The "Conte du Graal" Cycle: Chrétien de Troyes's "Perceval", the Continuations, and French Arthurian Romance 1843842858, 9781843842859

Citation preview

Conte du Graal:Romance 02/11/2011 11:27 Page 1

THE CONTE DU GRAAL CYCLE

THE CONTE DU GRAAL CYCLE Chrétien de Troyes’s Perceval, the Continuations, and French Arthurian Romance

Thomas Hinton

Thomas Hinton

Chrétien de Troyes’s late twelfth-century Conte du Graal has inspired writers and scholars from the moment of its composition to the present day. The challenge represented by its unfinished state was quickly taken up, and over the next fifty years the romance was supplemented by a number of continuations and prologues, that came to dwarf Chrétien’s text. In one of the first studies to treat the Conte du Graal and its continuations as a unified work, Thomas Hinton considers the whole corpus as a narrative cycle. Through a combination of close textual readings and manuscript analysis, Hinton argues that the unity of the narrative depends on a balanced tension between centripetal and centrifugal dynamics. He traces how the authors, scribes and illuminators of the cycle worked to produce coherence, even as they contended with potentially disruptive forces: multiple authorship, differences of intention, and changes in the relation between text, audience and book. Finally, this book tackles the long-held orthodoxy that places the Perceval Continuations on the margins of literary history. Widening the scope of enquiry to consider the corpus’s influence on thirteenth-century verse romances, this study re-situates the Conte du Graal cycle as a vital element in the evolution of Arthurian literature. THomAs HInTon is Junior Research Fellow in modern Languages at Jesus College, oxford.

BOYDELL & BREWER Ltd PO Box 9, Woodbridge IP12 3DF (GB) and 668 Mt Hope Ave, Rochester NY 14620-2731 (US) www.boydellandbrewer.com

Gallica

Gallica Volume 23

The Conte dU Graal Cycle Chrétien de Troyes’ Perceval, the Continuations, and French Arthurian Romance

Gallica ISSN 1749–091X

General Editor: Sarah Kay

Gallica aims to provide a forum for the best current work in medieval French studies. Literary studies are particularly welcome and preference is given to works written in English, although publication in French is not excluded. Proposals or queries should be sent in the first instance to the editor, or to the publisher, at the addresses given below; all submissions receive prompt and informed consideration. Professor Sarah Kay, Department of French, New York University, 13–19 University Place, 6th floor, New York, NY 10003, USA The Editorial Director, Gallica, Boydell & Brewer Ltd., PO Box 9, Woodbridge, Suffolk IP12 3DF, UK

Previously published titles in this series are listed at the end of this volume.

The Conte dU Graal Cycle Chrétien de Troyes’ Perceval, the Continuations, and French Arthurian Romance

Thomas Hinton

D. S. BREWER

©  Thomas Hinton 2012 All Rights Reserved. Except as permitted under current legislation no part of this work may be photocopied, stored in a retrieval system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner The right of Thomas Hinton to be identified as the author of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988 First published 2012 D. S. Brewer, Cambridge ISBN  978–1–84384–285–9 D. S. Brewer is an imprint of Boydell & Brewer Ltd PO Box 9, Woodbridge, Suffolk IP12 3DF, UK and of Boydell & Brewer Inc. 668 Mt Hope Avenue, Rochester, NY 14620, USA website: www.boydellandbrewer.com A catalogue record for this title is available from the British Library The publisher has no responsibility for the continued existence or accuracy of URLs for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate Papers used by Boydell & Brewer Ltd are natural, recyclable products made from wood grown in sustainable forests This publication is printed on acid-free paper

Printed in Great Britain by CPI Group (UK) Ltd, Croydon, CR0 4YY Disclaimer: Some images in the printed version of this book are not available for inclusion in the eBook. To view these images please refer to the printed version of this book.

CONTENTS List of Illustrations vi Acknowledgments vii List of Abbreviations viii Manuscript sigla ix Introduction 1 1. Narrative Aesthetic and Cyclic Formation

29

2. Manuscripts, Memory and Textual Transmission

70

3. Authorship, Kinship and the Ethics of Continuation

111

4. Rereading the Evolution of Arthurian Verse Romance

163

Conclusion 218 Appendix 1: Narrative Summaries Appendix 2: Lengths and Dates of Texts Appendix 3: Manuscripts of the Conte du Graal Cycle Appendix 4: Full Contents of Conte du Graal Cycle Manuscripts Appendix 5: Arthurian Verse Romances: Dates and Manuscripts Appendix 6: Contents of Arthurian Verse Romance Manuscripts

229 244 245 246 250 251

Bibliography 255 General Index

268

Index of Manuscripts

277

ILLUSTRATIONS Figures 1. Mons, BU, 331/206 (MS P), p. 371. Historiated initial depicting Perceval. Source: Bibliothèque de l’Université de Mons. 2. Paris, BNF, fr. 1453 (MS S), fol. 72v. Guiromelant watches Clarissant pleading with Gauvain. By courtesy of the Bibliothèque Nationale de France. 3. Paris, BNF, fr. 12576 (MS T), fol. 1r. Opening miniature. By courtesy of the Bibliothèque Nationale de France. 4. Paris, BNF, fr. 12577 (MS U), fol. 1r. Opening miniature. By courtesy of the Bibliothèque Nationale de France.

75 99 100 102

Tables I. Lines marked by large initials in at least two of EPQTV 84 II. Lines from the First Continuation marked by large initials in EQ 88

Disclaimer: Some images in the printed version of this book are not available for inclusion in the eBook. To view these images please refer to the printed version of this book.

ACKNOWLEDGMENTS It is a pleasure to be able to thank those without whom this study would exist only as a collection of scribbled notes and half-baked ideas. My first, and greatest, debt is to Simon Gaunt, whose encouragement and acute intelligence have been a constant source of inspiration. He has had a great influence on the development of this book, both as a scholar and as a friend. My work has benefited from numerous conversations over the past few years. I would like to thank Bill Burgwinkle and Jane Gilbert for their careful reading of the text and stimulating suggestions for improvement. Luke Sunderland read and commented on more of my work than seems fair, and has been a patient and supportive friend. Several people have offered advice at various stages of the project, whether through informal chat, direct or indirect correspondence: Matilda Bruckner, Emily Butterworth, Linda Gowans, Roger Middleton, Alex O’Brien, Karen Pratt and Julian Weiss have all contributed to the book’s enhancement. I am also grateful to the anonymous readers for Boydell & Brewer whose recommendations were of great help. Further down the production line, I have benefited from the editorial experience of Caroline Palmer, the meticulous copy-editing of Diane Wardle, and the heroic indexing work of Freya Verstraten Veach. Portions of Chapter 3 are reworked from an article previously published in Arthurian Literature 26 (2009): 91–108, and I thank Boydell & Brewer Ltd for permission to reproduce the material here. Jesus College Oxford generously contributed to the cost of publication, while funding provided by the Arts and Humanities Research Council made this project possible in the first place. I hope that they continue to provide similar opportunities to researchers in the Humanities. Manuscript study, meanwhile, depends on access to library collections, and I am grateful to the National Library of Scotland, the Bibliothèque de l’Université de Mons, the Bibliothèque interuniversitaire de Montpellier and the Bibliothèque nationale de France for their assistance. I would like to finish with two acknowledgments that go especially deep. The first is to my family: my parents, sister and grandmother. Their belief in my work has been a constant source of strength, and I hope that the results of my research go some way towards justifying it. Finally, Lucie Hinton has lived with this project for several years, during which time our lives have changed beyond recognition. I dedicate this book to her, in anticipation of adventures to come.

ABBREVIATIONS Texts CdG

Chrétien de Troyes, Le Roman de Perceval ou Le Conte du Graal, ed. Busby C1 [LR] First Continuation [Long Redaction], ed. Roach (1950 [II])1 C1 [MR] First Continuation [Mixed Redaction], ed. Roach (1949 [I]) C1 [SR: A] First Continuation [Short Redaction of MSS APS], ed. Roach (1952 [III, 1]) C1 [SR: L] First Continuation [Short Redaction of MS L], ed. Roach (1952 [III, 1]) C2 [SR] Second Continuation [Short Redaction], ed. Roach (1971 [IV]) C2 [LR] Second Continuation [Long Redaction], ed. Roach (1971 [IV]) C2 Second Continuation, ed. Roach (1971 [IV])2 CG Gerbert Continuation, ed. Williams (tome I: 1–7020; tome II: 7021–14078) and Oswald (tome III: 14079–end)3 CM Manessier Continuation, ed. Roach (1983 [V]) All translations are my own unless stated otherwise.

1 Volume references are to William Roach’s five-volume edition of The Continuations of the Old French Perceval of Chrétien de Troyes (1949–83). 2 The Second Continuation is edited in two redactions in Roach (1971 [IV]) up to C2 [SR] 10268 / C2 [LR] 20689; from this point onwards, a single version of the text is given, with line numbers following that of the Long Redaction. 3 The Gerbert Continuation is edited in three separate volumes by Mary Williams (1922 and 1925) and Marguerite Oswald (1978).

Manuscript Sigla A Paris, Bibliothèque nationale de France, fr. 794 B Bern, Burgerbibliothek, 354 C Clermont-Ferrand, Bibliothèque municipale et interuniversitaire, 248 D Rappoltsteiner Parzifal: German translation interpolated into Wolfram von Eschenbach’s Parzival (two extant manuscripts) E Edinburgh, National Library of Scotland, Advocates 19. 1. 5 F Florence, Biblioteca Riccardiana, 2943 G 1530 prosification (black letter edition for Galiot du Pré) H London, College of Arms, Arundel XIV K Bern, Burgerbibliothek, 113 L London, British Library, Add. 36614 M Montpellier, Bibliothèque interuniversitaire, Section Médecine, H 249 P Mons, Bibliothèque universitaire, 331/206 Q Paris, Bibliothèque nationale de France, fr. 1429 R Paris, Bibliothèque nationale de France, fr. 1450 S Paris, Bibliothèque nationale de France, fr. 1453 T Paris, Bibliothèque nationale de France, fr. 12576 U Paris, Bibliothèque nationale de France, fr. 12577 V Paris, Bibliothèque nationale de France, n. a. fr. 6614 For dating and contents of the manuscripts, see Appendices 3 and 4.

Introduction The subject of this book is a literary corpus whose character is a challenge to definition, beginning with the matter of its title. What I refer to throughout as the Conte du Graal cycle is a constellation of texts, found in different combinations in different manuscripts, which tell the story of Perceval from his first encounters with knighthood and the mysteries of the Grail to his eventual succession to the Fisher King’s throne and death as a hermit, interspersed with the adventures of other knights, chiefly Gauvain. The first move in this narrative game is Chrétien de Troyes’s Conte du Graal, believed to have been written between 1180 and 1195. That romance, left unfinished, provoked a number of writers to continue the tale, their work generally identified today as the First, Second, Manessier and Gerbert Continuations.1 The first two Continuations are thought to have appeared by around 1200; the Manessier Continuation is dated between 1214 and 1227; and the Gerbert Continuation between 1225 and 1230. Two prologues, the Elucidation and the Bliocadran, also appeared in the early thirteenth century, so that all the constituent parts of the corpus were in existence (in their earliest forms) within fifty years of Chrétien’s text.2 1 The difficulty of identifying the constituent parts of this corpus is amply demonstrated by the history of the titles assigned to these texts. The Manessier Continuation is often labelled the ‘Third Continuation’, though this title is occasionally used for the Gerbert Continuation, which may at other times be referred to as the ‘Fourth Continuation’; meanwhile, ‘Gauvain Continuation’ and ‘Perceval Continuation’ are terms still used for the first two Continuations (especially in Francophone criticism). In the first half of the last century, these two units were treated as a single item under the heading ‘Wauchier Continuation’; when a division into two was accepted, the appellation ‘Pseudo-Wauchier’ was adopted for a time for the former text, and the original title retained for the latter. (See Roach 1956: 107–9 for a summary of the debate over this last point, but note that the attribution of the Second Continuation to Wauchier de Denain has subsequently been revived.) Whichever title one adopts, there is an additional question relating to typography: ‘First Continuation’, First Continuation or First Continuation? I have chosen to give the names of the individual texts in roman type, in order to emphasise my conception of the cycle as a unit whose constituent parts function together, rather than as separable, individual works. I make an exception with regard to Chrétien’s Conte du Graal, which is found on its own in a small number of manuscripts, and also in deference to established critical practice, although my goal in this study is to emphasise its more common transmission as a constituent part of the cycle. 2 Dates are from Pickens, Busby and Williams (2006).

2 THE

CONTE DU GRAAL CYCLE

Apparently popular with contemporary audiences, yet often neglected by modern critics, the Conte du Graal cycle played a role in the development of medieval Arthurian narrative which so far has not been fully appreciated. Instead, the vast majority of commentators have preferred to treat the unfinished Conte du Graal in isolation from its sequels, too frequently dismissed as confused or prolix by critics dreaming of the conclusion that Chrétien might have written. Jean Frappier, for instance, concludes his analysis of the text as follows: Nous n’en savons pas plus. La mort a arrêté en cet endroit (v. 9234) l’oeuvre du maître Chrétien de Troyes. Cette brusque interruption nous laisse à nos incertitudes, car nous ne pouvons nous fier à ses continuateurs pour connaître le veritable dessein du grand romancier.3 [We know no more. At this point (v. 9234), death put an end to the work of the great Chrétien de Troyes. This sudden interruption leaves us to grapple with our uncertainties, for we cannot rely on his continuators to know the true intentions of the great writer.]

Frappier’s interest is explicitly displaced from the extant corpus onto the question of how the ‘master’ might have brought his romance to an end, and the Continuations are read as merely one possible (and inauthentic at that) attempt to answer this riddle. These approaches to the Continuations echo that of Gustave Cohen, who warned in 1931 that ‘il faut, pour interpréter correctement Chrétien et son Perceval, faire abstraction de cette évolution ultérieure’ [In order to interpret Chrétien’s Perceval correctly, one must ignore this subsequent evolution of the material].4 The perceived inadequacies of these texts are summed up in Frappier’s study during a general presentation of French Grail romance: they lack coherence and they display insufficient fidelity to the parameters of the narrative as set out in the Conte du Graal: Aucun plan d’ensemble ne paraît avoir dominé cette production diffuse; d’un texte à l’autre, on relève des contradictions; chaque continuateur a inventé selon sa fantaisie, en perdant souvent de vue l’aventure du Graal… Le tout représente plus de soixante mille vers.5

Frappier (1972: 252–3). Cohen (1931: 445). See also Le Rider (1978: 7): ‘J’ai voulu oublier dans toute la mesure du possible les continuations médiévales du Conte du Graal… Le confondre dans un “cycle du graal”, l’interpréter à partir des allégories et des exégèses ritualistes de ses épigones eût été, m’a-t-il semblé, le trahir.’ [I have decided, as far as possible, to ignore the medieval continuations of the Conte du Graal… To merge it into a ‘grail cycle’ and interpret it through the allegories and ritualistic interpretations of its imitators would be, it seems to me, to betray it.] 5 Frappier (1972: 13). 3

4

INTRODUCTION 3

[It appears that no general plan underlay this uncoordinated activity; from one text to the next, one notes contradictions; each continuator wrote according to his whim, often losing sight of the Grail adventure… The whole thing comes to more than sixty thousand lines.]

The First Continuation in particular has been singled out for the lack of interest it shows in Perceval, preferring to narrate the adventures of Gauvain, Caradoc and Guerrehet. In Frappier’s opinion, ‘le récit manque… d’unité ou de conjointure. Il se fragmente en contes à peu près indépendants’ [The story lacks… unity or conjointure. It breaks up into a collection of more or less independent tales].6 And Keith Busby, in a study of Chrétien’s Conte du Graal aimed at students, similarly flags a lack of unity and coherence as a defining feature of the corpus: These Continuations are on the whole fairly loosely attached to Chrétien’s poem, and seem to consist of a number of tenuously related adventures, featuring Gauvain, Perceval, Caradoc, and Guerrehés (Gauvain’s brother), in which the Grail sometimes plays a role, and sometimes does not.7

Busby does, however, go on to note: ‘They reflect a manner of storytelling that requires further urgent investigation, although here is not the place.’ In recent years, he and other scholars have begun serious study of what he calls elsewhere ‘one of the most extraordinary products of medieval romance writing’.8 A brace of stimulating articles by Alexandre Leupin either side of 1980 represents the earliest attempt to come to terms with the aesthetics of the corpus.9 More recently, Mireille Séguy has devoted some pertinent pages to the subject in the context of a wider discussion of Grail romance, while Matilda Bruckner has produced a number of thoughtful pieces on what she calls its ‘poetics of continuation’, culminating in the first English-language monograph given over to it, published in 2009.10 Indeed, interest in the Conte du Graal cycle now appears to have reached critical mass, with a number of articles, books and doctoral theses on the subject appearing or soon to appear.11 Busby’s call for further investigation is beginning to find an answer in the forest of Arthurian scholarship.

6 Frappier (1973: 153). Note the choice of the term conjointure (coined by Chrétien to describe his authorial technique in Erec, v. 14), establishing an implicit contrast between the art of the ‘master’ and the supposed artlessness of his continuator. 7 Busby (1993a: 91). 8 Busby (1994b: 178). 9 Leupin (1979) and (1982). 10 Séguy (2001a), especially 286–342; Bruckner (1987), (1993), (2000), (2006) and (2009). 11 In the last five years alone, one may cite the doctoral theses of Sébastien Douchet, Massimiliano Gaggero, Etienne Gomez and Leah Tether.

4 THE

CONTE DU GRAAL CYCLE

The current study is an attempt both to understand the corpus’s cyclic aesthetics and to identify how medieval audiences and authors of Arthurian verse romance responded to it. Continuation as a literary practice is fundamental to medieval poetics, fuelled as much by an aesthetics and ethics of rewriting as by the material conditions of manuscript transmission that left the medieval text more vulnerable to alteration and addition than its modern counterpart. The period’s most popular text, the Roman de la Rose, was a product of continuation: the version transmitted in the overwhelming majority of manuscripts features the continuation of Jean de Meun, which seized upon the suspended narrative of Guillaume de Lorris’s original romance as an invitation to redirect and extend the work to more than twenty thousand lines, five times the initial length. Similarly, many medieval cycles developed by exploiting the potential for narratives to be re-opened through the addition of new material; in most cases, however, the new texts can be considered to exemplify the notion of ‘suite’ or ‘sequel/prequel’, rather than that of ‘continuation’ proper: that is, where the latter notion suggests the prolongation of an unfinished narrative, the former responds to a desire for further development of material initially considered to be complete.12 The decision as to what is and is not complete is clearly a delicate one, but it is impossible to avoid, as argued by David Hult: Any critical evaluation of literary continuations must in some way deal with the question of a text’s relative openness or closure – the extent to which a ‘first’ text can be considered complete and unified in an ideal way or, conversely, in need of further additions.13

The existence of the Conte du Graal Continuations in itself is proof that medieval authors and audiences read Chrétien’s text as open, a judgment that is confirmed by the most superficial engagement with its ending. Whether one considers such irresolution to be deliberate or accidental, the end of the text clearly constitutes an interruption of, rather than conclusion to, the narrative.14 Gauvain has dispatched a messenger to ask King Arthur to witness his duel; just as the messenger arrives at court, Arthur, despairing of obtaining news of his nephew, faints before the eyes of Lore, one of the queen’s ladies in waiting, who runs to inform her mistress:

12 For a discussion of the distinction between suite and continuation in relation to modern literature, see Genette (1982: 181–3); for consideration of the terminology in the specific context of a discussion of the First Continuation, see the questions and answers at the end of Roach (1956). 13 Hult (1984: 248). 14 Rider (1998: 18–19) observes that ‘ending the story with an incomplete or unanswered question is… entirely in keeping with the story’s logic’, and suggests that this may have been a deliberate decision on the author’s part.

INTRODUCTION 5

Et quant la roïne le voit, Si li demande qu’ele avoit  (CdG 9233–4) [And when the queen sees her, she asks her what is wrong]

The text thus breaks off in a manner that invites continuation on the levels of both micro-development (a question awaiting a response) and macro-development (the audience’s questions about whether Arthur will recover, how he will react to the messenger, and what will happen if and when he travels to Gauvain’s side). We have been deflected from the main narrative (Gauvain’s conflict with Guiromelant) to a subsidiary narrative (the messenger’s journey to court), which is in turn interrupted by Arthur’s faint; the narrative viewpoint then switches again to another character, Lore, who moves us into the queen’s chamber, only to be interrupted herself before she can relate her tale to her mistress. As we will see repeatedly in this study, such interruptions are part and parcel of the textuality of this first Grail romance, to which the Continuations will respond in their different ways. But the essential point for now is that, in ending in this manner, the text invites continuation; the extraordinary production of continuatory narrative by the continuators, and the popularity of the resultant corpus, provide confirmation that Chrétien’s text on its own appeared insufficient to many readers. Similarly, Guillaume de Lorris’s allegorical Roman de la Rose ends at a point of narrative crisis, with the figure of Amant [Lover] cursing the frustration of his desire; both the rose that he wishes to pluck, and the figure of Bel Acueil [Fair Welcome] who had encouraged him earlier in the narrative, are secured behind a fortress built by Jalousie [Jealousy]. Guillaume’s text is followed in the vast majority of manuscripts (over 250) by Jean de Meun’s continuation; in addition, eight manuscripts preserve a different and much shorter continuation, generally referred to as the Anonymous Conclusion.15 As with the Conte du Graal, both continuations operate on the level of micro-development (reading Amant’s final line as an unfinished sentence to be continued) and of macro-development (concluding the narration of Amant’s attempt to capture the rose). Despite the similar mechanics at work, the two continuations could hardly be more different: the Anonymous Conclusion wraps up the narrative in the space of seventy-eight lines, while Jean’s text swells the original four thousand lines written by Guillaume into an encyclopaedic romance of over twenty thousand verses, which covers a dazzling range of discourses and subjects. If the Roman de la Rose and the Conte du Graal corpus both demonstrate the power of a perceived suspended ending to generate additional narrative, one may wonder what advantage is to be gained in speaking of one as a (primarily) dual-authored romance and the other as a cycle. The term ‘cycle’

15

On the Anonymous Conclusion, see Brook (1995).

6 THE

CONTE DU GRAAL CYCLE

as used in modern literary criticism is a coinage of the early nineteenth century, when it was applied to the body of romances dealing with Arthur and the Round Table, or to the group of chansons de geste relating to Charlemagne.16 In other words, it is a concept born of modern reading practices. A more detailed discussion of cyclicity will be provided later in this Introduction, but it is necessary to pre-empt that argument a little in order to explain the use of the term in relation to the Conte du Graal corpus. The essential point for the time being is that a narrative cycle is a collection of texts read in sequence according to a uniting principle of coherence. A cycle is at once one and many; as David Staines puts it: ‘the unity of a cycle… is the thematic pattern imposed upon the multiplicity of the cycle’s parts’.17 In this way the codices of the Guillaume cycle contain a sequence of texts relating the enfances (youth) of its heroes, their mature exploits and the final events of their lives; the biographic sweep that binds such a cycle together can be extended into a genealogical principle by the inclusion of material relating to the heroes’ ancestors or descendants.18 The Roman de la Rose, on the other hand, has been transmitted and is usually read as a single, albeit complex and polyphonic, text. Individual critics may place emphasis on the unity binding its two unequal halves, or instead on its refusal to coalesce into a single viewpoint;19 but in either case, as Hult observes, ‘the two parts have always been designated by a single title and thus, more often than not, viewed as two components of one text’.20 In contrast, as we have seen, most scholarship on the Conte du Graal corpus has placed emphasis on the disparate character of its different parts; the overwhelming omission of the Continuations in analyses of Chrétien’s romance says much about the (lack of) relevance that these later texts are still generally perceived to have by critics of the Conte du Graal. To some extent, this reaction is provoked by the corpus’s complex and shifting narrative aesthetics. Yet there is considerable evidence that, in large measure, medieval audiences received the corpus as a single textual entity. The tension between the manifest cohesion of the manuscript tradition and the diverse directions of the narrative is encapsulated by Busby’s remark that ‘this “other Grail cycle” does not work by means of any sort of “unity”’.21 For a history of the term’s use, see Staines (1994) and (1996). Staines (1994: 110). 18 On the manuscript tradition of the Guillaume cycle, see Tyssens (1967) and Sunderland (2010). 19 See Huot (2010: 3–4) for a summary of critical debate over how to conceptualise the general aesthetics of the Rose. 20 Hult (1984: 249). It is of course still open to scholars to analyse Guillaume’s text in abstraction from Jean’s continuation, as a ‘finished’ work in its own right. Indeed, this is precisely the project of Hult (1986); but this critical move must be performed knowingly as a deliberate counter-narrative to the general thrust of Rose scholarship, which draws on the whole, dual-authored textual tradition. 21 Busby (1994b: 178). 16 17

INTRODUCTION 7

As indicated by Busby’s use of scare quotes, the real issue at state is the kind of unity to be found in the corpus; clearly, the application of modern notions of textual coherence dooms the Continuations to critical contempt. This is where I believe a cyclic reading of the corpus is valuable. The coherence which produces cyclic unity is always provisional and sufficiently precarious for multiple versions and recombinations of a corpus to form and co-exist. Unlike the Roman de la Rose tradition, then, the Conte du Graal cycle rests on a unity that runs in tension with the potential for its parts to be broken up or interpolated into: less ‘essential’ parts, such as the Elucidation and Bliocadran prologues, or the Gerbert Continuation, orbit around the cyclic core; Chretien’s text is sometimes transmitted by itself; the Second Continuation is transmitted on its own in one manuscript.22 To read the corpus as a cycle is simultaneously to acknowledge this potential for its constituent parts to be separated and combined in different ways while maintaining the focus on the features that bind them together into a single textual whole. The Conte du Graal corpus is thus unusual in representing an ‘incomplete text plus continuation’ ensemble that invites a cyclical reading. Perhaps for this reason, it has rarely been studied in relation to the phenomenon of cyclicity. Bruckner alludes in a 1987 article to the importance of understanding ‘the particular character of this cycle within the larger context of thirteenth-century cyclicization’ but does not pursue the question; Busby’s short article entitled ‘The Other Grail Cycle’ sketches out possible avenues for further research.23 Aside from such isolated instances, however, the Conte du Graal corpus has been largely absent from debates about the nature of medieval narrative cyclicity, presumably because its narrative aesthetics do not correspond neatly to critics’ expectations of what a cycle is supposed to be.24 The purpose of this Introduction is to make the case for approaching the text as a cycle in two complementary ways: first, to suggest limitations to the traditional scholarly habit of isolating the Conte du Graal from its Continuations in order to study it alongside Chrétien de Troyes’s other romances; secondly, to give an account of how the Conte du Graal cycle fits into the

22 For the purpose of comparison, it is worth reiterating more precisely the manuscript tradition of the Roman de la Rose: one manuscript preserves only Guillaume de Lorris’s text; one contains Guillaume’s text with the Anonymous Conclusion (AC); a further six sandwich the AC between Guillaume’s text and Jean de Meun’s continuation; the overwhelming majority (more than 250) contain Guillaume’s text followed by Jean’s continuation. No manuscript preserves Jean’s portion on its own. 23 Bruckner (1987: 262); Busby (1994b: 176–8). Similarly, Bruckner (2009) frequently uses the term ‘cycle’ to describe the corpus, but never explicitly considers the nature of this cyclicity. 24 See however Gomez (forthcoming), whose decision to refer to the corpus as the ‘cycle du Conte du graal’ coincided with my choosing the appellation ‘Conte du Graal cycle’.

8 THE

CONTE DU GRAAL CYCLE

broader context of cyclification in the late twelfth and thirteenth centuries, and what it can add to critical study of the phenomenon.

The Conte du Graal Corpus and Chrétien de Troyes It is not difficult to find points of contrast between the Conte du Graal and the other Arthurian romances of Chrétien de Troyes. Though unfinished, it stands at well over nine thousand lines by the time it breaks off, with sufficient material left unresolved to fill several thousand more, whereas Chrétien’s other four romances all run to around seven thousand lines. Rupert Pickens identifies the following additional features as making this romance unique within Chrétien’s oeuvre: (i) both the prologue and the main tale deal explicitly with religious sentiment and doctrine; (ii) the romance is thoroughly informed by prediegetic matter (e.g. accounts of British history, events in communities in exile from Arthur’s kingdom); (iii) it interacts with Wace’s Brut to a greater degree than any other Chrétien romance (with the possible exception of Erec); (iv) it offers a doubled narrative which allots to two heroes, Perceval and Gauvain, an independent and equally elaborated set of adventures; and (v) the alleged source text is given a name by its narrator, ‘li contes del graal’ (CdG 66).25 ‘Les différences’, concludes Barbara Sargent-Baur after comparing the Conte du Graal’s prologue with those of Chrétien’s other romances, ‘semblent annoncer un nouveau départ’ [The differences appear to announce a new beginning].26 This new direction, exemplified in the Gauvain adventures that dominate the second half of the text, has so disturbed some modern readers that at one point debate raged over whether the Gauvain part was by Chrétien at all.27 Discussing the second half of the romance, Per Nykrog wryly sums up the feeling of disquiet that has affected several of its critics: ‘these adventures engage Chrétien on a road that must fill the reader with apprehension’.28 The choice of words is apposite: the reader’s ‘apprehension’ equates to a fear of the unknown, as Chrétien’s knights lead him progressively away from the narrative models established by the earlier romances and the horizon of expectation disappears from view. Certainly, the evidence of the surviving manuscripts suggests that the singularity of the Conte du Graal was not lost on medieval audiences. Of Chrétien’s other four Arthurian romances, three (Erec, Cligés and Yvain) survive in twelve copies or fragments, and the Lancelot in only eight; by contrast, there are eighteen extant complete or fragmentary copies of the Conte du 25 Pickens (2005: 170–1). On the Conte du Graal’s interaction with Wace, see Pelan (1931) and Sturm-Maddox (1984). 26 Sargent-Baur (2000: 12). 27 See the summary of various positions on this issue in Busby (1993a: 51). 28 Nykrog (1973: 269).

INTRODUCTION 9

Graal.29 Beyond the question of number, the manuscript tradition points to another important difference between the Conte du Graal and its predecessors. These latter texts are almost always collected in manuscripts with other romances, either singly or in groups of two or three.30 Only two manuscripts (Paris, BNF, fr. 794 and fr. 1450) collect all of Chrétien’s texts together, and in both cases these make up less than half of the total number of lines in the manuscript.31 There are thus no manuscripts which can comprehensively be described as Chrétien compilations. The Conte du Graal, in contrast, is generally transmitted along with one or more of the Continuations. This is the case for eleven of the fifteen complete or largely complete manuscripts of the text, as can be seen from the table in Appendix 3. In the vast majority of cases the texts are copied by a single hand without any indication of a break between the different parts of the corpus; the only instance where any explicit indication is given of the transition between Chrétien’s contribution and those of other authors is A (Paris, BNF, fr. 794), where the scribe (Guiot) indicates the end of Chrétien’s part with the colophon ‘Explycyt perceuax le ueil’ [Here ends the old Perceval].32 The eight manuscripts which contain the Manessier Continuation (EMPQSTUV) all originally contained no other texts, though the blank folios in MT were used for later additions; moreover, it is likely that manuscript L (London, BL, Add. 36614) initially contained only the Conte du Graal, with the first two Continuations, the Bliocadran and the Vie de Sainte Marie l’Egyptienne being added before the manuscript left the workshop. These nine codices bear witness to the popularity and stability of a Conte du Graal corpus, which (anticipating on fuller discussion below) I will be describing as cyclical. One may also adduce two additional pieces of evidence. First, the Rappoltsteiner Parzifal, a redaction of Wolfram von

See Busby (2005). As David F. Hult (1998: 20) notes: ‘The first four romances, which have survived… in roughly similar numbers of manuscripts, are associated with each other in a uniform and mostly random fashion… The Perceval, for its part, developed a tradition of its own.’ 31 The so-called Annonay fragments, from a manuscript containing Cligés, Yvain, Lancelot and the Conte du Graal, provide a further possible instance of a manuscript including all of Chrétien’s works, though there is no hard evidence that Erec was contained in this codex, nor of what other texts might have been present. On these fragments, see Nixon (1993b: 20–2). 32 In L, the Conte du Graal is copied in a different hand from the Continuations; similarly, in T (Paris, BNF, fr. 12576) the end of the First Continuation and beginning of the Second Continuation (folios 95 to 121) are copied by a different hand from the rest of the text. In both cases, however, the hands are so similar that they probably belong to scribes from the same workshop; given the otherwise homogeneous presentation of the text, the change of hand is most likely a consequence of practical considerations, rather than expressing any desire to signal a change of authorship or of textual unit. For more details, see the Introductions to Roach’s edition of the Mixed Redaction of the First Continuation (Roach 1949 [I]) and Busby’s edition of Chrétien’s text (Conte du Graal); on manuscripts TV in particular, see Busby (1993b) and Gaggero (2008b). 29 30

10 THE

CONTE DU GRAAL CYCLE

Eschenbach’s Parzival by Claus Wisse and Philipp Colin known as siglum D in the Conte du Graal manuscript tradition: Wisse and Colin translated and interpolated into Wolfram’s text a corpus comprising the Elucidation, Conte du Graal, First, Second and Manessier Continuations. The original manuscript (Karlsruhe, Landesbibl., Cod. Donaueschingen 97) is dated between 1331 and 1336; a copy is also preserved in Rome (Bibl. Casanatense, Mss. 1409). The second additional witness to a unified corpus is the 1530 prosification printed by Galiot du Pré (siglum G): this contains the Elucidation and Bliocadran prologues, Conte du Graal, First, Second and Manessier Continuations. Conversely, all four manuscripts which transmit Chrétien’s romance independently (BCFH) show signs of the partner-shifts which affect the transmission history of the whole Conte du Graal manuscript tradition, where the use by scribes of multiple sources leads to periodic swapping and pervasive instability in the relations of the different manuscripts; these copies may therefore ultimately be derived from manuscripts presenting either all or part of the corpus.33 Busby has shown this to be the case for manuscript H (London, College of Arms, Arundel XIV). It contains two interpolations (428 lines after CdG 3926 and 116 lines after CdG 3994) relating to the breaking of the sword given by the Fisher King to Perceval; in the first, after the Fisher King has had the broken pieces recovered, he has them placed on a bier, declaring that the knight who can heal him will be the one who can mend the sword. This passage is modelled on, and designed to harmonise with, the broken sword test introduced in Branches I and V of the First Continuation, and therefore a manuscript containing at least this Continuation must have been involved at some point in the composition of H.34 MS H, along with MS B (Bern, Burgerbibliothek 354), are the only surviving codices in which the Conte du Graal is found on its own as part of a compilation. The cyclical manuscripts, by contrast, go to significant lengths to emphasise the unity of their texts. An extreme instance of this is provided by MS P (Mons, BU 331/206), whose scribe deliberately encourages the illusion of single authorship. Opening with the Elucidation and the Bliocadran (this is the only codex in which both texts appear), Chrétien’s prologue is removed from the beginning of the Conte du Graal and its last eight lines reworked into a prologue to the Bliocadran, which thereby becomes absorbed into the Conte du Graal as its first chapter;35 meanwhile, 33 See Van Mulken (1993) and the Introduction to Busby’s edition, where he summarises the situation as follows: ‘Il serait possible de démontrer sans trop de peine à l’aide des variantes l’existence de presque chaque filiation concevable entre deux ou plusieurs manuscrits’ [It would be relatively easy, with the help of the variants, to demonstrate the existence of almost any conceivable filiation between two or more manuscripts] (Conte du Graal: xlii). 34 Busby (1993c). 35 Chrétien’s prologue (the whole thing this time) also directly precedes the Blio-

INTRODUCTION 11

the name of Manessier at the end of the cycle is replaced by that of Chrétien, who becomes both the initiator and completer of the narrative. The success of this strategy can be gauged from the inside cover of the manuscript, where an unidentified modern hand has made the following note: Ce roman est du douzième [‘treizième’ crossed out] siècle. Son auteur est Chrestien de Troyes. Il se nomme au pénultième vers de tout l’ouvrage [where Manessier’s name has been removed] et deux fois dans les cinq derniers de la première colonne de la page 6 [the prologue to Bliocadran]. [This romance is from the twelfth century. Its author is Chrestien de Troyes. He names himself in the penultimate line of the whole work and twice in the last five lines of the first column of page 6.]

A further indication of cohesion in the Conte du Graal corpus is the rarity of manuscripts which include other texts from outside the corpus. Thus the Conte is found with another Chrétien text only in the two manuscripts which include all of his romances, A and R (Paris, BNF, fr. 1450, which interpolates the romances into Wace’s Brut) – and, in both cases, a portion of the Continuations is also included. In the case of MS R, which includes only Branch I of the First Continuation, it seems plausible that the scribe retained this Branch principally in order to bring to a close the episode involving Gauvain and Guiromelant begun by Chrétien, before moving on to the next romance. This scribe may or may not have been aware that Chrétien’s part of the text was left unfinished, but his decision shows at the very least that it was not felt necessary to signal a change of authorship to the manuscript’s intended users. MS A, meanwhile, includes the whole First Continuation and the first 800 lines of the Second. Guiot’s manuscript is made up of three units, each with its own quire structure, with Chrétien’s other four romances constituting the first, and the Conte du Graal and its two Continuations coming at the end of the third. Terry Nixon argues that the Chrétien unit was intended to come after that containing the Conte du Graal, creating a continuous textual entity of Chrétien works.36 Yet, even if this conjecture is correct, the two Continuations would still sit between the Conte du Graal and Erec, demonstrating that – even in these most ‘authorly’ of Chrétien manuscripts – the desire to create a unified ‘Chrétien corpus’ was less powerful than the cohesion of the Conte du Graal ensemble. As Busby argues, the tendency for modern critics to edit and study Chrétien’s final romance in isolation from its Continuations therefore ‘does violence to the cohesive nature of the codicological whole’.37 cadran in MS L, suggesting that this text owes its survival to the strategy of integration into the Conte, without which it may not have been preserved by the versions of the cycle which include it. 36 Nixon (1993a: 22). 37 Busby (2005: 70).

12 THE

CONTE DU GRAAL CYCLE

This raises the question of why critics have been so resistant to reading the Conte du Graal corpus as the unified text presented by the manuscripts. The answer lies in large part in the investment of French Medieval Studies in the name ‘Chrétien de Troyes’ and the corpus of Arthurian romances which has coalesced around it. The author-figure ‘Chrétien’ is frequently made to serve as the key reference point for accounts of the development of medieval romance, of twelfth-century vernacular literature in general, and even of the modern novel.38 Further, the name ‘Chrétien de Troyes’ has functioned for some time as a useful standard to be brandished by medievalists seeking admittance for their texts to the inner chamber of literary respectability, a name with which to communicate with colleagues (and a reading public) more used to apprehending literature through the lens of modern authorship. This strategy is quite explicit in Michel Zink’s preface to the Livre de Poche Pochothèque collection of Chrétien’s oeuvre: Il ne faut pas s’étonner de voir Chrétien de Troyes figurer dans une collection consacrée aux Classiques Modernes. Il peut être dit classique, car s’il existe un canon des grands auteurs, il y figure à coup sûr. Et il est moderne… parce qu’il est le premier dans notre littérature à avoir donné ses lettres de noblesse à ce genre éminemment moderne qu’est le roman… il écrit avec une fluidité, une densité et une légèreté, un art de mêler le mystère à la limpidité, qui en font l’un des plus grands écrivains français.39 [One should not be surprised to find Chrétien de Troyes in a collection dedicated to Modern Classics. He can be called classic since, if a canon of great authors exists, he is certain to be included. And he is modern… because he is the first in our literature to have made respectable that eminently modern genre which is the roman… he writes with a fluidity, a gravity and a lightness of touch, a talent for presenting the enigmatic in a limpid style, which make him one of the greatest French authors.]

Recent years have also seen the publication of a Pléiade Œuvres complètes, as well as volumes with titles such as A Companion to Chrétien de Troyes, The Manuscripts of Chrétien de Troyes and The Legacy of Chrétien de Troyes. All this activity has ensured that, despite a more-or-less total absence of available biographical data, the author-figure Chrétien looms larger in both scholarly and popular culture than perhaps any other authorial name in medieval French literature. And a corollary of this success, as Virginie Greene ob38 On Chrétien as precursor to the modern novel, see the citation from Zink (1994) below. Jewers (2000) takes Chrétien as the starting point (and Cervantes’s Don Quixote as the end point) for an account of the role of chivalric romance in the history of the novel. A similar account of the development of romance is implicit in the title of the collaborative volume Romance: Generic Transformation from Chrétien de Troyes to Cervantes (Brownlee and Brownlee 1985). 39 Zink (1994: 5).

INTRODUCTION 13

serves, is that ‘Chrétien de Troyes’ has become ‘an author’s name functioning as authors’ names do in our times, that is, as a marketing tool’.40 The popularity of the Chrétien de Troyes romances both within and without the medievalist community explains why the Conte du Graal, despite having its own distinct manuscript tradition, is so often severed from its codicological context in order to be studied alongside his other romances. Yet it is worth reminding ourselves that what we think we know about this medieval author is informed far more by modern literary criticism than by contemporary medieval documentation. The capacity for scholarly conjecture about the author Chrétien to harden into fact is noted by Sarah Kay: From an effect derived from these texts we postulate an entity that precedes them. The biographical approach makes a backdoor return in efforts to fathom an author’s literary personality, which then sets limits on the interpretation of ‘his’ works and conditions on admission to ‘his’ canon.41

Kay’s article speculates that the name ‘Chrétien de Troyes’ may have functioned as an anonym, a descriptive nom de plume adopted by a range of authors in order to signal their participation in a shared literary debate. In doing so, it offers a salutary reminder that the decisions we make about authorship and attribution shape our reading habits, often blinding us to alternative ways of understanding and interpreting medieval texts. Hult makes a similar point in his Morrison Library Inaugural Address, where he questions the assumption in so much critical literature that Chrétien’s reputation among medieval audiences was comparable to that which he enjoys today. Clearly, Chrétien’s name was associated with Arthurian romance at an early stage, as evidenced notably within the Conte du Graal continuations and prologues, where the name appears up to four times in the text depending on the manuscript (see below for more details). Yet the vast body of scholarship on this author today is not commensurate with their level of medieval popularity, a fact brought out by Hult’s comparison of manuscript survival for other twelfth-century texts: Aimon de Varenne’s Florimont has survived in seventeen manuscripts, while there are more than twenty extant copies of Wace’s Brut and over thirty-five of Benoît de Sainte-Maur’s Roman de Troie.42 To imply, as do the editors of Les Manuscrits de Chrétien de Troyes, that the more than forty surviving manuscripts containing at least one of his romances demonstrate that Chrétien’s pre-eminence was recognised by twelfth-, thirteenth- and fourteenth-century audiences, is to filter the phenomenon of medieval textual transmission through the distorting lens of modern assumptions about the cohesiveness of an authorial œuvre and the 40 41 42

Greene (2006: 218). Kay (1997: 2–3). Hult (1998: 14).

14 THE

CONTE DU GRAAL CYCLE

reciprocal relation that might obtain between an author’s reputation and that of his various works.43 It is worth emphasising again that the evidence for even a single ‘Chrétien collection’ manuscript is scant, if not inexistent. Manuscript R compiles Chrétien’s romances (with the opening episode of the First Continuation attached to the Conte du Graal) as an interpolation into Wace’s Brut; it also includes the Roman de Troie, the Roman d’Enéas and the Dolopathos. Manuscript A also contains the Brut and the Roman de Troie, along with Athis et Prophilias and Les Empereurs de Rome (see Appendix 4 for the order of texts in the manuscripts). As Hult points out, the way in which the romances have been compiled alongside (and, in R, within) more obviously historicising narratives, contextualises Chrétien less as an eminent author and pioneer of Arthurian fiction than ‘as a contributor of fragments to what amounts to a universal history’.44 If we are able to abstract the scholarly author-construction ‘Chrétien’ from our consideration of the medieval manuscript evidence, a different picture of the development of medieval Arthurian literature begins to emerge, one in which the Conte du Graal cycle as a textual unit plays a major role. For instance, the earliest illustrated ‘Chrétien manuscripts’ are cyclical Conte du Graal codices, dating from the last quarter of the thirteenth century. Busby suggests that the fact that some Conte du Graal manuscripts were illustrated at all is down to the popularity of the prose Vulgate Cycle manuscripts;45 this may then in turn have encouraged the illustration of Chrétien’s other romances. In this case, one might have to turn on its head Busby’s comment that ‘the evidence suggests that parts of the Continuations were very well-known indeed and one may surmise that this was at least in part due to their association with Chrétien’s Perceval’;46 we might say rather that the popularity of Chrétien’s other romances may be due at least in part to the association of his name with the Conte du Graal cycle, Continuations and all.47 This assertion is supported by data collated by Colette-Anne Van Coolput on textual references to Chrétien in medieval literature. Of the thirteen references which she lists, four come from the Conte du Graal

43 See Alison Stones’s ‘General Introduction’ (1993a: 3): ‘The number of surviving manuscripts and fragments… is already some indicator of the relative popularity of the texts.’ The same logic underpins Nixon’s argumentation in his article on romance collections (1993a: 17): ‘The more than forty manuscripts and fragments which contain one or more of the romances of Chrétien de Troyes form one of the largest legacies of any known author of medieval romance.’ 44 Hult (1998: 23). 45 Busby (1993d: 365–6). 46 Busby (1993d: 367). 47 Bruckner (2009: 3) similarly sees the Continuations as playing a potential significant role in the popularity of Chrétien’s work: ‘It might be argued that Perceval survives in many more copies than other romances by Chrétien… precisely because of the interest generated by the continuations.’

INTRODUCTION 15

cycle itself (C1 [MR] 1234 and 4116, CM 42641 (a variant from MS P), CG 6984, Elucidation 475). The cycle thus makes a habit of referring to its originary author, and its success can plausibly be considered to have contributed towards investing the name ‘Chrétien’ with a certain amount of literary capital. Of the nine other references, only two make explicit reference to any of Chrétien’s works: one (Miracle d’une none tresoriere) invokes him as author of Perceval and Cligés, the other (Sarrasin’s Roman de Hem) as author of Perceval.48 Meanwhile, Arthurian verse romances which allude to events in Chrétien’s works turn their attention exclusively to the events of the Conte du Graal cycle.49 Indeed, as I argue in Chapter 4, the ways in which thirteenth-century authors interacted with the cycle shows that they were far more attuned to its narrative aesthetics than modern readers have been, treating it as a coherent textual tradition with or against which to write. A major goal of this study will therefore be to read the Conte du Graal and its Continuations as the unified corpus which the manuscript evidence argues for, and by so doing to bring into view the significant role that it played in the reception and evolution of Arthurian romance, especially verse romance. In order to achieve this, it will be necessary to understand the aesthetic principles which underlie the corpus’s cyclic coherence. Before beginning to examine these, however, I return to two terms which have featured heavily already in the course of this Introduction – ‘continuation’ and ‘cyclicity’. By probing these notions further, greater light can be shed on the particularity of the Conte du Graal cycle, and on what study of this corpus can add to our understanding of medieval literature.

The Conte du Graal Corpus and Cyclicity The cycle as a literary form emerged in the late twelfth and early thirteenth centuries. Existing verse narratives, expanded and adorned with prequels and sequels, became ordered narrative compilations transmitted in cyclical manuscripts. The earliest extant cyclical codices date from the thirteenth century, and reflect a growing interest both in the production and ownership of manuscripts in general and, more specifically, in reading the cyclical narratives themselves. Fundamental to grasping the cyclicity of the Conte du Graal corpus is the crucial point that its elaboration, from the final quarter of the twelfth century to the second quarter of the thirteenth, is contemporary with the development of the major epic cycles, and of the prose Vulgate Cycle, such that one can place it alongside these vast works as part of a ‘cyclic turn’ in medieval French literature. Van Coolput (1987: 333–7). See Schmolke-Hasselmann [Middleton] (1998: 213), and pages 185–200 below, where I deal in more detail with the ramifications of this. 48 49

16 THE

CONTE DU GRAAL CYCLE

Recent years have seen much attention paid to the question of how to recognise and characterise cyclicity in medieval narrative, with two conferences in the early 1990s producing edited volumes on the subject.50 More recently still, Bart Besamusca has discussed cyclicity in some detail in his study of the Middle Dutch Lancelot Compilation, drawing heavily on five criteria proposed by Povl Skårup as entering into the definition of a cycle.51 Since Skårup’s ideas and terminology provide a good example of both the possibilities and limits of a general definition of cyclicity, it is worth briefly restating them here. We can speak of a cycle, he suggests, only in cases where we are dealing with (1) a sequence of originally independent texts, (2) contained in the same manuscript and ordered according to the progression of events, (3) which share the same protagonist or of which the protagonists are related. In addition, (4&5) there must be ‘cyclic signals’ both between texts and within texts in the ensemble. Skårup has in mind, on the one hand, linking passages which highlight the sequential nature of the different texts (such as an announcement of the next text that will follow); on the other, passages within texts which refer backwards or forwards to events in other parts of the cycle (‘allusions’) or which show signs of having been altered in order to avoid discrepancies with events elsewhere in the corpus (‘adaptations’). In other words, cyclic signals manifest a desire on the part of a compiler, redactor or author to establish the coherence of the narrative sequence. For Besamusca, ‘the features listed by Skårup are incontrovertible’. Yet he himself acknowledges that the Middle Dutch Lancelot Compilation, which he considers ‘a classic example of a narrative cycle’, ‘cannot be said to comply with the [third] criterion entirely: two romances… invalidate it’.52 While Skårup’s work might constitute a starting point for discussion of cyclicity, any attempt definitively to limit, either by inclusion or by exclusion, the essential features of a group of texts is liable to work less well the further away one moves from the particular case or cases on which the delimitation is based. Skårup’s criteria are no exception to this tendency: they are developed in the context of an analysis of the Karlamagnús saga, a cycle of Scandinavian translations of French chansons de geste, and define a model of cyclicity drawn ultimately from the chanson de geste, which maps less successfully onto other groups of text. For instance, as Besamusca himself notes, the genealogical aspect of Skårup’s third criterion is more appropriate to epic cycles, where the frequent narrative concern with feuding between families makes filiation (or, conversely, ancestry) a natural motor for cyclical 50 Besamusca, Gerritsen, Hogetoorn and Lie (1994); Sturm-Maddox and Maddox eds. (1996). 51 See Besamusca (2003) and Skårup (1994). The influence of Skårup’s definition is also apparent in the editors’ introductory comments on the nature of cyclicity in Besamusca, Gerritsen, Hogetoorn and Lie eds. (1994: 1). 52 Besamusca (2003: 146 and 141).

INTRODUCTION 17

prolongation;53 as for the Conte du Graal cycle, it clearly breaks the first rule, since each Continuation and prologue is explicitly conceived as an addition to the existing body of text. This quality of our cycle sets it apart from the majority of others, which are typically formed of sequences of individual narratives, each of which contains its own relative unity. One might usefully speak of its composition as embodying a principle of cyclic continuation. Thus, although Skårup’s criteria provide an insightful and productive way into thinking about cyclicity, the attempt to produce a single, comprehensive and synchronic, definition already begins to break down before the multifarious nature of the literary evidence. It may be more fruitful to look at cyclification as a process in diachrony. Here we can distinguish between three broad phases, beginning with the reconfiguration of originally independent texts to create longer narratives. The Guillaume cycle, one of the first cycles to be recognised as such by critics, is based around a core of three such twelfth-century texts that seem to have been brought together around 1190: the Couronnement de Louis (c. 1130), the Charroi de Nîmes (c. 1130–40) and the Prise d’Orange (c. 1140– 50). The second phase, which becomes visible around the turn of the thirteenth century, is the deliberate composition of new texts to add to an already established cycle, demonstrating a desire to fit the new narratives into the perceived cyclical frame. The Guillaume material was thus expanded through the addition of the Enfances Guillaume (c. 1200–1225), as well as the creation of new texts relating the youthful exploits and further adventures of other characters from the narrative: Les Enfances Vivien (c. 1200–1225), La Chevalerie Vivien (c. 1200), La Bataille Loquifer (c. 1200–1210) and Le Moniage Rainouart (c. 1190–1200) all date from this period. Two distinct phases can also be discerned in the cyclification of the Conte du Graal material, the first in the extension of Chrétien’s unfinished narrative by the first two Continuations, and the second in the addition of the two later Continuations and the two prologues to the now-established cycle; a significant difference in the case of our corpus is that the initial phase already involves the composition of new material in order to continue the interrupted narrative, rather than the bringing together of previously independent texts. The third stage of cyclification is the compilation of material into cyclical manuscripts, a process which implied a new phase of editing and rewriting to make the constituent parts fit together more harmoniously. As a result of this process, each cyclical manuscript potentially creates a different configuration of the corpus. Within the Conte du Graal cycle tradition, this phase begins concurrently with the second, as the First Continuation in particular is subjected to numerous recastings which its editor sorts into three families: the Short, Long and Mixed Redactions. Indeed, another notable feature of

53

Besamusca (2003: 141).

18 THE

CONTE DU GRAAL CYCLE

cyclicity is the precarious nature of the unity it creates; the constituent parts of cycles may be pulled apart, left out or transmitted independently. As a result, each cyclical manuscript represents a potentially unique realisation of cyclicity, based on its choice and arrangement of elements. An archetypal instance of this is MS BNF, fr. 112, completed by Micheau Gonnot in 1470. Gonnot wove bits from various prose romances, some already cyclical, others not – namely, the Lancelot proper, the prose Tristan, three versions of the Queste del Saint Graal, La Mort le roi Artu, the Suite du Merlin, Palamède and the Prophéties de Merlin – into what is effectively a continuous Arthurian narrative spread over four books (of which the first is now lost).54 The Gonnot manuscript thus exemplifies the totalising drive that characterises a certain kind of cyclicity in its attempt to produce the most complete history of the Arthurian world possible. A manuscript such as BNF, fr. 112 demonstrates that the third stage of cyclification – the collection of material into cyclical codices – can operate on material not previously subjected to the first two, creating one-off cyclical compilations through the imposition of a binding principle of coherence on its disparate texts. MS BNF, fr. 1450, which as we have seen contains Chrétien’s romances and a Branch of the First Continuation interpolated into Wace’s Brut, has frequently been described as cyclical in this sense.55 It contains Benoît de Sainte Maure’s Roman de Troie, the Roman d’Eneas, the Brut with its Chrétien interpolation and the Roman de Dolopathos, a version of the Sept Sages de Rome. This sequence of texts articulates a model of translatio imperii, moving from Troy to Rome and then to Britain (a trajectory personified in the figures of Eneas and his descendant Brut), which becomes the manuscript’s cyclical narrative, above and beyond the narrative movements of each individual romance. The circular implication of the term ‘cycle’ is thus materialised in the meta-narrative of the rise and fall of civilisations.56 The scribal handling of the material manifests a conscious desire to create this architectonic. The prologue to the Roman de Troie is detached from the rest of the text and thereby made to serve as the prologue to the whole collection; Chrétien’s romances, meanwhile, are shorn of their prologues and made to read like a continuous narrative, with the exception of the Cligés prologue which, famously, includes a celebration of translatio studii et imperii. This

On this manuscript, see Pickford (1960). Walters (1985); Maddox (1996: 40); Taylor (1994: 63–4); Besamusca (2003: 150–1). 56 Jane Taylor describes this narrative model as ‘organic’ cyclicity, where the fictional structure is modelled on a conception of time itself as a cycle. She views organic cyclicity as dominant from the fourteenth century onwards, exemplified by cycles such as the Perceforest. See Taylor (1994). 54 55

INTRODUCTION 19

prologue, though buried in the middle of the textual sequence, thus becomes a kind of mission statement for the whole manuscript.57 The Conte du Graal cycle manuscript tradition testifies to the coherence of the corpus as perceived by its audience, but this coherence admits of a degree of instability. The most common grouping (eight out of eleven copies) is CdG-C1-C2-CM, forming a central core into whose orbit the other texts are pulled more occasionally. This might thus be termed the canonical form of the cycle, as Bruckner also suggests; however, as she goes on to note, ‘“canonical” in this context admits of many different actualizations, given the three different redactions of the Gauvain Continuation (Short, Long, and Mixed), the addition of materials preceding Chrétien’s romance, as well as the inclusion of the Fourth Continuation in TV’.58 Indeed, there is a limited amount of evidence for contemporary appreciation of the corpus as made up of separable segments: the rubric in MS A distinguishing the ‘old Perceval’ from the rest of the ensemble, the four MSS (BCFH) where Chrétien’s text is presented on its own and the one instance of a Continuation treated independently, Roach’s MS K (Bern, Burgerbibl. 113), where the Second Continuation appears on its own as part of a romance compilation, with a unique ending designed to make it a self-sufficient narrative. We can therefore conclude that the cyclic version of this corpus was more popular than other permutations of the material, even though its unity remained precarious. The preference for cyclic coherence is itself built into the overarching narrative of the corpus, which bears out certain of Skårup’s criteria: the order of its constituent parts is broadly guaranteed by the progression of the central narrative, and the various parts of the corpus are insistently tied together by what Gomez calls ‘effets de cycle’: signs of a relationship between an individual unit and a cyclic whole, which may be perceivable in the content or the structure of the work.59 Moreover, the oldest surviving manuscripts of our cycle are contemporary with the earliest extant cyclical manuscripts in any genre. The first surviving manuscripts of the Vulgate Cycle material are dated to around 1250, with production really taking off in the fourteenth century. Our MS L, located by Nixon in the first quarter of the thirteenth century, thus predates the earliest extant copies of a prose romance cycle by some time.60 The fact that this manuscript does not contain the later parts 57 See Huot (1987: 28): ‘Enfolded in the heart of the book… this classic statement of translatio extends its significance throughout the collection.’ 58 Bruckner (2009: 188–9). 59 Gomez (forthcoming). The notion of ‘effet de cycle’ is broader in scope than Skårup’s concept of the ‘cyclic signal’ in that the latter is defined narrowly as a linking passage which establishes continuity with preceding or succeeding texts; Gomez’s ‘effet de cycle’ refers more generally to any element in a text that invites the audience to read it as part of a larger textual whole. 60 Note however that Alison Stones (1977) dates one manuscript containing the Estoire, Merlin and Lancelot (Rennes, BM 255) to the 1220s. Busby’s edition of the

20 THE

CONTE DU GRAAL CYCLE

of the cycle is less significant than the fact that (aside from the short Vie de Sainte Marie l’Egyptienne included at the end) it contains only the sequence Bliocadran-CdG-C1-C2, presented as a single textual unit. This succession of texts, even without the ending provided by the Manessier Continuation, was thus already recognised as a cyclical corpus, and the absence of the later Continuations suggests simply that their existence was not known to the planner(s) of the manuscript; indeed, depending on the precise date of the manuscript’s production, they may not even have been written yet. The Conte du Graal cycle thus provides direct evidence, both in its composition and transmission, of crucial stages in the evolution of medieval cyclicity. Yet, despite these indices of unity, a number of factors have led scholars to impose boundaries onto the various parts of the corpus. The simplest of these is one of scale: each Continuation is longer than an average Arthurian verse romance (including all of Chrétien’s), and the length of the resultant corpus obliges some kind of breaking up of the text into parts in order for it to be accessible to modern audiences. With the exception of Potvin’s inaugural 1866 edition of Perceval le Gallois, which reproduces the whole contents of MS P as a single textual entity, editorial practice has always favoured the separation of the corpus into constituent parts, determined by the principle of presumed authorship.61 This practice is supported by the belief that each of these parts was written separately, and so scholars have felt justified in considering each part as separable in recognition of the ‘multiple authorship and considerable differences of style and intent’ between the different Continuations.62 And the reverence shown by medievalists towards the ‘Chrétien corpus’ means that one is more likely to find the Conte du Graal edited or studied together with other Chrétien romances than with the rest of the cycle from which it has been extracted. As a result, an excessive focus has been placed on the search for boundaries between discrete textual units; the ensuing debate over who wrote what has effectively precluded the unified reading demanded by the cycle’s aesthetics of coherence.63 What is missed in this insistence on defining textual limits is the way in which the different parts of the corpus are interrelated through a play of difference and similarity, where filiation between a particular section of the cycle and its predecessors is under constant renegotiation.64 Indeed, one of the most distinctive features of the Conte du Graal cycle is the tension between the corpus’s propensity Conte du Graal dates L to the second half of the thirteenth century, but it also dates E to the first half (rather than third quarter as given by Nixon) and the same observations could be made about this manuscript, mutatis mutandis, as are made here with regard to L. 61 See Potvin (1866–71). Oddly, Potvin also published the prose romance Perlesvaus (not found in MS P) as Volume I of his edition. 62 Pickens, Busby and Williams (2006: 222). 63 On the debates around authorship and how to divide up the various parts of the text, see Wilmotte (1930), Lot (1931), Roach (1956), Vial (1978), Corley (1984) and (1987). 64 As Bruckner (2009: 46) warns: ‘If we aim to get a sense of what the process of

INTRODUCTION 21

to elide the transitions between its constituent parts on the one hand and the tendency for each part to assert its own specific character on the other. In this cycle of continuations, each new part is designed to follow organically on from the last; yet, with the exception of the anonymous First Continuation, each new addition also manifests a desire to assert a measure of independence from Chrétien’s original text. This is felt most forcefully in the last two Continuations, whose authors demonstrate a concern for helping their readers to identify the extent of their contributions. Indeed, both texts refer back to the same moment in the narrative as the starting point for their work. Manessier declares at the end of the cycle that his work ‘conmença au soudement / De l’espee’ [began at the point where the sword was mended] (CM 42660–1), while Gerbert specifies in his Continuation that he takes up the story from the point at which Perceval ‘La bone espee rasalda / Et… du Graal demanda’ [mended the good sword and… asked about the Grail] (CG 7011–12).65 On the strength of these two authorship claims, and the fact that two manuscripts (KL) end here, the end of the Second Continuation has been located at the moment when Perceval mends the broken sword kept by the Fisher King in the Grail Castle. In the two manuscripts (TV) which contain Gerbert’s contribution, the final fourteen lines of the Second Continuation feature twice, as lead-in to both the Gerbert and Manessier Continuations, with Gerbert’s text therefore sending Perceval on a looping journey which leads him back to the same situation in which he began.66 I will look at this phenomenon further in Chapter 1, but for the time being the relevant point is that Gerbert’s authorship claim allows the alert reader to work out the part of the cycle for which he is responsible. Similarly, in the Second Continuation, Wauchier de Denain – rather than Chrétien – is invoked by the narrator as his authority: Gauchiers de Dondain, qui l’estoire Nos a mis avant en memoire, Dit et conte que Perceval, Li bons chevalier, li loial, Erra bien pres de quinze dis  (C2 31421–5) [Gauchiers de Dondain, who gave us the story to remember, tells and recounts that Perceval, the good and loyal knight, wandered for close to two weeks] continuation is about, trying to tease apart what has been so assiduously written together seems counterproductive.’ 65 Most studies conclude that Gerbert and Manessier had no knowledge of each other’s work, and that Gerbert’s conclusion is now lost, cut off by the scribes of TV in order to include both Gerbert and Manessier’s contributions to the narrative. For a conflicting viewpoint, see Stephens (1996). 66 On the idea of the Gerbert Continuation as a loop which returns to the place it began, see Bruckner (2000) and (2009: 192–7).

22 THE

CONTE DU GRAAL CYCLE

Though this comment does not allow the audience to know the extent of the narrative to which Wauchier (through the narrator) is laying claim, it does constitute the cycle’s first official challenge to the authorship of Chrétien. And in case there was any room for doubt about the corpus’s multiplicity of authors, the Gerbert Continuation identifies Chrétien as the man who ­‘Perceval comencha’ [began Perceval] but who died before completing it (CG 6984–7), thereby inscribing its own authorship in a post-Chrétien era. Authorship in this kind of cycle thus implies a process of negotiation between fidelity to the narrative past and assertion of independence. The movement from one author’s contribution to another’s constitutes a potential moment of narrative crisis, in which the direction which has previously been established and defended may be challenged or even abandoned. The presentation of the corpus as a single textual unit makes the identification of these moments of transition a matter of critical conjecture, dependent in large part on one’s view of who was involved in writing what: in Bruckner’s words, ‘the authorial identities present are decidedly irregular, unpredictably placed, and often as invisible as the universe’s dark matter: three named continuators certainly do not account for all the writing gathered together’.67 Indeed, the most vexed question, and least susceptible of finding an answer, is how to identify and quantify the contribution of the remanieurs and scribes through whose hands the text passed during the period of its elaboration. Nevertheless, we can at least work with the broad transitions which have found consensus among scholars, and a brief analysis of these will suffice to show that, in general, the transition from one textual unit to the next is effected in the middle of an episode, so that each continuator’s first task is to resolve or conclude the situation inherited from his predecessor (the exception is that between the First and Second). It is as if each writer is attempting to set the parameters within which the next will have to operate, and thus defend the body of work that he himself has contributed. The first identifiable transition takes us from Chrétien’s portion to the First Continuation. Chrétien’s narrative ends with both Arthur and his queen in a faint, and a messenger about to arrive to request their presence at Gauvain’s duel with Guiromelant. The continuator was therefore forced to follow this narrative strand until the duel was concluded, before being able to make his own narrative choices. The position of the break between the First and Second Continuations is more controversial. It has traditionally been placed after the conclusion of the Guerrehet adventure (Branch VI, Episode 8), the new text opening with a return to Perceval for new adventures; it would in this case constitute an exception to the tendency described above. However, Corin Corley has suggested that the boundary between the texts should in fact be the point in Episode 5 at which Perceval, having been sent to kill a stag

67

Bruckner (2009: 41).

INTRODUCTION 23

with a white dog belonging to a damsel he is attempting to seduce, has the stag’s head and dog stolen by an unknown knight (C2 [SR] 10268); if Corley is correct, the end of the First Continuation also attempts to chart the course of its own succession.68 Significantly, each of these points – the disarray of Arthur and his court, the dismay of Perceval struggling to understand what has befallen him – is a moment of crisis within the narrative, as if reflecting the potential crisis in narrative direction which new authorship implies. This is also the case in the transition between the Second Continuation and the Gerbert (in TV) or the Manessier Continuation (in the other manuscripts, which omit Gerbert’s text). Perceval joins the pieces of the broken sword but a crack remains, and he is warned by the Fisher King that he is not yet sufficiently virtuous to pass the test fully. Here, as before, the diegetic crisis makes necessary the provision of a solution by the new author, locking him temporarily into the narrative programme of his predecessors. Yet, in general, these attempts to control the parameters of the succeeding text work only at a local level, that of the episode which forms the point of contact; each continuation, over the several thousand lines of its length, chooses its own distinctive narrative path. The periodic shifts in narrative dynamic thereby occasioned are one of the most fascinating aspects of the cycle, and these are emblematised in the corpus’s deployment of a variety of heroes, with the handling of their different trajectories often decisive in defining the narrative aesthetic of each individual part. The addition of secondary heroes is a common feature of cyclic narrative, as Luke Sunderland emphasises, often resulting in a tension between the desire for narrative progress (movement towards the conclusion of the story/life of the central hero) and for narrative proliferation (delaying forward progression via the introduction of counter-narratives): Alternation between heroes is part of the drive to increase narrative, but the figure of the hero – the central protagonist – defines and limits the entire scope. Biography and cyclicity thus clash.69

These two conflicting narrative dynamics I refer to in the current study as centripetal and centrifugal. Centripetal narrative is governed by a narrative centre, often the principal protagonist, to which the movements of the narrative are subsumed, and periodic returns to which guarantee its coherence. There is a clear hierarchy between central narrative and subplot. Centri­fugal narrative may gesture towards a centre, yet its movement tends to spin outwards to generate new narratives, none of which is given prominence over the others. This terminology is adapted from Bruckner, who applies the distinction to the field of Arthurian intertextuality. She characterises the relation 68 69

Corley (1987: 4–5, 37–41). Sunderland (2010: 12).

24 THE

CONTE DU GRAAL CYCLE

between Chrétien de Troyes’s Lancelot and the Prose Lancelot as one of centripetal intertextuality, meaning that Chrétien’s text is absorbed by the prose romance as its central episode: From the condensed version of that episode, the rest of this enormous prose cycle seems to radiate in all directions: the matter of Lancelot’s story becomes the model through displacement and amplification for countless reinventions that help map out the chronology of Lancelot’s life and Arthur’s reign.

By contrast, she perceives the intertextual relation between Chrétien’s Conte du Graal and its Continuations as centrifugal: Chrétien’s romance… functions as a point of departure from which the continuators move away in apparently linear fashion, only picking up the threads of Chrétien’s plot here and there, while introducing a great deal of material unanticipated by his already lengthy 9000 verses.70

In her monograph Chrétien Continued: A Study of the Conte du Graal and its Verse Continuations, Bruckner returns to and refines this concept. The principal adjustment is a greater emphasis on the repeated re-absorption and re-invention of the originary text as the narrative proceeds: Centrifugal intertextuality thus figures the way Chrétien’s model serves as a repeating center throughout the series by remaining off-center at the narrative beginning, the place where all continuators return for inspiration and reinvention in order to set out anew, even as they pick up the linear thread of narration wherever their immediate predecessor left it.71

As this comment suggests, Bruckner argues throughout her study that Chrétien’s originating romance functions as a ‘master text’ exerting continuing tutelage over the expanding narrative, as ‘successive authors and continuations lead back repeatedly to the original author and his text’.72 This approach undeniably yields significant insight into how medieval readers and writers responded to the Conte du Graal, and as we will see there are many points of convergence between her analyses and the argument of this book; however, Bruckner’s model obscures to a certain extent the way in which each successive extension of the narrative produces a new ‘master text’ with or against Bruckner (1987: 225–6). Bruckner (2009: 16). See also in this regard Annie Combes’ comment on the Second Continuation’s cannibalising of the motifs and themes of Perceval’s initial visit to the Grail Castle in the Conte du Graal: ‘de multiples façons, la soirée vécue chez le Roi Pêcheur hante le récit’ [in various ways, the evening spent at the Fisher King’s castle haunts the narrative] (Combes and Berthin 2000: 48). 72 Bruckner (2009: 17). See also pp. 2, 14–16, 73, 227. 70 71

INTRODUCTION 25

which the next author in the sequence will write. Moreover, as I argue in Chapter 1, the choices made by each continuator impact retroactively on the inherited textual mass to which he is responding, so that Chrétien’s very authority as originator is partly a creation of the later texts. To write of Chrétien’s continued ‘authorship’ extending across the Continuations tells only half the story; for Chrétien’s text is itself ‘rewritten’ through each twist and turn to which the cyclic narrative is subjected by his successors.73 Whereas Bruckner sees Chrétien’s text as the operative centre of the cycle, I argue that it is the principal hero, Perceval, and his gradual progression towards the status of ‘best knight in the world’ that provides what I call its centripetal core. There are moments in the cycle where the centrifugal dynamic predominates, especially in the First Continuation; as we saw above, the multiplication of heroes in this text is one of the features that have led to its being singled out as the least coherent part of the corpus. However, we must also keep in mind that the core of the cycle is a biographical narrative, which begins with Perceval’s first contact with knighthood and ends with his death. Indeed, from the Second Continuation through the Gerbert and Manessier Continuations, the counter-narratives of other heroes are inscribed squarely within the dominant thread of Perceval’s adventures. The Conte du Graal cycle is thus informed by both centripetal and centrifugal principles; it is the productive tension of these two dynamics which defines the nature of our cycle. Centripetal, biographical narrative provides the core without which the corpus would not cohere as a textual unit; centrifugal, digressive counter-narratives provide the material which swells the text to cyclic size and delays the end for as long as its authors desire. Indeed, a distinctive overall feature of this cycle is what Bruckner calls its ‘incompletability’.74 Many cycles are predicated on a drive for completeness, with closure the declared ultimate goal. The Vulgate Cycle, for instance, is structured around the desire of both characters and audience for the fulfilment that knowledge about the Grail would bring; as Sunderland remarks, ‘we spend the entire text waiting for the appearance of the one hero who will be worthy of the sublime Grail’.75 The Grail itself, as Miranda Griffin has argued, functions as both ‘beginning and end’ of the cycle, the object which both calls forth and terminates the adventures of the text.76 Moreover, the Grail storyline is interlinked throughout the cycle with the account of the 73 In the same way, it makes a real difference to the aesthetic experience of reading Guillaume de Lorris’s Rose whether one encounters it on its own or (like the vast majority of readers medieval and modern) as part of the conjoined text created once Jean de Meun’s continuation is appended to it. 74 Bruckner (1987: 262). 75 Sunderland (2010: 20). 76 Griffin (2005) sees the Grail, the book and the body as the objects around which the cycle is structured. All three offer a promise of wholeness, although they simultaneously flag this as a fantasy masking a troubling fragmentation.

26 THE

CONTE DU GRAAL CYCLE

rise and fall of Arthurian chivalry; the death of the king and the annihilation of his society provide the cycle’s eventual endpoint. In the Conte du Graal cycle, the Grail narrative is a personal matter, of interest to Perceval (and, from the First Continuation, to Gauvain) but not to wider Arthurian society. Further, whereas the events of the Grail quest are endlessly prefigured in the Lancelot, and then remembered in the Mort Artu, Perceval’s itinerary is constantly modified as he progresses through the text, so that the audience is kept guessing as to the nature of his next task. The precise nature of his involvement with the Grail is famously opaque; yet, as I argue in Chapter 1, by the time we return to him in the Second Continuation the initial questions which he was supposed to ask in the original romance have already been answered. New challenges are needed in order to give new impetus to the text. The ‘grail question’ test becomes replaced by the ‘broken sword’ test, and when he passes this at the end of the Second Continuation, Gerbert and Manessier must find new excuses to send him away from the Fisher King’s castle long enough for his return to constitute an achievement. Thus, although the text does end with Perceval’s death and ascension to Heaven along with Grail, there is something almost gratuitous about this conclusion; one feels that the cycle could have rattled along indefinitely, sending its hero on a succession of new adventures away from, and back to, the Grail Castle.77 The symbol of the cycle’s endlessness is surely the broken sword kept by the Fisher King at the Grail Castle, which only one knight will be able to mend successfully. At the end of the Second Continuation, Perceval achieves this goal, but a small crack remains: Les pieces prist a rajoster, L’un[e] a l’autre delivremant, Et li aciers ansamble prant Si bellemant et si a droit Que lou jor qu’elle faite estoit Ne sambla estre plus novelle, Ne miauz forbie ne plus belle; Mais que tot droit an la jointure Fu remese une creveüre Petitet[e], non mie granz.  (C2 32550–9)

77 Manuscript K, which contains only the Second Continuation, offers a unique conclusion which demonstrates in slightly grotesque fashion that reaching the end is really beside the point in this cycle: having passed the sword test, Perceval heals the Fisher King of his wound, is made his heir and succeeds to the throne, all in the space of fifty-seven lines (see Appendix XI in Roach 1971 [IV]: 591–2). It can therefore be considered a demonstration in negative form that the individual parts of this corpus make best aesthetic sense when read together. In her taxonomy of continuation, Tether (forthcoming, ‘Conclusion’) classifies this short text as a Conclusion offering Short-Term Gratification.

INTRODUCTION 27

[He quickly took the pieces to put them together, and joins the steel so well and so rightly that on the day it was made it did not look newer, nor better-made, nor more beautiful, with this exception: at the point where the pieces were joined, a small, not very substantial crack remained.]

Has the adventure been completed? For Manessier, the answer is yes, and so he devises a new quest around the need for the crippled Fisher King to be avenged. Gerbert, by contrast, seizes upon the sword as an excuse to send his hero back out into the Arthurian world; the crack that remains becomes symbolic of the spiritual improvement still required of Perceval. The ambivalence of this sword, perfectly joined yet visibly cracked, seems to me to be richly evocative of how comfortable this cycle is with the notion of incompleteness. Throughout the corpus, gaps are deliberately opened, other narratives evoked, as the text appears to revel in its inability to make whole the fractured world of Arthurian narrative. And in its preference for middles over endings, and its flagrant disregard for canonically approved principles of coherence, it may have hitherto unappreciated relevance to discussions of narrative theory often reserved for far more recent experimental fiction.78 The analysis of the cycle offered here approaches the text from a variety of perspectives in order to gain a multi-faceted understanding of its aesthetics, before undertaking a reconsideration of its importance to the evolution of Arthurian romance. Chapter 1 begins by studying the cyclification of the corpus through its different versions. I argue that the Second Continuation plays a crucial role, both in imposing a retroactive cyclic coherence on the earlier material, and in establishing an aesthetic model for the later Continuations centred on the figure of Perceval; I also demonstrate how the First Continuation is rewritten and expanded in order to fit better with the cyclic aesthetics of the whole. In the second half of the chapter, I study how the motif of questing is used to generate narrative, with interlace emerging as a key structural feature. In light of these observations, Chapter 2 studies how the surviving manuscripts of the corpus received and re-interpreted its narrative aesthetic. I argue that paratextual information such as large initials, rubrics and illustrations function to highlight and strengthen the two principles of coherence identified in Chapter 1: the interlacing of adventures and the primacy of Perceval as central hero. I also trace a shift between the earlier and later manuscripts, with the later ones exploiting more fully the possibilities of the material object as a visual text, relating this change to the emergence of literate modes of thought. 78 Bruckner also views resistance to closure as central to the dynamics of the cycle: ‘The perception and power of the incomplete middle may in fact be the motor force that drives the narrative structure as a whole’ (2009: 187–204; 189 for the citation).

28 THE

CONTE DU GRAAL CYCLE

Chapter 3 focuses on the interaction between the cycle’s different authors, showing how the semantic fields of fertility, genealogy and recognition allow the authors to explore the ethical stakes of the relationship between textual (and authorial) past and present. The final chapter draws on the conclusions of the first three to establish the influence of the Conte du Graal cycle on the aesthetics of Arthurian verse romance. Where Chapter 2 focused on the response of one group of thirteenth- and fourteenth-century audiences (the makers and users of the manuscripts), Chapter 4 considers how another kind of response, that of the writers of later Arthurian verse narratives, conditioned the ways in which the genre developed in the thirteenth century. Attention is paid to the use in these later texts of those motifs and narrative techniques identified as crucial to the narrative aesthetics of the cycle, as well as to the phenomenon of manuscript compilation and its relationship to the manuscript transmission of the cycle. The book concludes by discussing a parallel development in Arthurian literature, the Vulgate Cycle, which was ultimately to eclipse the Conte du Graal cycle in popularity; my conclusion considers how the findings of this study might be carried forward into further work on the interaction between the two cycles, an interaction characterised by mutual influence but also by competition. It is a critical orthodoxy to view the development of Arthurian romance as branching out from the œuvre of Chrétien de Troyes in two directions: the short verse narratives on the one hand and the prose cycles on the other. Literary history has consecrated chivalric romance as a genre invented by Chrétien and reworked by the French prose romance tradition into a form which was to find its apotheosis in extensive fifteenth-century narratives such as Malory’s Morte d’Arthur, Martorell’s Tirant lo Blanc and Boiardo’s Orlando Innamorato, and ultimately to play a role in the beginnings of the modern novel through the farcical adventures of Don Quixote. As this book demonstrates, the Conte du Graal cycle needs to be acknowledged as a part of this story. Its influence on thirteenth-century Arthurian romance is too infrequently noted; further, there are good reasons for thinking that it played a significant role in establishing Chrétien’s own literary reputation with subsequent generations. Appearing at a crucial moment in both the evolution of cyclicity and of Arthurian literature, this unique literary corpus deserves to be studied for what it can add to our understanding of both of these phenomena.

1

Narrative Aesthetic and Cyclic Formation In this chapter I will examine how narrative aesthetic is created, promoted and contested at different points in the Conte du Graal cycle. Such a long text, with multiple authors, will inevitably change as it progresses, with successive authors choosing how to develop the narrative project inherited from their predecessors. Moreover, parts of the corpus were substantially rewritten, perhaps several times over, during the text’s manuscript transmission. The following discussion will track this complex process, arguing that the cyclification of the Conte du Graal material is characterised by the imposition of two related narrative structures onto earlier, more centrifugal material: interlace, where multiple knights’ adventures are woven together, and biography, with Perceval’s chivalric and moral progression becoming the text’s central concern. Particular attention will be paid to material which recurs or is reworked several times through the corpus, taking account of the whole manuscript tradition of the cycle. I leave until Chapter 2 a consideration of what the manuscript presentation of the cycle can tell us about its reception, focusing here on how narrative structures work to produce and promote specific textual aesthetics. The structure of this chapter acknowledges a distinction between the diachronic and synchronic study of cyclicity – that is, the study of cyclic formation through time (the diachronic dimension) and that of the cyclic properties of an ensemble at a given stage of that formation (the synchronic dimension).1 Diachronic analysis traces the development of non-cyclic into cyclic material, often through successive stages in manuscript transmission; synchronic analysis considers a given cyclic manuscript or set of manuscripts as a specific and self-sufficient instance of cyclicity. This chapter will accordingly be divided into three parts. The first section investigates the genesis of the cycle, paying particular attention to the existence of multiple redactions of a large chunk of the corpus. It identifies two distinct stages in the initial response to Chrétien’s original text: the centrifugal narrative of the First Continuation and the centripetal dynamic imposed retroactively by the Second Continuation onto the sequence CdG-C1-C2. This new dynamic, which takes Perceval as the cycle’s biographical core, sets a standard followed by the

1

See Maddox (1994b: 102–3).

