¹ These various corpora are Adam de la Halle's polyphonic songs (Everist, M., ‘The Polyphonic Rondeau c. 1300: Repertory and Context’, Early Music History, 15 (1996), pp. 59–96CrossRefGoogle Scholar and ‘“Souspirant en terre estrainge”: The Polyphonic Rondeau from Adam de la Halle to Guillaume de Machaut’, Early Music History, 26 (2007), pp. 1–42); motets with vernacular secular tenors (M. Hasselman, ‘The French Chanson in the Fourteenth Century’ (Ph.D. diss., University of California at Berkeley, 1970), i, pp. 97–136); and Everist, M., ‘Motet, French Tenors and the Polyphonic Chanson c. 1300’, Journal of Musicology, 24 (2007), pp. 365–406CrossRefGoogle Scholar), the eight rondeau-motets preserved in fr. 12615 (Everist, M., ‘The Rondeau-Motet: Paris and Artois in the Thirteenth Century’, Music & Letters, 69 (1988), pp. 1–22CrossRefGoogle Scholar), and a handful of isolated motets (Arlt, W., ‘Machaut in Context’, in Cerquiglini-Toulet, J. and Wilkins, N. (eds.), Guillaume de Machaut, 1300–2000 (Paris, 2002), pp. 147–62Google Scholar). Christopher Page targets a performance practice rather than a repertory: the hotchpotch of polyphony and courtly song through the performance of songs accompanied on an instrument in the learned Paris of Jerome of Moravia (Voices and Instruments of the Middle Ages: Instrumental Practice and Songs in France 1100–1300 (London and Melbourne, 1987), pp. 50–76). Conversely, the phenomenon of the penetration of motet principles into the repertory of trouvère song has been identified several times through the mensural notation of a number of courtly songs in a few chansonniers (Page, C., ‘Tradition and Innovation in BN fr. 146: The Background to the Ballades’, in Bent, M. et Wathey, A. (eds.), Fauvel Studies: Allegory, Chronicle, Music, and Image in Paris, Bibliothèque Nationale de France, MS Français 146 (Oxford, 1998), pp. 353–94, at 359–1Google Scholar, and O'Neill, M., Courtly Love Songs of Medieval France: Transmission and Style in the Trouvère Repertoire (Oxford, 2006), pp. 45–52Google Scholar) along with the ‘Parisian’ presentation of some chansonniers, especially fr. 846 (Everist, M., Polyphonic Music in Thirteenth-Century France: Aspects of Sources and Distribution (New York and London, 1989), pp. 200–3Google Scholar). The extent of the impact of this phenomenon should not be overemphasised, however, given the sporadic nature of these measured songs and the superficiality of the ‘Parisian’ features of fr. 846.

² Amongst the repertories mentioned in n. 1, only the rondeau-motets of chansonnier fr. 12615, presumably copied between 1253 and 1277 (see Everist, Polyphonic Music, pp. 171–87), can safely be dated earlier.

³ On this theory of reversal, see Earp, L., ‘Lyrics for Reading and Lyrics for Singing in Late Medieval France: The Development of the Dance Lyric from Adam de la Halle to Guillaume de Machaut’, in Baltzer, R., Cable, T. and Wimsatt, J. (eds.), The Union of Words and Music in Medieval Poetry (Austin, Tex., 1991), pp. 101–31Google Scholar. Page identifies the sporadic – and still mixed with popularisant registers – penetration of high registers in the polyphony as early as a few thirteenth-century motets (Voices & Instruments, pp. 74–6).

⁴ This distinction is most evident between great polyphonic sources copied in France, all coming from Paris (F, W2, Mo, Ba, Cl) and the main French sources of the trouvères, coming from Artois or Picardy (M, K, P, X, fr. 12615, f. fr. 844), Burgundy (O) or the north-east (U).

⁵ This dichotomy regarding measure is notably conceptualised by John Stevens in terms of a bipolarisation between musica metrica and musica ritmica (Words and Music in the Middle Ages (Cambridge, 1986); see parts II and III, esp. pp. 414–34), i.e. between the measured music of the dance-like songs and the unmeasured music of the high-register chansons. The interpretation of intermediate registers such as the pastourelle, on the other hand, turns out to be more problematic (pp. 471–6).

⁶ This difference in treatment is glaring in the chansonnier W at the very end of the thirteenth century. On this, see Stevens, J., ‘The Manuscript Presentation and Notation of Adam de la Halle's Courtly Chansons’, in Bent, I. (ed.), Source Materials and the Interpretation of Music: A Memorial Volume to Thurston Dart (London, 1981), pp. 29–64Google Scholar, and Page's commentary in ‘Tradition and Innovation’, pp. 367–8.

⁷ See Page, C. (ed., trans. and comm.), ‘Johannes de Grocheo on Secular Music: A Corrected Text and a New Translation’, Plainsong and Medieval Music, 2 (1993), pp. 17–41, at 19–21CrossRefGoogle Scholar.

⁸ There is scarcely anyone but Hans Tischler who has recently proposed, along the line of Pierre Aubry, Friedrich Ludwig and Jean Beck amongst others, a modal interpretation of trouvère song; see his editions: with Rosenberg, S., Chanter m'estuet: Songs of the Trouvères (London and Boston, 1981)Google Scholar and Trouvère Lyrics with Melodies: Complete Comparative Edition (Neuhausen-Stuttgart, 1997). Various alternative theories on rhythmic interpretation of this repertory have been suggested, particularly Hendrik van der Werf's free rhythm based on the flow and meaning of poems (The Chansons of the Troubadours and the Trouvères: A Study of the Melodies and their Relation to Poems (Utrecht, 1972), pp. 35–45), and Stevens's isosyllabic interpretation, proposing a balanced relationship between music and text, conducive to conveying the numerical balance of the phrases/lines of the songs (Words and Music, pp. 500–4). For a discussion of Grocheo's ‘non ita praecise mensuratam’ music, see van der Werf, H., ‘The Not-so-precisely Measured Music of the Middle Ages’, Performance Practice Review, 1 (1988), p. 42–60CrossRefGoogle Scholar and Stevens, Words and Music, pp. 429–34.

