Related sources

Other sources

by first line
   by composer



Papers and notes

General Index of music editions
by first line
   by composer


Editions and papers
on this site:

Complete Works of Gilles Mureau

Amiens MS 162 D

Sacred music of the 15th century

Uppsala MS 76a

Peter Woetmann Christoffersen

Papers on

Basiron’s chansons
Busnoys & scibes PDF
Chansons in Fa-clefs
Chansoner på nettet
Fede, Works
Dulot’s Ave Maria
Open access 15th c.
MS Florence 2794


The French musical manuscript in Florence, Biblioteca Riccardiana, Ms. 2794,
and the ‘Loire Valley’ chansonniers

Peter Woetmann Christoffersen - June 2012

This article is also available as a PDF-file including the list of contents of the MS Florence 2794.

Sections in the article:

Date and place
Structure and genesis
Relations between scribal hands
How could this have happened?
A hypothetical chronology concerning Florence 2794 and the Dijon and Laborde chansonniers

This preliminary discussion of Florence 2974 builds on the available literature and a microfilm, as I have not yet been able to study the manuscript in place. The material has been supplemented by close readings of selected concordances with the Dijon and Laborde chansonniers. (1) The readings prompt some intriguing hypotheses concerning the genesis and early history of the manuscripts.


Florence 2794 is a parchment manuscript of medium size, 240-245 x 166-170 mm, ordered in ten fascicles, of which the last is missing a bifolio (between ff. 75 and 76). Modern foliation is stamped in the upper right corner of the recto pages, and the manuscript is bound between modern covers of reddish-brown tooled leather with brass locks. One principal scribe entered the original layer of music and texts, and several later hands added music and supplemented texts – see the List of contents, which also specifies the hands and the fascicle structure.

The main scribe (hand A or FlorenceA) designed the manuscript as an entity with a planned musical repertory. In the long run he was not able to carry through his plans. He ruled all pages with spaces left for initials and decorations, but none of them were ever filled in (a later hand (C) drew some inked initials in the spaces in the superius parts on ff. 72v and 73v). For most openings he used a basic pattern, which consists in on the left-hand page five or six staves for the upper voice and the additional text, and four and five staves respectively for the tenor and contratenor (bassus) parts on the right-hand page. The position of certain pieces in the manuscript was clearly prearranged, as he here modified his staff-pattern to fit their special needs. For example, the un-texted motet »Gregorius presul« certainly was meant to open the finished volume as it begins on the only opening with spaces left for initials, which in all three voices should cover the beginnings of two staves (ff. 1v-2). Similarly, he had planned to enter the two four-part songs by Busnoys (nos. 22-23, ff. 25v-27), which have voices of more equal length than the three-part songs, at the start of the 4th fascicle and adjusted the layout accordingly. In other cases he just modified the basic pattern by adding staves on the left-hand pages etc. All the pages he left without music are ruled according to his basic pattern of 5-6 staves on the left side of the opening and four and five staves (or just nine) for the lower voices on the right-hand page.

In 1973 Joshua Rifkin published his identification of the main scribe with the hand of the scribe who added some songs to the Laborde Chansonnier (hand LabordeC: ff. 101v-104, 105v-106, and 120v-121), and furthermore, that also the second scribe of Florence 2794 added some songs to the Dijon Chansonnier (hand DijonB: ff. 187v-192). (2) A few years later the same author published his finding in Florence 2794 of the signatures and handwriting of the French singer and composer Pietrequin Bonnel – ff. 67v-68, the song “Qu’en dictez vous, suis je en danger”, and on the folios 3v, 4v, and 7v he added his name above the songs. (3)

Date and place

In the musicological literature Florence 2794 has been regarded as a French source because of its repertory and the native French orthography of its texts, and its has been dated in the last quarter of the 15th century. (4) More specifically, the manuscript has been connected with the French royal chapel and dated in the early part of the 1480s, (5) and the contribution of Pietrequin Bonnel has been placed before 1488 when he left France for Savoy and later Italy – and possibly brought the MS with him to Italy. (6)

Structure and genesis

The main part of the manuscript must be designated as a chansonnier; also its small format identifies it as a secular music manuscript. However, sacred music turns up in the first fascicles in a way uncharacteristic of a chansonnier. Not that an opening motet is foreign to a chansonnier; on the contrary, an introductory short sacred song was welcome here, but the two four-part motets by Compere and Ockeghem which open the second fascicle belong to another category of music, and »Gregorius presul« is not quite comparable to, for example, Frye’s »Ave regina celorum« which opens the Laborde and Wolfenbüttel chansonniers. In fact, the two first fascicles do not seem to belong to the main scribe’s original plan for his collection.