30 THE

CONTE DU GRAAL CYCLE

two later Continuations, and prompts the recasting of the First Continuation into a Long Redaction which is tied more closely to the cyclic agenda. The second and third sections will examine synchronically how the treatment of a key narrative motif – the quest – changes through the cycle, accompanying and at times underpinning changes in narrative structure. Again, the Second Continuation is revealed as the central creative force behind the imposition of the cycle’s biographical centre, with Perceval gradually acceding to the status of best knight in the world.

Cyclification: the Long and the Short of It The first three volumes of William Roach’s edition of the Continuations are given over to four different versions of the First Continuation. He identifies three redactions: the Mixed (represented by T in volume 1), the Long (represented by E in volume 2) and the Short (for which he prints the texts of L and A on facing pages, identifying L as differing significantly from the other Short Redaction manuscripts). In essence, we can say that there are two distinct versions, one of which is generally much shorter than the other, and that (as its name indicates) the Mixed Redaction found in TV is an amalgam of the two, containing little not found in either. Indeed, the tripartite model proposed by Roach somewhat obscures the important point that the behaviour of TV is not all that different from that of P or of U, which include some episodes from the Long Redaction and others from the Short. Pierre Gallais, meanwhile, has suggested that E, which Roach takes as his base manuscript for the Long Redaction, is copied from an exemplar close to MQ but provides interpolations from an L-type Short Redaction manuscript.2 It was clearly standard practice for scribes to cull material from multiple exemplars when transcribing the text; as we will see, the specific choices of what to include and what to leave out made by each scribe have a direct impact on the narrative dynamic presented by a particular version of the cycle. Considerable attention has been paid to the question of which of the two main redactions is closer to the ‘original’ text of the Continuation, though it seems impossible to tell how many author-redactors were involved in the creation of either version as it exists today.3 The desire to identify one version as the ‘original’, implicitly more authentic, text, is partly generated by the profusion of similarities and differences between the various manuscripts Gallais (1964: 183). I would therefore second Roach’s caution in the Introduction to Volume 1 of his edition (1949 [I]: xl): ‘the goal of the present edition of the First Continuation is not to provide the original text nor the “genuine” text of the First Continuation, because it has not been demonstrated that there ever existed any single state of the text which lay back of all the extant copies at however remote an earlier stage.’ 2 3

NARRATIVE AESTHETIC AND CYCLIC FORMATION

31

and the critic’s attendant fear of becoming lost in this textual maze, but once again excessive focus on how to divide up and classify the evidence may obscure the essential point that the text’s openness to elaboration is a crucial part of its aesthetic.4 It is perfectly possible, and may prove more fruitful, to work with a model of textual development that does not depend on identifying an Ur-text; thus, without pronouncing on what is or is not ‘original’, I concur with the now widespread agreement that the Short Redaction is anterior.5 As I argue below, the Long Redaction has stronger coherence than the Short with the material coming before and after it in the corpus, and a number of the details which distinguish it from the Short Redaction appear to be inspired by the Second Continuation. I therefore see the Long and Mixed Redactions as a more or less stable series of elaborations of and interpolations into this Short Redaction material. The question then immediately arises as to why the Short Redaction of this Continuation should have provoked such extensive reworking. An answer may lie in Volume IV of Roach’s edition (the Second Continuation). The division of the text into two redactions is maintained by Roach beyond the end of the First Continuation and through the first five episodes of the Second. At this point, two striking things occur in the manuscript tradition. In the first instance, the manuscript relations break up: there are several distinct renderings of the theft of Perceval’s white dog and stag’s head, and three manuscripts (EPS) include three subsequent episodes which are not in the others; secondly, from the point that the manuscripts come back together, the different versions of the text are sufficiently homogeneous for Roach to switch to a single redaction with variants at the foot of the page. In other words, the redactor(s) of the Long Redaction continued to rewrite the first few episodes of the Second Continuation, where the narrative returns to Perceval for the first time since the Hermit episode of Chrétien’s section (a hiatus of over twelve thousand lines in the Short and over twenty-two thousand in the Long). Taken together with the fact that no manuscript ends immediately after the Guerrehet branch (where Roach ends his edition of the First Continuation), it appears that the manuscripts wish to play down any notion of rupture between these two texts. Indeed, the First Continuation is not found in any manuscript without at least these early episodes of the Second.6 This suggests that the Second Continuation was written shortly 4 Thus Gallais’ otherwise painstaking comparison of the manuscripts of the First Continuation is seriously marred by the unargued conviction that MS L, which he identifies as offering the earliest version of the text, preserves grosso modo the work of a primary author, to whom should be attributed the entire text of the Short Redaction in this manuscript. See Gallais (1988–89, I). 5 See e.g. Roach (1956: 110–11); Corley (1987: 2); Gallais (1988–89: I, xvii–xviii); Combes (2004: 4). 6 The exception, manuscript R, contains only the part of Branch I which concludes the Guiromelant section, left interrupted at the end of Chrétien’s contribution.

32 THE

CONTE DU GRAAL CYCLE

after the First, and associated with it more or less immediately. In light of all this, then, we might reformulate the question considered above: why, if the manuscript fates of the two Continuations are so closely linked, does the vast majority of the Second Continuation not undergo the extensive rewriting received by the material which precedes it? And what effect does the rewriting have on the way in which this material is read? Despite their close association in the manuscripts, the first two Continuations present quite different models of the process of continuation. As mentioned in the Introduction, many scholars have commented on the apparent lack of fidelity to Chrétien’s text displayed by the First Continuation. Jean Frappier’s description of the text as ‘un chapelet de lais ou de fabliaux courtois que seule relie la permanence du décor arthurien’ [a collection of courtly lais or fabliaux bound together only by the enduring Arthurian background] exemplifies the limitations of critical categories in attempting to relate the First Continuation to what comes before and after it.7 It also bears witness to scholarly perplexity at a continuation of Chrétien’s narrative which ignores Perceval, introducing instead heroes (Caradoc and Guerrehet) who do not appear in the earlier text. Roach’s division of this part of the text into six sections, each with its own set of episodes, when he divides the Second and Manessier Continuations into sequentially numbered episodes, reflects the critical consensus as to its exceptionality. Although Roach himself does not use the term ‘branch’ usually found in scholarship on the First Continuation, this terminology derives support from within the text itself: just before narrating the Guerrehet material which critics refer to as Branch VI, manuscripts LMQU insert the following passage (I give the version from MS M): Li contes faut ci [L: Li grans contes cange] entreset, A une autre branche revet Que vos m’orroiz sanz demorer Tout mot a mot dire et conter. Chascun de vos [U: nous] cuide savoir Du grant conte trestot le voir, Mes nu set pas, se Diex me gart, Tot en ordre, par grant esgart, [Q: la quarte part] [L: Covient] Toute la chose deviser; Ja ne m’en orroiz ainz parler S’en ordre non et a droit point, Einsi con li contes se joint.  (C1 [LR] 18363–74; [SR: L] 8299–310)8 Frappier (1973: 153). Roach corrects v. 18371 to MS L’s ‘Covient la chose deviser’, but I believe the reading given by all three Long Redaction manuscripts here to be intelligible, given the punctuation I suggest above. 7

8

NARRATIVE AESTHETIC AND CYCLIC FORMATION

33

[The tale breaks off here forthwith; it goes to another branch, which you will hear me, without delay, tell word for word in its entirety. Each of you think you know all the truth of the great tale; but none of you is able, God save me, to relate the whole matter rightly in the correct order. You will never hear me speak of it if not in order and correctly, just as the tale fits together.]

The manuscripts in which this passage occurs are MQU, those that Roach assigns to the Long Redaction (the other Long Redaction manuscript, E, is missing the folios which would have contained this section), and L’s version of the Short Redaction. Here, the narrator’s concern is to define the relation of the ‘branche’ which he is about to narrate to the ‘grant conte’ which, he suggests, is the focus of his audience’s interest. The same distinction is made in these manuscripts (and not in the others) just before Gauvain’s arrival at the Grail Castle in Branch V: Seignor, la branche se depart Du grant conte, se Dex me gart; Des ore orroiz conment il fu De ce qu’avez tant atendu  (C1 [LR] 17115–18; [SR: L] 7039–42) [Lords, the branch leaves the great tale, God save me; now you will hear how it was in the matter you have awaited for so long]

There is a certain obscurity here in the relation between ‘branche’ and ‘grant conte’, which can be clarified by a consideration of two other narrative ensembles which use the metaphor of the branche to describe their textual division. In the Roman de Renart, each branch is a self-contained narrative, which need not rely on other branches for its diegetic coherence (although in practice several branches do refer to events in another branch).9 What connects the branches, forming the unifying stem or trunk of the diegetic universe, is the recurrence of characters, setting and type of narrative intrigue: the central character is always the fox Renart, and the narrative is made up of the tricks he plays on a revolving cast of other animals and peasants. In the Vulgate Cycle, each branche is the adventure of a different knight within the same diegesis. The unifying notion of common intrigue, motif and setting is present, but added to this is the idea that the branches taken together form disparate strands of the same evolving narrative. The text interlinks the adventures of various knights, leaving one story suspended while pursuing another, in a technique known by critics as interlace (entrelacement). This 9 A small number of Renart manuscripts (the gamma family) organise the material along cyclical lines, beginning with Renart’s birth and ending with his death, thus gesturing towards the overarching diegetic coherence which the other manuscripts refuse. See Sunderland (2008).

34 THE

CONTE DU GRAAL CYCLE

term was coined by Ferdinand Lot in a study of the Prose Lancelot; it is borrowed from descriptions of the geometric motifs popular in medieval manuscripts and architecture, which despite having no identifiable centre are complex and coherent entities.10 Similarly, he suggests, the unity of the Prose Lancelot is created by the interweaving of its various narratives. The movement between different plot threads allows comparisons and connections to be made between them, so that the return to one adventure can be informed by material from the one that has just been interrupted. Each branch therefore impacts on and advances the progress of others, yet can still be read as a coherent totality within itself.11 Cedric Pickford points out that in Arthurian prose romance the term ‘branche’ may be used to identify a subdivision of the text. This use is found occasionally in the Vulgate Cycle, where the Merlin, the Queste del Saint Graal, the Mort Artu and the Charrette material are all described as branches in various manuscripts.12 Beyond this, the term is also used to designate the tales of heroes not included in the text. At such moments, the text indicates a hierarchy of subjects, with ‘branche’ referring to a secondary narrative sprouting from the central trunk.13 Elspeth Kennedy draws attention to one such instance in the non-cyclic Prose Lancelot: of twenty knights who leave court to seek Lancelot, the narrator elects to follow Gauvain, adding: ‘Et neporquant chascuns de ces vint chevaliers a son conte tot antier, qui sont branches de monseignor Gauvain, car ce est li chiés et a cestui les covient an la fin toz ahurter, por ce que il issent de cestui’ [And nevertheless each of these twenty knights has his own entire tale, which are branches of that of Sir Gauvain, for this tale is the principal one and in the end all the others

See Lot (1918: 17). The self-sufficiency of each plot thread is illustrated by the way in which Malory reorders his Arthurian source material in the Morte Darthur, unravelling the interlaced adventures into a series of tales, with each one completed before the narrative moves on to the next. See Vinaver (1971: 128–9). 12 Certain manuscripts of the Prose Tristan also occasionally refer to subdivisions of the text as branches. By far the most systematic use of the term is in the Perlesvaus, which is divided into eleven branches, all introduced as such in the text. See Pickford (1960: 144–53). 13 In the case of Perlesvaus, however, Mireille Séguy (2001a: 357) argues that ‘la subdivision en branches obéit moins au besoin d’ordonner les différentes aventures les unes par rapport aux autres qu’à celui de ménager des pauses dans ce roman exceptionellement dense… L’ensemble du récit n’adopte donc pas la figure d’un arbre, malgré le terme qui désigne ses subdivisions internes, mais celle d’un ensemble composé de différentes tranches d’histoire qui s’ajoutent les unes aux autres’ [The subdivision into branches represents not so much a need to establish an order among the different adventures as a need to create breaks in this exceptionally dense romance… Overall, the narrative cannot be likened to a tree, despite the term used to describe its internal subdivisions, but rather to a whole composed of different layers of narrative added one on top of the other]. 10 11

NARRATIVE AESTHETIC AND CYCLIC FORMATION

35

must come back to it, since they come from it].14 At another point, other knights’ tales are described as branches of Lancelot’s, and his as a branch ‘del Greal, si qu’il i fu ajostez’ [of the Grail, having been added to it].15 Each tale has its place within the hierarchy as both source and tributary of others. A similar gesture appears to be being made in the passages cited above, with the term ‘grant conte’ projecting a greater authority than that of ‘branche’. Yet, awkwardly for such an assumption, the syntax of the introduction to Gauvain’s arrival at the Grail Castle actually implies the opposite: the ‘branche’ is leaving the ‘grant conte’, and it is the content of the ‘branche’ which the audience has ‘tant atendu’ [awaited for so long]. This reading is confirmed in the same manuscripts LMQU when, after relating Gauvain’s vow to return to the Grail Castle and succeed, the narrator announces: Les batailles qu’il acheva Ne les merveilles qu’il trova, Ne vos puis ore pas retrere; Por ce si m’en covient a tere. Au droit conte voil reperier  (C1 LR 17855–9; [SR: L] 7789–93)16 [I cannot tell you now about the battles that he won or the marvels that he found; therefore I must be silent about them. I want to return to the true tale]

The Grail material, represented here by Gauvain’s attempts to improve his reputation and earn a return to the Grail Castle, is presented as distinct from the ‘droit conte’.17 Indeed, the subsequent passage leaves us with a much clearer idea of what is not in the ‘conte’ than what is, since they consist in an enumeration of narratives which cannot (by choice or necessity?) be included. The narrator’s invoking of his obligation to stick to the script ironically serves to draw attention to the First Continuation’s aesthetic of narrative dispersal. Presenting us with a list of narrative stubs which could have been – but will not be – expanded, the text both opens a door onto a multiplicity of diegeses and ostentatiously closes it in the face of the audience’s piqued interest. The ‘droit conte’, then, is the path steered by the text between competing diegetic possibilities; in other words, it acts as a placeholder, marking narrative progression but only becoming narrative itself by actualising itself Prose Lancelot, 365.35–366.3. Prose Lancelot, 571.24–31. 16 The other Short Redaction manuscripts (ASP) and the Mixed Redaction of TV include this passage, but with making reference to a ‘droit conte’. Again, LMQU manifest a greater concern for defining the relation between the different narratives of the text. 17 Throughout this episode, Gauvain’s behaviour is clearly repeating that of Perceval in Chrétien’s section. As we will see below, Perceval also responds to Grail failure by throwing himself into unnarrated ‘batailles’ and ‘merveilles’. 14 15

36 THE

CONTE DU GRAAL CYCLE

into branches. By coming into being, each branch ‘se depart’ from this hollow stem, as the narrator notes in the case of Branch V. The resultant model of textuality is centrifugal, with each branch representing a diegetic whole not reliant on the outcome of earlier or later branches.18 Despite the significant difference between this narrative structure and that of the Vulgate Cycle, these passages of the First Continuation selfconsciously evoke a higher, inscrutable unity, in terms which can usefully be read in relation to interlace patterns. The action of ‘se departir’, as I have hinted, suggests the organic growth and multiple directions of plants; Eugene Vinaver has pointed out that foliage, especially the acanthus leaf, is also a key referent of the artistic patterns on which the interlace analogy draws.19 ‘Se departir’ is also an action associated with chivalric errance: the moment in Arthurian romance when two knights part company is often conveyed by a formula along the lines of ‘ensi se departent et s’en vont’ [in this manner they part ways and leave]; this parting of the ways is frequently exploited in interlace technique to suspend one narrative and pursue another. What is intriguing here is the fact that the vocabulary of ‘departir’, so redolent of interlace, should be used to describe a text which makes almost no use of the technique, and that three of the four manuscripts to do so are those which transmit the Long Redaction at this point. In these manuscripts, the narrator effectively offers his audience a way of thinking about the challenging centrifugal aesthetic of the First Continuation: as in interlace technique, where coherence is produced at a higher level by juxtaposition, the branch structure boasts a centre (the ‘droit/grant conte’) which is constituted by the bringing together of disparate narrative strands ‘en ordre... et a droit point / Einsi con li contes se joint’ [in order and correctly, just as the tale fits together] (C1 [LR] 18373–4; [SR: L] 8309–10). It is striking, then, that these passages acknowledge similar anxieties to those expressed by modern critics over the relation of the First Continuation’s branch structure to that of the rest of the corpus. The audience may be unable to perceive the connection between the narrative material and the ‘grant conte’ which lies veiled behind it, holding out the promise of satisfying audience expectations. But the narrator is defiant: he will pursue his narrative project ‘mot a mot’, ‘en ordre’. This passage reads as an explicit avowal of

18 There are noticeable points of contact between the First Continuation’s ‘branche’ structure and the figure of the rhizome discussed by Gilles Deleuze and Felix Guattari (1980: 9–37). Despite its use of arboreal vocabulary, the text appears more rhizomatic than arborescent: fragmented, freely making connections across diegetically unrelated branches, rather than proceeding by division to define a central core. The later Continuations, with their repeated use of the motif of the hero choosing a different path to his peers at a crossroads, can be conceived of as imposing an arborescent framework on the earlier rhizomatic material. 19 Vinaver (1971: 79–80).

NARRATIVE AESTHETIC AND CYCLIC FORMATION

37

the potential for disappointed expectations, after the narrator’s promise at the start of Branch V that the audience will finally get to hear ‘ce qu’avez tant atendu’ [what you have awaited for so long] (C1 [LR] 17118; [SR: L] 7042). This announcement, prefacing the first reintroduction of the Grail material for nearly twenty thousand lines (ten thousand in the Short Redaction), dangles the promise of continuity with the past before the audience, showing awareness of possible impatience with the text’s centrifugal aesthetic. The narrator imagines his audience as having been on tenterhooks since Perceval’s visit to his hermit uncle, wondering when the Grail material was going to make a reappearance. Yet the anticipated continuity is only partially realized: it is Gauvain who is now the Grail hero, and he also fails to bring an end to the Fisher King’s travails. The courtiers of the Grail Castle recognise immediately that resolution is not to be forthcoming, greeting Gauvain’s arrival with a regretful ‘ce n’est il mie’ [he is not the one] (C1 [LR] 17249). When read in the cyclical manuscripts, the ‘autre branche’ of Guerrehet turns out to be the prelude to a sea change in narrative aesthetic. Perceval’s reappearance after the conclusion to this episode, following his lengthy absence from the text, forms a striking contrast to the narrative centrifuge which precedes it. This moment is powerfully dealt with in manuscript E, where the narrator reminds his audience of the Welshman’s quest for ‘la cort o la lance qui saine’ [the court of the bleeding lance] (C2 [LR] 19611), before beginning: Or conmance de Perceval: Perceval, ce conte l’estoire, A si perdue la mimoire Que de Dieu ne li sovint mes. Cinc foiz passa avril et mais, Ce sont cinc anz trestot antier, Ainçois qu’il antrast an mostier, Ne Dieu ne sa croiz n’aora. Tot ainsint cinc anz demora, Et por ce ne laissa il mie Qu’il ne queïst chevalerie; Et les estranges avantures, Les felonesses et les dures, Ala querant, si an trova Tant que molt bien s’i esprova.  (C2 [LR] 19616–30) [Now begins the tale of Perceval. Perceval, the story tells, had completely lost his memory, and never remembered God; five springs had passed, that is five full years, since he had entered a church or worshipped God or his cross. He remained in this state for five years; but he did not give up seeking chivalric glory. And he went around seeking strange, dangerous and difficult adventures; he found many, and acquitted himself well of them.]

38 THE

CONTE DU GRAAL CYCLE

These fifteen lines are taken more or less word for word from the beginning of the Hermit episode in Chrétien’s section, precisely from the point at which Perceval is reintroduced after extensive narrative focus on Gauvain. In contrast to the decentred dynamic of the First Continuation, with its multiple heroes, this loop back to Chrétien’s text marks a clear desire on the interpolator’s part to announce this moment of the narrative as the direct prolongation of Perceval’s previous appearance, and to make it clear that this is to be the centripetal core around which the narrative will now be organised. And in case the point was lost on the audience, the narrator continues: De ce ne voil dire an avant, Car raconté l’ai ça devant; Et qui deus foiz le conteroit Trop grant annuiz vos sambleroit.  (C2 [LR] 19631–4) [I do not wish to say more of this, since I have recounted it earlier; and it would be very dull for you if someone was to tell it twice.]

Having jogged the audience’s memory and established the continuity of the narrative voice with that which recounted Perceval’s earlier adventures (‘raconté l’ai ça devant’), the text now prepares the way for the development ‘an avant’ of new material. The nature of the text chosen for recycling by the E redactor is in this respect highly significant. By returning to Perceval, and persisting with him as the central hero from hereon in, the corpus makes the alternation of heroes, and that between Perceval and Gauvain in particular, the principal device for structuring the text into parts. The passage in manuscript E makes the point especially forcefully by re-using material from one such moment of alternation in order to write another. Indeed, the second half of Chrétien’s text, in which the Hermit episode occurs, manifests what can be described as a proto-interlace structure which, as I will show, provides the impetus for the very different narrative dynamics of both First and Second Continuations. Over its first 4500 lines, Chrétien’s part of the corpus follows the development of Perceval from naive bumpkin to accomplished knight. However, just as he is being welcomed into Arthur’s admiring court, a moment of crisis occurs as he learns the dire consequences of his failure to ask who is served by the grail he has recently seen at the Fisher King’s castle. While other knights, Gauvain first among them, announce their intentions to take up various adventures, Perceval vows never to sleep in the same place two nights in a row until he has learned the answers to the questions he had failed to ask. Against all expectations, the narrative leaves him at this apparently crucial moment to follow Gauvain, before suddenly fast-forwarding five years to rejoin Perceval, without filling in the detail of what has happened to him up to this point. We are then returned to Gauvain at the point we left him, five years previously, without knowing whether we will hear more of Perceval’s

NARRATIVE AESTHETIC AND CYCLIC FORMATION

39

a­dventures before the Hermit episode or how many more of Gauvain’s adventures in the intervening years we will be told about. Here, the interlace technique has a quite specific effect which we can best identify by comparing it with the prose Lancelot. In this text, interlace ‘serves both to link and to suspend’,20 allowing various simultaneous adventures to be followed, and therefore to inform each other. The time frame within which the different actions take place is quite rigorously adhered to, and the movement between protagonists thus takes on a chaptering function. In the second half of the Conte du Graal, as we have seen, the switch from Gauvain back to Perceval opens up a five-year gap, which ensures that each narrative takes place in a different time frame. Consequently, there is no opportunity for the Perceval part to affect the Gauvain part, and vice versa. The effect of interruption is not so much to link or suspend as to disrupt the continuity of narrative. If the proto-interlace of the Conte du Graal functions to loosen the links of narrative structure, the Continuations respond to this precedent in strikingly different ways. The Second Continuation rejects Chrétien’s invitation to disruption. There are few time gaps in a narrative which takes place over a relatively short number of consecutive days. Changes in perspective take place only within the present tense of Perceval’s itinerary, chiefly when a defeated enemy is sent to Arthur’s court.21 These brief departures from the hero serve as markers delineating the end of one episode and the beginning of another, as we rejoin our protagonist further along his route. There is an extended section involving Gauvain’s exploits as he searches for Perceval (Episodes 28–32), but again these events are carefully inserted into the time frame of the Welshman’s wanderings, so that when we return to him we are told that he had travelled ‘bien pres de quinze dis’ [close to two weeks] (C2 31425) since we were last with him, which corresponds to the time taken up by Bagomedés’s ride to Arthur’s court and Gauvain’s subsequent adventures. This rigour is typical of a text in which practically each day and night of the hero’s wandering is scrupulously recorded, with a particular insistence on the need to find food and shelter for the night: if Perceval has to sleep in the forest, we are invariably told that he goes hungry (C2 22263–5, 23539–41, 24236–40, 26107–9). Chapter-structure is therefore given both by the alternation of heroes and by the succession of day and night in each protagonist’s journey; in both cases temporal continuity guarantees the coherence of the whole.22 Kennedy (1986: 201). For instance: the Sire du Cor in Episode 2 (MSS EPT only); Abrioris in Episode 10; the White Knight in Episode 13; the Biau Mauvais in Episode 17; Garsallas and Criseuz in Episode 21; Brun de la Lande in Episode 31. For episode references, see Appendix 1. 22 Rosemary Morris’s general observation on Arthurian verse romance – ‘It is the fact of journeying which characterises the knight, and the continuity of the journey which gives the narrative its coherence’ – perfectly sums up the unity and structure of the Second Continuation. See Morris (1988: 261). 20

21

40 THE

CONTE DU GRAAL CYCLE

The Gerbert Continuation slightly loosens this tightness, often establishing separation between episodes by having Perceval ride for a week uneventfully, but the language of interlace is still a regular presence. Perceval is again the dominant figure, starring in more than three quarters of the episodes. The Manessier Continuation maintains a similar temporal logic to the Second Continuation throughout the diegesis, though here the sequential ritual which Matilda Bruckner has called the topos of ‘Hospitality’ (the hero’s welcome, disarming, feeding and lodging at the end of a day’s journey) plays a far less important role in the narrative structure.23 An alternative chaptering principle is found in the motif of injury: the hero, having incurred wounds in vanquishing his foe, is forced to rest for a certain period of time. The enforced pause in the hero’s itinerary during this recovery period allows the narrative to explore the adventures of another protagonist. For instance, after telling us that Sagremor ‘Jut… bien sis semeines / Ainz que ses plaies fusent saines’ [was bedridden.. for six good weeks before his wounds were healed] (CM 35029–30), the narrator announces: Mais ci de monseignor Gauvain Un petitet vos conterons Et de Sagremor vos lerons Qui malades se jut adés  (CM 35036–9) [But here we will tell you a little of Sir Gauvain, and leave Sagremor who lies wounded in bed]

He follows this announcement with the reassurance: ‘Mais bien retornerons arier / Quant tens sera dou repairier’ [But we will go back to him without fail when it is time to do so] (CM 35045–6).24 Indeed, when that time comes the narrator reminds us of the state in which we abandoned him: Or est il droiz que l’en vos die De Sagremor le Desrée […] Toz fu gariz et respassez Des granz maux qu’il avoit assez  (CM 39970–1; 39975–6) [Now it is right that you are told of Sagremor the Impetuous… He was completely cured and recovered from the great pains which he had suffered]

Bruckner (1980). The technique is used to suspend the adventures of Perceval (Episodes 2 and 19) and Sagremor (Episode 5), as well as providing a model for the conclusion of Gauvain’s revenge mission against Keu on behalf of the murdered Silimac (Episode 10, with confirmation of Keu’s recovery in Episode 11). 23 24

NARRATIVE AESTHETIC AND CYCLIC FORMATION

41

The First Continuation, by contrast, embraces the centrifugal possibilities of Chrétien’s interlace technique. The insertion of time gaps is here carefully managed in order to guarantee the autonomy of individual tales. Thus, while the beginning of each branch is made to follow chronologically from the end of the last, the length of time over which they each take place leads to gross chronological contradictions. Between Branches II and IV lies the tale of Caradoc, which spans the entire conception, youth and first chivalric exploits of the hero, yet when Gauvain recounts in Branch IV his adventure with the Pucelle de Lis in Branch II, he dates it only ten years in the past (C1 [LR] 13611 [SR: L] 4395).25 A few lines later, when the child born of that coupling appears, the narrator notes that ‘il n’avoit / Que cinc anz au plus trestot droit’ [He was no more than five years of age at the most] (C1  [LR] 14773–4 [SR:  L] 5036).26 Similarly, when Arthur chastises his knights for forgetting that Giflet is trapped at Chastel Orguelleus, he reminds them: ‘Trois anz l’ont tenu am prison’ [They have held him in prison for three years] (C1 [LR] 12768);27 yet not only Caradoc’s lifetime but the seven-year siege of Branlant have intervened since Giflet left court to make his way there. The significant point here is that the First Continuation shows little concern for chronology beyond its capacity for separating episodes off from each other. The loose chronology of the Arthurian environment tends to turn it into a kind of static background on which the passage of years has no other effect than to create narrative space for new stories to be told, and establishing a centrifugal narrative project which in diegetic terms pays only minimal attention to the material inherited from Chrétien. In light of this, the return to Perceval which opens the Second Continuation according to Roach (but which Corley sees as the last part of the First) marks the point at which the corpus can be said to become cyclical. By establishing the primacy of the abandoned Perceval strand, the dispersed, anthological focus of the earlier material – the siege of Branlant, the story of Caradoc, the journey to rescue Giflet at Castel Orguelleus and the tale of Guerrehet – is transformed into a series of digressions. The duality of heroes of Chrétien’s original romance has been restored.28 But more than this, the treatment of the two heroes now sets up a clear hierarchy: Gauvain’s adventures are relegated 25 These kinds of numerical value often fluctuate across different manuscripts, and the number of years is variously given as ten (EMQTV), five (PS), four (A) and one (U). In all cases, the temporal incongruity with the span of the Caradoc narrative is flagrant. 26 MQ give the boy ‘dis anz’ [ten years]; all other manuscripts agree on an age of five. 27 Variants: ‘quatre ans’ [four years] (PU); ‘Bien a deus ans u pres de trois’ [for a good two, nearly three years] (L). AS stay vague here, saying only that Giflet ‘molt a esté en prison’ [has been in prison for a long time]. 28 Bruckner (1987: 253) also notes the importance of the Second Continuation in forcing a return to the plot inherited from the Conte du Graal section: ‘While the First Continuation, especially in the Short Redaction, moves away from Chrétien’s romance by imitating its own centrifugal textuality, once the Perceval Continuation is joined to it, we

42 THE

CONTE DU GRAAL CYCLE

to the status of secondary adventures, operating in the shadow of the main narrative concerning Perceval. Indeed, in a move that will be much used in the interlace technique of Arthurian prose romance, the narrative strand in which Gauvain’s adventures play out is itself set in motion by a desire on the part of Arthur’s knights to find Perceval. The Manessier and Gerbert Continuations will continue to focus on Perceval as main hero, with the adventures of other knights (Gauvain first and foremost) used as a counterpoint to the main strand, consecrating this re-centring of the cycle onto the itinerary of Perceval. At the same time, it threatens to transform the material found in the Short Redaction into a yawning gap, disrupting coherent communication between the early and the later adventures of Perceval. We are now in a position to suggest an answer to the question, posed earlier, of why the recasting of material into a longer redaction affects only part of the cyclic corpus – the rewriting of the First Continuation into its Long Redaction is intended partially to bridge this gap, smoothing the cyclical coherence of the whole corpus; if the Second Continuation is not considered in need of substantial recasting, this is because it is precisely this part of the corpus which sets the tone for the writers to follow. The cyclifying process is manifest in the Long Redaction’s obvious desire to strengthen the links (and hence the narrative coherence) between Chrétien’s part of the corpus and the action of the First Continuation. Thus, in the extended version of Branch I, Gauvain finds himself at the Grail Castle in an anticipation of his visit there in Branch V of all the Redactions. Annie Combes has shown convincingly how this interpolation attempts to mediate between Chrétien’s account of the Grail adventure and the very different account given in Branch V of the First Continuation, where the Fisher King is no longer an invalid but ‘un chevalier grant et menbru’ [a tall, strong-limbed knight] (C1 LR 17333) who comes striding into the hall, the Grail procession is replaced by a self-moving dish which magically replenishes the guests’ plates as it passes among them, and mending a broken sword belonging to an unidentified dead knight becomes an additional test of election for the hero who is to complete the adventure. By retaining the sword test (and Gauvain as the knight concerned) while reintroducing most of the details of Perceval’s evening at the castle as given in Chrétien, the interpolation forms a transition between the two contradictory passages; refusing a total conciliation which would have involved the excision or denaturing of one of the passages, the interpolator has preferred to attenuate their contradictions through addition.29 Bruckner terms this ‘the non-Aristotelian logic of and/both’, identifying it as can see the duality of Chrétien’s romance also asserting itself across the larger text of the Continuations.’ 29 Combes (2004). On the differences between the accounts of the Grail ritual given in Chrétien’s text and the First Continuation, see also Roach (1966); Séguy (2001a: 27–33); and Tether (2009).

NARRATIVE AESTHETIC AND CYCLIC FORMATION

43

a constitutive element in a corpus which embraces contradictions and oppositions in its accumulation of material, rather than seeking to resolve them: ‘nothing is ever lost but more can always be added, as new threads are interwoven with old ones in a pattern of repetition and contrast, according to the logic of and/both.’30 Most of the strengthening of links with the Conte du Graal occurs in the five episodes added to Branch I in the Long Redaction; the rewriting of the other branches (which may or may not be the work of the same redactor) consists of expanded accounts of the same events as in the Short Redaction, with the exception of Branch III, where a number of episodes are added to the tale of Caradoc, including a tournament scene spanning 1500 lines. The strengthening strategy begins with a clear statement of narrative intent. When Gauvain’s messenger reaches the court at Orcanie at the start of the First Continuation, the Short Redaction has him explain to Arthur that his nephew has conquered a castle and now requires the king’s presence for his duel with Guiromelant (C1 [SR: L] 43–66; [MR] 47–68). Tellingly, in the Long Redaction he recounts in more detail Gauvain’s adventures since he passed beyond the border of Galvoie (C1 [LR] 78–130); the narrator moves into indirect speech to emphasise the impossibility of constructing a narrative which would be able to contain them all: ‘Car il dist que Gauvains a fet / Tant proësces que il meïsmes / N’am porroit pas dire le disme’ [For he told them that Gauvain had accomplished so many feats that he himself would not be able to recount a tenth of them] (C1 [LR] 96–8). In order to add scenes where material from Chrétien could be revisited, the interpolator(s) found a motivation for Gauvain to ride off alone: the marriage between his foe Guiromelant and his sister Clarissant, which he happily accepts in the Short Redaction, is here celebrated behind his back and without his explicit assent, prompting him to leave court in disgust. Aside from the extra Grail Castle visit, Gauvain’s promise to return to the court of Escavalon if he fails to find the bleeding lance (CdG 6187–93) is revisited for the dénouement of the Branch, with the double combat against Guingambresil and the new figure of Dinasdarés replacing that with Guiromelant which concludes the Branch in the Short Redaction. In both cases, Arthur’s intervention imposes peace and leads into the scene of homage-paying from which Brun de Branlant’s notable absence motivates the siege of Branlant in Branch II. The Long Redaction thus models its conclusion on that of the Short, while adding an extra narrative arc drawing together Gauvain’s encounters with the Grail theme – indirectly, via his quest for the lance in Chrétien’s text, and directly, via his visit to the Grail Castle later in the First Continuation.

30

Bruckner (2009: 206).

44 THE

CONTE DU GRAAL CYCLE

The most thorough engagement with Chrétien’s part of the corpus comes in the expanded version of the Long Redaction given by EU, where the aforementioned episodes are interwoven with the revival of another Gauvain promise, to free the besieged damsel of Montesclaire (CdG 4718–20). He is first confronted by a dwarf who berates him for forgetting the promise, an intervention which can be read extra-diegetically as a critique of the insouciance shown by the Short Redaction to its narrative inheritance. (Indeed, the dwarf, who is described as ‘cort, boçu et desfet’ [short, hunchbacked and disfigured] (C1 [LR] 2517) as part of a lengthy passage of descriptio, bears striking resemblance to the Hideous Damsel who had originally brought news of the Montesclaire adventure.) Then, before he is able to make good on his promise, he is attacked by a knight in the castle where he has taken refuge for the night. Having defeated him, he discovers a further connection with the diegetic past: his opponent’s aggression is motivated by a promise made to his lady to fight any knight who arrives at the castle; in this somewhat inefficient way, she hopes one day to take revenge on a scoundrel named Greoreas, who had raped her several years before (C1 [LR] 3307–97). This incident sends us back to the Conte du Graal, where Greoreas’s mistreatment of Gauvain was motivated by his punishment at the latter’s hands for this very crime (CdG 7109–31). In this way, the narrative background to the incident is enriched, encompassing not only events from Chrétien’s text but also a common past to which both allude, and in which both inhere. After the visit to the Grail Castle discussed above, Gauvain finally arrives in Montesclaire, rescues the damsel and takes the Espee as Estranges Ranges which was promised to the conqueror of this adventure in the Chrétien section (CdG 4712), before rushing off to keep his appointment at the court of Escavalon. The treatment of narrative structure in this part of manuscripts EU anticipates that in the Second Continuation, where Perceval’s various quests (to find the Grail Castle, bring back a white stag’s head to the Chessboard Castle and reach Mont Dolereus) are made to interrupt each other, before eventually being resolved one after another. A number of shared motifs reinforce the comparison, suggesting that the redactor of these episodes was as keen on continuity forward with the Second Continuation as backward with Chrétien’s material. For instance, elements of the Montesclaire episode are clearly modelled on Perceval’s deliverance of Blancheflor and Beaurepaire in Chrétien’s text: the besieger is a rejected suitor and the castellan with whom Gauvain stays before arriving at the castle is the besieged damsel’s uncle, just as Gornemant was Blancheflor’s. Further, when evoking the dismay of the townspeople at the departure of he who ‘…de mort / Nous a recous… / …et d’essil’ [has rescued us from death and ruin] (C1 [LR] 4818–20), the redactor seems to be remembering the townspeople of Beaurepaire pleading with Perceval: ‘Sire, qui nos a trait d’escil…’ [Worthy sir, who has rescued us from ruin] (CdG 2946). At the same time, at least one detail of the episode appears to be taken from the Second Continuation. When Gauvain has retrieved the

NARRATIVE AESTHETIC AND CYCLIC FORMATION

45

Espee as Estranges Ranges from the pillar on which it hangs, the damsel explains that many knights have previously tried to open the crypt where it was kept, but had returned home ‘Ou hors de sens ou maumené’ [either driven mad or in a bad way] (C1 [LR] 4721). The same fate, being driven ‘hors dou sens’, befalls the knights who fail to tether their horse to the pillar on Mont Dolereus: this is the case for Keu and his comrades, as reported by Bagomedés (C2 28309–13), and also the ami of the unnamed damsel whom Perceval meets on his own way up the mountain (C2 31548–60). While it is always difficult in such cases to ascertain the direction in which influence operates, two related points can be made in support of my contention that the Gauvain episode is drawing on the Second Continuation material. First, as we have noted, this Gauvain material occurs in an interpolation contained only in EU, whereas all manuscripts of the Second Continuation mention this detail of the Mont Dolereus story (in all but ES, the exact text at 28310 is ‘fors de lor sens’); it is far more likely that the interpolation drew on the material common to all manuscripts than vice versa. Secondly, the Mont Dolereus material is one of the major narrative arcs of the Second Continuation, and the insanity which awaits those who fail its test motivates the subplot of Bagomedés’s ill treatment at the hands of the deranged Keu and his subsequent desire for revenge; in comparison, the use of the idea in the First Continuation interpolation is far more superficial, being unveiled only at the end of the episode in order to confirm the value of Gauvain’s success. Other evidence of motifs directly borrowed from the Second Continuation is admittedly not abundant, but it does exist, again in the version of the Long Redaction given by EU. Prior to the dwarf’s accusation, Gauvain returns an ivory horn stolen from a damsel riding a mule, killing her enemy in the process. In gratitude, she gives him a ring which will allow him to defeat his enemies, even when outnumbered, but which he must return to her when she asks for it (C1 [LR] 2470–97). This detail recalls the other magic ring in the narrative, that given by another damsel on a mule to Perceval in the Second Continuation to help him find the Grail Castle (C2 25978–26047). Here, again, the ring is to be returned, and indeed the damsel later reappears to reclaim it (C2 27528–31). In the episode from the First Continuation’s Long Redaction, the ring is never reclaimed; the inclusion of this narratively superfluous requirement appears to have no function other than to harmonise with the later episode in the Second Continuation. Beyond these specific instances of motif transfer, the most interesting trace of cyclification in the Long Redaction of the First Continuation is its use of interlace formulae and episodic narrative structure which, as we have seen, are prevalent in the later Continuations, but almost totally absent from the Short Redaction of the First. The episodic structure of the interpolated material in Branch I has already been noted, with the interruption and resumption of the different narrative strands bearing more resemblance to the Second Continuation than to the majority of the First. The use of time in these

46 THE

CONTE DU GRAAL CYCLE

episodes, as in several of the episodes interpolated into Branch III, is also in keeping with that in the Second Continuation. The alternation of day and night is carefully observed, and the question of shelter and sustenance given considerable attention. Thus, when she hears that Gauvain has spent the night fasting in the forest, the damsel who offers him the ring immediately invites him to share her meal, noting: ‘Si eüssiez or grant mestier / Q’antre vos et vostre destrier / Fussiez un petit aaisié’ [For you must have great need, you and your steed, to rest a little] (C1 [LR] 2129–31). A similarly pragmatic approach to these matters is in evidence in Branch III, when the hermit sheltering Caradoc thoughtfully procures richer food for the knight, aware that Caradoc is not used to a hermit’s meagre fare: ‘[Car] Caradoc ne poïst mie / Vivre de si tres povres vie / Com li bons hom l’avoit ampris, / Ne il ne l’avoit pas apris’ [For Caradoc could not live the frugal life that the good man had undertaken, since he had not learned how to do so] (C1 [LR] 10511–14).31 It is also in Branches I and III of the Long Redaction that we find examples of the interlace formulae missing in the Short Redaction. Branch I contains two such instances. First, at C1 [LR] 326–7, switching his interest from the queen’s reaction to Gauvain’s messenger’s news to that of Arthur and his knights, the narrator tells us: ‘Mais ne voil plus parler des trois, / Ainçois vos dirai de la cort’ [But I do not wish to say more of those three; instead I will tell you about the court], a formula missing from the equivalent passage in the Short Redaction; secondly, Gauvain’s ride away from the Gué Perilleux (the point at which the interpolated material of Branch I begins) is announced with the words: ‘De l’ost vos lairons ci atant, / Si vos conterons de Gauvain’ [Here we will cease speaking of the army, and will tell you about Gauvain] (C1 [LR] 1956–7). Interlace formulae are also used to mark the start of a new section several times in the long version of Branch III, again with no equivalent appearing in the short. Thus, to pick out just one example, the transition from Episode 9 to Episode 10 in the Long Redaction (equivalent to that between Episodes 3 and 10 in the Short Redaction) prompts the following comment from the narrator: N’an [=Guinier’s virtue] voil pas ci endroit conter, Mais ça avant an conterai Quant leus et tans an troverai. A mon conte voil revenir Et ma matiere maintenir D’illuec ou devant la laisai. Moi est avis que dit vos ai De ma dame Ysave la chose Et ce por qu’elle fut anclose A Nantes an la haute tor.  (C1 [LR] 9610–19) 31 On the significance of food as a theme in the different parts of the cycle, see SturmMaddox (1996).

NARRATIVE AESTHETIC AND CYCLIC FORMATION

47

[I do not wish to tell you here about Guinier’s virtue, but will tell you about it later, when I find the right time and place to do so. I want to return to my tale and take up my story from where I had left it. I believe that I have told you about my lady Ysave, and why she was imprisoned in the high tower in Nantes.]

The affirmation of authorial agency, the assurance of future development of the interrupted thread (and indeed Guinier’s virtue proves to be central to two incidents later in the narrative – Caradoc’s cure and the horn test), together with the reminder of events leading up to the new point of departure, are all hallmarks of the interlace technique deployed in the Second Continuation. The effect, here as in the later Continuations, is triple: first, to reinforce narrative continuity; secondly, to signal a break between two portions of narrative; thirdly, and most subtly, to exemplify the principle of continuation as an act of fidelity to the narrative past – the narrator ‘voil revenir’ to his ‘conte’, taking responsibility for the interlace moment. These changes made in the Long Redaction suggest a desire on the redactor’s part to take better account of the narrative aesthetic imposed by the Second Continuation, characterised by the partial adoption of a sequential, episodic structure, as well as a constant dipping back into the diegetic past in order to prompt further development. The moves made by successive continuators and redactors to control the centrifugal tendencies of the Short Redaction of the First Continuation bear witness to a developing consciousness of the material as a cyclical corpus with a unified aesthetic. Cyclic continuation, it seems, can bear a certain amount of digression and novelty, as is attested by the integral presence of the varied and diegetically distinct narratives of the First Continuation included in all manuscripts, but it demands a narrative centre against which the digression can be measured and through which the novelty can be integrated into a coherent whole. Of the eleven manuscripts which contain the Conte du Graal along with one or more continuation, eight include the first two and the Manessier Continuation. This arrangement (along with the occasional inclusion of Gerbert’s text, the Elucidation and the Bliocadran) makes plain at once the coherence and the popularity of the cyclic ensemble over other possible permutations of the material. The Hideous Face of Adventure In describing the coming into being of the Conte du Graal cycle, we have stressed the developments between successive states of the corpus. I now propose to read the narrative synchronically, as it appears in the various manuscripts. I leave consideration of the material condition and specific mise en manuscrit of each version of the text until Chapter 2. My aim here is to examine one specific set of mechanisms used by all versions of the corpus to generate cyclic structure, namely the device of the quest and its atten-

48 THE

CONTE DU GRAAL CYCLE

dant shaping of the heroes’ actions. The accumulation of overlapping quests, I will show, provides the narrative impetus for textual accretion. Hanging plot threads create a repository of interrupted narratives to be reactivated by future writers. In such a structure, the audience can never be certain whether the promise of resolution implied by each of these threads will be kept or not; some narratives are elucidated and resolved while others are never revisited. For instance, Gauvain’s visit to the Grail Castle in Branch V of the First Continuation comes about after an unknown knight is mysteriously murdered, having made Arthur’s nephew promise to take up his journey in such an event; the name of the knight is eventually revealed to be Silimac when his sister appears in the Manessier Continuation. By contrast, we never learn the identities of either the young man murdered before Perceval’s eyes in the Second Continuation or the knight who kills him, going to his grave in mute defiance of our hero’s attempts to discover the reason for his crime. The potential for quests to re-open doors into the narrative past which had apparently closed makes it a natural device to map onto interlace structure; in this corpus, certainly, the impetus for interrupting one knight’s narrative to return to another’s is unfailingly provided by the need to check up on the returning character’s progress towards a given goal. It is therefore appropriate that the first major switch in central character – from Perceval to Gauvain – is also the first point at which explicit oaths to quest are sworn. Indeed, it becomes a repository of potential quest material, as the Hideous Damsel dangles the bait of adventure before the assembled knights of Arthur’s court: ‘Qui velt faire chevalerie, / S’il la le quiert, n’i faldra mie’ [He who wishes to perform chivalric deeds will not fail to find the opportunity for them there] (CdG 4699–700). The quest is, of course, one of the principal means for knights to define and enhance their identities. In the Second Continuation, it is repeatedly emphasised that the achievement of the Mont Dolereus adventure is reserved for the best knight in the world, and the temporary madness inflicted on those who fail the adventure represents punishment for their presumption to a status which they do not deserve.32 The treatment of this adventure in the Second Continuation serves to ratify the chivalric superiority of Perceval over all other knights, and thus makes other chivalric tests somewhat redundant. More precisely, the other chivalric adventures and tournaments which Perceval encounters become so many different vehicles for making the same point about his superior chivalry. By contrast, the Hideous Damsel’s performance in the Conte du Graal section suggests a world in which every knight can find his own allotted adventure. Certainly, her appearance prompts a whirlwind of oathswearing among Arthur’s knights: no sooner has she finished her speech and left than Gauvain leaps up to claim the Montesclaire quest as his own. This

32

See C2 [LR] 19916–22, C2 [SR] 9818–29, C2 28309–21 and C2 31614–22.

NARRATIVE AESTHETIC AND CYCLIC FORMATION

49

is followed by Giflet’s decision to go to Chastel Orguelleus and Kaherdin’s to go to Mont Dolereus, before Perceval makes his own vow to seek the truth about the grail and lance, discussed later in this chapter. What is interesting about Kaherdin’s words is that, unlike the other knights who speak here, he is not responding directly to anything said by the Hideous Damsel, who has made no mention of Mont Dolereus.33 Rather, it appears that her words have stirred in him a spontaneous desire for adventure, and after Perceval’s oath the mood seems to catch the whole assembly: Et bien ensi jusqu’a cinquante En sont levé, et si creante Li .i. a l’autre et dist et jure Que merveille ne aventure Ne savront qu’il ne l’aillent querre, Tant soit an felenesse terre. Et que que il s’apareilloient Parmi la sale et il s’armoient…  (CdG 4741–7) [And in the same manner up to fifty knights stood up, and swore to each other that whenever they heard of a marvel or an adventure they would travel to find it, however dangerous the area in which it was situated. And as they were getting ready and arming themselves in the hall…]

At this point, just as the knights are arming themselves for their quests, Guingambresil bursts in to accuse Gauvain of treasonous murder. Gauvain’s own oath to free the damsel of Montesclaire is laid aside pending the result of his journey to clear his name. Likewise, the text abandons the questing oaths just made by all of his companions in order to follow him on his journey to Escavalon; surprisingly, even Perceval’s intention to learn the truth about the grail and lance is forgotten. As we saw above, the effect of this move is to disrupt the narrative continuity, promoting an aesthetic of postponement. Yet abandoned quests are always open to resumption; the Hideous Damsel episode may gesture towards narrative’s ability to frustrate audience expectations, but the Continuations rewrite it retrospectively as a reservoir of source material. Each of the individual quests announced here receives elaboration later in the corpus. As mentioned above, the Mont Dolereus narrative becomes one of the main plot strands of the Second Continuation, despite (or perhaps because of) the lack of detail provided by Kaherdin when he mentions it in the Conte du Graal: ‘Et je sor le Mont Dolerous,’ Fait Keendins, ‘monter irai, Ne dusques la ne finerai’  (CdG 4724–6) 33

The name is given as ‘Mont Perilleus’ [Perilous Mountain] in manuscripts AHS.

50 THE

CONTE DU GRAAL CYCLE

[‘And I,’ said Kaherdin, ‘will go to the Dolorous Mountain, and will not rest until I reach it’]

With no further indication of what this adventure entails, the Second Continuation fills in the content, imagining a pillar to which only the world’s best knight can tether his horse. The minor Arthurian character Kaherdin drops out of the picture, to be replaced by Perceval, the only candidate to the title of ‘best knight in the world’ countenanced by this Continuation. The confirmation of Perceval’s knightly pre-eminence through the Mont Dolereus quest is a gradual process. The first mention of the quest comes in the first episode following the reintroduction of Perceval, after he defeats the Sire dou Cor, lord of the Castle of the Ivory Horn. When the Mont Dolereus is mentioned at the castle later that day, it is presented as an adventure for any knight to try; in the Short Redaction, use of the third person plural emphasises the multiple identities of those to whom the challenge is proposed: Icel jor oïrent noncier Qu’el haut pui del Mont Dolereus Avoit un pilier merveilleus  (C2 [SR] 9818–20) [That day they heard tell that on the high peak of the Dolorous Mountain there was a marvellous pillar]

Perceval immediately claims this quest for himself in terms which echo those used of it by Kaherdin in the Hideous Damsel episode: ‘Certes, jamés ne finerai / Devant que la venuz serai’ [In faith, I will not rest until I have arrived there] (C2 [LR] 19925–6). In doing so, he intends to determine ‘Se seroie boens chevalier’ [whether I will be a good knight] (19929) – the conditional tense indicating the potential at this stage for a negative answer to the question. It is important to bear in mind that this passage comes at the end of the first episode involving Perceval for over twenty-two thousand lines (twelve thousand in the Short Redaction); at this stage, Perceval has had fewer lines of narrative than Gauvain (or Caradoc, for that matter) and his status as central character of the text is not secure. By the next time that the Mont Dolereus is mentioned, this centrality is no longer in question, as the narrative structure of the Second Continuation has reduced other characters’ appearances to digressions. In so doing, it revives a prophecy made during Perceval’s first visit to Arthur’s court in the Conte du Graal; as he leaves the court, a maiden laughs and tells him: Vallet, si tu vis par eage, Je pens et croi en mon corage Qu’en trestot le monde n’avra, N’il n’i ert n’en ne l’i savra, Nul meillor chevalier de toi  (CdG 1039–43)

NARRATIVE AESTHETIC AND CYCLIC FORMATION

51

[Young man, if you live long enough, I think and believe in my heart that there will never be, nor will anyone hear of, any knight greater than you]

The first half of Chrétien’s text appears to bear out the narrative programme set out here: Perceval takes to chivalry like a second skin, as a series of defeated opponents are sent to surrender at Arthur’s court. The king, perceiving the utility of this knight’s service, sets out to find him with an army in tow; a chance encounter ensues almost immediately and the hero, having demonstrated his prowess on the bodies of Keu and Sagremor (unhorsed in violent confrontation), is warmly welcomed into the king’s entourage. The queen describes him as ‘chevaliers esprovez / De haute proëce et de bele’ [a proven knight of high and worthy prowess] (CdG 4594–5), and Perceval seems well on the way to fulfilling the destiny predicted for him at the start of his adventures. It is at this point that the Hideous Damsel arrives at court and the narrative programme centred on Perceval’s gradual elevation towards chivalric perfection is derailed. The Damsel’s harangue overwrites the prophecy of the laughing maiden with a new, negative character assessment of the one she calls ‘li maleüreus’ [the wretched one] (CdG 4665). The only future she describes is one of suffering for the inhabitants of the Fisher King’s lands, and the operative developmental notion in relation to Perceval is that of opportunity missed: ‘tu ne la retenis mie, / Fortune, quant tu l’encontras’ [You did not hold on to Fortune when you met her] (CdG 4650–1). As the audience struggles to come to terms with the implications of this condemnation, Guingambresil bursts onto the scene to accuse Gauvain of treason, and Perceval is cast aside by the narrator until further notice, to be revisited only briefly nearly two thousand lines later. In rehabilitating Perceval’s trajectory towards chivalric excellence, the Second Continuation chooses to read the first half of Chrétien’s romance as the programmatic motor of the text and sets about fashioning a narrative that will repeatedly perform the necessary identification of Perceval as first among knights. Prophecy hardens into definite confirmation in an episode which sees the hero arrive at an unfinished bridge which can only be passed by the worthiest knight in the world (C2 26772–82). The castellan Briol brings Perceval here, explaining that a return to the Fisher King’s court can only be effected if Perceval can prove his mastery of chivalric values (26238–40); the bridge, created by a lady expert in the magical arts (26714–15), provides an appropriate means for him to do so.34 We glimpse Perceval’s safe arrival on the other side through Briol’s eyes:

34 Séguy (2001a: 289–90) points out that the unfinished bridge, which ‘failloit’ [ended] half-way across the river (C2 26534), feeds into the thematics of incompletion operating through the cycle.

52 THE

CONTE DU GRAAL CYCLE

Briol voit bien et aparçoit Que c’est li miaudres chevalier, Li plus hardi et li plus fiers Qui soit ou mont  (C2 26808–11) [Briol can see and tell that this is the best, the bravest and most valiant knight in the world]

Briol, in a moment of obvious symbolism, is left behind on the other bank as Perceval rides on to a tournament – at Chastel Orguelleus. It is no coincidence that Perceval’s forward movement, both literally and in terms of his developing identity as the pre-eminent knight of the Arthurian world, brings him to another location drawn from the treasure trove of the Hideous Damsel episode. Having outshone all others at the tournament, Perceval’s next adventure ends with a robber knight telling him: je sai bien trestot de voir Que vos iestes li plus proisiez, Miauz apris et miauz anseigniez, Et miaudres de chevalerie Que chevalier qui soit an vie. Et se vos volez seurmonter Tout le mont d’armes et oustrer, Si tenez cel chemin erbous Au grant pui dou Mont Dolerous.  (C2 27482–90) [I well know that you are the worthiest, most skilled and most well-educated knight, and best in chivalry, alive today. And if you wish to outdo all others in the world in the field of arms, keep to this grassy path all the way to the great peak of the Dolorous Mountain.]

This time the quest is explicitly proposed to Perceval in his capacity as best knight in the world. When he later meets Bagomedés, Perceval reiterates his desire to go to Mont Dolereus ‘Por savoir et por esprover / Se ja boens chevaliers sera’ [to know and to find out whether I will be a good knight] (C2 28352–3), demonstrating his recognition of the symbolic capital associated with the adventure. When the interlaced narrative then leaves Perceval in order to discuss other knights, he is making his way ‘Tot droit vers le Mont Dolerox’ [directly to the Dolorous Mountain] (C2 28395). The suspension of this narrative is, for once, not compounded by further obstacles; when we return to Perceval we learn that, over two weeks of travel, he has not encountered a single adventure worthy of narration (31428–9). Perceval’s own adventures, which have until the meeting with Bagomedés been interlaced around each other, are now sorted into sequential order; in particular, the Mont Dolereus adven-

NARRATIVE AESTHETIC AND CYCLIC FORMATION

53

ture now becomes the means through which the Grail Castle can be reached, rather than the two quests running in competition with each other. This is confirmed by the child, sitting on the branch of a tree, who is Perceval’s final interlocutor before he reaches the mountain; refusing to answer our hero’s questions about the Fisher King, his parting words are instead: Mais une chose sai ge bien, Que vos porroiz demain aler Au Mont Doleroux au piller, O vos orroiz, ce cuit, novelles Qui vos seront plaisanz et belles.  (C2 31484–8) [But I well know one thing, that tomorrow you will be able to go to the pillar on the Dolorous Mountain, where I believe you will hear pleasing news.]

This is the most explicit statement yet of the predestined nature of Perceval’s arrival at Mont Dolereus: there is news waiting for him there. Sure enough, the damsel he meets after successfully tethering his horse to the pillar announces: Or si vos poëz bien vanter Que vos plus grant honor avez Que chevalier de mere nez Eüst onques jor de sa vie  (C2 31698–701) [Now you can boast that you have achieved a greater honour than any knight born of a mother ever had in his life]

before showing him the way to the Fisher King’s court. The Second Continuation’s primary narrative project for Perceval thus appears to be his confirmation as pre-eminent in chivalry, the top knight of the Arthurian world. In an act of mutual confirmation, Perceval’s gradual appropriation of the Mont Dolereus quest validates its prophecies while simultaneously endorsing his status as cyclical hero. We examined above the Montesclaire adventure that appears in the interpolation to the Long Redaction found in EU; its development of Chrétien’s narrative stub into a fully fledged quest drew on an impressive number of episodes from the Conte du Graal, while the punishment of madness meted out to those who failed on the quest was taken from the Second Continuation’s Mont Dolereus material. The EU interpolation was apparently not known by Gerbert de Montreuil, nor by the scribes who transmitted his continuation in manuscripts TV, as this text contains an entirely different elaboration of the material from the Hideous Damsel episode. The conclusion of the Tristan segment of the text sees the Cornish hero reconciled to Marc’s court and Arthur’s men preparing to depart. Perceval announces that

54 THE

CONTE DU GRAAL CYCLE

he will be heading off on his own to seek the Grail, and this appears to trigger Gauvain’s memory of the respective oaths they swore at Arthur’s court, nearly thirty-eight thousand lines previously: Mesire Gavains ot et voit Que Perchevaus aler s’en velt Le Graal querre: or le conselt Li vrais Dieus, qui bien le puet faire! ‘Et je al Pui de Montesclaire Irai’, fait mesire Gavains.  (CG 4798–803) [Sir Gauvain hears and sees that Perceval wants to go and seek the Grail; may he be aided by the Good Lord, in whose power it is to do so! ‘And I will go to the Mountain of Montesclaire’, says Sir Gauvain.]

This reintroduction of the Montesclaire adventure is far more low key than that found in EU; indeed, when we later hear of Arthur’s sadness for his nephew ‘Qui a fait et promesse et veu / D’aler al Pui de Montesclaire’ [who has promised and vowed to go to the Mountain of Montesclaire] (CG 4854– 5), it is not clear whether the vow referred to is that made in the Hideous Damsel episode or the one made just a few lines previously. Gerbert seems less interested than the interpolator of the EU version in criticising Gauvain’s previous failure to honour his vow. Indeed, despite Gauvain’s renewed oath here, the Montesclaire narrative is presented as it was by the Hideous Damsel, as a quest open to any knight. In the end it is Perceval whose adventures lead him to Montesclaire, where he frees the besieged damsel from her unwanted suitor. When Gauvain later reaffirms his intention to honour his vow, he is told that the damsel has already been freed, though the identity of her saviour is unknown (CG 12252–60). Since this information is hearsay, he persists in his quest until he meets an eyewitness of the event and learns that it is Perceval who came to the damsel’s rescue (13976–84). Gauvain’s reaction on learning that his quest has been completed by his fellow knight is one of joy, even relief, since Perceval’s success has freed him from his vow: Quant me sire Gavains oï La pucele, molt s’esjoï Car bien entent et set et voit Que Perchevaus li preus avoit Assomé l’aventure estraigne Por coi il parti de Bretaigne, Que voé ot qu’il i venroit.  (CG 13985–91) [When Sir Gauvain hears what the damsel says, he is overjoyed, for he understands, knows and sees that worthy Perceval had completed the strange adventure on account of which he had left Britain, having vowed that he would arrive there.]

NARRATIVE AESTHETIC AND CYCLIC FORMATION

55

At this point Gauvain’s narrative ends, and the text returns to Perceval for the remainder of Gerbert’s continuation. This part of the text therefore allots centre stage to Gauvain only insofar as he is questing for Montesclaire; when he learns that this ‘strange adventure’ (CG 13989) has been completed, he is free to disappear back into the collective stasis of the Arthurian court. The function of his own quest is ultimately to underline Perceval’s superiority, bearing out a prophecy made in Chrétien’s part of the text: the Hideous Damsel’s original words advertised the Montesclaire adventure as that which would garner its conqueror ‘le pris… de tot le mont’ [the glory of all the world] (CdG 4701–2). For Gerbert, working with the Second Continuation’s reconfiguration of the narrative programme, the elect knight can only be Perceval; whereas the Long Redaction of the First Continuation could countenance Gauvain succeeding in such qualifying adventures, and even reaching the Grail Castle, Gerbert’s text requires the quest to be transferred from Arthur’s nephew to Perceval before it can be brought to a conclusion. Like the Montesclaire episode, the Hideous Damsel’s reference to Chastel Orguelleus finds its initial development in the First Continuation, though this time in all versions of the text. Giflet’s vow to test himself there, laid aside by Chrétien in favour of the binary Gauvain–Perceval structure, is accommodated more easily by the First Continuation’s centrifugal aesthetic. As with the Montesclaire episode of EU, the narrative thread is reactivated by an accusation of dereliction of duty: this time it is Arthur who berates his assembled knights for having forgotten Giflet, who has languished in prison at Chastel Orguelleus for several years. The quest to reach Chastel Orguelleus and free Giflet then becomes the narrative motor for the material which critics refer to as Branch IV. When Arthur’s rescue party eventually reaches the castle, the knights are able to verify at first hand the Hideous Damsel’s assertion that ‘la ne faut nus qui i aille / Qu’il ne truisse joste ou bataille’ [he who goes there will not fail to find a joust or a battle] (CdG 4697–8); the custom of jousting allows several Arthurian knights to take centre stage, with Lucan, Bran de Lis, Keu, Yvain and eventually Gauvain each taking their turn to duel. In its succession of champions, then, the jousting custom at Chastel Orguelleus offers a précis of the First Continuation’s centrifugal heroic structure. The Gerbert Continuation offers a similar scenario to that which motivates Branch IV, but whose contrasting development says much about the differing narrative structures of the First Continuation and those which follow it. Gauvain’s renewed oath to go to Montesclaire, made at the end of the Tristan material, leads him to move temporarily out of the narrative, which elects to follow Perceval. A rationale for his disappearance from the text is later given: he has been imprisoned, as we learn when Perceval frees him from his captors. Later, when Gauvain has learned of Perceval’s success at Montesclaire and renounced his quest, he encounters Arthur’s court, on its way to free him from his imprisonment. Gauvain and Arthur

56 THE

CONTE DU GRAAL CYCLE

are here in analogous positions: both have set out on quests which have been rendered unnecessary by Perceval’s chivalric prowess. In the Gerbert Continuation, as in the Second, quests initially associated with other characters are either reassigned to Perceval or made redundant by him. Thus, while the Chastel Orguelleus branch of the First Continuation exemplifies an aesthetic within which adventure is the business of the whole court, things are quite different when the setting is revisited for a tournament scene in the Second Continuation. An interpolation in manuscript T, keen to promote the coherence of the whole corpus, reminds the audience that this is the place where Giflet spent a long time imprisoned, but there is no disguising the fact that one knight now supersedes all the others.35 Perceval, having left Briol on the opposite bank of the river, dominates both days of the tournament and rides off to follow his destiny as best knight in the world. Thus, of the three quest destinations mentioned by the Hideous Damsel, Perceval finds success at Mont Dolereus, Chastel Orguelleus and (in the manuscripts which include the Gerbert Continuation) Montesclaire. The narrative reservoir of this moment in the cycle is ultimately exhausted by the tyrannical appetite of its biographical core.

This Quest Which Is Not One: Perceval and the Grail Perceval’s own oath and the Hideous Damsel’s harangue which precedes it have been the subject of much close analysis. It is a widespread assumption of many such discussions that Perceval’s oath to seek the truth about the grail and lance will eventually lead him back to the Grail Castle and redemption of his earlier mistake.36 Yet this notion is at odds with the Damsel’s actual words, which emphasise Perceval’s missed opportunity without suggesting that a second chance is achievable: Molt est maleürous qui voit Si bel tans que plus ne coviegne, S’atent encor que plus biax viegne. Che iéz tu, li maleüreus, Qui veïs qu’i[l] fu tans et leus De parler et si te teüs [...] Et ses tu qu’il en avendra See Roach (1971 [IV], Appendix VII, vv. 134–8). See for example Roach (1956: 113): ‘Il ne restait donc que quatre choses à continuer dans tout ce poème de Chrétien: Perceval devait retrouver le Roi Pêcheur et le guérir en posant la question sur le Graal…’ [There were thus only four things in all of Chrétien’s poem left to continue: Perceval was to find the Fisher King once more and heal him by asking the question about the Grail…]. 35 36

NARRATIVE AESTHETIC AND CYCLIC FORMATION

57

Del roi qui terre ne tendra Ne n’iert de ses plaies garis?  (CdG 4662–7; 4675–7) [Wretched indeed is he who sees better weather than he is entitled to and waits for better weather still. Thus are you, the wretched one, who saw that it was the right time and place to speak, yet kept quiet… and do you know what will come of it, of the king who will not be able to rule his land, nor be healed of his wounds?]

The ensuing catalogue of woes to come, cast in the future tense, suggests an irrevocable destiny rather than a hypothetical outcome; the accent is very much on the idea that Perceval’s time has come and gone. While Frappier is undoubtedly right to warn against taking the Damsel’s words ‘pour argent comptant’ [at face value] (1972: 145),37 a further complication to the notion of a redemptive Perceval questing for the grail occurs when we consider his response to her harangue: Et Perchevax redist tout el: Qu’il ne gerra en .i. hostel .II. nuis en trestot son eage, Ne n’orra d’estrange passage Noveles que passer n’i aille, Ne de chevalier qui miex vaille Qu’autres chevaliers ne que dui Qu’il ne s’aille combatre a lui, Tant que il del graal savra Cui l’en en sert, et qu’il avra La lance qui saine trovee Et que la veritez provee Li ert dite por qu’ele saine; Ja nel laira por nule paine.  (CdG 4727–40) [But Perceval said otherwise: that for the rest of his life he would not spend two nights in the same lodgings; he would not fail to investigate any news he should hear of some faraway event; nor, if he heard tell of a knight of greater valour than the others (even should he be worth two), would he fail to measure himself against him: all of this until he could learn whom the grail served, and until he could find the bleeding lance, and learn the truth of why it bled. He would not, he said, abandon this undertaking, whatever it might cost him.]

Is Perceval actually announcing a quest and, if so, what is its nature? Frappier’s gloss on this passage illustrates the selective emphasis which critics have tended to bring to it: 37 See also Brigitte Cazelles (1996: 35), who warns that ‘[the Damsel’s] interpretation of Perceval’s behaviour as a failure to act is neither necessarily correct nor impartial’.

58 THE

CONTE DU GRAAL CYCLE

cinquante autres chevaliers échangent des serments, jurent de se jeter dans les aventures les plus périlleuses. Seul Perceval déclare qu’il ne couchera pas deux nuits de suite au même endroit et qu’il ne prendra nul repos tant qu’il n’aura pas pénétré le mystère du graal et de la lance qui saigne.38 [Fifty other knights make vows to each other, swearing to throw themselves into the most perilous adventures. Only Perceval declares that he will not sleep two nights in the same location, nor rest, until he has elucidated the mystery of the grail and bleeding lance.]

Undeniably, Perceval’s stated desire to find the bleeding lance implies an intention to return to the Grail Castle. Frappier, along with many others, has drawn attention to the importance of the first line of this passage: the fact that Perceval ‘redist tout el’ appears to mark out a destiny set apart from that of Arthur’s knights. In his paraphrase, Frappier contrasts these knights’ vainglorious appetite for dangerous adventure with Perceval, who alone sets out after the mysteries of the grail and lance. But the implication that he is reacting against the worldly concerns of Arthur’s knights misrepresents the order of events in the episode. In fact, Perceval’s own oath follows, and is thus to be understood as opposed to, those made by Gauvain, Giflet and Kaherdin; it is itself followed by the vows of the fifty knights. The words ‘et bien ensi’ [and in the same manner] (CdG 4741) which introduce these vows suggest a continuity between Perceval’s oath and those made by this group of fifty-odd knights, rather than a contrast.39 When we examine the part of Perceval’s oath which Frappier’s paraphrase omits, this impression is confirmed. For Perceval does not say that he will seek the Grail Castle; indeed, he does not use the word ‘querre’ at all. As Keith Busby (1993a: 49) suggests, ‘he seems to know that the Grail Castle cannot be sought as such’. What Perceval instead proposes is (1) not to spend two consecutive nights in the same dwelling, (2) to test himself in any strange land he hears of, and (3) against any worthy adversary he hears of, until he learns the secrets of the grail and lance. The force of the ‘tant que’ at v. 4735 is ambiguous: it could indicate a causal link (a belief that chivalric deeds will make him worthy of finding the Grail Castle again) or merely an acknowledgment of the time which may pass before such a return might occur (in which case his vow of chivalry is to be understood as a decision to keep busy in the meantime). In either case, parts (2) and (3) of his vow as outlined above are manifestly similar to those of his fellow knights, who swear ‘Que merveille ne aventure / Ne savront qu’il ne l’aillent querre, / Tant soit en felenesse terre’ [that whenever they heard of a marvel

Frappier (1972: 143). The ‘tout el’ might then be understood as signalling Perceval’s rejection of the Hideous Damsel’s presentation of his failure as irrevocable, whereas the three knights who speak before him all respond positively to her speech. 38 39

NARRATIVE AESTHETIC AND CYCLIC FORMATION

59

or an adventure they would travel to find it, however dangerous the area in which it was situated] (CdG 4744–6). When we next meet Perceval, in the Hermit episode, his actions have been perfectly in keeping with his vow: Tot einsi .v. ans demora, Ne por che ne laissa il mie A requerre chevalerie; Et les estranges aventures, Les felenesses et les dures, Aloit querant, et s’en trova Tant que molt bien s’i esprova  (CdG 6224–30) [He continued in this manner for five years, and never stopped seeking chivalric glory; he went around looking for strange, dangerous and difficult adventures, and found many, so that he proved his valour well]

The vocabulary here – ‘requerre’, ‘estranges’, ‘aventures’, ‘felenesses’, ‘querant’ – places this behaviour squarely within the remit of both Perceval’s oath and those of his fifty fellow knights. Yet maintaining his oath has led Perceval into what is presented as a greater sin: forgetting God. Accordingly, in the subsequent conversation with his hermit uncle, Perceval’s desire to learn what he failed to ask has been replaced as a primary concern by the fresh realisation that his shame has caused him to lose his way: Et del graal que je i vi Ne soi pas cui on en servi, S’en ai puis eü si grant doel Que mors eüsse esté mon wel, Et Damedieu en oblïai, Ne puis merchi ne li crïai Ne ne fis rien, que je seüsse, Por coi jamais merchi eüsse  (CdG 6379–86) [And I do not know who was served by the grail that I saw; since then I have been so tormented by this that I would rather have died. And because of this I forgot God; I never asked him for forgiveness, nor did anything, as far as I know, by which I might have had his forgiveness]

The hermit is able to explain to Perceval that his failure at the Grail Castle was linked to his abandonment of his mother, but the terms of reference for the hero’s personal advancement have shifted. There is no advice on, nor even evocation of, the possibility of a future return to the Fisher King’s court. No mention is made of the unfortunate subjects of the lands around the Grail Castle, nor of any potential deliverance they might hope for from the

60 THE

CONTE DU GRAAL CYCLE

destinies which the Hideous Damsel had outlined for them. We have come a long way from the narrative of knightly education set in motion by the maiden’s prophecy at the start of the text, a prophecy that no longer appears to hold any proleptic force in the wake of the Hideous Damsel’s searing condemnation and the narrative’s appropriation of Gauvain as an alternative hero. Instead, the hermit’s teaching redirects Perceval’s focus onto his own salvation, enjoining him to atone for his sin against his mother and to come back into contact with God. Perceval also learns that the grail serves a second invalid king, father of the first, who subsists on nothing but the communion wafer which is brought to him in it every day. Perceval has now acquired understanding of his earlier failure, been forgiven for the sin which caused it, and learned half of the information he set out from Arthur’s court to uncover. The narrative leaves him at his uncle’s hermitage, in a movement of interlace which functions like the Hideous Damsel episode to confound any expectations of narrative progression. It is hard to detect any sense of urgency in the possible development of the Perceval narrative from here, and the question is raised of what would be achieved by a return to the Grail Castle. Perceval has not yet learned the secret of the bleeding lance (which immediately prior to this episode has become associated with Gauvain, who has sworn to seek it), but there is little suggestion that much would be added to his reputation or to his personal development by such revelations. Perceval’s quest, if it ever really existed, no longer has a viable object. Perhaps this interpretation of these key scenes from Chrétien’s part of the text helps to explain why the First Continuation sees no point in returning to Perceval’s adventures. Five years of chivalry (and sixty defeated knights sent to Arthur’s court) have been dispatched in under twenty lines of text, as if Perceval’s loss of memoire [memory] in turning his back on God had infected the ability of the estoire [story] to record his adventures (CdG 6217–19).40 Faced with the apparent satisfaction of Perceval’s desire to learn about the grail, the Grail Castle material in the First Continuation explores alternative characters and objects as means of returning to the Fisher King’s court. In the Grail Castle visit of Branch V, common to all manuscripts of the cycle, Gauvain is tried, and rejected, as the Grail knight, becoming instead a new Perceval in his failure to fulfil the required criteria. He does, however, learn that the bleeding lance was that with which the crucified Christ was struck (C1 [LR] 17513–18).41 The mysteries of the procession would now appear to be resolved: the audience has learned whom the grail serves with Perceval, and with Gauvain why the lance bleeds. However, this episode has so extensively rewritten the original Grail scene that new possibilities are opened up Chapter 2 will develop further the significance of this rhyme pair estoire/memoire. As this part of the text is more stable than the earlier parts of the Continuation, line references are given only to the Long Redaction, with exceptions and alternative readings signalled as such. 40 41

NARRATIVE AESTHETIC AND CYCLIC FORMATION

61

of a subsequent return. First, there is no longer a procession. The Grail, whose mysterious trajectory from one room to another had aroused Perceval’s curiosity, is now disassociated from the lance, which Gauvain notices standing in a silver vase. Indeed, although the Grail now behaves in an extraordinary manner, flying around the table serving the assembled diners, there is no longer any mystery about its destination, and Gauvain’s questions to the Fisher King do not concern it.42 Most significantly, new elements have been added to the scene: the body of a knight on a bier and a broken sword whose mending now becomes the prerequisite for some unspecified future development which appears to involve vengeance on behalf of the dead knight and a possible redemption for the area and its inhabitants (C1 [LR] 17427–30; [SR: A] 7305–10). As Gauvain’s imperfections prevent him from mending the sword, narrative potential is created for a future, perfect, knight to succeed.43 This knight’s identity is unclear at this stage. It could be Gauvain, who is told by the king that:   se Diex vos avançoit tant Vostre proëce ça avant Que ça vos lessast retorner, Bien le porrïez achever.  (C1 [LR] 17467–70) [if God were to improve your valour enough that he allowed you to return here, you might well be able to complete the task.]

Perceval certainly received no such encouragement in the earlier scene. However, Gauvain’s words upon waking far from the castle the next morning strongly echo those of Perceval before him, as he swears:   que il feroit Tant d’armes et se peneroit, Que se Diex li donnoit trover La cort, que bien porroit souder L’espee, et bien acheveroit La besogne por qu’il iroit  (C1 [LR] 17807–12)

42 Manuscripts ALMQU contain an explanation of the Grail’s origins by the Fisher King, but even this interpolation is provided spontaneously by the king, before answering Gauvain’s questions: ‘Mes avant vos voil fere sage / De ce donc en vostre corage / Doutastes ceanz, biax amis’ [But first, dear friend, I want to instruct you about that which, in your heart, you feared here] (C1 [LR] 17553–5). 43 In manuscripts AS of the Short Redaction, this knight to come is described as he who ‘De tot le mont le pris avra’ [will have the praise of the whole world] (C1 [SR: A] 7352); similarly, in TV, as the Fisher King hands the broken sword to Gauvain for him to try to mend it, he tells him that if he succeeds no one will doubt ‘Que li mieldres estes del monde’ [that you are the best in the world] (C1 [MR] 13387). It may be that these lines were interpolated with Perceval’s Second Continuation adventures in mind.

62 THE

CONTE DU GRAAL CYCLE

[that he would strive so hard in chivalry that, were God to allow him to find the court again, he would be able to mend the sword and complete the task that had led him there]

Gauvain has in one important sense gone further than Perceval, since he has partially healed the land by asking why the lance bleeds. Yet there is little indication that his plan to busy himself with chivalry in order to become deserving of the Grail quest again will be any more fruitful than it was for Perceval. The narrator, echoing the move made at the end of the Hideous Damsel episode to undercut Perceval’s oath by abandoning his narrative, warns us that he will not be discussing Gauvain’s adventures for the foreseeable future (C1 [LR] 17855–8). As happened to Perceval, the adventures which Gauvain hoped would increase his worth fail to leave a mark in the estoire. But, the narrator continues, nor will he be developing material around the other potential Grail knight mentioned in this episode: Si [n]’orroiz ja du chevallier Qui fu el paveillon ocis, Dont il fu ne de quel païs;  (C1 [LR] 17860–2) [You will never hear again about the knight killed in the tent, nor where he was from]

It is this unnamed knight’s mission that Gauvain takes on after the stranger is killed under his protection; he appears to be the knight of whom the Fisher King says: Bien sai, cil qu’i l’avoit empris Est remés en vostre païs; Ne sai qui si l’a retenu, Mes molt l’avions atendu  (C1 [LR] 17475–8) [I well know that he who had undertaken the task has remained in your land; I do not know what has kept him there, but we have waited a long time for him]

The audience is aware that the knight has been killed and will not be returning to pursue the task, marking another narrative dead end in the Grail adventure. The narrator’s refusal to consider either his back-story or Gauvain’s subsequent wanderings as part of his main tale is, as remarked earlier, consistent with the narrative aesthetic of disruption and centrifuge which we have identified at work both in the second half of Chrétien’s part and the First Continuation in general. The later Continuations, however, have little sympathy for this aesthetic in their drive to cyclify. Like the majority of critics, they read Perceval’s oath in the Hideous Damsel scene as announcing an eventual return to the

NARRATIVE AESTHETIC AND CYCLIC FORMATION

63

Fisher King’s court, with one significant difference: they are bound also to bear in mind the way in which the Grail Castle scene in Branch V of the First Continuation has altered the parameters of what happens there. Since the grail and lance questions have more or less been answered by now, the sword test becomes the new focus of attention. However, there is still the awkwardness of the Hermit episode to be dealt with, both in its marginalisation of (and refusal to narrate) Perceval’s chivalric adventures and in its lack of impetus for further development of his narrative. Thus, when Perceval reappears after his long absence from the First Continuation, the text needs a way to minimise the importance of the time spent with his hermit uncle earlier on. Of the ten manuscripts which transmit this material, seven (AKLMQSU, representing the Short Redaction here) opt not to make mention of the hermit at all, saying only that Perceval:   ot an tante terre esté Et trové tantes avantures Et feites batailles molt dures, Dom il l’enor avoit eüe.  (C2 [SR] 9460–3) [had been in many lands and found many adventures, and been in fierce battles, all of which had brought him honour.]

