EG caG aFE ac (etc.)

Be 40078: v. 1 “[sessionem] MEam” (identical at v. 2 “[directionem] MEam”): DFGFEDCD F FD FGF GE FD CFDCD FF FD FGF GFG FF

Gr 807: v. 1 [sessionem] “MEam” (identical at v. 2 [directionem] MEam): CFGFDCD FFD FGF GE FD CFDCD FFD FGF GFGFE

Mod 7:

  • v. 1 “DOmine”: FGaba FGaba cccG aFbaGa

  • v. 1 “cognoviSTI”[me]: DEDC

  • v. 1 “ressurectiOnem”: GbaGGD FGa Ga bcba

  • v. 3 “cognoVI/STI”: Fa/abaG

  • v. 3 “Omnia”: GaFE FGaGa GaFE FGaGaG (etc.)

Pst 120:

  • v. 1 “DOmine”: EFG aGF EFG aGGG cc (etc.)

  • v. 1 “MEam” (same in v. 2): DEFEDCD F FD FGFGF EDEDC D FFFD FGFGF GFE

RoV 52: The far right side of the microfilm on the first page (the respond and beginning of v. 1) was not possible to read clearly.

respond, “SUNT”: FaGFG

  • v. 1 “ME”: CD FFaGF FD GF E EC GFGFFD GFGFFD (etc.)

  • v. 1 [sessionem] “MEam” (same in v. 2, MEam): DF GFEDCD FF FD DGFGFFD CEDC DFF FD FGFGFE. (This version of the melisma is closer to Pa 776 than to the majority reading of Mo 159.)

  • v. 3 “Omnia”: GaGF FGaGa GaFE FGaGaG (etc.)

Roman MSS: Bod 74 lacks this offertory. The Roman version lacks the second verse of the Gregorian, Intellexisti. In the respond, RoS 22 has the following variants:

  • “auTEM”: Gcb

  • “NImis”: acbaba

  • “SUNT”: aGFEFG

  • “deUS”: EGF GaGFEDEDF

  • “princiPAtus”: FGaGbaGF

Oratio mea (87)

This piece exhibits very few variants. The Beneventan MSS transmit an independent reading in two passages. In the respond, the words “in eternum” read “in excelsis” in most sources, with the same melody. “In eternum,” however, is found in Compiègne, Cha (p.417) 47, and all Beneventan MSS, suggesting that this variant entered the tradition early on. In the verse, on the final syllable of “custodivi,” all non‐Beneventan MSS have a short melisma, as in Mo 159: G FG(quil.)aG ab‐flat a.

Lo 4951 presents a very different reading on “AU/RUM”: de ccd ede ccd aGa b(quil.)cd accca ccca Gab(quil.)c/bc abaG aG.

Cai 61: respond, “loCUS”: a bcbaG

Pa 1235:

  • respond, “loCUS”: aba/aG

  • respond, “ME/A”: cac cdaG GaGF ac dcdc cac cac caG aba aG

  • verse, “PRO/BAvit”: Ga/cd

Be 40078: verse, “PRO/BAvit”: Fa/cd

Gr 807:

  • respond, “MEa”: ca cc daG

  • verse, “AUrum”: deb cd eb cddc def ge ff fd fffdcb cde fd(liq.)

Roman MSS: This offertory is lacking in Bod 74. RoS 22 has the following variants:

  • “et”: cdc

  • “ME/US”: FGaGbaGFEF/FGFE

  • “exCELso”: cdc bcbabcba Gacba Gacba bcdcbabG (5319 is missing the notes in boldface)

  • “aSCENdat”: Gac

  • “domiNUM”: GabaG

  • “meA”: ba cabaG abaG aGFGaG FaG cdcb GbaGF bacbaG

Confessio et pulchritudo (88)

This offertory is sung with the Cantate domino verses of the Christmas offertory Laetentur caeli (7). In most MSS, the verses are simply cued.

