On the face of it, Turnus should have been a character that a mid-twelfth-century audience could have identified with and admired. Virgil's ‘ingens Turnus’ (‘mighty Turnus’, XII, 927) stands poles apart from the villains of the first phase of Old French literature, who are generally constituted as such by their status as traitors (Ganelon in the Chanson de Roland, Modret in Wace's Brut, Hoel in Ille et Galeron, numerous antagonists of the Normans in Benoît's Chronique de ducs de Normandie), cowards (Tedbalt and Esturmi in the Chanson de Guillaume), perjurers (Etioclés in the Roman de Thèbes), or outsiders whose beliefs and/or values place them firmly beyond the pale of courtly Christian society (the Saracen leaders of the early chansons de geste, plus the fils à vilain figures of Partonopeus de Blois and the Alexander romances). As a man who attempts to defend his homeland against alien incomers and his own honour against an outrageous breach of faith by his overlord, Turnus would appear to conform to the vernacular audience's conception of a hero rather than its opposite. Virgil himself calls Turnus a hero, and it is not hard to imagine the Rutulian prince portrayed in Books VII to XII of the Aeneid as the protagonist of a chanson de geste from the cycle des barons révoltés, resisting injustice and a royal volte-face after the fashion of Girart de Roussillon.

A toy-boy and an older woman may be part of the stock-in-trade of modern romantic fiction, but they are not the first thing that comes to mind when we think of the heroes and heroines of twelfth-century French romance. A more typical pairing might be Erec and Enide, the one in his early twenties and the other still young enough to be unmarried at the start of the romance that bears their names. Or, again, the hero and first heroine of Ille et Galeron, who meet when Ille has been knighted and has spent three years establishing himself, and when Galeron is almost certainly no older than he is. In Partonopeus de Blois, however, the hero is thirteen at the start of the action, and within a matter of days he has been seduced by the empress of Byzantium, a woman several years his senior. The narrator draws attention to the question of the hero's age by stressing that he was only thirteen years old and by insisting that, despite his youth, Partonopeus is fully developed, both morally, physically and socially:

Except where otherwise indicated, quotations and line numbers have been taken from the Collet-Joris Lettres gothiques edition, whose text of the main body of the romance is based on MS A. This edition also presents the alternative, shorter ending and the Continuation based on MSS B, G and T. Where alternative text exists, the editors have chosen a split-page format, with the text and translation of A printed on the upper half of each verso–recto pair and the alternative text on the lower half. On split pages, line numbers for A are given in standard parentheses ( ); line numbers for BGT are given in curly brackets { }. Elsewhere, line numbers appear in square brackets [ ] for text from B used to fill lacunae in A and for vv. 12117–4590 of the Continuation. I have not followed this practice because the combination of three different styles of brackets can be very disconcerting when line numbers are integrated into continuous prose. It is normally clear from my discussion whether I am referring to the ending of MS A or MS B, or to the Continuation (readers can check the positioning of the major lacunae in A by referring to the synopsis in Appendix 2). One potentially confusing feature of the Lettres gothiques edition is that the Continuation follows on directly from the shorter B ending; this means that the first 1,460 lines of the Continuation are printed on the lower half of pages whose upper half still presents the longer A ending of the main romance. All references here to this part of the Continuation therefore indicate the lower halves of pp. 662–825.

This synopsis is designed as a guide for readers who may not be familiar with details of the plot of the Old French romance, or with the alternative versions that exist for certain sections of the narrative. It also indicates the positioning of the narratorial interventions that are such a prominent feature of the text. Except where otherwise indicated, the narrative is broadly similar in each of the seven complete or near-complete manuscripts listed in Appendix 1. Where versions differ significantly, or where lacunae have been filled with text from other manuscripts, changes of base manuscript are indicated by prefixes in square brackets. Where the Collet-Joris edition gives both the text of MS A and the alternative version found in one or more other manuscripts, the prefix [A] indicates those passages that are printed on the top half of the page (pp. 618–25 and pp. 636–825); the prefixes [B], [G] and [T] indicate the text printed on the bottom half of the same pages. The short passage of alternative text from MS B on pp. 250–63 is summarised in a footnote.

