Quotation—not a very promising subject, you might think. Like the footnote, with which indeed it shares a common history, quotation seems more of an academic obligation than a creative act.1 Poets may allude to or imitate one another, but they will not repeat an earlier text except to pastiche it, as if repetition was in itself already faintly comic. The lyricist’s emblem is the nightingale, not the parrot. In the twelfth century the prestige of poetic nightingales is at an all-time high.2 Around the middle of the century, the troubadour Marcabru wrote a brace of parodic love songs in which a foolish lover sends a starling—not quite a parrot, but the point is the same—to deliver his message to a tart who then turns him down.3 In an outrageous mock oration, which Marcabru surely knew, Ovid summons the world’s birds, nightingale included, to mourn the death of his girlfriend’s parakeet: the choice of bird implies both irony toward his theme and mimicry of Catullus’s well-known lament for his girlfriend’s loss of her pet sparrow.4 Both Ovid and Marcabru come across as anti-parrot.
But parrots have always had their promoters, as we will see; and quotation has had its poetic high points too. One was in late antiquity when there developed the genre of the cento, a work put together entirely from fragments of classical texts.5 Another was from the late twelfth to the early fourteenth centuries when, in a similar spirit, many new forms of writing prospered that centered on quotations from a literature that was fast assuming classical stature: that of the troubadours. With the exception of its best-known practitioner—no less a poet than Dante Alighieri—this “secondhand” deployment of troubadour song has been largely overlooked.6 The task of documenting it has been relegated to editors’ introductions and notes to the individual works concerned; and critical attention has been diverted to the related but distinct practice of lyric insertion that is typical of Northern French romance.7
Dante’s troubadour quotations, however, are just a famous but fairly small-scale instance of an art that, by his time, had been flourishing for over a century. Appendix 1 records the presence in writings of this period of over 600 passages of quotation from more than 350 known poems by nearly 100 different troubadours, as well as from roughly 50 other unidentified compositions. That is, nearly a quarter of all known troubadours are quoted at some time; and about 15 percent of the corpus is quoted by other writers. The number of actual instances of quotation is of course higher than 600 but not by much since, as Appendix 1 also shows, the same troubadour passage is rarely quoted by more than one author or work. Almost all the works in question are in the same language as the troubadour lyrics they quote, a language I call Occitan.
No one questions the centrality of the troubadours to the development of European culture and, indeed, of European sensibilities; much has been written about the importance of their legacy. But this book is the first to restore to visibility the phenomenon of troubadour quotation and to study it as a whole. The dozen or so lines that Dante quotes have not previously been set in this wider context. Nor has it been appreciated to what an extent the act of repetition on which quotation (like many other scholastic practices) depends is, paradoxically, an engine of change.8 If troubadour poetry conveys knowledge of desire, quoting it exhibits desire for knowledge. This passage through the discourse of knowledge—or supposed knowledge—is, I shall argue, influential in transforming desire from its various elaborations as fin’ amor (or courtly love) in the Middle Ages to new and more learned formulations that culminate in what is often called the Renaissance.
This project has led me to revive a distinction that is not regularly upheld between citing and quoting. By citing, I understand referencing an author, a work, or an opinion. By quoting, I mean something much more textually precise: the verbatim repetition, in its original form, of a passage that can be anything in length from a complete line of verse to a sequence of several stanzas. For my study, a core of at least a line must be accurately repeated, even if (as often happens) this core is accompanied by recapitulation of a more allusive kind, for a passage to qualify as a quotation. (I have set the minimum at a full line since the troubadours’ stylized rhetoric abounds in reiterations of only a few syllables that may reflect convention more than intended allusion; and the network of connections between poets that this produces has been thoroughly studied by Jörn Gruber.)9 Clearly it can happen that a text is quoted but not cited, as when a passage is repeated verbatim without any apparent markers of its provenance; and vice versa, a text can be cited without being quoted, for example, when a troubadour is named and an opinion imputed to him but no attempt is made to capture its verbal expression. Equally, a text may be both quoted and cited, as when an excerpt from a lyric is repeated, attributed to a celebrated poet, and hailed as an authoritative source of sound opinion. Differentiating in this way between quoting and citing makes it easier to grasp the extraordinary novelty of troubadour quotation. Whereas citation of Latin authorities is commonplace, Occitan lyrics are probably the first corpus in any language (including Latin) to be quoted at such length verbatim in medieval Europe.10 The passages listed in Appendix 1 are all quotations in this sense.
Many twelfth-century troubadours allude to one another’s songs, reprising phrases or rhyme schemes; they also reiterate material from their own songs, in tornadas, for example, or to link successive songs on the same theme together. But as far as I know, the earliest author to quote entire lines (or more) from troubadour songs is the Catalan poet and grammarian Raimon Vidal de Besalú, and the first work to contain quotations is his Razos de trobar (Rational principles of poetic composition). Dating from the turn of the twelfth to the thirteenth century, this, the oldest vernacular grammatical treatise-cum-art of poetry, quotes famous troubadour songs as models from which Raimon’s fellow Catalans can learn to appreciate or compose lyrics themselves. There follow many other such treatises, some of them developments of the Razos de trobar, others independent. Raimon Vidal is also credited with the authorship of three short didactic verse narratives (novas), two of which are extensively peppered with quotations from troubadour lyric and were probably composed by 1209 or 1213. Abril issi’ e mays intrava (April was ending and May beginning) is about how the successful joglar, or performer of lyric, can negotiate varying levels of connoisseurship in his audience; So fo e.l tems c’om era gais (It was at that time of merriment) dissects the rival claims of members of a love triangle.