There are no significant variants among core MSS. They may be found only in the following sources:

  • Pa 1235: “conSPECtu”: FF FDGF

  • Cai 61: “Eius”: GaGa

  • Pst 120:

    • “conSPECtu”: FF FD FGF

    • “Eius”: GaGa

RoV 52: The whole offertory appears to be written a fourth below its usual position, with a final of low B. The reasons for the notation at this position are unclear. At “PULchritudo,” a b-natural was probably employed at the level of Ben 34 (as it is indicated in Mo 159). At the position a fourth lower, there is no equivalent of this pitch (it would be F-sharp). There is otherwise only one significant variant: “magnificenTIa”: CDECD.

Roman MSS: This offertory is lacking in Bod 74. RoS 22 has no significant variants.

(p.418) Desiderium (89)

This offertory shows variants in pitch level that arise from an irregular tonal structure and the need for low B-flat in the final verse. Pa 776, transcribed at the affinal position, was chosen as the base version because its pitch-level profile, with all verses in the same (transposed) plagal range, is clearly the preferred reading. If this version were transcribed in the untransposed range, the low F's in the final verse would be B-flats. In the final melisma, moreover, it is likely that the b's were intended to be sung as b-flats, the equivalent of E-flat at the lower position.

The final verse of the offertory, with its cadence on F, has a fifth-mode melodic structure. This offertory thus presents a problem of modal structure similar to that of other sixth-mode offertories: a structure typical of the mode 5 is projected a fifth below the final c. A few sources included in the sampling present a structure more consistent with a traditional understanding of the modes: the respond and first two verses are notated in the untransposed range, closing on F, and the second verse is written in the normal authentic range. This version, however, is clearly an emendation, as evidenced by the status of Pa 776 as the majority version and the presence of the singular readings discussed below.

Pitch-level profile: = Pa 776: Pa 780, Pa 1134, 1135, 1137, 1137, Pst 120, Mod 7, RoV 52, Gr 807. (In Pa 1135, Pa 1136, and Pa 780, the pitch-level profile of Pa 776 is confirmed by a notated repetendum cue to “posuisti,” which begins on c in the untransposed range. In Pa 1121, the verse starts on a new line and lacks a repetendum cue.)

The following sources notate the respond and first two verses in the untransposed plagal range, with F as the final. The final verse (Inveniatur), however, is written in the normal authentic range, with the tonal focus on c: Mo 159, To 18, Rei 264, Cai 61.

Several sources present distinctive readings of the final verse: Ben 34, Lo 4951, Pa 1235, Be 40078. These sources require some explanation. In Ben 34, the respond and verses are notated in the untransposed range until the word “tuis,” where the low B-flat first appears. A new line starts on the final syllable of “tuis.” While a custos indicates that the syllable “is” will begin on C, implying that the next note is low B-flat, a new F clef is introduced in the next line that moves the range of the verse up a fifth on the last syllable of “tuis.” It is doubtful that the new clef indicates a sudden transposition in performance and more likely that it reflects a “correction” of the clefing, a retrospective decision that the whole offertory should be written at the affinal position. This impression is confirmed by the repetendum cue to “posuisti,” which is notated to begin on c rather than F (its position in the respond).

In Lo 4951, both the final verse and repetendum cue begin on new lines. The final melisma of the verse, however, is written a fifth lower than its position in the other sources (relative to the rest of the verse), beginning at the same pitch as the cadence on “oderunt.”

In Be 40078, the transposition to the higher register begins at v. 1, Vitam petiit.

Pa 1235 notates the offertory in the untransposed range until the latter part of the final verse. It presents a unique reading of the final verse, in which a transposition to the higher register is made gradually and transitionally in the course of the third verse. By the cadence on “oderunt,” the verse is in the authentic range, so that the cadence falls on F.

There are very few significant melodic variants among Mo 159, the core group of Aquitanian and Beneventan sources, and MSS from German-speaking areas. Some Italian MSS, however, present a shortened version of the final melisma of v. 3. Pst 120, for example, closes the melisma with the figure acccabaFacdcc, lacking equivalents for the last thirteen notes of Pa 776. To 18 presents an even shorter form of the melisma: accdcdededcdcccdedecccedccca.