An explanation of this could be that the collection was commissioned by a wealthy patron who wanted a chansonnier, and that the conditions for the project had to be changed quite early in the process of compilation. Maybe the intended receiver fell away, so that the main scribe had to try out other measures in order to attract a new prospective owner; apparently he did not succeed. This may be the point where the opening motet »Gregorius presul« appeared in the picture. The main scribe also added a fascicle with the two attractive four-part motets by Compere and Ockeghem – at this stage the collection existed in the form of loose fascicles. Obviously he did not count on laying his hands on any further comparable motets, as he filled out the remainder of the pages in the fascicle with staves destined for three-part music. Fascicle 3, too, was a separate entity containing only four songs by composers connected to the royal chapel or well known in these circles (Du Fay, Hayne, Ockeghem and Fresnau, nos. 14-17). Fascicles 4-10 make up the proper chansonnier and were presumably ruled in one operation and filled out successively by the main scribe. It is very common for chansonniers to contain at the end one or more fascicles with musical staves drawn in and no music entered. They function as repositories where the owner may get new favourite songs copied. But in this case the main scribe apparently planned to open a new section of the chansonnier with a four-part song, namely the double chanson “Adieu mes amours on m’attent / Adieu mes amours” by “Josequin”, and just before this song he entered the final song of the section containing mainly three-part songs, namely the “Rondeau royal”, “Ung aultre l’a n’en queres plus”, “De okeghem“, which connects fascicles 8 and 9. (7)

All his work on fascicle 3 and most of fascicles 4-10 was probably done before he added fascicles 1-2. The textless tribute to Pope Gregory I, »Gregorius presul« (no. 1; see further the edition) was probably planned as a sort of “dedication” piece, and most probable first and foremost as a musicians’ motet, a tribute to music and the collegium of musicians, rather than to the myth of Gregory as the creator of the yearly cycle of plainchant. It is remarkable that only when the plainchant tune reaches the words “musice artis” in bars 92-96 the setting becomes declamatory chordal. We cannot know whether the text he intended to use was in fact the old trope or a new text fashioned to honour a wished-for receiver of the manuscript or a contemporary musician (Ockeghem, the leader of the court chapel, could be a candidate for praise as the father of modern music!). Maybe the planned text failed to appear – and the main writer gave up his project. His last entry in the chansonnier was a textless version of the rondeau “En effait se ne reprenes” (no.  52, ff. 60v-61), for which a later hand added the beginning of the poem on f. 1.

At this stage the main scribe abandoned his project. It was a collection of fascicles: a chansonnier with a lot of empty, ruled pages at the end (fascicles 4-10); a fascicle with four songs and empty pages (fascicle 3); one fascicle containing an unfinished three-part motet and another with two fully texted four-part motets, both fascicles having many empty pages (fascicles 1-2). The main scribe may have died or left his position at the musical institution. In any case, a second scribe (hand B) very soon took over work on the collection, had it bound, and copied music on most of the empty pages – probably in collaboration with other scribes working in the same institution (mainly hands C and D), any temporal difference between the hands is not discernible, and they are like the main scribe all professionals and educated in similar institutions. Their repertory is basically of the same sort and generation as the main scribe’s. It encompasses, however, no more songs by Ockeghem and Busnoys; the foremost names are now Agricola, Compere, Hayne and Pietrequin Bonnel, but hand C also knew some old songs by Du Fay and Binchois, which he placed alongside a song by a younger composer, Prioris (nos. 19-21). The last pages in the manuscript were filled by guests, each adding a single song (hands E-H, including Pietrequin).

Relations between scribal hands

Joshua Rifkin’s identification of the scribal hands working on Florence 2794 and the Laborde and Dijon chansonniers respectively has been accepted in the musicological literature. (8) My investigation based on photographic reproductions of the pages in question fully supports Rifkin’s results (I do not believe that it ever will be permitted to bring these sources together in one location in order to compare the hands directly). The information that two scribes who in turn worked on Florence 2794, both supplemented the repertory of two older chansonniers, represents a nearly unbelievable lucky chance, which demands a closer inspection. However, in the existing literature this information is just recorded without much further discussion.