The mention of ‘avantures’ and ‘batailles’ confirms that Perceval has followed up on his oath at Arthur’s court to seek out faraway adventures and challenging combats; unlike the opening of the Hermit episode, however, there is no criticism of this conduct, nor suggestion that it has led him into sin. Entering a forest, he meets a huntsman, who repeats the Hideous Damsel’s condemnation of his failure at the Grail Castle. Like the Damsel, he pointedly refuses to greet Perceval (C2 [SR] 9493–4; cf. CdG 4642–4), calling him ‘Cheitis’ and presenting his failure as irrevocable. He also warns him to turn back; if he continues, and reaches the place where a horn hangs from a gate, ‘s’avrïez perdu / Vostre pris’ [you will have lost your worth] (C2 [SR] 9520–1). Perceval rides on ‘[l]a ou ge cuit morir’ [to the place where I think I will die] (C2 [SR] 9525), reasoning that his shame is such that he has nothing to lose. This re-enaction of the Hideous Damsel’s harangue effectively excises Perceval’s visit to his hermit uncle, leaving him back where he was immediately after the Damsel’s appearance. By introducing the episode of the Ivory Horn in this way, it becomes a test of the validity of the huntsman’s words, and hence of his presentation of Perceval’s failure as irrecoverable. Sure enough, far from losing him his ‘pris’, this adventure sees the first evocation of a glorious chivalric destiny for Perceval: even before the combat, the way he blows the horn prompts the Sire du Cor to describe him as the best knight in the world (C2 [SR] 9617), and when he has been defeated he tells Perceval: ‘Vos seroiz li miaudres vasax / … Et li plus renomez / Qui onques fust de mere nez’ [you will be the best knight and the most famous ever born of a mother]

64 THE

CONTE DU GRAAL CYCLE

(C2 [SR] 9776–8). This utterance displaces the irrevocable future predicted by the Damsel and the hunter with another, more glorious one, which (as we have seen) the Second Continuation will eventually confirm.44 Of the three other manuscripts, which represent the Long Redaction here, E’s interpolation of fourteen lines from the opening of the Hermit episode has already been discussed. Whereas the Short Redaction marginalises the hermit material by returning intertextually to the Hideous Damsel material which precedes, E and the other two Long Redaction manuscripts (PT) choose to acknowledge briefly the meeting between Perceval and his uncle. All three dramatise the moment of departure not portrayed in the original episode, shifting emphasis away from the past mistakes which were the subject of the hermit’s sermonising. In ET, Perceval (newly cleansed of his sins in readiness) rides ‘[p]ar mains païs et si trova / Maint mau pas et mainte avanture / Qui ne sont pas an escristure’ [through many lands and found many challenges and adventures that are not preserved in writing] (C2 [LR] 19640–2). This type of formula is by now familiar: both Perceval and Gauvain found earlier in the cycle that their attempts to redeem their Grail Castle failures through chivalry made no impression on the narrative record. However, the redeployment of the same semantic field here carries no such negative implication, since it introduces the Ivory Horn episode, the first of many adventures related by the ‘escristure’ of the Second Continuation. Transformed into a mere linking device, the topos loses the sting of condemnation. Finally, manuscript P offers yet another means of circumventing the fallout from the Hermit episode. Opting for concision, it returns us to ‘Perceval, / Qui est montés sor son ceval / Et s’est partis de son hermite’ [Perceval, who has mounted his horse and bidden farewell to his hermit] (C2 [LR] 19610 variants: MS P). It goes further than any other manuscript in acknowledging the irrevocable failure plotted by the textual material it has inherited, having its hero internalise the Hideous Damsel’s prophecy in a display of self-criticism: ‘mais ne quide trover jor / La court au rice Pesceour’ [but he does not think he will ever find the court of the Rich Fisher].45 The use of the word

44 The Sire du Cor also notes Perceval’s young age: ‘Molt vos voi juesne’ [I see you are very young] (C2 [SR] 9769), further emphasising his potential to develop into the world’s best knight. 45 For William Burgwinkle (2004: 97), the recurrent moral judgments to which Perceval is submitted by others throughout the Conte du Graal cycle are designed to create him as a subject within the discourse of elite chivalric masculinity: ‘This move to assumption of responsibility for one’s own subjection… describes quite accurately the process by which Perceval is transformed from a young naïf who checks his every action against his mother’s commandments to a more “mature” individual who has replaced the mother by an Other: an internalized sense of culpability that directs his quest.’ In a similar vein, Ben Ramm (2007: 120) opines that ‘Chrétien’s Perceval is always already the scapegoat, and he confirms himself in this role after his silent failure at the Grail castle’.

NARRATIVE AESTHETIC AND CYCLIC FORMATION

65

‘quide’ raises the possibility that this assessment might be wrong. The text has subtly reversed the situation: rather than a determined Perceval battling against a narrative uninterested in his attempts to shore up his reputation and looking elsewhere for sustenance, we now have a Perceval wracked by doubts thrust centre stage by a narrative in search of a unifying figure whose exploits it might recount. In this version, after leaving the hermit it takes our hero a mere three days of wandering in the forest before he arrives at the Castle of the Ivory Horn. Thus each manuscript, whether by omission or by reconfiguration, ultimately finds a way to make the audience believe that the Grail Castle was always Perceval’s eventual destination. As we have seen, the Second Continuation, having re-established the validity of Perceval’s association with the Grail Castle, makes progression in chivalry the goal of its narrative logic. Being the world’s best knight, the honour which the Sire du Cor predicts for Perceval and which he achieves after crossing the unfinished bridge, becomes the condition for an eventual return to the Fisher King’s court. Thus, whereas the Hermit episode portrays Perceval’s chivalric exploits as misguided attempts to atone for the unredeemable, the Second Continuation’s narrative project endorses and realises the oath which he makes at Arthur’s court after the Hideous Damsel’s passage there. The narrative structure which makes the Grail Castle the final destination of an episodic journey is adopted by the two later Continuations, creating a looping pattern whereby Perceval’s departure is necessary for his later return. Yet such a structure relies on the invention of new reasons to send the hero out and new conditions for him to return. Each successive Grail Castle visit therefore becomes the opportunity for a new narrative agenda to be launched. In the Gerbert Continuation, it is Perceval’s moral and spiritual blemishes, symbolised by the crack left in the sword after he has joined it, which provide the rationale for his departure. In the Manessier Continuation, the motif of revenge associated with the sword in the First Continuation is developed, as Perceval sets out to kill the man responsible for wounding the Fisher King and murdering the knight on the bier (revealed to be the king’s brother). In each case, our hero leaves the Grail Castle with a clear notion of what he must do before he can return, which is a world away from the situation he faces in the Chrétien section of the corpus. Other adventures function either to delay or to assist his inexorable progression back to the Fisher King’s court. The cyclical agenda is served in these later texts by a newly developed clarity about the direction the narrative is headed. At the same time, the purpose of returning to the Grail Castle becomes somewhat opaque; ostensibly revelatory, these visits serve in fact to confirm what we as readers have already learned (the lance is that of Longinus; Perceval is the world’s best knight…). The notion of return, and deferral of return (its inseparable twin), becomes from the Second Continuation onwards the structural principle of

66 THE

CONTE DU GRAAL CYCLE

both Perceval’s errance and the narrative progression.46 But while key locations are revisited, the hero proves resistant to others’ desires to fix him in place. The urgency to move on is, however, mitigated by promises that he will return when he has the opportunity. The intratextual precedent for such situations is given in Chrétien’s section. The first time Perceval finds his way to Biaurepaire, he meets his beloved Blanchefleur and ends the crippling siege of her lands by the suitor Clamadeu, but promises to return when he has learned of his mother’s fate. This promise is not kept immediately, but Perceval does return to Biaurepaire in the Second Continuation to find the area decisively changed from siege-inspired wasteland to prosperous mercantile territory; subsequent visits in the Gerbert and Manessier Continuations confirm this impression. At the end of each of these episodes, the same sequence is played out: request to stay–refusal–promise to return. This sequence occurs sufficiently often throughout the cycle that it takes on for the audience a rather conventional character. Nevertheless, the hero’s itinerary does have the capacity to loop back on itself; Perceval’s wandering at times brings him unexpectedly back to one of the places he has promised to revisit, rather fortuitously making good on his promises. In a sense, the audience finds itself in the same position as Perceval, unsure whether the next path will lead forwards to new adventures or back down memory lane. This effect is heightened by the disorientating combination of strangeness and familiarity in a journeying environment made up of literary tropes – the forest, dark and unending; the river, deep, fast-flowing and unfordable. This environment is in one sense frustratingly concrete, interfering with the character’s journey – unable to cross the river, the knight is obliged to follow it. Yet it is also semantically abstract – each river, each forest is built from the same pool of adjectives, so that markers of geographical or topographical specificity are rare. Against such a background, it is little wonder that neither audience nor knight has a clear idea of where they will end up next. The partial exception to this logic of the accidental return is the Gerbert Continuation, which makes more transparently deliberate use of textual memory.47 In this part of the text, Perceval recovers his faculties of recognition; riding through a forest, he recognises it as the area where he:

46 As well as the various visits to the Grail Castle and the reuse of settings from the Hideous Damsel episode discussed above, the text also returns to Biaurepaire (CdG, C2, CG, CM), Perceval’s family home (CdG, C2, CG), the Chapel of the Black Hand (C1, C2, CG, CM), the Castle of Maidens (C2, CG, CM), the castle of Gornemant (CdG, CG) and the smith Trebuchet’s workshop (CG, CM). 47 Bruckner (2009: 61–2) neatly describes Gerbert’s art both in this Continuation and in the Roman de la Violette as ‘not mere imitation but rather reinvention that seriously replays and rethinks his predecessors. Gifted bricoleur, he takes obvious pleasure in using elements from a variety of sources, including his own narrative, to come up with something new based on the “already said”.’

NARRATIVE AESTHETIC AND CYCLIC FORMATION

67

Encontra les cinc chevaliers A cui il demanda premiers S’il estoit Dieus ou angelos, Quant il portoit ses gavelos.  (CG 2609–12) [had met the five knights whom he had asked, right at the start, whether they were God or angels, when he was carrying his javelins.]

Having nudged the audience’s thoughts right back to the opening of the cycle, the narrator then draws on Perceval’s subsequent return to the region in the Second Continuation – nearby is his family home, where he had been ‘l’autre fois, quant il i fu / Et sa suer l’ot reconneü’ [the other time, when he had arrived there and his sister had recognised him] (CG 2617–18). Three memories are being engaged simultaneously across the whole cycle: that of Perceval, that of the text and that of the audience. Similarly, the return to Biaurepaire in this text is, for once, not accidental but a deliberate journey undertaken from the castle of Blancheflor’s uncle Gornemant in order to celebrate the wedding which both Perceval and text have been avoiding until this point. The purposeful movement from Gornemant’s castle to Biaurepaire retraces the hero’s steps in the early part of Chrétien’s section, harnessing this fragment of narrative to its cyclical agenda. Even here, however, the past has the capacity to re-emerge unexpectedly, further embedding the Gerbert Continuation into the rest of the corpus. The brother of the giant killed by Perceval in the Second Continuation comes looking for revenge; similarly, an adventure involving a locked chest eventually reveals itself to be another vendetta tale, as the opening of it by Perceval in front of four brothers identifies him as the murderer of their father – the Scarlet Knight who was Perceval’s first adversary. If the Gerbert Continuation performs perhaps the most extensive structural engagement with textual and audience memory, such play is at work throughout the cycle. A final striking instance is provided in the closing scenes of the Manessier Continuation. In the lead-up to Perceval’s concluding visit to the Fisher King, the narrator explains that his hero: ne sot ou le chastel querre Car ne sot mie bien la terre De la entor ne le païs Et si en estoit il naïs. De la Gaste Forest ert nez, Mes n’estoit mie bien senez, Quant de sa mere se parti Ne onques puis n’i reverti.  (CM 41853–60) [did not know where to seek the castle for he did not know well the surrounding area, nor the land. And yet he had been born

68 THE

CONTE DU GRAAL CYCLE

there: he was born in the Waste Forest, but he had not been very wise when he had left his mother, nor had he returned there since.]

The revelation that the Grail Castle stands in the same region where Perceval grew up is an innovation of this passage, which seems attracted to the concept of bringing the hero back to where he started. Yet perhaps even more surprising for the audience is the claim in the final line that Perceval had never returned to the area since he had left his mother as a young man. We are asked to forget, not only his return to his mother’s house in the Second Continuation (and a second return in the Gerbert Continuation of TV), but also his two (three in TV) previous arrivals at the Grail Castle on which the cyclic coherence of the narrative has in part depended. In imposing its innovation on the narrative the text chooses to forget the design which it inherited, wiping its hero’s memory clean. When Perceval does eventually see and identify the castle, the narrator is explicit: ‘par aventure la avint’ [he came there by chance] (CM 41867). Our hero’s errances thus bring him back one final time, not only to the sought-after setting of the Grail procession ritual, but also to the unrecognisable landscape of his birth, the place where both he and the cycle began. In the Introduction to Mille Plateaux, Gilles Deleuze and Felix Guattari suggest an analogy between two ways of thinking about texts and two types of memory. One approach to textuality, like short-term memory, works under conditions of discontinuity, rupture and multiplicity; the other, on the contrary, tends towards systematisation.48 This distinction provides a helpful means of tying together the observations we have made about narrative structure in the Conte du Graal cycle. The whole corpus displays a capacity to choose to remember or to forget what has come before, depending on whether the desire is to privilege innovation or tradition. Yet a definite change of aesthetic is apparent as one moves forward in the corpus. The Chrétien section, followed by the First Continuation, is centrifugally structured, promoting an aesthetic of dispersal and postponement. The join between these two parts is exemplary in this respect: the First Continuation respects continuity at a local level, displaying a short-term concern with remembering and completing the interrupted Guiromelant thread, but its aesthetic allows it comfortably to accommodate multiple heroes and narratives. The Second Continuation, followed by the two later Continuations, contests this aesthetic by rewriting the earlier material into a more tightly controlled structure, using devices such as interlace and quest to promote Perceval as the pre-eminent knight of the Arthurian world. In order to do so, forgotten plot threads are revived and made to serve the purpose of demonstrating the central hero’s superiority. Likewise, the Long Redaction of the First Continuation uses references to 48

See Deleuze and Guattari (1980: 24–5).

NARRATIVE AESTHETIC AND CYCLIC FORMATION

69

earlier and later material in the narrative to tie it more closely to the rest of the corpus. The effect of this later aesthetic is to create a biographical cycle around Perceval, with other characters’ adventures recast as digressions, and in doing so the Continuations unavoidably alter the way in which the Conte du Graal is read and transmitted. The following chapter will examine the manuscript tradition of the corpus in order to demonstrate how its thirteenthand fourteenth-century audiences received and reinterpreted this composite and contradictory text, and how the long-term memory of the cycle finds material form on the parchment of the codex.

2

Manuscripts, Memory and Textual Transmission In the prologue to his Roman de Troie, Benoît de Sainte-Maure recounts how Cornelius, nephew of Sallust, came across Dares Phrygias’s eyewitness account of the fall of Troy one day while browsing for grammar books in a bookcase. As Benoît tells it, Dares felt compelled to leave a written record of the events he was witnessing: Por ço qu’il vit si grant l’afaire Que ainz ne puis ne fu nus maire, Si voust les faiz metre en memoire; En grezeiz en escrist l’estoire.1 [Seeing that these were the most significant events that there had ever been, or would ever be, he wanted to consign the facts to memory: he wrote the story in Greek.]

The fiction of narrative genesis that Benoît presents involves a fairly complex interaction of oral and written authority, encapsulated in the apparently straightforward act of ‘metre en memoire’: events are written down by ocular witnesses and handed down the generations via a double translation, first into Latin and then French (thus mirroring the trajectory of learning and empire described by Chrétien de Troyes in the prologue to Cligés). Significantly, the latter translation is also that of an ancient source document into a performed romance narrative. This vignette illustrates some of the issues that pertain to both the reality and the representation of textual transmission in the Middle Ages, and which I will be discussing in this chapter. In particular, the rhyme pair of memoire/ estoire encapsulates the instability of distinctions made between ‘performative oral’ and ‘documentary written’ modes of thought and practice in medieval literature. This rhyme pair is relatively common in Old French literature, appearing overwhelmingly in the context of discussions of cultural heritage and its transmission.2 A cursory gloss on the words might associate the notion Roman de Troie, vv. 97–100. The entries for ‘estoire’ and ‘memoire’ in Tobler–Lommatzsch (3.1402–4 and 5.1378–83 respectively) contain a number of examples of this rhyme pair, some of which will be the subject of further comment in this chapter. It is fairly clear that, apart from the 1 2



MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

71

of memoire with oral and estoire with written transmission, but both terms are in fact riddled with ambiguities which testify to the complexities of transmission and reception of medieval texts. As Mary Carruthers and Brian Stock have both shown, increased reliance on written record from the eleventh century onwards implied neither a lessening in the status of memory, nor an abandonment of oral discourse.3 On the contrary, oral and literate modes of expression informed and influenced each other throughout the medieval period. Both the creation and transmission of texts were frequently conceived of as functions of the faculty of memory in terms which might draw on both oral and written frames of reference. Thus Dante, explaining the genesis of his Vita Nuova, describes finding in the ‘book of my memory… a rubric that says: Incipit vita nova. Under this rubric I find written the words that I intend to copy into this little book…’4 As Carruthers demonstrates, the image of the writing surface as a metaphor for memory is a structuring model for medieval discussions of the subject, beginning with the ancients. However, Dante’s actualisation of this model goes further than merely updating the metaphor to reflect late-thirteenth-century writing practices; it speaks of the growing importance of elements of manuscript design such as rubrication, and an awareness of their ability to convey information to audiences interacting directly with the manuscript page. The manuscript context of a work engages important issues around literary creation and reception. These issues acquire an extra layer of complexity in a cyclical corpus, where the process of mise en cycle is fundamentally a function of manuscript transmission and creates such enormous textual masses that the artificial memory of the codex seems like the only means of keeping a handle on the whole. Sandra Hindman’s comment about manuscript T of our cycle that ‘the book acquires privileged status as a repository of memory’ has general validity for all such cyclical manuscripts.5 As we saw in Chapter 1, the process of scribal accretion, editing and rewriting leads to variations between individual manuscripts of a cycle, with each one providing a different configuration of the cyclical material. Each of these configurations constitutes one possible way of representing, and transforming, the narrative facility of the rhyme, the texts that show a particular taste for it are those which attempt substantial interrogations of cultural heritage and heredity. In particular, texts such as Florimont and the Conte du Graal cycle share a tendency to use the same semantic field in discussing the genealogy of the narrative as when discussing that of its participants, diegetic (characters) or extra-diegetic (patrons). 3 See Carruthers (1990) and Stock (1983). On the rise of literacy and its connection to testimony, see Michael Clanchy’s seminal study (1979). 4 ‘In quella parte del libro de la mia memoria dinanzi a la quale poco si potrebbe leggere, si trova una rubrica la quale dice: Incipit vita nova. Sotto la quale rubrica io trovo scritte le parole le quail è mio intendimento d’assemplare in questo libello’ (Vita Nuova; quoted in Singleton 1949: 26). 5 Hindman (1994: 44).

72 THE

CONTE DU GRAAL CYCLE

dynamic of the text. Further, because of the material aspect of a manuscript, visual elements such as the placement and nature of initials and miniatures can work to create a sort of spatial palimpsest, a second textual layer made up of a cycle of images and material information. This material text can be read in different ways. It can be neglected altogether, as we did in the previous chapter – an audience listening to the text may not necessarily be interested in visual engagement with its manuscript. For those who do choose to read it on the page, it acts as a commentary on the diegesis, but it also mediates and therefore helps to produce the diegesis for its reader. In addition, the very practice (not to mention expense) of manuscript decoration creates an implied reader who may choose to browse the spatial text of the manuscript without engaging with its diegetic text. An owner of an Arthurian romance manuscript is not necessarily an avid reader of the genre. However, even the most superficial engagement with the material text is bound to produce meaning, just as flicking through a book in a shop, glancing at its cover and so on, is often sufficient for the browser to form opinions about its content. As Sylvia Huot argues, the visual and aural uses of manuscripts must have been susceptible of informing each other: Through a visual appreciation of author portraits, recurring iconographic motifs, and general page layout, even an illiterate ‘reader’ could have been conditioned to certain concepts of authorship, codicological continuity, or literarity, which in turn might inform his or her appreciation of texts received aurally.6

Thus, while the paratextual apparatus may at first glance appear semantically slight, a sustained interrogation of the manuscript tradition of the Conte du Graal cycle can yield significant insights into the nature of medieval romance textuality. This chapter is divided into four parts. The first part pursues further the implications for our corpus of the suggestive coupling of estoire and memoire, arguing that the cycle engages both written and oral modes of communication in developing an account of its own creation and transmission. The next two sections demonstrate how the developing discourse of text-as-livre affected the presentation of the Conte du Graal cycle in its surviving manuscripts, with a demonstrable evolution in the use made of the paratextual possibilities of the manuscript page. Simultaneously, the analysis of the cycle’s codicological context reveals two complementary ways in which thirteenth- and fourteenth-century readers were invited to conceive of the corpus’s coherence. Drawing primarily on the unilluminated manuscripts, the second part of the chapter examines the ways in which the placement of large initials and the wording of rubrics respond to and develop the interlace which we showed in 6

Huot (1987: 7).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

73

the previous chapter to be an important aspect of the text’s cyclic structure. In the third part I will argue that the illustrated manuscripts underline a different aspect of the corpus’s cyclicity, tending through their miniatures and rubrics to insist on the text as a biographical ‘roman de Perceval’. Models of cyclic coherence, as we saw in the introduction, tend to demand some notion of at least provisional completeness, of the self-sufficiency of the cyclic ensemble as separable from its wider intertextual context; completeness, however, is itself a far from straightforward notion, susceptible of taking different forms. Accordingly, the final section of the chapter will consider different models of cyclic completeness generated by the textual configurations, exclusions and inclusions found in the manuscripts, revealing how the biographical cyclic core causes alternative models of completeness to be marginalised in the manuscript tradition, before concluding with a discussion of how the paratextual frame provided by incipits and excipits guides the reader’s understanding of the nature of the text being read.

‘Si com le tesmoingne l’estoire’: History and Memory Our eight cyclical manuscripts contain a variety of different codicological features, and there is no sense of a standardised programme existing for the material text, as regards either the placement or the content of decoration. This makes these manuscripts especially valuable, since each one can be read as offering a distinct approach in guiding the reader’s response to the text; features repeated across different manuscripts, while they might be partly explained by shared models, can also provide evidence of similar reading patterns on the part of their scribes. Two manuscripts, PT, have historiated initials throughout and a small number of miniatures (just an opening miniature in P and four in T). Three, MSU, have extensive programmes of illumination (over fifty miniatures); S also has twelve historiated initials, while M’s miniatures are almost always accompanied by large initials. Of these five illuminated manuscripts, PSU feature rubrics; T does not; M has rubrics identified as later additions. The other three, EQV, have large decorated initials (between four and twelve lines high) and no miniatures.7 Roger Middleton acknowledges the contribution of the material text in creating a visual aesthetic, but he expresses caution as to the ability of decorated initials and miniatures to yield textual information: What the miniatures, historiated capitals, gilded capitals, ornamented capitals, and coloured capitals all have in common is the intention to enhance the appearance of the book as an object to be seen (perhaps even dis7 As V has extensive lacunae, it is hard to make general pronouncements on its ­material text, and for this reason it is mentioned sparingly in the present discussion.

74 THE

CONTE DU GRAAL CYCLE

played). Whether they have a second function, of marking structural divisions of some sort, is much less clear.8

While it is true that the smaller, two-line initials appear both at moments of narrative change and of narrative stasis, the ambiguous distribution of this lower-level decoration should not deter us from recognising the import of occasions where the material text does appear to be marking textual structure. The annotation of one folio of manuscript P (Figure 1 below) provides evidence that at least some readers understood the larger-scale decoration of the Conte du Graal cycle as implying a structural division. Line 32265 of the Second Continuation (‘Tant a alé, ce m’est avis’ [He travelled for so long, it seems to me]) begins the narration of Perceval’s second visit to the Grail Castle, and in manuscript P it is allotted a ten-line historiated initial. By this initial, a reader has written: ‘Nota / tout ce capitle’ [Note this whole chapter], apparently intending to flag up the account of the hero’s arrival (the next historiated initial occurs just after the Fisher King’s explanation of the Grail’s sacred nature). The notion that the historiated initial marks the opening of a chapter is therefore shared by both this anonymous reader and the text’s editor, Roach, who chooses this line as the first of the Second Continuation’s final episode. Table I on pages 84–5 shows the lines marked by large initials in at least two of manuscripts EPQTV (the five codices which do not contain a substantial programme of miniatures). A striking feature of this initial placement is a recurrent tendency to highlight couplets containing the rhyme pair estoire/ memoire. Of the thirteen instances across multiple manuscripts listed in the table, three mark a couplet that rhymes these two words (four in manuscript E due to its use of the same couplet at CdG 6217–18 and C2 [LR] 19617–18). By inflating and decorating the openings of these lines, the manuscripts highlight the semantic pairing of story and memory: Perceval, ce conte l’estoire A si perdue la miemoire  (CdG 6217; marked in EPQTV) [Perceval, the story tells us, has so lost his memory] Gauchiers de Dondain, qui l’estoire Nos a mis avant en memoire  (C2 31421; marked in EPTV) [Wauchier de Denain, who gave us this story to remember] Gauvain, si raconte l’estoire Qui de lui est mise en memoire  (CM 35051; marked in EPT) [Gauvain, so tells the story which is recorded of him] 8

Middleton (1993a: 157).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

75

Disclaimer: Some images in the printed version of this book are not available for inclusion in the eBook. To view the image on this page please refer to the printed version of this book.

Figure 1. MS P (Mons, BU 331/206), p. 371. C2 32265. A historiated initial depicts Perceval arriving at the Fisher King’s castle. Rubric: ‘Pierchevaus li galois’. The marginal annotation reads: ‘Nota tout ce capitle’.

The rise of prose Grail romance has often been linked to a perception that prose is a more trustworthy and veridical form for narrative, typified by the oft-quoted assertion in a prologue to a translation of the Pseudo-Turpin dated around 1200 that ‘nus contes rimé n’est verais’ [no rhymed tale is true].9 On this account, the shift from verse to prose is part of a game whereby the Grail romances present themselves ostentatiously as true stories, dictated by angels and devoid of artifice. Yet the Conte du Graal corpus, a Grail romance cycle completed in verse during the very period that saw the rise of prose Arthurian romance, muddies the waters of such an analysis. What is intriguing here is that the rhymed verse form of the text generates a level of insistence on the pairing of estoire (with its undertone of historical truth) and memoire which could not be achieved in prose. Indeed, Chrétien uses the rhyme in the prologue to his first Arthurian romance, Erec, in a well-known passage which plays on the idea of the story surviving for posterity:

9 Cited in Woledge and Clive (1964: 27). On the fiction of the prose form as an inherently suitable vehicle for historical truth, see Godzich and Kittay (1987); Spiegel (1993); and Baumgartner (1994b).

76 THE

CONTE DU GRAAL CYCLE

Des or comencerai l’estoire Que toz jors mais iert en memoire Tant con durra crestïentez. De ce s’est Crestïens ventez.10 [Now I will begin the story that will always be remembered, as long as Christendom lasts; this is Chrétien’s boast.]

The scribe of manuscript Q was evidently familiar with Erec, since he introduces this passage (though without marking it with an initial) into the account of the marriage of Caradoc and Guinier after she has saved him from death. A vast number of gifts were given to the happy couple, assures the narrator, Si com le tesmoingne l’estoire Qui tot jors mes iert an mimoire Tant con durra crestïantez: De ce s’est Crestïens ventez.  (C1 [LR: Q] 11906 +1–4) [As is testified by the story that will always be remembered, as long as Christendom lasts; this is Chrétien’s boast.]

It is quite likely that this scene of celebrations called to mind the similarly lavish coronation of Erec and Enide which concludes Erec, but by pulling these lines from the prologue of that text the scribe of Q shows particular concern for the pairing of story and memory. Moreover, in rewriting the first line of this quatrain, the estoire is presented as important less as an example of its author’s talent (as was the case with the Erec prologue) than as a witness to the veracity of the events recounted in this (hi)story. The cycle places itself from the outset in a hybrid space, creating a narrative about its own gestation that involves both orality and literacy. Chrétien’s prologue ascribes the tale’s origin to a book given to the author-figure ­‘Crestiens’: Ce est li contes del graal, Dont li quens li bailla le livre. Oëz comment il s’en delivre.  (CdG 66–8) [It is the tale of the grail, which the count gave him the book about. Listen how he acquits himself of it.]

The combination of written source and oral delivery sets the template for the repeated coupling of estoire and memoire found later in the cycle, with the written book standing behind, and guaranteeing, the performed tale. The

10

Erec, vv. 23–6.

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

77

figure of Wauchier de Denain, a writer of hagiography, is thus invoked at C2 31421 as the authority for the transmission of the story into memory: Mais de Perceval lou Galois Porroiz le conte avant oïr S’a gre vos vient et a plesir. Gauchiers de Dondains, qui l’estoire Nos a mis avant en memoire, Dit et conte que Perceval…  (C2 31418–23) [But now you will be able to hear the tale of Perceval the Welshman, if it pleases you. Wauchier de Denain, who gave us this story to remember, tells and recounts that Perceval…]

Long thought to be apocryphal, this authorial attribution is now generally held to be genuine, though Matilda Bruckner cautions against assuming widespread recognition by a medieval audience, given the many deformations of the name in the manuscript tradition.11 Nevertheless, whether recognised or not, the value of Wauchier’s name is to anchor the performed tale within a history of written transmission; the authority is not so much that of the individual named as that generated by recognition of that name’s function within a clerkly economy of writing. A similar move is found at the conclusion of the cycle where, as in the prologue, a written source is attested for the latter part of the text. This time it is the author-figure ‘Manessier’ who acts as guarantor for the truth of what has been recounted: Dame, por vos s’en est pené Manessier tant qu’il l’a finé […] Tant en a aconté et dit Con l’on a Salebiere en treuve, Si com l’escrit tesmoingne et preuve, Que li rois Artus seella. Encore le puet on veoir la, Tot seellé en parchemin, Cil qui errent par le chemin.  (CM 42657–8; 42662–8) [Lady, for you Manessier has worked until he has completed it… He has said as much as one finds written at Salisbury, as is testified and proven by the written text that Arthur had sealed there. It can still be seen there, sealed in parchment, by those who pass by.]

11 Bruckner (2009: 46–7). For a summary of the debate over the attribution of the Second Continuation to Wauchier, and a strong statement of the case for, see Corley (1984).

78 THE

CONTE DU GRAAL CYCLE

In these concluding lines to the cycle, the narrative of written transmission now stretches from the Arthurian past (contemporary with the events recounted) through the author-figure ‘Manessier’ to the narratorial voice speaking to the audience. As Hindman suggests, the cycle’s epilogue blurs the boundaries between literate and non-literate testimony: when in the imaginative rhymed couplet Manessier invites the reader to travel the road (par le chemin) to visit and ‘see’ the book (parchemin), he inventively transforms the analogy of the transaction of sealing into one that instead describes the act of reading.12

Similarly, the figure of ‘Manessier’ himself has a double function, at once a symbol of literate and of non-literate modes of apprehension: he is the transmitter of the written text, a link in the chain of transmission, yet at the same time he is implicitly presented as an ocular witness, whose testimony is lent authority from direct first-hand visual experience: Qui encore en cel païs va La sepoulture puet veoir Sor quatre pilers d’or seoir, Si com Manesier le tesmoingne, Qui met a chief ceste besoingne  (CM 42638–42) [Anyone who goes to that land can still see the tomb, sitting on four golden pillars: so Manessier, who is completing this task, testifies]

The reference to Manessier ‘who is completing this task’ in one important sense breaks with the aesthetic of narration found elsewhere in the corpus. By presenting the act of writing in the present tense, the temporal gap between writer and narrator is effaced, with narration and creation presented as simultaneous events. Conversely, the temporal immediacy of communication between text and audience projected elsewhere in the cycle disappears, since the present of the tale is now identical with that of its creation by the author. The audience is no longer included in the present tense of narration. Compare this with the passage cited above where Wauchier de Denain is named: here, the present tense in which the text unfolds is that in which the audience can ‘avant oïr’ [hear now] the tale, while the writer’s act is presented as past (‘nos a mis avant en memoire’ [gave us to remember]). As in the passage from the Roman de Troie cited at the beginning of this chapter, the act of ‘mettre en memoire’ is presented here as a literate one. The suggestion of textual commemoration in the form of narrative (‘estoire’)

12

Hindman (1994: 9–10).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

79

is found in a number of Old French texts which couple the two concepts of history and memory, as in the following passage from Wace’s Roman de Rou: Mult soleient estre onuré E mult preisé et mult amé Cil ki les gestes escriveient E ki les estoires faiseient. Suvent aveient des baruns E des nobles dames beaus duns, Pur mettre lur nuns en estoire, Que tuz tens mais fust de eus memoire.13 [Those who wrote histories and gestes used to be greatly honoured, prized and loved. They often received good presents from lords and noble ladies, in exchange for writing down their names in the narrative so that they would always be remembered.]

If the written form is generally represented as a repository for memoire, its transmission need not be textual. The late-twelfth-century romance Florimont, whose insistence on the pairing of history and memory rivals that of the Conte du Graal cycle, weaves a fiction of transmission which illustrates the complex interaction of oral and written modes of testimony. The author, we are informed, translated a Latin source into French; drawing on this chain of written transmission, the narrator announces: Por letz ancïens ramanbreir Vos veul issi dire et conteir Ensi com j’ai escris troveis D’une ystoire la veriteis.14 [In order to preserve the memory of the ancients, I want to tell you the truth of a story just as I found it written.]

The written source would appear to be a necessary, but not sufficient, condition for the ancients to be remembered; the story must also be told: Jel dirai, que l’ai en memoire; Or escouteiz mout riche istoire.15 [I, who have it in memory, will tell it – now listen to a very worthy story.]

The ambiguity of memoire here (is the narrator claiming to have access to 13 14 15

Roman de Rou, III, vv. 143–50. Florimont, vv. 43–6. Florimont, vv. 109–10.

80 THE

CONTE DU GRAAL CYCLE

the written document, or to have memorised the text for recitation?) appears irresolvable; the narrator himself insists more on the need for the ‘istoire’ to be heard by an audience, who in turn are asked to integrate it into their own cultural memory, where it can be of benefit to current and future generations: Et qui welt oïr ceste istore Et retenir en sa memore, Se em boen poent i welt antandre, Assez puet oïr et aprendre16 [And he who wishes to hear this story and keep it in memory can hear and learn much, if he is willing to understand it in the right way]

The crucial point is that it matters little whether the memorial process takes a literate or aural/oral form; the source document is invoked as originating authority, but it is the passing of the narrative into memoire through either oral or written means which truly authorises it as estoire. As Carruthers notes, it is this public process of ‘committing to memory’ which invests a text with cultural authority: It is also important to recognize that there are two distinct stages involved in the making of an authority – the first is the individual process of ‘authoring’, and the second is the matter of ‘authorizing,’ which is a social and communal activity. In the context of memory, the first belongs to the domain of an individual’s memory, the second to what we might conveniently think of as public memory. Texts are the primary medium of the public memory.17

What is innovative about the prose romances is not so much the claim to pseudo-historical status based on a written source as the means by which the text represents its ‘authorising’ process. The prose romances’ suppression of the narrator-figure, and consequently of the language of performance, presents authorising as a matter of written record, the business of reading communities. ‘Or dist li contes que…’ [Now the tale says that…], a narrative formula repeatedly encountered in prose Arthurian romance, reverses the pattern we have identified in twelfth-century verse romances such as Florimont or the Roman de Troie: where the livre in those texts lay behind the conte, in the prose Arthurian romances the livre is the narrative itself, reporting what the conte ‘dist’. Thus the final lines of La Mort le roi Artu announce that the author-figure Gautier Map ‘fenist ci son livre’ [finishes his book at this point]; the livre belongs to the present tense of narration-creation, and 16 17

Florimont, vv. 9263–6. Carruthers (1990: 189).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

81

the locative adverb ‘ci’ identifies the text with the space of the manuscript page.18 I would suggest that this tendency for thirteenth-century vernacular narratives to represent the text as a livre may be linked to an increase in the production of manuscripts for lay audiences through this period. The fact that we retain so many more copies of precisely those romances which present themselves (however playfully) as self-authorising books speaks of a change in how the relationship between audience, text and codex was conceived, with the textual object increasingly becoming conflated with, or represented as proceeding from, the material object.19 Nor can this phenomenon be isolated from that of cyclicity itself; the rise of the codicological discourse around the text reflects the same shift in literary mentality as the early thirteenth century’s new-found taste for vast textual ensembles, constituted and compiled by scribes into coherent cyclical manuscripts. As an Arthurian verse cycle, composed from the late twelfth to the early thirteenth centuries and overlapped by the creation of the major prose cycles, the Conte du Graal corpus is particularly useful for tracking this development. In Chrétien’s prologue to the Conte du Graal, it is clearly an auditory community of authorisers which the narrator has in mind when he describes Chrétien rhyming his conte from the count’s livre, inviting the audience to judge the quality of the text in aural terms: ‘Oëz comment il s’en delivre’ [Listen how he acquits himself of it] (CdG 68). By the time of Manessier’s epilogue, the text itself is conceived of as a ‘book’ (CM 42652), and Manessier invites his audience to travel to Salisbury and find the ‘escrit’ [writing] which guarantees the value of his narrative (CM 42663–4). The Conte du Graal cycle thus represents an accommodation between the performative discourse of twelfth-century verse romance and the text-as-livre discourse of thirteenthcentury prose romance. In the lead-up to the first ‘estoire’/‘memoire’ couplet at CdG 6217, we are told that: Ne d’aus ne del doel que il font Rien plus a dire ne me plaist. De monseignor Gavain se taist Ichi li contes a estal, Si comenche de Percheval.  (CdG 6212–16) 18 Mort Artu, p. 263. One manuscript (Brussels, Bibl. royale des ducs de Bourgogne, MS 11145) of another contemporary prose romance, the Perlesvaus, evokes in its colophon the ‘contes’ of which the ‘livres s’en tera ore atant’, thus making a similar distinction between the book, unfolding in the narrative present, and the contes from which it is drawn (Perlesvaus, I, p. 409). For further elaboration on the spatial dimension of the Vulgate Cycle’s narration, see Perret (1982), Baumgartner (1994a: 379–403) and Griffin (2005: 20–1); on the implications of spatial versus temporal deictic markers, see Fleisch­man (1990: 31–3). 19 Huot (1987) has convincingly demonstrated an analogous movement in lyriconarrative poetry ‘from a more performative toward a more writerly poetics’ (1), beginning in the thirteenth century.

82 THE

CONTE DU GRAAL CYCLE

[I do not wish to say anything more about them nor about their sorrow. Here the tale ceases speaking about Sir Gauvain, and begins to speak of Perceval.]

The two modes of performative, personified narration and the spatially defined ‘conte’ are both represented. The ambivalence of the text as transmitted in the manuscripts is manifest in their different readings of this passage. Manuscript B – not a cyclical manuscript – removes the suggestion of spatial narration by substituting the temporal adverb ‘atant’ for ‘ichi’ in line 6215; manuscript A, which contains the First Continuation and part of the Second, similarly eliminates the idea of a self-narrating ‘conte’, replacing ‘comenche’ with ‘parlerons’ at line 6216. The transmission of the Conte du Graal cycle in a relatively large number of manuscripts shows that the rhetorical strategies employed by texts such as the Pseudo-Turpin to discredit (and displace) verse were only partially successful. Yet the rise within both prose and verse texts of the discourse of text-as-livre mentioned above prompts us to consider a second stage in the cycle’s interaction with writing, that of its manuscript tradition. In the remainder of this chapter, I will study the paratextual information of this tradition in order to demonstrate what the material texts of the manuscripts reveal about how the cycle was understood by its second-generation audiences.

‘Ci vient li contes a Percheval’: Chapter and Verse A corpus as vast and diffuse as the Conte du Graal cycle raises challenges to its audience’s memory. As suggested above, the vast textual accumulation of cycles is only possible within manuscripts, and the way these manuscripts choose to present their narrative content has an important effect on the way the text is understood. If, as the narrator of Florimont claims, understanding a text requires its being committed to cultural memory, the way in which the Conte du Graal cycle enters its audience’s memoire both reflects and determines the value which the cycle is held to have. The manuscripts are thus bound to engage with the questions of how to remember the text and of what to remember. The material text provides fertile ground for tracing this engagement, since its features work explicitly and implicitly to guide the manuscript user’s response to the text. The placement of initials, rubrics and miniatures can have a bookmarking function, allowing the reader to move more easily between parts of the text, establishing parallels across the textual surface; the content of the paratextual apparatus, meanwhile, can serve as an aid to memory or to memorisation, foregrounding moments considered most significant or noteworthy. As Janet Doner points out, the initial, historiated or otherwise, is ‘a privi-

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

83

leged element of the verse text’.20 Almost always occurring at the opening of a rhymed couplet, it has the ability to perform an introductory function, and therefore has considerable influence over how the manuscript’s user apprehends the breaking up of the text into sections. Indeed, in our manuscripts there is a clear correlation between the placement of large initials and the beginning of an episode. In particular, the (re)introduction of a new hero is frequently marked. Interlace, which we identified in Chapter 1 as one of the key elements of the cycle’s narrative aesthetic, is thus fore-grounded by the placement of large initials. This is especially true of the unilluminated manuscripts EQV and the historiated manuscripts PT; illustration, especially when freed from the frame of a historiated initial as it is in MSU, can perform an ambiguous function, integrating material from one part of an episode into another. The large initials in the unilluminated manuscripts, being the only method used of marking certain parts of the manuscript space as worthy of special notice, are more securely anchored into the structure of the text which they present; the historiated initials in PT retain an organic relation with this structure in terms of placement, even when the illustrative content is drawn from elsewhere in the text. In EPQTV, the use of large initials thus suggests that the ‘mise en memoire’ of the Conte du Graal cycle is a matter of noticing and understanding its narrative dynamic of interlace. Table I below shows the lines marked by large initials in EPQTV. It is notable that almost all of these lines mark the introduction of new material, and most of them also mark a change of central character. The distribution of these initials further confirms several of the observations made in Chapter 1 about the constitution of the corpus’s cyclical drive. There is a clear difference, most notably, between the lines from the First Continuation marked by multiple manuscripts and those from the rest of the corpus. In the Conte du Graal section, there is agreement between the various scribes that the Hideous Damsel episode constitutes a significant development in the narrative, resulting in a change of narrative dynamic as the text leaves Perceval for the first time. The switch back to Perceval for the Hermit episode at line 6217 is also marked in all the manuscripts, including the illustrated ones. Equally, the key moment of the return to Perceval at C2 [LR] 19607 is marked in EPQ, despite the extremely varied versions of the text given in each copy.

20

Doner (1996: 77).

84 THE

CONTE DU GRAAL CYCLE

Table I Lines marked by large initials in at least two of EPQTV Context MS E Introduces the N/A Hideous Damsel’s (folios arrival at Arthur’s missing) court, and the narrative’s first switch of central character, leaving Perceval for Gauvain 4-line Introduces the 6217 initial ‘Perceval ce conte Hermit episode l’estoire’ (switch back from Gauvain to Perceval) 6519 Switch back to 4-line ‘Messire Gavains Gauvain from initial tant erra’ Perceval 9227 Final narrative 6-line ‘Ma dame Lore moment of Chrétien’s initial se seoit’ part of the text, continued in the first lines of C1 [C1] 5385 [LR] Transition from 7-line ‘A une Pantecouste Guiromelant material initial fu’ into Brun de Branlant story: Arthur assembles his barons to wage war on Brun 7037 [LR] Arthur holds court at 8-line ‘Ce fu an mai el Pentecost and knights initial tans d’esté’ Caradoc 10-line 11949 [LR] Arthur holds court; ‘A un jor d’un[e] introduces the initial episode of the golden Acension’ nipple given to Caradoc for Guinier [C2] 19607/17 Transition between 8-line [LR] C1 and C2: return to initial ‘Perceval ce conte Perceval after end of l’estoire’ (E) Guerrehet material ‘Or revenrons a Perceval’ (P) ‘Saingnor vos avez bien oi’ (Q) 21579 Narrative switches 10-line ‘Or vos diron de to Perceval after initial Perceval’ Abrioris’s arrival at Arthur’s court 29209 Narrative follows ‘Messires Gauvain after Gauvains sans Arthur’s knights arest’ separate on their quest for Perceval

Line [CdG] 4603 ‘Grans fu la joie que li rois’

MS P 10-line initial

MS Q 6-line initial

MS T MS V 7-line N/A (folios initial and missing) miniature

10-line initial

10-line initial

7-line initial and miniature

9-line initial

Small (2-line) initial

Small 8-line (2-line) historiated initial initial 10-line initial

8-line initial

Follows SR

8-line initial

Follows MR

Follows MR

SR

10-line initial

MR

MR

SR

8-line initial

MR

MR

10-line initial

8-line initial

Small (2-line) initial

Small (2-line) initial

Small (2-line) initial

10-line initial

Small (2-line) initial

8-line initial

10-line initial

7-line historiated initial

Small (2-line) initial

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION 31421 ‘Gauchier de Dondains qui l’estoire’ [CM] 35051 ‘Gauvains si raconte l’estoire’ 37141 ‘Ici de Gauvain vos lairons’

Narrative switches to Perceval after Gauvain’s return to Arthur’s court Narrative switches from the injured Sagremor to Gauvain Narrative switches to Perceval after Gauvain’s return to Arthur’s court

8-line initial

10-line initial

8-line initial

10-line initial

10-line initial

10-line initial

8-line historiated initial

85 8-line initial

8-line N/A N/A (folios (folios miniature missing) missing) 4-line 7-line N/A (folios initial historiated missing) initial

C2 [LR] 19617 is the point at which manuscript E re-inserts a passage from the Hermit episode of Chrétien’s text in order to re-impose Perceval’s story as the narrative focus. As shown in the table, in both appearances of the passage, the first line (‘Perceval, ce conte l’estoire’) is marked with a large initial. The material text thus supports and reinforces the narrative’s attempt to abolish the distance which the thirteen thousand or so intervening lines had built up, promoting continuity in the Percevalian core in spite of the First Continuation’s transfer of the narrative spotlight onto other characters. Thus, despite its lack of illumination, the mise en page of this manuscript provides an instructive example of how the material text can work to draw attention to specific features of narrative. Here, the insistence on marking the change of hero shows that the scribe of E was highly attuned to the cycle’s dynamic of interlace. A number of such moments in the narrative are marked in this manuscript by large initials that have no counterpart in other manuscripts. At C1 [LR] 6607, a space was left for a twelve-line initial to mark the line ‘[I]ci remaint de Bran de Liz / Et de sa seror au cler viz / Qui remaint anceinte d’enfant’ [Here the tale leaves Bran de Lis and his sister fair of face, who is left pregnant with child]. The text, having dropped the news that the Pucelle de Lis has fallen pregnant after her dalliance with Gauvain, returns to camp with the unsuspecting father-to-be to finish off the narration of Arthur’s siege of Branlant, leading into Branch III. The Lis family plot thread will remain suspended until Branch IV, when Gauvain resumes his interrupted duel with Bran de Lis and meets his son for the first time. The scribe of E clearly felt that the announced suspension of a narrative thread should be marked by a large initial, even though a nine-line initial appears less than seventy lines later, at 6671, to mark the opening of Branch III (‘En icel tans que je vos di’ [At this time which I am telling you about]). Another moment from the First Continuation where E alone places a large initial is a transitional passage from the tournament scene to the tale of Ysave and Eliavrés’s revenge on their son in the Caradoc branch. Line 9583 is marked by a seven-line initial: ‘Li rois fu a sejor grant piece’ [The king rested there for a long time]. The reason for this initial becomes clear in the ensuing lines, as the narrator laments:

86 THE

CONTE DU GRAAL CYCLE

De ma dame Ysave sa niece, La mere au vaillant Carados, Ne puis plus metre ariere dos. Mes deslais plus n’i puet monter Qu’il ne me coviegne a conter Tel chose dont molt me desplest  (C1 [LR] 9584–9)21 [I can no longer hold back telling about his niece, Lady Ysave, brave Caradoc’s mother. I cannot put off any longer what I must recount, a thing which greatly displeases me]

Unable to defer the tale of Ysave’s ‘folie’ (9593) any longer, he continues: A mon conte voil revenir Et ma matiere maintenir D’illuec ou devant la laisai.  (C1 [LR] 9613–15) [I want to return to my tale and take up my story from where I had left it.]

The initial thus introduces another switch of narrative subject, as the conclusion of the tournament – presented here as a valiant attempt to put off recounting morally dubious material, rather than as the bafflingly long digression most critics have felt it to be – leaves the way open for the suspended narrative of Ysave’s infidelity to resume. These lines were discussed in Chapter 1 as an instance of the introduction of interlace formulae into the First Continuation by the authors of its Long Redaction, and the placement of a large initial here in manuscript E provides confirmation that this narrative dynamic was understood and assimilated by its scribe, who in turn wanted to ensure that users of the manuscript would also notice and understand it.22 The snatches of interlace embedded in the text of the Long Redaction are thus brought to the fore by this manuscript’s presentation of the text. The tendency intensifies in the later Continuations, where interlace formulae are more integral to narrative structure. For instance, E is again unique in placing a six-line initial at 22225. The White Knight, defeated by Perceval, has arrived and been pardoned at Arthur’s court, and the narrator returns us to the hero’s adventures with the line: ‘Or dirons de Perceval ci’ [Now we will tell of Perceval here]. This narrative sequence – Perceval defeats his opponent, who arrives at court and is welcomed by Arthur, before the narrative returns to Perceval – is the same as that found a few hundred lines earlier, where the opponent in question is Abrioris. As shown in Table I, the switch from Abrioris to Perceval at 21579, effected in much the same terms (‘Or vos diron de Perceval’ [Now we will tell you of Perceval]), is 21 22

Manuscript V also marks this moment with a large (seven-line) initial. See pp. 45–7.

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

87

marked in EQ by a large initial and in P by a small initial. The principle of marking the return to Perceval evidently also guides the scribe of E at 22225; indeed, the absence of initials at this point in the other manuscripts may be attributable to the fact that the reading ‘Or dirons de Perceval ci’ is given only in ES, apparently called into existence by the logic of narrative transition outlined above (the other manuscripts offer variants on ‘Or nos taisons atant de li’ [Now we will cease to speak of him], placing emphasis more on the minor character being left behind than on the reintroduction of the hero).23 The rhetoric of interlace also triggers large initials in the other mostly complete unillustrated manuscript, Q, which similarly places them at lines not marked in other manuscripts. From the First Continuation, for instance, the six-line initial at 10109 (‘Atant ici vos laisserai / Et de Cador vos parlerai’ [Now I will leave (telling) you (about this matter) here and tell you about Cador]) and the eight-line initial at 10535 (‘Ici remaint de Quarados’ [Here the tale leaves Caradoc]) both mark narrative switches of hero during the account of Cador and Arthur’s search for Caradoc after the serpent has attached itself to his arm (III, 12–13). In the Second Continuation, a four-line initial at 25393 marks the line ‘Au chevalier vuel revenir’ [I want to return to the knight], as the narrator leaves Perceval briefly to follow another defeated opponent (the thief of the stag’s head) to court. The lack of attention given to these lines in E, and conversely the failure of Q to mark the lines discussed above, make it clear that we are dealing not with a systematic procedure but rather a general tendency. The mise en page of the unillustrated manuscripts thus gives interlace a prominent role in the division of the text; in each manuscript, the material text is partly structured around a selection of the interlace ‘moments’ given by the narrative text. I argued in Chapter 1 that from the Second Continuation onwards the cyclic dynamic of interlaced quest becomes a major principle of textual organisation, and a glance at Table I confirms that all five manuscripts EPQTV respond to this: from the opening of the Second Continuation, all six lines marked by large initials in at least two manuscripts occur at a narrative switch, involving narratorial intervention of the type we have been discussing in this study. As one might expect, interlace plays a much less frequent role in the placement of large initials in the First Continuation. Instead, large initials tend to be triggered by introductory phrases that define the temporal framework for the material to follow without connecting this narrative sequence back to the episode which has immediately preceded it. The placement of initials thus reflects the important difference we noted in Chapter 1 between the narrative aesthetic of the First Continuation and that of subsequent parts of the corpus: 23 A similar phenomenon is found a number of times in manuscript T. For instance, at CM 33761, T invents an appropriately worded couplet to go with the reintroduction of Sagremor and gives it a large initial: ‘Li contes dist que Saigremors’.

88 THE

CONTE DU GRAAL CYCLE

whereas the interlace of the latter is set within a fairly coherent temporal framework, the former approaches temporality on an episodic level only, preferring to set each episode off clearly from that which precedes it. There is little concern for temporal relation or consistency between episodes. Thus, as shown in Table I, the three lines of the First Continuation marked with large initials in both EQ have in common an introductory function and a lack of specificity about time; ‘A un jor d’une Acension’ [It happened on Ascension] (LR 11949) and ‘A une Pantecouste fu’ [It was on Pentecost] (LR 5385) both rely on the indefinite article to effect this, while ‘Ce fu en mai, el tans d’esté’ [It was in May, in summertime] (LR 7037) evokes a début printanier, the beginning of a new narrative rather than the continuation of an existing one. This small sample of three initials is indicative of another general tendency in the approach to the visual text of the First Continuation in EQ. By far the greatest concentration of large initials in these unillustrated manuscripts occurs in Branch III and the second half of Branch I, which are precisely those parts of the text rewritten most extensively by the Long Redaction. Of ten large initials in manuscript E’s version of the First Continuation, two are found in Episodes 5–10 of Branch I and four in Branch III; for Q, the respective figures are two and five out of a total of eleven. Table II below shows the lines marked by large initials in these sections. Table II Lines from the First Continuation marked by large initials in EQ Manuscript E C1 [LR] v. 1953

Episode Line ‘Ensis li rois pansis chemine’ I,6

5385

I,10

6607

II,7

6671

III,1

7037

III,3

9583

III,9

11949

III,15

12507 13611

IV,1 IV,5

15089

IV,7

Context Arthur regrets Gauvain’s flight; narrative switch to Gauvain is imminent ‘A une Pantecoste fu’ Transition from Guiromelant material into Brun de Branlant story: Arthur assembles his barons to attack Brun ‘Ici remaint de Bran de Lis’ End of Pucele de Lis episode – Gauvain returns to help with the siege of Branlant ‘En icel tans que je vos di’ Opening of Branch III (wedding of Ysave and King Caradoc at Arthur’s court) Arthur holds court at Pentecost and knights ‘Ce fu en mai, au tans d’esté’ Caradoc ‘Li rois fu a sejor grant piece’ End of tournament, now the text will speak of Ysave Arthur holds court; introduces the episode ‘A un jor d’un[e] Acension’ of the golden nipple given to Caradoc for Guinier ‘Ce fu an mai, au tans novel’ Opening of Branch IV (Arthur holds court) ‘Vos savez bien, or a .x. ans’ Gauvain recounts his adventure with the Lis family ‘Seignor, einsint com je vos di’ The battle between Gauvain and Bran de Lis has ended

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

89

Manuscript Q C1 [LR] v. 5281

Episode Line I,10 ‘A la cort est Artus venuz’

5385

I,10

‘A une Pantecoste fu’

6709 7037

III,1 III,3

‘Ce fu par .i. mardi matin’ ‘Ce fu en mai, au tans d’esté’

10109 10535

III,12 III,13

‘Atant ici vos laisserai’ ‘Ici remaint de Quarados’

11949

III,15

‘A un jor d’un[e] Acension’

14759

IV,7

‘Atant s’an cort la damoïlle’

15211

IV,8

‘Il n’orent pas mout sejorné’

17743

V,5

‘Oi avez une partie’

18375

VI,1

‘Icele nuit que je vos di’

Context Arthur arrives at the court of Escavalon to make peace between Gauvain and his two assailants Transition from Guiromelant material into Brun de Branlant story: Arthur assembles his barons to attack Brun King Caradoc marries Ysave Arthur holds court at Pentecost and knights Caradoc Narrative switch from Arthur to Cador Narrative switch from Caradoc to Cador and Arthur Arthur holds court; introduces the episode of the golden nipple given to Caradoc for Guinier The Pucelle de Lis fetches her son to break up the battle between Gauvain and Bran de Lis Arriving at Chastel Orguelleus, Arthur’s men hear a bell ring The Fisher King tells Gauvain about the Grail Opening of Branch VI (Arthur cannot sleep)

The mise en page of EQ is clearly responding to and reinterpreting the narrative dynamic given by the Long Redaction, and a consideration of the lines chosen for decoration demonstrates that it is largely doing so in two distinct ways. One, already discussed, is to highlight those lines which introduce narrative switches or make use of the language of interlace: this is the case in E at C1 [LR] 1953, 6607 and 9583, and in Q at 10109 and 10535. The other, more prevalent, tendency is to mark lines that perform (or appear to perform) the kind of generic introductory function described above. Both manuscripts, for instance, choose to mark the account of Caradoc’s parents’ marriage, though they select different lines as the starting point for this episode.24 In manuscript E, it is King Caradoc’s arrival at Arthur’s court to ask for a wife which is treated as the opening line, a decision followed by Roach, who makes this the beginning of Branch III: En icel tans que je vos di Que li rois iert a Quinilli, Vint a la cort uns chevaliers  (C1 [LR] 6671–3) [At this time which I am telling you about, while the king was at Quinilli, a knight came to court]

24 In order to avoid confusion between the two identically named characters, I shall refer to the father as King Caradoc and the son as Caradoc.

90 THE

CONTE DU GRAAL CYCLE

The manuscript emphasises a certain continuity with the action of the preceding episodes, since it was in the aftermath of Brun’s surrender that Arthur had brought his defeated opponent to Quinilli. This continuity, however, is only loosely defined in relation to time, as we are told that Brun was given his freedom by the king ‘dedanz les huit jors’ [within eight days] (C1 [LR] 6659), but know nothing about how long the king remained there before the arrival of King Caradoc. The new chapter, announced by the large initial of E, is thus anchored at only the most local level into the conclusion of the preceding material. In Q, the moment chosen for decoration is 6709, the opening line of the description of the wedding day: ‘Ce fu par un mardi matin’ [It was on a Tuesday morning]. As noted above, the indefinite article removes any specificity about time that might guide the reader to relate this material to the temporal framework of the text up to this point. Moreover, by choosing to mark this line, Q declines to relate the new section of text to the preceding episode, even at a local level. The narrative dynamic which emerges from the mise en page of the First Continuation in these two manuscripts is noticeably different from that of the later Continuations. The insistence on marking the kinds of opening phrases we have been discussing suggests reading the text as a series of new beginnings, with only minimal diegetic connections to be made between the different episodes. One is reminded of Frappier’s description of this Continuation as ‘un chapelet de lais ou de fabliaux courtois que seule relie la permanence du décor arthurien’ [a collection of courtly lais or fabliaux bound together only by the enduring Arthurian background];25 taking the term ‘décor’ literally, it is clear that the visual coherence of the mise en page does provide a certain unity to these various narrative moments. Yet, while the First Continuation lacks the diegetic and temporal cohesion of the later Continuations, it has been suggested that its coherence can be found in the redeployment of specific themes and images across narrative segments not explicitly related.26 Indeed, I would argue that the unillustrated manuscripts present this continuation as a series of diegetically isolated but analogically related episodes, and in doing so open up a way of understanding the narrative of Caradoc in terms of the overall cyclical aesthetic of the corpus. The Caradoc branch has struck many of its readers as a separate romance parachuted into the corpus. Albert Thompson calls it ‘a conspicuous digression’, arguing that ‘it must have existed originally as a separate work, the Livre de Caradoc’.27 John Grigsby also opts to give the branch its own title, describing it as ‘the complete, harmonious Roman de Caradoc, which could be removed and studied separately on its own merits... with no harm to the

25 26 27

Frappier (1973: 153). See for instance Grant (1986); Leupin (1979) and (1982); Busby (1994b). Thompson (1955: 127).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

91

story line of the remaining narrative’.28 Similarly, Marguerite Rossi claims that ‘l’épisode de Caradoc, par sa structure, contraste avec l’ensemble de la Continuation Gauvain’ [the Caradoc episode, in its structure, contrasts with that of the First Continuation as a whole].29 While the kind of scholarly scalpel-work suggested by Grigsby is indeed possible (and has been done),30 it is clear from our analysis of the mise en page that the scribes of EQ saw the Caradoc material as an integral part of the corpus they were copying. The expansion of the Caradoc narrative to over five thousand lines in the Long Redaction brings it to a length to rival many Arthurian verse romances, and not far off the length of Chrétien’s other texts. Yet the manuscripts respond to this expansion not by emphasising the various stages of Caradoc’s development, but rather by marking the branch itself as a series of new, isolated departures. Instead of presenting us with a ‘complete, harmonious’ romance, EQ offer us two ways of reading Caradoc’s tale as fully integrated into its surrounding text: either as a series of discrete episodes exemplifying the narrative dynamic of the First Continuation, or as a minor part of the interlaced structure of the whole corpus. Interlace is a structuring device shared by both the Conte du Graal cycle and its prose cycle contemporaries. It is therefore not surprising that the manuscripts, in drawing attention to the interlace dynamic of the corpus, should do so in a way that is reminiscent of the discourse of prose cycles. The production of our surviving manuscripts, which date mostly from the mid thirteenth century to the mid fourteenth, is contemporaneous with the composition and widespread manuscript transmission of the major prose Grail cycles. The popularity of both cycles makes it likely that the workshops in which our manuscripts were created would also have produced prose cycles. For instance, Alison Stones has identified a cluster of Parisian manuscripts illuminated by the same artist in the early fourteenth century (probably c. 1315–20), designated as the Master of fr. 1453 for his work on this codex, our manuscript S; the same artist worked on manuscripts of the Estoire del saint Graal and the prose Tristan, as well as other items such as the Guillaume cycle and the Ovide moralisé.31 This glimpse into fourteenth-century manuscript production suggests that there was ample scope for scribal and presentational practices from Arthurian prose and verse cycles to inform and affect each other. The process is apparent in the rubrication of manuscripts PU, which in its mediation of the cyclical text evinces two related characteristics: a sympathy with the interlace dynamic of the narrative and an adoption of the spatial Grigsby (1988: 384). Rossi (1980: 247). 30 See for instance Brodman (1991), McCracken (1998: 135–43) and Lacy (2001), all of which treat the ‘Livre de Caradoc’ as an independent text. 31 Stones (1993b: 264–5). 28 29

92 THE

CONTE DU GRAAL CYCLE

language of the prose tradition. Thus, of forty rubricated historiated initials in P, eight make use of a formula along the lines of ‘Ci vient li contes a…’ [Here the tale comes to…]. The impersonal model of narration implied by this formula, whereby the role of moving forward the narrative is ascribed to the tale itself, is evidently closer to that displayed in the prose romances than to the preference for strongly personalised narrator- and author-figures found in the Conte du Graal cycle. At the same time as they undermine the performative flourishes of the narrative, however, the rubrics work to reinforce the dynamic of interlace which we have identified as so important to the coherence of the corpus. The formulae followed in these rubrics suggest a network of points at which the story is always returning to the hero, translating the narrative logic of interlace into the spatial framework of the manuscript page. In five of the eight ‘Ci vient li contes a’ cases in P, the rubric carries interlace information even though the diegesis makes no provision for having left the hero in the first place: the historiated initials at CdG 4603, C1 [SR: A] 1575, C1 [LR] 8035, C2 21909, 22637 mark a continuity, rather than a switch, in the identity of the central hero, whereas the rubrics insist on interlace movement as the initiating moment of the narrative sequence.32 As we saw in Chapter 1, in the cyclical dynamic of this corpus the notion of a narrative return to the hero is often bound up with the hero’s return to a location he has previously visited. This aspect of the cycle’s textuality is also acknowledged in these rubrics. At CdG 4603, the initial which begins the Hideous Damsel episode is rubricated thus: ‘Ci vient li contes a Percheval. Ensi come il revint a court dont il se fu parti’ [Here the tale comes to Perceval. How he came back to the court which he had left]. The rubric doubles the tale’s return to Perceval with the hero’s own return to Arthur’s court. Yet Perceval’s only previous experience of Arthur’s court was the briefest of visits at the beginning of his career; his arrival here is presented somewhat misleadingly as the end of a circular journey, a predestined return. The distortion is apparently motivated by the rubricator’s awareness of the looping structure of Perceval’s adventures in other parts of the corpus. The rubric fits the narrative movement more accurately at C2 22637, where Perceval’s first return to Biaurepaire is labelled: ‘Ciendroit vient li contes a Piercheval. Ensi come il vint au chastiel de Biel Repaire’ [Here the tale comes to Perceval. How he came to the castle of Beaurepaire]. Meanwhile, the rubric to 23121 (which does mark genuine interlace movement in the text) places the accent on the open quest, another device which I identified in the previous chapter as key to the corpus’s cyclic structure: ‘Ci vient li contes a Piercheval. Ensi come il

32 P’s version of the First Continuation contains a small portion of the Long Redaction during Branch III, which Roach prints as an Appendix at the end of his edition of the Short Redaction (1952 [III, 1]). The line number given here as C1 [LR] 8035 is v. 501 of his Appendix.

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

93

ert en la queste dou brachet’ [Here the tale comes to Perceval. How he was hunting for the brachet]. In U, all fifty-two miniatures are marked by rubrics beginning ‘ci devise’ [Here (the tale) tells of] or similar; the manuscript is again invoked as a spatial structure consisting of a series of discrete areas (‘ciendroit’ [in this place]), within each of which a particular narrative segment occurs.33 As with manuscript P, the implied subject of the verb ‘devise’ is the selfnarrating tale, as is made clear in the opening rubric: ‘Ci conmence le romans de Perceval le Galois et devise de moult de aventures qui li avindrent et conment il conquesta les armes vermeilles’ [Here the romance of Perceval the Welshman begins and tells of many adventures that happened to him and how he won the red arms]. There is however a significant difference between P and U, in that the latter appends far more extensive narrative summaries to the lines it is marking. For instance, the miniature representing Gauvain’s success in the challenge of the Marvellous Bed is introduced at CdG 7817 in U by the following rubric: Ci devise conment mesire Gauvain se coucha ou Lit Perillieus l’escu au col et conment on traioit a li saietes qu’il ne savoit dont il venoient. Et estoient fichiees en son escu. Et apres ce .i. lyon issi d’une chambre a qui il se combati et tant avint que le lyon fu ocis et demoura .i. de ses piez dehors l’escu et l’autre par dedens. [Here (the tale) tells how Sir Gauvain lay down on the Perilous Bed with his shield around his neck and how arrows were fired at him without him knowing from whence they came. And they were stuck in his shield. And after this a lion came out from a room and he fought it, and in the end the lion was killed and one of its feet stayed on the exterior of the shield and the other was stuck through it.]

The level of detail in this rubric allows it to mention the two separate parts of Gauvain’s feat – weathering the storm of arrows and defeating the lion – where the equivalent rubric (also placed at CdG 7817) in P simply states: ‘Ensi come Gauwains sist ou lit de la Merveille’ [How Gauvain sat on the Marvellous Bed]. This expansion of the rubric corresponds to an expansion in the illustration. The material text has taken on more ample proportions, such that it represents an alternative means of experiencing the events of the narrative.34 Hindman, in a discussion which focuses especially on U, argues 33 The Rappoltsteiner Parzifal also displays this feature of mise en page, with almost all of the rubrics being cast in spatial language; for instance: ‘hie het Kador Karadossen funden’ (=C1 [MR] 7495); ‘hie kumet künig Artus für kastel Orgalus’ (=C1 [MR] 11206). 34 A useful modern analogy for this kind of multi-level text is provided by the stories of Rupert Bear. The pictures are accompanied by three distinct kinds of text: a single-line page-header summarising the narrative thrust of the page; a two-line rhyming couplet accompanying each picture; and a prose narrative running along the bottom of the page,

94 THE

CONTE DU GRAAL CYCLE

that these prose rubrics turn the verse narrative into an ‘outmoded referent to the original text rather than... an integral part of the reading process’.35 Quite apart from the awkwardness of explaining away the transcription of fifty thousand lines of text as a mere ‘outmoded referent’, this argument fails to do justice to the degree of ambivalence displayed by the manuscripts. As I have suggested above, they are rather caught between verse and prose, anchored in and responding to the performative effects and personified interlace of the narrative text, yet influenced in the manner of their transmission by the prose cycle manuscripts with which their scribes were doubtless familiar. If verse had really been superseded by prose for the creators and users of these manuscripts, then the manuscripts themselves would not have been required, and would certainly not have survived; there was, after all, no shortage of Grail material in prose for scribes to copy. The Conte du Graal cycle survived, then, because there was interest in reading and transmitting its thousands of lines of verse. The infiltration of the aesthetics of Arthurian prose manuscripts into the manuscript tradition of our cycle has consequences for the way in which the codex suggests its user might apprehend the text and commit it to memory. The detailed rubrics we have seen in manuscript U, and the content selected for illustration, guide the reader’s response in different ways from the large initials given pride of place in the unillustrated manuscripts. As the following section will suggest, the material text created by the extensive illustration and rubrication of the later manuscripts places less emphasis on structure and aesthetic, and more on narrative content.

‘Ci conmence le romans de Perceval’: Illustration and Biography The expanded rubrication of manuscripts SU relative to P goes hand in hand with a concurrent expansion in the size of illustration used. In the example discussed above, the historiated initial of P, depicting Gauvain sitting on which provides a more extensive treatment of the material. As with our manuscripts, different types of reading become possible, and the differing stylistic constraints and possibilities of verse and prose guide the users of the text towards particular aesthetic experiences. 35 Hindman (1994: 194). It should be pointed out that Hindman does consider, as I do, that these manuscripts are ambivalent in their combination of verse and prose codes; where I disagree is in the emphasis she places on conflict between the two, seeing the expansion of the material text in U as a sign of ‘disquietude’ with the status of verse. As I hope to have shown, there is a more complex and less conflictual interplay going on between performative and written modes of discourse, with codes found more commonly in prose also in evidence in both the material and diegetic texts of this cycle. As Hindman herself asks (123): ‘Couldn’t we… see the persistence of verse romance as evidence that, serving the same function, it met some of the same needs that gave rise to prose historiography, with which it was therefore compatible instead of at odds?’

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

95

the bed, is by its format unable to carry the same amount and complexity of information as the double-column miniature of U, which faithfully depicts the two incidents mentioned, one on each side of the page. A comparison of the Conte du Graal cycle manuscripts allows us to track a development in the complexity of visual information from the thirteenth to the fourteenth centuries, as planners and makers of manuscripts began to give over more space on the page to illustration. In rough chronological order, we can see three stages exemplified by our manuscripts: large initials but no illustration (EQ); an intermediate stage, represented by M (miniatures concentrated in the first half of the text, with rubrics added by a later hand), P (historiated initials and rubrics) and T (historiated initials and miniatures, but no rubrication); and a final stage seen in SU (longer rubrics, two-column miniatures, disappearance of large initials). An important consequence of this gradual dissociation of miniature from text is that images no longer function primarily as structural markers, achieving a certain autonomy as a visual narrative in their own right. Within this visual narrative, the opening miniature carries particular weight. Usually larger and more detailed than those which follow, it is likely to be the first aspect of the text to catch the reader’s attention. In our fourteenth-century manuscripts (SU), most of the opening page is given over to this opening image. That of manuscript U leaves room for only six lines of text below it, while Terry Nixon has calculated that a miniature probably occupying twentyone out of thirty-six lines over two columns adorned the missing first folio of S.36 The opening miniatures of our earlier manuscripts are more modest. That in M is barely bigger than the manuscript’s other miniatures, occupying twelve lines of the left-hand column; P opens with a ten-line miniature which spans both columns (the only double-column image in this manuscript); T, meanwhile, has a half-page miniature taking up two of its three columns (see Figures 3–4, pp. 100 and 102, for the opening miniatures of TU). The amount of the opening page given over to words thus shrinks over time, as the expansion of the visual text eats up the space available. Perhaps the best illustration of the increasing importance of the visual at the outset of the text is provided by manuscript Q. The first folio of this thirteenth-century manuscript must have gone missing quite early, since it was replaced in the fourteenth century by an updated folio, of which only a stub now remains. Whereas Q, as we have seen, is unillustrated, the fourteenth-century folio contained a decent-sized opening miniature which has been almost entirely lost. This fourteenth-century folio apart, the differences separating the material texts of EQ from those of SU speak of a change in the conception of the manuscript’s function which is not unconnected to the move from a performative to a writerly poetics discussed above, with

36

Nixon (1993b: 77).

96 THE

CONTE DU GRAAL CYCLE

the text increasingly becoming identified with the book. The manuscripts of the Conte du Graal cycle thus bear witness both textually and materially to a significant shift in the relation between text, manuscript and audience: within the text, Chrétien’s entreaty to the audience to judge his text as an oral performance is replaced fifty years (and thousands of lines) later by Manessier invoking his narrative as a book and inviting his audience to see the evidence of his narrative for themselves. Similarly, the impressive presentation and more developed material texts of our fourteenth-century manuscripts suggest that they were increasingly seen as textual objects in their own right, rather than as aids to performance, and that the passing of the story into cultural memory could now legitimately be effected by private readers (either in groups or individually) for whom possession of the text involves possession (and display) of the book. To view the book, as Manessier’s conclusion emphasises, is to be in the presence of the textual ‘truth’; where Chrétien’s prologue asks the audience to evaluate the narrative aurally, Manessier’s epilogue posits visual apprehension as the means by which authorisation of the text is to be achieved.37 Within this transition, manuscripts MPT in their different ways represent an intermediate stage. This is signalled primarily by a certain amount of inconsistency. Manuscript T, for instance, alternates between using historiated initials and miniatures, which has led Hindman to describe it as being ‘at a crossroads of communication’.38 Manuscript P carries rubrics which vary greatly in the level of narrative detail which they provide (and hence in the extent to which they supplement the text). Certain rubrics give specific information which would allow a reader immediately to identify the relevant passage: ‘Ensi con mesire Gauvains se combati a Giromelant le fier’ [How Sir Gauvain fought proud Guiromelant] at C1 [SR: A] 797, for instance, or ‘Ensi come Bagomede apiele Kes le seneschal de traison’ [How Bagomedés accused Keu the seneschal of treachery] at C2 28649. At other times, the rubric adds little to the information contained in the image, as in the following from CM 33789: ‘Saigremors ki se combat a .i. chevalier’ [Sagremor fighting a knight]. In these cases, the rubric’s function as an aid to memory is weaker, and the reader is called upon to provide the contextual detail to identify the scene and make the material text comprehensible. A similar example occurs at CM 36375: ‘Ensi come mesire Gavains se giut en 37 Reis (2010: 386) similarly argues that the development of paratextual elements in the codex would have facilitated its visual reception: ‘As the text and page layouts of manuscripts became more reader-friendly, the reader of Chrétien’s thirteenth-century manuscripts evolved from an aural reader to a more visual reader.’ 38 Hindman (1994: 45). Hindman incorrectly claims that historiated initials are replaced wholesale by miniatures in the middle of Gerbert’s Continuation (there is more interpenetration than this: for instance, the first and fourth images are miniatures and the penultimate image is a historiated initial). However, this fact would tend to support, rather than undermine, several of her conclusions.

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

97

la forest’ [How Sir Gauvain slept in the forest]. The number of times that characters are forced to sleep in the forest makes this rubric fairly useless as a bookmark; it merely reproduces the information contained in the image, of Gauvain lying under a tree. The material text here prefers to draw attention to a certain narrative aesthetic, underlining the importance of the day/night succession of knightly wandering to the structure of the diegesis. We are thus closer to the function and significance of the large initials in the unillustrated manuscripts. An even more striking instance of this is the series of three historiated initials found towards the end of the Second Continuation: at C2 29209, the rubric ‘Mesires Gavains’ introduces an image of Gauvain riding a horse; at C2 31421, ‘Piercheval’ is the rubric accompanying a similar depiction of Perceval on horseback; finally, an initial showing Perceval riding up to a castle at C2 32265 is annotated ‘Pierchevaus li Galois’ (see Figure 1, p. 75). These rubrics, little more than labels, are matched by images of a similarly unspecific nature. Carrying no narrative content, they draw attention away from specific moments in the text and towards its general interlace dynamic, less explicitly but just as effectively as the ‘ci vient li contes a...’ rubrics discussed above. If, as was often the case, the rubrics functioned as instructions to the illustrator, the lack of engagement with diegetic detail was bound to produce images whose generality offers little help to any reader lost in the forest of the heroes’ adventures. Underlying this variation in the function of the rubrics is a sense of ambivalence on the part of the illuminators to their work. A greater concentration of the less specific labels and images occurs in the latter part of the text, as if the sheer size of the narrative encouraged a change in their engagement with its detail. The sense of occasional disengagement is heightened by three instances where the rubrics and illustrations are at odds with the text. At C2 24241, the rubric ‘Ensi come Perchevaus jua as eschas et il le materent’ [How Perceval played against the chess pieces and was defeated by them] fits the picture of Perceval playing chess, yet it comes several thousand lines after the relevant episode, and instead introduces his arrival at the Castle of Maidens. The other two instances are even more telling. At C2 [SR] 9457 a generic picture of a knight riding up to a castle is rubricated with the label ‘Ensi que Percheval vint a la court le riche Roi Pescheour’ [How Perceval came to the court of the rich Fisher King] when it is in fact at the Chastel dou Cor that our hero is arriving. At CM 32807, meanwhile, the rubric ‘Ci raconte Pierchevaus au roi Artut dou Graail’ [Here Perceval tells King Arthur about the Grail] accompanies a miniature of a knight in conversation with a king; yet in the text it is the Fisher King who is telling Perceval about the Grail. The rubricator appears to have lost his way in the succession of similar episodes; apparently, invoking the Grail adventure was his way of hedging his bets. In doing so, his actions draw attention to the importance of this recurring scene in the text’s structure. The rubrics of manuscript M, added by a fourteenth-century hand at the

98 THE

CONTE DU GRAAL CYCLE

bottom of each page containing a miniature, clearly do not have the same significance as those in P. Their function can hardly be said to be decorative; rather, they are supplementary, intended to augment the material text constituted by the programme of illustration. As Angelica Rieger has argued, these rubrics are generally more accurate in their representation of narrative events than their corresponding miniatures, displaying a concern for harmonising text and image.39 This feature of M allows us to glimpse the transition from a thirteenth- to a fourteenth-century conception of manuscript function from another angle: a fourteenth-century user of this manuscript has updated its material text to make the codex a more extensive and cohesive textual object, in the way that SU undoubtedly are. In their different ways, MPT all bear witness to an ambivalent stage in this transition.40 Another paratextual feature which testifies to the development of the codex as a textual object is the nature of its programme of illustration. Manuscripts MT, like P, contain a noticeable concentration of generic images: knights fighting, riding and reaching castles, or a court seated for a meal (a favourite of M). The images in these manuscripts are also generally fairly static, with characters depicted in what Hindman (discussing manuscript P) has called ‘stock hieratic’ poses.41 By contrast, the miniatures of SU tend to present more detailed and dynamic scenes.42 Both manuscripts feature fiftytwo images, which is considerably more than any of the earlier manuscripts with the exception of M, which has fifty-five, though these are much smaller and less detailed than those in SU; their visual texts are thus able to incorporate a greater portion of the diegesis. As Nixon (1993b: 75) notes, the twocolumn miniatures of U in particular ‘are almost always subdivided so that the opportunity for extensive representation of action is sacrificed in favor of more scenes’. These miniatures usually depict two scenes each, allowing a greater amount of the narrative to find its way into the visual text. S, meanwhile, squeezes a remarkable amount of detail into its miniatures, with more figures involved in each image than in any other manuscript. Indeed, an especially striking feature of the visual text of S is the recurrence in around half of the miniatures of characters observing, rather than participating in, the action of the miniature. This is particularly noticeable in cases where the position of the observing character is in contradiction with the text itself. Thus, in the miniature at C1 [SR: A] 1043, Guiromelant watches from the battlements as Clarissant beseeches Gauvain to allow her to See Rieger (1993). One should note, however, Keith Busby’s recent assertion that the rubrics of M are contemporary with the rest of the manuscript, and may be instructions to the illuminator. See Busby (2002: 346). 41 Hindman (1994: 117). 42 On this issue I disagree with Hindman (1994: 196), who implies that the miniatures of SU are what she terms ‘prose miniatures’, ‘often static, whereas those in verse texts are more directional’. 39 40

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

99

Disclaimer: Some images in the printed version of this book are not available for inclusion in the eBook. To view the image on this page please refer to the printed version of this book.