(p.419) Roman MSS: This offertory, without verses, is found in two of the three Roman MSS, Vat 5319 and RoS 22. The version of Vat 5319 closes with a repetition of the opening words that is lacking in the Gregorian, along with an apparent cue to a verse, Vitam, with a melodic incipit that is not recognizable as the beginning of a sixth-mode offertory verse. Hucke identified the incipit as a reference to introit and communion psalmody. The reasons for its appearance here are unclear, and it is not been included in the transcription of Vat 5319.1

RoS 22 is notated in the untransposed range rather than the affinal position and exhibits some minor differences from Vat 5319 in its use of Formula B, particularly its final cadential segment, where it shows a typical preference for the longer form: GaGFEFG occurs in the place of Vat 5319's GFEFG several times. RoS 22 also has the following variants:

  • “ei”: GaF-GaGFEFG

  • “eiUS”: GaGFEFG

  • “eUM”: GaGFEFG

  • “pretioSO”: GaGFEFG

Domine deus in simplicitate (90)

In phrase 6, Pa 776 has no note to accommodate the second syllable of “filii,” a probable scribal error. I have supplied the F from the other Aquitanian MSS and indicated it with brackets in the transcription.

The variant modal assignment of Domine deus in simplicitate was addressed long ago by Bomm.1 The internal phrases of the melody suggest the sixth mode. This melody is, in fact, similar to offertories in the sixth‐mode family.2 In many MSS of the core group, however, the respond closes on low C, as in the base reading of Pa 776. In Mo 159, the respond is written a fifth higher than it is in Pa 776 as I have transcribed it, closing on G. In Mo 159, however, the repetendum after the second verse, to “deus israhel,” is written a fifth too low, starting on a rather than e, which implies that the respond was intended to be sung a fifth lower and close on low C. The notated level of the repetendum in Mo 159 implies that the Aquitanian sources should be transcribed to close on low C, rather than G, in order to maintain the intended relative pitch level between the respond and verses. The repetenda in the Aquitanian sources confirm this impression.

Low C, of course, is not a legitimate final. This inconsistency with modal theory explains the higher notation of the respond in Mo 159 and, in part, the variant endings in other sources. In some MSS, the end of the respond is emended to close on F, or, less frequently, D. In other cases, it closes with a cue to repeat the opening of the respond, “domine deus,” which brings the respond to a close on F. The latter practice is mentioned by John of Afflighem, who disparages it.3 The wide variety of endings in the respond can be summarized as follows: (p.420)

  • Ending on G, without return to opening: Mo 159 (notated a fifth above the other MSS)

  • Ending on low C, without return to opening: [Pa 776, Pa 1121, Pa 780] Pa 1235, Cai 61

Ending on F, without return to opening:

  • Lo 4951: “voluntaTEM” is FacaGbaGF

  • Be 40078: “voluntaTEM” is acGF GFD DDF

  • Pa 1132: “voluntaTEM” is FacGFGFDCFF

  • To 18: “voluntaTEM” is FabaGF GaGaFED FFF

  • RoV 52: FacGFGFEDF

  • Frutolf tonary: Fac[G]F G[FD]CF

Ending with return to opening “domine deus” (thus on F): Ben 34, Ben 35, Ben 39, Mod 7, Pst 120, Pad 47, RoV 52 (crossed out by a later hand).

Ending on D: Tri 2254: “voluntaTEM” is FacGFED

Given the sixth‐mode traits of the melody, the C‐ending in Aquitanian MSS (and the G ending in Mo 159) comes as something of a surprise and is probably to be regarded as the lectio difficilior. In contrast to the melodic consistency of the sources ending on C, the sources that close on F do so in different ways, suggesting that they are emendations made to close the respond on a legitimate final and perhaps also to bring the ending of the melody into conformity with the melodic traits of its internal phrases. The C‐ending is thus the preferred reading, and Pa 776 was accordingly chosen as the base version. The repetendum in Pa 776, “dicentes,” probably was intended to lead back to “deus israhel,” as indicated in the Roman version.

Other significant variants: There are few consequential variants, with the following exceptions.

  1. 1. The order of the verses varies.

  2. 2. The relationship between the closing melismas of the two verses. In Pa 776, the verse Fecit closes with an extended version of the melisma that closes the verse Maiestas, with the added notes acaG acaGa. The same reading is found in the other Aquitanian MSS examined and most MSS outside the core group, with the Italian exception mentioned below. This reading is also consistent with the adiastematic MSS La 239 and Ei 121. In the Beneventan tradition, the melisma in the Fecit verse is nearly identical to the Aquitanian one, but Beneventan MSS have a shortened version of the final melisma in the verse Maiestas (on “dominum”).