⁹ A typical example of this view is Gaston Raynaud's introduction to his edition of the poems of Mo motets, of which he notices, not without a certain pique, the irregularity and heterogeneity of language and the vagueness of registers (Recueil de motets français des XII^e et XIII^e siècles publiés d'après les manuscrits (Paris, 1881–3), i, pp. xvi–xx).

¹⁰ A few studies, however, discern the presence of repetitions or phrases with repetitive patterns in the upper voices of motets, but immediately discount their importance, observing their short-lived extent within the course of a work (Everist, M., French Motets in the Thirteenth Century: Music, Poetry and Genre (Cambridge, 1994), pp. 86 and 166Google Scholar) or in the course of the century (Tischler, H., The Style and Evolution of the Earliest Motets (to circa 1270) (Henryville, Ottawa and Binningen, 1985), i, pp. 62–4Google Scholar). For van der Werf, it is the very phraseological irregularity and the absence of any reiterated melodic material that account for the would-be erratic versification of the poems (The Chansons of the Troubadours and the Trouvères, pp. 71–2).

¹¹ Catherine Bradley, for example, has shown that phraseological and poetical regularity could occasionally be a desirable goal in the fabric of the early motet (‘The Earliest Motets: Musical Borrowing and Re-use’ (Ph.D. diss., University of Cambridge, 2010), pp. 49–98). On that question, see also G. Saint-Cricq, ‘Formes types dans le motet du xiii^e siècle: Étude d'un processus répétitif’ (Ph.D. diss., University of Southampton, 2010), i, pp. 91–5), <http://eprints.soton.ac.uk/203381/> (accessed Mar. 2013).

¹² See Everist, ‘The Rondeau-Motet’. The rondeau form is also identified by Sylvia Huot as the only case of formal resemblance between lyric forms and the motet, given the diversity of line lengths and the random distribution of rhymes in the latter (Allegorical Play in the Old French Motet: The Sacred and the Profane in Thirteenth-Century Polyphony (Stanford, Calif., 1997), p. 3).

¹³ This incompatibility of forms is clearly highlighted by van der Werf (The Chansons of the Troubadours and the Trouvères, pp. 71–2) and Page (‘Tradition and Innovation’, pp. 361–3) to justify the fundamental opposition of language between motet and chanson.

¹⁴ See Page's analysis of the poetic language of the motet, built on the remains of lower-register songs (Voices and Instruments, pp. 75–6). Mark Everist brings out the prevalence of the popularisants topoi of the rondet de carole according to Paul Zumthor's concept of ‘type-cadres’, which designates narrative formulae recurring in a given category of poems (French Motets, pp. 97–101). Beverly Evans identifies the literary topic of ‘bonne vie’ as being at the origin of most vernacular secular tenors in the motet (‘The Unity of Text and Music in the Late-Thirteenth Century French Motet: A Study of Selected Works from the Montpellier Manuscript’ (Ph.D. diss., University of Pennsylvania, 1983), pp. 96–129).

¹⁵ The near univocality of exchanges of the motet with choreographic songs was identified as early as 1935 by Handschin, Jacques (‘Die Modaltheorie und Carl Appels Ausgabe der Gesaenge von Bernart de Ventadorn’, Medium Aevum, 4 (1935), pp. 69–82)CrossRefGoogle Scholar, and copiously scrutinised through borrowings of styles, material, language or form (see e.g. Stevens, Words and Music, pp. 461–5; Everist, ‘The Rondeau-Motet’; and Page, C., The Owl and the Nightingale (London, 1989), pp. 118–20Google Scholar).

¹⁶ Gennrich identifies fifteen motetus parts matching trouvère songs in ‘Trouvèrelieder und Motettenrepertoire’, Zeitschrift für Musikwissenschaft, 9 (1926), pp. 8–39 (part I), and pp. 65–85 (part II). Nonetheless, four of them, preserved without tenor, are erroneously identified as motetus and simply consist of songs. A further motetus is devoid of any concordance as a song in sources. Conversely, a few further matches were unnoticed by Gennrich.

¹⁷ Grocheo's insistence on the opposition between measured polyphony and unmeasured monody, as well as the way the chansonnier W displays the difference of notation between the grands chants and the rest of Adam's repertory may thus be seen as the last-ditch struggle of a declining ethics (see Page, ‘Tradition and Innovation’, p. 368 and O'Neill, Courtly Love Songs, pp. 196–7).

¹⁸ Both in Table 1 and in this article, the number following vdB refers to the number of the refrain in Nico van den Boogaard's catalogue of refrains (Rondeaux et refrains du XII^e siècle au début du XIV^e: Collationnement, introductions et notes (Paris, 1969)). The number following RS refers to the number of the song in Gaston Raynaud's catalogue of songs revised by Spanke, Hans (G. Raynauds Bibliographie des altfranzösischen Liedes (Leiden, 1955)Google Scholar). The number following the incipit of motet parts designates the number of the voice in Friedrich Ludwig's catalogue of motets (Repertorium organorum recentioris et motetorum vetustissimi stili (Halle, 1910)).