Figure 2.  Paris, BNF, fr. 1453 (MS S), fol. 72v. C1 [SR: A] 1043. Guiromelant watches from the ramparts as Clarissant pleads with her brother Gauvain to give permission for their marriage. Rubric: ‘Coment Clarissent prie a son frere que il li doinst le Guiromelant’.

marry him (see Figure 2). In the text, she has just run onto the field of battle where both knights are fighting, giving no textual basis for Guiromelant’s position in the picture. Similarly, in the miniature at C2 21873, Perceval kills a giant while a gaggle of maidens looks on from the castle ramparts; in the text, there is only one damsel concerned. One could multiply these examples; the miniaturist’s preferred technique seems to be to use a character involved in the relevant episode but not directly involved in the action being represented, as in the Guiromelant episode above, or at CdG 9227, where Arthur watches Lore telling the queen that he has just fainted in the court below. The selection of this moment for depiction is symptomatic of manuscript S’s taste for scenes involving the act of witnessing. Lore has, after all, been sitting ‘En unes loges, si veoit / Le doel qu’en fist parmi la sale’ [in a gallery, and saw the grief being manifested in the hall] (CdG 9228–9), as emphasised by the rubric: ‘Coment dame Lore fu esbahie pour le duel qu’ele vit en la

100 THE

CONTE DU GRAAL CYCLE

Disclaimer: Some images in the printed version of this book are not available for inclusion in the eBook. To view the image on this page please refer to the printed version of this book.

Figure 3.  Paris, BNF, fr. 12576 (MS T), fol. 1r. Opening miniature taking up the upper quarter of the page. Top half: Perceval, at the door of his maternal home, kneels before the knights in the forest. Lower half: Perceval speaks to Yvonet at Arthur’s court (left), then kills the Scarlet Knight (right).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

101

sale’ [How lady Lore was dismayed by the grief that she saw in the hall]. In total, seven rubrics make use of the vocabulary of vision. Indeed, Perceval’s double witnessing of the bizarre events in the Chapel of the Black Hand is deemed sufficiently noteworthy to receive two, almost identical, miniatures with very similar rubrics: ‘Coment Perchevalz vit en la chapele .i. cors gisant sus l’autel un cierge ardant et une main noire’ [How Perceval saw in the chapel a body lying on the altar, a burning candle and a black hand] (C2 32133) and ‘Coment Perchevalz vit le cors sus l’autel et le cierge ardant et le main noire et le diable qui descendoit ardans de vers le ciel’ [How Perceval saw the body on the altar and the burning candle and the black hand and the devil on fire descending from the sky] (CM 37253). Tellingly, despite their evident visual potential, S is the only manuscript to represent both these scenes; of the others, U devotes a miniature to Perceval’s arrival at the chapel in the Manessier Continuation and another to the aftermath of his victory over the Devil, while T does show the confrontation with the Black Hand in a miniature at the start of the episode (CM 37141).43 That S marks them both betrays its interest in the process of visual testimony and, I would suggest, its awareness of the manuscript (and its material text in particular) as a visual witness to the narrative, a memoire for the estoire. As a visual memory of the narrative, the illustrations in our manuscripts generally opt to portray Perceval’s adventures. For instance, Chrétien’s part of the text allots slightly more lines overall to Gauvain than to Perceval, yet the distribution of miniatures is hugely skewed in favour of the latter: twenty out of twenty-five in M, seven out of eight in P, nine out of fifteen in S, five out of six in T and six out of eight in U. The effect of this is to relegate the adventures of Gauvain to the status of digressions, not worthy of special attention in the material text, and thus confirm the rereading of Chrétien’s part of the narrative effected by its incorporation into cyclical manuscripts with the Second, Manessier and Gerbert Continuations. The opening miniatures of TU similarly set the tone for the development of the visual narrative, depicting the early adventures of Perceval. The half-page image of T (Figure 3) represents three scenes: the top half is given over to Perceval kneeling before the knights in the forest, while the bottom half is split between his arrival at Arthur’s court and his killing of the Scarlet Knight. Meanwhile, the near-full-page spread of U (Figure 4) includes four scenes: beginning with Perceval riding out for the day from his mother’s house at the start of the text, it then moves to Perceval on his knees before the knights in the forest; in the bottom half of the picture, we see him leave his mother for good (as she falls to the floor), followed by the killing of the Scarlet Knight. 43 In this way, T is able to mark both the interlace dynamic (CM 37141 is the point at which the narrative switches from Gauvain to Perceval, with the words ‘De Gauvain ici vos lairons’ [Here we will leave off telling you of Gauvain]) and the visual potential of the text at this point.

102 THE

CONTE DU GRAAL CYCLE

Disclaimer: Some images in the printed version of this book are not available for inclusion in the eBook. To view the image on this page please refer to the printed version of this book.

Figure 4.  Paris, BNF, fr. 12577 (MS U), fol. 1r. Opening miniature taking up most of the page. Top half : Perceval takes leave of his mother for the day (left), then kneels before the knights in the forest (right). Lower half : As Perceval leaves for good, his mother falls in a faint at the door to the family home (left); we then see him killing the Scarlet Knight (right).

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

103

In both cases the key event is clearly held to be the transformative encounter in the forest which awakes in Perceval the desire to become a knight. In order to underscore this, U has Perceval kill his opponent with a lance, rather than the javelins he is wielding in the other parts of the picture (and that he actually uses in the text); T, meanwhile, includes a second, smaller miniature further down the page (after the prologue), which though badly rubbed may well be a repetition of the scene of Perceval kneeling before the knights.44 The next image in both manuscripts will depict our hero wearing armour and wielding a sword, demonstrating visually his new-found knightly status. The first interaction of either of these manuscripts with a reader is therefore to cast what is to follow as the story of Perceval’s education in chivalry and gradual ascension to the rank of best knight in the world. Again, this is the trajectory which in Chapter 1 we saw imposed on the text by the narrative agenda of the Second Continuation. These images provide a good example of how the visual memory of manuscript miniatures functions. Divorced of any primary bookmarking role, their memorial role is to provide this kind of visual gloss to the diegetic text. Through the selection of content, they create a supplementary narrative which mediates the reader’s experience of the diegesis. Of particular interest to this study is the creation of image series which serve to promote certain of Perceval’s adventures as forming the narrative core of the text. To remember this cycle, these miniatures suggest, is to remember the key stages of Perceval’s story. In manuscript U, the series of important returns in the Second Continuation to the environments of the past – Biaurepaire, Perceval’s mother’s home, the Fisher King’s castle – all receive illustration, as do (within the same narrative sequence) the episodes in which Perceval’s election as world’s best knight is confirmed: the Glass Bridge, the tournament at Chastel Orguelleus and the incident with the knight under the marble slab. In manuscript S, at roughly the same point in the text, a series of consecutive pictures illustrates Perceval’s encounter with the child in the tree, his adventure at the Chapel of the Black Hand and his viewing of the broken sword at the Grail Castle. All three of these episodes generate questions which Perceval puts to the Fisher King at the end of this narrative sequence. The notion that Perceval’s Grail-related curiosity represents a key feature in the unfolding of the story raises a related issue. The rubrics and illustrations may suggest that the text be read as a ‘roman de Perceval’ [romance of Perceval], but the prologue of course describes the text as the ‘conte

44 A figure can be made out standing and looking down; in the context of the above discussion, it appears more plausible that this is the leader of the knights looking down at Perceval, rather than Perceval himself looking down while saddling his horse as suggested in the index of illuminations in Les Manuscrits de Chrétien de Troyes (Busby, Nixon, Stones and Walters 1993: 275).

104 THE

CONTE DU GRAAL CYCLE

del graal’ [tale of the grail].45 The text accommodates this ambivalence by making the Grail Castle the principal connecting node in the network of Perceval’s wandering, and our hero’s ultimate worldly fate, to take up the Fisher King’s crown, successfully fuses the two themes. This balancing act in turn is reflected in the illustrations. Manuscripts ST both fail to illustrate any of the Grail material until late in the text; by introducing this material into the visual text at such a late stage, they associate the Grail with the final stages in Perceval’s development (conversely, they decline to include any picture associating Gauvain with either the Grail or the Chapel of the Black Hand). More, the treatment of the Grail itself in the illustrated manuscripts casts it as the culmination of this narrative: with the exception of U, no manuscript depicts the Grail itself until the final images; where the Grail procession is rendered, other elements such as the lance or broken sword are included instead.46 The visual texts of these manuscripts thus behave less like the Conte du Graal cycle itself, where seeing the Grail is unproblematic (the challenge is to understand what has been seen), and more like the Queste del Saint Graal from the Vulgate Cycle, where the ultimate goal of the quest is to see the divine object. The reader is treated like the elect Arthurian knights, forced to wait until the end of the tale for what he or she ‘tant [a] desirré a veoir’ [wanted so badly to see].47 Perhaps the most original version of this technique is in manuscript T, where none of the Grail scenes are represented visually. Instead, after the final words of the text, an extra-diegetic image has been included of Perceval kneeling before the Grail, as an angel reaches down to carry it off to the heavens. That there is no direct textual basis for this image speaks of the exceptional status which the illustrator gave to the Grail, saving a place for its depiction after the conclusion of the narrative. Through this image, the manuscript is able to suggest that the ‘roman de Perceval’ is also the ‘conte del graal’.

45 As Bruckner (2009: 10) notes, this duality in medieval naming of the corpus is reflected in a modern critical hesitation between two titles for Chrétien’s romance; in both cases, the choice is not entirely anodyne, since ‘the two titles – Le Conte du Graal and Perceval – tell us rather different things about this romance and how it is viewed’. 46 U, surprisingly, does things in reverse, illustrating the Grail at every opportunity only to leave it out of the final miniature of an angel descending between Perceval and Hector, despite a rubric that reads: ‘Ci devise conment Perceval et Hector se furent tant conbatu que il cuidoient bien morir… et .i. angre vint atout le saint Graal qui les conforta’ [Here (the story) tells how Perceval and Hector had fought so long that they both thought they would die… and an angel came down with the Holy Grail and gave them strength]. It is as if the Grail (figured as a ciborum) is treated in this programme of illumination as a cipher for divine presence, and thus rendered superfluous by the presence of the real thing. 47 See Queste, p. 253, v. 14; p. 277, vv. 31–2; p. 278, vv. 9–10.

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

105

‘Ci fenist de Perceval le Galois’: Models of Completeness Our interrogation of the material texts of the Conte du Graal cycle manuscripts has uncovered two centres of coherence underpinning the codices’ organisation of textual material. The illuminations in the illustrated manuscripts tend to promote the notion of the corpus as a ‘roman de Perceval’, whereas the placement of large initials, especially in the unillustrated manuscripts, more strongly underlines the interlacing of different narratives featuring different heroes. As I suggested in Chapter 1, the agenda set by the Second Continuation subjugates the cycle’s interlace dynamic to its biographical drive, with adventures promised to other characters eventually drifting into Perceval’s narrative. Yet the interlace dynamic does not disappear, nor is it entirely subsumed by Percevalian biography. Indeed, it makes a partial return in the Manessier Continuation, where a new hero, Sagremor, is introduced to share space on the page with Gauvain and Perceval; the adventures of the first two, along with a self-contained narrative involving Lionel and Boort, are allowed to defer and interrupt the development of Perceval’s story to a much greater extent than in the Second Continuation. Similarly, Gerbert makes room for an extended sequence centred on Tristan, as well as including the usual portion given over to Gauvain’s adventures. The two principles of biography and interlace thus interact in a way that is more complementary than contradictory, suggesting that cyclic coherence can be generated in a number of simultaneous ways; the corpus can be at once a ‘Conte du Graal’ and a ‘roman de Perceval’. Moreover, in such a long and varied text it is not surprising that interlace and biography do not exhaust the possibilities for generating structure or meaning. Neither of these narrative dynamics can account for what we have already discovered about the quite different aesthetic in the First Continuation. The remaining part of this chapter will consider how alternative centres of coherence are proposed by individual manuscripts of the cycle. The inclusion in some codices of texts which are marginal to the tradition, I will show, allows us to consider different ways of reading the cyclical material. Finally, I will return to the issue of the ambiguity of the corpus’s central focus by way of a consideration of the explicits and incipits found in the manuscripts. While interlace continues to operate as a narrative dynamic, the paratextual frames created by these opening and closing rubrics overwhelmingly argue for biography as the centre for the cycle’s completeness and coherence. The Elucidation and the Bliocadran, thought to be roughly contemporary with the later Continuations, are generally referred to as prologues to the Conte du Graal. In terms of cyclification, there is nothing surprising about the addition of prequels to an existing cyclical mass. The Mort Artu ends with the assertion that ‘aprés ce n’en porroit nus riens conter qui n’en mentist de toutes choses’ [afterwards no one will be able to tell anything more without

106 THE

CONTE DU GRAAL CYCLE

every word being a lie],48 a message which the later Estoire del Saint Graal and Merlin parts of the Vulgate Cycle assimilated by recounting the events that preceded the events of Arthur’s reign. Several chanson de geste cycles similarly recount the enfances (childhood adventures) of their central heroes, written to satisfy the question: ‘What came before?’ When the accretion of material in a forward direction becomes problematic (because of an insurmountable narrative obstacle, such as a death, or simply because of the sheer amount of narrative baggage that has accumulated), the past can offer untapped territory for additional narrative. What is notable is the evident difficulty which the Conte du Graal prologues had in achieving an equal status with the Continuations as part of the cyclical core. Compared with the prequels to other cycles, these texts appear as afterthoughts, dwarfed by Chrétien’s text and its Continuations: the Elucidation is under five hundred lines long and the Bliocadran barely eight hundred lines, and of the fifteen Conte du Graal manuscripts, the Bliocadran appears in two (LP) and the Elucidation in just one (P). The Elucidation has long been a difficult text for critics to assimilate, considered of little value even in relation to the Continuations. Its editor describes it as ‘pedestrian in style, confused in motivation, and… garbled in transmission’.49 More recently, Norris Lacy has warned that ‘[r]eaders who know Chrétien’s uncompleted Grail romance will… find the Elucidation perplexing because it offers information not included in Perceval, and some passages simply contradict Chrétien’s romance’.50 For instance, the Fisher King’s castle is first introduced as the object of a concerted search by Arthur’s knights, rather than being the familial destiny stumbled upon by Perceval; even more strikingly, we are told that Perceval successfully asks the Grail question (here rendered as ‘de coi li greaus servoit’ [what the grail served], Elucidation, v. 248), though this is not sufficient to heal the Fisher King and his lands. The critical comments on the value of this short text are remarkably similar to some of those we have seen aimed at the First Continuation in the past – accused of abandoning Chrétien’s narrative programme, introducing contradictory material and lacking clear structure. I would suggest that we could best understand the Elucidation as a text that inscribes itself within the centrifugal aesthetic which we saw characterised the First Continuation’s Short Redaction in particular.51 The clear inconsistencies with and additions to Chrétien’s narrative found in that text – such as the rewriting of the Grail ritual and the introduction of the sword test – are similar to those in the Elucidation. Moreover, details taken from the Continuations, such as the broken Mort Artu, p. 263. Thompson (1959: 207). 50 Lacy (2007), at http://www.library.rochester.edu/camelot/elucidation.htm [last consulted 17 September 2011]. 51 I pursue this argument in more detail in Hinton (2011). 48 49

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

107

sword and dead knight in the Grail procession, are themselves reworked in the Elucidation’s account. Perhaps the clearest statement in the text of this centrifugal aesthetic project is its invoking of seven guards, who will relate the seven branches of the story (and who correspond to seven occasions that the Fisher King’s court was found). The narrator titillates his public with the names of these seven branches, in a way reminiscent of the narrator of the First Continuation evoking the narratives which he does not have room to tell: some of these are cryptic in the extreme (the Story of the Great Sorrows, the Adventure of the Shield). What is suggestive for our discussion of cyclic coherence is that it is the Grail itself which is described as the source for the textual material (Elucidation, v. 382). The centrifugal model is, by definition, one without a centre, but it derives a certain unity from its originary point. By positing the Grail in this position, the Elucidation attempts to represent the cycle it introduces as one or several ‘tales of the Grail’. In contrast to the Elucidation, the Bliocadran sets itself one specific narrative task: to recount the events leading up to the opening of the original narrative. Chrétien’s text begins with the introduction of the figure of the ‘fix a la veve dame / De la gaste forest soutaine’ [son of the widowed lady of the isolated Waste Forest] (CdG 74–5); the Bliocadran undertakes to explain what the ‘gaste forest’ is, what Perceval is doing there and why his mother is a widow. The last of these (and to some extent, the first) was already explicated in the early part of Chrétien’s text, but the author of the Bliocadran clearly saw enough scope in the subject for significant narrative development. In doing so, he created a text whose addition to the cyclic core is based on a genealogical model of cyclicity. This kind of cyclic model is found in a number of chanson de geste cycles, and operates in the Vulgate Cycle through the transferral of the identity of Grail knight from Lancelot to his son Galaad. Yet it does not sit especially well with the strong insistence of the Conte du Graal cycle on Percevalian biography as the centre of narrative coherence. In contrast to the case of Guinglain in the Bel Inconnu, who is brought up in similar ignorance of his identity, Perceval’s paternity is not made a crucial issue by the text; in fact, after the opening episode no reference is made to his father at all.52 The important family members whom he meets (the hermit uncle, the Fisher King) are related to him maternally. The effect is to present Perceval as a self-made knight, one whose chivalric identity is defined by his deeds. No doubt he inherits his abilities from noble stock on both sides, but his qualities are not comparable to his father’s own prowess, as is the case for Guinglain. The Bliocadran, by foregrounding 52 Wolfgang (1980) analyses the motif of Perceval’s father in Arthurian romance, finding it ‘amazing’ that, the Bliocadran aside, the hero’s paternity largely fails to interest the continuators (29). See Chapter 3 for a more lengthy consideration of the treatment of Perceval’s filiation by the authors of the cycle.

108 THE

CONTE DU GRAAL CYCLE

Perceval’s father, threatens to destabilise the image of Perceval as a hero sui generis, and it may be partly for this reason that it failed properly to integrate not only the Conte du Graal’s cyclic core but also wider Arthurian tradition – Perceval’s father appears in a number of later texts, but nowhere is his name given as Bliocadran. In a similar way, the Elucidation’s centrifugal dynamic was always going to make it difficult to reconcile with the strongly centripetal nucleus of the Conte du Graal cycle. While the First Continuation was recuperated into this nucleus by the extensive rewriting that we discussed in Chapter 1, this very recuperation recast its original aesthetic as marginal, and therefore the Elucidation, a centrifugal prologue to a centripetal cycle, was never destined to be widely disseminated. Prologues and continuations represent a means of recasting the material being added to, and the number of internal contradictions found in medieval narrative cycles (and not least the one under discussion) provides ample proof that, in the area of rewriting, medieval literary aesthetics preferred addition to suppression. Yet prologues and continuations are themselves subject to being recast by incipits and explicits, which when present constitute literally the first and last word on how the manuscript user interacts with the text. If aspects such as rubrication or illustration posit potential answers to the question of how the story is to be remembered and understood (what kind of memoire is to be made of the estoire), this paratextual frame asks and answers the inverse question, namely that of the nature of the estoire which is to be committed to memoire, the text which is beginning and ending. On this matter, as hinted above, the incipits and explicits of our cycle tend to support the visual text in framing the text as a biographical romance. It is regrettable that the relevant leaves have been lost in several of our manuscripts: EQSV are all missing their opening and closing folios. Nevertheless, the evidence of the remaining codices is fairly compelling. The explicit to manuscript M reads, ‘Explicit de Perceval le galois le galois’ [Here ends the tale of Perceval the Welshman the Welshman (sic)], and the rubric added by the fourteenth-century hand to its opening miniature follows this in declaring: ‘Ci comance Perceval le Galois’ [Here begins Perceval the Welshman]. We have already discussed the closing miniature of T which implicitly reconciles the two major themes of Perceval and the Grail; just above this miniature the text’s final rubric reads: ‘Explicit li romans de Percheval’ [Here ends the romance of Perceval]. Manuscript U, as one might expect by now, insists especially strongly on biography as the primary textual focus. The large opening miniature of Perceval’s early adventures is rubricated ‘Ci conmence le romans de Perceval le Galois et devise de moult de aventures qui li avindrent’ [Here begins the romance of Perceval the Welshman, which tells of many adventures that befell him], subsuming the whole subsequent cycle into the category of the ‘many adventures’ concerning the central hero. Once again, the manuscript with the most highly developed material text uses it to insist on Percevalian biography as the narrative core. In case the point was

MANUSCRIPTS, MEMORY AND TEXTUAL TRANSMISSION

109

not sufficiently strongly made, the explicit is a veritable epitaph to the greater glory of Perceval: Ci fenist le roumans De Perceval le Galois Lequel fu moult preus et courtois Et plain de grant chevalerie Pour lamour Dieu feni sa vie [Here ends the romance of Perceval the Welshman, who was very brave and courtly and full of chivalry; he ended his life in the service of God]

This colophon aims to guide the reader’s memory of the text around the concept of heroic biography, even providing its own extra-diegetic testimony as to the worth of the Welsh knight. The sole dissenting voice among the paratextual frames comes in manuscript P. Here, although there are no incipits or explicits per se, several features of the opening folios argue for a different conception of the corpus. P is the only manuscript to include both prologues, and it is also the only one whose rubrics describe the text as the ‘story of the Grail’. The relevant rubric comes in between the two prologues, just after the point where the scribe has inserted the last few lines of Chrétien’s prologue to lead into the Bliocadran. The label ‘Ci comence li contes del saint greal’ [Here begins the tale of the holy grail] serves as paratextual reinforcement to the identification of the text as the ‘contes del graal’ made just a few lines previously in the passage from Chrétien’s prologue. The Elucidation, placed outside of the ‘conte’, functions as an introduction to the corpus. It is significant, then, that this short text presents Perceval as only one of several knights to find the Fisher King’s court, insisting in its opening lines on the Grail as the central theme of the narrative: Pour le noble comencement Comence .i. romans hautement Del plus plaisant conte qui soit, C’est del Graal…  (Elucidation, vv. 1–4) [A romance is beginning worthily, for a noble beginning, of the most pleasant tale there is; it is (the tale) of the Grail…]

The Elucidation’s evocation of the seven Grail guardians, each of whom represents one branch of the story to be told, suggests a reading of the cycle which would pay more attention to the centrifugal possibilities embodied in its earlier stages (it is worth noting here that P carries the Short Redaction of the First Continuation rather than its more cyclical Long Redaction). P’s concern to downplay Perceval’s centrality appears to extend to its opening

110 THE

CONTE DU GRAAL CYCLE

miniature, which is quite unlike the corresponding opening images of any other Conte du Graal manuscript. Nine figures are visible – eight seated at a table, of whom one appears to be a king, and one on horseback on the lefthand side. Given that this miniature introduces the Elucidation, it is intriguing that the number of courtiers seated with the king comes to seven, the same as the number of guardians and of branches of the narrative which we are subsequently promised.53 The miniature can thus be read as a figure of the manuscript’s attempt to emphasise the cycle’s centrifugal tendencies and inexhaustible narrative possibilities at the expense of its biographical core. In this chapter, I have examined several aspects of the mise en manuscrit of the Conte du Graal cycle – the significance of large initials, rubrics, illustrations and paratextual frames, as well as the consequences of inclusion and exclusion of texts. I have argued that different elements are better adapted to convey and highlight particular information about the text, and hence guide the user’s response to the diegesis. For a modern reader, there is a striking contrast between experiencing the text in the manuscripts and reading it in published editions. In particular, the critical tendency to divide the text into independent texts is confounded by the manuscripts’ insistence that the Story of the Grail is to be taken as a single entity, whose movements to and from the central narrative of Perceval’s wanderings produce coherence over the length of the cycle. If the authors of the text were keen to signal their contributions to the whole, there is little sense that the creators and users of the manuscripts were as interested in acknowledging them as we are today.54 I have also attempted to situate our manuscripts within a period of important cultural transition, where the spread of literate modes of expression to vernacular literature altered the conception of the function of the codex from being a support for the text (the book behind the textual object) to being an object for visual engagement (the book as textual object). If this change in mentality was already noticeable in thirteenth-century literature, its effects are equally apparent from a comparison of our thirteenth- and fourteenthcentury manuscripts. The developments we have traced in the presentation of the cycle from EQ through MPT to the later manuscripts SU testify to the growing importance of the written document in vernacular French literary (and literate) culture through this period.

53 The identification of the figures as representing in some way the multiple narratives of the Elucidation has been made a number of times, most recently by Hindman (1994: 90), but also by Thompson (Elucidation: 67–78) and Potvin (1865–71: II, 2). 54 Godzich and Kittay (1987: 48) suggest that the scholarly tendency to seek to define and delimit the structure of texts should be understood as an attempt to assert mastery over them. Clearly, the purpose and nature of our interaction with the text are appreciably different from the use made of them by the contemporary audiences and users of medieval manuscripts.

3

Authorship, Kinship and the Ethics of Continuation Thus far I have deliberately set aside in-depth discussion of the authors of our cycle. This has allowed consideration of the whole corpus as a coherent and self-sufficient textual body; by stripping away the literary capital that has accrued (or failed to accrue) to the cycle’s various authors, Chrétien’s contribution could be considered on an equal footing with that of the continuators, offering an alternative to the overwhelming critical tendency to study the corpus through the specific lens of Chrétien scholarship. Nevertheless, as Matilda Bruckner reminds us, romance intertextuality can be understood as ‘a dynamic play operating between anonymous and named authors’.1 At moments in the Conte du Graal cycle, the figure of the author bursts into the textual frame to claim responsibility for the words being transmitted; at other times, the play of authorial communication can be tracked through the analysis of certain particularly suggestive semantic fields. The Conte du Graal opens with images of fertility and generation, quickly linked to the process of textual dissemination via the authorial figure of Chrétien, who ‘semme et fait semence / D’uns romans que il encomence’ [sows the seeds of a romance that he is beginning] (CdG 7–8). This opening gambit suggests that the semantic field of fertility and sterility might fruitfully be ploughed for signs of textual positioning around the notion of authorship, a connection which has also been made by Sébastien Douchet: Sur l’arbre planté par Chrétien viennent s’enter quatre continuations aux modes d’écriture fort différents. La question de la filiation des héros y est posée avec une acuité dont nous pensons qu’elle est en relation directe avec la façon dont ces romans perpétuent le lignage littéraire de Chrétien.2 [On the tree planted by Chrétien came to be grafted four continuations with quite different modalities of writing. Within them, the question of filiation

1 Bruckner (2006: 13). The present chapter intersects in a number of ways with Bruckner’s work on authorship (2006; 2009, Chapter 1) and motherhood (2009, Chapter 3) in the cycle; though our analyses differ in emphasis and interpretation on many points, the productive interactions between them demonstrate the importance of these issues in the corpus. 2 Douchet (2007: 232).

112 THE

CONTE DU GRAAL CYCLE

is posed so insistently that I believe it bears a direct relation to the way in which these romances perpetuate the literary lineage of Chrétien.]

Indeed, the alacrity with which the interrupted text was taken up by subsequent authors can be understood as a response to Chrétien’s prologue, which presents his act of writing explicitly as the beginning (‘que il encomence’) of a romance in which he sows the seeds of continuation by those who know how to follow: Et qui auques requeillir velt, En tel liu sa semence espande Que fruit a .c. doubles li rande  (CdG 2–4) [And he who wishes to reap something should cast his seeds in a place which will yield what he sows two hundred times over]

Philological discussions of textual dissemination have accustomed us to connecting the vocabulary of organic growth with that of kinship, so that we speak of the different branches of a family tree of manuscripts. The genealogical analogy on which such conversations depend can naturally be extended to discussion of the relation of filiation implicit in the process of continuation. Like Chrétien’s Veve Dame, or Caradoc’s mother Ysave in the First Continuation, the mother text may exert authority over its progeny, returning to haunt and shape the progress of the next generation; but it is simultaneously vulnerable to the actions of these progeny.3 Whether the child-text positions itself in opposition to its mother-text, or represents itself as faithfully honouring the family name, the inherited past is always partially an invention of the present that grapples with it. As we saw in Chapter 1, the game of fidelity is a retroactive one: the way in which the Second Continuation positions itself as the logical progression of the Perceval narrative transforms the dynamic of Chrétien’s original text into something quite different; the newly cyclic textual unit then makes necessary the rewriting of the First Continuation into a more ‘faithful’ Long Redaction. Manessier’s conclusion to the cycle, which will be discussed in more detail below, argues for a filiation of patronage, dedicating his contribution to the great-niece of the ancestor in whose name the text began (CM 42652–3). Manessier’s authorial act is thus represented as establishing a genealogical continuity in the commissioning of the text, displacing attention from the issue of authorial agency. Continuation, this passage suggests, is a mode of writing centred less on fidelity to previous authors and more on a sense of duty to the cyclic text; yet at the same time, his act of self-naming and 3 The term ‘mother text’ is a productive coinage by Bruckner (2009: 118), who notes that ‘generational relations and stresses operate as significantly on the structural and intertextual levels as they do on the level of content’.

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

113

concern for defining the limits of his contribution intimate that the tensions of balancing collective and individual authorship are not so easily ignored. This brief example demonstrates how representations of the relation between past and present authors need to be read carefully for the insight they provide into the textual dynamics at work in the cycle. Similarly, the ethical positioning of the cycle’s heroes in relation to their past experiences offers another means of establishing intratextual dialogue between different parts of the cycle. For instance, the making of a promise establishes an expectation of later narrative enactment, and subsequent authors must choose how to develop the narrative programmes they inherit. The author thus has a responsibility towards the audience’s horizon of expectation: while the Continuations differ in the narrative aesthetics they offer the audience, a certain amount of continuity is necessary in order for the audience to recognise the text as continuation; to ignore proleptic elements earlier in the narrative can appear a betrayal of the audience’s expectations, and conversely the choice to return to an interrupted thread can be read as a statement of allegiance to the pre-existing narrative. Yet responsibility for narrative cuts both ways. An ethical imperative also governs the audience’s interaction with the text. If the author must decide how to deal with the weighty inheritance of narrative past, the audience too has to choose how to respond to the text’s invitation to engagement. We saw in the previous chapter that the general tendency of the Conte du Graal cycle’s manuscript tradition, transmitted by and for its thirteenth- and fourteenth-century audiences, is to endorse the centripetal cyclical agenda by mobilising paratextual elements such as rubrication and illustration. The attempts by the different continuators to wrest control of the narrative agenda depend for their success on mobilisation of the audience as witnesses to their textual project: text becomes ‘estoire’ as it enters into ‘memoire’, but just as important for the shape of the narrative is what the audience is allowed (allows itself) to forget. The survival of the corpus as a codicological unit in so many copies attests to the collective success of the continuators in negotiating the fine line between textual tradition and innovation. The present chapter will investigate the ethics of continuation in the Conte du Graal cycle along the lines suggested by this introduction. The method chosen for this discussion is analysis by accretion of certain key themes susceptible of conveying an ethical stance on the process of continuation. The first section considers the thematic of filiation as set up in the Conte du Graal’s fixation on Perceval’s abandonment of his mother, arguing that the handling of this plot strand in the Continuations allows them to define their relation to the mother-text. The second section extends the argument to the theme of fertility, linking back to the issue of kinship via the cycle’s treatment of sexuality and procreation. As we will see, the different continuators make use of these themes to present their contributions as both viable and necessary developments of the narrative. In the final section, the theme of recognition, which lurks in the background throughout the preceding discus-

114 THE

CONTE DU GRAAL CYCLE

sion, is brought centre stage. Recognition scenes abound in this cycle, and I see these scenes as mirroring the process by which the audience apprehends the text, invited alternately to forget and to remember earlier narratives. It is through such moments that we can gain an appreciation of the ethical stakes of continuation, glimpsed in the dynamic interplay between memory and forgetting which sculpts the contours of the cyclic narrative.

Parental Guidance During Perceval’s visit to his hermit uncle in the Conte du Graal, he is told that the reason for his catastrophic silence at the Grail Castle is an anterior sin committed in the first lines of the text: Et dist: ‘Frere, molt t’a neü .I. pechiez dont tu ne sez mot, Ce fu li doels que ta mere ot De toi quant tu partis de li, Que pasmee a terre chaï Al chief del pont devant la porte, Et de cel doel fu ele morte. Por le pechié que tu en as T’avins que tu ne demandas De la lance ne del graal’  (CdG 6392–401) [And he said: ‘Brother, a sin of which you know nothing has harmed you greatly; this was the grief your mother suffered on your account when you left her. She fell to the ground in a faint at the foot of the bridge, in front of the gate; and she died of this grief. It is because of the sin you carry for this deed that you were unable to ask about the lance or the grail’]

The sense of ‘dont tu ne sez mot’ in v. 6393 has been debated by critics, since the hermit appears to present the death of Perceval’s mother as new information when both hero and audience had already learned of this event earlier in the narrative.4 It must be noted that the troublesome lapse in narrative memory implied by this passage is hardly unique in Arthurian literature: it is one of the narrative hallmarks of the Vulgate Cycle that characters are told something many times over. In the final text of the cycle La Mort le roi Artu, Arthur is warned of the adultery between Lancelot and Guenièvre several times, and frequently he responds to this information as if hearing it for the first time; it is only towards the end of the cycle that he begins to treat the 4 See for instance Busby (1993a: 68): ‘The hermit is wrong when he says that Perceval knows nothing of his sin (although “sez” in line 6393 may mean “comprehend”), since he has already been informed of it by his cousin.’

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

115

adultery as real and act accordingly.5 As Miranda Griffin notes, ‘[t­]­he characters’ individual memories appear to have frozen at the point just before a dramatic or traumatic revelation is made, resulting in their reiterated surprise each time the revelation is re-enacted’.6 Such moments in prose Arthurian romance are part of a game played with the audience, challenged to remember previous episodes, often across thousands of lines, that the characters and the narrative itself appear to have forgotten. What this suggests is that an important aspect of the Vulgate Cycle’s narrative project is already present in Chrétien’s text, an aesthetic of amnesia that goes hand in hand with the use of interlace, an interruptive technique in Arthurian literature of which the Conte du Graal was also an early exponent, as we saw in Chapter 1. Perceval’s unwillingness to commit his mother’s death properly to memory is visible in his reaction to learning of it for the first time from his cousin, directly after his visit to the Grail Castle. With the casual aphorism ‘les mors as mors, les vis as vis’ [the dead to the dead, the living to the living] (CdG 3630), he calls off his journey back to the family home, clear in his mind that, now dead, his mother is no longer any concern of his. This offhandedness is all the more surprising given that it runs in direct contradiction to the promise he made to the monks and nuns of Biaurepaire as he was leaving their town: Je revenrai… Que ja por rien nel laisserai; Et s’ele est vive, j’en ferai Nonain velee en vostre eglise; Et s’ele est morte, le servise Ferois por s’ame chascun an  (CdG 2960–5) [I will return… nothing will prevent me from doing so; and if (my mother) is alive, I will have her take the veil as a nun in your church; and if she is dead, I will have a service performed for her soul every year]

Indeed, from his meeting with Gornemant onwards, the obsessive image of the mother he had left collapsed at her doorstep returns repeatedly to interrupt his encounters with other characters. To Gornemant’s entreaty for him to stay on with him for a month, or even a year, and gain further instruction in chivalry, Perceval replies that he wishes only to see his mother again: Car pasmee le vi cheoir Al chief del pont devant la porte, Si ne sai s’ele est vive ou morte.

5 6

See Mort Artu, pp. 5–6; 61–4; 107–12. Griffin (2005: 89).

116 THE

CONTE DU GRAAL CYCLE

Del doel de moi quant le laissai, Chaï pasmee, bien le sai; Et por ce ne porroit pas estre Tant que je seüsse son estre  (CdG 1584–90, my italics) [For I saw her fall in a faint at the foot of the bridge, in front of the gate, and I do not know whether she is alive or dead. She fainted with grief because I left her, this I well know; and for this reason I cannot stay until I know how she is]

The repeated image of the mother fallen in a faint in front of the family home picks up directly on the vocabulary of the original incident: Quant li vallés fu eslongiez Le get d’une pierre menue, Si se regarde et voit cheüe Sa mere al chief del pont arriere, Et jut pasmee en tel maniere Com s’ele fust cheüe morte.  (CdG 620–5, my italics) [When the youth had moved a stone’s throw away, he looked back and saw his mother fallen at the foot of the bridge, and she lay in a faint as if she had fallen down dead.]

Perceval’s desire to learn his mother’s fate is the narrative impetus which drives the hero forward to new adventure; however, instead of finding his way home, he now arrives at Biaurepaire, where he saves Blancheflor from the siege of her suitor Clamadeus.7 As the narrator explains, Perceval has won the right to both maiden and land, but again the image of the fallen mother intervenes to drive the narrative onwards: Et si fu soie toute quite La terre, se il li pleüst Que son corage aillors n’eüst; Mais d’autre chose plus li tient: De sa mere li resovient Que il vit pasmee cheoir, S’a talent qu’il l’aille veoir  (CdG 2914–20, my italics) [And the land could have been his undisputed, had he so desired, if his thoughts were not elsewhere. But he is more interested in another thing: he remembers seeing his mother fall in a faint, and desires to go and see her]

7 As Bruckner (2009: 218) comments, ‘it is this desire to go back to the mother that paradoxically leads Perceval elsewhere’.

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

117

The quest to return home has thus taken over from the quest to become a knight as the motor for narrative movement; in both cases, situations of potential narrative stasis are unblocked by the hero’s desire for what he does not have, or for what he had and lost. Shortly after leaving Biaurepaire, Perceval’s preoccupation with his mother’s fate rears its head once more, when he comes to the deep river along whose banks he will find the Fisher King’s castle: L’eve rade et parfonde esgarde Si ne s’ose metre dedans, Et dist: ‘Ha! Sire toz puissans, Se ceste eve passer pooie, Dela ma mere troveroie, Mien escïent, se ele est vive.’  (CdG 2988–93) [He looks into the deep, fast-flowing water which he dares not enter, and says: ‘Oh! Almighty Lord, if only I could pass over this river I would, I believe, find my mother on the other side, if she is alive.’]

This time the narrative economy is organised somewhat differently: it is the unfordable river which interrupts the quest for the mother and makes possible the meeting with the Fisher King. Indeed, this is the final time that Perceval thinks of his mother; as he leaves the Grail Castle his thoughts are now turned towards the mysteries he has just experienced. In this opening part of the text, different narratives appear to compete for retention within the memories of both hero and narrative. When the goal is to become a knight, the mother is forgotten; when the goal is to find the mother, the beloved is forgotten; when the goal is to learn about the Grail procession, the mother is forgotten again. When the hero’s thoughts turn towards Blancheflor, he forgets everything else, including himself (CdG 4202). And yet, when his reverie is disturbed and he is eventually welcomed into Arthur’s court, his beloved once again disappears from the narrative. The relation of this narrative technique to the interlace which will come to play such an important role in the cycle’s aesthetic is worthy of comment. With a single hero, the text appears to struggle to contain all of the narrative threads it has put into play, and the hero’s progress becomes a regression, always attempting to move towards the thing that has been left behind. Interlace, as we suggested above, demands the periodic forgetting and remembering of narrative threads by text, audience, and often characters themselves, and therefore allows multiple, competing threads to be accommodated within the narrative economy. But as a consequence of this, every interruption contains a promise of resumption, a fact acknowledged by the narrator when he comments on his own use of the technique:

118 THE

CONTE DU GRAAL CYCLE

De Percheval plus longuement Ne parole li contes chi, Ainz avrez mais assez oï De monseignor Gavain parler Que rien m’oiez de lui conter.  (CdG 6514–18) [The tale does not speak any more of Perceval here; indeed, before you hear me tell of him again, you will hear many things about Sir Gauvain.]

When Perceval meets his cousin, the text is still in its initial phase of attempting to define a single trajectory for its hero from the various competing possibilities, and his reaction to the news of his mother’s death (‘Autre voie m’estuet tenir’ [I must follow another path], CdG 3625) appears to signal that the narrative is steering irrevocably away from the domestic sphere of the Waste Forest. Instead, other issues crowd onto the stage for consideration: the hero’s love for Blancheflor, which plunges him into a meditative absence; his integration into Arthur’s court, sealed by Gauvain’s offer of friendship; and then the revelation by the Hideous Damsel of the consequences of his failure to speak at the Grail Castle. It is of course at this point that the text introduces interlace by forgetting its initial hero and instead following Gauvain. When we do finally return to Perceval, five years on, we learn that while the text was ignoring him, he too has lost his ‘miemoire’ (CdG 6218). A kind of transference is at play between character and text, both of which display this periodic capacity to fail to follow through on the narrative past. It becomes clear in the course of his meeting with his hermit uncle that Perceval’s aphorism ‘les mors as mors, les vis as vis’ is quite inaccurate, as the dead return to haunt the destinies of the living. The ‘other path’ of narrative progress turns out to lead back to the stories of the past. In searching for his chivalric identity, Perceval attempts to leave his old, familial identity behind, and his uncle’s words place the blame for his subsequent failure squarely on this misguided impulse. So little did Perceval care for his matrilineal inheritance that he failed to turn back upon seeing his mother collapse as he left home. This, Chrétien’s romance suggests, is not a viable strategy for making your way in the world, and so the would-be selfmade hero, who believes he has become a knight by taking the arms of his defeated opponent, is forced to share his romance with Gauvain, and made to spend five aimless years wandering in the wilderness. And yet the presentation by the hermit uncle of the play of cause and effect suggests that poor Perceval could hardly have done things differently; his destiny is derailed by a narrative paradox neatly summarised by Bruckner: ‘he must leave his mother to achieve the Grail; he cannot leave his mother and achieve the Grail’.8 Moreover, as I argued in Chapter 1, by the end of the Hermit episode 8

Bruckner (2009: 125).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

119

there seems little left for him to achieve within the narrative parameters set up by the Hideous Damsel’s condemnation. The orphaned Perceval is left without a narrative to call his own, unmoored from literary precedent, in an extraordinary state of apparent narrative superfluity. An illustrative contrast can be drawn with Renaut de Beaujeu’s Bel Inconnu, whose protagonist, like Perceval, is brought up by his mother in ignorance of his chivalric pedigree. Renaut’s text turns on the revelation, made around the halfway point in the narrative, that the hero’s name is Guinglain, and his father Gauvain.9 Donald Maddox takes this moment as an exemplary instance of what he calls the ‘specular encounter’, a moment of revelation where the central character either learns or reveals his or her identity, and around which the narrative is organised. He argues that ‘[r]­estoration of interrupted communication between the filial and the paternal generations is thus the fulcrum of the entire romance, the vital turning point that confers heroic identity and initiates the long process culminating in the hero’s enculturation in Arthurian society.’10 Moreover, by tying its hero to the family tree of one of Arthurian literature’s most celebrated characters, and indeed establishing a kinship relation between Guinglain and the king himself, the Bel Inconnu implicitly attempts to authorise its own place within the genealogy of Arthurian romance. If Guinglain’s integration into Arthurian society is made all the more forceful by the revelation of his genealogical connection to the king via his favourite nephew, Perceval’s case is clearly quite different. Chrétien’s hero learns, or rather intuits, his identity during the same conversation with his cousin where he learns of his mother’s death, but no genealogical significance is attached to this. The name of the father in this encounter is never revealed: Et cil qui son non ne savoit Devine et dist que il avoit Perchevax li Galois a non, Ne ne set s’il dist voir ou non; Mais il dist voir et si nel sot.  (CdG 3573–7) [And he who did not know his name guessed it, and said that he was called Perceval the Welshman, though he did not know whether what he was saying was correct or not; in fact, he was telling the truth and did not know it.]

Although the wording of this passage, stressing the truth of the hero’s words, suggests that this is a significant moment, there is nothing here to guide Bel Inconnu, vv. 3212–42. Maddox (1994a: 38). See also Maddox (2000) for a more detailed discussion of the ‘specular encounter’ in medieval French romance. 9

10

120 THE

CONTE DU GRAAL CYCLE

the reader as to what its precise significance might be. As a consequence, Perceval’s intuiting of his name serves only to deepen the mystery around his identity, and the incomprehension of the audience about the significance of the moment is mirrored by the hero’s own lack of insight: three times in the passage the narrator underlines the fact that Perceval ‘ne set’ what he is saying, prefiguring his uncle’s observation that he ‘ne set mot’ of his mother’s death. Charles Méla has argued that the figure of the father haunts Chrétien’s text as an absence, a void which is to be filled imperfectly by a succession of surrogate father-figures: Arthur, Gornemant, the Fisher King, the Hermit Uncle.11 We could add to this, following our discussion of Perceval’s desire to return to his mother in the early part of the text, that his abandonment of her and her subsequent death create a second absence; it is his failure to acknowledge this absence which leads directly to the catastrophic consequences of his Grail Castle failure. Discussing the Perceval portion of Chrétien’s text in isolation from the rest of the cycle, critics have often described it as a medieval Bildungsroman;12 yet, until the Continuations take up this narrative thread again, it appears on our analysis more like the tale of a little boy lost, his path towards greatness derailed by an insufficient grasp of his filial responsibilities. The character begins life as the ‘widow’s son’, defined in relation to his mother (CdG 74), yet his yearning for chivalry sets him on a course to emulate the deeds of his father, whom the text reveals to have been a knight himself.13 As Bruckner remarks, the decisive excision of the maternal from the boy’s speech occurs during his knightly instruction with Gornemant. Chastising him for constantly quoting his mother’s advice, Gornemant advises him henceforth never to say that she has taught him anything; instead, he is to attribute his education to ‘Li vavasors… / Qui vostre esperon vos caucha’ [the worthy man who gave you your spurs] (CdG 1686–8). The gradual erasure of the mother from Perceval’s identity is thus presented as ‘the price paid for his socialization’ into the public, very male world of Arthurian chivalry.14 Symbolically, when Perceval’s real ‘knightly’ 11 Méla (1977) and (1984: 85–94); on the absence of the paternal as a trauma in the text, see also Huchet (1980); Kleiman (2008). Bruckner (2009: 135) observes that critical interest in Perceval’s undiscussed paternity has overshadowed the more explicit theme of the quest for the mother: ‘Many critics tend to replace the puzzle of the mothers’ presence by a search for the absent father.’ 12 See for instance Micha (1951) or Ribard (1976a). An original variation of this view is proposed by Bloch (1983: 204), for whom the story is ‘an upside-down Bildungsroman, one in which learning is essentially a process of unlearning – an undoing of the obfuscating signs of the mother’. 13 An interpolation found in MSS AL underlines further the importance of the maternal to the construction of the hero’s identity at the outset of the text. Asked his name by the leader of the knights he meets in the forest, the boy’s first reply is that he is called ‘Biax Filz’. See the notes to Busby’s edition of the Conte du Graal, p. 429. 14 Bruckner (2009: 131).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

121

name is revealed, the only reference to provenance in it is the geographical suffix ‘le Gallois’, which is immediately subtracted by his cousin and replaced by alternatives highlighting the consequences of his own actions: Tes nons est changiés, biax amis. – Comment? – Perchevax li chaitis! Ha! Perchevax maleürous  (CdG 3581–3) [Your name is changed, dear friend. – In what way? – Perceval the Unfortunate! Ha! Perceval the Unhappy]

The mark of the past, represented by Perceval’s Welsh origins, gives way to descriptors which posit his Grail Castle misadventure as a newly defining feature of his identity. Proleptically, it anticipates what the Hideous Damsel will say during her dramatic appearance at Carlion: that Perceval, whom she also calls ‘li maleüreus’ (CdG 4665), has missed his opportunity for glory.15 We saw in Chapter 1 how the Second Continuation rescues Perceval from the predicament in which Chrétien had left him, making it possible for him to attain the status of best knight in the world which was prophesied for him during his first visit to Arthur’s court (CdG 1061–2). In doing so, Wauchier de Denain, along with the contributors to the cycle who follow him, transforms the relationship between Perceval and his lineal past. Wauchier authorises his hero’s return to the family home, no longer an irrecoverably lost space, but now a locus of redemption. The episode begins with the staging of a return from amnesia of the figure of the dead mother which had haunted the text in its early stages. Without warning, the topography of the romance’s opening lines irrupts into the audience’s consciousness and then, after a delay for recognition, into Perceval’s: Percevaux monte sans targier, Tant que vers tierce chevaucha. An la Gaste Forest antra, Ou mainte foiz esté avoit; Molt petit s’i reconnoissoit, Car molt a lon tans qu’il n’i fu. (C2 23546–51) [Without delay, Perceval mounted his steed, and rode until mid morning. He entered the Waste Forest, where he had been many times before; he recognised very little, for it was a long time since he had last been there.]

15 As Emmanuèle Baumgartner (1999: 110) observes, ‘[l]a cousine au reste ne touche pas au nom de “Perceval”, … mais barre le lieu d’origine, la terre de Galles, comme si le héros n’y avait plus droit’ [The cousin leaves intact the name ‘Perceval’ but excises the place of origin, the land of Wales, as if the hero no longer had any claim to it].

122 THE

CONTE DU GRAAL CYCLE

The line ‘Molt petit s’i reconnoissoit’ is susceptible of two, ultimately convergent readings. On the one hand, it speaks of Perceval’s lack of recognition of his surroundings, due to the amount of time elapsed since he left; on the other, the genealogical thrust of the passage invites a reading of ‘molt petit’ as a reference to Perceval’s age when he left the Waste Forest, hinting at the reconciliation between the hero and his childhood to be effected in the ensuing scene. The overdetermination of meaning in this line plays on a similar ambivalence in the sense of ‘reconnaitre’, which can range from recognition to acknowledgment; clearly, both meanings are highly relevant to Perceval. In a wide-ranging study of anagnorosis in literature, Terence Cave has argued that recognition ‘reaches back most often to painful or problematic narrative events hidden in the past’, an assertion exemplified by the passage under discussion.16 The moment of recognition thus functions as a moment of acknowledgment, a taking on of responsibility for the narrative past that reconfigures the hero’s identity. As if to underline this fact, the recognition scene here is drawn out over several stages: first Perceval identifies his surroundings, then he recovers his memory of the past, and then he himself undergoes the same routine of nonrecognition followed by identification at the hands of his mother’s household. The penny initially drops for the hero when he recognises a tree as that under which he had spoken to Arthur’s knight in the cycle’s opening episode, back when he was identified only as his mother’s son. Understandably, his first reaction is one of disbelief: ‘Diex,’ fait il, ‘ou suis ge asenez? Je cuit pres dou manoir ma mere … Si m’aïst Diex, molt me mervoil Conmant je suis ici venuz’  (C2 23556–7; 23564–5) [‘Lord,’ he said, ‘where have I landed? I think I am near my mother’s house… God help me, I have no idea how I arrived here’]

Having reminded himself (and, by the same token, the audience) of the early events of the narrative, Perceval’s next words suggest that he is coming to a redemptive understanding of the importance of the filial bond: ‘Chose ne desirroie plus Que savoir verté de ma dame, Que Diex face merci a l’ame.’ Lors a ploré molt tandremant; De sa mere pitié li prant, Car il li venoit de nature.  (C2 23566–71) 16

Cave (1988: 22).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

123

[‘There is nothing I would wish more than to know the truth about my mother, may God have mercy upon her soul.’ Then he cried tenderly; he was seized by pity for her, for he was her natural child.]

This new-found desire to reconnect with the people and places of his past is underlined again in terms that will be echoed by Galaad before the Grail in the Queste del Saint Graal; catching sight of his mother’s manoir, Perceval exclaims: ‘Je voi ce qu’ai tant desirré’ [I see what I have desired for so long] (C2 23579).17 Indeed, his wish to ‘savoir verté’ [know the truth] about his mother and her death itself recalls his curiosity about the events of the Grail procession, while also picking up on the vocabulary of ‘savoir’ which we have traced as central both to his encounter with his hermit uncle and to the moment when he intuits his own name. The disconnection between the widow’s son of the cycle’s opening and the knight that Perceval has become is given voice by his sister. This character, introduced in the Second Continuation, provides an account of the cyclical narrative as perceived from within the family domain, a narrative which begins and ends with Perceval’s departure from home. Thus, unaware that she is speaking to her brother, she recounts his own actions back to him: Qui qu’an eüst duel ne annui, A la cort le roi an ala. Ne sai conmant il esploita, C’onques n’an oï puis parler.  (C2 23662–5) [He went to the king’s court, regardless of the suffering and sadness he might cause in doing so. I do not know what happened to him, for I have never heard speak of him since.]

Perceval’s subsequent adventures, though they may have made him the talk of the Arthurian world, have failed to reach the isolated manoir in the Waste Forest. Indeed, he has changed so much in the intervening period that the household greets the stranger without recognising him: Mes ne l’ont mie ravisé Ne couneü ne tant ne quant A son vis ne a son samblant, Avec ce ne leur an sovint.  (C2 23594–7) [But they did not give him a second look, and did not recognise him in the slightest; they remembered neither his face nor his appearance.]

17

Cf. Queste, p. 278, vv. 9–10.

124 THE

CONTE DU GRAAL CYCLE

Having evoked this disconnection between the hero and his family, the Second Continuation now proposes a reconciliation, which is at the same time a reconciliation of the cyclical Perceval narrative with the unfinished yet abandoned business of Chrétien’s part of the text. Thus the image of the mother fallen in a faint at the door is recalled from the textual past by the sister (C2 23667), who reveals that her body lies in a nearby hermitage, where it was buried by an uncle of theirs who lives there. Interestingly, this uncle is related to Perceval through his father, making him the only paternal relative encountered in the cycle, and not therefore identifiable with the hermit uncle of Chrétien’s part of the text. Nevertheless, the hermitage functions in both cases as a space of penitence for the hero, in this instance overlaid by the redemptive presence of the maternal tomb, and it is therefore no surprise that Perceval feels the need to go there: Percevaux dist a sa seror Qu’a l’ermitaige viaut aler, A son oncle voudra parler Qu’il n’avoit veü des s’anfance, Et si panra sa penitance De ses pechiez, que c’est droiture; Et si verra la seposture Ou sa mere est ansevelie, Qui por lui seul pe[r]di la vie.  (C2 23736–44) [Perceval told his sister that he would like to go to the hermitage and talk to his uncle, whom he had not seen since his childhood; there he would do penance for his sins (as was right) and see the tomb where his mother, who had lost her life on his account, was buried.]

This episode stages a number of powerful scenes of recognition and penitence, as Perceval finally admits that his mother’s death is his responsibility. When he reveals his identity to his sister, he tells her ‘que c’estoit ses frere / Et que por lui morust sa mere’ [that he was her brother, and that her mother had died because of him] (23699–700); later, at the hermitage, he weeps over the tomb, and the memory, of the mother (23932–3). On a narrative level, the reconnection of the family narrative with that of the hero is completed when the hermit takes his nephew by the hand and has him narrate his adventures from the point at which the filial rupture occurred: Tot li conte quanque fait a Des lors que li rois l’adouba Qui li dona armes vermoilles […] Aprés li a trestot conté Qu’il a fait et com a ouvré, Si com avez ou livre oï  (C2 23943–5; 23975–7)

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

125

[He told him everything he had done since the king, who had given him the arms of the Scarlet Knight, had knighted him… Then he told him everything that he had done and how it had gone as you have heard in the book]

The domestic domain is no longer a space erased from, and inaccessible to, the narrative, and conversely Perceval’s narrative, which for the Second Continuation is the central thread of the cycle, is no longer unknown in the Waste Forest. The cyclic chivalric narrative is thus symbolically accepted into the hermetic environment which had attempted, before the start of the narrative, to resist the incursion of knights. The opposition set up by Perceval’s mother at the outset between chivalric chaos and domestic stability, which nourished itself on the absent figures of her dead husband and elder sons, is now resolved by the penitence of the hero weeping over his mother’s grave. In doing so, Wauchier replaces the anxiety of maternal influence suffered by Chrétien’s hero with a more ‘healthy’, more patriarchal genealogical order. The substitution of a paternal doppelganger for the maternal hermit uncle is no innocuous detail; it allows the Second Continuation to underline that the voice of familial and religious authority need not speak exclusively from the maternal, anti-chivalric perspective. Narrative dynamic is visible in the deployment of the domestic theme again in the Gerbert Continuation, whose entire narrative project is predicated on the need for the hero to achieve spiritual self-improvement before he will be able to understand all of the mysteries he has seen and gain the full approval of the Fisher King. At the heart of this project is a reading of the end of the Second Continuation – where Perceval joins two halves of a broken sword, leaving just a chink between the pieces – as symbolically announcing an abiding spiritual imperfection in the hero requiring correction. The chink is understood by Gerbert as an ‘inviting gap’, and to underline this new narrative dynamic Perceval’s first adventure after leaving the Grail Castle sees him refused entry into Earthly Paradise after breaking his sword hammering on the gate.18 As discussed in Chapter 1, Gerbert goes out of his way to send Perceval back down paths he has trodden earlier in the cycle, and it is therefore no surprise that he should bring his hero once more to the family abode. This time, communication between the different textual memories of the cycle is presented as unproblematic. In contrast with the Second Continuation account, particular emphasis is put on the ease with which the protagonists recognise each other, beginning with Perceval’s identification of the location at the same time as the reader:

18 On the notion of the ‘inviting gap’, a narrative lacuna that invites or allows further development, see Brandsma (1995).

126 THE

CONTE DU GRAAL CYCLE

Bien reconoist le terreoir, Car pres d’iluec ert le manoir Ki fu sa mere, bien le voit.  (CG 2613–15, my italics) [He recognised the area well, for nearby was the house that had been his mother’s, as he could well see.]

In similar fashion, upon greeting him, Perceval’s sister ‘[l]e reconnoist bien’ (2643); and when they meet their hermit uncle, we are told that he ‘bien les a reconneüs / Si tost come il les a veüs’ [recognises them well as soon as he saw them] (2733–4). If, as suggested above, the semantics of ‘reconnaissance’ are susceptible of binding together notions of recognition and acknowledgment, the easy and natural communication between Perceval and his relatives can be read as a scene of acceptance by all parties of his place within the family structure. In accordance with Gerbert’s project of moral improvement for Perceval, whereas the Second Continuation appeared to suggest that the sin of the mother had been redeemed, on this visit it emerges that Perceval’s guilt has not yet fully been explored. He again arranges with his sister to visit the hermitage where his mother is buried, and is soon to be found weeping over her grave once more. The difference comes in the words he speaks as he weeps, which draw a much more explicit link than previously in the cycle between the abandonment of the mother and the sinful state that prevents Perceval from full knowledge of the Grail secrets: Lors pleure et dit a vois auçor: ‘Ha, dolce mere, li pechié Que j’ai por vous m’ont si chargié Que je espenis nes verrai Ne a l’amour Dieu ne venrai S’il ne me regarde em pité.’  (CG 2740–5) [Then he cried and said in a louder voice: ‘Oh! gentle mother, the sins I have committed towards you have weighed so heavily upon me that I will never see them expiated, nor gain God’s love, if He does not take pity on me.’]

In TV, the only manuscripts to transmit the Gerbert Continuation, Perceval must perform an additional circuit, proving himself again within the explicitly Christianising framework devised by the author. Yet whether redemption comes here, or in the Second Continuation episode as in the other manuscripts, the continuators show a desire to demonstrate that the aborted narratives of Chrétien’s romance can be reconciled with and incorporated into their own textual projects. The inheritance of the textual past no longer hangs unacknowledged over the narrative; instead, it is assimilated back into the narrative present. Indeed, the Gerbert Continuation takes this exorcism one

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

127

step further, since it suggests that Perceval is now able finally to leave behind his textual beginnings. As he arrives at the family home, we learn that he has a particular project in mind for his sister: Dont dist que sa suer jecteroit D’iluec, et si le meteroit Hors de cele forest salvage  (CG 2619–21) [Then he said that he would take his sister away from there, and out of that wild wood]

If the socialisation of Perceval in Chrétien’s text depended on the erasure of his rustic upbringing, communication between the chivalric knight and his kin is now so harmonious that he can even consider bringing his family out into wider Arthurian society. Moreover, Perceval has made the transition from mother-murderer to sister-protector: predictably, on their journey from the family home, he is obliged to defend her from abduction. Arthurian chivalry, which in Chrétien’s romance had threatened the stability of the familial sphere, is now presented as essential to its security. Once again, this narrative sequence refers back to the textual past of the Second Continuation, where another knight had already attempted to carry off the hero’s sister, as well as many other instances which thrust him into the role of protector of the defenceless. In the Gerbert Continuation, one particular episode stands out as especially significant. Perceval arrives one day at the castle of his former mentor Gornemant to find him transformed into a vulnerable old man; every day, he and his sons are forced to kill forty enchanted soldiers, who return to wage war the following day after resuscitating overnight. Worn down by the attritional campaign, Gornemant and his sons will be reliant on Perceval’s intervention to save their lives and their land. The image of the master reliant on the pupil suggests one way of conceiving of the relation between the different authors of the cycle. The notion of apprenticeship is often found more or less explicitly in discussion of Arthurian verse romance, with Chrétien de Troyes projected as the founder of a school of epigones. John Grigsby makes the point acerbically in summing up his perception of the literary value of the Continuations: ‘[t]hough the four continuators, especially Wauchier, lacked Chrétien’s art, he obviously remained their master’.19 Yet Gerbert’s account of his role in the authorship of the cycle shows more self-confidence than this comment allows. It is Gerbert who claims that death intervened to prevent Chrétien from completing the Conte du Graal:

19 Grigsby (1988: 53). See Chapter 4 for a re-evaluation of the view that thirteenthcentury Arthurian verse romance is primarily modelled on Chrétien’s corpus of Arthurian narratives.

128 THE

CONTE DU GRAAL CYCLE

Ce nous dist Crestiens de Troie Qui de Percheval comencha, Mais la mors qui l’adevancha Ne li laissa pas traire affin  (CG 6984–7) [This is what Chrétien de Troyes, who began the story of Perceval, tells us; but death, which outran him, did not let him bring it to an end]

While this passage makes Gerbert the first scholar to argue that the Conte du Graal was Chrétien’s last romance, what interests me is that it presents the continuation of the text as a necessary move, a gesture of faithfulness to the first author. At the same time as claiming literary filiation and hence authority for his Continuation, Gerbert implies that the original author’s death creates the need for a new author to take responsibility (and credit) for the text. Having reported the father’s death, he now announces that he has made the project his: Mais or en a faite sa laisse Gerbers, selonc le vraie estoire; Dieus l’en otroit force et victoire De toute vilenie estaindre Et que il puist la fin ataindre De Percheval que il emprent  (CG 7000–5) [But now Gerbert has made it his work, according to the true story; may God give him the strength and victory in quashing all baseness, so that he may reach the end of the story of Perceval, which he has undertaken]

The language of this passage, asking God for ‘strength and victory’ in order to ‘reach the end’, invites the audience to parallel the author’s quest for an ending with that of the hero: in both cases, Gerbert represents success in this quest as contingent upon moral self-discipline and divine election. Pushing the comparison further, the notion of Gerbert coming to the aid of Chrétien’s interrupted text puts one in mind of Perceval rushing to the rescue of Gornemant. Simultaneously, the image of the writer Gerbert acknowledging the death of the master Chrétien calls up once again the image of Perceval weeping over his mother’s tomb, exorcising his debt to her before riding free from her influence. By establishing his filiation with the story’s originating author, Gerbert demonstrates that the discourses of intergenerational communication and literary continuation are closely intertwined.20 To the anxiety over the maternal influence displayed by Chrétien’s text and its 20 For further reflections on how the interruption and restoration of communication play out as themes in the cycle, see Hinton (2009).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

129

hero, he opposes a reassuring narrative of clerkly continuity, casting Chrétien as progenitor of the cycle. In smoothing out the tensions in genealogy explored by the originary romance, Gerbert valorises the masculine authority of clerkly textuality over the maternal authority insistently and provocatively placed by Chrétien in opposition to the values of Arthurian chivalry. The Manessier Continuation approaches the notion of authorial communication from a different direction. Manessier doctors the realities of textual transmission, excising all trace of the intervening authors to present himself as the continuator of the original author’s work: Si com Manesier le tesmoingne, Qui met a chief ceste besoingne El non Jehanne la contesse Qu’est de Flandres dame et mestresse […] Ai en son non finé mon livre. El non son aiol conmença Ne puis ne fu des lors en ça Nus hons qui la main i meïst Ne du finner s’antremeïst  (CM 42641–4; 42652–6) [As Manessier testifies, who brings this task to an end in the name of the countess Jeanne, lady and mistress of Flanders… I have completed my book in her name. It was begun in the name of her ancestor, and no man since then has turned his hand to it, nor ­attempted to complete it]

Where Gerbert commemorates Chrétien’s name through the epitaph inscribed within his own narrative, Manessier omits any mention by name of his predecessor. Instead, he focuses attention on his patroness Jeanne de Flandres, in whose name, he declares, he has completed his book, author and patron thus sharing responsibility for the text. Chrétien’s prologue dedicates the text to Philippe de Flandres, Jeanne’s paternal great-uncle, whom he claims provided him with the source book for his tale. Manessier’s epilogue mirrors the prologue in linking the value of the text with that of the patron, whose qualities he extols at some length. Yet while it chooses to remember this part of the prologue, it tellingly does not refer to Chrétien by name; instead, the impersonal ‘conmença’ [began] plays down any notion of agency on the part of the original writer. Thus the game of authorial communication is carefully obscured behind a narrative of patronal heredity.21 And where Gerbert posits the clerk Chrétien as having sowed the seeds of the narrative project, Manessier reserves this place for Philippe, the uncle of his patroness’s father. 21 See Séguy (2001b: 219–20) and Bruckner (2009: 55) for congruent analyses of this closing signature.

130 THE

CONTE DU GRAAL CYCLE

In both cases, as throughout the Continuations, the realignment of textual genealogy from the maternal (anti-chivalric) to the paternal (chivalric or clerkly) order appears to be a requirement of the restoration of harmonious intergenerational communication. At the outset of the text, Perceval’s mother argues for the hero to remain at home, and not to engage with the narrative of Arthurian chivalry; her death, and its subsequent redemption later in the corpus, suggest that in this cycle, narrative is a matter for men.22

Textual Generations: Fertilising the Conte Perceval is not, of course, the only character in the cycle whose forward momentum sees him confronted by his own past. In Chrétien’s part of the text, which gives over more lines to narrating Gauvain’s adventures than those of Perceval, Arthur’s nephew is repeatedly interrupted or prevented from finishing tasks he has undertaken by the repercussions of events occurring before the beginning of the text. His intention to rescue the damsel of the Mont Perilous is undermined by the arrival at court of Guingambresil and the latter’s accusation of treasonous murder (CdG 4747–65). Later, his progress is hindered by the vengeful and ungrateful actions of a knight whom Gauvain meets without realising that they have already crossed paths, and whose health and eyesight he restores: S’a monseigneur Gavain veü; Lors primes l’a reconneü.  (CdG 7065–6) [He saw Sir Gauvain; then he recognised him for the first time.]

Recognition leads to revelation of another pre-diegetic event, as the stranger steals Gauvain’s horse, reminding him of the shame he has previously suffered at his hands: Ne te sovient il de celui Cui tu feïs si grant anui Qu’il li covint oltre son pois Avec les chienz mengier .i. mois, Les mains liies triers le dos?  (CdG 7111–15) [Do you not remember the man whom you harmed so greatly when, against his will, you forced him to eat with the dogs for a month, his arms bound behind his back?] 22 Bruckner (2009, Chapter 3 and Conclusion) reads the gendering of the process of continuation in a different way, arguing that the notion of continuation is ‘built into the role of mothers and their links with sons, just as the opening parable of the sower links the fertility of the word to the fruitfulness of mother earth’ (227).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

131

Gauvain is now able to name his interlocutor as Greoreas, a knight he had previously punished for raping a damsel, but his own recognition comes too late to spare him the humiliation of losing his horse. Instead of the valiant Gringalet, he will be obliged to ride a squire’s nag, an event which threatens to endanger his knightly identity and quite literally to slow his progress: Si ne set que il puisse faire De son ronchin, qu’il n’en puet traire Cors ne walos por nule paine.  (CdG 7215–17) [He did not know what he could do with his nag, which he could induce neither to run nor to gallop, whatever he tried.]

Finally, as Gauvain ventures into the unknown land beyond the Galvoie border, he finds himself at the Castle of Marvels, reigned over by two queens who are later revealed to be his mother and grandmother. Once again, maternal authority is represented as inimical to chivalric narrative: once Gauvain has passed the test of the Marvellous Bed to become lord of the castle, he is warned that he will be required to remain there permanently. The words of Ygerne, his grandmother, make clear that she shares the concerns expressed by Perceval’s mother about the negative effect of knightly adventure on the security of the domestic sphere: Vos n’en devez issir jamés, Se vos tort ne nos volez faire.  (CdG 8332–3) [You should never leave here if you wish to stay safe from harm.]

The female order of the castle, with which Gauvain will have to negotiate, spans three generations: his mother, her mother and his sister. The doubling of the mothers is reinforced further by Ygerne’s status as mother not only of Gauvain’s own mother, but also of King Arthur. Recognition is delayed for all parties involved, including for the audience, who initially have no idea who these mysterious women are. Yet filiation becomes an insistent theme from the first meeting between Gauvain and Ygerne, who asks him a number of questions pertaining to the progeny of the kings at Arthur’s court, to which Gauvain gives quite detailed replies: ‘Mais or me dites del roi Loth, De sa feme quanz fix il ot. – Dame, .iiii. – Or les me nomez. – Dame, Gavains est li aisnez, Et li seconz est Engrevains, Li Orgueilleus as Dures Mains; Gaherïés et Guerrehés Ont non li autre dui après.’

132 THE

CONTE DU GRAAL CYCLE

Et la roïne li redist: ‘Sire, si Damediex m’aït, Einsi ont il non, ce me samble […] Or me dites, connoissiez vos Le roi Urïen? – Dame, oïl. – Et a il a la cort nul fil? – Dame, oïl, .ii. de grant renon: Li .i. mesire Yvains a non […] Et li autres ra non Yvains, Qui n’est pas ses freres germains; Por che l’apele l’en Avoltre’  (CdG 8135–45; 8148–52; 8157–9) [‘But now tell me about King Lot and how many sons he has. – Lady, he has four. – Tell me their names. – Lady, the eldest is Gauvain, the second is Engrevain the Proud, of the Hard Hands; Gaheriet and Guerrehet are the names of the other two.’ And the queen replied: ‘Lord, may God help me, I believe that these are indeed their names. Now tell me: do you know King Urien? – Lady, yes. – And does he have sons at court? – Lady, yes, two of great renown: one is called Sir Yvain and the other is also named Yvain, though he is not his full brother; for this reason he is known as the Bastard’]

Gauvain, along with the audience, will learn the significance of these questions from Guiromelant, who himself is initially unaware that he is speaking with Arthur’s nephew, against whom he bears his own grudge. The mystery deepens when Gauvain declares that both his mother and grandmother have been dead for a number of years. Guiromelant insists that they are still alive, having repaired to the castle in question at the death of King Uther. This detail situates Ygerne’s disappearance from the Arthurian environment as contemporary with the self-imposed exile of Perceval’s family in the Waste Forest (CdG 442–54). Moreover, as with the manoir in the Waste Forest up until Perceval’s return in the Second Continuation, the Castle of Marvels appears to have been a space hermetically sealed from the passage of narrative time: Quant Unterpandragon ses [=Arthur’s] pere Fu mis en terre, issi avint Que la roïne Ygerne i vint En cest païs, si aporta Tot son tresor et si frema Sor cele roche cel chastel  (CdG 8740–5, my italics)23 23 Baumgartner (1999: 93) draws attention to the apparent suspension of time within the Castle of Marvels when she describes Gauvain’s mission there as being to ‘faire (r)

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

133

[When Utherpendragon, Arthur’s father, was buried, it happened that Queen Ygerne came to this land; she brought all her treasure and established this castle on this rock]

Here, however, the emphasis is reversed, so that stress is placed on the knowledge which has failed to pierce the Arthurian sphere: having removed themselves to a safe place away from the vagaries of chivalric competition, the women have ceased to exist even for their own kin. One detail in Guiromelant’s account exemplifies the informational void that has opened up between the Castle of Marvels and the Arthurian court: Gauvain’s mother was pregnant when she arrived there, and the daughter to whom she gave birth, a sister of whose existence Gauvain had not been aware, is the beautiful young damsel who had greeted him upon his arrival at the castle. The risks inherent in such communicational breakdowns are made apparent as the text flirts with the spectre of incest. For, having apprised the audience (and the hero) of Clarissant’s identity, Chrétien then stages a heavily ironic scene in which the two mothers are made inadvertently to imagine an incestuous marriage. Watching Gauvain transmit Guiromelant’s love message to Clarissant, as he had promised to do, they misinterpret the scene as one of courtship: Et la vieille roïne sist Dalez sa fille, si li dist: ‘Bele fille, que vos est vis De ce seignor qui est assis Les vostre fille et lez ma niece? Conseillie a a li grant piece; Ne sai de coi, mais molt [me] siet […] Et pleüst Dieu que il l’eüst Espousee, et tant li pleüst Com fist a Eneas Lavine.’  (CdG 9045–51; 9057–9) [And the old queen sat next to her daughter and said to her: ‘Dear daughter, what do you think of this lord who is sitting next to your daughter and my granddaughter? They have been in private conversation for a long time; I don’t know what they are talking about, but it pleases me… and might it please God for him to marry her, and that he should find her as pleasing as Aeneas did Lavine.’]

The insistent foregrounding of the kinship ties linking the women highlights entrer le château et ses habitants des deux sexes dans le cycle du temps, dans le cours normal du monde des vivants’ [bring the castle and its inhabitants of both sexes back into the temporal cycle and the normal course of the world of the living].

134 THE

CONTE DU GRAAL CYCLE

the one crucial piece of information missing in this scene – Gauvain’s own relationship to them – and the irony is further heightened when the younger queen expresses the wish: Qu’il soient comme frere et suer Et qu’il l’aint tant et ele lui Qu’il soient une chose andui  (CdG 9062–4) [That they should be like brother and sister and that he should love her and she him, so that they should be as one being]

In case anyone has missed the point, the narrator explains that: En sa proiere entent la dame Qu’il l’aint et qu’il le praigne a fame. Cele ne reconnoist son fil  (CdG 9065–7) [By this wish, the lady means that she would like him to love her and take her as his wife. She does not recognise her son]

The queens’ misapprehension is understandable, since a private, hushed conversation between a beautiful damsel and a valorous visiting knight is often the precursor to the flowering of love in Arthurian romance. As Roger Dragonetti notes, ‘ce dialogue auquel assistent de loin les deux reines, Ygerne et la mere de Gauvain… prend toutes les allures d’une déclaration d’amour, bien que le son et le sens des paroles ne parviennent pas à leurs oreilles’ [this dialogue which the two queens, Ygerne and Gauvain’s mother, witness from a distance has the appearance of a love declaration, although the sound and sense of the words are out of earshot].24 Perhaps fearing a similar misapprehension on the part of his audience, the narrator is unusually explicit in ruling out the possibility of incest, insisting that: Come frere et suer seront il, Que d’autre amor point n’i avra  (CdG 9068–9)25 [They will indeed be like brother and sister; there will be no other kind of love between them]

If the possibility of romance is so insistently suggested and denied in this passage, it is because it forms part of a wider pattern in Chrétien’s text, whose heroes are repeatedly presented with love-scenarios which refuse to play out

Dragonetti (1980: 212). This passage takes on extra, intertextual irony when one considers that the younger queen, Gauvain’s mother, will give birth in the Vulgate Cycle to Mordred, the fruit of unintentional incest with her own brother, Arthur. 24 25

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

135

properly. On Perceval’s side, the tone is set by his mother’s advice as he sets out from home: S’ele le baiser vos consent, Le surplus je vos en desfent, Se laissier le volez por moi  (CdG 547–9) [If she allows you a kiss, I forbid you to take anything more, if for my sake you are willing to forsake it]

This advice almost immediately backfires, as Perceval’s first encounter, with a damsel in a tent, is the source of incomprehension all round – he misapplies his mother’s commands which he doesn’t understand, the damsel is upset by his actions and the damsel’s ami misinterprets the situation as adultery and punishes her. The simulacrum of rape which occurs here replaces any possibility of a genuine sexual encounter. Later, after winning Blancheflor’s love, Perceval turns his back on the possibility of marriage and legitimate sexual activity (unlike Chrétien’s earlier heroes, Erec and Yvain) and leaves her to seek the mother who had forbidden him the ‘surplus’.26 Gauvain, meanwhile, fares even worse in Chrétien’s part of the text, since the damsels whom he encounters are all for one reason or another romantically inappropriate, beginning with the Pucelle as Petites Manches, far too young for his relationship to her to be anything more than a parody of love service. Things appear more promising when he is left alone with the sister of his enemy, the king of Escavalon, and they begin speaking of love (CdG 5817–31); but, just as she has accepted Gauvain’s suit, his identity is uncovered and he is forced to leave. The next damsel to cross his path is the Orgueilleuse de Logres, whose hatred for knights means that his courtly platitudes fall on deaf ears. During these adventures, even the pursuit of a white doe ends in failure, with his horse symbolically losing a shoe in the chase.27 This inability to instigate sexual encounters needs to be understood as one aspect of a generalised sterility which permeates Chrétien’s text and its various settings. The Waste Land of the Grail Castle is prefigured in the opening lines by the Waste Forest of Perceval’s family domain. In the same way, the wound which forced Perceval’s father into exile will be paralleled by that of the Fisher King – Perceval’s father was struck ‘parmi les jambes’ [through the legs] (CdG 436), the king ‘parmi les hanches ambesdeus’

26 Balsamo (1993: 281) sees Perceval as embodying ‘a perpetual, unremitting stagnation of desire’. 27 As Peter Haidu (1968: 249) observes, the text repeatedly generates irony ‘at the expense of Gauvain’s habit of involving himself with ladies from whom he can obtain no satisfaction’.

136 THE

CONTE DU GRAAL CYCLE

[through both thighs] (CdG 3513), so that both are paralysed.28 Moreover, the location of the wound in each case suggests a sterility in symbiosis with that of the lands: the father, we are told, was subsequently unable to reign effectively, with ruinous consequences: Sa grant terre, ses grans tresors, Que il avoit comme preudom, Ala tot a perdition, Si chaï en grant povreté.  (CdG 438–41) [His extensive lands and many possessions, which he held as befits a worthy man, all went to ruin; he fell into great poverty.]

The same connection between reproductive and environmental sterility is insisted upon by the Hideous Damsel who berates Perceval for his failure at the Grail Castle, forecasting that as a result: Dames en perdront lor mariz Terres en seront essilliees Et puceles desconseillies  (CdG 4608–10) [Ladies will lose their husbands, lands will be lain waste and damsels abandoned]

As we have seen, the Continuations refuse the narrative programme suggested in the Hideous Damsel’s harangue, which would see Perceval condemned to irrevocable failure in his Grail Castle adventure. As part of the project of rewriting, the theme of sterility receives short shrift; on the contrary, elements of Chrétien’s narrative which had foregrounded the theme are targeted for transformation. Thus, in Gauvain’s visit to the Grail Castle found in Branch V of the First Continuation, the Fisher King strides into the room to greet his guest, a ‘chevalier grant et menbru’ [tall, strong-limbed knight] (C1 LR 17333) rather than the bed-ridden cripple of the first Grail scene. This sudden, unannounced change to the inherited schema suggests a deliberate aim on the part of the First Continuator to demonstrate the extent to which continuation can free itself from the strictures of narrative precedent in its drive to generate additional material. The Grail is no longer solemnly carried before the hero’s gaze, tantalisingly out of reach; instead it appears, self-moving, to fill the diners’ plates with food: Lors vit parmi un huis entrer Le riche Graal, qui servoit 28 On the basis of this parallelism, some critics have suggested that either the Fisher King or his father might be identifiable as Perceval’s father, the hero’s lineage thus being tainted by incest. See Poirion (1973); Rider (1998); Duggan (2001: 79–81).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

137

Et mist le pain a grant esploit Par tout devant les chevaliers.  (C1 [LR] 17348–51) [Then he saw the rich Grail enter through a door to serve the food; it promptly put the bread down in front of the knights.]

Like the Fisher King before it, and in contrast to the more ceremonial character of Chrétien’s Grail scene, the Grail enters unaided from an adjoining room. The First Continuation is not the only text to imagine the Grail as a self-moving, cornucopian dish, but its vitality, mobility and fecundity sit especially well with the conception of narrative dynamic found in this text. Moving from one diner to the next, the Grail imitates the movement of the continuation from hero to hero, story to story, across its different branches; the proliferation of narrative thereby effected is matched by the proliferation of food that the Grail provides, and which exceeds the boundaries of the text: Toz les mes qu’il lor aporta, Seignor, ne vos conterai ja, Quar qui touz les aconteroit, Certes molt vos ennuieroit, Fors qu’il mengierent a loisir.  (C1 [LR] 17371–5)29 [Lords, I will not describe to you every dish that it brought them, for if someone were to recount them all it would surely bore you; I will merely say that they ate what they wanted.]

