4. PadB: Padua, Biblioteca Universitaria MS 1115

PadB comprises a single parchment bifolio of secular polyphony used as front flyleaves for a 15th century manuscript, I-Pu 1115. This manuscript, rebound in modern covers, contains the Sermones of Hieronymus and treatises on morality.97 The two folios contain the complete music for three polyphonic vocal compositions and single parts for three additional pieces. There are three pieces with Italian texts and two with French texts, with one, Ay si, whose language is unclear. PadB testifies that musicians in Padua at this time were interested in music from other parts of Northern Italy, France, and very probably Sicily.

The first folio in the present ordering, designated folio A, bears on its recto side the call number of the manuscript from a catalogue of the manuscripts and books in the library of Saint Giustina in Padua compiled from 1453-84.98 The call number, found in the top and bottom margin of the page, is "yy.2.n.23" with "AC 3" added as an additional designation in the bottom margin. In the space below the final stave, probably at the same time as the call number was added, judging by the handwriting style, is an index of the works of the manuscript as it currently survives. The index is signed by Bachinus, who also indexed I-Pu 658 and I-Pu 684 and probably prepared the library catalogue. The cover of the manuscript has been replaced by modern cardboard and the spine of the manuscript has become detached from the end gatherings of the manuscript, allowing easy examination of the gathering structure. There are thirteen gatherings in the main manuscript. The manuscript lacks a consistent layout, suggesting it was the work of several scribes.

The two folios of music give no indication as to their original foliation. The current foliation found in the upper-right corner of the versos is A and B. Because the majority of polyphonic music manuscripts of this time bear some sort of original foliation, it is possible that the folios have been trimmed along their outer edges to the width of the main corpus of the manuscript, losing this valuable information. The top and bottom edges of the folios have probably not been trimmed since the fragment is smaller than the parchment of the rest of the manuscript. Folio A is 23.0 cm in width and 31.6 cm in height. Staves measure 1.4 cm across the five lines.

Each page is ruled with 10 five-line staves. The staff lines begin at nearly the same distance from the left margin on every staff of each page. The exception to this is found on folio BR where the first staff has been indented to allow room for the large initial letter E, which was never added. There are also slight variations which indicate the left margin was not carefully laid out. For example, the tenth stave on folio AR begins several millimeters farther in than the other staves on the page.

The staff lines on the right sides of the first folio end at various distances from the right margin. The sixth staves of folio AR and AV, for example, end farther from the right than any of the other staves on their respective pages. The composition at the top of f. AR, Se per dureça, ends on the sixth line of the manuscript. It is possible that the scribe was more careless with this staff, knowing that it would not be used completely. The same argument can be made for the similarly shortened sixth staff of the verso of folio A. However, in this case, the scribe of the music has managed to fit all the music onto the top five staves though he had to extend the fifth staff into the margin when he added the music.

Each of the four pages in this manuscript has a different musical arrangement. Because the arrangement of musical voices is important to the discussion of Aler m’en veus, a staff by staff description of the manuscript follows:

Two of the pages, f. AR and f. BV have a single complete composition at the head of the page, with a single part of a second piece at the bottom of the page. Folio BR contains a single composition which takes up all but the final two staves of the page. Except for some of the shortest tenors, a voice from another piece could not fit in the space at the bottom of BR. Folio AV is the only page not to contain a complete composition and is the only page to leave more than two staves blank. Explaining this layout will only be possible after a detailed look at the style, notation, and paleography of each of the compositions in the fragment.

The gathering structure is as follows:

1. Se per dureça (anonymous)

Folio AR begins with the two voice anonymous ballata Se per dureça. This unicum composition has most of the features of a siciliana, a compositional style involving the recasting of text and music of Sicilian (or southern Italian) unwritten traditions into the more conventional Tuscan and Northern Italian form of the ballata.99 These pieces, many of which can be found in the fourth gathering of the old section of the Reina Codex, were probably composed based on second or third hand knowledge of the Sicilian songs.100 Although I cannot comment on the provenance of the text, the prominent use of parallel octaves and fifths, especially in the beginnings and endings of phrases, the varied repetition of material, and the use of short phrases separated by rests points toward a connection to this tradition.101 The simultaneous declamation of text at the beginnings of phrases, though shorter than in the examples Pirrotta gives, is also characteristic of music of this tradition.

