II The simple polyphony in Amiens 162
A part of the repertory of Amiens 162 belongs to the type that has been characterized as cantus planus binatim by Alberto Gallo (Gallo 1966, 1989 and 1984, p. 304). His terminology was inspired by a remark by Prosdoscimus de Beldemandis in his Expositiones tractatus pratice ... of c. 1404, and it describes according to Gallo a polyphony consisting of unmeasured plainchant that is “shadowed” by an extra voice, whose function it is to enhance the sound and solemnity of the chant. The term “cantus planus binatim” has been criticized by Margaret Bent for being too exclusive, not including music just as simple, which has mensural elements in its notation or is right-out mensural (Bent 1989), and Christian Berktold has challenged that Prosdoscimus was at all describing this sort of singing (Berktold 2001). I prefer the term “simple polyphony”, which includes all the songs copied with an intention of practical use in Amiens 162, measured or unmeasured, for two to four voices.
I understand this term as a neutral label, which underscores the music’s technical simplicity in comparison with contemporary complex polyphony, not necessarily ‘retrospective’ and certainly not ‘primitive’. (1) As we have seen in the preceding chapter, what Caulaincourt was looking for was contemporary music of a specialized sort, and that he needed help from an expert to succeed in finding it. Its technical basis is traditional, founded in musical skills for improvising a solemn sound for certain services, and it reaches back to the oldest traces of polyphony; it has to be regarded on the background of improvised polyphony and the different traditions of this art flourishing through the centuries. However, in the world of sacred music, many participants in the services were able only to read and sing the monophonic chants of the liturgy. Therefore, also simple polyphony had to be circulated in a notation, which was readable to everyone, as the numerous and geographical widespread sources testify. In this way the ‘simple polyphony’ can be regarded as a sort of counterpart to ‘plain chant’.
Setting aside all the terminological differences, the literature gives us an overview of the geographical spread and composition of this repertory supplemented by analyses of selected segments. A great number of the sources are catalogued in the volumes of Répertoire International des Sources Musicales (series B IV, vol. 3-5). The German-speaking regions, France, Hungary and Scandinavia were surveyed in the classic studies by Arnold Geering and Theodor Göllner, and by many others (Geering 1952 and 1961, Göllner 1961 1969, Stenzl 1972; Göllner 1993, Bergsagel 1989, Rajeczky 1961). The Italian repertory has been the theme of several conferences and collective volumes (Corsi & Petrobelli 1989, Dalla Vecchia 1996, Cattin & Gallo 2002); and simple polyphony has also been highlighted as separate themes in conferences of broader aims (Husmann 1966, Stenzl 1980, Meyer 1993). All of these publications open the doors to a wealth of other bibliographical references.
The two-part settings in Amiens 162 are all in strict note-against-note style. The surveys of the German repertory by Geering and Göllner conclude that the settings of this type largely conform to the rules laid down in the organum treatises, which can be dated around 1100, but which obviously reflect older performance traditions. (2) Göllner based his analysis on two manuscripts, whose main sections can be dated in the first half of the 15th century. He found that these songs in general showed the following characteristics (Göllner 1961, pp. 40-59):
The liturgical tune is enhanced by a counter voice in principally the same range; the restricted range of the setting forces many voice crossings and a preference for contrary motion.
Every musical unit begins and ends on a perfect concord, octave or fifth preferably; fifths above or below the tune are the preferred concords, and concords of thirds mostly appear in passing when the voices cross on a prime concord (in the typical sequence of intervals 5-3-1-3-5).
The melodic figures of the counter voice are determined by the vocabulary of the plainchant tunes; in some cases the melodic shape of the voice takes precedence over the strict application of consonance rules.
The datings of the sources analysed by Geering represent a longer span of time, from the 13th to the 16th centuries. We can extract the following observations to complement the description (Geering 1952, pp. 44-51):
Settings in strict parallel motion are unknown in the notated repertory. Geering has found only one setting with moments of parallel fourths; the fifth is the favourite concord, more thirds and sixths crop up in the younger repertory, and among the youngest we can find songs with parallel motion only in thirds and sixths, and parallel fifths are excluded.
Predominant contrary motion belongs to the settings copied during the 15th century; in the earlier centuries such settings are not common.
Geering found 20 three-part songs in his repertory, most of them from the 15th-16th century.
