¹ In MS C they occur separately in the disordered section which Ursula Günther has termed CII; see the information given in Earp, L., Guillaume de Machaut: A Guide to Research (New York and London, 1995), p. 78Google Scholar. They are also non-adjacent in MS E. The manuscript sigla used here are the standard sigla as detailed in Earp, Guillaume de Machaut.

² MS C contains twenty-four balades in total which correspond to the first twenty-four in the modern numbering. The authorial ordering for these, however, is contained in the prescriptive index to MS A. The main departure from the modern order (based on a different source, Vg) is that the canonic balade Sans cuer / Amis / Dame (B17) is placed ninth. Further evidence for Machaut's interest in order can be found in the Prologue (contained in Guillaume de Machaut: poésies lyriques, ed. Chichmaref, V. (Paris, [1909]), i, pp. 1–13)Google Scholar and is manifest in his work, especially as contained in those manuscripts which he is thought to have overseen; see Huot, S., From Song to Book: The Poetics of Writing in Old French Lyric and Lyrical Narrative Poetry (Ithaca, 1987), pp. 211–301Google Scholar.

³ The topic of the actions of Fortune, and her personification, are traditional and had a long history by the fourteenth century; see Patch, H. R., ‘Fortuna in Old French Literature’, Smith College Studies in Modern Languages, 4/4 (1923), pp. 1–45Google Scholar and idem, The Goddess Fortuna in Medieval Literature (Cambridge, Mass., 1927).

⁴ The Machaut texts are transcribed from the music section of the earliest source, MS C. I am very grateful to Kevin Brownlee for helping me with this translation in the early stages of this study.

⁵ Johnson, L. W., Poets as Players: Theme and Variation in Late Medieval French Poetry (Stanford, Calif., 1990), pp. 41–54Google Scholar. Of the musical balades he also discusses De fortune (B23) and De toutes flours (B31).

⁶ Ibid., p. 43.

⁷ C. Heppleston's unpublished paper ‘The Louange des Dames: Misreading a Poem and Reading the Collection’ makes a similar point against Johnson's view, although Heppleston is arguing from the point of view of B22's place in the Louange (hereafter Lo and a standard number taken from Earp, Guillaume de Machaut), where it appears as Lo188, and its connections with Douce dame savoir de puis noir (Lo187) and Helas pour ce que fortune mest dure (Lo189). I am grateful to Charles Heppleston for allowing me access to this material.

⁸ Charles Heppleston (personal communication) has pointed out that ‘The 1st line is ambiguous; syntactically it could read either “je doy me plaindre de F. et je doy me loer” which would mean “I must blame F. and rely on her” or “je doy plaindre de F. et je doy la loer” – “I must blame F. and praise her”’.

⁹ Again I am grateful to Kevin Brownlee, and also to Charles Heppleston, for assistance in the initial translation and interpretation of this text.

¹⁰ Nature does not feature in the notated balades preceding B22. ‘Desconfiture’ previously appears only in De desconfort (B8). That ‘desconfiture’ is a word of some significance for Machaut is shown by the extremely dissonant treatment of it at the end of Motet 15; see Bent, M., ‘Deception, Exegesis and Sounding Number in Machaut's Motet 15’, Early Music History, 10 (1991), pp. 15–27CrossRefGoogle Scholar.

¹¹ Of the musical balades preceding B22, the sententious announcement of opinion occurs (in a slightly different form) only in Amours ne fait (B1), Pour ce que tons (B12), Esperance (B13) and De petit po (B18).

¹² Johnson comments that ‘we are not told exactly what her “desconfiture'” is, but one can only assume that it is some amorous misfortune, perhaps her lover's absence’ (45). That it is more than this is, I think, connoted by the relationship between the two texts, which implies a change (reported, if not real) in the man's love.

¹³ Johnson, , Poets as Players, p. 44 (emphasis added)Google Scholar.