In the first decades of the thirteenth century, new forms of troubadour quotation develop in the north of Italy too. Some or all were initiated by Uc de Saint Circ, a troubadour who took up residence in the Marca Treviziana, in the Veneto. Uc launched the genres of the vida and razo, short prose introductions to the lives of individual troubadours and explanations of how individual songs came to be composed. Only a few vidas contain quotations, but razos almost invariably quote at least the opening line of the song they accompany and may contain excerpts from several songs. This region and this period also sees the compilation of the first manuscript collections of the troubadours, known as chansonniers, which seemingly from an early date contained anthologies of excerpted stanzas (coblas triadas) alongside whole songs. Such collections have found their way into about a third of the surviving troubadour songbooks; there may have been many others that are now lost.11 It is not easy to distinguish systematically between excerpted stanzas, stanzas that exist only in isolation (coblas esparsas), and truncated copies, but we do have the vida of one anthologist, Ferrarino da Ferrara, which credits him with extracting from whole songs the elements in them that are the most significant, thereby generating what has been called a “Sentenzensammlung”—a kind of lyric equivalent to a collection of Latin authorities.12 In the manuscript that preserves the 226 extracts attributed to Ferrarino (the part of chansonnier D known as Dc), and in some others, these stanzas are accompanied by rubrics that clearly indicate their status as excerpts; and one, chansonnier H, even offers guidance on their meaning and potential application. The extracted stanzas are effectively being earmarked—or solicited13—as future quotations, the more structured the florilegium, the closer its foreshadowing of our modern dictionary of quotations. No other European poetry is explicated or anthologized in this way. Appendix 2 details the contents of six of these florilegia. Comparison with Appendix 1 shows that for some troubadours (like Folquet de Marselha, Raimon de Miraval, or the anonymous stanzas) there is a very close correlation between quoted passages and extracted stanzas, whereas for others (like Aimeric de Belenoi or Giraut de Bornelh), even though they are both extensively quoted and represented in collections of anthologized stanzas, there is less direct overlap between the actual passages concerned. If these excerpted stanzas are nevertheless regarded as quotations-in-waiting, and added to the count of actual passages quoted, the total more than doubles.
The treatment of Occitan quotations and anthologized stanzas in Italy and Catalonia provides a context in which to reconsider the troubadour passages inserted into the works of the northern French romancers Jean Renart (Le roman de la rose, more commonly known as Le roman de Guillaume de Dole, ca. 1210–1230) and Gerbert de Montreuil (Le roman de la violette, ca. 1230). Jean Renart has hitherto been studied as the inventor of lyric insertion; however, this “invention” might be better described as a response to lyric quotation (more on this distinction below). The practice of quotation may also shed light on the star role assigned to a parrot in an amusing verse novas of the mid-thirteenth century, the Novas del papagai (Tale of the parrot), and its possibly fourteenth-century successor, Frayre de Joy e Sor de Plazer (Brother of Joy and Sister of Pleasure).14
In the last third of the thirteenth century, we find lyric poets reaching for quotations in ways that have implications for the melodies of songs as well as their texts; I examine two examples of this practice, one Italian (Bertolome Zorzi) and one Catalan (Jofre de Foixà). Between 1288 and about 1292, Matfre Ermengau of Béziers writes the Breviari d’amor (Love’s summary), a long verse encyclopedia describing the universe as emanating from God’s love. One section of it is studded with over 260 lyric quotations, which Matfre interprets as a means to know the true nature of this love and so reverse our fall into carnal desire. The Breviari marks a high point in the work done by troubadour quotation to transform our relation to desire and knowledge. Unusually for the history of troubadour quotation, it is not only in Occitan, but composed in Occitania by an Occitan for Occitan compatriots. Although there is no evidence that Dante knew Matfre’s text when composing his Commedia, there are striking similarities between the ambitious hermeneutic, cosmological, and theological frameworks in which the two authors place the troubadours. Dante certainly did know the Occitan treatises of grammar and poetics, and modeled on them his De vulgari eloquentia (composed between 1303 and 1305). It is because of this debt that I include the De vulgari, despite the fact that it is in Latin. (A few other Latin works of this period quote the troubadours but are not discussed in this study.)15
One of the last in date of the texts I consider is, like the first, an Occitan treatise of grammar and poetics, Guilhem Molinier’s Las leys d’amors (The laws of love). This summa of poetic knowledge was compiled for the poetic academy in Toulouse that styled itself the Consistori de la sobregaia companhia del gai saber (Consistory of the most joyous company of the gay science), henceforth shortened to Consistori de gai saber. It went through various redactions between about 1323 and 1356, in the course of which quotations from the classical troubadours recede from prominence and new styles of quotation are established. Although the Leys and the activities of the Consistori inspired further Occitan works of grammar and poetics in both Occitania and Catalonia, and the practice of quotation continued sporadically in Catalonia into the fifteenth century,16 I shall not follow these developments but instead end with Petrarch, in one of whose songs Occitan is quoted only in order to be superseded.
The overview I have just given indicates the trajectory of this book and the works covered in the individual chapters that follow. An opening group examines the pioneering forms taken by the scholastic address to troubadour poetry: the beginnings of troubadour quotation in the grammatical tradition (Chapter 1); the translatio of troubadour poetry to Catalonia in the novas of Raimon Vidal (Chapter 2); the role of quotation in vidas and razos in relation to the design of chansonniers (Chapter 3); and the influence of florilegia of excerpted stanzas (Chapter 4). Chapters 5 and 6 consider the reflections on quotation offered by Jean Renart’s Guillaume de Dole and by the Novas del papagai, taken as representing, respectively, what I call the nightingales’ way and that of the parrots. Chapters 7 to 11 deal with works written from the latter part of the thirteenth century onward. The pioneering days behind them, authors treat quotation ambitiously and self-consciously in works that progressively transform the preceding tradition. The playful use of lyric quotation in the lyric (Chapter 7) contrasts with the theological transformation of florilegia in the Breviari (Chapter 8). While Dante tends to occlude the earlier history of quotation, he nevertheless achieves with it some dazzlingly innovative effects (Chapter 9). Chapter 10 shows how the Leys d’amors goes further in this direction, wiping the slate clean so that vernacular poetry can begin again. This situation of semi-erasure of troubadour quotations is the starting point for Petrarch’s “Lasso me.” In Chapter 11, I contend that Petrarch’s song illustrates how quotation has enabled the poetic subject to desire to desire differently. The Middle Ages of the troubadours are relegated to the past, and the birth is confirmed of a poetry that will sweep across Europe under the new banner of Renaissance. As an engine of change, quotation has revolutionized the first-person subject of poetry and with it the kind of poetry that can be produced and enjoyed.