¹⁹ The term ‘canso form’ is sometimes used by scholars to name this form (see e.g. Apel, W., ‘Canso’, Harvard Dictionary of Music, ed. Apel, (Cambridge, Mass., 1944; 2nd edn., 1970), p. 132Google Scholar) in a tradition initiated by Gennrich, (Grundriss einer Formenlehre des mittelalterlichen Liedes als Grundlage einer musikalischen Formenlehre des Liedes (Halle, 1932)Google Scholar). The use of this Occitan word, inferring that AAB form originates in the courtly song of the troubadours, nonetheless turns out to be problematic: on the one hand, neither documentary evidence nor Occitan song poems authenticates the use of this word in this particular context (Apel, W., ‘Rondeaux, Virelais, and Ballades in French 13th-Century Song’, Journal of the American Musicological Society, 7 (1954), pp. 121–30, at 124CrossRefGoogle Scholar and Perrin, R., ‘Some Notes on Troubadour Melodic Types’, Journal of the American Musicological Society, 9 (1956), pp. 12–18, at 12–16CrossRefGoogle Scholar). On the other hand, the repertory shows that Occitan courtly song breaks up in a variety of different forms (Aubrey, E., The Music of the Troubadours (Bloomington, Ind., 1996), pp. 146–7Google Scholar and H. van der Werf and S. Haynes, ‘Canso’, Grove Music Online, <http://www.oxfordmusiconline.com>, acc. 19 July 2012). Finally, the use of the term ‘canso’ in the context of the AAB form is a source of confusion with other forms and senses of the word (van der Werf, The Chansons of the Troubadours and the Trouvères, p. 64, n. 1), and most obviously when it designates the courtly register or genre of the Occitan song.

²⁰ On the Parisian origin of W2, see Everist, Polyphonic Music, pp. 97–110. For MüA, see Dittmer, L., Eine zentrale Quelle der Notre-Dame Musik: Faksimile, Wiederherstellung, Catalogue raisonné, Besprechung, und Transcriptionen (New York, 1959), pp. 12–20Google Scholar, and Everist, Polyphonic Music, pp. 137–49. For Mo, see Everist, ibid., pp. 110–34 and Wolinski, M., ‘The Compilation of the Montpellier Codex’, Early Music History, 11 (1992), pp. 263–301CrossRefGoogle Scholar.

²¹ Only Tischler identifies a group of forty-four motets exhibiting an AAB formal pattern and relates these works to the trouvère chanson (The Style and Evolution, pp. 68–71). Nevertheless, his group is defined without rigour and includes too many formal variants for the formal type to be clearly identified.

²² Including motets nos. 2, 6, 7, 8 and 11, which are preserved in the chansonniers fr. 12615, fr. 844, fr. 845 or fr. 846 (see Table 1).

²³ See the investigations relevant to the dating of the manuscripts which serve here as chronological landmarks: for W2, Everist, Polyphonic Music, pp. 97–110 ; for Mo, Y. Rokseth, Polyphonies du XIII^e siècle, 4 vols. (Paris, 1935–9), iv, pp. 21–35 and Everist, ibid., pp. 110–34 and French Motets, p. 11; for Tu, Everist, ‘Motets, French Tenors’, pp. 370–1, n. 18.

²⁴ In addition to Dante Alighieri's De vulgari eloquentia, the treatises dealing with the composition of vernacular poetry are Li livres dou tresor, written in the second half of the thirteenth century by Brunetto Latini, the unattributed Las leys d'amors, written in 1356 but offering a retrospective view on troubadour lyric poetry, and the Delle rime volgari trattato written in 1332 by Antonio da Tempo. They are the subject of a survey in Stevens's Word and Music (pp. 22–5 and 45–6). More closely related to the AAB motet corpus, Johannes de Grocheo's De musica lingers over vernacular chanson (Page, ‘Johannes de Grocheo’, pp. 22–35), but turns out to be more concerned with proposing a taxonomy of the ethos of the genres than with its technical composition. He nonetheless supplies useful examples.

²⁵ The original edition of the treatise, along with translations into Italian and English, has been put online by the Societa Dantesca Italiana (<http://www.danteonline.it>, acc. 24 July 2012). The passages relevant to the description of the music-poetic forms of chanson are to be found in Book 2, chs. 10–14.

²⁶ Forms without diesis correspond to the oda continua of the troubadours. Dante himself relates this structure to the Occitan repertory since he gives a song by Arnault Daniel as an example (Book 2, ch. 10, §2)

²⁷ Dante provides the general set of the formal conventions in Book 2, ch. 10.

²⁸ Whilst Karp, Theodore (‘Interrelationships between Poetic and Musical Form in Trouvère Song’, in Clinkscale, E. and Brooks, C. (eds.), A Musical Offering: Essays in Honor of Martin Bernstein (New York, 1977), pp. 137–61, at 147)Google Scholar and Robert Perrin (‘Some Notes’, p. 13) only state that the AAB pattern forms a vast majority in the trouvère repertory, van der Werf (The Chansons of the Troubadours and the Trouvères, p. 64) reckons its proportion to be 70% of the whole corpus.

²⁹ Dante first claims that each pes may be made up of several lines (Book 2, ch. 11), in even or odd number (ch. 13, §9), then twice evokes the versification related to the three-line pes or versus (ch. 12, §§9–10 and ch. 13, §10). He does not explicitly mention the two-line pes, but we understand that the latter stands as a norm, as he treats the three-line pes as a particular and difficult case. The one-line pes is not specifically mentioned in the treatise, but is theoretically fully possible.

³⁰ Here and in the column ‘form’ of Table 1, as well as in the musical analyses, the symbol * designates the differentiated or ‘ouvert-clos’ cadences at phrase endings, whereas the symbol ′ indicates a more important variant, as in the block BB′ of motet no. 1 (Example 2). In all the transcriptions, the sign / is the translation of a punctus placed within the poem. See also n. 77.