Prologue and genealogy of the hero (vv. 1–498)

The poet praises God and his patron; he evokes spring and the singing of the lark and the nightingale, but declares that he prefers the subtler song of the oriole. He defends his decision to compose a narrative in the vernacular about the distant past, promising that his story will be of benefit to those who understand what it has to say. Authorities divide the world into three parts, the largest of which is Asia, where Troy is to be found.

The evidence presented in the previous three chapters reveals that the Partonopeus Continuation is undisputably the work of more than one individual, and that it took a considerable period of time for it to evolve into the form in which it has come down to us. That process of evolution seems to have consisted of several separate stages and to have involved at least four continuators: the original poet himself, who was responsible for the Anselot episode and part of the invasion narrative; one or more twelfth-century interpolators (who may have included Walter Map); a thirteenth-century remanieur who added the ‘Salut d'amour’, and the fourteenth-century scribe of manuscript T. One of the aims of this chapter is to ask whether we can put any more flesh on these chronological bones and come to a more detailed understanding of how and when the various components of the Continuation were assembled. We will also need to look further into why the romance was continued in the first place, and why this sequel to the adventures of Partonopeus and Melior took on the shape it did. The question of patronage provides a fitting conclusion to this chapter and to the book, for without the impetus provided by the shadowy figure who inspired and rewarded the composition of Partonopeus de Blois, Old French romance might not have developed in the ways with which we are familiar. We begin, however, with questions of poetic form and intertextual influence which may shed some more light on the genesis of this most disconcerting section of the narrative.

Bruckner was one of the first critics to draw attention to the experimental nature of much of the writing in Partonopeus de Blois, a work that she saw as setting a standard ‘for beauty and pleasure, for experimentation in form and fusion, that becomes exemplary for the romancers who follow’. Noting the mingling of different styles and genres in the Continuation, she comments in passing that ‘Anselot's story itself is a kind of lai told in the first person.’ This brief description of the 600 or so lines that contain Anselot's account of what happened to him after he was abandoned by Partonopeus in the Ardennes is rich in insights. The use of the term lai captures the self-contained nature of the story, which could quite easily be detached from the episode that frames it to stand alone as a tale of court intrigue and thwarted love. The idea of a lai which is also a first-person narration points to the startling originality of this passage: there are no other lais or equivalent short narratives from this period that present themselves in this way. When we look more closely at the form and structure of the story we find that it offers, in miniature, a summa of the experimental approach that Bruckner sees as characterising the romance as a whole. Multiple narrative techniques – extended first-person narration, écriture en abyme, doubling, repetition with variation of schemas from the main body of the romance, fusion of different intertexts – are brought into play and combined with the political subtext that we identified in Chapter 2 to create a truly innovative piece of fiction.

Although their primary focus lies elsewhere, the previous two chapters have both engaged obliquely with questions of narrative closure. Walter Map's possible involvement in expanding Anselot's tirade against the fils à vilain reminds us that interpolation is not something that ‘just happens’ to medieval texts. Rather, it is a conscious process, undertaken by a specific individual who either finds the original composition lacking in some way or wants to add his own contribution to it. Interpolation of this kind is based on the premise that a vernacular narrative is ‘open’ in a way that scriptural or classical texts are not: it can legitimately be expanded or modified to suit different tastes or circumstances. A different kind of non-closure is associated with Anselot's story. Our account of its development features an author who composes a second narrative in parallel to his first but leaves it unfinished, before incorporating this incomplete text into a continuation of the original romance. Even without the strong probability that the process of continuation required a significant rewriting of the first ending, this points to a general sense that vernacular fiction is not bound by the same concept of completion as other works with which readers and copyists came into contact. Precisely what our poet and his contemporaries thought about the nature of closure is more difficult to determine, however.

For much of the critical history of twelfth-century French romance, Partonopeus de Blois has been one of a group of texts that have existed on the periphery of mainstream scholarship (others include the romances of Hue de Rotelande, especially Proteselaus, Aimon de Varennes's Florimont and the anonymous Guillaume de Palerne). Writing in 1953, Sikko Pieter Uri lamented the fact that ‘the romance of Partonopeus seems nearly forgotten in the twentieth century’. Even in 1999 a critic could still claim, with some justification, that ‘the medieval bestseller Partonopeu […] is scarcely known today even by specialists in the field’. It is only in the last twenty years that it has really started to come out from under the critical radar and take its rightful place in discussions of how the genre evolved. The marginalising of Partonopeus is less a function of its intrinsic quality – the great Edmond Faral described it as one of the most elegant of all Old French romances, while Jean Frappier pronounced it a masterpiece – than of its complex manuscript tradition and its chequered history as a published text.

