The Anglo-Norman Language and its Contexts 1903153301, 9781903153307

Citation preview

spine 20mm 14 jan 10

Chapman, Anne Curry, Eric Haeberli, Richard Ingham, Andy King, Anthony Lodge, Jean-Pascal Pouzet, William Rothwell, David Simpkin, Louise Sylvester, David Trotter, Laura Wright. Cover: Part of a thirteenth century letter in Anglo-Norman from Gillian de Tregoz to W. de Merton, from National Archives Ancient Correspondence SC1/7, no. 198. Photograph by Emma Cavell; reproduced by kind permission of the National Archives.

Boydell & Brewer Ltd PO Box 9, Woodbridge IP12 3DF (GB) and 668 Mt Hope Ave, Rochester NY 14620-2731 (US) www.boydell.co.uk / www.boydellandbrewer.com

INGHAM (ed.)

YORK MEDIEVAL PRESS

The Anglo-Norman Language amd its Contexts

The development of Anglo-Norman (the variety of medieval French used in the British Isles), together with the role it played in the life of the medieval English kingdom, is currently a major topic of scholarly debate. The essays in this volume examine Anglo-Norman from different perspectives and in different contexts, though with a concentration on the theme of linguistic contact between Anglo-Norman and English, seeking to situate such contact more precisely in space and time than has hitherto been the case. Overall they show how Anglo-Norman retained a strong presence in the linguistic life of England until a strikingly late date, and how it constitutes a rich and highly valuable record of the French language in the middle ages. Contributors: Adrian Bell, Paul Brand, Mark Chambers, Adam

YORK MEDIEVAL PRESS

The

Anglo-Norman Language and its Contexts Edited by RICHARD INGHAM

The Anglo-Norman Language and its Contexts The question of the development of Anglo-Norman (the variety of medieval French used in the British Isles), and the role it played in the life of the medieval English kingdom, is currently a major topic of scholarly debate. The essays in this volume examine it from a variety of different perspectives and contexts, though with a concentration on the theme of linguistic contact between Anglo-Norman and English, seeking to situate it more precisely in space and time than has hitherto been the case. Overall they show how Anglo-Norman retained a strong presence in the linguistic life of England until a strikingly late date, and how it constitutes a rich and highly valuable record of the French language in the middle ages.

AN-1a.indd 1

11/03/2010 12:00:07

YORK MEDIEVAL PRESS York Medieval Press is published by the University of York’s Centre for Medieval Studies in association with Boydell & Brewer Limited. Our objective is the promotion of innovative scholarship and fresh criticism on medieval culture. We have a special commitment to interdisciplinary study, in line with the Centre’s belief that the future of Medieval Studies lies in those areas in which its major constituent disciplines at once inform and challenge each other. Editorial Board (2005–2010): Professor J. G. Wogan-Browne (Dept of English and Related Literature) Dr T. Ayers (Dept of History of Art) Professor P. P. A. Biller (Dept of History) Dr J. W. Binns (Dept of English and Related Literature) Dr Gabriella Corona (Dept of English and Related Literature) Professor W. M. Ormrod (Chair, Dept of History) Dr K. F. Giles (Dept of Archaeology) Consultant on Manuscript Publications: Professor Linne Mooney (Dept of English and Related Literature) All enquiries of an editorial kind, including suggestions for monographs and essay collections, should be addressed to: The Academic Editor, York Medieval Press, University of York, Centre for Medieval Studies, The King’s Manor, York, YO1 7EP (E-mail: [email protected]). Publications of York Medieval Press are listed at the back of this volume.

AN-1a.indd 2

11/03/2010 12:00:07

The Anglo-Norman Language and its Contexts Edited by

Richard Ingham

YORK MEDIEVAL PRESS

AN-1a.indd 3

11/03/2010 12:00:08

© Contributors 2010 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 First published 2010 A York Medieval Press publication in association with The Boydell Press 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 and with the Centre for Medieval Studies, University of York ISBN 978 1 903153 30 7 A CIP catalogue record for this book 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.

Typeset by Word and Page, Chester Printed in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne

AN-1a.indd 4

11/03/2010 12:00:08

CONTENTS List of Figures

vi

Acknowledgements

vii

List of Contributors

viii

List of Abbreviations 1. Anglo-Norman: New Themes, New Contexts (Editor’s Introduction) Richard Ingham

1

2. Later Anglo-Norman as a Contact Variety of French? Richard Ingham

8

3. The Sources of Standardisation in French – Written or Spoken? Anthony Lodge

26

4. Husbonderie and Manaungerie in Later Medieval England: A Tale of Two Walters William Rothwell

44

5. Bridging the Gap: The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts David Trotter

52

6. From Apareil to Warderobe: Some Observations on Anglo-French in the Middle English Lexis of Cloth and Clothing Mark Chambers and Louise Sylvester

63

7. Languages in the Military Profession in Later Medieval England Anne Curry, Adrian Bell, Adam Chapman, Andy King and David Simpkin 8. The Language of the English Legal Profession: The Emergence of a Distinctive Legal Lexicon in Insular French Paul Brand

AN-1a.indd 5

ix

74

94

9. Mapping Insular French Texts? Ideas for Localisation and Corre Dialectology in Manuscript Materials of Medieval England lated Jean-Pascal Pouzet

102

10. A Pilot Study on the Singular Definite Articles le and la in Fifteenth-Century London Mixed-Language Business Writing Laura Wright

130

11/03/2010 12:00:08

11. Investigating Anglo-Norman Influence on Late Middle English Syntax Eric Haeberli

143

12. The Transmission of Later Anglo-Norman: Some Syntactic Evidence Richard Ingham

164

Index



183

FIGURES Maps illustrating dialect mixing in Parisian French

37–38

vi

AN-1a.indd 6

11/03/2010 12:00:08

ACKNOWLEDGEMENTS The volume represents the published proceedings of two workshops on AngloNorman and medieval language hosted by the School of English, Birmingham City University, in February 2007 and January 2008. The support of the British Academy in funding the research network on Anglo-Norman and Middle English made the workshops possible, and is gratefully acknowledged, as are helpful suggestions made by an anonymous external reviewer of the volume.

vii

AN-1a.indd 7

11/03/2010 12:00:08

CONTRIBUTORS Adrian Bell: University of Reading Paul Brand: University of Oxford Mark Chambers: University of Westminster Adam Chapman: University of Southampton Anne Curry: University of Southampton Eric Haeberli: University of Geneva Richard Ingham: Birmingham City University Andy King: University of Southampton Anthony Lodge: University of St Andrews Jean-Pascal Pouzet: University of Limoges William Rothwell: Swansea David Simpkin: University of Reading Louise Sylvester: University of Westminster David Trotter: University of Aberystwyth Laura Wright: University of Cambridge

viii

AN-1a.indd 8

11/03/2010 12:00:08

ABBREVIATIONS ALF AND DMF DMLBS EELR L1 L2 LAEME LALME MED OE OED PROME

pl. sg. TNA YCOE

J. Gilliéron and E. Edmont, Atlas linguistique de la France. Paris, 1901–11. Anglo-Norman Dictionary, ed. W. Rothwell, L. Stone and T. Reid. London, 1977–92, and at http://www. anglo-norman.net/gate/ Dictionnaire du moyen français (1330–1500), at http:// atilf.atilf.fr/dmf.htm Dictionary of Medieval Latin from British Sources, ed. R. Latham. London and Oxford, 1975–. The Earliest English Law Reports , vols. I–IV, ed. P. Brand. Selden Society 111, 112, 122, 123. London, 1996, 1996, 2005, 2007. first language second language M. Laing and R. Lass, A Linguistic Atlas of Early Middle English, 1150–1325. Edinburgh, 2007–, at http://www. lel.ed.ac.uk/ihd/laeme1/laeme1.html A. McKintosh, M. L. Samuels and M. Benskin, A Linguistic Atlas of Late Middle English. Aberdeen, 1986. Middle English Dictionary, ed. H. Kurath and S. McAllister Kuhn, Michigan, 1952–2001, at http://quod.lib. umich.edu/m/med/ Old English Oxford English Dictionary, at http://www.oed.com/ The Parliament Rolls of Medieval England, ed. C. GivenWilson, P. Brand, A. Curry, R. E. Horrox, G. Martin, W. M. Ormrod and J. R. S. Phillips, CD ROM version. Leicester, 2005. plural singular The National Archives York–Toronto–Helsinki Parsed Corpus of Old English Prose

Online resources referred to in this volume have been checked as current in December 2009.

ix

AN-1a.indd 9

11/03/2010 12:00:08

AN-1a.indd 10

11/03/2010 12:00:08

chapter one

Anglo-Norman: New Themes, New Contexts Editor’s Introduction

Standard accounts of the Anglo-Norman language still represent it as an emigrant language that flourished in England for a hundred years or so, and thereafter rapidly degenerated, to be replaced by Middle English in all its functions except its use in law. This volume shows that English–French bilingualism remained a central fact of the linguistic life of England well into the late medieval period, and gives detailed consideration to profiling what later Anglo-Norman was like, and how it functioned. At the heart of our subject is a challenging paradox: if certain contemporary reports are to be believed, at one point early in the fourteenth century French looked like becoming the official language of England. Over two hundred years after the Norman Conquest, it had become adopted across virtually the whole range of written registers: prose and verse literature, technical writing, commerce, central- and local-government business, legal proceedings and private correspondence. Yet most authorities agree that by this time the descendants of the Norman conquerors of 1066 were no longer ethnically distinct from the native population, and hardly ever spoke French as a mother-tongue. Languages without native speakers are supposed to die, or else they survive, like Latin, as highly specialised formal written codes. However, French existed to a considerable extent as a spoken vernacular in later medieval England: even its continental detractors, who mocked it as ‘le faus franceis d’Engleterre’, derided specifically its spoken features. In short, the paradox is that French in England, conventionally referred to as ‘Anglo-Norman’, had few or no native speakers, yet was quite commonly spoken and enjoyed the status of a prestige written language. To the extent that Anglo-Norman continued to be spoken outside royal and seigneurial court circles, this is likely to have been within particular specialised professional domains such as law and estate management, where we have technical manuals in Anglo-Norman written from the later thirteenth century onwards. As David Trotter’s work emphasises, the textual record shows that French was in fact used across a whole range of contexts where accurate and efficient communication was essential, and for purposes where major financial and economic issues were at stake. How this state of affairs can be reconciled with the still commonly purveyed notion that Anglo-Norman was no more than ‘bad French’, imperfectly learned as a foreign language, remains unclear.

1

AN-1a.indd 11

11/03/2010 12:00:08

The Anglo-Norman Language and Its Contexts There is still a dearth of studies that consider the textual evidence itself as regards levels of attainment in French. This particularly concerns the areas of lexis and syntax, where no single-volume reference works exist to compare with the authoritative study of Anglo-Norman phonology and graphemics of Short (2007). Although Rothwell (1996) showed the potential for fruitful investigation of the Anglo-Norman lexicon as regards word-formation, the study of Anglo-Norman lexical semantic fields has not been systematically attempted. Nor do we know whether the sort of French written in England in the thirteenth and fourteenth centuries shows a marked decline in linguistic competence at the level of syntactic rules, as one would expect if by this time it was no longer acquired as an L1. There has been work on the relation between Insular French and English lexis in the late Anglo-Norman period, to which this volume contributes further, but outside the lexical domain the effect of Anglo-Norman on the development of English has typically been minimised in the research tradition. Yet it may be that, in syntax as well as in lexis, boundaries between languages in medieval England were somewhat porous. Producers of late-fourteenth- and early-fifteenth-century written English were often ‘code-crossing’, calquing English on French, so syntactic structures of the written code may have crossed over with them. The themes pursued by the chapters in this volume draw on the above background. A pervasive issue throughout much of the research tradition is inevitably whether the quality of later Insular French makes it worthy of serious study. Assessments of Anglo-Norman as ‘bad French’ have typically measured it against the central Old French of the thirteenth century. Before that time, however, many editors of Anglo-Norman texts note how similar it is to Continental French. Lodge’s chapter provides an important perspective in setting Anglo-Norman in the context of continental developments, particularly the emergence of Parisian French as a prestige variety. The impression of a growing gulf between Insular and Continental French is explicable within a shift in the locus of prestige models for linguistic usage. The Plantagenet courts of western France, notably in Normandy and Anjou, ceased to exist in the thirteenth century, while during this period the wealth of the Picard towns initially created alternative prestige models in the north of France. Subsequently the growing demographic and economic weight of Paris caused an apparently irreversible shift in perceptions about where the ‘best French’ was to be heard. Lodge adopts the hypothesis that the speech-norms of what was to become standard French derive from the koineisation which occurred in Parisian speech thanks to dialect mixing between the mid-twelfth and midthirteenth centuries. He finds that norms in Old French stabilised only in the fourteenth century: prior to that date Parisian French, like that of every region, was variable, as can be seen from the written texts produced in the city. An intriguing issue prompted by Lodge’s work is that Anglo-Norman was also, up to a point, affected by dialect contact, as Western French and Picard influences are usually thought to have been important in its development, yet it does not show koineisation comparable with the outcomes produced by

2

AN-1a.indd 12

11/03/2010 12:00:08

Anglo-Norman: New Themes, New Contexts dialect contact in the Paris region. This was presumably because of the absence of a point of focus for intense interaction between the dialects going into the mixture. Following the Introduction Ingham’s chapter seeks to situate Anglo-Norman within a wider context of variability in pre-modern (non-standardised) languages. It outlines some of the atypical grammatical features of AngloNorman, and shows that they are not unique to it, and not necessarily to be accounted for on the basis of imperfect learning of French by English speakers. Rather, they are features which resemble those known to arise in contact varieties of contemporary languages, particularly emigrant ones, where native speakers depart from home-variety forms under pressure from a dominant language used in the same geographical area. The contribution by Rothwell considers the importance of the use of Anglo-Norman in estate management, attested to not only by the treatises on that subject composed in French by Walter of Henley and others, but by the vocabulary focused on in Walter de Bibbesworth’s instructional poem known as Le Tretiz. This study notes the wide diffusion these works must have attained, judging by the large number of surviving manuscripts. This implies a ‘widespread understanding of the technical French of agriculture among the landed classes’ in the later medieval period. Nevertheless, Rothwell observes that in the domain of agricultural activities, relatively little lexical transfer took place into English, suggesting a lesser degree of language mixing than in other domains. His study reminds us of the role of professional contexts in the maintenance of French in England, and in its survival in the vocabulary of English. It has long been assumed that to all intents and purposes Anglo-Norman did not display regional variation in England, indeed, that its relative invariance was precisely why it remained useful as a medium of communication across a country subject to problems of mutual intelligibility among Middle English dialects. Pouzet considers the intriguing possibility that in fact there was a degree of regional variation in Insular French. To investigate whether there might be any substance in this notion, a need for a correlated approach to French and English is argued for, following the analytic approaches recently taken by Middle English dialectology. His chapter proposes categories of ‘profiling’ for Insular French texts, in terms of scribal features and what they might reveal about insular forms of French. He notes the existence of multilingual fourteenth-century manuscripts which would allow comparison of French and English data. One can envisage levels of language at which such a comparison could be revealing one way or the other, especially where localisable characteristics of Middle English are known, as is broadly the case with phonology. While Pouzet’s chapter is essentially programmatic in nature, those by Trotter and Wright are both resolutely data-based. They are similar in that they open up the fascinating domain of mixed-language records as a resource for analysing the interaction between English, French and Latin in the later medieval period. Trotter investigates the language of late-fourteenth- and fifteenth-

3

AN-1a.indd 13

11/03/2010 12:00:08

The Anglo-Norman Language and Its Contexts century bridge accounts, showing they can profitably be analysed in terms of current approaches to code-switching among bilingual language users. He identifies three types: two matrix languages, Latin and French, in either of which the text can switch to English items. A third type has Latin as the matrix language, switching to items that are of indeterminate language-membership status. This poses an interesting problem for code-switching theory in that the latter generally assumes that the language of the items in the matrix or the embedded language is determinate. There is a suggestive asymmetry displayed by the typology of code switching patterns documented in Trotter’s chapter. Matrix-language English, switching to French or Latin items, is apparently not found in the bridge accounts. Language users were apparently not tempted to switch codes once the matrix language for account documents was English. The point where a change took place to English as a matrix language may thus be seen as marking the end of the period of medieval bilingualism. A true bilingual would have been able to switch from a matrix English to an embedded French item. Trotter also notes the curious phenomenon of apparently trilingual passages in the same sentence, such as: Et de vjd solutis pro le clapsyng, ‘And for 6d paid for (book-)clasping’. He proposes that the French definite article le functions to signal the shift out of the matrix Latin to the vernacular English item clapsyng. This suggestion merits further investigation, as Trotter says, and is followed up in Ingham (2009). Trotter’s chapter raises a corner of the veil that has for so long covered the vast amounts of documentary material that have been left unexamined by linguists in the main, and which would provide empirical support for new conceptualisations of the language ecology of the late mediaeval period. Wright investigates the extent of variable practice among writers of fifteenthcentury matrix-language Latin accounts, taking further Trotter’s observation of the definite article le used in these accounts preceding English etymons. She reports on an examination of fifteenth-century mixed Latin–English accounts in which she finds plentiful use of le but very little use of the feminine article la in this function. Furthermore, the use of the French article blocked the appearance on an English noun of a Latinising suffix which was otherwise possible. Likewise in mixed French–English accounts, le rather than la is virtually always used to precede an English item. Her interpretation, following Trotter’s suggestion, is that in mixed-language accounts the French article preceding an English lexeme is a code-switching marker. The unmarked form of the article may thus be seen as having been appropriate to this function, which is quite unrelated to gender-marking. Wright finds a curious echo of the late medieval mixed-language practice of le-determination in the Tudor period, when certain London churches became known as, for example, St Mary le Bow despite the French element not having been part of the medieval name. The contribution by Chambers and Sylvester presents the aims and some initial findings of an Arts and Humanities Research Council project on the lexis of cloth and clothing, begun in 2007 and directed by Professor Gale Owen-

4

AN-1a.indd 14

11/03/2010 12:00:08

Anglo-Norman: New Themes, New Contexts Crocker of the University of Manchester. They seek to identify inter-language developments whereby the lexis of one language influenced that of the other. They are particularly interested in cases where lexemes adopted into English come from Insular rather than Continental French. They illustrate the kind of lexicographical scholarship that will build on existing dictionary resources and allow a fuller picture to be established of the amount of overlap between medieval English and French in this semantic field. An important issue arising from their work is the question of the source of Romance lexical items used in England, and whether insular usage simply maintained dialect lexis, perhaps from Normandy and western France, perhaps from the economically blossoming Northern/Picard dialect area, or whether it innovated as it did with morphological word-formation (Rothwell 1996). From lexis we move to grammar. Haeberli considers whether features of later Middle English could be accounted for in terms of contact with French, given the existence of substantial bilingualism. He concludes that French influence on Late Middle English syntax may plausibly show the effect of contact with French as regards the loss of then as a distinctive trigger for inversion, and the temporary increase of inversion with subject pronouns. The existence of extensive syntactic variation among different Middle English texts may also be seen as a differential effect of French influence. While acknowledging that the dearth of relevant information about the sociolinguistic background to surviving texts makes contact analyses difficult to establish for medieval languages, he considers that hypothesis of French influence seems a serious candidate for the majority of the structures he analyses. A common belief regarding the grammar of later Anglo-Norman is that it showed widespread disregard for continental norms of gender-marking and pronoun forms. Ingham subjects this view to a critique on the basis of AngloNorman petitions dated c. 1300–90 in the PROME database. With respect to gender, it is claimed that this opinion has arisen from consideration only of gender-marking on determiner and adjective forms, where the distinction would normally have been a matter of the presence or absence of final ‘e’. Spelling and metrical evidence indicates that this was no longer pronounced in Anglo-Norman by the second half of the thirteenth century at the latest. Yet this did not necessarily mean that Insular French speakers did not learn nouns specified for gender. An examination of around 7,000 instances of the possessive modifiers mon, ma, son and sa shows almost no gender violations with these modifiers until the 1370s, when they become noticeably more common. As regards object pronouns, it has been thought that as French became a second language, the atonic pronoun system became vulnerable to error in favour of use of the strong forms instead. However, factors to do with the loss of phonological oppositions in later Anglo-Norman again complicate the picture. Where phonological interference can be excluded, as with eux by comparison with les, frank errors do not appear until the late fourteenth century, at roughly the same time as the appearance of common gender errors with possessives. The conclusion drawn is that if care is taken to control for

5

AN-1a.indd 15

11/03/2010 12:00:09

The Anglo-Norman Language and Its Contexts phonological interference, later Anglo-Norman seems to have observed continental grammar in the respects studied in this chapter until a relatively late date. Curry et al. raise the question of whether knowledge of French among English speakers was in terminal decline in the earlier fifteenth century, when it falls out of use as the language of record in England. Their focus, however, is on language used by the English in France during the period of occupation lasting nearly four decades following Agincourt. Unlike the fourteenth century, when English attacks on France took the form of fairly brief raids, the English presence in France in the fifteenth century was relatively long-term, or at least intended to be so. The occupation was secured by thousands of English soldiers residing in Normandy and other occupied areas. Curry et al. ask how much evidence there is from the ample documentary material surviving from this occupation for the linguistic abilities of the English occupiers in French. The situation is complicated because the English made extensive use of local Frenchmen as administrators, who seemed to have written most of the available documents. However, French was used in musters, payment orders and receipts by English captains in France. The evidence of spoken French amongst the English soldiery obviously has to be indirect; they are known to have had functions which required interaction with the local Francophone population. Proclamations were made only in French from the mid-1420s. The authors argue that the military occupation of northern France would have required extensive knowledge of French on the part of at least the upper echelons of the English military presence, and possibly among soldiers as well. However, once the French recaptured their territory, the English army was forced to return to base, and by this time (the mid-fifteenth century), all public use of French was pretty much extinct in England. Indeed the French language would by then have been stigmatised as that of the (now successful) enemy. Hence any linguistic skills acquired in France would have found no scope for being maintained in the home country. Brand’s chapter takes up the themes prefigured in his earlier work on the languages of English law (Brand 2000). He shows that when legal texts first appear in French a little after 1250, a wide range of specialist senses of French was already in existence, pointing to a fairly long tradition of the use of French in English law at that point. A true legal profession did not come into existence until the thirteenth century. The difficult question Brand concludes with is to what extent the rich technical register of the law predated or followed that development: did lawyers take up pre-existing specialist senses of words such as mutacion or avowerie, or did they invent them? Future work on this question probably depends on a close analysis of Insular French and Latin terminology. In any case, it is clear from Brand’s study that the keen semantic sophistication provided by the French language in its insular guise in the later thirteenth century and afterwards affords ample evidence of its continuing vitality as a medium in which to transact major questions that would have been of decisive financial importance to litigants: Insular French as used in the law courts could

6

AN-1a.indd 16

11/03/2010 12:00:09

Anglo-Norman: New Themes, New Contexts not have been the moribund ritual, replete with obsolete and barely understood jargon, of certain all-too-facile textbook caricature. In their differing ways the authors contribute to a better characterisation both of Anglo-Norman itself and of the ways in which it made contact with the other languages of medieval England. It thus adds to a growing body of knowledge detailing the ways in which Anglo-Norman survived into the later medieval period as a functioning medium of communication adequate to the tasks for which it was required, and as a linguistic matrix for the newly emergent English of the Early Modern period. Studies of the diachronic evolution of English have all too often neglected the multilingual reality of medieval Britain, in which Insular French played a central role; linguistic developments affecting the structure of English were treated wherever possible within a paradigm of internally driven change. The thesis of this volume, coming at a time of greatly increased interest in language contact, present and past, is that such an approach is unsustainable. The paradigm has shifted: contact between English and other languages is no longer an add-on to linguistic histories, a kind of last resort when language-internal solutions fail, but forms the centrepiece of explanatory accounts of change. The contributors to the volume document the continuing vitality of Anglo-Norman in the linguistic ecology of later medieval England, and offer fresh perspectives on how to work within this new conceptual framework.

Bibliography Brand, P. (2000), ‘The languages of the law in later medieval England’, in D. Trotter (ed.), Multilingualism in Later Medieval Britain, Cambridge, pp. 63–76. Ingham, R. (2009), ‘Mixing languages on the manor’. To appear in Medium Aevum. Rothwell, W. (1996), ‘Playing “follow my leader” in Anglo-Norman studies’, Journal of French Language Studies 6, 177–210. Short, I. (2007), Manual of Anglo-Norman. Anglo-Norman Text Society, Occasional publications 7. London.

7

AN-1a.indd 17

11/03/2010 12:00:09

chapter two

Later Anglo-Norman as a Contact Variety of French? Richard Ingham

1. Introduction Any assessment of the status of French in England in the later medieval period has to contend with a widespread perception that it was not really genuine French. Nineteenth-century editors set the tone, considering it was merely ‘une manière imparfaite de parler français’ (Paris and Bos 1881), ‘le mauvais français qu’on parlait, et surtout qu’on écrivait, en Angleterre’ (Meyer and Toulmin-Smith 1889). Pope (1934) followed suit, declaring that in its later stages Anglo-Norman became a ‘jargon’ barely understood by those who used it. This notion of deviance or incorrectness has generally been associated with the ‘foreign-language’ status of later Anglo-Norman, characterised by Pope (1934: 424) as a ‘period of degeneracy in which insular French . . . gradually became a dead language that . . . always had to be taught’. Later writers such as Price (1984) took the same line: It is clear from the kind of French that was being written in England in the thirteenth and, even more so, in the fourteenth century that the writers had less than total command of the language . . . (it) was indeed a language in an advanced state of decline. Grammatically it was often little more than ‘bad French’. . . ‘Late Anglo-Norman is characterised by so many and such marked deviations from any other kind of French at the time as to lead one to the view that what we have before us is not just another authentic speaker French but incorrect French written by people for whom it was a foreign language and whose command of it was inadequate. (Price 1984: 224)

Kibbee (1996) similarly formulated the position in terms of an ‘essential difference’ between Continental and Insular French, which he located in the distinction between a mother tongue and a non-native language. He argued that apparent noun gender errors and syntactic constructions showing English influence provide evidence that, as time went on, Anglo-Norman became an imperfectly learned second language. With French specialists making such claims, it is no surprise to find historical linguists such as Thomason and Kaufman (1988) asserting that by the mid-thirteenth century there were ‘virtually no competent users of French’ in England.

8

AN-1a.indd 18

11/03/2010 12:00:09

Later Anglo-Norman as a Contact Variety of French? Those with close knowledge of later Anglo-Norman see this view as overstated, however. To quote Rothwell (1993: 45): ‘For a century and more from around 1260 French was not only spoken on a daily basis by large numbers of English citizens carrying out their professional duties, but, more importantly, it was written in great quantity to keep the records needed by any advanced society’. Trotter (2003a, b) likewise commented on the amply documented use of Anglo-Norman in the later thirteenth and fourteenth centuries, for purposes of communicating matters of political and financial importance both within and outside England. As Trotter points out, it is hardly plausible that those responsible for government administrative documents, charters, petitions, business letters, wills, accounts and the like would have been happy to conduct their affairs in a language they did not know very well. Rothwell (2001) drew attention to the continued functional use of Insular French well beyond the point where, according to traditional interpretations, French was supposed to be defunct. Nevertheless, the numerous differences between Insular and Continental French are real and contemporaries were well aware of them (see for example Brereton 1939), so the notion that later Anglo-Norman was merely an imperfectly learned second language will remain influential until a more fully satisfactory account is forthcoming. The remarkable growth of research into language contact in recent decades now provides an opportunity to reassess the issues. The scenario envisaged by Kibbee and others, in which English as a native language influenced French learned as a non-native language, is one form that contact influence may take, but it is certainly not the only one. Current models of language contact have thrown light on various ways in which a language whose speakers come into regular contact with those of another may develop. Recent research into contact language varieties has made available much greater empirical evidence, and a much better understanding of the parameters involved, than were available to previous commentators on Anglo-Norman. In this article we explore a contact perspective on Anglo-Norman, considering certain of its supposedly deviant grammatical characteristics which up to now have lent support to the non-native-language-learning account propagated by authors such as Price and Kibbee. The alternative possibility is considered that Anglo-Norman was maintained by bilingual speakers into the fourteenth century, rather than being taught as an instructed second language, and that contact via bilingualism was the factor that brought about departures in Anglo-Norman from Continental French. We begin by surveying common grammatical traits observed in Anglo-Norman that have been held to distinguish it from Continental French, and go on to ask whether they are in fact signs of imperfect L2 learning, or whether they might have arisen within a native-speaker minority-language variety in close contact with a majority language. Our procedure will be on the one hand to interrogate a substantial range of later Anglo-Norman materials now available electronically through the Anglo-Norman Hub, and on the other hand to make comparisons with continental varieties of French, particularly

9

AN-1a.indd 19

11/03/2010 12:00:09

The Anglo-Norman Language and Its Contexts peripheral ones, in order to illustrate what commonalities might be expected with them if Anglo-Norman was indeed a contact variety.

2. Atypical traits of Anglo-Norman morphosyntax The following have been noted as among distinctively Anglo-Norman morpho­syntactic traits; they are ‘atypical’ (Buridant 2000) with respect to Continental French: (1) i. Confused gender-marking, le for la (Hunt 2004: 29). ii. Confusion between syntactic contexts for strong and weak forms of pronouns (Grant 1978: 36–7, Johnston 1961: xix). iii. Confusion of direct and indirect object case-marking: les for leur (Grant 1978: 36, Johnston 1961: xix). iv. Second and third conjugation –ir and –re infinitive endings modelled analogically on first conjugation –er (Pope and Wright 1943: lxvii, Johnston 1961: xx). v. The use of the imperfect instead of the preterite (Buridant 2000: 368, Short 2007: 130).

The issue for us in the present article is whether to interpret these areas of grammatical difference from Continental French as evidence for imperfect non-native language learning. Initially in favour of the imperfect-learning hypothesis is the fact that the traits in (1) neutralise contrasts existing in Old French but not in Middle English. (i) Middle English did not have grammatical gender; (ii) did not distinguish direct and indirect object morphologically; (iii) had no verb conjugation classes; (iv) had no distinction between tonic (strong) and atonic (weak) pronoun forms; and (v) lacked a contrast between imperfect and preterite tenses. Grammatical contrasts drawn by an L2 grammar, but not by the L1, are generally thought to be more difficult to master than areas where grammatical contrasts in the two languages happen to coincide. Errors involving (1 i–v) might well be expected on the part of ‘fossilised’ learners, i.e. those whose ultimate attainment in the L2 stopped short of target-like mastery (see for example Han 2004, DeKeyser 2005). Given the deviations from continental practice reported by Anglo-Norman text editors that we have noted above, the argument for seeing later AngloNorman as an imperfectly learned second language is bound to seem persuasive. In this study it will be shown, however, that all these features would also be expected under a scenario where French continued to be a primary language of a bilingual minority, within a community speaking English as a majority language. Secondly, the co-existence of these features with evidence of accurate acquisition of other formal contrasts will be reviewed, a situation which makes a bilingual contact-language interpretation at least as a plausible an overall account of the language ecology of later medieval England as the conventional story of imperfect L2 learning.

10

AN-1a.indd 20

11/03/2010 12:00:09

Later Anglo-Norman as a Contact Variety of French? 3. Language contact As a working definition of a contact variety, let us adopt the position that it is a language variety (which may, but need not, be learned as a mother tongue) influenced by a socially dominant language, where social dominance can take the form either of demographic weight or of cultural prestige. It makes some use of formal patterns calqued upon forms of the dominant language which speakers of the contact variety regularly encounter in their ordinary language use. Such formal patterns are not found in non-contact varieties of the language in question. Contact varieties arise in different ways. A helpful typology of contact situations is found in Winford (2003), who refers to the variety undergoing change by contact as the Target Language: Language maintenance: Source Language (SL) and Target Language (TL) are maintained. ▪ Borrowing: TL borrows lexical items from a SL, leaving core linguistic systems unaffected. ▪ Structural convergence: ‘In cases involving bi- or multilingualism within the same speech community, the results of language convergence are often manifested in increasing structural convergence between the languages involved’ (Winford 2003: 13). Language shift: abandonment of Source Language in favour of Target Language ▪ TL is adopted as an L1 by incoming group, who may or may not abandon the original L1; TL may be influenced by incoming group’s L1. ▪ TL is introduced into a new geographical area by colonisers, adopted as an L1 by formerly L2 speakers, and shows considerable L1 substrate effects, e.g. Hiberno-English. In the case of Anglo-Norman, the conventional analysis (see for example Baugh and Cable 2002) is in terms of the second of the shift scenarios: the French speakers are supposed to have abandoned French and taken up English as their everyday vernacular. This accounts for the eventual elimination of French as a functioning language in England, but does not address the sustained use of French in England for three centuries and more after the Conquest. By the later twelfth century, there is historical evidence that people of Norman and of English origin were no longer readily distinguishable (Short 1995), thanks to intermarriage between these ethnic groups, so a plausible period in which to date this shift from French to English might seem to be the twelfth century. Far from coinciding with a sharp reduction in French use resulting from a move to English, the thirteenth and fourteenth centuries saw a striking expansion of French into a whole range of written genres for which it had not previously been used in England, hardly a sign of a shift away from the language. An alternative scenario, to be considered in this study, is therefore to approach the issue of contact between Anglo-Norman and English in terms of the secondlanguage maintenance scenario, lasting into the fourteenth century. In view

11

AN-1a.indd 21

11/03/2010 12:00:09

The Anglo-Norman Language and Its Contexts of the well-known divergences from Continental French during this period, Anglo-Norman cannot be said to have remained ‘structurally unaffected’, as in the first of Winford’s two contact-with-maintenance scenarios. Hence its status will be evaluated on the basis of whether it can be shown to have displayed structural-convergence features. It will be considered whether, on the basis of other documented cases of structural convergence, these features could have arisen as part of the natural development of a native variety in contact with another language, rather than being second-language learner errors.

4. Structural convergence in contemporary contact varieties There is now a large and rapidly expanding literature on language contact showing grammatical convergence affecting a variety spoken as a native language. Let us consider some examples taken from recent work in this domain of grammatical processes that take place affecting the morphosyntax of a language variety in contact with another language. Clyne (2003: 124) shows how syntax may ‘converge’ in bilingual usage in Australia, with, for example, emigrant German verb positions adopting English order. Likewise, morphological simplification takes place, e.g. the accusative case is used for the dative in US emigrant varieties of German (Clyne 2003) and French (Maher 1991). Similarly, emigrant Dutch speakers in Australia overgeneralised one article form, while emigrant German speakers in Australia overgeneralised one gender (neuter) for inanimates (Clyne 2003: 122). In their study of contact French in Ontario, Mougeon and Beniak (1991: 97) observed some neutralisation of the 3sg./3pl. distinction marking on the present tense of verbs, especially mettre, pouvoir, vivre, connaître. Jones (2005) has shown that native speakers of Jèrriais may neutralise noun-gender agreement on adjectives. In all these cases, the contact variety is1 acquired as a home language transmitted to children by primary caregivers, so the grammatical traits in question cannot be seen as foreign-language-learner errors. Other forms of innovation taking place in a contact variety are described in Heine and Kuteva (2005). A Croatian minority language in Italy has a new auxiliary expressing necessity and future, unlike homeland Croatian, but conforming to the model of Italian. The Maltese preposition fuq, which originally meant ‘on’, in a spatial sense, now replicates the use of the English preposition on in Maltese expressions meaning ‘on television’, ‘on holiday’, ‘on the advice of’ etc. Pennsylvania German speakers have extended their use of the ‘have’ perfect at the expense of the ‘be’ perfect, following the model of English. Likewise, native speakers of Swedish in Finland have developed an interrogativeparticle construction which uses an initial complement åm meaning ‘whether’. This replicates a pattern of Finnish, which has an interrogative particle. It is important to note that Heine and Kuteva do not draw a fundamental distinction between native and second language status in their discussion of contact1



Or else was, in the case of Jèrriais, now a dying language.

12

AN-1a.indd 22

11/03/2010 12:00:09

Later Anglo-Norman as a Contact Variety of French? induced changes. They discuss contact-induced phenomena in a second language, for example where Singaporean speakers of English use a possession verb to denote existence, as in their L1, Chinese. The same process can occur in a target variety regardless of its native or non-native status: American Indian language speakers have created an evidential morpheme in their L2 English, modelled on the expression ‘they say’. And Bulgarian L1 speakers also have an evidential morpheme, which is derived from contact with Turkish. In short, there is plentiful contemporary evidence of the influence of a linguistically dominant source variety on the grammar of a contact variety, whether or not imperfect foreign-language learning is involved. In the next section, we apply this perspective to later Anglo-Norman.

5. Structural convergence in Anglo-Norman 5.1 gender-marking Erratic gender-marking on articles and modifiers has often been considered a sign of the ‘degeneracy’ and ‘deviation’ that has been held against later AngloNorman. Short (2007: 123) takes a more circumspect view, attributing the variability to phonetic factors, especially assimilation of la to le, as in Picard. However, aberrant gender-marking in later Anglo-Norman undoubtedly went beyond article use. As argued by Ingham (this volume), it seems to have been a matter of the wholesale loss of the weak ‘e’ sound as part of the phonemic system. From later-thirteenth-century texts onwards, the letter is no longer reliably used to mark gender contrasts on adjectives and other modifiers such as nul (‘no’) and cest (‘this’), e.g.: (2) a. Ceus Templeris funt arrer e semer saunt nul maner de en cloysture.   PROME 1290 ‘The same Templars have had them ploughed and sown without any kind of enclosure.’ (2) b. Lexecucion du jugement renduz en la court le roi meintenaunt sanz nule delay se deit faire.  PROME 1307 ‘The execution of the judgment given in the king’s court ought to be carried out immediately without any delay.’ (3) a. En temoynaunce de cest chose nus avoms fet fere noz lettres patentes. PROME 1291 ‘In evidence of this matter we have had made our letters patent.’ (3) b. E out le vant maner, e plusours terres ke ceste Robert tent uncore.  PROME 1277 ‘And had the aforesaid manor and several other lands which this Robert still holds.’

13

AN-1a.indd 23

11/03/2010 12:00:09

The Anglo-Norman Language and Its Contexts The loss of gender-marking by the presence or absence of the grapheme , however, cannot be taken as proof that users of Insular French were acquiring the language as second-language learners, and failing to learn noun gender accurately. As noted by Pope (1934: 437), the weak ‘e’ sound was also lost in Old Walloon, and was often omitted graphemically as well. A Walloon text such as the Temporaliteit of de Hemricourt (later fourteenth century) well illustrates this tendency. As in Anglo-Norman, agreement on modifiers appears to be graphemically random, forms with or without final appearing indifferently with both masculine and feminine nouns, as in the following examples from Temporaliteit: (4) Tout la terre 4, Tout le bien 8, Chest noble citeit de Liege 17, Chest homaige 209 ‘all the land’ ‘all the property’ ‘this noble city of L.’ ‘this homage’ (5) Toute la dicte maison 95, Toute le paais 200, De nulle cas queilcunque 38, Nulle personne queilcunque 44 ‘All the said house’ ‘All the country’ ‘ Of no case whatever’ ‘No person whatever’

Contracted forms such as al and del similarly ignored noun gender: (6) Del creation d’un noveal evesque 25, Del temps 43 ‘Of the creation of a new bishop?’ ‘Of the time’ (7) Al loy de Liege 52, Ly emperreur . . . revenroit al paiis 21 ‘To the law of L.’ ‘The emperor would come back to the country’

Thus although Walloon was a native-speaker variety of medieval French, in its written form it ceased to distinguish noun gender agreement reliably where graphemic forms including weak ‘e’ were involved. Pope (1934: 437) attributed the loss of weak ‘e’ to the retention of heavy word stress in AngloNorman and Walloon. There also seems to be a commonality in that both were in contact with Germanic languages, which had heavy word stress. According to Geschière (1950: XI), Walloon syntax likewise displays the influence of Dutch, which would suggest that it can be considered as to that extent a contact variety. In any case, if we simply follow Pope’s phonological account, the effect of heavy word stress operated regardless of whether the affected variety was unquestionably a native language (Walloon) or – supposedly – not (Anglo-Norman).

5.2 tonic and atonic object pronouns Anglo-Norman texts offer instances where the strong form of a personal pronoun precedes a finite verb, e.g.: (8) Et si le bercher sey poet aquiter  Sénéschaussée ch. 27 ‘And if the shepherd can clear himself’

14

AN-1a.indd 24

11/03/2010 12:00:09

Later Anglo-Norman as a Contact Variety of French? Such constructions certainly diverge from central French usage, which required the clitic form of the object pronoun. However, the same trait can be found in various native-speaker dialects, e.g. (Ingham, forthcoming): (9) a. Je moy ay mis al repouz.  Hesbaye (c. 1380) (Walloon) ‘I put myself to rest.’ (9) b. Jo tei preio.  Stimm 10.13 (c. 1250 ?) (Franco-Provençal) ‘I beg you.’ (9) c. Dous quaus deners le dit vendeor ssey tint pleinerement por paiez. Poitou I, 50 (1295) (Poitevin) ‘Of which moneys the said vendor holds himself to be fully paid.’

All such cases where we have been able to locate examples come from peripheral areas of the medieval French-speaking domain, as in (9). Walloon, FrancoProvençal and Poitevin were all spoken in language communities where French bordered on other languages, in particular Germanic and Provençal, where there was no contrast between tonic and atonic forms of object pronouns, at least in the 1sg., 2sg. and 3sg. forms where tonic and atonic reflexes of Latin me, te, se developed. The texts providing examples in (9) never used eux in place of les in the 3pl., nor did the Anglo-Norman texts discussed in Ingham (forthcoming).2 We may understand the phenomenon better if we place it in a wider historical context. The tonic/atonic object personal pronoun distinction seems to have evolved as a feature of French, rather than of other Romance varieties, and became absolute in central French. In common with peripheral Frenchspeaking areas on the continent, Anglo-Norman seems to have retained the possibility for either form to be used in prefinite position. Cases such as (9) accordingly do not demonstrate imperfect non-native language learning. The fact that the use of the strong form here is attested in early Anglo-Norman texts, according to Short (2007: 125), also argues against seeing it merely in terms of confusion by English second-language learners. Its continued presence in peripheral varieties of Continental French suggests that it is a candidate for a contact-influenced form.

5.3 pronoun case The Modern French personal pronoun system distinguishes 3pl. direct object forms le/la from indirect object lui, and the 3pl. direct object form les from indirect object leur. Old French had essentially the same system. In later AngloNorman, instances of the direct object used for the indirect object form are found, e.g.:

2

As argued by Ingham (2007), pre-finite lui instead of Central French le in AngloNorman texts appears for lu, an alternative clitic object form found in peripheral varieties of Old French. In later Anglo-Norman the spelling form was commonly used for /u/.

15

AN-1a.indd 25

11/03/2010 12:00:10

The Anglo-Norman Language and Its Contexts (10) E jeo trovai le clerc le weucunte . . . e jeo le demaundai ce ke il aveit receu de la vile. H III 320 ‘And I found the sheriff’s clerk . . . and I asked him what he had got from the town.’ (11) Et le rei les dona truage de xvi. M. liveres.  Le livere de reis de Engleterre (late 13th century?) ‘And the king gave them tribute of £16.’

Using the direct object form for the indirect object can be seen as case neutralisation. It could be argued that it constitutes evidence of English L1 influence, since in English the accusative/dative pronoun distinction no longer existed, so the direct object forms in these examples could be seen as L2 transfer errors. However, they also correspond to the accusative-for-dative case uses we have seen above in contact-variety German, especially in situations where the majority language is English. They are thus interpretable as contact-variety features, in line with the trend towards simplification of morphological contrasts in such varieties.

5.4 conjugation class Anglo-Norman had a reported tendency to extend the –er ending of the first conjugation to other classes. Examples taken from the Anglo-Norman Hub textbase are: (12) a. . . . tanqe les ditz creditours serront molt leez de prender une petite parcelle de lour dette & relesser le remanant.  Statutes I, 398 (1376–7) ‘Whereas the said creditors will be most happy to take a small portion of their debt and release the remainder.’ (12) b. . . . les verreis heirs, qi serrount engittez de tenementz, qe eus cleimerount tener par la ley de Engleterre  Britton I, 289 (c. 1300) ‘. . . the true heirs, who will be cast out of tenements that they will claim to hold by the law of England’

These could be seen as failures on the part of learners of French to assign a verb to the correct conjugation class. However, they are also in line with the tendency in contact varieties mentioned in section 4 above to reduce allomorphy. It is not clear from the existing literature how strong this tendency actually was in later Anglo-Norman: did it amount to a wholesale abandonment of conjugation distinctions, or was it a matter of merely occasional divergence which left them intact most of the time? In order to address this question, use was made of the Anglo-Norman Hub textbase, so as to establish how commonly second and third conjugation infinitives were remodelled on the first conjugation ending. Prose texts from the later Anglo-Norman period (later thirteenth century to 1400) were selected for this purpose. Frequencies of the ten most common –re and –ir verbs used in this textbase with the conventional inflexion were compared with frequencies of overgeneralisation to the first

16

AN-1a.indd 26

11/03/2010 12:00:10

Later Anglo-Norman as a Contact Variety of French? conjugation infinitive inflexion. All spelling variants of the relevant verbs identified by the search procedure were included, e.g. acumplir, acomplir etc. The results are shown in Table 1. Table 1. Frequencies of overgeneralised forms and conventional forms in later Anglo-Norman administrative texts and chronicles, ten most common simplex second and third conjugation verbs, Anglo-Norman Hub textbase, prose works 1250–1400 -er inflexion prender

-re/-ir inflexion 4

total

prendre

1335

1339

1103

1119

metter

16

mettre

responder

10

respondre

472

482

render

6

rendre

573

579

vender

10

vendre

357

367

sub-total

46 (2.4%)

vener

44

tener

155

3,840 (97.6%)

1,886

venir

1644

1688

tenir

1661

1816

accompler

22

accomplir

338

360

server

22

servir

488

510

parter

2

partir

234

236

sub-total

245 (5.3%)

4,365 (94.7%)

4,610

overall total

291 (3.4%)

8,205 (96.6%)

8,496

Evidence that the conjugation-class distinctions were being eroded in later Anglo-Norman, though it does exist, seems slight, approaching 10% of occurrences only with the verb tenir, and in most other cases being negligible. On this basis, departures from continental conjugation class membership were not at all common in later Anglo-Norman. At over 96% overall, the very high level of accuracy with –re and –ir infinitives is noteworthy, and belies an account of later Anglo-Norman as grammatically chaotic. It seems compatible with a contact account in showing sporadic morphological simplification, against a background of generally correctly acquired formal distinctions.

5.5 preterite and imperfect in anglo-norman In Old French verbal marking distinguished a punctual past-time event (preterite tense) from a past-time event seen as having duration (imperfect tense):3 3



This is a somewhat oversimplified account: other factors such as discourse foregrounding and backgrounding also seem to have been present.

17

AN-1a.indd 27

11/03/2010 12:00:10

The Anglo-Norman Language and Its Contexts (13) a . . . li traitor,/ Ki nus jugat devant l’empereur  Ch. Roland 1024–5 ‘The traitor, who accused us before the emperor’ (13) b. Jo atendeie de te bones noveles.  St Alexis 479 ‘I was waiting for good news from you.’

The verb estre formed an exception, in that the preterite and imperfect could be used interchangeably to denote a past state of affairs (Foulet 1930: 224). In Anglo-Norman, the imperfect was sometimes used to denote a punctual past-time event, as noted by Buridant (2000: 368 ff.), who labelled this the ‘imparfait atypique’. Tense confusion is a well-known L2 learner problem, and the atypical imperfect in Anglo-Norman can plausibly be seen as a matter of inadequate learning, especially since medieval English did not distinguish punctual and durative aspect in past-time events; the –ed(e) tense inflexion was used for both, e.g.: (14) a. (durative) Al folc him luuede.  ASC 1140 ‘Everyone loved him.’ (14) b. (punctual). . . & bebyried him heglice in þe minstre.  ASC 1137 ‘. . . and buried him solemnly in the cathedral’

As foreign-language learners usually experience difficulty where the foreign language makes a distinction not existing in the L1, errors on the part of AngloNorman users acquiring French as a non-native language could certainly have been expected here. The extent to which later Anglo-Norman texts used the atypical imperfect has not to our knowledge been systematically investigated. To deal with the issue, we need to be able to interpret the meaning of a verb form as denoting a completed versus an incomplete/habitual event. Supposing we were dealing with a Modern French text written by a non-native speaker, to assess the appropriateness of e.g. Paul mangeait une pomme or Paul mangea une pomme depends on knowing the meaning intended, for which evidence would be sought in the context. In roughly the same way, we sought contextual cues in the Anglo-Norman texts from adverbs favouring a completed event reading that a verb form should be given this interpretation. Three common adverbs were selected for this purpose: puis, lors, and meintenant (the last of which meant ‘straightaway’ in Old French). They were normally used to introduce a new narrative episode, and thus strongly favoured a foregrounded complete-event reading rather than a background ongoing event or a habitual event, the typical uses of the imperfect in Continental Old French. They could thus be taken as contexts that would typically elicit the preterite tense. Texts which would be likely to include the frequent use of these three adverbs were selected. They were seven narrative prose texts, mostly chronicles, with composition dates ranging from c. 1250 to the late fourteenth century. They were searched using the concordancer facility of the Anglo-Norman Hub; using this procedure, 250 past-referring contexts were identified. Of these, 240 uses of the preterite were

18

AN-1a.indd 28

11/03/2010 12:00:10

Later Anglo-Norman as a Contact Variety of French? recorded, as against ten uses of the imperfect. Of seventy-eight verb types used overall, only five occurred with these adverbs in the imperfect, four of them voloir, povoir, savoir, avoir,4 which in fact may co-occur in Old French with lors in a stative sense. Thus far, it can be said that the choice of tense form was highly native-like. However, the Anglo-Norman texts displayed imperfect as well as preterite forms of the estre auxiliary in these passive clauses, e.g.: (15) a. Et lors estoit sertein jour assigné.  Cron Lond 71 ‘And then a particular day was assigned.’ (15) b. Monsire William de Mountagu, qi puis estoit fait counte de Salesburi Anon Chr2 168 ‘Sir W. de M., who later was made Count of Salisbury’ (16) a. . . . ki engendra de ly Henri, ke fu pus fet rey de Engletere  Reis Britt 180 ‘Who begat from her Henry, who later was made king of England’ (16) b. Et lors furent maundez messagers à le . . .   Cron Lond 70 ‘And then messengers were sent to the . . .’

Almost all the adverb contexts of the present study were for actional passives, since the three target adverbs tended to be used as discourse markers introducing new events in a narrative, rather than backgrounded states. In this construction, imperfect and preterite were not interchangeable in Continental French: a stative passive was formed with the imperfect of estre and an actional passive with the preterite. Four instances of the imperfect with estre were observed, as in (15), out of twenty-four actional passive contexts, the other cases having the preterite as in (16). The clearly atypical uses of the imperfect were thus limited to these passive constructions. Since Anglo-Norman users never confused the preterite and the imperfect with over seventy ordinary verb types, faulty learning of the Old French tense system by L2 learners does not offer a satisfactory account of these results. In an L2 setting, learners struggle to master aspectual uses in the target language across a whole range of verb types, not just one. On the other hand, the imperfect/preterite alternation with estre as an actional passive auxiliary was admittedly ‘atypical’, in Buridant’s (2000) sense, and requires explanation. In our view, it can be seen as a contact phenomenon, in which a target language intensifies a pre-existing trait, under the pressure of a source language. The trait in question was the alternation between preterite and imperfect in Old French in the case of stative estre, as mentioned above. Contact with English, in

4

A single instance of a non-stative verb, jurer, was used with the imperfect. However, the imperfect is commonly used in Old French with verbs of communicating, such as dire, repondre, clamer etc., where modern French would avoid it in a foregroundedevent meaning. The imperfect found with jurer can therefore not be taken as a divergence from Old French.

19

AN-1a.indd 29

11/03/2010 12:00:10

The Anglo-Norman Language and Its Contexts which there was but a single past form, favoured an extension of the imperfect tense beyond the stative cases to an actional passive sense of estre where in Continental French the alternation did not apply.

6. Contact influence of Anglo-Norman grammar on English: the modal perfect The influence of French on Middle English grammar has been minimised by those, such as Thomason and Kaufman (1988), who seek to locate contact influence on English largely in the lexical domain.5 Clearly, English was demographically dominant, so it can plausibly be argued that the linguistic structures of the vast majority of the people would have been resistant to French. However, the prestige of French can be seen as exerting a kind of cultural dominance in reverse, so perhaps allowing linguistic structures to be permeable to French influence via contact convergence, as outlined above. In this section, we should like to renew consideration of potential French influence on English grammar, occurring at a time which appears to have fallen within the later Anglo-Norman period, from 1250 onwards, according to available textual evidence. In Early Middle English, according to Molencki (2001), the apodosis clause of counterfactual past conditional sentences consistently avoided the Modern English modal + auxiliary combination, in favour of the past perfect, i.e. the past tense of auxiliary have + past participle, e.g.: (17) & ʒef an miracle nere . . . . ha hefde iturpled wið him . . . . dun into helle grunde. Anc. Riw II.195.2804 ‘And were it not for a miracle she would have toppled with him down to the bottom of hell.’

The pattern exemplified by hefde iturpled is retained in other Germanic languages to this day, though they tend to use the past subjunctive with their have auxiliary. Molencki showed that in southern Middle English around 1300 a different pattern developed in this context, featuring the modal perfect construction, i.e. a modal + non-finite auxiliary have, e.g.: (18) a. He wald haue forced me in . . .   Cursor Mundi MS C 4399 (southern) ‘He would have forced me in . . .’ (18) b. Wold he neuer haue ghyuen to rede . . .   Cursor Mundi MSS T and A 10788 (southern) ‘He would never have given his advice’ 5



Though French influence on English grammar was traditionally acknowledged, e.g. by Mustanoja (1960), regarding the of genitive versus the –’s genitive.

20

AN-1a.indd 30

11/03/2010 12:00:10

Later Anglo-Norman as a Contact Variety of French? This is essentially the present-day English structure. However, northern manuscripts of the same poem retained the past perfect in the apodosis, e.g. (cf. 6b): (19) Ne had he neuer gyuen to rede . . .   Cursor Mundi MS G 10788 (northern) ‘He would never have given his advice . . .’

Fischer (1992: 257) claimed ‘it is unlikely that either Latin or French played any role’ in the development of the modal perfect construction in English. This is a surprising assertion, since in fact Old French commonly used the modal + perfect infinitive with avoir to express counterfactual past notions, e.g.: (20) . . . que je ne le deüsse pas avoir refuse (Joinville sec. 426) ‘. . . that I should not have refused it’  Brunot and Bruneau (1949)

We have found numerous further examples from the later thirteenth century, e.g.: (21) b. Il deussient avoir receu les deniers . . .   Haute Marne 77 (c. 1270) ‘They should have received the money . . .’ (21) a. Se je le demandoie par les resons que je peusse avoir dites devant le jugement . . . .g’irois contre le jugie’  Beaumanoir I 126 (c. 1280) ‘If I asked for it according to the reasons I could have said before the hearing . . . I would go against the sentence.’

