A concordance for an early fourteenth-century motet: Exaudi melodiam/Alme Deus/TENOR revisited

by Elizabeth Eva LEACH

In 1970, Jürg Stenzl published an account of an unusual polyphonic motet copied in a manuscript of chant from a the Abbey of Saint-Maurice in the Valais, Switzerland, together with a transcription of the motet and images of the manuscript. The manuscript, Saint-Maurice MS 4, is a notated gradual of unknown provenance, which has been in the abbey since before 1920.[1] Stenzl noted that the feasts in the gradual suggest that the book as a whole came from Franciscan circles and is not originally from the Valais.[2] On f.123v, a blank page at the end of the Temporale, a hand that Stenzl diagnoses as being a little later than the main hand has added what Stenzl neologizes as a ‘Sanctus Motet’, whose two upper voices, notated in the mensural notation of the ars nova, in tempus perfectum with major prolation, both have texts starting with a triple ‘Sanctus’ invocation and then proceed respectively ‘Exaudi melodiam’ (Triplum) and ‘Alme deus’ (Motetus); the tenor designation is not given.

Certain aspects of the notation led Stenzl to the conclusion that the Saint-Maurice source (hereafter StM) had been ‘translated’ from an earlier notational format, specifically that kind of notation typified by the motets in the seventh and eighth fascicles of the Montpellier motet manuscript (F-MO 196). All three voices of the motet in StM notate perfect breves in the form of a long, whereas imperfect breves have their normal form and perfect breve rests following line ends are also correctly written (a line through one space). The tenor notation, for example, has groups of three longs which are to be read as perfect breves, each group separated by a correctly written perfect breve rest.[3]

Stenzl’s hunch was correct. Since he wrote his article another copy of this motet has come to light, in a fragment now in Dijon, which is indeed in an earlier notational style. And the Dijon version (hereafter Dij) shows that his diagnosis about the notation was not his only correct surmise on the basis of the single copy of the motet he had in front of him. Given that the motet texts make no further reference to the Sanctus after the initial section’s identical triple invocation in the upper voices of the StM version, Stenzl considered the initial reference to the Sanctus a later textual addition to an initial melisma.[4] In the Dij notation, there is indeed no text for this initial section, which is presented as a textless melisma.

Dij forms the flyleaf to a book, F-Dm 35, from the first half of the thirteenth century, which contains the Gospels, Canticum canticorum, and Glossa ordinaria. The book itself has been at Cîteaux since the thirteenth century, although its binding dates from the fifteenth century, according to Barbara Haggh’s 2002 description.[5] Given that the flyleaf dates from the early fourteenth century, that is, after the date of the contents of the book that it serves, it was probably added at the time of binding, by which time the fourteenth-century music (but not the contents of the thirteenth-century book) had fallen out of use. The flyleaf will be viewable on DIAMM via this specific page once the library grants permissions.

Haggh classes the motet among those she discusses as ‘previously unknown’, and her brief description does not therefore note the concordance in StM.[6] This omission leads her to diagnose the untexted melisma at the top of each of the upper voice columns as the tenor of a different, preceding motet. (This is the section text with triple ‘Sanctus’ invocations in both voices in StM.) Comparison with StM provides a shortcut to seeing these untexted lines instead as the opening part of this motet, despite their textless status in the Dijon leaf, and they fit perfectly with the opening of the motet’s tenor part.

Following the motet, the flyleaf also contains the tenor Hec dies, noted by Haggh as identical to the chant tenor used in several thirteenth-century motets.[7] Although she notes that it extends ‘on four staves written across the bottom of the page’, this is not completely accurate. These four staves are written on two lines with a central division, as if they were originally designed for parallel (equal) upper voices of a new motet. If this were the plan, it was clearly abandoned in favour of copying not one but two motet tenors, each entered on a single line of the left hand column, leaving both right hand stave lines blank. The Haec dies tenor is only the first of these two staves. The second staff (the bottom-most line of the page as it survives) might originally have had a textual identification that is now cut away, but from the pitches given it appears to be another tenor found in thirteenth-century motets, its pitches taken from an ‘Alleluya’, and in roughly the same rhythm as that found in the short refrain motet in Montpellier, Endurez, endurez / ALLELUIA (Mo248). The music hand for both these motet tenors looks similar to that for the rest of the leaf, but the hand for the only visible text (‘Hec dies’) is a less careful one, perhaps that of the music scribe who chose to enter some additional music in space left after the text scribe had already finished work.[8]