We saw in Chapter 1 how the cyclic, centripetal dynamic of the Second Continuation and the Long Redaction of the First led them to imagine Grail scenes which either ignored this visit’s departures from Chrétien, or else attempted a reconciliation through amalgamation of elements drawn from the different versions. Yet Wauchier de Denain is just as keen as the First Continuator to shake off Chrétien’s theme of sterility. The Second Continuation, however, achieves this aim by different means, consistent with its concern for coherence between narrative development and the textual past. The First Continuation used the Grail scene, the narrative centre of Chrétien’s text, to make a statement of its willingness to countenance inconsistencies with the narrative programme inherited from the pre-existing text. The Second Continuation avoids inconsistency by setting its rewriting within the context of the narrative of redemption which it uses to rehabilitate its hero. Thus, as we saw above, Perceval returns to his mother’s home to lay to rest the filial guilt which has been haunting him since the start of the cycle, preventing him from achieving his destiny. This return home is preceded by another

29 For a discussion of the ongoing association of the Grail with food in thirteenthcentury Arthurian narratives, see Berthelot (1996).

138 THE

CONTE DU GRAAL CYCLE

chance arrival in a location loaded with diegetic significance, his return to Biaurepaire and his beloved Blancheflor. Perceval’s return to Biaurepaire, like his return to the family home, allows the text to demonstrate the progress he has made towards the status of preeminent knight in the world. On his first visit there, both the castle and the land around it are described as ‘gaste’, hewn from the same vocabulary as the Waste Forest or the Fisher King’s domain: Car s’il eut bien defors trovee La terre gaste et escovee, Dedens rien ne li amenda […] Ensi trova le chastel gaste, Qu’il n’i avoit ne pain ne paste Ne vin ne sydre ne cervoise.  (CdG 1749–51; 1771–3) [For just as he had found the land devastated outside, so the interior was no better… Thus he found the castle lain waste: there was neither bread nor meat; neither wine, nor cider, nor beer.]

When he returns, the landscape has changed so much that he fails to recognise it: Molt se mervoille Percevaux An quel païs il est antrez, Que bien avoit deus anz passez Que n’avoit mes terre veüe Qui de toz biens fust si vestue, Ne replanie ne pueplee   (C2 22560–5) [Perceval wondered what land he had entered, for it had been two good years since he had seen fields so filled with crops and abundantly peopled]

The rivers are full of fish; the town is full of people and all sorts of goods. The sterile environment of Chrétien’s text has been rewritten into a fertile community, but here is no inconsistency – on the contrary, the transformation is the logical consequence of Perceval’s previous passage at Biaurepaire, where he had freed the town from a crippling siege. This successful restoration of fertility and life to the previously barren castle of Biaurepaire is thus made to stand as a precursor to the transformation that Perceval is destined to work on the lands and on the body of the Fisher King. Simultaneously, it serves as a metaphor for the transformation wrought by the continuator upon the textual mass which he had inherited, turning the centrifugal Conte du Graal and First Continuation corpus into a fully fledged cycle centred on the successes of its hero, Perceval.

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

139

Perceval’s returns to the landscapes of the past in the Second Continuation invest these locations – Grail Castle, family home, Biaurepaire – with a structural importance in underpinning the cyclic dynamic. The later Continuations, by repeating the trick, further reinforce this dynamic. The Gerbert Continuation, once again, shows a particular concern for building its narrative from the material of the textual past, and thus it not only returns its hero once more to Biaurepaire, but places this visit immediately after a return to the castle of Gornemant. The itinerary mirrors that followed by the hero in Chrétien’s part of the text, but again the passage of time allows the continuator to mark the progress which both narrative and hero have made in the intervening period: having begun as Gornemant’s student, Perceval now becomes the saviour of his bloodline, leading his former mentor’s sons into battle against their enchanted foes. As Blancheflor’s uncle, Gornemant then travels with Perceval to Biaurepaire to witness and condone the marriage of his niece to his protector, which offers Gerbert an opportunity to underline once more the beneficial changes effected by Perceval on the landscape. If ‘gaste’ had been the keyword in Chrétien’s description, ‘plein’ and its derivatives play a similarly striking role here: Et les forés grans et plenieres Sont d’autre part, de bestes plaines Et li pré et les terres plaines  (CG 6200–2) [And the large and fertile forests are full of animals, as are the fields and the plains]

Alongside the vocabulary of plenitude, the word ‘riche’ is employed to describe the castle (6192, 6196), the orchard (6206) and the vineyard (6210). None of this comes as a surprise to either audience, hero or narrator, as the last of these reminds us in now-familiar terms: Autre fois l’avons devisé. Quant Perchevaus l’ot avisé, Bien a reconut le païs.  (CG 6211–13, my italics) [We have mentioned all this earlier. When Perceval saw it, he recognised the region.]

For Gornemant, however, all this is new information: Mais Gornumans fu esbahis, Car ainc puis n’i avoit esté Que Clamadeus avoit gasté La terre et le païs d’entor.  (CG 6214–17) [But Gornemant was stunned, for he had not returned there since Clamadeus had lain waste to the land and the surrounding region.]

140 THE

CONTE DU GRAAL CYCLE

The vavassor’s surprise is explicitly contrasted with the prior knowledge enjoyed by audience and hero, as the narrator again reminds us of Perceval’s Second Continuation visit to Biaurepaire: Mais ore est de si riche ator Con vous par mon dit le savez Et autre fois oï l’avez Deviser et la grant richece Et de la vile la noblece, Si seroit anuis du retraire Autre fois, por che m’en veil taire.  (CG 6218–24) [But now it had become very prosperous, as you know from my account, and previously you have heard both the wealth and the splendour of the town described; it would be boring to recount it again, and so I wish to say no more on the matter.]

Gornemant, a character not seen since early in Chrétien’s narrative, now becomes a witness to the transformations effected by the continuators upon the text. His joy at the restoration of the land’s fertility, along with his approval of the marriage of Blancheflor and Perceval, create a sense of harmonious transition between the different generations of textual material, condoning the changes operated by the continuators. This, of course, is all in keeping with the way that Gerbert presents his own work as a faithful continuation of the original author’s, a necessity born out of the master’s untimely death.30 Curiously, in Gerbert’s conception, marriage between Perceval and Blancheflor does not lead to consummation, as the two agree on their wedding night to maintain their chastity. The next morning, Perceval hears a divine voice condone his decision, telling him that if he keeps his virginity, his descendants will achieve great things. The words used are ‘de ta lignie’ [of your lineage] (CG 6906), making it unclear whether what is envisaged is a miraculous virgin birth or an indirect descendance via his sister, for instance. This lack of clarity appears quite deliberate, since the whole passage seems designed to titillate the audience’s curiosity and suggest a radical new narrative direction in which the text could move. From Perceval’s lineage, the voice declares, a daughter will be born who will marry the ‘riche roi’ (6909), but who will find herself in danger of being burned or exiled (6911–12). She will be saved by her son, and will have other children who will conquer many lands (6913–16). From here on in it becomes apparent that Gerbert has in mind some form of the Crusade cycle, which recounts Godefroy de Bouillon’s successes in the First Crusade and the exploits of his supposed ancestor the Swan Knight: one son will be changed into a bird, to his parents’ chagrin 30 See Bruckner (2009: 60–1) for a consonant reading of this episode, focusing especially on the relationship between narrator, authors and audience.

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

141

(6917–21); the eldest brother, following in the footsteps of his ancestor Perceval, will marry a damsel whose lands he will have won back for her in battle (6922–6); this marriage will produce a daughter who will have a fruit which pleases all (6927–30); finally, most gloriously, this daughter will have three sons who together will conquer Jerusalem (6931–3). Thirty lines of text have sufficed to dispatch the narratives of four generations of descendants, and threaten to drag the Arthurian narrative into the ambit of chanson de geste. This moment offers the audience a glimpse of a different Conte du Graal cycle, one more akin to Perceforest’s successions of genealogies and civilisations. It hardly needs saying that such a genealogical tradition failed to materialise within the Conte du Graal cycle’s various permutations, its scale being far vaster than that on which the cycle was imagined, and its narrative dynamic quite different from the Percevalian biography which forms the centripetal core of the corpus.31 Nevertheless, Gerbert deliberately entertains the possibility of such a text. Nor is this the only moment in the cycle in which such alternative Conte du Graal narratives are evoked. We saw in the previous chapter that the Elucidation and Bliocadran suggested other models of cyclicity, and that perhaps for this reason they remained marginal to the cycle’s textual tradition. The narrative value of progeny, it seems, is doubleedged. It provides potential for the production of new material, free from the programmatic strictures attached to the figure of Perceval, the established narrative centre of the cyclic project from the Second Continuation onwards. However, because of this distance from the cyclic centre, the narratives of descendants or ancestors can only ever be interludes – excessive development would recast Perceval’s adventures as merely one generation of many, and risk unbalancing the entire edifice. If Perceval’s progeny were to become a narrative focus, then the resolution of the hero’s ongoing Grail adventure would no longer be the end point towards which the cycle constantly claims to be striving. The First Continuation, of course, dispenses with Perceval altogether, finding that the adventures of other characters provide more scope for its 31 While agreeing that Perceval’s life forms the hub of the corpus, Gaggero (2008a) points out that the genealogical sweep of this passage fits into a general tendency in the Gerbert Continuation to explore the hero’s genealogy. As well as this prophesied conquest of Jerusalem by Perceval’s descendants, he discusses passages which explore Perceval’s ancestry, especially his mother (here named as Philosophine), who was responsible for bringing the Grail from the Holy Land to Britain. Consequently, Gaggero considers the TV narrative to gesture towards ‘una prospettiva compiutamente ciclica’ [a completely cyclical perspective] (277): the wheel of history will ultimately return the Percevalian lineage to the land of its origins, by the same token restoring Christian dominion over the Holy Land. Yet, Gaggero also notes, this genealogical, organic cyclicity exists only as a series of intertextual allusions: no manuscripts, for instance, attempt to combine the Conte du Graal cycle and the Crusade cycle.

142 THE

CONTE DU GRAAL CYCLE

project of narrative invention from scratch. Frequent focus on Gauvain ensures heroic continuity with Chrétien’s text, but this is evidently not considered incompatible with extended narratives involving other protagonists. The clearest instance of this tendency is the story of Caradoc which makes up Branch III of Roach’s edition. No previous version is known of this tale, which begins with the hero’s namesake and father and ends with his own coronation and reign. The central narrative is concerned chiefly with exploring the theme of filial responsibility; like Chrétien’s Perceval, Caradoc is punished for showing insufficient loyalty to his mother (though retribution here is on the scale of petty human vengeance, rather than that of cosmic providence as with Perceval). Where the later Continuations engage with Chrétien’s thematics of maternal authority by developing and defusing Perceval’s guilt, the First Continuation transfers the issue across to another character, situating it within a different problematic altogether. Through Gauvain, the First Continuation also develops the theme of filiation on a larger narrative scale, one which transcends the boundaries of individual branches and provides a surprising amount of narrative continuity. In Branch II, the tale of Arthur’s siege of the castle of Branlant turns into a quite different narrative when Gauvain leaves the army camp to go riding. He comes to a tent where he meets a beautiful damsel, whom he later discovers belongs to a family whose relative he had killed; the episode thus recalls both Perceval’s mock-courtly ineptitude with the damsel in the tent at the start of Chrétien’s narrative and Gauvain’s encounter with the sister of the king of Escavalon.32 The deviation from Chrétien’s pattern of blocked sexuality is then all the more significant, as the maiden reveals that she has been saving her body for Gauvain, with whose description she has fallen in love. Indeed, the sterile, mock-sexual encounters of Chrétien’s section are replaced with what might be termed an aesthetics of consequence. The damsel’s brother, swearing to avenge this perceived disgrace to the family’s honour, follows and almost kills Gauvain. After they have agreed to resume their combat at a later date, the text follows Gauvain back to Arthur’s camp, announcing: Ici remaint de Bran de Lis Et de sa seror au cler vis Qui remaint anceinte d’anfant  (C1 [LR] 6607–9) [Here the text leaves Bran de Lis, and his sister with the fair face who is left expecting a child]

The casual way in which this information is let slip initially conceals its significance for the narrative; after all, the text thus far has accustomed its 32 On the damsel in a tent as a recurrent motif through the cycle, see Bruckner (2009: 86–115). She suggests reading this motif as a metaphor for the narrative architecture of the corpus (220).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

143

audience to various blind alleys, with all sorts of narrative threads remaining resolutely undeveloped. Yet this time the First Continuation chooses not to forget. In Branch IV, seven thousand lines after the conception is announced, Gauvain finds himself in a castle belonging to Bran de Lis, and their enmity is revived. In a memorable scene, the resumption of hostilities is interrupted by the mother’s introduction of the child between the two opponents. As Gauvain’s son and Bran’s nephew, the child represents a blood bond between the men, and the mother deliberately reminds the antagonists of the kinship network which connects them when she tells her son: Et por Dieu proiez vostre pere Qu’il ait merci de vostre mere, Que por Dieu lait que il n’ocie Mon frere, plus l’ain que ma vie.  (C1 [LR] 14951–4) [And beg your father, in the name of God, to have mercy on your mother, and not to slay my brother, whom I love more than my life.]

Her refusal to take sides in the dispute, underlined by the homophony of the terms pere–mere–frere, is rewarded by the intercession of the boy’s greatuncle, Arthur: the sight of the child reaching out to touch the shining swords convinces the king to intercede and make peace. The narrative of Bran’s vendetta having been concluded, Gauvain’s son disappears once more into the background, his narrative function apparently accomplished. Yet it seems that children, once they have been born into fiction, refuse to be ignored. At the conclusion of the expedition to Chastel Orguelleus which makes up the subject of the rest of the branch, the boy is kidnapped, prompting Arthur and his knights to mount a search and rescue operation. The only characters to opt out of this quest are Keu and, surprisingly, Gauvain himself, as if the father fears becoming too closely associated with – even eclipsed by – his son’s narrative. Gauvain’s Laïus-like reflex is initially endorsed by the text, which weaves the tale of his visit to the Grail Castle out of his return to court to attend the queen. Yet later, following Gauvain’s Grail Castle failure, the narrator re-activates the narrative possibilities embodied by the mysterious disappearance of the son, in a long passage of which the following extract gives a flavour: Ne [ne m’orroiz parler] de monseignor Bran de Liz, Qui fu de son neveu marriz, Si sai molt bien de veritez C’onques par euls ne fu trovez. Por ce qu’il me covient entendre A la grant matire comprendre, Qui l’enbla ne qui le norri Ne m’orroiz ja parler ici;

144 THE

CONTE DU GRAAL CYCLE

Ne de celui qui l’escola, En quel maniere l’enseingna Sor totes riens a tenir chier Ses armes et son bon destrier  (C1 [LR] 17865–76) [Nor (will I tell) of sir Bran de Lis, who was sad for his nephew; indeed, I know very well that they (the search party) never found him. Because I have to concentrate on including the whole of the great tale, you will not hear me speak here of who hid him, nor who fed him, nor how or by whom he was taught to value his arms and good horse above all things]

The narrative technique here is similar to that employed by the Gerbert Continuation in prophesying the future of Perceval’s lineage: in the space of a few lines, the boy has become a young man learning the basics of knighthood. The speed with which these narratives are developed is all the more striking in comparison to the thousands of lines that will be devoted to Perceval’s looping journeys to and from the Grail Castle.33 Moreover, in this instance, the narrator is quite explicit that these matters which he deigns to evoke will not be developed further. They are to remain potential narratives, a sign of the overabundance of material at the disposal of the cornucopian, centrifugal First Continuation. Again the text plays cat and mouse with its audience, as it begins despite these omissions to recount the early exploits of Gauvain’s son. At this point, he shows all the signs of becoming a new Perceval around whom a new narrative may be built: he displays an inherited natural aptitude for chivalry; he has grown up without a father and does not know his own name; and he shares the young Perceval’s tendency to naivety, refusing to use his shield in combat in case it should get damaged. Yet this narrative potential is once more nipped in the bud by Gauvain, who arrives on the scene and recognises his son, bringing him back to court but in the process overshadowing him: if the young man is loved at court, it is ‘por le preudome qui l’avoit / Engendré’ (C1 [LR] 18324–5) [for the worthy man who had fathered him]. As we know, while it may flirt with the idea of a genealogical romance centred on the lineage of Gauvain (the final narrative of the First Continuation concerns his brother Guerrehet), the cycle ultimately chooses Perceval as its central character, and from the Second Continuation onwards other heroes are relegated to a secondary role. Thus, when Gauvain’s son reappears in the Second Continuation (as he does twice), he functions as a representative 33 Mireille Séguy, commenting on further instances of this narrative technique in the First Continuation, captures well its singular effect on the reader: ‘Le résultat est un curieux mélange d’effets d’annonce, de dénis d’écriture, d’allusions obscures et d’abrupts raccourcis, dont le lecteur ne vient pas à bout sans un léger tournis’ [this results in a strange mixture of hype, refusals to write, obscure allusions and jarring shortcuts, which leaves the reader feeling somewhat dizzy] (Séguy 2001a: 300–3; 301 for the citation).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

145

of the Arthurian court, communicating information to the relevant hero – first to Perceval, then to his own father. These appearances are nevertheless of considerable interest, since they demonstrate an aspect of the narrative ethics of this Continuation which we have not yet discussed. We have seen Wauchier’s concern with promoting his text as the ‘faithful’ development of the initial Perceval material, and it seems he was also conscious of a duty of fidelity towards the wider Arthurian intertext. For as we saw above, Gauvain already had a son in Arthurian romance, whose earliest surviving appearance is in Renaut de Beaujeu’s Bel Inconnu. Dated before 1200, this text would apparently be roughly contemporary to Chrétien’s Conte du Graal, and it or some version of the same story was known to Wauchier. Thus, in the Second Continuation, Gauvain’s son is fully cognisant of his own identity: Li Biaux Desconneüz ai non; Einsint m’apellent li Breton. Messires Gauvains est mom pere.  (C2 22387–9) [My name is the Fair Unknown; this is what the Britons call me. Sir Gauvain is my father.]

The reference to ‘li Breton’ is doubly significant, encompassing both Arthurian society within the text and the literary tradition of the matière de Bretagne within which the character exists. The second time he appears, meeting Gauvain, his self-identification is even more comprehensive: ‘Sire,’ fait il, ‘j’ai non Ginglains, Vostre filz, que li roi Artuz Mit non li Biaux Desconneüz’  (C2 31070–2) [‘Sire,’ he said, ‘I am called Guinglain, your son, to whom King Arthur gave the name Fair Unknown’]

The First Continuation either did not know, or did not care, about the Bel Inconnu intertext, and the only name given to Gauvain’s son there is the nickname ‘Lioniaus’, which he receives for his wildness and his beauty (C1 [LR] 18082–4).34 The narrative dynamic of each continuation therefore seems well embodied in its presentation of Gauvain’s son. The First Continuation, the unpredictable, untameable offspring of the cycle, describes him as every inch the bold young man: En la chambre com hom sauvages Se porfichoit, et fu trop biax  (C1 [LR] 18082–3) 34 The awkwardness of attaching a different name to a character well known by the time the surviving manuscripts were written is suggested by three codices (MQU) which give him the name ‘li oisiaux’ [the bird] at this point.

146 THE

CONTE DU GRAAL CYCLE

[He wandered through the chamber like a wild man, and was very handsome]

Like Lioniaus loitering in the communal chamber, the First Continuation lies confidently in the manuscripts, unapologetically interrupting the adventures of Perceval. Similarly, it seems fitting that, by the time of the Second Continuation, Guinglain should be represented as calmly accepting his filial and feudal duties, understanding and embracing his place in Arthurian genealogy.

Narrative Memory and the Scandal of Recognition Terence Cave reminds us that recognition scenes have the capacity to provide both reassuring closure and troubling misgivings. Reading Odysseus’s scar in the homecoming scene of the Odyssey as ‘a sign that the story, like the wound, may always be reopened’, he provocatively asks: What if Odysseus were a false messenger, a well-equipped spy or conman? What if Oedipus were not the murderer of Laïus (the evidence is only circumstantial), or what if he were some other exposed baby substituted for Oedipus by incompetent or artful shepherds?35

Such questions go to the heart of the contract between audience and text: the ‘willing suspension of disbelief’ is entered into on the understanding that reliable narrative satisfaction will be provided by the text. They appear particularly well suited to the Conte du Graal cycle; after all, as many critics have noted, Perceval’s cognitive abilities, his capacity for reading signs correctly, are both a central theme and a narrative motor of Chrétien’s part of the cycle.36 Moreover, the audience’s understanding of events is allowed to develop in tandem with that of Perceval. The narration of the Grail scene, for instance, is framed through his gaze, so that the audience’s apprehension of the central mystery of the text is mediated by the limits of the hero’s understanding.37 Similarly, the repeated staging of meetings between the hero (whether this be Perceval or Gauvain) and members of his family forces the audience into the position of learning gradually, along with the character, about his kinship ties. In order to find the origin of this tendency, we must return once more to the foundational early exchanges between Perceval and his mother. Having detailed the young man’s chivalric inheritance (on both maternal and paternal sides) and the events that brought the family to live in Cave (1988: 24 and 43). See especially Pickens (1977); Burns (1987/88); Bloch (1983, Chapter 6). 37 Cf. Cazelles (1996: 11): ‘Because Chrétien narrates his tale through the eyes of Perceval, and because Perceval is ignorant, the reader of the romance must pursue clues whose full meanings are not entirely clear to the protagonist himself.’ 35 36

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

147

the Waste Forest, she goes on to reveal that Perceval had two brothers, who themselves became knights: Et vos, qui petis estiiez, .II. molt biax freres aviiez; Petis estïez, alaitans, Peu aviiez plus de .ii. ans.

(CdG 455–8)

[And you, who were young, had two very handsome brothers; you were young, not yet weaned, barely two years old.]

In a single move, the text introduces the characters of the brothers and accounts for the hero’s lack of knowledge of them. The insistence on Perceval’s infant status at the time, with the repetition of ‘petis estiiez’ [you were young] and the unusual precision in establishing his age, attempts to explain why these narratively significant events should have left no mark on his memory.38 Similar excuses are made during Gauvain’s encounter with his grandmother, mother and sister at the Castle of Marvels. In this case, the hero’s inability to recognise his own mother is implicitly ascribed to his belief that she has been dead for years. Gauvain, like Perceval, is shown that ‘les mors as mors, les vis as vis’ can only ever be a temporary solution to the problem of living with the past. Meanwhile, we learn from Guiromelant that the existence of Gauvain’s sister could not have been known to him since she was still in her mother’s womb when they passed beyond the ken of the Arthurian court (CdG 8757–63). Clarissant herself maintains the pattern of recognising (acknowledging) the importance of kinship ties while explaining the hero’s failure to recognise (identify) them when she tells the incognito Gauvain: Lasse! il ne set se je sui nee Mes freres, n’onques ne me vit.  (CdG 9038–40) [Alas! My brother does not know whether I was born, and has never seen me.]

By repeatedly staging and excusing these lapses in narrative memory, Chrétien’s text appears to be protesting too much, deliberately foregrounding the troubling phenomenon. Here again, as we have found throughout this chapter, issues of literary descendance and fidelity to the textual past constituted by audience memory and anterior narrative are never far from the surface, lurking in the shadows of the diegetic theme of kinship. 38 As might be expected, the manuscripts vary in the detail, but not in the spirit, of v. 458. L gives ‘N’aviez pas plus de .ii. ans’ [you were not more than two years old], R has ‘N’aviés encore pas .iii. ans’ [you were not yet three]. Only B betrays the image of Perceval ‘alaitans’, with the line: ‘Peu aviiez plus de .v. ans’ [you were barely five years old].

148 THE

CONTE DU GRAAL CYCLE

In the homecoming scene from the Odyssey, the identifying scar is not quite able to provide authentic recognition; Odysseus is required to undergo further tests, reveal memories, demonstrate techniques only he could know. Acceptance could, it appears, be deferred indefinitely (like Penelope’s shroud), and it is only when his wife is finally convinced by one last piece of evidence that the narrative is allowed to move on, and take for granted the stranger’s identification as Odysseus. As Cave notes, this produces an uneasy sort of resolution: ‘The narrative has to be saturated with signs in order to exorcise the threat [of imposture], but the saturation sufficiently reveals that it cannot wholly be exorcised.’39 I would suggest that something similar is happening in these scenes, where Perceval’s kin insist on his lack of knowledge of them, as if to absolve him of the awkward questions that might be lurking in the audience’s mind: is it really possible that Perceval has never met, or heard of, any of his family? And if so, how secure are the identifications and recognitions which these characters offer to the hero, and to us?40 As Cave suggests, such speculation takes us beyond diegesis into the realm of audience reception, where the artful conventions of literature rub up against the messy uncertainties of life. He argues that the ‘real’ drama of recognition and imposture constituted by the well-known case of Martin Guerre ‘is shaped ab initio by existing narrative structures and interpretations. Amphitryon is one of his siblings, Odysseus another, perhaps Oedipus another again.’41 And since his story touches on both the diegetic and extradiegetic content of this chapter, a further place might be found in this unlikely family tree for Baudouin IX de Flandres, first Catholic emperor of Constantinople. Baudouin represents the missing link in the narrative of succession between Philippe and Jeanne, the two bookending patrons of our cycle: he inherited the county of Flanders from Philippe, his maternal uncle, and bequeathed it to his daughter Jeanne. The term ‘missing link’ is especially appropriate to this example, since Jeanne (like Perceval) was robbed of her father by chivalry and warfare. The Fourth Crusade took Baudouin away from Flanders in 1202, at which time Jeanne, again like Perceval, was barely two years old. Baudouin is believed to have died in captivity after the Frankish defeat at Adrianople in April 1205, and when news eventually filtered back to Flanders the succession passed to Jeanne as the count’s elder daughter. The recognition drama which interests us occurred in 1225, towards the end of the period in which it is thought Jeanne may have commissioned

Cave (1988: 24). Cazelles (1996) bears witness to the kinds of critical doubt these issues can raise; reading the romance as a factional dispute over the heroic subject Perceval, she pays close attention to Perceval’s supposedly authoritative Grail interlocutors (the Fisher King, his cousin, the Hideous Damsel and his hermit uncle) in order to demonstrate that their aims are less noble and more suspect than may appear. 41 Cave (1988: 15). 39 40

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

149

Manessier to complete the Conte du Graal cycle.42 In his Chronique rimée, Philippe Mouskés tells how, in this year, there appeared in Flanders a hermit around whom developed the rumour that he was the returned Baudouin. He was visited by various members of the local aristocracy, several of whom initially confirmed the identification while others noticed troubling discrepancies. For instance, this Baudouin was shorter than remembered; the fact was explained by age and the time he had spent in captivity. Intriguingly, given the links we have observed between recognition and literature, one cause for concern was the library which the hermit kept in his dwelling, when Baudouin, ‘ce sçet-on bien, / Ne sot laitres qui valust rien’ [as is well known, had no learning to speak of].43 Indeed, Mouskés himself draws the analogy between contemporary events and the matière de Bretagne, comparing the popular fervour that seized the urban populations of Flanders with the conviction of the Bretons that Arthur would one day return: A Valenciènes l’atent-on, Ausi comme funt li Breton Artu, qui jà ne revenra.44 [He is awaited at Valenciennes, just as the Bretons await Arthur, who will never return.]

Matters came to a head when Jeanne, chased from her lands by the popular movement in favour of ‘Baudouin’, appealed to Louis VIII of France. The king had the pretender brought before him, and proceeded to ask him a number of questions to which Baudouin could have been expected to know the answers. Having found him unable to answer these, he deferred the interrogation overnight. ‘Baudouin’ took flight, was captured and eventually publicly executed in Lille by Jeanne, who had the body exposed for several days. This act calls for a different kind of recognition, that of the ruler’s violence inflicted on the aberrant body. Jeanne’s public punishment of imposture is intended for the urban audience, admonishing them to recognise her authority as uncontested.45 The conclusion drawn by Cave in connection to Martin Guerre applies just as strongly in this case:

See Roach’s Introduction in Roach (1983 [V]: xiii). Mouskés, vv. 24735–6. 44 Mouskés, vv. 25201–3. 45 These warnings were addressed to very real issues of class struggle. Valenciennes, for instance, took advantage of the uproar to declare itself a commune, and was severely punished as a result. Robert Lee Wolff (1976: 300) emphasises the socio-political import of the involvement of the menus gens in the debate: ‘What had presumably begun as a conspiracy among the aristocracy appears to have become almost a proletarian revolution.’ 42 43

150 THE

CONTE DU GRAAL CYCLE

Recognition, then, in this ‘true story’, is not the recovery for good or ill of certain knowledge, nor the reassuring restoration of the co-ordinates of kinship and social position. It unmasks a crisis, a perpetual threat of imposture on the one hand and arbitrary law and coercion on the other.46

Indeed, even the coercion of political authority was not sufficient to silence the whispers that the hermit really was Baudouin and that Jeanne had committed parricide.47 The stakes of recognition in the anecdote are bound up with issues of power and inheritance: to be recognised is to be powerful; to be powerful is to have the authority to recognise. Similarly, in the Conte du Graal cycle, recognition scenes allow us to read each author’s anxiety over the status of his own contribution to the corpus. If, as the afterlife of the ‘False Baudouin’ demonstrates, the narrative is never entirely closed to an audience’s capacity for re-interpretation, we can legitimately look to the recognition scenes of the corpus for a sign of how its authors went about convincing their audiences that their work was an authentic part of the text, that it merited a place in the literary genealogy of the cycle. In the First Continuation, as ever, we find traces of rebellion against the potentially stifling poetics of filial duty represented in and by Chrétien’s material.48 Two such instances of insubordination stand out as especially worthy of comment. As we saw above, the narrative of Gauvain’s dalliance with the Pucelle de Lis wends its way through the First Continuation, constituting its only recurring diegetic thread. In the initial encounter between the pair, the damsel is greeted by Arthur’s nephew, whose identity is initially unknown to her. She replies: Sire, cil qui fist soir et main Saut et gart monseignor Gauvain, Et vos autresi beneïe.  (C1 [LR] 6243–5) [Sir, may He who made the evening and morning watch over Sir Gauvain, and bless you also.]

Pressed as to why she has greeted Gauvain first, when ostensibly he is not there, she clarifies that ‘Tot ainsint respon je mon pere / Con je vos fait, et a mon frere’ (6251–2) [I respond to my father and brother just as I have done Cave (1988: 15). See Wolff (1976: 299): ‘Despite his alleged confession and suspicious behavior, the populace remained most reluctant to believe that the impostor was not Baldwin. Down through the centuries, Jeanne has often been accused of parricide, and remains one of the most controversial figures of Flemish history.’ 48 Tether (forthcoming) terms the First Continuation an ‘Exploitative Prolongation’, one which takes advantage of the unfinished state of the text it is continuing in order to follow its own narrative agenda. 46 47

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

151

to you]; she has fallen in love with Gauvain through the good she has heard of him, and ‘Por ce li ranz saluz avant / Qu’a mom pere ne a autrui’ [for this reason I wish him well before my father or any other person] (6266–7). In a deft reversal, the motif of the hero’s encounter with a hitherto-unknown family member is transformed; here, the character encountered is a hitherto-unsuspected love-interest who has been waiting patiently for Gauvain to arrive. A certain amount of comedy is provided by Gauvain’s desperate desire, having heard her story, to be recognised by the damsel as the one she has been expecting. He explains carefully the oath that prevents him from revealing his name and yet, in the right circumstances, will allow him to do so: Franche pucelle, ainz a nelui Mes nons ne fu ancor nomez S’il ne fu avant demandez; Mais a nelui nou voil celer Puis qu’i[l] le plot a demander.  (C1 [LR] 6268–72) [Noble maiden, I have never revealed my name to anyone if I were not asked it first; but I would never wish to hide it once I have been asked.]

Taking the hint, the damsel asks him to name himself, and he is only too happy to oblige. Yet mere force of conviction is not enough, and the damsel must confirm his identity, which she does by comparing his features to those on a picture she has had made of him. Unlike Penelope, she finds physical resemblance to be sufficiently convincing. Yet this analogy obscures the complexities of the comparison being made in each case. Penelope, confronted with the stranger claiming to be Odysseus, must compare his features with those she remembers her husband as having. Her demand for further proof is an implicit condemnation of the reliability of her faculties of memory, after twenty years of separation. The Pucelle de Lis, on the other hand, places her trust in a visual confection, albeit one created by a former maidservant of Guinevere.49 The ‘real’ Gauvain is required to match up to the standards imposed by the artefact. Thus, where Chrétien’s text interrogates the possibility of achieving authentic recognition, the First Continuation resolves this issue by problematising the notion of authentication by which such recognition is to be achieved. Alexandre Leupin, noting that the portrait used for identification is itself a mere accessory to the body of tales concerning Gauvain which had first stoked the damsel’s interest, points out that:

49 Interestingly, this unnamed servant is said to be a ‘Sarrazine’ (C1 [LR] 6297). The authentication of the male chivalric subject is thus made to depend on the testimony of a character doubly marginal (as a female Saracen) to the world in which his name is made.

152 THE

CONTE DU GRAAL CYCLE

la jeune fille ne confirme pas l’adéquation du nom et du corps par une nature, mais par quelque chose qui est déjà, et à double titre, pris dans la secondarité suspecte de l’imitant…le portrait est traversé de rhétorique, c’est le texte d’autres textes: un simulacre de simulacre.50 [The maiden confirms the identity of name and body not through anything essential, but through something which is already, twice over, caught in the suspect secondarity of imitation… the portrait is constructed out of rhetoric, it is the text of other texts: a simulacrum of a simulacrum.]

The damsel’s belief in her image is shown to be a leap of faith, of the same kind as that required of the audience confronted with the circular logic of the text. We know that the picture will match up to Gauvain because the narrator has informed us that the maidservant Si propremant avoit portraite L’imaige a lui an samblant faite, Que nus hom vivant n’i faillist A lui connoistre qui veïst La portraiture et lui ansamb[le], Einsint finement la resamble.  (C1 [LR] 6307–12) [had crafted the image so well in his likeness that no living man, having seen him together with the portrait, would have failed to recognise him, so closely did he resemble it.]

Yet we only recognise this man as Gauvain because the narrator has identified him as such. Our investment in the narrative requires us to believe that we are not the victims of a literary imposture. A similar conclusion regarding the act of faith required for narrative sense to operate may be drawn from the third branch of the First Continuation. The tale of Caradoc hinges on questions of filial responsibility. Having received the name of his father, king of Vannes, Caradoc serves his knightly apprenticeship at Arthur’s court. During the course of an adventure where he proves his valour, he meets the enchanter Eliavrés, who reveals himself to be the young Caradoc’s real father, having conceived him in the course of an ongoing liaison with the king’s wife, Ysave. Caradoc’s reaction is both surprising and revelatory. Initially, he rejects Eliavrés’s account as false, and in the Long Redaction account makes a deliberate statement of filial identification with the king: ‘Chevalier,’ fait il, ‘vos vantez De tel chose dont vos mantez,

50

Leupin (1979: 114).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

153

Qu’ainz mom pere ne deceüstes Ne o ma mere ne geüstes’  (C1 [LR] 7397–400)51 [‘Knight,’ said he, ‘this thing of which you boast is a lie, for you never deceived my father nor lay with my mother’]

Yet his first actions following this encounter belie his words: he heads to Nantes, where King Caradoc greets him as his son. Now, however, he finds himself unable to accept this identification. In the Short and Mixed Redactions, the qualifications involved are most instructive: ‘Sire, bien me devez beisier, Qu’an cest monde n’a rien vivant Cui j’aim plus de vos n’autretant, Mes vostre filz ne sui ge pas. – Si es! – Non sui. – Diz tu a gas? – Biax sire, ainz vos di verité.’ (C1 [SR A] 2506–11) [‘Sire, it is right for you to kiss me, for there is nothing in this world that I love more than you, but I am not your son. – Yes, you are! – I am not. – Are you joking? – No, sire, I am telling you the truth.’]

In the Long Redaction, the same idea is expressed a little later in the passage: Por ce ne vos tien pas a pere, Mais tote voie elle est ma mere, Et si nou di ge pas por ce Qu’il n’a home ou monde que ge Ain tant con je faz vostre cors.  (C1 [LR] 7467–71) [For this reason I do not consider you my father, but she is still my mother; I say this even though there is no man in the world that I love as much as I do you.]

Although Caradoc is clear with the king that he is not his son, he is also aware that the emotional ties built up over a lifetime of believing in their kinship have not suddenly vanished. And whatever the nomenclature of his relationship with the king whose name he bears, Caradoc makes the choice to side with him over the issue of his mother’s adultery, advising him to enclose her in a tower so that Eliavrés cannot get at her (C1 [LR] 7516–19). This move to suppress the scandalous infidelity of Ysave also appears to

51 The Long Redaction here is represented by EMQU; the other manuscripts contain Caradoc’s rejection of Eliavrés’s account, but not the identification of King Caradoc as his father.

154 THE

CONTE DU GRAAL CYCLE

represent an attempt by Caradoc to conceal the scandal of his own parentage. The tactic meets with little success, since Eliavrés’s magical powers allow him to visit Ysave at will, and their merry-making (complete with music and performance) is audible to all around, with no attempt at discretion (C1 [LR] 9656–65). In wishing to bury the secret, the two Caradocs have succeeded only in bringing it into the spotlight. It is tempting to read this scene at a further level, as a performance of the equally scandalous infidelity of the First Continuation to the text which it purports to continue. Caradoc’s belief is that imprisonment will suffice to maintain the lineal purity of the narrative; in the Short Redaction, this is brought out explicitly when he tells the king that, once Ysave is in the tower: Se il avient que ele ait oir, Ne seroiz pas an sopeçon Qu’il soit d’autrui se de vos non  (C1 [SR A] 2540–2) [If it happens that she has a son, you will not need to doubt that he is someone else’s and not yours]

Yet this strategy is shown to be hopelessly misguided, as the potency of the sorcerer permits him to circumvent all the obstacles placed in his way. He has already engendered one illegitimate heir in Caradoc, as well as three monstrous offspring from the animals with which he is forced to mate as punishment for his transgressions (C1 [LR] 9773–88). Michelle Szkilnik argues perceptively that Eliavrés’s powers to produce both offspring and intrigue are intimately bound up with the production of narrative: ‘Eliavrés… joue un rôle moteur: il est le maître d’œuvre du récit, une métaphore de l’écrivain’ [Eliavrés plays a key role: he is the architect of the narrative, a metaphor for the author].52 This potency contrasts sharply with King Caradoc’s inability, when placed in the same situations, to engender anything. Narrative, according to this episode, thrives through illegitimacy. Caradoc’s blood parents experience his role in Eliavrés’s capture and humiliation as a betrayal, and concoct a filicidal revenge plan; having tricked her son into life-threatening symbiosis with a serpent, Ysave goes so far as to reject him, describing herself in the Short Redaction as ‘moi qui estoie ta mere’ [I who was your mother] (C1 [SR A] 2652). Ultimately, though, faced with her son’s suffering, she repents and convinces Eliavrés to tell her how he may be cured. That the antidote requires the selfless actions of a damsel who loves Caradoc truly (C1 [LR] 11170–6) demonstrates again that the First Continuation tends to place the bonds of romantic love above those of familial duty. Moreover, Caradoc’s decision to side with the father he grew

52

Szkilnik (1988: 278).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

155

up with, rather than the one who engendered him, is rewarded explicitly when the king recognises his namesake as his heir. Caradoc’s tale suggests that the recognition of filiation is a performative, transformational act, one requiring an act of faith: a father is constituted by recognition as such by his children. By refusing to believe Eliavrés’s narrative, by taking on the inherited identity of King Caradoc, the hero forges his own genealogical destiny. Perceval’s attempts to do so in Chrétien’s text were thwarted by the haunting presence of his matricidal departure; the First Continuation, in presenting Caradoc as a positive counterexample to Chrétien’s character, implicitly contests its own dependence on the mother-text. Yet, at the same time, it appears to acknowledge that to shut Chrétien’s narrative away inside a tower, at safe remove from its own diegetically independent tales, can only ever be a temporary measure. Eventually, the scandal of filial recognition must emerge; eventually, the narrator concedes, he must return to the unfinished matter of the Grail Castle, ‘ce qu’avez tant atendu’ [what you have awaited so long] (C1 [LR] 17118). I argued in Chapter 1 that the Long Redaction of the First Continuation represents a rationalisation of the Short Redaction material in light of the cyclic model retroactively imposed by the subsequent Continuations. Traces of this are present in the different Redactions’ handling of the theme of filiation in the Caradoc tale. The first thing to note is that the Long Redaction’s additions to this branch function to draw attention temporarily away from Ysave’s adultery. The Short Redaction moves immediately from her imprisonment in the tower to the narrative of her ongoing nocturnal trysts with Eliavrés. The Long Redaction, however, interpolates two thousand lines between the two events, creating the initial impression that the infidelity theme has been successfully suppressed. Instead, narrative focus moves to Caradoc’s chivalric exploits, as he first rescues Guinier from an abductor and then triumphs, along with his friends, in a tournament at Arthur’s court which occupies 1500 lines of text. Once the dust has settled, the narrator appears to admit that the unusual length and level of detail in this scene were a diversion designed to distract the audience from remembering Ysave in her tower. The return to this plot thread is jarringly sudden, following on from the departures of the heroes from Arthur’s court: Li rois fu a sejor grant piece. De ma dame Ysave sa niece, La mere au vaillant Carados, Ne puis plus metre ariere dos.  (C1 [LR] 9583–6) [The king rested there for a long time. I can no longer hold back telling about his niece, Lady Ysave, brave Caradoc’s mother.]

The narrator presents himself as unwilling to proceed, but forced to bow to the exigencies of narrative integrity:

156 THE

CONTE DU GRAAL CYCLE

Pleüst a Dieu que ma matire Peüsse ci androit lessier Sanz mon conte trop ampirier.  (C1 [LR] 9600–2) [If only I could abandon my subject at this point without doing harm to my tale.]

The evocation here of the ‘conte’ creates an image of textual continuity quite in keeping with the Long Redaction’s concern for harmony between the different parts of the cyclic corpus. The insertion of the extra episodes provides the opportunity for the use of interlace technique, a move underlined by the vocabulary of v. 9613: ‘A mon conte voil revenir’ [I wish to return to my tale]. By introducing this structure, it integrates the Caradoc tale into the interlace dynamic of the cycle, and in doing so exemplifies the very act of long-term textual memory that it seeks to promote. Whether or not we take at face value the narrator’s claims to be ill at ease with his subject matter, the Long Redaction also marks itself out in its approach to the relation between Caradoc, King Caradoc and Eliavrés. The Short Redaction accommodates the paradoxical paternity of a hero with two fathers by having the narrator refer to the king as Caradoc’s father, while the characters are left to explore Eliavrés’s paternity claim. When the king hears about Ysave’s trysts in the tower, he sends for ‘his son Caradoc’ (C1 [SR A] 2577), yet when chastising him Ysave reminds Caradoc that he has betrayed his father Eliavrés (2642–4). Caradoc himself expresses regret for the ‘mesfet / Qu’il a a pere et mere fet’ [wrong he had done to his father and mother] (2665–6), and yet when the king dies, the narrator tells us that: Li reaumes, si com il dut, A son fil Caradoc revint  (C1 [SR A] 2848–9) [The kingdom, as was right, came to his son Caradoc]

By refusing to describe Eliavrés as the hero’s father, by stressing the rightness of the succession, the narrator invites the audience to participate in legitimising Caradoc’s rejection of blood ties, as well as in the elision of his shameful birth. That this could be troubling on both a thematic and a metaphorical level is demonstrated by the lengths to which the Long Redaction goes to clarify the situation of Caradoc’s paternity. Following the young man’s return to Nantes to tell the king that he does not consider him to be his father (C1 [LR] 7467), the narrator, unlike his counterpart of the Short Redaction, finds himself unable to ignore the significance of blood ties: Eliavrés is described as ‘li anchanteres / Qui a Caradoc estoit peres’ [the enchanter who was Caradoc’s father] (9621–2) when the narrator picks up the tale again after the tournament scene, and in almost identical terms as ‘[l]’anchanteor qui ert ses pere’ [the enchanter who was his father] (9771) during the description of his capture in flagrante delicto. Caradoc’s actions are thus stigmatised explicitly

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

157

as a betrayal of his own father. For further emphasis, the Long Redaction concludes the scene of Eliavrés’s punishment by mentioning Caradoc’s intercession with the king to save his life, ‘[c]ar tot de voir iert il ses pere’ [for after all he was his father] (9793), a detail not present in the Short Redaction. More emphasis on the bond between blood father and son is produced during the lovers’ discussion of how they might avenge themselves. In the Short Redaction, the enchanter is quick to respond to Ysave’s urging for reprisals: ‘Dame,’ fet il, ‘s’il ne vos poise, Une vangence molt cortoise Prandrai de lui isnelemant. Je vos di bien veraiemant Jel ferai vivre et non valoir.’  (C1 [SR A] 2607–11) [‘Lady,’ said he, ‘if it does not displease you, I will take sweet revenge on him immediately. I tell you truly, I will make his life worthless.’]

Eliavrés’s lack of hesitation, conveyed by the adverb ‘isnelemant’ [immediately], is in marked contrast to his attitude in the Long Redaction version of the scene. Here, his first reaction is revolt at the idea of harming his own flesh and blood: Dame, trop seroit grant pechiez Et trop grant cruauté feroie, Se je mon anfant ocioie.  (C1 [LR] 9814–16) [Lady, it would be a grave sin, and I would commit a great cruelty, if I killed my child.]

He relents in the face of Ysave’s insistence, but in terms which place the burden of intention on the mother’s shoulders: ‘Certes,’ fait il, ‘ses peres sui; Si ne troveroie an mon cuer Que je l’oceïsse a nul fuer, Mais que sol por vostre voloir Le ferai vivre et non valoir, Se vos volez ja antremetre.’  (C1 [LR] 9834–9, my italics) [‘In truth,’ said he, ‘I am his father, and could not find it in my heart to kill him on any account; only because you desire it, I will make his life worthless, if you wish to undertake this thing.’]

The weaving of textual supplement around the semantic hub of ‘le ferai vivre et non valoir’ demonstrates the distance between the two accounts. For

158 THE

CONTE DU GRAAL CYCLE

the Long Redaction, the bonds of biological filiation entail a set of ethical responsibilities which the text explores at much greater length than in the Short Redaction. The notion that Caradoc’s actions constitute a betrayal of his parents is discussed by other characters, notably when his friend Cador comes before Ysave to petition her for a cure for the serpent’s poison. His rhetoric recognises Ysave’s right to punish her son for his action, but also insists on the need for indulgence with one’s children: La mere doit bien maistroier Son anfant por lui chastoier […] Et après le chastoiement Le doit atrere doucement.  (C1 [LR] 11119–20; 11125–6) [The mother must punish her child in order to instil discipline… and after the punishment she should treat him tenderly.]

Following Caradoc’s recovery, the Long Redaction inserts a scene of cathartic reconciliation between mother and son, not found in the shorter version. Caradoc’s first act is to free his mother from the tower where he had imprisoned her, begging and receiving forgiveness for what the text now calls his ‘mesprison’ (11649): Mais a ce point l’a deserree Son fil, qui vers lui s’umelie Et humblement merci li crie Dou grant mal que il avoit fait. Et l’an pardone le mesfait  (C1 [LR] 11652–6) [But then her son freed her, prostrating himself before her and humbly begging her indulgence for the great harm he had done to her. And she forgave him the misdeed]

Finally, the death of King Caradoc and the succession of his namesake is handled quite differently in the two Redactions. As we saw above, the narrator of the Short Redaction recognises both a sense of entitlement (‘si com il dut’) and a relation of filiation (‘son fil Caradoc’) in this event (C1 [SR A] 2848–9); in the Long Redaction, these lines become: Et tant que Caradoc de Vegne, Ses sires, morut, et son resne Li a laissié com a son oir.  (C1 [LR] 11713–15) [And one day his lord Caradoc of Vannes died, and left him his kingdom as his heir.]

A feudal relation has replaced the genealogical bond of the earlier text, and

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

159

all notion of entitlement is omitted. Instead, the bequeathing of the kingdom is presented as a gift, and the formulation ‘com a son oir’ brings out explicitly the illegitimacy of the succession: Caradoc will reign as if he were the king’s heir.53 If the First Continuation underlines the possibility of choosing and defining one’s own filiation, the Long Redaction suggests that the resultant effect cannot entirely replace biological destiny as given by our progenitors – legitimate inheritance is not so lightly dismissed. One might suggest that, on the Long Redaction’s reading, the First Continuation is like a continuation, whereas the Second – by bringing back Perceval and making his Grail Castle quest the narrative centre of the cycle – is the real deal. This, of course, is precisely what the Second Continuation attempts to suggest, deliberately returning the hero to locations already visited in Chrétien’s text. Following the traces of kinship and recognition through the corpus brings us back to ground already trodden, as we find ourselves making good on our promise to return to the encounter between Perceval and his sister in the Second Continuation. In light of the preceding discussion, we may be on the lookout for signs of textual anxiety around the recognition scene. As happened to both Perceval and Gauvain in Chrétien’s text, our hero finds himself confronted by a relative whose existence was hitherto unknown to him; indeed, Wauchier is keen to underline this, having Perceval reflect before the encounter that he is: pres dou manoir ma mere, Mes je n’i ai seror ne frere, Mien escïent, ne autre ami.  (C2 23557–9) [near my mother’s home, but I have neither sister nor brother there, as far as I know, nor any friend.]

As we saw above, the tone of the passage is one of reconciliation and reassurance; the hero’s guilt, like his family’s anxiety over his fate, is dissolved by a narrative of mutual acceptance. In an important departure from the pattern established by Chrétien, it is Perceval himself who seems to intuit the identity of the damsel who greets him upon his arrival at the manoir: Percevaux son salu li rant, Qui bien savoit a escïent Qu’elle estoit [sa] germaine suer  (C2 23605–7)

53 The ambivalence of the conjunction ‘com a’ in Old French allows this phrase to carry simultaneously the senses ‘as his heir’ and ‘as if he were his heir’. The uneasy slippage between these two senses perfectly captures the Long Redaction’s discomfort before this narrative of doubled paternity. Rockwell (2000: 440–4) notes a similar set of ambiguities around the use of ‘si com’ in the Chevalier as deus espees, a text we will examine in Chapter 4.

160 THE

CONTE DU GRAAL CYCLE

[Perceval returned her greeting, for he well knew that she was his sister]

The last time he had experienced such an intuitive moment was, of course, in divining his name during the conversation with his cousin after leaving the Grail Castle. However, this self-identification was immediately contested by his cousin, and the moment of self-discovery became a revelation of loss and failure. Here, by contrast, Perceval remains in control of the situation. And unlike Gauvain before the Pucelle de Lis, it is enough for Perceval to speak his name for his sister to recognise him. Subsequently, when the servants arrive to lay the table, verbal identification is again sufficient to allay their concerns about the familiarity their lady is displaying towards this stranger: Molt se mervoille[nt] quant il voient Que lor damoiselle besoit Li homs qui estranges estoit; Antr’aus lou tenoient a mal. Qant el leur dit de Perceval Que c’estoit ses freres germains, Vers Dieu an tandirent lor mains Tuit cil qui sont an la meson. Einsint grant joie ne vit om Com tuit ansamble demenerent.  (C2 23706–15) [They marvelled greatly when they saw their mistress embracing the stranger; privately they disapproved. When she told them that Perceval was her brother, all those in the household raised their hands to the heavens. Never before had such great joy been seen.]

There is no need here for narrative saturation with signs, proofs, tests: Perceval is taken at his word. In this display of mutual trust between the characters, the text appears to be inviting the audience to trust it, and its retroactive recasting of Chrétien’s material as that to which it is being faithful. The motif of the ‘unknown kin’, which in Chrétien’s text appeared as the symptom of textual doubt about the hero’s capacity to create his own literary destiny, becomes in the Second Continuation a gesture of fidelity towards its parent text. Perceval’s new ability to take control of the moment of revelation invites the audience to recognise the new confidence displayed by the future ‘best knight in the world’ as the realisation of a narrative programme always present in the earlier, originary moments of the cycle. R. Howard Bloch’s stimulating analysis of the Conte du Graal converges with mine in diagnosing a ‘simultaneous problematization of paternity and of narration’ in the text, which he sees as typical of romance in general.54 54

Bloch (1983: 206).

AUTHORSHIP, KINSHIP AND THE ETHICS OF CONTINUATION

161

The issues of genealogy and narration in turn emerge as closely connected with questions of memory and doubt, both within the diegesis and without it, in the relation between author, text and audience. We have seen that this nexus of issues is of especial relevance to the Conte du Graal cycle, with its changes of narrative direction and aesthetic, and its multiple authors, for whom the status of what they are writing in relation to the whole is a recurrent concern. The moments in which the different writers invoke their role in the cycle’s creation bear witness to a desire for their contribution to be recognised as legitimately continuing the given textual past. By defining a relation of authorial succession between himself and Chrétien, Gerbert thus creates himself as an author at the same time as he confirms his predecessor’s status as originator, ‘semeur’ of the cycle. Retroaction applies here insofar as the death of Chrétien is required to enter into narrative memory in order for Gerbert to be consecrated as his heir. Gerbert’s innovations are thereby transformed into acts of narrative loyalty, with the unfinished business of Chrétien’s interrupted composition portrayed as the lack which justifies, even demands, the existence of an authorial descendant. Wauchier de Denain similarly positions his Continuation retroactively as a project in fidelity, by masking the break in authorship behind the renewal of narrative interest in the heroic subject, Perceval. The Continuation opens in most of the manuscripts with a confrontation between Perceval and a huntsman, who reminds the knight of his Grail Castle failure in similar terms to those used by his interlocutors in Chrétien’s section of the text.55 He addresses Perceval as ‘Cheitis!’ [Unfortunate!], reprising the surnom that his cousin had coined for him, while the general tone of his harangue is strongly reminiscent of that delivered by the Hideous Damsel. If Wauchier displaces attention from authorial succession onto narrative continuity, Manessier’s epilogue explores the family tree of the cycle’s patrons as an alternative means of promoting the unity of the work, and justifying his own intervention into it as sanctioned by the living heir of the text’s originator: Philippe de Flandres. Manessier’s act of completion, announcing that he has completed his book (CM 42652), functions further as an act of appropriation of the cycle. Gerbert, too, tells us that ‘or en a faite sa laisse / Gerbers’ (CG 7000–1, my italics). These later authors thus appear comfortable staking a claim to partial authorship of the cycle, displaying an overt authorial presence not readily perceptible in the earlier Continuations (especially the First Continuation, whose author(s) remain(s) unknown). Moreover, 55 The exceptions are MSS EPT. It is worth remembering Corley (1987), which argues strongly that the traditional break between First and Second Continuations is erroneous; in this case, this initial return to the Perceval thread would be the work of one of the authors of the First Continuation, rather than of Wauchier. Even if this is the case, Wauchier’s response to this ‘Perceval’ stub has transformed it retroactively into the point at which the cyclic dynamic begins to take hold.

162 THE

CONTE DU GRAAL CYCLE

where the authors of the earlier Continuations allow their narrative projects to be read implicitly, through their aesthetic and diegetic decisions, Manessier and Gerbert both allude explicitly to the desire for completion. Manessier claims to have finished the text, a claim which the manuscript transmission of the cycle was to endorse; Gerbert, meanwhile, at least gestures towards his desire to accompany Perceval to journey’s end, asking God to ensure that he reaches his goal. Others had already left their literary mark on the text, by turning an interrupted romance first into a wider collection of Arthurian adventures, and then into a cycle around Perceval. It was perhaps natural that these two later authors should have seen the opportunity for innovation in offering to give the cycle the thing it had so carefully avoided: narrative closure.

4

Rereading the Evolution of Arthurian Verse Romance The narrator of the thirteenth-century Arthurian verse romance Hunbaut presents himself as having been engaged in friendly rivalry with Chrétien de Troyes: Ne dira nus hon que je robe Les bons dis Crestïen de Troies Qui jeta anbesas et troies Por le maistr[i]e avoir deu jeu, Et juames por ce maint jeu.  (Hunbaut, vv. 186–90) [No man will say that I am stealing the good words of Chrétien de Troyes, who threw twos and threes to gain mastery of the game; we played many games in this way.]

By juxtaposing a reference to Chrétien’s textual production with the description of a gambling contest, this passage invites the audience to consider the composition of Arthurian verse narrative as a game played out between its different authors. The use of the concept of ‘maistrie’ to characterise this relationship speaks of an ambivalence between companionship and challenge, fidelity and domination, which we have seen at work throughout the Conte du Graal Continuations. This conflict is doubtless less acute in Hunbaut and other thirteenth-century romances, since these are not attempting to pursue an interrupted, inherited narrative, but are free to build their plots from scratch. Thus the Hunbaut narrator is able to state more explicitly than those of the Continuations both his admiration for the ‘bon dis’ of his predecessor, and his determination to avoid following too closely in his footsteps. Indeed, he goes so far as to boast of the quality of writing that such freedom has allowed him to produce: Ja mais ne vos erent dis vers De nule rime qui cels sanblent. Or entendés con il asanblent Et con il sont a dire fort!  (Hunbaut, vv. 34–7) [You will never hear verses which rhyme like these. Now hear how they fit together, and how powerful they are when spoken aloud!]

164 THE

CONTE DU GRAAL CYCLE

Despite the vehemence of this claim to innovation, scholars working on thirteenth-century Arthurian verse romance have long been alive to the later romances’ debt to the work of Chrétien. Indeed, for a long time the critical orthodoxy was to consider these romances as pale imitations of Chrétien’s works, and their authors as inferior epigones.1 By far the most influential recent work on the subject is Beate SchmolkeHasselmann’s Der arturische Versroman von Chrestiens bis Froissart, published in German in 1980 and translated into English in 1998 as The Evolution of Arthurian Romance: The Verse Tradition from Chrétien to Froissart. Schmolke-Hasselmann issues an important corrective to the traditional view of post-Chrétien verse romance, showing the romances to be much more than servile imitations. Instead, she proposes that each text should be read as an act of creative reception, and that attention should be paid to the ways in which these romances dialogue with, and transform, the material of their predecessor. The one point on which she agrees with the scholarly tradition is in seeing the work of Chrétien de Troyes as the primary intertext for these romances. As she puts it in the conclusion to Part 1 of her study: The creator of the first Arthurian romances founded a genre which subsequently made a deliberate policy of continually referring back to his works. This study has attempted to convey a picture of the many possible ways of coming to terms with Chrétien’s works, with regard to both poetic techniques and content.2

However, as the Introduction to this study has argued, the manuscript evidence available to us today suggests that the critical emphasis on Chrétien’s romances as a corpus needs to be nuanced by an appreciation of the significant popularity of the Conte du Graal cycle, and that Chrétien’s reputation for thirteenth-century audiences may have been partly a result of the cycle’s success. If this is correct, then Schmolke-Hasselmann’s reading of the corpus of thirteenth-century Arthurian verse romance through the lens of Chrétien’s œuvres complètes needs to be supplemented with an account of how the Conte du Graal cycle functions as a part of the horizon of expectation for

1 Alexandre Micha, for instance, in discussing the verse romance compilation MS Chantilly, Musée Condé 472, describes the non-Chrétien texts as ‘des récits de caractère conventionnel… les productions des écrivains disciples de notre maître romancier’ [conventional narratives… the products of authors who were disciples of our master romancer] (Micha 1939; rpt. 1966: 271). This view of later verse romances as derivative is in line with Gaston Paris’s judgment of fifty years previously: ‘les ouvrages de Chrétien de Troyes ont été, pour la plupart des auteurs qui sont venus après lui, les modèles qu’ils ont suivis jusque dans les plus petits détails’ [the works of Chrétien de Troyes were, for most of the authors who came after him, the models which they followed down to the smallest detail] (Paris 1888: 38). 2 Schmolke-Hasselmann [Middleton] (1998: 215).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

165

the writers and audiences of these texts.3 Indeed, the Perceval Continuations constitute an avowed blind spot in her study, described in her Introduction as ‘a specialized off-shoot from the main body of Arthurian verse’.4 Yet, in choosing to leave them to one side, Schmolke-Hasselmann forces herself on occasion into the position of misrepresenting the textual record. For instance, she claims that Chrétien’s ‘unique literary supremacy’ cast such a spell over his contemporaries and successors ‘that for a while after his death no other poet had the temerity to embark on anything comparable to the first Arthurian romances’.5 This diagnosis of anxiety of influence stifling the shortterm evolution of Arthurian romance fits neatly into the critical orthodoxy of Chrétien’s pre-eminence, but there is little evidence available of his works having enjoyed massive diffusion. In any case, the delay between the writing of the Conte du Graal and its first two Continuations may have been as little as five years, a fact grudgingly acknowledged by Schmolke-Hasselmann almost as an after-thought: ‘At the same time [the period of initial reaction to Chrétien] the First and Second Continuations were produced in an attempt to make the transition to longer forms.’6 Most of the romances discussed in this chapter are dated to the first half of the thirteenth century, a period in which (to judge from the scant surviving manuscript evidence) a cyclical Conte du Graal corpus was circulating which contained at least CdG-C1-C2.7 While most of the surviving Conte du Graal cycle manuscripts date from the second half of the thirteenth century, early-thirteenth-century witnesses such as manuscript L (containing the cyclic sequence Bliocadran-Cd-G-C1-C2), coupled with the clear points of contact with the corpus in the Arthurian verse tradition to be discussed in this chapter, make a case for significant circulation of cyclical Conte du Graal material among readers and writers of French Arthurian narrative from much earlier in the century. The phenomenon of the Conte du Graal Continuations clearly needs to be taken into account in a description of the developing ‘response to Chrétien’, to cite the title of the first part of Schmolke-Hasselmann’s book; as precocious responses to the unfinished Conte du Graal, and as a substantial and apparently successful body of text, they cannot have failed to impact on the evolution of the genre. I would like to suggest, therefore, an alternative hypothesis: that Arthurian verse romances began to appear in substantial numbers, and that Chrétien de Troyes’s romances became models for later authors, partly because of the success of the Conte du Graal cycle. Thus I 3 To my knowledge, the only study to entertain this idea is Atanassov (2000), which considers the First Continuation alongside several of the thirteenth-century romances discussed in this chapter. 4 Schmolke-Hasselmann [Middleton] (1998: 6). 5 Schmolke-Hasselmann [Middleton] (1998: 18). 6 Schmolke-Hasselmann [Middleton] (1998: 18). 7 See Appendix 5 for the dating of the romances discussed in this chapter and Appendix 6 for details of the manuscripts in which they are preserved.

166 THE

CONTE DU GRAAL CYCLE

am not attempting to deny that the authors of these texts knew and responded to Chrétien’s romances, as important work by Schmolke-Hasselmann and others has demonstrated, and there are of course other intertextual interactions at work in the thirteenth-century romances, notably with the Arthurian prose tradition.8 Rather, I contend here that the use made of Chrétien’s other works by these romances is conditioned by their reception of the Conte du Graal cycle, just as the author-name ‘Chrétien’ and the fictional world which it evoked are far more closely associated with this particular corpus than hitherto appreciated. In this chapter, I will demonstrate how a reading of thirteenth-century Arthurian verse romance in relation to the Conte du Graal cycle brings to light the significant role played by this corpus in setting the agenda for the aesthetics of the genre. I will begin by considering why Gauvain is the hero of so many of these texts, and how the repeated staging of a confrontation between reputation and reality relates to the treatment of the character in the Conte du Graal. The second section re-opens the discussion of recognition scenes conducted in the previous chapter. There, we saw that these scenes contain reflections about the narrative contract between audience and text, and that the authors of the Continuations use them to discuss their differing conceptions of the process of continuing narrative. Similarly, in the thirteenth-century romances, recognition scenes are fertile textual ground for analysing the ways in which these texts might position themselves in relation to literary antecedents. As Norris Lacy has argued, the phenomenon of motif transfer (the reuse and reconfiguration of motifs across different texts so familiar to readers of Arthurian verse romance) is a crucial element of the aesthetics of the genre.9 A text, in transferring a motif from another romance, can demonstrate by its deployment of the motif a particular attitude towards the generic tradition within which it exists. Attention will therefore be paid in the third part to the transfer of motifs from the Conte du Graal cycle into other romances, in order to elucidate how these later texts understand their relationship to the cycle, which emerges as an important source of material, but from which the verse romances are careful to retain a certain distance, carving out their own distinctive narrative logic. The final part of the chapter will bring us back to the issues discussed in the first chapter of this study: the use of interlace and centripetal narrative structures within each text to provide narrative coherence. It is in these centripetal structures, I argue, that the influence of the Second Continuation can be felt most strongly on the later romances. The development of interlace within a verse romance cycle narrated in the first person is clearly distinct from the model of interlace 8 For a robust statement of Chrétien’s influence on later romance, see the contributions to Lacy, Busby and Kelly eds. (1987/88), and Baumgartner (2005). More generally, see the entries for Section P (‘Influences’) in Kelly (2002). 9 Lacy (1996).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

167

found in prose romance, and its widespread use in thirteenth-century Arthurian verse texts testifies to the important role that the Conte du Graal cycle played in the evolution of the genre.

Gauvain Continuation: What’s in a Name? Le Chevalier à l’épée, a short early-thirteenth-century Gauvain narrative, contains in its prologue a well-known justification for choosing Arthur’s nephew as its hero: Je cont de monsaignor Gauvain, Qui tant par ert bien ensaigniez Et qui fu des armes prisiez Que nus reconter ne savroit. […] L’en ne doit Crestïen de Troies, Ce m’est vis, par raison blamer, Qui sot dou roi Artu conter, De sa cort et de sa mesniee Qui tant fu loee et prisiee, Et qui les fez des autres conte Et onques de lui ne tint conte. Trop ert preudon a oblïer. Por ce me plet a reconter Une aventure tot premier Qui avint au bon chevalier.  (Chevalier à l’épée, vv. 8–11; 18–28) [I speak of Sir Gauvain, who was so well mannered and so esteemed for feats of arms that no one would know how to tell it all… One should not, it seems to me, blame Chrétien de Troyes, who knew how to tell about King Arthur, his court and his retinue which was so highly praised, and who told of the deeds of others but never took him into account. He was too worthy a man to be forgotten. For this reason, it pleases me to recount, to begin, an adventure that befell the good knight.]

Despite the attention it has received, this remains a complex passage to interpret. Debate has largely focused on the desirability of amending ‘ne’ in v. 18 to ‘en’.10 For D. D. R. Owen, the manuscript reading demands that the narrating subject be identified with the third-person referent ‘Crestien de Troies’; 10 The text survives in only one manuscript (Bern, Burgerbilbliothek 354). Several scholars have followed E. C. Armstrong’s argument that vv. 18–19 make better sense as ‘L’en en doit Crestïen… / …blamer’. See Chevalier à l’épée (1900); Paris (1900); Schmolke-Hasselmann [Middleton] (1998: 105–6). For D. D. R. Owen’s reading of ‘Crestïen’ as a third-person authorship reference, to be identified with the narratorial

168 THE

CONTE DU GRAAL CYCLE

the passage is to be understood as Chrétien defending himself from the accusation of forgetting Gauvain (on the basis of which he suggests that Chrétien may have written this short text at the same time as the Conte du Graal). For the partisans of amendment, the sense would be something like: ‘Chrétien failed to tell of Gauvain’s adventures, but since he is such a worthy knight, I will remedy this omission.’ Yet few scholars have commented on the oddness of this claim.11 For Chrétien’s contribution to the Conte du Graal, as we have seen, devotes more time to Gauvain’s adventures than to following Perceval, and the First Continuation stands as testimony to the potential for narrative accumulation around the character bequeathed by Chrétien. Is it possible that the author of the Chevalier à l’épée was unaware of the Gauvain part of Chrétien’s text, or of his final romance altogether, while claiming familiarity with the author’s work? Nothing in the manuscript record authorises such a conclusion. The narrative itself, in which Gauvain’s attempts to have his way with a beautiful damsel are complicated by a magic sword suspended over the bed in which they lie, demonstrates on the contrary a direct engagement with his adventure on the Marvellous Bed in the Conte du Graal. Read in this way, what the Chevalier à l’épée prologue does suggest is that the Conte du Graal is being treated as a narrative about Perceval. One might see its author as initiating the scholarly tradition that has bestowed the alternative title of Perceval on Chrétien’s final romance, condemning Gauvain’s adventures to the status of subplot. Yet, when one takes into account the evidence of reception and cyclic formation I have described in the preceding chapters, it becomes clear that something more complex is happening. It is, after all, the cyclification of the Conte du Graal by its Continuations (and first and foremost by the Second Continuation) which ensures the centrality of Perceval’s narrative to the tale; the ‘roman de Perceval’ that contemporary audiences would have engaged with was thus more likely to be some configuration of the cyclic work than Chrétien’s text alone. I would suggest, in fact, that it is only in light of the cyclification of the Conte du Graal material into a Conte du Graal cycle that we can make sense of the Chevalier à l’épée narrator’s puzzling claim. It is in the context of the Second Continuation’s recalibration of the narrative that Gauvain finds himself playing second fiddle to Perceval, who emerges as the future ‘best knight in the world’. Moreover, if this is correct, it does indeed seem that the name ‘Chrétien de Troyes’ becomes associated very quickly with the Conte du Graal cycle as a whole,

‘je’, see Chevalier à l’épée (1972, Introduction, pp. 7–9 and Notes, pp. 91–2) and Owen (1971). 11 One exception is Schmolke-Hasselmann [Middleton] (1998: 107), who notes: ‘For the modern reader there is a certain element of contradiction in the more critical statements made by Chrétien’s successors, since we know (or think we know) that Chrétien did in fact give detailed attention to the king’s nephew; after all, several thousand lines of the Conte du Graal are devoted to him.’

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

169

which must have been circulating among aficionados of Arthurian narrative by the first decades of the thirteenth century. In choosing to write about Gauvain, and by accusing ‘Chrétien’ of not taking him seriously as a character, the Chevalier à l’épée proposes to transform the supporting actor of the Conte du Graal cycle into the star of his own, fully fledged romance. Here, I suggest, lies one explanation for the striking interest of many thirteenth-century verse romances in Gauvain: whereas the cycle ultimately turns Perceval into the character who (according to his tombstone at CM 42636–7) ‘brought to an end the adventures of the Holy Grail’, Gauvain’s own adventures appear to follow no such totalising logic. The First Continuator flirts with the idea of making him the Grail knight, but the subsequent texts of the cycle show no interest in coupling Gauvain with the concept of election. The character of Gauvain therefore allows thirteenthcentury Arthurian verse romances to dialogue with, and evoke, the important intertextual precedent of the Conte du Graal cycle, but without becoming part of it. As we will see below, several romances allude to the events of the cycle, but always at a distance, preferring to send their heroes off in new, unexplored (and therefore undetermined) narrative directions. The question of how to position oneself in a relationship of filiation with a textual tradition, while simultaneously staking a claim to originality, is one with which we were much concerned in the previous chapter, and it is not surprising to find that the themes and motifs which emerged as especially rich for producing metaphors of authorial agency and responsibility are equally central to the thirteenth-century verse romance corpus. Chief among these, as noted by several critics, is the subject of Gauvain’s reputation. As Keith Busby observes, ‘[a] good deal of space in the romances is devoted to the exploration of the relationship between Gauvain’s reputation (especially as a valiant knight and as a ladies’ man) and the reality of his performance, between expectation and the moment of truth’.12 For Stoyan Atanassov, these romances are obsessed with the potential for questioning and decoupling the semantic pair of nom and renom, which he theorises in terms of signified (the private subject designated by the name Gauvain) and signifier (the public reputation that has attached itself to that name).13 Time and time again, Gauvain is confronted by strangers who know him, or rather believe they do: knights who desire to match themselves against him in order to prove their worth and damsels who have fallen in love with what they have heard about his qualities. The model for this second type of encounter is the Pucelle de Lis episode of the First Continuation, and the romances’ engagement with this scene will be considered in the following section. These two types of encounter are in fact closely related. The damsels fall in love with Gauvain because of his reputed valour and courtliness; conversely, 12 13

Busby (1980: 305). Atanassov (2000).

170 THE

CONTE DU GRAAL CYCLE

knights who seek him out to fight him often do so because they have been scorned by a damsel who prefers Gauvain. In both cases, there is a public investment in the monument of his reputation, even when this is manifested as a desire to put it to the test. Indeed, the public figure ‘Gauvain’ often functions to support the values of the Arthurian community, as is brought out in a revealing scene from Meraugis de Portlesguez. Having watched him surrender to his opponent Meraugis, Arthur’s men feel the defeat as a collective shame and declare that Gauvain is no longer worthy of his name: Et dïent: ‘Gauvains a perdu Son non! Onques nul jor du monde Ne fu mes la table reonde Deshonoree fors par lui […] Quant il est vis recreanz, donques Somes nos trestuit recreant’  (Meraugis, vv. 5490–3; 5496–7) [And they said: ‘Gauvain has lost his name! Never was the Round Table dishonoured except by him… if he lives as a coward, then we are all cowards’]

The strength of the community’s attachment often has dangerous consequences for the man himself. Here, in their anger, Arthur’s knights threaten to cut off Gauvain’s head and send it to the king (vv. 5500–1). In Le Chevalier as deus espees, meanwhile, Arthur’s nephew is attacked and badly wounded by an assailant he has never met. The villain is a knight called Brien, whose love for the Dame des Iles is thwarted by her desire to marry Gauvain. Her desire is directed towards the public persona rather than the private subject: although she has never met him, the good she has heard spoken of him makes her certain that only he can be a suitable match for her. Again, ardent attachment to the reputation of Gauvain boils over into a desire to kill the person: the queen agrees to marry Brien if he can prove himself more worthy than Arthur’s nephew, by cutting off his head or defeating him in battle. When the two knights meet, Gauvain is unarmed but Brien is determined to fight immediately. Gauvain resolves not to flee, for fear of being called a coward: Bien dist ke ja jor k’il soit vis Et soie soit la poesté Ne sera a cort conté Nus contes de sa couardie.  (Chevalier as deus espees, vv. 3010–13) [He said that as long as he was alive and able to prevent it, no tale would ever be told at court of his cowardice.]

He is thus trapped into putting his life in danger by the need to maintain his

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

171

reputation.14 Believing he has killed Gauvain in the uneven contest, Brien returns to his lady to announce that the task is accomplished. Although the deception is unintentional, Atanassov is right to stress that what Brien wins here is the possibility of recounting the appropriate narrative that will allow him his lady’s hand in marriage: Tout se passe comme s’il s’agissait moins de mettre à mort l’adversaire que de fabriquer les signes d’un meurtre et de mettre en récit un fait qui s’était produit sans témoins. Ce qui compte pour Brien, c’est sa propre version de sa victoire.15 [It appears to be less a matter of killing the adversary than of fashioning the signs of a killing and of turning a deed which took place without witnesses into a narrative. What is important for Brien is his own account of his victory.]

The proof of Gauvain’s death, and of Brien’s own worth, requires others to bear witness to, and authorise, his narrative, and so it is no surprise that the preliminaries to his wedding ceremony resemble a legal case. The archbishop rises before the assembled knights and ladies and recounts once more the agreement reached, and now apparently held, between Brien and the Dame des Iles. His final words – ‘Et nous issi le vous disons, / K’ele vielt que vous le saciés’ [And we tell it to you here because she wishes you to know it] (vv. 5522–3) – explicitly interpellate this diegetic audience as witnesses. Brien then gets to his feet to testify: ‘Gauvains est ocis voirement’ [Gauvain is truly dead] (v. 5526). Gauvain’s subsequent intervention acknowledges the need for public recognition of his existence (he is still alive) and of his reputation (he has not been defeated by Brien); he enforces his countermanding narrative in a judicial duel, and Brien is forced to concede before all present that he has lost his ‘onnor’ (v. 5845). The theme of Gauvain’s apparent death is also a major plot thread in L’Atre périlleux, and again the proof that he survives is simultaneously a demonstration of his chivalric prowess, as he defeats an opponent who claims to have killed him. Here, a murdered knight is wrongly identified as Gauvain; the body, dismembered and taken away by the murderers, is unavailable for inspection, and Arthur’s nephew is therefore forced to identify himself as ‘cil sans non’ [the nameless one] until he can reconquer his identity. Critics have puzzled over the fact that the substitution of ‘cil sans non’ for the name

14 A similar prioritising of reputation over wellbeing is evident in Le Chevalier à l’épée, where Gauvain risks being killed by a magical sword hanging over the bed, rather than let it be said that he spent the night next to a naked damsel and failed to take advantage: ‘Miaus vient il a anor morir / Qu’a honte vivre longuement’ [Better to die honourably than to live a long life in shame] (Chevalier à l’épée, vv. 588–9). 15 Atanassov (2000: 91).

172 THE

CONTE DU GRAAL CYCLE

‘Gauvain’ does not take place as soon as the news of his death is introduced into the narrative, but only once he has completed his business at the Perilous Cemetery and returned to the matter of pursuing his supposed murderers.16 It seems to me that this feature of the text can be explained in terms of the determining role played by public opinion. When Gauvain is first confronted with the narrative of his death, he is engaged in pursuit of a knight who has abducted a damsel from Arthur’s court. He is forced to choose which of these two competing adventures to follow (Atre, vv. 641–2), and decides to leave the matter of overturning the false rumour until he has completed the quest he has already begun. In the course of this adventure, he meets and reveals his name to a number of people; none of them express surprise that he is alive. Apparently, the rumour has not yet spread. But as Gauvain travels further in both distance and time from the initial incident, the probability of its crossing his path again increases. When it eventually does, it is embedded in a further narrative in which Gauvain’s reputation plays a central role. He meets a knight, whose name we learn is Espinogre, who tells him how, after three years of courting a damsel, he has finally obtained her favours against a promise that Gauvain (whom neither has met) will guarantee his future fidelity. She has heard of Gauvain’s virtues and is confident that they will suffice to keep Espinogre in line; he, however, has heard that Gauvain is dead and so intends to break his promise that very night. When he hears this, Gauvain is reminded for the first time of his earlier encounter (vv. 3245–7). It is from here onwards that he will identify himself as ‘nameless’, beginning with his reply to the defeated Espinogre’s request to reveal his name: ‘Je ne vous puis le mien non dire,’ Fait Gavains, ‘que je l’ai perdu, Si ne sai qui le m’a tolu.’  (Atre, vv. 3450–2) [‘I cannot tell you my name,’ said Gauvain, ‘for I have lost it, and do not know who has taken it from me.’]

It seems clear from this account that it is when the initial misapprehension has become generalised misinformation that Gauvain must bow to public opinion and give up temporarily the right to his name.17 Identity is portrayed not as a matter of psychological self-apprehension, but rather of recognition as subject by others.

16 See for instance Lacy (2006: 115): ‘What is most striking about subsequent events, however, is that, although his life and identity, and his name as well, have been ostensibly taken from him, he does not immediately abandon that name.’ 17 This narrative logic is complicated in BNF, fr. 1433 by an interpolated episode preceding the meeting with Espinogre, in which Gauvain tells another defeated opponent, Brun sans pitié, that he is ‘le chevalier sans non’ [the knight without a name] (v. 549).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

173

Overturning the weight of public opinion is made more difficult by the campaign of disinformation waged by the murderers, whose wooing of the sisters they desire depends on convincing the community that they have killed the king’s nephew. Since the damsels themselves refuse to recognise the dismembered body as being that of Gauvain, the knights turn to the wider public for validation: Car les damoiseles disoient Ke autre fois veü l’avoient, Et ce n’ert mie illuec son cors. Et li chevalier dissent lors Ke si ert et sel prouveroient; Car as marciés crier feroient, Par tox les castiax du païs, K’il avoient Gavain ocis, Et se nus osoit ce desdire K’il feroient crier et dire, K’il seroient prest du prouver.  (Atre, vv. 5287–97) [For the damsels were saying that they had seen him since, and that this body was not his. And the knights said that it was, and that they would prove it: they would have it proclaimed in the markets held in all the castles of the land that they had killed Gauvain, and if anyone dared to gainsay this proclamation, they would be ready to prove it in battle.]

The elder sister has fallen in love with Gauvain’s renom and intends to ask Arthur to arrange for her to marry him; the suitors therefore determine to kill him, and obtain the damsels’ agreement that they will have them should they succeed. Once again, desire for ‘Gauvain’ the public figure has led to the threat of death for Gauvain the private individual. Indeed, at times in this text, Arthurian society appears to prefer a dead Gauvain to a live one: the ‘real’ Gauvain, after all, is always at risk of failing to live up to his reputation, whereas a dead hero is both reassuringly fixed and usefully malleable in what he represents to his community.18 Emblematic of the Arthurian community’s willingness to mourn Gauvain as a lost icon is the remarkable scene in which our hero is persuaded to spend the night at the home of Tristan-qui-ne-ment, a vavasor whose ardent desire to tell his guests the sad news culminates in the revelation that he is now the proud owner of Gauvain’s right arm. He gave shelter one night to the murderers, who recounted their adventure to him; unable because of the

18 Jane Gilbert (2008: 29) sees the rhymed tradition of the Chanson de Roland as envisaging the same role for its hero as that suggested here for the myth ‘Gauvain’: ‘Not Roland but his idealized image is to be kept alive, a socially useful resource.’