Ulrike Hascher-Burger has looked at this type of music from a different angle. (3) She has analysed two Dutch song collections dating from the end of the 15th century, that is, more or less contemporary with the repertory of Amiens 162, and both of them were written in houses belonging to the reform movement Devotio moderna. The male and female communities of the Devotio moderna were in spiritual outlook probably far removed from the conservative monks of Corbie who fought for the preservation of their feudal rights, but their musical outlook reflect similar monastic traditions. Hascher-Burger divides the two-part songs in five groups, of which we can disregard the fifth group, songs with traces of art music:
1) Songs in predominant contrary motion with many perfect concords,
2) with parallel motion primarily in fifths,
3) with more dissonances than average, including fourths, and
4) with many imperfect concords, thirds and sixths.
This is, of course, a quite rough classification, as all the songs show characteristics of more than one group. The differences in the numbers belonging to each group that she has found in the two manuscripts are remarkable. In MS Berlin, Staatsbibliothek Preussischer Kulturbesitz, oct. 190, a songbook from a male convent in Utrecht, Hascher-Burger counts 19, 10, 5 and 7 respectively in the four groups, while the other manuscript from a female house in Zwolle, Utrecht, Universiteitsbibliotheek, MS 16 H 34, has 4, 1, 4 and 23 respectively in the same four groups. This is a near reversal of the counts in the two songbooks. Hascher-Burger explains the differences in style by pointing at the influences of different localities and traditions, of levels in education and of gender.
It is of great importance to our project that Hascher-Burger documents the existence of widely differing styles and aesthetics in contemporary sources of simple polyphony. That the repertory incorporated the solemn sound of parallel perfect concords and the often quite disjunct motion of the counter voice as well as the smooth enjoyment of parallel thirds and sixths. This wider stylistic spectrum puts a new perspective on the description given in Geering’s survey. The differences seem less to depend on different stages of musical ‘evolution’ than they signal differences in localities and contexts of tradition, and as we shall see in the Amiens 162 repertory the different stylistic models may co-exist within extended settings.
How do we place the two-part songs in Amiens 162 in relation to these analyses of contemporary repertories? Most of them are extended, multi-section settings, and they comprise one song in the group of funeral compositions, four sequences and a trope, which all belong to the repertory collected by Hand A, and finally a Credo-setting, which Caulaincourt (Hand C) later added.
Already in the first two-part song in Amiens 162 we meet an aspect of simple polyphony not stressed in the research just mentioned, namely the evidence of a clear plan for the course of the setting. A ‘compositional intention’ is discernible in the three stanzas of the “Libera me”-trope »Quando deus filius virginis« on ff. 13v-16. It opens inconspicuously with the voices circling each other within the range of the fifth d-a. The peak of the tenor tune on d’ is reached at the start of the fourth line, while the highpoint of the contra voice comes at the end of the fifth line on e’ with a tenth between the voices. It happens in the repeat of the tune’s b-element, which was first heard with the second line of text (cf. ex. 5a-b). This leads to a full repeat of the tune’s a- and b-elements with the contra voice in high range singing the refrain-like lines starting with the exclamation “O, O, O” and emphasized by fermatas. This development towards the climax was prepared in lines three and four (the c- and d-elements), which both end on a concord of a third, f-a (in stanzas 2 and 3, which give us the original version, stanza 1 has been revised), and thus evade cadential points of rest and keep up the flow. The repeats of the b-element, which appears as lines two, five and seven, demonstrates the planned development. The all begin in an identical way, but in the last two lines a fifth concord is simply changed into a sixth resulting in a higher curve for the rest of the line.
Example 5a-b, “Quando deus filius virginis”, second stanza (ff. 14v-15), notes 15-25 (line 2) and notes 53-68 (line 5 and beginning of line 6)
The contra voice is mainly in contrary motion and avoids parallel perfect concords entirely in the original version; parallels are in thirds and sixths, and mostly only two in a row. When the tune is in the low end of its range, the contra is above it, and when high, it goes below – up to an octave below. The composer appreciated the change of sonority from fifth to third on repeated notes, and for the “O, O, O”-passage’s repeated notes he used in the contra the in improvised polyphony ‘safe’ progression of third, fifth and sixth (ex. 5b). This song represents a ‘modern’ type of contrary motion two-part note-against-note setting and is a proof of Hand A’s ability to procure the newest and most effective music of this sort for his collection of funeral music.