The remainder of this introduction provides a general historical and theoretical framework within which to situate these chapters.
Appendix 1 is set out so as to show which troubadours, which songs, and which passages are quoted, using the standard Pillet-Carstens (PC) system of identifying authors and songs.17 The most-quoted troubadours date from the third quarter of the twelfth century to the first quarter of the thirteenth, from Bernart de Ventadorn to Aimeric de Peguilhan. The fact that the later troubadours simply have not yet begun their careers in time to be quoted by the earliest works containing quotations obviously affects the frequency with which they appear in this table; nevertheless, a kind of golden age seems to persist in the imagination of later writers, coloring which troubadours are perceived as most quotable at least to the early fourteenth century. The most prevalent genre to be quoted is the canso, or love song, which can be regarded as the canonical form of troubadour lyric even though cansos in fact make up only about 40 percent of the surviving corpus. Bernart de Ventadorn is the most popular troubadour, not only in terms of the number of passages and songs that are quoted but also in the variety of works that quote them. His “Ab joi mou lo vers e.l comens” (70.1) is the most quoted song in the corpus: not exactly a predictable winner, though it was certainly widely copied and imitated in the Middle Ages.18 Arnaut de Maruelh, Folquet de Marselha, Gaucelm Faidit, Giraut de Bornelh, Peire Vidal, Raimbaut de Vaqueiras, and Raimon de Miraval are also all extensively mined. Alongside the cansos, the nonlyric verse ensenhamens of Arnaut de Maruelh (“Razos es e mezura”) and Garin lo Brun (“El termini d’estiu”) are widely aired, preparing a place for the similarly nonstrophic verse of the later At de Mons.
Some troubadours with long entries in Appendix 1 owe most of their apparent prominence to a single work or group of works that quote them. This is the case for Bertran de Born (because of the many razos devoted to his sirventes), Aimeric de Peguilhan (favored by the Breviari), and At de Mons (chief recourse of Guilhem Molinier in both redactions of the Leys). The emergence of a preferred troubadour is one of the hallmarks of the art of quotation; Dante’s preference for Arnaut Daniel is as unique to him as Raimon Vidal’s highlighting of Raimon de Miraval in So fo e.l tems, but the fact of having a favorite seems, by Dante’s time, to be an established convention. While some troubadours are inevitably exploited more than others, the breadth of reference overall is impressive, with virtually every work reaching for some little-quoted poet or song. The tabular form of Appendix 1 makes it easy to see how little convergence in the choice of extracts there is. Quite the reverse: there appears to be a concerted effort at dispersal. One of the more amusing aspects of the authors’ drive to personalize their selections is the tendency to slip in, alongside more celebrated songs, snippets of their own composition. Thus of Raimon Vidal’s five lyrics, three are known only from self-quotations in So fo;19 and Matfre Ermengau secures his own posterity as well as that of other troubadours in his family by quoting his own and their songs in the Breviari. The Leys contain verses likewise seemingly confected by Molinier himself, on an unprecedentedly large scale. Dante’s predilection for self-quotation also indicates his adherence to earlier usage.
Despite the broad range of quotation, some troubadours seem noticeably absent or underrepresented in Appendix 1. The tiny number of quotations from the very early troubadours Guilhem de Peitieu, Jaufre Rudel, and Bernart Marti, and the total absence of Cercamon and Alegret, may not be surprising given their low profile in the chansonniers. More surprising are the very small number of quotations from Marcabru and the quasi absence of Peire d’Alvernha, since the lyrics of both troubadours circulated widely. The case of Peire is especially striking since his poems open a substantial group of Italian songbooks and his name features on lists of famous troubadours compiled by Dante and Petrarch. Another shortfall, which may go some way to explain the relative absence of Marcabru and Peire, affects songs we might classify as clus (composed in an obscure or hermetic style). As will be abundantly confirmed throughout this study, those who quote the troubadours tend to home in on nuggets of thought; they seek illumination from what is quoted, rather than aiming to throw light on it. Thus it is that, as well as Marcabru and Peire d’Alvernha, Raimbaut d’Aurenga, the hermetic poems of Giraut de Bornelh, and indeed the whole corpus of Arnaut Daniel are underrepresented in Appendix 1 in comparison with their standing in the transmission elsewhere.
An absence of a quite different kind is that of songs by women and dialogue songs, which perhaps do not seem sufficiently substantial or sententious. Women’s voices make no appearance at all apart from the stanzas included within prose razos, the song attributed in manuscripts to Raimon Jordan (404.5), and the likely fictitious dialogue between Aimeric de Peguilhan and a lady (10.23). The fact that dialogue pieces are often composed for specific circumstances may also make them less mobile and transferrable than cansos. Songs by nonnative Occitans are also extremely rare, maybe because the grammarians, who inaugurate the practice of quotation, have scruples about linguistic authenticity (a notable exception is the showcasing of songs by Catalan poets in the Ripoll treatises).20 With only minor differences of emphasis, the same generalizations hold true for the entries in Appendix 2.
Thus, amid the diversity of quotation, certain conventions develop and some of the ways in which quotations are used in earlier works—the early grammars, novas, vidas, razos, and early florilegia—are adopted by the later ones, including those of Dante.