174 THE

CONTE DU GRAAL CYCLE

rules of hospitality to chase them from his home, he settled for owning a little piece of the legend, which he intends to have embalmed and kept as a relic. When Gauvain and his companion Espinogre ask him to keep it safe until its true provenance can be established, his reply contains a clear rejection of their implication that this is not in fact the arm of Arthur’s nephew: Il lor respont qu’isi ert il: Ja rien de lui n’ert tenu vil, La u il soit, ne bras ne main; Car bien set que c’est de Gavain, Si doit estre moult cier tenu. (Atre, vv. 5203–7, my italics) [He replied that this would be done; no part of the knight’s body would be mistreated in his presence, neither arm nor hand, for he well knew that this was Gauvain’s, and must therefore be deeply cherished.]

The words by which Tristan begins his tale speak of a desire to pass on what he has learned, which he justifies by pointing out that it has already passed into common knowledge: ‘Segnor,’ fait il, ‘je vous veul dire La dolor et l’anui et l’ire K’avenue est novelement […] Se on peüst celer et taire Ce que je vous veul ci conter, Nus hom nel deüst reconter, Mais l’uevre est ja par tout seüe, K’el ne porroit estre teüe.’  (Atre, vv. 4977–9; 4984–8, my italics) [‘Lords,’ he said, ‘I want to tell you about the pain, the suffering and the sorrow that have recently occurred… If what I want to tell you could have been hidden and hushed up, no man would have had the right to reveal it; but the deed is already widely known, and cannot be kept quiet.’]

Tristan’s readiness to believe in Gauvain’s death echoes the attitude of the little group that first informs our hero of his demise. Alerted by the sounds of damsels in distress, Gauvain comes across three maidens who claim to have witnessed the event at first hand. The narrative recounted to him anticipates its dissemination through the land, in similar terms to those used by Tristan: Car li damages est si grant, Onques si grans ne fu veüs, Quant il sera par tout seüs. (Atre, vv. 522–4, my italics)

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

175

[For the outrage is so great that, when it is widely known, no greater sorrow will ever have been seen.]

When he objects that they must be mistaken, explaining that he has just come from court where he saw Gauvain seated at the table, a squire who has been travelling with the damsels intervenes: ‘Je fui,’ fait il, ‘vallés Gauvain Antan a un tornoiement, Et si sai tout certainement Que c’est il qui est decaupés.’ (Atre, vv. 610–13) [‘I was,’ said he, ‘Gauvain’s squire at a tournament once, and I know for certain that this dismembered body is his.’]

The lure of a dead ‘Gauvain’ is sufficiently strong to blind even those who should be able to recognise him, a blindness made shockingly literal in this case, as the squire’s reward for attempting to prevent the murder was to have his eyes gouged out. If the attachment of the characters encountered by Gauvain in L’Atre périlleux to the idea of Gauvain’s death is slightly disturbing, the most explicit statement of this tendency occurs in La Vengeance Raguidel. Here, a damsel, once spurned by Arthur’s nephew, intends to kill him and have their bodies laid side by side in a splendid tomb, in a simulacrum of the relationship which she had desired to have with him. Again, a dead Gauvain is shown to be more dependable than the live version. Even if he had married her, opines the damsel, he would soon have been off with another woman (vv. 2341–6); but if she kills him, they will stay together forever: Qant mort serïons ambedui, en cest sarcu serïons mis boce a boce, vis contre vis. Issi me feroit compagnie mors qant nel puis avoir en vie !  (Vengeance Raguidel, vv. 2296–300) [When we are both dead, we will be placed in this coffin with our mouths and faces joined. In this way he will keep me company in death, since I cannot have him alive!]

The notion that Gauvain’s public image represents some kind of ideal for the community is a feature of the character from early in the tradition. The Gauvain of Chrétien de Troyes’s earlier Arthurian romances has variously been described as a ‘yardstick’ (Busby 1980: 53) and a ‘personnageréférence’ (Ribard 1976b: 10); as the Yvain narrator puts it, he is the sun of the Arthurian world:

176 THE

CONTE DU GRAAL CYCLE

Chil qui des chevaliers fu sire Et qui seur tous fu renommés Doit bien estre soleil clamés.  (Yvain, vv. 2400–2) [He who was best of the knights, and renowned above all others, should indeed be called the sun.]

However, the Gauvain of the thirteenth-century romances has more in common with the character as he appears in the Conte du Graal cycle than with the Gauvain(s) of Chrétien’s first four romances. In the earlier texts, Gauvain stays in the background; even in Yvain and Lancelot, where his role in the narrative is more developed, his adventures find meaning in relation to those of the central hero. In the Conte du Graal, for the first time, he is sent on adventures that concern only him; as Jacques Ribard puts it, the ‘personnage-référence’ has become a ‘personnage-itinéraire’.19 And it is only once Gauvain has acquired this independence as a character, once he is endowed with a personal past, that the confrontation between public and private identities becomes possible. Indeed, a recurrent motif in Gauvain’s adventures in the second half of the Conte du Graal is the capacity for his name to get him into trouble. In this sense, a clear contrast is established with Perceval: where the new knight’s narrative involves the making of a reputation and the discovery of his identity, Gauvain’s reputation is made, and it is his revelation of his identity to others which leads him into trouble. More specifically, he is accused on multiple occasions of having committed some crime against his interlocutor. Both hero and audience are then brought to wonder how the image of Gauvain as the paragon of Arthurian virtues, pre-eminent in chivalry and courtesy, can be reconciled with the accusations of treason (Guingambresil), mistreatment (Greoreas) and murder (Guiromelant) to which he is subjected by private parties throughout the romance. The Conte du Graal, like the later verse romances, thus makes Gauvain the focus for an insistent questioning of the gap between appearance and reality. Indeed, the first episode following his departure from court sees him arrive at a tournament where, because he refuses to take part, he is mistaken first for a merchant, then a thief, and mocked by a group of ladies (CdG 5060–90). As in the texts we have been examining, Gauvain’s identity is constructed and deconstructed by the gaze of others. Despite, or perhaps because of, the intense scrutiny to which his deeds past and present are subjected, the final lines of Chrétien’s text afford us a clear view of Gauvain’s value to the community. A messenger sent to Arthur by his nephew arrives in the city of Orquenie, where the king is holding court, and finds the population in turmoil:

19

Ribard (1984: 7).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

177

  ‘nous celui avons perdu Qui por Dieu toz nos revestoit Et dont toz li biens nos venoit Par almosne et par charité.’ Einsi par tote la cité Monseignor Gavain regretoient Les povres gens qui molt l’amoient.  (CdG 9208–14) [‘We have lost he who clothed us in God’s name, and from whom possessions came to us through alms and charity.’ In this manner, throughout the city, the poor folk who loved Gauvain greatly mourned his loss.]

The extraordinary grief of the ‘povres gens’ over Gauvain’s absence reflects that of Arthur, who falls into a swoon when he notices that his nephew is not among the assembled barons at court. Even in absentia, Gauvain thus functions as a common denominator linking the king to his people, binding the community together. And it is precisely this aspect of the scene that will be foregrounded in the opening lines of the First Continuation, where Gauvain’s envoy delivers his message, and the court’s mood is immediately lifted: Corz ne fu ainz si esjoïe Quant la novele fu oïe. Or n’ont il mes de rien anvie, Quant il sevent celui an vie Qui corteisie referoit S’ele del tot perdue estoit.  (C1 [SR: A] 67–72)20 [A court was never as joyful as when the news was heard. Now they were lacking nothing, since they knew that he who would restore courtliness to the world, were it ever lost, was alive.]

The court’s dependence on Gauvain, emphasised here, is explicitly acknowledged by Keu in a speech which is transmitted essentially unaltered in all manuscripts of the text: Voir dit qui dit, se Dex me voie, Que nus ne set que prodom valt Jusqu’a cele ore qu’il defaut […] Or nos esclarcist et ajorne,

20 As Séguy (2001a: 299) suggests, the sudden swing in the court’s mood, from depression to delight, gives immediate momentum to the narrative motor of the text. The absent presence of Gauvain thus creates links, not only across the community, but also between the Conte du Graal and its first Continuation.

178 THE

CONTE DU GRAAL CYCLE

Quant cil est heitiez et an vie Qui flors est de chevalerie.  (C1 [SR: A] 174–6; 182–4) [As God is my witness, he speaks the truth who says that nobody knows the worth of a man until he is gone… now a new day has risen for us, since the flower of chivalry is alive and well.]

It is no coincidence that the author has placed these words in the mouth of the court’s most misanthropic character. Keu’s willingness to admit what Gauvain valt to the community holds all the more force coming from the man whose usual response to new arrivals at court is withering sarcasm. The insistent linking of Gauvain, reputation and identity sets the Conte du Graal apart from Chrétien’s other Arthurian romances, none of which develops the Gauvain-as-public-property theme to anything like the same extent. This theme persists and develops through the Continuations, supporting the manuscript evidence that the Conte du Graal partakes in a textual tradition which sets it apart from Erec, Cligés, Yvain and Lancelot, within the codices of the Conte du Graal cycle. Our investigation of the attention paid to Gauvain by the authors of thirteenth-century verse romance therefore suggests that they are dialoguing especially closely with this textual tradition, and that the Conte du Graal cycle constitutes a primary model for how the later romances conceive of Arthurian verse romance.

‘Ce n’est il pas, mais il le sanble’: Recognition and Intertextuality Gauvain’s encounter with the Pucelle de Lis, discussed in Chapter 3, demonstrates how the Continuations continue to probe the issue of recognition, reputation and reality in relation to Gauvain.21 This adventure is subsequently reworked in the Second Continuation, where a damsel named Tanree confesses that she has never loved a man but him, ‘Car vos estes de tel renon / Que je vos ai amé pieç’a’ [for you have so great a reputation that I have loved you for a long time] (C2 29848–9). Having given him her virginity, she soon realises that Gauvain has lost interest in her, ready to move on to the next adventure, the next conquest: Je puis tres bien de fi savoir Que, ainz que soit passé li mois, An avra il ou deus ou trois, Qui plus belles de moi seront Et autant com je l’ameront. Qu’an porra donc, se il m’oublie?  (C2 30468–73)

21

See pp. 150–2.

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

179

[I can be quite certain that, before the month is out, he will have two or three lovers who are more beautiful than I, and who love him just as much. How, then, can he fail to forget me?]

Tanree’s error was to believe that she might be able to get the ‘real’ Gauvain to stay in place, despite the reputation for amorous dalliances that had accrued to him. Were she a good ‘reader’ of the rules of Arthurian textuality, the renom with which she fell in love, and with which the audience of the cycle is by now familiar, should have warned her of the outcome of her love. The issue of mapping nom onto renom, here as in the examples discussed in the previous section, thus invites comparison between the reception of the narrative’s events by its characters and by its audience. As we noted in the previous chapter, scenes of recognition and identification have the potential to raise troubling concerns for audiences. Peregrine Rand has demonstrated how a reading in manuscript context of the romances contained in Chantilly, Musée Condé 472 can lead one at least to entertain doubts over the authority of what is being narrated. He singles out L’Atre périlleux, which features in the manuscript, as especially troubling: [D]oubting those things which must be, and usually are, taken for granted in a romance (that the characters are indeed who the narrator says they are), is like opening Pandora’s Box. Doubts have been raised concerning the knight formerly known as Gauvain, doubts which, furthermore, have been expressed even by ‘Gauvain’ himself, and nothing has been done, indeed nothing can be done to restore the absolute credibility of the narrator’s word. The possibility that is raised is that the romance that has just been read is a far from authoritative alternative version of events which are represented nowhere else, and of which, unsatisfactory though it may be, it is thus the only version available to the reader.22

As we saw in Chapter 3, the scene of recognition between Gauvain and the Pucelle de Lis in the First Continuation seeks to allay such concerns by acknowledging the artificial and adventitious nature of narrative, and inviting the audience to enjoy the story in full cognisance of this fact. The fact that the scene is reworked by several of the thirteenth-century verse romances underlines the importance of the extradiegetic conversation between text and audience to the aesthetics of the genre. In the Chevalier as deus espees, the Dame des Iles is not the only woman in love with Gauvain. On his way to interrupt Brien’s wedding, he successfully defends the Castiel du Port from Gernemant de Norhombellande, a powerful knight who is waging war in retribution for having been refused the hand of the local lord’s daughter. Even before the battle, Gauvain falls in love with the damsel, and arranges for her to be his reward, should he defeat Gernemant. At the moment of con22

Rand (1998: 145).

180 THE

CONTE DU GRAAL CYCLE

summation, however, the damsel bursts into tears: she explains that she had been saving her virginity for Arthur’s nephew, Gauvain, about whom she had heard so much praise; she is crying because she has just been reminded of the news that he has been killed by Brien. Her interlocutor is pleased to inform her that Gauvain is alive, a revelation that delights her. She asks this knight’s identity, which he proudly declares: ‘Je sui, se Diex me voie, / Icil Gauvains dont vous dissiés’ [I am, God help me, this Gauvain of whom you speak] (vv. 5032–3). All seems ready for a resolution similar to that in the First Continuation, with Gauvain passing some kind of identity test. Amusingly, however, the damsel refuses to believe this part of his revelation, objecting that he may be saying this simply in order to have sex with her, now that he knows she is waiting for Gauvain. She resolves to travel to the court of Arthur, ‘Le bon roi par tout renommé’ [the good king of universal repute] (v. 5047) to ask whether this knight she has dealt with is indeed his nephew. Intriguingly, then, the point of comparison here will be not a visual reference but the word of the king guaranteed by his renommée. On an extradiegetic level, one might suggest that the faith of the audience in Arthurian narrative is mirrored in that eventually placed by the damsel in the word of its eponymous king. In La Vengeance Raguidel, the motif is given a different twist. The damsel in question, the Pucele du Gaut Destroit, has set up a whole system of traps in order to capture and kill Gauvain; as part of this, she keeps a maidservant named Mahot whose sole function is to greet knights as they arrive and identify the one whose arrival is so fervently desired. Mahot’s ability to recognise Gauvain comes from her childhood spent in Carlion, where she got to know all of Arthur’s knights. Sure enough, as soon as she sees the king’s nephew, she knows with whom she is dealing: La pucele l’a encontré qui ert a la porte venue. Messire Gavains le salue et ele lui delivrement. Bien le conoist, certainnement set que c’est messire Gavains. (Vengeance Raguidel, vv. 1910–15) [The damsel met him at the door. Sir Gauvain greeted her, and she him. She recognised him, without a doubt – she knew that this was Sir Gauvain.]

Rather than pass on what she has seen, though, Mahot immediately warns Gauvain that he is in grave danger, and arranges to introduce him to her mistress as Keu. Mahot’s betrayal of her mistress ironically marks her as a character that the audience can trust, since it creates an axis of collusion for the following scene in which she, Gauvain and the audience are aware of the deception, while the Pucele de Gaut Destroit reveals exactly how she intends

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

181

to murder Arthur’s nephew, blithely unaware that the one she wants within her grasp is standing beside her. The unequivocal nature of Mahot’s support for Gauvain therefore serves to enhance the authority of her testimony, an impression heightened by the narrator’s use of the qualifiers ‘bien’ and ‘certainnement’ in the passage cited above. Mahot’s recognition of Gauvain puts him, and the audience, in a position of superior knowledge for the duration of the ensuing scene, while the Pucele de Gaut Destroit is shown to be a thoroughly inattentive reader: Qant Gavains l’ot, la color mue, ke bien s’en fust aperceüe cele, se l’eüst esgardé, mais ne s’en a garde doné  (Vengeance Raguidel, vv. 2303–6) [When Gauvain heard this, he changed colour; she would easily have noticed this if she had looked at him, but she did not pay attention to it]

Thus, although the Pucele’s case conforms to the motif we have been discussing, the woman who falls in love with Gauvain without knowing him, the development of this motif in La Vengeance Raguidel is quite unusual. In the First Continuation, or in Le Chevalier as deus espees, Gauvain (and the audience who traverse the Arthurian world via his itinerary) encounters a damsel whose existence was hitherto unknown to him, yet who has been waiting for him to arrive in her life. Here, he is in the presence of a damsel whose love has turned to hatred, whom he has previously met (albeit briefly) and who fails to discover until after he has left that the event for which she has waited so long has been and gone. The narratorial voice throughout this scene works to reassure the audience that we will not be left out of the loop, that the bond of trust between audience and text will be respected, and that all aspects of the scene will be explained. Thus, in explaining the existence and function of the guillotine-window with which the Pucele intends to decapitate Gauvain, the narrator twice makes direct reference to the gaps in our knowledge that he proposes to plug:    Volés savoir le leu par la ou le veoit? Une fenestriele i avoit  (Vengeance Raguidel, vv. 2118–20) [Would you like to know how it was possible to see into it (the tomb)? There was a small window] Vos ne savés, se jo nel di, por qoi la fenestre i fu faite.  (vv. 2126–7) [You will not know, if I do not tell you, for what purpose the window had been made.]

182 THE

CONTE DU GRAAL CYCLE

We are a long way from the enigmatic description of the Grail procession in Chrétien’s Conte du Graal, where the audience is allowed to apprehend the event only through the mystified gaze of Perceval. And yet the contrast between our understanding of this scene and the Pucele’s lack of insight appears less clear-cut when one considers her initial scepticism regarding the identity of the knight before her. This cannot be Keu, she objects to Mahot, since the feats he has performed prior to his arrival are beyond Arthur’s seneschal. Mahot, taking on the reassuring tone of the narrator, insists that she is telling the truth. Indeed, she invites her mistress to see the ‘truth’ with her own eyes: Dame, por qoi le dites vos? C’est Kex li senescax por voir. Venés la fors por lui veoir et si sarés bien se je ment. Molt puis avoir le cuer dolent qant vos de rien me mescreés.  (Vengeance Raguidel, vv. 1972–7) [Lady, why do you say this? Truly, it is Keu the seneschal. Come outside to see him and you will know whether I am lying. It makes me heavy-hearted indeed that you do not believe me.]

Mahot, whose sole use to the Pucele is to identify and narrate the arrivals at her castle of Arthurian knights, shows how the voice of authority can be used to impose falsehood as well as truth. She uses the same verb – ‘savoir’ – used in the extracts cited above to define the transmission of information from narrator to audience. In both cases, the audience is thereby incited to feel confident about the truth-value of the information being imparted; yet, in this case, Mahot’s narratorial ventriloquism functions to conceal the fact that she is duping her mistress. The unfortunate Pucele, then, functions in this recognition drama both to invite a feeling of dramatic superiority in the audience, and then more subtly to suggest that this confidence in our ability to understand and anticipate the tale may itself be an effect of narratorial rhetoric. If the scenes we have been studying play with the themes of recognition and communication on a diegetic (between characters) and an extradiegetic (between audience and narrator) level, their use of the ‘damsel in love with Gauvain without knowing him’ motif puts them in intertextual communication with the Conte du Graal cycle, inviting the audience to identify and contrast the two episodes. The final case I will consider here explicitly opens up the issue of intertextual recognition. In Hunbaut, Arthur sets out with a number of his knights on a quest to find Gauvain’s sister, who has been kidnapped (Gauvain himself is travelling independently). One day they arrive at a castle where they find the châtelaine surrounded by her knights and

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

183

ladies, listening to a romance: ‘D’un roumant oent .I. biaus dis, / La pucele le faissoit lire’ [They were listening to a worthy tale from a romance; the damsel was having it read] (Hunbaut, vv. 3052–3). The presence of this ‘roumant’ at the outset prepares the audience for the intertextual resonances of the scene. For the damsel keeps in her room a wooden statue carved in the likeness of Gauvain, whom she has never seen: Si ert entaillie de fust Si adroit et en tel sanblant Que ja nus hon por nul sanblant, Tant fist ne percevans ne cointe Qui tant fust [de] Gauvain acointe, Se l’imagene esgardast el vis Qu’il ne [li] fust mout [bien] avis Qu’il veïst Gauvain en apert.  (Hunbaut, vv. 3112–19) [It was sculpted from wood so skilfully and with such close resemblance that the most perceptive of men, or most shrewd, however well acquainted he might be with Gauvain, looking the statue in the face, would have been under the impression that he was seeing the real Gauvain.]

The remarkable likeness of the statue recalls the picture used by the Pucelle de Lis to validate Gauvain’s identity in the First Continuation. But here, rather than resolving ambiguities, the work of art causes them: when the chamberlain accidentally leaves the bedroom door open, Keu and Gifflet catch sight of the statue and mistake it for the real thing. Scandalised, they tell their companions that Gauvain, who should be out rescuing his sister, is instead dallying in the damsel’s bed. An argument ensues between those who believe what their eyes have seen and those whose knowledge of Gauvain’s virtues tells them that there must be some misidentification at work: Et respondire[nt] tuit ensanble: ‘Ce n’est il pas, mais il le sanble Iche ne poroit il pas estre’  (Hunbaut, vv. 3209–11) [And they all replied as one: ‘It is not Gauvain, though it resembles him; it could not be him’]

Ironically enough, eyewitness accounts prove less reliable than convictions based on one’s own judgment of character. Moreover, the irony is compounded by the fact that one can quite easily imagine Gauvain becoming distracted from the task at hand by an amorous encounter: this is, after all, what happens in La Vengeance Raguidel when his entanglement with Ydain temporarily displaces the central quest on which he is engaged; or, in Hunbaut itself, when his earlier attraction to another damsel threatens to jeopardise the

184 THE

CONTE DU GRAAL CYCLE

mission on which Arthur has sent him. The statue incident makes the same point as Mahot’s invocation to her mistress in La Vengeance Raguidel to open her eyes in order better to be duped, or the case of the squire in L’Atre périlleux whose proximity to the murder does not prevent him from misrecognising the victim. Recognition in Arthurian romance is a process created out of rhetoric, and narrative ‘truth’ is ultimately a matter of who can tell the most convincing narrative. If this episode begins with a figure for intertextuality, the ‘roumant’, it ends with an even clearer invitation to the audience to read Hunbaut in relation to other romances: as Arthur prepares to leave, his parting words are to ask the damsel’s name. He receives the following reply: Sire, volentiers le dirai Ne ja ne le vos celerai. J’ai non ‘Cele du Gaut Destroit’, Qui vo [ne]veu Gauvain covoit.  (Hunbaut, vv. 3407–10) [Lord, I will tell you it willingly, and would never hide it from you. I am the Maiden of the Narrow Wood, the one who longs for your nephew Gauvain.]

This name appears to mean little to Arthur, since he departs without further comment. Clearly, then, it is intended more for extratextual audiences familiar with the character from La Vengeance Raguidel. It invites such readers to identify themselves with the group of initiates who would have appreciated the transformations operated by the Hunbaut author on his predecessors’ work. To give but one example among many: in the earlier text, Gauvain is introduced to the Pucele de Gaut Destroit as Keu; in Hunbaut, it is the same Keu who is sent ahead of Arthur’s party to greet her, and who will later be a victim himself of a failure of recognition. These kinds of game played with the audience’s knowledge and memory of other texts demonstrate the fundamentally intertextual nature of Arthurian verse romance, and the delight of recognition must be regarded as a central aspect of its aesthetic appeal to its audiences past and present. The statue of Gauvain is a sign to be read (and misread) by Arthur’s knights, but also a sign full of intertextual significance for the initiated audience. It invites recognition of Hunbaut’s position in the canon of Arthurian literature. Busby suggests that ‘Hunbaut, and the corpus of Arthurian verse romance in general, is fashioned in the sanblance of, and embodies features from, Chrétien’s oeuvre, but, like the statue, it is not the real thing.’23 I hope to have shown that the intertextual signals generated by the scene with the Pucele de Gaut Destroit in Hunbaut place the text and its antecedents (which should be conceived as encompassing much more than Chrétien’s works) 23

Busby (1994a: 64).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

185

in a more complex relation than that of model and counterfeit copy. Rand observes of this scene that ‘[i]t is not the statue that will be judged against Gauvain (should he ever turn up) but Gauvain who will be judged against the statue’;24 similarly, Hunbaut invites us to reverse the usual logic of identification of the copy with its original. Just as the Second Continuation can be held to have ‘created’ the Conte du Graal cycle in the form that we have been studying in this book, so each new Arthurian romance has the capacity to modify the horizon of expectation for the genre. And as a corpus, in their delight in placing Gauvain in situations designed to explore the frailty of the fit between nom and renom, the verse romances rewrite Chrétien de Troyes’s Arthurian universe through the prism of the Conte du Graal cycle.

Ceci n’est pas un cycle: Responding to the Conte du Graal Cycle If the Pucelle de Lis encounter from the First Continuation provides the model for the scenes we discussed in the previous section, this is part of a general tendency for the thirteenth-century Arthurian verse romances to refer back to, or rework material from, events in the Conte du Graal cycle. We can identify two distinct trends in this regard, which I will discuss in turn. First, the elements from the Conte du Graal cycle which find their way into these romances tend to be material that originates in Chrétien’s part of the text, before being taken up and reused in the first two Continuations. Indeed, in this respect the First Continuation appears to have played a significant role in defining certain of the more recurrent aspects of the genre. Secondly, a minority of romances dialogue especially closely with the cyclic intertext, and these tend to work more with motifs and themes specific to Chrétien’s portion and the Second Continuation (that is, the biographical ‘Perceval’ narrative around which the cycle coheres). Thus, in both cases, the nature of the response to Chrétien’s text is determined by the manner in which the later romances choose to engage with the cycle, whether they interact more with its centripetal (CdG-C2) or centrifugal (CdG-C1) impetus. The first case of transference from the Conte du Graal cycle to the later verse romances which I will discuss is Arthur’s custom not to sit down for dinner until an adventure has arrived. It is frequently noted that this is one of the most common motifs of the genre.25 It makes its first appearance in Arthurian literature in Chrétien’s Conte du Graal, when Clamadeu, who has been defeated by Perceval and dispatched to Arthur, arrives at court (CdG Rand (1998: 174). See for instance Schmolke-Hasselmann [Middleton] (1998): ‘The most important tradition, which becomes very forcefully established, is the custom of waiting for aventure’ (88); and: ‘As the genre approaches the end of its development the motif of waiting for adventure gradually emerges as the really central Arthurian motif’ (42). 24 25

186 THE

CONTE DU GRAAL CYCLE

2785–826). It is Pentecost; Keu comes through the hall before the king and informs him that dinner is ready. Arthur is in the process of rebuking his seneschal, insisting that he will not eat until he receives some new information, when Perceval’s prisoner arrives. Because this motif originates in Chrétien’s romance, it seems to fit nicely into the critical narrative that makes Chrétien’s œuvre the primary model for the later texts. Scholarly work on the Occitan romance Jaufre, for example, has noted the parallel between this passage and the opening of the later romance, in which the motif of waiting for adventure is coupled with close verbal reminiscences of the description of Keu coming before the king (all emphases in what follows are my additions): Et Kex parmi la sale vint Ab tant Quexs per la sala venc, Trestoz desaffublez et tint Desenvoutz, et en sa man tenc En sa main destre .i. bastonet, Un baston parat de pomier. El chief .i. chapel de bonet Dont li chaveil estoient blont. N’ot plus bel chevalier el mont, Et fu trechiez a une trece; Mais sa biauté et sa proëce Mais siei gab e siei villain ditz Empiroient si felon gap. Li tolon de son pretz granren Sa cote fu d’un riche drap De soie tote coloree; Çains fu d’une çainture ovree, Dont la bocle et trestuit li membre Estoient d’or, bien m’en ramembre, Que l’estoire ensi le tesmoigne. Chascuns de sa voie s’esloingne E anc no-i ac pro cavalier, Si comme il vint parmi la sale; Que volentiers no-ill fezes via; Ses felons gas, sa langue male Car cascun sa lenga temia Redoutent tuit, si li font route; Per ses vilains gaps que gitava Car n’est pas sages qui ne doute, Ou soit a gas ou soit a certes, Felonnies trop descovertes. Ses felons gas trop redoutoient Trestot cil qui laiens estoient, C’onques nus a lui ne parla. Et il devant toz s’en ala Ab aitant denan lo rei ven Jusqu’al roi la ou il seoit, E ditz: ‘Seiner, sazon seria Et dist: ‘Sire, s’il vos plaisoit, De manjar oimais, si-us plazia.’ Vos mengeriez des ore mais. E-l reis es se ves el giratz: – Ke, fait li rois, laissiez m’en pais; ‘Quexs, per enugz a dir fos natz Que ja par les oex de ma teste Et per parlar vilanamentz; Ne mengerai a si grant feste, Que ja sabetz vos veramentz Que je cort esforchie tiegne, E avetz o vist mantas ves, Que non manjaria per res Jusque novele a ma cort viegne.’        (CdG 2793–826) Que cort tan esforsada tenga, Entro que aventura mi venga

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

187

O qualque estraigna novella De cavalier ou de pulcella.’ (Jaufre, vv. 123–5; 138–9; 126–9; 140–52)26 [Conte du Graal: And Keu came through the hall without his cloak; he had a small stick in his right hand and a hat of blond felt on his head. There was no more handsome knight in the world, and his hair was tied in a ponytail. But his beauty and prowess were spoiled by his cruel jokes. His tunic was made of rich coloured silk; it was girded with a belt of which the buckle and loops were made of gold, as the story testifies, I well remember. Everyone moves out of his way as he came through the hall; they all fear his cruel jokes and wicked tongue, and so stay out of his way. For it is unwise not to fear direct calumny, whether intended seriously or in jest. All those who were in the hall greatly feared his cruel jokes, so that no one dared speak to him. And he came before everyone, right up to where the king was sitting, and said: ‘Sire, if it pleased you, you could eat now.’ ‘Keu,’ said the king, ‘leave me in peace; for by the eyes in my head I will never eat, when holding a great court on such an important feast day as this, until news comes to my court.’] [Jaufre: Then Keu came through the hall in a carefree manner, carrying a stick made from applewood… But his jokes and wicked words detracted greatly from his worth… There was no worthy knight who did not willingly get out of his way, because everyone feared his tongue for the cruel jokes it spat out… Now he comes before the king and says: ‘Sire, it is time to eat now, if it pleases you.’ And the king turns to him and says: ‘Keu, you were born to say disagreeable things and speak in a base manner; for you well know and have seen many times that I will not eat for anything, when holding such a great court, before an adventure has come to me, or some strange news concerning a knight or a damsel.’]

While these verbal echoes have been used as evidence for the Jaufre poet’s intimate textual knowledge of Chrétien’s romances, such discussions have not taken into account two very similar passages which appear in the Caradoc branch of the First Continuation. I have reproduced the full extracts below for ease of comparison; the first prefaces the arrival of Eliavrés at court to propose the beheading contest, while the second introduces the Chastity Horn episode which closes the branch: Et Qex li senescaus s’en ist D’une cambre el palais ça fors, Tos desfublés, en pur le cors, Vestus d’un fres blïaut ermin, Covers d’un rice baudequin;

26

On the parallels between these two passages, see Baumgartner (1978: 623–34).

188 THE

CONTE DU GRAAL CYCLE

Et il ert grans et biaus et genz. Parmi outre totes les gens Au dois devant le roi s’en vint; En sa main une verge tint. Il s’agenolle belement, Si li a dit cortoisement: ‘Sire, l’ive porïés prendre, S’il vos plaisoit, sans plus atendre, Car tos est pres vostres mangiers. – Non fera[i], Qex, biaus amis ciers. Ne place Diu que ja aviegne C’a haute feste rant cort tiegne, Por c’aie corone portee, Qu’ive soit prise ne dounee Devant que estrange novele Ou aventure laide o bele I soit, voiant tos, avenue. La costume ai ensi tenue Tote ma vie entresque ci.’  (C1 [SR: L] 2216–39)27 [And Keu the seneschal came out of a room in the palace, without his cloak, simply dressed, wearing a tunic of ermine covered with a rich sheet; and he was tall and handsome and becoming. He came through the people to the dais before the king; in his hand he held a stick. He knelt graciously and said courteously: ‘Sire, you could take the water now if it pleased you, without further delay, for your meal is ready.’ ‘I will not do so, Keu, dear friend. May it never please God, when I hold court and wear my crown for an important feast, that the water be taken or given before some strange news or adventure, whether good or bad, has arrived in full sight of all. I have upheld this custom all my life until this point.’] D’une cambre est Qex [fors] issus, Tos desfublés, si est venus Devant le roi et si li dist: ‘Sire, sire, se Dex m’aït, Les graisles feroie souner, S’il vos plaisoit, por le laver, Car tos est pres vostres mangiers.’ – Non ferois, Qex, biaus amis ciers; D’iaue doner ne parlés ja. Ma costume savés pieç’a: Il ne m’avint onques encore, No fera il, se je puis ore, 27 The Long Redaction gives a shorter version of the first passage, but otherwise all Redactions give essentially the same text.

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

189

Que mangase a cort que tenise Devant c’avenir i veïsse Mervelle estrange u aventure.’  (C1 [SR: L] 3125–39) [Keu came out of a room without a cloak; he came before the king and said to him: ‘Sire, sire, God help me, I would have the trumpets sounded to announce the water, if it pleased you, for your meal is ready.’ ‘I will not do so, Keu, dear friend; do not speak yet of giving the water. You know that this is a long-standing custom of mine: it has never yet happened to me when holding court, nor will it if I can help it, that I eat before seeing some strange marvel or adventure arrive here.’]

Certain details are missing in the First Continuation extracts, mostly concerning Keu’s sharp tongue and the fear he inspires in others. On the other hand, where Chrétien’s text has Arthur refuse to eat ‘Jusque novele a ma cort viegne’ [until news comes to my court] (2826), the Jaufre extract follows the first passage from the Caradoc material in expanding the notion of ‘novele’ into ‘estrange novele’ [strange news], and the second passage in supplementing this idea with that of ‘aventura’ [adventure]. In similar fashion, the detail in Jaufre of Quexs coming ‘denan lo rei’ [before the king] is found in both First Continuation extracts but not in the Conte du Graal; in the earlier passage, as in the Occitan text, the relevant line ends with the rhyme-word ‘vint’. Moreover, the notion that Arthur’s waiting for adventure represents a well-worn habit is at most implicit in Chrétien’s text. Indeed, the comparison of the four passages appears to trace the development of the motif into a fully fledged custom. First enunciated in Chrétien’s section, it is in the First Continuation that it becomes a ‘costume’ [custom] which the king has upheld ‘tote ma vie’ [all my life] (vv. 2238–9); subsequently, Arthur is able to remind Keu (and the audience) that the custom has already been stated (‘Ma costume savés pieça’ [you know that this is a long-standing custom of mine], v. 3134). The developing popularity of the motif, which can be found in most of the romances discussed in this chapter, and several others besides, means that by the time the Jaufre author is writing, Arthur’s comment to Quexs that ‘avetz o vist mantas ves’ [you have seen this many times] (v. 147) is a statement the audience can agree with.28 If the verbal parallels between the passages in Jaufre and the Conte du Graal suggest that the Occitan author had written

28 There has been disagreement over whether Jaufre is earlier or later than the Conte du Graal, and therefore over which of the texts influenced the other. The two dates around which the debate has raged are c. 1180 and c. 1220–30, depending on how one resolves the Occitan romance’s references to the ‘rei d’Aragon’ [king of Aragon] (vv. 61–84; 2623– 40). A general consensus now appears to have been reached in favour of the later date, which would mean that Jaufre also post-dates the first two Continuations. For a summary of the debate which endorses this view, see Gaunt and Harvey (2006: 535–6).

190 THE

CONTE DU GRAAL CYCLE

copies of Chrétien’s poem before him,29 then the manuscript tradition of the text makes it quite likely that these would have included the First Continuation material. The First Continuation’s reuse of the motif may help to explain why this specific moment in Chrétien’s text, which is not flagged as especially significant, found its way into the Occitan romance. A similar process of assimilation into the textual tradition can be detected in relation to the name ‘Melian de Lis’. In Chrétien’s Conte du Graal, he is a minor character who has the misfortune to be injured by Gauvain in a tournament. The confrontation between the two men is eclipsed by that between the two damsels whom they have championed, the Pucele as Petites Manches and her sister. The First Continuation introduces the niece and nephew of Melian (who is revealed to have died of his injuries), these characters having a far more important role to play in the narrative: the Pucele de Lis, whose avatars throughout Arthurian verse romance we have been studying, and her brother Bran, whose vendetta against Gauvain provides the connecting thread between Branch II and Branch IV of the First Continuation. The name ‘Melian de Lis’ (or sometimes simply ‘Melian’) is then used for a character in Meraugis, also mentioned in La Vengeance Raguidel, for a dead knight in Le Chevalier as deus espees and for a minor character in Jaufre. An incidental detail from Chrétien’s text, having received amplification in the First Continuation, thereby enters into the common stock of Arthurian material. If these borrowings indicate that the material from the Conte du Graal is being viewed through the lens of the First Continuation, it is also possible to identify instances of direct borrowing specifically from the Continuations. One clear case is the motif of the mysterious boat which appears at court bearing a knight whose death must be avenged. This event provides the initial momentum for both Branch VI of the First Continuation and La Vengeance Raguidel, and is possibly being alluded to in the Elucidation’s mention of a tale about ‘li chevaliers mors del calan’ [the dead knight of the skiff] (Elucidation, v. 363). Another motif found in more than one romance appears to have its basis in Branch IV of the First Continuation. Keu, sent ahead by Arthur to find food, is involved in an altercation with a dwarf who beats him with a peacock on a spit, badly burning him. In Hunbaut, a hungry Gauvain takes similar action against a knight who refuses to share his venison; in Fergus, the eponymous hero also has to mete out punishment with a charged spit in order to get something to ease his hunger; finally, in Jaufre, Quexs is reminded by Melian of the knight ‘que-us feri ab lo paon / El col’ [who struck you in the neck with the peacock] (vv. 6653–5), a clear reference to the incident from the First Continuation. D. D. R. Owen has argued that Guillaume le Clerc, author of Fergus, ‘used his knowledge of Chrétien and the Continuations in almost equal measure in

29

See Gaunt and Harvey (2006: 538).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

191

building up his romance’, and indeed Guillaume is perhaps the only writer of this period to turn explicitly to the Second Continuation for material.30 The most evident instance of this is the episode in which Fergus kills a giant, freeing two damsels from the castle where he has been keeping them; in doing so, he is repeating an exploit carried out by Perceval in the Second Continuation. Apparently, Guillaume wished to signpost his interaction with that text, since he introduces the episode by having Fergus lose his way and end up at the Mont Dolereus (Fergus, vv. 4457–61). As we saw in Chapter 1, this place name features prominently on Perceval’s itinerary in the Second Continuation, as the setting for a culminating demonstration that he is the best knight in the world. By having Fergus, who is on a quest to rescue his sweetheart Galiene, lose his way, the author seems to suggest that he has stumbled out of his own narrative and into Perceval’s. The intertextual links between Fergus and Perceval are in any case remarkably strong, since Fergus is explicitly a neo-Perceval: not only are his childhood and arrival at Arthur’s court closely modelled on those of the Welshman, but a richly symbolic moment occurs when Perceval appears during the young man’s knighting ceremony: Es vos Perceval le Galois Par miliu de la sale errant. Une espee en sa main trençant Tint Percevals que li donna Ses bons ostes qui [l’]herberga.  (Fergus, vv. 1410–14) [Then Perceval the Welshman came into the middle of the hall; in his hand he held a sharp sword that had been given to him by his good host who had sheltered him.]

This last line is, of course, a reference to the moment in Chrétien’s Conte du Graal when the Fisher King offers Perceval a sword. Schmolke-Hasselmann speaks of Perceval’s ‘fatherly manner’ towards the hero, and in having him hand his sword to Arthur so that he may gird it around Fergus there is an obvious implication that the baton of adventure is being handed down to the next generation.31 In light of the metaphors for intertextual communication which we have studied in the preceding chapter and this one, this moment would seem to figure a further filiation between the Perceval cycle and its literary descendant. If Fergus draws heavily on the early adventures of Perceval, Le Chevalier as deus espees seems especially interested in the Hideous Damsel’s prediction of dire consequences resulting from Perceval’s failure to assuage his curiosity at the Grail Castle: 30 31

See Owen (1987/88: 3). Schmolke-Hasselmann [Middleton] (1998: 159).

192 THE

CONTE DU GRAAL CYCLE

Dames en perdront lor mariz Terres en seront essilliees Et puceles desconseillies Qui orfenines remandront, Et maint chevalier en morront.  (CdG 4608–12) [Ladies will lose their husbands, lands will be lain waste, damsels left orphaned and abandoned, and many knights will die because of it.]

We saw in the previous chapter how the vocabulary of sterility, and its rejection by the Continuations, fed into an intertextual debate around the nature and feasibility of continuation. Le Chevalier as deus espees is not a continuation, but its interest in the notion of filiation, and in attaching itself to textual precedent, is apparent in its recycling of this semantic field. However, whereas the Hideous Damsel’s harangue draws a link between Perceval’s failings and the continuing, even worsening, sterility of the land, the relation between social catastrophe and personal responsibility is quite different in the later text. Its hero, called the ‘chevalier as deus espees’ [knight with the two swords] because he does not know his real name, learns that his father Bleheris has been killed by a wicked knight named Brien de la Gastine (a different Brien from the one who claimed to have killed Gauvain). On his way to exact vengeance, he comes to a convent where he meets his father’s sister. Upon learning his identity, she embraces her nephew, describing her pain at Bleheris’ death: La dame dist: ‘Biaus tres dous niés, Tant me fu et pesans et griés Ma vie, puis ce ke je soi La mort vostre pere. Et je n’oi De vous nule certaineté, Car on vous ot desireté, Et la vostre mere escillie; Në ele n’estoit conseillie Par nul homme ki fut en vie, Car Brïens par sa trecherie Li ot tous ses hommes ocis.’ (Chevalier as deus espees, vv. 8363–73) [The lady said: ‘Dear nephew, my life has been difficult and painful since I learned your father’s death. And I knew nothing for certain about you, for you had been disinherited and your mother exiled; she had no guidance from any living man, for Brien in his treachery had killed all of her men.’]

This speech, picking up on the rhyme pair ‘(des)conseillie / escillie’ from the Hideous Damsel’s harangue, bears witness to two important transformations in the theme. First, the hero is no longer made responsible for the events that

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

193

have occurred. In the Conte du Graal cycle, an overlapping series of issues contrive to force Perceval to admit responsibility for the Fisher King’s injury, his mother’s death and even the collapse of the social order. Here, the sin is attributed to Brien, a figure on whom revenge can (and must) be taken. The development of the hero as a subject is thus made to depend, not on exorcising the demons of personal guilt, but on performing the appropriate ritual of vendetta on a designated enemy. Indeed, Bleheris’ final instructions to his wife were that their son should not learn his real name (which is Meriadeuc) until he has avenged his death (vv. 6831–8). Filiation is thus conceived as a duty to pursue the father’s unfinished business, and the taking of revenge on the family foe as a precondition for recognition as a son. Concurrently, the scope of social disorder is reduced. In the Conte du Graal, what appeared to be a family matter (Perceval’s abandonment of his mother) turns out to impact on a whole generation of Arthurian men, the ‘maint chevalier’ [many knights] whose deaths are predicted by the Hideous Damsel. In Le Chevalier as deus espees, family matters remain just that. The specific points of linguistic contact between the two texts on this point might encourage us to see the later romance as set implicitly in the wake of Perceval’s adventures, but the point I want to stress here is precisely that, if such momentous events are unfolding, they do not impact on this narrative. The heroes’ adventures are of importance to them, but they have limited impact on the lives of other knights. This point is of general importance in understanding how the aesthetics of Arthurian verse romance evolved out of, and in part against, the cyclic dynamics of the Conte du Graal and (subsequently) the Vulgate Cycle corpuses. Both of these cycles propose a narrative in which individual adventures are subsumed to an overarching central trajectory. In the Vulgate Cycle, this is provided by the eschatological framework of the Grail material, to which the fall of Arthur’s kingdom provides a kind of secular coda; in the Conte du Graal cycle, as I have argued throughout the present study, it is Perceval’s progress towards the status of ‘world’s best knight’ which tends to function as a magnet to attract the adventures of others into his itinerary. The ostensible narrative motor is desire for the end point, with individual adventures functioning to prolong the text by delaying the achievement of this desire for as long as possible. Thirteenth-century verse romances, by contrast, limit themselves to recounting ‘une aventure’ [an adventure] (Atre, v. 2). There is no need for their heroes to be established as pre-eminent among all knights. Indeed, in most of these texts, that honour is reserved for Gauvain. Thus, in describing the knighting of Fergus, Guillaume le Clerc’s narrator remarks that he is armed better than any man alive, as long as one excepts Gauvain: Cil ne trova onques son per Ne par home ne fu matés; Por ço vel qu’il en soit ostés  (Fergus, vv. 1430–2)

194 THE

CONTE DU GRAAL CYCLE

[He was never equalled, nor defeated by any man; for this reason I wish to exclude him]

Similarly, in Le Chevalier as deus espees, Meriadeuc objects to Brien’s claim that his defeat of Gauvain makes him the best knight in the world, on the grounds that that title should now go to him: Quant mesire Gauvains est mors Si con vous dites, après lui Di je ke li chevaliers sui Mieudres et li plus biaus du mont (Chevalier as deus espees, vv. 5544–7) [If Sir Gauvain is dead as you claim, I say that after him I am the best and most handsome knight in the world]

In essence, Gauvain’s function is analogous to that of a monarch in a modern parliamentary democracy: by holding the post unchallenged, without exercising the executive powers associated with it (such as ‘achever les aventures’ [ending the adventures], which is the explicit prerogative of Galaad in the Queste del Saint Graal and Perceval in the Manessier Continuation), he assures the stability of the system of errance and chivalric adventure. These romances, whose heroes Schmolke-Hasselmann describes as functionally ‘interchangeable’, may change the generic horizon of expectation, as I argued above, but the events they recount rarely leave a trace on the diegetic Arthurian environment.32 Indeed, in certain cases, the conclusion of the text is accompanied by a sense that everything is back to how it was at the beginning, an aesthetics of no-consequence. Jaufre opens and closes with episodes in which Arthur is abducted by a marvellous beast, which turn out to have been orchestrated for the court’s amusement by one of the king’s knights, who has powers of metamorphosis. Jean-Charles Huchet and Caroline Jewers have both identified this knight-magician as a figure for the author, designed to highlight the fictionality of the narrative and the artificiality of the conventions of chivalric adventure.33 I would like to add two further observations to this line of analysis. First, as Schmolke-Hasselmann notes, that ‘this criticism of aventure, perceptible everywhere and recognizable as a persistent motif throughout the Arthurian verse romances of the thirteenth century, is still not capable of rocking the foundations of the traditional narrative structures of Arthurian romance: aventure is much too vital an element in romances of chivalry’.34 Secondly, that the position of the scenes at either extreme of the text, as well as their circular narrative logic, provides a précis of Arthurian 32 33 34

See Schmolke-Hasselmann [Middleton] (1998: 14). See Huchet (1989) and Jewers (2000). Schmolke-Hasselmann [Middleton] (1998: 87).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

195

verse romance aesthetics: no consequences arise from either episode, nor do the adventures and exploits of Jaufre leave their mark on the development of the genre. The pleasure of the Arthurian court, when the magician’s transformation is revealed, mirrors that of the extradiegetic audience who know that each romance is a new experience to be enjoyed, a new game to be played. As Busby argues, this conflation of the audience within and without the text is a typical feature of the genre: Aventure exists in general in Arthurian romance to provide matière for entertainment at the king’s court. Knights return from their travels and relate their experiences to audiences within the romance; they are, of course, in their turn being manipulated by authors telling the tales to an extradiegetic audience.35

Adventure, then, exists to bring the pleasure of narrative back to the static court, usually at mealtime, as if the courtiers depended on it for their nourishment. This connection between stories and sustenance is brought out especially forcefully in the opening scene of La Vengeance Raguidel, where the failure of adventure to arrive eventually prompts Arthur to send his knights to eat without him. Starved of both food and narrative, he spends a restless night until he is rescued by the arrival of the mysterious boat. L’Atre périlleux is another romance which begins with the ‘refusal to eat’ motif, and here the hunger for adventure is soon sated, as a damsel arrives at court and asks Arthur for the privilege of serving the wine at the following day’s meal. When she is abducted, it is up to Gauvain, in whose protection she was supposed to be, to give chase. But it soon emerges that here again the adventure is not what it seems: the whole incident was orchestrated by the abductor, Escanor, the ami of the damsel, in order to have reason to test himself against Gauvain. In the Conte du Graal, Perceval’s family business opens out onto a social catastrophe; here, what had appeared to be a challenge to the whole court turns out to concern only Gauvain. As in Le Chevalier as deus espees, the scope of ‘aventure’ is limited, its consequences operating on a reduced scale. Indeed, the end of the romance presents the audience with a narrative of no-consequence similar to those which bookend Jaufre, as Gauvain confronts and defeats the Orguelleus Faé, the knight behind the murder of the false ‘Gauvain’. As his name implies, we are in the presence of another magician-figure, who once again appears to have arrogated the author-function to himself, in that he can erase the supposedly indelible effects of the narrative’s events: Ne doit estre en tel mal retraite, Puis qu’ele puet estre amendee. 35

Busby (1994a: 53).

196 THE

CONTE DU GRAAL CYCLE

Voiant cels de vostre contree, Vous renderai cel chevalier, Et ses armes et son destrier, Tout sain si com onques fu plus.  (Atre, vv. 5830–5) [The deed should not be considered as so bad, since it can be reversed. Before all of your fellow countrymen I will restore this knight, his arms and his horse, to the best health he ever had.]

Not only will the dismembered body be reassembled and reanimated, but the blinded squire will get his vision back – and this will all happen before an audience of Gauvain’s compatriots (v. 5832), providing another opportunity for spectatorship. Alexandre Leupin reads the broken sword of the Conte du Graal cycle, which Perceval can repair only imperfectly, leaving a small chink between the two pieces, as a symbol of the logic of la faille, the tip of an iceberg of incompletion that pervades the narrative.36 By contrast, the reassembly of the Orguelleus’ victim (eventually identified as Cortois de Huberlant) in L’Atre périlleux is described as a simple operation. Having returned to the dwelling of Tristan-qui-ne-ment to recover the corpse’s arm, before the astonished gaze of his travelling companions, the Orguelleus works his magic: Et le Faé sans delaier A le bras de l’escrin osté, Et el cors ariere posé; Puis fu plus sains que nul poisson.  (Atre, vv. 6378–81) [And the Faé, without delay, took the arm from the box and placed it back on the body; then the knight was as healthy as any fish.]

Each author of a new romance has the ability to wave his hand and make the adventures of the past disappear from narrative memory; each text constitutes a new Arthurian subject to be cut up and reassembled.37 It is not that the Conte du Graal cycle does not contain these kinds of adventure. Even 36 See Leupin (1982: 238): ‘les figures du manque, dans le récit, renvoient sans cesse de l’une à l’autre, dans un perpétuel tourniquet de substitutions’ [in the narrative, the figures of lack refer ceaselessly back and forth to each other, in an unending cycle of substitutions]. 37 Michelle Szkilnik notes a tendency for verse romances in which Gauvain plays an important role to insist on the theme of forgetting: ‘L’oubli est constitutif du personnage de Gauvain et c’est une condition nécessaire à la prolifération des romans dont il est le protagoniste’ [forgetfulness is essential to the character of Gauvain, and a necessary condition for the proliferation of romances in which he is the hero] (Szkilnik 2007: 92). Put in terms of the argument of this chapter, the failure of memory to carry through from one text to another can be seen as the guarantee of the independence of individual narratives, though Szkilnik rightly insists that the theme of forgetting itself becomes an element in the generic horizon of expectation with which each romance dialogues, and a means for

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

197

the Second Continuation, which we have consistently identified as the text which creates the cyclic coherence of the whole, is so built that ‘many of the episodes… could be removed without fundamentally affecting the development of the main themes’, as Corin Corley has demonstrated.38 But these episodes take their place, and find their meaning, in the context of the cycle’s generally centripetal drive. This, I believe, is what makes the chink in the mended sword so significant: it signals a kind of ‘imperfect’ coherence which attempts to accommodate centrifugal impulses within a broadly centripetal narrative dynamic. For Manessier, this was an entirely possible state of affairs, and he clearly considered the sword theme to have run its course; for Gerbert, the accession of Perceval to the rank of best knight in the world could not be satisfactory until the chink was filled in, and this imperative was enough to send Perceval on a looping, seventeen-thousand-line journey with plenty of room for more ‘non-essential’ adventures along the way. Yet the central Perceval dynamic does leave the Arthurian world altered in a way that the kind of aventure we have been examining simply does not. This is true within the cycle, where other characters’ adventures become reassigned to Perceval, or where Gauvain’s dalliance with the Pucelle de Lis leaves him with a son who will cross his path in later episodes. It is also the case that the events of the Conte du Graal cycle are taken into account by a number of other texts, though they do so subtly, as if to maintain sufficient distance from the corpus to avoid becoming subsumed into it. This important aesthetic difference between the cyclical narrative and the verse romances thus needs to be understood as complementary rather than oppositional: the aesthetics of no-consequence deployed in the verse romances to an extent depend on locating consequence elsewhere, in the cyclical corpus, with Perceval. As we have seen, Fergus makes direct reference to Perceval’s visit to the Fisher King; he is also seen riding the horse ‘C’au vermeil chevalier toli’ [that he had seized from the Scarlet Knight] (Fergus, v. 147). Even more interesting for our purposes is the reference in the prologue to ‘Perceval / Qui tant pena por le graal’ [Perceval who strove so arduously for the grail] (vv. 13–14); although the reference remains vague, the idea of striving for the Grail seems to fit better the events of the later Continuations (C2-CG-CM) than those of Chrétien’s romance, where the significance of the object within the narrative economy only becomes apparent towards the end of Perceval’s narrative. Since we know that Guillaume le Clerc was familiar with the Second Continuation, this provides further support for the contention that the ‘Perceval’ story referred to by these other texts should be understood as some configuration of the Conte du Graal cycle, and not just Chrétien’s portion its author to remind his audience of the intertextual aesthetics of the genre within which he is writing. 38 Corley (1987: 42–5, 45 for the quotation).

198 THE

CONTE DU GRAAL CYCLE

of the text. As argued above, Fergus sets up a relation of filiation between the hero and his illustrious antecedent, which can be read as establishing an analogous relation between the romance and the Conte du Graal cycle. As Schmolke-Hasselemann claims: ‘here Perceval is the one who in his time has already experienced and can remember everything that now awaits the youthful hero. In this sense the romance is a historical continuation of the earlier work’.39 By setting its events some time after the events of Perceval’s adventures, Fergus grafts itself onto the Conte du Graal cycle as a kind of suite. In L’Atre périlleux, the younger sister of the damsel whose love for Gauvain causes the murder of his false double is said to be in love herself with another of Arthur’s knights. She intends to ask the king to arrange a match between herself and the ‘Chevalier Vermel’ (v. 5124): Ki a la cort le roi Artur Vint antan estre chevalier, Ki les armes et le destrier Conquist par son cors seulement, Desarmés, par grant hardement, Et li rois Artus li dona Del chevalier qui en porta Sa coupe d’or de devant lui.  (Atre, vv. 5128–35) [Who came, long ago, to King Arthur’s court to become a knight; who conquered the arms and horse unarmed through great bravery, with only his body as a weapon; King Arthur gave him those of the knight who had carried off his golden cup.]

It is telling that Perceval is not named directly here; indeed, the damsel is forced to admit: ‘Je ne sai pas nomer le mien’ [I do not know what the name of my knight is] (v. 5108). The narrator invites the audience to recognise the reference to the opening of the Conte du Graal, and thereby sets his romance in the same diegetic universe inhabited by the cycle’s characters. At the same time, he keeps this intertext at arm’s length, as a narrative whose effect is felt at a distance by the younger sister. And despite its appearance at the heart of the events that motivate the central narrative, it has no effect whatsoever on the development of these. The plot to kill Gauvain entirely overshadows the mention of Perceval, who passes under the narrative radar as far as the romance’s villains are concerned. Two other romances, Beaudous and Durmart le Galois, make similarly fleeting mention of Perceval, but play with the intertextual past of the char39 Schmolke-Hasselmann [Middleton] (1998: 160). See also Zemel (1994: 339), who argues that Guillaume ‘reduces Chrétien’s Perceval to prior history to be continued and fulfilled in the romance about Fergus’.

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

199

acter in markedly different ways. In Beaudous, he wears a sleeve that the ‘pucele de Belrepaire / Qui vout de lui son amin faire, / Li envoia par druerie’ [the damsel of Beaurepaire, who wanted to make him her lover, had sent him as a token of her devotion] (Beaudous, vv. 3944–6); in Durmart, his appearance at the tournament disappoints the damsels who have come hoping to marry the victor, since they are sure they will have no joy from him: ‘Ne esperance n’i avoient / Por la queste del saint Graal; / Quar il [ert] castes et loial’ [nor did they have any hope, because the Quest for the Holy Grail had made him chaste and virtuous] (Durmart, vv. 7376–8). This may be a reference to the Perceval who accompanies Galaad in the Queste del Saint Graal, or it may refer to the character as he evolves in the two later Continuations (Gerbert’s Continuation is particularly insistent on the importance of chastity, while Manessier has him become a hermit at the end of his life). What is significant in both cases, however, is that Perceval’s narrative past prevents him from being a love interest in the romance at hand; his intertextual baggage bars him from significant involvement in the events of these texts. In other romances, Perceval is conspicuous by his absence, or else is included only briefly in a list of knights; he warrants such brief mentions in Le Bel Inconnu (vv. 5501–2), La Vengeance Raguidel (v. 4298+1; see ‘Variantes’ to Roussineau’s edition, p. 340) and Le Chevalier as deus espees. This last case is especially interesting: we have seen how closely the text dialogues with the Conte du Graal cycle, yet the only mention of Perceval nods deliberately to the prose tradition represented by the Didot-Perceval and the Perlesvaus, identifying him as ‘Perceval, li fil Alain / Le Gros des Vaus de Kamelot’ [Perceval, son of Alain the Fat of the Vales of Camelot] (vv. 2608–9). It is tempting to read this reference to a competing Perceval tradition as a deliberate obfuscation of the obvious filiation of the narrative with the Conte du Graal cycle. In the case of La Vengeance Raguidel, Perceval’s name has been added into the text by a scribe eager to expand a list of knights at Arthur’s court. For Perceval to be readmitted into the text in this way, as just another piece of the Arthurian background, only serves to highlight further the initial exclusion. The refusal to discuss Perceval is another point of contact between these romances and the First Continuation, where the only mentions of the eventual hero of the cycle come in the long tournament scene included in the Long and Mixed Redactions of the Caradoc branch and in an interpolation contained in most of the manuscripts (Episode V, 5 in Roach’s schema) in which the Fisher King reveals to Gauvain that Perceval belongs to the Grail lineage. Two broad approaches are thus discernible in the use made by the thirteenth-century romances of the events narrated in the Conte du Graal cycle. Some authors reduce Perceval to an anonymous Arthurian figure, stripping him of the aura of predestined saviour of the world which he has in the cycle, or else they omit him entirely. Others deliberately attempt to tie their

200 THE

CONTE DU GRAAL CYCLE

romances into the space-time of the cycle, acknowledging its importance to the genre but setting it firmly in the past as an event whose resonances no longer need to be taken into account. I would also suggest that in so doing these authors demonstrate an appreciation for the Conte du Graal cycle’s logic of narrative consequence and biographical coherence, which privilege the impression of the events of the cycle altering Arthurian verse romance ‘history’. The accretion of the Conte du Graal cycle, I have suggested throughout this study, is the result both of its core centripetal dynamic and of a centrifugal impulse towards the accumulation of extra-curricular adventures, best exemplified in the material of the First Continuation. Both the sheer size and, crucially, the coherence of the corpus worked to ensure its manuscript transmission as a single unit. The centrifugal intertextuality of the verse romances discussed in this chapter, by contrast, did not encourage their collection into stable entities. Instead, these romances (like Chrétien’s other romances, for that matter) are usually found collected along with other material in disparate anthologies (see Appendix 6). Where romances are found together, they form shifting alliances of meaning, encouraging their readers and audiences to consider the points of contact and difference thereby formed. The aesthetics of the genre, as analysed above, clearly favour such interactions, as Busby notes: ‘[t]he nature of the genre is such that it would be easy enough to make a random selection of half-a-dozen medieval French romances, place them in any order, and find thematic similarities, parallels and contrasts between them to justify that order’.40 However, whereas Busby seems to imply that such a course of action would be somewhat fatuous, I would suggest that this is precisely the kind of audience reaction which these romances invite. The one manuscript which does appear to present a deliberate collection of Arthurian romance material is Chantilly, Musée Condé 472. This manuscript contains the following texts: Les Merveilles de Rigomer, L’Atre périlleux, Erec, Fergus, Hunbaut, Le Bel Inconnu, La Vengeance Raguidel, Yvain, the Charrette, a portion of the Perlesvaus and several branches of the Roman de Renart; of these, Rigomer, Hunbaut and Le Bel Inconnu are unicae, making this a precious document for modern readers of Arthurian romance. Lori Walters has argued in a series of articles that this collection should be considered a Gauvain cycle, but if so its cyclicity is clearly very different from that of our Perceval cycle.41 The failure of the individual romances to take account of the events occurring in the preceding material makes fragmentation and narrative centrifuge a defining feature of the codex. Indeed, Rand convincingly demonstrates that narrative closure is shown to be problematic in a number of ways throughout the compilation, concluding that ‘it is this concern with types of closure, among other things, that gives the compilation 40 41

Busby (1994a: 51). See especially Walters (1994) and Walters (2006).

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

201

a forceful sense of unity’.42 Paradoxically, such unity is always provisional, always open to being supplemented by more narrative, and in this respect there are strong points of contact with the more centrifugal tendencies of the Conte du Graal cycle. A copy of L’Atre périlleux contained in another manuscript (BNF, fr. 2168) interpolates an episode in which Gauvain fights and defeats Brun Sans Pitié, king of the Rouge Cité, sending him to Arthur’s court to surrender. Whether one particular manuscript of the Atre includes this episode or not makes little difference to the basic structure of the text; the aesthetic of the text is sufficiently open to accommodate one more or one less adventure. The Conte du Graal cycle is full of similar instances of episodes added or removed, shortened or lengthened, as we have had cause to observe while picking our way through its different manifestations and configurations. Corley’s point about the non-essential nature of many of the Second Continuation’s episodes thus finds its exemplification in the manuscripts of these romances, whose scribes both appreciated and exploited the contingent nature of the texts in order to generate more narrative. Chantilly, Musée Condé 472 is a more or less unique example of a verse romance collection, while conversely the Conte du Graal Continuations are virtually never found in isolation (the exception is MS K, Bern, Burgerbibliothek 113, which includes the Second Continuation in a compilation of diverse literary texts). The coherence of the Conte du Graal cycle manifestly gives it a codicological unity that the other verse romances cannot attain, nor do they seek to.43 Yet, as the concluding section of this chapter will show, while Arthurian verse romance may be intertextually centrifugal, the influence of the cyclic dynamic deployed in the Conte du Graal cycle can be felt within individual texts in their organisation of narrative into central theme and digression, and in their use of interlace to move from one to the other.

The Centre and the Periphery: Organising Arthurian Narrative In Old French epic, vengeance and vendetta are common motors of narrative development, most strikingly when a new generation is called upon to pursue the feud which opposed their fathers. Raoul de Cambrai, for example, surfs along on a veritable chain reaction of vendettas: the rival claims of the Rand (1998: 19). Paris, BNF, fr. 1433, which contains only L’Atre périlleux and Chrétien’s Yvain, may be considered the exception that confirms the rule: too short to constitute a collection, it manifests rather a desire to pair these two texts off against each other, creating what Walters (1991) calls a ‘super-romance’. It thus demonstrates the potential for thirteenthcentury Arthurian verse narratives to be read against Chrétien’s other romances, while simultaneously confirming that a Chrétien text such as Yvain was perceived as susceptible to intertextual centrifuge in a way that the Conte du Graal, wedded to its cyclical context, was not. 42 43

202 THE

CONTE DU GRAAL CYCLE

eponymous hero and Ybert de Ribemont to the Vermandois lead to a conflict between Raoul and Ybert’s son Bernier; when Raoul is killed, his nephew Gautier swears to avenge him. Throughout the text, deaths call forth the need for vengeance, which allows the development of more narrative, providing yet more opportunities for new deaths and new vengeance claims.44 Arthurian romance also makes use of the theme in order to generate narrative. In Branch VI of the First Continuation, Guerrehet is given the task of avenging the dead knight whose corpse has been brought by boat to Arthur’s court; the central narrative of the Manessier Continuation, meanwhile, concerns Perceval’s desire to exact revenge on the knight who has crippled the Fisher King, and in so doing heal the king of his wound. In this last case, the vengeance quest has evidently been introduced into the story in order to replace the ‘grail questions’ and ‘broken sword’ themes, which had been used in the previous Continuations to motivate Perceval’s movement away from and then back to the Grail Castle.45 The revenge theme also provides a central thread in the discovery by Meriadeuc of his identity in Le Chevalier as deus espees, as well as giving its title to La Vengeance Raguidel. As a recent article on this last romance observes: ‘Pour se venger, il faut avoir la mémoire longue… ne pas quitter de vue le passé de l’injure ou du préjudice, pour préserver intact un désir tourné vers le futur de la revanche’ [To exact revenge requires a long memory… one must not lose sight of the past insult or injury, in order to keep intact the desire of future revenge].46 In other words, vengeance lends itself naturally to use as a centripetal narrative dynamic. Yet it can also appear as a threat to the narrative’s progress to its goal, as a potential distraction to the central quest: in Chrétien’s section of the Conte du Graal, Gauvain’s determination to quest, first for the Espee as Estranges Renges and later for the lance that bleeds, is derailed largely by his repeated meetings with figures who claim he has wronged them (Guingambresil, the king of Escavalon, Greoreas, Guiromelant). One fascinating way in which Arthurian verse romances show their interest in the narratological potential of the vengeance theme is by evoking and then rejecting the possibility of exploring vengeance claims. A clear case of this occurs in an episode found in manuscripts EPS of the Second Continuation, when Perceval meets an old knight whose first words to him are: Dont venez? Ou alez ainsi Par ce païs? Et si savez Que duremant mespris avez See Sarah Kay’s Introduction to her edition of Raoul de Cambrai, p. lxxi. La Mort le roi Artu also chooses to introduce vengeance as a central theme once the Grail theme has been thoroughly exhausted. In this case, Gauvain’s and Arthur’s need for revenge against Lancelot is the mechanism through which the narrative becomes locked into its irrevocable endgame of mutually assured destruction. 46 Wolf-Bonvin (2007: 103). 44 45

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

203

Et vers moi et vers mon lignaige? Certes vos faites grant oustraige  (C2 20928–32) [Where are you coming from? Where in this land are you going? And do you know that you have gravely sinned towards me and my lineage? Certainly, you have committed a great outrage.]

We soon learn that he is the brother of the Scarlet Knight killed at the beginning of Chrétien’s Conte du Graal. As he reminds Perceval: ‘Ainz puis n’an fu prise vanjance’ [revenge has never been taken] (20943). Arthurian convention might lead us to expect Perceval to face a battle here, but instead he finds forgiveness: ‘Voirs est,’ fait il, ‘quites soiez, Por ce que pas nou renoiez. Jamés de ce jor an avant N’an orroiz parler tant ne quant.’  (C2 20951–4) [‘Truly,’ said he, ‘I will absolve you of the deed, since you do not deny it. You will never hear speak of it again from this day.’]

Having renounced any claim to revenge, the knight goes on to provide Perceval with helpful information about the quest on which he is engaged to recover the white stag and hound that were stolen from him earlier on. He even points him in the right direction to find the perpetrators of the crime, though Perceval falls into a daydream and soon finds himself going the wrong way. The purpose of this episode, beyond creating a new link between the hero’s adventures in the Second Continuation and those of the first part of the cycle, appears to be the highlighting of the narrative games open to the author of romance. A vendetta situation, rather than hold the hero up, is made to generate the means for him to reach his goal more quickly; another device (the daydream) is then introduced in order to foil this expectation and delay the arrival of Perceval at his destination.47 In Le Chevalier as deus espees, a similar game is played out during Gauvain’s journey to take his own revenge on Brien. Again and again, he is offered hospitality by strangers who turn out to be his enemy’s relatives, and who are dismayed to learn that he is still alive. First he eats with a pilgrim who asks him, upon learning that he has come from Britain, where Gauvain’s body is resting. It emerges that this is Brien’s father, and his reaction to Gauvain revealing his identity is a classic piece of double-think which recalls Caradoc’s response to learning that Eliavrés is his father in the First Continu-

47 By contrast, the Gerbert Continuation derives a fairly lengthy narrative from the introduction into the text of the Scarlet Knight’s sons, and their desire for revenge on the knight who killed their father.

204 THE

CONTE DU GRAAL CYCLE

ation.48 On the one hand, he expresses disbelief: ‘ne quic a mon ensïent / Ke tu cil Gauvains ne soies’ [I do not believe that you are this Gauvain] (vv. 3674–5); on the other, he admits that ‘or endroit / Vauroie kë ocis fuissiés’ [right now I wish that you had been killed] (vv. 3702–3). That evening, Gauvain stays with a hermit who turns out to be Brien’s uncle. When Gauvain falls asleep, the hermit plots with his manservant to kill his sleeping enemy, but his religious scruples intervene at the crucial moment and he decides not to follow through on his plan. The following night, Gauvain stays at the castle of a cousin of Brien. The lord of the castle is out all night, and learns only the following morning that his wife had failed to ascertain the identity of her guest. He rides after Gauvain and, learning his name, attacks him. He is, of course, soon defeated, but his men, who have arrived on the scene, are keen to pursue Gauvain further. Understanding that the itch for vengeance is only likely to lead to more painful humiliations, and further items on the list of scores to be settled, Brien’s cousin orders his men not to scratch: Et s’aficent mout que cis maus Revenra mout prochainement Sor celui ki si laidement Avoit lor seignor atorné. Et ains k’il s’en soient torné, Li castelains les apiela Et lor dist: ‘Seignour, d’aler la Ne vous mellés ne tant ne quant […] n’i porïés valoir rien Vers lui. Si le laissiés ester.’ (Chevalier as deus espees, vv. 4166–73; vv. 4178–9) [And they all declared that this outrage would promptly be revisited upon he who had left their lord in such a piteous state. But before they could leave to pursue him, the castellan called after them and said: ‘Lords, do not concern yourselves with this matter… you would not be able to stand up to him, so let him be.’]

Gauvain again finds himself the potential subject of revenge attacks in Hunbaut, though this time as a result of his own actions. During the first half of the text, as Hunbaut and Gauvain journey to deliver a message on Arthur’s behalf to a rival king, Gauvain’s inability to follow his friend’s advice threatens to derail their mission. At one point, having beaten a knight who refused to allow him to share his meal, he learns that this same knight’s permission will be required in order for the two companions to take the boat that they will need in order to reach their destination. Hunbaut’s diplomacy is needed

48

See pp. 152–3.

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

205

to soothe the knight’s wounded pride, and the heroes are soon on their way. Hunbaut and Le Chevalier as deus espees both set up revenge scenarios in order to generate drama and raise the possibility of the narrative moving off in a new direction. But these scenarios are soon defused in order to allow the main narrative thread to progress. A similar technique is used at the end of both Le Chevalier as deus espees and L’Atre périlleux, this time in order to stress the fact that the story is coming to a close. In both texts, those responsible for spreading the false rumours of Gauvain’s death are present at court at the end of the tale. When this is known, a desire arises to exact revenge for the trouble they have caused, and it requires all Gauvain’s diplomacy to smooth things over. In the case of Brien, who has lived at court for some time using the nickname ‘biel prison’ [handsome prisoner], the revelation of his identity threatens to turn his friends into enemies: Lors ot il assés anemis K’ot eü ains a amis. Mais mesire Gauvains li prous A force l’acorda a tous (Chevalier as deus espees, vv. 12039–42) [Then he had as many enemies as he had previously had friends. But brave sir Gauvain forced them to be reconciled with him]

In L’Atre périlleux, Arthur’s reaction to hearing Gauvain recount what has happened to him is to evoke the need for vengeance: Mais je voel ains estre vengié De cil qui tant m’a coroucié, Qui as gens se vantoit a tort Qu’il vous avoit ocis et mort De ce voil jou venjance avoir.  (Atre, vv. 6591–5) [But I want to take revenge first on the one who has so upset me by boasting falsely to everyone that he had killed you; I want vengeance on him.]

Luckily for the Orguelleux Faé and Goumeret, Gauvain has already promised to guarantee their safety at court, and their eventual fate is to be married to the damsels they had been courting all along. It is surely not coincidental that it should be Gauvain, the knight who throughout these romances best exemplifies the aesthetics of no-consequence, who insists on the need to forgive and forget. He is at it again in Fergus, when the Noir Chevalier arrives at Arthur’s court as Fergus’ prisoner. Reflecting that the Noir Chevalier has cost him the lives of many of his knights, Arthur asks his nephew to devise a suitable punishment. He advises the king:

206 THE

CONTE DU GRAAL CYCLE

Qu’il pardoinst deboin[ai]rement Au chevalier son mautalent. S’il l’ocio[i]t que li vauroit? Ja por cho ne recoverroit Iceles qui estoient ocis. Les mors as mors, les vis as vis.  (Fergus, vv. 3545–50) [That he should generously forgive the knight for the anger he had caused. What good would it do to kill him? It would not bring back those who had been killed. The dead to the dead, the living to the living.]

The proverbial advice of the final line is, of course, also a citation from the Conte du Graal. There, it summarised everything that was wrong in Perceval’s attitude towards his past – the belief that he could leave it behind as he made his way through the formative stages of knightly development. By contrast, its very presence here is an avowal by Guillaume le Clerc that the narratives of the present can only work within the framework bequeathed by those of the past; that every new literary generation must come to terms with its textual inheritance. At the very moment that Gauvain scotches the vendetta, thus aborting the possibility of more tales involving relatives of the Noir Chevalier, his choice of words invites the audience to supplement the narrative themselves by submitting the romance to intertextual readings. So far, we have discussed cases where vengeance is seen as a threat to centripetal narrative progress and therefore rejected as a plot thread. In La Vengeance Raguidel, however, putting off vengeance is at the heart of the text’s narrative dynamic. We know when Gauvain first leaves court that he has forgotten to take with him the lance tip without which he cannot punish the murderer of the dead knight in the boat: Messire Gavains s’est montés, l’escu au col, sor le destrier et vielt aler celui vengier qui ert venus el car ocis, mais il ne set en quel païs il estoit ne dont il ert nés. Issi est Gavains esgarés qu’il ne set u querre celui qui vengier le doit avuec lui, ki emporta les .V. aniax. Illueques vint uns damoiseax qui sa lance li aporta. Gavains le prent, sil commanda a Deu atant et puis s’en vet. Mais tant se haste que il let le tronçon dont il doit vengier le mort. Ci a grant destorbier,

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

207

car sans le tronçon de la lance n’en prendroit il nul jor venjance. Il s’en vait, si l’a oblïé.  (Vengeance Raguidel, vv. 528–47) [Sir Gauvain mounted his horse, his shield round his neck, with the intention of avenging the murdered knight whose body had arrived on the chariot. But he did not know what land he was from. Gauvain was indeed lost, since he did not know where to seek the man with whom he was supposed to exact revenge, who had taken away the five rings. A young man came to him carrying his lance. Gauvain took it, commended him to God and left. But he went in such a hurry that he left the lance tip with which he was to avenge the dead man. This is a great calamity, for without the lance tip he would never take vengeance. He set off, having forgotten it.]

He does not know whom he is avenging, who he is to kill, nor where he is to go, and he has left behind the vital instrument of vengeance. As SchmolkeHasselmann pithily observes: ‘Gawain in this romance is literally a knight errant.’49 By the time Gauvain does return to court again to fetch the lance tip, he has met and fallen for Ydain, an attachment which will intervene to force postponement once again of the initial task. As Gauvain prepares to set off to fight Druydain for the hand of his amie, Keu’s sarcasm demonstrates a keen appreciation of the mechanics of the text: Or avons plus a faire, or est tot au recommencier! De la venjance au chevalier, ce me samble, est pris li respis. (Vengeance Raguidel, vv. 4364–7) [Now we have much to do; now everything must begin again! It seems to me that the revenge of the knight will have to wait.]

The main quest, to avenge the knight in the boat, is indeed begun three times. Every time the text returns to it, the audience is reminded of this ultimate goal, the adventure which launched the narrative and sated Arthur’s initial desire for marvellous narrative. This quest thus forms the centripetal core of the romance, unifying the text and simultaneously allowing other narratives to co-exist as digressions. The narrative dynamic of La Vengeance Raguidel is thus eminently comparable to that of the Second Continuation, which pulls the same trick of periodically reminding its audience of its ultimate destination. Wolf-Bonvin’s observation on the importance of a long memory for achieving revenge could apply equally well to the function of the principal 49

Schmolke-Hasselmann [Middleton] (1998: 130).

208 THE

CONTE DU GRAAL CYCLE

quests in this text. The intermittent reminders of Perceval’s desire to return to the Grail Castle, reach Mont Dolereus and recover his stolen animals serve to reassure the audience that the text is getting ‘somewhere’, which allows extra material in the form of ‘adventure’ to be generated almost indefinitely. Perhaps the most thorough-going instance of this technique in Arthurian verse romance comes in a section of Les Merveilles de Rigomer which must rank as one of the most literal transpositions of the tension between centripetal and centrifugal narrative dynamics to be found in medieval French literature. Having learned that Lancelot is being held prisoner at the castle of Rigomer in Ireland, Gauvain assembles an impressive roll call of Arthurian knights to form a search party: in total, fifty-one knights are named over forty-two lines (Rigomer, vv. 7059–100), which is virtually the entire search party of fiftyeight. Once they reach Ireland, it becomes apparent that their minds are not entirely focused on the quest to rescue Lancelot. When a knight challenges them to a joust, Gauvain is quick to claim the adventure for himself, but he takes the trouble to reassure his companions that there will be plenty more for them to enjoy: Ceste prumiere aventure Vel jou avoir oltreement; Mais çou ert par tel convenent Que cascuns la soie avera, Si con Dex li envoiera.  (Rigomer, vv. 7260–4) [I want to have this first adventure all to myself; but I make this promise: that each of you will have his own, as long as God sends him one.]

Gauvain’s words are proved correct, as seven of his companions peel off from the main group, one by one, to follow centrifugal adventures. One of these, Engrevain, describes the journey to Rigomer in terms which fail even to mention Lancelot’s captivity, but instead make the whole thing sound like a knights’ holiday: Nos en alons esbanoier Pres de .lx. chevaliers En la terre de Rigomer Pour les mervelles esgarder.  (Rigomer, vv. 8151–4) [We are some forty knights off to enjoy ourselves in the land of Rigomer to see the marvels.]

A similar use of the quest motif to introduce narrative centrifuge is found in the Gauvain interlude of the Second Continuation, which takes up Episodes 29 to 32 of Roach’s edition of the text. This lengthy digression is prompted by a quest to seek Perceval, a task to which many knights commit. Leaving

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

209

court, Gauvain meets up with his companions at a crossroads, and informs them that he intends to find his way back to the court of the Fisher King in order to find out more about the Grail. Already his quest to seek Perceval (last seen heading towards the Mont Dolereus) has become doubled by a quite different concern, specific to him. This admission that the search for Perceval does not exhaust the reasons for leaving court seems to liberate the other knights, as they immediately split up, heading off to seek adventure: ‘Chascuns s’an vet querre avanture / La ou il la cuide trover’ [each of them leaves to seek adventure where he thinks he will find it] (C2 29156–7). The narrator now warns us that only Gauvain’s adventures will be included in this tale: Ne des autres ne dirai ja Quiex avantures lor avindrent. Des voies, des chemins qu’il tindrent, Ne m’orroiz vos avant conter; Mais de Gauvain vos voil parler Ce que l’estoire nos an conte.  (C2 29194–9)50 [Nor will I tell of the others what adventures happened to them. You will not hear me speak more about the paths they took; instead I want to relate to you what the story tells us about Gauvain.]