¹⁴ The viewpoints of Lo 187–95 include a man who has heard that he is refused but does not wish to believe it (Douce dame savoir ne puis noir; Lo187), a man abandoned by his lady and his friends (the rondeau Helas pour ce que fortune mest dure; Lo189), a balade which vents anger on ‘tongue’, a synecdochal (but feminine) personification which enables a depersonalised invective (Langue poignant aspre amere et ague; Lo190), a man who suffers nobly, is of good cheer and pledges loyalty, come what may (La grant doucour de vostre biaute fine; Lo191), a man who councils that one ought not to believe lightly the false reports, especially that his lady should not credit those circulating about him, with the implication that if she were to do so she would be unworthy (On ne puet riens savoir si proprement; Lo192), a man who laments his lady's perfidy in a poem very similar in terms of rhyme words and diction to both B22 and B23 (Il ne mest pas tant dou mal que jendure; Lo193), and a woman suffering because she is far from her sweet loyal lover to whom she pledges loyalty and to put aside all other men (Il nest dolour desconfors ne tristece; Lo194).

¹⁵ MS G lacks the larger sequence from the Loange within which B22 and B23 normally occur. MS E transmits the sequence but omits B22 and B23 from the Loange. This posthumous source also has its own particular version of order within all the music sections in which B22 and B23 are no longer adjacent.

¹⁶ This comparison is made by Machaut more directly in the Voir Dit, where he compares his lady to Fortune, and is then himself compared to a different (‘pagan’) type of Fortune by a messenger who berates him for believing gossip too readily (‘tout par legierement croire’). The refrain in the Voir Dit ‘Et tout par legierement croire’ is similar to that of Lo192, ‘De legier croire encontre son ami’. See Machaut, , Le Livre dou Voir Dit, trans. and ed. Leech-Wilkinson, D. and Palmer, R. B. (New York and London, 1998), pp. 564–617Google Scholar and also the extended discussion of both the ‘feminisation of Machaut’ as Fortune, and the equation of Fortune with ‘fame’ (i.e. both ‘woman’ and ‘renoun’) in Cerquiglini, Jacqueline, “Un engin si soutil”: Guillaume de Machaut et l'écriture au XIV^e siècle (Geneva and Paris, 1985), pp. 139–55Google Scholar.

¹⁷ Johnson, , Poets as Players, p. 43Google Scholar.

¹⁸ I have considered De toutes flours in a paper read at the 24th Medieval and Renaissance Music Conference, York, 1998, entitled ‘Counterpoint, Norms and Semitone Placement in Machaut's Ballades’. See also Fuller, S., ‘Guillaume de Machaut: De toutes flours’, in Everist, M. (ed.), Music before 1600 (Models of Musical Analysis; Oxford, 1992), pp. 41–65Google Scholar.

¹⁹ See Johnson, , Poets as Players, pp. 51–4Google Scholar.

²⁰ The most reliable modern edition for the Machaut balades is de Machaut, Guillaume, Musikalische Werke: Balladen, Rondeaux und Virelais, ed. Ludwig, F., i (Publikationen älterer Musik; Leipzig, 1926)Google Scholar. The edition by Schrade, L., The Works of Guillaume de Machaut (Polyphonic Music of the Fourteenth Century, 2–3; Monaco, 1956)Google Scholar, in which the balades are in vol. iv, is probably more widely available, for which sole reason its bar numbers will be used here. Readers are strongly advised to refer back to the original sources wherever feasible and at least to take note of the corrections to Schrade proposed by various authors and summarised in Earp, , Guillaume de Machaut, p. 588Google Scholar.

²¹ In my usage, ‘cadence’ is not coterminous with ‘directed progression’, a term adopted from Fuller. I reserve the former term specifically for closural protocols which are often, but not always, effected by directed progressions. Directed progressions, conversely, often mark the opening of phrases and are not therefore always cadential in the modern sense. For further clarification see Leach, E. E., ‘Counterpoint and Analysis in Fourteenth-Century Song’, Journal of Music Theory, 43 (2000), forthcomingGoogle Scholar.