³¹ In order to form this control corpus, fifty-eight songs were chosen at random, provided they exhibited an AAB form. They include twenty-two grands chants, thirteen pastourelles and twenty-three songs of various genres. The detailed list is given in Saint-Cricq, ‘Formes types’, i, p. 123, n. 102). The analyses of these songs have been made after the editions of Tischler and Rosenberg, Chanter m'estuet, and Tischler, Trouvère Lyrics with Melodies, and are then dependent on the choice of the manuscript witnesses in these editions.

³² On the question of the terminology ‘ouvert-clos’ and its relevance for the thirteenth-century music, see Saint-Cricq, ‘Formes types’, i, pp. 52–6.

³³ This aspect is precisely treated both in Latini's essay, the Ley d'amors (see Stevens, Words and Music, pp. 22–5), and Grocheo's De musica (Page, ‘Johannes de Grocheo’, p. 28). Dante broaches that question in Book 2, ch. 13.

³⁴ Fascicle VI of Mo has been used for comparison, being chronologically close and structurally comparable to our corpus with its two-voice French motets from the third quarter of the century. The total of sixty-six motets was obtaineded by subtracting the motets also appearing in the AAB corpus.

³⁵ Apel, (‘Imitation in the Thirteenth and Fourteenth Centuries’, in Essays in Music in Honor of Archibald Thompson Davison by his Associates (Cambridge, 1957), pp. 25–38)Google Scholar and Harbison, Denis (‘Imitation in the Early Motet’, Music & Letters, 45 (1964), pp. 359–68)CrossRefGoogle Scholar observe that a great deal of melodic imitation between the upper parts of motets is led by textual imitation. Everist (French Motets, pp. 171–2) identifies melodic repetition linked to that of the poem as an important mode of melodic repetition in the motet.

³⁶ On the issue of poetic treatment in both French and Latin repetitive motets, see Saint-Cricq, ‘Formes types’, i, pp. 63–70.

³⁷ See Book 2, ch. 11, §13 and ch. 12, §9.

³⁸ The exception is motet no. 3, but it should be specified that both pedes have the same overall length, the discrepancy appearing between equivalent elements.

³⁹ Book 2, ch. 12, §2.

⁴⁰ Book 2, ch. 11, §13: ‘In their mutual relationship the pedes should be equal, in both number of lines and number of syllables, as well as in their organisation; for otherwise it will not be possible to repeat their melody exactly. And I hold that this principle is also to be observed in the versus.’ Dante reiterates this explanation in ch. 12, §9.

⁴¹ Only eleven AAB motets are written in mensural notation in at least one source, but the synchronisation between motetus and tenor, as well as other various features in motets nos. 1 and 10, secures a mode 1 or 2 in the motetus. Motets nos. 7, 8 and 9 are more ambiguous in that respect and are not taken in account, but let us note that they are very likely to exhibit the same musical length between their A phrases, given the isosyllabism and the strict melodic repetition in the AA block.

⁴² Dante tackles the issue of rhyme regarding the relationship between the blocks (Book II, ch. 13, §7), but he eventually remains evasive on that point, only remarking that ‘certain poets’ shift rhymes after the pedes, while ‘other poets’ do not.

⁴³ In order to draw a valid comparison of the lengths of the sections in the unmeasured corpus of the chanson with the measurable corpus of the motet, the lengths are calculated by number of syllables in both repertories. Note that in the cases of the measurable motets, the ratio of AA is almost the same in terms of musical perfections (41, 8%) as in terms of number of syllables.

⁴⁴ Stevens (Words and Music, p. 461, n. 3) has already briefly broached this issue: when he considers Gennrich's fifteen parts common to the motet and chanson, he takes a one-sided position by claiming, on the basis of style, that they most certainly come from polyphony rather than from the chanson repertory. Amongst Gennrich's – and Stevens's – fifteen parts we find four out of the five motets under examination here; motet no. 1 was unnoticed by Gennrich, probably because the motetus and the song carry different texts.

⁴⁵ See the sources of these songs in Table 1, column ‘sources’. The contrafactum between the motet and the songs was discovered by the musical ensemble Diabolus in Musica (dir. Antoine Guerber) in its album Rosarius (Studio SM, 2000).

⁴⁶ The text of the motetus and that of Isaiah 11:1–11, from which the Marian tenor comes, include the same vegetal metaphors and theme of peace and reconciliation within divine love.

⁴⁷ The responsory SANCTE GERMANE (O27) is used as tenor in one three-voice organum copied in W1 (fol. 5^r), F (fol. 34^v), W2 (fol. 10^r) and Mo (fol. 13^r) and one two-voice organum copied in F (fol. 82^r). These organa use the conventional melisma as found in thirteenth-century Parisian chant sources, which radically differs from the tenor of motet no. 7.

⁴⁸ The tenor word varies between the different sources, and the latter never have the same melody for this melisma, which seems erroneous in all versions. For a detailed analysis of this tenor, see Saint-Cricq, ‘Formes types’, i, pp. 202–4.

⁴⁹ Payne, Thomas (‘Associa tecum in patria: A Newly Identified Organum Trope by Philip the Chancellor’, Journal of the American Musicological Society, 39 (1986), pp. 233–54, at 252–3)CrossRefGoogle Scholar notices that the name GERMANE and melisma O27 had a generic function in the thirteenth century: the melisma could be used for all saints, and GERMANE was equivalent to the undetermined N, intended to be substituted by the name of the saint to be celebrated. Husmann, Heinrich (‘The Enlargement of the Magnus Liber Organi and the Paris Churches of St-Germain-l'Auxerrois and Ste-Geneviève-du-Mont’, Journal of the American Musicological Society, 16 (1963), pp. 176–203, at 193–5)CrossRefGoogle Scholar specifies that one could write N, GERMANE or MARCELLUS in some Parisian breviaries. This certainly explains why, in the W2 version of this motet, the scribe contents himself to write SANCTE under the tenor.

