12 Vernacular poetry and music
Prima la musica, poi le parole : this assertion, the title of an opera by Antonio Salieri, wittily alludes to the conundrum faced by any poetic and musical collaboration. In the story of the opera the music is already written and the harassed poet is told he must write the verse to fit the music in just four days. It is not important, according to the musician, for the music to convey the meaning of the words. But of course this is a joke that works by inverting the usual expectations of any text–music relationship, especially in opera. One of the primary aims of this chapter will be to assess the character of this relationship in its earliest formation in the medieval period. Poetry and music come together in vernacular song to create some of the most subtly exquisite survivals of medieval music. The art of the troubadours in the twelfth century, closely followed by that of the trouvères in the thirteenth, persists in our time as one of the most vividly enduring images, not only of the medieval singer, but also of song tout court , and of the Middle Ages in general.
Yet many questions remain about the character of this art. It seems not only paramount but impossible to decide which comes first, the poetry or the music. In communicating so strongly across the centuries, medieval song teases us with the question of what it is communicating and whether what we hear or perform as we re-create it bears any relation to what was heard or performed in the Middle Ages. This matters for several reasons. Many editors and performers since the nineteenth century have based their reconstructions of medieval songs on the assumption that the words drive the melody, especially with regard to rhythm. 1 More recently, others have argued that this assumption undervalues the importance of the music (Aubrey, 244) and that ‘music's elements follow rules of their own, regardless of which poetic elements are present’ (Aubrey, 253). That this debate exists shows how little modern agreement there is about what constitutes the ‘poetic’ or the ‘musical’ in the period. It also shows that it is a vital and continuing point of inquiry for any attempt to understand the nature of medieval song. Whether the words or the music have primacy turns out to have powerful consequences for the way we interpret the songs, how we understand them to have been composed, and what their larger function and significance was for medieval poet-composers and their audiences.
The chapter will begin by describing this seminal stage, the age of the troubadours, in the composition of song and trace its origins, character, function and generic variety. It will then give an account of the multiple ways in which the relationship between poetry and music fundamentally changed throughout the thirteenth century and into the fourteenth, most significantly in the development of new genres and new ways of perceiving genre; in the polyphonic motet; and in song's relationship with narrative. The concluding sections will turn to wider questions about the figure of the poet and of the composer and how ideas about poetry and music were articulated, practised and debated.
Vernacular song
Origins
The idea of an origin for vernacular song has haunted many scholars. It remains elusive partly because of the conundrum I have just outlined. The very first example of vernacular song, by definition, is rather like the very first word: we are never going to know when it was uttered and its utterance would have long preceded any thought of writing it down. But in mentioning writing, we have to distinguish the writing down of the words from the writing down of the melody. With just one exception – a song with an Occitan text of around 1100 from St Martial de Limoges – the melodies of vernacular song do not survive in written form until the thirteenth century. 2 Moreover, even when we reach the thirteenth century, only some of the surviving copies of vernacular songs contain music. Of the forty manuscripts (including fragments) of twelfth- and thirteenth-century Occitan literature, ranging from epic and didactic works to over 2,500 lyrics, just two contain music. 3 There are two more that date from the fourteenth century. Immediately, then, we are presented with a view of song that – on the face of it – is highly prejudiced towards the words.
For some scholars, these survival patterns are proof that troubadour songs were appreciated in the thirteenth century for their words rather than for their melodies. Others are quick to point out that the relative paucity of music tells us more about the shortage of skills for copying music (as true in the modern period as it was in the medieval) than about the allegedly ‘literary’ character of the songs. But the argument goes deeper than this. For instance, there is much debate about whether troubadour song was meant to be written down. Since the written versions that we have date from 100 to 150 years after the songs were composed, how can we be sure that they give us an accurate impression of the songs as they were first composed and performed? It seems quite likely that, in a situation rather like the game of consequences, what survives in writing has travelled a long way from its oral genesis. There are several important issues here: one is that the oral version of a song is always going to be different from its written version. This is true even when the act of writing is contemporary with the performance, and the differences are naturally exacerbated when the act of writing postdates the performance by more than a century. It follows that the act of writing down a song probably bears more relation to the act of writing down a play or a piece of poetry than to the activities of singing or speaking. It is the activity of writing itself, in other words, that helps to make a work ‘literary’. A second issue concerns the passage of time. It is crucial to bear in mind when we seek to interpret troubadour song that our knowledge of it is shaped by thirteenth- and fourteenth-century writers. 4 We cannot find its origins because it is always already distanced from its own time by having been modernized as well as turned into writing.