The two texts of the motet are given in Figure 1.[9]

The ten-line triplum is fairly regular and can be understood as two stanzas of five lines of which the rhyme scheme of the second is an inverted palindrome of the first: i.e. the lines of the second stanza both reverse the pattern of the first and have the other kind of rhyme in each case, with a-rhymes as b-rhymes and vice versa (overall, abbaa bbaab). The two rhymes used are the same in each stanza. The triplum text references its status as a sung imprecation as it asks the Paraclete to ‘hear the melody of Thy servants’ and make them good enough in this life to be as disciples in the heavenly fatherland.

The motetus text is made up of three stanzas, each of four lines and also shows a complex but regular concern for patterns of rhyme. Each stanza here has different rhymes and a different pattern. The first stanza interlaces (abab), the second has couplets (ccdd), and the third has a palindromic rhyme scheme (effe). The text here is less grammatically assured, but asks God to send his Spirit down so that, as in Acts 2:6, each may hear the other speak in his own language.[10]

Although the texts are in Latin, stress is ignored, with the triplum’s ‘vite’ (paroxytonic, i.e. p) happily rhyming with ‘paraclite’ (proparoxytonic, i.e. pp), and the motetus’s ‘deus’ (p) rhyming with ‘spiritus’ (pp). Instead, the poetry works by syllable count (in the triplum 7-syllable lines, except lines 5 and 7, which have 4 syllables; in the motetus with a regular patterning of 4 and 7-syllable lines in each stanza). This strongly suggests a poet more familiar with the norms of Romance poetry, probably a speaker of some variety of French.

The motet’s music

The tenor of the motet is not identified as a chant fragment and can be understood as being in three sections, corresponding to the three sections of the motet, separated by marked section breaks in both sources and here labelled sections A, B, and C.[11] All three of the tenor sections present melodies based on the pitch F, and mainly occupying the fifth up to c, although rising to d in section C. Figures 2A, 2B, and 2C gives editions of these sections.

• The A section, Figure 2A, has one 18-pitch color in the tenor. The upper voices are melismatic in Dij, but with the added text ‘Sanctus, sanctus, sanctus’ in both voices in StM. The tenor is rhythmicized in six groups of three perfect breves, each followed by a perfect breve rest.
• The B section is rhythmicized exactly as in section A, with each 15-pitch color organized into fives groups of three perfect-breve notes and a perfect-breve rest. Its 15-pitch color comes three times (each color is shown on a different system in the Figure 2B, which must be read across both pages). The first two colores terminate in an ‘open’ tonal ending to a, the third in a ‘closed’ tonal ending to F. The two upper voices of this part of the motet are each fully texted with the mutually independent (but thematically related) texts given in Figure 1.
• In the C section the tenor proceeds entirely in perfect breves with no rests grouping them. The 24 pitch sequence is repeated (although the last part of the tenor is missing in StM), giving two complete, identical colores. I would agree with Stenzl’s diagnosis that the piece was composed ab initio as a three-part piece because of upper voices’ similarity in range, exchange of musical material in equivalent parts of different tenor colores (cf. B section bars 4–9, and 24–29, boxed in Figure 2B), and overlapping phrase lengths.[12] Stenzl also draws attention to the striking use of ‘isolated’ perfect breves in both voices and the presence of rests in the middle of verbal phrases (lines) as further evidence of the necessity of the piece being in three voice parts. Given that the addition, substitution, and subtraction of voice parts and texts is absolutely standard in thirteenth-century motets but relatively rare in the motets of the named composers of the fourteenth century (Vitry and Machaut in particular), this feature might tend to support a dating that places this particular motet in the fourteenth century.

The Dijon copy is cut away on the right hand side, which means that parts of the motetus and tenor are missing. Nonetheless, the surviving music shows few variants excepting the greater use of ligatures in this notational style, which clarifies the texting of the upper voices. The variants and small number of errors are listed in the Appendix below.