The music for Se per dureça is written in Italian notation with puncti divisionis but without divisio letters. The composition uses semibreves maior and minor. The maior type of semibreve is twice as long as the minor. Under standard Marchettan notational rules, when two semibreves are used to fill a tempus which takes six minims (three beats), the "natural way," via naturae, to sing this rhythm is minor then maior:

· · =

However, a downward pointing tail or cauda can be added to the first semibreve indicating that the via naturae should not be used and instead the "artificial," via artis, rhythm of maior followed by minor should be used:

· · =

The scribe of the tenor voice of the ballata places an unnecessary cauda on a semibreve in the antepenultimate tempus of Se per dureça:

· · · · = | |

The line would have the same rhythm even if the cauda were not present. Choices like this can been seen as evidence of a scribe unfamiliar with the notational system. But because this is an isolated violation of Marchettus’ rules, I wish to look for other reasons why the scribe could have chosen to write this cauda. One compelling explanation is that in the previous four instances of two semibreves in a divisio, extending back to the second line of text, are all to be sung via artis. Thus, the scribe could have added the cauda as a warning to the singer that the rhythm of the measure was different than what the singer might expect given what came before.

Since it is notated in Italian notation, Se per dureça has minims of fixed duration. The ballata does not use any semiminims but is free in the superius part with its use of triplet minims, often directly before or after duple minims and once, on "uero" in the superius part, syncopated by a duple minim rest (m. 17). The flags on the triplet minims hang to the right in this piece. Kurt von Fischer considers this less usual in Northern Italian manuscripts. In his Studien, he writes that the left-hand tail on the triplet is a northern idiosyncrasy while the right-hand tail is more often seen in Tuscan manuscripts.102 In the superius, the scribe often grouped the normal minims in groups of three (e.g., mm. 2 and 14) temporarily giving a triple, senaria perfecta () feel to a piece which is otherwise senaria imperfecta ().

While the positioning of the puncti divisionis makes clear that the meter is senaria imperfecta, this piece, like all of the pieces in PadB, lacks divisio letters. These indications became less common as the fourteenth century gave way to the fifteenth. The scarcity of divisionis indicators may be owing to the decline in prominence of the madrigal after the 1360s, a poetic form that when set to music often had a change in meter before the ritornello.103 Without this internal change in time signature letter indications became less necessary. Pieces with complex changes of meter in the late fourteenth and early fifteenth centuries were more often written in French notation with changing prolation signatures and coloration.

The music on the first staff, though not the staff itself, is indented, to leave room for a large S. No initial capitals are found in PadB. Instead, there is a small S inked into the margin, probably as a guide for the illuminator.

It is most likely that a single scribe copied the entire ballata and the text. This scribe will be designated Scribe A. The width of Scribe A’s strokes are generally uniform. The spacing is tight between minims within a group; between a group of minims and the notes which precede and follow, though, there is usually a small gap. However, the third and fourth custos on the page seem to have fatter front sides than the first two do. The writing of the capital letters is very similar between the superius and tenor voices. The scribe places light double-slashes at the ends of lines 2, 3, and 4 of the text in the cantus. The scribe only uses such marks in the tenor at the end of the ballata. After the first three lines, he uses a punctus in the text to divide the lines, a mark used at the end of the first line of text by the scribe of the top voice. In the music, the four lines are divided with vertical lines running through the staff. Again, there is an inconsistency in usage between the voices. The upper part uses a single vertical line at the end of the third line of text and thin double vertical lines at the end of lines 1, 2, and 4. The tenor part uses these same lines, but in exact opposite order: single lines appear at the ends of lines 1, 2, and 4 while double lines are present only after the third line.

The paleography of the text gives some difficulties. The scribe of Se per dureça writes, with a few exceptions, letters that are unornamented. The scribe gives the "h" of "chel" on the second staff an ornamented stem which is not seen in the corresponding text in the tenor voice. In the tenor, the letter "d" of dureça has a similarly ornamented stem. The fanciest and most seemingly out of place "d" comes at the end of the text in both voices on the syllable "da." Here the top of the "d" extends leftward nearly horizontally before returning to the right in an exaggerated flourish. The flair of the "d" may be because it is at the end of the voice and can be seen as similar to the decorated double-longs with which scribes ended pieces. Inconsistencies within the writing of a single scribe such as these make scribal identification in general difficult.