²² The anticipation of rhymes in the first stanza is also found in De toutes flours (B31).

²³ Johnson, , Poets as Players, p. 53Google Scholar comments on the caesural placement of ‘Fortune’ in De toutes flours (B31), saying that ‘by its position, followed in each case … by a lyric caesura … it becomes the epicenter of the poem’. This is also true of B22, which he thinks conventional. In both balades Fortune is emphasised by the musical disruption of the verse structure.

²⁴ For B22 all the sources which transmit the work in a performable version do so with all four parts, the counterpoint of which confirms that the triplum (Tr) cannot be performed without the contratenor (Ct). The triplum assumes a discant relationship with the lowest voice at places where the contratenor supports a tenor–cantus (T–Ca) progression by falling below the tenor; the tenor–triplum (T–Tr) discant relationship is briefly abandoned (although the T–Ca and T–Ct relationships are retained). The manuscript copy of B22 in PR has a contratenor which does not take on tenor function at these places. I have argued elsewhere that the version of the song in PR works in the combination T–Ct–Ca but not in the combination T–Ct–Ca–Tr, transmitting the earliest surviving version of the contratenor that should have appeared in C; see Leach, E. E., ‘Machaut's Balades with Four Voices’, Plainsong and Medieval Music, 9 (2000), forthcomingGoogle Scholar.

²⁵ Although no copy transmits a three-part version with contratenor, examination of the counterpoint of those later sources with four parts copied (Ch, PR and E) reveals that a T–Ct–Ca performance of B23 is possible, whereas a four-part one is not. Any single source in which all four parts are transmitted together represents, therefore, a compendium of three possible performance versions, one two-part and two different three-part combinations. That the contratenor and triplum are alternatives is exemplified by their readings of the tenor ouvert note e. In the ouvert cadence the triplum takes ć−mi above the tenor e, and below the cantus é. This relates the open and closed endings as an imperfect sonority and its perfect resolution respectively (an unusual feature shared with B22). The contratenor reads the ouvert ending more normatively, effecting a T–Ct directed progression simultaneously with the T–Ca progression, giving a perfect resolution to e5/8. Like the triplum, the contratenor recognises the repeat of ouvert material at the end of the B section so that this cadence is replicated there in bars 49–50. The contratenor version thus takes away one of the few musical parallels with B22 – an imperfect sonority in the ouvert. This fact and its source situation combine to support those commentators who have denied that the contratenor is by Machaut.

²⁶ See also Berger, C., ‘Tonsystem und Textvortrag: Ein Vergleich zweier Balladen des 14. Jahrhunderts’, in Alte Musik als ästhetische Gegenwart (Internationaler musikwissenschaftlicher Kongress 1985, Stuttgart; Kassel, 1987), ii, pp. 202–11, esp. music example 2Google Scholar.

²⁷ Arguably musical links within the balade section actually cut across the paired nature of B22 and B23, reaching out to de-isolate them from their lonely pairing and ensconce them in the heart of the closing of the sequence of twenty-four musical balades in MS C. B22, for example, is musically more linked to Se quanque amours (B21) since they are jointly the only two four-part balades in the first twenty-four, sharing the same mensuration and the rhythmic patterns that accompany it. In this case the musical similarities, in combination with manuscript order, can forge comparisons which at the level of text alone might be overlooked. In the dark and shadowy light of Fortune, the happiness of B21's lover appears as hubris.

²⁸ Le Jardin de Plaisance et Fleur de Rethoricque (Paris: Ant. Vérard, [1501])Google Scholar. Later editions were also made until c. 1527; see Earp, , Guillaume de Machaut, pp. 114–15, sect. 3.5 [49]Google Scholar. All the Machaut works included are those that also have a musical setting, including B23 (which appears in a male-voiced version; see C. Berger, ‘Tonsystem und Textvortrag’, p. 208, n. 25).

