¹ Sadie, S., ed., The New Grove Dictionary of Music and Musicians, 20 vols. (London, 1980)Google Scholar.

² Rome, Biblioteca Casanatense, Cod. 2856; ed. Wolff, A., ‘The Chansonnier Biblioteca Casanatense 2856: Its History, Purpose and Music’ (Ph.D. diss., North Texas State University, 1970)Google Scholar; accepted by Atlas, A. in Robert Morton: The Collected Works, Masters and Monuments of the Renaissance 2 (New York, 1981), p. xxviiGoogle Scholar, and by Brown, Howard Mayer in A Florentine Chansonnier from the Time of Lorenzo the Magnificent: Florence, Biblioteca Nazionale Centrale MS Banco Ran 229, Monuments of Renaissance Music 7 (Chicago, 1983), p. 193CrossRefGoogle Scholar. Lewis Lockwood associates this source with one described in a Ferrarese document and suggests a date of 1480 based on the career of the specified scribe: see Music in Renaissance Ferrara 1400 – 1505: The Creation of a Musical Center in the Fifteenth Century, Studies in the History of Music 2 (Cambridge, Mass., 1984), pp. 224–6Google Scholar. Lockwood's proposal remains controversial, as not everyone agrees that the description which he cites is appropriate to this particular manuscript. For a partial defence of Wolffs position, see below.

³ New Haven, Yale University, Beinecke Rare Book and Manuscript Library, MS 91; ed. Perkins, L and Garey, H., The Mellon Chansonnier (New Haven and London, 1979) See vol. i, pp. 17–32Google Scholar.

⁴ Dyon, Bibliothèque Municipale, MS 517; facs. ed. Plamenac, D., Dijon: Bibliothèque Publique: Manuscript 517, Publications of Medieval Musical Manuscripts 12 (Brooklyn, 1970)Google Scholar; ed. Barret, C. E, ‘A Critical Edition of the Dijon Chansonnier. Dijon, Bibliothèque de la Ville, MS 517 (Ancien 295)’ (Ph.D. diss., Vanderbilt University, 1981)Google Scholar; part ed Droz, E., Thibault, G., Rokseth, Y, Trois chansonniers français du xv^e; siècle (Paris, 1937, repr. 1977)Google Scholar. Washington, D.C, Library of Congress, MS m2 1 l252 Case (‘Laborde Chansonnier’); edn in preparation by J. Pruett Copenhagen, Kongelige Bibliotek, MS Thott 291, 8°; ed. Jeppesen, K., Der Kopenhagener Chansonnier (Copenhagen and Leipzig, 1927, repr. 1965)Google Scholar. Wolfenbüttel, Herzog August-Bibliothek, Cod. Guelf. 287 Extravagantium. Paris, Bibliothèque Nationale, MS Rés. Vmc 57 (‘Nivelle Chansonnier’); facs. ed. Higgins, P (Geneva, 1984)Google Scholar.

⁵ Strunk's, Oliver datings were reported by Helen Bush in ‘The Laborde Chansonnier’, Papers of the American Musicological Society 1940, pp. 60–3Google Scholar.

⁶ Paris, Bibliothèque Nationale, MS Rothschild 2973; part. ed. Kottick, E., The Unica in the Chansonnier Cordiforme, Corpus Mensurabilis Musicae 42 (1967)Google Scholar.

⁷ Perkins, and Garey, , Mellon Chansonnier, ii, p. 156Google Scholar; Atlas, , Morton, p. xxivGoogle Scholar; Brown, , A Florentine Chansonnier, p. 185Google Scholar. On the controversial dating of the Nivelle Chansonnier, see my forthcoming study ‘The Chanson Repertory at the Northern Courts: France and Burgundy during the Era of Ockeghem’, Appendix 1.