Stylistically, Se per dureça shows an interesting recurring voice exchange in measures 5, 17, 29, and 39; occurring once per phrase near the middle of the phrase, this gesture, along with the descending three minims in the second measure of each phrase unifies the melodic material of the composition. The rhythm of a three triplet minims syncopated by a normal minim rest (m. 17) is also unusual and not seen anywhere else in the fragments examined. Two of the cadence formulas in this ballata are rather normal and made by condensing a 3rd to a unison (third phrase, mm. 32-33; final phrase, mm. 43-44), or by expanding a 3rd to a 5th (first phrase, mm. 10-11), while a third cadence type by parallel motion to a unison after the condensation of a 3rd to a unison (second phrase, mm. 22-23). The ballata begins on a C-G fifth and cadences on A, C, E, and D, making it nearly impossible to speak about modality in this piece.

Although as said above, I cannot comment about the provenance of the text, I have presented it below, with all variations and some general observations in hopes of aiding others in the classification of this ballata.

Se per dureça tu morir me fay
Dona da cui merçe104 trouero may
E me credua chel mio gran fer uire
Ta uese uenta se may fusti105 cruda.

The rhyme scheme is aabc.106 The ballata may be classified as a ballata minore since it has a two-line ripresa (the first two lines).107 The ballata is also missing the text for the second piede and the volta (which would have the same music as the ripresa). Each line of text can be read as having eleven syllables, though it is possible that by reading "mio" in the first half of the first piede and "may" in the second half of the first piede as having two syllables each (as indeed one must when reading "may" at the end of the second line) the piedi can be seen as rarer, but still known, twelve syllable lines. When performed, the first two lines of text with their accompanying music would be repeated.

transcription: [page 1] [page 2]

2. Ay si (anonymous)

The second piece on folio AR which has the incipit Ay si, is of a very different style than Se per dureça. It is a composition in two parts, the first of which is definitely repeated, since it has both open and close endings (first and second endings in modern parlance). The second part is designated "Secunda pars."109 It is the only composition in this study to use both void notes and red notes, and it uses both types. Because only the contratenor survives, it is difficult to discern much about the over-all style of the composition. The incipit suggests that it was a French texted piece, possibly a virelai, though the incipit could also be Italian (a variant of "Ahi si"). This piece may have been an ars subtilior composition, especially if the cantus voice was more active and had rhythmic cross-relations (duple vs. triple) with the other voices.

Ay si begins in duple time, expressed through the use of red, void notes. The note values used are entirely breves and semibreve notes and rests. These void notes are often combined into ligatures the most common of which is the two-note ligature cum opposita proprietate yielding two semibreves. At the end of the first line of music, the notation switches to black notation with the heretofore unseen minim suddenly becoming the dominant note value. The relationship between the value of the void notes and the value of the black notes can only be conjectured, since the usage of void notes varied from composition to composition in this time. A constant semibreve ( = ) can be ruled out, since otherwise there would be no need for the change in notation. The most logical relationship between the two note values would seem to be a 3:2 ratio ( = ). However, if this were the case, there would be no distinction between the red, void notation used at the beginning of the composition and the solid red notation used later in the piece. While there are some pieces which use two different symbols to mean the same interpretation, I am reluctant to apply that here without more evidence. The next most probable interpretation, and the one which I have used in the transcription is a 2:1 ratio between the void semibreve before and the solid semibreve later (implying a 2:3 ratio between the void semibreve and the solid minim).110 As mentioned above, solid red notes are used to indicate a 2:3 ratio with the black semibreve. The six notes beginning the second part of the composition are all red and occupy the space normally filled by four semibreves. More interesting is the use of a single red note in a c.o.p. ligature in m. 42. A punctus before the minim rest prior to the ligature indicates that the rest along with the ligature must be fitted into a single breve. Thus the first note of the ligature is perfect while the second is imperfect. When this pattern reoccurs at mm. 56 and 75, the second note of the ligature is not red. It is an interesting dilemma which note should be imperfected. On the one hand, the use of a fully black ligature rather than a half red ligature should signal a difference. On the other hand, imperfection of a ligature a post ante is unheard of in the other Paduan sources and would violate the rule of similis ante similem perfecta (see chapter 3). In the end, I decided on the first interpretation aided by evidence in m. 38, where the scribe presses two minima together to form the same note value as a red semibreve, that the scribe’s command of red mensural notation might not be very great. Further evidence of this will be seen in the discussion of Dolçe fortuna.