Clearly, the Second Continuation is willing to countenance centrifuge, but only in controlled doses. William Kibler has coined the term ‘phantom adventures’ to describe the unnarrated exploits of characters in Arthurian romance, which take place away from the manuscript page.51 The term seems apt, as these events are sometimes felt in the effects they produce on the diegesis (as when a knight is found in prison but we are not told what sequence of events led him there) even though they are not seen. Evoking the phantom adventures of other knights is a technique which allows narrators, as in the Second Continuation, to give an impression of narrative abundance while maintaining a focus on the particular narrative thread they wish to pursue. In Le Chevalier as deus espees, Meriadeuc arrives in a forest where Arthur has come to search for him. Near a spring, he notices the signs of recent combat on the ground: fresh blood, hoof prints and fragments of shields (Chevalier as deus espees, vv. 8710–19). As he soon learns, Arthur’s knights have been treating the quest as an opportunity to engage in jousting. However, in an apparent rejection of the generic tradition that the quest for the missing knight should lead to narrative centrifuge, the knights have made a pact: 50 The Manessier Continuation makes use of the ‘quest for Perceval’ motif to generate narratives involving Boort, Lionel and Calogrenant. 51 See Kibler (2000), especially p. 287. Peregrine Rand discusses the same phenomenon in greater depth under the label ‘unnarrated romances’. See Rand (1998, Chapter 3, especially pp. 111–12).

210 THE

CONTE DU GRAAL CYCLE

C’au cief de fois ainsi avoient Convent ke tous les jors iroient Querre aventures, et la nuit As tentes revenroient tuit, Por k’i ne lor fust destorné D’aucun, et isi atorné.  (Chevalier as deus espees, vv. 7853–8) [For in the end they had made a pledge that they would spend every day seeking adventures, and then all return to the tents at night, so that none of them would be lost and thereby deflected from their purpose.]

The phantom adventures of these knights keep criss-crossing the main narratives: Meriadeuc meets a knight who has been defeated by Garehet, wounds Gifflet in single combat and eventually sets out with Gauvain, once their other narrative business has been tied up, to rescue about two hundred knights who (despite their best intentions) have allowed themselves to be captured and imprisoned by Rous du Val Perilleus, an enemy of Arthur. Les Merveilles de Rigomer does not content itself with phantom adventures. The journey to Rigomer, as well as being an excuse for the knights to test their mettle, is also a pretext for the author to recount other narratives, and for the audience to hear them. Remarkably, for a brief time, it appears as if the romance may be about to turn into an Arthurian version of the Thousand and One Nights or the Canterbury Tales, those texts in which a central narrative thread is used as a jumping-off point for a dizzying variety of narratives. More specifically, one can feel the Rigomer author exploring the different avenues of Arthurian narrative: Sagremor rapes a damsel, leaving her pregnant with the child who will one day kill him (the tragic mode); Gauvain is disarmed and made prisoner by an army of maidens and ladies against whom he is unable to defend himself (the comic mode); Waheriés kills a tyrant and spends the night with his opponent’s long-suffering wife (the romantic mode); Yvain and Engrevain rescue damsels in distress (the heroic mode). At the end of each adventure, however, the narrator is careful to remind his audience that the main group is still travelling towards Rigomer, so that the text never moves too far away from the centripetal dynamic. Even within each mini-narrative, the main quest makes unexpected intrusions into the course of events: Cligés meets a knight who was wounded at Rigomer and Waheriés hears from a messenger that his companions have routed a band of Irish knights. Even more importantly, the end of each knight’s adventure sees him carry on alone to Rigomer, so that each centrifugal narrative is reabsorbed back into the centripetal core.52 52 Norris Lacy, while not exactly a fan of this part of the text, acknowledges its meticulous logic: ‘This method, as I suggested, may strike the reader as exceedingly tedious, but tedium does not necessarily render it ineffective. While it takes random adventure

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

211

The only exception to this procedure is Gauvain’s narrative, the first, which is halted when he is captured by his enemies. At the moment of Lancelot’s own imprisonment at Rigomer, an implicit connection was drawn between the loss of memory caused by an enchanted ring he is given to wear and the narrator’s inability to pursue any further the tale of a knight whose body and mind are caught in the stasis of captivity: Quant ou doit fu li aniaus mis, Dont fu Lanselos si sopris, Ne li menbre de nule rien, D’armes porter ne d’autre bien  (Rigomer, vv. 6327–30) [When the ring was placed on his finger, Lancelot was so overcome that he forgot everything about bearing arms or any other good action] Entendu avons et öi De Rigomer dont jo vos di Et de la prison destraignant, Ou tant bon chevalier vaillant Sont devenu malvais et sot. La convient estre Lanselot Desi adonc q’eure venra, Que autre aventure avenra. […] Or le laisons de lui ester, Que n’i poons rien conquester.  (vv. 6403–10; 6413–14) [We have heard and listened to what I have told you about Rigomer and the fearsome prison where so many worthy knights have become weak and foolish. Lancelot will have to remain there until such time as another adventure comes to befall him… Now let us leave him be, for we can get no more from him.]

A knight who was no longer able to bear arms (v. 6330) had little left to offer a chivalric romance. Nothing more, for the time being, could be done with the character, and so the text had to begin again with a new hero, Gauvain: Chi commence nostre romans, Qui mout est riches et vaillans, Des aventures qi avinrent A chiaus qui a Rigomer vinrent. Gavains, li preus, s’en entremist  (Rigomer, vv. 6421–5)

and discursive technique almost to their logical conclusion, it also carefully ties all the adventures to the central direction of the romance’ (1993: 85–6).

212 THE

CONTE DU GRAAL CYCLE

[Here begins our romance, which is rich and worthy, about the adventures that befell those who came to Rigomer. Brave Gauvain took up the challenge]

Now, having left Gauvain temporarily halted in his progress while pursuing the adventures of other knights, the narrator again draws on the vocabulary of memory: Or vos en ai grant ment conté Et ne pourquant ai oblié Mout del millor et del plus bel, Mais par celui qui fist Abel, Or ai talent que me ravoie A ço que oblié avoie; Car sans cestui oubliement Ne poons traire a finement De chou que avons commencié.  (Rigomer, vv. 10583–91) [Now I have related many things to you, and yet I have forgotten a good deal of the best and most interesting material; but, in the name of he who made Abel, I wish to return now to what I had forgotten; for without this digression we cannot bring to a close what we have begun.]

The digressions which the narrator allowed himself in the preceding three thousand lines are characterised in this passage as a kind of forgetting, while the return to the Gauvain narrative is conceived as a ‘ravoiement’, a return to the straight path. Even more interestingly, if ‘cestui oubliement’ is taken to refer to the act of forgetting/digressing, the last three lines can be read as suggesting that digressions are an essential part of the romance’s aesthetic, without which it cannot be completed. In order to become a narrative worthy of the name, a centripetal core requires occasional excursions into centrifuge. The Conte du Graal cycle, because its cyclic aesthetic is predicated on continuity, explores at length the tension between centrifugal and centripetal narrative dynamics. One important way in which it negotiates this tension is in its use of interlace. Narrative structure, narratorial vocabulary and paratextual information all bear witness to the significance of this practice to the coherence of the cycle. Given that the verse romances we have been studying explore the same issues of narrative dynamic as the Conte du Graal corpus, it is not surprising to find that they also make frequent use of the technique; interlace vocabulary is present in virtually all of the texts mentioned in this chapter. In light of these facts, it seems odd that interlace has often been seen essentially as a feature of prose Arthurian romance.53 As Lacy warns, with 53 For instance, Peggy McCracken (2008: 38) defines narrative interlace as ‘the technique of weaving in and out multiple storylines that is typical of prose romance’.

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

213

specific reference to Rigomer: ‘in its use of interlace, the text offers some parallels to the prose cycle, suggesting that we should not be too hasty in assuming that the esthetics of verse and of prose romances of the period must have been discrete.’54 A number of scholars have considered the significance and functions of formulae such as ‘or dist li contes que…’ [now the tale says that…], with which the Vulgate Cycle introduces a change of narrative.55 Yet the form and function of interlace in the verse tradition is studied far less frequently.56 In contrast to the impersonal narrative voice of the ‘conte’ in prose romance, interlace formulae in the Conte du Graal cycle almost always make use of first-person singular or plural forms: ‘D’eus vos lairai ore a itant’ [Now I will leave off telling you about them] (C2 [LR] 19607), ‘Ici vos dirons…’ [Here we will tell you…] (CM 35047), and so on. Even later in the cycle, when the text comes increasingly to be conceived of as a ‘livre’, the performative voice of a personified narrator is preferred as the agent of communication between text and audience. The relation between this narrative voice and the implied second-person audience (‘vos’) places the balance of power on the side of the narratorial je, whose decisions on what he will or will not recount the audience is obliged to accept. This attitude was evident in the first instances of interlace suspension in the Conte du Graal, where the narrator warns his audience that he will be abandoning Perceval for an unspecified period of time, and where the effect was to disrupt narrative continuity in favour of an aesthetic of interruption. It remains in place throughout the cycle, even as the narrative aesthetic shifts from centrifugal interruption to centripetal alternation between central hero and secondary characters. The Conte du Graal cycle thus bears out in this respect the results of Sophie Marnette’s analysis of narratorial presence in verse romance: Compositeur du récit, garant de sa vérité, il joue constamment sur son pouvoir d’auteur. En effet, il ne dit que ce qu’il veut bien dire mais fait aussi allusion à ce qu’il aurait pu raconter.57 [Composer of the tale, guarantor of its truth, he plays constantly on his authorial power. Indeed, he says only what he is willing to say, while alluding to what he might have recounted.]

Lacy (1993: 90, n. 19). See, notably, Kennedy (1986: 156–201); Baumgartner (1987: 175–7); Chase (1994); Marnette (1998: 44–6). 56 See however Bruckner (2009: 59–61), who discusses the implications of firstperson plural pronouns in interlace formulae in the Conte du Graal cycle. 57 Marnette (1998: 96). Marnette’s conclusions are based on linguistic study of Chrétien’s Erec and Yvain, as well as the Lais attributed to Marie de France. Roberta Krueger’s work on the narratorial je in thirteenth-century verse romance reveals a similar approach to narratorial presence. See Krueger (1987). 54 55

214 THE

CONTE DU GRAAL CYCLE

The Conte du Graal cycle’s use of interlace, a mechanism not present in the earlier verse texts on which Marnette draws, provides an ideal mechanism by which the narrator can comment on his technique and draw attention to his power as narrator. The same tendency is immediately apparent when one considers interlace in thirteenth-century verse romances, as the following examples will show: Ci lais du roi Artu ester Car ci avant m’estuet conter Comment mesire Gauvains oirre (Chevalier as deus espees, vv. 3535–7) [I leave off telling about King Arthur here, for from hereon in I must narrate how Sir Gauvain is getting on] Or redison, car il est drois, Con li doi compaignon le font, Ki entrecompaigné se sont (Chevalier as deus espees, vv. 6128–30) [Now let us tell again, for it is right to do so, how the two companions who have joined forces are faring] Au roi et a sa conpaignie Me convient des or reperier  (Atre, vv. 298–301) [I must now return to the king and his retinue] Or me restuet dire conment Cil qui aloit por son non querre En aventure par la terre Puet traire a cief de son afaire.  (Atre, vv. 4894–7) [Now I must tell how he who was travelling the land seeking his name managed to bring the matter to a conclusion.] Or vos vaurai du roi ester Et de Gauvain vaura conter, Qui s’en va querre sa seror

(Hunbaut, vv. 3415–17)

[Now I want to leave off telling you about the king, and wish instead to tell you about Gauvain, who is leaving to seek his sister] Del roi Artu ne des barons Plus avant ne vos conterons Ains revenrai a mon traitié La u je l’ai primes laissié  (Fergus, vv. 3639–42) [We will not tell you anything more about King Arthur or his barons; instead, I will return to my subject at the point I had originally broken off]

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

215

Or le laisons de lui ester, Que n’i poons rien conquester, Ains nos convient prendre conroi De Gavain, le neveu le roi  (Rigomer, vv. 6413–16) [Now let us leave him be, for we can get no more from him; instead it is right that we deal with Gauvain, the king’s nephew] Des .ij. ne vos vel jo plus dire, Si averai parlé premiers Des lvj. chevaliers Qui sont errant vers Rigomer.  (Rigomer, vv. 7984–7) [I do not wish to speak of those two any more; first I will tell of the sixty-one knights who are travelling towards Rigomer.]

The contrast with the interlace found in the Vulgate Cycle is both striking and significant. The impersonal voice of prose romance contributes to its attempt to create the impression of an authoritative, divinely inspired pseudohistory. The game played by verse romance with its audience, embodied in the deployment of a personified narrative voice, is different. Since each verse romance represents a fragment of the Arthurian world, there is no need to gesture towards the universalising historical models that the authors of the Vulgate Cycle use to make and undo their claims to authority.58 Instead, these authorial interventions serve to foreground the mechanics of narration and reception: the audience depends on the narrator’s choices (as signified by verbs such as vel/vaura [want] in the above examples), but the narrator himself makes a show of his dependence on the generic conventions that govern what he can do and what his audience will expect from him (as signified by terms such as m’estuet, me convient [I must], car il est drois [for it is right]). This highlighting of the narrator’s performance combines in thirteenthcentury verse romance with the emergent discourse of text-as-book, leading to a complex tension in the generation of narrative authority. On the one hand, the use of verbs such as ‘conter’ [to tell] and ‘oïr’ [to hear] by a firstperson narrative voice evokes the context of reception as a performance before an audience. On the other hand, the narrator of Hunbaut describes himself as engaged in making ‘.I. livre’ [a book] (v. 31), thereby identifying the narrative voice with that of the author, and opening a gap between the narrative present tense, in which the book is written, and the temporal space in which the text is experienced by its audiences or readers. The narrator of Les Merveilles de Rigomer expresses himself in vigorously performative language (for instance: ‘Entendu avons et öi / De Rigomer dont jo vos di’ [We have heard and listened to what I have told you about Rigomer], Rigomer,

58

See Burns (1985: 35–54) on fictions of authorship and authority in the cycle.

216 THE

CONTE DU GRAAL CYCLE

vv. 6403–4), but he also refers to a manuscript source in unusually material terms: ‘Ce dist la laitre ou parcemin, / Que par un venredi matin / Se dormoient li chevalier’ [The letter on the parchment tells us that one Friday morning the knights were asleep] (vv. 6707–9). These texts, like the Conte du Graal cycle, bear witness to a hybrid stage in the material history of literature. Their concern for issues of reception and influence, their avowed desire to entertain and nourish their audiences with narrative, makes them fertile ground for tracing the complexity of the thirteenth-century literary experience. The Pucele de Gaut Destroit in Hunbaut, seen listening to a courtier read aloud from ‘.I. roumant’, suggests one form that this experience might have taken.59 Aside from the performative nature of the narrator’s voice, one other aspect of these formulae seems especially intriguing; namely, that the language of interlace is present even in texts where there is no real need for it. L’Atre périlleux, for instance, revolves around the character of Gauvain, who collects adventures and travelling companions as he proceeds through the text. The only points at which he leaves the front of the stage are when Keu sets out before him to catch the abductor Escanor and when a number of his travelling companions head back to the knight Codrovain’s castle in order to arm themselves for a skirmish. Sure enough, at these moments the text makes use of the interlace switch formulae cited above, even though the change of character in both cases is only fleeting. It is as if the narrator of L’Atre périlleux is taking any opportunity he can to flex his narratorial muscles and signal his presence. This is, once again, a characteristic we have encountered in the Conte du Graal cycle – the Long Redaction of the First Continuation inserts a number of interlace formulae where there is no actual interlace, in order to strengthen the coherence of this section with the rest of the narrative. It may be hypothesised, then, that the use of such formulae becomes a recognised feature of the genre, so that even those texts whose structure does not make great use of interlace feel compelled to include them, just as every inclusion of a forgetful knight or a damsel in a tent serves to strengthen the links with the general Arthurian intertext to which each individual textfragment belongs.60 To read thirteenth-century Arthurian verse romance in relation to the Conte du Graal cycle, rather than the ‘oeuvres complètes’ of Chrétien de Troyes,

59 Coleman (1996) argues for the category of aurality (defined on p. xi as ‘the reading of books aloud to one or more people’) as a way of escaping the problematic binary of ‘oral/written’, neither side of which offers a satisfactory means of discussing medieval interactions with literary texts. 60 The term ‘texte-fragment’ is a coinage of Paul Zumthor (1977) which captures well the way in which each medieval text exists and signifies within a wider intertextual framework.

REREADING THE EVOLUTION OF ARTHURIAN VERSE ROMANCE

217

reflects more accurately the manuscript transmission of these texts as they have come down to us today. It reveals that the authors of Arthurian verse romance were careful readers of the cycle. On a surface level, the influence of the names, themes and motifs of the cycle is felt to varying degrees in the different romances; in the case of romances such as Fergus and Le Chevalier as deus espees, the reuse of themes from the cycle comes with a sense of filiation, of contestation but also of a kind of continuity. In a general sense, the Conte du Graal cycle established a horizon of expectation against which thirteenth-century authors read Chrétien’s and each others’ works. What I have called its aesthetics of consequence encouraged several of these authors to set their narratives within the framework of its spacetime, though this often has the effect of highlighting the somewhat different aesthetics of no-consequence developed by the shorter romances. In these texts, the completion of an adventure or a narrative leaves the generic Arthurian environment fundamentally unchanged: Keu never learns his lesson and Gauvain never quite makes peace with the world. The open, centrifugal nature of these stories finds exemplification in their manuscript transmission, since they have come down to us collected in a wide variety of codices, of which only one could be termed a ‘verse romance compilation’. By contrast, the material coherence of the Conte du Graal cycle, as demonstrated by the surviving manuscript tradition, helped to ensure its spread and popularity. Yet although they neither seek nor attain the cyclic coherence of the Conte du Graal corpus, the thirteenth-century romances show an appreciation of and willingness to use narrative techniques developed at length in the cycle, such as interlace, the multiplication of heroes, and the mise en scène of the tension between centrifugal and centripetal narrative dynamics. The centrifugal drive of the First Continuation made it a popular source of material from which to build an Arthurian narrative, but it is in the structural techniques mentioned above (interlace, separation of the text into central and peripheral narratives) that the influence of the cycle qua cycle can most keenly be felt.

Conclusion I have seen no book of chivalry that creates a complete tale, a body with all its members intact, so that the middle corresponds to the beginning, and the end to the beginning and the middle; instead they are composed with so many members that the intention seems to be to shape a chimera or a monster rather than to create a well-proportioned body.1

Cervantes’ canon gave his summary judgment on chivalric romance over four hundred years ago, but the distaste shown for the narrative structure of such texts is strikingly similar to the reservations expressed by many modern readers of thirteenth-century Arthurian literature. Indeed, it was not until the second half of the twentieth century that scholars began to develop a taste for, and an appreciation of, the lengthy and often repetitive adventures of Arthurian narratives. Norris Lacy offers the case of John Steinbeck’s protracted attempts to modernise Malory as an example of the aesthetic challenges that medieval romance can pose today.2 Beginning his project with the conviction that most modern readers (and scholars) had failed properly to appreciate Malory, Steinbeck soon found himself making substantial changes to his source material: out went the repetitions, the temporal dislocations and the interlace. Becoming progressively dissatisfied with both the aesthetic and the ethos of Malory’s Arthurian world, Steinbeck eventually ended his book at the point when Lancelot and Guinevere kiss, leaving the rest of his source text unnarrated. If the taste for Arthuriana is as strong today as it has ever been, the medieval romances on which contemporary versions depend remain stubbornly resistant to the aesthetic norms of modern literature. As noted in the Introduction, recent years have seen a marked increase in scholarly enthusiasm for medieval narrative cycles. In the Arthurian domain, most attention has been paid to the corpus of prose romances known as the Vulgate Cycle, whose five constituent texts are found in various combinations in around 150 manuscripts.3 This cycle is thought to have been elaboDon Quixote, p. 412. See Lacy (2009: 132–3). 3 See Alison Stones’s list of manuscripts at http://vrcoll.fa.pitt.edu/stones-www/ LG-web/Arthur-LG-LibraryList.html [last consulted 19 September 2011]. 1 2

CONCLUSION 219

rated c. 1210 to 1230, its period of composition thus overlapping that of the Conte du Graal cycle. It has frequently been noted that the two later Continuations, by Gerbert and Manessier, betray the influence of prose Grail romance, both in the presence of specific episodes found in the prose texts, and in the overtly Christian interpretation placed on the adventures associated with what is now explicitly ‘le Saint Graal’ (CM 32717), rather than ‘un graal’ as it appears in Chrétien’s text (CdG 3220), or ‘le riche Graal’ as in the First Continuation (C1 [LR] 17349).4 It is all too easy in hindsight, in view of the respective numbers of extant manuscripts, to see the verse cycle as an epiphenomenon in relation to its prose counterpart. But, if literary history ultimately decided in favour of the prose version, it does not automatically follow that this preference was enshrined from the moment of its appearance on the scene. On the contrary, it is legitimate to wonder what influence the earlier part of the corpus (CdG-C1-C2) may have had on the development of the Vulgate Cycle; it does, after all, represent some of the earliest cyclical material in the Arthurian tradition, and provides a specific model for Grail romance which subsequent authors were free either to follow or to challenge. Indeed, given the enthusiasm with which both textual traditions were pursued and developed well into the thirteenth century, it is reasonable to posit a period of cohabitation, perhaps competition, during which the cycles were most likely read side by side, and from which the Vulgate Cycle eventually emerged victorious.5 The recent attention paid to prose Arthurian romance is in part a result of a renewed interest in the manuscript context of medieval textual transmission. The contemporary success of the Vulgate Cycle challenges modern scholars to understand what it can tell us about the aesthetic and ideological concerns of medieval audiences, as well as what it has to offer to modern readers. Scholars such as Bruckner and Busby have begun to probe the more modest success of the Conte du Graal cycle in relation to questions of medieval textuality, and the present study is intended as a contribution to the burgeoning critical conversation on these texts. The following pages pursue my argument a little further, tying together the findings of the different chap-

4 See Séguy (2001b: 217); Pickens, Busby and Williams (2006: 27, 34); Bruckner (2009: 15, 56, 179–86). Marx (1963) suggests that Manessier and the author of the Queste del Saint Graal were working from a common source, and did not know each other’s work, but Corley (1986) makes a strong case for the anteriority of the prose episodes, arguing that Manessier’s reworking of the material demonstrates more artistry than he is generally credited with. 5 As well as the evidence of production of prose and verse cycle manuscripts in the same workshops mentioned on p. 91, many fourteenth-century inventories list Conte du Graal cycle manuscripts alongside prose cycle codices. See Busby (2002), who discusses (among other examples) a 1373 inventory of the Louvre library (pp. 647–8) and a 1384–85 inventory of Richard II’s books (pp. 682–3).

220 THE

CONTE DU GRAAL CYCLE

ters and sketching out a rough picture of what this remarkable corpus might have to contribute to a wider account of the evolution of Arthurian romance.

The Poetics of Continuation in the Conte du Graal Cycle The Conte du Graal cycle accumulated in a very specific way, as a series of successive moves in an evolving narrative game. Unlike the sequential cycles which have dominated critical discussion of the phenomenon, it developed as a cycle of continuations, each author picking up the hanging narrative threads of his predecessors and writing forwards from the point at which he found the text. Each new addition altered the nature of the whole, both shaping the responses of subsequent writers and, retroactively, rewriting the textual tradition it had inherited in its own image. Chrétien’s open text became the semence, initially, for a centrifugal collection of narratives which showed no interest in pursuing Perceval’s adventure, finding in each tale told an excuse for jumping on to the next. One wonders how the corpus might have evolved had this aesthetic been pursued. Instead, Perceval’s narrative was revived, now re-interpreted as unfinished business, and thus was plotted the centripetal course followed by the subsequent contributors to the cycle. Alternation of characters now became a technique for digression from the central tale, a process of cyclification mapped out in Chapter 1 of this book. As one reads through the Continuations, it is noticeable that authorship becomes an ever more explicit concern, with those responsible increasingly keen to assert and define their involvement in the textual enterprise. Thus we move from the anonymous (and multi-authored) First Continuation to the Second, whose reference to Wauchier de Denain is now accepted, after some debate, as a genuine authorship-claim. Meanwhile, the later Continuations, known by the names of their authors, demonstrate a patent desire to delimit the scope of what each has written, even though the manuscripts themselves refuse to acknowledge these invitations to signal changes of authorship. The conflict between authorial signature and scribal practice illustrates the principal theme of Chapter 3, that continuation involves negotiating between innovation and fidelity.6 As I argued, recognition and genealogy play a major role both in the diegesis of the corpus and in extra-diegetic concerns with the status of writing, its relation to what has gone before, and the contract between story and audience which allows a narrative to be recognised as such. The intertextual dialogue instigated by the Gerbert and Manessier Continuations with the Vulgate Cycle poses further questions which I will evoke 6 Renate Blumenfeld-Kosinski (1986: 443) identifies a similar tension at work in the Paon Cycle, referring to the ‘mixture of pride and deference’ shown by one of its continuators.

CONCLUSION 221

briefly in the following section. This interaction with the prose tradition may go some way towards explaining a feature which sets the later Continuations apart from their predecessors. Both aim for a kind of closure shunned by the earlier texts of the cycle: Manessier congratulates himself on having ‘finé mon livre’ [completed my book] (CM 42652), while Gerbert asks God to give him the strength ‘que il puist la fin ataindre’ [to reach the end] (CG 7004). It is not clear whether Gerbert’s text ever had an ‘ending’ other than the provisional one it closes on in the two surviving manuscripts, but as Mireille Séguy argues, the whole text exhibits a concern for explication and resolution of ambiguity, where the earlier texts fed on the perpetuation of enigmas.7 Manessier’s ending, meanwhile, introduces into the cycle a motif, the epitaph, which is widely used in the Vulgate; moreover, its very wording appears explicitly modelled on the vocabulary of the Grail quest in the prose cycle: Ci gist Perceval Le Galois, qui du Saint Graal Les aventures acheva. (CM 42635–7)8 [Here lies Perceval the Welshman, who brought to an end the adventures of the Holy Grail.]

At the same time as they display their awareness of the prose tradition, however, both the Manessier and Gerbert Continuations reinforce the cyclic agenda set by the Second Continuation. They maintain the focus on Perceval, and they continue and develop further the narrative techniques deployed by that text: the interlacing of the central tale with digressions involving other knights and the motif of the hero’s return to locations visited earlier in the cycle. In doing so, they strengthen the corpus’s coherence, but they also multiply the opportunities for an ever-deepening network of reflections on the nature of continuation, as Perceval finds himself brought back repeatedly to the landscapes of the narrative past. The study of the cycle’s manuscript tradition in Chapter 2 revealed that thirteenth- and fourteenth-century readers and manuscript makers identified the text as a ‘roman de Perceval’, and interlace as the principal mechanism for structuring the different movements of the narrative, from central to peripheral hero and back. The manuscripts also bear witness to the validation of the corpus as a cyclic whole, with a preference for the combination See Séguy (2001a: 327–41). Other points of contact between this ending and that of the Queste del Saint Graal in particular include the hero’s election to a secular kingship, preceding his ascension to Heaven; the disappearance of the relics of the quest at the time of the hero’s death, never to be seen again on Earth; and the preservation of the record of his adventures in a bookcase at Salisbury on the orders of Arthur. 7 8

222 THE

CONTE DU GRAAL CYCLE

CdG-C1-C2-CM, found in eight manuscripts (including the two which interpolate CG), the Rappoltsteiner Parzifal interpolation and the 1530 French prosification. The stability of this configuration implies the marginalisation of the Elucidation and Bliocadran prologues, whose aesthetics do not sit easily with the principal cyclic dynamic of the corpus; nevertheless, a small number of manuscripts (including the German and prose versions) do contain one or both of these texts, demonstrating a further ambivalence in cyclic codices between maximising coherence and maximising content. Our examination of the Conte du Graal cycle’s manuscript tradition also suggests that it may have a good deal of relevance to discussions of the rise of vernacular literacy, and the attendant changes in the status and function of manuscripts. Reading the text from end to end takes one from Chrétien’s narrator summoning his listeners to hear his composition through to Manessier announcing that his book is finished, and inviting the audience to view the parchment, kept at Salisbury, that guarantees the truth of his tale. The conception of literary production and reception thus moves progressively towards a closer identification of the text with the book. The cyclical manuscripts themselves, which date from the mid thirteenth to the mid fourteenth centuries, testify to an emerging conception of the potential for signification in the material text, with the later codices developing rubrics and illustrations into supplementary narratives. More work in this field would be of use in establishing how the development of vernacular literacy impacted on the composition and transmission of Arthurian texts. It has been a contention throughout this study that the Arthurian romances of Chrétien de Troyes, so often treated as the collective model for the form and content of these later romances, were not on the whole considered a coherent corpus by medieval readers, with the Conte du Graal in particular being read and transmitted as a subsection of the Conte du Graal cycle. Indeed, this separate tradition is apparent right into the sixteenth century, as evidenced by the phenomenon of prosification which Chrétien’s romances underwent at this time: while Erec, Yvain and Cligés were all de-rhymed, there is no prose Conte du Graal; rather, as we have seen, it was prosified as part of the cyclic whole. I therefore agree with David Hult’s conception of the differing manuscript fortunes of Chrétien’s works: ‘In the medieval imaginaire, the Grail story [i.e. the Conte du Graal cycle] formed a book, whereas the other romances seem rather to have been pieces requiring some higher informing principle.’9 Having studied the nature of the corpus in the first three chapters, I argued in Chapter 4 that the Conte du Graal cycle needs to be taken into account in research on the evolution of Arthurian verse romance. Far from being an incidental off-shoot, the cycle appears to have been of significant impor-

9

Hult (1998: 21).

CONCLUSION 223

tance as a lens through which the Arthurian innovations of Chrétien were refracted, developed and modified for thirteenth-century audiences. Analysis of both content and structure of the thirteenth-century verse romances has revealed that their authors were especially interested in dialoguing with the textual tradition constituted by the Conte du Graal and its Continuations. Our cycle therefore occupies an intermediary space, between the non-Grail verse romances on the one hand and the prose Grail romances on the other. In the concluding pages of this study, I would like to offer a few further thoughts on the relation between these three phenomena, and in particular on that between the verse and prose cycles.

Closing Arguments Studies of cycles tend to insist on the desire of such corpuses for totality. Gérard Genette equates the production of ‘continuations en chaîne’ [endless continuations] with an ‘archarnement “cyclique”, c’est-à-dire totalisant’ [a stubborn cyclical drive, that is, a drive for completeness], while more recently Nathalie Koble has evoked ‘[les] exigences de totalisation inhérentes à l’écriture cyclique’ [the totalising demands inherent to cyclical writing].10 The textual and manuscript traditions of the Vulgate Cycle, a mass of additions, amplifications, interpolations and re-combinations, testify eloquently to this impulse, such that it has been described as the ‘classic’ cycle.11 This desire for totality rests on a strong textual dynamic directed towards closure, best exemplified in the final lines of the cycle’s final text, La Mort le roi Artu: Si se test ore atant mestre Gautiers Map de l’Estoire de Lancelot, car bien a tout mené a fin selonc les choses qui en avindrent, et fenist ci son livre si outreement que après ce n’en porroit nus riens conter qui n’en mentist de toutes choses.12 [Now master Gautier Map ceases to tell the Story of Lancelot, for he has brought everything to a conclusion in accordance with the events that occurred, and here finishes his book so completely that afterwards nothing but lies could be told.]

The sense of conclusion in these lines rests upon the claim that it has recounted the true history, the whole history, of the Arthurian realm. Yet, as a number of critics have shown, such claims contain their own undoing: closure can only be provisional, because by designating its current state as complete, the text implies that the previous configuration was lacking some10 11 12

Genette (1982: 244); Koble (2007: 9). Sunderland (2010: 8). Mort Artu, p. 263.

224 THE

CONTE DU GRAAL CYCLE

thing; once this model of completion is set up, there is nothing to prevent it from being applied indefinitely.13 The emphatic, almost excessive finality of the narrator’s language at the end of the Mort Artu reveals the insecurity of the text in the knowledge that the very totalising dynamic which sustains it makes it vulnerable to further completions, further additions, from and for readers whose desire to have the whole story may not have been sated. The Mort Artu itself opens with an admission of the relativism inherent in the distinction between ‘complete’ and ‘incomplete’: Aprés ce que mestres Gautiers Map ot mis en escrit des Aventures del Seint Graal assez soufisanment si com li sembloit, si fu avis au roi Henri son seigneur que ce qu’il avoit fet ne devoit pas soufire, s’il ne ramentevoit la fin de ceus dont il avoit fet devant mention et conment cil morurent dont il avoit amenteües les proesces en son livre; et por ce commença il ceste derriene partie.14 [When Master Gautier Map had put into writing what seemed to him a sufficient account of the Adventures of the Holy Grail, his lord King Henry felt that what he had done was not sufficient, unless he recorded the end of those whom he had mentioned previously, and how those had died whose deeds he had narrated in his book; and for this reason, he began this final part.]

The desire for a more and more complete story leaves every new version of the cycle open to supplementation. At the same time, the overall shape of the cycle follows a model of organic cyclicity – the rise and fall of Arthur’s kingdom, interlaced with the eschatological sweep of the Grail narrative – which offers the reader a doubled sense of closure. The end of the Queste del Saint Graal, which sees the Grail and the world’s best knight leave the world forever, signals the end of aventure in Arthur’s kingdom; in the Mort Artu, left with no external challenges to meet, violence becomes focused inwards, tearing the court apart, and culminating in the simultaneous parricide/filicide of Arthur and Mordred on the plain of Salisbury. Arthurian society, having lost the grace of God at the end of the Queste, now meets its secular end in bloody carnage, as the armies of father and son achieve mutual annihilation. While this model of totalising cyclicity fits the Vulgate Cycle perfectly, it does not appear to reflect so well the aesthetics of the Conte du Graal cycle, where the desire both for closure and for completeness, while present, are far less pressing than in its prose counterpart. While both the Queste del Saint Graal and the Conte du Graal cycle feature what might be termed a Grail quest, the two phenomena are really not commensurable. In the former case, the quest has a necessary character, signalled by the empty seat at the 13 14

See especially Rockwell (1996); Griffin (2005); Sunderland (2010: 63–100). Mort Artu, p. 1.

CONCLUSION 225

Round Table announcing Galaad’s quasi-Christic arrival, and its object is a mysterious object to be desired, found and possessed. In our cycle, the sense of mystery is more diffuse; the hero’s stated desire, to return the Grail Castle, has by the Second Continuation lost its connection to any significant restorative function and takes on a more adventitious quality; even Manessier’s ending, which has Perceval imitate Galaad in ascending to Heaven with the Grail, has little significance beyond confirming once again the exemplary nature of Perceval’s career, and it is hard to see any organic connection between hero, object and environment. To put it another way, Arthur’s knights in the Queste know what they are seeking, but not where to seek it; once the Grail is found by Galaad, the narrative well of adventure runs dry, and the remaining characters are left to play out a deadly endgame in the parochial world of the Mort Artu. By contrast, Perceval in the Conte du Graal cycle knows where he wants to go but without a clear sense of what purpose this will serve. By the time he has returned there, and left again, at the end of the Second Continuation, it is clear that this locus provides no more necessary a conclusion to his wandering than any of the other locations tried out along the way, from family home to Biaurepaire, and one feels as though the pattern of departure and return could play out indefinitely. Bruckner captures well the particularity of this narrative dynamic, writing of ‘the peculiar kind of endlessness that characterizes the Perceval cycle, which appears fragmentary no matter how many continuations are added’.15 Several of the features analysed in this book exemplify this alternative model of cyclicity. A taste for evoking narratives that could, but will not, be told is shared by more or less every text in the corpus, from Chrétien to Gerbert, passing through the first two Continuations and the Elucidation. We are not invited to witness the lifespan of Arthurian society, but rather an assortment of narratives involving a limited number of knights, first and foremost Perceval. Thus, when Tristan appears for a time in the Gerbert Continuation, his famous love affair with Iseut is used to motivate a journey to Marc’s court, but soon forgotten for the more martial pleasures of a tournament. The text shows no desire to pursue Tristan’s story further than the scope of its intersection with the narratives of Gauvain and Perceval. The Conte du Graal cycle thus aspires to tell us a fragment of the story of the Arthurian age, and we are invited to fill in the gaps from our intertextual knowledge of the genre. This impression of narrative fragmentation is enhanced in the earlier texts (CdG-C1) by the insertion of vague or inconsistent gaps in time between different narrative arcs. Meanwhile, even Perceval’s centripetal narrative is relatively open-ended: he keeps leaving and returning to a cluster of key locations, finding new reasons to stay on the move, in a looping movement which takes in a potentially infinite number

15

Bruckner (2009: 24).

226 THE

CONTE DU GRAAL CYCLE

of adventures. Manessier eventually breaks the circuit and gives Perceval an ending fit for a king, and a Grail knight, but even though this end is final, the whole nevertheless feels open. Rosemarie McGerr identifies a distinction made by both modern and medieval theorists between two kinds of closure: that concerning the manner in which a text ends and that concerning its audience’s sense that a satisfactory whole has been completed.16 Framed in this way, I would suggest that the Conte du Graal cycle offers the first kind of closure without the second. Or, more precisely, it finds its unity and coherence in the impulse to digress, to defer narrative progression to the extent that to reach the end becomes less significant than what happens on the way there.17 The Second Continuation’s narrator admits as much in a passage of narrative transition, as Arthur’s knights set out to seek Perceval. Reaching a crossroads, the companions separate, and the narrator warns us that his story has no pretensions to completeness: An l’estoire n’ai pas trové Que messires Yvains devint; Ne de Lancelot, qui bien tint Lou grant chemin o il antra, Ne des autres ne dirai ja Quiex avantures lor avindrent.  (C2 29190–5) [I did not find it told in the story what became of Sir Yvain, nor of Lancelot who kept to the main path down which he had started, nor will I ever tell what adventures befell the others.]

Only Gauvain’s adventures will be narrated, but even this limited narrative arc constitutes a digression from the central plot thread; after all, when we last saw Perceval, he had been on his way ‘Tot droit vers le Mont Dolerox’ [straight to Mont Dolereus] (28395). However, the narrator now explains, going ‘tot droit’ anywhere is not how the aesthetics of this cycle function: Or escoutez avant le conte Qui molt fait bien a escouster; Que por l’estoire consomer Fait l’an lou conte durer tant.  (29200–3) [Now listen to the tale which is very pleasing to hear, for the tale is made to last a long time in order to use up the story.]

See McGerr (1989: 153). This analysis could equally be applied to other verse Arthurian romances, such as Chrétien’s Lancelot or Renaut de Beaujeu’s Le Bel Inconnu. What distinguishes the case of the Conte du Graal cycle is the unprecedented and composite scale on which this aesthetic operates. 16 17

CONCLUSION 227

The assertion is striking, almost paradoxical: in order to ‘consomer’ the story, to bring it to a definitive end, the tale is spun out as long as possible. The pleasure of the ‘conte’ is the purpose of the text, whether this is achieved through a centrifugal structure as in the First Continuation’s Arthurian frieze, or a centripetal one as in the Second Continuation’s accumulation of adventures and quests placed in the way of the droit chemin to the Grail Castle. We noted above that Manessier’s ending reflects the influence of the prose Grail cycles; however, the aesthetic effect of each conclusion is quite different. One important difference consists in the scope of the ending envisaged. In the Mort Artu, as we have seen, the end of the text corresponds with the end of the Arthurian world. Nothing can be added to it without lying, partly because there are no relevant characters left whose histories might be related, nor (after the end of the Grail quest) are there any more adventures to be had. In the Manessier Continuation, Perceval’s epitaph notes that he completed the adventures of the Holy Grail, and the vessel itself is never again seen on Earth after his death. Yet here this appears to entail no consequences for the future of knightly adventure, nor for the stability of Arthur’s realm. The adventures of the Grail are Perceval’s alone. Indeed, rather than a catastrophic rupture, Manessier posits continuity between the legendary Arthurian world and that of his contemporary audiences, embodied by the text sealed in parchment at Salisbury.18 If the Conte du Graal cycle has found its place neither within discussions of Arthurian narrative cycles, nor of Arthurian verse romance, one obvious reason for this is that its form does not correspond wholly to either category of text. It is a cycle, yet unlike other Arthurian cycles, it is composed in verse; it is a verse romance, but vastly longer than anything envisaged by the other thirteenth-century authors of Arthurian verse texts. Yet the question of form is intimately bound up with that of textual aesthetic and, as I hope to have established, the Conte du Graal cycle is also aesthetically intermediary between the two types of Arthurian literature. In some ways it functions like the prose texts. In its extensive use of metaphors of insufficiency to generate ever more narrative, in the dynamic tension between linear progression and digression, and in the proliferation of different manuscript redactions and configurations, it can be compared to prose cycles such as the Vulgate. But in other ways it exhibits features typical of the verse romances: like them, it is comfortable with the provisional nature of closure, relaxed about its status as a fragment of the Arthurian world, and there is little attempt to impose a 18 Bruckner (2009: 194–5) sees this as an opposition between two ways of resolving the tension between romance and history set up in Chrétien’s Conte du Graal: ‘While the Lancelot-Grail takes the path of romance as history with its two endings, the Queste messianic, the Mort apocalyptic, the Perceval Continuations remain on the path of estoire as romance: the messianic overtones of Perceval’s redemptive role do not develop as they do for Galaad into salvation history’ (195).

228 THE

CONTE DU GRAAL CYCLE

sense of historicity or of chronological development on the diegesis. Indeed, as we saw in Chapter 4, what little sense of wider chronology there is in many of the verse romances functions to place the events they recount within the space-time of the Conte du Graal cycle. Our corpus therefore remains irresolvably ambivalent, a foot in each camp, and all the more noteworthy for it. The lines of thought that we have sketched out in these pages leave us, appropriately enough, with new questions to replace the ones with which we began this study. Having demonstrated the centrality of the Conte du Graal corpus to an account of the evolution of French Arthurian romance, I would suggest that the three strands of thirteenth-century Arthurian literature – the verse romances, the prose Grail texts and the Conte du Graal cycle – might profitably be studied together, rather than set apart as separate phenomena. This is not a light undertaking, and the size and complexity of such a task are not to be underestimated. The scholarship done in this direction so far has confined itself to the identification of common motifs or the vexed question of relative chronologies. To begin to build on such work, by progressing to questions of aesthetics and intertextual communication, would allow us to move towards an understanding of the deeper interactions between the constituent parts of the Arthurian multiverse.

Appendix 1 Narrative Summaries Since the events of the Continuations are fairly complex and not widely known, I provide narrative summaries below to help orient the reader. The division of the First, Second and Manessier Continuations into episodes follows William Roach’s editions, which also provide more detailed summaries. For the Gerbert Continuation, no such editorial analysis exists, and so I have compiled my own list of episodes, giving line numbers from the CFMA editions.

First Continuation Branch I: Guiromelant (MR 1–2053; LR 1–5508; SR [L] 1–1070; SR [A] 1–1176) 1. Gauvain’s messenger requests Arthur’s presence at Gauvain’s duel. 2. Arthur and his followers arrive and the king meets his mother and sister, whom he had thought dead, at the Chastel des Merveilles. 3. Guiromelant arrives with his followers for the duel. 4. Gauvain and Guiromelant fight while Clarissant, torn between her brother and lover, begs Arthur in vain to intercede. 5. [ALSPR] Clarissant brings about a reconciliation and is married to Guiromelant with Gauvain’s blessing. 6. [DEGMQTUV] Clarissant’s intervention secures a postponement to the following day. By the next morning, Arthur has married Clarissant and Guiromelant. Furious, Gauvain leaves the court. 7. [EGU only] Gauvain kills a knight who has stolen a damsel’s ivory horn and is given a magic ring as a reward. A dwarf reminds him of his promise to rescue the besieged damsel of Montesclaire. After surviving a revenge attack from followers of the dead knight (the knight’s wounds having bled as soon as he approached the corpse), he defeats a knight sworn to fight all on behalf of a maiden intent on achieving revenge on Greoras, and releases twenty maidens from imprisonment by the couple. 8. [DEGMQTUV only] Gauvain finds himself at the Grail Castle where he is unable to join the pieces of a broken sword. 9. [EGU only] Gauvain rescues the damsel of Montesclaire and claims as his reward the Espee as Estranges Renges. 10. [DEGMQTUV only] On his way to keep the appointment made with Guigambresil at Escavalon, Gauvain meets a knight called Dinasdarés who accuses him of having killed his father.

230 APPENDIX

1

11. At Escavalon, he is ordered to fight both Dinasdarés and Guigambresil, until Arthur arrives and imposes a reconciliation. Branch II: Brun de Branlant (MR 2054–3082; LR 5509–6670; SR [L] 1071–2046; SR [A] 1177–2053)1 1. Arthur and his army lay siege to the castle of the rebel Brun de Branlant. 2. The siege is prolonged by two damsels in the castle who appeal to Yvain not to let them starve. 3. Gauvain is injured during a raid by Brun and forced to rest. 4. Almost recovered, Gauvain sets out into the forest to test his strength. 5. Gauvain meets a damsel, the Pucelle de Lis, who is in love with him despite never having met him. They have sex, and Gauvain departs, promising to return to her. 6. Gauvain is pursued by the damsel’s father, who attacks him and is killed. [MQ only] An unnamed brother suffers the same fate. 7. A brother of the damsel, Bran de Lis, attacks Gauvain, whose wound re-opens. They postpone the battle until the next time they meet. 8. Gauvain returns to camp to recover, and the siege is finally ended by Brun de Branlant’s submission. Branch III: Caradoc (MR 3083–8734; LR 6671–12506; SR [L] 2047–3271; SR [A] 2054–3254) 1. Arthur marries his niece Ysave to King Caradoc of Nantes, but her lover, the magician Eliavrés, dupes him into having sex with three animals in the shape of Ysave for the first three nights after the wedding; on the third night, Eliavrés and Ysave conceive a son. 2. King Caradoc, unaware of the deception, gives the child his own name; when he is old enough, he is sent to Arthur’s court. 3. Some years later, King Arthur announces a Pentecostal feast, on the eve of which Caradoc is knighted. 4. On the feast day, an unknown knight offers to exchange blows with any of the king’s men. Caradoc accepts and beheads the knight, who picks up his head and promises to return in a year to deliver the return blow. 5. A year later, the knight returns, but instead of killing Caradoc he takes him aside and reveals that he, Eliavrés, is the young man’s father. 6. Returning to Nantes, Caradoc has his mother imprisoned in a tower. 7. [DEGMPQTUV only] On his way back to Arthur’s court, Caradoc rescues Guinier from Aalardin, who has just wounded her brother Cador. 8. [DEGMPQTUV only] The three men become friends, and Guinier falls in love with Caradoc.

1

In MSS MQ, this branch is reduced to a short summary which omits Episodes 4–7.

NARRATIVE SUMMARIES

231

9. [DEGMQTUV] Arriving at Arthur’s court, the friends join a tournament at which Caradoc is declared the best knight; marriages are arranged for all three, including that of Caradoc and Guinier. [P] The three friends remain in Arthur’s company; no mention is made of the tournament or the marriages. 10. Her confinement in the tower does not prevent Ysave from seeing Eliavrés, until their son catches him and he is forced to endure the same indignity he had inflicted on King Caradoc after his wedding. 11. To exact revenge, the lovers cause an enchanted serpent to fasten itself to their son’s arm, which will kill him within two years. 12. [DEGMQTV] Hearing of his plight, Arthur sets out to see Caradoc; Guinier and Cador do likewise from Cornwall. Caradoc flees in shame, seeking holy men to do penance for his sin against his parents. [ASLPU] Caradoc departs to do penance for his sin. After a long search, Cador finds him and brings him back to Nantes. 13. [DEGMQTV only] Caradoc’s friends are distraught at his flight. After nearly two years of searching, Cador finds him. 14. Cador learns from Ysave how to cure his friend: Caradoc is to sit in a tub of vinegar, with Guinier sitting in a tub of milk. The serpent leaps at her and is killed by Cador, who cuts off the tip of his sister’s breast. King Caradoc dies, and Caradoc becomes king of Nantes, marrying Guinier. 15. Some time later, Caradoc receives the boss of a magic shield from Aalardin, which restores in gold the tip of Guinier’s breast, the secret of which he tells her to guard. 16. At another of Arthur’s Pentecostal feasts, a stranger presents a drinkinghorn to the king from which only a man whose sweetheart has been faithful can drink. All, including the king, are unsuccessful, except Caradoc. Branch IV: Chastel Orguelleus (MR 8735–12706; LR 12507–16836; SR [L] 3272–6765; SR [A] 3255–6748) 1. Arthur decides to hold a great feast for Pentecost, to atone for his recent inactivity. 2. At the feast, Arthur criticises his knights for forgetting that Giflet has been held prisoner at the Chastel Orguelleus for four years. All agree that he must be rescued. 3. Arthur leads a band of knights in a rescue party. Looking for food, Keu is punished for his rudeness by a knight who then welcomes Gauvain’s courtesy and lodges the party for the night. 4. Gauvain finds an empty castle with the table laid, and returns with the other knights. As they are about to eat, Gauvain notices a shield in the next room and dons his armour. Arthur demands that Gauvain explain his behaviour before he will eat. 5. [ALS] Gauvain tells the story of his adventure with the Pucelle de Lis, and

232 APPENDIX

1

his subsequent unfinished duel with her brother; he has just recognised Bran’s shield and is afraid to be caught unarmed. [DEMPQTUV] As above, except that Gauvain claims to have raped the damsel. 6. Bran bursts into the hall and the two men decide to fight at once. 7. During the fight, a damsel appears with a child of five years: Gauvain’s son. By placing the child between the fighting knights, she convinces Arthur to step in and reconcile them. Bran decides to join the rescue party. 8. The knights arrive at Chastel Orguelleus, where Lucan is awarded the first joust the next day. 9. Lucan wins his first fight, but is then taken prisoner. Bran de Lis is granted the next joust. 10. Inside the castle, Lucan tells Giflet of Arthur’s arrival. 11. Bran wins his fight, but Keu loses his for stepping outside the bounds of the field. Hearing church bells, Bran explains that no jousting will take place from Saturday midday to Monday prime, and advises a hunt the next day. 12. During the hunt Gauvain meets a seated knight who refuses to speak to him and threatens to kill him if he will not let him die; he then meets a damsel looking for this knight. Later, Bran explains that this is the Riche Soudoier, master of Chastel Orguelleus, and his amie. 13. Yvain wins his fight the next day, and the knights learn that the Riche Soudoier will be the next opponent. Gauvain is awarded the fight. 14. Gauvain defeats the Soudoier, but agrees to pretend he has lost to spare the Soudoier’s amie from disappointment. 15. The damsel is deceived and leaves the castle; the Soudoier then surrenders to Arthur, and Lucan and Giflet are freed. 16. On their way home, Arthur’s company learn that Gauvain’s son has been kidnapped. All, except Gauvain, Keu and Giflet, set off to find him, without success. Branch V: Gauvain’s Grail Visit (MR 12707–14118; LR 16837–18374; SR [L] 6766–8310; SR [A] 6749–8306) 1. A strange knight rides past Guenevere one day without greeting her. Keu tries and fails to force him to return; Gauvain is more successful with courtesy, swearing to accomplish the knight’s mission should he be unable to himself. 2. On their way back, the stranger is mysteriously killed by a javelin. Gauvain accuses Keu, takes the knight’s armour and horse, and leaves to fulfil the mission. 3. Gauvain takes refuge from a storm at the Chapel of the Black Hand; later, the horse takes him down a long causeway at the end of which is a castle. 4. Gauvain is greeted at the Grail Castle, where he sees a corpse on a bier with a sword fragment placed on it and the Grail, which serves food at the evening’s meal; later, he sees the bleeding lance standing in a rack. He is

NARRATIVE SUMMARIES

233

unable to join the sword fragment to that which had been carried by the knight he has replaced, but is still allowed to ask about what he has seen. He learns that the lance is that of Longinus. 5. [ALMQU only] The host explains that the Grail was used by Joseph of Arimathea to collect the blood of Christ on the cross. 6. Before he can learn about the knight on the bier, Gauvain falls asleep. The next morning he wakes far from the castle, and determines not to rest until he has found out more about what he has seen at the Grail Castle. 7. In a forest, a damsel orders the young man with whom she is travelling to find out the name of a knight who has passed without speaking to her; the young man accidentally kills the knight. When this happens again, the young man is wounded for not wanting to damage his shield by using it. The damsel takes him to guard a ford. 8. Gauvain comes to the ford and fights its guardian until he discovers that the young man is his son. All three go to Arthur’s court. Branch VI: Guerrehet (MR 14119–15322; LR 18375–19606; SR [L] 8311–9509; SR [A] 8307–9457) 1. Unable to sleep one night, Arthur sees a swan-boat bring a dead knight to shore. There is a lance fragment stuck in the knight’s breast, and a letter in the boat telling the king to have the body placed in the hall of his castle for a year and a day: he who draws out the lance must avenge the knight’s death by striking a blow with the same weapon in the same part of the body, or else suffer the same shame ‘that befell Guerrehet in the orchard’ (C1[LR] 18496). 2. The next morning, Gauvain and the other knights discover the body, and Arthur explains to them the letter’s demands, which they find impossible to achieve as the victim’s and murderer’s identities are both unknown. No one understands the reference to Guerrehet’s shame. 3. We flash back to Guerrehet seeking his brother Gauvain; he comes to a deserted castle, in the garden of which he finds a wounded knight in a tent, who is furious at his intrusion, which will be punished by the Petit Chevalier. 4. The Petit Chevalier defeats Guerrehet and forces him to choose between returning to fight again in a year’s time, or becoming a slave weaver in the service of the wounded knight – he chooses the former. 5. As he leaves, Guerrehet is mocked by the castle’s inhabitants; he returns to Arthur’s court hoping to bury the affair. 6. Guerrehet claims not to understand the reference in the letter, but later he accidentally draws the lance fragment from the body. Keu gets Arthur to make Guerrehet reveal the incident of the garden, and Guerrehet leaves, taking the lance with him. 7. Guerrehet kills the Petit Chevalier, who was on his way to find him; he then kills the Chevalier’s master, striking him through the breast with his lance.

234 APPENDIX

1

A damsel informs him that he has avenged the dead knight and takes him to a nearby castle where he is gratefully received. 8. The next morning, Arthur finds Guerrehet asleep in a swan-boat, and the damsel explains what has happened. She leaves in the boat with the body of the knight, who is revealed to be Brangemuer, son of a mortal and a fairy.

Second Continuation Introductory passage: The text returns to Perceval, who has left his uncle’s hermitage behind. (LR 19607–19653; SR 9457–9529) 1. Perceval defeats the Sire dou Cor at the Castle of the Ivory Horn and hears of the pillar on the Mont Dolereus to which only the world’s best knight can attach his horse. (LR 19654–19936; SR 9530–9855) 2. [EPT only] The Sire dou Cor surrenders to Arthur’s court. (LR 19937– 20006) 3. A damsel attempts to drown Perceval by inviting him to cross a river with her. (LR 20007–20106; SR 9856–9976) 4. Perceval enters the Castle of the Magic Chessboard and falls in love with the damsel there. She tells him he must bring her the head of a white stag before she will accept him, and gives him a white dog with which to hunt it. (LR 20107–20303; SR 9977–10161) 5. Perceval catches the stag, but his dog is stolen by a damsel who tricks him into fighting a knight in black armour. As they fight, the dog and stag’s head are stolen by another knight. (LR 20304–20689; SR 10161–10268 [end of SR]) 6. [EPS only] A hunter gives Perceval lodging for the night. (20690–20772) 7. [EPS only] Perceval avenges the murder of a young man without learning the name of either victim or killer, nor the reason for the deed. (20773– 20908) 8. [EPS only] Perceval meets an old knight, the brother of the Scarlet Knight he killed when he first came to Arthur’s court, who forgives him for the murder. (20909–21080) 9. At a castle, Perceval kills a lion and defeats its owner, Abrioris. (21081– 21481) 10. Abrioris surrenders to Arthur’s court. (21482–21578) 11. Perceval finds a dead knight and later his amie. (21579–21658) 12. Perceval kills a giant and frees a maiden held captive in the giant’s castle. (21659–21955) 13. Perceval defeats the White Knight guarding the Gué Amorous, but refuses to take up the custom of defending it. The White Knight travels to Arthur’s court. (21956–22224) 14. Perceval meets the Biau Desconneü and agrees to be at Arthur’s court by Christmas. (22225–22551)

NARRATIVE SUMMARIES

235

15. Perceval returns to Biau Repaire to spend three days with Blancheflor, promising to return later. (22552–23120) 16. Perceval meets an ugly damsel and her ami, the Biau Mauvais, whom he defeats in combat. (23121–23396) 17. The Biau Mauvais and his amie Rosete travel to Arthur’s court. (23397– 23532) 18. Perceval returns to his mother’s home, where he meets his sister and saves her from being carried off by another knight. (23353–23834) 19. Perceval and his sister journey to see their (paternal) hermit uncle. Perceval leaves his homeland, promising to return. (23835–24221) 20. Perceval spends a night at the Castle of Maidens. (24222–24731) 21. Perceval recovers the hound and stag’s head from the damsel and knight who had stolen them. After defeating the knight, he is told the story of the Black Knight whom he fought earlier. (24732–25432) 22. Perceval meets a damsel who lends him her white mule and a magic ring which will lead him towards the Grail Castle. With them, he crosses the treacherous Glass Bridge. (25433–26193) 23. Perceval meets Briol and crosses an unfinished bridge which can only be crossed by the world’s best knight. (26194–26824) 24. Concealing his identity, Perceval joins in a tournament at Castel Orguelleus against Arthur’s men. He is adjudged to be the best knight there. (26825– 27373) 25. Perceval frees a knight imprisoned in a tomb, who attempts to seal him inside but then releases him and directs him to the Mont Dolereus. He meets the damsel of the white mule, who takes back the mount and the ring. (27374–27600) 26. Perceval returns to the Castle of the Magic Chessboard with the stag’s head, where the damsel tells him the chessboard’s history before joining him in bed. He leaves in the morning, promising to return. (27601–28238) 27. Perceval rescues Bagomedés, who has been hung from a tree by Keu and three other knights returning, driven mad, from the Mont Dolereus. (28239–28408) 28. Bagomedés defeats Keu in judicial combat at Arthur’s court, and the king’s knights all set out to find Perceval. (28409–29208) 29. Gauvain meets the Petit Chevalier and his sister Tanree, whom he seduces. (29209–29953) 30. Gauvain and the Petit Chevalier fight incognito against Arthur’s men in a tournament in the Blanche Lande, winning all the plaudits. Afterwards, Gauvain departs, leaving Tanree to lament his loss. (29954–30507) 31. Gauvain meets the Pensive Knight and rescues his amie from Brun de la Lande, who surrenders to Arthur’s court. (30508–31040) 32. Gauvain meets his son Guinglain and returns with him to Arthur, who is at war with King Carras. (31041–31420) 33. Perceval sees a child in a tree, who directs him to the Mont Dolereus. He

236 APPENDIX

1

ties his horse to the pillar and a maiden greets him as the best knight in the world. The following day, she points out the way to the Grail Castle. (31421–32027) 34. Perceval is caught in a storm and sees a tree covered in lit candles; he then shelters in the Chapel of the Black Hand. (32028–32264) 35. Perceval arrives at the Grail Castle, where he eats with the king. The Grail procession passes before them and the Fisher King explains to him the significance of the child in the tree before inviting him to join the pieces of the broken sword. Perceval succeeds, leaving just a small gap in the blade, and the king announces that he has not yet achieved perfection but offers him all his worldly possessions. (32265–32594)

Manessier Continuation 1. The Fisher King tells Perceval the history of the elements of the Grail procession, and of the sword which had belonged to his brother, murdered by Partinal. Perceval promises to avenge this murder, and the Fisher King then explains the significance of the illuminated tree and the Chapel of the Black Hand. Perceval vows to return there to do battle with the Black Hand. (32595–33183) 2. Perceval meets Sagremor and is wounded in battle with robber knights. (33184–33757) 3. Sagremor pursues and kills the leader of the robber knights. (33758–34080) 4. Sagremor stays at the Castle of Maidens, where he defeats its besieger Tallidés. (34081–34725) 5. Sagremor rescues a damsel from attempted rape, but is wounded in the process. (34726–35050) 6. The sister of Silimac, the knight murdered under Gauvain’s protection in the First Continuation, arrives at Arthur’s court and takes Gauvain away with her. (35051–35299) 7. Gauvain rescues a damsel from a pyre and Dodinel from imprisonment. (35300–35745) 8. Gauvain defeats King Margon, who is besieging Silimac’s sister’s castle, and promises to wound Keu to avenge his murder of Silimac. Margon joins Arthur’s retinue, becoming known as the Roi des Cent Chevaliers. (35746–36363) 9. Gauvain is taken prisoner by Silimac’s niece and learns that Silimac’s sister is called the Sore Pucelle. (36364–36620) 10. Gauvain, incognito, engages Keu in a judicial duel and wounds him. (36621–36916) 11. Gauvain meets Agravain, his brother, and learns that Keu has recovered. (36917–37140) 12. Perceval defeats the Black Hand in the chapel. (37141–37862)

NARRATIVE SUMMARIES

237

13. Perceval is tempted by the Devil disguised as a black horse, then as Blancheflor. (37863–38409) 14. Perceval defeats the knight of Lindesores. (38410–38545) 15. Perceval rescues the amie of Dodinel from a would-be rapist and learns that Blancheflor is besieged by Aridés d’Escavallon. (38546–38922) 16. Perceval visits Tribuët the smith, who mends his sword. (38923–39026) 17. Perceval returns to Biau Repaire, where he defeats Aridés. He refuses to stay as he must be at Arthur’s court for Pentecost. (39027–39359) 18. The three knights recently defeated by Perceval surrender to Arthur and join the Round Table. (39360–39576) 19. Perceval meets the Biau Mauvais and is injured while rescuing two women from a pyre. (39577–39969) 20. Sagremor, Dodinel and other knights return to Arthur’s court for Pentecost, but there is no sign of Perceval. The knights set out to search for him. (39970–40182) 21. Boort abandons his brother Lionel to save a maiden. (40183–40401) 22. Gauvain rescues Lionel, who sets out to exact revenge on Boort for not saving him. (40402–40513) 23. The Devil attempts to fool Boort into thinking his brother is dead. (40514– 40624) 24. Lionel attacks Boort; attempting to stop him, Calogrenant is killed. The brothers are reconciled. (40625–40974) 25. Perceval and the Biau Mauvais fight in a tournament where they take the honours, and Perceval changes his companion’s name to Biau Hardi. (40975–41317) 26. Perceval and Hestor, not recognising each other, fight almost to death; the Grail appears and heals their wounds. (41318–41606) 27. Perceval defeats and beheads Partinal. (41607–41860) 28. Perceval brings Partinal’s head to the Grail Castle and heals the Fisher King. The king reveals that he is Perceval’s uncle and makes him his heir. (41861–42101) 29. Perceval defeats six brothers on his way to Arthur’s court, where he recounts his adventures. Arthur has the deeds of all his knights recorded in writing and sealed in a bookcase in Salisbury. Perceval receives word that the Fisher King has died. (42102–42468) 30. Arthur and his knights accompany Perceval for his coronation at Corbenic, where they witness the Grail procession. Perceval reigns in peace for seven years, then spends ten years as a hermit, sustained by the Grail. At his death, the Grail, lance and carving dish are carried away to Heaven. (42469–42637) Epilogue: Manessier dedicates his finished book to Jeanne, Countess of Flanders, and informs his audience that they can check his source text at Salisbury, where Arthur had the events of the narrative written down and kept. (42638–42668)

238 APPENDIX

1

Gerbert Continuation 1. Perceval is told by the Fisher King that repentance and atonement for his mother’s death are the key to mending the sword properly and learning the Grail secrets. He is woken at midnight by a voice telling him to return home and save his sister. (1–102) 2. Due to his sinfulness, he is unable to gain admittance to the Earthly Paradise, breaking his sword on the door. An old man in white tells him to go to its forger to have it mended, and gives him a sealed letter which provides protection from the Devil. (103–298) 3. He finds the surrounding lands restored to fertility and is gratefully received in a nearby castle by a lady named Escolasse. She shows him the smith Trebuchet’s manor, and he later rebuffs her advances. (299–672) 4. Perceval kills the serpents guarding Trebuchet’s manor and the smith mends the sword, though this will mean his death. (673–901) 5. Perceval frees two damsels hanging from a tree by their tresses, where they have been hung by Agravain and Sagremor, who have lost their minds at the Mont Dolereus. He brings them to their senses with the sealed letter. (902–1158) 6. Perceval leaves the two knights with a friendly castellan. In the forest of Carlion, he meets Arthur with his men, who are hunting the White Stag, and exchanges barbs with Keu. (1159–1332) 7. Perceval sits in the Siege Perilous, freeing six knights who had been swallowed up by the Earth after sitting there. They explain that they have witnessed the torments which await sodomites in Hell. (1333–1612) 8. Perceval’s cousin arrives at court looking for him to force her ami to honour his promise to marry her (he is about to marry another woman the next day). Without revealing his identity, Perceval leaves with her. (1613–1925) 9. Perceval defeats the recalcitrant ami, Faradien, and his cousin is rightfully married. The newlyweds leave to go together to Arthur’s court. (1926– 2481) 10. Perceval prays in a chapel that he might find the lance and Grail. Sleeping under a hawthorn bush, he is woken by a devil disguised as the Fisher King’s daughter on a black mule, who promises to teach him about the Grail and lance if he will lie down with her, but he refuses to be tempted into such sinful activity. (2482–2592) 11. After riding for a week, he recognises the place near his mother’s house where he saw the five knights he took for angels. Reunited with his sister, they visit the hermitage where their mother is buried and Perceval confesses his guilt at her death. He is sermoned by a hermit. The next day, he takes his sister away with him. (2593–2861) 12. Perceval defeats Mordret, who wants to abduct his sister, and sends him to Arthur. He rides on all week until they reach the Castle of Maidens. The lady there reveals that she is his mother’s cousin and that his mother’s

NARRATIVE SUMMARIES

239

name was Philosophine; they brought the Grail over to Britain together, though the people’s sinfulness led to its subsequent removal to the Fisher King’s castle. Perceval leaves his sister at the castle and sets off the next morning. (2862–3247) 13. Mordret arrives to surrender at Arthur’s court, followed by Faradien. (3248–3376) 14. A knight arrives at court and defeats Giflet, Lancelot and Yvain; during an even contest with Gauvain, a minstrel identifies him as Tristan, and Gauvain takes him joyfully off to his lodgings. (3377–3692) 15. Having become great friends with Tristan, Gauvain heads an elite band of Arthur’s knights who go with the Cornish knight to a tournament organised in Yseut’s honour by King Marc against the Roi des Cent Chevaliers, disguised as badly dressed minstrels. (3693–3854) 16. Tristan and his companions are hired as watchmen by Marc. Yseut recognises her lover when he plays the lai de Chievrefueil on a small flute (‘flaguel’). (3855–4110) 17. Tristan and Yseut resume their affair while Marc’s side suffers in the tournament. Gauvain suggests lending them a hand, and so the minstrels (led by Tristan) ride into the fray, turning the tide. Perceval arrives and, seeing the Roi des Cent Chevaliers’ men in trouble, joins in on their side, defeating several of the Round Table knights, including Tristan, before Gauvain and Perceval learn each other’s names. (4111–4726) 18. Marc forgives Tristan and honours Perceval. Leaving Tristan behind, the knights go their separate ways – Perceval onwards, Gauvain to Montesclaire and the others back to Arthur. (4727–4868) 19. Perceval reaches the castle of his mentor Gornemant and his four sons, who are forced every day to fight the same forty knights – killed one day, they are alive and fresh again the next. Gornemant is wounded and Perceval offers to take his place. (4869–5190) 20. Perceval and the four brothers kill the forty knights, but one of the sons is badly wounded. Perceval decides to spend the night on the battlefield to learn how the knights come back to life. (5191–5496) 21. He watches an old giantess revive the dead knights with ointment from two ivory barrels. She reveals that Gornemant is being punished by the Roi de la Gaste Cité, a follower of the Devil, for having knighted him, and that as long as she lives he will learn nothing of the Grail. He beheads her and kills the six knights who have been re-animated, being wounded in the process. He heals himself with the balm and returns to a hero’s welcome at Gornemant’s castle. (5497–6154) 22. Perceval and Gornemant head off with the barrels to Beaurepaire, and Perceval and Blancheflor are married, agreeing to preserve their virginity. The next morning, a voice announces that if they stay chaste, they will have an illustrious lineage: their great-great-grandsons will conquer Jerusalem

240 APPENDIX

1

for Christianity. Perceval appoints Gornemant as his regent and leaves to seek the Grail. (6155–7020) [End of Vol. 1 of Williams ed.] 23. Perceval, forgetting to take the healing balm, spends the night with a hermit. The next day, a knight rides up with an abducted damsel and tells the hermit to marry them. Perceval defeats him and sends him to Arthur. He himself later leaves with the damsel. (7021–7438) 24. They find her ami lying wounded under a tree, and Perceval sends him with the damsel to Beaurepaire to be healed by Blancheflor with the balm. (7439–7526) 25. Perceval meets a damsel bemoaning the treatment her ami is receiving in a nearby town. He rescues him, abolishing a custom whereby passing knights have to fight up to four knights plus their lord in succession and, if they lose, pull a cart through the streets while being pelted with refuse. The next day, he takes the dangerous left-hand path at a crossroads, sending the rescued knight and damsel the other way. (7527–8291) 26. Riding through a deserted land where he sees various knights with severe burn wounds, Perceval then meets two hermits by a wooden cross, one of whom is whipping it while the other worships it. He then follows the Beste Glatissante and sees it eaten by the foals bursting from its body. He arrives at the dwelling of the Roi Hermite, where dinner is interrupted by a damsel who arrives carrying a white shield with a red cross (containing a piece of the True Cross), which can only be taken by the world’s bravest knight, he who would find the Grail and lance. Perceval welcomes the damsel and takes the shield from her. The Roi Hermite (who is Perceval’s uncle) later explains the spiritual significance of the two hermits and the Beste Glatissante, urging Perceval to stay free of sin if he wants to earn Paradise and learn the Grail mysteries. The next day, Perceval and the damsel go their separate ways. (8292–8905) 27. Perceval meets the Pucele as Dras Envers dragging a cart containing her dead ami: he was burnt to death by the shield given to the Dragon Knight by the Devil, the dragon design on which breathes fire on his opponents. The ami died attempting to break the villain’s siege of Montesclaire, the damsel of which he wishes to force to marry him. After spending the night at a nunnery, they journey on and Perceval challenges the Dragon Knight; the sight of his holy shield frightens the demon out of the Dragon Knight’s shield and Perceval kills him, though his shield is taken in the process by an unidentified damsel. The siege is lifted and Perceval rides off after the damsel who took his shield, but instead meets the Pucele as Dras Envers, who intends to bury her ami before beginning life as a hermit. (8906–10192) 28. After a week’s journeying, Perceval stays at an abbey where he sees a priest saying Mass, with the help of an angel, to a bedridden man with weeping sores. He learns that this is Evalac, the former king of Sarras converted to Christianity and baptised Mordrain by Joseph of Arimathea, who came

NARRATIVE SUMMARIES

29.

30.

31.

[End

241

to Britain to free Joseph after its pagan king Crudel had imprisoned him. Punished for wanting to get too close to the Grail, Mordrain has awaited the arrival of the ‘true knight’ seeking the Grail and lance to heal his wounds and allow him to die. Perceval prays to God to help him recover his shield, find the Grail and lance, and free Mordrain from his torment. (10193–10613) He reaches a fortified castle where he sees a chest made of ivory, brought ten years previously by a swan-boat, which when opened will make the four brothers who live there both happy and sad. They have been capturing knights and forcing them to try to open it, including Gauvain, who languishes in their prison. Perceval opens the chest (destined for ‘the world’s best knight’) which contains the head of the brothers’ father – the Scarlet Knight – and a note explaining that the man who opened the chest is his murderer. Perceval captures the child of Leander (lord of the castle and eldest brother), who is forced to agree to fight him in the field – if Perceval wins, he must fight the other brothers; if he loses, he will die. They fight an inconclusive battle, adjourned to the following day, but in the night the brothers are forced to intervene to foil an assassination attempt by a small group of the castle’s residents. The traitors are executed and Perceval given a month to recover. He wins the subsequent rematch and peace is made, with the prisoners freed. Gauvain declares his intention to save the damsel of Montesclaire, but Leander informs him that someone else has already done so. The next day, Gauvain heads to Montesclaire anyway, and Perceval takes the most dangerous path. (10614–12380) Gauvain comes to a tent where he is welcomed by a beautiful damsel who asks him to spend the night naked in her bed, without touching her, and the next morning he can have his way with her; she intends to cut his throat with a knife, but he finds the knife beforehand and rapes her. In his escape (now aided by the damsel), he kills and maims one each of her two brothers and two cousins, along with three of her father’s knights. He rides to a castle where custom requires him to relate the adventures he has encountered – unknown to him, the lord of the castle (Urpin) is the damsel’s father, and to save his life she insists on guarding him overnight, until the safe conduct granted by the custom of hospitality expires. They stage a hostage situation, and Urpin offers Gauvain a private duel in the forest in exchange for sparing his daughter. Gauvain wins and Urpin refuses to ask for mercy; his daughter intervenes once more and his life is spared. Gauvain tells her to come and find him at Arthur’s court. (12381–13956) Gauvain rides away until he reaches a hermitage, where the Pucele as Dras Envers is living, and he learns that Perceval has freed the damsel of Montesclaire. The next day, he meets Arthur’s army on its way to free him from Leander’s prison and tells them of Perceval’s exploits, much to Arthur’s pleasure. (13957–14078) of Vol. 2 of Williams ed.]

242 APPENDIX

1

32. After parting from Gauvain (see ep. 29), Perceval rides for a week before staying with a hermit who warns him not to go further down the road, and hears his confession. (14079–14341) 33. Perceval is tricked by a demon into freeing it from a rock in which Merlin had trapped it, with dire consequences for the local area, but he then tricks it into going back into the rock, and traps it there. (14342–14556) 34. Sleeping by a portrait of a damsel, Perceval is woken by a knight who wants to fight him. When Perceval wins, the knight refuses to beg for mercy, but an unidentified damsel arrives, saves his life, and leaves. The defeated knight explains that his amie was killed on this spot, where he has buried the body and had the portrait placed on top which Perceval had seen. He vowed to kill anyone who passed by until he had avenged her death, but now wishes to be a hermit and atone for his sins. However, when he tries to speak his amie’s name, he dies of heartbreak. Perceval falls asleep, and wakes to find the body gone, replaced by a marble tomb commemorating the dead knight. (14557–14998) 35. Perceval rescues a damsel submerged in a fountain, who has been placed there by her ami for telling him that Perceval was braver and more courtly than he. The ami then arrives and is beheaded by Perceval, who falls asleep in the damsel’s lap. She tries to get a young man riding past to kill him as he sleeps, claiming he has abducted her, but he refuses, and Perceval (who has heard all) gets up and leaves her on her own. (14999–15268) 36. Having eaten with a pilgrim, Perceval meets a damsel who claims to have been abducted, asking him to take her with him; in fact, she is the head of a gang of brigands, who attack him as soon as he has dismounted. He kills them, and the damsel flees. Riding after her, Perceval comes to the robbers’ hideout, where he kills everyone inside (including the damsel) and sets fire to it. (15269–15615) 37. Perceval is received successively in a castle whose inhabitants are grateful for the death of the brigands and in a hermitage where he is criticised for his warlike behaviour. Turning right at a crossroads, he defeats Madieus a le Cote Maltaillie and sends him to Arthur. (15616–16049) 38. After spending the night being honoured in a castle, Perceval meets a giant leading a horse, the brother of the giant killed in the Second Continuation, who is looking for him in order to avenge his brother’s death. Perceval kills him, takes the giant’s horse and rides off. (16050–16415) 39. He is well received by a vavasor, who gives him new weapons and armour. (16416–16556) 40. Madieus reaches the Irish king’s court in Dublin, which Arthur is visiting, and is befriended by Lancelot. After a marvellous feast, Idiers ‘son of Nu’ suggests a quest for Perceval. Forty-two knights take part, each going his separate way. (16557–16718) 41. An attempt by Keu and Gollain li Chaut to take Perceval (whom they have not recognised) prisoner ends with humiliating defeat. (16719–16838)

NARRATIVE SUMMARIES

243

42. A wooden cross shows Perceval the way to the Fisher King’s abode, which he reaches at nightfall. He is welcomed by the king, whom he asks for explanations of several adventures: the body in the chapel, the demon in the rock, the child in the tree. The king says he will tell him after they have eaten. The Grail procession passes before the diners, including the broken sword, which the king invites Perceval to have another go at joining. This time he does so perfectly, and the king declares him ‘dignes… de savoir / de ces affaires tot le voir’ [worthy of knowing the whole truth about these matters]. The sword is taken away, ‘et Perchevax se reconforte’ [and Perceval takes heart] … (16839–17080) [End of Vol. 3, Oswald ed. The manuscripts containing this Continuation now return to the opening of the Manessier Continuation.]

Appendix 2 Lengths and Dates of Texts Elucidation Bliocadran Conte du Graal First Continuation1 Short Redaction Long Redaction Mixed Redaction Second Continuation Gerbert Continuation Manessier Continuation

early 13th century early 13th century 1180–1195

484 lines 800 lines 9212 lines

before 1200 early 13th century 13th century before 1200 1225–1230 1215–1227

9509 [L] or 9456 [ASP] lines 19606 lines 15322 lines 19607–32594 (12987 lines) 14078 lines 32595–42668 (10073 lines)

Notes 1. I have listed the constituent texts of the cycle according to the diegetic order in which they appear (prologues first, Gerbert between the Second and Manessier Continuations), rather than in date order. 2. Dates are taken from the principal edition used for each text (denoted by a star in the Bibliography). 3. The figures given in the lengths column refer to the line numbering used in the relevant editions. Roach’s editions of the Continuations use consecutive numbering as follows: from 1 for the various redactions of C1, then following on from the end of the Long Redaction of C1 for C2, and following on from C2 for CM. Roach edits two versions of the Short Redaction, one based on MS L and the other on ASP; both numberings are given here. CG, edited by Mary Williams, uses an independent numbering scheme. 4. Corin Corley (1987) disagrees with Roach’s placement of the boundary between the First and Second Continuations. For Corley, the Second Continuation does not begin until Roach’s line 20530, and thus spans 12064 lines (this would also extend the Short Redaction of the First Continuation to 10268 lines).

1

For a list of which manuscripts follow which Redaction, see Appendix 3.

Dates are based on those given in Nixon (1993b). Text in this column indicates which Redaction is followed in each manuscript – Short, Long or Mixed – with numerical figures referring to Roach’s division of the text.