Anglo-Norman counterfactuals with a modal and perfective avoir are attested as early as the mid-twelfth century: (22) a. Ke vus dussez aver dyst issy adeprimes  Jeu d’Adam 31 (c. 1150) ‘That you should have said this first’ (22) b. Ke nus ne purreiom aver heu autre bref  Jeu d’Adam 47 (c. 1150) ‘That we could not have had another writ’

They continue to be commonly used in later Anglo-Norman: (23) E pout adunke aver change sy yl vousit, e ne fit nent.  Petitions 1292 p. 53 ‘And could then have changed if he wanted, and did nothing.’

Old English lacked the modal and auxiliary have combination, so we see its appearance beginning in southern Middle English as a clear candidate for contact influence from French. Examples (22 a–b), showing that the source structure was present in Insular French at the beginning of the Middle English period, provide support for that scenario. It is particularly significant that the innovation, unlike most historical changes in English grammar, seems to have spread from the south to the north, given that the use of Anglo-Norman was more prevalent in the south

21

AN-1a.indd 31

11/03/2010 12:00:10

The Anglo-Norman Language and Its Contexts than in the north of England. Molencki’s work indicates that the modal perfect form had not yet affected the language in the early thirteenth century, but was a later Middle English development. Anglo-Norman influence on English in this period is not expected on conventional accounts of the medieval language ecology in England, which assume that by then Anglo-Norman was an instructed second language. On a contact account, on the other hand, the phenomenon can be understood: bilingual Anglo-Norman speakers, though numerically inferior, enjoyed cultural dominance, so their use of a Frenchbased structure in their English would plausibly have been regarded as a prestige variant to be adopted by the wider English–speaking community. It seems certain that this mechanism for the spread of contact induced change from French is the one that applied in the case of idiomatic French-based lexis, such as before hand calqued on devant/avant la main (Orr 1962), so we would propose the same route for the spread of the English modal perfect.

7. Summary and conclusions Incomparably more is known about what to expect when languages come into contact than a generation ago. In this article we have used some of the insights from studies of contact varieties to reconsider the status of later AngloNorman, examining it in relation to two broadly defined models,6 one as an imperfectly learnt instructed L2, the other as a contact variety. Our argument has been that the well-known Anglo-Norman grammatical phenomena listed in section 2 do not show that it was an imperfectly acquired L2. They are also consistent with a contact account in that the types of grammatical variation are found in emigrant native-speaker varieties of modern languages, and/or in native-speaker varieties from areas where medieval French was in contact with other languages. In addition, there are points of later medieval English grammar such as the counterfactual modal perfect which can be attributed to French influence, not at all an expected outcome if French was an artificially acquired second language. For Anglo-Norman to have influenced English, its users must surely have been balanced bilinguals with a high degree of competence in Insular French. They not only enjoyed high social prestige, but were also assured and fluent speakers of the language, whose French-influenced English presumably came to be diffused down through other social levels by imitation, as is typical of the sociolinguistic spread of prestige variants. We thus arrive at a view of later Anglo-Norman users that asserts their linguistic strengths as bilinguals, who were influenced by English but whose command of French was deeprooted and instinctive enough for it to serve as the source of some contact6



Numerous other supposedly English-influenced features of Anglo-Norman could be considered in this light, e.g vouloir as a future auxiliary paralleled in the Suisse Romande Tu veux tomber ‘Tu vas tomber’ (Knecht 1978: 252).

22

AN-1a.indd 32

11/03/2010 12:00:10

Later Anglo-Norman as a Contact Variety of French? induced structural change in English. This approach raises questions about how Anglo-Norman continued to be transmitted in the thirteenth and fourteenth centuries, if it was not ‘a dead language that always had to be taught’, to cite Pope again. We have elsewhere (Ingham 2007) outlined an account of how this may have taken place. For the present, we seek merely to present the case in favour of viewing later Anglo-Norman as a contact variety of medieval French, showing the influence of English but also to some degree itself influencing English. The departures from Continental French that it exhibited are consistent with its having continued as a naturally acquired dialect for much longer than has generally been supposed.

Primary sources ASC: Anglo-Saxon Chronicle, MS E (Peterborough Chronicle), ed. S. Irvine. Woodbridge, 2004. Ch. Roland: La Chanson de Roland, ed. J. Bédier. Paris, 1937 Jeu d’Adam: Le jeu d’Adam, ed. W. Noomen. Paris,1971. Beaumanoir: Les coutumes de Beauvaisis, ed. J. Renouard. Paris, 1842. H III: Royal and Other Historical Letters Illustrative of the Reign of Henry III, ed. W. Shirley, vol. II, 1236–1272, Rolls Series 27, London, 1866, pp. 320–2 (1268) Haute Marne: Documents linguistiques de la France: Haute-Marne, ed. J.-G. Gigot. Paris. 1974. de Hemricourt, Temporaliteit, at http://membres.lycos.fr/chockier/FILES/INSTITU/ Tempo.html Hesbaye: Jacques de Hemricourt, Le Miroir des nobles de Hesbaye, ed. C. de Borman. Bruxelles, 1910. Poitou: Chartes et documents poitevins du XIIIe siècle en langue vulgaire., ed. M. La Du. Archives historiques du Poitou, 1903. St Alexis: La vie de Saint Alexis, ed. M. Perugi. Geneva. 2000. Sénéschaussée, in Walter of Henley and Other Treatises on Estate Management and Accounting, ed. D. Oschinsky. Oxford, 1971. Stimm: Altfrankoprovenzalische Uebersetzungen hagiographischer lateinischer Texte: 1. Prosalegenden, ed. H. Stimm. Wiesbaden, 1955. The following text files from the Anglo-Norman Hub were accessed at http://www. anglo-norman.net: Anon Chr2: Anonimalle Chronicle 1333–81, ed. V. Galbraith. Manchester, 1927. Britton: The French Text Carefully Revised with an English Translation, Introduction and Notes, ed. Francis Morgan Nichols, 2 vols. Oxford, 1865. Cron Lond: Chroniques de London depuis l’an 44 Hen. III jusqu’à l’an 17 Edw. III, ed G. Aungier. Camden Society. London, 1844. Reis Britt: ‘Le livere de reis de Engleterre’: Le livere de reis de Brittanie e le livere de reis de Engleterre, ed J. Glover. Rolls Series 42. London, 1865. Statutes: Statutes of the Realm, vols. I and II. Record Commission. London, 1810–25.

23

AN-1a.indd 33

11/03/2010 12:00:11

The Anglo-Norman Language and Its Contexts Bibliography Baugh, A., and T. Cable (2002), A History of the English Language. 5th ed. London. Brereton, G. (1939), ‘Some grammatical changes made by two revisers of the Anglo-Norman version of Des grantz geantz’, in Studies in French Language and Mediæval Literature Presented to Mildred K. Pope by Pupils, Colleagues and Friends. Manchester. Brunot, F., and C. Bruneau (1949), Précis de grammaire historique de la langue française. 3rd ed. Paris. Buridant, C. (2000), Nouvelle grammaire de l’ancien français. Paris. Clyne, M. (2003), Dynamics of Language Contact: English and Immigrant Languages. Cambridge. DeKeyser, R. (ed.) (2005), Grammatical Development in Language Learning. The Best of Language Learning Series. Oxford and Malden, Mass. Fischer, O. (1992), ‘Syntax’, in N. Blake (ed.), The Cambridge History of the English Language, vol. II, 1066–1476, Cambridge, pp. 207–408. Foulet, L. (1930), Syntaxe de l’ancien français. Paris. Geschière, L. (1950), Élements néerlandais du wallon liègeois. Amsterdam. Grant, J. (éd.) (1978), La passiun de Seint Edmund Anglo-Norman Text Society 36. London. Han, Z. (2004), Fossilization in Adult Second Language Acquisition. Clevedon. Heine, B., and T. Kuteva (2005), Language Contact and Grammatical Change. Cambridge. Hunt, T. (2004), Le chant des chanz. Anglo-Norman Text Society 61–2. London. Ingham, R. (2007), ‘Bilingualism and language education in medieval England’, Seminar, World Universities Network, October 2007. Summary at http://www. wun.ac.uk/multilingualism/seminar_archive/07_08_program/ingham.html ——(forthcoming), ‘L’anglo-normand et la variation syntaxique en français medieval’, in P. Danler (ed.), Actes du XXVème CILPR. Innsbruck. Johnston, R. (ed.) (1961), The Crusade and Death of Richard I. Anglo-Norman Text Society 17. Oxford. Jones, M. (2005), ‘Transfer and changing linguistic norms in Jersey Norman French’, Bilingualism: Language and Cognition 8/2, 159–75. Kibbee, D. (1996), ‘Emigrant languages and acculturation: the case of AngloFrench’, in H. Nielsen and L. Schøsler (eds.), The Origins and Development of Emigrant Languages, Odense, pp. 1–20. Knecht, P. (1978), ‘Le français en Suisse romande: aspects linguistiques et sociolinguistiques’, in A. Valdman (ed.), Le français hors de France, Paris, pp. 249–58. de Kok, A. (1985), La place du pronom personnel régime conjoint en français. Une étude diachronique. Amsterdam. Maher, J. (1991), ‘A cross-linguistic study of language contact and attrition’, in H. Seliger and R. Vago (eds.), First Language Attrition, Cambridge, pp 67–84. Meyer, P., and L. Toulmin-Smith (1889), Les contes moralisés de Nicole Bozon. Société des anciens textes français. Paris. Molencki, R. (2001), ‘Counterfactuals in the different manuscripts of the Cursor Mundi’, Neuphilologiche Mitteilungen 102/1, 11–22. Mougeon, R., and E. Beniak (1991), Linguistic Consequences of Language Contact and Restriction. Oxford.

24

AN-1a.indd 34

11/03/2010 12:00:11

Later Anglo-Norman as a Contact Variety of French? Mustanoja, T. (1960), A Middle English Syntax. Helsinki. Orr, J. (1962), Old French and Modern English Idiom. Oxford. Paris, G., and A. Bos (1881), Vie de saint Gilles par Guillaume de Berneville. Société des anciens textes français. Paris. Pope, M. (1934), From Latin to Modern French, with Especial Consideration of AngloNorman: Phonology and Morphology. Manchester. ——and J. Wright (1943), La seinte resureccion. Oxford. Price, G. (1984), ‘Anglo-Norman’, in G. Price (ed.), The Languages of Britain, London, pp. 197–206.. Rothwell, W. (1993), ‘The legacy of Anglo-French: faux amis in French and English’, Zeitschrift für romanische Philologie 109, 16–46. ——(2001), ‘English and French in England after 1362’, English Studies 82, 539–59. Short, I. (1995), ‘Tam Angli quam Franci: self-definition in Anglo-Norman England’, Anglo-Norman Studies 18, 153–75. ——(2007), Manual of Anglo-Norman. Anglo-Norman Text Society Occasional Publications Series 7. London. Thomason, S., and T. Kaufman (1988), Language Contact, Creolization, and Genetic Linguistics. Berkeley. Trotter, D. (2003a), ‘Not as eccentric as it looks: Anglo-French and French French’, Forum for Modern Language Studies 39: 427–38. ——(2003b), ‘L’Anglo-Normand: variété insulaire ou variété isolée? Grammaires du vulgaire’, Médiévales 45, 43–54. Winford, D. (2003), An Introduction to Contact Linguistics. Oxford.

25

AN-1a.indd 35

11/03/2010 12:00:11

chapter three

The Sources of Standardisation in French – Written or Spoken? R. Anthony Lodge

1. Introduction For sociopolitical reasons Anglo-Norman was related closely to the Norman dialect during the eleventh and twelfth centuries. For commercial reasons it was also closely related to Picard, at least until the end of the thirteenth century. The importance of the Parisian dialect for Anglo-Norman developed only in the thirteenth century, which must be related to the fact that, by this time, Paris had emerged as the largest conurbation in western Europe; the power and wealth concentrated there gave its form of speech unrivalled prestige. Although it took several centuries for Parisian written forms to completely displace regional writing systems in the French provinces, it started to have a formative influence in the late thirteenth century and eventually of course became the French standard language. In order to appreciate the impact the French of Paris had on later Anglo-Norman, it is important to understand something about how it emerged. This question raises issues which are highly contentious in France, thanks to the role of the standard language in definitions of French identity and nationhood. Did the embryonic standard language develop initially in writing, and defuse subsequently through Parisian speech before moving out to the provinces? Or did it crystallise initially in Parisian speech before being reflected in the Parisian writing system? For reasons which are not clear, it took Paris a long time to develop a writing system of its own: very few non-Latin texts, known to have been produced in Paris, survive from before the mid-thirteenth century. This paper addresses those alternatives, investigating whether speech or writing was the matrix of the embryonic standard French language. The origins of standardisation in French raise some of the most refractory problems in the history of that language. In the case of Italian, German and, most spectacularly, modern Hebrew, analysts agree that a central role was played in the initial focusing of standard-language norms by a literary or written form of the language – the works of Dante in the case of Italian, Luther’s translation of the Bible in the case of German, and the Bible and Talmud in the case of Hebrew. Is this equally true of French? It is difficult to conceive of

26

AN-1a.indd 36

11/03/2010 12:00:11

The Sources of Standardisation in French – Written or Spoken? standardisation without the development of writing, but, in the case of French, did a supra-regional written standard emerge prior to the development of spoken norms, or was the whole process of standardisation grounded initially in a particular spoken dialect?1 It is a subject clearly relevant to this volume, as it offers something of an overview of the dialect situation in medieval (thirteenth-century) French, contextualising more localised studies presented later. At the same time it must be acknowledged that we cannot do justice to every aspect of the problem within the confines of the space available here.

1.1 the dialects of medieval gallo-romance Let us begin with an overview of the dialects of medieval Gallo-Romance. Histories of the French language routinely open with a map showing the whole of Gaul divided into three linguistic parts – Langue d’Oil, Langue d’Oc and Franco-Provençal (see for example Chaurand 1999: 36). Maps like this make indispensible descriptive tools, but we must be careful how we use them in the reconstruction of past states of the language. Strictly speaking, Chaurand’s map provides an image of the dialect situation not in medieval times, but in the late nineteenth century. It is derived notably from material presented in Gilliéron and Edmont’s Atlas linguistique de la France (1901–11). We cannot project such modern information back in time, in a simple, straightforward way, to reconstruct the dialect situation of the medieval past. Three important caveats have to be entered. Firstly, dialect differences were almost certainly less marked in the thirteenth century than they appeared to Gilliéron and Edmont, after six hundred years of linguistic change, and isoglosses can, and do, move with time. There is an increasing body of evidence which shows that the unity of Gallo-Romance, and, indeed, of the Romance world in general, persisted a good deal longer into the Middle Ages than earlier scholars believed (Wright 2002). Medieval Occitan and medieval French may have had different-looking writing systems, but there seem to have existed quite high levels of mutual intelligibility 1



In this paper I will be using dialect in the English sense of ‘variety of a language used by a group smaller than the total community of speakers’, and not in the more restricted sense in which dialecte is commonly used in French, and which is first found in Littré 1863: ‘DIALECTE, PATOIS. Tant que, dans un pays, il ne se forme pas de centre et, autour de ce centre, une langue commune qui soit la seule écrite et littéraire, les parlers différents, suivant les différentes contrées de ce pays, se nomment dialectes; on voit par là qu’il est tout à fait erroné de dire les dialectes dérivés de la langue générale; le fait est que la langue générale, qui n’est qu’un des dialectes arrivé par une circonstance quelconque et avec toute sorte de mélanges à la préséance, est à ce titre postérieure aux dialectes. Aussi quand cette langue générale se forme, les dialectes déchoient et ils deviennent des patois, c’est-à-dire des parlers locaux dans lesquels les choses littéraires importantes ne sont plus traitées. Avant le XIVe siècle il n’y avait point en France de parler prédominant; il y avait des dialectes; et aucun de ces dialectes ne se subordonnait à l’autre. Après le XIVe siècle, il se forma une langue littéraire et écrite, et les dialectes devinrent des patois’ (Littré 1863).

27

AN-1a.indd 37

11/03/2010 12:00:11

The Anglo-Norman Language and Its Contexts between speakers of the two languages. Secondly, although isoglosses are essential taxonomic devices, they remain methodological fictions. With genetically related vernaculars, dialect/language boundaries, such as those we see on Chaurand’s map, rarely correspond to sharply defined linguistic frontiers on the ground. Abrupt dialect/language boundaries reflect breaks in communication. Gaston Paris was at pains to point out a hundred and twenty years ago that the dialects of Gallo-Romance form a dialect chain, a continuum from the Channel to the Mediterranean without abrupt breaks (Paris 1889). If this was the case in the nineteenth century, it was probably even more so in the thirteenth, before standardisation. Thirdly, Chaurand’s map, like the ALF before it, concerns itself only with the traditional dialects of the countryside. It makes no provision for the presence of urban dialects in the modern world, and, a fortiori, precludes their existence in medieval times. Unencumbered by standardisation pressures, medieval speakers of Romance seem to have had a more flexible attitude to language variation than the one we possess today. The question arises of how and when language standardisation began.

1.2 language standardisation A highly influential typology of the processes involved in language standardisation was published by Einar Haugen some forty years ago (Haugen 1972: 110): Form

Function

Society

Selection of norms

Acceptance

Language

Codification

Elaboration

This typology was developed to handle the language-planning issues raised by standardisation in the modern, post-colonial world, where the four pro­ cesses usually happen in rapid succession as the result of conscious intervention by official bodies. We can usefully apply it to the standardisation of the old, national languages of Europe, provided we remember that, here, standardisation often took much longer, and it was organic, piece-meal and largely unconscious. This paper will be concerned with the beginning of the standardisation process in French, i.e. with the ‘selection of norms’.

2. The origins of standardisation in French In the case of certain languages – Italian, German and modern Hebrew, for example – standardisation was based initially on a written or literary form of the language. We note, however, that with these languages, standardisation happened relatively late and, crucially, after the spread of literacy to large swathes of the population. Does the beginning of standardisation in French,

28

AN-1a.indd 38

11/03/2010 12:00:11

The Sources of Standardisation in French – Written or Spoken? i.e. the ‘selection of norms’, which occurred a good deal earlier, follow the same written-based pattern? It is hard to conceive of language standardisation, involving the four processes identified by Haugen, taking place without the development of writing. The permanence and power of writing mean that a written standard will be influential in shaping new standards of speech. But, in the case of French, does this mean that a supra-regional written standard, created ex nihilo at an early date, initiated the whole process, that the ‘selection of norms’ began with the emergence of a written standard, with no spoken dialect behind it? The beginnings of new languages are invariably the least documented and the most difficult to trace. With the languages of the past we trace developments in speech principally through scrutiny of the surviving written documents, which only become prolific once the new language has become established. The most intractable problem we will have to deal with, then, will be the one of evidence, and the relationship between speech and writing in medieval texts. Do developments visible in written documents reflect what was happening in speech? Do such developments take place largely independently of developments in speech? Can they influence, even initiate, developments in speech?

2.1 the neogrammarian approach The Neogrammarians saw the origins of the French standard language in the dialect spoken in the Île-de-France. Gaston Paris (1889: 486) gave this dialect the label francien, and this continued to be used for much of the twentieth century in the standard manuals of the history of French (for example Pope 1934). For the Neogrammarians francien constituted a necessary stage in the soundlaw sequence leading from Latin to Modern French, and provided the nascent standard language with the ‘proper’ historical credentials: a ‘pure’ source-dialect (free from Germanic contamination) suitably located in the region of the capital. Ferdinand Brunot was at pains to emphasise the ‘purity’ of this dialect, declaring that ‘le francien ne doit pas être considéré comme un amalgame’ (Brunot 1905, t.2, p. 325). If we follow Haugen’s typology, the francien dialect was ‘selected’ in the twelfth century, ‘elaborated’ in the later Middle Ages, ‘codified’ in the seventeenth to eighteenth centuries and ‘accepted’ as standard by the French population at large in the nineteenth to twentieth centuries. How do we gain access to this medieval dialect? The Neogrammarians generally took what seems to us now a rather naive view of the relationship between speech and writing in the medieval period, despite the caveats entered by scholars like Gertrud Wacker (Wacker 1916). In thrall to a Romantic idea that medieval culture was more ‘natural’ and ‘spontaneous’, less governed by convention than modern society, the Neogrammarians had confidence in the capacity of medieval writing systems to reflect faithfully the regional variation present in the spoken language. During the 1980s, with the general post-’68 rejection of the positivism present in Neogrammarian thinking, the francien hypothesis was called into

29

AN-1a.indd 39

11/03/2010 12:00:11

The Anglo-Norman Language and Its Contexts question by a number of linguistic historians. Some were worried by the circularity implicit in tracing the origins of the standard language to a spoken dialect (le francien) which, they felt, could only be reconstructed with reference to that same standard language. They saw francien as a phantom dialect, ‘une variété hybride et peu distincte, dont nous ne savons rien de sûr avant le XIIIe siècle’ (Chaurand 1983: 91). Bergounioux (1989) maintained that a rural dialect requiring the label francien never existed in the Paris region, apparently believing that it was annihilated at birth by the linguistic influence of the city (though he declines to indicate how the speech of the city itself originated). Cerquiglini (1993: 124) declared that medieval Île-de-France had no dialect at all: ‘L’Île-de-France ne se distinguait par aucun dialecte. Jusqu’aux portes, et sans doute dans les rues de la modeste bourgade parisienne, on devait parler picard, normand ou orléanais.’ Various scholars came to see that the French standard language is a good deal less ‘pure’ than Gaston Paris had believed: fundamental elements in its phonology and morphology originated in dialects other than the dialect spoken in the Île-de-France. There survives a good deal of evidence allowing us to reconstruct the rural vernaculars of the Île-de-France in earlier times: literary and semi-literary representations of rustic speech written from the fifteenth century, metalinguistic observations made since the sixteenth century onwards, data present in the ALF and more recent regional atlases. All indicate the gulf which, from early modern times at least, separated the rapidly evolving speech of the metropolis from that of the more conservative rural hinterland. In an important paper, Claire Fondet (1995) isolated a number of spatial variables which are highly significant for our understanding of the relationship between the standard language and the dialect of Île-de-France. She demonstrated that while the French standard language is based on the speech of the Île-de-France, it is not based on a ‘pure’ form of that dialect, as the Neogrammarians believed. The French standard language is a mixed variety, a koine of some sort.2 If the French standard language began life as a sort of koine, following Trudgill’s definition, how could this particular mixture have come about?

2.2 the structuralist approach In contrast to the approach followed by the Neogrammarians, linguists working within the Saussurean paradigm lay stress on the conventional nature of linguistic systems. They make two distinctions which are important here: a. between langue and parole, and b. between internal and external elements of language. The linguist should strive to look beyond the variations present in parole in order to uncover the underlying, invariant system of langue. His 2



By koine I mean ‘a historically mixed but synchronically stable dialect which contains elements from the different dialects that went into the mixture, as well as interdialect forms that were present in none’ (Trudgill 1986: 107–8).

30

AN-1a.indd 40

11/03/2010 12:00:11

The Sources of Standardisation in French – Written or Spoken? prime concern should be with the internal structure of language rather than with externals (the socio-demographic context in which the language is used). Working in an intellectual climate strongly influenced by Saussure, C.-T. Gossen (Gossen 1962, 1967) drew attention to the level of conventionality present in the medieval writing systems of northern Gallo-Romance. He observed that the written forms found in medieval French manuscripts of different regional provenance show a good deal of commonality, and that they do not mirror very precisely the details of dialect variation visible in modern dialect atlases. He saw Old French spellings as possessing only a minimal anchorage in speech, i.e. they were detached from their local spoken dialect, and are to be read purely in visual terms. The basic unit in Gossen’s approach is the visual ‘grapheme’, a notion modelled upon that of the aural ‘phoneme’ of structuralist phonology. The Old French spelling system is seen, by the twelfth to thirteenth centuries, to have moved a long way along the cline from phonemic to logographic, and to be not very different from what we find in the standardised language of the modern world, where a uniform (-emic) system of spelling masks a wide range of spoken (-etic) realisations. From these ideas emerged what has come to be known as Skriptatheorie: each region of northern France had its own regional writing system or scripta, which admitted a small and varying proportion of regional features, but which, in the main, followed the norms of a supra-regional written koine. This approach has induced historians of French to regard the variation present in Old French writing systems as being a. relatively ‘free’ (i.e. not correlated with regional variation in speech), and b. relatively superficial, for, behind such variability as existed, there lay a supra-regional written koine reflecting the underlying langue, the standard language in embryo.3 Two questions arise. Firstly, at what date did a supra-regional written koine emerge in northern Gallo-Romance? Secondly, how did a set of conventions designed for the written language come, in the fullness of time, to influence the way people speak? Let us take these questions in order. The structuralist view has it that a conventional written koine emerged in northern France, independent of local variations in speech, at some time before the twelfth century. Hilty attached this supra-regional norm to the speech of Paris in the eighth century (Hilty 1973: 254). Delbouille suggested that such a variety developed as early as the ninth century, prior to the diversification of the langue d’oil dialects (Delbouille 1962: 124). Cerquiglini maintains that a conventional written koine was elaborated by an ‘enlightened’ group of scribes and literary authors as early as the Strasbourg Oaths: ‘C’est grâce à l’existence d’une société cléricale, guidée par une “lumière de raison”, animée par les litterati désirant illustrer un bel usage littéraire de l’idiome roman, que dès les premiers textes est fondé et pratiqué un “illustre françois”’ (Cerquiglini 1993: 124). This same author sees at work a long-term project devised in Carolingian

3

This view coincides with a long tradition of prescriptivist thinking which sees the written language as the quintessential form of the language.

31

AN-1a.indd 41

11/03/2010 12:00:11

The Anglo-Norman Language and Its Contexts times to bring about the linguistic and political unity of the country: ‘Une langue française transcendant la diversité des parlures, inscrite dans le projet d’une forme commune échappant pour des raisons politiques ou esthétiques, à l’échange local et quotidien’ (Cerquiglini 1993: 124). How did this koine, originally put together in the literary language, become ‘accepted’ as the norm for speech? Proponents of Skriptatheorie are largely silent on this point, but the assumption seems to be made that the norms embodied in the written koine were gradually diffused top-down, beginning, of course, in Paris. Selection of norms took place in the royal court, from where they trickled down the various strata of Paris society: ‘Ce fut, en France, la langue littéraire de la scripta, qui, sans être artificielle, s’élabora dans des conditions sociologiques différentes de celle des idiomes populaires. A partir du XIIIe siècle, le parler urbain, stratifié, certes, mais proche de celui de la classe aisée, n’a cessé de s’opposer au parler rural jusqu’à la Révolution’ (Fondet 1995: 201). The structuralist approach thus maintains not only that the various written scripta of Old French exist independently of variations in speech, but that they subsequently influenced, even initiated, developments in speech. This raises a number of problems, which I will call the ‘data and chronology problem’ and the ‘implementation problem’ (see Weinreich, Labov and Herzog 1968: 95–105). 2.2.1 The data and chronology problem Manuscript evidence does not exist in great quantities to support the case for the existence of a stable set of supra-regional written norms in northern GalloRomance before the thirteenth century (see Gsell 1995). Literary production was prolific in French during the twelfth century, but texts composed before 1200 are almost all preserved in manuscripts written many decades later. The number of manuscripts surviving from before 1200 is very small (see Pfister 1993). French did not come into widespread use as an administrative language until well into the thirteenth century (see Giry 1894: 464–72, Lusignan 2004: 79–94). Before then, it is impossible to find any institutional framework, like a royal chancery, which might have made some supra-regional norm possible. Gossen was, of course, correct in his observation that within thirteenth-century texts we find a sizeable proportion of forms which are invariant for region, but does it need a supra-regional written koine to account for this? Firstly, it is pretty certain that dialect differences in northern Gallo-Romance were less marked in the thirteenth century than they were six hundred years later at the time of the ALF. Secondly, medieval vernacular spellings were approximations and it is to be expected that only the more salient regional forms would be specially encoded in local writing systems. Thirdly, the vernacular writing systems which emerged in northern France and in Anglo-Norman England between the ninth and twelfth centuries were developed essentially for the recording of literary texts. The perambulatory nature of literary manuscripts inevitably caused a degree of dialect-mixing and the levelling of strongly marked regional forms. However, ad hoc levelling on an individual-text basis does not entail the existence, at this stage, of a stable written koine.

32

AN-1a.indd 42

11/03/2010 12:00:12

The Sources of Standardisation in French – Written or Spoken? The focusing of linguistic norms normally follows the pathway: local norms

→ supra-local norms → regional norms → supra-regional norms. It would be

unusual for the normalisation pathway in northern Gallo-Romance to have moved in the opposite direction, with supra-regional norms emerging from the outset. Antonij Dees was half-right when he asserted: ‘La notion de koinè écrite, ainsi que la notion corrolaire de scripta régionale, n’ont aucune adéquation observationnelle pour la période antérieure à 1300’ (Dees 1985: 113). Although regional norms appear to have emerged before 1300, notably in Picardy, in Normandy and in the east of the country, it is hard to show the existence of a supra-regional koine before that date. 2.2.2 The ‘implementation problem’ If we move forward, temporarily, to the end of the Middle Ages, it is clear that Parisian speech was indeed, by then, a mixed dialect (a koine) containing forms drawn from several regions of northern France. How did the supraregional written koine believed to have emerged before the twelfth century come to be accepted as the spoken language of most Parisians by the end of the Middle Ages? An acrolectal variety was almost certainly cultivated in the royal court as early as the twelfth century, setting the ‘courtois’ apart from non-noble elements in society. It is more or less certain that this acrolect exerted an influence on the forms used by the literary entertainers of the day: the much-quoted passage from Conon de Béthune, in which the poet apologises for showing traces of a dialect not approved of by the royal circle, is clear testimony to this fairly obvious fact (see Lodge 1994: 99). By the same token, however, it is most doubtful that strong influence was exerted in the opposite direction, that is, that members of the royal court were moved to imitate in their speech a set of linguistic forms presented to them by literary entertainers. Writing need not remain a mere record, and can influence the community of speech (Haugen 1972: 163), but it is doubtful whether, given levels of literacy in the thirteenth to fourteenth centuries, the prestige of vernacular written forms, and even familiarity with them, were sufficient for this to happen on a significant scale, even among the elites.

3. A sociolinguistic approach It is legitimate and interesting to consider the writing systems found in medieval French texts purely in visual terms, without seeking to correlate them with the underlying spoken language. Regional writing systems can act as powerful symbols of regional identity, even if they are seen as operating independently of regional variations in speech. However, since language is essentially spoken, to consider medieval scripta solely as autonomous visual systems contributes little to our understanding of the process of language change, which is the central concern of historical linguistics. To the extent that variation is present in medieval French texts, historical sociolinguists regard

33

AN-1a.indd 43

11/03/2010 12:00:12

The Anglo-Norman Language and Its Contexts it not as being random or ‘free’, but as being conditioned by relevant extralinguistic factors, which it is their job to identify. In order to pursue this sort of research, variationists have to estimate accurately how much regional variability exists in medieval French texts and how long it persists. It is clear that in the thirteenth century the amount of regional variation visible in the phonetics and morphology of written documents is very considerable and that in the following century it diminishes significantly (see Völker 2007). The thirteenth century, therefore, offers historical sociolinguists a precious ‘window of opportunity’ for using the writing systems of vernacular texts to trace the progress of phonetic and morphological change. Variability in thirteenth-century French cannot be analysed in terms of deviations from some central norm.4 There was, as yet, no central norm. It is to be analysed in terms of quantitative differences in the distribution of key dialectal variables, in the way pioneered, for example, by Dees (1985). How did the central norm emerge? We saw earlier that the French standard language is based on some sort of koine. We have seen the evidential problems which arise if we try to see this koine arising initially in the written language, and the problems of sociolinguistic plausibility which we run up against if we look at the implementation of change without taking the community of speakers properly into account. Can historical dialectology and historical sociolinguistics offer a more satisfactory narrative? Is it possible that the standard language koine came about on the basis of a ‘real-life’ koineisation which occurred in the speech of the city? This would overcome both the ‘chronology problem’ and the ‘implementation problem’ discussed earlier.

4. Koineisation and urbanisation Whenever speakers from non-adjacent sectors of the dialect continuum come into contact, temporary mixed varieties arise. Sometimes it is individual speakers who move around different parts of the country, levelling out their most salient local features in the different dialect areas to which they travel. This would be the case, for example with itinerant entertainers accommodating their pieces to suit different audiences. Sometimes it is groups of speakers migrating from different areas to converge on a central place, bringing together forms of different dialect provenance and eliminating the most strongly marked regional forms in their day-to-day interactions. Speakers of Gallo-Romance had been moving about the area all the time, accommodating their dialectal speech to that of their interlocutors at every encounter. However, these countless individual acts of accommodation could only result in

4

The term ‘norm’ is used here in the sense of descriptive, statistical regularities, not in the sense of explicit, prescriptive rules, such as those articulated in the process of codification.

34

AN-1a.indd 44

11/03/2010 12:00:12

The Sources of Standardisation in French – Written or Spoken? a stable koine when certain conditions had been met: above all, it required a period of regular and intense interaction, during which time individual acts of accommodation all came to point in roughly the same direction. For koineisation to take place, it is usually essential for a significant focusing of interactions to occur in some central place. The prime locations of such intensified levels of interaction are towns. In the centuries following the fall of the Empire, fragmentation of the urban network of Roman times was accompanied by the dialectalisation of Latin. It was only in the eleventh century that things began to change, with a remarkable upsurge of urbanisation, beginning in northern Italy and in the Low Countries, and eventually involving Paris. During the twelfth century Paris emerged as an urban giant, dwarfing all the other conurbations of western Europe. Table 1. The largest cities in Europe, 1000–1900 (population in thousands) (Hohenberg and Lees 1985: 11) 1000 Constantinople Cordoba Seville Palermo Kiev Venice Regensburg Thessaloniki Amalfi Rome

450 450 90 75 45 45 40 40 35 35

1400 Paris Milan Bruges Venice Granada Genoa Prague Caffa Seville Ghent

1700 Constantinople London Paris Naples Lisbon Amsterdam Rome Venice Moscow Milan

700 550 530 207 188 172 149 144 130 124

1900 London Paris Berlin Vienna St Petersburg Manchester Birmingham Moscow Glasgow Liverpool

275 125 125 110 100 100 95 85 70 70

6,480 3,330 2,424 1,662 1,439 1,255 1,248 1,120 1,072 940

An essential feature of the demography of medieval towns was that they could not sustain themselves, still less expand, on the sole basis of the fertility of their inhabitants. They were entirely dependent on attracting the surplus population of the surrounding rural areas. Population growth in twelfth- to thirteenth-century Paris was fed almost entirely by immigration.

35

AN-1a.indd 45

11/03/2010 12:00:12

The Anglo-Norman Language and Its Contexts Exceptional demographic developments on this scale cannot be ignored by historians of the language. One can predict with a strong measure of certainty that the intensity and regularity of social interaction which evidently occurred in twelfth- to thirteenth-century Paris would lead to dialect-mixing and the focusing of a new koineised variety. This in turn would raise the speech of the city above the dialect continuum not only of northern Gallo-Romance in general but also of its hinterland dialect, traditionally labelled francien. Claire Fondet’s paper cited above indicates the sort of dialect-mixing which provides the basis of the French standard language. The spatial configuration of elements going into the mixture can be determined with reference to the maps in the ALF. The results of this can be seen in the maps reproduced here (Maps 1–9). Table 2. Dialect-mixing in Parisian French (1) Northern boundary Palatalisation of [ka-] and nonpalatalisation of [ke-/ki-] Map 1 ALF 250

(5) Incursions from the east [wa] ~ [we] e.g. poire, droit, froid Map 5 ALF 1047

(2) Eastern boundary Presence of epenthetic consonants in the group [n’r] Map 2 ALF 1359

(6) [o] ~ [jo] e.g. manteau, seau, eau Map 6 ALF 1208

(3) Western boundary (7) Incursions from the west Differentiation of the [o] > [ow] > [Ew] -ent ~ -ont 3pl. present indicative Map 3 ALF 151 stressed on the ending Map 7 ALF 1064 (4) Southern boundary Raising of stressed [a] to [e] Map 4 ALF 992

(8) -ions ~ -iens 1pl. imperfect indicative, present subjunctive Map 8 ALF 512 (9) -aient ~ -aint 3pl. imperfect indicative Map 9 ALF 10

It needs to be emphasised that the isoglosses traced out here depict the dialect situation as it was found by Gilliéron and Edmont in the late nineteenth century, not as it might have existed in medieval times. Isoglosses can and do move over time. Indeed, maps 5 to 9 show precisely that. Maps 1 to 4 show the features marking the boundaries existing on the four sides of the central dialect area. These appear to have been relatively stable across the centuries. Maps 5 to 9, on the other hand, show the bulges in each isogloss around the city of Paris, suggesting that these particular forms entered Paris from the east or west, and that, having established themselves there, they subsequently

36

AN-1a.indd 46

11/03/2010 12:00:12

The Sources of Standardisation in French – Written or Spoken?

Map 1 (ALF 250)

Map 2 (ALF 1359)

Map 3 (ALF 151)

Map 4 (ALF 992)

37

AN-1a.indd 47

11/03/2010 12:00:13

The Anglo-Norman Language and Its Contexts

Map 6 (ALF 1208)

Map 5 (ALF 1047)

Map 7 (ALF 1064)

Map 8 (ALF 512)

Map 9 (ALF 10) 38

AN-1a.indd 48

11/03/2010 12:00:13

The Sources of Standardisation in French – Written or Spoken? began diffusing out on the other side of the city. If we are looking for a koineised variety to act as the prime source for language standardisation in French, it would be imprudent not to relate it closely to the real-life koine which almost certainly developed in the city in the twelfth century. Penny (2000: 43–4) reports an analogous development in Madrid when this city became the capital of Spain in the sixteenth century. Such a hypothesis does not fall foul of the ‘implementation problem’ we discussed earlier, but, at such a remote time, it inevitably raises problems of data.

5. Written evidence Do we have any documentary evidence to support the hypothesis that the speech-norms of standard French are derived historically from the koineisation which occurred in Parisian speech between 1150 and 1250? Unfortunately, Paris was slow to use the vernacular (as opposed to Latin) in precisely dated administrative documents: we find very few Parisian ‘charters’ before 1250, and it is only after 1300 that they become copious. However, they do exist (see Matzke 1880, 1881). Paul Videsott of the University of Bolzano has recently drawn attention to a charter emanating from the royal chancery in 1241 (source: Videsott forthcoming): . . . 7 Et se la pais/ n’est tenue cil Jahans de Thorete doit reseisir le Roi de Navarre de ces aquez et des issues sauf le droit/ de chacun qui demorra en toutes choses en tel point cum il estoit au jor que la pais fu porpallee. 8 Et est asavoir/ que li Templier ne porront aquerre en la conté de Champaigne et de Brye ne en ses fiez dedens le terme/ devant dit quar nos lor avons deffendu et il distrent que non feroient il. 9 Et li aquest que li Templier// avoient fait devant la pais qu’il firent au roi de Navarre loer doivent demourer à totes les issues 10 né/ por sael que nos metains en ces letres nulles choses ne lor sont comfermees se la pais ne se fait 11 Et ceste/ pais devant-dite doivent li Templier tenir se li grant maistres du Temple d’Outre-mer et li couvenz / le louent et li rois de Navarre la doit tenir s’il si acorde 12 Et se la pais n’est tenue, la chose sera / en tel point comme devant en totes choses Et se la pais ne tient ces letres ne vaudront rien ou // tre les Brandons devant diz. 13 Et se ceste pais est tenue li rois de Navarre et li Templier nos / ont requis que nos la tesmognein ou confermain sauve la guarde au seignor de Champaigne / et de Brye. 14 Ce fu feit à Pontoise l’an de l’incarnation nostre Seignor mil deus cenz et / quarante et un ou mais d’aoust

What is particularly interesting here are the forms metains, tesmognein and confermain (italicised above) which correspond to the variable forms of the 1pl. present subjunctive indicated in our Map 8. These examples show the persistence of pre-koineisation forms even in the writing system of the royal chancery. I undertook a quantitative analysis of orthographic and morphological variation in a corpus of approximately 180 charters emanating from the Prévôté

39

AN-1a.indd 49

11/03/2010 12:00:13

The Anglo-Norman Language and Its Contexts de Paris between the years 1249 and 1365. Here is an example, a Parisian charter of 1260 (source: Depoin 1925): A touz ceus qui ces letres verront et orront, je Nicolas de Castenoi, chevaliers, et ma dame Agnes, ma feme, saluz. En Jesucrit nous fesons a savoir a touz ceus qui sont et qui a venir sont que, comme contenz fust entre nous d’une partie et religieuse gent le prieur et le couvent de Saint Martin des Chans de Paris de l’autre partie, sur ce que nous et Ferri e le Juene mes freres et sa feme disions et demandions seur une piece de terre, qui siet ou terroir de Chastenoi par devers Puiseus, joignant de cele partie a la terre Renart le Tainturier, jadis bourgois de Paris, et de l’autre partie par desus au jardin Basile de Charni, lequel jardin elle tient de nous a une maalee de cens, et au jardin Aveline la Peletiere, et de une partie d’en costé au pré Ferri le Juene et de l’autre partie a la terre Ernoul Bridoul deux garbes de blé pour raison de champart, l’une par reson de nostre iretache et l’autre par reson d’achat d’Adam Morree, le quel achat nous aviens fait a Adam Morree. Et us aviens nous et nos devantiers d’avoir et de recevoir les devant dites jarbes par reson de champart en icele piece de terre par lonc tens. La quelle chose devant dite li religieus nous nioient nous enquise la verité, par le conseil de bone gent avons renoncié et renoncons du tout et expresseement a la chose et au content desus dit, et quitons et avons quiti du tout les garbes desus dites et queque reson et queque droiture nous aviens ou poienz avoir en icele piece de terre, sans riens retenir ne riens des ore en avant n’i reclamerons ne ferons reclamer par nous ne par autrui. Et en tesmoing de ce nous avons seelees ces letres de noz seaus, ce fu fait en l’an de l’Incarnation mil et ii cens et sessante ou mois de juin.

When we look at this corpus we see that, as with charters from any other region, the variable forms represent a mixture which is specific to the region of origin and which, far from being fixed, is subject to considerable change over the hundred years spanned by the corpus. Examination of the way this corpus handles the nine key phonetic and morphological variables we selected earlier suggests three broad conclusions: ▪ the Parisian charters contain a mixture of written forms, some originating in the central area (variables 1–4) and others imported from outside (variables 5–9). ▪ the in-mixing of exogenous elements affects the morphology (variants 7 and 9) earlier than the phonology (variants 5 and 6). ▪ the mixture of dialectal elements in our texts has not yet stabilised: the endogenous variants (1–4) are virtually categorical, whereas the exogenous variants (5–9) involve a degree of variation and change: changes implied in 7 (–ent ~ –ont 3pl. present) and 9 (–aient ~ –aint 3pl. imperfect) have been brought to completion, changes in 6 ([o] ~ [jo]) and 8 (–ions ~ –iens 1pl. imperfect) look like ‘changes in progress’, and the change in 5 ([we] > [wa]) is not yet visible. The evidence of this corpus suggests that, in the period considered here, the process of koineisation in Paris was on-going, not complete.

40

AN-1a.indd 50

11/03/2010 12:00:14

The Sources of Standardisation in French – Written or Spoken? 6. Conclusion This paper has been concerned with one of the two ‘social’ processes identified by Haugen in his analysis of standardisation, the ‘selection of norms’. I have argued that, whereas in Italian and German, which standardised later, the initial ‘selection of norms’ may well have settled on a written koine elaborated artificially by writers, in French, where standardisation began earlier thanks to the dominance of Paris, such a scenario is less plausible. It is hard to conceive of language standardisation taking place without the development of writing. The permanence and power of writing mean that a written standard will be influential in shaping new standards of speech. But, in the case of French, it is unwise to discount the sociolinguistic impact of demographic transformation on the scale which occurred in Paris in the twelfth to thirteenth centuries, it is unrealistic to suppose the creation ex nihilo, at an earlier date, of a stable, supraregional written standard and it is anachronistic to suppose that such a set of written norms could have materially affected the speech of the population at large before the advent of widespread literacy. It is more reasonable to suppose that the initial source of standardisation was a naturally occurring koine which developed in the speech of the city through large-scale in-migration, and which was subsequently represented in the written language of Paris.

Bibliography Balibar, R. (1985), L’Institution du français. Paris. Bec, P. (1971), Manuel Pratique de Philologie romane, vol. II. Paris. Bergounioux, G. (1989), ‘Le francien (1815–1914): la linguistique au service de la patrie’, Mots/Les langages du politique 19, 23–40. Brunot, F. (1905), Histoire de la langue française des origines à 1900, vol. I. Paris. Cerquiglini, B. (1993), La naissance du francais. Paris. Chaurand, J. (1983), ‘Pour l’histoire du mot “francien”’, in Mélanges de dialectologie d’oil à la mémoire de R. Loriot, Dijon, pp. 91–9. ——(ed.) (1999), Nouvelle histoire de la langue française. Paris. Cohen, M. (1987), Histoire d’une langue: le français. Paris. Dees, A. (1980), Atlas des formes et constructions des chartes françaises du 13e siècle. ZRP, Beiheft 178. Tübingen. ——(1985), ‘Dialectes et scriptae à l’époque de l’ancien français’, Revue de linguistique romane 49, 87–117. ——(1988), ‘Propositions for the study of Old French and its dialects’, in J. Fisiak (ed.), Historical Dialectology, Trends in Linguistics, Studies and Monographs 17, The Hague, pp. 139–48. ——(1989), ‘La reconstruction de l’ancien français parlé’, in M. Schouten and P. van Reenen (eds.), New Methods in Dialectology, Dordrecht, pp. 125–33. Delbouille, M. (1962), ‘La notion de “bon usage” en ancien français’, Cahiers de l’Association internationale des études françaises 14, 10–24. Depoin, J. (1925), Chartes et documents de l’abbaye Saint-Martin-des-Champs. Paris.

41

AN-1a.indd 51

11/03/2010 12:00:14

The Anglo-Norman Language and Its Contexts Dubuisson, P., and M.-R. Simoni-Aurembou (1990), ‘Französisch: Arealinguistik III. Zentrale Dialekte’, in G. Holtus, M. Metzeltin and C. Schmitt (eds.), Lexikon der romanistischen Linguistik, Tübingen, vol. V,1, pp. 637–53. Fondet, C. (1995), ‘Contribution à la question des origines du français: quelques aperçus à partir de la dialectologie de l’Essonne’, in M. Tamine (ed.), Ces mots qui sont nos mots. Mélanges d’histoire de la langue française, de dialectologie et d’onomastique offerts au professeur Jacques Chaurand, Charleville-Mézières, pp. 189–206. Giry, A. (1894), Manuel de Diplomatique. Paris. Gossen, C.-T. (1962), ‘Langues écrites du domaine d’oil’, Revue de linguistique romane 26, 271–308. ——(1967), Französische Skriptastudien. Vienna. Gsell, O. (1995), ‘Französische Koine’, in G. Holtus, M. Metzeltin and C. Schmitt (eds.), Lexikon der romanistischen Linguistik, Tübingen, pp. 271–89. Haugen, E. (1972), ‘Dialect, language, nation’, repr. in J. Pride and J. Holmes (eds.), Sociolinguistics. Harmondsworth. Hilty, G. (1973), ‘Les origines de langue littéraire française’, Vox Romanica 32, 254–71. Hohenberg, P., and L. Lees (1986), The Making of Urban Europe 1000–1950. Cambridge, Mass. Littré, E. (1863), Dictionnaire de la langue française. Paris. Lodge, R. (2004), A Sociolinguistic History of Parisian French. Cambridge. Lusignan, S. (2004), La Langue des rois au moyen âge. Paris. Matzke, E. (1880, 1881), ‘Der Dialekt von Ile-de-France im XIII und XIV Jahrhundert’, Archiv für das Studium der neueren Sprachen und Literaturen 64, 385–412 ; 65, 57–96. Michaëlsson, K. (1959), ‘Quelques variantes notées dans la prononciation parisienne au temps de Philippe le Bel’, in VIII Congresso Internazionale di Studi Romanzi, II.2, Florence, pp. 287–97. Milroy, J., and L. Milroy (1999), Authority in Language. 3rd ed. London. Muscatine, C. (1981), ‘Courtly literature and vulgar language’, in G. Burgess and A. Deyermond (eds.), Court and Poet. Liverpool. Paris, G. (1889), ‘Les parlers de France’, Romania 17, 475–89. Penny, R. (2000), Variation and Change in Spanish. Cambridge. Pfister, M. (1973), ‘Die sprachliche Bedeutung von Paris und der Ile-de-France vor dem 13. Jh.’, Vox Romanica 32, 217–53. ——(1993), ‘Scripta et koinè en ancien français aux XIIe et XIIIe siècles’, in Ecritures, langues communes et normes: formation spontanée de koinès et standardisation dans la Gallo-Romania et son voisinage, Neuchâtel/Geneva, pp. 17–41. Pope, M. (1934), From Latin to Modern French. Manchester. Siegel, J. (1985), ‘Koines and koinéization’, Language in Society 14, 357–78. Trask, R. (1997), A Student’s Dictionary of Language and Linguistics. London. Trudgill, P. (1986), Dialects in Contact. Oxford. Videsott, P. (forthcoming), Autour du plus ancien document en français de la chancellerie royale capétienne. Völker, H. (2007), ‘A “practice of the variant” and the origins of the standard. Presentation of a variationist linguistics method for a corpus of Old French charters’, Journal of French Language Studies 17, 207–23. Wacker, G. (1916), Über das Verhältnis von Dialekt und Schriftsprache. Halle.

42

AN-1a.indd 52

11/03/2010 12:00:14

The Sources of Standardisation in French – Written or Spoken? Weinreich, U., W. Labov and M. Herzog (1968), ‘Empirical foundations of language change’, in W. Lehmann and Y. Malkiel (eds.), Directions for Historical Linguistics, Austin, 95–195. Wuest, J. (2003), ‘Le rapport entre langue parlée et langue écrite: les scriptae dans le domaine d’oïl et dans le domaine d’Oc’, in M. Goyens and W. Verbeke (eds.), The Dawn of the Written Vernacular in Western Europe. Leuven.

43

AN-1a.indd 53

11/03/2010 12:00:14

chapter four

Husbonderie and Manaungerie in Later Medieval England: A Tale of Two Walters William Rothwell

Oschinsky (1971) took forward Elizabeth Lamond’s pioneering work on the history of agriculture in later medieval England (Lamond 1890), publishing a fuller study of the material used by her predecessor and adopting the title ‘Walter of Henley’ and Treatises on Estate Management. This new work consisted primarily of four texts of between eleven and sixteen pages each in AngloFrench accompanied by substantial explanatory commentary, all dealing with estate management in rural England in the second half of the thirteenth century, but all of them approaching their task from different perspectives. The earliest of these texts was written for the countess of Lincoln by Robert Grosseteste, bishop of Lincoln, and entitled Les reules ke le bon eveske de Nichole Robert Grosseteste fist [. . .] de garder e governer terres e hostel (i.e. estates and households) going back to about 1242. This was followed some twenty-five or thirty years later by an anonymous ‘Seneschaucy’, then by the ‘Hosbondrye’ of Walter of Henley, and finally an anonymous ‘Husbandry’ from around the turn of the century. The ‘Reules’ of Grosseteste show a landowner bringing to his lay responsibilities the same kind of approach that he brought as bishop of Lincoln to the governance of his bishopric. The structure of authority from the bishop down through his various officials to the peasants who worked his estates is clearly set out. The chief official who manages the whole estate for the landlord is the steward (seneschal), whose many responsibilities include the organisation of the numerous agricultural operations, the checking of all the accounts appertaining to them and also those referring to the running of the household. Moreover, at Michaelmas he must plan ahead for all the activities of the estate in the coming year. The role of the landlord himself is epitomised in the ceremony of dining, with its echoes of high mass in his cathedral. The food and drink, at their blessing before dinner in hall, are to be presented to the lord by his baker and butler wearing their master’s livery in spotless condition, the servants are allowed to enter only when the freemen and guests are seated, the correct positioning of the ale and wine on and under the tables is laid down and the marshal is on duty to supervise the seating and ensure good conduct at table with the assistance of two men specifically detailed for the task. The

44

AN-1a.indd 54

11/03/2010 12:00:14

Husbonderie and Manaungerie in Later Medieval England same kind of arrangement obtained in the bishop’s management of the estate in general, the whole atmosphere of the ‘Reules’ being one of a beneficent, almost fatherly control working down from the master to his household and estate staff. This personal characteristic of the ‘Reules’ is absent from the anonymous and more business-orientated Seneschaucy, which in its turn sets out in detail the duties of the seneschal but from a secular point of view, followed by the responsibilities of all the other staff on the estate in descending order of rank, who worked under the direction of the steward and were collectively responsible for the success of a complicated operation of land management. The steward in the ‘Seneschaucy’ had to be a man of many parts, first and foremost an experienced lawyer able to deal with the legal questions associated with the possession and exploitation of large areas of land. As chief administrator he had to be competent to estimate the extent and value not only of the arable land held on the estate, but also of pasture and meadow, and he must understand the techniques of ploughing different kinds of terrain at different times of the year and keeping the ploughs in good working order. He also had to be capable of estimating the potential yield of various crops in different conditions of ground, weather, and so on, and then of supervising the harvest closely, so as to thwart any attempt to steal the lord’s grain as it was moved around the estate or falsify the estate records. The steward also had responsibility for the welfare of the whole range of farm animals in terms of their reproduction, market value, protection from disease, and so on. Whilst keeping a check on his workers, he himself would be subject in his turn to scrutiny by other household officials regarding the conduct of his stewardship. Although based on the ‘Seneschaucy’, the third treatise, the Hosbondrye of Walter of Henley, ‘appealed to a wider audience’ (Oschinsky 1971: 7), surviving in thirty-five manuscripts. Instead of explaining the management of an estate by detailing the duties of the individual officials, as in the ‘Seneschaucy’, Henley’s text is set out ‘in the form of a sermon, the advice of a father to his son’ (Oschinsky 1971: 7), concentrating on the actual exploitation of the soil, the many basic problems posed for the landowner according to season and weather, together with the best practice needed for their solution, and also providing advice as to the most efficient way of handling the finances of the estate. This eminently practical approach may be responsible for the popularity of the work. Finally, the anonymous ‘Husbandry’ is closer to the Seneschaucy than to Grosseteste or Walter of Henley, being essentially a treatise on the manifold duties of the accountant who is responsible for the success or failure of the estate as a financial institution. Oschinsky (1971) also contains nine extracts from other treatises on accounting, mainly as applied to the economics of husbandry, eight in Latin, one in French. The contents of the four French treatises on the management of estates in the English countryside with their differing approaches have been sketched above in very summary fashion in order to show that, although products of the same economic society and dealing with the same subject, they are independent

45

AN-1a.indd 55

11/03/2010 12:00:14

The Anglo-Norman Language and Its Contexts works. When viewed as a whole they provide not only a wide range of detailed information regarding the agricultural techniques and financial workings of medieval land management as they would have been practised by a variety of landowners in the East Midlands in the later thirteenth century, but they also set out different aspects of the working relationships, duties and responsibilities of the officials and estate employees, illustrating how the whole economic and social structure of the medieval countryside was held together in a wellestablished written framework including both law and accounting as well as agricultural practice.1 In addition to being valuable for the study of the economic history of medieval England, these texts are also important for an understanding of its linguistic history on account of the specialised French lexis contained in their subject-matter, which their authors handled without difficulty. This may be seen by comparing the technical vocabulary of agriculture in French used in the four treatises with the ‘borrowing’ evident in Anglo-Latin as illustrated by the ‘Table of the Differences between the French and Latin Rules’ which follows them in Oschinsky (1971: 412–15), where the Latin version of Grosseteste’s work may be seen to abound in calques from its French counterpart. This linguistic contrast is visible again in the Appendix (pp. 459–78) where Latin extracts from a number of independent treatises on accounting are followed by a French text from a register of fees relating to lands in Dunster (Somerset) belonging to the barons Mohun. The Latin texts are peppered with transparently English or French words tricked out as Latin, but the Dunster text is on a par with the main agricultural treatises in its competent use of French and the absence of English. The esteem in which these French treatises on agricultural management were held in their day and the extent of their circulation may be judged not only by the statement that ‘Eighty-four manuscripts include one or more of our treatises’ (Oschinsky 1971: 10), but also by the copious variant readings given in the pages of ‘Critical Apparatus’ which follow the edition’s main texts, especially the one which accompanies Walter of Henley’s ‘Hosbondrye’. This has no less than forty-one pages of notes (pp. 344–85) containing linguistic material not found in the printed version of the text and so reflecting the dissemination of the treatise and pointing to a widespread understanding of the technical French of agriculture in the English countryside amongst the landed classes in the later Middle Ages.2 Oschinsky (1971) provides not only a list of the many manuscripts which 1



2



‘Just as the legislation of Edward I brought about the rise of a class of professional lawyers [. . .], so too the legal demands of the period [. . .] were the impetus for the creation of a class of professional estate officers trained for estate management and accounting [. . .], and receiving practical training on manorial estates’ (Oschinsky 1971: 73–4). See Oschinsky (1971: 10–50) for details of the manuscripts and pp. 50–74 in the chapter entitled ‘The Background to the Treatises’ for an insight into the readership of the different manuscripts.