Social and literary functions
The functions of the earliest surviving vernacular songs can only be appreciated by knowing something of the social and historical context of their poet-composers. Some 460 troubadours, including a few women known as the trobairitz , are known by name over the period from around 1100 to 1300: they lived and worked across a large area of what is now southern France, at first only in Poitiers, Ventadorn and Narbonne, and then more widely into the Auvergne, Limousin, Aquitaine, Gascony and Languedoc and eastwards into Provence. Some were based in the northern Italian courts of Lombardy and Piedmont, and in Catalonia. This very substantial poetic and musical movement ebbed and flowed in successive generations: the first includes Guillaume IX, Duke of Aquitaine (1071–1126), Jaufre Rudel (…1125–48…), and Marcabru (…1130–49…), followed by Bernart de Ventadorn (…1147–70…) and Peire d’ Alvernhe (…1149–68). Bernart and Peire are also often viewed as part of the central ‘classical’ generation, with such other figures as Raimbaut d'Aurenga (…1147–73), Bertran de Born (…1159–95, d. 1215), and Giraut de Bornelh (…1162–1199…). The later, third and fourth generations, include troubadours who worked into the second half of the thirteenth century alongside their northern French counterparts, the trouvères. Perhaps the most well known of these is the so-called last troubadour, Guiraut Riquier (…1250–92), whose musical output – unsurprisingly considering his dates – survives more fully than any other troubadour (48 melodies out of 87 poems). Otherwise, music survives far more patchily: 42 of these 460 authors have songs attributed to them with music, and for most of them just one or a small handful of songs in their total surviving output has music. 5 A single song by a female composer, the ‘Comtessa’ de Dia (second half of the twelfth century(?) 6 survives with music. 7
Our knowledge of the social standing of the troubadours is largely provided by short prose mini-biographies and commentaries called vidas and razos that accompany the songs in several of the manuscripts. A kind of marketing device to advertise and sell the songs and their authors, these are colourfully written with an air of dubious authority. Nonetheless, some of their information seems plausible enough and can sometimes be corroborated or supplemented by documentary records or references in the songs themselves. Many of the troubadours starting with Guillaume, at once ninth Duke of Aquitaine and seventh Count of Poitou, belonged to the nobility: many were minor nobles and younger sons of aristocrat families. Richard I (the Lionheart), eldest son of Guillaume IX's granddaughter Eleanor of Aquitaine and Henry II of England, is the famous example of a king turned poet and composer. Others, however, were of humble or middling origins: several rose to fame from being mere joglars or jobbing professional instrumentalists and singers, such as Cercamon and Albertet; others again were from urban or mercantile backgrounds, such as Aimeric de Peguilhan, said to have been the son of a draper; many more are likely to have been clerics who had abandoned their training or livings.
The evidence of such high status for certain troubadours has helped shape the (largely correct) impression we have of a type of music that belonged in exclusive circles. Yet it is an interesting feature of the vidas that they often like to create an image for a troubadour of a self-made man pulling himself up out of obscurity. 8 Social mobility was clearly important: many of the song texts develop a pose of detachment, of the clever, mocking outsider to love, which may well have reflected a genuine experience of social exclusion. This needs to be borne in mind when the function of the songs is considered, because it shows that the picture of the songs as being part of a high courtly ethos has more than one dimension. Courtliness, perhaps the key defining term with which to understand troubadour song, may be described as a complex relationship between social, cultural and aesthetic aspiration on the part both of lords and the members of their households. 9 The image of courtliness, in short, was as important as the social reality, and the social reality itself was not so much indulgently self-enclosed as loose, itinerant and profoundly involved with war (frictions between Angevin rulers and the French royalty, the Crusades), trade (burgeoning urban and maritime economies) and religion (the Albigensian heresy). The words and melodies of troubadour song help to define a discourse of courtliness by satirizing it as well as promoting it, by wild exaggeration as well as piercing social comment, by invention as well as description. At the same time, the very fact that many lords were themselves keen to compose indicates that vernacular song was more than a matter of courtly entertainment. It was an aspect of living the courtly life, or at least of ventriloquizing its most potent characteristics.
If there is a single overarching theme that defines courtliness in vernacular song, it is fin’ amors or, as Gaston Paris termed it in 1884, ‘amour courtois’ (courtly love). 10 This difficult, richly contentious notion of refined, rarefied, extreme love is the air which many genres of troubadour song breathe. Recent research has greatly deepened and enlarged our knowledge of how love and its associated themes in song, such as adultery, secrecy (celar ), the imagined power of the lady (domna ), the role of the lauzengier (slanderer) or gilos (the jealous ones) relate to the changing social and ecclesiastical representations of love and marriage in the medieval period. It has become possible to see that some of the contradictions in fin’ amors are part of a wider exploration of cultural fault lines in the portrayal of women as both passive and active in love, as silent and yet dominating, as objects of both worship (pretz, valor, onor ) and derision. Some of the specific functions of troubadour song thus include the creation of a discourse about women's conflicted roles in an aristocratic household alongside an explicitly male-voiced expression of frustrated desire. Less abstractly, perhaps, the large amount of surviving debate poetry indicates that song was a means of establishing a public forum for brilliantly fast-moving exchanges between lords and the shifting group of people in attendance on them, a platform for public argument about the qualities and values of courtly life.