    Ben 39: c dc ccc dc ca cbc ccbGaG cbcda FGFGF FGa c ca.

    RoV 52 v. Maiestas, “DOminum”: cdc ccc dca cb cdaGc dc ccc dca cb ccbGa cb cdaF GFGF FGac caG acaG acaGa (same melisma in v. Fecit)

  3. 3. v. Maiestas, “DOmini” (phrase 5): Different reading in Mo 159 and Pa 1235:

    Mo 159: FG aG Ga

    Pa 1235: FGaGa

Roman MSS:

Bod 74 lacks this offertory. In Vat 5319, the respond and both verses are written at the affinal position, with traits typical of other sixth‐mode offertories, which attests to a broad tonal continuity with the Gregorian version.

The Roman respond closes with a cue to the beginning of the respond, “domine deus.” It is unlikely, however, that the melody concluded at “deus,” as it did in Benventan and many Italian MSS, because “deus” occurs in the middle of a statement of Formula B. The performance is more likely to have ended at the cadence on “simplicitate.”

(p.421) RoS 22, notated at the regular tritus position, with the final on F, has the following variants:

  • “LEtus”: FGaGbaG

  • “POpulus”: FGaGaGF

  • “est”: GaGFEFG

  • “gauDIo”: GaGF

  • “cusTOdis”: FEFG

Memor sit dominus (91)

With an original assignment to the Mass for the ordination of a bishop, this offertory is found in few of the MSS examined. In some later MSS it appears as part of the common of saints for a bishop or confessor. The offertory is transmitted, without verses, in Vat 5319. Outside of the Aquitanian tradition, Memor sit is found in only in two other MSS included in the sampling, Mod 7 and Pad 47, in a distinctly variant version. A fuller examination of these variants, however, is outside the scope of this study.

Domine Iesu Christe (92)

It is doubtful that the Requiem Mass formed a part of the core repertory transmitted from Rome to Francia in the eighth century. It is lacking in the early sources indexed in Antiphonale Missarum Sextuplex, appearing for the first time in La 239. Domine Iesu Christe, then, is one of the few demonstrable instances of reverse, Frankish-to-Roman transmission in the offertory repertory. This piece is not consistently present in Gregorian MSS, some of which transmit Domine convertere as the offertory for the Requiem Mass.

Given the late introduction of this piece into the repertory, it is not surprising to find a varied melodic and textual tradition among Gregorian MSS. Although all versions examined employ a simple second-mode recitation formula, the versions differ in the lyrics of the opening of the respond and in melodic details. The Aquitanian version, for example, lacks the words “de sancto tui” at the opening of the respond, has a different melody for the verse Hostias, and has three additional verses. A full examination of these variants, however, is outside the scope of this project.

Stetit angelus (93)

Although Stetit angelus is not part of the core repertory, its shares a partial text, in the first verse, with the Roman St. Michael offertory, In conspectu, a piece also given as the Michael offertory in Mont-Blandin. As discussed in chapter 2, it is not clear whether the two versions are cognates. Both Michael offertories are included here, with the corresponding passages provided in side-by-side transcription.

(p.422) Aside from some pitch-level variants among Italian sources, Stetit angelus exhibits few variants between MSS. The Beneventan MSS present a reading very close to that of Mo 159 and the Aquitanian sources, with the following very small exceptions:

  • v. “in conspecTU”: In most MSS, the end of the melisma has a C that is lacking in the Beneventan MSS.

  • respond, “iuxta aram”: Most MSS have a different text underlay, with the first three notes of “aram,” DGF, given to the last syllable of “iuxta.”

  • v. “ANgelorum”: Mo 159 and most other MSS have CD.

  • v. “doMI/NE”: The text underlay indicated in the transcription is strongly suggested by the placement and grouping of the neumes, but in most sources the underlay is cabc/aG.

Some Aquitanian MSS, including Pa 776 and Lo 4951 in the core group, have an additional verse, Factum est silentium, which is based to some extent on melodic material from the first verse.