The notational system used in Ay si is primarily French. There are very few puncti divisiones and there are notes which are sounded across the division between two tempora. Alteration of minims is present as can be seen by the first two notes of the second staff of music (m. 31 in transcription). The same scribe who copied Se per dureça probably copied Ay si. However, some important paleographical differences leave open the possibility of different scribes. In Ay si, the long span of minims on the second line of music features equal spacing between the minims. The minims in Se per dureça are much more tightly spaced and arranged into groups of two, three, or more notes. Ay si uses lozenges that are more oblong than the rounder note heads of Se per dureça, particularly that scribe’s semibreves.

One piece of evidence that undermines a dual-scribal view of this page is the position of the clefs in relation to the left margin. After the fifth staff, the clefs begin drifting in from the left side of the staff (half of this is because the staves are also drifting to the left). One would expect if the scribe changed between staff 6 and staff 7 that the new scribe would not continue the trend of moving rightward. However, a single scribe writing down both of these pieces in the same sitting would logically account for this.

More significant are some of the differences in the writing of the text.111 Ay si’s letter "d" has a stem which is so bent forward it is nearly horizontal (see in "contratenor de" and "secunda"). The counter on the top of the "a" is so curved that it almost forms a closed bowl on the top of the letter. This curve is in contrast to the upper stroke of Se per dureça "a" which is like a horizontal arm. However, we have seen even within Se per dureça enough variations in the writing of the letter "d" that we must tolerate small inconsistencies in scribal writing unless we are willing to accept a twenty-scribe explanation for every fragment encountered.

transcription: [page 1] [page 2]

3. Aler m’en veus (Ciconia)

The verso side of folio A contains a single voice of a virelai, Aler m’en veus, attributed to "Johes" which on the basis of an attributed Latin contrafactum of the piece found in I-Bc 15 f. 266v-267 can be traced to Johannes Ciconia, who was in Padua in the first decade of the fifteenth century.112 The Latin version of this piece has two voices while only one voice survives in this French version. It has been presumed that the second voice is on the missing folio which would be adjacent to f. AV in the original manuscript. There are some reasons to question this assumption based on evidence of norms of layout in the Paduan fragments.

The piece is in two sections the second of which is repeated with open and closed endings. The closed ending is unusual in that it adds more text than the first ending. With the ars subtilior composition Sus unne fontaine (c.f. chapter 6), Aler m’en veus is one of only two known French-texted pieces by Ciconia. Its unusual extension of the closed ending causes Hallmark to identify it as in a class of its own.113

The text primarily consists of 8 syllable lines, with some lines, such as the first, probably being sung with 9 syllables and with a 6 syllable line ending the first half of the piece. The rhyme scheme of the text is aabbaab dbdb dbaba although more definite conclusions about the rhyme scheme can only be said when an appropriate reading for the end of the 6th line is found.

The music for Aler m’en veus is written without puncti divisionis.114 Because it is written in tempus imperfectum cum prolatione minore, there is only one form of the semibreve here, and all note values indicate exactly one duration (like common practice western music notation). Syncopation abounds in this piece, with minim-semibreve-minim or minim-semibreve-semibreve-minim being the most common forms of syncopation (1:8 reduction):



These syncopations occur both within a single long (as shown above) and less frequently between two longs, as at the beginning of the virelai. In this piece we begin to see a cadential suspension formula that would become popular centuries later. Measures 2-3 in the transcription show the descent from f through a syncopated semibreve e then a syncopated semibreve d then a minim c (Presumably the sharp on the breve at the beginning of the measure is still in effect) back to the cadential d.