Genres
Scholars have found many ways of discussing and defining genre in vernacular song. It is often difficult to know whether it is form, genre or style that is under consideration. For an older generation including Friedrich Gennrich and Hans Spanke, followed more recently by Pierre Bec, form was a key perspective, yet subsequent work influenced by literary considerations, by John Stevens and Christopher Page, for example, has preferred a broader notion of genre that includes social context and style. Where form is emphasized, monophonic song can be divided up into different structural types: the laisse , the refrain-song, the sequence and the strophic song. This offers an illuminating sense of the fundamental technical choices made by composers, but it provides no means of linking these choices of poetic and melodic arrangement to theme or content, tone, diction and choice of pitch range, register and, more broadly, social meaning. Stevens and Page develop ways of approaching the songs particularly through style, higher and lower, but also through matters of performance (rhythm and the use of instruments) and social history (Parisian music theorists and practitioners, and Cistercian anxieties about dance-song). Perhaps the most interesting newer work by such scholars as Margaret Switten and Elizabeth Aubrey has gone on to ponder how far the music and the texts correspond generically. The widespread use of contrafacta in early vernacular song means that there is no intrinsic connection between a melody and its textual partner, and provokes questions about the kinds of generic expectation carried by melodies, and whether they overlapped with, were independent of, or, conversely, were determined by the texts.
Here there remains a marked gulf between the thinking on genre carried out by literary scholars and that by musicologists. Whereas much research has gone into the literary and social contexts for the texts, in which scholars have developed approaches from psychoanalysis, intertextuality, gender studies, and ecclesiastical and economic history, the melodies have received less attention. In part, this is perhaps because it has seemed hard to talk about the melodies without the words, and harder still to comment on both together. As Switten has rightly remarked, we have yet to find a critical and scholarly language to talk of the relations between music and words. These relationships ‘do not constitute a discipline with a distinguishing ideology and approaches sanctified by use. There is even no satisfactory terminology to speak of both [disciplines] at once.’ 11 More positively, the lack of a hallowed critical language indicates how far the scope for further research into vernacular poetry and music extends, and its consequent potential as an innovative field of enquiry.
The pre-eminent genre of medieval vernacular song was the strophic troubadour canso, a term used after the 1150s (the earlier, more general term was vers ) to refer to songs on love. 12 Cansos explore the topic of fin’ amors in all its variety and complexity, through an equally rich and varied use of versification. The canso had a kind of counter-genre in the sirventes , a satirical and moralizing type of poem that became more widespread, and acquired this generic title, at a similar date. It was characterized by often being attached to a borrowed canso melody, as, for example, Bertran de Born's Un sirventes on motz no falh which, as he points out, uses a melody by Giraut de Bornelh. Other genres include the tenso, partimen, joc-partit or debate songs; pastorela , a song that narrates an encounter between a knight and a peasant girl; dansa , a song based on a dance; descort , a ‘discordant’ song in which the stanzas (unusually) vary in rhyme, metre, and sometimes language; enueg , a canso listing unpleasant subjects; planh : a lament on the death of a king or other dignitary; gap , a boasting poem.
In mentioning dance, it is worth remarking that there also existed a lively unwritten dance-song tradition. Textual sources from the twelfth century onwards ranging from chronicles, romances and sermons to town records contain many references (some sternly condemnatory) to round dances and processional dancing that clearly involved a lead singer and some kind of chorus response. 13 A vivid illustration in the Oxford Chansonnier, MS Douce 308, placed at the head of the section of balletes shows two figures dancing with linked hands in front of a pipe and tabor player. This kind of evidence hints at the much larger world of performance and improvised music that surrounded the relatively meagre written remains that we now possess both of texts and of melodies. Many texts refer, often obliquely, to dance movements, yet cannot be tied to particular surviving tunes, and indeed the song copied next to the picture just described seems to have no particular connection to dance in its text, nor does this chansonnier contain music. As with instrumental music more generally, it is important to be aware that the written sources represent only a fraction of what was played and composed throughout the medieval period.
Moving back from genre to form, we find in troubadour song perhaps the most artful exploitation of sound patterns in any short genre, modern or medieval. Metre (number of syllables per line) and rhyme are used through the careful placing of words to partner pitch, intervallic relationships, motifs, incipits and cadences in the melody. The fundamental structural unit is the strophe, organized in its most general outline for words and music in the form AAB. Within this outline, each song, of five, six or seven stanzas ( coblas ) with one or more shorter envoys (pendant stanzas) or tornadas , is crafted to contain sequences of repetition and variation in both words and melody of often dazzling virtuosity. Remarkably, very few songs have the same pattern: over half the poems with extant melodies have a unique rhyme and verse structure. 14 In one of the earliest and most influential analyses of troubadour song, De vulgari eloquentia (ca 1303–5), Dante describes the two-part structure of the canso strophe as the frons and the cauda, each frons usually divided into two pedes (singular pes). The distribution of rhymes across the frons and cauda falls into several common patterns: the same rhyme sounds in each strophe (coblas unissonans ) but with different rhyme words; coblas doblas and coblas ternas where the rhyme sound changes every two or three stanzas; coblas capfinidas where the first line of a stanza repeats a word from the last line of the previous stanza; and coblas retrogradas where the rhyme sounds are reordered in a different sequence from stanza to stanza.