⁵⁰ Certainly revealing is the fact that all four versions of the same song RS 1188 Qui bien aime, a tart oublie, which uses the citation at its incipit, show very divergent melodies for this proverb. See the details in Saint-Cricq, ‘Formes Types’, i, pp. 194–5.

⁵¹ The different sections of chansonnier U are dated by Robert Lug from 1231 for the oldest to c. 1260 for the latest, with an intermediary section c. 1232–3 (see ‘Katharer und Waldenser in Metz: Zur Herkunft der ältesten Sammlung von Trobador-Liedern (1231)’, in Rieger, A. (ed.),Okzitanistik, Altokzitanistik und Provenzalistik: Geschichte und Auftrag einer europäischen Philologie (Frankfurt, 2000), pp. 249–74Google Scholar and ‘Politique et littérature à Metz autour de la guerre des Amis (1231–1234): Le témoignage du Chansonnier de Saint-Germain-des-Prés’, in Chazan, M. and Regalado, N. (eds.), Lettres, musique et société en Lorraine médiévale (Geneva, 2012), pp. 451–83)Google Scholar. The song RS 505 is included in the c. 1260 section.

⁵² This cycle extends from motet no. 17 (fol. 18^v) to no. 23 (fol. 24^r).

⁵³ See the analysis in Saint-Cricq, ‘Formes types’, i, pp. 261–3.

⁵⁴ For that matter, this part has so many songlike characteristics that, paradoxically, this challenging case is the only one in Gennrich's list that John Stevens endorses as coming from a chanson rather than polyphony (Words and Music, p. 461, n. 3).

⁵⁵ See the analysis in Saint-Cricq, ‘Forme types’, i, pp. 198–9.

⁵⁶ For that matter, Gennrich was so convinced that the motetus was a citation of a song that, despite the lack of concordance, he added this part to his list, precisely on the twofold ground of its ‘Liedform’ and the irregular pattern of the tenor (‘Trouvèrelieder’, pt. II, pp. 78–80).

⁵⁷ A digitisation of the chansonnier is available online on the site of the BnF: <http://gallica.bnf.fr/ark:/12148/btv1b6000950p> (accessed 25 July 2012).

⁵⁸ For the repertory of the motet, it is certainly preferable to use the term ‘short citation’ rather than the conventional ‘refrain’, owing to the considerable ambiguity of the latter term in the context of the motet, and the discrepancy with its meaning in the chanson repertory. On that issue, see Saint-Cricq, ‘Formes types’, i, pp. 149–53.

⁵⁹ Motets nos. 7 and 14, which also include short citations, are disregarded here because these citations are in fact already incorporated in the songs quoted by these motets as their upper part.

⁶⁰ On the issue of the formal integration of the citations in the motet, see Rokseth, Polyphonies, iv, pp. 209–12; Evans, ‘The Unity of Text and Music’, chs. 9, 10 and 11 and Everist, French Motets (1994), pp. 158–66. Short citations may have a stronger formal impact on motets, yet in rather idiosyncratic and scattered works such as motets entés and cento motets. For discussions of the motet enté, see Everist, French Motets, pp. 75–89; Peraino, J., ‘Monophonic Motets : Sampling and Grafting in the Middle Ages’, Musical Quarterly, 85 (2001), pp. 644–80CrossRefGoogle Scholar; Butterfield, A., ‘Enté : A Survey and Reassessment of the Term in Thirteenth and Fourteenth-Century Music and Poetry’ , Early Music History, 22 (2003), pp. 67–101CrossRefGoogle Scholar and Saint-Cricq, ‘Formes types’, i, pp. 167–86. For discussions of the cento motet, see Evans, , ‘The Textual Function of the Refrain Cento in a Thirteenth-Century French Motet’, Music & Letters, 71 (1990), pp. 187–97CrossRefGoogle Scholar and Everist, French Motets, pp. 109–25.

⁶¹ See e.g. Anne Ibos-Augé's analysis of some refrain songs from Gautier de Coinci's Miracles de Nostre-Dame: ‘La Fonction des insertions lyriques dans des œuvres narratives et didactiques aux XIIIème et XIVème siècles’ (Ph.D. diss., University of Bordeaux, 2000), i, pp. 237–41.

⁶² For an examination of the formal impact of citation in motets nos. 3 and 12, see Saint-Cricq, ‘Formes types’, i, pp. 158 and 159–60 respectively.

⁶³ vdB 1605 is cited among others in the songs RS 1660 Pour li servir en bonne foi (with music in M, fol. 156^v, without music in fr. 12615, fol. 170^v) and RS 1975 Pour ce que mes cuers souffre grant doulour (C, fols. 183^v and 186^v, without music). The pre-existence of this citation in the song RS 1660 is suggested by its transposition to the upper fifth in the motet, which can be accounted for by the necessity to adhere to the polyphonic complex. Above all, the hypermetric four-perfection and seven-syllable norm are found throughout the motetus of motet no. 6, with the notable exception of the second phrase citation that stretches out over eight syllables, which is the standard in song RS 1660.

⁶⁴ The translation of the poem of the motetus is that of Tischler's edition, The Montpellier Codex, iv (translations by S. Stakel and C. Relihan) (Madison, Wis., 1985), p. 78.

⁶⁵ The assumption that the citation of vdB 539 pre-exists the motet is based on manuscript evidence. This citation is found with music and text in various sources prior to the sixth fascicle of Mo, where motet no. 4 is preserved, and notably in the chanson de rencontre RS 575 in question, recorded in chansonnier U (fol. 56^v). The section of U relevant to RS 575 might be dated as early as 1231 (see above, n. 50). The first citation at the beginning of the motetus occurs in the pastourelle RS 1365 Pensis chief enclin, but applies only to the text. This pastourelle is also preserved in a source pre-existing Mo VI: chansonnier M (fol. 102^v), which can be dated between 1253 and 1277 (see Everist, Polyphonic Music, pp. 171–87).