A ‘peripheral’ motet?

Thus far, Stenzl’s perceptions about the motet’s notation and texting have largely been confirmed by the discovery of a concordance. His further conclusions about the place of the motet in music history, however, are more questionable, if only because they rest on the assumptions of much older scholarship in which the imprint of a nationally (or at least patriotically) driven music history is clearly perceptible. Stenzl’s correct claim about the irregular use of long forms for perfect breves in StM betraying the translation of an earlier notational style into ars nova notation is accompanied with the thought that the notator of this version was perhaps located somewhere ‘peripheral’, because he hadn’t quite mastered the art of ars nova notation.[13] While Stenzl might mean that this particular notator was working somewhere less geographically central, his claim folds easily into the material in the second part of the article, which argues that the motet itself is from such a periphery, specifically from England, on the basis of features that it shares with other motets that had, at that time, been identified as English. Stenzl’s point of departure is Besseler’s 1927 identification of certain pieces in which he claimed the motet and conductus styles had become so enmeshed that it was possible to view them as leading towards the ars nova.[14] Besseler’s list, as repeated and expanded by Stenzl, consists of eight motets from the last two fascicles of Montpellier. In listing them again here, I not only summarize the features that Stenzl draws out from them as indicative of Exaudi melodiam/Alme Deus’s place in this group, but also signal my doubts about the tightness of the grouping by offering thoughts on how unalike some of these features really are.