  • Gr 807: Many of the b-flats in Ben 34 are c's, at “habens,” “ascendit,” “conspectu.”

    • “Thurbilem” is written a whole tone lower than it is in other MSS, beginning on the second syllable.

  • Be 40078: respond, “alleluia” (words and music) omitted

  • Pa 1235:

    • respond, “iuxta aram”: same text underlay as Ben 34.

    • respond, “ascendit”: No b-flats indicated.

Cai 61: The verse is written a fourth below the majority reading of Ben 34, joining at “et confitebor.” Because the b's in the majority version are usually indicated as flats (matching the version of Mo 159) the two readings are intervallically identical.

Italian MSS: For reasons that are unclear, many Italian MSS notate the respond and both verses at the affinal position. A fuller sampling of Italian MSS is presented in Justmann's dissertation.1 In one version transcribed by Justmann, Rossi 76, the respond is notated with the final on G, retaining protus characteristics with the consistent use of b-flat. Further variants occur at the end of the verse, beginning with “sanctum tuum”; the Italian sources in his transcriptions end the verse variously on high g, a, D, or low G. These endings hint at a possible nondiatonic practice in the Italian tradition. In most MSS, the melodic figure on “tuum” articulates a minor third, G-b-flat (a flat is indicated in all MSS with a sign for it). In several of Justmann's MSS, however, the interval articulated is a major third, f-a in RoV 52, and c-e in Vercelli 161 and Vercelli 162.

  • RoV 52: The whole offertory is written at the affinal position, a fifth above the majority version of Ben 34.

    • respond: text reads “super” in place of “iuxta.”

    • respond “aSCENdit” (5↑Ben 34): melisma begins ae dee gge (etc.)

    • v. “domine”: text underlay is the same as the Beneventan MSS, contrasting with most other MSS. At the end of the verse, the transpositional level changes to an octave plus a third above the majority reading.

  • (p.423)
  • Pst 120:

    • “turiBUlem”: ecdbaGa. The whole offertory is written at the affinal position, a fifth above the majority version of Ben 34.

    • v. “domine”: text underlay is the same as the Beneventan MSS, contrasting with most other MSS.

  • Roman MSS: Bodmer 74 lacks In conspectu angelorum. RoS 22 has the usual variant in the cadential segment of Formula B: GFEFG in 5319 is GaGFEFG in RoS 22 on tiBI and tuUM.

    • “temPLUM”: FEFG

    • “doMIni”: FGaGaGFG

Beatus es Simon (94)

As discussed in chapter 2, this offertory was probably added to the Roman repertory after the transmission of the tradition to the Franks. Although it is not part of the standard Gregorian repertory, it is found in two Beneventan MSS, Ben 39 and Ben 35. The Beneventan version is a contrafact of Angelus domini, and the two MSS exhibit a pitch‐level variant at the end of the verse that is also found in the Beneventan tradition of Angelus domini. The end of the melisma on “Christus” is written a fifth lower in Ben 35. In the case of Angelus domini, the version of Ben 35 is an anomaly, whereas that of Ben 39 is found in all MSS outside of the Beneventan sphere.

In phrase 5, the bracketed notes over “filium” are not visible on the microfilm of Ben 39 and are supplied from Ben 35.

Roman MSS: The text underlay in Vat 5319 is unclear at “Petre.” The syllable “tre” is written under last four notes of the melisma on “pet.” I consider this to be an error because of the usual cadential function of the notes acbaG, and have emended the transcription accordingly. RoS 22 has the expected underlay.

Bod 74 lacks this offertory. RoS 22 shows the following variants from Vat 5319:

On “PETre”: varied reading of the melisma: cbabaG aGacba dcdbcba GbaGF GFEFGFEFEDE (etc.)

  • “SANguis”: G ac cbcabaGF acbabG

  • “revalaVIT”: aaGFaGFG

  • “PAter”: Gabac

  • “EST”: G acabaGF GbaGaG

  • “CElis”: cbabaGaG acba dcdbcba GbaG FGFEFGFEFEDE G ababaGaF acbabaGa

Notes:

(1.) Sidler (1939), 11, Pitman (1973), 82, Frasch (1986), 1:161, and Hankeln (1999), 1:94.