The fact that quotation, unlike citation, is constituted by verbatim repetition means that it promotes the troubadours’ language as well as their poetry. And indeed, a striking feature of the list of works containing troubadour quotations is the dominance among them of treatises of Occitan grammar and poetics. Occitania is an area corresponding to roughly the bottom third of what is now France and extending down into the valleys of Piedmont. It comprises many regional dialects, but the Occitan of the troubadours is already fairly standardized in the twelfth century and is not identified with any one of them. The grammarians thus have an already poetic, to some extent codified, idiom with which to work. But they innovate decisively when, in a move mirroring the respect for its poetry that quotation enacts, they elevate this idiom to a grammatical language on the model of Latin. Indeed, it is noticeable that all the texts containing quotations, although divergent among themselves, have Latin analogues (treatises, didactic poems, debate poems, accessus ad auctores, hymns, encyclopedias), and that their use of quotation bears an edgy relation to the quotation by Latin authors of extracts from the Bible or from ecclesiastical or classical texts. Is such proximity to Latin broadly parodic, or merely a vernacular parallel? In any case, its effect is increasingly to consecrate Occitan as a kind of neo-Latin whose poets are authorities and to identify its literature as an opportunity to extend scholasticism to the vernacular. Scholastic models are likewise explored relative to vernacular texts by thirteenth-century authors from northern France (Henri d’Andeli, Gautier de Coinci, Richard de Fournival, Jean de Meun), but the mode of textual quotation is unique to writers in Occitan.21
It is striking that many of these Occitan writers are either Italian or Catalan, and/or have Italian or Catalan audiences and readers in mind, even though they all quote the troubadours in the original and, except Dante and Petrarch, adopt Occitan for all their known compositions. Only Matfre Ermengau and Guilhem Molinier are Occitans writing in Occitania seemingly solely for fellow Occitans, and even their works quickly give rise to Catalan traditions. Clearly, Occitan enjoys a status that today we would call “international,” though the term is hard to justify at a time before the development of the modern nations. Perhaps the best way of describing it is as a mobile, literary standard. As such, Occitan has a significantly different status from other, emergent, vernacular literary languages in that it is not grounded in the political realm.22 It was not and never has been the language of a nation or a people and in this respect again it has a quasi-classical status, resembling Latin, Hebrew, or Arabic more than do, for example, German and French.
Why did those who copied and repeated troubadour poetry assist in consecrating Occitan as such a language? And why was such a language so widely copied and repeated? In trying to answer these questions, I have been helped by Jacques Derrida’s controversial assertion, in Le monolinguisme de l’autre, that culture as such is colonial. In his own experience of growing up as an Algerian Jew, Derrida recounts how French appeared as a prestige literary standard (albeit one with all too many political associations). One passage in particular suggests a parallel with the way literary Occitan seems to have attracted so many speakers. Here Derrida describes the seductive lure of identification offered by an alien, written language, even though the forms that promise a sense of identity by their very nature deprive one of it: “People always imagine that a man or a woman who writes must know how to say I. At any rate, the potential for identification must seem assured: both assured within language and assured of it.”23 In this respect, prestige languages are revealing of language in general: the effect of simultaneously finding a self and losing one’s self within a foreign language is only the most visible instance of the inevitable impact of any language and of language as such: “This I would have formed itself, then, at the site of a ‘situation’ which cannot be located but instead is constantly referred on to somewhere else, to something else, to another language, to the Other in general. Its site would be in an experience of language as such, language in the broad sense of the word, one that cannot be situated.”24 Derrida’s remarks can help us to figure why so many medieval speakers quote the first-person poetry of the troubadours. Assuming the prestigious discourse of the lyric promises to constitute an I that would enable them to situate themselves alongside other authors in that discourse; it beckons with the imaginary assurance of finding themselves at home in a language that, because it is a mobile standard, has no home; it offers to identify them in a language that, in itself, lacks identity.
Buoyed up by prestige and unencumbered by political or ecclesiastical baggage, literary Occitan offers the dream of a genuinely new, secular and lay, poetic subjectivity. Quotations from its poetry are like seeds from which a new, secular and lay poetry can grow.
Derrida’s remarks help draw attention to one of the most fundamental properties of the Occitan first-person lyric: its capacity to inspire the desire to repeat it, to assume one’s place in it, and to assume a personal relation to it, however self-defeating such a desire for identity may be. Repetition of this poetry takes many forms, of which quotation is just one. Viewing quotation in this wider context will help focus its significance.
In twelfth-century Occitania the troubadours are involved in intense interactions with one another, debating and reworking one another’s songs. Probably the most elaborate technique of imitation devised in this period is contrafacture, a technique whereby a new song is modeled on the stanza form of an existing one and can be sung to the same music. Such formal and musical quotation is the counterpart of the textual quotation studied here, in which words are typically detached altogether from music. As if aware of this complementarity, some of the works that exploit textual quotation—like Raimon Vidal or Jofre de Foixà—seem playfully to pit it against contrafacture. Contrafacture enables troubadour song to percolate—not as text, but as form or melody—into the cultures of all the regions neighboring Occitania.25
Other ways of recycling the troubadours present different emphases from region to region. North of Occitania, in northern France and Germany, the closest form of repetition to be widely practiced is the translation of Occitan lyrics into German or French.26 (Some are also translated into Italian by the Sicilian school, seemingly under the influence of the German Frederick II’s Sicilian court.)27 Literal quotation of the troubadours in the original Occitan is found only rarely in France, and the original is regularly Gallicized (that is, assimilated linguistically to Northern French). This is the fate of the troubadour songs quoted in Guillaume de Dole and the Roman de la violette, and also of the troubadour stanzas copied in manuscripts recording Northern French texts.28 In France and Germany, and later in Sicily, the troubadours also serve as models for imitation by native poets in their own vernacular; the Occitan originals are rapidly absorbed into evolving domestic traditions. Because of this emphasis on recreation rather than verbatim repetition, I call this path the nightingales’ way; the importance of the Sicilian school makes such a label preferable to defining this mode of reception as exclusively “northern.”
Among the troubadours’ immediate Mediterranean neighbors, in what are now parts of Spain or Italy, there is a much greater tendency for native poets to imitate troubadour poetry in Occitan.29 Among the earliest examples of romance vernacular song from Aragon, Catalonia, and Italy are works of poets whose first language is Catalan and/or Italian, but who choose to compose using the language as well as the forms of the troubadours, and they are intended for readers or listeners whose first language is likewise Catalan and/or Italian. It is these same regions, and the Occitanian homeland in between, that give rise to the works studied in this book: works composed in Occitan that quote the troubadours very expertly likewise in Occitan. In these regions, then, the desire to repeat operates first of all at the level of language; either the troubadours are repeated verbatim, or more broadly their language is mastered so as to provide a recognizable subject position (Derrida’s monolangue). Although this path does not exclude song, it tends to identify lyric above all as language and as text, and I call it the parrots’ way.