46

AN-1a.indd 56

11/03/2010 12:00:14

Husbonderie and Manaungerie in Later Medieval England contain versions of one or other of the agricultural treatises, but also identifies the texts which accompany them, giving the reader a balanced picture of the literary context in which they are found. In the Bodleian manuscript Selden Supra 74, the French text of Walter of Henley’s treatise sits together with what she calls ‘the poem by Walter of Bibbesworth on learning French, with English glosses’ and also two religious works in French by Nicole Bozon. A similar reference is made to the copy of Walter of Henley’s ‘Hosebondrye’ in the British Library manuscript Harley 3860 which also houses Grosseteste’s religious Chateau d’amour and William de Waddington’s Manuel des péchés (Oschinksy 1971: 14), whilst another copy of Henley’s work from c. 1300 referred to on p. 132 contains Bozon’s Proverbes de bon enseignement together with another version of Bibbesworth’s Tretiz. An example of this connection between Bozon and Bibbesworth found by Paul Meyer in the then Phillipps manuscript 83363 had been published as early as 1884, but had gone unnoticed: ‘la portion la plus considérable, celle qui contient les poésies de Bozon et de Gautier de Bibbesworth [. . .] n’a pas été étudié jusqu’à ce jour’ (Meyer 1884: 499), confirming Oschinsky’s (1971: 44) statement that ‘the texts by Walter of Bibbesworth, Walter of Henley, and Nicole Bozon were considered in some circles as equally “good” reading and [.  .  .] the comparison of the three treatises shows the authors knew each other’s work’. These references all indicate that for the readership of the day such French texts were all of a piece and were appreciated accordingly. Nor did the writers necessarily confine themselves to one type of work. Grosseteste the landowner who compiled the Reules for estate management in French was also the bishop better known to historians of literature as the author of several religious works, also in French, including an Oraciuncula post Prandium Gallice (Thomson 1940: 14), which is reminiscent of his ordering of the ceremony of dinner in hall mentioned above. Likewise, Bibbesworth’s poems referred to by Meyer and another published by Wright (1857) are quite different in character from his Tretiz, as will be shown later. Modern scholarship, however, tends to treat the contents of medieval manuscript compilations as individual items to be edited separately and without reference to any other compositions by their authors, with the result that the works of the writers mentioned above have been separated into distinct categories to be studied by different groups of scholars. The religious writings of Grosseteste and Bozon are regarded as belonging to medieval French literature and as the preserve of literary historians, whilst Grosseteste’s agricultural treatise in French is put in the category of medieval English history. In the twentieth century the minor literary works by Bibbesworth in French referred to above, although accessible in print and clearly relating to England, one of them being a dialogue between Bibbesworth and a nobleman named Henry de Lacy, were not taken into consideration in the assessment of his Tretiz, so that, despite all its manuscripts being written in England for an English readership, 3



Olim MS Philipps 8336, now British Library Add. 46919.

47

AN-1a.indd 57

11/03/2010 12:00:14

The Anglo-Norman Language and Its Contexts this main work was judged to be an independent exercise in French grammar, to be measured by the standards of medieval French current on the continent, instead of being correctly situated with the agricultural works and treated as belonging to English history and the history of English. Oschinsky’s brief description of the Tretiz as a ‘poem [.  .  .] on learning French, with English glosses’ mentioned above does not explain its true purpose, its contents or its connection with Walter of Henley’s ‘Hosbondrye’. It would be more accurately described as a lengthy and detailed exposition in French of the vocabulary of the countryside intended to enable the children of Bibbesworth’s patroness to administer their estates in French when they reached maturity. Moreover, it has survived in numerous manuscripts which vary considerably in length, linguistic quality and the amount of their glossing, some of them being heavily glossed, others not at all. A knowledge of the vocabulary in the Tretiz would be essential for readers of Henley’s work on the practical administration of an estate in French, what Bibbesworth’s Tretiz terms husbonderie and manaungerie so that the two works are in a very real sense complementary. Indeed, the Bibbesworth and Henley texts in the Bodleian manuscript referred to by Dr Oschinsky are in such close physical proximity on the parchment that the final verse of Bibbesworth’s Tretiz – Blaunche poudre ove groce drage – ending his description of a lavish feast runs on directly to the heavily smudged initial line of Henley – Ici comence la useboundrye. This faulty assessment arose at least in part from what may be described as the transfer of the Tretiz from England to France. One of its manuscripts had been edited by Wright (1857) but the edition does not appear to have been widely circulated, perhaps because it antedated Elizabeth Lamond’s work that brought the agricultural texts into the public domain and perhaps because it contained many errors which make it difficult to follow in places. It is absent from Oschinsky’s bibliography. In 1929 a new edition appeared, based on a different manuscript and published not in London, as the Wright edition had been, but in Paris. This new edition was submitted by Miss Annie Owen for a doctorat d’université at the Sorbonne (Owen 1929). Her introduction, notes and glossary were all in French, and whilst passing reference was made in her introduction to a number of Anglo-French texts of the period in which the Tretiz was mentioned, none of them was connected with the treatises on estate management. The thesis was examined by three eminent French philologists who regarded the opening lines of the Prologue to Owen’s base manuscript ‘G’ – Le tretiz ki munseignur Gauter de Bithesweth fist a madame Dyonise de Mountechensi pur aprise de langage – as a declaration that he intended to write a study of medieval French grammar, without regard to the remaining ten lines of the Prologue, which explain in detail that the purpose of the work was to provide the children of his patroness with all the French vocabulary of the countryside that they would later need in order to manage their own estates by husbondrie e manaungerie. Along with the rest of the agricultural vocabulary of the Prologue in prose, these two key terms that link the Tretiz to the treatises on estate

48

AN-1a.indd 58

11/03/2010 12:00:14

Husbonderie and Manaungerie in Later Medieval England management went unnoticed by the examiners in their review article.4 In line with the prevailing practice of the time, they concentrated their attention on the verse text of the Tretiz which provided rhymes necessary for the pursuit of phonology, considered to be the principal element in philology, measuring the work against the yardstick of the supposed ‘correct’ French of the continent, as if it were a faulty grammar of French produced in Paris for French readers rather than a vocabulary of the English countryside in French, composed in Hertfordshire for the training of the English offspring of landowners of French descent in an increasingly English environment.5 However, if the background to Bibbesworth’s work is taken into consideration, his Tretiz may be seen in a different light. Wright and Halliwell (1841: 134– 5) contains one of Bibbesworth’s shorter works, a dialogue in verse between the writer and Henry de Lacy on crusade. In this exchange of some seventy verses in perfectly acceptable Anglo-French he attempts unsuccessfully to persuade the count of Lincoln to go off to the crusade, but Lacy is reluctant to part from his beloved. Later, Meyer (1884) published short extracts from the beginning and end of two other pieces in verse by Bibbesworth, one in honour of the Virgin, the other in praise of women, both found in the manuscript from which Owen took her ‘C’ text of the Tretiz.6 Meyer (1884: 501) dismissed the poems as worthless, hence their severely truncated form as printed, calling the first one ‘cette ridicule composition’, but failing to appreciate the knowledge of French that must lie behind Bibbesworth’s ability to play on words, a skill evident even in the few lines retained by Meyer and reminiscent of similar passages in his Tretiz that would later be unnoticed in their turn by Owen’s examiners. Yet Meyer (1884: 501) has a footnote that would identify Bibbesworth’s patroness, dame Deonyse de Mountchensy,7 referred to in the Prologue to the Tretiz, as being the grandmother whose grand-daughter of the same name married into the de Vere family, counts of Oxford. It is highly unlikely that a personage with such an elevated French family background would have lent her name to a thoroughly incompetent piece of work. Like Lacy and the authors of the agricultural treatises, these people belonged to the aristocracy, who retained their French language along with their English lands for generations after 1066. As a consequence of the modern segregation of different kinds of writing referred to above, the subject and language of these minor pieces of Bibbesworth’s writ

4

5



6 7

Romania 55 (1929), pp. 575–9. The examiners attributed most of the manifold errors that they found in the Tretiz to Bibbesworth himself, in line with the disparaging view of later Anglo-French prevailing at that time on both sides of the Channel. However, close comparison of the edition with the individual manuscripts on which it is claimed to be based shows no evidence that they checked the text as printed, its Middle English glosses or its variant readings against any of the manuscripts, so that their findings cannot be accepted at face value, many of the errors cited in their review being of editorial, not authorial origin. Olim MS Philipps 8336, now British Library Add. 46919. Meyer prints dame Denise de Monchensy.

49

AN-1a.indd 59

11/03/2010 12:00:14

The Anglo-Norman Language and Its Contexts ing have not usually been taken into account when passing judgement on the more substantial and more important Tretiz. On a larger scale, the relationship between the agricultural treatises and the Tretiz, in particular between Henley and Bibbesworth, was similarly not recognised, the key terms husbondrie and manaungerie used in the Prologue of the Tretiz to define the purpose of the work being ignored or misunderstood. Godefroy’s dictionary (1881), quoting Bibbesworth as given by Meyer (1884), mistranslates husbondrie as ’mariage’, recognising the root of the term as being the English ‘husband’, but not the meaning ‘to store, manage prudently’ attaching to the verb ‘to husband’, or, consequently, the agricultural context of husbondrie. Bibbesworth’s accompanying synonymous manaungerie is not listed in the dictionary. Miss Owen gives ’mangerie’ (‘feast’) for manaungerie in her glossary, without any of her examiners querying this totally aberrant and obsolete term (recorded in Godefroy but absent from the dictionaries of modern French). The more recent Altfranzösisches Wörterbuch (Tobler-Lommatzsch 1958), quotes husbonderie from Bibbesworth, giving the correct gloss but, like Godefroy, omits manaungerie. Husbondrie is obviously an English word used in a French context, and manaungerie would seem to be a form of the French mesnagerie, which on present evidence is found in England earlier than in France, its attestation in Bibbesworth’s Tretiz antedating considerably its earliest entry in the dictionaries of French. Godefroy (1881) records mesnagerie no earlier than 1552, Tobler-Lommatzsch (1958) do not list it at all and the Trésor de la langue française has no example of it before 1530, but in England a non-Ciceronian managium is attested in the DMLBS for 1296 and glossed as ‘management’ in a context of land tenure. Since Anglo-Latin ‘borrows’ liberally from the two vernaculars, Anglo-French and Middle English, this would suggest that forms of the word may have been current in England from that time onwards. The attestations of it in the dictionaries of English are no earlier than Bibbesworth’s Tretiz. The Middle English glosses available to the dictionaries in works such as Bibbesworth’s Tretiz and in far greater number in Hunt (1991) are the work of a small literate minority, and it must be noted that the scribes of the Bibbesworth manuscripts from the fourteenth and early fifteenth centuries often gloss a French term in the body of the text not by an English word but by another French term regarded as being English.8 What must remain in the realm of conjecture concerning the agricultural treatises is the point at which the language of spoken communication changed from French to English as instructions were passed down from the literate landowner to his officials and from them to his illiterate peasant workers. The survival of bilingual pairs of terms such as ‘barn/grange’, ‘earth/soil’, ‘grass/herb’, ‘plum/ prune’ into modern English shows that there must have been some mingling of the languages in the countryside, a development that can be glimpsed in the

8

For example bercere/norice ‘A’ fol. 297rb, recercelez/crispe ‘A’ fol. 297va, garteres/garterys ‘A’ fol. 300ra, aveir/catel ‘G’ v.308, eschale/schelle ‘G’ v. 201, ‘G’ hume/soupe ‘G’ v. 201, hirchoun/hirchoun ‘G’ v.777 (i.e. English ‘urchin’= ‘hedgehog’) etc.

50

AN-1a.indd 60

11/03/2010 12:00:15

Husbonderie and Manaungerie in Later Medieval England French treatises, where the common English ‘acre’ is routinely found as acre9 together with the French perche.10 Yet, of the sixteen farming activities listed in Bibbesworth’s Prologue – arer, rebigner, waretter, semer, searcler, syer, fauger, carier, muer, batre, ventre e mouwere, pestre, brescer, bracer, hautefeste araer,11 only carier, batre and feste survive today as ‘to carry’, ‘to beat’ and ‘feast’. Thus the works of the two Walters and the other treatises show the importance of French in the history of agriculture in medieval England, but the failure of so many of their French terms dealing with the physical cultivation of the land to take root in English is an indication that the great contribution made by French to the lexis of modern English is not to be found equally in all branches of the language. The persistence of widespread illiteracy in England amongst the lower classes until the Education Acts of the later nineteenth century means that there can be no accurate overall picture of the language used by the large proportion of the English population that remained unlettered long after the end of the medieval period.

Dictionary sources Trésor de la langue française. Dictionnaire de la langue du XIXe et du XXe siècle (1789– 1960), ed. Paul Imbs. Paris, 1960–4.

Bibliography Godefroy, F. (1881–1902), Dictionnaire de l’ancienne langue française et de tous ses dialectes du IXe au XVe siècle. 10 vols. Paris. Hunt, T. (1991), Teaching and Learning Latin in Thirteenth-Century England. Cambridge. Lamond, E. (1890) Walter of Henley’s Husbandry: together with an anonymous Husbandry, Seneschaucie, and Robert Grosseteste’s Rules. London. Meyer, P. (1884), ‘Notice et extraits du MS 8336 de la Bibliothèque de Sir Thomas Phillips à Cheltenham’, Romania 13, 497–541. Oschinsky, D. (1971), Walter of Henley and Other Treatises on Estate Management and Accounting. Oxford. Thomson, S. (1940), The Writings of Robert Grosseteste, Bishop of Lincoln, 1235–53. Cambridge. Wright, T., and J. Halliwell (1841), Reliquiæ Antiquæ. London. Wright, T. (1857), A Volume of Vocabularies. London. Tobler-Lommatzsch (1958 [1936]), Altfranzösisches Wörterbuch. Berlin. Owen, A. (ed.) (1929), Le traité de Walter de Bibbesworth sur la langue française. Paris. 9



10

11



The dictionaries of medieval French attest acre as an English word taken into French. Perche is found on p. 264 cA of the Seneschaucy (‘written between 1260 and 1276’, p. 89), with acres on the same page. The French perche is attested in Middle English with this sense in about 1290 (OED) or 1300 (MED). Respectively, ‘to plough, hoe, cover with straw, sow, weed, scythe, mow, cart, stack, beat, winnow, grind, knead, malt, brew, organise a splendid feast’.

51

AN-1a.indd 61

11/03/2010 12:00:15

chapter five

Bridging the Gap: The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts* David Trotter

Bridges are and were complex constructions, not least because they were often either in or adjacent to water, and because structural failure had spectacular and disastrous consequences. The proximity of water virtually guaranteed a constant maintenance problem. At the same time, bridges were political, military, commercial and economic assets, since they could be used for defensive purposes as well as for the collection of tolls and rents. They are thus relatively well documented. On the one hand there are various ‘high-level’ documents (for example in the Ancient Petitions, SC 8, series in TNA) where, for example, the burgesses of towns or local magnates solicit royal assistance or royal permission to levy tolls and so forth. These, typically, are monolingual. But those documents regarding more immediate, practical matters like construction and repairs are routinely multilingual, both in England and elsewhere in those numerous regions of medieval Europe where more than one language was in use (cf. for example Coutant 1994). Precise technical instructions needed to be given to local, probably monolingual workmen, but often on the authority of their social superiors who were themselves doubtless either multilingual, or monolingual but in a different and more prestigious language. Except in the case of London Bridge (Wright 1996), very little attention has been given to these texts by linguists (cf. Trotter 2006). This paper will look at a sample of documents concerning the construction and repair of medieval bridges in Exeter, Rochester and Leicester. Analysis may enable us to draw conclusions about the respective roles of the languages concerned, of how they interact and of how they were deployed to produce practical, effective documents. Bridges, unlike castles and cathedrals, were typically in the hands of municipalities, looked after by the bourgeoisie, not the aristocracy; and unlike cathedrals or other ecclesiastical buildings, their custodians were laymen. These may perhaps seem to be obvious points, but they do have a linguistic and sociolinguistic significance. Bridges constructed, maintained and taxed * I am grateful to my uncle, John Moore, a retired civil engineer, for correcting a few technical errors in this paper.

52

AN-1a.indd 62

11/03/2010 12:00:15

The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts by municipalities were at one remove from the higher level of language use in the form of medieval church Latin; the activities which went on around and on them involved all sorts of different social classes, and thus, probably, a number of different languages. The conventional language of accounting might have been some form of Latin (or, more probably, the type of mixed business language on which Laura Wright has written with such perspicacity; see also her article in the present volume), but the workmen, the town officials, the tradesmen and the tax collectors were not necessarily themselves able to read and understand Latin. For them, then, it would have been necessary to communicate by using other languages, or at the very least by using (where appropriate and necessary) technical terminology in a language or languages with which they would have been familiar. The pattern is a very recognisable one, and by no means restricted to medieval England. Studies of mill construction in Flanders (Coutant 1994), of repairs to towns in the Pyrenees (Trotter 2000b), and of any number of buildings, castles and cathedrals across medieval Europe, all demonstrate the same, pragmatic and commonsense approach which simply uses the language which is necessary for the purpose for which it is needed. Thus it is that bridges, which have in addition the advantage of being well documented, not least because of their importance to the localities where they were found, provide a useful and productive source of data for investigation of the question of language contact, language mixing and multilingualism in general. The work of Laura Wright has already been mentioned. Concerned essentially with the vocabulary of the River Thames, Wright’s (1996) study inevitably touches on the most important bridge in medieval England, London Bridge, and there is a section of her book which deals specifically with the terminology of construction. However, whilst the most prestigious and the most important, London Bridge is by no means the only such structure, and in this paper I propose to look briefly at three others for which documentation survives, and for which that documentation is of a type which is of interest for the purposes of the study of multilingualism. These three bridges are the Exe Bridge in Exeter, and the bridges at Leicester and at Rochester. There are doubtless many more, and clearly this study could be extended. Clearly, also, whilst the bridge documents (typically in the form of the accounts of bridge wardens or similar officials) form a convenient subclass of medieval document, they are in many regards almost certainly not exceptions, and it therefore follows that any tendencies and patterns which we can discern in them may well be generalisable to a wider class of materials. The excerpts below attempt to classify, in a rather simplistic manner, the mixed-language elements in the documents which I have looked at. The underlying hypothesis is that in each case there is what specialists would now call a matrix language, within which are embedded (and the terminology varies) either loanwords or single-lexeme switches. That is to say, we are confronted here with either borrowing, or code-switching. The distinction, in any case not one which is easy to make with any certainty, is virtually impossible in the case of medieval documents, since we are now quite unable to retrieve any

53

AN-1a.indd 63

11/03/2010 12:00:15

The Anglo-Norman Language and Its Contexts information regarding the attitude of the writer. That is to say, in none of the cases which I am dealing with here is there any explicit indication of tagging or marking to indicate that the writer was either particularly concerned at the element being inserted into his matrix language, or necessarily (and this is the more contentious point) even cognisant of the fact that a change of language had taken place. There is an exception (possibly) to this, and I will return to it below; but, as we shall see, it is an exception which, far from clarifying the issue, in fact substantially complicates it. The elements admitted from languages other than the base or matrix language (and this is entirely predictable) are almost invariably substantives. This is of course a general point: in most languages, at most times, and in most types of text, the element borrowed is a substantive. There are some exceptions: the first example under (I) Latin with English elements, which features apparently a verb (rammare). But every single other instance is of a substantive.

I. LATIN WITH ENGLISH ELEMENTS 1. (L) Et ij. s. in una carrecta conducta ad carriagium gravellae1 pro dicto opere rammando per unum diem. 2. (L) Et vj. d. in pollis emptis pro uno scaffoldo2 ibid faciendo pro dicto opere. 3. (L) Et vij s. iij d. tam in lx latthis3 emptis clx clavis ad eosdem Pynnes4 pro tegulis emptis quam in tegulis carians de aula comunitatis 4. (R) Item in Maeremio quercino empto ad faciendum unum collistrigium j Thewe5 j schevyng stol6 et j par de stokkys ad custodiendum libertatem dominii de Tyllebery . . . precium xs. (1399–1400; Becker 94) 5. (R) In iiijxx xij longis petris vocatis Endestones mense Junii emptis . . . lxjs iiijd. (1425–6; Becker 75)

1



4 5 2 3

6



DMLBS gives this (‘gravel’) as [Anglo-Norman, Middle English] from 1209; both gravella (f.) and gravellum (n.) are attested, so the expansion of the inflexion (here, suspension mark = lł) is conjectural. Middle English < Anglo-Norman. Old/Middle English from 1130 onwards (DMLBS 2 lata). RMLWL 2 pinna, ‘tile-pin’, from 1316. ‘A punitive device for female offenders consisting of a raised platform with an attached post to which the transgressor would be tied’: MED theu (n.(2)). ? ‘shedding-’ or ‘showing-’? The latter seems more likely.

54

AN-1a.indd 64

11/03/2010 12:00:15

The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts 6. (R) Item eidem Willelmo (Champeneys, lathomo) pro j c Carectata de frestoun7 empto . . . vjs. viijd. (1426–7; Becker 75) 7. (R) In viiij tontigh8 de Rag emptis cum Johanne Cartere pro ponte paviando (1435–6; Becker 75) 8. (R) Et soluti Johanni Smyth de Cheteham pro xxviij pilshoun (ed.: ‘pile shoes’)9 ponderantibus ccccij libris pro libra jd. ob. (1438–1438; Becker 76) 9. (R) Et de jd. soluto pro factura unius claps10 pro libro in Capella (1443–4; Becker 26) 10. (R) Et de xxd. solutis pro factura et scriptura unius rotulis [sic] vocati A bederolle11 de nominibus benefactorum pontis tam vivorum quam defunctorum (1444–5; Becker 21)

Conventional accounts of borrowing (if that is what this is, which is of course itself a matter of some debate) would probably see in borrowings of this type recourse to another language because the matrix language was lexically deficient. Since, invariably, what we are dealing with is, in sociolinguistic terms, the incorporation into a higher matrix language of an element from a lower target language, the idea that there might be any prestige associated with the borrowings is to be rejected. If a Latin text uses English or French, or a French text uses English, then I think we can safely conclude that the prestige borrowing thesis can be dismissed. A question then might be whether we can plausibly see this as a case where the borrowing fills a lexical gap. It is possible that the lexical items under (4) fall into this category of necessary borrowing, if, that is, we accept that medieval Latin did not have any word to correspond to the devices of punishment being constructed or established here, or that (5) ‘endstones’ were unknown to Latinate masons. Harder to believe is that some of the other terminology, within the same group or elsewhere, was really lacking in the matrix language. It is certainly the case for example that in II (French with English elements), none of the terms which have been used in the texts and taken from English, with the possible example of the penultimate (not attested elsewhere and therefore inherently problematic), would have been unavailable in Anglo-Norman:

9 10 11 7

8

‘A fine-grained limestone or sandstone, freestone’: MED frēstōn. ‘A unit of weight equal to the amount contained in a full tun’: MED tŏnne-tight. Not attested MED although sub shō, the compound with plough- is featured. ‘A clasp for the covers of a book’: MED claspe. Cf. MED bēde, but the compound is not attested.

55

AN-1a.indd 65

11/03/2010 12:00:15

The Anglo-Norman Language and Its Contexts II. FRENCH WITH ENGLISH ELEMENTS 11. (E) Item pour la fesure de xij flakes;12 viij s. (‘Item, for the making of two hurdles, 12 s.’) 12. (E) Item pour ij weigges de feer; vij d. (‘Item, for two iron wedges, 7d.’) 13. (E) Item a un garsoun lowé de aler a Chuddelegh aprés le charetter j d. Item pour ij berewes; iij d. (‘Item, for a servant hired to go to Chudleigh with the carter, 1d. Item, for two barrows [barrowfuls], 3d.’) 14. (E) Item j saghiere (?); xiiij d.Item pour cariours de (flun?); pour j jour (‘Item, a sawyer, 14d. Likewise for carriers on the river [?], for a day.’) 15. (E) Item a Wauter Godale standelvere;13 pour iiij xx piés de piere xx s. [. . .] Item a Roger Polman standelvere pour xxvij piés de piere vj s. (‘Item, for W. G. quarryman; for 80 [cubic] feet of stone, 20s. [. . .] Item, to R. P. quarryman for 27 [cubic] feet of stone 6s.’) 16. (R) Item payé in expense pur abater Brounche coper et scharper (ed.: ‘sharpen’)14 prest a Caryage ccclxxv olmewiz15 pernaunt pur le pese (ed.: ‘for the piece’) iijd. (1409–1410; Becker 69) (‘Item, paid in expenses to cut down branches, cutting and trimming ready for transport, 375 elm trees, charging per tree 3d.’) 17. (R) Item payé a x tydemen pur chacer16 pylles par xij Tydez et ij a ayder par un tyde pernaunt checum par le Tyde iijd. (1409–10; Becker 82–3) (‘Item, paid for ten tidemen for driving piles during twelve tides, and two men to help per tide, each one paid 3d. per tide’) 18. (R) Item payé pur un Batew achaté de Nicolaus Hobbe le quel fut Mellys de Strode, pris xxvjs viijd. Item en Goddisselver17 et en bevereche iijd. ob. (1409–1410; Becker 84) (‘Item, for a boat bought from N. H., formerly belonging to M. de S., price 26s. 8d. Item, for God’s penny and for drinks 3d.’) 19. (R) Item demandé pur [. . .] le fforsure del drawbrege ové Merym et iiij Gros boltz de ferre (1416; Becker 110) (‘Item, requested for the reinforcing of the drawbridge with timber and four big iron bolts’)

Anglo-Norman could happily have coped with all of this. Indeed, the interesting composite term in quotation 12 (from Exeter) indicates that we are faced 14 15 16 17 12 13

‘Hurdles, wattled frames’: MED flēke. ‘Stone-digger, quarryman’: MED stōn + delver. Or < shāpen (MED, s.v.) which seems more likely. ‘Elm tree’, i.e. wych elm, ulmus montana: MED elm(e, wich(e (1). AND only has this sense for ‘to drive (nails)’. Cf. OED God’s penny: ‘A small sum paid as earnest-money on striking a bargain, esp. on concluding a purchase or the hiring of a servant’?

56

AN-1a.indd 66

11/03/2010 12:00:16

The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts here not with a choice out of necessity, but with a choice plain and simple. Weigges could perfectly well have been expressed as (Anglo-Norman) coings. The adjacent de feer is indicative of the extent to which the author was probably operating in Anglo-Norman, for him the ‘matrix language’. The most likely explanation for this category is that these were technical terms, known to the workmen, and consciously or unconsciously used because of that. (I shall come back to this question of whether the process was conscious or unconscious.) The same might well be true for the English elements in the Latin texts (I), for example the technical terms about types of stone, or the use of the word scaffoldo (2), or latthis (3), or even the pins used to fix the tiles (3). It is probable that the necessary terminology existed in Latin as well, but it is also probable that it would have been far more readily understood by the men who actually had to do the work if it was expressed in English. English in Latin, and English in French, are clear cases where a low-prestige language (as far as writing is concerned) is incorporated into one or other of the legitimate, traditional post-Conquest languages of record. I come now to the middle component (in sociolinguistic terms), that is, apparently French elements within a Latin matrix-language document. This is a little more complex because here we have the two accepted and acceptable languages of record intermingling:

III. LATIN WITH ENGLISH OR (APPARENTLY) FRENCH ELEMENTS 20. (L) Et ij d tamen in sceynture pro le archam supponendam quam in stipendiis diversorum pro lapides colligendas ibidem in aqua ad dictum opus 21. (R) Item in ferro empto ad ligandum le18 Stokkys ponderanti xxxv libras cum clave ante facta precium le clavis ijd., vjs ijd., cum clave (1399–1400; Becker 94) 22. (R) Et soluti Johanni Cartere pro M. Assheler ab eo empto pro le Eastbrest ad finem pontis . . . viiij li. xvjs viijd. [. . .] Item pro xxij pedibus de vawsour19 xviijs iiijd. (1430–1431; Becker 75) 23. (R) Et de xxs. solutis Johanni Hassok pro le Stoling20 Capelle [i.e., the Bridge Chapel on the river-bank] predicte ultra iijs iiijd. datos per capellanos (1443–4; Becker 20)

Or: lé = les? MED vousŏur, cf. AND vousure, ‘wedge-shaped stone (for building an arch)’. 20 Gerund of the (rare) MED stōlen (one quotation only of the transitive verb from 1475, which this one antedates), a derivative of stōle, itself of Anglo-French origin. 18 19

57

AN-1a.indd 67

11/03/2010 12:00:16

The Anglo-Norman Language and Its Contexts 24. (R) Et de vs. solutis a lez verymen21 ibidem pro regardo22 per batellagio meo pro diversas vices ij annos (1444–5; Becker 104) 25. (R) Et de vjd. solutis pro le clapysing23 unius libri (1449–50; Becker 26) 26. (R) Et de xiiij li. iijs iijd. solutis xviijcim hominibus conductis ad trahendum le Ram24 pro affirmatione pilarum videlicet cuilibet eorum per lviij tides et dimidium quilibet capit per le Tyde iijd. Et ijd. ob. solutis pro ij libris cepi pro unctione magni hauuser.25 Et de vjd. pro terryng26 eiusdem. Et de ijs viijd. solutis pro ix herdell’27 emptis pro scafoldes pro le Ram (1457–8; Becker 82)

I stress that these words are only apparently French because the distinction between a French or an English lexeme in a Latin text is far from obvious, to such an extent that in the MED, many such attestations are only questionable instances of ‘English words’. Often, there is a suspicion that they are so designated because of an inherently teleological method in the MED: the words are certainly found in English, and are ‘English’, later on, so they are regarded as being ‘English’ from the outset even if the context in which they appear is not English and even in the absence of incontrovertible (or indeed, any) proof that they were perceived at that time as English. In the final example here (26), there are a number of problems: the text is clearly in Latin, almost certainly abbreviated (but the editor has concealed this), but the words which have been incorporated into that Latin are themselves apparently drawn from both English and French. There is a very real difficulty in determining whether a word like scafoldes (as it has been edited here) or hauuser (or hawser?) is English or French, or even Latin, which depends critically on whether a final suspension mark has been expanded (see Wright, passim). Pretty obviously, the same problem could be held to surface in the first example in this section (20), where I have chosen to expand the suspension mark on the end of the word arch’, and thus to turn it decisively into a Latin word. Implicit in the analysis of whether a given word is a loanword or not are certain interpretations (more often, assumptions) about whether its incorporation is conscious, and whether, indeed, for the putatively bilingual speaker/ writer concerned, a real distinction existed between the lexical items in what are now, for us, categorically distinct languages or lexical sets (for the general point in a modern analysis, cf. Glessgen 2007: 115–17). The amount of research now available from a neurological perspective, and the variety of conclusions The form is not listed sub MED feri-man. Clearly a calque of Anglo-Norman regard (AND, s.v.). 23 ‘Furnishing (a book) with clasps’: MED clasping, although all the quotations have the form here, clapsyng. 24 ‘Pile-driver’, one of the senses listed in MED ram. 25 ‘Hawser’: MED hauser < Anglo-Norman. 26 I.e. ‘tarring’, to protect it against rust(?): MED terren. 27 MED hirdel, with this sense (i.e., ‘hurdles’ used in the construction of scaffolding). 21

22

58

AN-1a.indd 68

11/03/2010 12:00:16

The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts reached, is truly bewildering (for example, Crinion 2006; Perani and Abutalebi 2005). To this must be added the explosion in so-called ‘cognitive’ linguistics (e.g. Grzega 2003). That this work is often difficult is not a compelling reason for it not to be integrated into work by linguists on bilingualism and on related phenomena such as lexical borrowing/code-switching. Finally, the examples in the last category (III) raise also an unexpected and apparently unexplored problem: the status of the article ‘le’. Logically this would be an indicator of a following French (Anglo-Norman) substantive, but it is in fact followed in every case here by an English noun. This is directly comparable to the similarly striking case of English compound place-names of the ‘Chester-le-Street’ type. An examination of a list of these, derived from the material available on the website of the Survey of English Place-Names Project in the Institute for Name-Studies, Nottingham28 shows that almost without exception, ‘le’ as an element in place-names introduces an English word. ‘La’, however, is occasionally followed by a word which is French: Adwick le Street (Yorkshire WR); Appleton le Moors (Yorkshire NR); Appleton le Street (Yorkshire NR); Aston le Walls (Northamptonshire); Barnetby le Wold (Lincolnshire); Barnetby le Wold (Lincolnshire); Barnoldby le Beck (Lincolnshire); Barton le Street (Yorkshire NR); Barton le Willows (Yorkshire NR); Bolton le Sands (Lancashire); Brampton en le Morthen (Yorkshire WR); Burgh le Marsh (Lincolnshire); Carlton le Moorland (Lincolnshire); Chapel en le Frith (Derbyshire); Chester le Street (Durham); Clayton le Dale (Lancashire); Clayton le Moors (Lancashire); Clayton le Woods (Lancashire); Dalton le Dale (Durham); Font Le Rio (Dorset); Gayton le Marsh (Lincolnshire); Gayton le Wold (Lincolnshire); Hamble le Rice (Hampshire); Haughton le Skerne (Durham); Hetton le Hole (Durham); Holton le Clay (Lincolnshire); Holton le Moor (Lincolnshire); Houghton le Side (Durham); Houghton le Spring (Durham); Hugglescote and Donington le Heath (Leicestershire); Hutton le Hole (Yorkshire NR); Kirmond le Mire (Lincolnshire); Laughton en le Morthen (Yorkshire WR); Maltby le Marsh (Lincolnshire); Mareham le Fen (Lincolnshire); Marton le Forest cum Moxby (Yorkshire NR); Marton le Moor (Yorkshire NR); Newton le Willows (Yorkshire NR); Normanby le Wold (Lincolnshire); Normanton le Heath (Leicestershire); Norton le Clay (Yorkshire NR); Poulton le Fylde (Lancashire); Preston le Skerne (Durham); Stainton le Street (Durham); Stainton le Vale (Lincolnshire); Stretton en le Field (Derbyshire); Sturton le Steeple (Nottinghamshire); Sutton le Marsh (Lincolnshire); Thornton le Beans (Yorkshire NR); Thornton le Clay (Yorkshire NR); Thornton le Fen (Lincolnshire); Thornton le Moor (Lincolnshire); Thornton le Moor (Yorkshire NR); Thornton le Moors (Cheshire); Thornton le Street (Yorkshire NR); Thorpe le Street (Yorkshire ER); Thorpe le Willows (Yorkshire NR); Walsham le Willows (Suffolk); Walton le Dale (Lancashire); Welton le Marsh (Lincolnshire); Welton le Wold (Lincolnshire); Wharram le Street (Yorkshire ER); Whittle le Woods (Lancashire); Witton le Wear (Durham); Capel-le-Ferne (Kent); Kirby-le-Soken (Essex); Stanford-le-Hope 28

http://www.nottingham.ac.uk/~aezins//kepn.php

59

AN-1a.indd 69

11/03/2010 12:00:16

The Anglo-Norman Language and Its Contexts (Essex); Thorpe-le-Soken (Essex); Walton-le-Soken (Essex); Ashby de la Launde (Lincolnshire); Ashby de la Zouche (Leicestershire); Fisherton de la Mere (Wiltshire); Kirkby la Thorpe (Lincolnshire); Layer de la Haye (Essex).

One possible interpretation (within my class (III) list) is that le, far from introducing French (as one might expect from a French definite article), simply signals a switch from Latin to vernacular – and vernacular is a category which includes both French and English, perhaps (and this is where the cognitive and neurological aspect resurfaces) not distinguished by the writer. Documents of this type, in other words, function not with a ternary (Latin–French–English) system, but a binary (Latin–vernacular) mechanism. This hypothesis perhaps merits further investigation.

Bibliography primary sources (E) = Exe Bridge Wardens’ Account of 1349 (Devon Record Office, Exeter) (L) = Leicester Borough Records BR III/1/48: Expenses of the West Bridge (Mayor’s Account, 1365/6) (cf. M. Bateson (ed.), Records of the Borough of Leicester, Leicester, 1899, vol. II, 140) (R) = Rochester Bridge accounts: Excerpts from the accounts of the Bridge Wardens, 1398–1479, from M. Janet Becker, Rochester Bridge: 1387–1856. A History of Its Early Years, Compiled from the Wardens’ Accounts. London, 1930.

secondary sources Coutant, Y. (1994), Middeleeuwse molentermen in het Graafschap Vlaanderen. Terminologie du moulin médiéval dans le comté de Flandre. Tongeren. Crinion, J., et al. (2006), ‘Language control in the bilingual brain’, Science 312, 1537–40. Glessgen, M.-D. (2007), Linguistique romane. Domaine et méthodes en linguistique français et romane. Paris. Grzega, J. (2003), ‘Borrowing as a word-finding process in cognitive historical onomasiology’, Onomasiology Online 4, 22–42, at http://www1.ku-eichstaett.de/SLF/ EngluVglSW/OnOn.htm Hunt, T. (2000), ‘Code-switching in medical texts’, in Trotter (2000a), pp. 131–47. Perani, D., and J. Abutalebi (2005), ‘The neural basis of first and second language processing’, Current Opinion in Neurobiology 15/2, 202–6. Rothwell, W. (1980), ‘Lexical borrowing in a medieval context’, Bulletin of the John Rylands University Library of Manchester 63, 118–43. ——(2000), ‘Aspects of lexical and morphosyntactical mixing in the languages of medieval England’, in Trotter (2000a), pp. 79–92. Schendl, H. (2000), ‘Linguistic aspects of code-switching in medieval English texts’, in Trotter (2000a), pp. 79–92. Taylor, K. (2005), Social Aesthetics and the Emergence of Civic Discourse from The Shipman’s Tale to Melibee, The Chaucer Review, 39, 298–322.

60

AN-1a.indd 70

11/03/2010 12:00:16

The (Socio)linguistic Evidence of Some Medieval English Bridge Accounts Trotter, D. (ed.) (2000a), Multilingualism in Later Medieval Britain. Cambridge. ——(2000b), ‘Le Prince Noir et les dominicains de Bagnères-de-Bigorre’, Annales du Midi 114, 365–8. ——(2006), ‘Language contact, multilingualism, and the evidence problem’, in U. Schaefer (ed.), The Beginnings of Standardization: Language and Culture in Fourteenth-Century England, Frankfurt, pp. 73–90. Wright, L. (1996), Sources of London English: Medieval Thames Vocabulary. Oxford. ——(2002), ‘Code-intermediate phenomena in medieval mixed-language business texts’, Language Sciences 24, 471–89.

Annexe: (Tentative) transcriptions of unpublished documents (E and L) E = Exe Bridge Wardens’ Account of 1349 (Devon Record Office, Exeter) (recto a) Paié pour merym a faire le gortz; xi s. [. . .] Item paié a diz est esteymours pour (par?); mesme la semaigne xx s. iiij d.pernaunt (?); chescun par la semaigne ij s. Et le mestre iiij d. outre [. . .] Item paié pour cariage de cj summe de merym xij s. vij d. [. . .] Item pour iiij charges de merym v s. iiij d. pernaunt (?) pour la charge xvj d. [. . .] Item paié a Wauter Godale masoun de Silferton pour le oeveraigne des piqs en la quarere pour cxiij piés xxviiij s. ix d. pernaunt (?) pour le pee iij d. [. . .] Item paié pour le cariage des dites piers c’est a savoir pour ix charges de wayn xiij s. vj d. pernaunt (?) pour le charge xviij d.Item pour le cariage de xlij summes de piere a chival [. . .] Item lour cortesie; v d. ½ [. . .] Item paié a trois homes lowés d’abatre verges a bois pour ij journeez [. . .] Item pour la fesure de xij flakes; viij s. [. . .] Item a un home et j chival lowé de carier merym tote la semaigne entier ij s. [. . .] Item paié a les oeverours et carpenters pour lour cortesye tote la semaigne entier xiij s. iij d. ½ [. . .] Item pour ij weigges de feer; vij d. [. . .] Item a un garsoun lowé de aler a Chuddelegh aprés le charetter j d. Item pour ij berewes; iij d. (recto b) Paié a sese esteymours pour vj jours xxxiij s. x d.Item un messager qui ala aprés les esteymours ij d.Item pour cariage de xxxvij summes de piere [. . .] Item pour cariage d’un charge de piere xviij d. [. . .] Item pour cariage de iiij xx summe de merym et de verges x s. pernaunt (?) pour la summe 1 d ½. Item pour xxxij altres oeverours allowés mesme la semaigne xliiij s. xj d. [. . .] Item vij carpenters mesme la semaigne ix s. Item j saghiere (?); xiiij d.Item pour cariours de (flun?); pour j jour [. . .] Item pour diverse apparail achaté al charette viij s. iiij d. [. . .] pour cariage de iij charges de piere iiij s. vj d. [. . .] Item a Wauter Godale standelvere; pour iiij xx piés de piere xx s. [. . .] Item a Roger Polman standelvere pour xxvij piés de piere vj s. [. . .] Item paié a Roger Roche mestre des masouns xx s. en partie de paie de xli s. de son lower [. . .] Item paié pour le courtesie a toutz les oeverours de mesme la semaigne xj s viij d. ½ [. . .]

61

AN-1a.indd 71

11/03/2010 12:00:16

The Anglo-Norman Language and Its Contexts L = Leicester Borough Records BR III/1/48: Expenses of the West Bridge (Mayor’s Account, 1365/6) (cf. translation in Bateson, Records of the Borough of Leicester ii 140) Expense pontis occidentalis29 [1]30 Et xxx s. in stipendiis iij. cementariorum conductorum per tres septimanas operantes circa pontem occidentalem quia capiunt per septimanam iiij s. iv d [2] ex conventione ad tascum. Et xxij s. in stipendiis iiij. hominum serviendum ibidem per xvj dies dimid. in facta [?] tempore p’tnu [?] [3] et ij in quarrera et ij. servientes cementarios predictos quia capiunt iiij d. in die ad tascum. Et x s. in viiij quartaribus [4] calcis emptis cum carriagio pro predicto ponte videlicet quam ad xv d. Et ij s. in sabilone carriata ad idem. [5] Et xij. s. datur diversis laboratoribus Abbatis Leycestrensis in quarrera ibidem ut adiuvarent in dicta quarrera per [6] diversas vices. Et vj. d. datur Johanni de Sleford pro labore suo circa dictum opus pro supervisione. Et ij. s. in [7] una carrecta conducta ad carriagium gravellae pro dicto opere rammando per unum diem. Et vj. d. in pollis emptis [8] pro uno scaffoldo ibid faciendo pro dicto opere. Et ij. d. in clavis emptis ad idem opus. et viij s. in ?? [9] carrecta conducta pro lapidis cariandis de quarr[era] per iiij dies videlicet per diem ij. s. ex conventione Et iiij s. in [stipendiis?] [10] diversorum conductorum pro dicto ponte pavendo in grosso. Et vij s. iij d. tam in lx latthis emptis clx clavis [11] ad eosdem Pynnes pro tegulis emptis quam in tegulis carians de aula comunitatis una cum stipendio diversorum te[12]gulatorum cooperientes cameram super dictum pontem cum diversis neccessariis ut supra sic ex conventione facta in grosso [13] Et ij d. in stipendiis ffabrorum pinctorum ferramentum cementariorum predictorum infra predictum tempus Et ij d tamen in [14] sceynture pro le archam supponendam quam in stipendiis diversorum pro lapides colligendas ibidem in aqua ad dictum opus.

Written in left margin. Line numbers relate only to the selected paragraph, not to the folio.

29 30

62

AN-1a.indd 72

11/03/2010 12:00:17

chapter six

From Apareil to Warderobe: Some Observations on Anglo-French in the Middle English Lexis of Cloth and Clothing* Mark Chambers and Louise Sylvester

The origins of the present project lie in Gale Owen-Crocker’s research into the dress and textiles of the Anglo-Saxon period (Owen-Crocker 2004; OwenCrocker and Netherton 2005–6). The project is entitled ‘The Lexis of Cloth and Clothing in Britain c. 700–1450: Origins, Identification, Contexts and Change’. As the title suggests, it has a twin focus: dress and textiles; and the vocabulary that existed for them in medieval Britain. What we are interested in is the question of how textiles and garments were named. The vocabulary of the various languages spoken and written in the British Isles is documented in different specialist dictionaries, yet geographical proximity and interaction through labour and trade indicate that this evidence should be categorised and analysed together. The Lexis of Cloth and Clothing project prompts us to think both semasiologically (what did this word mean?) and also onomasiologically (how was this object described in language?). In practice, this means that investigators on the project will analyse terms and their citations from both documentary and literary texts in the light of surviving textiles and dress accessories, and graphic images in medieval art, investigating the complex relationships between vocabulary, artefact and image. We will also include definitions in modern English of medieval technical processes and artefacts. This orientation towards interdisciplinary material will, we hope, enable a re-examination of the medieval vocabulary and potentially the redefinition and refinement of existing information found in established medieval dictionaries. We will catalogue the sources of the lexis, questioning whether the same naming practices governed the Celtic, Germanic and Romance languages of Britain. For example, medieval English garment-names derive variously from the part of the body they cover, from the cut or shape of the garment, from the texture of its fabric; cloth may be named from its technique, place of origin, * We should like to acknowledge our gratitude to the Arts and Humanities Research Council for the funding with which to undertake this project.

63

AN-1a.indd 73

11/03/2010 12:00:17

The Anglo-Norman Language and Its Contexts real or supposed, or from its trade route or the merchants who carried it. Even allowing for the limitations of relevant texts, regional and dialectal variations within the languages may emerge. We will try to contextualise the introduction of loan words within historical and cultural developments; for example, the Flemish textile trade or the popularity of French, Italian and Latin literary texts. Loan words are a particular focus of interest for questions of semantic shift: did loan-words (as well as other lexical items) regularly or rapidly become divorced from their original sense? If this is so, can any socio-economic reasons explain this? At the heart of the project is the assembly and examination of textiles and clothing lexis in the early languages of Britain, and an investigation of the genesis and subsequent development of the vocabulary. Our plan is to produce a data set combining Old and Middle English supplemented by data from Anglo-Norman and the other languages in use in the British Isles between c. 700 and 1450.1 This data will be examined comparatively and etymologically, with a view to identifying relationships between the early languages of the British Isles, and to examining inter-language developments in relation to social, political and cultural factors and contextually, identifying, re-examining and cataloguing by genre the texts in which terms occur, and comparing the data derived from different sets and subsets. Finally, data, definitions, etymologies and images will be combined in a searchable database held on the internet for the use of scholars and for anyone who is interested in any of the nomenclature associated with dress and textiles in the British Isles in the medieval period. It is important that we strive to save Anglo-Norman from being underacknowledged, as has occurred with many similar collections of British medieval lexica in the past (for a useful critique of English lexicography in this area see Rothwell 2001). The efforts of scholars working in Anglo-French lexicology provide a valuable resource for a hitherto largely under-appreciated facet of the British medieval lexicon, and the Lexis of Cloth and Clothing project intends to take this research into consideration while collecting and refining its data. In our title, we employ the Middle English forms of appareil and wardrobe, two words dealing with clothing and both carrying senses that were coined in the French of medieval Britain. In an article arguing that Anglo-French provides the missing link in English etymology, William Rothwell discusses the Anglo-Norman provenance of apareil: ‘In Medieval French on both sides of the Channel apareil etc. meant “preparation(s)”, “thing(s) prepared”, hence “equipment”, “furniture”, “gear” of various kinds’ (Godefroy 1888, I, 317–18; AND I, 32): only in Anglo-French, however, was the term applied specifically to clothing (Rothwell 1991: 175). Here, a common word in modern English for 1



Much of the Old English vocabulary has been collected by Owen-Crocker. We are grateful to Frances McSparran and staff working on the Middle English Dictionary and to David Trotter and the team at the Anglo-Norman Dictionary for their scholarly generosity towards our project.

64

AN-1a.indd 74

11/03/2010 12:00:17

From Apareil to Warderobe clothing (apparel) has drifted quite substantially from the semantic valence of its modern French cognate, appareil. As Rothwell has demonstrated, both the OED and the MED fail to acknowledge the Anglo-French development of the word. In a later article written with Lisa Jefferson, Rothwell draws our attention to the office of garderobier or warderoper, which are both mentioned in the AngloNorman accounts of the London Merchant Taylors’ Company, for Henry IV and Henry VI respectively (Jefferson and Rothwell 1997: 289). The MED notes that the word warderobe (or a variant) could mean anything from ‘a room for storage’ to ‘a badger’s excrement’ to a very important royal office, and it suggests that the word made its way into English from Old French garderobe through the north-eastern variety. The dictionary does not, however, provide details of the Anglo-Norman aspect of the royal office. After demonstrating the deficiencies in attestations provided and/or definitions offered, Jefferson and Rothwell suggest that the major dictionaries leave their readers ‘in ignorance of the great extension of the powers and scope of the Wardrobe and its officials and those who served in it under a number of medieval English monarchs’ (1997: 290; see also Tout 1928 cited there). This is precisely the sort of extra-dictionary, lexicological assessment that the Lexis of Cloth and Clothing project is keen to take on board and to develop as it seeks to provide a full, multi-disciplinary corpus of the language of cloth and clothing in medieval Britain and to make it as accurate and up to date as possible. David Trotter has long been a proponent of the need for research of the lexis of the Middle Ages to take into account the multilingual situation of Britain in this period (see, for example, Trotter 2000: 1–5). He has recently commented on the ‘movement and instability in the lexis in the fourteenth century’, calling for ‘multilingual, interlinked electronic databases [. . .] or [. . .] genuinely polyglot new dictionaries’, both of which he admits present considerable challenges (2006: 81). While not pretending to offer a full etymological dictionary, our project aims to provide evidence of usage in all of the languages of medieval Britain (including, but not limited to, Anglo-Norman, Cornish, Flemish, Irish, Medieval Latin, Manx, Middle English, Norn, Norse, Old English, Old Scots, Scots Gaelic and Welsh), to include etymological notes, and to provide links to and from other major electronic dictionaries (such as the MED and AND). In its final, public version, we intend to present the lemmata in such a way as to demonstrate with accuracy the polyglot nature of medieval usage and to break down modern notions of strict language classification, the ‘possibly anachronistic metalinguistic designation’ (Trotter 2006: 80) which would have had little meaning to the medieval speaker or writer. Given the scope of the project’s aims, we decided to begin with a general search and comparison of the various dictionaries available, and then to refine this information through comparison with later specialist research, historical records and archaeological evidence. In order to fulfil our first remit to generate a working list of Middle English headwords and to begin refining definitions, etymologies and orthographies, we conducted initial searches for definitions

65

AN-1a.indd 75

11/03/2010 12:00:17

The Anglo-Norman Language and Its Contexts containing various parts of the body, such as hand, neck, forearm, etc., as well as various modern terms for garments, accessories and textiles. It was immediately apparent that Middle English words for armour, in particular, often develop names from Anglo-French, rather than the French of medieval Paris. The MED notes, for example, that the adjective quisheld – which describes a pair of armoured boots or leggings in the late-fourteenth-century York Memorandum Book – seems to stem from Anglo-French quissel (see AND, s.v. quisser), or modern cuisse, armour for the thighs.2 Likewise, most of the dictionaries note the Anglo-French nature of rēredos, meaning a decorated screen behind an altar or else a decorated wall panel. In Middle English, however, it comes to mean a piece of armour worn on the back, as noted in the fifteenth-century Fabric Rolls of York Minster, and by Lydgate in his Troy Book.3 Thus the names for the trappings of the knight speaking in fourteenth-century Britain often have their etymological origins in insular rather than continental varieties of French. Searching the dictionary definitions for specific parts of the body gives an immediate taste of the Anglo-French flavour of the clothing lexis in use in medieval Britain. In the late-fourteenth-century Wycliffite Bible, for example, we are told that ‘Judith bound up the tresses of her hair with a coronal’, meaning a sort of coronet or a diadem (Forshall and Madden 1950: ‘Judith’ 16.10). The MED suggests that the Anglo-French word coronal may be in play here, but with the sense of ‘the head of a spear or tilting lance’ or else ‘the capital of a column’, but the AND cites a contemporary, anonymous Anglo-French Chronicle which uses the word coronal to mean a headdress, and we find the word in Anglo-French applied to Christ’s crown of thorns as early as the thirteenth century (Childs and Taylor 1991: 146).4 Lexicographical scholarship indicates that particular Middle English usages in the semantic field of cloth and clothing stem from Anglo-French. For example, the Middle English noun cressaunt, in its sense of a crescent-shaped ornament – rather than referring to the moon itself – seems to derive from Anglo-French usage. According to both the MED and the AND, it appears in this sense in Anglo-French manuscripts of the fourteenth century, and again in Anglo-Latin accounts of the fifteenth. The verb ‘to line’, meaning to add material to a garment, is apparently an Anglo-French creation. In his discussion of the French vocabulary of the

2



3



4

quisheld, adj., in MED. The only attestation noted looks to be a ‘lexeme switch’ from a Latin entry in the York Memorandum Book from the 1420s or 1430s: ‘Item, ubi solvere solebant pro factura xij parium ocrearum in mundum, que linate fuerint, quyssheld, lased vel clasped’ (Sellers 1912: 194, our emphasis). Lydgate writes: ‘Some chose . . . For to be sure myd of al her foos An hol brest-plate with a rere-doos Be-hynde schet’ (Bergen 1908: 62). ‘coronal, coronet: le corinal la roigne cheist de sa teste et fust debrusé’ (Child and Taylor 1991: 146); for the use of coronal to describe the Crown of Thorns, the AND cites Jean de Howden’s Le Rossignol (Corpus Christi College, Cambridge, MS 471) suggesting a date before 1275.