(2.) The corresponding note at “erubescam” in Gr 807 is not an E but a D.

(1.) Sidler (1939), 23–25.

(1.) See Bomm (1929), 162–64; Frasch (1986), 1:207–8.

(2.) Sidler (1939), 28–31; Frasch (1986), 1: 207–8; 2:293–303; Hankeln (1999), 1:95–96; 2:135.

(3.) Hankeln (1999), 1:95–96.

(4.) Sidler (1939), 29.

(5.) Frasch (1986), 1:173–76.

(1.) Hansen (1974), 420–21; Hankeln (1999), 3:18.

(2.) Bomm (1929), 175.

(3.) Sidler (1939), 45–46.

(4.) Ibid.

(5.) Hankeln (1999), 1:100.

(6.) See the summary in Hankeln (1999), 2:213.

(7.) Hankeln transcribes the end of the melisma at the level of Mo 159 and 776, but it seems to be written a tone higher, consistent with the repetendum on E. See Hankeln (1999), 3:72.

(1.) See Hankeln (1999) 1:101–2; 2:137; 3:20–21; and Hankeln (1995), 539–60.

(1.) Sidler (1929), 60–65.

(1.) In Pa 776, this passage marks a change in the value of the dry point line, indicated with an “equaliter” sign, probably because of lack of writing space. While the “e” does not function as it normally does here, the pitch level of the following passage (the second “nos”) suggests a reading of this cadence on G, the solution adopted by Hankeln and Frasch. See Hankeln (1999), 3:23; and Frasch (1986), 2:338.

(2.) Sidler (1939), 68–70; Pitman (1973), 264–69; Hankeln (1999), 1:102–3.

(3.) The pitch level of Lo 4951 is not clear because the verse starts on a new line without a custos. The verse proceeds in a way that suggests a beginning at the level of Pa 776.

(4.) See Sidler (1939), 69.

(1.) Justmann (1988), 2:1042–43.

(2.) Justmann s transcriptions show a similar reading in two other sources, Mod 13 and Ver 161. Justmann (1988), 1:492.

(1.) The notated level (in relation to the first verse) is that of Lo 4951.

(2.) Stäblein and Landwehr‐Melnicki ed. (1970), 273.

(1.) Mü 10086 indicates a b‐flat as the top note on “confiteri” in the respond (corresponding to c in most other MSS), followed by a natural on “domino.”

(1.) Sidler (1929), 44–47; Hankeln (1999), 1:103–5; 2:139, 3:31–33.

(2.) Hankeln (1999), 1:104; 2:139.

(1.) I was not able to read the version of Mod 7 on the microfilm.

(1.) According to Pitman (1973), 224, it is also found in Pa 1135.

(1.) See the summaries in Hankeln 1 (1999), 105–6; 2, 139; and Pitman (1972), 107–8.

(2.) Hankeln 2 (1999), 139.

(3.) Mo 159 has one b‐flat in this verse, on the last syllable of “perveniet,” which would be E‐flat at the position a fifth lower, but in Mü 10086, the corresponding note is F. Like most later German MSS, Be 40078 lacks repetendum cues.

(1.) Hankeln (1999), 1:105–6; 2:140. Most Aquitanian MSS appear to start the verse a fourth lower but lack clear repetendum cues.

(2.) In additional to the summary below, see Pitman's descriptions of the transitional “bridges” in other MSS. Pitman (1973), 77–81.

(3.) The repetendum in Pa 1121 suggests a performance level equal to Ben 34 and Pa 776. There is an “equaliter” sign functioning as a custos between verses, however, that suggests a performance a fourth lower. As Pitman suggests, this sign might be an error. Ibid.

(4.) See Hankeln (1999), 2:140.

(1.) “Hoc offertorium plurimi male emittunt vitantes heptaphonum qui in fine est, quia eis absonus videtur. Unde et in quibusdam libris in principio eius invenitur et in fine repetunt.” Johannes Affligemensis, De musica cum tonario, ed. J. Smits van Waesberghe, Corpus scriptorum de musica, vol. 1 ([Rome]: American Institute of Musicology, 1950), 188. Translation is that of Warren Babb in Palisca, ed. (1978), 180.