1

2

Clermont-Ferrand, BMI 248 London, BL, Add. 36614

Paris, BNF, fr. 794 Bern, Burgerbibliothek 354 Florence, Bibl. Riccardiana 2943 Paris, BNF, fr. 1450 Edinburgh, NLS, Adv. 19. 1. 5 Montpellier, BI, Sect. Méd., H249 Mons, BU 331/206

C

A B

Paris, BNF, fr. 1453

Paris, BNF, fr. 12577

London, College of Arms, Arundel XIV

S

U

H

TOTAL

Paris, BNF, fr. 1429 Paris, BNF, fr. 12576 Paris, BNF, n. a. fr. 6614

Q T V

P

M

15

14th (mid)

14th (2/4)

14th (2/4)

13th (4/4) 13th (4/4) 13th (4/4)

13th (4/4)

13th (4/4)

13th (2/4) 13th (3/4)

13th (2/4)

15

X

X

X

X X X

X

X

X X

X

X X

X

13th (1/4) 13th (2/4) 13th (2/4)

X

11

LR (SR from III/11)

10

2

8

X

X X X X

X X

X

X

X

2

Elucidation & Bliocadran

Bliocadran

CM Prologues

X

X X X

LR MR MR

CG

X

X

SR (LR in part of III)

SR

X

X

X

X

C2

LR

SR of I LR

SR

SR

CdG C12

13th (1/4)

Date1 12

R E

F

L

Shelfmark

MS

Appendix 3 Manuscript of the Conte du Graal Cycle (excluding fragments)

Appendix 4 Full contents of Conte du Graal Cycle Manuscripts MS A (Paris, BNF, fr. 794) – 13th c. (2/4) Chrétien de Troyes, Erec Chrétien de Troyes, Lancelot Chrétien de Troyes, Cligés Chrétien de Troyes, Yvain Athis et Prophilias Benoît de Sainte-Maure, Le Roman de Troie Wace, Le Roman de Brut Calendre, Les Empereurs de Rome Chrétien de Troyes, Conte du Graal First Continuation (SR) Second Continuation MS B (Bern, Burgerbibliothek, 354) – 13th c. (2/4) Seventy-four brief works in verse (including Le Chevalier à l’épée, La Mule sans frein, Cort Mantel and the Folie Tristan; some didactic works; a large number of fabliaux). For a full list, see Hagen (1875: 338–45) Le Roman des Sept Sages de Rome (incomplete) Chrétien de Troyes, Conte du Graal MS C (Clermont-Ferrand, BMI, 248) – 13th c. (1/4) Chrétien de Troyes, Conte du Graal (ends at CdG 9212) MS D [German translation] (Karlsruhe, Landesbibl., Cod. Donaueschingen 97; and Rome, Bibl. Casanatense, A. I. 19 [copy]) – c. 1331–1336 Wolfram von Eschenbach, Parzival, Books 1–2 Elucidation Wolfram von Eschenbach, Parzival, Books 2–14 First Continuation (MR) Second Continuation Manessier Continuation Wolfram von Eschenbach, Parzival, Books 15–16 MS E (Edinburgh, NLS, Adv. 19. 1. 5) – 13th c. (3/4) Chrétien de Troyes, Conte du Graal (opening folios missing – begins CdG 5493) First Continuation (LR) Second Continuation Manessier Continuation (final folios missing – ends CM 40772)



FULL CONTENTS OF CONTE DU GRAAL CYCLE MANUSCRIPTS

247

MS F (Florence, Biblioteca Riccardiana, 2943) – 13th c. (2/4) Chrétien de Troyes, Conte du Graal (final folios missing – ends CdG 8608) MS G [Prosification] (Black letter edition for Galiot du Pré) – 1530 Elucidation Bliocadran Chrétien de Troyes, Conte du Graal First Continuation (LR) Second Continuation Manessier Continuation MS H (London, College of Arms, Arundel XIV) – 14th c. (mid) Wace, Le Roman de Brut Gaimar, L’Estoire des Engleis Le Lai d’Haveloc Pierre de Langtoft, Chronique La Lignee des Bretons et des Engleis Chrétien de Troyes, Conte du Graal (added) Walter de Henley, Hosebonderie (added) ‘Art d’aimer’ (allegorical poem on love) MS K (Bern, Burgerbibliothek, 113) – 13th c. (mid) A variety of texts, including the Second Continuation (with a unique ending) and Durmart le Galois. See Appendix 6 for a full list of contents. MS L (London, BL, Add. 36614) – 13th c. (1/4) Chrétien de Troyes, Conte du Graal (CdG 1–68) Bliocadran Chrétien de Troyes, Conte du Graal (CdG 69 onwards) First Continuation (SR) Second Continuation La Vie de Sainte Marie l’Egyptienne MS M (Montpellier, BI, Sect. Méd., H 249) – 13th c. (4/4) Chrétien de Troyes, Conte du Graal First Continuation (LR) Second Continuation Manessier Continuation (added) ‘Salut d’Amour’ (lyric poem) MS P (Mons, BU, 331/206) – 13th c. (4/4) Elucidation (followed by CdG 61–69) Bliocadran Chrétien de Troyes, Conte du Graal (CdG 69 onwards)

248 APPENDIX

4

First Continuation (SR – LR in part of III) Second Continuation Manessier Continuation MS Q (Paris, BNF, fr. 1429) – 13th c. (4/4) Chrétien de Troyes, Conte du Graal (opening folio missing – begins CdG 117) First Continuation (LR) Second Continuation Manessier Continuation (ends CM 42490 – folios may be missing) MS R (Paris, BNF, fr. 1450) – 13th c. (2/4) Benoît de Sainte-Maure, Le Roman de Troie Le Roman d’Enéas Wace, Le Roman de Brut (vv. 1–9798) Chrétien de Troyes, Erec Chrétien de Troyes, Conte du Graal First Continuation (to the end of Branch I, Episode 5, SR) Chrétien de Troyes, Cligés Chrétien de Troyes, Yvain Chrétien de Troyes, Lancelot Wace, Le Roman de Brut (vv. 9799–end) Dolopathos (incomplete) MS S (Paris, BNF, fr. 1453) – 14th c. (2/4) Chrétien de Troyes, Conte du Graal (opening folio missing – begins CdG 103) First Continuation (SR) Second Continuation Manessier Continuation (final folio missing – ends CM 42539) MS T (Paris, BNF, fr. 12576) – 13th c. (4/4) Chrétien de Troyes, Conte du Graal First Continuation (MR) Second Continuation Gerbert Continuation Manessier Continuation (added) La Mort du Comte du Henau (added) note of debts Renclus de Moiliens, Miserere Renclus de Moiliens, Le Roman de Carité MS U (Paris, BNF, fr. 12577) – 14th c. (2/4) Chrétien de Troyes, Conte du Graal First Continuation (LR – SR from III/11) Second Continuation Manessier Continuation



FULL CONTENTS OF CONTE DU GRAAL CYCLE MANUSCRIPTS

249

MS V (Paris, BNF, n. a. fr. 6614) – 13th c. (4/4) Chrétien de Troyes, Conte du Graal First Continuation (MR) Second Continuation Gerbert Continuation Manessier Continuation This manuscript has many folios mutilated or missing throughout, and ends at CM 35073. See the Introduction to Busby’s edition of the Conte du Graal (1993: xxxiii–xxxv) for further information about the missing passages.

Appendix 5 Arthurian Verse Romances: Dates and Manuscripts Dates are taken from those given by Kelly and contributors (2006) in Chapter X (‘Arthurian Verse Romance in the Twelfth and Thirteenth Centuries’) of Glyn S. Burgess and Karen Pratt eds., The Arthur of the French. Manuscripts containing fragments of texts have not been included. L’Atre périlleux

mid 13th BNF, fr. 2168 BNF, fr. 1433

Chantilly, Musée Condé 472

Beaudous

c. 1250

BNF, fr. 24301

Le Bel Inconnu

1191–1225

Chantilly, Musée Condé 472

Le Chevalier a l’épée

late 12th/early 13th Bern, Burgerbibliothek, 354

Le Chevalier as deus espees first third 13th

BNF, fr. 12603

Durmart le Galois

Bern, Burgerbibliothek, 113

c. 1220

early 13th Fergus

Chantilly, Musée Condé 472 BNF, fr. 1553

Hunbaut

Chantilly, Musée Condé 472

1250–1275

c.1220–1230 Jaufre

BNF, fr. 2164 BNF, fr. 12571

Vatican City, Bibl. Apost. Meraugis de Portlesguez 1200–1215   Vaticana, Reg. Lat. 1725 Vienna, Õsterr. Nationalbibl.,   2599 Turin, Bibl. Naz., L. IV. 33 Les Merveilles de Rigomer first half 13th

Chantilly, Musée Condé 472

first third 13th Chantilly, Musée Condé 472 La Vengeance Raguidel Nottm. Uni Lib., Middleton  LM.6

Appendix 6 Contents of Arthurian Verse Romance Manuscripts Bern, Burgerbibliothek, 113 Garin le Loherain and Gerbert de Metz Second Conte du Graal Continuation (with a unique ending) Pierre de Beauvais, Les Olimpiades Biography of Robert Guiscard Ernoul, Chronique d’Outre Mer Description of Jerusalem Letter of Prester John Prose lapidary Histoire de la Franchise de France Extract of Jacques de Vitry, Historia Orientalis (translated into French) Genealogy of the kings of France Alart de Cambrai, Le Livre de Philosophie et de Moralité Huon de Saint Quentin, La Complainte Jérusalem contre Rome Helinand de Froidmont, Vers de la Mort (fragment) Henri d’Andeli, La Bataille des Vins Prose version of La Riote du Monde (incomplete) Le Doctrinal (moral treatise) Verse sermon on the Last Judgment Partonopeus de Blois Durmart le Galois Two fragments of Perlesvaus Fragments of a decasyllabic translation of Maccabees Bern, Burgerbibliothek, 354 Seventy-four brief works in verse (including Raoul de Houdenc’s Le Songe d’Enfer, Le Chevalier à l’épée, La Mule sans frein, Le Lai du Mantel and the Folie Tristan; some didactic works; a large number of fabliaux). For a full list, see Hagen (1875: 338–345). Le Roman des Sept Sages de Rome (incomplete) Chrétien de Troyes, Conte du Graal Chantilly, Musée Condé, 472 Jehan, Les Merveilles de Rigomer L’Atre périlleux Chrétien de Troyes, Erec Guillaume le Clerc, Fergus Hunbaut

252 APPENDIX

6

Renaut de Beaujeu, Le Bel Inconnu Raoul de Houdenc, La Vengeance Raguidel Chrétien de Troyes, Yvain Chrétien de Troyes, Lancelot Perlesvaus Several branches of the Roman de Renart Nottingham, University Library, Middleton LM 6 Benoît de Sainte-Maure, Le Roman de Troie Gautier d’Arras, Ille et Galeron Heldris de Cornualle, Le Roman de Silence Lambert li Tors and Alexandre de Bernay, Le Fuerre de Gadres La Chanson d’Aspremont Raoul de Houdenc, La Vengeance Raguidel Various fabliaux (including several by Gautier le Leu) Paris, BNF, fr. 1433 L’Atre périlleux Chrétien de Troyes, Yvain Paris, BNF, fr. 1553 Fifty-two works in verse and prose, including Benoît de Sainte-Maure’s Roman de Troie, Gui de Cambrai’s Barlaam et Josaphat, Gerbert de Montreuil’s Roman de la violette, a number of saints’ lives and religious pieces, Guillaume le Clerc’s Fergus, and a series of lais and fabliaux. For a full description, see Lepage (1975). Paris, BNF, fr. 2164 Jaufre Paris, BNF, fr. 2168 L’Atre périlleux Li Vilains de Farbu Five lais: Yonec, Guigemar, Lanval, Narcise, Graalent Aucassin et Nicolete Four fabliaux Gautier de Metz, L’Image du monde Le Vie Carlemaine, si com il ala en Espaigne Marie de France, Fables Short verse texts, including fabliaux and a bestiary Paris, BNF, fr. 12571 Jaufre



CONTENTS OF ARTHURIAN VERSE ROMANCE MANUSCRIPTS

253

Paris, BNF, fr. 12603 Thirty-one texts, including: Le Chevalier as deus espees Chrétien de Troyes, Yvain Le Roman d’Eneas Wace, Le Roman de Brut Adenet le Roi, Les enfances Ogier Fierabras Short verse texts, including fabliaux, Jean Renart’s Le Lai de l’ombre, Raoul de Houdenc’s La Voie d’enfer and Huon de Cambrai’s La Male honte Marie de France, Fables La femme qui conquie son baron (fabliau) Paris, BNF, fr. 24301 Vie des Pères Chronicle of French kings from Clovis to Louis VII Passion Christi Herbert, Dolopathos Robert de Blois, Beaudous Robert de Blois, religious poems Turin, L. IV. 33 Le Roman de Troie (in prose) La Chronique du Pseudo-Turpin Le Clerc de Voudroy, Le Dit des drois Chronique d’Outremer D’une aventure du roi Artu (extract from Jehan, Les Merveilles de Rigomer) Melion Gliglois Raoul de Houdenc, Meraugis de Portlesguez Le Roman de Thèbes (in prose) Vatican City, Biblioteca Apostolica Vaticana, Reg. lat. 1725 Chrétien de Troyes, Lancelot Chrétien de Troyes, Yvain Jean Renart, Le Roman de la Rose [Guillaume de Dole] Raoul de Houdenc, Meraugis de Portlesguez Vienna, Österreich Nationalbibliothek, 2599 Raoul de Houdenc, Meraugis de Portlesguez

BIBLIOGRAPHY Primary texts Where multiple editions of a text are referenced, an asterisk after an entry indicates that this is the principal edition used for citation. Atre (1936). L’Atre périlleux, ed. Brian Woledge. Paris: Champion. Beaudous (1889). Robert von Blois: Sämmtliche Werke, ed. Jakob Ulrich, vol. 1. Berlin: Mayer and Müller. Bel Inconnu (1992). Renaut de Bâgé, Le Bel Inconnu, ed. Karen Fresco. New York: Garland. Bliocadran (1976). Bliocadran: A Prologue to the Perceval of Chrétien de Troyes: Edition and Critical Study, ed. Lenora D. Wolfgang. Tübingen: Niemeyer. Chevalier à l’épée (1900). Le Chevalier à l’Épée: An Old French Poem, ed. E. C. Armstrong. Baltimore: J. Murphy Co. Chevalier à l’épée (1972). Two Old French Gauvain Romances: Le Chevalier à l’épée and La Mule sans frein, ed. D. D. R. Owen and R. C. Johnston. Edinburgh and London: Scottish Academic Press.* Chevalier as deus espees (2006). Le Chevalier as deus espees, ed. and trans. Paul Vincent Rockwell. Cambridge: D. S. Brewer. Chrétien de Troyes (1994). Chrétien de Troyes: Romans, La Pochothèque. Paris: Livre de Poche. Conte du Graal (1993). Chrétien de Troyes, Le Roman de Perceval ou Le Conte du Graal, ed. Keith Busby. Tübingen: Niemeyer.* Don Quixote (2005). Miguel de Cervantes, Don Quixote, trans. Edith Grossman. London: Vintage. Durmart (1965–66). Durmart le Galois, ed. Joseph Gildea. Pasadena: Villanova University Press. 2 vols. Elucidation (1931). The Elucidation: A Prologue to the Conte del Graal, ed. Albert Wilder Thompson. New York: Institute of French Studies. Erec (1994). Chrétien de Troyes, Erec et Enide, ed. Jean-Marie Fritz, in Chrétien de Troyes (1994). Fergus (1983). Guillaume le Clerc, The Romance of Fergus, ed. Wilson Frescoln. Philadelphia: William H. Allen. First Continuation (1993). Première Continuation de Perceval (Continuation-

256 BIBLIOGRAPHY Gauvain), ed. William Roach, trans. Colette-Anne Van Coolput-Storms. Paris: Livre de Poche. Florimont (1932). Aimon de Varennes, Florimont, ein altfranzösischer Abenteuer-­ roman, ed. Alfons Hilka. Göttingen: Gedruckt für die Gesellschaft für roman­ ische Literatur. Gerbert Continuation (1922–78). Gerbert de Montreuil, La Continuation de Perceval, ed. Mary Williams (vols. 1 and 2) and Marguerite Oswald (vol. 3). Paris: Champion. 3 vols. Hunbaut (1914). Hunbaut: Altfranzösischer Artusroman des XIII. Jahrhunderts, ed. W. Foerster, J. Stürzinger and H. Breuer. Dresden: Gedruckt für die Gesellschaft für romanische Literatur. Jaufre (2006). Jaufre, ed. Charmaine Lee. Rome: Carocci. Meraugis (1897). Raoul von Houdenc: sämmtlichte Werke, ed. M. Friedwagner, vol. 1. Halle: Niemeyer. 2 vols. Mort Artu (1996). La Mort le roi Artu, ed. Jean Frappier. Geneva: Droz. 3rd edition. Mouskés (1836–38). Chronique rimée de Philippe Mouskés, ed. the Baron of Reiffenberg. Brussels: Hayez. 2 vols. Perlesvaus (1932–37). Le haut livre du graal: Perlesvaus, ed. William Albert Nitze and T. Atkinson Jenkins. Chicago: University of Chicago Press. 2 vols. Potvin (1866–71). Perceval le Gallois: le poème de Chrestien de Troyes et de ses continuateurs d’après le manuscrit de Mons, ed. Charles Potvin. Mons: Dequesne-Masquillier. 6 vols. Prose Lancelot (1980). Lancelot do Lac: The Non-Cyclic Old French Prose Romance, ed. Elspeth Kennedy. Oxford: Clarendon. 2 vols. Queste (1980). La Queste del Saint Graal, ed. Albert Pauphilet. Paris: Champion. 2nd edition. Raoul de Cambrai (1992). Raoul de Cambrai, ed. Sarah Kay. Oxford: Clarendon. Rigomer (1908–15). Les Merveilles de Rigomer, ed. W. Foerster and H. Breuer. Dresden: Gedruckt für die Gesellschaft für romanische Literatur. 2 vols. Roach (1949–83). The Continuations of the Old French ‘Perceval’ of Chrétien de Troyes, ed. William Roach. Philadelphia: University of Pennsylvania Press (I– II), American Philosophical Society (III–V). 5 vols. I: The First Continuation, Mixed Redaction of mss. TV [1949]; II (with R. H. Ivy, Jr): The First Continuation, Long Redaction of mss. EMQU [1950]; III, 1: The First Continuation, Short Redaction of mss. ALPRS [1952]; III, 2: Glossary to the First Continuation by Lucien Foulet [1955]; IV: The Second Continuation [1971]; V: The Third Continuation by Manessier [1983].* Roman de Rou (1970). Wace, Le roman de Rou, ed. A. J. Holden. Paris: A. & J. Picard & Cie. Roman de Troie (1904–12). Benoit de Sainte-Maure, Le Roman de Troie, ed. L. Constans. Paris: Firmin-Didot. 6 vols. Vengeance Raguidel (2004). Raoul de Houdenc, La Vengeance Raguidel, ed. Gilles Roussineau. Geneva: Droz.

BIBLIOGRAPHY 257

Yvain (1994). Chrétien de Troyes, Le Chevalier au Lion (Yvain), ed. David F. Hult, in Chrétien de Troyes (1994).

Secondary works Atanassov, Stoyan (2000). L’Idole inconnue: le personnage de Gauvain dans quelques romans du XIIIe siècle. Medievalia 31. Orléans: Paradigme. Balsamo, Gian (1993). ‘Son, Knight, and Lover: Perceval’s Dilemma at the Castle of Beaurepaire’, Exemplaria 5: 263–81. Baumgartner, Emmanuèle (1978). ‘Le roman aux XIIe et XIIIe siècles dans la littérature occitane’, in Grundriß der romanischen Literaturen des Mittelalters, IV, 1, 627–44. Heidelberg: Carl Winter Universitätsverlag. —— (1987). ‘Les techniques narratives dans les romans en prose’, in Lacy, Kelly and Busby eds., The Legacy of Chrétien de Troyes, I, 167–90. —— (1994a). De l’histoire de Troie au livre du Graal: le temps, le récit (XIIe– XIIIe siècles). Orléans: Paradigme. —— (1994b). ‘Vers, prose et fiction narrative’, in Karen Pratt ed., Shifts and Transpositions in Medieval Narrative, 1–9. Cambridge: D. S. Brewer. —— (1999). Chrétien de Troyes: Le Conte du Graal. Paris: PUF. —— (2005). ‘Chrétien’s Medieval Influence: From the Grail Quest to the Joy of the Court’, in Lacy and Grimbert eds., A Companion to Chrétien de Troyes, 214–27. Berthelot, Anne (1996).‘The Atypical Grails, or the Ravages of Intertextuality in the Thirteenth Century’, in Norris J. Lacy ed., Text and Intertext in Medieval Arthurian Literature, 209–18. New York: Garland. Besamusca, Bart (2003). The Book of Lancelot: The Middle Dutch ‘Lancelot’ Compilation and the Medieval Tradition of Narrative Cycles, trans. Thea Summerfield. Cambridge: D. S. Brewer. Besamusca, Bart, Willem P. Gerritsen, Corry Hogetoorn and Orlanda S. H. Lie, eds. (1994). Cyclification: The Development of Narrative Cycles in the Chansons de Geste and the Arthurian Romances. Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen Verhandelingen. Bloch, R. Howard (1983). Etymologies and Genealogies: A Literary Anthropology of the French Middle Ages. Chicago: University of Chicago Press. Blumenfeld-Kosinski, Renate (1986). ‘The Poetics of Continuation in the Old French Paon Cycle’, Romance Philology 39: 437–47. Brandsma, Frank (1995). ‘Opening Up the Narrative: The Insertion of New Episodes in Arthurian Cycles’, Queeste 1: 31–9. Brodman, Marian Masiuk (1991). ‘The Livre de Caradoc’s Chastity Test’, Neophilologische Mitteilungen 92: 471–84. Brownlee, Kevin and Marina Scordilis Brownlee, eds. (1985). Romance: Generic Transformation from Chrétien de Troyes to Cervantes. Hanover and London: University Press of New England. Bruckner, Matilda Tomaryn (1980). Narrative Invention in Twelfth-Century

258 BIBLIOGRAPHY French Romance: The Convention of Hospitality. Lexington: French Forum Publishers. —— (1987). ‘Intertextuality’, in Lacy, Kelly and Busby eds., The Legacy of Chrétien de Troyes, I, 224–65. —— (1993). ‘The Poetics of Continuation in Medieval French Romance – From Chrétien’s Conte du Graal to the Perceval Continuations’, French Forum 18: 133–49. —— (2000). ‘Looping the Loop Through a Tale of Beginnings, Middles, and Ends: From Chrétien to Gerbert in the Perceval Continuations’, in Busby and Jones eds., ‘Por le soie amisté’, 33–51. —— (2006). ‘Authorial Relays: Continuing Chrétien’s Conte du Graal’, in Virginie Greene ed., The Medieval Author in Medieval French Literature, 13–28. New York and Basingstoke: Palgrave Macmillan. —— (2009). Chrétien Continued: A Study of the Conte du Graal and its Verse Continuations. New York: Oxford University Press. Burgess, Glyn S. and Karen Pratt, eds. (2006). The Arthur of the French: The Arthurian Legend in Medieval French and Occitan Literature. Cardiff: University of Wales Press. Burgwinkle, William (2004). Sodomy, Masculinity, and Law in Medieval Literature: France and England, 1050–1230. Cambridge: Cambridge University Press. Burns, E. Jane (1985). Arthurian Fictions: Rereading the Vulgate Cycle. Columbus: Ohio State University Press. —— (1987/88). ‘Quest and Questioning in the Conte du graal’, Romance Philology 41: 251–66. Busby, Keith (1980). Gauvain in Old French Literature. Amsterdam: Rodopi. —— (1993a). Chrétien de Troyes: Perceval (Le Conte du Graal). London: Grant and Cutler. —— (1993b). ‘The Scribe of Manuscripts T and V of Chrétien’s Perceval’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 49–65. —— (1993c). ‘The Text of Chrétien’s Perceval in MS London, College of Arms, Arundel XIV’, in Ian Short ed., Anglo-Norman Anniversary Essays, 75–85. London: Anglo-Norman Text Society. —— (1993d). ‘Text, Miniature, and Rubric in the Continuations of Chrétien’s Perceval’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 365–76. —— (1994a). ‘Hunbaut and the Art of Medieval Arthurian Romance’, in Keith Busby and Norris J. Lacy eds., Conjunctures: Medieval Studies in Honor of Douglas Kelly, 49–68. Amsterdam: Rodopi. —— (1994b). ‘The Other Grail Cycle’, in Besamusca, Gerritsen, Hogetoorn and Lie eds., Cyclification, 176–8. —— (2002). Codex and Context: Reading Old French Verse Narrative in Manuscript. Amsterdam: Rodopi. 2 vols.

BIBLIOGRAPHY 259

—— (2005). ‘The Manuscripts of Chrétien’s Romances’, in Lacy and Grimbert eds., A Companion to Chrétien de Troyes, 64–75. Busby, Keith and Catherine M. Jones, eds. (2000). ‘Por le soie amisté’: Essays in Honor of Norris J. Lacy. Amsterdam: Rodopi. Busby, Keith, Terry Nixon, Alison Stones and Lori Walters, eds. (1993). Les Manuscrits de Chrétien de Troyes / The Manuscripts of Chrétien de Troyes. Amsterdam: Rodopi. 2 vols. Carruthers, Mary (1990). The Book of Memory: A Study of Memory in Medieval Culture. Cambridge: Cambridge University Press. Cave, Terence (1988). Recognitions: A Study in Poetics. Oxford: Clarendon. Cazelles, Brigitte (1996). The Unholy Grail: A Social Reading of Chrétien de Troyes’s Conte du Graal. Stanford: Stanford University Press. Chase, Carol (1994). ‘“Or dist li contes”: Narrative Interventions and the Implied Audience in the Estoire del Saint Graal’, in William W. Kibler ed., The Lancelot-Grail Cycle: Text and Transformations, 117–38. Austin: University of Texas Press. Clanchy, Michael (1979). From Memory to Written Record: England, 1066– 1307. Oxford: Blackwell. Cohen, Gustave (1931). Chrétien de Troyes et son œuvre. Paris: Boivin. Coleman, Joyce (1996). Public Reading and the Reading Public in Late Medieval England and France. Cambridge: Cambridge University Press. Combes, Annie (2004). ‘Nouer les fils de la réécriture: Une Visite interpolée au château du Graal’, Cahiers de Civilisation Médiévale 47: 3–16. Combes, Annie and Annie Bertin (2001). Écritures du Graal. Paris: PUF. Corley, Corin (1982). ‘Réflexions sur les deux premières continuations de Perceval’, Romania 103: 235–58. —— (1984). ‘Wauchier de Denain et la deuxième continuation de Perceval’, Romania 105: 351–9. —— (1986). ‘Manessier’s Continuation of Perceval and the Prose Lancelot Cycle’, Modern Language Review 81: 574–91. —— (1987). The Second Continuation of the Old French Perceval: A Critical and Lexicographical Study. London: MHRA. Deleuze, Gilles and Felix Guattari (1980). Mille Plateaux. Paris: Éditions de Minuit. Doner, Janet R. (1996). ‘Scribal Whim and Miniature Allocation in the Illustrated Manuscripts of the Continuation-Gauvain’, Medium Aevum 65: 73–95. Douchet, Sébastien (2007). ‘Généalogie et filiation littéraire dans la Continuation de Wauchier de Denain’, in Christine Ferlampin-Acher and Denis Hüe eds., Lignes et lignages dans la littérature arthurienne. Actes du 3e colloque arthurien organisé à l’université de Haute-Bretagne 13–14 octobre 2005, 231–44. Rennes: Presses Universitaires de Rennes. Dragonetti, Roger (1980). La Vie de la lettre au Moyen Age: le Conte du graal. Paris: Le Seuil.

260 BIBLIOGRAPHY Duggan, Joseph J. (2001). The Romances of Chrétien de Troyes. New Haven and London: Yale University Press. Fleischmann, Suzanne (1990). ‘Philology, Linguistics, and the Discourse of the Medieval Text’, Speculum 65: 19–37. Frappier, Jean (1972). Chrétien de Troyes et le mythe du Graal: Étude sur Perceval ou le Conte du Graal. Paris: Société d’édition d’enseignement supérieur. —— (1973). Amour Courtois et Table Ronde. Geneva: Droz. Gaggero, Massimiliano (2008a). ‘La genealogia di Perceval: interferenza tra materie e prospettiva ciclica nella Quarta Continuazione del Conte du Graal di Gerbert de Montreuil’, Medioevo Romanzo 32: 262–88. —— (2008b). ‘Mise en texte e riscrittura nelle Continuazioni del Conte du Graal di Chrétien de Troyes’, Moderna 10.2: 61–82. Gallais, Pierre (1964). ‘Formules de conteur et interventions d’auteur dans les manuscrits de la Continuation-Gauvain’, Romania 85: 181–229. —— (1988–89) L’Imaginaire d’un romancier français de la fin du XIIe siècle: description raisonnée, comparée et commentée de la Continuation-Gauvain, première suite du Conte du Graal de Chrétien de Troyes. Amsterdam: Rodopi. 4 vols. Gaunt, Simon and Ruth Harvey (2006). ‘The Arthurian Tradition in Occitan Literature’, in Burgess and Pratt eds., The Arthur of the French, 528–45. Gaunt, Simon and Sarah Kay, eds. (2008). The Cambridge Companion to Medieval French Literature. Cambridge: Cambridge University Press. Genette, Gérard (1982). Palimpsestes: La littérature au second degré. Paris: Seuil. Gilbert, Jane (2008). ‘The Chanson de Roland’, in Gaunt and Kay eds., The Cambridge Companion to Medieval French Literature, 21–34. Godzich, Wlad and Jeffrey Kittay (1987). The Emergence of Prose: An Essay in Prosaics. Minneapolis and London: University of Minnesota Press. Gomez, Etienne (forthcoming). ‘Les effets de cycle dans le cycle du Conte du graal’, in Danièle James-Raoul ed., Les Genres en question au Moyen Âge. Grant, Marshal S. (1986). ‘The Question of Integrity in the First Continuation of Chrétien de Troyes’ Conte du Graal’, Proceedings of the PMR Conference 11: 101–25. Greene, Virginie (2006). ‘What Happened to Medievalists after the Death of the Author?’, in Virginie Greene ed., The Medieval Author in Medieval French Literature, 205–27. New York and Basingstoke: Palgrave Macmillan. Griffin, Miranda (2005). The Object and the Cause in the Vulgate Cycle. London: Legenda. Grigsby, John L. (1988). ‘Heroes and their Destinies in the Continuations of Chrétien’s Perceval’, in Lacy, Kelly and Busby eds., The Legacy of Chrétien de Troyes, II, 41–53. Haidu, Peter (1968). Aesthetic Distance in Chrétien de Troyes: Irony and Comedy in Cligés and Perceval. Geneva: Droz. Hindman, Sandra (1994). Sealed in Parchment: Rereadings of Knighthood in

BIBLIOGRAPHY 261

the Illuminated Manuscripts of Chrétien de Troyes. Chicago: University of Chicago Press. Hinton, Thomas (2009). ‘The Aesthetics of Communication: Sterility and Fertility in the Conte del Graal Cycle’, Arthurian Literature 26: 91–108. —— (2011). ‘New Beginnings and False Dawns: A Reappraisal of the Elucidation Prologue to the Conte del Graal Cycle’, Medium Aevum 80: 41–55. Huchet, Jean-Charles (1980). ‘Mereceval’, Littérature 40: 69–94. —— (1989). ‘Le roman à nu: Jaufre’, Littérature 74: 91–9. Hult, David F. (1984). ‘Closed Quotations: The Speaking Voice in the Roman de la Rose’, Yale French Studies 67: 248–69. —— (1986). Self-fulfilling Prophecies. Readership and Authority in the First Roman de la Rose. Cambridge: Cambridge University Press. —— (1998). Manuscript Transmission, Reception and Canon Formation: The Case of Chrétien de Troyes. Berkeley: University of California Press. Huot, Sylvia (1987). From Song to Book: The Poetics of Writing in Old French Lyric and Lyrical Narrative Poetry. Ithaca and London: Cornell University Press. —— (2010). Dreams of Lovers and Lies of Poets: Poetry, Knowledge, and Desire in the Roman de la Rose. London: Legenda. Jewers, Caroline (2000). Chivalric Fiction and the History of the Novel. Gainsville: University Press of Florida. Kay, Sarah (1997). ‘Who Was Chrétien de Troyes?’, Arthurian Literature 15: 1–35. Kelly, Douglas ed. (1996). The Medieval Opus: Imitation, Rewriting, and Transmission in the French Tradition. Amsterdam: Rodopi. —— (2002). Chrétien de Troyes: An Analytic Bibliography. Supplement 1. London: Tamesis. Kelly, Douglas and contributors (2006). ‘Arthurian Verse Romance in the Twelfth and Thirteenth Centuries’, in Burgess and Pratt eds., The Arthur of the French, 393–459. Kennedy, Elspeth (1986). Lancelot and the Grail: A Study of the Prose Lancelot. Oxford: Oxford University Press. Kibler, William W. (2000). ‘Sagremor in the Arthurian Verse Romances’, in Busby and Jones eds., ‘Por le soie amisté’, 283–92. Kleiman, Irit Ruth (2008). ‘X Marks the Spot: The Place of the Father in Chrétien de Troyes’s Conte du Graal’, Modern Language Review 103: 969–82. Koble, Nathalie (2007). ‘Introduction. L’impossible épuisement des cycles arthuriens’, in Nathalie Koble ed., Jeunesse et genèse du royaume arthurien. Les Suites romanesques du Merlin en prose, 7–14. Orléans: Paradigme. Krueger, Roberta L. (1987). ‘The Author’s Voice: Narrators, Audiences and the Problem of Interpretation’, in Lacy, Kelly and Busby eds., The Legacy of Chrétien de Troyes, I, 115–40. Lacy, Norris J. (1993). ‘Les Merveilles de Rigomer and the Esthetics of “PostChrétien” Romance’, Arthurian Yearbook 3: 77–90.

262 BIBLIOGRAPHY —— (1996). ‘Motif Transfer in Arthurian Romance’, in Kelly ed., The Medieval Opus, 157–68. —— (2001). ‘Jealousy, Fidelity, and Form in the Livre de Caradoc’, in Joan Tasker Grimbert and Carol J. Chase eds., Philologies Old and New: Essays in Honor of Peter Florian Dembowski, 281–9. Princeton: Edward C. Armstrong Monographs. —— (2006). ‘L’Atre Perilleux and the Erasure of Identity’, Arthurian Literature 23: 109–16. —— (2007). ‘The Elucidation: Introduction’, The Camelot Project at the University of Rochester. URL: http://www.library.rochester.edu/camelot/elucidation. htm. Last consulted 17 September 2011. —— (2009). ‘Arthurian Texts in their Historical and Social Context’, Arthurian Literature 26: 131–48. Lacy, Norris J. and Joan Tasker Grimbert, eds. (2005). A Companion to Chrétien de Troyes. Cambridge: D. S. Brewer. Lacy, Norris J., Douglas Kelly and Keith Busby, eds. (1987–88). The Legacy of Chrétien de Troyes. Amsterdam: Rodopi. 2 vols. Le Rider, Paule (1978). Le Chevalier dans le Conte du Graal de Chrétien de Troyes. Paris: Société d’Edition d’Enseignement Supérieur. Leupin, Alexandre (1979). ‘Les enfants de la mimésis: différence et répétition dans la Première continuation du Perceval’, Vox Romanica 38: 110–26. —— (1982). ‘La Faille et l’écriture dans les Continuations du Perceval’, Le Moyen-Age 88: 253–69. Lot, Ferdinand (1918). Étude sur le Lancelot en prose. Paris: Champion. —— (1931). ‘Les auteurs du Conte du Graal’, Romania 57: 117–36. Maddox, Donald (1994a). ‘Lévi-Strauss in Camelot: Interrupted Communication in Arthurian Feudal Fictions’, in Martin B. Schichtman and James P. Carley eds., Culture and the King: The Social Implications of the Arthurian Legend, 35–53. Albany: State University of New York Press. —— (1994b). ‘Notes Toward a More Comprehensive Approach to Medieval Literary Cycles’, in Besamusca, Gerritsen, Hogetoorn and Lie eds., Cyclification, 102–7. —— (1996). ‘Cyclicity, Transtextual Coherence, and the Romances of Chrétien de Troyes’, in Sturm-Maddox and Maddox eds., Transtextualities, 39–52. —— (2000). Fictions of Identity in Medieval France. Cambridge: Cambridge University Press. Marnette, Sophie (1998). Narrateur et points de vue dans la littérature française médiévale. Bern: Peter Lang. Marx, Jean (1963). ‘Étude sur les rapports de la 3e continuation du Conte du Graal de Chretien de Troyes avec le cycle du Lancelot en prose en général et la Queste del Saint Graal en particulier’, Romania 84: 451–77. McCracken, Peggy (1998). The Romance of Adultery: Queenship and Sexual Transgression in Old French Literature. Philadelphia: University of Pennsylvania Press.

BIBLIOGRAPHY 263

—— (2008). ‘The Old French Vulgate Cycle’, in Gaunt and Kay eds., The Cambridge Companion to Medieval French Literature, 35–47. McGerr, Rosemarie P. (1989). ‘Medieval Concepts of Literary Closure: Theory and Practice’, Exemplaria 1: 149–79. Méla, Charles (1977). ‘Perceval’, Yale French Studies 55/56: 253–79. —— (1984). La Reine et le Graal. Paris: Le Seuil. Micha, Alexandre (1939). La Tradition manuscrite des romans de Chrétien de Troyes. Paris: Champion; rpt. 1966, Geneva: Droz. —— (1951). ‘Le Perceval de Chrétien de Troyes, roman éducatif’, in René Nelli ed., Lumière du Graal, 122–38. Paris: Les Cahiers du Sud. Middleton, Roger (1993a). ‘Coloured Capitals in the Manuscripts of Erec et Enide’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 149–93. —— (1993b). ‘Additional Notes on the History of Selected Manuscripts’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, II, 177–243. Morris, Rosemary (1988). ‘Aspects of Time and Place in the French Arthurian Verse Romances’, French Studies 42: 257–77. Nixon, Terry (1993a). ‘Romance Collections and the Manuscripts of Chrétien de Troyes’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 17–25. —— (1993b). ‘Catalogue of Manuscripts’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, II, 1–85. Nykrog, Per (1973). ‘Two Creators of Narrative Form in Twelfth-Century France: Gautier d’Arras and Chrétien de Troyes’, Speculum 48: 258–76. Owen, D. D. R. (1971). ‘Two More Romances by Chrétien de Troyes?’, Romania 92: 246–60. —— (1987/88). ‘The Craft of Fergus: Supplementary Notes’, French Studies Bulletin 25: 1–4. Paris, Gaston (1888). ‘Romans en vers du cycle de la Table Ronde’, in Histoire Littéraire de la France 30: 1–270. Paris: Imprimerie Nationale. —— (1900). Review of Chevalier à l’Épée (1900, ed. Armstrong), Romania 29: 593–600. Pelan, Margaret M. (1931). L’Influence du Brut de Wace sur les romanciers français de son temps. Paris: Droz. Perret, Michèle (1982). ‘De l’espace Romanesque à la matérialité du livre. L’espace énonciatif des premiers romans en prose’, Poétique 50: 173–82. Pickens, Rupert T. (1977). The Welsh Knight: Paradoxicality in Chrétien’s Conte del Graal. Lexington, KY: French Forum. —— (2005). ‘Le Conte du Graal: Chrétien’s Unfinished Last Romance’, in Lacy and Grimbert eds., A Companion to Chrétien de Troyes, 169–87. Pickens, Rupert T., Keith Busby and Andrea M. L. Williams (2006). ‘Perceval and the Grail: The Continuations, Robert de Boron and Perlesvaus’, in Burgess and Pratt eds., The Arthur of the French, 215–73.

264 BIBLIOGRAPHY Pickford, Cedric Edward (1960). L’Evolution du roman arthurien en prose vers la fin du Moyen Age d’après le manuscript 112 du fonds français de la Bibliothèque Nationale. Paris: Nizet. Poirion, Daniel (1973). ‘L’ombre mythique de Perceval dans le Conte du Graal’, Cahiers de civilisation médiévale 13: 191–8. Ramm, Ben (2007). A Discourse for the Holy Grail in Old French Romance. Cambridge: D. S. Brewer. Rand, Peregrine John (1998). ‘Narrative Closure in a Thirteenth Century French Manuscript: Chantilly, Musée Condé 472’, Doctoral thesis, Cambridge University. Reis, Levilson C. (2010). ‘From Aural Reception to Visual Paratext: The Reader in the Manuscripts of Chrétien de Troyes’s Romances’, Neophilologus 94: 377–89. Ribard, Jacques (1976a). Chrétien de Troyes: Le Conte du graal. Paris: Hatier. —— (1976b). ‘Les romans de Chrétien de Troyes sont-ils allégoriques?’, Cahiers de l’Association internationale des Etudes françaises 28: 7–20. —— (1984). ‘Un personnage paradoxal: le Gauvain du Conte du graal’, in Lancelot, Yvain, et Gauvain: Colloque arthurien belge de Wégimont, 5–18. Paris: Nizet. Rider, Jeff (1998). ‘The Perpetual Enigma of Chrétien’s Grail Episode’, Arthuriana 8: 6–21. Rieger, Angelica (1993). ‘Le programme iconographique du Perceval montpelliérain, BI, Sect. Méd. H 249 (M)’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 377–435. Roach, William (1956). ‘Les Continuations du Conte du Graal’, in Les romans du Graal dans la littérature des XIIe et XIIIe siècle, 107–17. Paris: Centre national de la recherche scientifique. —— (1966). ‘Transformations of the Grail Theme in the First Two Continuations of the Old French Perceval’, Proceedings of the American Philosophical Society 110: 160–4. Rockwell, Paul Vincent (1996). ‘“Je ne suiz mie soffisanz”: Insufficiency and Cyclicity in the Lancelot-Grail Cycle’, in Sturm-Maddox and Maddox eds., Transtextualities, 71–91. —— (2000). ‘Appelation contrôlée: Motif Transfer and the Adaptation of Names in the Chevalier as deus espees’, in Busby and Jones eds., ‘Por le soie amisté’, 435–52. Rossi, Marguerite (1980). ‘Sur l’épisode de Caradoc de la Continuation Gauvain’, in Mélanges de langue et littérature françaises du Moyen Age et de la Renaissance offerts à monsieur Charles Foulon, II [=Marche Romane 30.3/4], 247–54. Rennes: Université de Haute-Bretagne. 2 vols. Sargent-Baur, Barbara (2000). La Destre et la senestre. Amsterdam: Rodopi. Schmolke-Hasselmann, Beate [Middleton] (1998). The Evolution of Arthurian Romance: The Verse Tradition from Chrétien to Froissart, trans. Margaret and Roger Middleton. Cambridge: Cambridge University Press.

BIBLIOGRAPHY 265

Séguy, Mireille (2001a). Les Romans du Graal ou le signe imaginé. Paris: Champion. —— 2001b). ‘Le Sceau brisé: L’Impossible fin de la Troisième continuation du Conte du Graal’, Bien dire et bien aprandre 19: 213–24. Singleton, Charles (1949). An Essay on the Vita Nuova. Cambridge, Mass: Harvard University Press. Skårup, Povl (1994). ‘Un cycle de traductions: Karlamagnús saga’, in Besamusca, Gerritsen, Hogetoorn and Lie eds., Cyclification, 74–81. Spiegel, Gabrielle M. (1993). Romancing the Past: The Rise of Vernacular Prose Historiography in Thirteenth-Century France. Berkeley: University of California Press. Staines, David (1994). ‘The Misreading of a Trope’, in Besamusca, Gerritsen, Hogetoorn and Lie eds., Cyclification, 108–10. —— (1996). ‘The Medieval Cycle: Mapping a Trope’, in Sturm-Maddox and Maddox eds., Transtextualities, 15–37. Stephens, Louise D. (1996). ‘Gerbert and Manessier: The Case for a Connection’, Arthurian Literature 14: 53–68. Stock, Brian (1983). The Implications of Literacy: Written Language and Models of Interpretation in the Eleventh and Twelfth Centuries. Princeton: Princeton University Press. Stones, Alison (1977). ‘The Earliest Illustrated Prose Lancelot Manuscript?’, Reading Medieval Studies 3: 3–44. —— (1993a). ‘General Introduction’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 1–8. —— (1993b). ‘The Illustrated Chrétien Manuscripts and their Artistic Context’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, 227–322. —— (no date). ‘Lancelot-Grail Intranet Site: List of Manuscripts by Library’. URL: http://vrcoll.fa.pitt.edu/stones-www/LG-web/Arthur-LG-LibraryList. html. Last consulted 19 September 2011. Sturm-Maddox, Sara (1984). ‘“Tenir sa terre en pais”: Social Order in the Brut and the Conte del Graal’, Studies in Philology 81: 28–41. —— (1996). ‘Food for Heroes: The Intertextual Legacy of the Conte del Graal’, in Norris J. Lacy ed., Text and Intertext in Medieval Arthurian Literature, 117–31. New York: Garland. Sturm-Maddox, Sara and Donald Maddox (1996). ‘Cyclicity and Medieval Literary Cycles’, in Sturm-Maddox and Maddox eds., Transtextualities, 1–14. Sturm-Maddox, Sara and Donald Maddox, eds. (1996). Transtextualities: Of Cycles and Cyclicity in Medieval French Literature. Binghamton, N.Y.: Medieval and Renaissance Texts and Studies. Sunderland, Luke (2008). ‘Le Cycle de Renart: From the Enfances to the Jugement in a Cyclical Roman de Renart Manuscript’, French Studies 62: 1–12. —— (2010). Old French Narrative Cycles: Heroism between Ethics and Morality. Cambridge: D. S. Brewer.

266 BIBLIOGRAPHY Szkilnik, Michelle (1988). ‘Les deux pères de Caradoc’, Bulletin Bibliographique de la Société Internationale Arthurienne 40: 268–86. —— (2007). ‘Le chevalier “oublieux” dans le roman arthurien en vers’, in Patrizia Romagnoli and Barbara Wahlen eds., Figures de l’oubli (IVe–XVIe siècle) [=Études de lettres 1–2], 77–97. Lausanne: Université de Lausanne. Taylor, Jane H. M. (1994). ‘Order from Accident: Cyclic Consciousness at the End of the Middle Ages’, in Besamusca, Gerritsen, Hogetoorn and Lie eds., Cyclification, 59–73. Tether, Leah (2009). ‘(Dis-)continuing Gauvain’s Lance Quest in the First Continuation of the Old French Perceval’, Reading Medieval Studies 35: 99–113. —— (forthcoming). The Continuations of Chrétien’s Perceval: Content and Construction. Cambridge: D. S. Brewer. Thompson, Albert Wilder (1955). Review of William Roach ed., The Continuations of the Old French ‘Perceval’ of Chrétien de Troyes, Speculum 30: 124–9. —— (1959). ‘Additions to Chrétien’s Perceval – Prologues and Continuations’, in R. S. Loomis ed., Arthurian Literature in the Middle Ages, 206–17. Oxford: Clarendon. Tobler, Adolf and Ernest Lommatzsch (1915–2002). Altfranzösiches Wörtebuch. Berlin and Stuttgart. 11 vols. Van Coolput, Colette-Anne (1987). ‘Appendice: Références, adaptations et emprunts directs’, in Lacy, Kelly and Busby eds., The Legacy of Chrétien de Troyes, I, 333–42. Van Mulken, Margot (1993). ‘Perceval and Stemmata’, in Busby, Nixon, Stones and Walters eds., Les Manuscrits de Chrétien de Troyes, I, 41–8. Vial, Guy (1978). ‘L’auteur de la deuxième continuation du Conte du Graal’, in Mélanges d’études romanes du Moyen Âge et de la Renaissance offerts à Monsieur Jean Rychner [=Travaux de linguistique et de littérature 16], 519–30. Strasbourg: Centre de Philologie et de littératures romanes. Vinaver, Eugene (1971). The Rise of Romance. Oxford: Clarendon. Walters, Lori J. (1985). ‘Le rôle du scribe dans l’organisation des manuscrits de Chrétien de Troyes’, Romania 106: 303–25. —— (1991). ‘The Creation of a Super-Romance: Paris, Bibliothèque Nationale, fonds français, MS 1433’, Arthurian Yearbook 1: 3–25. —— (1994). ‘The Formation of a Gauvain Cycle in Chantilly MS 472’, Neophilologus 78: 29–43. —— (2006). ‘Manuscript Compilations of Verse Romances’, in Burgess and Pratt eds., The Arthur of the French, 461–87. Wilmotte, Maurice (1930). Le Roman du Gral d’après les versions les plus anciennes. Paris: La Renaissance du Livre. Woledge, Brian and H. P. Clive (1964). Répertoire des plus anciens textes en prose française: depuis 842 jusqu’aux premières années du XIIIe siècle. Geneva: Droz. Wolf-Bonvin, Romaine (2007). ‘Voir tue. Scènes de l’oubli dans la Vengeance

BIBLIOGRAPHY 267

Raguidel’, in Patrizia Romagnoli and Barbara Wahlen eds., Figures de l’oubli (IVe–XVIe siècle) [=Études de lettres 1–2], 99–117. Lausanne: Université de Lausanne. Wolff, Robert Lee (1986). Studies in the Latin Empire of Constantinople. London: Variorum. Wolfgang, Lenora D. (1980). ‘Perceval’s Father: Problems in Medieval Narrative Art’, Romance Philology 34: 28–47. Zemel, Roel (1994). ‘The New and the Old Perceval: Guillaume’s Fergus and Chrétien’s Conte du Graal’, Bibliographical Bulletin of the International Arthurian Society 46: 324–42. Zink, Michel (1994). ‘Avertissement’, in Michel Zink ed., Chrétien de Troyes: Romans, 5–7. Paris: Livre de Poche. Zumthor, Paul (1977). ‘Le texte-fragment’, Langue française 40: 75–82.

Index References to illustrations are in bold, references to tables are underlined. adventure (aventure) 68, 193–7, see also phantom adventures aesthetics of (no-)consequence 142, 194, 197, 205, 217 Arthur 4–5, 6, 22–3, 41, 43, 55–6, 77, 88–9, 97, 114–15, 120, 131–3, 134 n.25, 143, 145, 149, 167, 173, 176–7, 184, 194, 198, 202 n.45, 205, 207, 209–10, 214, 221 n.8, 224 court of 39, 48–51, 55, 84–5, 86, 88–9, 89–90, 92, 100, 101, 117–18, 131–2, 155, 191, 198, 199, 205 refuses to eat 185–9, 195, see also food Atanassov, Stoyan 169, 171 L’Atre Périlleux 171–5, 179, 184, 193, 195–6, 198, 200–1, 205, 214, 216 authorship 10–13, 20–3, 25, 72, 111–13, 127, 161, 220

as textual motif 25 n.76, 70, 77–8, 96, 124–5, 129, 183, 215–16, 223–4 text as a book (livre) 72, 80–2, 95–6, 110, 213, 215, 222 branche 33–7 Brien (‘Biel Prison’) 170–1, 179–80, 194, 203–5 Brien de la Gastine 192–3 broken sword 10, 21, 23, 26–7, 42, 61–3, 65, 103, 104, 125, 196–7, 202 Bruckner, Matilda 3, 7, 14 n.47, 19, 20 n.64, 22, 23–5, 27 n.78, 40, 41 n.28, 42–3, 66 n.47, 77, 104 n.45, 111, 112 n.3, 116 n.7, 118, 120, 130 n.22, 142 n.32, 219, 225, 227 n.18 Burgwinkle, William 64 n.45 Busby, Keith 3, 6–7, 10, 11, 14, 58, 98 n.40, 114 n.4, 169, 184, 195, 200, 219

Bagomedés 39, 45, 52, 96 Balsamo, Gian 135 n.26 Baudouin de Flandre 148–50 Baumgartner, Emmanuèle 121 n.15, 132 n.23 Beaudous 198–9 Bel Inconnu (Renaut de Beaujeu) 107, 119, 145, 199, 200 Besamusca, Bart 16–17 Biaurepaire (Beaurepaire) 44, 66–7, 92, 103, 115–17, 138–40, 199, 225 Blancheflor 44, 66–7, 116–18, 135, 138–40, 199 bleeding lance 37, 43, 56–8, 60, 65, 104, 114 Bloch, R. Howard 120 n.12, 160 Blumenfeld-Kosinski, Renate 220 n.6 book, see also manuscript libraries 149, 219 n.5

Caradoc 32, 41, 43, 46, 50, 76, 85–7, 89, 90–1, 112, 152–9, 187, 189 paternity 89–90, 152–9, 203–4 relationship with mother Ysave, 85–6, 112, 142, 153–8 Carruthers, Mary 71, 80 Cave, Terence 132, 146, 148–50 Cazelles, Brigitte 57 n.37, 146 n.37, 148 n.40 centrifugal, see narrative dynamic centripetal, see narrative dynamic chanson de geste 6, 16, 106, 107, 141, see also Chanson de Roland; Guillaume cycle; Paon cycle; Raoul de Cambrai Chanson de Roland, La 183 n.18 Chevalier as deus espees, Le 159 n.53, 170–1, 179–80, 181, 190, 191–4, 199, 202–5, 209–10, 214, 217 Chevalier à l’épée, Le 167–9, 171 n.14

270

GENERAL INDEX

Chrétien de Troyes 2, 11, 22, 25, 75–6, 81, 96, 106, 111–12, 127–9, 161, 163, 167–9 reputation 12–15, 28, 111, 164–6, 222–3 romances 8, 20 Cligés 8, 9 n.31, 15, 18, 70, 178, 210, 222 Conte du Graal, Le (Perceval), see Conte du Graal corpus Erec et Enide 3 n.6, 8, 9 n.31, 11, 75–6, 178, 200, 213 n.57, 222 Lancelot (Le Chevalier de la Charrette), 8, 9 n.31, 24, 34, 176, 178, 200, 226 n.17 Yvain (Le Chevalier au Lion) 8, 9 n.31, 175–6, 178, 200, 201 n.43, 213 n.57, 222 Chronique rimee (Philippe Mouskés) 149 Clarissant 43, 98–9 [99], 133–4, 147 Cligés, see Chrétien de Troyes, romances Cohen, Gustave 2 Coleman, Joyce 216 n.59 Combes, Annie 24 n.71, 42 Conte du Graal corpus, passim considered as a cycle, 5–8, 9–11, 15–27, 47, 197, 220–2, see also cyclicity, of the Conte du Graal corpus prosification (1530) 10, 222 reception of medieval 2, 8–9, 13–15, 19–20, 73–110, 164–6, 221–2 modern 2–3, 6–7, 13, 20, 106, 164–5 texts Elucidation 1, 7, 10, 45, 105–10, 141, 190, 222, 225 Bliocadran 1, 7, 9, 10–11, 20, 47, 105–9, 141, 165, 222 Conte du Graal, Le (Perceval [CdG]) 1, 4–5, 8–11, 14–15, 22, 24, 35 n.17, 38–9, 42–4, 47–51, 56–60, 66, 74, 76, 81, 83–5 [84], 96, 100, 101–3 [102], 109, 111–21, 130–6, 145, 146–7, 160–1, 165, 168, 176–8, 182, 185–93, 195, 197–8, 201 n.43, 202–3, 206, 213, 220, 222–3, 227 n.18 First Continuation [C1] 1 n.1, 3, 4 n.12, 15, 17, 22–3, 25, 29–37, 41–7, 48, 55–6, 60–3, 68–9, 76,

83–91 [84, 88–9], 92 n.32, 99, 105–7, 109, 112, 136–8, 141–6, 150–9, 161, 165 n.3, 168, 169, 177–8, 185, 187–90, 199, 200, 202, 216, 217, 222, see also Continuations Second Continuation [C2] 1 n.1, 7, 9 n.32, 19, 21–2, 22–3, 24 n.71, 25, 26–7, 29–30, 31–2, 37–8, 39, 41–2, 44–6, 47, 48, 49–53, 56, 61 n.43, 63–6, 67, 68, 74, 75, 77, 83–5 [84–5], 87, 97, 101, 103, 105, 112, 121–5, 126–7, 132, 137–40, 141, 144–6, 159–60, 161 n.55, 165, 166, 168, 178–9, 185, 191, 197–8, 201, 202–3, 207–8, 208–9, 221, 225, 226–7, see also Wauchier de Denain; Continuations Gerbert Continuation [CG] 1, 7, 21–2, 23, 25–7, 40, 42, 47, 53–6, 65–8, 96 n.38, 101, 105, 125–8, 139–41, 144, 199, 203 n.47, 219, 220–22, 225, see also Gerbert de Montreuil Manessier Continuation [CM] 1, 9–10, 15, 19–20, 21, 23, 25, 26–7, 32, 40, 42, 47, 48, 65, 66, 67–8, 74, 77–8, 81, 85, 87 n.23, 101, 105, 112, 129, 161–2, 169, 194, 199, 202, 209 n.50, 213, 219 n.4, 220–1, 222, 225, 226, 227, see also Manessier transitions between different parts 20–3, 84, see also Corley, Corin translated into German, see Rappoltsteiner Parzifal continuation (practice of) 4–5, 22, 26 n.77, 32, 150 n.48, 220; see also family relations, literary filiation; fidelity, textual Continuations, see also Conte du Graal corpus, texts criticisms of, 2–3, 32, 36, 86, 90, 106, 127, see also Conte du Graal corpus, reception of redactions of the First Continuation 17, 19, 30–8, 42–7 Short Redaction (C1 [SR: A and L]) 30–1, 32–3, 35, 41 n.28, 43, 45–6, 50, 61 n.43, 92 n.32, 106, 109, 153–8



GENERAL INDEX

Mixed Redaction (C1 [MR]) 15, 30–1, 35 n.16, 43, 61 n.43, 153, 199 Long Redaction (C1 [LR]) 30–1, 32–3, 35, 42–7, 53, 55, 68–9, 84, 86, 88–91 [88–9], 92 n.32, 109, 112, 137, 152–9, 188 n.27, 199, 216 redactions of the Second Continuation 31–2, 42 Short Redaction (C2 [SR]) 31, 50, 63–4 Long Redaction (C2 [LR]) 31, 37–8, 64, 84 Corley, Corin 22–3, 41, 161 n.55, 196–7, 201, 219 n.4 crossroads 36 n.18, 209, 226 cyclicity 5–6, 15–20, 29, 81, 108, 200–1 genealogical 6, 16, 107, 141, see also family relations, literary filiation models of, 15–17, 23, 223–7 of the Conte du Graal corpus, 7, 17–27, 41–3, 47, 105–8, 141 n.31, 196–7, 220–2 Deleuze, Gilles (with Félix Guattari) 36 n.18, 68–9 Dolorous Mountain, see Mont Dolereus Don Quixote (Miguel de Cervantes) 12 n.38, 28, 218 Doner, Janet, 82–3 Douchet, Sébastien 3 n.11, 111 Dragonetti, Roger 134 Durmart le Galois 198–9 Erec et Enide, see Chrétien de Troyes, romances Estoire del Saint Graal, La, see Vulgate Cycle Estoire de Merlin, La, see Vulgate Cycle family relations, see also unknown kin; recognition father-son 85, 89, 89 n.24, 107–8, 120, 124, 135–6, 142, 144–5, 152–9, 192–3, 203–4, 224 mother-son 46–7, 59–60, 64 n.45, 66–8, 85–6, 101, 102, 103, 107, 112, 113, 115–28, 130 n.22, 131–4, 135, 137, 141 n.31, 142, 146, 152–9, 192–3 siblings 3, 43, 67, 85, 98, 99, 123–7, 131–4, 140–1, 142–3, 144, 146,

271

147, 159–60, 182, 183, 190, 203, 214 literary filiation 20, 22–3, 71 n.2, 111–13, 119, 127–9, 160–2, 164–5, 169, 191–2, 198–9, 217, see also mother-text; cyclicity, genealogical Fergus (Guillaume le Clerc) 190–1, 193–4, 197–8, 200, 205–6, 214, 217 fertility 111–12, 130 n.22, 138–40, see also sterility fidelity (and lack thereof) between characters 86, 152–9, 172, 175, 178–9, 180–2 textual 2–3, 21–2, 32, 36–7, 106, 112, 128, 145, 154, 160, 161, 220, see also memory, textual Fisher King, 1, 10, 21, 23, 26–7, 37, 42, 51, 53, 56 n.36, 61, 62, 65, 67, 74, 89, 97, 103, 104, 106, 107, 117, 120, 125, 135–8, 148 n.40, 191, 193, 197, 199, 202, 209, see also Grail Castle Florimont (Aimon de Varenne) 13, 71 n.2, 79–80, 82 food 39, 46, 98, 190, 204, see also Arthur, refuses to eat; Grail, serves food Frappier, Jean 2–3, 3 n.6, 32, 57–8, 90 Gaggero, Massimiliano 3 n.11, 151 n.31 Galaad 107, 123, 194, 199, 225, 227 n.18 Gallais, Pierre 30, 31 n.4 Gauvain, passim and the Grail 33–5, 37, 42, 60–2, 89, 136–7, 199, 208–9 children (incl. Guinglain) 41, 85, 107, 119–20, 142–3, 145–6 identity in question 131–4, 147, 150–2, 171–5, 179–85, see also recognition; unknown kin reputation 35 embodiment of chivalric values 169–78, 193–4 resulting in him becoming the object of love 142, 150–2, 169–70, 173, 178–80, 182–3 sexuality frustrated relationships 133–4, 135, 179–80 lack of fidelity 175, 178–9 genealogy, see cyclicity, genealogical; family relations Genette, Gérard 4 n.12, 223 Gerbert de Montreuil, 21, 26–7, 53–5, 66 n.47, 125–6, 127–9, 139–41, 161–2,

272

GENERAL INDEX

197, 221, 225, see also Conte du Graal Corpus, Gerbert Continuation Gilbert, Jane 183 n.18 Giflet 41, 48–9, 55, 56, 58 Godefroi de Bouillon 140–1 Godzich, Vlad 110 n.54 Gomez, Etienne 3 n.11, 7 n.24, 19 Gonnot, Micheau 18 Gornemant de Goort 44, 66 n.46, 67, 115, 120, 127–8, 139–40 Grail (the object, the vessel), 1, 3, 21, 25–6, 38, 42, 49, 54, 56–61, 63, 74, 76, 97, 103–4, 106–7, 109, 114, 118, 123, 126, 136–7, 141 n.31, 197, 202, 209, 224–5, 227 Holy Grail 104 n.46, 109, 169, 193, 199, 221, 224, 227 procession 42, 60–1, 68, 104, 106, 117, 123, 146, 182 serves food 42, 60, 136–7 Grail Castle, 21, 24 n.71, 26, 33, 35, 37, 38, 42–3, 44, 45, 48, 53, 55, 56, 58–65, 66 n.46, 67–8, 74, 75, 103–4, 106–7, 114, 115, 117, 118, 120–1, 125, 135–6, 139, 143–4, 155, 159, 160, 161, 191, 202, 208, 225, 227, see also Fisher King Greene, Virginie, 12–13 Greoreas 44, 130–1, 176, 202 Griffin, Miranda 25, 115 Grigsby, John 90–1, 127 Guattari, Felix, see Deleuze, Gilles Guerrehet 3, 32, 41, 131–2, 144, 202 Guillaume cycle 6, 17, 91 Guiromelant 5, 11, 22, 43, 96, 98–9 [99], 132–3, 147, 176, 202 Haidu, Peter 135 n.27 Hermit episode 38, 59–60, 63–5, 83–5 [84], 107, 114, 118–19, 148 n.40 Hideous Damsel 44, 48–58, 63–5, 66 n.46, 83, 84, 92, 148 n.40 condemnation of Perceval 51, 56–7, 58 n.39, 60, 63, 118–19, 121, 136, 161, 191–3 adventures offered to Arthur’s knights 48, 54–6 Hindman, Sandra 71, 78, 93–4, 96, 98, 110 n.53 Huchet, Jean-Charles 194 Hult, David 4, 6, 9 n.30, 13–14, 222 Huot, Sylvia 19 n.57, 72, 81 n.19

Hunbaut 163, 182–5, 190, 200, 204–5, 214, 215–16 incompleteness 2, 4, 7, 21 n.65, 25–7, 51 n.34, 56 n.36, 196–7, 221, 223–8, see also continuation, practice of interlace 33–42, 45–8, 52, 60, 68, 83–94, 97, 101 n.43, 105, 115, 117–18, 156, 166–7, 212–16, 218, 224 interpolation 9–10, 30–1, 38, 42–6, 56, 61 n.42–43, 120 n.14, 172 n.17, 199, 201, 223, see also inviting gap; Rappoltsteiner Parzifal; Wace, Roman de Brut inviting gap (Frank Brandsma) 125 Jeanne de Flandre 129, 148–50, see also patronage Jewers, Caroline 12 n.38, 194 Jaufre 186–90, 194–5 Kaherdin 48–50, 58 Kay, Sarah 13 Kennedy, Elspeth 34, 39 Keu 40 n.24, 45, 51, 55, 96, 143, 177–8, 180–4, 186–90, 207, 216, 217 Koble, Natalie 223 Kibler, William 209 Kittay, Jeffrey, see Godzich, Vlad Lacy, Norris 106, 166, 172 n.16, 210 n.52, 212–3, 218 Lancelot 24, 34–5, 107, 114, 202 n.45, 208, 211, 218, 226, see also Prose Lancelot; Chrétien de Troyes, romances, Lancelot Lancelot Compilation (Middle Dutch) 16 Le Rider, Paule 2, n.4 Leupin, Alexandre 3, 151–2, 196 Lore 4–5, 84, 99–101 Lot, Ferdinand 34 Maddox, Donald 119 Manessier 11, 21, 26, 27, 77–8, 81, 96, 112–13, 129, 148–9, 161–2, 197, 199, 219 n.4, 221, 222, 226, 227, see also Conte du Graal Corpus, Manessier Continuation manuscript, see also interpolation illustration 14, 72–3, 82, 83, 91, 93, 94–104, 109–10, 113, 222 characters as witnesses in 98–101 [99], 103



GENERAL INDEX

large initials 82–91 [84–5, 88–9], 105 rubrication 19, 71, 72–3, 75, 82, 91–4, 96–8, 99–101 [99], 104 n.46, 108, 109, 113, 222 incipits and excipits 105, 108–9 scribal activity 9–11, 21 n.65, 22, 30, 53, 73, 76, 83–94, 199, 201 Marnette, Sophie 213–14 marriage 43, 76, 88–9, 89–90, 98–9 [99], 133–5, 139–41, 171, 175, 205 Marvellous Bed 93, 131, 168 Marx, Jean 219 n.4 McCracken, Peggy 212 n.53 McGerr, Rosemary 226 Méla, Charles 120 memory of audience 38, 66–8, 82, 96, 109, 113–15, 147, 184 of author/narrator 167–8, 212 cultural 80, 82–3, 96 manuscript as an aid to 71, 82, 94, 96, 101, 103, 119, see also manuscript, illumination memoire/estoire 60, 70–2, 74–81, 101, 108, 113 of romance characters 37, 41, 44, 54, 55, 59–60, 67–8, 114–15, 117, 122–4, 147, 151, 196 n.37, 202, 205, 206–7, 211–12 textual 49, 66–9, 117–18, 121, 143, 147, 156, 161, 196, 207–8, see also fidelity, textual Meraugis de Portlesguez 170, 190 Merveilles de Rigomer, Les 200, 208, 210–13, 215 Micha, Alexandre 164 n.1 Middleton, Roger 73–4 Mont Dolereus (Dolorous Mountain) 44–5, 48–50, 49 n.33, 52–3, 56, 130, 191, 208, 209, 226 Morris, Rosemary 39 n.22 Mort le roi Artu, La, see Vulgate Cycle Morte d’Arthur, Le (Thomas Malory) 28, 34 n.11, 218 mother-text, 112–13, 155, see also Conte du Graal; family relations, literary filiation narrative dynamic 30, 71–2, 101 n.43, 105, 125, 139, 156, 193 centrifugal/centripetal 23–5, 29, 36–8, 41, 47, 55, 68–9, 90–2, 106–10, 113, 137–8, 141, 144, 161 n.55,

273

166, 185, 197, 200–2, 206–8, 210, 212–13, 217, 220, 225, 227, see also interlace Nixon, Terry 11, 14 n.43, 19–20, 95, 98 Nykrog, Per 8 oaths 35, 38, 48–9, 54–5, 56–9, 61–3, 65, 151, see also promises Odysseus 146, 148, 151 Orlando Innamorato (Matteo Maria Boiardo) 28 Ovide moralisé 91 Owen, D. D. R. 167–8, 190–1 Paon cycle 220 n.6 paratextuality 71–3, 82, 96 n.37, 108–10 Paris, Gaston 164 n.1 patronage 71 n.2, 112, 129, see also Philippe de Flandre; Jeanne de Flandre Perceval, passim best knight in the world 25, 30, 48, 50–3, 55, 56, 61 n.43, 63–4, 65, 68, 103, 121, 138, 160, 168, 191, 193, 197 biography 25, 29–30, 38, 56, 69, 73, 100, 101–3 [102], 105, 107–9, 141, 185, see also narrative dynamic, centripetal condemnation of, see Hideous Damsel paternity 107–8, 120, 124, 135–6 relationship with mother 59–60, 64 n.45, 68, 101–3 [102], 107, 113, 114–27, 128–9, 130, 135, 137, 141 n.31, 142, 159, 193 sexuality (blocked) 135, 140–1, 199 Perlesvaus 34 n.12–13, 81 n.18, 199, 200 phantom adventures / unnarrated romances 35 n.17, 209–10, 209 n.51 Philippe de Flandre 129, 148, 161, see also patronage Pickens, Rupert 8 Pickford, Cedric 34 Potvin, Charles 20, 110 n.53 promises 36–7, 43, 44, 48, 54, 66, 110, 113, 115, 117–18, 133, 159, 172, 205, 208, see also fidelity; oaths prophecy 50–1, 55, 60, 64, see also Hideous Damsel, condemnation of Perceval prose compared with verse 19, 28, 75, 93–4, 98 n.42, 212–15, 219, 219 n.5, 223, 227–8

274

GENERAL INDEX

romance 18, 28, 34, 42, 75, 80–2, 91–2, 94, 115, 166–7, 199, 212–13, 215, 219, 222, 227–8, see also Vulgate Cycle Prose Lancelot, see Vulgate Cycle Prose Tristan 18, 34 n.12, 91 Pseudo-Turpin (Prologue) 75, 82 Pucelle de Lis 41, 85, 88–9, 142–3, 150–2, 160, 169, 178–9, 183, 185, 206 quest 26, 27, 37, 43, 47–50, 52–60, 62, 64 n.45, 68, 87, 92–3, 104, 117, 120 n.11, 128, 143, 159, 172, 182, 183–4, 191, 199, 202, 203, 207–10, 221, 224–5, 227 Queste del Saint Graal, Le, see Vulgate Cycle Ramm, Ben, 64 n.45 Rand, Peregrine 179, 185, 200–1, 209 n.51 Raoul de Cambrai 201–2 Rappoltsteiner Parzifal 9–10, 93 n.33, 221–2 recognition 37, 66–7, 68, 77, 113–14, 121–6, 130–1, 134, 138–9, 144, 146–52, 155, 159–60, 166, 171–5, 178–85, 193, 198, 220 redactions of the First and Second Continuations, see Continuations, redactions of Reis, Levilson 96 n.37 retroactivity 25, 29, 112, 155, 160, 161, 220 rewriting 17, 32, 42–3, 68, 71, 76, 106–8, 136–40, 220, see also Continuations, redactions of revenge 27, 40 n.24, 44–5, 61, 65, 67, 130–1, 142, 154, 157, 192–3, 201–8 Ribard, Jacques 175–6 Rider, Jeff 4 n.14 Rieger, Angelica 98 Roach, William 30–3, 41, 56 n.36, 74, 89, 229, 244 Roman de Renart 33, 200 Roman de la Rose (Guillaume de Lorris and Jean de Meun) 4–7, 7 n.22 Roman de Troie, Le (Benoît de SainteMaure) 13–14, 18, 70, 78, 80 Rossi, Marguerite 91 Rupert Bear 93 n.34 Sagremor 40, 51, 87 n.23, 96, 105, 210

Sargent-Baur, Barbara 8 Scarlet Knight 67, 100, 101–3 [102], 124–5, 197 brother of 203 sons of 67, 203 n.47 Schmolke-Hasselmann, Beate 164–6, 168 n.11, 185 n.25, 191, 194, 198, 207 scribes, see manuscripts, scribal activity Séguy, Mireille 3, 34 n.13, 51 n.34, 144 n.33, 177 n.20, 221 sexuality, see Gauvain, sexuality; Perceval, sexuality Skårup, Povl 16–17, 19 Staines, David 6 Stones, Alison 14 n.43, 19 n.60, 91 Sunderland, Luke 23, 25, 223 sterility 111, 135–8, 192, see also fertility; Waste Forest; Perceval, sexuality Szkilnik, Michelle 154, 196 n.37 Taylor, Jane 18, n.56 Tether, Leah 3 n.11, 26 n.77, 150 n.48 Thompson, Albert 90, 106, 110 n.53 Tirant lo Blanc (Joanot Martorell) 28 translatio 18, 19 n.57, 70 Tristan 53, 55, 105, 225 unknown kin 67, 107, 119, 123–5, 131–4, 142–3, 147, 152, 159–60, 192–3, 202–4, see also recognition Van Coolput, Colette-Anne 14–15 Valenciennes 149 Vengeance Raguidel, La 175, 180–4, 190, 195, 199, 200, 202, 206–8 Veve Dame, see Perceval, relationship with mother Vie de Sainte Marie l’Egyptienne, La 9, 20 Vita Nuova, La (Dante Alighieri) 71 Vulgate Cycle 14, 15, 19, 25–6, 33–4, 36, 81 n.18, 106, 107, 114–15, 134 n.25, 193, 213, 215, 218–19, 220–1, 223, 224, 227 Estoire del Saint Graal 19 n.60, 91, 106 Estoire de Merlin 19 n.60, 34, 106 Prose Lancelot 18, 19 n.60, 24, 26, 34–5, 39 Queste del Saint Graal, La 18, 34, 104, 123, 194, 199, 219 n.4, 221 n.8, 224–5, 227 n.18



GENERAL INDEX

Mort le roi Artu, La 18, 26, 34, 80–1, 105, 114–15, 202 n.45, 223–5, 227 Wace Roman de Brut 13 Chrétien’s romances interpolated into 8, 11, 14, 18 Roman de Rou 79 Walters, Lori 200, 201 n.43 Waste Forest 67–8, 107, 121–3, 125, 132, 135, 138, 147 Wauchier de Denain 1 n.1, 21–2, 74, 77, 78, 121, 125, 127, 137, 145, 159, 161, 220, see also Conte du Graal corpus, Second Continuation

275

witnesses, see manuscript illuminations Wolff, Robert Lee 149 n.45, 150 n.47 Wolfgang, Leonora 107 n.52 Wolfram von Eschenbach 9–10 Ysave 46–7, 85–6, 88–9, 112, 152–8, see also Caradoc Yvain (Le Chevalier au Lion), see Chrétien de Troyes, romances Zemel, Roel 198 n.39 Zink, Michel 12 Zumthor, Paul 216 n.60

Index of Manuscripts For Conte du Graal manuscripts, the relevant siglum follows the manuscript’s name. Annonay fragments: 9 n.31 Bern, Burgerbibliothek, 113 [K]: 19, 21, 26 n.77, 63, 201 Bern, Burgerbibliothek, 354 [B]: 10, 19, 82, 147 n.38, 167 n.10 Brussels, Bibliothèque royale des ducs de Bourgogne, MS 11145: 81 n.18 Chantilly, Musée Condé 472: 164 n.1, 179, 200–1 Clermont-Ferrand, Bibliothèque municipale et interuniversitaire, 248 [C]: 10, 19 Edinburgh, National Library of Scotland, Advocates 19. 1. 5 [E]: 9, 20 n.60, 30, 31, 33, 37–8, 39 n.21, 41 n.25, 44–5, 53–4, 55, 64, 73, 74, 83–91 [84–5, 88–9], 95, 108, 110, 153 n.51, 161 n.55, 202–3 Florence, Biblioteca Riccardiana, 2943 [F]: 10, 19 Karlsruhe, Landesbibliothek, Cod. Donaueschingen 97: 9–10 London, British Library, Add. 36614 [L]: 9, 11 n.35, 19–20, 21, 30, 31 n.4, 32–3, 35, 41 n.27, 61 n.42, 63, 106, 120 n.13, 147 n.38, 165 London, College of Arms, Arundel XIV [H]: 10, 19, 49 n.33 Mons, Bibliothèque universitaire, 331/206 [P]: 9, 10–11, 15, 20, 30, 31, 35 n.16, 39 n.21, 41 n.25, 41 n.27, 64–5, 73–4, 75, 83, 84–5, 87, 91–3, 94–5, 96–8, 101, 106, 109–10, 161 n.55, 202–3 Montpellier, Bibliothèque interuniversitaire, Section Médecine, H 249 [M]: 9, 30, 32–3, 35, 41 n.25–26, 32, 61 n.42, 63, 73, 83, 95, 96, 97–8, 101, 108, 110, 145 n.34, 153 n.51

Paris, Bibliothèque nationale de France, fr. 112: 18 Paris, Bibliothèque nationale de France, fr. 794 [A]: 9, 11, 14, 19, 30, 35 n.16, 41 n.25, 41 n.27, 49 n.33, 61 nn.42–43, 63, 82, 120 n.13 Paris, Bibliothèque nationale de France, fr. 1429 [Q]: 9, 30, 32–3, 35, 41 n.25–26, 61 n.42, 63, 73, 74, 76, 83, 84–5, 87–91, 95, 108, 110, 145 n.34, 153 n.51 Paris, Bibliothèque nationale de France, fr. 1433: 172 n.17, 201 n.43 Paris, Bibliothèque nationale de France, fr. 1450 [R]: 9, 11, 14, 18–19, 31 n.6, 147 n.38 Paris, Bibliothèque nationale de France, fr. 1453 [S]: 9, 31, 35 n.16, 41 n.25, 41 n.27, 45, 49 n.33, 61 n.43, 63, 73, 83, 87, 91, 94–5, 98–101 [99], 103, 104, 108, 110, 202–3 Paris, Bibliothèque nationale de France, fr. 2168: 201 Paris, Bibliothèque nationale de France, fr. 12576 [T]: 9, 21, 23, 30, 35 n.16, 39 n.21, 41 n.25, 53, 56, 61 n.43, 64, 68, 71, 73, 74, 83, 84–5, 87 n.23, 95, 96, 98, 100, 101–3, 104, 108, 110, 126, 141 n.31, 161 n.55 Paris, Bibliothèque nationale de France, fr. 12577 [U]: 9, 30, 32–3, 35, 41 n.25, 41 n.27, 44–5, 53–4, 55, 61 n.42, 63, 73, 83, 91, 93–4, 95, 98, 101–3 [102], 104, 108–9, 110, 145 n.34, 153 n.51 Paris, Bibliothèque nationale de France, n. a. fr. 6614 [V]: 9, 21, 23, 30, 35 n.16, 41 n.25, 53, 61 n.43, 68, 73–4, 83, 84–5, 86 n.21, 87, 108, 126, 141 n.31



INDEX OF MANUSCRIPTS

Rennes, Bibliothèque municipale, 255: 19 n.60

277

Rome, Biblioteca Casanatense, Mss. 1409: 9–10

Already Published 1.  Postcolonial Fictions in the ‘Roman de Perceforest’: Cultural Identities and Hybridities, Sylvia Huot 2.  A Discourse for the Holy Grail in Old French Romance, Ben Ramm 3.  Fashion in Medieval France, Sarah-Grace Heller 4.  Christine de Pizan’s Changing Opinion: A Quest for Certainty in the Midst of Chaos, Douglas Kelly 5.  Cultural Performances in Medieval France: Essays in Honor of Nancy Freeman Regalado, eds Eglal Doss-Quinby, Roberta L. Krueger, E. Jane Burns 6.  The Medieval Warrior Aristocracy: Gifts, Violence, Performance, and the Sacred, Andrew Cowell 7.  Logic and Humour in the Fabliaux: An Essay in Applied Narratology, Roy J. Pearcy 8.  Miraculous Rhymes: The Writing of Gautier de Coinci, Tony Hunt 9.  Philippe de Vigneulles and the Art of Prose Translation, Catherine M. Jones 10.  Desire by Gender and Genre in Trouvère Song, Helen Dell 11.  Chartier in Europe, eds Emma Cayley, Ashby Kinch 12.  Medieval Saints’ Lives: The Gift, Kinship and Community in Old French Hagiography, Emma Campbell 13.  Poetry, Knowledge and Community in Late Medieval France, eds Rebecca Dixon, Finn E. Sinclair with Adrian Armstrong, Sylvia Huot, Sarah Kay 14.  The Troubadour Tensos and Partimens: A Critical Edition, Ruth Harvey, Linda Paterson 15.  Old French Narrative Cycles: Heroism between Ethics and Morality, Luke Sunderland 16.  The Cultural and Political Legacy of Anne de Bretagne: Negotiating Convention in Books and Documents, ed. Cynthia J. Brown 17.  Lettering the Self in Medieval and Early Modern France, Katherine Kong 18.  The Old French Lays of Ignaure, Oiselet and Amours, eds Glyn S. Burgess, Leslie C. Brook 19.  Thinking Through Chrétien de Troyes, Zrinka Stahuljak, Virginie Greene, Sarah Kay, Sharon Kinoshita, Peggy McCracken 20.  Blindness and Therapy in Late Medieval French and Italian Poetry, Julie Singer 21.  Partonopeus de Blois: Romance in the Making, Penny Eley 22.  Illuminating the Roman d’Alexandre: Oxford, Bodleian Library, MS Bodley 264: The Manuscript as Monument, Mark Cruse

Conte du Graal:Romance 02/11/2011 11:27 Page 1

THE CONTE DU GRAAL CYCLE

THE CONTE DU GRAAL CYCLE Chrétien de Troyes’s Perceval, the Continuations, and French Arthurian Romance

Thomas Hinton

Thomas Hinton

Chrétien de Troyes’s late twelfth-century Conte du Graal has inspired writers and scholars from the moment of its composition to the present day. The challenge represented by its unfinished state was quickly taken up, and over the next fifty years the romance was supplemented by a number of continuations and prologues, that came to dwarf Chrétien’s text. In one of the first studies to treat the Conte du Graal and its continuations as a unified work, Thomas Hinton considers the whole corpus as a narrative cycle. Through a combination of close textual readings and manuscript analysis, Hinton argues that the unity of the narrative depends on a balanced tension between centripetal and centrifugal dynamics. He traces how the authors, scribes and illuminators of the cycle worked to produce coherence, even as they contended with potentially disruptive forces: multiple authorship, differences of intention, and changes in the relation between text, audience and book. Finally, this book tackles the long-held orthodoxy that places the Perceval Continuations on the margins of literary history. Widening the scope of enquiry to consider the corpus’s influence on thirteenth-century verse romances, this study re-situates the Conte du Graal cycle as a vital element in the evolution of Arthurian literature. THomAs HInTon is Junior Research Fellow in modern Languages at Jesus College, oxford.

BOYDELL & BREWER Ltd PO Box 9, Woodbridge IP12 3DF (GB) and 668 Mt Hope Ave, Rochester NY 14620-2731 (US) www.boydellandbrewer.com

Gallica