The short trope »Sospitati dedit egros« (ff. 41v-42) exhibits another type of ‘modern’ simple polyphony. Its first line opens in contrary motion, but then it changes to parallel motion in mainly thirds, four in a row. The second line starts in a similar way, progresses in parallel fourths and thirds, but ends in parallel fifths, six in a row (ex. 6)! The next line is nearly consistently in contrary motion with the contra above the tenor, while it in the fourth line, which repeats the tune, goes below the tenor and uses some parallel thirds and fifths. In the last two lines where the tenor is high, the contra follows it in parallel octaves below and ends by shadowing it in fifths. This piece of simple polyphony is clearly more influenced by parallel motion than “Quando deus”, and its composer obviously accepts parallel motion in perfect concords as well as in imperfect concords as equal means to vary the sound and to direct the course of the counter voice. The very compact, sequence-like setting demonstrates the way in which the multi-section tune in double versicles develops and intensifies its melody by introducing a contrasting semitone (in the fifth line “Vas in mari ...”) or by changing into a higher range (in the seventh line “Ergo laudes ...”. The composer of the counter voice has taken great care to vary the sound by using its full range in the first two double and varied versicles (lines 1-4), in which the tune is a bit restricted in expression. This is not needed in the last two pairs (lines 5-8), where the tune attracts more attention, so they are just identical repeats.
Example 6, “Sospitati dedit egros” (ff. 41v-42), notes 1-30 (lines 1-2).
This interpretation of a traditional technique, which we safely can assume was ‘modern’ in Caulaincourt’s circles, characterizes the sequence settings collected by Hand A. For the long compositions the whole spectrum of techniques are used to keep the music alive – in varying selections. »Virgini Marie laudes« (ff. 28v-30) is a contrafactum of “Victime pascali laudes” and has the well-known tune in its tenor. The counter voice follows it mainly in thirds while constant crossing above and below it; parallel fifths and octaves appear when needed. The setting is smooth in sound and the contra is easy to sing. The original version has very little variation in the double versicles; in the long third section the internal cadences are reversed, from first unison then fifth to fifth-unison – simple, and quite effective. Caulaincourt has later introduced a bit more variation by changing a third into a fifth sonority in section 2b (at “Mors et vita”). This was a conscious choice, because when the exactly same melodic figure returns in section 4b, he made the same change (at “Scimus Christum”). In this way he made the sound of the second versicle in sections 2 and 4 slightly different from the first ones. »Veni sancte spiritus« (ff. 35v-37) is a similar setting of a well-known tune, tending a bit more towards contrary motion, but also containing long rows of thirds. As previous mentioned this was probably the first long composition, which Caulaincourt had the sole responsibility of entering in the manuscript, and he tried his hand in revising the music. Originally this setting probably was completely regular with identical settings of the double versicles.
One of the most interesting creations in this genre is the sequence »Veneremur virginem« (ff. 37v-41), which as earlier discussed probably baffled Caulaincourt in its use of key signatures – he did not touch it later. It is a combination of two different Marian sequences: The Mixolydian “Veneremur virginem” is set complete in five stanzas (1-4 and 6), but before the last a stanza from the more widespread Dorian sequence “Hodierne lux diei” has been interpolated. It is the last stanza of “Hodierne lux diei”, beginning “Salve splendor firmamenti”, which in its concluding praise of Maria fits perfectly into the meaning of “Veneremur virginem”, and it creates a nice contrast to the longer setting.
The setting of the five stanzas of “Veneremur virginem” is in itself a well-formed and consistent sequence in the ‘modern’ style (ex. 7a). It involves much contrary motion, but every phrase tends to contain also shorter or longer parallel passages, and it avoids any use of parallel perfect concords, and most versicles use internal repeats. It is euphonious, placed within the Hypomixolydian range d-f’, and variation in sound is obtained by letting the stanzas 1-2, 4 and 6 be dominated by the sound of parallel thirds with the parts quite close together, while stanza 3 stands out by its parallel sixths and up to a tenth between the parts, and in stanza 4 its pace is slowed down a bit by the introduction of double values and ligatures.
Stanza 5 uses a probably local version of the tune for “Hodierne lux diei”, with its range compressed from a ninth to a sixth, but it is still recognizable. The setting differs audibly from “Veneremur virginem”. Its range is higher, f-g’, it demands a key signature of one flat, and the two first line endings involve parallel fifths (ex. 7b); and its three-line versicles and different rime pattern contrast the steady flow of the four-line versicles of “Veneremur virginem”. Coming from the very stable sound universe of “Veneremur”, this set of versicles stand out as if belonging to a different tradition. However, it is not too far away: The two identical versicles also repeat most of their first line, and they end in five concords of thirds in a row before the final fifth!
Example 7a-b, “Veneremur virginem” (ff. 37v-41), notes 1-17 (start of versicle 1a) and notes 264-284 (start of versicle 5a).