66

AN-1a.indd 76

11/03/2010 12:00:17

From Apareil to Warderobe Archive of the London Grocers, Rothwell notes that the grocers were lining hoods – liner chaperons – and using a verbal form of the noun lin well before any continental examples (1992: 35). The verb appears in English in 1440 in the Promptorium Parvulorum.5 Rothwell also discusses the noun lineure, which apparently refers to a type of linen garment in the 1454 Proclamations of the Privy Council, but which appears as early as 1345 in the grocers’ Anglo-French accounts (1992: 35). The Anglo-French adjective estreit apparently gives us Middle English streit in the sense of ‘tight-fitting clothing’ or ‘tightly pulled sewing or binding’ (see MED and AND); and, in the late fourteenth century, we find the similarly unprefixed Middle English word volupēre used to describe a headdress, derived from the Anglo-French form meaning bandage (see MED, s.v. volupeure). These examples cannot help but appear atomistic, but our hope is that in assembling all the lexis of this semantic field in use in all the languages of the British Isles in this period, we will be able to draw some wider conclusions about the different languages and their valence in this period. Before we can do that, however, the vocabulary must be assembled and checked. Below we set out some initial evidence of Anglo-French lexical contact in Middle English words for cloth and/or clothing, using the major dictionaries available to us alongside specialist research of which we have been made aware. These words do not represent finished entries and are certainly not treated comprehensively. They are, rather, indications of the kind of work the project is conducting in its opening phase. We hope that they and subsequent published efforts will generate interest and feedback from scholars who are interested in the lexicography of medieval Britain, and we welcome suggestions or emendations.

Blēaunt Costume historians note that in the medieval period, the French word bliaut typically (but not exclusively) referred to a long, fitted garment with wide sleeves, skirted at the bottom, worn by aristocratic males and females and probably brought from the East during the crusades (see Norris 1999: 31–41; Houston 1996: 220; see also Gardner 1950: 63–5). The MED records that the Middle English noun bleaunt, meaning ‘a costly silk fabric’ or a ‘garment’ or ‘a bedspread made from this garment’, finds its way into the English lexicon through Anglo-Norman usage, with its first recorded attestation appearing in the thirteenth century. The OED does not mention an Anglo-Norman connection, but refers to Old French bliaut/bliaud/bliat/blialt, and cites cognates in many of the Romance languages. The editors note the presence of the n before the final t in the English variants, which is not present in the Old French or the other earlier Romance languages but does appear in a few Middle Low Ger5



‘Lynyn clothys: Duplo, duplico’, Promptorium Parvulorum (c. 1430) (British Library, Harley 221) (Way 1843: part 1, 306).

67

AN-1a.indd 77

11/03/2010 12:00:17

The Anglo-Norman Language and Its Contexts man examples and perhaps once in a mid-fourteenth-century text from medieval Picardy.6 The AND records a variant for bliaut which contains an n (bliaunt), but does not provide the variant’s context. We find the word first attested in the French of Britain in the late twelfth century. In the Anglo-Norman Romance of Horn it is spelled bliaut (‘E il iert bien vestu d’un bliaut de cendal’, 1964: 575) reflecting the Old French spelling and carrying the sense of a tunic or garment, probably like the one described above, in this case made from sendal, a rich, silken material. While we must allow that unpublished, often non-literary, material is certain to include attestations to which we do not now have access, a general survey of the Middle English texts presently available demonstrates the word bleaunt was popular in Middle English literature, particularly in the thirteenthand fourteenth-century romances.7 It could refer to either the material or the garment, depending on the context; in fact, the author of Sir Gawain and the Green Knight refers to both: ‘a meré mantyle watz on þat mon cast / Of a broun bleeaunt, enbrauded ful ryche’, and, ‘He were a bleaunt of blew þat bradde to þe erþe’ (lines 879 and 1928). The term bleaunt obviously signalled a garment or material of some quality, as its presence on ‘aristocratic’ characters in the French and English romances regularly attests. It seems likely, furthermore, that the Middle English form bleaunt (blihand, blehand) finds is origins in the Anglo-French, as none of the other Romance languages show regular evidence of the –nd ending, and available attestations in Middle English are primarily in texts with earlier Anglo-Norman connections.8 While this is far from conclusive, bleaunt surely represents a native form of Old French bliaut, finding its way into English usage through Anglo-Norman rather than through continental borrowing.

Brouderer Jefferson and Rothwell have demonstrated that the unprefixed form of ‘embroiderer’ in Middle English, brouderer – formed from the infinitive of the Old French verb brosder/brouder etc. with the addition of the –er ending to signal the ‘doer’, ‘worker’, or ‘tradesperson’ – is also found in the Latin 6





7



8

The DMF records a form bliant (read bliauf), from Li Romans de Bauduin de Sebourc (c. 1350), cf. DMF bliaud, subst. masc. Such as Sir Gawain and the Green Knight (late fourteenth century); see also Sir Tristrem (c. 1300) lines 410 and 450; The Romance of Guy of Warwick (c. 1300) line 5787; and (as a bedspread) The Wars of Alexander, an Alliterative Romance (a. 1400) line 4912. Also see the entry for blidalt, Französisches Etymologisches Wörterbuch (Wartburg 1928); and for bliaud, bliaut, Trésor de la langue française (Imbs 1975). The –nt ending is apparently also found in Welsh by the thirteenth century: see the entry for bleaunt in Geiriadur Prifysgol Cymru: A Dictionary of the Welsh Language (1950–67); we are grateful to our colleague, Dr Stuart Rutten (University of Manchester) for making this known to us.

68

AN-1a.indd 78

11/03/2010 12:00:17

From Apareil to Warderobe (brouderarius),9 and the Middle English (brouderer) and is certainly the result of Anglo-Norman influence (1997: 285).

Canvass When Chaucer’s Canon’s Yeoman tells us that he had ‘on the floor ycast a canevas’ (line 939, our emphasis) to collect the debris from his master’s failed experiment, he uses a word which is not necessarily of Anglo-French origin, but which certainly has an Old Northern French background. Under the entry for the noun canevas, meaning ‘a fabric made from flax or hemp, canvas’, or ‘a piece of canvas; a canvas covering’, and so forth, the MED attributes an AngloNorman origin to the form, related to central French chanevaz. The OED more precisely attributes the Middle English development to Old Northern French, and we find Chrétien de Troyes, or perhaps one particular scribe, referring to canevas in the late-twelfth-century Roman de Perceval (Roach 1956: 499). The MED’s earliest recorded attestations of the word are generally in multilingual, usually Latin-based accounts, where it is often impossible to assign a particular language to ‘technical’ terms or to full entries. In the accounts rolls of the executors of Thomas, bishop of Exeter, for the year 1310, for example, we find reference to ‘iiij canevaz continentibulus 1 ulnas venditis’ (Hale and Ellacombe 1874: 9, our emphasis). Likewise, the charters of endowment and so forth for the Priory of Finchale (County Durham) for the year 1354 make reference to ‘ij canevaces’, using the word to mean a sort of bedspread (Raine 1837, our emphasis). The DMLBS records the form canevatios as early as 1207, attributing the word to Anglo-Norman. In Anglo-Norman, we find the word canavaces glossing the Latin word carentivillas in a thirteenth-century text (see Hunt 1991: 149). As would be expected, none of the variants of the noun recorded in the AND (canevas, canavace, canefas, canevace, canevaz, canevesce, canewas, etc.) have the initial ch- frequent in Central French forms, and the Middle English shares this feature.10 It also appears in the variant forms canevel and canevele in a couple of Anglo-Norman glosses (see attestations and references in the entry for canevel in the AND). It is quite clear, furthermore, that Middle English and Anglo-Norman canevas (if, indeed, they can be considered separately) reflect the Anglo-Norman and Old Northern French tendency to retain initial ca– where Central French would regularly develop cha– or chie–.11 We can conclude with some confidence, then, that, as 9



10

11





The DMLBS records broudararius from the Rotuli Litterarum Patentium (Patent Rolls) for 1395. It does not suggest an etymology, but it does derive the verb broudare from Old French (cf. broudare and brouderearius, DMLBS, fasc. 1, 119). The dictionary does record one occurrence of the form chanevaz, but it is used only as an adjective, meaning ‘hempen, hemp-coloured’. Compare Middle and Modern English caitiff (Anglo-French caitiff) versus Central French chaitif, cauldron verses chaudron, etc. (See Bush 1922: 161–72; a more basic discussion may be found in Baugh and Cable 1993: 171). Of the few forms of the

69

AN-1a.indd 79

11/03/2010 12:00:18

The Anglo-Norman Language and Its Contexts with many words beginning ca–, the Middle English word canevas has made its way through Anglo-Norman, rather than having been borrowed from a Central French variety. In Modern English, the word canvas has retained its versatility in describing a textile made from hemp or flax, a garment or sheet of such material, and even something that resembles such a garment or sheet, such as sail-cloth or oil painting (see the OED entry for canvas, canvass, n).

Jaumber An Anglo-French word for part of the leg armour, jamber, formed from the French word for leg, jambe, and from Old French jambiére, is imported directly into Middle English, and the MED’s earliest attestation (1322) is clearly AngloFrench: ‘j pare de guyssens et j pare de jambers’ (see Report on Manuscripts in Various Collections, vol. VII, 344).12 Within a few decades (a. 1388) we find Robert Mannyng of Brunn’s King Arthur wearing pieces of armour of the same name: ‘Hym self was armed fynly wel Wyþ sabatons, & spores, & iaumbers of stel’ (line 9880).13 The AND offers an attestation from 1419, from the Munimenta Gildhallae Londoniensis: ‘aketours, jambers, haberjons’ (vol. I, 12). Like many French words for armour and its accoutrements (pourpoint, spaudeler, the adjective quisheld, etc.), the Anglo-Norman word jamber finds itself in lists of items which make their way into Middle English through common usage.

Meniver Meniver (modern miniver) is a common Middle English word describing the fur of a squirrel, often the winter underbelly fur of grey squirrels, used to line or trim garments, particularly in the fourteenth and fifteenth centuries (see MED). The MED correctly acknowledges an Anglo-Norman derivation, related to the later Central French menu vair, which Frédéric Godefroy identifies as ‘fourrure tirée du petit-gris’, first attested around 1380 in an inventory of Charles V. The OED does not acknowledge an Anglo-Norman derivation, but does helpfully indicate that, in Early Middle English senses, ‘etymology would suggest that it must have been artificially spotted or variegated fur, with a smaller pattern than that of vair’. The AND offers an attestation from the

12 13



word appearing with ca- in Middle French, most are found in northern dialects. For example, the form canevach appears in Picardy in Phillip de Mézières’s latefourteenth-century Testament; see the entry for canevas in the Trésor, and the DMF (canevas, subst. masc.). This form carries on into Middle French: cf. jambiere, subst. fem., DMF. This reading is from the Lambeth Palace Library manuscript (Lambeth MS 131). Sullens suggests that the Innter Temple manuscript (Petyt MS 511, vol. VII) is earlier and more accurately represents Mannynge’s dialect. It reads, ‘gode chambres of iren and stele’ (our emphasis).

70

AN-1a.indd 80

11/03/2010 12:00:18

From Apareil to Warderobe Divisiones Mundi which it dates to c. 1300. The MED’s earliest entries are largely Anglo-Norman, and there is at least one ‘vernacular’ attestation in a Latin inventory for the year 1299.14 There is a reference to ‘covertures de menuvoyr’ in the Rolls of Parliament for 1314–15, and ‘revers de menevoir’ for 1363,15 in forms that look to be developing away from menu vair in Anglo-Norman contexts. The first ‘Middle English’ attestation in the MED comes from Floris et Blancheflor (c. 1250). In the Auchinleck Manuscript (MS Advocates 19.2.1, National Library of Scotland, Edinburgh, dated c. 1330–40), Floris gives a helpful innkeeper a ‘mantel of scarlet / Ipaned al wiʒ meniuer’ (lines 130–1).16 For the Central French form menu-vair, the Trésor de la langue française offers an earlier date than Godfroy, suggesting 1306 (see s.v. menu-fair).17 It is certain (and largely accepted) that the Middle English form meniver (menivoer, menivieir, meniwere, menever, etc.), used to describe a particular type of fur lining, developed through Anglo-Norman usage and was not ‘borrowed’ from Central French. It is unfortunate that the OED’s current etymological data does not acknowledge this development.

Bibliography Baugh, A., and T. Cable (1993), A History of the English Language, 4th ed. London. Benson, L. (ed.) (1987), The Riverside Chaucer, 3rd edn. Oxford. Bergen, H. (ed.) (1908), Lydgate’s Troy Book, part 2, book 3. EETS, es 103. London. Bush, S. (1922), ‘Old Northern French loan-words in Middle English’, Philological Quarterly 1, 161–72. Childs, W., and J. Taylor (eds.) (1991), The Anonimalle Chronicle 1307–1334. Leeds. Forshall, J., and F. Madden (eds.) (1950), The Holy Bible: Containing the Old and New Testaments, with the Apocryphal Books, in the Earliest English Versions, Made by John Wycliff and His Followers. Oxford. Gardner, R. (1950), ‘A note on Old French bliaut’, Romance Studies 12, 63–5. Godefroy, F. (1888), Dictionnaire de l’ancienne langue française. Paris. Hale, W., and H. Ellacombe (eds.) (1874), Account of the Executors of Richard, Bishop of London, 1303, and of the Executors of Thomas, Bishop of Exeter, 1310. Camden Society, ns X. London. 14



15



16



17



‘de menivieyr ad capuscia’, Extracts from the Account Rolls of the Abbey of Durham from the Original MSS., ed. C. Fowler, vol. II, Surtees Society Publications 100. Durham, 1899, p. 495. Cited in MED. Rotuli Parliamentorum, et Petitiones, et Placita in Parliamento, ed. J. Strachey (London, 1783), for the years 1314–15, L.292a; and the year 1363, 2.278b. Vries offers the readings ‘a pane of menuuer’, from Cambridge Univ. Libr. MS. Gg.iv.27.2, and ‘A mantyl of scarlet with menyuere’, from MS. Egerton 2862, both of which are later manuscripts (1966: 86–7). The Trésor (Imbs) follows the Altfranzösisches Wörterbuch in citing Joinville’s Saint Louis as the earliest attestation (also known as ‘Le livre des saintes paroles et des bonnes actions de St Louis’, or Histoire de Saint Louis, early fourteenth century).

71

AN-1a.indd 81

11/03/2010 12:00:18

The Anglo-Norman Language and Its Contexts Houston, M. (1996), Medieval Costume in England and France: the 13th, 14th and 15th Centuries, 2nd ed. Mineola, NY. Hunt, T. (1991), Teaching and Learning Latin in Thirteenth-Century England, vol. II: Glosses. Cambridge. Imbs, P. (ed.) (1975–), Trésor de la langue française. Paris. Jefferson, L., and W. Rothwell (1997), ‘Society and lexis: a study of the AngloFrench vocabulary in the fifteenth-century accounts of the Merchant Taylors’ Company’, Zeitshrift für französische Sprache und Literatur 107, 273–301. McNeill, G. (ed.) (1886), Sir Tristrem. Scottish Text Society, series 1, vol. VIII. Edinburgh. Norris, H. (1999), Medieval Costume and Fashion. Mineola, NY. Originally published as Costume & Fashion, Volume Two: Senlac to Bosworth, 1066–1485. Owen-Crocker, G. (2004), Dress in Anglo-Saxon England. Revised and enlarged edition. Woodbridge. —— and R. Netherton (eds.) (2005–6), Medieval Clothing and Textiles, vols. I–II. Woodbridge. Pope, M. (ed.) (1955), The Romance of Horn by Thomas, vol. I, Anglo-Norman Text Society 9–10; T. Reid, T. ref. and compl. 1964, vol. II, Anglo-Norman Text Society 12–13. Oxford. Raine, J. (ed.) (1837), The Charters of Endowment, Inventories, and Account Rolls of the Priory of Finchale, in the County of Durham. Surtees Society 6. London. Report on Manuscripts in Various Collections (1914). Historical Manuscripts Collection 55, vol. VII. London. Riley, T. (ed.) (1859–62), Munimenta Gildhallae Londoniensis, 3 vols, vol. I: Liber albus. London. Roach, W. (ed.) (1956), Chrétien de Troyes: Le roman de Perceval ou Le conte du Graal. Geneva. Rothwell, W. (1992), ‘The French vocabulary in the archive of the London Grocers’ Company’, Zeitschrift für französische Sprache und Literature 102. ——(2001), ‘OED, MED, AND: the making of a new dictionary of English’, Anglia 119, 527–53. ——(1991), ‘The missing link in English etymology: Anglo-French’, Medium Ævum 60, 173–96. Sellers, M. (ed.) (1912), York Memorandum Book: Part 1 (1376–1419). Surtees Society 120. London. Skeat, W. (ed.) (1886), The Wars of Alexander, an Alliterative Romance, EETS, es 47. London Sullens, I. (ed.) (1996), Robert Mannynge of Brunne, The Chronicle. Binghamton, NY. Thomas, R. (ed.) (1950–67), Geiriadur Prifysgol Cymru: A Dictionary of the Welsh Language. Cardiff. Tobler A., and E. Lommatzsch (eds.) (1963), Altfranzösisches Wörterbuch. Wiesbaden. Tolkien, J., and E. Gordon (eds.) (1928, rev. Norman Davis 1967), Sir Gawain and the Green Knight. Oxford. Tout, T. (1928), Chapters in the Administrative History of Medieval England, vol. IV. Manchester. Trotter, D. (ed.) (2000), Multilingualism in Later Medieval Britain. Cambridge. ——(2006), ‘Language contact, multilingualism, and the evidence problem’, in U. Schaefer (ed.), The Beginnings of Standardization: Language and Culture in Fourteenth-Century England, Frankfurt.

72

AN-1a.indd 82

11/03/2010 12:00:18

From Apareil to Warderobe Vries, F. (ed.) (1966), Floris and Blanchefleur. Groningen. Wartburg, W. von (ed.) (1928), Französisches etymologisches Wörterbuch. Bonn and Leipzig. Way, A. (ed.) (1843–65), Promptorium Parvulorum, 3 parts. London. Zupitza, J. (ed.) (1966),The Romance of Guy of Warwick. EETS, es 42, 49, 59. London.

73

AN-1a.indd 83

11/03/2010 12:00:18

chapter seven

Languages in the Military Profession in Later Medieval England Anne Curry, Adrian Bell, Adam Chapman, Andy King and David Simpkin

From the reign of Edward I onwards, the English were almost constantly involved in warfare both with their near neighbours, the Welsh, Scots and Irish, and with the French and their continental allies. This had a marked effect on the nature of English armies. Under Edward I, the proportion of paid, as opposed to unpaid and feudally provided, troops began to increase substantially. By the 1330s all soldiers serving the English crown, whether at home or abroad, received wages. The escalating need for funds to pay military wages determined the concurrent development of parliament. In their turn, the intensity of warfare and the opportunities for paid service, especially during the Hundred Years’ War with France (c. 1337–1453), which involved occupation of territory as well as periodic campaigning, both stimulated and facilitated the rise of the professional soldier. We are currently exploring this phenomenon in detail1 by collecting the names of all known soldiers serving the English crown in the second and third phases of this war between 1369 and 1453.2 We are already detecting lengthy military careers, performed over several theatres of war – we are not limiting ourselves to campaigns in France – and extending from the various ranks of nobility through the men-at-arms, whose status was described as armiger/scutifer/escuier/esquire, to the archers, who were commonly described as sagitarii/valetti/valets/archiers/archers. As will be seen from this mention of terminology, the sources we are exploiting can be in any of three languages in use in later medieval England – Latin, Anglo-French and Middle English. To date, however, there has been little assessment of the use of languages in a specifically military context.3 It has 1



2



3



The present essay derives from work carried out within ‘The Soldier in Later Medieval England’ project, funded by the Arts and Humanities Research Council and based at the Universities of Reading and Southampton. This period was chosen because of the wealth of suitable sources for nominal data, but it is hoped in due course to extend the time span both backwards and forwards. For nominal data collected to date see http://www.medievalsoldier.org. Contamine (1980: 102–13) discusses levels of literacy amongst the French military.

74

AN-1a.indd 84

11/03/2010 12:00:18

Languages in the Military Profession in Later Medieval England often been suggested that language was considered an expression of national identity in the context of England’s late medieval wars. On five occasions, the crown called on the support of parliament with the justification that the French intended to invade England and destroy the English language. The first is found in the writ of summons in 1295 (Stubbs 1913: 480) when Edward I was at war with the French as well as the Scots and facing a rebellion in Wales (linguam anglicam . . . omnino de terre delere proponit). The other four are to be found within the parliament rolls themselves: in 1344, four years after Edward III had declared himself king of France (a destruire la langue Engleys); 1346, just after his victory at the battle of Crécy (a destruire et anentier tote la nacion et la lange Engleys); 1377, when English successes in France were at a low ebb and when the French were preparing to launch raids on the south coast (et d’ouster de tout la langue Engleys); and under similar circumstances in 1380 (d’ouster oultreement la lange Engleise).4 There is some irony that the argument was expressed on the first occasion in Latin and on the next four in Anglo-French. In 1431 the notion appeared in English, not with direct reference to the threat of French invasion but calling to mind the Welsh rebellion of Owain Glyndŵr in the first decade of the fifteenth century.5 It is possible, of course, that on earlier occasions the arguments had been rehearsed in parliament in English.6 Here, as in so many other contexts in the later middle ages, we face a major problem – that the language of record is not necessarily the language of use. The intensity of war with France raises particular questions about both record and use. At face value, we might expect the Anglo-French conflict to have increased both the need for and desire of Englishmen to have knowledge of French, whether through the French of England or through contact with the French of France. According to the chronicler, Jean Froissart, as soon as Edward III withdrew his homage to Philip VI of France in 1337 (thereby starting the Hundred Years’ War), there was an immediate response to the linguistic issues which it raised through an order made in the parliament at Westminster of October 1337. ‘Encorres fu il ordonné et arresté que tout seigneur, baron, chevalier et honnestes hommes de bonnes villes mesissent cure et diligence de estruire et aprendres leurs enfans la langhe françoise par quoy il en fuissent plus able et plus coustummier ens leurs gherres’ (Froissart, vol. II, 1867: 419).7 Modern commentators, based on the evidence of manières de langage produced in England between 1396 and 1415, have concluded that the war of that period provided a similar stimulus to acquisition of French.

4

5



6





7

PROME, 1344 parliament, item 6; 1346 parliament, item 7; 1377 parliament item 12; 1380 parliament, item 16. PROME, 1431 parliament, item 32. ‘Consideryng also that the saide insurrections, rebellions and tresons, ymagined and done be the saide Owen were . . . fynally in distruction of all Englissh tonge for evermore’. We know that at the opening of the 1362 parliament the reasons for its summons was expressed in English (montre en Englois) even though this event was recorded in Anglo-French (Ormrod 2003). Cited in Lusignan (1987: 107).

75

AN-1a.indd 85

11/03/2010 12:00:18

The Anglo-Norman Language and Its Contexts Witness the comment of Briggs: ‘There is evidence that the English nobility began to study the French of Paris in order to serve more effectively as soldiers and diplomats and as administrators in English-held parts of France’ (Briggs 1999: 84).8 Some caution needs to exercised with both quotations. Froissart’s remark cannot be substantiated by the parliament roll since none survives for that assembly. Remember too that he wrote it between 1376 and 1380, at least forty years after the date to which it refers.9 Froissart was apparently in England around the time war resumed in 1369, visiting again in 1395.10 Since there is no evidence Froissart knew English,11 any information he gleaned must have come from those with whom he could converse in French. He himself tells us of a lengthy conversation with Sir William Lisle during his visit in 1395, implying incidentally that Sir William’s French was good enough. Was Froissart’s comment on the supposed parliamentiary order of 1337 therefore reflecting contemporary concerns rather than historical reality? In the same period (1385 to be precise) John Trevisa had commented that ‘in all the gramere scholes of Engelond, children leveth Frensche and construire and lerneth in English’, adding later that ‘also gentil-men haveth now moche i-left for to teche her children Frensche’. Although he did not explicitly mention military groups, he saw this as a disadvantage for those going overseas: ‘that is shame for them and they that schulle passe the see and travaille in strange landes and in many other places’ (Higden, vol. II, 1865–86: 159–61). Briggs’s comment takes us to the heart of the matter to which this essay is devoted. The dating of the first two manières to 1396 and 1399 places them in the period of long truce when commercial intercourse between England and France was boosted: their vocabulary is appropriate to that context. That for 1415 includes a small amount of vocabulary on weapons and armour compared with the earlier texts (Kristol 1995: 79). It also contains a model conversation on the battle of Agincourt but this is generic and has interesting parallels with vernacular chronicling of the event (Kristol 1995: 70).12 We can imagine that conversation would go down rather better in the school room than in a meeting with a real Frenchman. Furthermore, there are relatively few surviving

8 9

10

11





12



For the texts see Kristol (1995). This is the dating given for the Amiens manuscript in which this passage appears (Froissart, ed. Diller, vol. I, 1998: 221). I am grateful to Professor Peter Ainsworth for discussion on this point. See his reflection on his later visit: ‘I was anxious, therefore, to visit that country for it ran in my imagination that if I once again saw it, I should live the longer. For twenty-seven years past I had intentions of going thither and if I should not meet with the lords whom I had left there, I should at least see their heirs, who would likewise be of great service to me in the verification of the many histories I have related of them’ (as translated in Johnes 1874, vol. II, p. 568). No evidence has yet been found to confirm the popular view that he was at court in the service of Queen Philippa (personal communication, Dr Gottfried Croenen). For translation and discussion see Curry (2000a: text D8). Similarly the mentions in the 1399 manière of the deposition of the king (Richard II) and of the new king’s (Henry IV’s) intentions to campaign in Scotland (Kristol 1995: 65–6).

76

AN-1a.indd 86

11/03/2010 12:00:18

Languages in the Military Profession in Later Medieval England manuscripts of any of these works. On the whole, it has been assumed that knowledge of French on the part of English people diminished in the fifteenth century. Data on the decline of the use of French in both governmental and private record-keeping was established in the seminal article of Suggett (1946). The last known will in French is dated to 1431. There were no more deeds in French enrolled on the Close Rolls after 1434, and the last French petitions were presented to the privy council in 1441. In parliament (where the first English petition is found in 1388), English entries are few and infrequent until 1422, but between 1433 and 1443, the decline of French parliamentary petitions was extremely rapid. Although Suggett points out that English was beginning to be used in these and other contexts in the 1390s, the evidence which she and subsequent commentators cite indicates that the final shift from French to English was abrupt. In terms of petitions to the privy council, for instance, English had appeared very sporadically until 1436 but dominated thereafter. Yet this period of the decline of French in England corresponds exactly with the period of greatest military presence. Between 1415 and 1450 the English occupied much of northern France, including, from 1420 to 1436, the French capital itself (Curry 1994).13 By the treaty of Troyes of May 1420 Henry V became heir to the French throne. His premature death meant that it was his son, Henry VI, who actually acceded to the double monarchy of England and France on the death of Charles VI in October 1422, and who was crowned king of France in Paris in December 1431. Although the extent of English territorial holding began to diminish from the spring of 1429 when Joan of Arc inspired Charles VII’s armies to relieve Orleans, the English held on to Maine until 1448, Normandy until 1450, and Gascony, the southern territory held since the late twelfth century, until 1453. Mention of the latter reminds us that English contacts with France were long-standing. Indeed, they go back even further – to 1066. Henry V deliberately visited the tomb of William the Conqueror when he took Caen in September 1417, and made many other efforts to stress the ‘Norman inheritance’ of the English. He went so far as to give lands in the duchy to his administrators and soldiers, a policy which continued throughout the occupation. The English secured their conquest by sending over expeditionary forces virtually every year between 1415 and 1450, and by installing garrisons in key locations. The English lost little time in taking over existing French administrative systems, including those applying to military organisation. Since these native systems used only French in their written record, the English occupation was largely conducted in the French of France. Never before had so many Englishmen been involved or resident in France for such long periods, or needed to be so much in communication with the local French population. Ample evidence survives to show that such Englishmen 13



Serge Lusignan (1987: 110) made a prescient observation, however: ‘Car n’oublions pas que la période la plus prolifique en manuels d’enseignement correspond d’assez près à l’époque où la royauté anglaise par la personne de Henry V a paru le plus proche de réaliser le project de régner sur la France.’

77

AN-1a.indd 87

11/03/2010 12:00:19

The Anglo-Norman Language and Its Contexts also purchased land, became bourgeois of the towns in which they lived, married local women and made donations to local churches.14 At the end of the occupation of Normandy, we know that some ‘stayed on’. For instance, John Edwards, captain of La Roche Guyon, ‘se fist Franchois’, and Richard Merbury, his counterpart at Gisors, ‘se renderoit Franchois’, in order to continue to enjoy the landed inheritance of their French wives (Stevenson 1863: 88, 289).15 There had been, of course, English military involvement in France in earlier centuries. Gascony had seen both English soldiers and landholders. During the fourteenth-century phase of the Hundred Years’ War, the English had held from time to time garrisons in other parts of France – Ponthieu, Brittany, Poitou, and Lower Normandy. From 1347 onwards Calais had been in English hands. There was thus extensive exposure to various regional forms of French, although within English Gascony, Anglo-French was used in administration as well as between Englishmen stationed there (Trotter 1997: 202, 204–10; 1998).16 Most English soldiers in the fourteenth century did not find themselves based in France for long periods but rather passing through quickly on raiding campaigns known as chevauchées. We can imagine that little time was spent engaging in conversation with inhabitants the soldiers were about to pillage. This was hardly an incentive to the learning or use of French. The scale of this earlier presence was nowhere near that of the fifteenth-century occupation of northern France. Yet, as noted earlier, it was precisely in this period of greatest and most socially widespread contact that the use of French declined in England. The questions we must ask are therefore two-fold: did war with France play any role in stimulating or preserving the learning and use of French in England, and if so, when? And why, since the English presence in France was at its height in the first half of the fifteenth century, did the learning and use of French diminish at that very point? We need to emphasise that we are approaching this from a historical, not linguistic, perspective. We believe that there would be much potential in subjecting documents generated in our military contexts to fuller linguistic analysis, and also in setting our findings within a comparative context of the impact of war and occupation on language acquisition and use in other periods and locations.17 14



15



16



17



For general discussion, see Allmand (1983), and, more specifically on soldiers, Curry (2000b). In the north wall of the church at Caudebec, a window proclaims itself in French to be the gift of Fulk Eyton, a Shropshire soldier who was garrison captain in the town from the mid-1430s until its surrender in 1450. Other Englishmen are known to have served in French garrisons after the English lost Normandy (Curry 1992: 163 n. 72). Gascon soldiers served in Edward I’s invasions of Wales and Scotland. They were also to be found in the garrisons which the English maintained in Normandy in the fifteenth century (Curry 1992: 152). The Arts and Humanities Research Council is also now supporting a project on the issue in post-1945 Germany and 1990s Bosnia, led by Professor Hilary Footitt at Reading and Professor Michael Kelly at Southampton.

78

AN-1a.indd 88

11/03/2010 12:00:19

Languages in the Military Profession in Later Medieval England Let us look first at the evidence of the documents of military administration produced in England between 1369 and 1453 to see whether they follow the pattern which Suggett observed. There is no lack of information here since, as noted at the outset, armies were paid and hence administered by the royal government. There was no separate war office in the middle ages. Armies were administered through the existing structures, using the same officials. It is not surprising, therefore, that language use follows the basic trends of the departments which produced them. So, for instance, throughout the period (and indeed beyond) the accounts which captains had to provide to the Exchequer after service on expeditions to France were in Latin, since that was the standard practice of that department for its financial administration. By contrast, from the late thirteenth to early fourteenth century onwards, the French crown used French in all of its financial transactions as in other aspects of government.18 Responsibility for taking the muster of troops before embarkation also lay with the Exchequer. Although the musters were usually taken by specially appointed knights and esquires with experience in military activity, the recording task fell to Exchequer clerks. Musters could either be in Latin or French. Taking the period from 1369 to 1450, based on the evidence of the rolls used in our project, ninety-one are in Latin and sixty-eight in French.19 The chronological distribution is as follows: Table 1. Chronological distribution of rolls used in project 1369–79 1380–9 1390–9 1400–9 1410–19

1420–50

Latin

41

9

3

9

24

5 (+3 in both)

French

24

11

0

5

23

5 (+3 in both)

In other words, whilst Latin predominates over the period as a whole, there is not a simple trend. Rather, the proportion of French rolls increases in the 1380s but that of Latin rolls rises again until there is a second increase in the use of French in the fifteenth century. In the material for 1441 we can also see differences between the two surviving rolls for the duke of York’s expeditionary army. One was drawn up in the customary Anglo-French before the army left England, the other in the French of France when part of the army landed at Honfleur.20 There is a further complexity in that the documents were often the work of more than one clerk. One supervisor might be responsible for the basic layout of the muster roll – the heading, the subheadings and the names – but other clerks might add additional comments, usually in the marginalia of The fourteenth-century trends in the language of royal charters are discussed in Lusignan (2004: ch. 3). 19 These rolls are to be found in TNA E 101. For a listing see Public Record Office Lists and Indexes no. XXXV (London, 1912). 20 See, for instance, the two rolls of the retinue of Sir Thomas Percy 1377–8 (TNA E 101/37/28, and E 101/36/39). 18



79

AN-1a.indd 89

11/03/2010 12:00:19

The Anglo-Norman Language and Its Contexts the documents beside the soldiers’ names. It is clear that the muster rolls, and clerks, often accompanied the campaign and were therefore updated, as men died, or were dubbed knights and so on. Tallying or updating might be done in French even if the roll was initially drawn up in Latin, or vice versa.21 This provides evidence of code-switching since a study of the handwriting shows that the same scribe was operating in both languages. On occasion, when a roll is mainly in Latin the terms it employs for the different kinds of soldiers are given in French: this is so for the muster of the expeditionary force sent to France in 1420, which employs the headings of escuiers, lances, and valletz.22 There are no musters of expeditionary armies in English, but from around 1400 some personal names begin to be expressed in that language.23 This may suggest something about how the roll call was actually undertaken. Only very occasionally are there any annotations in English: once in 1424 where a Latin roll for a company to serve at the siege of Le Crotoy includes English annotation in the same hand,24 and again in the muster of the duke of Somerset in 1443, written in Latin with comments in French and English.25 Whilst the Exchequer was Latin-based, documents issued in the Chancery under the privy seal used Anglo-French. Therefore the indentures or contracts which the captains entered into, and which were written in duplicate so that both the crown and the indentee could keep a copy to avoid fraud, were in this language. So too were the warrants issued to the Exchequer to authorise the release of pay to the indentees for themselves and their men (soldiers were not paid directly but via their captain).26 There are no exceptions to this for the fourteenth century nor for the first twenty years of the fifteenth-century campaigns. In 1435 we begin to see indentures and warrants in English.27 Over the next six years practice was mixed, but from 1441 onwards, English was the norm. This trend began earlier in the case of private indentures of retainer. As the collection edited by Jones and Walker shows, from at least the 1290s such indentures were composed in Anglo-French. Their first indenture in English was drawn up in October 1426 (Jones and Walker 1994: 146–7). From 1433 all of their examples are in English. A similar pattern is seen with subcontracts for the expeditionary armies, where French is used until the late 1420s.28 This Respectively, TNA E 101/53/33 (26 March 1441) and E 101/54/9 (20 April 1441). TNA E 101/49/36. 23 For example, ‘John’ rather than ‘Johannes’ or ‘Jean’ although it is usual to find a mixture of forms in the same document. 24 TNA E 101/51/16: ‘launcez the whych were at Crotay, x sperys’, along with references elsewhere to ‘bowes’. 25 TNA E 101/54/5. 26 Indentures are to be found within TNA E 101, and warrants for issue in E 404. For the period 1399–1485, there is a useful catalogue in Lists and Indexes Supplementary Series no. IX (London, 1964). 27 See for instance the warrant for payment to Richard Wastness, November 1435 (TNA E 404/52/156). 28 For examples in French from 1374 see Northumberland Record Office ZSW 1/58, 4/43–50, printed in Walker (1985). Sub-indentures for the army raised by Thomas 21

22

80

AN-1a.indd 90

11/03/2010 12:00:19

Languages in the Military Profession in Later Medieval England chronology is therefore very similar to that seen by Suggett in other areas of government. The change is therefore not necessarily the result of military considerations but broader governmental ones, linked also to the standardisation of English noted by Fisher (1977) and others (Richardson 1980), and described as ‘Chancery English’. Yet as noted earlier, whilst this linguistic shift was happening, the English were more embroiled in France than ever before. Let us look now at the administrative materials produced as a result of the occupation of the fifteenth century. The early administration of Harfleur, the conquest made in Henry V’s first campaign of 1415, initially followed practices at Calais. Account books were in Latin, as were the muster rolls down to 1421, and the staff who ran the scaccarius installed in the town were English, as they were at Calais (Curry 2003).29 In the second invasion of Normandy, however, Henry soon sought to exploit French administrative arrangements (Curry 1996). The result was hybrid-language use. Enrolments in the Norman Rolls, begun at the landing on 1 August 1417, and totalling 1,063 entries to Henry’s death on 31 August 1422, were 90% (932) in Latin, following the practice of the Chancery enrolments in England, and were written by English clerks who had crossed with the army. The remaining 10% (129) were in French. Since these concerned surrenders, weights and measures and other issues involving the local community the language choice was dictated by audience (Angers 1993: 133 n. 44).30 Most musters for the troops installed by Henry in garrisons up to the first months of 1421 were written in Latin by distinctly English hands, suggesting that there was again a reliance on English clerks, but there were also some in French. Furthermore, there were many French documents concerning financial and military administration produced by the chambre des comptes which Henry established at Caen by November 1417 and which was staffed by a mixture of English and French personnel. A linguistic analysis of documents produced in this phase of the occupation would be useful to measure the relative influences of Anglo-French and of native French usages.31 Take, for example, the first surviving document from the Caen chambre des comptes, an order issued on 24 November by the newly installed president et tresorier of the duchy of Normandy, the English knight, Sir John Tiptoft. It was probably written by an English clerk unaccustomed to French practice, as is revealed by some of the spellings but also by the use of the tutoyer form, something not found in Montague, earl of Salisbury, in 1428 were in English (TNA E 101/71/2/826–868B) although his own indenture with the crown was in Anglo-French (Stevenson, vol. I, 1861: 404–14). 29 TNA E 101/47/39, 48/6, 48/17, 48/19. 30 There were only two acts in English. 31 Lusignan notes that treaties of surrender of places in Normandy reveal signs of Anglo-French. The surrender of Meaux (1421) was the only one written in English, however. (‘Parler en français: alterité et identité des Anglais et des Français à la fin du Moyen Age’, p. 11, note 60. I am grateful to Professor Lusignan for providing access to this unpublished paper.).

81

AN-1a.indd 91

11/03/2010 12:00:19

The Anglo-Norman Language and Its Contexts documents produced in Normandy before the invasion nor discovered, so far, in any other document produced during English rule (Curry 1998: 98–9).32 It did not take long, however, for the administration of the English conquest to be ‘frenchified’, witnessed not only by the near exclusive use of French forms and language from late 1420 and early 1421, but also by a similar trend towards the dominance of French clerks and local officials. In other words, the English dropped their imported English forms of administration (there is a marked decline in the number of enrolments on the Norman rolls in the last two years of Henry V’s reign), and relied solely on French administrative procedures and structures. This trend had already commenced once Henry established the chambre at Caen and began to appoint both Englishmen and Normans to existing French offices, but it was cemented by the treaty of Troyes of May 1420. This not only made Henry heir to the French throne but also guaranteed preservation of French institutions, laws and customs in the whole of France.33 When Henry died, Normandy, by virtue of the treaty terms, reverted wholly to his son’s French crown. The chambre des comptes at Caen was closed and its responsibilities merged with that of the French crown in Paris. When the English lost Paris in 1436, they established a new chambre at Rouen but along wholly French lines. In English-held northern France, therefore, indentures, orders to take musters, orders to pay, acknowledgments of receipt of pay, and the musters themselves were wholly in French for the remainder of the occupation through to the loss of Normandy in 1450.34 Given the trends noted above for England, there was therefore an increasing gulf between the languages used in military contexts on either side of the Channel as the years passed. Two contradictory themes therefore emerge. On the one hand, given the thousands of soldiers who served in northern France between 1415 and 1450, knowledge of French was surely more widespread than it had ever been before. Yet on the other, the use of French in England diminished considerably. Can these two themes be reconciled? Whilst explanations of both must be speculative, the second has received attention from previous scholars. Much has been made of influence of the king’s own preference for the use of English, as witnessed by the well-known statement in the Brewer’s Company records for 1422 (Chambers and Daunt 1931: 139). .  .  . whereas our mother tongue, to wit the English tongue, hath in modern days begun to be honourably enlarged and adorned, for that our most excellent lord, King Henry V, hath in his letters missive and divers affairs Dated 24 November 1417. Note, too, the requirements of the treaty not only that all persons in Charles VI’s personal and household service should have been ‘born in the realm of France, or in places where the French language is spoken’, but also that the king should reside ‘within our obedience and not elsewhere’, Curry (2008a). 34 At the departure of the English in 1450, these archives remained in France as part of the archive of the chambre des comptes in Paris. For the history of the archive see Nortier (1965) and Curry (1994: 49–50). 32 33

82

AN-1a.indd 92

11/03/2010 12:00:19

Languages in the Military Profession in Later Medieval England touching his own person, more willingly chosen to declare the secrets of his will, and for the better understanding of his people, hath with a diligent mind procured the common idiom (setting aside others) to be commended by the exercise of writing: and there are many of our craft of brewers who have the knowledge of writing and reading in the said English idiom, but in others, to wit, the Latin and French, before these times used, they do not in any wise understand. For which causes with many others it being considered how the greater part of the Lords and trusty Commons have begun to make their matters to be noted down in our mother tongue so we also in our craft, following in some manner their steps have decreed to commit to memory the needful things which concern us.

Although this extract does not make the link, scholars have argued that Henry’s language policy was a result of his reopening the war with France (Lusignan 2004: 202–4). This is not wholly substantiated for the first campaign. The newsletter which the king sent to the city of London from Harfleur after its capture was in Anglo-French, following the practice of the campaigns of the fourteenth century (Delpit 1847: 216–17).35 After his landing at the beginning of the 1417 campaign, however, Henry chose to communicate his initial successes in English. The reply from London was penned as usual in Anglo-French. The king’s eldest brother, Thomas, duke of Clarence, had sent a letter to the city in that language around the same time as his brother’s English letter (Sharpe 1909: 183–5; Delpit 1847: 219–20). Thenceforward, however, newsletters sent to England reporting the campaigns were always in English. This was not an anti-French gesture on the part of the king but a way of ensuring that news of his successes, and even more of his need for more supplies, were publicised to as wide an audience as possible. Henry was not pursuing a ‘language policy’ as such, but was demonstrating his desperate need for support. The conquest of Normandy, involving as it did a six-month siege of Rouen, the largest place ever to be successfully taken by siege by the English over the whole of the Hundred Years’ War, was a huge effort. Not surprisingly, Henry’s reign was the most heavily taxed of the whole medieval period. His use of English was therefore closely linked to a deliberate widening of the political community at a time of unparalleled military investment. It was further fanned by reservations within England on the terms of the treaty of Troyes, which envisaged a double monarchy.36 In the parliament of December 1420 the Commons demanded confirmation of the statute of 1340 (itself the result of concerns about Edward III’s pretensions in France), which guaranteed that England would never be subordinated to France (Curry 2002: 20–1). The xenophobia which Henry had encouraged in order to generate support for his campaigns now took on a life of its own in opposition to royal Taken from the Letter Book I of the City of London. It is notable that Henry had ordered the proclamation of the treaty in English throughout every county in England, effectively commissioning a translation of the French text in order to do so (Ormrod 2003: 786).

35 36

83

AN-1a.indd 93

11/03/2010 12:00:19

The Anglo-Norman Language and Its Contexts ambitions in France. Wisely, the crown did not request taxation of the Commons for another nine years, forcing the English war-leaders to rely instead on financial levies within their conquests. The war was never wholly popular in England. Subsequent events in the war fanned this further. There is ample evidence of a propaganda campaign conducted in English in the later 1430s in the wake of the defection of the duke of Burgundy from his English alliance and the resulting threat to Calais (Doig 1995). This defection ended the hope expressed in the treaty of Troyes of 1420 that ‘from this point onwards and forever more . . . all dissensions, hatreds, rancours, and conflict between the two kingdoms and their peoples should cease’ with ‘peace, tranquillity, concord, affection and firm friendship’ being established. It is not, therefore, a coincidence that the abrupt ending of the use of French came in the wake of the disasters of 1435–6. For the English, the French had now fully reverted to their perfidious stereotype, as witnessed not only by Burgundian defection, exacerbated by the duke’s subsequent ban on importation of English imports into Flanders, but also by the loss of Paris and rebellion in Upper Normandy. With Dieppe and Harfleur now in French hands, England itself was vulnerable for the first time since the early fifteenth century. Out of this came the Libelle of Englyche Polycye, expressing xenophobia against all foreigners, and complaining that the government had spent too much on the defence of Normandy rather than on that of Calais and of the Channel (Warner 1927; Holmes 1961).37 There was therefore a strong patriotic element in the use of English, fanned by Henry V but then taken up as means of emphasising the separate identity of the English in the double monarchy. A strong practical dimension also played a part in the decline of Anglo-French and the preference for English. As a mid-fifteenth-century commentator pointed out, some had Latin or French but all had English, the lewd as well as the learned.38 In contrast to the expeditionary armies of the late fourteenth century, those of the fifteenth contained much larger proportions (and also numbers) of archers, and men-at-arms of the lower squirearchy. In other words, in the face of demands for manpower the military community had widened. English was the appropriate universal language to use. Whilst under Richard II ordinances for military discipline were published in Anglo-French,39 under Henry V and VI, they were in Eng Intriguingly too, Jean Juvenal des Ursins invoked a linguistic element in his attack on the English in Audite Celi, written around the same time: ‘le temps doit venir que Angleterre doit souffrir grandement et tellement que a Londres trouvera l’an a peine home qui ose parler anglois’ (Lewis 1978: 261). 38 He justified writing in English as follows: ‘For that is your kynde langage / That ye have most her of usage / That kan eche man understonde / That is boren in Engelond / For that langage is most schewed / As well among lesed [learned] as lewed’ (Middleton 1911: 239). 39 British Library Cotton Nero DVI. The ordinances produced for the Scottish army in response to English attack were in the French of Scotland, which was not surprising given that they were written for a joint Franco-Scottish army (Thomson and Innes, vol. I, 1814–75: 190–1). 37

84

AN-1a.indd 94

11/03/2010 12:00:19

Languages in the Military Profession in Later Medieval England lish, with written copies being ordered to be given to captains so that they could proclaim them to their men (Curry 2008b). The use of English eliminated the possibility of misunderstanding. As early as 1362 it was agreed that pleading in royal and seignorial courts should be made in English (Ormrod 2003: 750–1). The use of English in the parliamentary record was also linked to the need for certainty. A good example is seen in 1426 when oaths taken by the lords at the reconciliation of the duke of Gloucester and Henry Beaufort, bishop of Winchester, were not only made in English but also written verbatim in the record so that there could be no doubt as to the guarantees made.40 The strong culture of spoken English contributed further to this situation, and to the preference for English in any situation where misunderstanding could occur. This can be seen in an early (and, according to Suggett 1946: 69, the first known) private letter in English. Given the subject of this essay, it is significant that it was written from Florence in November 1392 by (or more precisely for) the mercenary captain, Sir John Hawkwood.41 It requested those back home in England to give credence to what John Sampson, carrier of the letter, would communicate ‘by mouthe’. For soldiers such as Hawkwood, attempting to deal with their business at a distance, there could be no risk of complications, nor of interceptions. This was irrespective of their competence in other languages. Letters home in the fifteenth-century campaigns were also commonly in English (Suggett 1946).42 This brings us to the question of soldiers’ language use, and an assessment of the impact of the fifteenth-century phase of the Hundred Years’ War. In the late fourteenth and early fifteenth century, Anglo-French was the language of preference for the military leadership. Virtually all their letters from the front, whether the French, Scottish, Irish or Welsh front, and whether issued by the crown or by individuals, were in this language. Witness, for instance, the letters collected into a formulary by John Stevens which tell us so much about the campaigns in Wales in the reign of Henry IV (Legge 1941). The gentry in the northern Marches also routinely used Anglo-French in their correspondence,43 as did their counterparts in Wales (Trotter 1994: 470).44 As noted earlier, the TNA C 65/87 m. 2. The oaths are set within a linking Latin narrative of the events and context, showing that the choice of verbatim recording was deliberate. 41 In fact we know of the letter only from its enrolment in the Plea and Memoranda rolls of the city of London in June 1411 (London Metropolitan Archives, CLA/024/01/02 42, roll A41, at http://www.british-history.ac.uk/report.aspx?compid=36713). It is followed by Sampson’s written record of Hawkwood’s wishes, also in English, dated at Boreham, Essex, 20 April 1393. 42 Based on the evidence of the Llanthony cartulary. 43 See, for instance, the letter from Sir Thomas Gray and Sir William Swinburne to John de Middleham, constable of Alnwick Castle regarding a private enterprise into Scotland (undated, but written before Gray’s premature death in 1400) (Northumberland County Record Office ZSW 1/104, printed in Tuck (1968: 31). Gray’s father (another Sir Thomas, d. 1369) had written memoirs in Anglo-French, King (2005). 44 Citing examples concerning the military service of Sir Degory Sais. 40

85

AN-1a.indd 95

11/03/2010 12:00:19

The Anglo-Norman Language and Its Contexts first-known set of disciplinary ordinances for the English army, issued for the 1385 campaign to Scotland, were in Anglo-French. This campaign generated disputes subsequently heard in the Court of Chivalry, whose proceedings were recorded in the same language (Nicolas 1832; Ayton 1998; Rogers 1962).45 Therefore, even if Trevisa expressed concerns about the general lack of French in the 1380s, the language seemed to be thriving within the military context. The muster-roll evidence does not negate this. The level of diplomatic activity between England and the French crown, as well as other rulers within France, remained high. It involved men who also served as soldiers as well as royal chaplains like John Stevens, who was commissioned shortly before the departure of the 1415 expedition to transcribe copies of earlier diplomatic agreements with the French, and who then accompanied the king to France.46 If there was an area where Anglo-French usage was sustained into the fifteenth century, therefore, it was in the military and diplomatic context. The conquest and occupation of northern France could not have been successfully undertaken without it. Lusignan (1987: 4) makes a telling point when he notes that the English had a certain advantage in their dealings with the French since several of them understood the language of their enemies. Even if their pronunciation was different, or even if their formulations did not reach the heights of elegance of the French diplomats,47 this did not stand in the way of communication in effecting a conquest and occupation. Furthermore, with pre-existing continental possessions, and with diplomatic relations with other parts of France, they were already accustomed to different varieties of written and spoken French.48 There is space here to mention only a few areas where we can detect linguistic competence during the occupation. At the highest level of policy-making, commanders sat on the Grand Conseil alongside those French nobles and clergy who accepted English rule. The earl of Warwick spoke in French on behalf of the young king, Henry VI, when the latter was presented in Paris after his coronation in December 1431. Whilst the earl was captain of Rouen, he and his wife regularly entertained the bourgeois of the city and their wives, inviting along other English captains and officials with their spouses. Such a situation Note, however, that in the case of Grey v. Hastings, heard in 1408–10, we have evidence that statements were made in English (Jack 1965: 15; Keen 1992). 46 His authorship of the main Latin narrative of the campaign has also been speculated: Taylor and Roskell (1975: xxiii, 16). 47 Lusignan (1987) provides several examples of English reluctance to use French in negotiations. This was not simply an anti-French stance but admission that neither their aural understanding nor their own speech was as eloquent as that of the French, and that they feared the French would exploit this situation to their diplomatic advantage. 48 For instance, Henry IV received letters from the duchess of Brittany (later to be his wife) and the king of Scotland in French. Sir Richard Aston, lieutenant of Calais both sent letters to, and received them from, the duchess of Burgundy in French. The king’s diplomatic correspondence to the queen of Denmark and the count of Juliers, however, was conducted in Latin (Hingeston 1860). 45

86

AN-1a.indd 96

11/03/2010 12:00:19

Languages in the Military Profession in Later Medieval England needed even small talk to be conducted in French. Peers, knights and esquires who occupied command functions had enough linguistic capacity to carry out duties such as taking declarations of allegiance from the local population. Throughout the occupation virtually all captaincies and lieutenancies were held by men of English birth. Personal interactions with townspeople are well evidenced. For instance, Sir John Cressy, who died in office as captain of Lisieux in 1444, was popular enough to have a tower subsequently named after him.49 In the early phases of the occupation, the English involved in the financial administration, alongside their French counterparts, had trained within the English Exchequer, which still used French. William Allington, appointed treasurer of Normandy after the fall of Rouen, had previously served as receiver of Brest when that place was garrisoned by the English in the mid-1390s, before becoming treasurer of Calais (1398–9) and subsequently of Ireland in the 1400s (Roskell, vol. II, 1992: 27–9). Linguistic capacity may therefore have been a factor in appointments. In the case of the appointment of Thomas Wylmerden – a man of low status – to the sergenterie of St Lo in 1422, it was verified in advance that he ‘entendait bien le langage francois’ (Angers 1993: 133, n. 45). Throughout the occupation, the baillis who administered local government and justice were almost exclusively Englishmen: it is difficult to believe that they did not have a high level of competence in French. Their functions would have raised problems if they could not read as well as speak and understand, and, significantly, they were required to take an oath in French at the point they took up office. The controllers installed in garrisons from the mid-1420s, responsible for keeping records of the absences of soldiers and of war booty, were also English, and needed the same range of linguistic skills as the baillis. Englishmen occupied a range of other functions which required interaction with the native population, such as gaolers, gatekeepers, provisioners. Again they surely had some facility in spoken, and even written, French. In many cases this linguistic competence must have pre-dated service in France, and have been the result of family immersion or schooling, or both. The evidence of these office-holders would suggest that competence penetrated some way down the social scale. Those who had not received this moulding in England picked up the language simply by being in France. That soldiers could arrive without any knowledge is demonstrated by some incidents early in the occupation where English soldiers turned to violence when they could not make the inhabitants understand what they wanted, and ended up being killed themselves. We know of these through later pardons to their killers, who justified their behaviour on the grounds of linguistic confusion. Robert le Paumier, for instance, explained that ‘deux hommes parlant langage estrange’ had come to his village near Caen in November 1417 ‘que on ne entendoit point, et ne savoit on se c’estoient Bretons, Anglois, Escocoys ou autres gens’ (Cacheux, Archives communales de Lisieux, CC 25, p. 181.