Already by a close reading of the songs, which the scribes each entered in two manuscripts (Florence 2794 and Laborde, and Florence 2794 and Dijon respectively), the relationships becomes even more amazing: As one of his first efforts the main scribe entered Fresnau’s »De vous servir m’est prins envye« into the third fascicle of Florence 2794 (no. 17, ff. 20v-21). When he copied the same song into the last, nearly empty section of the Laborde chansonnier (no. 84, ff. 103v-104), he used a better exemplar or he revised the one he had – most significantly by correcting the key signature in the tenor as well as the text (see further the discussion and edition of the song). And exactly the same happened when the second scribe (FlorenceB) copied Compere’s »Dictes moy toutes vos pensees« into Florence 2794 as part of his completion of the first fascicle (no. 7, ff. 8v-9) and into the Dijon chansonnier (no. 159, ff. 191v-192) – again the improvements concern a misleading key signature in the tenor and a faulty text line (see further the discussion and edition of the song).  

Considering the professional status of the scribes it seems impossible that they made the entries in Laborde and Dijon before their work on Florence 2794. They either got hold on better exemplars or revised the ones they had, possibly after consulting with the composers. The obviously higher status of the two big chansonniers, both much nearer completion than Florence 2794 and partly illuminated, may also have contributed to a sharpening of the scribes’ attention. Of course, we are here on speculative ground, but I think that we may safely assume that the additions to Laborde and Dijon by the Florence scribes were made later than the two scribes’ entries of the same songs in the Florence MS.

How could this have happened?

The first and simplest explanation must be that the Dijon scribe did in fact deliver the two chansonniers Dijon and Laborde to patrons close to the French court, and that their owners then some years later independently approached the court chapel’s suppliers of music in order to have new pieces added to the manuscripts. In this case, the work of the two Florence scribes is evidence that Dijon and Laborde were still in royal court circles when the additions were made. (9) But this story does not seem very convincing. On the other hand, it does not either sound plausible that two manuscripts of their size and with so much work and money invested just lay fallow in the Dijon scribe’s atelier during a long period of time – if we adopt the current datings of the MSS involved –, until his successor, the Florence scribe and his associates, had them supplemented with extra music, had the index of Laborde updated and some existing pieces corrected, (10) and finally disposed of.

However, what speaks for the last hypothesis is that Florence 2794, Dijon and Laborde all probably represent different sorts of ‘failed projects’, which remained in the possession of their scribes for some time, and during their career passed through the same hands. While the small Copenhagen chansonnier (Copenhagen, The Royal Library, MS Thott 291 8°) stands as a finished product, which was handed over to the person who commissioned it from the Dijon scribe’s workshop, the scribe was not able to close the deals on the Dijon and Laborde MSS even if a lot of expertise and expenditure was invested in their production. On the other hand, they were hardly available to the Florence scribes through an extended period. Nothing indicates that the main scribe had access to the repertory when he made his part of Florence 2794. The shared repertory in fact only includes a small number of songs already present in Dijon and Laborde, that is Florence 2794 nos. 15, 16, 31 and 48 – all except Basiron’s song (no. 48) belonging to the standard repertory. Neither the songs later added to the three MSS show any close mutual dependency (Florence 2794 nos. 4, 12, 18-20, 32, 37, 39, 50, 57, and 65), though the still later added songs in Laborde and the songs, which Hand C entered in Florence 2794, are closely related. Likewise the main scribe’s entries in Laborde (as Hand LabordeC) do not show any signs that he had Dijon’s versions of the same songs before his eyes, cf. the editions and the comments on Laborde nos. 82 and 86, Busnoys’ »A une dame j’ay fait veu« and Ockeghem’s »Les desloyaulx ont la saison«.

Maybe the proposed datings of the MSS concerned have to be revised somewhat. Jane Alden places the activities of the Dijon scribe in the 1470s, and Rifkin and Litterick date Florence 2794 in the early 1480s. Fresnau was a singer in the French court chapel before he went to Milan and became a colleague of Compere. Duke Galeazzo of Milan was murdered in 1476, and the following year his heir disbanded his chapel. It is a possibility that Fresnau accompanied by Compere more or less directly went back to the French court chapel where both of them stayed on for a long time – as it is well known nearly all archival documentation concerning the personnel of the chapel has disappeared. (11) If they were present in the last years of the 1470s, and the dating of Florence 2794 is moved backwards correspondingly – and possibly also the main activities of the Dijon scribe may be dated a bit later – then the whole process appears more plausible.