It is impossible to know if this stanza was lifted from a complete setting of “Hodierne lux diei” or was made for insertion into “Veneremur” – with the traditional tune modified for the occasion. As it stands in Amiens 162, the setting challenges us to consider that the composer had a clear idea of the different styles in simple two-part polyphony and used them to create some contrast and a culmination in an otherwise slightly monotonous sequence, and that the aesthetic deliberations put into the work of creating simple polyphony followed the same paths as in the working out of mensural art music. Or, maybe our aesthetic appreciation of his choices is irrelevant, because the extra stanza was inserted for textual reasons, and its musical effect was purely accidental. The second alternative is, of course, possible, but appears less probable.
Hand A had envisioned the long »Stabat mater dolorosa« (ff. 30v-35) as the opening piece of his collection of sequences. Its ten double versicles set an otherwise unknown tune in a style similar to »Virgini Marie laudes«, but with a more disjunct counter voice. The majority of its double versicles are varied to some degree, and during its course it displays the full range of possibilities of the genre in contrary and parallel motion in perfect and imperfect concords. Originally, parallel fourths and unisons between the parts appeared in several places – most of them were since eliminated in the revision. In this way all four types found by Hascher-Burger in shorter contemporary songs are present in one or more of its sections. The setting has been carefully phrased with the help of vertical strokes and fermatas, and rows of fermatas are found on important words. It ends with a long “Amen”, which appears to be freely composed with the two voices conceived simultaneously in more or less strict contrary motion.
Caulaincourt has put some work into a revision of the counter voice. He has erased and rewritten notes in nearly all the sections, and he has drawn his own custos on top of those of Hand A at the end of all staves. All the erasures, which I have been able to discern on the pages, are shown in ‘Erasures in Stabat mater dolorosa’, which is appended to the edition (no. 7b). There we can compare the black notes, the erased ones, with his corrections, the white notes. His primary goal seems to have been the elimination of what he regarded as errors in the counter voice and in some passages to shift it into a lower range. For example, he has removed unisons in sections 3a-b and 6a-b; and in section 9a, the strict contrary motion at “inebriari” ended in the progression seventh-none – he has replaced it with the parallel motion of section 9b. In section 5a the counter voice started on a fourth concord and had a short unison passage near the end, and it was mostly above the tenor; all this has been changed into a counter voice mostly below the tenor. In 5b he has changed the beginning into a variant of the low version, but has retained the remainder of the high one, only correcting parallel fourths at “sibi” into contrary motion. Here he has created a greater difference of sound between the two halves of the stanza than before. He did not appreciate concords of fourths. In sections 10a-b, he has changed fourths into unisons, but in 6b he left the fourth on “pati”, and erased the corresponding third on “plagas” in 6a, and put in a fourth instead in order to make the two halves identical. Possibly the words “plagues” and “suffer” justified the fourths.
It is not easy to find a principle for the changes not provoked by ‘errors’. In sections 1 and 5 he has introduced new differences between the double versicles, while in sections 3 and 6-9 he has attempted to make them more similar. Probably the notes as they stand were not Caulaincourt’s final word on the counter voice. Above it he has on all the right hand pages written “vacat” and drawn a vertical line in the left margins. Maybe he had the amended piece performed and decided that the song would make a better figure without the counter voice, and therefore discarded it from further deliberations by deeming it “vacat” (it is useless!). If we compare the varied and quite engaging “Stabat mater” tune in the tenor when sung alone with impact of the two-part setting, where the counter voice tends to blur the contours of the melody, we may easily be convinced to agree with him.
A piece that Caulaincourt added after he had returned to Corbie, is the two-part »Credo in unum deum« (ff. 121v-124); the exemplar for it may have been lent to him by somebody passing by, or he may have found it in Amiens. It sounds old-fashioned compared with the other two-part simple polyphony songs. Parallel fifths and octaves dominate its sound, and we even find a passage in parallel fourths (at “Crucifixus”) and quite a few single concords of fourths. The setting is like the “Stabat mater” phrased by vertical strokes and fermatas, and important passages like “Et homo factus est” and “Et vitam venturi seculi. Amen” are emphasized with fermatas above all notes. Also this song has been extensively revised by Caulaincourt, and the erased notes tell a curious story.
His exemplar was presumably not of the highest quality – that he suddenly should have become quite sloppy in his copying seems unlikely in the light of his careful writing. The version originally copied had passages where strong dissonances appeared, and somewhere in the process the positions of the voices had been reversed, so that the tenor carrying the pre-existent tune stands on the right hand pages in Amiens 162. Normally, the left side of the openings are reserved for this voice. This reversal was of consequence for his corrections of the faulty passages.