The setting of the text is primarily one to three notes to a syllable, except for a single long melisma on "merchi". With regard to setting the end of a line of text, this piece is more true to its French text and structure than to its Italian provenance: long melismatic flourishes on the penultimate syllable are not to be found.

The top voice of Aler m’en veus is laid out on the first five staves of f. AV while the remaining five staves are blank. Since the two voices are equal, one would presume that the second voice would occupy five staves as well and therefore could easily have been placed on f. AV as well. Unless the text residuum for the missing fifth section (same music as the first section) was particularly long, it could be made to fit on the final staff by an enterprising scribe. Even if it were long, the lower margin would have provided ample room for additional text.

One might suggest that the layout of one voice on top of each of the pages would be a logical layout for a composition. However, this arrangement of voices was not used elsewhere in the Paduan fragments studied. A composition was not broken across a page in any of the five fragments unless it was necessary because of the length of the composition or the presence of other compositions (on a single page) above it on the same page. While this might be a meaningless aberration, one would think the cost of parchment would have kept the scribe of a utilitarian manuscript, such as this one, conscientious of ways of optimizing layout. There is only one other page in the fragments studied which contains a substantial amount of unused space, f. A of I-Pu 658 (PadC).

Particularly puzzling is why, if the two voices were laid out one on top of each page, would another piece not be added across the bottom of the two pages, as A piançer l’oche was to f. BV or Alta regina was to f. 60v of PadA. One explanation for this is that there was a third voice to Aler m’en veus below the second voice on the facing page, and thus the composition could neither fit on one page, nor fit another composition on the two pages combined. There are three problems with this explanation though. First, it relies on an unequal distribution of voices which favors the second page, which would also be unusual. Second, the top two voices of the composition are lower than most upper voices and would require a quite low tenor voice. Finally, although Bent and Hallmark remark that imitation and rhythms of the top voices is similar to the upper voices of a motet, the two voices show no signs of requiring a lower voice. There are no perfect fourths on strong beats, which would suggest the need for a voice a perfect fifth lower. Many cadences are bare octaves, but this is not uncommon and the imposition of a third voice between these two voices would be hard to believe. The possibility of two more voices relieves the harmonic problems but exacerbates the imbalance of voice parts on the page. It may be possible that only the top voice was ever present in the manuscript. Imitative as the second voice is though, it could not have been generated canonically from the first voice.

In lieu of strong evidence for any alternate theory, for now the standard two voice explanation seems most likely. One further piece of evidence to support this is the presence of the first name "Johes" at the top of the page, possibly suggesting "Ciconia" at the head of the second.

transcription: [page 1] [page 2] [page 3]

4. En ce gracieux (Senleches)

The French texted three-part composition, En ce gracieux relies on complex syncopations far more than any other work in PadB (only Sus unne fontaine comes close in the music studied) and yet is much less rhythmically complex than the rest of Senleches’s output. The composition consists of a texted cantus with untexted contratenor and tenor voices.

The composition is notated in French notation, without mensuration signs or points of division. There are a number of errors where notes which need to be dotted are found without, possibly indicating a this piece was never performed from this manuscript by people who were not already familiar with the work. The top two voices have a b-flat in their key signature while the tenor voice has a single e-flat as its signature. It is very probable that the b’s are flatted as well (c.f. m. 3 where a b would cause a melodic augmented-fifth, and m. 13 which would have an augmented octave between outer voices). However, since the scribe indicates the b in m. 7, and similarly dissonant intervals appear at other places in the score, the possibility of b’s in the tenor voice cannot be fully ruled out. The scribe of the composition is Scribe A, with features very similar to Aler m’en veus.

The piece is written in French notation, tempus imperfectum cum prolatione minore, so there is no need for imperfection or alteration. Points of addition are used, even on ligatures c.o.p. Interestingly, the second ending is marked with the Italian "chius" rather than the French "clos."