⁶⁶ Moreover, the motet and the song share a significant amount of musical material at different points: see Saint-Cricq, ‘Formes types’, i, pp. 165–6 and Saint-Cricq, ‘Transmitting a Compositional Process: Two Thirteenth-Century Motets’, paper read at the conference ‘The Gothic Revolution: Music in Western Europe, 1100–1300’, Princeton University, 4–6 November 2011, to be published in Musica Disciplina.

⁶⁷ Huot demonstrates the intertextual function of citations, allowing the works to be confronted dialogically (Allegorical Play, esp. pp. 56–84). Ardis Butterfield approaches the refrains as autonomous entities that mediate between genres, works and authors, especially within a given cultural context (Poetry and Music in Medieval France from Jean Renart to Guillaume de Machaut (Cambridge, 2002), esp. pp. 75–121 and 133–50). Both the intertextual potential and the mediating function of short citations between the motet and other repertories involving refrains are explored in Clark, S., ‘“S'en dirai chançonete”: Hearing Text and Music in a Medieval Motet’, Plainsong and Medieval Music, 16 (2007), pp. 31–59CrossRefGoogle Scholar and Saltzstein, J., ‘Relocating the Thirteenth-Century Refrain: Intertextuality, Authority and Origins’, Journal of the Royal Musical Association, 135 (2010), pp. 245–79CrossRefGoogle Scholar. The roles and modes of transmission of the citations are also scrutinised in Butterfield, , ‘Repetition and Variation in the Thirteenth-Century Refrain’, Journal of the Royal Musical Association, 116 (1991), pp. 1–23CrossRefGoogle Scholar and Peraino, , ‘“Et pui conmencha a canter”: Refrains, Motets and Melody in the Thirteenth-Century Narrative Renart le Nouvel’, Plainsong and Medieval Music, 6 (1997) pp. 1–16CrossRefGoogle Scholar.

⁶⁸ This chanson texture, found in a portion of the late-century repertory of the motet, was spotted by Ernest Sanders, for whom the most striking aspect was a dissolution of the tenor modality, the rhythmic flow of which tends to keep pace with that of the upper voices (‘The Medieval Motet’, in Arlt, W.et al. (eds.), Gattungen der Musik in Einzeldarstellungen: Gedenkschrift Leo Schrade (Bern, 1973), pp. 497–583, esp. 534–5Google Scholar). Everist (French Motets, p. 174 and ‘Motet, French Tenors’) also perceives a levelling and cohesion of parts around a formal project based on the repetitive forms of secular tenors.

⁶⁹ Amongst the liturgical melismas used in Parisian polyphony, only the Alleluya V Letabitur iustus in Domino (M49) and the Alleluya V Domine in virtute (M66) include the word LETABITUR, but the melodies do not match the tenor of the motet. For that matter, this tenor appears to be a source of scribal confusion in all the manuscript sources of this work: named LETABITUR in W2 and Mo, LETABIMUR in fr. 844 and fr. 12615, or even LABIMUR in Her, this tenor is moreover the subject of countless variants, lacunae and further discrepancies in all these versions. Here again, the word LETABITUR – or LETABIMUR – (‘he/we shall rejoice’) might have been chosen on intertextual grounds, since the motetus text promises the true lover to have his ‘joie doubler’ (‘joy redoubled’). The word LABIMUR (‘we shall sink’) cannot be explained but by an amusing scribal confusion.

⁷⁰ The constraints of the tenor are precisely invoked by Page in order to explain the difficulty in building repetitive melodic patterns in the upper parts of the motet (‘Tradition and Innovation’, p. 369). Everist retains melodic repetition amongst the possible compositional modes at work in the motet, but he also specifies that it is dependent on the tenor pattern and, as such, limited in the whole repertory (French Motets, pp. 170–3). Ronald Voogt claims that repetitive structures are comparatively much more widely spread in conductus than in motet because conductus is an entirely composed genre (‘Repetition and Structure in the Three- and Four-Part Conductus of the Notre-Dame School’ (Ph.D. diss., Ohio State University, 1977), pp. 278–9). Sanders (‘The Medieval Motet’, pp. 499–500) and Rebecca Baltzer (‘Notation, Rhythm and Style in the Two-Voice Notre-Dame Clausula’ (Ph.D. diss., Boston University, 1974), pp. 58–61) point out the same causal link between tenor structure and duplum repetitions in the repertory of the clausula.

⁷¹ This irregularity is found in motet no. 10, where contrapuntal necessity forces the author to modify one pitch between A and its repetition on perfection 9 of the motetus (see Saint-Cricq, ‘Formes types’, i, p. 117 and Saint-Cricq, ‘Transmitting a Compositional Process’).

⁷² In motet no. 3, the author superimposes two AAB forms while overlapping them. See detailed analysis in Saint-Cricq, ‘Formes types’, i, pp. 44–5.

⁷³ Amongst the various tendencies he identifies in the motet repertory of the second half of the thirteenth century, Sanders detects, nonetheless without clearly referring to a significant corpus, two-voice motets with invented or arranged tenors designed to adhere to the motetus. For Sanders, this feature suggests that they should be named ‘duets’ or ‘accompanied songs’ rather than ‘motets’ (‘The Medieval Motet’, p. 552).

⁷⁴ See detailed analysis of these works and of the arrangement of their tenors in Everist, ‘The Rondeau-Motet’.

⁷⁵ In the polyphonic rondeau c. 1300, the middle voice turns out to be the principal voice: whenever a monophonic concordance appears in a part of a polyphonic work, it is to be found in the middle voice; and whenever a polyphonic rondeau refrain is found elsewhere with music, the music is always that of the motetus. On that question, see Everist, ‘The Polyphonic Rondeau’, pp. 84–5.