When we turn to the music, the language of description immediately becomes less straightforward. It is not so easy to plot patterns of repetition in melodies as it is in texts, or at least the patterns are more open to diverse forms of analysis. Scholars who have worked closely with troubadour texts and music together (Stevens, Switten and Aubrey) have drawn attention to the way that verbal rhyme and musical repetition, while they may be subtly juxtaposed, rarely simply coincide. Their relationship is poised and balanced, intimate yet also independent. Nonetheless, as remarked above, the practice of contrafacta puts a brake on attempts to see a song as necessarily an organic text–music whole. Where individual texts are copied with more than one melody we cannot be sure in any single manuscript version that the same person composed the text and the music, and therefore that there is a single intentio governing the operation of words and melody. Or again, a poet-composer may have been more gifted as a musician than as a poet, or vice versa: the vida of Jaufre Rudel somewhat waspishly complains that ‘he composed many songs about [his lady] with good tunes and poor words’. A melody by Guiraut de Bornelh reused by Peire Cardenal (Aubrey, 153) suggests further creative possibilities in showing how some troubadours deliberately reworked material.
The celebrated canso Can vei la lauzeta (‘When I see the lark’) by Bernart de Ventadorn well illustrates the interpretative challenges posed by troubadour song ( Example 12.1 ). The challenges arise in the first instance out of the text but then need to be considered alongside the music. The melody of this canso is more widely transmitted than any other troubadour melody and with fewer differences between versions, but this relative stability in musical terms is not matched by the words. As is fairly common among the troubadour repertory as a whole, the seven stanzas plus envoy are arranged in more than one order in the various manuscript copies. For an older generation of text editors, it was a matter of selecting the ‘best’ order and printing that once, thus creating a fixed song of which three stanzas are given below:
Can vei la lauzeta moverde joi sas alas contral rai,que s'oblid'e.s laissa chazerper la doussor c'al cor li vai,ai! tan grans enveya m'en vede cui qu'eu veya jauzion,meravilhas ai, car desselo cor de dezirer no.m fon.Ai, las! tan cuidava saberd'amor, e tan petit en sai!car eu d'amar no.m posc tenerceleis don ja pro non aurai.tout m'a mo cor, e tout m'a me,e se mezeis e tot lo mon;e can se.m tolc, no.m laisset remas dezirer e cor volon.Anc non agui de me poderni no fui meus de l'or'en saique.m laisset en sos olhs vezeren un miralh que mout me plai.miralhs, pus me mirei en te,m'an mort li sospir de preon,c'aissi.m perdei com perdet selo bels Narcisus en la fon.
When I see the young lark moving its wings for joy in the ray of the sun so that it forgets itself and lets itself fall for the sweetness that goes to its heart, alas! such great envy comes to me of anyone whom I may see rejoicing, I wonder that forthwith my heart does not melt with desire.
Alas! so much did I think that I knew about love, and so little do I know, for I cannot abstain from loving her from whom I shall never have reward. She has stolen my heart from me and has stolen myself from me and herself and the whole world. And when she stole herself from me she left me nothing except desire and a heart filled with longing.
Never did I have control over myself, nor was I my own master from that moment when she let me look into her eyes, into a mirror that pleases me greatly. Mirror, since I beheld myself in you, my deep sighs have slain me, for so did I lose myself as fair Narcissus lost himself in the fountain. 15
More recent approaches to text editing, by contrast, have encouraged scholars to consider the different text orders as potentially interesting in their own right. 16 Rather than seek a single definitive version of ‘the’ song, we are now invited to think of Can vei la lauzeta as having several possible realizations, perhaps indicative of its being reworked for different audiences and occasions. Thus in one stanza order shared by five manuscripts, two of the stanzas, including the one on Narcissus, come later and two others are shifted earlier. 17 The result in this case is the stark difference between a speaker/singer who is despairing but ultimately reconciled to the nature of love, and one that a recent scholar has called ‘an angry and recalcitrant woman-hater’. 18
Example 12.1 Can vei la lauzeta (‘When I see the lark’) by Bernart de Ventadorn (Hendrik van der Werf and Gerald A. Bond, eds., The Extant Troubadour Melodies: Transcriptions and Essays for Performers and Scholars [Rochester, NY: authors, 1984], pp. 62–71)
What then of the music? In general terms, the strophic melody is characteristic of Bernart's restrained, mellifluous style, moving largely in adjacent steps, with short ornamental passages not so much interrupting as easing the beautifully composed flow of sound. This quiet control in the music is balanced by a similarly restricted palette of rhymes and rhythmic patterns: 8 syllables per line, 8 lines per stanza and only 4 rhyme sounds (although again this varies slightly across the manuscripts). Just one abrupt descent of a whole fifth occurs in the whole melody at the end of the fifth line. Is there a connection here with the text? For instance, it is tempting to argue that this feature, coming just after the mid-point of the strophe, has a structural parity with the singer's declaration in stanza 5 just after the mid-point of the song that he will renounce his destructive lady and leave her forever. 19 But in another version of the song this stanza occurs not fifth but seventh, being the last before the envoy. Clearly the same argument could not be made with the text–music relationship in this version. In part this is a matter of reading each version responsively and individually. But it also speaks more widely of the need for a critical approach to the mutual support of words and music in troubadour song that can take account in some way of the structural asymmetry of a linear text aligned with a strophic melody. It must also take account of the distance as well as the intimacy of text and music in any one song.