²⁹ MS Pa copies from MS E; see Earp, , Guillaume de Machaut, pp. 115–18Google Scholar. On this issue see also Y. Plumley, ‘Lyrics for Reading and Lyrics for Singing: The Relationship between the Chanson and Lyric Poetry Repertories in the Late Fourteenth Century’, unpubl. paper read at the annual meeting of the American Musicological Society, Boston, 1998. I am grateful to Yolanda Plumley for allowing me access to this paper. I have also noticed similarities between the copies of some of the lyrics in MSS I and E. I copies the balade texts of B18, B25, B29 (Ca 1, complete and first seven lines of Ca 2), B32, B34 (both texts), B39 and B42. Of these all but the Voir Dit balades B32 and B34 show evidence of I copying from E's music section (which lacks the Voir Dit balades).

³⁰ This relationship has been previously noted by several scholars; see Günther, U., ‘Zitate in französichen Liedsätzen der Ars Nova und Ars Subtilior’, Musica disciplina, 26 (1972), pp. 53–68Google Scholar. The tenor of the opening of B23 in the refrain of Dame qui fust is changed fairly significantly in its detail. The underlying counterpoint remains similar, however. The relationship between these balades has been examined by C. Berger, ‘Tonsystem und Textvortrag’, pp. 202–11.

³¹ Dame qui fust and Se je ne suy can be found in French Secular Compositions of the Fourteenth Century, ed. Apel, W., ii (Anonymous Ballades) (Corpus mensurabilis musicae, 53; American Institute of Musicology, 1971)Google Scholar, as numbers 131 and 174, respectively. They can also be found in A Fourteenth-Century Repertory from the Codex Reina, ed. Wilkins, N. (Corpus mensurabilis musicae, 36; American Institute of Musicology, 1966)Google Scholar.

³² As I have argued in Leach, ‘Machaut's Balades with Four Voices’.

³³ The contratenors of E and PR re-converge in bar 35, as can be seen in Example la.

³⁴ There are several problems with the first quatrain. Line 2 is a syllable short and probably corrupt and I do not see how line 4 follows from lines 1–3; I have therefore left it untranslated. The overall sense is, in the first half of the stanza, that the woman has been very fortunate and, in the second half, that (good) Fortune has now abandoned her.

³⁵ It is possible that, were the rest of the poem preserved, this technique would continue – the third strophe returning to the praise of Fortune and possibly the lover's success. Certainly the verb ‘loer’ only makes sense if some comparison to a formerly happy fortunate state were made in later stanzas, unless it here means (as ‘me loer’) that the lady must complain of Fortune and yet rely on her (see n. 8 above). This remains speculative: as with many other texts – B22 and B23 amongst them – PR does not transmit the second and third stanzas of the poem.

³⁶ See Berger, ‘Tonsystem und Textvortrag’, pp. 210–11, esp. examples 1b and 4.

³⁷ Line 1 is the same as B23, line 8 (bars 1–14 based on B23, bars 51–64). Line 2 in each starts with d12 (bar 15; cf. B23, bar 13) and proceeds through a descending chain of sixths to the interval e6, held for an entire long and resolved by directed progression to d8 (bars 17–20; cf. B23, bars 14–17). In both songs the ouvert cadence has the T–Ca duet resolving to e8, and the clos cadence to d8, preceded by three descending sixths. The clos ending in each is longer than the ouvert. The B section of both balades opens with the perfect sonority a–é, which is then imperfected and resolves – to f8 in Dame qui fust and g8 in B23. The passage from the f8 sonority of bar 39 to the d8 resolution of bar 44 in Dame qui fust resembles the outline of line 6 (bars 36–42) in B23 (perhaps suggesting that the text underlay of lines 5–6 in the former should be adjusted). Line 7 in each resembles the ouvert cadence and both resolve to e8. The refrain, bars 54–65, quotes the T–Ca duet from the opening of B23 (bars 1–12) but with ornamentation in the tenor.

³⁸ The fifteenth-century manuscript copy New York, Pierpont Morgan Library, MS M.396 (Pm) does indeed transmit the song in only two parts.