⁷⁶ Odington, Walter, Summa de speculatione musicae, ed. Hammond, F. (Rome, 1970), p. 143Google Scholar: ‘the middle part should receive great attention, so that it is ornamented in itself.’ Odington's treatise was presumably written around 1300, by the time of Adam's polyphonic songs. Even if little is known on the biography of this English theorist, the music he describes and the examples he provides in his treatise show that he was familiar with French music.

⁷⁷ The role of palaeographic signs related to phraseology, i.e. the tractus for the music and the punctus for the text, also have an important role in the conception and notation of the form. See Saint-Cricq, ‘Formes types’, i, pp. 71–8.

⁷⁸ Synchronism and hypermetry on the other hand must be moderated in three-voice motets. In motet no. 12, these characteristics are clear in the relationship between tenor and motetus, but the triplum retains the traditional overlapping features. Motet no. 15 shows a more subtle reading of the form by the polyphonic whole, while in motets nos. 13, 14 and 16, asymmetry and overlapping of parts are still active processes. See the analysis of these works in Saint-Cricq, ‘Formes types’, i, pp. 191–9.

⁷⁹ This tabulation only includes motets whose AA block follows the sixteen-perfection norm. Motet no. 2, which provides two different versions of the tenor melisma, is counted twice.

⁸⁰ Motets nos. 8 and 9 are not counted because the former is preserved without tenor, and the latter lacks the last motetus phrase. The end of motet no. 1 is problematic, but the last pitch of the motetus testifies that the harmony would anyway have been different from that heard at the end of the A phrases.

⁸¹ The term ‘structural tones’ is used by van der Werf to designate the pitches around which the formal types of trouvère chansons are articulated (The Chansons of the Troubadours and the Trouvères, p. 65). A comparison between detailed cadential schemes in the AAB motets and in the corpus of songs shows the proximity of these patterns (Saint-Cricq, ‘Formes types’, i, pp. 126–7).

⁸² Amidst other attempts, Mathiassen suggests a tonal system based on Sachs's theory of thirds (The Style of the Early Motet (c.1200–1250): An Investigation of the Old Corpus of the Montpellier Manuscript, trans. J. M. Stochholm (Copenhagen, 1966), pp. 56–70). Tischler provides a colourful catalogue of ‘harmonic styles’ of the motet (‘The Evolution of the Harmonic Style in the Notre-Dame Motet’, Acta Musicologica, 28 (1956), pp. 87–95). The analysis of the IN SECULUM motets by Klaus Hofmann suggests that polarities in the motet are organised according the structural tones of the tenor (Untersuchungen zum Kompositionstechnik der Motteten im 13. Jahrhundert durchgeführt an den Motteten mit dem Tenor In Seculum (Stuttgart, 1972)). Pesce, Dolores (‘A Case for Coherent Pitch Organization in the Thirteenth-Century Motet’, Music Analysis, 9 (1990), pp. 287–318)CrossRefGoogle Scholar postulates a tendency of the motet to enhance the structural tones supplied by the tenor by such means as prolonged sonorities or ‘directed progressions’. One can finally cite the exotic endeavours of Schenkerian analysis applied to the motet by Saltzer, Felix (‘Tonality in Early Medieval Polyphony: Towards a History of Tonality’, Music Forum, 1 (1967), pp. 35–98)Google Scholar.

⁸³ This motet, moreover, includes a fascinating complex play of reiterated melodic cells, transposed and interlocked according to various modes, which structure and link motetus and tenor. See the analysis in Saint-Cricq, ‘Formes types’, i, pp. 104–5.

⁸⁴ In his catalogue of conductus (‘Notre-Dame and Related Conductus: A Catalogue Raisonné’, Miscellanea Musicologica, 6 (1972), pp. 153–229 (part 1); 7 (1975), pp. 1–81 (part 2)), Gordon Anderson identifies eleven concordances between Parisian polyphonic conducti and courtly songs; among them eight have an AAB form and a further one an AABB form. Anderson also makes an inventory of all types of songs cited in the conductus repertory, with a total of forty-six French songs, nine Occitan, four English, twelve Germanic, one cantiga and twenty rondeau refrains. See also the new online comprehensive catalogue of the conductus Cantum pulcriorem invenire, created and hosted by the University of Southampton (<http://catalogue.conductus.ac.uk/>), acc. Mar. 2013.

⁸⁵ In his investigation of repetitive structures in the three- and four-part conductus, Voogt brings to light several cases of conducti including a trouvère song in its parts, but also points to the influence of the forms of trouvère chansons in conducti even without proven citations (Repetition and Structure, p. 70). The link between AAB form and the tenors in contrafacta coming from the vernacular repertory is also noticed by Falck, Robert (The Notre Dame Conductus: A Study of the Repertory (Henryville, Ottawa, et Binningen, 1981), pp. 28–33)Google Scholar.

⁸⁶ Examples of trouvère chansons otherwise found as clausulae are the Marian song Quant voi le dous tens venir (RS 1485) or the pastourelle Pour conforter mon courage (RS 19). See also note 88 below.

⁸⁷ A well-known case of contrafacta between Parisian genres and the chanson de trouvère is Gautier de Coinci's Miracles de Nostre-Dame of c. 1220. Several sequences and parts of polyphonic conducti and motets are used in order to build songs. In addition, Everist notes a great permeability in style and forms between Miracles interpolations and Notre-Dame genres (Polyphonic Music, p. 43–4).