The sheer variety of musical devices and compositional features makes it hard to generalize about the music of the troubadours. Yet the attempts by scholars such as Aubrey and Switten to discuss the style and musical language of individual composers have made important steps towards our reaching a sense of the expectations that circulated about song on the part both of the composers themselves and of their audiences.
The changing contexts of song: grand chant , motet and narrative
If genre is a multivalent concept for the twelfth century, then it has even greater flexibility and inventiveness in the thirteenth. It is worth recalling that the very first written record of a troubadour song, Jaufre Rudel's Lanquan li jorn son lonc en may (‘In May when the days are long’), occurs in an early-thirteenth-century romance by Jean Renart known as Le Roman de la Rose . Romance is a key resource for vernacular song: Renart's Le Roman de la Rose alone contains the first surviving examples and extracts from a range of genres – the chanson de toile, rondet de carole , and the independent refrain – and two further genres – the chanson d’éloge and tornoi de dames – which are unique to the work. 20 Altogether some seventy narratives cite song from the early thirteenth century to the early fifteenth, in several hundred manuscripts. Music is provided in just under a quarter of these copies. 21 If the genre of romance seems an unusual place to start in describing vernacular song then this is an accident of modern scholarly emphases. Recognizing the importance of narrative as a context for the writing down of song throughout the thirteenth century helps us to grasp how broadly song participated in, and contributed to, the increasingly sophisticated explorations of vernacular literacy through the medium of the book.
The trouvère chanson, traditionally seen as the northern cousin of the canso, is equally traditionally described as a kind of appendage to its southern predecessor. It undoubtedly shares much common ground, especially in its treatment of love. However, placing it here in a separate section from the troubadours is a way of emphasizing that in several ways the account needs to be reversed. The trouvères invented the troubadours rather than merely followed after them. Moreover, the work they performed on song was part of a much wider cultural movement that included new prominence given to the vernacular in France through increasingly centralized legal practice, a rapid rise in the production of vernacular books, the growing international importance of the university in Paris, extensive urban expansion and a marked increase in ecclesiastical and monastic institutional activity. In musical terms, there were far-reaching changes in notational practice and an extraordinary proliferation of compositional techniques in which musical material from liturgical sources was juxtaposed with, cut through and transformed by the increasingly widely transmitted secular monophony of both troubadours and trouvères. It was the trouvères who effected these changes, partly by the very process of writing down the songs of their predecessors, and partly by creating many new directions for the relationship between text and music.
In broad terms, the genres practised by the trouvères were similar: the principal kind, comparable to the canso, is the chanson d'amour or grand chant courtois (‘grant chant’ is a resonant classifying term used in the late thirteenth-century chansonnier Oxford, Bodleian Library MS Douce 308). 22 Trouvères such as the great Châtelain de Couci (d. 1203) or Gace Brulé (ca1160–after 1213) built on the success of the troubadours by creating through the grand chant a powerfully pervasive courtly discourse of text and melody that dominated the thirteenth century and the two centuries following. Yet alongside these heights of refined expression, trouvère writing is distinctively characterized by a fascination with ‘lower-style’ genres. Apparently humble genres such as the rondet de carole (later rondeau), chanson de toile, chanson de femme, chanson de mal mariée, rotrouenge and sotte chanson are newly recorded from the thirteenth century, and crossover genres taken over from the troubadours such as the pastourelle or the jeu-parti are greatly expanded. Interspersed among, and to a large extent at the basis of, many of these types is the refrain, a brief formulaic tag of verse and (in many cases) melody cited in thousands of permutations. 23
The trouvères, in short, were not exclusively interested in creating single poetic–musical artefacts of superlative craft. They were also drawn towards breaking down that creative product into its constituent elements. This can be illustrated by two contrasting pieces. One is a song by the Châtelain de Couci often performed and recorded for modern listeners, the first stanza of which is given in Example 12.2 . This song through its music and its words distils all the courtliness one could hope to find in a medieval love song: a limpid spring setting impelling the lover to sing of his intense desire yet also of his pain as he is forced to leave his lady to go ‘outremer’ or beyond the sea. The haunting melody follows a classic AAB schema, rising in the B section to the highest pitch of the song rapidly followed by the lowest, yet with an unbroken, flowing continuity of melody. 24 At the opposite end of the spectrum of medieval song comes this second example, a short, technically simple exclamation: Jolietement m'en vois; jolietement (‘Happily I go; happily’). 25 This little phrase has an almost cheekily insignificant character when set beside the long, exquisitely expository eroticism of Li noveaus tanz . It implies accompanying movement, and for an older generation of scholars was a surviving trace of a world of lost popular dance. Yet what happens to this phrase is revealing of the many changes in focus that take place in the treatment of song during the thirteenth century. We find it cited in four different pieces: a song text in the motets and rondeaux section of the unnotated chansonnier Douce 308; in the tournament romance Le Tournoi de Chauvency , also copied in Douce 308; at the end of a stanza in a short strophic narrative, La Chastelaine de Saint Gille , again without music; and finally in a notated polyphonic motet (Mo 260) copied into the seventh fascicle of a huge motet compilation, the Montpellier Codex. 26 In each of these contexts it has a different form: in Douce 308 it is split in two around a single-stanza text, with ‘Jolietement m'en vois’ at the start and ‘jolietement’ at the end. La Chastelaine de Saint Gille simply quotes it ‘in full’ at the end of the stanza. In the Montpellier motet, by contrast, it is the refrain of a rondeau which has itself become the tenor of the three-voice motet ( Example 12.3 ).