⁸ London, British Library, MS Royal 20.a.xvi; ed. Litterick, L., ‘The Manuscript Royal 20.a.xvi of the British Library’ (Ph.D. diss., New York University, 1976)Google Scholar. Florence, Biblioteca Riccardiana, MS 2794; ed. Jones, G., ‘The “Second” Chansonnier of the Biblioteca Riccardiana, Codex 2794: A Study in the Method of Editing 15th-century Music’ (Ph.D. diss., New York University, 1972)Google Scholar. Paris, Bibliothèque Nationale, MS fonds fr 2245. Paris, Bibliothèque Nationale, MS fonds fr. 1597: ed. Shipp, C. M., ‘A Chansonnier for the Dukes of Lorraine: The Paris Manuscript fonds français 1597’ (Ph.D. diss., North Texas State College, 1960)Google Scholar On the interrelationship of these sources, see Litterick, pp. 39–82.

⁹ Montagna, ‘The Chanson Repertory’.

¹⁰ Hamm, C., ‘The Manuscript San Pietro b80’, Revue Belge de Musicologie 14 (1960), p. 48CrossRefGoogle Scholar.

¹¹ Fallows, D., Dufay (London, 1981), p. 77Google Scholar.

¹² Ibid., p. 78.

¹³ Archives du Nord, Series B, 3661, fol. 28^v; quoted by Marix, J., Histoire de la musique et des musiciens de la cour de Bourgogne sous le règne de Philippe le Bon (Strasbourg, 1939), p. 200Google Scholar.

¹⁴ Fallows, , Dufay, p. 78Google Scholar.

¹⁵ Planchart, A., ‘Guillaume Dufay's Masses: Notes and Revisions’, The Musical Quarterly, 58 (1972), pp. 20–3Google Scholar; accepted by Fallows, , Dufay, p. 78Google Scholar.

¹⁶ Vatican City, Biblioteca Apostolica Vaticana, San Pietro MS b 80.

¹⁷ Finscher, L., ‘Loyset Compère and his Works’, Musica Disciplina, 12 (1958), p. 107Google Scholar; Rifkin, J., ‘Compère, Loyset’, part 1, The New Grove Dictionary, iv, pp. 595–6Google Scholar.

¹⁸ Schavran, H., ‘The Manuscript Pavia, Biblioteca Universitaria, Codice Aldini 362: A Study of Song Tradition in Italy circa 1470–80’ (Ph.D. diss., New York University, 1978), pp.226–36Google Scholar.

¹⁹ Rifkin, ‘Compère’, p. 596.

²⁰ It has recently been discovered that Compère does contribute a song to the early layer of the Laborde Chansonnier: see Winn, M., ‘Some Texts for Chansons by Loyset Compère’, Musica Disciplina, 33 (1979), p. 44Google Scholar. However, neither the style of the era nor the internal chronology of Compère's songs permits a date in the 1460s for Puisque bien m ' est advenu; the song must surely date from the 1480s, and therefore be a later interpolation into the source.

²¹ Rifkin, ‘Compère’, pp. 595–6.

²² El Escorial, Real Monasterio de San Lorenzo del Escorial, Biblioteca y Archivo de Musica, MS iv.a.24; ed. Hanen, M. K., The Chansonnier Escorial iv.a.24, Musicological Studies 36 (Henryville, Ottawa and Binningen, 1983); see vol. i, p. 47Google Scholar.

²³ For a revised dating of Escorial iv.a.24, see Planchart, A., review of Hanen, Journal of the American Musicological Society, 38 (1986), p. 364Google Scholar. On the complex of sources dating from the early 1460s, see Montagna, , ‘Johannes Pullois in Context of his Era’, Revue Beige de Musicologie, 42 (1988)Google Scholar. De tous biens plaine is an addition to Pavia 362.

²⁴ Presque transi is certainly one of Ockeghem's earliest songs, as is D' ung aultre amer. Pullois's Si ung bien peu d ' esperance is found in Trent, Museo Provinciale d'Arte, MS 90, as are most of his identifiable songs. Dufay's Malheureulx cueur is thought to have been composed in December 1455 (see n. 22), and his Par le regard is in Trent 90, Pavia 362 and the main layer of Escorial iv.a.24. The disputed chanson Au travail suis is stylistically consistent with 1450s procedure, as reflected in the sources of that era.

²⁵ Fallows, , Dufay, pp. 70, 152Google Scholar.

²⁶ On the style of the Ockeghem circle and its chronological development, see Montagna, ‘The Chanson Repertory’.