49

87

AN-1a.indd 97

11/03/2010 12:00:19

The Anglo-Norman Language and Its Contexts vol. I, 1907–8: LXX). In another case there arrived in a cider house ‘un Anglois, Galois, normant ou autre ne scet de quelle langue ou pays il estoit’, who was subsequently killed by brigands (Cacheux, vol. I, 1907–8: LXXX). Local women were raped by ‘deux Anglois ou autres dont en l’entendoit aucunement le langage’ (Cacheux, vol. I, 1907–8: XC). One English soldier in April 1418 could not understand a Norman’s plea that he was ‘bulleté’ (i.e. that he had formally sworn allegiance to the English, Cacheux, vol. I, 1907–8: CV). Throughout the case the soldier is described as ‘un anglais ou autre homme tel qu’il estoit duquel il ne savoit le nom ne le seurnom ne la nascion dont il est natif bonnement’. Since the French would rarely have had any knowledge of English,50 it is easy to see how such situations arose.51 In 1425 there were complaints that captains were using ‘mots etrangers’ (presumably English words) as passwords so that the inhabitants who were obliged to assist in the watch could not understand them. As a result of local protest, captains were ordered henceforward to give ‘la nom de la nuit en langage francois . . . tel que ceulx qui feront le guet pussent raisonnablement entendre’.52 Some Englishmen continued to behave much as modern tourists: two English servants out drinking in Rouen at Epiphany 1427 showed their boorish nature by shouting to the locals who were annoying them ‘parle anglais, tu scez bien parler anglais’ (Cacheux, vol. II, 1907–8: CLXII). Would that we knew whether they really shouted this in French, or whether a French clerk was simply translating their English words! The local nickname for the English soldiers was ‘goddams’, an expression apparently derived from their propensity towards swearing in English. This reminds us that to each other they spoke English. Some stayed for only short tours of duty, and perhaps had little contact with the local population; others served for twenty years or more and achieved a high level of integration with the local population. Intermarriage and involvement in local commerce and landholding have already been mentioned. In these contexts they engaged with registration of deeds and transactions in the Tabellionage, where records were made, and cases heard, in French. Some were elected to the Estates where debates were exclusively in French. That language acquisition increased as the occupation continued is suggested by the fact that only in the early years There is a case where a Norman trying to recover his horse from an Englishman takes another Norman with him to do the talking ‘pour ce quil entendoit et parloit bien anglois’ (Luce, vol. I, 1879–83: 237–41). Since the treaty of Troyes had stimulated Anglo-French trade, there were presumably some locals who had gained some knowledge of English. Furthermore, as the occupation continued and there was intermarriage, the children of mixed marriages may have been bilingual. 51 For a fourteenth-century parallel, see the case of a young knight, the son of the king of Sicily, captured at Southampton in October 1338 near the start of the war. Clubbed to the ground by an Englishman, ‘he cried rançon’. ‘Yes’, the local replied, ‘I know you are a Françon’, and therefore killed him (Geoffrey le Baker 1889: 62–3). 52 Paris, Bibliothèque nationale de France, manuscrit français 26048/520, printed in Luce, vol. I (1879–83: 225–8) and noted in Angers (1993: 137). 50

88

AN-1a.indd 98

11/03/2010 12:00:20

Languages in the Military Profession in Later Medieval England were proclamations ordered to be made in both English and French. From the mid-1420s onwards they were exclusively in French (Angers 1993: 132–3). Furthermore, garrisons often contained men of other nationalities: Lombards, Germans, Spanish, Gascons, as well as French and Normans. Whilst it was not uncommon for these to constitute small groups within a garrison, others served as singletons. The implications of this for language acquisition are significant: did English or French operate as the lingua franca under such circumstances? From time to time the government required that nationality should be recorded on the muster rolls (Curry 1992). Welsh origins did not need to be noted but some musterers chose to do so, the implication being that the soldiers noted as Welsh were Welsh-speaking and, not surprisingly, serving together.53 What implications does all of this have? First, for the militarily active noble and knightly classes, there is little reason to believe there was any diminution in knowledge of French whilst the war continued. They already had competence in the language when they arrived, and their linguistic capacity was enhanced by their service in France. Secondly, given the presence of thousands of English soldiers in northern France, most of whom returned home at the end of their service, some knowledge of oral and aural French was more widespread amongst the English population than at any other previous time.54 That said, some caveats need to be expressed. The first is that as the occupation became established the majority of written records were drawn up by French clerks, and the financial administration passed almost exclusively into local hands. This diminished the need for written competence on the part of English administrators.55 As a result, exposure to the French of France, as well as the declining use of Anglo-French in administration at home, combined to spell the death knell of the latter in its a written form. Ironically, as a broader spectrum of Englishmen than ever were in a position to speak and hear French by virtue of their presence in France, England itself became an increasingly monolingual culture save in the specialist contexts of law French and the ecclesiastical and administrative contexts in which Latin persisted. The fact that For the campaign to Gascony in 1325 a company of 93 men from Caernarfon and Flint had their own doctor and proclamator, the assumption being that they needed Welsh speakers in these roles (British Library Add. MS 9967 fol. 90). It was also common to find Welsh chaplains accompanying Welsh troops in the armies of Edward I (e.g. British Library Add. MS 7965, five chaplains accompanying the men of North Wales under Gruffyd Llywd in the 1297 Flemish campaign). The archers of Swydd Yr Allt in northern Brecon serving in the 1415 army were accompanied by William Waldebesse, chaplain, one of the few non-Welsh names found in the list, but presumably bilingual (TNA E 101/46/20). 54 Colette Beaune suggested that only 20% of the English soldiers ‘furent longtemps capables de parler français’ (Angers 1993: 74) but this is a figure plucked from the air. 55 Even the government based in England started employing French-born secretaries, especially for liaison with France. By 1425 there were eight of them (Otway Ruthven 1939: 21, 91). 53

89

AN-1a.indd 99

11/03/2010 12:00:20

The Anglo-Norman Language and Its Contexts knowledge of French generated by service in France was predominantly oral may have contributed to the decline in grammatical standards of French in England, such as the diminution of gender distinctions.56 The second caveat is that whilst more Englishmen might have been able to speak and understand French, they always had locals to rely on for more complex matters. Spying, for instance, was an activity in which native speakers, amongst them women, were preferred to English soldiers. The vicomtes, the tier of local officers below the baillis, were almost exclusively local men. Many garrisons had a nativeborn trompette responsible for carrying messages. The ordnance retinue always had a French clerk who arranged purchases of equipment and powder. French notaries were used when anything needed to be written down.57 The third caveat is that levels of knowledge of French would have varied considerably dependent upon length of service and integration with the local communities. It is also noticeable that as the occupation continued, the participation of the higher nobility and knightly class diminished. This was for reasons linked to political circumstances at home, but it may have served as a disincentive to learn or keep up a knowledge of French. Even when active soldiers returned home they would have had no incentive in England to maintain their French or to pass it to the next generation. The decline in French usage in England compounded this situation. Once most of the territories were lost in the early 1450s, it may have been unpatriotic even to admit to a knowledge of the language of the enemy. Soldiers would no longer hear it or speak it, so their memory of it could diminish. The real period of the triumph of English is therefore the second half of the fifteenth century, when the English held only a Gibraltar-like enclave at Calais where, as at home, English was dominant in oral culture, with a residual use of Latin in accounting. The loss of Normandy and Gascony by the English in the mid-fifteenth century was effectively the loss of French as a language of the English. From that point onwards French was a distinctly foreign language associated with the hereditary enemy.

Paper by Richard Ingham, delivered to the WUN Multilingualism seminar, October 2007. It would be interesting to know whether experience of the French of Paris affected pronunciation of the French of England. On the problem in general, see Kristol (1994). 57 A good example is the agreement between Nicholas Molyneux and John Winter sealed in the church of St Martin at Harfleur on 12 July 1421, which was aimed at providing mutual support if they were captured by the French or if either died whilst their children were under age. Significantly, it required the survivor to support his partner’s children at school. The agreement is written in the French of Paris (Magdalen College Oxford, Southwark 213, printed in McFarlane 1963: 307–8). 56

90

AN-1a.indd 100

11/03/2010 12:00:20

Languages in the Military Profession in Later Medieval England Primary sources  archives Archives Communales de Lisieux The British Library BL Additional MS 9967 fol. 90 BL Additional MS 7965 BL Cotton Nero DVI London Metropolitan Archives CLA/024/01/02 42, roll A41 The National Archives C 65 – Chancery Parliament Rolls E 101 – Exchequer Accounts Various E 404 – Exchequer Warrants for Issue

published sources Acts of the Parliament of Scotland, ed. T. Thomson and C. Innes, 12 vols. Edinburgh, 1814–75. Cacheux, P. (ed.), Actes de la chancellerie de Henri VI concernant la Normandie sous la domination anglaise (1422–1435). Société de l’histoire de Normandie, 1907–8. Chambers, R., and M. Daunt (eds.), A Book of London English, 1384–1425. Oxford, 1931. Delpit, J. (ed.), Collection générale des documents français qui se trouvent en Angleterre. Paris, 1847. Froissart, Chroniques, ed. K. de Lettenhove, 25 vols. Brussels, 1867. Froissart, Chroniques. Livre 1: Le manuscrit d’Amiens, Bibliothèque municipale no. 486, vol. I, ed. G. T. Diller. Geneva, 1998. Geoffrey le Baker, Chronicon, ed. E. Maunde Thompson. Oxford, 1889. Higden, Polychronicon Ranulphi Higden Monarchi Cestrensis, Together with the English Translations of John Trevisa and of an Unknown Writer of the Fifteenth Century. Rolls Series. London, 1865–86. Hingeston, F. C. (ed.), Royal and Historical Letters during the Reign of Henry the Fourth, 2 vols. Rolls Series. London, 1860. Johnes, T. (ed.), The Chronicles of England, France and Spain, 2 vols. London and New York, 1874. Jones, M., and S. Walker (1994), ‘Private indentures for life service in peace and war, 1278–1476’, in Camden Miscellany XXXII,Camden Fifth Series, London, 1994. King, A. (ed.), Sir Thomas Grey, Scalacronica 1272–1363. Surtees Society 209. London, 2005. Kristol, A. (ed.), Manières de langage (1396, 1399, 1415). Anglo-Norman Text Society. London, 1995. Legge, M.-D. (ed.), Anglo-Norman Letters and Petitions from All Souls College MS 182. Anglo-Norman Text Society. Oxford, 1941. Lewis, P. (ed.), Ecrits politiques de Jean Juvenal des Ursins. Société de l’histoire de France. Paris, 1978. Luce, Siméon (ed.), Chronique du Mont-Saint-Michel, 1343–1468, 2 vols. Société des anciens textes français. Paris, 1879–83.

91

AN-1a.indd 101

11/03/2010 12:00:20

The Anglo-Norman Language and Its Contexts Nicolas, N. H. (ed.), The Scrope and Grosvenor Controversy, 2 vols. London, 1832. Report on the Manuscripts of Lord Middleton. Historical Manuscripts Commission. London, 1911. Roskell, J. S., et al. (eds.), History of Parliament. The Commons 1386–1421. Stroud, 1992. Sharpe, R. (ed.), Calendar of Letter Books of the City of London, Letter Book I. London, 1909. Stevenson, J. (ed.) (1861), Letters and Papers Illustrative of the Wars in France during the Reign of Henry VI. Rolls Series. London, 1861. ——(1863), ‘Le Recouvrement de Normendie par Berry, herault du roy’, in Narratives of the Expulsion of the English from Normandy. Rolls Series. London, 1863. Stubbs, W., Select Charters, 9th ed. London, 1913. Taylor, F., and J. S. Roskell (eds.), Gesta Henrici Quinti. Oxford, 1975. Warner, G. (ed.), The Libelle of Englyche Polycye. Oxford, 1927.

secondary sources Allmand, C. (1983), Lancastrian Normandy. The History of a Medieval Occupation. Oxford. Angers, D. (1993), ‘La guerre et la pluralisme linguistique: aspects de la Guerre de cent ans’, Annales de Normandie 43, 125–39. Ayton, A. (1998), ‘Knights, esquires and military service: the evidence of the armorial cases before the Court of Chivalry’, in A. Ayton and J. Price (eds.) The Medieval Military Revolution, London, pp. 81–104. Briggs, C. (1999), Giles of Rome’s De Regimine Principum. Reading and Writing Politics at Court and University, c. 1275–c. 1525. Cambridge. Contamine, P. (1980), ‘L’écrit et l’oral en France à la fin du moyen âge: note sur l’ “alphabetisme” de l’encadrement militaire’, in W. Paravicini and K. Werner (eds.), Histoire comparée de l’administration (IVe–XVIIIe siècles), Munich, pp. 102–13. Curry, A. (1992), ‘The nationality of men-at-arms serving in English armies in Normandy and the “pays de conquête”, 1415–1450: a preliminary study’, Reading Medieval Studies 18, 135–63. ——(1994), ‘English armies in the fifteenth century’, in A. Curry and M. Hughes (eds.), Arms, Armies and Fortifications in the Hundred Years War, Woodbridge, pp. 39–68. —— (1996), ‘L’administration financière de la Normandie anglaise: continuité ou changement’, in La France des principautés. Les chambres des comptes xive et xve siècles, Paris, pp. 83–103. —— (1998), ‘La chambre des comptes de Normandie sous l’occupation anglaise, 1417–50 (textes et documents)’, in P. Contamine and O. Mattéoni (eds.), Les chambres des comptes en France aux XIVème et XVème siècles, Paris, pp. 91–125. ——(2000a), The Battle of Agincourt. Sources and Interpretations. Woodbridge. ——(2000b), ‘Isolated or integrated? The English soldier in Lancastrian Normandy’, in S. Rees Jones, R. Marks and A. J. Minnis (eds.), Courts and Regions of Medieval Europe, York, pp. 191–210. ——(2002), ‘Le traité de Troyes (1420). Un triomphe pour les Anglais ou pour les Français?’, in D. Couty, J. Maurice and M. Guéret-Laferté (eds.), Images de la Guerre de cent ans, Paris, pp. 13–26. ——(2003), ‘Harfleur et les Anglais, 1415–1422’, in V. Gazeau (ed.), La Normandie et l’Angleterre au moyen âge, Paris, pp. 173–87.

92

AN-1a.indd 102

11/03/2010 12:00:20

Languages in the Military Profession in Later Medieval England Curry, A. (2008a), ‘Two kingdoms, one king: the treaty of Troyes (1420) and the creation of a double monarchy of England and France’, in G. Richardson (ed.), Contending Kingdoms, London, pp. 23–41. ——(2008b), ‘The military ordinances of Henry V: texts and contexts’, in C. GivenWilson and A. Kettle (eds.), War, Government and Aristocracy in the British Isles c. 1150–1500, Woodbridge, pp. 214–49. Doig, J. (1995), ‘Propaganda, public opinion and the siege of Calais in 1436’, in R. Archer (ed.), Crown, Government and People in the Fifteenth Century, Gloucester, pp. 79–106. Fisher, J. (1977), ‘Chancery and the emergence of standard written English in the fifteenth century’, Speculum 52, 870–99. Holmes, G. (1961), ‘The Libel of English Policy’, English Historical Review 76, 193–216. Jack, I. (1965), ‘Entail and descent: the Hastings inheritance’, Bulletin of the Institute of Historical Research 38, 1–19. Keen, M. (1992), ‘English military experience and the court of Chivalry: the case of Grey v. Hastings’, in P. Contamine, C. Giry-Deloison and M. Keen (eds.), Guerre et société en France, en Angleterre et en Bourgogne XIVe–XVe siècle, Lille, pp. 123–42. Kristol, A. (1994), ‘La prononciation du français en Angleterre au XVème siècle’, in Mélanges de philologie et de littérature médiévales offerts à Michel Burger, Paris, pp. 67–87. Lusignan, S. (1987), Parler vulgairement, Les intellectuels et la langue française aux XIIIème et XIVème siècles. Montreal. ——(2004), La langue des rois au moyen âge. Paris. McFarlane, K. (1963), ‘A business partnership in war and administration, 1421– 1445’, English Historical Review 78, 200–10. Nortier, M. (1965), ‘Le sort des archives dispersés de la chambre des comptes’, Bibliothèque de l’école des chartes 123, 460–537. Ormrod, W. (2003), ‘The use of English. Language, law and political culture in fourteenth-century England’, Speculum 78, 750–87. Otway Ruthven, J. (1939), The King’s Secretary and the Signet Office in the Fifteenth Century. Cambridge. Richardson, M. (1980), ‘Henry V, the English Chancery and Chancery English’, Speculum 55, 726–50. Rogers, A. (1962), ‘Hoton versus Shakell: a ransom case in the Court of Chivalry, 1390–5’, Nottingham Medieval Studies 6, 74–108. Suggett, H. (1946), ‘The use of French in England in the later middle ages’, Transactions of the Royal Historical Society, 4th series, 28, 219–39. Trotter, D. (1994), ‘L’anglo-français au pays de Galles: une enquête préliminaire’, Revue de linguistique romane 58, 461–88. ——(1997), ‘“Mossenhor, fet metre aquesta letra en bon francés”: Anglo-French in Gascony’, in De mot en mot. Aspects of Medieval Linguistics. Essays in honour of William Rothwell, Cardiff, pp. 199–222. —— (1998), ‘Some lexical gleanings from Anglo-French Gascony’, Zeitschrift für romanische Philologie 114, 53–72. Tuck, J. (1968), ‘Richard II and the border magnates’, Northern History 3, 27–52. Walker, S. (1985), ‘Profit and loss in the Hundred Years War: the subcontracts of Sir John Strother 1374’, Bulletin of the Institute of Historical Research 63, 100–6.

93

AN-1a.indd 103

11/03/2010 12:00:20

chapter eight

The Language of the English Legal Profession: The Emergence of a Distinctive Legal Lexicon in Insular French Paul Brand

The first proto-professional lawyers active in the English common-law courts can be identified in the early thirteenth century and the first fully professional during the reign of Henry III (1216–72), but it is only in the reign of Edward I (1272–1307) that professional lawyers emerge fully into the light of day and enough evidence survives to allow us to talk of the existence of an English ‘legal profession’. That profession was divided from the beginning into two separate groups: serjeants, who were professional specialists in pleading in court for their clients, and attorneys, who were professional specialists in making appearances in court on behalf of clients and in briefing serjeants. The centre of the emerging legal profession was the Common Bench, the main central royal court for civil litigation, which was normally resident at Westminster. There were around thirty serjeants practising there in years around 1300 and they are also known to have practised in other royal courts. There were also over two hundred professional attorneys practising in the court in 1300 and smaller groups in other royal courts. Other serjeants are to be found in each county court and in the courts of the city of London and in other towns (Brand 1992b). In the 1270s serjeants began to be appointed as royal justices but they did not gain a monopoly of judicial appointments to the main royal courts until the 1340s (Brand 1992a: 157–67). Even before that merger had taken place, and while the judiciary was still a rather more heterogenous body, the two groups (of professional lawyers and professional royal justices) which had long been in close contact with each other formed a single ‘professional community’ for socio-linguistic purposes. Also members of that community were the clerks of the royal courts, the most senior members of which themselves played a significant part in the judicial process, not only through authorising process but also in actual adjudication, as well as compiling the official record of litigation. By 1307 it is known that there were as many as forty-three different clerks writing the main plea roll of the court (Brand 1992a: 169–201). One of the phenomena that allows us to talk about the emergence not just of professional lawyers but more specifically of a legal profession in the later

94

AN-1a.indd 104

11/03/2010 12:00:20

The Language of the English Legal Profession thirteenth century is the evidence indicating that various types of formal instruction in the English common law were being given, probably in and around the Common Bench at Westminster. By the 1290s the court of Common Bench gave some degree of official recognition to the status of ‘apprentice of the Common Bench’ for men attached to the court for the purpose of learning the law, for whom a special enclosure (the ‘crib’) was provided from 1291 onwards and whom the justices of the court might address in the course of a hearing for educational purposes. The existence of such a recognisable group of law students may have encouraged the provision of other educational facilities round court. It is known that from an early date law reports were also being used for instructional purposes. There is also some evidence to suggest that already by the early fourteenth century lectures were being given on the statutes, which were forerunners of ‘readings’ later given at the inns of court (themselves not created intil the 1340s). Some, if not all, of the surviving common-law quaestiones seem to be connected with legal education. There is also a little evidence to suggest the existence of learning exercises to give apprentices practice in pleading. By the late 1270s there is also evidence of basic instruction on common-law litigation being given in lectures of which we have two different versions, one given in 1278 and the other some time shortly after 1279 (Brand 1999a). Although Latin had been the language of record for official records of the English royal courts from the time of the first surviving plea rolls in the mid1190s, and almost certainly from the creation of the first official records in the mid-1170s (as they were from the first of the final concords made in the new royal courts), this was certainly not the spoken language used in court. It is now clear that by the reign of Edward I French was the language used in the king’s courts not just for the formal initiation of pleadings but also in the subsequent argument conducted there. Woodbine’s suggestion in a 1943 article that prior to the middle of the thirteenth century English continued to be the language of the royal courts is difficult to disprove definitively but there is some evidence to suggest it is wrong. On the balance of probabilities it seems more likely that French was the language of the new royal courts from their establishment in the reign of Henry II onwards. Latin was also the language of formal record in county courts, city courts and even hundred and manor courts but the few surviving unofficial law reports of pleading in county courts (from around 1300) suggest that pleading in them too by then was in French, and the same goes for the few reports of pleading in city courts from around the same time. The evidence does not allow us to determine whether this practice too necessarily went back to the twelfth century (Brand 2000). It is against this background of the extensive use of French in royal courts, probably from the time of the the creation of the new royal courts in the last quarter of the twelfth century, and the subsequent creation as much as a century later of a group of professional lawyers in the royal courts in England that we need to view our surviving evidence for the creation and use of a specialist legal French lexicon in England. Our evidence for that lexicon comes from

95

AN-1a.indd 105

11/03/2010 12:00:20

The Anglo-Norman Language and Its Contexts a variety of sources. Law-reporting in England seems to begin around 1270 (Brand 1999b). The reports allow us to hear lawyers and justices in the direct speech they used in court, and within a decade they normally indicate which particular lawyers and justices were involved in the pleading of particular cases. During the initial stage (down to the summer of 1291) law-reporting was only on a limited scale. Just over four hundred law reports have been identified as belonging to this first stage. None survive in collections assigned to specific terms or years or eyre sessions. There was then a step change in law-reporting after the summer of 1291, apparently associated in the Common Bench with arrangements made to set aside a specific area in the courtroom for the use of the ‘apprentices of the Bench’, which allowed them to take notes on what they heard. Thereafter law-reporting was generally on a much larger scale, and substantial collections survive of reports assigned to particular terms or particular sessions of the eyre. Initially a significant minority of law reports were in Latin, but the majority were always in French, and French soon settles down as the normal language of law-reporting. This is hardly surprising if French was indeed the language of the discourse reporters were hearing in court. A notable feature of law-reporting in its early phases, and something which helps to confirm its wholly unofficial status and the absence of any kind of formal or even informal organisation lying behind it, is the existence of multiple surviving independent reports of the same case (in some cases as many as five or six) which, taken with the related plea-roll enrolment, help to create a much fuller, even kaleidoscopic, picture of what was said in court, much fuller than any one report could possibly do. It is not known for certain who made the earliest law reports or their reasons for doing so. Some may have been made by court clerks for the information of individual justices or to assist themselves in compiling the Latin enrolments; others by attorneys for the benefit of their clients or for their own use; others by clerks in the service of serjeants or serjeants not acting in the case; some by apprentices of the Common Bench both for their own instruction and for the instruction of others. The lectures of the 1270s were given in legal French, and other instruction seems sometimes to have been given in French and sometimes in Latin or was at least recorded in Latin. Of special importance here is the ‘treatise’ known as Brevia Placitata which may derive also from instruction, seems to date in its earliest version from around 1260 and provides us with evidence of Insular French legal vocabulary a good decade earlier than the first surviving law reports. Since the second half of the eleventh century English legislation had been issued in Latin. The earliest legislation to be promulgated in French comes from 1275, when not only the statute of Westminster I but also the statute of Jewry and the statutes of the Exchequer were issued in that language: Statutes of the Realm I (26–39, 197–8, 221–221a). But Latin was used for the statute of Mortmain of 1279 and the statute of Westminster II issued in 1285 (Statutes of the Realm I, 51, 71–95) and over the reign of Edward I as a whole the two languages had almost equal prominence in legislative use. With the reign

96

AN-1a.indd 106

11/03/2010 12:00:20

The Language of the English Legal Profession of his son, Edward II, French became the predominant but still not the only language of legislation, and it was to be more than a century before the last Latin legislation was issued (in the reign of Henry VI). Even earlier is the use of French as the language of legislative drafting. French is the language of the two early drafts of parts of the Provisions of Westminster of 1259 as well as of the penultimate draft of the whole of that same legislation, although the Provisions themselves are in Latin. It is also the language of the working draft of the statute of Acton Burnel of 1283, complete with deletions and interlineations, though the eventual language of the statute this time is also French (Brand 1992a: 325–67; 2007). These drafts help to confirm the supposition, which would in any case seem likely, that the language of discussion of legislation was also French. What do these sources have to tell us about the existence and development of a specialist legal lexicon in Insular French? Let me start with evidence relating to the legal profession itself. The older word for the person who appears in court to plead on behalf of a litigant is contur or in Latin narrator. For the former the earliest evidence comes from Brevia Placitata, of which the earliest element goes back to c. 1260, and from Hengham Magna of around the same date (Brevia Placitata 6, 20, 25 etc.; Hengham 34). The plea-roll evidence for narrator goes back to 1220, but the earlier date seems simply to reflect the fact that our Latin sources start earlier than our French ones (Brand 1992a: 48). The term itself seems to be derived from one (but only one) of the functions performed by such an agent, that of making the preliminary ‘count’ (conte or narracio), and the term exists also in Norman French, but not in other French dialects. The term used for pleaders in the law reports and also in the late-thirteenth-century treatise Britton, is serjant, which also has a Latin twin serviens, though this is much less commonly encountered (Brand 1992b: 94). There is also evidence of the portmanteau term serjant contour in c. 29 of the statute of Westminster I of 1275 and serviens narrator in the later legal treatise Fleta (Brand 1992b: 94). Legal historians are uncertain as to how the word for ‘servant’ came to be applied to what was the elite of the English legal profession, but what seems clear is that this is a distinctively insular lexical development, and one that seems to be connected with the emergence of a professional elite performing this function. What was to become the other main group of professional lawyers were the ‘attorneys’. Here the Latin word attornatus is attested from 1200 onwards and the French word only from the second half of the thirteenth century. However, the Latin word developed from a French root and the word was long used for someone performing the function of a representative appointed by a litigant, whether or not he was a professional (Brand 1992b: 46, 96). The terms ‘common’ (communis) or ‘general’ (generalis) attorney were sometimes used to distinguish the professional from the non-professional but do not seem to have entered into common currency (Brand 1992b: 86–7). The special term used initially for law students attached in some way to the Common Bench, aprentiz in French or in Latin apprenticius, found from the late 1280s onwards, is also obviously a development from an immediately French (and not Latin) root (apprendre) and

97

AN-1a.indd 107

11/03/2010 12:00:20

The Anglo-Norman Language and Its Contexts also seems distinctively insular in form as well as meaning (Brand 1992b: 110). A second area which clearly witnessed the development of a specialist French legal lexicon, though one we can only begin to see with the survival of the first French-language legal texts beginning in the 1250s, is the general area of litigation (EELR IV, xxxii–xxxvii). Civil litigation in the king’s courts was almost always initiated by royal writ (bref le roy) an insular term, and there were special verbs for acquiring a writ (se purchaser) and for using that writ to initiate litigation (a specialist sense of the common verb porter). The general term used for a law suit seems generally to be la parole. Within the law suit there were special terms for the initial oral claim or complaint: conte (or, in Latin, narracio) with conter (or, in Latin, narrare) for the related verb well attested from the time of Brevia Placitata onwards but probably much older. Defence (Latin defensio) is used for the corresponding opening speech (of general denial) for the defendant with the related verb defendre. Part of this defence was the formal denial of ‘wrong and force’ (tort e force) or whatever else had been alleged by the plaintiff before detailing his claim or complaint, and by 1277, and probably long before, these specific words in the count had come to be described as the ‘words of the court’ (mos de la court or verba curie) or just les paroles. The defendant (or his lawyer) would often try to argue that the plaintiff was not entitled to an answer to his claim or complaint as formulated. This was the specific context for a number of other specialist terms. One commonly found in the early reports is responable (Latin responsalis) meaning ‘entitled to receive an answer’, generally used in a context where an opponent is suggesting that a litigant is not so entitled. Striking also is the use of the verbal phrase gesir en sa bouche, ‘lies in his mouth’, or, more commonly, the negative ‘does not lie in his mouth’ (ne gist pas en sa bouche and variants), meaning something like ‘he is or is not allowed (by the rules of pleading) to say because only someone else may say it’. This is another commonly used phrase with an obviously specialised meaning. We also find a related use of gesir in relation to ability to bring an action in the phrase le bref gise en lor person (EELR II, 313). The invariable word for argument in the legal usage of the period seems to be reson, a specialist meaning of a much more general term; the term for an analogy used for the purposes of argument is similitude (EELR I, 87). The semi-technical phrase for a verbal assertion unsupported by the requisite documentation is either simple voice, simple dit or simple vent (EELR I, 103, 147). Pleading would generally end in one of two ways: either by the offer and acceptance of an issue for jury trial or by adjournment for, or the immediate giving of, judgment. The special phrase used for an offer of jury trial was to say that the litigant ‘is ready to prove’ (prest est de(l) averrer or prest est de mettre aveyr, or variants), though this is sometimes to be found in rather more elaborate forms. An alternative formulation for the same thing, though one normally used only retrospectively, and not when making the offer, was the phrase to ‘offer averement’ (tendre averement). If the offer was accepted (and the positive assertion denied) the technical term for this is transverser or traverser, a non-technical verb being used in a specific technical sense (EELR I, 34; EELR II,

98

AN-1a.indd 108

11/03/2010 12:00:21

The Language of the English Legal Profession 218). For there to be a joinder of issue between the parties it was (as we discover from discussions of the point) necessary that the two be maintaining genuinely contrary positions, that they be contrarius to each other, and that there be contrariusete between them, again technical terms. The process of reaching such an issue and proceeding to jury trial seems to have been described in the court’s own technical language as ‘descending’ (descendre) to jury trial. When the jury had assembled to give its verdict it would be ‘charged’ (charge) to do so by the presiding justice, which seems generally to mean ‘told (specifically) what it was to give its verdict on’, again a general word being used in a specific technical sense. This would then be followed by a presentation by the parties or their counsel to the jury, which could be described either as an evidence, or as an aveyement. The verdict itself was of course the veirdit, another specialist technical term. Pleading was, to judge from the reports, peppered with requests for judgment, usually with the formulation of demandom jugement, but sometimes using alternative and apparently synonymous nouns for judgment such as descrecion or agars or some combination of those terms. When both parties or their serjeants agreed that the court’s (considered) judgment on a particular point of law should determine the outcome of the case, the term used for this in the reports was to ‘demur for judgment’ or perhaps literally ‘wait for judgment’ (demorer en jugement). This soon becomes a technical term. The corresponding phrase used by a justice when adjourning the parties for a future judgment was ‘await your judgments’ (attendez vos jugemenz). A third area where we can also see the development of a specialist lexicon is in substantive law. One particularly rich area is land law. Already by the mid-thirteenth century fee had come to have a variety of technical meanings in law, one of which was a hereditary interest in land. Developments in land law in the third quarter of the thirteenth century made it useful to distinguish between different basic types of hereditary estate (those capable of being inherited by all heirs and freely alienable; those which could only be inherited by special classes of descendants of the original grantee) and we begin to see the developoment of new terms for the former: fee simple (EELR III, 288), or its early competitor, fee pure (EELR III, 237), and of a new term for the type of fee from which they were being distinguished, fee taille (EELR III, 237, 288). I have also found one instance of the term ‘full feoffment’ (pleyn feffement) for a grant in fee simple, as opposed to the grant of a more limited interest in land (EELR III, 147). Another term increasingly used in a specific technical sense (which is distinct from its earlier and wider use) is franc tenement for an interest lasting only for a lifetime. Also developing specific technical meanings in the context of rights in land and their devolution are the non-technical verbs remeyndre, reverter and descendre. Evidently also part of the technical language of land law are the terms mutacion or transmutacion to denote the successful transfer of an estate (for an example see EELR III, 237), or of possession (EELR III, 296; EELR IV, 547); feffement (and the associated verb feffer) for a grant as well as the more technical alienacion (EELR I, 49), and fet and especiaute in the specific technical

99

AN-1a.indd 109

11/03/2010 12:00:21

The Anglo-Norman Language and Its Contexts meaning of written deeds (EELR I, 176). A special technical term for those who have acquired their interest through alienation is estrange purchasur (EELR II, 214), for those in land by descent prive (EELR II, 234). Technical terms related to the hereditary descent of property rights and challenges to hereditary claims include (for)estranger for the process of alleging or showing that a supposed relative was not of the blood of the person concerned (EELR III, 97; EELR IV, 454), as well as bastarder for proving that a litigant or someone involved in litigation was a bastard (EELR II, 212); resort and resortyr for the process of tracing a line of inheritance upwards in order then to follow it down to a collateral heir (EELR II, 263, and III, 170; EELR I, 41, and IV, 574); and use of the term general heyr to distinguish what would later be called an ‘heir general’ (EELR I, 99). Other areas where specialist terms developed include the law of distraint, where the process of taking a distress comes to be called destreindre and the plea in justification of a distraint an avowerie (with the associated verb avower), and the term aver develops the quite specialised legal meaning of an animal of any kind when taken in distraint. A second is the law of obligations where the need to distinguish between an heir whose ancestor had placed him under an obligation and who had also inherited assets from him from their heir who was under the obligation but did not have the benefit led to the emergence of the specialist legal term heir du sank for the latter. The term asset is itself another such technical term in the law of obligations where its context can be this one but may also be a number of others. It was originally adverbial but was easily construed as a noun and it had certainly turned into a noun by 1300, when a pleader said mes le abbe asset aset en a par quey estre justice et athache a respundre del ravisement (BL MS Add. 37657, fol. 31r). A third is the the law of tort or wrongs where trespas develops the specific legal meaning of offence (with the related verb trespasser and agent noun trespassur). It is unfortunate that our evidence for a specialist French legal vocabulary in England only really begins around the middle of the thirteenth century. The evidence for a specialised English legal vocabulary in Latin begins before 1200 with the earliest surviving plea rolls and treatises. Behind much of this it is possible to detect or suspect a specialist legal vocabulary in French from which the Latin has been derived, but it is not possible to demonstrate it at all until around 1260 and it is only in the last decade of the thirteenth century that our evidence becomes really plentiful. The emergence of that evidence is itself closely connected with the emergence of a legal profession, but its late emergence makes it difficult to distinguish clearly between vocabulary in specialist use that predates the emergence of a legal profession and that which developed within that profession. The phenomenon of a specialist legal vocabulary in Insular French certainly predates the emergence of the profession and cannot owe its existence to that profession: so it is probably right to think and talk of a specialist vocabulary or jargon that came to be the distinctive domain of the legal profession and came to owe its further development to that profession, rather than one originally developed by the legal profession.

100

AN-1a.indd 110

11/03/2010 12:00:21

The Language of the English Legal Profession Primary sources Brevia Placitata, ed. G. Turner and T. Plucknett. Selden Society 66, 1951. EELR, vols. I–IV. Hengham: Radulphi de Hengham Summae, ed. W. Dunham Jr. Cambridge, 1932. Statutes of the Realm I, ed. A. Luders and others. Record Commission, 1810. British Library MS Add. 37657.

Bibliography Brand, P. (1992a), The Making of the Common Law. London. ——(1992b), The Origins of the English Legal Profession. Oxford. ——(1999a), ‘Legal education in England before the Inns of Court’, in J. Bush and A. Wijffels (eds.), Learning the Law: Teaching and the Transmission of Law in England, 1150–1900, London, pp. 51–84. —— (1999b), ‘Observing and recording the medieval bar and bench at work: the origins of law reporting in England’. London. ——(2000), ‘The languages of the law in later medieval England’, in D. Trotter (ed.), Multilingualism in Later Medieval Britain, Cambridge, pp. 63–76. ——(2007), ‘The making of English thirteenth century legislation: some new evidence’, in A. Lewis, P. Brand and P. Mitchell (eds.), Law in the City: Proceedings of the Seventeenth British Legal History Conference, London (and Dublin, 2005), pp. 42–53. ——(2008), ‘A working draft of the statute of Acton Burnell’, in P. Brand and S. Cunningham (eds.), Foundations of Medieval Scholarship: Records Edited in Honour of David Crook. York and London. Woodbine, G. (1943), ‘The language of English law’, Speculum 18, 395–436.

101

AN-1a.indd 111

11/03/2010 12:00:21

chapter nine

Mapping Insular French Texts? Ideas for Localisation and Correlated Dialectology in Manuscript Materials of Medieval England* Jean-Pascal Pouzet

Chaucer’s Prioress’s ‘Frenssh’, spoken ‘after the scole of Stratford atte Bowe’, has been taken as an icon of insular Francophone degeneracy over time, when judged by the reconstructive standards of the ‘French of France’ – especially the phantom category of ‘Francien’ – applied to the ‘eccentric cousin’. But as the French of England is now reinstated in a continuum of evolution and variability with the dialects of ‘French French’, it may be worth emphasising that for all the irony in the Chaucerian depiction, there is no overt derogatory pronouncement on the French ‘form of speche’ itself: rather, this French is spoken ‘ful faire and fetisly’, and simply in ignorance of Paris French – as might be customarily expected in any late medieval English religious setting, in this case a Benedictine nunnery dedicated to St Leonard. It may well be, rather, that such a form of French stands in potential contradistinction to other English local realisations of French – here perhaps, as once suggested, the geographically close nunnery of St Leonard at Barking – and that the topical references to ‘Stratford atte Bowe’ and ‘Parys’ serve more essentially to project forms of French onto a linguistic map where issues of locality and situational sociolinguistic difference are brought into relief.1 * My thanks are due to R. Ingham for his patience and support regarding the revised form of this article; to J. Wogan-Browne for offering an opportunity to present a slightly revised version at the York ‘French of England’ Conference (‘Linguistic Accommodation and Cultural Hybridity, c. 1100 – c. 1500’; 13–16 July 2007); and to the members of audiences in Birmingham and York who, while underlining the difficulty of the task, expressed interest and offered valuable response, particularly D. Burrows, A. Butterfield, N. Cartlidge, M. Chambers, G. de Wilde, J. Gilbert, A. McAllister, D. Russell, N. Romanova, and D. Trotter. T. Hunt kindly read the essay in draft, and saved me from stylistic errors. My sister Guylène has likewise offered help and good sense. Much has been put into this article, but I remain solely responsible for, perhaps, failing to take enough into account. Some of the issues developed here are also broached in Pouzet (2009a). 1



Chaucer, Canterbury Tales, General Prologue, I.A.124–6 (1988: 25; see pp. 803–4 for classic bibliography and discussion). On some of the resonances behind ‘fetisly’, see

102

AN-1a.indd 112

11/03/2010 12:00:21

Mapping Insular French Texts? This article is a report on early work planned for a long-term project on just one area of that topic – insular forms of the French language investigated according to their locality, or indeed, localisability – with the purpose of highlighting the assumptions underlying this project, the limitations it may face, together with some of the potential solutions and results attainable through flexible yet rigorous approaches spanning various bibliographies and methodologies. The substance of this essay relies on a corpus of manuscript data which is slowly being assembled and interpreted, and some relevant illustrative manuscripts are briefly mentioned. For reasons which are discussed below, this corpus is concerned with the manuscript and scribal context of French texts in an insular situation of production and transmission, and what it can tell us if we look at socio-cultural and book-making (codicological and palaeographical) features, and simultaneously at spellings and spelling variants, the relationship between spelling and underlying phonetic data, lexicography, and syntax. On account of the format of this contribution, of its work-in-progress nature, and of its attempted multi-disciplinary dimension, the bibliographical references supporting the theoretical framework are far from exhaustive for each of its classic constituents (manuscript studies, linguistic theory and literary history), and its disciplinary subcategories (‘French’, ‘Anglo-Norman’2 and ‘Middle English’3 studies) which are still insufficiently interactive.4 In the light of this desirable heightened interaction, it is appropriate to posit definitions for the forms of language considered and the concepts used to describe them. As shown by recent fascinating work on glosses, macaronic texts in general and mixed-language administrative and business records in particular, materials with Latin as a matrix language hosting vernacular English and French words (one sort not always easily distinguishable from the other) are duly taken into account.5 The notion of ‘Middle English’ as a matrix language straddles either side of the divide ‘early’ vs. ‘late’ recently re-theorised in the new online Linguistic Atlas of Early Medieval English, and the Linguistic Atlas of Late Medieval English; spanning therefore, respectively, the periods c. 1150 to c. 1325, then c. 1300 to c. 1530 – with a grey zone of borderline cases for some of the materials associated with the decades around 2



3





4

5

Williams (2004). The benefits derived from taking this new look at Insular French are shown by Trotter (2003a; ‘eccentric cousin’ at p. 427). In the field of Anglo-Norman, significant overviews of general or specific areas of research touching on issues of localisation include Vising (1923), Legge (1963), Sinclair (1965) and Dean (1972), to which, more recently, Dean and Boulton (1999), Harper-Bill and van Houts (2003) and Short (2007) must of course be added. In the area of Middle English language, and dialect studies in particular, further relevant references are to be found in the body and bibliographical sections of Kristensson (1967–2002), McIntosh et al. (1986), Laing (ed.) (1989), Blake (ed.) (1992), and more recently Machan (2003), Mugglestone (ed.) (2006; especially in relation with chapters 3 and 4), and Laing and Lass (2006). See, however, Clanchy (1993). References include: Hunt (1991), Wenzel (1994), Rothwell (1992), Schendl (2002), and the work of Wright (1997, 2000, in this publication).

103

AN-1a.indd 113

11/03/2010 12:00:21

The Anglo-Norman Language and Its Contexts 1325. Typically, ‘Insular French’ is taken to cover all material for which the matrix language is recognisably French, and which geographically stems from the British Isles as opposed to the continent, over roughly the same global timeframe of c. 1150 to c. 1500. There is, by inference, an analogous distinction of earlier and later texts, but with time-lines not necessarily having to replicate those for English – particularly since they should usefully serve also to shed new light on the classic diagnosis of the ‘precocity’ and ‘obsolescence’ syndromes of insular production in French throughout the medieval period.6 This definition – which accepts ‘Anglo-French’ (preferably) or ‘Anglo-Norman’ as interchangeable terms – is therefore meant to include material which can be shown to be, with the greatest degree of certainty and consistency:7 A. Insular copies of genuinely, or exclusively, insular artefacts: e.g. all documentary evidence provided by local archival documents (in particular economic and business materials: cartularies, account-books and rolls, rentals, custumals);8 or such a ‘religious/literary’ creation as the early-fourteenth-century Chant des chanz, known to survive uniquely in Oxford, Bodleian Library, MS Rawlinson Poetry 234 (fols. 57r–105r; second half of 13th c.; matrix Insular French; probably Yorkshire), on which see Hunt (ed.) (2004). B. indisputably insular copies of continental texts: e.g. late-twelfth- to midfourteenth-century insular copies of the ‘Assumption de Notre Dame’ section of the originally eastern French biblical poem known as Herman de Valenciennes’s Bible or Roman de Dieu et de sa mere; that section seems to have enjoyed distinct insular dissemination, from an initial archetype probably not unlike Paris, Bibliothèque nationale de France, nouv. acq. fr. 4503 (fols. 1r–11r; last quarter of 12th c.; matrix Insular French; no clear origin or provenance established yet),9 and is found notably (but not exclusively) in: ▪ London, British Library, MS Egerton 2710 (fols. 136r–138v; second half of 13th c.; matrix Latin and Insular French; provenance: later inscription fol. 83v (late 14th to early 15th c.?) associates it with the priory of Benedictine nuns at Derby, Derbyshire) ▪ MS Harley 2253 (fols. 23ra–33va; datable 1330 x 1340; matrix Latin, English and Insular French; its English constituents multi-layered, but copying by scribes A and B associated with Ludlow, West Midlands; and probably as well

6



7



8



9

Such discussion cannot take place here. Suffice it to suggest that the classic study by Legge (1963) has been followed by studies insisting accordingly on precocity, e.g. Legge (1965), or terminal stage, e.g. Berndt (1976). The prevalence of a literary perspective, or of a corpus of literary texts, must be noted in most cases, unlike that of, for example, Baker (1979). A more elaborate version of the following categories is presented in Pouzet (2009a). The availability of such databases as PROME, ed. C. Given-Wilson et al. (2005 or 2006) is happily enlarging the research facilities of important corpora which generally have relatively well-defined locality. A description, to be published in Careri et al. (forthcoming), is available (as a sample of the book) at http://www.irht.cnrs.fr/recherche/programme_catlitfr_english.htm.

104

AN-1a.indd 114

11/03/2010 12:00:21

Mapping Insular French Texts? for passages in MSS Harley 273 and Fouke le Fitz Waryn in Royal 12. C. xii copied by scribe B) ▪ MS Harley 5234 (fols. 153ra–155va; late 13th to early 14th c.; matrix Latin and Insular French; provenance: associated with Durham Cathedral) ▪ Oxford, Bodleian Library, MS Digby 86 (fols. 169r–177r; datable 1330 x 1340; matrix Latin, English and Insular French; its English constituents multi-layered, but mainly assigned to Worcestershire) ▪ Paris, Bibliothèque nationale de France, fr. 19525 (fols. 8r–12r; end of 13th c.; matrix Insular French only; no clear origin or provenance established yet).10 This definition of ‘Insular French’ is naturally set against continental texts variously circulating in the British Isles (for example the royal booty of ‘French of France’ manuscripts after the Battle of Poitiers in 1356; or Cambridge, University Library, MS Ff.i.33, most of which was written, as a colophon reveals (fol. 76v), by Lepainteur in Bourges in May 1420, and procured at an unknown date (ante 1440) by John Shirley on account of his engagement with translations from French),11 or continental copies of Insular French texts – for example some manuscripts of the so-called ‘Anglo-Norman Brut’ studied by Matheson (1998: especially 11–12 and 30–7); and possibly some manuscripts of Mandeville, on which see recently Bennett (2006). This implies a new look at a timehallowed and thorny issue of ‘Anglo-Norman’ studies: the rehearsal of a clear set of criteria to tell Insular from Continental, and an idea of how such criteria must combine to build cumulative evidence; such a discriminatory approach has been attempted for a long time, most prevailingly in the classic linguistic studies by Menger (1904) and Pope (1934, revised edition 1952, re-issued 1966).12 Precisely of what nature those parameters must be (morphology, phonology, lexis and syntax), and how their respective degrees of discriminatory certainty can be assessed, are currently matters subject to renewed attention, and certainly central issues in any attempts at linguistic qualifications of Insular French materials. In my approach, a convergence of all parameters is crucial; but for reasons of space, what follows attempts to concentrate on what has been recently termed (in a context of English dialectology: Laing 1999) the ‘sound–symbol correspondence’ of the insular forms of the French language, with a consideration of the scribal realisations of the graphemic/phonemic interface in lexis as features aiding or permitting localisation. The other crucial factor to mention is the essential inspiration behind this project, which is predicated on two interrelated strands of argument. The first 10

11





12

This is not a complete list of insular manuscripts. On Herman’s poem, see Spiele (ed.) (1975). It is to be regretted that only the older bibliography is mentioned in relation to it in the important publication of Parkes and Tschann (ed. and intr.) (1996: xxix). For a recent discussion of some aspects of the ‘Assomption’ in Harley 2253 in relation to some other Insular manuscripts (including those mentioned here), see Thompson (2000). On CUL Ff.i.33, see Connolly (1998: 120–44). Vising (1923: 27–33) had also contributed a précis of Anglo-Norman linguistic traits or tendencies.

105

AN-1a.indd 115

11/03/2010 12:00:21

The Anglo-Norman Language and Its Contexts takes into account the fact that recent work in the fields of Middle English literary, cultural and linguistic studies has laid emphasis on the ‘situatedness’ of insular Latin and vernacular English manuscript production, in an age of pre-national, highly localised textual creation, with high potentials of applicability to all forms of insular textual creativity.13 Ideally, this approach would be concerned with integrating Insular French more systematically into a global collaborative attempt to map out a ‘literary geography’ of medieval England from c. 1150 until c. 1500. This notion of a ‘literary geography’ is borrowed from scholars who – following Elisabeth Salter’s magisterial sketch of English poetry in the fourteenth century (1983: 52–85) – have offered prolegomena or more to achieving such a critical panorama: notably, Thorlac Turville-Petre for the later medieval north-east Midlands (1983), Richard Beadle for Norfolk during the whole of the Middle English period (1991) and Ralph Hanna for fourteenth-century Yorkshire (2003, 2008).14 Naturally, it has been shown that any global concept of an English ‘literary geography’ must be predicated on the identification and analysis of linguistic, specifically dialectal, data in available manuscripts. The relations between the fixed nature of local records and the distance covered by English texts ‘dialectally translated’ from shire to shire create heuristic tensions between local and trans-local (if not pre-’national’) archives which emerge with particular acuteness (though along diverging lines) in the larger narratives (Turville-Petre 1996; Hanna 2005a; Scase 2007; Horobin forthcoming). Thus the qualification and appreciation of discursive configurations depend heavily on the localisation of such available manuscript data as is within the scope of dialectal investigation. But this has proved all the more fruitful when achieved in close combination with archival evidence linking a manuscript or a portion of it securely to one area: ideally a centre of production, sometimes matching the place of preservation, if the material has not travelled too far away from its original dialectal area. Michael Benskin has appropriately characterised such an imagined locality of archival evidence for English, together with the benefits that can be derived from it: The one large body of material for which local origins are in most cases either explicit or readily deduced are the local documents. ‘Local documents’ is, of course, merely a convenient label for a mass of materials of very diverse origins. For our purposes the term covers personal correspondence, whether great collections like the Paston Letters or stray survivals from persons otherwise unknown; municipal records, like the Dublin Assembly Rolls and the Coventry Leet Book, records of corporation meetings and enactments; the records of courts, secular or ecclesiastical (though the latter are commonly in Latin), of manors, liberties and municipalities; and legal instruments – 13

14





Rigg (1992) and Sharpe (2001; together with its Supplements, paper and online) can also be read for all the hints of localising circumstances they provide for Latin Insular writing. To these references may be added Beadle (1994) and Lawrence-Mathers (1995, 2003, in a context of dominant Latinitas).

106

AN-1a.indd 116

11/03/2010 12:00:21

Mapping Insular French Texts? depositions and indentures, conveyances and arbitrations. All of these can be expected to have quite explicit indications of their local origins, and in general they can be trusted to represent the written language if not of the stated place, then somewhere near it. (Benskin 1977: 501–2)

Although Benskin also recognises ‘exceptions’ to that mechanism of localisation, in ‘the work of scribes whose linguistics habits were acquired at greater or lesser distances from the places whence their documents are related’, that caveat has simply been an incentive for devising ‘principles by which scribal dialects can be assigned more or less confidently to different areas on linguistic grounds alone’, which LALME has impressively identified and LAEME is revising as appropriate. Furthermore, such local documentary materials listed by Benskin (to which the vast body of local religious and administrative records may be added) may not exclude Middle English ‘literary’ texts, which are not so rarely found in localisable archival depositories.15 It is argued here that the methodology of monitoring them and analysing their language in more exact topical circumstances could now be extended to and tried out on Insular French materials. Together with this first strand of thinking, an interest in mapping materials is also kindled by a growing sense of dissatisfaction with a scenario of uniformity of Insular French, because that scenario has declared it impossible to identify any local, let alone dialectal, realisation of the French language in the British Isles (and England in particular). Received wisdom has been that communication in a supra-regional context could be effortlessly conducted either in Latin or French, as it could not in English without inter-dialectal difficulties, thus lending a stability to the paradigms of ‘Anglo-Norman’ that enhanced it almost to the status of a pre-national koine, and served to forge a notion of monolithic insularity in the use of French.16 From the start up to the late fifteenth century, this presupposes – but never demonstrates – that not only did the factors of dialectal diversity in continental France fail as French crossed the Channel (or at best survived in a negligible way), but that the homogeneity of French essentially counteracted the fragmented distribution of English in its regional realisations – witness, for instance, the famous late-fourteenthcentury elaborations on a passage of Ranulf Higden’s Polychronicon (Book I, chapter LIX), in manuscripts of John Trevisa’s translation:

15



16



For instance, even with the greater part composed of Latin materials, some of the cartularies listed by Davis (1958) are in matrix Insular French, or include nonmatrix French syntagms (the language(s) of record are generally not reported in the entries; a revised list is in preparation). For a recent view on the ‘exploitability’ of local archives for literary texts, see Boffey (2000). Essays edited by Lavezzo (2004) have almost nothing on Insular French in their fascinating avoidance of factors of cultural or linguistic homogeneity; for instance, A. Galloway’s ‘Latin England’ (pp. 41–95) provocatively eschews the ‘fossilisation syndrome’ often attached to the use of Latin in late medieval England, but does not discuss Anglo-Norman.