1. The first motet on Stenzl’s list, Iam nubes dissolvitur/Iam novum sydus oritur/SOLEM (Mo275), the only widely copied motet in the list, appearing in five further sources, beyond its rather fragmentary appearance in fascicle 7 of the Montpellier manuscript.[15] Stenzl notes various similarities to Exaudi/Alme: it has a melismatic opening (although it lacks a closing melisma); one of the similar single upper voices sings during tenor rests; there are isolated notes in the upper voices; and it is ‘isoperiodic’. Unlike Exaudi/Alme, though, the color and talea are coextensive (12 notes), although the rhythmic organization can be viewed as being made up of two different rhythmic cells of two and three notes respectively, arranged ababa, which together form the 12-pitch tenor color. Unlike Exaudi/Alme, the color remains the same throughout the motet, coming four times in total, the first time for the opening melisma and the following three times for the upper-voice texts. There is a high degree of musical repetition in the upper voices between colores, and the kind of exchange seen in the boxed section of Figure 2B above is much more pervasive, strongly confirming that this motet, too, was written in three parts from the start. The upper-voice texts, however, are rather unlike those in Exaudi/Alme, but more similar to several other motets in this list, in that they share significant text between them, which they present in short phrases after one another so that one voice typically has its isolated notes (sung melismatically) while the other has a line of text, and then the two swap round, often with the same text (from single words to whole lines), although not always (as is the case, for example, in Mo339) with identical musical material. The aural effect is thus one of rather shorter-range antiphonal exchange, at least texturally, overriding the ‘isoperiodic’ effect of the color-repetitions’ ’stanza’.
2. The second motet, Salve, virgo virgunum/Salve, sancta parens/OMNES (Mo300; images here and here), is, like Exaudi/Alme, in three sections, the first of which is a melismatic introduction and the final section of which is virtually melismatic, setting only an ‘Amen’. The upper voices have overlapping phrases and frequent isolated notes in the middle of phrases. As in Mo275 but not as in Exaudi/Alme, the different sections of the motet all have a same tenor pitch sequence, which here comes once each in the introduction, once in the Amen, and twice in the middle, texted section. Each color has 10 pitches (6 of them the final pitch, F) but the very short tenor talea presents each color statement as five two-note groups, separated by rests. The upper voices have the same text save for the initial line in the Triplum, which is unique to that voice. The effect is less of color-stanza structure, as arguably in Exaudi melodiam/Alme Deus, and more of an antiphonal pattern, at least in the texted, central section.
3. Number three in Stenz’s list, Marie assumptio/Huius chori suscipe cantica/TENOR (Mo322; images here, here, and here), also has three sections, the opening and closing two being melismatic, and, like Exaudi/Alme, each section has a different tenor color. But in Mo322 both of the first two sections repeat their colores, while the third is made up of four different, three-note groups, each immediately repeated. The first color, as was the case throughout in Mo275 above, is coextensive with its talea and has a palindromic grouping of rhythmic cells within it (ababa); it has 21 pitches. The second section’s color has 15 pitches organized into three five-note rhythmic groups. As in the third section of Exaudi/Alme, the tenor of the third section of Marie assumptio/Huius chori suscipe cantica proceeds in equal notes with no rests. The upper-voice texts are differentiated from each other, much as those of Exaudi/Alme are, although there is no voice-exchange or stanza-like repetitions in the music of the upper voices when the colores repeat.
4. The fourth motet, Alle–psallite cum–luya/Alle–psallite cum–luya/ALLELUIA (Mo339; images here and here), is in three voices with a single text and uses ‘voice exchange’, which Stenzl notes as if it is like that seen in Exaudi/Alme. However, the scale and aural effect in Mo339 is rather unlike Exaudi/Alme. Alle–psallite cum–luya has three rather short tenor segments, each a single-talea color, each increasing in length, and each immediately repeated with the upper-voice melodies and text exchanged.[16] In each of these three short two-color sections, one voice sings the text, the other has a melisma. The text is an increasingly long trope within the word ‘alleluia’, completely unlike the more complex texts of the other motets on this list. These three two-color sections are rounded off by a single color with the text ‘Alleluia’ in both voices at once. Even more than in Mo300, the aural effect is that of a short-range (phrasal) antiphonal voice-exchange.
5. The fifth motet, Balam inquit vaticinans/Balam inquit vaticinans/BALLAAM (Mo340; images here, here, and here)–like Mo275 and Mo300 but unlike Exaudi/Alme–uses the same tenor color throughout, which is repeated four times.[17] Here there is one tenor color for a melismatic introduction, the second for the texted part of the motet, and two further repetitions in a melismatic hocket at the end. The tenor color has 22 pitches but pitches 1–9 are identical in pitch and rhythm with pitches 10–18 and the upper voices directly exchange their musical material between these two sections, causing each color section to have an aab structure, not just in the tenor part, but in overall sound. In the texted section, the text is sung first in the triplum while the motetus has an untexted melisma, and then the voices swap over, giving a strong aural effect of antiphonal imitation. The upper-voice texts are identical except for the slightly longer final line in the triplum, which allows the motetus to catch up so that both upper voices sing their last three text syllables together.
6. Huic ut placuit, tres magi mistica/Huic ut placuit, tres magi mistica/HUIC MAGI (Mo341; images here, here, and here) is the inverse of the sectional situation in Mo275 in that it does have a terminal melismatic section, but lacks an initial melisma. Its 21-pitch color is internally organized into three seven-note tenor groups and comes twice in the texted part and twice more in the terminal hocket melismatic section. In each section, the color repeat involves an exact switching of the pitch content of the two upper voices. In the texted part, only one voice has the text at any given moment and it is the same text in both parts, rather like a larger-scale version of the effect in Mo339. The initial text line even does a similar internal troping of the tenor text as seen in a much more thoroughgoing way there.[18]
7. Stenzl’s first addition to Besseler’s list, Virginis eximie/Nostra salus oritur/CERNERE DIVINUM LUMEN GAUDETE FIDELES (Mo343; images here and here), has the three broad sections familiar from Exaudi melodiam/Alme Deus but they all have the same tenor color and each has it only once. As in Exaudi melodiam/Alme Deus, the first two sections rhythmicize their tenor pitches in the same way, here with a very short, three-note group coming 12 times. In the first section this is followed by a two note terminal unit so that the extensive 38-pitch color ends before the texting of the upper voices starts. In the second section, the tenor rhythmic units continue so that the third presentation of the color, which starts in the middle of the second, texted section, offsets the rhythmic presentation of the pitches as compared to the first two of their presentations. The second section thus has the complete tenor color 2 and the first 22 notes of color 3 now differently rhythmicized. The final section, the terminal melisma, completes the third color, with the tenor organized in three new five-note rhythmic groups. The two different texts of the upper voices are presented simultaneously without the isolated notes and melismas found in many of the other motets listed here. This seems a rather different kind of motet from any of the others in the original list.
8. Stenzl’s final addition is Quant se depart li jolis tans/Hé, cuer joli!/IN SECULUM (Mo324; images here, here, and here), a motet which lacks most of the features of the others and is, moreover, texted in French. It lacks any melismatic sections and has two tenor colores of 34 pitches, organized in 11 x 3 equal-note rhythmic groups separated by rests and then a terminal pitch (F). There is no sustained reflection of the color repeat in the upper voices, no antiphonal text or music effect, no voice exchange, and only very brief moments of hocket. It is unclear as to why this motet was added to Stenzl’s list.