²⁷ Fallows, , ‘Johannes Ockeghem: The Changing Image, the Songs and a New Source’, Early Music, 12 (1984), pp. 218–30CrossRefGoogle Scholar.

²⁸ Fox, C W., ‘Barbingant’, The New Grove Dictionary, ii, p. 140Google Scholar, Fallows, ‘Johannes Ockeghem’, pp. 222–5.

²⁹ Until the recent discovery of a new document (see n. 32), it was often assumed that van Ghizeghem died in 1472. Although such a theory is logically incompatible with another which explains his poor showing in the ‘Burgundian chansonniers’ by an assumption that they were compiled early in his career, nevertheless these two theories seem to have coexisted.

³⁰ For example, Thomson, J., An Introduction to Philippe(?) Caron (Brooklyn, 1964), pp. 14–15Google Scholar, and Reese, G., Music in the Renaissance (revised edn, New York, 1964), p. 104Google Scholar.

³¹ Hudson, B., Hayne van Ghizeghem: Opera omnia (Brooklyn, 1976), pp. xvi–xxGoogle Scholar. The idea that van Ghizeghem might have had a post-Burgundian career, perhaps at the Bourbon court, was cautiously advanced in Marix, J., ‘Hayne van Ghizeghem: Musician at the Court of the 15th-century Burgundian Dukes’, The Musical Quarterly, 28 (1942), pp. 283–4Google Scholar.

³² Higgins, P., ‘In hydraulis Revisited: New Light on the Career of Antoine Busnois’, Journal of the American Musicological Society, 39 (1986), p. 40 n. 17CrossRefGoogle Scholar.

³³ Thibault, G., ‘Caron, Philippe’, The New Grove Dictionary, iii, p. 816Google Scholar, and Brown, , A Florentine Chansonnier, pp. 48, 105Google Scholar. In his review of The Mellon Chansonnier (Journal of the American Musicological Society, 34, 1981, p. 142)Google Scholar, Atlas strongly criticised this view for ignoring the prevalence of northern music in Italian sources.

³⁴ The assignment to Morton of a post-Burgundian career in Italy was rejected by Fallows, , ‘Robert Morton's Songs: A Study of Styles in the Mid-fifteenth Century’ (Ph.D. diss., University of California at Berkeley, 1978)Google Scholar and Atlas, , Morton, p. xxGoogle Scholar. The alleged Italian career of Busnois is rejected in Montagna, ‘The Chanson Repertory’.

³⁵ The concept that Busnois had quite separate careers at the French and the Burgundian courts is advocated in Higgins, ‘In hydraulis Revisited’, and Montagna, ‘The Chanson Repertory’.

³⁶ Johannis Tinctoris opera theoretica, ed Seay, A., Corpus Scriptorum de Musica 22, vol. iia (1978), p. 10Google Scholar

³⁷ This brings up the possibility that Caron could be the figure who contributed a tale to the Cent nouvelle nouvelles, a collection compiled at the Burgundian court in the first half of the 1460s.

³⁸ Thomson, , Caron, pp. 4–5Google Scholar.

³⁹ Wright, C., ‘Dufay at Cambrai: Discoveries and Revisions’, Journal of the American Musicological Society, 28 (1975), p. 205CrossRefGoogle Scholar. Brown mentions this association, but takes no sides on the issue. Yet he consistently refers to Caron's residence in Cambrai as fact, and works from dates specifically associated with Jean Caron: see A Florentine Chansonnier, pp. 48, 49, 105.

⁴⁰ Paris, Bibliothèque Nationale, MS fonds fr. 15123 (the Pixérécourt Chansonnier).

⁴¹ Seville, Biblioteca Colombina, MS 5-i-43 and related portions of Paris, Bibliothèque Nationale, MS nouv. acq. fr 4379, fols. 1–42 only; ed. Moerk, A., ‘The Seville Chansonnier: An Edition of Sevilla 5–1–43 & Paris N. A. FR. 4379 (PT. 1)’ (Ph.D. diss., University of West Virginia, 1971)Google Scholar.