107

AN-1a.indd 117

11/03/2010 12:00:21

The Anglo-Norman Language and Its Contexts Hit semeþ a greet wonder how Englische, þat is þe burþe tonge of Englisshemen and her owne langage and tonge, is so dyuerse of sown in þis oon ilond, and þe langage of Normandie is comlynge of anoþer londe, and hath oon manere soun among alle men þat spekeþ hit ariʒt in Engelond; for a man of Kente, southern, western and northern men speken Frenssche al lyke in sowne and speche, but they can not speke theyr Englyssh so. Treuisa. Neuerþeles þere is as many dyuers manere Frensche in þe reem of Fraunce as is dyuers manere Englische in þe reem of Engelond. (Babington 1869, vol. II: I, LIX: 160–1)17

Although Trevisa’s own addition, concerning the proportionate comparison of the dialectal diversity of the kingdoms of France and England for their respective nominal vernaculars, is perhaps more directly congruent with axiomatic facts of socio-linguistic reality and geographic fragmentation, Higden/Trevisa seemed to be more directly concerned with correctness (‘spekeþ [. . .] aright’) which, if not in itself a chimera, glosses over the (less ‘perfect’?) realisations of the local varieties of Insular French, whether oral or written, which must also have existed. Even Legge (1967: 44–5), after quoting that passage from Trevisa in a brief study on the origins of literary Anglo-Norman – in method a follow-up from her magisterial study of 1963 – was inspired to express an awareness of non-standardisation in the Frenches of England – also drawing on early work by Paul Zumthor, significantly concerned with variance. But the almost ubiquitous theoretical suppression of Anglo-Norman local variability has caused any localisation of a manuscript containing Insular French to be taken to rely on an examination of non-linguistic data, making language an intractably unsearched and underestimated parameter for those purposes. This may serve as background to the desirability of resituating the Insular French language at the core of language-historical enquiry, alongside book production and literary history: the corpus must consist primarily of manuscripts, while recourse to critical editions may be substantial. Many recent efforts in Middle English literary history have been aimed at grappling with individual texts or ‘textual situations’ in their manuscript, specifically codicological and scribal contexts, so as to apprehend better the implications of their associations (with regard to their gathering in booklets, for instance), and of the cultural analogies and differences of their materials in comparison with other artefacts.18 Alongside these efforts, the idea is indeed to look at manuscripts again and to revisit manuscript data alongside critical editions wherever they exist, in order to interpret them according to principles of systematic data-collecting; some of these principles are described in the 17



18



The italicised clause is found in a manuscript which is not the base manuscript of the Rolls Series edition (for details of manuscripts, see Babington 1869, vol. I). The phrase (and part of the advocacy of) ‘textual situations’ is from Taylor (2002), which includes in particular a chapter on an ‘Anglo-Norman’ celebrity, Oxford, Bodleian Library, MS Digby 23. On these efforts, see – stemming here essentially from departments of English – the general and particular issues raised (for later medieval England) in Griffiths and Pearsall (1989) and in the work of Hanna (1996a, b, 2001, 2004, 2005a); this spirit still inspires Gillespie and Wakelin (forthcoming).

108

AN-1a.indd 118

11/03/2010 12:00:22

Mapping Insular French Texts? later part of this article. This is not to dispute the intrinsic value of critical editions (generally strong in publications under the auspices of the AngloNorman Text Society), but to take fully into account the necessary historical relativism of their various persuasions (‘Lachmannism’, ‘Bédierism’, or their prevailing pragmatic descendant, ‘best-textism’),19 and the implications those have on the presentation and treatment of scribal lemmata.20 Such a critical return to the manuscripts has been gaining ground in particular in the field of Old and Early Middle English editing, especially (but not exclusively) in the case of multiple copies of a text, whether religious, literary, scientific or ‘documentary’ (meaning essentially administrative or legal):21 beyond the ‘obvious’ realms of matters relating to the editing of Langland and Chaucer, good cases in point are the recent work on the Middle English biblical poem La estorie del evangelie and the dialect of its original version, ‘very little to the south of south Lincolnshire’22 – or the salvaging of ‘bad texts’.23 The other service this does is to double-check what the critical apparatuses of editions may report only unobtrusively, selectively, or not at all. Although we have a valuable array of sound editions to go by, what matters for the linguist-cum-historian of book production may not always shine through. Additionally, there is always the case of faulty readings in some of them, whether they are bona fide errors or reconstructive blunders – on the latter sort, William Rothwell has mounted a meticulously researched vindication of the achievements of insular scribes of French material against the judgements of their editors (Rothwell 2004).24 It is argued that, beyond a salutary assault on the excesses of modern prescriptive 19





20

21





22



23



24

Landmark discussions on such issues include the Kane and Donaldson Piers Plowman A-Version and B-Version editorial procedures, together with some of their fruitful descendants; some of the issues concerned are addressed for instance in some of the essays by Hanna (1996a). A useful assessment of an applicability of best-text policies to Chaucer texts is found in Fisher (1994). Studies with direct focus on, or applicability to, the vernaculars of medieval England where those issues are addressed include Fleischmann (1996), Greetham (1996) and McCarren and Moffat (1998). An excellently argued approach – beyond the field of medieval studies per se, but with fruitful reflection on it – is McGann (1983, reprinted 1992, re-issued 1996). The return to manuscripts has marked the last two decades of Old English studies as well, since the important Tokyo seminar whose proceedings have been published as Sato (1997). For practical and scientific reasons, my corpus has started essentially with religious and literary material, but on account of the miscellaneous nature of a substantial portion of manuscripts, there is also an early intake of scientific and administrative/legal material – the latter category being represented by some multilingual manuscripts of Registers of Writs in particular (and their codicological and textual associates), on which see de Haas and Hall (1970). This qualification was established by McIntosh (1987: 191); see also Pickering (1972) and Millward (1997; ed. 1998). On this issue, see for instance Harris (1983). The preference for manuscripts or their facsimiles over editions is made clear and radical by Laing and Lass (2006: 426–7). The English and French texts published by Reichl (1995) from Oxford, Bodleian Library, MS Digby 2 unfortunately provide an example of editorial errors.

109

AN-1a.indd 119

11/03/2010 12:00:22

The Anglo-Norman Language and Its Contexts philologies of Medieval French in general, such a vindication may also serve to expose a priori assumptions about the non-localisability of the Frenches of England on linguistic grounds – although Rothwell does not seem prepared to address this issue. Taking Tony Hunt’s and Michael Benskin’s joint edition of three fourteenth-century medical texts in Insular French including substantial English items (Hunt and Benskin 2001), he repeats that ‘of the three [languages in use in later medieval England], only Middle English can be said to lend itself to the kind of orthographical/phonological exercise undertaken for [spellings in] the Ancrene Riwle, because Middle English alone is grounded in a living vernacular’. Whilst it may be received as axiomatic that definitions of what a ‘living vernacular’ may have been in medieval England must be taken to vary according to a complex of factors, Rothwell continues: In the Three Receptaria from Medieval England, the spelling in the Middle English sections is certainly not ‘regular’, but the links with one specific regional dialect are so close that Michael Benskin is able to establish the precise area of it composition; no such localisation, however, is possible for the Anglo-French element in the texts, although the range of spelling variants is broadly of the same order as in the Middle English sections, being based, however, not on geographical differences, but on the training or social position of those who used it. (Rothwell 2004: 16)

But precisely issues of scribal ‘training’, and their potential relation to locality – scriptoria, or ‘scriptorial’ venues or amenities of any description in local households, shall we say – might repay further work. Furthermore, in that very publication, in acknowledging that ‘regional variation within AngloFrench is yet uncharted’, Benskin (in Hunt and Benskin 2001: 194) may also be hinting at the desirability of attempting just such charting. My project is thus an attempt to localise Insular French manuscript materials on a broad basis, represented not only – as has been attempted already – by a range of non-linguistic factors (the features of the manuscript and what they can tell us about its socio-cultural contexts), but also all aspects of scribal evidence (from the scripts to the forms of language which they embody). More specifically on the linguistic front, this project is also in fact a plea for what I call tentatively ‘correlated dialectology’ – a critical gesture which is practically non-existent in the field of Anglo-Norman. For all its achievements, nineteenthcentury Neogrammarian scholarship in Romance philology, when applied to Insular French, virtually declined to confront the issue – one of its significant products, Menger’s Manual published in 1904, inaugurated a concerted attempt at establishing what made the ‘Anglo-Norman dialect’ a distinct province of ‘Gallo-Romania’, at the macro-insular level only (see also Legge 1965). In an annual public address delivered at the Académie des Inscriptions et BellesLettres, André Crépin (2004) has restated that the enthusiastic endeavours of the Cambridge ‘Anglo-Normanists’ of the early 1930s were short-lived; one important lead by O. H. Prior (1924: xx), which was never seriously taken up, was the possible localisation of Insular French materials by means of patterns

110

AN-1a.indd 120

11/03/2010 12:00:22

Mapping Insular French Texts? of diphthongisation influenced by distinctive contacts with Middle English.25 Vising (1923: 39–40) took an unsettling approach, declaring that ‘concerning the locality to which Anglo-Norman works belong it is in most cases impossible to state anything with certainty’, but he immediately hastened to add that it was ‘possible, however, to make certain inferences’, regrettably not substantiated, wherewith he produced a list of some Insular French texts classified by shire; it was certainly not compiled on the strength of linguistic evidence. The idea behind this embryonic repository has been developed in some significant measure by M. D. Legge’s concern with the cultural, but not linguistic, origins of Insular French materials, together with passing ‘inferences’ in the introductions to several (but far from all) Anglo-Norman Text Society editions – naturally richer when the published work is by a known Insular French writer. With ‘correlated dialectology’, basically, the backbone of my argument is to ask whether it might not be possible to apply to the study of Insular French material the theoretical framework developed for the assessment of dialectal forms of the Medieval English language in LALME and LAEME, and to combine any results more systematically with the socio-cultural and material, ‘book-oriented’ (codicological and palaeographical) evidence yielded by individual manuscripts. In the field of medieval French (its insular variety in particular) it would be inconsiderate not to heed the intellectual achievements that Middle English dialect studies have made over the last fifty years or so, given the sheer methodological profit lying there.26 But beyond this general sort of emulation, there is also a question central to my approach, though the need to ask it has hardly been felt so far: whether there is a potential transferability of patterns of analysis from advanced dialect studies in English to its vernacular semblable, Insular French; and if so, under precisely what terms and conditions. In this perspective, the more decisive argument in favour of correlated dialectology applied to insular manuscripts lies in Edinburgh’s most recent theoretical stance for LAEME, in the process of collecting and interpreting Early Middle English material. A very brief presentation of that stance cannot do justice to the fascinating act of labour and imagination it performs, but should nevertheless point out that ongoing work has offered an opportunity for a necessary theoretical revision of LALME, because LALME principles, palpably the now classical questionnaire and ‘fit technique’ in particular, cannot be straightforward because of the greater paucity of Early Middle English See Crépin (2004: 17 n. 48; 18, and n. 51), quoting O. H. Prior’s linguistic introduction to the first (and only) volume of Cambridge Anglo-Norman Texts (Prior 1924, xx). As Sinclair (1965: 114–15) reminds us, Prior’s introduction had appeared the year before in France (Prior 1923). 26 Those achievements started in particular with the pre-war pioneering work of Moore, Meech and Whitehall (1935), and Kaiser (1937); they continued with such landmarks preceding LALME (and LAEME) as McIntosh (1956, 1989, 1963), Samuels (1963) and the studies assembled by Benskin and Samuels (1981). Pouzet (2009a) comments on the awareness of the difficulties concerning the localisation of Old French literary texts underlying the work of Dees (1987) and Busby (2002). 25

111

AN-1a.indd 121

11/03/2010 12:00:22

The Anglo-Norman Language and Its Contexts materials, and more importantly of the non-linear charts of data survival for any given dialect area. One other, logically related revision concerns the less directly ascertainable role played by Mischsprache in the overall regime of Early Middle English material scanned. For a correct interpretation of data included in LAEME, Margaret Laing (1999: 255–7), following Michael Benskin, emphasises the usefulness of resorting to medieval explanatory concepts for the correspondence between a given sound and the symbol that is meant to transcribe it: When one wishes to ‘get inside the head’ of a medieval copyist, reference to graphemes and allographs seems neither appropriate nor useful. The doctrine of littera includes the concepts figura, nomen and potestas. Littera is the abstract idea of ‘the letter’, which may be symbolised by a number of different shapes [. . .] and referred to by nomen – the letter-name. Figura is the letter-symbol, its realisation or shape. Potestas is the sound value implied by the littera.

She further argues that, in a period when the relatively greater consistency of later Middle English writing practices and dialectal differences was not yet so consistently established, it is even more possible that ‘a single figura may do service for three litterae and therefore may imply also any of the different possible potestates of those litterae. The logical extension of this phenomenon is the reverse process that many different figurae, or combinations of figurae, may be employed to realise the same postestas.’ This model is not always as bewildering as it may seem, but has to take account of both multi-dimensional variability and non-symmetrical consistencies. It is therefore further predicated on a distinction between ‘economical’ and ‘complex’ scribal strategies for sound–symbol correspondence: ‘economical systems will restrict themselves to very few variants’ within the littera–figura–potestas chain, and ‘complex systems make freer use of the possible choices’; but ‘however anarchic a complex system looks’, it remains systematic. From the theoretical viewpoint, this is because any factor indicating such variability in the sound–symbol correspondence practised by a given scribe is treated as an inbuilt pattern which is encoded within a larger chart of scribally attested correspondences between littera and potestas in Middle English writing systems. This explains the greater emphasis laid on a systematic ‘tagging’ of scribal realisations lifted from direct authentic contact with the manuscripts (or their reliable facsimiles and digital images): only at the level of manuscript evidence can each singular scribal system reveal anything consistent about letter–symbol correspondence procedures, in other words, about a scribe’s degree(s) of cognitive and morphological decoding of scripts from received exemplars, and re-encoding in his own scribal production – in a context of earlier inter-dialect circulation. Larger patterns of correspondence for a configuration of forms within one dialect area, or from one area to another, may thereafter prudently emerge from the cumulative tagging of

112

AN-1a.indd 122

11/03/2010 12:00:22

Mapping Insular French Texts? individual systems.27 Are these rashly summarised procedures in any way transferable to an assessment of the writing systems of Insular French? And since the divide is between earlier and later Middle English, should we also at any rate rethink the distinction between earlier and later Insular French? And if so, where do we place the divide? Though there can be no short answer to these fundamental questions which the project wishes to raise, a small example may serve to introduce some of the main issues at stake in an attempt at a transfer process. Some of the spelling variants of are recorded in the Anglo-Norman Dictionary (both paper and online versions) as .28 Any of those spellings in a given text may be taken to correspond to a scribe’s special encoding strategy for a given sound–symbol correspondence. Supposing that, analysing scribal evidence, we can work out parameters of all theoretical realisations of that correspondence, and we can match them again against all scribal realisations, we may form a clearer picture of what the scribes were at when they wrote a given sequence of letters for an Insular French lexeme: this could include suggesting that the spelling may either reflect the scribal realisation of the velar glide /j/ + rounded /u/ + /r/, or of an initial dental + affricate cluster /dʒ/ (with the same vowel grade); with we could have the realisation either of /j/ or /dʒ/ + unrounded /o/ – unless of course the distinction in spelling between and is not phonemically discriminative enough; and that that system is probably more ‘economical’ with the majority spelling than with the more ‘complex’ spelling – the exact scribal context of which should be examined with care. Taken at the level of minute observation and micro-analysis within a clearly defined sound– symbol template and a well-established palaeographical cadre, such patient investigation could be location-driven: an as yet unrecorded spelling in AND (printed or online), , contained in the Middle English Genesis and Exodus (considered briefly below) is from West Norfolk, and could be checked against the distribution of spellings for the scribal reflexes of in manuscripts presumed to be from the same shire, as well as in manuscripts further afield. The last important related factor is that, as regional variability in Insular French materials is predictably to be detected not in isolated features (which might prove insignificant), but – as for English dialectology – in an examination and appreciation of a combination of traits, an investigation of them needs to be conducted along clear lines of strict yet imaginative sampling and tagging procedures; so the phonological traits behind the allographs should be For ideas in this paragraph, see Laing (1999; quotations at 255–7), and more recently Laing and Lass (2006). Before these publications, on such issues Laing and McIntosh (1995) is also especially important. On Mischsprache, see the classic discussion by Benskin and Laing (1981). 28 AND (printed), Fascicule 3 (F–L), s.v. (1983: 372). As of 1 March 2009, the online AND can only offer partial revision (disposing of the spelling and adding further lexical collocations for ), because the letter J has not yet been fully covered by the AND online team. 27

113

AN-1a.indd 123

11/03/2010 12:00:22

The Anglo-Norman Language and Its Contexts matched against other significant phonemic/graphemic features. A list of such pertinent combinatorial features is certainly delicate to establish a priori, as it might have to be revised as we go; but it might be deemed safe to start from such a list of ‘obviously insular characteristics’ as the eleven features analysed by Hunt (1994: 268–9) in his critical edition of Cambridge, Trinity College, MS B.1.45 (fols. 19v–24),29 to which I would suggest adding hiatus vs. coalescence (e.g. vs. ), and regressive assimilation vs. non-assimilation in the case of a nasal following a diphthong (e.g. vs. ). The combinatorial process then could work at the level of the unambiguous encoding of scribal realisations of sound–symbol correspondences for lexemes which display any of those 11+ criteria, in carefully chosen hands and manuscripts. In this respect then, the notion of ‘correlating’ Insular French and Middle English material could take several forms resting on critical assumptions and involving methodological choices, which I could sum up as: correlating French and English as two diachronically and diatopically discrete linguistic entities, taking simultaneously into account both distinctive features and areas of complex overlap and dependency – in lexicography, phonetics, semantics and syntax. The fact that it is not always possible to tell which language influenced the other first for a given set inhering in any of those four discursive parameters in a given scribal context conditions their ongoing adjacency in time and place, and triggers the necessity for careful close correlation in their linguistic examination.30 If we accept the notion of correlating language studies and manuscript studies more closely (in the sense further developed below), a possible consequence is that, if we can identify the hand of a scribe as copying both French and English material in one or more than one manuscript, and if we can pin down the Middle English dialect of the scribe with any degree of certainty, then it may be possible to state those terms and conditions under which it would be possible to transfer the benefits of the analysis of Middle English to French material copied in the same hand, with a view to establishing its own linguistic profile. In short, the project is a reappraisal of areal linguistics for the two vernaculars of medieval England.31 However, the problems I see with correlated dialectology presented under these terms are numerous; although they may not be adequately addressed here, some of the main theoretical difficulties and areas of future investigation My reference to Hunt’s criteria is, purposefully, to philological wisdom in practice (rather than in theoretical grammar). My two suggested additions are here substantiated with consistently diverging realisations by two different scribes of the same Insular French text, Guillaume le Clerc de Normandie’s Vie de Tobie, respectively Oxford, Jesus College, MS 29 (II) (fols. 195va–198rb) and Oxford, Bodleian Library, MS Rawlinson Poetry 234 (fols. 1ra–9rb), discussed briefly below. 30 In this respect, see for instance the useful work of taxonomy performed by Rothwell (1997) in his analysis of the forms of languages used in the York Memorandum Book (with entries 1376 x 1493). 31 For a recent, lucid presentation of the main issues at stake in areal linguistics, see Campbell (1998, 2nd ed. 2004: 330–43). 29

114

AN-1a.indd 124

11/03/2010 12:00:22

Mapping Insular French Texts? could be identified. First and foremost, correlating Middle English and Insular French data cannot easily solve difficult questions inherent in the distinct analysis of either language; but in the long run cross-perspectives on the local realisations of either may serve to shed new light on some theory-related or corpus-related limitations of existing approaches on the ‘French of England’ as well as on Early and Late Middle English. To limit discussion here to just some major crucial impediments, I would suggest that two inherent obstacles remain in both LALME and LAEME bibliographical and taxonomic procedures. In LALME, the Repertory List of manuscripts (as well as its reflex, the County List) does not take issues of contents into account, so that scribal contexts of multilingualism, wherever they exist, go unnoticed. Moreover, the choice of terms and features in the subsequent establishment of the questionnaires and linguistic profiles serves to process data of the English language that reflect a conception of English as looked at from its purely Germanic components, with their emphasis on specific grammatical categories such as pronouns or adjuncts, or on some incontrovertibly Germanic lexical words. Even in the area of spelling variants, the potential influence of (Insular) French analogical forms is thus not considered as a parameter – it is worth noting that, on the side of ‘Anglo-Norman studies’, Mildred K. Pope had taken them into account in her magisterial study (1934; revised 1952, re-issued 1966), and that line of pursuit has been impressively developed in the work of W. Rothwell. Likewise, the precious Catalogue for Sources of a Linguistic Atlas of Early Medieval English, published by Margaret Laing in 1993 as a precursor of LAEME, is essentially concerned with the preliminary localisation of English materials in manuscripts, but fails to offer systematically a complete list of contents, thus again disregarding some crucial scribal adjacencies with materials in matrix Insular French; nor does the slant towards quintessentially Germanic etymons and lexemes seem to be redressed (in Laing and Lass 2006, for instance). Several multilingual and ‘multi-generic’ manuscripts (irrespective of whether they are ‘pure’ or mixed genre: for example religious and scientific and medical contents in Oxford, Bodleian Library, MS Digby 2) do, however, constitute a precious first-hand corpus that is largely unstudied from this perspective. Some of them could serve as complex ‘anchor’ texts to start with, and that an appropriately revised ‘fit technique’ could thereafter serve purposes of localisation, in the treatment of apparently ‘unlocalisable’ texts in matrix Insular French. As to those first categories of manuscripts, three possibilities are considered and presented below. They have three common denominators: they are inspected regardless of whether they are predominantly or exclusively ‘literary’ or ‘documentary’ in content;32 they take distinctive account It is assumed that the ratio of circulation of ‘administrative’ documents of most sorts squares, or at least is not at odds, with that of ‘literary’ ones: for instance, the unique or multiple copies of some post-conquest Latin/vernacular charters spanning a generous timeline, as assembled by Pelteret (1990) can be taken to compare with the

32

115

AN-1a.indd 125

11/03/2010 12:00:22

The Anglo-Norman Language and Its Contexts of both matrix and significant non-matrix realisations of Insular French and English; and more importantly, they have their origin, or at least provenance, established or sufficiently clarified from certain non-linguistic grounds (facts related to book production at large, including principally but not exhaustively binding, fascicular layout and quiring system, evidence of medieval pressmarks or other non-onomastic book-owning signs; textual layout, formal characterisation of script and ductus, decorative style; exact situational data such as names, places, dates in margins, rubrics or colophons) which may permit dating or datability, and/or location or localisation; or also on dialectal grounds wherever there are occurrences of English as a matrix language or as non-matrix materials being significantly represented: A. Monolingual manuscripts containing only Insular French, or multi­ lingual manuscripts with any of Latin, English and French as matrix languages, of which at least a significant portion in Insular French is autograph, or demonstrably at just one or two removes from the autograph copy, and with or without a proportion of non-matrix materials in either language or both languages inserted in any of the matrix languages, whether in the same autograph hand(s) or in others (sharing a sufficient number of similarities in character): e.g. Paris, Bibliothèque nationale de France, fr. 24766, containing in particular the autograph copy of the work of an Osney canon, Angier of St Frideswide’s Dialogues de Saint Gregoire (finished in 1213).33 B. Multilingual manuscripts with one and the same scribe having copied at least ten lines each of French and English matrix materials, with or without any proportion of non-matrix materials in either language or both languages (or in Latin) inserted in any of the matrix languages, whether in the same hand(s) or in others (sharing a sufficient number of similarities in character): e.g. Cambridge, University Library, MS Gg.I.1, Trinity College, MS B.14.39, London, British Library, MSS Arundel 292, MS Harley 2253, Oxford, Bodleian Library, MSS Digby 2, Digby 53, Digby 86, Oxford, Jesus College MS 29 (II).34 C. Monolingual manuscripts containing only Insular French as the matrix language, or bilingual manuscripts without matrix English, i.e. containing Latin and Insular French as matrix languages, with or without any proportion of non-matrix materials in Latin, French, English or all, inserted in or adjacent to either matrix language, whether in the same hand(s) or in others (sharing, however, a sufficient number of similarities in character): e.g. Cambridge, Trinity College, MS R. 16. 2 (the Trinity Apocalypse);35 or copies of the ‘Rules of dissemination of unique or multiple copies of most ‘non-documentary’ manuscripts reported in Dean (1999). 33 Ker (1964: 140). On Angier’s work and the manuscript, see Dean (1999), entries 512 (and 513, 838.1, 870). 34 For reasons of space, no attempt here is made at bibliographical referencing for each manuscript; this would naturally have to be done in subsequent studies. Relevant references for some of them, however, are found passim in this article; for some, issues of origin and/or provenance have been dealt with extensively. 35 See recently McKitterick et al. (2005).

116

AN-1a.indd 126

11/03/2010 12:00:22

Mapping Insular French Texts? St Augustine’ tailored to specific Augustinian houses: Cambridge, Trinity College, MS B. 1. 45 (fols. 19v–24), the whole manuscript associated with Augustinian canonesses at Canonsleigh, Devon (end of 13th c.), with the English elements in its booklet 1 being in the hand of scribe D of the English Ancrene Riwle in London, British Library, MS Cotton Cleopatra C VI; and Oxford, Bodleian Library, MS Bodley 57 (fols. 2r–4r), the whole manuscript perhaps associated with the Augustinian canons of St-Mary-in-the-Fields, Leicester, Leicestershire, or at least ‘made locally’.36 These three suggested possibilities of complex ‘anchoring’ are classified by a decreasing degree of (provisionally) estimated preference. The issue of dating/ datability may require a word of explanation. As is well known, essential starting points to assess dated and datable manuscripts are obviously the English ‘Dated and Datable Manuscripts’ projects, significant branches of which have been completed by A. G. Watson (1979, 1984) and P. M. Robinson (1988, 2003); those catalogues include a non-negligible number of manuscripts containing Insular French texts, either matrix or non-matrix. It must be recalled, however, that their rationale excludes most business documents, making it therefore an object of further research to investigate contiguity or identity in ‘literary’ and ‘business’ hands, and the materials they copy; a relaxation of that principle (in favour of considering more business documents) is, however, observed (by Hanna 2005b) in the latest catalogue by Robinson covering the disparity of London libraries. Obviously, the first category is the most desirable one, but unfortunately the rarest;37 the third category, more than the first, confronts us with the pitfalls of equating origin and provenance. The choice of the second category requires slightly more discussion. It is thought to be especially desirable, on account of the inherently greater prospects of linguistic localisability of the English materials in manuscripts, and the subsequently higher confidence rate of localisability of shared Insular French contents; but it may raise serious objections, pertaining to just how the ‘anchor text’ technique can be shown to work also for Insular French material; and to what extent an ‘anchor’ location of Insular French materials can be assumed from a confident ‘anchor’ location of Middle English materials. One such objection is that matrix materials in English and French copied by the same hand are likely to reflect predominantly an influence of English scribal practices even when copying French, therefore potentially skewing the parameters of recognition of matrix Insular French per se. It is likely, moreover, that such a slant in favour of prevalent English scribal habitus would increase as we move towards the later medieval period, because it is (presumably On these, see the presentation and edition by T. Hunt (whom I thank for the gift of off-prints), respectively Hunt (1994) and Hunt (1995). On Bodley 57’s association with Leicester, or at least its Leicester locality, see Webber (2006: 145) and Gullick and Webber (2006: 190–1). 37 The one extant copy of Rossignos, Cambridge, Corpus Christi College, MS 471, is not an autograph manuscript, pace Sinclair (1965: 141–2): see now Hesketh (ed.) (2006). 36

117

AN-1a.indd 127

11/03/2010 12:00:22

The Anglo-Norman Language and Its Contexts rightly) assumed that in time Insular French was less and less a living, multidimensional vernacular, in the sense that it became more and more globally confined to the written sort of transmission only. But in addition to this realistic scribal bias being already built in within the proposed parameters of investigation, deliberate relegation of such adjacencies in precise scribal contexts may obfuscate one of the vital dimensions of Insular French, namely that of a language in contact, not just lexically, as has been amply documented and theorised in the work of Tony Hunt, William Rothwell and David Trotter, but also morpho-graphemically, i.e. also in terms of the surface graphical realisation of the ‘sound–symbol correspondence’ of French as influenced by English – consider for instance the identity of spelling, i.e. ‘sound–symbol’ realisation, of the adjective by the same scribe transcribing French and English texts in MS Jesus college 29 (II): English at the end of the Peines d’enfer (fol. 200v), and French at the beginning of the Doctrinal Sauvage (fol. 201r). For all its difficulty, the systematic identification and registering of such Englishinfluenced traits on the transcriptional techniques of the French of England would precisely serve the cause of an immense task in areal linguistics, that of making an essential characteristic of Insular French palpable.38 In addition here, a ‘control category’ could be provided by manuscripts, with English as the matrix language, of established dialectal provenance, which incorporate a substantial number of French loanwords, representing a corpus of analogical forms comparable to materials in matrix Insular French suspected to share adjacent locality. Indeed the forms of non-matrix English or French words can be compared to those of the same words found in texts in (correspondingly) matrix French or matrix English which are localisable. To take a very small example: in his edition of Peter Langtoft’s Chronicle, and in his recent work, Jean-Claude Thiolier considers the text preserved in Oxford, Bodleian Library, MS Fairfax 24 (booklet 1) to be closest to the writer’s original, thus to be of non-northern origin, more precisely perhaps Lincolnshire or slightly further south, not on account of the forms of the French in it (they are not considered as discriminatory), but, classically, either on the linguistic-dialectal grounds of the English metrical songs inserted, or on the (codicological and palaeographical) presence of a tenso on the crusade of 1270, apparently written in collaboration between Henry of Lacy, earl of Lincoln (?1249–1311) and Walter of Bibbesworth, unique to that manuscript (fols. 19r–20v, i.e. the last item in quire 1).39 In Fairfax 24 one such English segment in one of the metrical songs

I thank D. Trotter for discussions implicitly forcing me to think hard (and here, far from conclusively, of course) about that point. 39 Thiolier (1989: 69–75); see also Thiolier (1993, 1998, 2005). The ultimately southern origin of Fairfax 24 (booklet 1) was demonstrated in the 1989 publication; its very probable association with Lincolnshire is deduced from the more recent work (1993 and 1998) of Professor Thiolier, whom I thank for the gift of off-prints of the three latest publications mentioned. On Fairfax 24 more generally, see also Pouzet (2009b). 38

118

AN-1a.indd 128

11/03/2010 12:00:23

Mapping Insular French Texts? includes the phrase ,40 a collocation and a spelling which, for instance, compare with , found in the Middle English poem Genesis and Exodus, known to be extant only in Cambridge, Corpus Christi College, MS 444, a manuscript assigned a location in west Norfolk by M. Laing – one that corroborates a southern origin for this first booklet of Fairfax 24, therefore presumably not just for the non-matrix English texts, but also for the matrix French in it. Genesis and Exodus in turn includes many Insular French etymons grafted onto its English matrix structure, and even, en français dans le texte, ;41 all the spellings they take would undoubtedly reward comparative scrutiny with other localisable scribal realisations.42 One feature which may further contribute to an areal perspective on dialectal studies of Middle English and Insular French is the phenomenon of Middle English words with local distribution, such as have been sketched for the two west Yorkshire manuscripts of the Cursor Mundi (Edinburgh, Royal College of Physicians, and London, British Library, MS Cotton Vespasian A iii) both datable to c. 1340; those lexemes include etymons or whole loanwords of Insular French provenance.43 But a notion of ‘Anglo-Norman word-geography’ can be helpful there only if lexemes assumed to have exclusive regional distribution also pass a negative test of comparison (‘no hits’) with well-located documentary evidence from elsewhere. If they do, then their occurrence might be established as a distinctive factor of localisation, at the level of lexis; if they do not and are checked by attestation elsewhere, then a study of the respective spellings of occurrences might tell us something about the circumstances of their local realisations.44 On all counts, it is therefore desirable that, as for LAEME, all the phonemic traits (potestates) of a given form of a lexeme (littera) should be encoded in correspondence with their scribal realisations (figurae) witnessed in manuscripts containing Insular French. Another objection (only lightly touched on here) is that, even if this was achieved with confident localisation credentials, we are not safe from the possibility that a French scribal text transmitted in a given manuscript may be multi-layered in its own respects, just as English materials are. The common scribe of Harley 273, Harley 2253 and Royal 12.C.xii may have tapped into as varied regional sources for his French texts as for his English ones. That phrase features as an editorial variant duly reported by Thiolier (1989: 300). On some of the linguistic and literary/cultural implications of the segment, see Pouzet (2005). 42 Here for instance, has a spelling for which may reflect the scribal realisation of the velar glide /j/ + rounded /u/, and contrast with other perfectly imaginable ‘Anglo-Norman’ spellings and sound-symbol correspondences for the same lexeme, indeed ‘attested’ by taxonomic sanction: see AND printed and online. 43 On the Edinburgh manuscript, see in particular Hörning (1906) and Ross (1971); on the Cotton manuscript, see Barth (1903) and Shinoda (2003) (Shinoda does not refer to Barth). 44 In this respect, it would be fruitful, for instance, to place side by side such studies as Bately (1988) and Trotter (2003b). 40 41

119

AN-1a.indd 129

11/03/2010 12:00:23

The Anglo-Norman Language and Its Contexts Co-incidentally, if the ‘localisability hypothesis’ of Insular French materials is predicated on their geographical and dialectal variability, are we not going to end up with too many distributional maps not only rubbing shoulders but eventually collapsing into non-legibility and global undecidability? It may well be that Insular French differences from region to region were never so wild as to approximate to what we understand by ‘dialect’ in the Middle English sense, and we are aware of this factor. But even so, it is still thought worth trying to check whether it was not just that Higden–Trevisa, and thereafter the philological tradition in their majority, have simply never been attuned to the thinner and sparser, yet perhaps nonetheless significant, grains of difference. After such a brief overview of ideas, problems and prospects, it is hoped that the need to address seriously the ‘situatedness’ of Insular French materials in an areal perspective, in particular through more sustained developments in the consequences of the morphological and phonological overlap of located varieties of Middle English with Insular French, may emerge as timely and crucial, in tandem with continued attention given to Insular French lexis and syntax.45 One of the virtues this may have is to serve to further dispel ‘the myth of the Anglo-Norman scribe’ so expertly and gallantly resisted by the late Cecily Clark.46 Naturally, not all ideas have been rehearsed here, and collaborative perspectives (desired) would undoubtedly broaden the scope of imagination and rigour, while probably bringing a salutary degree of caution to the enterprise. A significant resource is represented by place-names with Insular French ingredients in them – the general resilience of local toponymy predictably lending support to the ‘anchor’ hypothesis once more extended from English forms to realisations of French. Further insightful theoretical and pragmatic benefits may be derived from linguistic and neurological investigation into L2 learning (on which see, notably, Myles and Towell 2004), the assumption being that much in the acquisition and practice of L2 French by native English speakers may correspond to some of the cognitive processes whereby scribal inscription rendered itself palpable in manuscripts of medieval England. This possibility may be assessed alongside sociological and anthropological approaches concerned with defining ‘local’ forms of literacy (see for instance Barton and Hamilton 1998, re-issued 2000). To return finally to a heuristic dimension – to repeat a salient feature in Laing and Lass’s account – that such a project may bring to manuscript studies per se, it may be suggested that, failing a substantial amount of autograph manuscripts with matrix French, any establishment of ‘anchor texts’ of Insular French may be profitably predicated on the recognition of identical hands An illustration of such areal perspectives on English is provided notably by Benskin (1994) and Pilch (1997), the latter indirectly taking French phonemic elements into account. 46 Clark (1992a: 126 n. 1) explains that this study expands on certain arguments presented in Clark (1987), and constitutes the third panel of a ‘trilogy’ commenced with Clark (1991) and finished with Clark (1992b). 45

120

AN-1a.indd 130

11/03/2010 12:00:23

Mapping Insular French Texts? copying Latin, English and French materials in multilingual manuscripts. In reciprocity with palaeography, this may also serve to contribute additions, or complementary information, to a list of identifiable scribes, specially (but not exclusively) those whose hand is located in more than one manuscript, notably by extending its scope from Latin or Middle English, to both Middle English and Insular French in particular.47 For instance, we could obtain a conjoined scribal profile of the English and French passages of the Jesus 29 (II) scribe, or of all the passages of scribe B of MSS Harley 2253, Royal 12.C.xii, and Harley 273. Conversely, extant copies of the same text copied by different scribes could furnish a corpus to test the relative stability/variability of Insular French sound–symbol realisations: as briefly suggested above, a good case in point could be the Vie de Tobie composed by Guillaume le Clerc de Normandie for another William, prior of Kenilworth (Warwickshire) in the first third of the thirteenth century, known to be extant in three insular manuscripts later in this century: Oxford, Jesus College 29 (II) (fols. 195va–198rb; the scribe also copied English texts), Bodleian Library, MS Rawlinson Poetry 234 (fols. 1ra–9rb), and Paris, Bibliothèque nationale de France, fr. 19525 (fols. 129r–132r).48 Examples of scribes of Insular French materials newly identifiable not just by linguistic profile but also by name may include such individuals as Nicholas ‘Brounfeld/ Brownfeld’, who copied at least the first quire (and perhaps more) of MS Fairfax 24 discussed above (signing his name on fol. 20v). But in fine, one may still wish to ask, ‘what is correlated dialectology’, or what may it hope to achieve? In ‘What is comparative literature?’, George Steiner (1995: 1) writes that ‘every act of the reception of significant form, in language, in art, in music, is comparative’, in the sense that ‘interpretation and aesthetic judgement [. . .] arise from an echo-chamber of historical, social, technical presuppositions and recognisance’, and that under those terms and conditions, ‘a certain contract of eventual decipherment, or informed evaluation underwrites the encounter between our sensibility and the text or work of art.’ In essence, the correlated approach defended here is and aims to be comparative in an areal perspective; such an act of reconstructive reception also probes our notions of what ‘Anglo-Norman’ has been thought to be, a broadly homogeneous language not liable to regional variability. This approach promises an ‘eventual decipherment’ of dispersed ‘significant forms’ of Insular French.

Alongside a wealth of important data in various monographic studies, see in particular the concerted work of Gullick (1998), and Mooney (2000) (continuing the achievements of the late J. Griffiths). 48 The Vie de Tobie was edited by Reinsch (1879); see also Dean (1935). 47

121

AN-1a.indd 131

11/03/2010 12:00:23

The Anglo-Norman Language and Its Contexts Primary sources manuscripts Cambridge University Library, MS Gg. I. 1 Trinity College, MS B. 1. 45 Trinity College, MS B. 14. 39 Trinity College, MS R. 16. 2 Edinburgh, Royal College of Physicians London, British Library MS Arundel 292 MS Cotton Vespasian A. iii MS Egerton 2710 MS Harley 273 MS Harley 2253 MS Harley 5234 MS Royal 12. C. xii Oxford, Bodleian Library MS Bodley 57 MS Digby 2 MS Digby 53 MS Digby 86 MS Rawlinson Poetry 234 Oxford, Jesus College, MS 29 (II) Paris, Bibliothèque nationale de France fr. 19525 fr. 24766 nouv. acq. fr. 4503

printed sources Babington, C. (ed.) (1869), Polychronicon Ranulphi Higden Monachi Cestrensis; Together with the English Translations of John Trevisa and of an Unknown Writer of the Fifteenth Century, vol. II. Rolls Series. London. Chaucer, Geoffrey (1988), Canterbury Tales, ed. L. Benson, in The Riverside Chaucer, 3rd ed. Boston. Hesketh, G. (ed.) (2006), Rossignos by John of Howden, ed. G. Hesketh. Anglo-Norman Texts 63. London. Hunt, T. (ed.) (1991), Teaching and Learning Latin in Thirteenth-Century England, 3 vols. Cambridge. ——(ed.) (1994), ‘An Anglo-Norman treatise on the religious life’, in P. Monks and D. Owen (eds.), Medieval Codicology, Iconography, Literature, and Translation: Studies for Keith Val Sinclair, Litterae Textuales, Leiden, New York, Köln, pp. 267–75. ——(ed.) (1995), ‘An Anglo-Norman Rule of St. Augustine’, Augustiniana 45, 177–89. —— and M. Benskin (ed.) (2001), Three Receptaria from Medieval England: The Languages of Medicine in the Fourteenth Century. Oxford. ——(ed.) (2004), Le chant des chanz. Anglo-Norman Texts 61–2. London.

122

AN-1a.indd 132

11/03/2010 12:00:23

Mapping Insular French Texts? McKitterick, D., N. Morgan, I. Short and T. Webber (eds.) (2005), The Trinity Apocalypse (Trinity College Cambridge, MS R.16.2). London, Toronto and Buffalo. Millward, C. (ed.) (1998), La Estorie del Evangelie: A Parallel-Text Edition. Middle English Texts 30. Heidelberg. Parkes, M., and J. Tschann (eds. and intr.) (1996), Facsimile of Oxford, Bodleian Library, MS Digby 86. EETS, ss 16. Oxford. Prior, O., et al. (eds.) (1924), Cambridge Anglo-Norman Texts. Cambridge. Reinsch, R. (ed.) (1879), ‘La vie de Tobie de Guillaume le Clerc de Normandie. Nach der Pariser und Oxforder Hs. herausgegeben und mit einer Einleitung versehen’, Archiv 62, 375–96. Ross, C. (ed.) (1971), ‘An Edition of Part of the Edinburgh Fragment of the Cursor Mundi’, B.Litt. thesis, University of Oxford. Spiele, I. (ed.) (1975), Li Romanz de Dieu et de sa mere d’Herman de Valenciennes, chanoine et prêtre (XIIe siècle). Leiden.

Bibliography Baker, J. (1979), Manual of Law French. Avebury. Barth, C. (1903), Der Wortschatz des Cursor Mundi. Ein Beitrag zur Kenntnis der mittelenglischen Dialekte. Königsberg. Barton, D., and M. Hamilton (1998, re-issued 2000), Local Literacies: Reading and Writing in One Community. London. Bately, J. (1988), ‘On some aspects of the vocabulary of the West Midlands in the early Middle Ages: the language of the Katherine group’, in E. Kennedy, R. Waldron and J. Wittig (eds.), Medieval English Studies Presented to George Kane, Cambridge, pp. 55–77. Beadle, R. (1991), ‘Prolegomena to a literary geography of later medieval Norfolk’, in F. Riddy (ed.), Regionalism in Late Medieval Manuscripts and Texts, Cambridge, pp. 89–108. —— (1994), ‘Middle English texts and their transmission, 1350–1500: some geographical criteria’, in Laing and Williamson (1994), pp. 69–91. Bennett, M. J. (2006), ‘Mandeville’s Travels and the Anglo-French Moment’, Medium Ævum 75, 273–92. Benskin, M. (1977), ‘Local archives and Middle English dialects’, Journal of the Society of Archivists 5/8, 500–14. ——(1994), ‘Description of dialect and areal distributions’, in Laing and Williamson (1994), pp. 169–87. ——and M. Samuels (eds.) (1981), ’So meny people longages and tonges’: Philological Essays in Scots and Mediaeval English Presented to Angus McIntosh. Edinburgh, 1981. ——and M. Laing (1981), ‘Translations and Mischsprachen in Middle English manuscripts’, in Benskin and Samuels (1981), pp. 55–106. Berndt, R. (1976), ‘The period of the final decline of French in medieval England (fourteenth and early fifteenth centuries)’, Zeitschrift für Anglistik und Amerikanistik 20, 341–69. Blake, N. (ed.) (1992), The Cambridge History of the English Language, II: 1066–1476. Cambridge.

123

AN-1a.indd 133

11/03/2010 12:00:23

The Anglo-Norman Language and Its Contexts Boffey, J. (2000), ‘Prospecting in the archives: Middle English verse in record repositories’, in D. Pearsall (ed.), New Directions in Later Medieval Manuscript Studies: Essays from the 1998 Harvard Conference, York, pp. 41–51. Busby, K. (2002), Codex and Context. Reading Old French Verse Narrative in Manuscript, 2 vols. Amsterdam. Campbell, L. (1998, 2nd ed. 2004), Historical Linguistics: An Introduction. Edinburgh. Careri, M., C. Ruby and I. Short (forthcoming), Illustrated Catalogue of Twelfth-Century French and Occitan Literary Manuscripts. Rome. Presentation at http://www. irht.cnrs.fr/recherche/programme_catlitfr_english.htm. Clanchy, M. (1993), From Memory to Written Record, England 1066–1307. Oxford. Clark, C. (1987), ‘Spelling and grammaticality in the Vespasian Homilies: a reassessment’, Manuscripta 31, 7–10. —— (1991), ‘Towards a reassessment of Anglo-Norman Influence on English PlaceNames’, in G. Broderick and S. Ureland (eds.), Language Contact in the British Isles: Eighth International Symposium on Language Contact in Europe, Tübingen, pp. 275–93. ——(1992a), ‘The myth of the Anglo-Norman scribe’, in M. Rissanen, O. Ihalainen, T. Nevalainen and I. Taavitsainen (eds.), History of Englishes. New Methods and Interpretations in Historical Linguistics, Berlin and New York, pp. 117–29. —— (1992b), ‘Domesday Book – a great red-herring: thoughts on some eleventhcentury orthographies’, in C. Hicks (ed.), England in the Eleventh Century: Proceedings of the 1990 Harlaxton Symposium, Stamford, pp. 323–37. Connolly, M. (1998), John Shirley: Book Production and the Noble Household in FifteenthCentury England. Aldershot. Crépin, A. (2004), ‘Quand les anglais parlaient français’, Comptes rendus de l’Académie des inscriptions et belles-lettres (4th fascicule), 3–22. Davis, G., and C. Davis (1958), Medieval Cartularies of Great Britain: A Short Catalogue. London. Dean, R. (1935), ‘A missing chapter of the Vie de Tobie’, Modern Philology 33, 13–19. —— (1972), ‘The fair field of Anglo-Norman: recent cultivation’, Medievalia et Humanistica 3, 279–97. ——, with M. Boulton (1999), Anglo-Norman literature. A Guide to Texts and Manuscripts. London. Dees, A. et al. (1987), Atlas des formes linguistiques des textes littéraires de l’ancien français. Tübingen. Fisher, J. (1994), ‘Historical and methodological considerations for adopting ‘best text’ or ‘usus scribendi’ for textual criticism of Chaucer’s poems’, TEXT 6, 165–80. Fleischmann, S. (1996), ‘Medieval vernaculars and the myth of monoglossia: a conspiracy of linguistics and philology’, in S. Lerer (ed.), Literary History and the Challenge of Philology: The Legacy of Erich Auerbach, Stanford, pp. 92–104. Gillespie, A., and D. Wakelin (eds.) (forthcoming), The Production of Books in England c. 1350–c. 1530. Cambridge. Greetham, D. (1996), ‘Phylum – tree – rhizome’, Huntington Library Quarterly 58, 99–126. Griffiths, J., and D. Pearsall (eds.) (1989), Book Production and Publishing in Britain, 1375–1475. Cambridge. Gullick, M. (1998), ‘Professional scribes in eleventh- and twelfth-century England’, English Manuscript Studies 1100–1700 7, 1–24.

124

AN-1a.indd 134

11/03/2010 12:00:23

Mapping Insular French Texts? Gullick, M., and T. Webber (2006), ‘Summary catalogue of surviving manuscripts from Leicester Abbey’, in J. Story, J. Bourne and R. Buckley (eds.), Leicester Abbey: Medieval History, Archaeology and Manuscript Studies, Leicester, pp. 173–92. de Haas, E., and G. Hall (eds.) (1970), Early Registers of Writs. London. Hanna, R. (1996a), Pursuing History: Middle English Manuscripts and Their Texts. Stanford. ——(1996b), ‘Miscellaneity and vernacularity: conditions of literary production in late medieval England’, in S. Nichols and S. Wenzel (eds.), The Whole Book: Cultural Perspectives on the Medieval Miscellany, Ann Arbor, pp. 37–51. ——(2001), ‘Analytical survey 4: Middle English manuscripts and the study of literature’, in W. Scase, R. Copeland and D. Lawton (eds.), New Medieval Literatures 4, Oxford, pp. 243–64. ——(2003), ‘Yorkshire writers’, Proceedings of the British Academy 121, 91–109. —— (2004), ‘Middle English books and Middle English literary history’, Modern Philology 102, 157–78. ——(2005a), London Literature, 1300–1380. Cambridge. ——(2005b), Review of Robinson (2003), Medium Ævum 74, 122–3. ——(2008), ‘Some North Yorkshire scribes and their context’, in D. Renevey and G. Caie (eds.), Medieval Texts in Context, London, pp. 167–91. Harper-Bill, C., and E. van Houts (eds.) (2003), A Companion to the Anglo-Norman World. Woodbridge. Harris, K. (1983), ‘John Gower’s Confessio Amantis: the virtues of bad texts’, in D. Pearsall (ed.), Manuscripts and Readers in Fifteenth-Century England: the Literary Implications of Manuscript Study, Cambridge, pp. 27–40. Hörning, W. (1906), Die Schreibung des Hs. E des Cursor Mundi. Berlin. Horobin, S. (forthcoming), ‘Mapping script and word’, in Gillespie and Wakelin (forthcoming). Kaiser, R. (1937), Zur Geographie des mittelenglischen Wortschatzes. Palestra 205. Leipzig. Ker, N. (1964), Medieval Libraries of Great Britain: A List of Surviving Books. London. Kristensson, G. (1967–2002), A Survey of Middle English Dialects 1290–1350, 5 vols. Lund. Laing, M. (ed.) (1989), Middle English Dialectology: Essays on Some Principles and Problems. Aberdeen. —— (1993), Catalogue for Sources of a Linguistic Atlas of Early Medieval English. Cambridge. —— (1999), ‘Confusion wrs confounded: litteral substitution sets in Early Middle English writing systems’, Neuphilologische Mitteilungen 100, 251–70. —— and R. Lass (2006), ‘Early Middle English dialectology: problems and prospects’, in A. van Kemenade and B. Los (eds.), The Handbook of the History of English, Oxford, pp. 417–51. ——and A. McIntosh (1995), ‘Cambridge, Trinity College, MS 335: its texts and their transmission’, in R. Beadle and A. Piper (eds.), New Science out of Old Books: Studies in Manuscripts and Early Printed Books in Honour of A. I. Doyle, Aldershot, pp. 14–47. ——and K. Williamson (eds.) (1994), Speaking in Our Tongues: Medieval Dialectology and Related Disciplines. Cambridge. Lavezzo, K. (ed.) (2004), Imagining a Medieval English Nation. Minneapolis and London.

125

AN-1a.indd 135

11/03/2010 12:00:23

The Anglo-Norman Language and Its Contexts Lawrence-Mathers, A. (1995, 2003), ‘A northern English school? Patterns of production and collection of manuscripts in the Augustinian houses of Yorkshire in the twelfth and thirteenth centuries’, in Yorkshire Monasticism, British Archaeological Association Conference Transactions 16, pp. 145–53; revised as chapter 8, ‘The Augustinians and their Libraries’, of her Manuscripts in Northumbria in the Eleventh and Twelfth Centuries, Cambridge, pp. 177–93. Legge, M.-D. (1963), Anglo-Norman Literature and Its Background. Oxford. —— (1965), ‘La précocité de la littérature anglo-normande’, Cahiers de civilisation médiévale 8, 327–49. ——(1967), ‘Les origines de l’anglo-normand littéraire’, Revue de linguistique romane 31, 44–54. Machan, T. (2003), English in the Middle Ages. Oxford. Matheson, L. (1998), The Prose Brut: The Development of a Middle English Chronicle. Tempe. McCarren, V., and D. Moffat (eds.) (1998), A Guide to Editing Middle English. Ann Arbor. McGann, J. (1983, repr. 1992, re-issued 1996), A Critique of Modern Textual Criticism. Originally published Chicago, reprinted Charlottesville and London. McIntosh, A. (1956, 1989), ‘The analysis of written Middle English’, Transactions of the Philological Society pp. 26–55; reprinted in M. Laing (ed.), Middle English Dialectology: Essays on Some Principles and Problems, Aberdeen, pp. 1–21. ——(1963), ‘A new approach to Middle English dialectology’, English Studies 44, 1–11. ——(1987), ‘The Middle English Estorie del Euangelie: the dialect of the original version’, Neuphilologische Mitteilungen 88, 186–91. ——, M. Samuels and M. Benskin, with M. Laing and K. Williamson (eds.) (1986), A Linguistic Atlas of Late Mediaeval English, 4 vols. Aberdeen. Menger, L. (1904), The Anglo-Norman Dialect: A Manual of Its Phonology and Morpho­ logy, with Illustrative Specimens of the Literature. New York. Millward, C. (1997), ‘The medieval scribe as editor: the case of La Estorie del Evangelie’, Manuscripta 41,155–70. Mooney, L. (2000), ‘Professional scribes? Identifying English scribes who had a hand in more than one manuscript’, in D. Pearsall (ed.), New Directions in Later Medieval Manuscript Studies: Essays from the 1998 Harvard Conference, Woodbridge, pp. 131–41. Moore, S., S. Meech and H. Whitehall (1935), Middle English Dialect Characteristics and Bialect Boundaries. Essays and Studies in English and Comparative Literature 13. Ann Arbor. Mugglestone, L. (ed.) (2006), History of the English Language. Oxford. Myles, F., and R. Towell (eds.) (2004), Journal of French Language Studies 14/3. Special Issue on the acquisition of French as a second language. Pelteret, D. (1990), Catalogue of English Post-Conquest Vernacular Documents. Woodbridge. Pickering, O. (1972), ‘A newe lessoun off Chrystys ressurrectoun: a second manuscript, and its use of La estorie del evangelie’, English Philological Studies 13, 43–8. Pilch, H. (1997), ‘Middle English phonetics: a systematic survey including notes on Irish and Welsh loanwords’, in J. Fisiak (ed.), Studies in Middle English Linguistics, Berlin, pp. 439–67. Pope, M. (1934), From Latin to Modern French with Especial Consideration of AngloNorman: Phonology and Morphology. Manchester.