Partonopeus survives in seven complete or near-complete manuscripts, three fragments of lost manuscripts and thirty-three extracts – probably from three more lost manuscripts – incorporated into an anthology codex and two later vernacular texts (a checklist of all these witnesses, together with the sigla that will be used to refer to them, is given in Appendix 1). The romance also exists in three different versions.

This book set out to explore both the making of a romance – Partonopeus de Blois – and its impact on the making of Old French romance as a genre. In the process, it has become clear that one of our poem's key contributions to the development of twelfth-century fiction was to validate the notion of vernacular romance as a form that is always in the making, never definitively made. In some ways, then, it seems to go against the spirit of open-endedness embodied in Partonopeus to try to impose on this book the conventional mark of closure represented by a chapter headed ‘Conclusion’. Perhaps it would have been better to call this chapter the ‘Epilogue’, on the tacit understanding that, as in our romance, a first epilogue may be succeeded by a second prologue, in a potentially endless chain of continuation. On the other hand, an ending of sorts has been reached, in that the results of many years of work are finally to appear in print, and a number of conclusions – some of them tentative, others less so – have emerged about facets of this text that have remained under-researched until now.

As anticipated, exploring the highways and byways of the manuscript tradition has proved essential for understanding how Partonopeus took on the form(s) in which we know it. Complex and fragmented as it is, this tradition has helped us to tease out answers to questions about the likely configuration of the first version of the text as well as the type and extent of interpolation in subsequent additions.

The fils à vilain theme examined in Chapter 2 is only one of a number of unifying threads that run through Partonopeus de Blois, helping to tie its various rewritings more closely together. Others include the frequent interventions of the ‘lyric’ narrator who constantly aligns the progress of his own extra-diegetic love-affair with the changing fortunes of the hero and heroine, and the regular appearance in the central section of the narrative of women characters who act as as adjuvants and obstruants for the hero (Urraque, Persewis, Armant's wife versus the hero's mother and the niece of the king of France). One thread that runs consistently through the whole poem is the presence of animals and birds, and the use of animal imagery. The first section of the prologue features a skylark, a nightingale and an oriole that represent three different approaches to literary creation. The lark re-appears at the end of the A version of the text, alongside a corn bunting, a quail and a raven, in a metaphorical prelude to the triple wedding of Melior, Urraque and Persewis. The great forest of the Ardennes is the scene for three important episodes, each of which stresses the variety of creatures to be found there: elephants, lions, snakes, dragons, boars, leopards, bears, tigers and wolves are all mentioned, in a careful mingling of the familiar and the unexpected. In Chapter 1 we saw how two wild boars in the Ardennes are used to add layers of meaning to the portrait of the hero and to position the narrative at the intersection of the Breton lai and the dynastic romance.

The binary opposition between cortois and vilain underpins a whole range of medieval French texts intended for aristocratic audiences. A concern with what is courtly and what is not manifests itself in a variety of ways, from the less than flattering portraits of those outside court circles found in Chrétien de Troyes's Yvain and Le Conte du Graal to more explicit debates such as that dramatised in Raoul de Houdenc's Meraugis de Portlesguez. In this romance, two knights fall in love with the same lady: one, Gorvain, on account of her physical beauty alone, and the other, Meraugis, ‘por sa cortoisie,/ Por sa bonté sanz vilonie’ (‘on account of her courtliness, her goodness untouched by ill-breeding’, vv. 573–4). Raoul's hero makes it quite clear that if Lidoine had shown any hint of uncourtly qualities, then he would have found it impossible to love her:

Counter-arguments are sometimes made, as in a jeu-parti by Pierre de Beaumarchais in which the poet and a lady discuss whether it is better to love a man who has prouece but no cortoisie, or a courtly man who demonstrates no prowess. The lady opts for the former, declaring that a noblewoman should be prepared to overlook any lack of courtly qualities in a knight and be content with his prowess: ‘Si s’en doit bien bele dame paier/ Et oublier tote sa vilenie’ (‘So a fair lady ought to be satisfied with this and turn a blind eye to all his uncourtliness’).