One reason for the division between the two ways lies in Occitan political and cultural history. Occitania was not politically unified, since most Occitan-speaking territories were nominally held either from the French king or the Empire, and all were relatively autonomous both from their overlords and from one another. Their strongest cultural ties were with Iberia and Italy, and these are also the regions the languages of which Occitan most resembles; it is especially close to Catalan and indeed may not always have been distinguished from it. Iberian courts were among the very earliest to extend patronage to troubadours, followed by those of northern Italy in the late twelfth century; troubadours had far less direct contact with northern France.30 Alfonso II, king of Aragon from 1162 to 1196 and also, as the count of Barcelona, ruler of Catalonia, adopted Occitan as the literary language of his court in order to bolster his claim to Provence.31 Catalonia was an important sea power with aspirations to dominate the northern Mediterranean; the Crown of Aragon helped to diffuse Occitan, and with it troubadour culture, across Catalan-held territories, which included at different times Sardinia, Sicily, Naples, and the Balearics.
If Catalan imperial ambition boosts the status of Occitan, in other respects the prominence of troubadour parlance beyond the Alps and the Pyrenees seems rather to result from the vulnerability of Occitania. By the end of the twelfth century the troubadours were frequent visitors to both regions, but their presence seems to have increased in the thirteenth in response to the Albigensian Crusade of 1209 to about 1229. Begun as a holy war against Cathar heretics in the Midi, this war was rapidly interpreted by the Occitans as a campaign for the closer control of the south by the kingdom of France. The crusade’s impact on troubadour poetry has often been overstated; it may not have precipitated widespread literary decline;32 nevertheless it resoundingly confirmed Occitania’s propensity always to be more intimately involved with its southern neighbors than with its northern overlords. Troubadours like Uc de Saint Circ relocated in the Veneto at a time when French barons were seizing control of southern lands, exiling native landholders, and introducing the Inquisition. The survival of a hitherto primarily oral culture may have seemed precarious, the need correspondingly urgent to compile and preserve it in writing.
Quotation, then, forms a strand in troubadour reception that is influenced by both cultural and political factors. It takes its place among a whole complex of modes of recycling that range from contrafacture to compilation, but of which different forms predominate in different regions. The main differences overall are that the parrots’ reception emphasizes the Occitan language (and, concomitantly, grammatical and textual expertise in Occitan works) and that it stresses the value as knowledge, sententious or revelatory, of the troubadours, whereas the nightingales’ reception is more interested in assimilation, affect, and song. The differences between these two ways are epitomized by the point at which they are closest to one another: in the practice of lyric insertion as compared with that of lyric quotation.
The importance and interest of lyric insertion have been brilliantly demonstrated in recent years.33 Inaugurated in northern France in the early thirteenth century, the art of including lyrics within a longer work is traced from Jean Renart through the thirteenth century to the great dits amoureux of the fourteenth and on. Some of the lyrics that are inserted could, indeed, also be said to be quoted. But while, for example, Jean Renart identifies some of his inset pieces as having been composed by others, he leaves the origins of others deliberately opaque (see Chapter 5). And by the fourteenth century, rather than quoting other people’s poetry, northern French authors like Guillaume de Machaut insert their own formes fixes in their mesmerizingly sophisticated dits. Although the authorship of these lyric pieces is sometimes attributed to characters within the fiction, there are very few cases where they are clearly the work of anyone other than the poets of the dit concerned.34 In some cases the presence of the same songs in the lyric sections of the same manuscripts as the dits further dispels doubt as to their authorship.
An important implication of this development toward single authorship in lyric insertion is the assurance of having distanced the memory and authority of the troubadours, of having created one’s own tradition. The creative skill and poetic versatility of the nightingales’ way can be dazzling. In the parrots’ way of quotation, by contrast, not only are excerpts from other poets carefully preserved and repeated, but the sense of a multilingual environment is fostered, since in the majority of cases the authors and implied readers of the works that quote them are speakers of Catalan and/or Italian. If lyric insertion aspires to a national culture, quotation is exiled, diasporic, and cosmopolitan.35
Lyric insertion is found from the start in texts that are largely narrative (though they may have a didactic dimension), whereas verbatim quotations occur in texts that are more instructive or moralizing in character, and often in what are straightforwardly treatises. Insertion serves various narrative functions but the most typical, as Maureen Boulton shows, is to reflect or voice emotion; songs are also sung to jollify social occasions or to accompany social activities (such as women’s work).36 By contrast, as I have already said, quotations tend to home in on the moral, reflective, or other sententious aspects of troubadour poetry, inscribing it less in the field of affect than of knowledge. While lyric insertion typically imports songs into a fictional world where they are sung by characters within the narrative, accompanying their dancing or needlework, most Occitan texts that quote troubadour lyrics more closely resemble compilations of diverse voices. The differences between insertion and quotation are well exemplified by Dante. His own lyrics are inserted in Vita nuova and are a far cry from the lists of first lines from other poets reeled off in De vulgari eloquentia. (The fact that Dante uses both techniques confirms that the nightingales’ way is not exclusive to a particular region; the parrots’ way is less widely diffused.)