126

AN-1a.indd 136

11/03/2010 12:00:23

Mapping Insular French Texts? Pouzet, J.-P. (2005), ‘“un iur natural”: réflexions autour du vers 82 de Genesis and Exodus’, in C. Stévanovitch (ed.), L’articulation langue – littérature dans les textes médiévaux anglais, Nancy, pp. 277–95. ——(2009a), ‘Lieux et présence de l’anglo-français insulaire dans l’écriture religieuse anglaise (fin XIIe – fin XIVe siècles)’, in A. Crépin and Y. Leclant (eds.), Actes de la Journée d’études anglo-normandes, Académie des inscriptions et belles-lettres. Paris, pp. 39–80. ——(2009b), ‘Augustinian canons and their insular French books in medieval England: towards an assessment’, in J. Wogan-Browne et al. (eds.), Language and Culture in Medieval Britain: The French of England c. 1100-c. 1500, York, pp. 266–77. Prior, O. (1923), ‘Remarques sur l’anglo-normand’, Romania 49, 161–85. Reichl, K. (1995), ‘“No more ne willi wiked be”: religious poetry in a Franciscan manuscript (Digby 2)’, in R. G. Newhauser and J. Alford (eds.), Literature and Religion in the Later Middle Ages: Philological Studies in Honor of Siegfried Wenzel, Binghamton, pp. 297–317. Rigg, A. (1992), History of Anglo-Latin Literature 1066–1422. Cambridge. Robinson, P. (1988), Catalogue of Dated and Datable Manuscripts in Cambridge Libraries c. 737–1600. Cambridge. ——(2003), Catalogue of Dated and Datable Manuscripts c. 888–1600 in London Libraries, 2 vols. London. Rothwell, W. (1992), ‘The French vocabulary in the archive of the London Grocers’ Company’, Zeitschrift für französische Sprache und Literatur 102, 23–41; now available on the Anglo-Norman On-Line Hub at http://www.anglo-norman.net/ articlesA/grocers.xml. —— (1997), ‘Aspects of lexical and morphosyntactical mixing in the languages of medieval England’, in S. Gregory and D. Trotter (eds.), De mot en mot: Aspects of Medieval Linguistics: Essays in Honour of William Rothwell, Cardiff, pp. 213–32. ——(2004), ‘Ignorant scribe and learned editor: patterns of textual error in editions of Anglo-French texts’, on the Anglo-Norman On-Line Hub at http://www. anglo-norman.net/articlesA/scribe.xml. Salter, E. (1983), Fourteenth-Century English Poetry: Contexts and Readings. Oxford. Samuels, M. (1963), ‘Some applications of Middle English dialectology’, English Studies 44, 81–94. Sato, S. (ed.) (1997), Back to the Manuscripts: Papers from the Symposium ‘The Integrated Approach to Manuscript Studies: A New Horizon’ Held at the Eighth General Meeting of the Japan Society for Medieval English Studies, Tokyo, December 1992. Tokyo. Scase, W. (ed.) (2007), Essays in Manuscript Geography: Vernacular Manuscripts of the English West Midlands from the Conquest to the Sixteenth Century. Turnhout. Schendl, H. (2002), ‘Mixed-language texts as data and evidence in English historical linguistics’, in D. Minkova and R. Stockwell (eds.), Studies in the History of the English Language: A Millennial Perspective, Berlin, pp. 51–78. Sharpe, R. (2001), Handlist of the Latin Writers of Great Britain and Ireland Before 1540. Turnhout. Shinoda, Y. (2003), ‘On rare northern words found in Cursor Mundi (Cotton Vespasian A iii British Library), with special reference to nouns in rime [sic]’, in A Pilgrimage through Medieval English Literature – Essays Presented to Professor Isamu Saito on the Occasion of his Sixtieth Birthday, Tokyo, pp. 201–26.

127

AN-1a.indd 137

11/03/2010 12:00:24

The Anglo-Norman Language and Its Contexts Short, I. (2007), Manual of Anglo-Norman. Anglo-Norman Text Society, Occasional Publications Series 7. London. Sinclair, K. (1965), ‘Anglo-Norman studies: the last twenty years’, Australian Journal of French Studies 2, 113–55 and 225–78. Steiner, G. (1995), ‘What is comparative literature? An inaugural lecture delivered before the University of Oxford on 11 October 1994’, Oxford. Taylor, A. (2002), Textual Situations: Three Medieval Manuscripts and Their Readers. Philadelphia. Thiolier, J.-C. (ed.) (1989), Edition critique et commentée de Pierre de Langtoft, Le règne d’Edouard Ier, vol. I. CELIMA (Université de Paris XII). Créteil. ——(1993), ‘Pierre de Langtoft: historiographe d’Edouard Ier Plantagenêt’, in I. Short (ed.), Anglo-Norman Anniversary Essays, Anglo-Norman Text Society, Occasional Publications Series 2, London, pp. 379–94. ——(1998), ‘L’itinéraire de Pierre de Langtoft’, in J.-C. Faucon et al. (eds.), Miscellanea Mediaevalia. Mélanges offerts à Philippe Ménard, vol. II, Paris, pp. 1329–53. ——(2005), ‘Interférences lexicales chez Pierre de Langtoft’, in D. Billy and A. Buckley (eds.), Études de langue et de littérature médiévales offertes à Peter T. Ricketts à l’occasion de son 70ème anniversaire, Turnhout, pp. 337–51. Thompson, J. (2000), ‘“Frankis rimes here I redd, / Communlik in ilk[a] sted . . .”: The French Bible stories in Harley 2253’, in S. Fein (ed.), Studies in the Harley Manuscript: The Scribes, Contents, and Social Contexts of British Library, MS Harley 2253, Kalamazoo, pp. 271–87. Trotter, D. (2003a), ‘Not as eccentric as it looks: Anglo-French and French French’, Forum for Modern Language Studies 39, 427–38. ——(2003b), ‘The Anglo-French lexis of Ancrene Wisse: a re-evaluation’, in Y. Wada (ed.), A Companion to Ancrene Wisse, Cambridge, pp. 83–101. Turville-Petre, T. (1983), ‘Some medieval English manuscripts in the north-east Midlands’, in D. Pearsall (ed.), Manuscripts and Readers in Fifteenth-Century England: The Literary Implications of Manuscript Study, Cambridge, pp. 125–41. ——(1996), England the Nation: Language, Literature, and National Identity, 1290–1340. Oxford. Vising, J. (1923), Anglo-Norman Language and Literature. London. Watson, A. (1979), Catalogue of Dated and Datable Manuscripts c. 700–1600 in the Department of Manuscripts, The British Library, 2 vols. London. ——(1984), Catalogue of Dated and Datable Manuscripts c. 435–1600 in Oxford Libraries, 2 vols. Oxford. Webber, T. (2006), ‘The books of Leicester Abbey’, in J. Story, J. Bourne and R. Buckley (eds.), Leicester Abbey: Medieval History, Archaeology and Manuscript Studies, Leicester, pp. 127–46. Wenzel, S. (1994), Macaronic Sermons: Bilingualism and Preaching in Late-Medieval England. Ann Arbor. Williams, D. (2004), The French Fetish from Chaucer to Shakespeare. Cambridge. Wright, L. (1997), ‘Medieval Latin, Anglo-Norman and Middle English in a civic London text: an inquisition of the River Thames, 1421’, in S. Gregory and D. A. Trotter (eds.), De mot en mot: Aspects of Medieval Linguistics: Essays in Honour of William Rothwell, Cardiff, pp. 223–60. ——(2000), ‘Bills, accounts, inventories: everyday trilingual activities in the business world of later medieval England’, in D. A. Trotter (ed.), Multilingualism in Later Medieval Britain, Cambridge, pp. 149–56.

128

AN-1a.indd 138

11/03/2010 12:00:24

Mapping Insular French Texts? Wright, L. (this volume), ‘A pilot study on the singular definite articles le and la in fifteenth-century London mixed-language business writing’.

129

AN-1a.indd 139

11/03/2010 12:00:24

chapter ten

A Pilot Study on the Singular Definite Articles le and la in Fifteenth-Century London Mixed-Language Business Writing Laura Wright

Between the Norman Conquest and the evolution of Standard English in the late fifteenth to early sixteenth centuries, a mixed-language system, incorporating Middle English into a matrix of Medieval Latin and/or Anglo-Norman, was used for the text type of accounts and inventories. The mixture of [Medieval Latin + English] and [Anglo-Norman + English] was ordered in a principled manner – although all of its complexities have yet to be understood. The system changed over time, as is to be expected, and the purpose of this paper is draw attention to the introduction and distribution of the definite article, expressed as le, la and lez, les, in the [Medieval Latin + English] variety (which, despite having Medieval Latin as a matrix, is nonetheless vastly informed by Anglo-Norman, particularly in the wordstock). I survey here documents dated between 1420 and 1460 in the London records of the Bridge House Estate, and confine my comments to that period. The principles that applied in the mixed-language business text type were informed by those that governed Middle English, Anglo-Norman and Medieval Latin, but the resulting linguistic code was governed by principles that were specific to itself (see Wright 1995, 1997a, 1997b n. 38, n. 39, 1998, 2000, 2001, 2002, 2005 for a more detailed discussion). They included: ▪ do not write in monolingual Medieval Latin or Anglo-Norman, but include English nouns, stems of verbs, adjectives and -ing forms, variably (e.g. Itm � soluť Rīco Plastrer ꝓ plastryng de leʒ ffrontes domoɚ apud le horn xxxvjs viijd, ‘And paid to Richard Plasterer for plastering of the house fronts at the Horn 36s 8d’, fol. 283, BHWP 3, 1429 x 1430, where x indicates text internally dated) ▪ calque Romance nouns with English nouns (e.g. the words gigantum and giant in the following two entries: Itm � ij ħoibʒ ꝓ custoɖ Gigantm � in aduenť Regꝭ viijd, ‘And to two men for guarding the giants at the coming of the king and queen 8d’, p. 462, BHWP 2, 1429 x 1430; Itm � ꝓ imposicōe des Pysens ꝓ les Geauntes ixd, ‘And for positioning of the pisanes for the giants 9d’, p. 465, BHWP 2, 1429 x 1430 (OED pisane, n., from post-classical Latin pisana, ‘a mail

130

AN-1a.indd 140

11/03/2010 12:00:24

A Pilot Study on the Singular Definite Articles collar forming part of a cape extending over the shoulders and upper part of the breast; such a cape itself’). ▪ use both Germanic and Romance word-orders (e.g. the ordering of colours after tartarin and buckram but before buckram, fringe and damask in the following two entries: Iť soł ꝓ Tarteryn rouge ynde & bloy & bokeram rouge & bloy empť, ‘And paid for red indigo and blue tartarin and red and blue buckram bought’, p. 466, BHWP2, 1429/30; Itm � soł ꝓ Oerfreys ꝓ j Chesiple cū iiij virg̉ de bloɖ bokeram � ac cū ryba�n ꝓ eoɖ & panno lineo ꝓ alba frenge & ał apparať cū fcūra vni9 vestimenti de viriɖ damask ꝓ Capella, ‘And paid for gold embroidery for 1 chasuble with 4 yards of blood buckram, and with ribbon for the same, and linen cloth for white fringe, and other material with the making of one vestment of green damask for the chapel’, p. 473, BHWP 2, 1430 x 1431) ▪ variably apply number concord within the noun phrase (e.g. ꝓ le gynnes and also ꝓ leʒ gynnes, ‘for the gins’, fol. 34v, BHA 3, 1460 x 1461) ▪ use a multiplicity of suffixes to indicate verbal nouns (e.g. ꝓ les weres destruenɖ, ‘for (the) destroying (of) the weirs’, ꝓ excambio vasoɚ veter̉ de peauter, ‘for (the) exchanging (of) old vessels of pewter’, ꝓ imposicōe j Exiltre, ‘for (the) positioning (of) 1 axletree’, cū fcūra biłł, ‘with (the) making (of) a bill’, all p. 495, ꝓ sowdyng inde, ‘for (the) soldering there’, p. 494, BHWP 2, 1421 x 1422) ▪ visually merge any material that can be merged with the abbreviation and suspension system, but do it variably (e.g. the Latin word carpentarius and the English word carpenter in the entry: Itm � Joħi Brys ał Carpenť ꝓ sepť iijs vjd, ‘And to John Brys another carpenter for the week 3s 6d’, p. 451, BHWP 2, 1429 x 1430) ▪ use le to qualify a vernacular noun (e.g.  ꝓ le shoying dicť  Rotaɚ xijd, ‘for the shoeing the said wheels 12d’, fol. 16, BHA 3, 1460 x 1461, from OE scōgan, OED shoe, v.) ▪ be categorical about variation. With hindsight, commentators have said that this is a system in decline, that scribes were no longer competent in Anglo-Norman or Medieval Latin and so could no longer maintain correct Romance morphology or wordstock (see Wright 1997a for a discussion). My view is that a system that was used in an orderly fashion for several generations cannot have been faulty; rather, it was simply different from its antecedents, and at different points in time its rules varied. Romance-language nouns were marked for grammatical gender (see Ingham (this volume) for a demonstration of the decline c. 1370 of Anglo-Norman noun-gender assignment in insular texts); English nouns, by the fifteenth century, were not. For articles, Medieval Latin and Anglo-Norman used le for masculine nouns, la for feminine nouns, and les or lez to mark plural nouns. How, then, was gender assigned in a Medieval Latin text which obligatorily contained much English vocabulary? The answer is that the default article for English words was le (sg.), les or lez (pl.). This matched the English system, which marked plurality but not gender, and the default singular marker le was the variant which most closely resembled the invariable English article the phonologically. In the [Medieval Latin + English] variety, the definite article le

131

AN-1a.indd 141

11/03/2010 12:00:24

The Anglo-Norman Language and Its Contexts was used to mark an English noun rather than a Latin one1 Over time, la also came to be used, as well as le + plural nouns and lez + singular nouns, although always with a lower frequency. Variation remained vital throughout. What follows is concerned with the introduction of la in the [Medieval Latin + English] variety, the when and the how, in the London Bridge House Estate records. It should be emphasised that this is a preliminary investigation into the distribution of definite-article forms, and that other archives may show a different patterning. In the Bridge House Account Roll of 1381 x 1382, the only nouns marked with le are leʒ Stockos (2 tokens), leʒ Stockes (52 tokens) (the Stocks was the name of a large food market at the eastern end of Cheapside by the present-day Mansion House), and one token each of le loke and leʒ Stathelynges (the Lock was the name of a meadow at Deptford, which is the sense used here, and was also a common noun signifying the area of riverwater underneath the arches of London Bridge, used elsewhere in this archive; the starlings were the outworks of wooden piles driven into the riverbed around the piers of London Bridge upon which the piers were built). Bridgemasters’ Account Roll 17, written in the year 1404 x 1405, shows an increase in nouns qualified by le, but they are still not numerous: le hemming (the hemming of cloth), les Pyles, le Milende (Mile End, a placename), le Ramme (part of a ramming machine), le Countynghous, le drawbrigg, le Raven in Chepe, le Crowne in Suthwerk (both tenement names), le latys pontꝭ v̉tibilis (the drawbridge lattice), le Pultrie (the Poultry, part of the Stocks market), le Brighous, le swerde (a part of a ramming machine), le ffother (a weight), les masons, le tontyght (a weight). Thereafter, however, the tokencount rises, reaching a distribution in the following decade which was to be held steady until the switchover to monolingual English, which in this archive happened in 1479 x 1480. To focus on prepositional phrases, the 1420s seem to have been a decade of innovation. At the start of the decade, le prevailed. Surveying the financial year 1420–1 in the Bridge House Weekly Payments, I noted no tokens of la at all. The definite article le occurs thirty-seven times in this year, qualifying place-names and proper names. The default construction in the fifteenth century in this text type when using an English noun in a prepositional phrase was to precede it with a Latin preposition, and to affix no postposed Latin suffix to the noun, as the examples below show, taken from the Bridge House Weekly Payments (unless otherwise stated):2

1



2



I am grateful to David Trotter for this observation: see his article in this volume for a discussion. The gin was a ramming machine. The Unicorn, Cock on the Hoop and the Horn were all tenement names, the Cock on the Hoop being a brewhouse (the phrase on the Hoop frequently signified a brewhouse or tavern at this date), and the Strand is a street.

132

AN-1a.indd 142

11/03/2010 12:00:24

A Pilot Study on the Singular Definite Articles 1. cū le vernisshing, ‘with the varnishing’ (p. 467, 1421 x 1422) ad le tileculn, ‘to the tile kiln’ (fol. 11v, 1421 x 1422) in le gardin, ‘in the garden’ (fol. 11v, 1421 x 1422) apɖ le vnicorn, ‘at the Unicorn’ (fol. 212, 1426 x 1427) ꝓ le Briggehous, ‘for the bridge house’ (fol. 210v, 1426 x 1427) subtus le scaffehald, ‘underneath the scaffold’ (fol. 198v, 1426 x 1427) le Cok su le hoop, ‘the Cock on the Hoop’ (fol. 304, 1429 x 1430) usqʒ le horn, ‘towards the Horn’ (fol. 294v, 1429 x 1430) de le Gyn, ‘from the gin’ (fol. 306v, 1429 x 1430) iuxta le Strande, ‘next to the Strand’ (fol. 4, 1460 x 1461, BHR3)

The prepositions ad and de could join with the following definite article le to form al, del (sg.) and des, dez (pl.), although the default remained de, ad + le(s): 2. de Bryggehous, ‘from (the) bridge house’ (fol. 213v, 1426 x 1427) de le briggehous, ‘from the bridge house’ (fol. 69v, 1423 x 1424) del briggehous, ‘from the bridge house’ (fol. 69v, 1423 x 1424) de lez dietz ffloodgates, ‘from the said floodgates’ (fol. 34, 1422 x 1423) des ffloodgates, ‘from the floodgates’ (fol. 33, 1422 x 1423) ad lez cabanes, ‘to the cabins’ (fol. 221v, 1426 x 1427) al briggehous, ‘to the bridge house’ (fol. 100, 1423 x 1424)

However not all scribes used the contracted forms, and scribes who did also used the uncontracted forms (e.g. fol. 69v). Similarly, use of the article itself was variable:3 3. ad rammam, ‘to the ram’ (fol. 30, 1422 x 1423) ad le Raam, ‘to the ram’ (fol. 296, 1429 x 1430) de huksterys, ‘of the hucksters’ (fol. 218, 1426 x 1427) de lez huksteres, ‘of the hucksters’ (fol. 275, 1429 x 1430)

We can conclude that English nouns could be preceded by the definite article le, which then blocked a following Latin suffix on the noun. However, prepositional constructions without le did take Latin suffixes: contrast the word wharf as preceded by apud, pro and supra alone, with the word wharf when preceded by le:4 4. apɖ wherfam de Briggehous, ‘at the Bridge House wharf’ (fol. 210, 1426 x 1427) ꝓ wharfagio lyggyng & lathyng maeremis de Rundale, ‘for wharfage, laying and

3



4

The ram was part of the ramming machine (OE ram(m, OED ram n.1, Wright (1996: 38), and hucksters were peddlers (origin obscure, possible cognates in Middle Dutch, although the English word is attested earlier). Note in the fourth example the spelling typical of Scottish writing, but occcasionally found in London writing also.

133

AN-1a.indd 143

11/03/2010 12:00:24

The Anglo-Norman Language and Its Contexts lathing timber from Rundale’ (fol. 286v, 1429 x 1430) supa� wharfam de la Breggehous, ‘on the Bridge House wharf’ (fol. 308, 1430 x 1431) usqʒ le briggequerff, ‘towards the Bridge wharf’ (fol. 100v, 1423 x 1424) apud le ffresshwharf, ‘at the Fresh Wharf’ (fol. 291v, 1429 x 1430)

We can deduce that the article le signalled a switch into English, basing our deduction not only on the etymology of the noun, but also on its lack of Latin morphology. Moving now to what in Medieval Latin and Anglo-Norman was the feminine singular definite article, la; la is infrequent (if not entirely absent) in the Bridge House Estate records until the late 1420s, when it makes a first appearance in the phrase de la: 5. Itm � soluť Thom � Bloye j cisterna & j alegyste & j plank facť ꝓ j eddił ad hostm � de la Cowpe ad macelł scī Nichī p̉c̉ in toto xxvjs viijd, ‘And paid to Thomas Bloye for 1 cistern and 1 alegist and 1 plank made for 1 porch at the door of the Cowpe at St Nicholas Shambles price in all 26s 8d’ (fol. 223, 1428 x 1429) Itm � soluť Thom � Hamon emendanť gladiū de la Raam  dimiɖ die iijd, ‘And paid to Thomas Hamon for mending the sword of the ram by half a day 3d’ (fol. 271, 1429 x 1430)

Cisterns were commonly part of door and window-sill furniture, especially in brewhouses (see Wright 1992). An alejoist was a stand for a cask of ale (see MED giste (n.(3)), and see also Salzman (1952: 553), which describes a brewhouse to be built at Charing Cross in 1492/3 containing alejoists). In fact, practically all (if not all) the tokens of la noted in the Bridge House Estate Weekly Payments and Annual Rentals into the 1460s are embedded in the prepositional phrase de la. The definite article la did not have the function of signalling a following feminine noun when it preceded an English word, and such nouns did not have inherent grammatical gender. Contrast the following pairs of prepositional phrases, as recorded in the Bridge House Estate Weekly Payments (and annual Accounts where signalled): 6. de la horn, ‘from the Horn’ (fol. 267, 1429 x 1430) apɖ le horn, ‘at the Horn’ (fol. 267, 1429 x 1430) de la Brodshelde, ‘from the Broad Seld’ (fol. 42v, 1434 x 1435) tentȏ voc̉ le Brodsheeld, ‘a tenement called the Broad Seld’ (fol. 47, 1434 x 1435) de la Castell, ‘from the Castle’ (fol. 12v, 1461 x 1462) apud le Castell, ‘at the Castle’ (fol. 17v, 1461 x 1462) de la lymekylne, ‘from the limekiln’ (fol. 36v, BHA3, 1461 x 1462) a le lymekylne, ‘to the limekiln’ (fol. 37v, BHA3, 1461 x 1462)

Variability is the hallmark of the late medieval accounts text type, and de le continued to be used too, but only occasionally, as in 7.

134

AN-1a.indd 144

11/03/2010 12:00:24

A Pilot Study on the Singular Definite Articles 7. de le briggehouse, ‘from the Bridge House’ (fol. 69v, 1426 x 1427) de le Gyn, ‘from the gin’ (fol. 306v, 1430 x 1431)

The default became, from the 1420s onwards, de la when preceding an English noun. What, then, of the [Anglo-Norman + English] variety? I surveyed an [AngloNorman + English] text, the Merchant Taylors’ Company Accounts, from 1402 to 1408, and from 1428 to 1434, when that account begins to incorporate more and more English, prior to its switch into monolingual English. Both le and la were used to qualify French nouns, but all the non-French lexemes – that is, words of non-French etymology or non-French word-order or with meanings developed in Britain rather than France – were qualified by le. Examples include le keruyng, ‘the carving’, le refresshyng, ‘the refreshing’, del shudde, ‘from the shed’, al pasterie place, ‘at the pastry place’, del doublinge, ‘of the doubling’, le talwynge, ‘the tallowing’, le picchynge, ‘the pitching’, le Belle en chepe, ‘the Belle in Cheap’ (a tenement name), le kichen ʒerde, ‘the kitchen yard’, les galeymen, ‘the galley men’, le scolemaisť, ‘the schoolmaster’, le cook, ‘the cook’. The only instance I found where la was used with a word that was in any sense not French was la mendure, where mendure was not used in the sense given in the Anglo-Norman Dictionary of ‘amending, putting right something that has gone wrong’, but in the English sense of ‘mending, repairing’: la mendur̉ de vn glasynwyndow (Merchant Taylors’ Company Accounts, fol. 189v, 1434; AND mend2). As with the [Medieval Latin + English] variety, le served to signal a following English noun, but unlike the [Medieval Latin + English] variety, it was also used with French nouns. La was not used to signal a following English noun but a French one. Neither did the articles le and la signal grammatical gender, as the same noun could be modified by both, as demonstrated by fountain, chapel and course in 8 (examples taken from the Merchant Taylors’ Company Accounts): 8. po~ lamendement del fonteigne, ‘for the mending of the fountain’ (fol. 41v, 1407) po~ vn grate p̉s la fonteigne, ‘for a grate next to the fountain’ (fol. 47v, 1407) a vn goter al fyn del chapel, ‘at a gutter at the end of the chapel’ (fol. 41v, 1407) po~ iij formes en la chapel, ‘for 3 forms in the chapel’ (fol. 64, 1412) la prm̉er Cours, ‘the first course’ (fol. 12v, 1407) le prmer cos, ‘the first course’ (fol. 212, c. 1429)

The observation that the same noun could take both feminine and masculine modifiers has been made before (see Wright 1997a and Ingham, this volume), but usually with the conclusion that the scribe was ignorant of Anglo-Norman, not with the conclusion that Anglo-Norman itself had changed. Confining my remarks to the [Anglo-Norman + English] mixed-language business text type, I suggest that le and la ceased to mark grammatical gender in the fifteenth century, but continued on with a new function, that of distinguishing that which

135

AN-1a.indd 145

11/03/2010 12:00:24

The Anglo-Norman Language and Its Contexts was French from that which was not French (bearing in mind that the [AngloNorman + English] variety required material from both language systems). At this point it is worth noticing that Anglo-Norman naming practices seem to have followed the pattern of using de la but not de le, del or du. For example, a check of the index in Keene and Harding (1987) reveals that in Cheapside, residents’ surnames followed the pattern of [de + place-name] and [le + occupational name] as a default, with [de la + name element] occurring sporadically, but with no tokens at all of de le, del or du. A search of McHardy’s (1977) tax lists revealed the same pattern with just one de le, John de le Ewe (1380), and a search of Ekwall (1951) revealed the same pattern with just William del Abbaie and Thomas du Boys as exceptions, both on tax lists of 1319. In conclusion, there is systematicity in the mixed-language business text type, but it is complex. The principles governing the [Medieval Latin + English] variety were not the same as those governing Medieval Latin, and the principles governing the [Anglo-Norman + English] variety were not the same as those governing Anglo-Norman. To rehearse the points made here with regard to definite articles: in the [Medieval Latin + English] variety, the definite article le signalled a following English noun rather than a masculine one, and it blocked a following Latin suffix on that noun. Definite article la occurred only in the prepostional phrase de la, and did not carry the property of signalling a following feminine noun. Thus the articles le and la as evidenced by the Bridge House Estate records were not marked for gender and did not have gender implications for the rest of the noun phrase. My remarks are confined to the Bridge House Estate records from the 1420s to the 1460s only; it remains to be seen how widespread this practice was, and for how long it was operative. In the [Anglo-Norman + English] variety, the definite article le signalled both a following English noun as well as a French one, but the definite article la was not usually followed by an English noun. French nouns could be qualified by both le and la in close proximity, demonstrating that signalling grammatical gender was not relevant and was not the aim. Rather, I suggest, the purpose of retaining both le and la was to signal membership of language system, either French or English. Again, my remarks on the [Anglo-Norman + English] variety are confined to the fifteenth century and to London Merchant Taylors’ Company documents; the wider distribution remains to be determined. There is, however, one wider usage still in use: le in London church-names, such as St Martin’s le Grand, St Mary le Bow, St Mary le Strand and the placename Marylebone – although the le particle was introduced at different dates in these names, often appearing quite late. Zachrisson (1924: 95) remarks: ‘By some popular notion le, later on, came to be looked upon as a preposition with the sense of “on”, “with”, or “by”’. For example, the church by the old Stocks market and known in modern times as St Christopher le Stocks was written in a sixteenth-century English context (Guildhall Library MS 4424: St Christopher le Stocks The Book of Records from 1483) as:

136

AN-1a.indd 146

11/03/2010 12:00:25

A Pilot Study on the Singular Definite Articles 9. Seynte chrystophers fol. 2, 1559 Saint Cristofer fol. 57, 1483 Saynte xp� ofer at ye stockis fol. 69, 1540 Saynte xp� ofer nyghe the stokkꝭ fol. 69, 1540 Saint xp� ofer beside ye stokkꝭ fol. 73, 1536 Scī xp�  oferi prope Le Stoks fol. 80 (Latin context) St Christopher neere the stockꝭ fol. 88, 1612

The church known in Latin as St Michael ad Bladum and in English as St Michael le Quern (a quern is a grindstone, of Old English derivation (OED quern 1.) – the name soon became confused with the noun corn) stood at the western end of Cheapside, between Blow Bladder Street north and Paternoster Row south. It is found written in early documents, and then in an English context in the Churchwardens’ Accounts (Guildhall Library MS 2895/1: St Michael le Quern Churchwardens’ Accounts 1514–1604) as: 10. St Michael que fundata est ante portam Sancti Pauli twelfth century St Michael de Foro twelfth century St Michael le Quern 1260 Sancti Mich’ ad bladum 1271 St Michael atte Corn 1297–8 Saynte Mych< > on the quern fol. 1, 1514 Seynt Mighell at quern fol. 1v, 1522 Seynt Michell in the querne fol. 22, 1521 Sent Mikkellꝭ at the Corn fol. 29, 1522 Scī Michis ad bladum fol. 36v, 1523 Sentte Mykellꝭ querne fol. 59, 1529

The London historian Derek Keene has this to say about the name of St Mary le Bow, Cheapside (Keene and Harding 1987): Throughout the Middle Ages, after the 12th century, the church’s name is almost invariably given as St. Mary de arcubus in Latin or as des arches in French. In the English vernacular of the 14th and 15th centuries the name usually appears as St. Mary atte Bowe, a form which persisted into the 16th century. The names ‘Bowchurch’, ‘St. Mary Bowe’ (or occasionally ‘St. Mary de Bowe’), and ‘Bowchurchyard’ were commonly used in the 16th century and later. The name ‘St. Mary le Bow’ does not appear to have been introduced until the later 16th century, or to have been widely adopted until the 17th.

The place-name of Stratford at Bow also had an unetymological le particle given to it in the sixteenth century; it was written (Gover, Mawer and Stenton 1942: 134):

137

AN-1a.indd 147

11/03/2010 12:00:25

The Anglo-Norman Language and Its Contexts 11. Stratford atte Bowe 1279 Stratford of the Bowe 1449 Stratford on the Bow 1593 Stratford le Bowe c. 1560

The church of St Peter le Poer used to stand in Old Broad Street. It was written as (Guildhall Library MS 4093/1: St Peter le Poer General Register 1561–1723): 12. Sainct Peeter the poore fol. 1v, 1561, and thereafter

The monastery and college of St Martin’s le Grand stood in the street that bears its name. It was referred to in early documents (TNA, Dean and Chapter Archive, Canterbury Cathedral, CCA-DCc-ChAnt/Z/36, 12 Mar 1291) as: 13. ecclesia beati martini King William’s 1068 charter, Latin context mynster [sancta Martine] King William’s 1068 charter, Old English context sanctus martinus magnus London 1291, Latin context saint martin de Londres 1321 judicial inquiries, Anglo-Norman context

I have yet to determine when the first le particle entered, but it does not seem to have been early. The church of St Mary le Strand was built anew in the Strand in 1714 and so named; before that it was referred to (Gover, Mawer and Stenton 1942: 166) as: 14. Parochial de Innocentibus in villa Westm. c. 1220 Eccl. Sanctorum Innocentium 1241 parochia Ste Marie de Stronde 1274 Our Lady at Stronde 1489

The village of Marylebone was originally named Tyburn after the River Tyburn until the fifteenth century, when a new church was built and dedicated to St Mary, and the village migrated upstream to the church and became known as Marybourne. The le particle is not found until the seventeenth century (Gover, Mawer and Stenton 1942: 137): 15. Tiburne 1086 (and so named until 1453) Maryburne 1453 Marylebone 1626

When written in a Latin context in the Bridgewardens’ Annual Accounts, the church known as in parochia Sancti dunstani (‘in the parish of St Dunstan’, BHA 3) was rendered in English contexts until 1496 as St Dunstan’s, St Dunstan Water Lane, and thereafter as St Dunstan in the East (BHA 4 and 5). In the

138

AN-1a.indd 148

11/03/2010 12:00:25

A Pilot Study on the Singular Definite Articles Bridgewardens’ Annual Accounts the phrase ‘in the East’ is spelt on fols. 183v (1496 x 1497), 290 (1507 x 1508), 19 (1510 x 1511); and on fols. 220 (1501 x 1502), 239v (1502 x 1503), 253 (1504 x 1505), 277v (1505 x 1506), 303v (1507 x 1508), and again on fol. 303 (1524 x 1525). But in between 1513 and 1524 it is written on fols. 56 (1513 x 1514), 114 (1516 x 1517), 136 (1517 x 1518), 155 (1517 x 1518), 175v (1518 x 1519), 195v (1519 x 1520), 215v (1520 x 1521), 238v (1521 x 1522), 259v (1522 x 1523), 279 (1523 x 1524). 16. in parochia Sancti dunstani pre-1480 (Latin context) seynt Dunston fol. 323, BHR3, 1480 seint Dunstoon in water lane fol. 377v, BHR3, 1483 S. dunstane in thest fol. 183, BHR4, 1497 in thest 1497–1510 in le Est 1513–24

Hence, the particle le was introduced into the written version of the churchname some thirty-three years after the records had switched to monolingual English. I have lifted these tokens from the Bridge House Estate’s annual rental, which was incorporated into the accounts in most, but not all, years. The subsequent entry each year was for the church of Alhallowe in the Walle (as spelt on fol. 144, BHA 5, 1516 x 1517), in which the definite article was always the and never le, so these le particles were not generalised but were specific to individual church-names. Contracted forms were common: the when followed by a word beginning with a vowel was often written in a cliticised form (as in in the 1496, 1507 and 1510 tokens), and the prepositional phrase at the was often written , as in atte feest of Seint Mighell tharchaungell (fol. 2, BHA 4, 1484 x 1485), ‘at the feast of St Michael the Archangel’. But it could also simply be a spelling for ‘at’: Atte horse le downe (fol. 290v, BHA 4, 1507x 8). This is another London place-name in which le was introduced, this time as a back-formation, the etymology being OE hors + OE ēg, ‘island, marsh’ + OE dun, ‘slope, hill’, first attested as Horseidune c. 1175 (Mills 2001: 115). The meaning of the hors element is not entirely clear. Gover in The Place-Names of Surrey (1934: 32) gives the etymology as OE hors, ‘horse’ + OE ēg, ‘island’ + OE dūn, ‘slope, hill’, which he interpreted as ‘hill by the horse island’. But Reaney in The Place-Names of Essex derives the hors element from OE horsc, which he glosses as ‘mud’, although the dictionaries list it as an adjective meaning merely ‘foul’. OED disagrees and lists it under the headword for the equine mammal: OED horse n. 11. b. A mud or sand bank. dial. 1926 H. A. TRIPP Suffolk Sea Borders vi. 109 Below Waldringfield is a ‘horse’ in mid-channel ‘horse’ being the name given to banks that crop up with rounded backs like the back of a horse. 1929 E. A. ROBERTSON Three came Unarmed ix. 149 Now the shoalwater of this coast is . . . full of under-water mud-banks or ‘horses’ which come dry or are barely covered at low tide.

139

AN-1a.indd 149

11/03/2010 12:00:25

The Anglo-Norman Language and Its Contexts Further downstream in Essex lies Fobbing Horse (so-called in a document of 1203) and Upper Horse and Lower Horse, which are all islands in the Thames estuary. On Canvey Island in the Thames estuary were now-lost place-names Wodeham Horse, Sea Horse, and Wyllyspitt Horse and a few other names in hors- were found nearby (Reaney 1935: 151–2). There may be more ‘horse’ names closer to the City, as the 1992 Ordnance Survey Pathfinder Map TQ48/58 shows that the spit of mud where the River Roding enters the Thames at Barking Creek is named Horse End, but I have been unable to trace the age of this name. An interpretation of hors meaning ‘mudbank in a stream’ + ēg, ‘island’ + dūn, ‘hill, slope’ would make good topographical sense for Horsleydown, and if so, the first attestation of Horseidune in c. 1175 provides an antedating of hors meaning ‘mudbank’ by 750 years. Before 1507, Horsleydown is recorded without the le particle, and it continued without the le particle as Horssey downe as late as the Copperplate map of c. 1553–9. The le particle was introduced unetymologically in or by 1507, which incidentally antedates the English Place-Name Society’s attestations with (‘t. Eliz.’, Gover, Mawer, Stenton 1934: 32). 17. Horsei dune c. 1175 (Mills 2001: 115) horse a downe fol. 200v, BHR4, 1501 x 1502 horse le downe fol. 290v, BHR4, 1507 x 1508

What, then, are we to make of all these late instances of London churches and place-names in le, well after monolingual English names – indeed, a multiplicity of English names – had been recorded for them? Despite records in general switching to monolingual English by 1500, the le element in certain church-names and place-names must have been felt to be integral and thus exempt from the switchover, even though the definite article in the names of others did switch to English, such as All Hallows in the Wall. I speculate that the reason that certain church-names preserved their le particle (or had one assigned to them) may have been because their names were frequently heard spoken aloud in Latin, Latin continuing to be the spoken language of church services – but if so, it must have been Latin of the mixed [Medieval Latin + English] variety, where le signalled a contiguous English word in a Latin context. This would explain why such formations found written in an earlier Latin context as, for example, St Mary de Arcubus, Sanctus Martinus Magnus and St Michael ad Bladum shifted to St Mary le Bow, St Martin’s le Grand and St Michael le Quern, the shift being from Classical Latin to the local, [Medieval Latin + English] variety, with the church-names all following the pattern identified by Trotter of [le + English word]. Alternatively, we would have to posit that the churchwardens understood the older system of [le + English word] long after it had fallen out of use and reintroduced it, perhaps for reasons of tradition.

140

AN-1a.indd 150

11/03/2010 12:00:25

A Pilot Study on the Singular Definite Articles Bibliography manuscripts London Metropolitan Archives MS Bridgemasters’ Account Roll 1, 1381 x 1382. MS Bridgemasters’ Account Roll 17, 1404 x 1405. MS Bridge House Accounts vol. II, (BHA 2), 1423 x 1460. MS Bridge House Accounts vol. III, (BHA 3), 1460 x 1484. MS Bridge House Accounts vol. IV, (BHA 4), 1484 x 1509. MS Bridge House Accounts vol. V, (BHA 5), 1509 x 1525 MS Bridge House Weekly Payments, first series, vol. I, (BHWP 1), 1404 x 1412. MS Bridge House Weekly Payments, first series, vol. II, (BHWP 2), 1412 x 421. MS Bridge House Weekly Payments, first series, vol. III, (BHWP 3), 1421 x 1430. Guildhall Library, London MS 34048/1, Merchant Taylors’ Company Accounts, 1397 x 1445. MS 4424, The Book of Records of St Christopher le Stocks from 1483. MS 2895/1, St Michael le Quern Churchwardens’ Accounts 1514–1604. MS 4093/1, St Peter Le Poer General Register 1561–1723 (microfilm).

printed works Ekwall, E. (1954), Street-Names of the City of London. Oxford. Gover, J., A. Mawer and F. Stenton. (1934), The Place-Names of Surrey. English PlaceName Society, vol. XI. Cambridge. ——————(1942), The Place-Names of Middlesex apart from the City of London. English Place-Name Society, vol. XVIII. Cambridge. Ingham, R. (this volume), ‘The transmission of later Anglo-Norman: some syntactic evidence’. Mills, A. (2001), Oxford Dictionary of London Place-Names. Oxford. Reaney, P. H. (1935), The Place-Names of Essex. English Place-Name Society, vol. XII. Cambridge. Salzman, L. (1952), Building in England Down to 1540 A Documentary History. Oxford. Wright, L. (1992), ‘OED’s tabard, 4’, Notes and Queries, ns 39/2, 155–7. ——(1995), ‘Middle English -ende and -ing: a possible route to grammaticalization’, in J. Fisiak (ed.), Linguistic Change under Contact Conditions, Trends in Linguistics Studies and Monographs 81, Berlin, pp. 365–82. ——(1996), Sources of London English: Medieval Thames Vocabulary. Oxford. ——(1997a), ‘The records of Hanseatic merchants: ignorant, sleepy or degenerate?’, Multilingua 16/4, 339–50. ——(1997b), ‘Medieval Latin, Anglo-Norman and Middle English in a civic London text: an inquisition of the River Thames, 1421’, in S. Gregory and D. Trotter (eds.), De mot en mot. Aspects of medieval linguistics: Essays in Honour of William Rothwell, Cardiff, pp. 223–60. ——(1998), ‘Mixed-language business writing: five hundred years of codeswitching’, in E. Håkon Jahr (ed.), Language Change: Advances in Historical Sociolinguistics, Trends in Linguistics Studies and Monographs 114, Berlin, pp. 99–118.

141

AN-1a.indd 151

11/03/2010 12:00:25

The Anglo-Norman Language and Its Contexts Wright, L. (2000), ‘Bills, accounts, inventories: everyday trilingual activities in the business world of later medieval England’, in D. Trotter (ed), Multilingualism in Later Medieval Britain, Woodbridge, pp. 149–56. —— (2001), ‘The role of international and national trade in the standardisation of English’, in I. Moskowich-Spiegel Fandino, B. Crespo Garcia, E. Lezcano Gonzalez and B. Gonzalez (eds.), Re-interpretations of English. Essays on Language, Linguistics and Philology (I), Coruna, 189–207. ——2002. ‘Standard English and the lexicon: why so many different spellings?’, in Mari C. Jones and Edith Esch (eds.), Language Change: The Interplay of Internal, External and Extra-linguistic Factors, Berlin, pp. 181–200. —— (2005), ‘Medieval mixed-language business texts and the rise of standard English’, in J. Skaffari, M. Peikola, R. Carroll, R. Hiltunen and B. Wårvik (eds.), Opening Windows on Texts and Discourses of the Past, Pragmatics and Beyond New Series 134, Amsterdam, pp. 381–99. Zachrisson, R. (1924), ‘The French element’, in A. Mawer and F. Stenton (eds.), Introduction to the Survey of English Place-Names. English Place-Name Society, vol. I, part I. Cambridge.

online sources British History Online: https://www.british-history.ac.uk/, https://www.britishhistory.ac.uk/report.aspx?compid=63231 Ekwall, E. (1951), Two Early London Subsidy Rolls: http://www.british-history.ac.uk/ report.asp?compid=31958&strquery=subsidy%20rolls Keene, D., and V. Harding (1987), Historical Gazetteer of London before the Great Fire: Cheapside; Parishes of All Hallows Honey Lane, St Martin Pomary, St Mary le Bow, St Mary Colechurch and St Pancras Soper Lane: http://www.british-history.ac.uk/ report.asp?compid=8467 McHardy, A. 1977. The Church in London 1375–1392. London Record Society. http:// www.british-history.ac.uk/report.asp?compid=36013&strquery=mchardy Oxford English Dictionary Online: http://dictionary.oed.com/ TNA: http://www.nationalarchives.gov.uk/A2A/records.aspx?cat= 054-cadchant_8&cid=-1&Gsm=2008–06–18#-1

142

AN-1a.indd 152

11/03/2010 12:00:25

chapter eleven

Investigating Anglo-Norman Influence on Late Middle English Syntax* Eric Haeberli

1. Introduction This paper focuses on some word-order developments in the Middle English period and explores whether these developments might have been influenced by contact with Anglo-Norman and/or Continental French. The issues to be considered are related to what has generally been referred to (somewhat misleadingly) as the Verb Second (V2) phenomenon. As extensively discussed in the literature, Old English exhibits frequent subject–verb inversion when a non-subject is in clause-initial position. Such word orders are reminiscent of the V2 phenomenon as found in all the modern Germanic languages with the exception of present-day English. In the Middle English period, the Old English subject–verb inversion syntax is lost to a large extent, but the development exhibits certain peculiarities that have remained unexplained. The aim of this article is to consider whether some or all of these peculiarities could be related to Anglo-Norman/French influence. The paper is organised as follows. Section 2 provides a brief description of the subject–verb inversion syntax found in Old and Early Middle English. In section 3, the developments in Middle English are discussed and four open issues with respect to these developments are identified. Section 4 then considers the plausibility of addressing these issues with reference to AngloNorman/French influence. It is argued that for three of these questions, contact with Anglo-Norman/French may have played a role whereas such an account seems less likely for the fourth issue. Finally, in section 5, some further points are discussed that bear on the question of Anglo-Norman/French influence on Middle English syntax, and section 6 summarises the paper.

* I would like to thank the participants in the Workshop on Anglo-Norman and Middle English held at the University of Central England, Birmingham (now Birmingham City University) for comments and discussion. Special thanks go to Richard Ingham for suggestions that led me to explore the issues presented in this paper.

143

AN-1a.indd 153

11/03/2010 12:08:45

The Anglo-Norman Language and Its Contexts 2. Subject–verb inversion in Old and Early Middle English The V2 property as found in all the modern Germanic languages except present-day English is characterised by the general occurrence of the finite verb right after the clause-initial constituent (i.e. in second position) regardless of what the nature of this constituent is.1 A consequence of this is that when the clause-initial constituent is not a subject the order of the subject and the finite verb is inverted, which leads to the characteristic inversion property of V2 languages. In Old and Early Middle English, cases of subject–verb inversion can regularly be found when some other constituent is fronted (see for example van Kemenade 1987, Pintzuk 1999). This is illustrated in (1) (fronted constituent in brackets, finite verb in bold print, subject in italics). (1) a. [ðæt] wat ælc mon  Boethius, 36.107.17 that knows every man ‘Everyone knows that.’ b. And [egeslice] spæc Gregorius be ðam . . .  Wulfstan, 202.46 And sternly spoke Gregorius about that ‘And Gregorius spoke sternly about that . . .’

In (1a), an object is in initial position whereas in (1b) it is an adverb that has been fronted. In both cases the subject follows the finite verb, and we thus get word orders that are reminiscent of languages characterised by the V2 property. However, the syntax of inversion in Old and Early Middle English does not fully correspond to that found in genuine V2 languages. In particular, there are two ways in which Old English/Early Middle English differ from a typical V2 system. First, a distinction between pronominal and full NP subjects has to be made in Old English/Early Middle English as inversion is possible with full NP subjects (cf. 1) but not with pronominal subjects (cf. 2a). The only exception to this observation can be found in some specific contexts (henceforth ‘genuine V2’ (GV2) contexts) such as questions, negative clauses and clauses introduced by some short adverbs (in particular þa, þonne, ‘then’) where subject–verb inversion also occurs with pronominal subjects (cf. 2b/c). The second property that distinguishes Old English/Early Middle English from typical V2 languages is the fact that even with full NP subjects inversion is not systematic, as example (2d) illustrates. (2) a. [þæt] [þu] meaht swiðe sweotole ongitan (XSV . . .)  Boethius, 88.14 that you can very easily understand 1



Languages may vary as to whether V2 is available in main clauses only or both in main clauses and subordinate clauses. Here we will focus on V2 in main clauses as early English does not seem to have had productive V2 in subordinate clauses (see for example van Kemenade 1997).

144

AN-1a.indd 154

11/03/2010 12:00:25

Investigating Anglo-Norman Influence on Late Middle English Syntax b. [hwi] sceole we oþres mannes niman (XVS . . .)  ÆLS 24.188 why should we another man’s take ‘Why should we take those of another man?’ c. [ða] aras he hal & gesund. (XVS . . .)  Bede 4: 32.380 then arose he uninjured and healthy ‘Then he got up uninjured.’ d. [ðone] Denisca leoda lufiað swyðost (XSV)  Wulfstan, 223.54 that Danish people love most ‘The Danish people love that one most’

The above observations are confirmed by quantitative evidence based on ten Old English text samples from The York–Toronto–Helsinki Parsed Corpus of Old English Prose (henceforth YCOE; Taylor et al. 2003). In contexts other than GV2, subject–verb inversion occurs in only six out of 391 (1.5%) of the clauses examined with an initial non-subject and a pronominal subject (Haeberli 2002a). Four of the six cases of inversion are from the same text (Orosius), and seven out of the ten text samples show no inversion at all. Thus, non-inversion as in (2a) is nearly compulsory with pronominal subjects in Old English. In the same kind of context, non-pronominal subjects invert with the finite verb in 75.3% (1437/1909) of the cases found in another sample taken from the YCOE. Non-V2 orders of the type shown in (2d) therefore occur with a non-negligible frequency of about 25%. Finally, it should be pointed out that the above quantitative observations also hold to a large extent for Early Middle English (cf. Kroch and Taylor 1997).

3. The loss of the Old English/Early Middle English subject–verb inversion syntax After the Early Middle English period, the Old English/Early Middle English subject–verb inversion syntax starts being lost. This development has been extensively discussed in the literature under the label ‘loss of V2’ (cf. e.g. Haeberli 2002a, b and references cited there). A central question in the literature has been why the productive Old English/Early Middle English subject–verb inversion syntax was lost during the Middle English period. Various potential causes for this change have been proposed, ranging from external ones (language and dialect contact (Scandinavian/northern English); see for example Kroch and Taylor 1997, Kroch, Taylor and Ringe 2000) to internal ones (e.g. loss of empty expletive subjects; Haeberli 2002b). It would go beyond the scope of this paper to review these proposals in detail. What is essential for our purposes is that there are at least four aspects of the development of subject–verb inversion in Middle English that remain difficult to explain in terms of any of the accounts proposed so far. These four aspects are listed below: (I) In what I called ‘genuine V2’ (GV2) contexts above, the syntax of inversion

145

AN-1a.indd 155

11/03/2010 12:00:25

The Anglo-Norman Language and Its Contexts remains more or less stable throughout the history of English. Thus, in questions and negative contexts, fronting of a non-subject element still gives rise to inversion in present-day English (e.g. When will he leave? Never again would I do that.). What has changed over time is the nature of the element that inverts with the subject (any type of verb in Old English/Middle English, auxiliaries only in present-day English), but the basic inversion property has been maintained. However, there are some elements that were mentioned among the GV2 contexts above that do not form part of this group any more in presentday English. Whereas adverbs like þa, þonne, ‘then’ (and to a lesser extent nu, ‘now’) systematically gave rise to subject–verb inversion in Old English/Early Middle English even with pronominal subjects, their descendants no longer do so in present-day English (*Then did he leave). One of the unanswered questions with respect to the syntactic developments in Middle English is why þa or þonne, contrary to interrogative and negative elements, lost their ability to trigger systematic subject–verb/auxiliary inversion. (II) A second question is based on the observation that the frequency of subject–verb inversion, although declining substantially outside GV2 contexts during Middle English, never drops to 0% even today. Some examples of present-day English subject–verb inversion are given in (3) (from Bresnan 1994: 78, Schmidt 1980: 6/8/9, Stockwell 1984: 581). (3) a. [Another very generous person] is Mr. McDonald. b. [In this rainforest] can be found the reclusive lyrebird. c. [Round the corner] came a big red bus. d. [Thus] ended his story. e. [In the year 1748] died one of the most powerful of the new masters of India.

Present-day English inversion can be found for example with be in contexts of predicate fronting (3a/b) and with main verbs in passive and unaccusative2 constructions, in some very restricted contexts (e.g. locative inversion (3c), and with certain clause-initial adjuncts (3d–e)). However, there are contexts in which there was a complete loss of subject– verb/auxiliary inversion in the history of English. Thus, as shown in (4), inversion with transitive verbs and inversion of the type ‘auxiliary–subject–verb’ are now entirely ungrammatical in present-day English while they systematically occurred in Old English/Early Middle English. (4) a. *[In this rainforest] can find a lucky hiker the reclusive lyrebird. b. *[In this rainforest] can the reclusive lyrebird be found.

Table 1 shows the status of subject–verb inversion in various contexts in Old English and Late Middle English (1350 to 1500). The Old English data are based on seven texts from the YCOE ((Boethius, Chronicle, Cura Pastoralis, Ælfric’s 2



Unaccusatives are typically verbs of change of state and change of location.

146

AN-1a.indd 156

11/03/2010 12:00:26

Investigating Anglo-Norman Influence on Late Middle English Syntax Table 1. Main clauses with an initial constituent (except question words, negation, þa, þonne, nu) preceding subjects in texts from Old English and from 1350 to 1500 Text (date)

Inversion with Inversion with transitive V and other V and full full NP subject NP subject

Inversion with pronominal subject

Old English text samples

59.5% (314/528)

81.3% (1123/1381) 1.5% (6/391)

30.4% (7/23)

73.7% (60/92)

25.4% (16/63)

40.3% (39/97)

65.2% (60/92)

15.4% (6/39)

0.0% (0/28) 0.0% (0/19) 1.1% (2/181) 2.7% (5/190) 44.1% (126/286) 22.0% (22/100) 34.0% (55/162) 71.0% (76/107)

3.0% (2/66) 10.2% (9/88) 20.8% (44/212) 25.6% (124/485) 40.5% (161/398) 68.0% (198/291) 80.1% (431/539) 73.6.% (265/360)

2.1% (1/47) 0.0% (0/103) 0.0% (0/25) 0.0% (0/48) 15.1% (13/86) 7.1% (6/85) 3.1% (1/32) 50.0% (95/190)

m2 (1250–1350) Earliest English Prose Psalter (c. 1350) m2/4 (comp. 1250–1350, MS 1420–1500) Richard Rolle (c. 1440/50 (a. 1348/9)) m3 (1350–1420) Old Testament (a. 1425 (a. 1382)) New Testament (c. 1388) Purvey, Prologue to the Bible (c. 1388) Trevisa, Polychronicon (a. 1387) Wycliffite Sermons (c. 1400) Brut/Chronicles of England (c. 1400) Mandeville’s Travels (?a. 1425 (c. 1400)) Chaucer (Boethius, Melibee, Parson, Astrolabe; c. 1380/1390) Cloud of Unknowing (a. 1425 (?a. 1400)) Mirror of St Edmund, MS Vernon (c. 1390) TOTAL m3

50.0% (6/12) 78.0% (64/82) 19.9% (42/211) 80.6% (29/36) 89.4% (59/66) 15.4% (23/149) 28.6% (321/1121) 52.5% (1357/2587) 18.5% (181/976)

m3/4 (comp. 1350–1420, MS 1420–1500) ME Sermons, MS Royal (c. 1450 (c. 1415)) Mirk’s Festial (a. 1500 (a. 1415)) Mirror of St Edmund, MS Thornton (c. 1440 (?1350)) TOTAL m3/4

9.1% (1/11) 21.0% (41/195) 69.4% (25/36)

31.4% (11/35) 51.4% (197/383) 83.3% (50/60)

6.6% (4/61) 3.6% (1/28) 52.5% (105/200)

27.7% (67/242)

54.0% (258/478)

38.1% (110/289)

m4 (1420–1500) Life of St. Edmund (c. 1450 (1438)) Book of Margery Kempe (c. 1450) Malory, Morte Darthur (a. 1470) Gregory’s Chronicle (c. 1475) Siege of Jerusalem (c. 1500) Capgrave’s Chronicle (a. 1464) TOTAL m4

0.0% (0/20) 5.9% (8/136) 14.9% (34/228) 3.9% (5/129) 12.5% (3/24) 20.0% (44/220) 12.4% (94/757)

10.8% (4/37) 36.5% (101/277) 36.7% (202/550) 44.4% (190/428) 50.0% (27/27) 74.0% (553/747) 51.5% (1077/2093)

0.0% (0/72) 12.7% (16/127) 12.9% (30/233) 0.0% (0/59) 4.4.% (4/91) 51.7% (31/60) 12.6% (81/642)

Letters, Ælfric’s Lives, Apollonius, Wulfstan). The Middle English data are taken from the Penn–Helsinki Parsed Corpus of Middle English 2 (henceforth PPCME2; Kroch and Taylor 2000), and more specifically from twenty-one texts or text samples containing more than fifty main clauses with a non-pronominal subject that is preceded by some constituent (except subordinate clauses,

147

AN-1a.indd 157

11/03/2010 12:00:26

The Anglo-Norman Language and Its Contexts question words, negation, þa, þonne, nu). For both Old English and Middle English, smaller samples were used for the section on subject pronouns. The labels m2, m3 and m4 in Table 1 refer to the Middle English periods introduced in the Helsinki Corpus of English Texts. Columns 2 and 3 of Table 1 show the contrast between verb types with respect to subject–verb inversion in clauses with non-pronominal subjects. Whereas the frequency of inversion with transitive verbs drops from nearly 60% in Old English to an average of 12.4% at the end of the fifteenth century, inversion with other verbs remains at a relatively high level until the end of the Middle English period (average of 51.5%). These data and the observations related to present-day English in examples (3) and (4) thus raise the following questions, to which no answers have been given so far in the literature: Why was subject–verb inversion not lost in all contexts? Why was it maintained mainly with unaccusative verbs and be? (III) If we now turn to column 4 of Table 1, we can observe that non-negligible frequencies of subject–verb inversion can be found with pronominal subjects in many Middle English texts. This phenomenon is illustrated in (5). (5) a. [On þe same maner] schalt þou do wiþ þis lityl worde GOD.    Cloud, 78.323 b. and [þe cherch of Lincoln] gaue he to Herry Beuforth . . .    Capgrave, 210.11 c. & [many tymes] haue I feryd þe wyth gret tempestys of wyndys Kempe, I, 51.110 d. And [many mervayles] shall he do  Malory, 47.79 e. [þis question] wolde I knowe of you  Private Letters, Mull, I, 126.623

Such orders were to a large extent ruled out in Old English. Thus, we find an increase with subject pronouns in the Middle English period that goes against the general trend of a decrease in inversion. The question that remains to be answered in this context is why subject–verb inversion emerged with pronominal subjects in Middle English. (IV) Table 1 also shows that the frequencies of inversion vary considerably across texts. For example, while some texts in the period m3 have reached a present-day English-like stage with hardly any inversion, others from the same period still have inversion rates of well over 50% even with transitive verbs. The final question that therefore arises is why authors vary so much in their use of subject–verb inversion in Late Middle English.