⁴² Higgins concludes that Busnois could not have come to the court of Burgundy or of the Count of Charolais before early 1467: ‘In hydraulis Revisited’, pp. 46–7, p. 69 n. 102.

⁴³ Brown, , A Florentine Chansonnier, ii, p. 635Google Scholar; and Perkins, and Garey, . Mellon Chansonnier, ii, no. 22Google Scholar.

⁴⁴ The triple-metre version from the Dijon Chansonnier is published in Marix, J., Les musiciens de la cour de Bourgogne au xv^e siècle (Paris, 1937), p. 86Google Scholar; the duple-metre version of Florence 229 is in Brown, , A Florentine Chansonnier, ii, p. 146Google Scholar.

⁴⁵ Atlas, , Morton, p. xxiGoogle Scholar.

⁴⁶ Droz, Thibault and Rokseth, Trois chansonniers français, no. 2; Hanen, , The Chansonnier Escorial, iii, p. 473Google Scholar.

⁴⁷ Florence, Biblioteca Nazionale Centrale, MS Banco Rari 229; ed. Brown, op. cit. (n. 2), vol. ii.

⁴⁸ Brown, , A Florentine Chansonnier, i, pp. 9–15Google Scholar.

⁴⁹ Se brief puis and Cent mille escus are usually given to Caron, and Je ne fay plus to Gilles Mureau. On the two songs usually given to Busnois, see n, 56.

⁵⁰ Peripheral manuscripts which tend to misattribute to one famous composer are not uncommon Three works of Jean Japart are given to Busnois in Bologna, Civico Museo Bibliografico Musicale, MS Q 17; see Atlas, , The Cappella Giulia Chansonnier, i, p. 59Google Scholar. The Segovia Manuscript distinctly favours Loyset Compère (see n. 29). In a future study I shall try to show that the unconfirmed Dunstable attributions in Modena, Biblioteca Estense, MS a x. 1.11 (‘Modena B’) are often equivalent to the ‘Anglicanus’ or ‘de Anglia’ of other manuscripts This theory assumes that fascicle-manuscripts usually circulated without any means of identification. For a different view, see Atlas, A., ‘Conflicting Attributions in Italian Sources of the Franco-Netherlandish Chanson, c. 1465–c. 1505: A Progress Report on a New Hypothesis’, Music in Medieval and Early Modern Europe: Patronage, Sources and Texts, ed, Fenlon, I. (Cambridge, 1981), pp. 249–93Google Scholar.

⁵¹ Agricola's Si dedero has 24 concordances, Caron's Helas que pourra 20, Ockeghem's Ma bouche rit 17, Morton's N'araige jamais 16, and Josquin's Adieu mes amours 15.

⁵² Noblitt, T., ‘The Ambrosian Motetti Missales Repertory’, Musica Disciplina, 22 (1968), pp. 77–103Google Scholar.

⁵³ Since Busnois could not have gone to Italy before 1483, post-Burgundian works would have to emerge in the same sources which first attest Isaac. In point of fact, the vast majority of Busnois's songs can be found in Italian sources pre-dating the work of Isaac.

⁵⁴ The song also occurs in Montecassino, Archivio della Badia, Cod. 871N, a source dating from around 1480 and possibly having a direct association with the Milanese musician Franchinus Gaffurius; see Pope, I. and Kanazawa, M., The Musical Manuscript Montecassino 871: A Neapolitan Repertory of Sacred and Secular Music of the Late Fifteenth Century (Oxford, 1978), pp. 31–2Google Scholar.

⁵⁵ On the list of musicians to be released from service, see Lowinsky, E. E., ‘Ascanio Sforza's Life, a Key to Josquin's Biography and an Aid to the Chronology of his Works’, Josquin des Prez Proceedings of the International Josquin Festival-Conference, New York, 21 – 25 June 1971, ed. Lowinsky, E E. and Blackburn, B. J. (London, 1976), pp. 40–1Google Scholar Lowinsky also demonstrated that Josquin must have already been the chief ornament of Milanese music, an inference from his lucrative benefices: see pp 33–6.

⁵⁶ For example, Brown, , A Florentine Chansonnier, p. 112Google Scholar.