By these adjustments we can establish a picture of a group of sources for the history of the French chanson, which is closely interweaved and shows continuity – quite different from the traditional picture of scattered sources. Concurrently a picture emerges of a setting for the production of musical artefacts, which is bigger and more institutionalized than what we earlier have been able to imagine. The best proposals for places of origin for the ‘Loire Valley’ chansonniers have been scribes active at the cathedral in Tours, at the palace church in Bourges or similar. Our picture depicts a more extended group of scribes or musicians who in turn or according to their periods of employment handled these functions. The role of scribe to the French court chapel seems to be an alternative proposal.

Persons attached to the court chapel had access to all the musical circles represented in the manuscripts, the local circles in Bourges, Orléans, Blois or Tours as well as the international scene of Burgundy or Paris as they followed the king around between the palaces in the Loire Valley, to diplomatic meetings in his extensive realm or on war campaigns

However, the sources do not tell us much about the function of the scribe in connection with the court chapel. And it is hardly likely that projects involving elegant, commissioned chansonniers were hauled around with the chapel on its journeys. Maybe we should rather imagine that the court chapel was associated with a supplier based in a big city, where access to affluent customers was steady. Paris, residence of for the central administration with its many highly educated, rich and socially ambitious officials, is an obvious guess, but also Tours, which during the decades following the end of the Hundred Years War for long periods de facto functioned as the capital of the kingdom, (12) and other localities must be considered. Centrally placed book entrepreneurs or libraires organised productions of luxury goods such as the costly illuminated books of hours, magnificent missals, or collector’s editions of literary works by financing the purchasing and preparation of materials, by ordering the copying of texts (and in these cases of music), and they had business relations with illuminators in several cities. If the so-called  ‘Loire Valley’ chansonniers result from such activities, which in some way or another involve the expertise present in the royal chapel, it may explain many of the obscure points in their genesis, and also their relations with Florence 2794.

The lack of research on these topics of course makes all this hypothetical. But a simplified working scenario could be that the Laborde and Wolfenbüttel scribes in some way or another were colleagues who occasionally borrowed exemplars from each other. The Dijon scribe probably ran a different business in the same circles, and he had suddenly to take over the Laborde scribe’s project – and the Florence scribe and his workshop succeeded the Dijon scribe.

A hypothetical chronology concerning Florence 2794 and the Dijon and Laborde chansonniers

Some time during the late 1470s the main scribe (Hand A) began a chansonnier, probably on commission from a member of the French court. The 3rd fascicle, which originally contained four attractive songs only, might have been a sort of trial pages to show his patron; he never got around to get this fascicle integrated into the manuscript. Having obtained the approval of the recipient he then produced the fascicles 4-10 and most of their contents.

A change of plan; probably the original recipient was no longer available. Hand A tries to make his collection more attractive to a new customer by sketching two fascicles with sacred music and a dedication motet; the manuscript may now be intended as a gift.

The Dijon and Laborde MSS pass into the care of the Florence scribe’s workshop; these not quite finished projects were transferred or inherited from the Dijon scribe. Fresnau and Compere enter the royal chapel, and their music becomes easily accessible to the scribes before 1480.

Hand A adds some songs to Laborde including Fresnau’s “De vous servir” in an amended version. Another scribe (Laborde D) helps out with the complement of repertory, the index is updated, and finally Laborde is delivered to a buyer.

Hand A abandons his own project without finishing the copying of no. 52 – he dies or changes position – or the intended receiver definitely drops the project; the opening dedication motet never receives a text.

Hand B assumes responsibility for Florence 2794 and supplies chansons by first and foremost Pietrequin, Compere and Agricola (in due course assisted by hands C and D); he probably has the manuscript bound in some form.

Hand B copies three songs into Dijon, two by Compere, and one of them, “Dictes moy”, in an amended version in comparison with the same scribe’s entering of it in Florence 2794. The Dijon chansonnier is as a finished product handed over to an owner.

Other hands gradually fill out the majority of the empty pages in Florence 2794, but the manuscript was never “completed” with painted illuminations and decorations and released to a wealthy owner. At some point it ends up in Florence, possibly by the mediation of Pietrequin. If this is the case, the development of the manuscript in France may have ended before 1488.