Example 12.2 Li noveaus tanz et mai et violette (‘The new season, May, the violet’) by the Châtelain de Couci, (F-Pn fonds fr. 12615, fol. 155r)
Example 12.3 Jolietement m'en vois; jolietement (‘Happily I go; happily’) (The Montpellier Codex , ed. Hans Tischler, trans. Susan Stakel and Joel C. Relihan, 4 vols., Recent Researches in the Music of the Middle Ages and Early Renaissance 2–8 [Madison, WI: A-R Editions, 1978–85]; hereafter Mo ; Mo 260, vol. III, pp. 78–9)
These three views of this song take us into a seemingly very different compositional world from Li noveaus tanz . From song as a highly wrought self-contained structure, designed to evoke a specific and thoroughly imagined emotional landscape in which the act of singing such a song is self-consciously a part, we have moved to a much more fragmentary, protean notion of song as a brief yet key motif, able to move readily between genres and function as a complex organizing tool in the creation of the rhythmic and motivic patterns of a polyphonic motet. Rondeau-motets are a relatively small subgenre of the motet, yet they illustrate a much broader general preoccupation with juxtaposing distinct and sometimes seemingly incompatible registers, styles and genres. Modern distinctions between song and narrative, strophic and non-strophic, and sacred and secular often do not seem adequate ways of describing the pleasure in dissonance and hybridity so evident in thirteenth-century song composition. ‘Song’ is overturned in these explorations and reinvented as a set of smaller elements, ready to be recombined, grafted, strung in sequence cento-style or layered polytextually and polymusically in ways that often seem more interested in disruption than in communication.
A motet that seems to make lack of communication a central theme of its comic textual and musical manipulations occurs within a cluster of Robin and Marion motets from fascicle 7 of the Montpellier Codex. These motets draw on texts taken from the pastourelle genre in which the feelings felt by an attractive shepherdess for her sturdy but not over-bright peasant lover are threatened by an adventuring chevalier. En mai, quant rosier sont flouri / L'autre jour, par un matin / Hé, resvelle toi [Robin] (‘In May with rose bushes blooming / The other day, in morningtide / Hey, wake up Robin’), Mo 269, is unusual in bringing together three texts of this type, not just in the upper two voices but in the tenor as well ( Examples 12.4a , 12.4b ). For instead of the more usual Latin chant extract, the tenor is here a refrain:
Hé, resvelle toi [Robin] ,Car on enmaine Marot ,Car on enmaine Marot .
[Hey, wake up Robin, for someone is taking Marot away, for someone is taking Marot away!]
Like several of the other Robin and Marion motets, Mo 269 has a link with the Arras poet and composer, Adam de la Halle: in this case the refrain also occurs with both text and music in Adam's Le jeu de Robin et de Marion .
Example 12.4a and 12.4b En mai, quant rosier sont flouri / L'autre jour, par un matin / Hé, resvelle toi [Robin] (‘In May with rose bushes blooming / The other day, in morningtide / Hey, wake up Robin’) (Mo 269, vol. III, pp. 93–5)
It is worth looking more closely at the context of this refrain in Adam's Le jeu de Robin et de Marion because it provides insight into the cleverly allusive practices of the motet's composer, who may even have been Adam himself. First, we may appreciate the comic timing of the juxtaposition of the two main texts. The triplum (the second texted upper voice above the tenor), given like nearly all pastourelle narratives from the perspective of the chevalier, recounts that Marion is sitting near a wood lamenting that Robin has abandoned her for Margot, ‘la fille Tierri’ (Tierri's daughter). But as soon as her lament is uttered, Robin, who has overheard it, comes running up and takes her away to play. In a neat reversal, the motetus (the first texted upper voice above the tenor) has the chevalier see Robin sighing for Marion, who then, having likewise overheard him , runs up quickly and assures him that he has ‘conquered’ her love. Musically, the two texts interwine to play off each character's self-enclosed (and, as we learn retrospectively, misguided) misery, most comically when first Marion and then Robin sings ‘Aymi!’ The composer places the identical words ‘et disoit: “Aymi”’ in direct succession across the two parts, so that just as Marion is complaining in the triplum that ‘Robin, mise m'avés en oubli’ (Robin, you have forgotten me), Robin is simultaneously exclaiming ‘Aymi! Quant vendra la bele au cuer joli?’ (Alas, when will she come, the fair one with the gay heart), each cry reinforced sonically by the chiming ‘i’ rhymes. At their most self-enclosed, in short, the composer brings the two characters into tight musical and verbal congruence.