⁵⁷ Thomas Noblitt has made the controversial suggestion that Josquin's paired-voice motet Ave Maria … Virgo serena was copied into a German source by the year 1476: see ‘Die Datierung der Handschrift Mus. ms. 3154 der Staatsbibliothek Munchen’, Die Musikfor-schung, 27 (1974), p. 36Google Scholar. Rifkin went on to hint at a Milanese school of paired voices in ‘Josquin in Context: Towards a Chronology of the Motets of Josquin’, paper read at the forty-fourth annual meeting of the American Musicological Society, Vancouver, 1978.

⁵⁸ According to Brown's proposal for the first owner of Florence 229, Agricola might have been in Florence when the manuscript was being compiled: see A Florentine Chansonnier, pp. 32–41. His alternative theory (pp. 41–51) would place the compilation prior to Agricola's arrival: see Atlas, , review of A Florentine Chansonnier, Journal of the American Musicological Society, 38 (1985), p. 159Google Scholar. A number of Agricola's works in this source seem to have been composed in Florence, based on patterns of concordances, texts and style. On Agricola's visit to Naples, see Atlas, , ‘Alexander Agricola and Ferrante i of Naples’, Journal of the American Musicological Society, 30 (1977), pp. 313–19CrossRefGoogle Scholar; the Segovia Chansonnier appears to be a good source for souvenirs of this visit.

⁵⁹ Concerning these two songs, Brown has presented an ingenious theory that they were originally two halves of a motet: see A Florentine Chansonnier, i, pp. 132–8Google Scholar. If so, Busnois would have been the composer of the motet, but this raises a possibility that the attribution from Rome 2856 to van Ghizeghem is not incorrect; he may have been the one responsible for turning the motet into a pair of chansons.

⁶⁰ The best source for Agricola's work during the 1490s is Florence, Biblioteca del Conservatorio Luigi Cherubini, MS Basevi 2439 (‘Basevi Codex’); ed. Newton, P. G. (Ph.D. diss., North Texas State University, 1968)Google Scholar.

⁶¹ In this connection, another puzzling repertory is that of Bologna, Civico Museo Bibliografico Musicale, MS q 16; an inventory is given in Pease, E., ‘A Report on Codex q 16 of the Civico Museo Bibliografico Musicale (formerly of the Conservatorio Statale Di Musica “G.B. Martini”), Bologna’, Musica Disciplina, 20 (1966), pp. 57–84Google Scholar, and clarified in Fuller, S., ‘Additional Notes on the 15th-century Chansonnier Bologna q16’, Musica Disciplina, 23 (1963), pp. 87–103Google Scholar.This source includes no post-Burgundian van Ghizeghem, and yet its repertory of Agricola and Compère might still be mixed between two phases.

⁶² Perkins, L., ‘Musical Patronage at the Royal Court of France under Charles vii and Louis xi (1422–83)’, Journal of the American Musicological Society, 37 (1984), pp. 542, 552CrossRefGoogle Scholar.

⁶³ Litterick, ‘The Manuscript Royal’, p. 82 n. 69, and ‘Hayne van Ghizeghem’, The New Grove Dictionary, xi, p. 786Google Scholar; apparently accepted by Brown, , ‘A “New” Chansonnier of the Early Sixteenth Century from the University Library of Uppsala: A Preliminary Report’, Musica Disciplina, 37 (1983), p. 179Google Scholar.

⁶⁴ Atlas, ‘Conflicting Attributions’, p. 275.

⁶⁵ Montagna, ‘The Chanson Repertory’.