It is a setting of the “Credo cardinalis” tune, which probably originated in France around 1300, and during the next centuries it became widely circulated all over Europe, always in rhythmically differentiated notation (canto fratto), and in several sources in two-part polyphony. (4) Also in Amiens 162 it is notated in square and rhomboid notes that permit a rhythmical interpretation. The counter voice, however, is notated in undifferentiated square notation, and everything point at that the rhythmical elements were not appreciated at Corbie. The tenor tune, as originally copied, was identical with the in the 15th century well-known Credo-tune with a few lesser variants. (5) Only three passages, “vivos et mortuos” (notes 257-262), “Et in spiritum … ex patre” (272-294) and “Amen” (398-401), are different.
The two-part structure was remarkably dissonant in two places, that is at “Qui ex patre ... procedit” (notes 289-301) and “Et vitam venturi seculi” (389-397). Here the many seconds and fourths between the voices give the impression that the counter voice was not at all constructed with the tenor tune in mind (cf. the “Erasures in Credo in unum deum”, edition no. 11b, and Example 1a-b in the comments). They called for Caulaincourt’s intervention. In fact, the Credo-tune was easy to recognize in these passages, but he chose to erase it and compose a new tenor instead of changing the counter voice. He placed it in a higher range and avoided all dissonances. For example, in “Et vitam venturi” the concords were 5-4(3)-2-3-4-4-2(3)-2-5, which he changed into 8-8-5-5-6-8-8-8-3. The majority of his other changes concern avoidance of fourths at line endings; he has three times changed the progression 4-4 into 3-5 by reversing the tenor notes, but still he left many other fourths untouched. Also, he did not like the ending in parallel thirds in the fermata passage “Et homo factus est” (167-172); this was replaced by parallel fifths, by changing both voices.
It is quite obvious that he was convinced that the voice part placed on the left-hand pages was the main voice, and consequently he made by far the greatest number of changes in the voice carrying the Credo cardinalis. Furthermore, these changes helped to pull the tune even stronger towards the Dorian central tones D and A, and thus the tenor became further and further removed from the well-known tune.
It seems amazing that he did not recognize the Credo tune. But it must be so. A good example can be found at the word “Visibilium” (note 31), where the Credo-tune on “-um” has the note b, which then is repeated for “omnium”. The original had the perfectly consonant sixth d-b here, which progressed to the fifth e-b. But the sixth was transformed into a fifth d-a by erasure and rewriting in the tenor, and thereby a distinguishing feature of the tune was lost and only more parallel fifths gained.
It is, of course, thinkable that the tune was not much in use in the Benedictine monastery. But it is more credible that the setting already appeared somewhat corrupted in Caulaincourt’s exemplar, with the placement of the voices reversed and with many wrong notes, which may have deceived him into regarding the tenor tune as the variable counter voice. In any case, he succeeded in producing a usable Credo without any disturbing dissonances, which was solemn in its old-fashioned sound of fifths and octaves.
The two-part settings of Amiens 162 show quite a wide stylistic spectrum. From the Credo added by Caulaincourt to the most modern settings collected by Hand A. In some cases we find an awareness of the structuring of sound and development in the settings, which points away from the spontaneity of improvised polyphony and from its monotony of sound. At the same time we see a tendency to let parallel thirds and sixths dominate the sound picture at the expense of fifths and octaves – they are still present, but not as prominent. And they sound different from the more traditionally oriented settings. Simple two-part polyphony lived on for centuries, but when we browse the sources from the 17th-19th centuries we find mostly settings whose parallelism is based on imperfect consonances. (6) Not much research has been done on this change into a preference for third-sixth sonorities; the repertory of Amiens 162 may be an indicator of what was happening.
1) I cannot agree with Thomas Schmidt-Beste who wrote that these repertories were “formerly given the pejorative label ‘simple polyphony’” (Schmidt-Beste 2011, p. 156).
2) For example, the treatise known as Ad organum faciendum and the De musica by Johannes, see further Eggebrecht 1984, pp. 40 ff.
3) Hascher-Burger 2008. I am grateful to Dr. Ulrike Hascher-Burger for lending me a copy of her article. On this repertory see also Blachly 2003.
4) Gozzi 2006 and Sherr 1992, pp. 183 ff; a modern version of the chant is Credo IV in GR, p. 67*.
5) Cf. Sherr 1992, pp. 186-187, which reproduces the version of the Giunta Gradual, Venice 1499-1500.
6) See, for example, the articles by Annunziato Pugliese, Cesare Ruini and Antonio Lovato in Cattin & Gallo 2002.