³⁹ Cerquiglini, , “Un engin si soutil”, pp. 23–32Google Scholar proposes a typology of insertions of lyrics into narrative. Her first type is that which works by collage, using refrains and incipits from lyrics by other authors which are incorporated in a fragmentary way into the narrative. For musical balades, although it is the insertion into another lyric frame of other lyric fragments, it remains the use of incipits and refrains which enable the reader's ‘plaisir de la reconnaissance, plaisir de classe, comme souligne Jean Renart, “que vilains nel porroit savoir”’ (p. 28).

⁴⁰ See Brownlee, K., ‘Authorial Self-Representation and Literary Models in the Roman de Fauvel’ in Bent, M. and Wathey, A. (eds), Fauvel Studies: Allegory, Chronicle, Music, and Image in Paris, Bibliothèque Nationale de France, MS français 146 (Oxford, 1998), pp. 73–103Google Scholar for a discussion of the use of the Rose as a model for authorial self-naming at the mid-point of a text. See also a similar argument about the mid-point as a site of musical citation in Bent, M., ‘Polyphony of Texts and Music in the Fourteenth-Century Motet: Tribum que non abhorruit / Quoniam secta latronum / Merito hec patimur and Its “Quotations”’, in Pesce, Dolores (ed.), Hearing the Motet: Essays on the Motet of the Middle Ages and Renaissance (New York and Oxford, 1997), pp. 82–103Google Scholar.

⁴¹ Nevertheless, my own suspicions about Se je ne suy's use of B22 as a model were initially awakened by the textual links. The quotation of an incipit text is usually significant in this repertoire. Once the text had directed my attention to the music I found the putative link more than adequately confirmed.

⁴² Brownlee, K., ‘Literary Intertextualities in 14th-Century French Song’, in Musik als Text: Bericht über den Internationalen Kongreβ der Gesellschaft für Musikforschung Freiburg im Breisgau 1993 (Kassel, 1998), pp. 295–9, at p. 295Google Scholar.

⁴³ Arlt, W., ‘Einführung’ in Musik als Text, pp. 287–90, at p. 287Google Scholar.

⁴⁴ See S. Huot, From Song to Book.

⁴⁵ The work of Jacques Boogaart has shown that the Machaut motets are peppered with quotations from trouvère poets (see Boogaart, J., ‘Love's Unstable Balance’. Muziek & Wetenschap, 3 (1993), pp. 1–33Google Scholar). Whilst many refrains are known and catalogued from the thirteenth century, the song repertory in the fourteenth century is yet to be fully investigated in this regard; Machaut's songs have turned up textual and text–music quotations (see the notes to Nen fait nen dit (B11) and Pour ce que tous (B12) respectively in Earp, Guillaume de Machaut for a summary). The refrain of Esperance (B13), long suspected to be a quotation, is found in Rondeau 139 from Oxford, Bodleian Library, MS Douce 308.

⁴⁶ PR's contratenor for B22 lacks figure z at the very opening although the figure still pervades the balade as a whole, especially in the cantus refrain.

⁴⁷ Rankin, S., ‘Observations on Senleches' “En attendant esperance”’, in Musik als Text, pp. 314–18, at p. 318Google Scholar.

⁴⁸ For recent examples treating related fourteenth-century songs see Arlt, W., ‘Machaut, Senleches and der anonyme Liedsatz “Esperance qui en mon cuer s'embat”’, in Musik als Text, pp. 300–310Google Scholar; S. Rankin, ‘Observations on Senleches' “En attendant esperance”’, Ibid., pp. 314–18; L. Welker, ‘Weitere Beobachtungen zu “Esperance”’, Ibid., pp. 319–21; idem, ‘“Soit tart tempre” und seine Familie’, Ibid., pp. 322–34; C. Berger, ‘Modus und Intertextualität’, Ibid., pp. 335–6; and Plumley, Y., ‘Citation and Allusion in the Late Ars nova: The Case of Esperance and the En attendant Songs’, Early Music History, 18 (1999), pp. 287–363CrossRefGoogle Scholar.