As he goes through this list, Stenzl notes earlier discussions of this group, notably that by Handschin, who is mentioned as considering all of the six motets on Besseler’s original list as likely to be English.[19] Stenzl’s conclusion is that, given the company it keeps, his ‘Sanctus motet’ seems rather likely to be English. He notes furthermore that it seems to conform to several of Apfel’s ‘traits of Englishness’, namely no.20 (lack of tenor identification), no.22 (change of tenor cants firmus in the repeat, i.e. Paraphrase), no.51 (melodic repeat in the upper voices), no.52 (rhythmic imitation), no.53 (melodic imitation), no.56b (isoperiodism), but also cautions that the question must remain open.[20]

There are many things to question here. As will be clear from my brief comments on the list above, I am not convinced that all of the motets have more than a few elements in common and wouldn’t like to offer them as a strongly defined group given their mutual differences. Stenzl does set aside Mo339 and stresses that the question of origin remains open, but his case for associating these piece and their Englishness is nonetheless made, and no alternative interpretation is offered. For me the listing of features focuses too much on the structural and textural and not enough on the performed effect of the pieces, their texting, and their likely meanings; there are certainly some points of comparison and similarity, but the whole set of issues seems to need revisiting.

Stenzl was writing in 1970, when there was less self-conscious reflection on music-historical method and when the recent turn toward the need for an understanding historiographical details in order to strip away certain kinds of ‘master narrative’ from music history had not yet occurred. From a perspective in 2011, I find it surprising that motets can be considered ‘English’: most of them appear uniquely in a Parisian source (Montpellier); Mo275 is widely copied in Europe; and Mo275 and Mo340 and Mo341 appear in a fragment now in Oxford whose Englishness is asserted by RISM with no grounds given.[21] I’m not sure what ‘English’ means in itself, nor what implication it might have for how we are to understand these pieces of music. Are these motets ‘by’ English composers, or were they merely sung in England, or were they sung by English singers? And what does ‘English’ signify when a piece is clearly being transmitted in mainland Europe, sung by singers who are from various parts of Europe, and in a period before the establishment of the modern nation state? The identity of ‘Englishness’–given the government of the non-continental part of England by members of the French royal house and the presence of large parts of ‘England’ (if we call England that which is governed by the English King) on the continental mainland of what is now France–seems rather hard to define.

Stenzl’s dating of the motet, too, relies on assumptions that now might be open to historiographically motivated question. The motet’s closeness to the motets in the last two fascicles of Montpellier in terms of its notational style and its musical style (as posited in its membership of the group outlined above) might place it around 1300 (depending on the notoriously problematic dating of the last two fascicles of Montpellier). Stenzl claims, too, that Exaudi melodiam/Alme Deus is similar to the motets in the interpolated version of Fauvel, (which dates from the late thirteen-teens but whose pieces, especially the motets, draw on a repertoire from a little time before that date). Judging by its texts, the motet can be assumed to have had a liturgical use, but Stenzl, quoting Besseler, affirms that the group within which Exaudi melodiam/Alme Deus seems to belong marks it as a representative of a new kind of motet, which extended the old tradition of liturgical music so as to introduce a newer, more audience-focused ‘artwork’, paving the way for the individuality of the motets of Philippe de Vitry and the subtilitates artium.[22] Again, the notational and stylistic teleology on which this conclusion depends dates back at least to the early twentieth century and now needs revisiting.