⁶⁶ Whether this is a French or Savoyard source, its repertory is clearly central to the French tradition, and the musical readings are consistently close. Rifkin has suggested that two of the scribes of Florence 2794 contributed to the Laborde Chansonnier: see ‘Scribal Concordances for Some Manuscripts in Florentine Libraries’, Journal of the American Musicological Society, 24 (1973), pp. 318–19Google Scholar; this identification of hands has been criticised by Gutièrrez-Denhoff, Martella, ‘Untersuchungen zu Gestalt, Entstehung und Repertoire des Laborde Chansonnier’, Archiv für Musikwissensckaft, 41 (1984), p. 123Google Scholar. Litterick makes an excellent argument for the direct dependence of attributions in Florence 2794 upon London 20.a.xvi: see ‘The Manuscript Royal’, pp. 111–15. The possibility that Florence 2794 was taken to Savoy and Italy by Pietrequin Bonnel is raised in Rifkin, J., ‘Pietrequin Bonnel and Ms. 2794 of the Biblioteca Riccardiana’, Journal of the American Musicological Society, 29 (1976), p. 288CrossRefGoogle Scholar, and especially Litterick, ‘The Manuscript Royal’, pp. 66–76.

⁶⁷ This manuscript tends to provide attributions for French figures, whereas it is not always helpful for Compère. On this basis, Jones suggested that Compère had not yet arrived on the French scene: see ‘The “First” Chansonnier’, pp. 14–17. Litterick dismisses the pattern as ‘meaningless’: ‘The Manuscript Royal’, p. 72.

⁶⁸ Vatican City, Biblioteca Apostolica Vaticana, Cappella Giulia xiii.27; part. ed. Atlas, op, cit. (n. 27), vol. ii.

⁶⁹ Harmonice Musices Odhecaton A (Ottaviano Petrucci. Venice, 1501); ed. Hewitt, H. (Cambridge, Mass., 1942; 2nd edn, rev., 1946)Google Scholar.

⁷⁰ Brown, , A Florentine Chansonnier, p. 103Google Scholar.

⁷¹ It is surprising that none of the other French sources have any unica of van Ghizeghem. Perhaps the two unica in Paris 2245, clearly among his latest works, were composed only shortly before his death, and never actually got into circulation.

⁷² See Hewitt, , Odhecaton, p. 65Google Scholar, or Reese, , Music in the Renaissance, p. 137Google Scholar.

⁷³ Hudson, , ‘Compère, Lovset’, part 2, The New Grove Dictionary, iv. p. 597Google Scholar.

⁷⁴ Ibid., p. 597

⁷⁵ Brown, , A Florentine Chansonnier, p. 119Google Scholar.

⁷⁶ Nevertheless, this explanation is not wholly adequate even for the three-voice songs. The Compère repertory in Paris 2245 is also anomalous in the sense that most of it is unique and stylistically unprogressive. Thus, the question is not so much one as why should other three-voice songs be found in Italian sources, but rather why are these particular songs found in Paris 2245?

⁷⁷ Brown, notes that ‘both the concern for rhetoric and the poem's less formal and courtly tone identify it as one of the most modern compositions in the manuscript’: A Florentine Chansonnier, p. 119Google Scholar.

⁷⁸ Finscher, L, Loyset Compère (ca. 1450–1518): Life and Works, Musicological Studies and Documents 12 (1964), pp. 211–22Google Scholar.

⁷⁹ On the date of Bologna q 17 (full citation n. 50) see Wright, C., ‘Anthoine Brumel and Patronage at Paris’, Music in Medieval and Early Modern Europe, p. 52, n. 38Google Scholar.

⁸⁰ Brussels, Bibliothèque Royale de Belgique, MS 228; ed Picker, M., The Chanson Albums of Marguerite of Austria (Berkeley and Los Angeles, 1965)Google Scholar.

⁸¹ Winn, M., ‘Regrets in French Chanson Texts of the Late xvth Century’, Fifteenth Century Studies, i, pp 193–217Google Scholar.

⁸² Compare the language with that in Dufay's En triumphant de cruel dueil, the piece Fallows believes to be an elegy for Binchois: Early Music, 3 (1975), pp. 358–9.

⁸³ Hewitt, , Odhecaton, p. 72Google Scholar.

⁸⁴ Picker, , The Chanson Albums, pp. 87–8Google Scholar.

⁸⁵ Reese, , Music in the Renaissance, p. 225Google Scholar.