Boulton pinpoints the distinctiveness of lyric insertion when she observes that “the heart of the device seems to lie less in quotation or in borrowing than in combining genres.”37 Although the earliest examples seemingly involve only short excerpts (stanzas or refrains) set in verse narrative, the practice becomes more and more flamboyantly heterogeneous as entire lyric poems are included, they become increasingly intricate formally, and/or prose is chosen for the surrounding narrative (as in the Prose Tristan and the Perceforest). Up to the dits of Guillaume de Machaut, inset lyrics are also usually represented as sung, and seemingly were intended actually to be so; they would thus stand out in performance from the remainder of the work, which would be recited or read aloud. In Jean Renart’s Guillaume de Dole, characters behave as if they were in a karaoke bar, delivering spirited if not always convincing performances of recent hits. The hybrid nature of texts with lyric insertions is reflected in their transmission, with many manuscripts including musical notation or marking the lyrics off from their narrative setting in some other way.38
Lyric quotations, by contrast, are usually assimilated to their context as much as possible. They seem never expected to be sung except when they are found in other lyrics. Formally, they tend to be treated by didactic poets as if they were passages of didactic verse; authors of prose treatises often opt for quotations of a single line, that is, of a passage so short that its metrical form is hard to detect in the surrounding prose. This tendency also persists in transmission. Manuscripts of works in rhyming couplets usually set out quotations in the same format as the rest of the text, in columns and line by line; and manuscripts of prose texts include the quotations within the flow of the prose.39 The influence of Latin practice in calling attention to quotations is found in razos in some chansonniers where quoted incipits (first lines) are written in red ink, and in some Breviari manuscripts where the name of cited troubadours may be underlined or copied in colored ink.40 But none of the texts I have looked at adopts the more flagrant signaling of quotations found in some Latin manuscripts, where the whole of a quoted passage may be underlined (often in red), overlined (by a pen stroke, often red, through the middle of the words concerned), or picked out by a column of diples (little s-like marks) in the adjacent margin.41 If lyric insertion involves a poetics of disruption, as Boulton aptly calls it,42 the poetics of quotation is rather one of mise en abyme in the basic sense defined by Lucien Dallenbach of “any insertion grounded in a relation of similarity to the work which contains it.”43
Such similarity gives rise to indeterminacy in identifying the limits of quotation. In fact, as already suggested, quotation in general plays with expectations of knowledge and recognition, whereas lyric insertion speaks more to the expressive qualities of the lyric text. Focusing on recognition (in quotation) or expressivity (in lyric insertion) has implications for subjectivity in the texts concerned. The first person implied by lyric insertion is typically a subject of desire—for love or for social involvement—Identified as a character within the fiction. This subjectivity, strung between the discursive poles of narrative and lyric, invites analysis in terms of the relations between these respective genres. How successfully do romance and lyric discourses combine to constitute a single subject? Do they complement or ironize one another? By contrast, in the more academic climate of the Occitan texts, we find not a single subjectivity defined (or not) between several genres, but rather a discourse of knowledge that connects together a network of potential subjects—minimally the one who quotes, the one who is quoted from, and the reader trying to fathom what is going on. While the concept of intertextuality is obviously still relevant,44 it offers less critical leverage here, I think, than that of intersubjectivity, by which I mean the relations forged between and thereby constituting subjects. When a writer chooses to quote some troubadours and not others, for example, he indexes a position relative to other subjects (including the troubadours he quotes and his readers) in a supposed knowledge community in which every subject is potentially both connected to and differentiated from the others. Subsequently, as Anthony Grafton has wittily shown, footnotes similarly found an economy of acknowledged and unacknowledged knowledge that is understood differently by different subjects.45
French lyric insertion exhibits the same innovative brio as the trouvères when they recreated troubadour lyrics in the langue d’oïl. The dynamism of this nightingales’ way seems to guarantee its future success. And yet the kinds of lyric verse that are essayed in lyric insertion sputter out in the sixteenth century, when the tide of taste turns in favor of Petrarchan models imported from Italy. The parrots’ way, although less glamorous than the nightingales’, wins out in the end. Lyric quotation is not of course solely responsible for this success; but it contributes to it by helping to keep alive a sense of poetry’s value as a mode of knowledge as well as of desire. The troubadours are mined for their insights by the authors who quote them; quotations test readers’ competence to recognize them; and the whole practice of quotation identifies troubadour poetry as something worth knowing about. For over 150 years before the Consistori at Toulouse decided to test candidates’ mastery of poetics and award them degrees and diplomas for it, writers emulated one another in exhibiting their command of the Occitan lyric tradition, its poets, their lives and opinions, as well as their actual words. The Occitan lyric corpus is at once deferred to as a source of knowledge and displayed as a culturally desirable object of knowledge.
It is in the interplay between intersubjectivity, desire, and knowledge that I believe the answer lies as to why quotation helps to produce subjective and cultural change.
From the very beginning, troubadour lyrics intertwine desire and knowledge. They express desire; they lay claim to knowledge of desire; they inspire, in different proportions in different listeners, the desire to desire and the desire to know; these in turn fuel the desire for lyric as both a vehicle and an object of knowledge; and a discursive field is created in which different subjects of desire/knowledge are supposed. This introduction has already spoken of the knowledge conveyed by quotations, the knowledge displayed by those who quote, and the challenge to knowledge that quotation poses to the reader or audience. Here I would like to sketch the transferential framework as a result of which these cognitive exchanges generate change, understood as a new alignment of the subject with the symbolic fabric that enables a new articulation of knowledge and desire.
I have proposed that the challenge to recognize quotation differentiates it from other forms of textual repetition. Similarly, Herman Meyer suggests that we distinguish between borrowing from a work and quoting it because, unlike quoting, “borrowing . . . has no referential character.”46 He means, I take it, that quoting inevitably involves referencing the extract’s origin in some way. How, though? Derrida’s well-known essay “Signature événement contexte” suggests that the answer is more elusive than we realize: “Every sign . . . can be quoted, put between quotation marks; in so doing it can break with every given context, engendering an infinity of new contexts. . . . This quotability, this duplication or doubleness/duplicity [duplicité], this repeatability of the mark is not an accident or an anomaly. . . . What would a mark be that could not be quoted? And whose origin could not be lost en route?”47 Given that any utterance is indefinitely repeatable and hence indefinitely quotable, we cannot rely on there being a stable origin that can be “referenced.” Perhaps we should say, then, that what characterizes quotation is that it foregrounds this repeatability (what Derrida calls its doubleness/duplicity [duplicité]), and provokes the reader to recognize it?48 Quoting will always be perilously close to other procedures of writing that rely on iteration; to differentiate it from other kinds of copy, a quotation will depend only on this tantalizing supposition that it has an original (or at least a second) home. Responding to quotation requires an act of recognition that—Derrida implies—is always only provisional and liable to misprision, since all these other forms of repetition are also capable of manifesting their own doubleness, or of having it variously recognized by different readers.