1) Concerning the Dijon (Dijon, Bibliothèque Municipale, Ms. 517) and Laborde (Washington D.C., Library of Congress, MS M2.1 L25 Case) chansonniers, see the detailed discussions in Jane Alden, Makers of a Songbook: The Scribes of the Laborde Chansonnier (Ph.d.-diss., Univ. of North Carolina) 1999, and idem., Songs, Scribes, and Society.  The History and Reception of the Loire Valley Chansonniers. New York 2010.

2) Joshua Rifkin, ‘Scribal Concordances for Some Renaissance Manuscripts in Florentine Libraries’, Journal of the American Musicological Society 1973, pp. 305-326 (at pp. 318-326).

3) Joshua Rifkin, ‘Pietrequin Bonnel and Ms. 2794 of the Biblioteca Riccardiana’, Journal of the American Musicological Society 1976, pp. 284-296.

4) “Le fait qu'aucune pièce italienne n'y figure et que les textes français y soient particulièrement corrects, laisse penser que la copie a été faite en France par des scribes du pays […], copie que l'on peut situer dans le dernier quart du XVe s., époque qui correspond avec le répertoire […].” (RISM BIV/5, p. 222),

5) George Morton Jones, The “First” Chansonnier of the Biblioteca Riccardiana, Codex 2794: A Study in the Method of Editing of 15th Century Music. PhD-dissertation, New York University 1972, pp. 16-18, dates the MS 1475-85 and proposes that it was commissioned by or for a member of the Sforza family in Milan! Louise Litterick, The Manuscript Royal 20.A.XVI of The British Library. PhD-dissertation, New York University 1976, pp. 66-76, dates the MS before 1488, and confirms its connection with the French court by a comparison with physical traits and repertories of other court MSS.

6) Rifkin 1976, p. 288. This dating may very well be correct, but Rifkin’s reasoning does not hold up for scrutiny. He writes: “We can, however, almost definitely rule out the Florentine phase of his career. “Qu’en dictez vous” belongs to a series of works written by several different scribes in the last gatherings of Florence 2794; it falls in the middle of a gathering, which no doubt means that Pietrequin entered it as part of the series, not as an addition made subsequently to fill a gap left by the other scribes.” This means, that Rifkin presumes that a wide selection of scribes were at work concurrently on the pages following f. 61, that is nearly all the hands found in the manuscript: A, B, C, E, F, G (Pietrequin?), and H (cf. Contents, and see Litterick (cf. note 5) pp. 78-79). This is not likely. As the present analysis shows, the main scribe (A) left his work unfinished with lots of empty pages, which were filled by later hands at several occasions. There is nothing to tell us when Pietrequin made his additions; it might have in Italy, or later in the 1490s when he sang in the chapel of Queen Anne de Bretagne. The last hypothesis would explain his underlay of all parts with text.

7) The appearance of Josquin Desprez here alongside Ockeghem confirms David Fallows’ dating of “Adieu mes amours” as an early work in the style of Ockeghem, and that the young “Josequin” at an early date was well known in French court circles, cf. David Fallows, Josquin. Turnhout 2009, pp. 41-43.

8) The rejection of Rifkin’s identifications by Martella Gutiérrez-Denhoff builds on her incorrect identification of the different hands found in the Laborde chansonnier; cf. Der Wolfenbütteler Chansonnier. Wolfenbüttel, Herzog August Bibliothek, Codex Guelf. 287 Extrav. Untersuchungen zu Repertoire und Überlieferung einer Musikhandschrift des 15. Jahrhunderts und ihres Umkreises. (Wolfenbütteler Forschungen 29) Wiesbaden 1985, pp. 101-102.

9) Cf. Alden 2010, p. 127.

10) Most of the songs entered by hands LabordeC (FlorenceA) and LabordeD have been added to the index in Laborde, cf. Alden 1999 pp. 78-79 and 245-253, and Alden 2010 pp. 91-93.

11) Concerning Fresnau and Compere, see Allan W. Atlas/Jane Alden, “Fresneau, Jehan” and Joshua Rifkin, Jeffrey Dean & David Fallows, “Compère, Loyset” in Grove Music Online; and Jean Fresnau (O. Carrillo & A. Magro, eds.), Messe et chansons. Turnhout 2004, pp. vii-xii.

12) Bernard Chevalier, Tours ville royale (1356 - 1520). Origine et développement d’une capitale à la fin du Moyen Age. Louvain 1975.