⁸⁶ Gutièrrez-Denhoff placed Paris 1597 at c. 1510 under the assumption that it was presented to Antoine de Lorraine during his time as duke: see ‘Untersuchungen’, pp. 124–5. However, Jonathan Couchman recognised that it could have been presented while he was being raised at court, beginning in 1501: see ‘The Lorraine Chansonnier: Antoine de Lorraine and the Court of Louis xii’, Musica Disciplina, 34 (1980), p. 115Google Scholar. Considering the absence of Mouton and Fèvin, a date around 1500 seems preferable, and Dr Litterick informs me that this is also her opinion.

⁸⁷ Canti B numero cinquanta (Venice, 1502), 1502Google Scholar; facs. ed. H. Hewitt, Monuments of Renaissance Music 2 (Chicago and London, 1967). Canti C numero cento cinquanta (Venice, 1504)Google Scholar.

⁸⁸ The repertory of Petrucci's Odhecaton is a careful mix of the familiar and unfamiliar, and some of the latter provide indications of direct links to the north. Two excellent examples are the unicum by Jean Mouton, who was quite new to the French scene, and the anonymous unicum in praise of Marguerite of Austria.

⁸⁹ Lockwood, , ‘Music at Ferrara in the Period of Ercole i d'Este’, Studi Musicali, 1 (1972), pp. 115–16, 129–30Google Scholar; Music in Renaissance Ferrara, pp. 123, 224–6.

⁹⁰ On the north-east Italian sources and the relationship of Petrucci to this group, see Atlas, , The Cappella Giulia Chansonnier, i, pp. 237, 251–2, 257–8Google Scholar.

⁹¹ Vatican City, Biblioteca Apostolica Vaticana, Cappella Sistina, Cod. 15, see Rifkin, ‘Compere’, pp. 596–7.

⁹² It is not beyond the realm of possibility that Compère did not return to France with the army. In the chaos at the battle of Fornovo the king's baggage train was lost, and the fate of his non-combatant servants is unknown. Gentil patron petitions the king to ransom a prisoner, being held in a strange country: that this plea was on behalf of a noble is unlikely, since Compère refers to the prisoner as ‘vostre instrument’. On the other hand, it would be difficult to place Compère in this situation since he praised Fornovo as a victory in Reveille toy. Finally, Gentit patron is the only one of Compère's songs not to first appear later than c. 1505, being unique to the Strozzi Chansonnier,

⁹³ There is increasing thought, mostly still at the stage of discussion, that music composed for specific patrons became that patron's property and a valued part of his collection, in the same class as a painting or piece of sculpture. If so, it would be irrelevant whether or not Compère did not return to the French court in the autumn of 1495, as his music might have become part of the court repertory at the time of composition.

⁹⁴ It is not surprising that the four-voice repertory of Paris 1597 is largely anonymous, since that manuscript provides no attributions at all, but what is surprising is its paucity of concordances with other French or Burgundian sources; in fact, half of the songs are unica. The implications of this repertory are as yet far from clear.

⁹⁵ Uppsala, Universitetsbiblioteket, Vokalmusik i Handskrift MS 76a; inventory in Brown, ‘A “New” Chansonnier’, pp. 171–233. Copenhagen, Kongelige Bibliotek, MS Ny kgl. Samling 2° 1848; ed. Christofferson, P., ‘Musikhandskriftet Ny kgl. Samling 1848 2°, Det Kgl. Bibliotek, Kobenhavn’ (Ph.D. diss., University of Copenhagen, 1978)Google Scholar.

⁹⁶ If the disputed chanson Lourdault, lourdauit is by Compère, it would belong with the second or third phase. Lourdault has sometimes been awarded to Ninot le Petit on the assumption that his attribution (in Bologna Q 17) comes from the earlier source (Compère's being the Odhecaton), an assumption that is no longer valid (see n. 79); the late ascription to Josquin is usually discounted. The song bears no resemblance to the secure work of le Petit, nor do the sources seem to fit his usual pattern; I feel strongly that Ninot le Petit is not the composer of Lourdault or Mon seul plaisir (disputed with Josquin). The song does have some significant points of contact with the work of Compère, and the pattern of concordances is very supportive, but the style has some eccentricities that make one hesitant to give it to any known composer at this time. The most interesting question is actually not who composed Lourdault, but what is the significance of the attributions?