When it comes to recognizing the original home of a quotation, the instability of medieval texts creates particular difficulties. Quotation typically involves partiality, in the sense both of division and of preference. Troubadour songs are not long, but it is exceptional for them to be quoted in full. Different works have different preferred lengths of quotation: in the novas it is the stanza, in razos the incipit (first line); the Breviari and the Leys are alone in including long extracts from didactic poems. Antoine Compagnon has baptized “solicitation” this capacity of a text to differentiate itself internally in such a way that a part of it calls out to be removed from its context and redeployed in another.49 His vocabulary of seduction softens what others have identified as a more violent process. “To quote a text is to break into it, to ‘tear’ something out of it,” writes Marjorie Garber.50 Yet how can we know that a passage is indeed a part when the notion of “whole” is problematic?
In some cases, the effect of internal division is undoubtedly important. While a portion of the text is manifest in the quotation, the remainder is latent alongside it. This is so when what is quoted is an opening line since the incipit doubles as a title, and so stands for the whole work from which it is taken, it is both part and whole (see below). Isolated excerpts can likewise require supplementing with other material from the same song, as I contend happens in Raimon Vidal de Besalú’s novas (Chapter 2) or Dante’s use of Arnaut Daniel (Chapter 9). On the other hand, quotations were not always hewed purposively from whole songs; as we have seen, they might be found ready-made in cobla collections (Chapter 4). Readers of these anthologies may well have been unaware of the rest of the songs from which they were taken and unable—as we are—always to be able to distinguish coblas triadas (excerpted stanzas) from coblas esparsas (freestanding stanzas).51 Manuscript copies, in any case, preserve shorter and longer versions of individual songs. In sum, it is virtually impossible to specify the exact outlines of the original context of any given quotation (supposing such an original existed) or to assess how much knowledge needs to be assumed to supplement it. This uncertainty is not only experienced differently by different subjects, it also actively fosters the idea that there are other possible subjects possessing a different level of knowledge.
Another difficulty in the way of recognizing quotations is grammatical. Verbatim quotation is closely related to direct speech, what we call “quotation marks” being used in modern punctuation to identify both. The way manuscripts usually signal the beginning and end of quotations, with enlarged initials or paraphs or touches of red ink, corresponds with how direct speech is marked in other vernacular texts, rather than with how quotations are marked in Latin ones.52 The persistence, or the shadow, of the unquoted around that which is quoted means, however, that pragmatically quotation is closer to free indirect discourse, in which two different discourses—the one quoting and the one quoted—are so superimposed as to make it difficult to establish the boundaries between them, or between the subjectivities involved.53
Quotation therefore plays with expectations of knowledge and recognition; it summons subjects of knowledge and recognition into existence; but it does not necessarily ratify them. In trying to get to grips with this phenomenon, I have adopted Jacques Lacan’s concept of “the subject supposed to know” (“le sujet supposé savoir”), because it means both that knowledge presupposes a series of subjects that are difficult to locate, and that subjects are supposed to have knowledge that is difficult or impossible to specify. Contending that everything we call “knowledge” is articulated against the background of the unknown of the primary repression and designed to protect us against whatever that unknown contains, Lacan postulates the subject supposed to know as the condition of all symbolic transactions and the reason why they are shot through with so much confusion and deceit. In particular, he views it as a crucial part of the psychoanalytical transference, where the subject supposed to know is both a necessary vehicle for the exchanges between analyst and patient and that which makes it possible for the patient to benefit from therapy. By observing the feints and displacements of the subject supposed to know, the patient is enabled to see that the subject as such is merely supposed, hypothetical; this perception enables him or her to disconnect from the symbolic web in which all subjects are entangled and experience a transforming interval of radical freedom before reconnecting to it once more, potentially in an altogether new way. Outside the domain of therapy, this book contends, the same cycle of repetition, (mis)recognition, freedom, and reconnection can also enable subjective change: the rearticulation of the subject of poetry from medieval courtly lover to Petrarchan poet.
I will not repeat here the theoretical complexities of this argument, which I have worked through elsewhere (and see Chapter 8).54 Instead I exemplify how knowledge and subjects are supposed in quotation, and how such suppositions can be transformative, in the case of a particular form of quotation—that of an incipit—which we will encounter many times in the following chapters.
The principal use of the incipit, in an age when works did not usually have titles, was to act as one. In many cases the incipit effectively became the title (e.g., Genesis, meaning “in the beginning”), or one of several alternative titles. D. Vance Smith observes that a commonly quoted incipit is “In principio” (In the beginning), the opening words of the Gospel of Saint John, and thus, like the Genesis example, an incipit that is itself about the act of beginning.55 Medieval scholars were expected, then, to recognize a text by its beginning and to identify it with its beginning. This gives the incipit enormous symbolic value, both as inaugurating something and, at the same time, as being that which is inaugurated.56
In addition to acting as titles, incipits were also used to introduce quotations in an abbreviated form. Well-known texts such as biblical passages could be quoted by their opening words alone, and readers would be expected to supply the remainder from their own knowledge. Here the incipit is not a fixed beginning like “In principio,” but a mobile one whose position depends on where in a text the decision to start quoting falls. The implication of quoting from within a text by means of an incipit is that the text is replete with beginnings—that it could begin again at any point. The challenge thrown down by quotation of an incipit is that of recognizing and being able to recall (enough—how much?—of) what follows. For D. Vance Smith, an incipit is not so much a point in time when something starts as the opening up of an enigma.57 Quite literally, incipital quotation is an invitation to begin again: to initiate (even if one appears to repeat it) a symbolic realm. Edward Saïd has identified this initiation as a specifically secular form of innovation. “Beginnings,” he writes, “inaugurate a deliberately other production of meaning—a gentile (as opposed to a sacred) one. It is ‘other’ because, in writing, this gentile production claims a status alongside other works: it is another work, rather than one in a line of descent from X or Y.”58 Saïd does not write from a psychoanalytic standpoint but his formulation captures exactly the mechanism of rupture and potential alteration that I have just described.