⁸⁸ The question of the provenance of the material – in particular short citations – circulating between song and polyphony remains a controversial issue, still in need of further investigation. The traditional view of the origins of the citation in the song tradition, persistent from pioneers such as Bartsch or Jeanroy until more recent scholarship (for an account of this pervasive view, see Saltzstein, ‘Relocating’, pp. 245–8) has rightly been challenged and tempered by Everist (French Motets, pp. 54–71), Clark (‘“S'en dirai chançonete”’) and Saltzstein (‘Relocating’, pp. 252–3), asserting that short citations could originate – and even remained confined – in polyphony, especially the motet. Everist (pp. 66–71) and Clark (pp. 46–7) moreover leave open the possibility for a refrain to be derived from a clausula. On the other hand, Bradley has recently demonstrated that a refrain could originate in the chanson and make its way into liturgical polyphony via the transcription of a motet as a clausula (‘The Earliest Motets: Musical Borrowing and Re-use’, p. 168–230; see also her article ‘Contrafacta and Transcribed Motets: Vernacular Influences on Latin Motets and Clausulae in the Florence Manuscript’ above). In strong resonance with this complex issue are notably motets nos. 2 and 3 in the AAB corpus, which are simultaneously clausula-related motets and carry a pedes-cum-cauda form including a short citation for the latter.

⁸⁹ These twenty-seven works are the subject of the investigation in Saint-Cricq, ‘Formes types’, and the detailed corpus is given in vol. 2, pp. 5–7.

⁹⁰ In addition to the works of fr. 12615, further rondeau-motets are included in Mo, fifth fascicle (nos. 138, 169), seventh fascicle (nos. 260, 265, 271), and eighth fascicle (nos. 312, 324), as well as in Fauv (fol. 42^v).

⁹¹ This collection is much more restricted, but also more concentrated in sources, time and influences: these motets are found in the seventh fascicle of Mo (nos. 256, 290, 295), three are recorded in the eighth fascicle (nos. 309, 333, 323), and a further work is preserved in Fauv (fol. 9^v). All seven works exhibit secular tenors; amongst them six carry the ABA form.

⁹² The number of parts having a formal type is compared to the overall number of 1,195 parts of the motet repertory according to Ludwig's catalogue (Repertorium, ii).

⁹³ Already identified by Ludwig (Repertorium, i/1, p. 338) and Handschin, (‘Über Voraussetzungen, sowie Früh- und Hochblüte der mittelalterlichen Mehrstimmigkeit’, Schweizerisches Jahrbuch für Musikwissenschaft, 2 (1927), pp. 5–42, at 31)Google Scholar as a fourteenth-century-style polyphonic or accompanied song rather than a motet, motet no. 11 is characterised as a ‘strophic composition in ballade form’ by Sanders (‘The Medieval Motet’, p. 552), compared to a polyphonic ballade by Everist (‘Geographies of Polyphonic Song’, Paper read at the Annual Meeting of the American Musicological Society, Nashville, 6 Nov. 2008, unpublished), and Arlt, but Arlt above all targets motet no. 7, in which he thinks he identifies ‘la disposition formelle musico-textuelle de la future ballade’ (‘Machaut in Context’, p. 154, n. 18). Motet no. 7 is also retained by Everist as being one of the ‘sources for the earlier polyphonic ballade’ (‘“Souspirant”’, pp. 4–5, n. 11).

⁹⁴ Between the rondeau-motets of fr. 12615 and Machaut's first polyphonic rondeaux, we find about sixty examples of rondeau-form polyphonic settings (see the detailed list in Saint-Cricq, ‘Formes types’, i, pp. 229–30).

⁹⁵ Manuscript fr. 1586, which includes Machaut's first polyphonic ballades, was presumably compiled in the early 1350s (F. Avril, ‘Les Manuscrits enluminés de Guillaume de Machaut: Essai de chronologie’, in Guillaume de Machaut: Colloque – Table ronde, Reims (19–22 Avril 1978) (Paris, 1982), pp. 117–33). However, the composition of these works may be earlier: Reaney, Gilbert (‘A Chronology of the Ballades, Rondeaux and Virelais set to Music by Guillaume de Machaut’, Musica Disciplina, 6 (1952), pp. 33–8)Google Scholar and Earp (‘Lyrics for Reading’, p. 111) suggest the 1340s. On the basis of a refrain citation play between two ballades by Machaut and Jehan de le Mote's Li regret Guillaume, David Maw assumes that Machaut is as likely to have borrowed this material from Jehan as the opposite. The former possibility would consequently date the ballades before 1339, i.e. the date indicated for Li regret Guillaume in the source Paris, BnF, n. acq. fr. 7514 (‘“Trespasser mesure”: Meter in Machaut's Polyphonic Songs’, Journal of Musicology, 21 (2004), pp. 114–16).

⁹⁶ There exist only two extant collections of monophonic ballades before Machaut, both preserved in Fauv: Jehan de Lescurel's fifteen works and six further ballades interpolated in the Roman de Fauvel. Other collections are purely poetic settings, making up a corpus of sixty-eight works, coming from Jean Acart de Hesdin's La prise amoureuse, Jehan de le Mote's Li regret Guillaume and Parfait du paon, the unattributed La dame à la licorne, and the six ‘ballades mythologiques’ resulting from an exchange of esoteric poems between de le Mote, Philippe de Vitry and Jehan Campion. For the details of these repertories (sources, dates and editions), see Saint-Cricq, ‘Formes types’, i, p. 231, n. 12).

⁹⁷ Noticing the lack of musical sources for the early ballade, Earp consequently accounts for the growth of the genre as being the fruit of poetic procedures only. He judiciously adds that this musical vacuum with which we are confronted certainly does not result from the loss of sources. On the one hand, the layout of the manuscripts for the first poetic ballades did not allow the writing of any music alongside the poems; on the other hand, the writing of the first musical polyphonic ballades in fr. 1586 demonstrates difficulties in layout faced by scribes obviously confronting a new notational challenge (‘Lyrics for Reading’, pp. 110–13).