⁹⁷ Mary Winn established conclusively that Reveille toy does celebrate the Battle of Fornovo (5 July 1495), while noting that the inclusion of the Florentines among the enemy is unaccountable: see ‘Some Texts for Chansons by Loyset Compère’, p. 49. Despite its reference to a prisoner and a king who sails on the sea, Gentil patron has not yet been pinned down, and may or may not relate to the events of 1495.

⁹⁸ Kellman, H., ‘Josquin and the Courts of the Netherlands and France: The Evidence of the Sources’, Josquin des Prez: Proceedings of the International Josquin Festival-Conference, New York, 21–25 June 1971, ed. Lowinsky, E. E. and Blackburn, B.J. (London, 1976), pp. 181–216Google Scholar.

⁹⁹ Blackburn, B., ‘Josquin's Chansons: Ignored and Lost Sources’, Journal of the American Musicological Society, 29 (1976), pp. 54–5CrossRefGoogle Scholar. Blackburn's article shows how difficult is the problem of sources of Josquin's late chansons. The Burgundian sources are currently being reassessed by Flynn Warmington at Brandeis University. The important source of late Josquin is Vienna, Österreichische Nationalbibliothek, Cod. 18746.

¹⁰⁰ It would be pointless to attempt to locate post-1495 Compère left in anonymity, since there is no foundation on which to build; Compère was continually developing over the years, and it is unlikely that one could use traits found in his four-voice songs of 1490–5 to identify his style in either five- and six-voice songs or three-part folksong arrangements from 1495–1510. Even to limit our search to Paris 1597 would not overcome this problem, particularly since the origin of this anomalous repertory remains unknown, as discussed earlier.

¹⁰¹ The classic work on the transition period is Brown, H. M., ‘The Transformation of the Chanson at the End of the Fifteenth Century’, Report of the Tenth Congress of the International Musicological Society, Ljubljana, 1967 (Kassel, 1970), pp. 78–96Google Scholar. Brown outlined a picture of the stages involved, but the new availability of reliable source datings and provenances should enable us to begin soon to attach names and locations to Brown's stages.

¹⁰² Brown, , A Florentine Chansonnier, pp. 112–13Google Scholar.

¹⁰³ Atlas, , The Cappella Giulia Chansonnier, p. 107Google Scholar.

¹⁰⁴ Hewitt, , Odhecaton, pp. 62–3, 101–4Google Scholar.

¹⁰⁵ Hudson, ‘Compère’, 598; Hewitt makes a similar comment in Odhecaton, p. 102.

¹⁰⁶ According to his current biography, Ninot le Petit died in 1502, based on his disappearance from papal records and the transfer of a benefice in 1502; if so, he would have been an outstanding pioneer of the four-voice chanson. However, there are serious difficulties reconciling this theory with the pattern of the musical sources. Aside from the two disputed songs, both of which have styles and patterns of concordances incompatible with the secure work of Ninot, all but one of his songs occur after the Odhecaton. Neither of the two sources from Florence, where he is known to have been esteemed (Bologna q 17 and Florence, Biblioteca Nazionale Centrale, MS Magl. xix.178), has any of his secure works from around the putative date of his death. Most of his songs come from later Florentine sources, and he stands very close (thirteen of his fourteen secure songs) to the Strozzi Chansonnier (Florence, Biblioteca del Conservatorio Luigi Cherubini, MS Basevi 2442), which was compiled after 1518: see Brown, , ‘Chansons for the Pleasure of a Florentine Patrician’, Aspects of Medieval and Renaissance Music: A Birthday Offering to Gustave Reese, ed. La Rue, J. (New York, 1966), pp. 56–66Google Scholar. There is, moreover, evidence beyond that of the sources. Pierre Moulu's motet Mater floreat florescat is divided into two sections, the first in praise of composers who were prominent in the fifteenth century (mostly deceased), and the second listing figures active at Florence or the French royal court during the 1510s: ‘Nynot’ is listed with the second group! See Hewitt, , Odhecaton, p. 103Google Scholar. Clearly, Ninot le Petit was composing primarily during the sixteenth century, probably at Florence from 1512, when Medici rule was restored.