The refrain adds a further comic layer, indeed layers, to the situation being played out in the upper voices. Hé, resvelle toi Robin! / car on enmaine Marot is sung in Adam's Le jeu de Robin et Marion by Robin's cousin Gautier whose warning that the chevalier has carried Marion off comes somewhat gratuitously since poor Robin has just been thoroughly beaten up in his attempt to stop this from happening. Adam presents Robin throughout as bested not only by the chevalier but also by Marion, who briskly rescues herself straight afterwards. So it is appropriate that the tenor of the motet should keep reiterating that Robin should wake up. But the refrain has a further, darker context in another pastourelle text ‘Hier main quant je chevauchoie’ (‘Yesterday morning when I was riding’) (attributed to Huitace de Fontaine). 27 In this song, Marion is alone and in a remote place – ‘pres de bois et loing de gent’ (near a wood and far from any people; line 4). She anxiously sings ‘Dex, trop demeure; quant vendra? / loing est, entroubliee m'a’ (‘God he is too long; when will he return? He is far away, I am anxious’; lines 9–10). But Robin is asleep, and does not awake even when Marion screams as loudly as she can: Hé! resveille toi, Robin! / car on en maine Marot . He only appears in the final strophe after she has been raped by the chevalier. If the composer of Mo 269 had this pastourelle in mind then the repetitions of Hé, resvelle toi Robin! would have a redoubled intensity, mocking Robin for his somnolent failure to break out of his textual trap. In this artful piece, the composer seems to use the overlapping motet voice structure to draw attention to the verbal miscomprehension it promotes: the characters overhear each other singing within each text, yet caught in simultaneously uttered voice parts they can neither understand what each other is expressing nor communicate with each other. It is in spite of their efforts rather than because of them that the motet ends happily, with a song of triumph (‘Robins, conquis avés l'amour de mi’ [‘Robin, you have conquered my love’]) and a silvan frolic (‘au bois sont alé pour deporter’ [‘they went off to the woods to play’]).
These contexts for song, motet and narrative were of crucial importance to the fundamental changes that occurred as the thirteenth century moved into the fourteenth. The high point of monophonic song in the art of the troubadours and trouvères passed, and polyphony assumed in vernacular song composition the kind of importance it already had in sacred contexts. At the same time, the relatively lowly genres associated with dance , rondeau, ballade and virelai, by a mysterious alchemy became elite courtly fare. Two factors may be singled out here, both of which are central to modern scholarly questions about these changes: rhythm and textuality. The celebrated Roman de Fauvel manuscript, F-Pn fonds fr. 146, exemplifies both. It contains a unique 1316–18 revision of this satiric narrative, in which 169 pieces of music were interpolated. 28 No other medieval narrative contains such a remarkable number and range of musical pieces. They are of particular interest to music historians for being a mélange of disparate genres, some newly recorded, others known from older sources and traditions but recorded here with updated notation in the newer style, old genres reworked into unusual forms, old texts given a new language or context. The manuscript provides substantial evidence of change, yet in terms that are hard to understand without a much more detailed context than we possess from other surviving sources. The challenge for current scholars is to bring together detailed work on such evidence that does exist, principally the notated and unnotated motets and other song forms from the Montpellier Codex and the Douce 308 chansonnier, together with the work of such turn-of-the-century authors of song and narrative as Adam de la Halle, Jehan de Lescurel, Nicole de Margival and Watriquet de Couvin. 29
In all these sources questions of rhythm and textuality are at once intensely problematic and intertwined. It is one thing to recognize that a sea-change occurs in high-style song, from the largely unmeasured notation of a chanson courtoise to an intricately measured polyphonic ballade by Guillaume de Machaut, but another to track the practical, creative and cultural reasons for this change. Scholars have drawn attention to several factors: the importance of urbanization, of increasing literacy and awareness of the creative potential of writing for musical transmission and composition, the new first-person love narrative, and the rise of the puys or cultural societies in northern French-speaking towns which encouraged competitive song production in a spirit of bourgeois social solidarity. One of the most unexpected symbols, perhaps causes of change in the character of song is the refrain, an example of which would be set as the basis for an annual compositional lyric challenge for members of puys . On the edges of strophes, creating changes of rhythmic mode in motets, marking divisions in a narrative, refrains are a connecting thread between genres, styles and registers. They liaise between orality and literacy, and between the languages of music and poetry.