The text of the composition describes good times in which birds are heard to sing to the speaker. The songs of the nightingale (oci oci) and cuckoo (cocu cocu) prominently figure into the composition. Indeed, the text of the entire second half almost seems an excuse to work in the cuckoo’s song: this is pure poesie pro musica. The song of the nightingale does not seem to have a definite pattern to it, except a slight predilection for rising minor thirds and major seconds from the first to the second syllable. The song of the cuckoo, on the other hand, is presented in hockets between the top two voices. The cuckoo is represented by a minim rest, minim, and a semibreve a minor third lower, from g to e. This is the same melodic pattern used by the cuckoo in Italian madrigals of later centuries and given in the first collection of notated birdsong, Musurgia universalis by Athanasius Kircher (1650).115 Below the hockets, the tenor sustains a double-long C. Typical for compositions of this time, the second ending is one note lower than the first ending, though there is a voice exchange in the upper parts so that the music is not literally moved down a step.

The bottom two voices are untexted probably indicating instrumental performance. In the cuckoo section of the contratenor, the text "cocu cocu cococu cocu" is written. It is unclear whether this is a guide to the player as to what the top voice is performing; one would think this section to be least likely to need such a guide. Is the player supposed to stop playing and sing this section? This does not seem likely either. An adequate explanation for the presence of this text has yet to be found.

transcription: [page 1] [page 2] [page 3]

5. Dolçe fortuna (Ciconia)

The second composition by Ciconia in PadB, the ballata Dolçe fortuna (or Dolçe dolçe fortuna in the manuscript) is stylistically very different from Aler m’en veus. The composition is in a triple meter, either .i. or tempus imperfectum cum prolatione maiore.117 The French use of alteration and imperfection and lack of points of division are present in this composition. The main interior cadence gives a textbook example of the so-called Landini cadence: sixth, decorated by an incomplete lower neighbor in the top voice, expanding outward to an octave with the rhythm of alternating minims and semiminims in the upper voice.118 Many of the other cadences approach a unison through the motion inwards of a third. The top voice often has a lower auxiliary, creating a temporary minor second (or in cadences on e, major second, since the d would not be raised to d) as an added touch of dissonance before the perfect consonant resolution.

Where dragma, or double tailed semibreves (), are used in PadB, the concordance in the later manuscript F-Pn 4379 (PC II) has void semibreves.119 Because the same scribe used red and red void notation in Ay si, one wonders why he would use dragma here unless he were merely copying the pieces and not acting as a scribal-editor which would, as Nádas has suggested, be more common. Textual variants occur between the two voices which would not be worthy of comment were it not for their consistency. Three times, the text of the upper voice uses "y" (may, costey, y foragi) where the tenor voice has "i". This raises the possibility that perhaps the scribe was copying from a source where the tenor voice was untexted as it is in F-Pn (PC II).

transcription: [page 1] [page 2] [page 3]

6. A piançer l’oche (Antonelus Marot)

Only the top voice of A piançer l’oche by Antonelus Marot survives, copied at the bottom of f. BV of PadB. The piece is copied in Italian notation, in octonaria though with no divisiones letters. Semibrevis caudata () is used extensively in the piece to denote the maior semibreve, even in the one instance (m. 3) where it would not have been necessary by the rule of via naturae. Oblique stemmed semibreves make their only appearance in PadB and are equal to 3 minims, as expected. In m. 49, the group of four semiminims should probably be four minims to complete the tempus.

The fragmentary composition A piançer l’oche was probably copied by a different scribe than the other compositions in PadB, though the similarities between the decoration of the final longs, the angle of the custos, and the shape of the clefs makes a firm determination difficult. The lighter hand and extensive text decoration are the principal arguments for the designation of a separate Scribe B. These arguments are summarized in the table below:

   scribe A    scribe B

range of copying AR-BV (Se per dureça - Dolçe fortuna) BV (A piançer l’oche)
accidentalsflats are dotted, slope upward. sharps written as sharps (not naturals). none


single, straight, 45 degree (plus or minus) similar to A
triplets/Sm trip- right flag
Sm - right flag also (never in same piece)
trip - flared right flags
Sm - left flags, not flared
pen size moderately large lighter on text, somewhat smaller on noteheads
ink color varies, even within a composition (Se per dureça has C: black, T: light brown) similar to dark ink of A
character features hooked stem on "h". some flourishes on "d" extensive decoration of almost all letters. flourishes on "d" the rule. Long tails on "s". Fancy cedilla.

transcription: [page 1] [page 2]