Directions for further research

Several research questions arise from reviewing Stenzl’s article in the light of the identification of the same motet in the Dijon fragments. Stenzl’s focus was on then-traditional questions of date and provenance, although it is unclear whether he considers the latter to be defined by the geographical place where the composer composed the piece or the national origins of the composer; what is clear is the emphasis on a composer rather than the diffusion of the motet as a performative item, meaningful to singers and listeners. The relatively high textual stability of the ‘work’ between the two surviving copies might point to assumptions about the motet’s ‘composerliness’ as a reasonable starting point, but consideration of the reception of this piece would nevertheless aid any understanding of its place in music history.

One might now ask questions about the wider diffusion and use of this motet. The Dijon fragment seems to emanate from Cîteaux so can’t readily be associated with the Franciscan practice of the StM book. The other music in the leaves from Cîteaux that feature in Haggh’s article are of a similar date and have been shown to have concordances with hockets quoted in a thirteenth-century theory source.[23] Perhaps this motet, too, can be associated with the thirteenth-century growth of Cistercian foundations in Europe.

In terms of date, the Dijon manuscript clearly shows an earlier notational style, but this doesn’t necessarily mean that it transmits an earlier copy of the piece; assessing that would rely on dating the fragmentary leaves themselves. The easy teleological progress of both notational and musical styles seems difficult to rely upon, as questionable in its older musicological values as the easy positing of centre and periphery. A more nuanced music-analytical study of motets might go some way to refining the grouping provided by Besseler and Stenzl, especially if its taxonomy were to rely less on listable structural features of pitch and rhythm and more clearly on the kind of integrated musico-textual analyses of motets that have recently begun to appear.[24] Certainly much work remains to be done in a newer musicological climate with an emphasis on singers, receiving communities, use, and meaning.

Appendix: variant readings

Section A

Triplum StM no ligatures; triple ‘Sanctus’ invocations starting in bars 1, 9.2, 17.

Motetus StM no ligatures; triple ‘Sanctus’ invocations starting in bars 1, 9, 19. bb.22–4: Dij missing (cut away)

Tenor bb. 21–2: StM 2 notes

Section B

Triplum b.4: StM no lig b.10–11: StM no lig b.14: StM no lig b.16: StM no lig b.22: StM no lig b.26: StM no lig b.32: StM no lig b.36: StM no lig b.41: StM no flat b.42: StM no lig b.44: StM lig only for notes 1–2 b.50: StM lig only for notes 3–4 b.53: StM lig only for notes 1–2 b.54: StM lig only for notes 1–2 b.56: StM lig only for notes 1–2 b.58: StM lig only for notes 1–2

Motetus b.2: StM no lig bb. 7–8: Dij missing (cut away) b.17: StM g-d-g [as in color 2] b.18: StM lig 1–2 bb. 18.1–2: Dij missing (cut away) b.20: StM d-e b.24: StM lig only for notes 1–2 bb. 28.3–29: Dij missing (cut away) b.31: StM no lig b.37: StM lig only for notes 1–2 b.37.2: StM e b.40: StM lig only for notes 1–2 bb. 41–44: Missing in Dij b.45: StM lig only for notes 1–2 b.46: StM lig only for notes 1–2 b.51–2: StM no lig b.56: StM no lig b.57: StM no lig bb. 57.3–58.3: Dij missing (cut away)

Tenor bb.1–2: StM 2 longs bb.9–10: StM 2 longs bb.17–18: StM 2 longs bb.21–22: StM 2 longs bb.27–28: Dij missing (cut away) bb.29–30: StM 2 longs bb.37–38: StM 2 longs bb.41–42: StM 2 longs [missing in Dij but assumed to be maxima by comparison with other colores] bb.41–48: Dij missing (cut away) bb.49–50: StM 2 longs bb.57–58: StM 2 longs

Section C

Triplum StM no ligatures

Motetus StM no ligatures bb.16–18 (up to the rest) and 38–42 (up to the rest): Dij missing (cut away)

Tenor bb.6–7: StM 2 longs bb.28–29: StM 2 longs bb.30–31: StM 2 longs b.26: Dij b erroneously a maxima bb.45–48: Dij missing (cut away)