By being so pervasive, refrains are part of the very fabric of thirteenth-century textuality. The key generating elements of works from so many different genres, they are the core materials of courtly discourse. Tracing the course of refrain citation across different generic contexts is like seeing the raw material of courtly speech and song being shaped, divided, combined, amplified and structured before one's eyes. Refrains are the ultimate examples of literate songs: their citation makes a song part of a larger textual whole, just as it also makes a text confront or absorb song. 30
Poets and composers
Voice and authority
That the new love narratives of the early fourteenth century, such as Nicole de Margival's Dit de la panthère d'amours and Jehan Acart de Hesdin's La Prise amoureuse , are the first surviving locations for the new formes fixes is worth further comment. It alerts us to the importance of narrative as a means of framing song throughout the later Middle Ages. In this final section I want briefly to outline some of the ways in which the relationship between words and music in medieval vernacular song was influenced by narrative practices. One of the earliest trouvères, Chrétien de Troyes, who was a master of Arthurian narrative as well as a composer of chansons, reminds us that the two genres were closely connected and, indeed, that narrative was the earliest source in the early thirteenth century for both secular and sacred vernacular song: in the Roman de la Rose by Jean Renart and the Miracles de Nostre Dame of Gautier de Coinci. By being set into narrative, song gained a certain visibility and distinctiveness: the practice stimulated and was no doubt itself provoked by a desire to promote authorship. Perhaps the most explicit example of authorial promotion is the Roman du castelain de Couci by Jakemès (ca1300) in which the Châtelain de Couci's songs are woven into a biographical romance that tells the story of the love affair allegedly expressed in the songs.
In these decades at the end of the thirteenth century and into the fourteenth, song authorship gradually became once more a matter of self-proclaimed courtly and aesthetic status as it had been for the troubadours. But where trouvère poet-composers tended to look back or sideways for models of vernacular authorship, the early-fourteenth-century love poets put forward their own compositions and their material representation of them in specially produced books as evidence of an authoritative approach to the art of love. To cite a song in a larger narrative frame is to assign it a voice, and many subtleties are available to the composer of a song who wishes to appropriate another's poetic voice in the service of his own authorial identity.
Citation
Citation was a key technique in the development of new approaches to song and authorial power. We can see this in a key transitional composer, Adam de la Halle. Associated primarily with Arras, but also with Paris and Naples, Adam's surviving works exhibit great versatility: he wrote grands chants , motets, polyphonic rondeaux, jeux-partis , two dramatic jeux – Le Jeu de Robin et Marion and Le Jeu de la Feuillée (a third is attributed to him, Le Jeu du Pelerin , that may have been posthumous) – an incomplete chanson de geste in laisses – Le Roi du Sicile – and a strophic congé . All of these compositions except the last two survive with music; in the case of the jeux , both for the refrains and for the dramatic songs. Adam's work is marked by widespread citation, not only of well-known refrain material but also of his own texts and melodies, including his own reworkings of refrain material. Motet 279 from the Montpellier Codex, for example, contains refrains from his polyphonic rondeaux (Dieus, coument porroie [‘God, how could I’], vdB, refr.496, and De ma dame vient [‘From my lady comes’], vdB, refr.477), as well as one that is sung in Le Jeu de la Feuillée . 31 This master of citation became noted as a figure of authority: his chansons are cited with respect by the narrator of Nicole de Margival's Dit de la panthère . It was a short step from here to the work of Guillaume de Machaut in which the practice of auto-citation received eloquent expression, most notably in his sequence of set-piece songs in the Remede de Fortune and in Le Voir Dit . Adam is evidence of the public importance attached to citation in the period, and also of the continuities of practice across the border of the centuries. New research into the nature and uses of citation from the thirteenth to the fifteenth centuries is already producing a substantial body of information about the intensive cross-referencing of poetic and melodic phrases among song composers throughout that period. 32
Music and literature
It is often remarked that the arts of music and poetry start to diverge in the late fourteenth century. It cannot be denied that after Machaut, perhaps the last great trouvère, no single author appears to have a comparable control over music and text together. And even Machaut has a bulkier poetic than musical output. But what continues into the fifteenth century is an increasing interest in the textuality of music as well as the musicality of the text. The production of commentaries and handbooks on the art of song from Eustache Deschamps's L'Art de Dictier (1392) to the Arts de seconde rhétorique (fifteenth century) testifies to a desire to give song an ever greater specificity of textual control. Books of music paradoxically give music more transmissive freedom than ever: through being written down and in due course printed a song can be known, remembered and reworked across a wider audience rather than being confined to a relatively narrow life as a performance. The transformation of the scribe into the poet, and of poetry into literature, gave music yet more reason to find rhetorical heights in the art of song.