(2.) Bomm (1929), 155–58.

(3.) See Alexander Rausch, Die Musiktractate des Abtes Bern von Reichenau: Edition und Interpretation (Tutzing: H. Schneider, 1999), 277; Frutolfi Breviarium de Musica et Tonarius, ed. Coelestin Vivell (Vienna: Akademie der Wissenschaften Sitzungberichte, 1919), 157; Ugolini Urbevetanis Declaratio musicae disciplinae, ed. Albert Seay, Corpus scriptorum de musica, vol. 7/1 ([Rome]: American Institute of Musicology, 1959), 168.

(1.) On the problem of modal assignment, see Bomm (1929), 169–71. Both problems are also discussed in Frasch (1986), 1:189; and Hankeln (1999), 1:108.

(1.) See also Pitman (1973), 103.

(1.) Levy (1971), 40–61.

(2.) Hankeln (1999), 2:224.

(3.) Ben 34 and 35 move to the higher register on “iusticiam”; ci/am is a/Ge(liq.).

(1.) Pitman (1973), 269–72, 339–41.

(2.) Pa 1121 exhibits some differences leading up to this passage, very similar to Trier 2254 described below. Beginning within the melisma on “iuSTIciam,” it is a whole tone below Pa 776, with recitation on b–flat at “iuSTIciam’” and ci/am is F/DF. “Non abscondi” begins on a new line and is preceded by a custos (“following “tuam”) that places “non abscondi” and the rest of the verse a whole tone higher than the other readings. This is probably to be regarded as an error.

(3.) Ben 34 and 35 move to the higher register on “iusticiam”; ci/am is a/Gc(liq.).

(4.) Trier 2254 also shows some differences leading up to this passage, with recitation on b–flat at “iusticiam”; ci/am is F/DF.

(1.) This passage is discussed in Hankeln (1999), 1:108; and 2:1434–4.

(1.) Bomm (1929), 163–64.

(2.) Possible explanations for the textual variants are proposed in Helmut Hucke, “Zur Aufzeichnung,” 300–302; Jeffery (1992), 29; and Dyer, (1998), 26–29. Hucke proposed that the compiler of the Old Roman sources used Gregorian sources in copying, an argument echoed by Leo Treitler. Jeffery suggested that the Roman respond and first verse were intended as alternate responds. On the basis of the close correspondence of the Roman respond with the Verona psalter and other Old Latin witnesses, Dyer proposed that the respond was discarded in the Gregorian tradition because the psalm translation was “archaic and unfamiliar.”

(3.) See note 2.

(1.) See Pitman (1973), 194–98; and Hankeln (1999), 1:108–9.

(2.) See Hankeln (1999), 2:142.

(3.) Ibid., 108–9.

(1.) Pitman (1973), 295–96.

(2.) See Domine in auxilium.

(3.) As in Pitman's consensus reading of the Aquitanian MSS. Ibid.

(1.) In Ps. 139:2, the PsG has “libera me” instead of “eripe me.”

(2.) Justmann (1988), 409–27; 688–705; 970–1006; 1274–91; and 1558–70.

(3.) According to Justmann's transcriptions, the two versions of the verse Qui dixerunt part ways toward the end of the verse, at “super laqueos mihi.” Provins 12 lacks the verse Dixi domino.

(4.) Justmann's transcriptions show the same reading in Modena (Mod) 0.1.3. Justmann (1988), 417–18; 423–24.

(1.) The PsG has “commedere” instead of “manducare” in Ps. 101:5.

(1.) On Domine convertere, see also Hankeln (1999), 108–9.

(1.) Justmann (1988), 447–53; 725–31; 1007–13; 1311–17; 1590–96.

(2.) Karp (1995), 151–64.

(3.) “Si autem multum sunt confusi et in affinibus facilius possunt emendari, eo dirigantur, quemadmodum illud offertorium.” Smits van Waesberghe (1950), 148. See the translation in Palisca, ed. (1978), 155. “If, however, they are much disordered and can be emended more easily in the kindred range, let them be rearranged there, as in this offertory.”

(1.) Bomm (1929), 160–62.