Closer to Lacan, psychoanalyst Jean-Bertrand Pontalis focuses on the potentially transformative value of inauguration when he takes as the title of his autobiography the “love of beginnings” (L’amour des commencements). As a therapist, his aim is precisely to help people to change (get better); to do so, their existing relation to reality needs to be temporarily suspended so that they can reconnect to it in a new way. As Pontalis puts it, “When words fail, it is because, without realizing it, one is about to touch a different earth.”59 A beginning institutes a break with what precedes, and it is this break where “words fail” that makes it possible to start afresh, “touch a different earth.” Fresh starts precipitated by moments of doubt and hesitation are what make up all lived experience, making “the love of beginnings” a fitting title for an autobiography.
Incipital quotation, the quotation of a beginning, creates the chance to “touch a different earth” because its demands provoke the momentary hesitation that enables the subject to begin again.60 It supposes a text that the one who quotes it knows; it supposes a knowledge enshrined in the text that is thereby supposed; it supposes knowledge of that text on behalf of the one who quotes, and it further supposes an act of recognition on the part of those who read it. In Saïd’s word’s again, “such a beginning authorizes; it constitutes an authorization for what follows from it.”61 If the incipit is attributed to a well-known author, then the fact that he is supposed to be knowledgeable enhances the value of the quotation; at the same time, quoting that author supposes that he is recognized as an authority, and helps to confirm him as one. But the fact of reaching for this quotation also supposes that the one who quotes is also an authority, one who commands the field of knowledge, not just the isolated incipit but troubadour poetry in general. It lies in the power of such an expert to make or break a poet’s authority and canonical status. Authority is supposed as being transferred away from the one quoted to the one who quotes. But this authority must be supposed to be recognized at least by some readers. Without a third person to do this at least some of the time, authority collapses, and so this third knowing subject has to be supposed. It can be supposed to such an extent that, even if the quotation appears banal or vapid, or not to mean quite what it is implied it means, it still carries weight.
Thus the reader is lured in by what Roger Dragonetti calls, in his book of the same title, “the mirage of the source”: “To recognize an author, praise him and take him as one’s guide may be an elegant means of creating an empty space behind the stage set of the source.”62 Incipits may be quoted and never expanded, as happens with those inserted into the biographical sequences in chansonnier N2.63 The incipit quoted in Garin d’Apchier’s biography in IK is from a song that is found nowhere else. This supposition that there may or may not be a knowing subject elsewhere infuses the act of quotation with suspicion. What looks like quotation could just as well be a fabrication, or it could be a misremembering or a misrepresentation to a point that could be called a suppression. Response to an incipital quotation is equally capable of mangling the text to which it might be supposed to refer.
All such exchanges mean that incipital quotation is capable of transforming the meaning of the text from which it is taken. Passing to and fro between incipit and the supposed expanded text does not stabilize interpretation but sends it spiraling. Concomitantly, a subject’s own position can veer dramatically. He or she need no longer be subject to the suppositions of others but recognize them as the suppositions that they are. Encountering supposed knowledge in its supposed form, a subject is free to forge his or her relation to it in a new way. Subjectivity changes as it passes through the detours of (supposed) knowledge. The following chapters will trace the paths of this detour, as a result of which the symbolic relation to desire is reforged.
The idea that the act of quotation is transformative was put forward some time ago by Umberto Eco. Using a sartorial metaphor, Eco shows how medieval Latin writers, when quoting their forebears, perennially succeeded in refashioning them, even though they always cut their garb from old cloth.64 Although the tailoring metaphor is undoubtedly medieval, in Old Occitan parrots are a better emblem of quotation. Despite being seemingly disadvantaged relative to nightingales by a supposed lack of creativity, medieval parrots have the capacity to sing as well as talk, whereas nightingales can only sing, they cannot speak.65 Thus parrots embody the fundamental human capacity for mimesis.66 Although nightingales’ mastery of song is natural, parrots have the advantage of being able to learn. Moreover, while they may learn to repeat what they have heard, parrots may also know what they are saying; in the Middle Ages some writers even attribute them “with [a] miraculous command of language and [a] fully developed human consciousness.”67 According to bestiaries, though some parrots are foolish and representative of sin, others are so wise and noble that they exemplify to human readers the qualities of a spiritual life.68 In both Occitan novas featuring parrots, the bird is considerably more resourceful and in some ways more human than the other protagonists. An exotic bird, the parrot is also an apt representation of the otherness of Occitan to the Catalan and Italian contexts of most of the works I examine. Their words may have an uncanny relation to the desires they seem to voice.
I do not intend my opposition between parrots and nightingales to be taken as discrediting nightingales, more as rehabilitating the kinds of imitation associated with parrots. Nor, in choosing to speak of nightingales’ way and a parrots’ way do I mean to insist too literally on the avian images. Nevertheless the central section of this book suggests that there were moments in the thirteenth century when writers did indeed resort precisely to the image of parrots and nightingales in order to characterize what was by then a divergence in the reception of troubadour poetry between northern France and the Empire on the one hand—represented in Guillaume de Dole (Chapter 5)—and Catalonia and northern Italy on the other—represented in various ways by the Novas del papagai (Chapter 6).69
This book is primarily about the parrots’ way. At its core is a practice of quotation, by which I mean not only verbatim repetition, but a repetition that calls into play a supposed act of recognition of a supposed act of knowledge, which, by supposing anchorage in anterior subjects, enables subjective renewal and hence, potentially, subjective change. It is this transformation of the lyric subject by means of quotation that I will now follow over roughly a century and a half, in Occitania, Catalonia, and Italy.