(2.) The problem spot is “in progeniis vestras.” In the respond, “in progenies” begins on low C and the following notes are either DFDF or EGEG. In both repetendum cues, the first two notes of “in progenies” appear to be CD, which would produce a reading like that of Pa 776.

(1.) The F's in the majority version of Ben 34, for example would be low B‐flats at the level of Mo 159 version 2. In Mo 159 version 2, they appear as C's on, for example, “terribilis” in v. 2.

(1.) See Frasch (1986), 1:197–202; 2: 408‐21; Hankeln (1999), 1:106–107 and 2:141.

(2.) See Snyder (2006), 99–106.

(3.) In Pa 1121 the verse begins with an equaliter sign. Here I have interpreted the sign as indicating that the horizontal axis is maintained between sections, and that the verse begins on a. The interpretation of the pitch level thus differs from that of Hankeln 2 (1999), 141.

(1.) Palisca (1978), 156.

(1.) Ei 121, La 239, and SG 339 simply have an unnotated “tunc” or “tunc moyses.”

(2.) Snyder (2006), 99–106.

(3.) Ibid.

(1.) Pfisterer (2007), 42–43.

(2.) Frasch (1986), 445–57; Justmann (1988); Hankeln (1999), 2:109–11.

(3.) Pfisterer (2007), 44–45.

(4.) Hankeln (1999), 2:109–11.

(5.) Hankeln (1999), 3:49–50.

(6.) The Italian MSS are Pistoia 199, Florence 44, and Milan S. 74. See Justmann (1988), 559, 1143.

(1.) Justman (1988), 1115–31; Maiani (1996), 216–17; Steiner (2007), 57–66.

(2.) Justmann (1988), 121–23. Two MSS, Mo and Pa 904, transmit Recordare quod steterim as a separate chant.

(3.) Variants in the Aquitanian verses are illustrated by Hankeln. Particularly notable are the differences in pitch level and melisma length in the Aquitanian second verse. See Hankeln (1999), 3:194–96.

(4.) Steiner (2007), 67–88.

(5.) See the summary in Hankeln (1999), 2:145.

(1.) On the past two syllables on “petiit” and on the second syllable of “saeculi.”

(2.) As Steiner showed, Ott “retransposes” the final melisma of Mo incorrectly, producing F‐sharp's that correspond to the b‐flats of Mo 159. See Steiner (1966), 166–68.

(1.) This argument is made in greater depth in Maloy (2007), 67–70.

(1.) Inveni is lacking in La 239.

(1.) The Compiègne version has one significant textual variant, with an incipit “virga tua est” rather than “virga recta est.”

(2.) Both verses also correspond to the Gaulois Psalter η as the respond does.

(1.) Sidler (1939), 32–42: Hankeln (1999), 1:96–97.

(2.) Hankeln (1999), 1:97.

(3.) And Pa 780 is unique among the sources examined in beginning the transposition earlier, at “quae virum,” where it is a fourth below Mo 159. See the summary in Hankeln (1999), 2:136.

(4.) Hesbert (1935), xxxix.

(1.) Pa 1137 lacks the melodically variant repetendum and instead includes a cue to “nimis” that is identical to the corresponding place in the respond, except written a fifth too high.

(2.) Ben 38 presents an exceptional reading here. At its written level, it appears to be a fourth below the written level of Pa 776, but this reading would be inconsistent with the end of the verse and repetendum to “nimis,” which are a fifth below the written level of Pa 776. This adjustment suggests that a scribal error has been made earlier on in the verse. Because of this lack of clarity, I have not included Ben 38 in the pitch‐level summary.

(1.) Hucke (1980), “Zur Aufzeichnung,” 300.

(1.) Bomm (1929), 158–60.

(2.) The common traits are described in chapter 3. In addition, the verse Fecit salomon shares substantial material, including the final melisma, with the verse Exspectans in Domine in auxilium (27).

(3.) “Hoc offertorium [Domine in auxilium] plurimi male emittunt vitantes heptaphonum qui in fine est, quia eis absonus videtur. Unde et in quibusdam libris in principio eius invenitur et in fine repetunt. Similiter in offertorio Domine deus in.” Smits van Waesberghe (1950), 188. For a translation see Palisca, ed. (1978), 180.

(1.) Justmann (1988), 541–54.