4. Exploring Anglo-Norman influence as a potential solution to (I) to (IV) As pointed out earlier, there have been generally no conclusive answers to the questions raised in (I) to (IV) in the literature so far. The goal of the remainder of this paper is to explore whether Anglo-Norman/French influence on Middle

148

AN-1a.indd 158

11/03/2010 12:00:26

Investigating Anglo-Norman Influence on Late Middle English Syntax English syntax could help us to account for these puzzles. But before considering each specific issue from this perspective, let us start by making some general observations. First, if we look at the literature on Middle English, we notice that, in stark contrast with the developments in the lexicon, contact with French has generally not been considered as a very important factor in the development of Middle English syntax. For example, in Fischer’s (1992) overview article of Middle English syntax, there are only four indexed references to potential French influence. The contexts in which French influence is mentioned is the rise of do (1992: 273), the development of the periphrastic genitive (1992: 226), the temporary emergence of postnominal adjectives (1992: 214), and the emergence of wh-relatives (1992: 299 ff.). However, one can also occasionally find reference to French influence in other contexts, as for example the emergence of indirect objects introduced by to (Allen 2006: 214/5). Furthermore, Ingham (2005) explores French influence on recessive features of Middle English syntax (including the one mentioned in issue (III) above). From the point of view of their basic syntactic properties, interaction between the two languages would not be implausible. Old French was a rather systematic V2 language, but subject–verb inversion began to weaken in Middle French, as shown by the increase in the frequency of ‘XP–Subject–V . . .’ orders in the fourteenth and fifteenth centuries (see for example Adams 1987, Roberts 1993, Vance 1997). Thus, Middle English underwent changes that are rather similar to those found in French. Furthermore, from the point of view of the status of French in fourteenth-century England, substantial influence that goes beyond the lexicon would certainly be conceivable as well. For example Rothwell (1998) points out that ‘[t]he scribal class of medieval England, responsible in large measure for the enrichment of later Middle English, was in varying degrees a trilingual one’. Transfer of syntactic features in the writing of such multilingual authors would not be unexpected. The continued importance of French in Late Middle English is also stressed by Kristol (2000: 38/9): ’Même si certains témoignages, en particulier un passage de la Manière de langage de 1396, affirment que le français est toujours la langue de conversation soignée dans certains milieux de la bonne société anglaise, .  .  . la situation linguistique en Angleterre médiévale doit sans aucun doute être décrite comme une diglossie codique: l’oralité appartient essentiellement à l’anglais, alors que le français occupe une partie importante des usages écrits.’3 And more specifically with respect to syntax, Ingham (2005: 22) speculates that ‘with late 14th century English we may not be looking at the product of an organic development of English from Early Middle English onwards, but rather at the reflex of

3

‘Even if some sources, in particular a passage from Manière de langage from 1396, affirm that French is still the language of refined conversation in certain circles of the English high society, . . . the linguistic situation in medieval England should without doubt be described as a code diglossia: orality essentially belongs to English whereas French occupies an important part of written usage.’

149

AN-1a.indd 159

11/03/2010 12:00:26

The Anglo-Norman Language and Its Contexts Anglo-Norman linguistic practices on which bilingual writers were calquing their English syntax’. Given these observations, it would in principle not be implausible that the four peculiarities of Late Middle English syntax observed in section 3 are related to contact with French. Note however that the types of influence that would be required for issues (I)/(II) on the one hand and issues (III)/(IV) on the other are not of exactly the same nature. Thus, issues (III) and (IV) (variation among authors, increase of inversion with pronouns) could simply imply occasional influence on the writing of Late Middle English authors that may not have profoundly affected the grammar of English (creating what, in the context of subject–verb inversion, may look like vestiges of a more productive inversion grammar; cf. Ingham 2005). Issues (I) and (II) (inversion with ‘then’, differences with respect to verb types), however, would imply more substantial influence on the grammar of English, i.e. even in the long term (loss of an option in (I) or introduction/maintenance of an option in (II)).

4.1. (I) why did þa/þonne (‘then’) contrary to interrogative and negative elements, lose the ability to trigger systematic subject–verb/ auxiliary inversion? As pointed out in section 2, the adverbs þa/þonne (‘then’) systematically trigger subject–verb inversion with both full NP and pronominal subjects in Old English (GV2, cf. example 2c). The later developments with respect to inversion with þa/þonne and their Middle English equivalents is shown in Table 2. Table 2 is based on the same Middle English texts as used for Table 1 and, in addition, includes the Early Middle English texts from the PPCME2 (period m1). Table 2. Inversion in main clauses with initial then in Middle English (same texts as in Table 1 + Early Middle English (m1)) Full NP subject

Subject pronoun

m1 (1150–1250)

94.5% (171/181)

86.5% (173/200)

m2 (1250–1350)

100% (3/3)

71.4% (5/7)

m24

37.5% (9/24)

50.0% (20/40)

m3 (1350–1420)

36.8% (127/345)

42.9% (146/340)

m34

50.6% (176/348)

39.5% (139/352)

m4 (1420–1500)

30.4% (219/720)

30.2% (191/632)

A minor decrease in inversion can already be observed in the Middle English period m1 as 13.5% of the clauses with a subject pronoun and 5.5% of those with a full NP subject exhibit non-inversion. By the end of fourteenth century, inversion has become a minority pattern, with frequencies that are similar to inversion in general, in particular with full NP subjects. For example, if we consider inversion with all verbs in period m3 in Table 1, we obtain a frequency

150

AN-1a.indd 160

11/03/2010 12:00:26

Investigating Anglo-Norman Influence on Late Middle English Syntax of 45.3% (1678/3708) whereas in Table 2 we have an even lower frequency of 36.8% in the same period. Inversion with pronominal subjects, however, is still more frequent with then in period m3 (42.9%) than with other clause-initial constituents (18.5% in Table 1). Could the decline of inversion with then in Middle English be related to French influence? Various pieces of evidence suggest that such a scenario is possible. As pointed out by Ingham (2006a), subject–verb inversion in Continental French seems to decline faster in contexts with an initial temporal adjunct than in other contexts. In support of this claim, Ingham gives the following frequencies for inversion with initial time adjunct in chronicles for three different periods: 1230–1275: 89%; 1290–1340: 30%; 1340–1400: 31%. This is in stark contrast with clauses with initial objects in the same texts. There, the rate of inversion is 100% in all periods. Ingham (2006b) makes very similar observations for chronicles written in Anglo-Norman, i.e. the variety that should be even more revealing from the point of view contact scenarios with Middle English. On the basis of the data provided by Ingham for chronicles from the second half of the thirteenth century and the first half of the fourteenth century (2006b: 38–40), we obtain a frequency of inversion in clauses with an initial time adjunct and a full NP subject of 55.8% (with unaccusatives 69.6%, with verbs other than unaccusatives: 27.2%). This rate of inversion is again considerably lower than with initial objects (85.7%) or with initial place adjuncts (100%) in the same texts. Thus, both in Continental French and in Anglo-Norman, initial time adjuncts seem to be the weakest triggers of inversion in the thirteenth and fourteenth centuries. Observations made by other authors point in the same direction. For example, with respect to a Middle French equivalent of then, Vance (1997) confirms that its capacity to trigger inversion was weakened early in Middle French. She notes that in the fifteenth–century text Saintré ‘the monosyllabic adverb lors, one of the first elements to participate in CSV4 in early MidF, has completely ceased to trigger inversion’ (1997: 347). More specifically in connection with French influence on English, we can also refer to Kroch and Taylor’s (1997) study of inversion in the Ayenbite of Inwit, a Kentish text from 1340 which is a fairly close translation of the French work La somme le roi. Kroch and Taylor (1997: 312) show that with clause-initial objects the Ayenbite of Inwit behaves very much like Old and Early Middle English: Inversion occurs in 82% of the clauses with a full NP subject and in 8% of the clauses with a subject pronoun. However, a completely different picture emerges for inversion with then. With full NP subjects the rate of inversion is as low as 25% and with pronominal subjects it is 58%. This deviation from the Old English pattern is unexpected at first sight. However, the observations made in the previous paragraph and the fact that we are dealing with a translation from French make an explanation in terms of French influence very likely. The translation context may then simply be one manifestation of a more general effect of contact with French.

4

I.e. non-inverted order with some constituent C in a pre-subject position.

151

AN-1a.indd 161

11/03/2010 12:00:26

The Anglo-Norman Language and Its Contexts In summary, whereas in Old and Early Middle English then distinguished itself from many other constituents in that it triggered systematic subject–verb inversion, Continental French and Anglo-Norman temporal adjuncts in the thirteenth and fourteenth centuries were distinctive in the opposite way as they were weaker triggers of inversion than other constituents. This distinctive property of temporal adjuncts in Continental French/Anglo-Norman could then be argued to have contributed to the decline in inversion with the temporal adjunct then in Middle English. It should also be pointed out that such Continental French/Anglo-Norman influence would have occurred within a Middle English context that seemed favourable to a weakening of inversion with then. As mentioned in section 2, GV2 is also found in Old English questions and negative clauses. It has therefore often been proposed that GV2 triggers can be unified by means of the semantic notion of operator. Thus, GV2 arises when an (overt or empty) operator occurs in clause-initial position. However, a temporal adverb like then does not form a natural class with operators, and it would therefore have had a marked status as a trigger of GV2 in Old and Early Middle English. Thus, the elimination of then from GV2 contexts would have been a natural development from a purely language-internal point of view, but contact with French may have provided the necessary impetus to set this development in motion.

4.2. (II) why was subject–verb inversion not lost in all contexts? Why was it maintained mainly with unaccusative verbs and be? As Table 1 shows, the Middle English period was one during which we can observe a considerable decline in subject–verb inversion. It is important to point out, however, that the establishment of the inversion syntax as we know it from present-day English continues in the Early Modern English period. There are indeed various types of inversion that can still regularly be found at the end of the Middle English period but are ungrammatical or very restricted in present-day English. For example, the subject in passives often occurs postparticipially, as shown in (6). (6) a. [Than] was mad pes on þis maner þat . . .     CapChr, 88.1704 b. and [with him] was coroned Helianore, doutir to þe kyng of Spayn CapChr, 127.2913

According to the data provided in Haeberli (2002c), full NP subjects occur with a frequency of over 13% in the post-participial position in main clauses during period m4 of the PPCME2 (1420–1500). In present-day English such orders can be found only in very restricted contexts (e.g. locative inversion) and these restrictions must have been introduced in the modern period. Although the developments in the subject–verb inversion syntax continue beyond the period during which French influence can plausibly be argued to be relevant, the emergence of some basic trends can nevertheless be situated within this period. In particular, as shown in Table 1, the loss of inversion with

152

AN-1a.indd 162

11/03/2010 12:00:26

Investigating Anglo-Norman Influence on Late Middle English Syntax transitive verbs is in clear progress throughout the Middle English period, whereas we find stagnation with other verbs in the fourteenth and fifteenth centuries. The issue that arises therefore from the point of view of potential French influence is why the Middle English decrease in inversion affected transitive verbs much more than other verbs. Could French have contributed to such a distinction between verb types? A positive answer to this question cannot be entirely ruled out, but it seems somewhat less plausible than in the context discussed in the previous subsection. Some support for the hypothesis of French influence may be obtained from the observation made in the literature that a distinction with respect to verb type also played a certain role in French inversion in the relevant periods. For example, according to Vance (1995), there is a gradual increase in the proportion of passive and unaccusative verbs among the clauses with the order ‘XP–verb–subject’ from the early thirteenth century to the late fifteenth century. In other words, the frequency of inversion with transitive verbs seems to decline.5 Furthermore, Ingham’s (2006b: 38) data based on Continental French chronicles from 1250 to 1350 suggest that when inversion is optional it is slightly favoured with unaccusative verbs. Thus, in clauses with an initial time adjunct, we find a frequency of 74.5% (143/182) inversion with unaccusatives as opposed to 61.0% (130/213) with other verbs. As for Anglo-Norman chronicles from the same period (1250–1350), Ingham’s (2006b: 38) figures indicate an even stronger contrast between verb types. Whereas inversion in clauses with an initial time adjunct occurs at a rate of 69.9% (243/349) with unaccusative verbs, the corresponding frequency for other verbs is 30.4% (46/169). Although verb type seems to play a role in the subject–verb inversion syntax in French, it is nevertheless doubtful whether this role was strong enough to influence Middle English. The data given by Ingham (2006b) concern only a very specific context, namely clauses with initial time adjuncts. In other types of clauses, the syntax of inversion in thirteenth and fourteenth century Continental French/Anglo-Norman is, as Ingham’s (2006b: 39/40) other data show, still very robust regardless of verb type. Even in the fifteenth century, inversion with full NP subjects remains fairly productive, as Vance’s (1997: 350) frequencies of inversion ranging from 50% to 73% suggest (cf. Table 3 in the next section for details). Furthermore, Vance’s (1995) data show that the increase in the proportion of unaccusative verbs in clauses with inversion is most striking in the fifteenth century, with frequencies rising from around 50% 5



Note that the distinction of verb types made by Vance (and also by Ingham 2006b, discussed below) is not exactly the same one as ours in Table 1. Our distinction is between transitive verbs and all other verbs (i.e regular intransitive (i.e. unergative) verbs as well as unaccusatives) whereas the distinction made for French is between unaccusatives and all other verbs (i.e. unergatives and transitives). The reason why no attempt is made in Table 1 to isolate unaccusatives is that it is notoriously difficult to delimit this verb class precisely. For the purposes of comparing the data, we will assume that figures for verbs other than unaccusatives in the French data reflect trends for transitive verbs even though they also include unergatives.

153

AN-1a.indd 163

11/03/2010 12:00:26

The Anglo-Norman Language and Its Contexts in the thirteenth and fourteenth centuries (Queste (1220) 49%; Joinville (1306) 56%) to around 70% in the fifteenth century (Saintré (1456) 69%; Commynes (1491) 78%). What these data suggest is that, although there is a development towards favouring subject–verb inversion with some verbs and disfavouring it with others in medieval French, the development may not be sufficiently advanced at what seems to be the latest relevant period for influence on Middle English syntax (i.e. before the fifteenth century). As a matter of fact, Ingham’s (2006b) comparison of Continental French and Anglo-Norman discussed in the previous paragraph may even suggest the opposite scenario. The contrast in inversion between unaccusative verbs and other verbs is stronger in AngloNorman (69.9% vs. 30.4%) than in Continental French (74.5% vs. 61.0%). So one might wonder whether it was not rather Middle English that influenced Anglo-Norman, as shown by the lower inversion rate with verbs other than unaccusatives in Anglo-Norman, rather than the other way round. In short, French seems chronologically to be lagging behind the developments in English with respect to the loss of subject–verb inversion. It therefore seems to be a rather unlikely source directing English towards a system in which inversion is ruled out with transitive verbs but survives with other verbs, in particular unaccusatives and be. A different explanation has thus to be found for this development. One possibility is that there are two fundamentally different ways to derive subject–verb inversion already in Old English, and only one of them is lost in Middle English (i.e. Germanic inversion, but not a kind of ‘free’ (Romance) inversion). But the question remains as to exactly how and why those cases of inversion survived that we now have in presentday English.

4.3. (III) why did subject–verb inversion with pronominal subjects emerge in the middle english period? As shown in Table 1, cases of subject–verb inversion with pronominal subjects can regularly be found in Middle English texts, although this option is generally ruled out in Old English. This is a surprising development given that the Middle English trend is towards eliminating inversion rather than towards increasing it. Compared to issue I examined in the previous subsection, this particular puzzle seems to be more amenable to an explanation in terms of French influence. Consider for example the quantitative data on subject–verb inversion in Old and Middle French provided by Vance (1997: 350), as set out in Table 3. Although there is a decrease in inversion with subject pronouns in the fourteenth century which is much more substantial than with full NP subjects, the frequencies remain high (i.e. higher than in almost all Middle English texts). These observations are to a large extent confirmed by Ingham’s (2006b) Continental French and Anglo-Norman chronicle data from 1250–1350. In Continental French, subject–verb inversion with a pronominal subject is entirely productive, with a rate of inversion of 61.9% (13/21) with an initial object and

154

AN-1a.indd 164

11/03/2010 12:00:26

Investigating Anglo-Norman Influence on Late Middle English Syntax Table 3. Main clauses with an initial constituent preceding subjects in Old French/Middle French texts Inversion with a pronominal SU

Inversion with a full NP SU

Queste (1225)

97% (97/100)

98% (122/126)

Joinville (1306)

59% (24/41)

79% (50/63)

Froissart (c. 1375)

37% (15/41)

73% (33/45)

Quinze Joies (1420)

38% (17/45)

68% (27/40)

Jehan de Saintré (1456)

24% (9/37)

50% (30/60)

Commynes (1491)

15% (11/60)

73% (80/109)

of 10.6% (5/47) with initial time adjuncts (the latter context being less favourable to inversion in general; cf. section 4.1 above). In Anglo-Norman, inversion with pronominal subjects occurs even more robustly. All clauses with an initial object feature inversion (20/20), and among clauses with an initial time adjunct 60.6% (20/33) invert the verb and the subject pronoun. In summary, we can find entirely productive subject–verb inversion with pronominal subjects throughout Old and Middle French, and in particular also in thirteenth- and fourteenth-century Anglo-Norman. The innovative Middle English inversion word order with subject pronouns could therefore clearly have been calqued on Continental French/Anglo-Norman usage. Although such an account would seem plausible, other factors cannot be entirely excluded as elements contributing to the temporary rise of subject– verb inversion with pronominal subjects. Thus, Kroch, Taylor and Ringe (2000) argue that dialectal variation may have had an influence on the development of inversion in Middle English. This proposal is based on the hypothesis that, due to Scandinavian influence, a more systematic V2 syntax that featured inversion with both full NP subjects as well as pronominal subjects developed in the north and that properties of this northern grammar spread southwards through dialect contact. The evidence for Kroch et al.’s hypothesis is rather limited (a single text from the north with systematic V2), but some support for a dialectal component in the Middle English developments in inversion comes from Warner’s (2005) observation that the frequency of inversion with pronominal subjects is indeed somewhat higher in northern texts than in southern texts.6 Finally, language-internal factors may also have played a role. Within the analysis of the decline of inversion in Middle English outlined in Haeberli (2002b), it is proposed that the rise observed with subject pronouns is a side effect of a more general attempt by language learners to accommodate inversion patterns in the input that could no longer be derived (2002b: 104).

6



However, contrary to what one might expect, no obvious dialectal contrast seems to be found for inversion with full NP subjects.

155

AN-1a.indd 165

11/03/2010 12:00:27

The Anglo-Norman Language and Its Contexts (IV) why do authors vary so much in their use of subject–verb inversion? Turning finally to the variation across authors observed in Table 1, French once again seems to be a plausible source of explanation for certain patterns. In this case both the presence and the absence of French influence may be relevant, the former as a factor favouring inversion and the latter as a factor disfavouring inversion. Consider for example the data given for period m3 (1350–1420) in Table 1, repeated here in Table 4. Table 4. Main clauses with an initial constituent (except question words, negation, ‘then’) preceding subjects in Middle English texts from 1350–1420 m3 (1350–1420) Old Testament (a. 1425 (a. 1382)) New Testament (c. 1388) Purvey, Prologue to the Bible (c. 1388) Trevisa, Polychronicon (a. 1387) Wycliffite Sermons (c. 1400) Brut/Chronicles of England (c. 1400) Mandeville’s Travels (?a. 1425 (c. 1400)) Chaucer (Boethius, Melibee, Parson, Astrolabe; c. 1380/1390) Cloud of Unknowing (a. 1425 (?a. 1400)) Mirror of St Edmund, MS Vernon (c. 1390) TOTAL m3

0.0% (0/28) 0.0% (0/19) 1.1% (2/181) 2.7% (5/190) 44.1% (126/286) 22.0% (22/100) 34.0% (55/162) 71.0% (76/107)

3.0% (2/66) 10.2% (9/88) 20.8% (44/212) 25.6% (124/485) 40.5% (161/398) 68.0% (198/291) 80.1% (431/539) 73.6.% (265/360)

2.1% (1/47) 0.0% (0/103) 0.0% (0/25) 0.0% (0/48) 15.1% (13/86) 7.1% (6/85) 3.1% (1/32) 50.0% (95/190)

50.0% (6/12) 78.0% (64/82) 19.9% (42/211) 80.6% (29/36) 89.4% (59/66) 15.4% (23/149) 28.6% (321/1121) 52.5% (1357/2587) 18.5% (181/976)

As Table 4 shows, the frequencies of inversion are particularly low in the translations of the Old and the New Testament, in Purvey’s Prologue to the Bible, and in Trevisa’s Polychronicon. In clauses with a transitive verb and a full NP subject and in clauses with a subject pronoun, the rates of inversion are 0% or very close to 0%. Regular occurrences of inversion can only be found with full NP subjects and intransitive/unaccusative verbs. In other words, the pattern of inversion in these four texts is very close to that found in present-day English. This is strikingly different from the other texts included in Table 4, which still have frequent occurrences of inversion in any type of context. The question that arises then is why the four texts mentioned above should be so much more advanced in the loss of subject–verb inversion than the others? The bible translations referred to in Table 4 are the late versions of the Wycliffite Bible. This revision is often assumed to have been led or possibly even carried out by John Purvey, the author of the third text sample with a low frequency of inversion (Prologue to the Bible). Thus, three out of the four texts with a low rate of inversion can be attributed to Wyclif and his followers, and maybe more specifically to John Purvey. In this connection, the following observations by Berndt (1972: 348) seem relevant for our purposes: Growing vernacular-consciousness and a more critical attitude towards customs and conventions favouring the use of French for specific purposes

156

AN-1a.indd 166

11/03/2010 12:00:27

Investigating Anglo-Norman Influence on Late Middle English Syntax or on special occasions . . . are not only reflected in the writings of Higden, Holkot or Pseudo-Ingulph. . . . Scarcely different in essence are utterances to be found in the works of Wyclif and his followers in the later fourteenth century who passionately defend the use of English in religious writings (Angli debent de racione in isto defendere lingwam suam) and, in pointing out the right of Englishmen to have the Holy Scriptures translated into English, declare that . . . Crist tau(te his disciples oute þis prayer; but be þou syker, noþer in Latyn, noþer in Frensche, but in þe langage þat þey usede to speke, for þat þey knewe best.

Given this context, it would be plausible to argue that Wyclif and his followers’ writings reflect a variety of English that is close to the vernacular and disfavours the use of features transferred from French. One consequence of this could be the avoidance of inversion patterns that may have become obsolete in the vernacular but are maintained through French influence. Let us then turn to the fourth text with a particularly low rate of inversion in Table 4, John Trevisa’s translation of Ralph Higden’s Latin text Polychronicon. There are two elements that can be mentioned as possible explanations as to why this text patterns with the Wyclif texts. First, it has been suggested that Trevisa was influenced by Wyclif. Thus, Fowler (1993) repeatedly links Trevisa to Wyclif, as the following citations show: ‘It is, however, possible to argue that Trevisa was influenced by Wyclif. . . . Wyclif’s ‘favorite historian’ was Higden; Trevisa translated Higden’s Polychronicon . . .’ (Fowler 1993: 5); ‘As a whole, however, the list is not inconsistent with Caxton’s statement that Trevisa translated the Bible, and the thesis that he worked with John Wyclif . . . on a translation of the Bible during the 1370s’ (Fowler 1993: 17); ‘Moreover, Trevisa unmistakably if obliquely defends the translation of scripture in his ‘Dialogue between a Lord and a Clerk upon Translation’, prefixed to the translation of the Polychronicon (1387). A similar defence, in almost the same words, recurs in the preface to the later version of the Wycliffite Bible itself’ (Fowler 1993: 18). According to these observations, Trevisa might be included in the account given above in terms of attitudes to the vernacular among Wyclif and his followers. But there may be another source of influence on Trevisa in language matters, and that is the author of the work he translated, Ralph Higden. Thus, Berndt (1972: 348) points out that Higden was opposed to ‘not only the use of French in this country but also the use of French elements in English itself, the tendency of even the common people to “frenchify (francigenare)” their language, which he, like others of his contemporaries, considers a “corruption of the native tongue (nativae linguae corruptio)”’. Given the above observations, absence of inversion in certain late-fourteenth-century texts may be due to the authors’ attempts to favour the vernacular and resist external influences such as those exerted by French. However, there is one text whose syntactic properties do not quite fit into what I have proposed in the previous paragraphs. The Wycliffite Sermons show a relatively high frequency of subject–verb inversion even with transitive verbs (44.1%) and a rather lower but still substantial rate with subject pronouns (15.1%).

157

AN-1a.indd 167

11/03/2010 12:00:27

The Anglo-Norman Language and Its Contexts This is unexpected if inversion was indeed a residual pattern maintained through French influence in Late Middle English and if such influence was disfavoured by Wyclif and his followers. At present, I can only mention one feature of the Wycliffite Sermons that may be relevant for an explanation of their unexpected syntactic behaviour. Warner (1982: 18) points out that the Wycliffite Sermons ‘represent a variety of English which has been influenced by contact with Latin’. This suggests that the Wycliffite Sermons may not reflect vernacular usage to the same extent as other texts we have considered. However, the high frequency of inversion could not be straightforwardly accounted for in terms of Warner’s (1982) observation because Latin did not have productive V2. In contrast to the texts discussed so far, we can also find texts with rather high frequencies of inversion in the period 1350–1420. The Brut or The Chronicles of England, Mandeville’s Travels, Chaucer’s prose texts, The Cloud of Unknowing and the Mirror of St Edmund (ms. Vernon) have frequencies of inversion ranging from 22.0% to 80.6% with transitive verbs, from 68.0% to 89.4% with verbs other than transitives and from 3.1% to 50.0% with subject pronouns. French seems a plausible source of influence for at least some of these texts. Thus, the first part of The Brut or The Chronicles of England, from which the PPCME2 sample is taken, is a translation of the French Brut d’Engleterre (cf. Kroch and Taylor 2000). Similarly Mandeville’s Travels is an anonymous translation of a French work. French inversion patterns may therefore have been transferred to the English translation. However, while the frequencies of inversion with transitive verbs (22% and 34% respectively) and other verbs (68% and 80.1%) are rather high in these two texts, the French influence does not seem to be sufficiently strong to lead to high frequencies of inversion with subject pronouns (7.1% and 3.1%).7 Furthermore, with respect to Mandeville’s Travels, Kroch and Taylor (2000) point out that ‘[t]he translator writes very good English, but often misunderstands the French text’. Although this observation suggests that we are not dealing with a proficient bilingual here, it does not entirely exclude the possibility of French interference in the translation. But a detailed comparison of the French sources and the English texts may be needed to shed more light on the extent to which the inversion patterns found in English are influenced by those found in French. I will have to leave this for future research. French influence can also be argued to play a role in Chaucer’s writing. As is well known, Chaucer had close links to France and French. He had French family connections and spent time in France on government business (see for example Benson 1987a). Furthermore, as pointed out by Rothwell (1998), ‘It must not be forgotten that Chaucer the administrator was in contact with both Anglo-French and Anglo-Latin for many years in his varied daily work . . .’. Finally, among the prose texts considered in Table 4, we find one translation (Melibee) and two texts (The Parson’s Tale and Boethius) for the writing of which

7

This is somewhat different in a text from the period 1420–1500 that can also be linked to French sources. In Malory’s Morte Darthur, the frequency of inversion with subject pronouns is almost identical to that with transitive verbs (12.9% vs. 14.9%).

158

AN-1a.indd 168

11/03/2010 12:00:27

Investigating Anglo-Norman Influence on Late Middle English Syntax French sources were probably also used (cf. Benson 1987b, Hanna and Lawler 1987). Thus, it would not be implausible to assume that French influence contributed to the strikingly high frequencies of subject–verb inversion in any context in Chaucer’s work (frequencies of 71% (transitive), 73.6% (other), 50% (subject pronouns). An alternative or complementary account of Chaucer’s frequent use of subject–verb inversion is briefly suggested by Kroch and Taylor (1997: 324 n 16). Based on their claim, mentioned already in section 4.3, that subject–verb inversion was more systematic in northern varieties of English than in the south, they propose that ‘Chaucer’s syntax may be of a piece with his East Midlands phonology, since the East Midlands were part of the Danelaw’ and that ‘[h]is language may, therefore, indicate a certain conservative regionalism compared to the developing London standard’. However, as mentioned in section 4.3 (see note 6), the dialectal contrasts with respect to inversion in Middle English are not as clear-cut as one might wish, and it may therefore not be sufficient to relate Chaucer’s use of inversion entirely to his dialectal origin. French influence could therefore still be considered at least as a factor reinforcing the use of inversion. Having considered cases in which resistance to French or contact with French can be argued to play a role in the absence or presence of inversion, let us conclude this section by pointing out that some of the variation observed across different texts cannot easily be related to French in any way. Thus, for example the two remaining texts from Table 4, The Cloud of Unknowing and the Mirror of St Edmund, have very high rates of inversion (50% – 78% – 19.9% for the Cloud and 80.6% – 89.4% – 15.4% for the Mirror) but they do not have any obvious links to French. The Mirror is a translation from Latin, but as pointed out above in connection with the Wycliffite Sermons, it is not clear how Latin could have influenced the use of subject–verb inversion. Some aspects of the variation in inversion in Middle English texts therefore remain to be investigated in future work.

5. The et V construction In the previous sections, I have discussed issues related to the syntax of subject–verb inversion in Middle English and the way in which French may have contributed to the loss of some options (systematic inversion with ‘then’), the retention of some options (inversion with unaccusatives), and the extension of some options (inversion extended to subject pronouns). If contact with French had had a very strong influence on syntactic features of Late Middle English texts, one could potentially also expect to find instances where French introduced an entirely new option. We should therefore consider constructions that are licensed in Old and Middle French but not (or only very marginally) in Old and Early Middle English and see whether this construction emerges (or increases in frequency) in later Middle English.

159

AN-1a.indd 169

11/03/2010 12:00:27

The Anglo-Norman Language and Its Contexts A case in point would be the so-called ‘et V’ construction, a cross-linguistically unusual construction commonly found in Middle French and to some extent also in Old French. In this construction, illustrated in (7), subject–verb inversion occurs right after et (‘and’) in clauses with full NP subjects (example from Vance 1997: 48). (7) et dona li quens bone seurté que ja mes nel guerroieroit. La queste del saint Graal 120, 21 And gave the count good assurance that never neg-him would-wage-war ‘And the count gave good assurance that he would never wage war with him.’

This construction is not characteristic for Old English,8 so if French influence on Middle English syntax was important, one might expect ‘et V’ to be transferred with a certain frequency to Middle English. This expectation does not seem to be borne out. Out of around 7,200 nonnegative conjoined main clauses with an NP subject in the texts shown in Table 1, only 43 (0.5%) show the order ‘and V’. Thus, ‘and V’ looks like a very marginal construction in Late Middle English. We can even observe a weakening of this word order option in the Middle English period, as its frequency was higher in Early Middle English (60 out of around 1400 clauses (4.3%)). Thus, the conclusion seems to be that French influence was not strong enough to establish ‘and V’ as a type of inversion used in Middle English with a certain regularity. However, some additional observations should be made in this connection. Although the ‘et V’ construction was a common construction in Continental French, it was much less so in Anglo-Norman. In the Continental French texts from 1250 to 1350 examined by Ingham (2006b), inversion occurs in 63.5% (66/104) of all clauses introduced by et and containing an unaccusative verb and in 38.1% (40/105) of the clauses with other verbs. In Anglo-Norman texts from the same period, Ingham (2006b) finds much lower frequencies of inversion: with unaccusative verbs, the rate of inversion is 20.1% (29/144) whereas with other verbs, clauses introduced by et are never inverted (0/109). The overall frequency of inversion in Anglo-Norman is thus only 11.5%, compared to 50.7% in Continental French. Hence, assuming that the main source of influence on Middle English was Anglo-Norman rather than Continental French, the rarity of ‘and V’ in Middle English is less surprising. This construction may not have been sufficiently salient in Anglo-Norman for writers to transfer it to Middle English. The ‘et V’ construction is therefore not as useful as expected as a testing ground for establishing the level of French influence on Late Middle English syntax. At present, it is not clear whether any other construction could be considered that would allow us to see whether French influence was 8



Cases of subject–verb inversion in the absence of a clause-initial constituent (i.e. verb first, V1) do occur in Old English declarative clauses, most frequently when the verb is negative. However, this option is independent of the presence of a conjunction.

160

AN-1a.indd 170

11/03/2010 12:00:27

Investigating Anglo-Norman Influence on Late Middle English Syntax sufficiently strong to lead to the emergence of some new option in the clausal syntax of Late Middle English writers.

6. Conclusion Starting from the observation that the development of the syntax of subject– verb inversion in the Middle English period shows some unexpected features, I examined the hypothesis that these features could be accounted for in terms of language contact within the context of the multilingual situation found in Middle English and more particularly in terms of contact with French. The discussion has shown that French influence on Late Middle English syntax is plausible as a factor contributing to: (i) the loss of then as a distinctive trigger for inversion; (ii) the temporary increase of inversion with subject pronouns; (iii) variation among different Middle English texts. However, French influence seems at best a minor factor in the preservation of inversion in some very restricted contexts. Due to the fact that information on the exact sociolinguistic context in general and more specifically on the context in which individual texts were written is sparse, contact analyses are difficult to establish conclusively for developments occurring in medieval languages. However, to account for three out of the four syntactic issues addressed here, French influence seems as likely a hypothesis as others that have been proposed in the literature.

Bibliography Adams, M. (1987), ‘From Old French to the theory of pro-drop’, Natural Language and Linguistic Theory 5, 1–32. Allen, C. (1995), Case Marking and Reanalysis. Grammatical Relations from Old to Early Modern English. Oxford. ——(2006), ‘Case syncretism and word order change’, in A. van Kemenade and B. Los (eds.), The Handbook of the History of English, Oxford, pp. 201–23. Benson, L. (1987a), ‘Introduction’, in L. Benson (ed.), The Riverside Chaucer. Boston. ——(1987b), ‘The Canterbury Tales’, in L. Benson (ed.), The Riverside Chaucer, Boston, pp. 1–22. Berndt, R. (1972), ‘The period of the final decline of French in medieval England (fourteenth and early fifteenth centuries)’, Zeitschrift für Anglistik und Amerikanistik 20, 341–69. Bresnan, J. (1994), ‘Locative inversion and the architecture of universal grammar’, Language 70, 72–131. Fischer, O. (1992), ‘Syntax’, in N. Blake (ed.), The Cambridge History of the English Language. Volume II: 1066–1476, Cambridge, pp. 207–408. Fowler, D. (1993), John Trevisa. Aldershot. Haeberli, E. (2002a), ‘Observations on the loss of verb second in the history of English’, in C. J. W. Zwart and W. Abraham (eds.), Studies in Comparative Germanic Syntax, Amsterdam, pp. 245–72.

161

AN-1a.indd 171

11/03/2010 12:00:27

The Anglo-Norman Language and Its Contexts Haeberli, E. (2002b), ‘Inflectional morphology and the loss of verb second in English’, in D. Lightfoot (ed.), Syntactic Effects of Morphological Change, Oxford, pp. 88–106. ——(2002c), ‘Subjects in early English passives’. Paper presented at the 12th International Conference on English Historical Linguistics, University of Glasgow, 26 August 2002. Hanna, R., and T. Lawler. (1987), ‘Boece’, in L. Benson (ed.), The Riverside Chaucer, Boston, pp. 395–7. Ingham, R. (2005), ‘Bilingualism and syntactic change in medieval England’, Reading Working Papers in Linguistics 8, 1–26. —— (2006a), ‘La syntaxe en transition entre l’ancien et le moyen français’. Paper presented at Diachro-3, September 2006. Paris. ——(2006b), ‘Syntactic change in Anglo-Norman and Continental French chronicles: was there a ‘Middle’ Anglo-Norman?’, French Language Studies 16, 25–49. Kemenade, A. van. (1987), Syntactic Case and Morphological Case in the History of English. Dordrecht. ——(1997), ‘V2 and embedded topicalization in Old and Middle English’, in van Kemenade and Vincent (1997), pp. 326–51. ——and N. Vincent (eds.) (1997), Parameters of Morphosyntactic Change. Cambridge. Kristol, A. (2000), ‘L’intellectuel “anglo-normand” face à la pluralité des langues: le témoignage implicite du MS Oxford, Magdalen Lat. 188’, in D. Trotter (ed.), Multilingualism in Later Medieval Britain, Cambridge, pp. 37–52. Kroch, A., and A. Taylor. (1997), ‘Verb movement in Old and Middle English: dialect variation and language contact’, in van Kemenade and Vincent (1997), pp. 297–325. ————(eds.) (2000), The Penn–Helsinki Parsed Corpus of Middle English 2. University of Pennsylvania, Department of Linguistics. ———— and D. Ringe (2000), ‘The Middle English verb-second constraint: a case study in language contact and language change’, in S. Herring, P. van Reenen and L. Schoesler (eds.), Textual Parameters in Older Language, Amsterdam, pp. 353–91. Pintzuk, I. (1999), Phrase Structures in Competition. Variation and Change in Old English Word Order. New York. Roberts, I. (1993), Verbs and Diachronic Syntax: A Comparative History of English and French. Dordrecht. Rothwell, W. (1998), ‘Arrivals and departures: the adoption of French terminology into Middle English’, English Studies 79, 144–65. Schmidt, D. (1980), ‘A History of Inversion in English’, Doctoral dissertation, Ohio State University. Stockwell, R. (1984), ‘On the history of the verb-second rule in English’, in J. Fisiak (ed.), Historical Syntax, Berlin, pp. 575–59. Taylor, A., A. Warner, S. Pintzuk and F. Beths (eds.) (2003), The York–Toronto–Helsinki Parsed Corpus of Old English Prose. York. Available from the Oxford Text Archive. Vance, B. (1995), ‘Verb second and the null-subject parameter’, in A. Battye and I. Roberts (eds.), Clause Structure and Language Change, Oxford, pp. 173–99. —— (1997), Syntactic Change in Medieval French. Verb-Second and Null Subjects. Dordrecht.

162

AN-1a.indd 172

11/03/2010 12:00:27

Investigating Anglo-Norman Influence on Late Middle English Syntax Warner, A. (1982), Complementation in Middle English and the Methodology of Historical Syntax: A Study of the Wyclifite Sermons. London. ——(2005), ‘Establishing the dialectal distribution of V2 in Middle English’. Paper presented at the University of Leiden. Wurff, W. van der (1990), ‘Diffusion and Reanalysis’. Doctoral dissertation, University of Amsterdam.

163

AN-1a.indd 173

11/03/2010 12:00:27

chapter twelve

The Transmission of Later Anglo-Norman: Some Syntactic Evidence* Richard Ingham

1. The status of French in medieval England At issue in this article is the linguistic competence of later Anglo-Norman users: whether their output profiles them as L2 speakers whose French was subject to L1 interference, or whether they should be seen as balanced bilinguals whose French was not usually influenced by English. This question matters as regards the status of Anglo-Norman in relation to other varieties of French: many earlier authorities ghettoised Anglo-Norman, considering it with Bruneau (1955) to have been ‘une langue à part’ because of the status it had of a second-language variety. This perspective was also adopted by Kibbee (1996: 7–11), who emphasised the ‘essential difference’ between AngloNorman and Continental French, citing Anglo-Norman gender errors and ‘special syntactic constructions’ that reflect English rather than French. ‘By the 13th and 14th centuries, [Anglo-Norman] shows its imminent death by exhibiting the standard features of a dying language’ (Kibbee 1996: 15). He cited among these ‘standard features’ a tendency for syntax to align itself with that of the dominant language, English in this case. Earlier commentators, notably Vising (1923), Meyer (1889), Pope (1934) and Tanquerey (1916), also observed a decline in the quality of French from c. 1250 onwards, so one might conclude that by that time Anglo-Norman was no more than an imperfectly learned second language, and can be disregarded as a French dialect. Berndt (1972: 354) followed this trend, claiming that later Anglo-Norman was ‘clearly a language learned at school’. Yet in the field of Anglo-Norman studies the perils of uncritically following the assertions of earlier writers regarding the relationship between French and English in the medieval period have been made clear by Rothwell (1996), so some caution is in order. As Rothwell (2001) observed, if one followed the * I am very grateful for the comments and suggestions of David Birdsong, Ann Curry, Gwilym Dodd, Serge Lusignan, and Andres Kristol, who in their different respects have helped to focus my thinking on this piece of research. All errors and misinterpretations remain my own.

164

AN-1a.indd 174

11/03/2010 12:05:49

The Transmission of Later Anglo-Norman conventional textbook versions of the history of English, it is a surprise to find later Anglo-Norman existing at all, at least after 1362, since in that year the use of French in the law courts, supposedly its last remaining stronghold, is supposed to have been banned by royal decree. So it is by no means clear in what sense Anglo-Norman really was a dying language in the thirteenth to fourteenth centuries. Lusignan (2004) has pointed out that it was in the fourteenth century that Anglo-Norman seems to have experienced its period of most widespread use as a medium of communication in England: this development is surely hard to reconcile with the earlier conventional notion of degeneration into a poorly understood jargon used by speakers whose native language was English. The paradox that Insular French apparently flourished when its users were native speakers of English can perhaps be resolved if we recall that the debate on the status of French in England has usually been predicated on an assumption that Trotter (2006) has dubbed ‘serial monolingualism’. Since English was apparently learned as a native language even by the upper echelons of society, it was assumed by Berndt (1972), Richter (1979), Clanchy (1993), Kibbee (1996) and numerous others that French no longer had the status of a native language. But this does not follow: a logical alternative is that both languages were learned in native-like fashion by at least a significant section of the educated community. That position was in fact adopted by Legge (1980: 110) though few other commentators have endorsed her point, perhaps in view of the very broad claims she made for the extent of such bilingual attainment. Trotter (2003a: 239) has argued against attempts to isolate Anglo-Norman from the medieval French dialect continuum, citing as evidence of its functional vitality the extensive use of Anglo-Norman by English merchants for international correspondence in the fourteenth century.1 In his view Anglo-Norman belonged to a medieval Francophone dialect continuum, ranging from Anjou to the north-eastern regions of present-day France. Placing Anglo-Norman within the medieval French dialect continuum would normally imply that its speakers possessed native-speaker-like competence, rather than acquiring it as an instructed second language. If so, their divergences from Continental French ought not to be labelled as L2 errors. Rather, we should speak of AngloNorman as a variety of French in its own right, whose users observed the particular norms of that variety.2 Consequently, it would be appropriate to regard such speakers, who also spoke English in most cases, as bilingual. Recent research on bilingualism and second-language acquisition, summarised in section 2 below, demonstrates the ability of individuals to acquire 1



2



Other authors may also explicitly include Anglo-Norman among medieval French dialects, though, as with Togeby (1973), this may be because early Old French literature is mostly written in Anglo-Norman, rather than in recognition of AngloNorman as an everyday functioning variety of French. An example of one such feature would be the systematic predeterminer placement in Anglo-Norman of meïsmes (‘same’) which is seemingly not attested in Continental French, and is not calqued on any English equivalent.

165

AN-1a.indd 175

11/03/2010 12:00:27

The Anglo-Norman Language and Its Contexts more than one language to a native-like standard if language exposure takes place in early childhood. It is therefore important to note that contemporary references to the learning of French in England mention that it took place early in the lifespan. According to Trevisa’s translation of Higden’s Polychronicon, gentlemen’s children learned it ‘as they were rocked in their cradles’, Froissart (de Lettenhove 1863), commenting on problems encountered by English negotiators using French with their continental counterparts during the Hundred Years’ War, said that the French of France was unlike the French that the English had learned in their childhood (italics mine), and Bibbesworth’s Tretiz illustrates how the childhood learning process might have been fostered, at least among the nobility: (1) E quant il encurt a tele age Qu’i[l] prendre se poet a langage, E[n] fraunceis lui devez dire Cum primes deit sun cors descrivre . . . Bibbesworth, Tretiz, lines 21–4 (Rothwell 1990)

For Bibbesworth, writing in the later thirteenth century (see Rothwell, this volume), the learning of French began as soon as the child could manage language, which indicates childhood bilingualism, rather than the acquisition of French as a second language in school. Commentators may naturally wish to reserve judgement over the evidential value of authorial anecdote, or individual passages in surviving sources. But if we wish to learn something about how French was acquired in medieval England, we have available to us another source of information, and one of massive proportions: the documentary record of texts written in Insular French over several centuries, running into thousands of printed pages and many more unpublished works. How far can the quality of the French be considered to shed light on the issue of whether Anglo-Norman was genuinely a dialect of medieval French? Remarkably few detailed investigations of this question have been attempted, perhaps inhibited by the scholarly consensus that developed in the early twentieth century that took Anglo-Norman to be disqualified from that status on the grounds that it was an acquired second language. Yet, as noted by Rothwell (1996), when scholars have considered the quality of Anglo-Norman, they have often remarked on how close the language was to continental models, and how rare it is in practice to identify the many clear cases of English influence that would be expected in an imperfectly learned L2. Two recent studies of Anglo-Norman syntax (Ingham 2006a, b) found that in prose texts between about 1250 and 1350, Anglo-Norman followed particular grammatical changes taking place in Continental French. Verb–subject order was being lost after temporal adjuncts (Vance 1997), but not yet after preposed direct objects, and clitic object pronouns in non-finite clauses were switching from postverbal to preverbal position (De Kok 1985). No support

166

AN-1a.indd 176

11/03/2010 12:00:28

The Transmission of Later Anglo-Norman was found for English influence on Anglo-Norman in either of these grammatical domains. These results, taken from a wide range of later Anglo-Norman texts, gave credence to the continued vitality of Anglo-Norman as a medieval French dialect well beyond the time when, according to Pope (1934), Berndt (1972) and others, it supposedly lapsed into terminal decline. Yet they also prompted the present enquiry: given that Anglo-Norman did eventually meet its demise, we wished to know when the textual record of Anglo-Norman revealed syntactic traits that indicated unmistakably that it was an imperfectly learned L2, at which point the case for its membership of the medieval French dialect continuum could no longer be sustained. That Anglo-Norman diverged in various ways from Continental French is unquestioned. Brereton (1939) presented evidence from the later thirteenth century that Continental French speakers identified as Anglo-Norman traits the use of a les for aux, de les for des and que used instead of qui as a subject relative pronoun. The issue for us is whether this should be interpreted as divergence in the L2 acquisition sense (Sorace 2003), in which grammatical patterns of the L2 tend in some respects towards the L1, or rather the diatopic divergence of one L1 dialect from others. The presence or absence of interference from English, as proposed by earlier scholars, would seem to have a major bearing on the question. In an L2-learning scenario, L1 English would be expected to influence L2 syntax as a systematic determining influence on the grammatical system, which typically produces ‘imperfect second-language acquisition’. A striking instance of this can be seen in later-fifteenth-century Law French, where the use of French object pronouns mimics the English pattern, e.g.: (2) a. Le chaunceler adiourne eux en leschekere Chambyr  SCEC 147 (c. 1458) (2) b. Ieo ne barre luy pur ceo mater  SCEC 151 (c. 1458) (2) c. . . . sicome le Roy graunt a moy toutz lez finez et amercimentz deinz vne certeine lieu coment que le Roy voille pardoner eux [= ‘les autoriser’] SCEC 146 (c. 1458)

Object pronouns are here placed postverbally, as in Late Middle English. Law French writers committing such gross solecisms, in terms of French norms, were unquestionably influenced by the grammar of their L1. In this article we seek to clarify the status of later Anglo-Norman by considering what its syntax may reveal about the nature of the linguistic competence of those who used French for professional purposes in medieval England. We follow the suggestions by Togeby (1974: avant-propos) and Lusignan (2004: 158) that syntax may be a more accessible indicator of medieval language variation than for example dialect phonology. A diachronic investigation is undertaken into two key variables, noun-gender agreement and object-pronoun forms, using a range of late-thirteenth-century and fourteenth-century data, especially the very substantial database of parliamentary petitions entitled the Parliamentary Roll of Medieval England (PROME). These two syntactic variables are selected for their ability to discriminate between instructed

167

AN-1a.indd 177

11/03/2010 12:00:28

The Anglo-Norman Language and Its Contexts second language and native-like performance. It will be shown that in both these respects, the point at which clearly non-native-speaker-like grammatical trends emerged was in the 1370s. Up to then, departures from continental patterns affect forms which were vulnerable to loss of the relevant sound contrasts, rather than from the loss of a grammatical distinction. In the late fourteenth century, however, Anglo-Norman forms appear which manifestly neutralise key grammatical distinctions between masculine and feminine gender, and clitic and strong-form pronouns. We interpret these findings as further support for the position that Anglo-Norman was, until the mid-fourteenth century at least, a variety of the medieval French dialect continuum used in England by bilingual speakers, and not an instructed second language.

2. Methodology and sources In recent decades, studies of the linguistic competence of advanced secondlanguage learners have posed the question of what linguistic traits distinguish them from mother-tongue L1 speakers (see Sorace 2003, Birdsong 2004 and references therein). Findings are accumulating that suggest there are qualitative differences between highly fluent speakers of a language who learned it bilingually, at roughly the same time as another language, and those who did so after acquiring their L1. The former usually appear to show native-like grammar in both languages (Meisel 2004), whereas the latter, no matter how advanced their apparent ability in their L2, are susceptible to occasional errors in certain formal aspects of it (Johnson and Newport 1989). In fact, for native-like competence to be achieved, it may not be necessary for the child’s acquisition of both languages to start together, provided the second language follows the first during approximately the first five years of life. Prévost (2003) and Rothweiler (2006) found that successive acquisition of German as an L2 within the age range of 3–5 years resulted in native-like syntax with respect to verb finiteness and word order, whereas comparable studies of children who began later indicate a non-native-like pattern of development. Advanced L2 users make some errors with grammatical gender (Bartning 2000, Long 2003) and with object pronouns (Valenzuela 2006). Balanced bilingual speakers, however, are usually able to keep the grammatical systems of their two languages separate (De Houwer 1990), using gender-marking almost entirely consistently when acquiring French (Granfeldt 2005: 173). In the present article we seek to uncover evidence from later Anglo-Norman as to when native-like competence in grammatical variables gave way to clearly imperfect L2 learning of French in particular respects. The two domains selected for investigation are object-pronoun syntax and noun gender, two areas where Middle English patterned differently from Old French. Contemporary research findings that they are domains which discriminate advanced L2 learners from bilingual speakers make them a suitable choice for our present purposes.

168

AN-1a.indd 178

11/03/2010 12:00:28

The Transmission of Later Anglo-Norman To address these issues we have made use of an electronic corpus of data covering the last two centuries in which French was used as a language of public administration in England. The PROME database consists of exactly dated texts taken from manuscripts closely contemporary with the parliamentary sittings they record. The Anglo-Norman material contained in PROME represents a rather substantial time period, from the later thirteenth to the later fifteenth centuries. In addition it is stylistically fairly uniform, thus avoiding the problem encountered, if one samples texts of various styles, that apparent changes in language may simply be preferences in different stylistic registers that happen to come from different periods, rather than being genuine diachronic shifts in a linguistic system.

3a. Noun gender in later Anglo-Norman No detailed study of noun gender-marking in Anglo-Norman is known to us, and the editors of AND have not indicated noun gender as a fixed or variable property of entries, so direct comparison with Continental French usage cannot be made by this means. However, various studies mention that AngloNorman noun gender-marking was problematic. Tanquerey (1916: lv) found sporadic gender errors with determiners that were ‘dus a l’ignorance’ as he put it, as well as failures of adjective–noun gender agreement that he took to be either spelling mistakes or due to English influence. In his edition of the Shropshire Bills in Eyre (c. 1290), in which he commented adversely on the standard of French, Bolland (1914) drew attention in particular to gender errors, e.g. cele Thomas, cely Alice. Gender errors were also noted by Kibbee (1996: 10) as a factor reflecting English influence. Since Old English noun gender distinctions had largely disappeared by Early Middle English, L1 influence would be expected to have neutralised gender-marking, either by using a single form to replace the two members of a gender opposition, or by using the two forms indiscriminately. However, Anglo-Norman gender errors in the thirteenth century such as those reported above are ambiguous as to whether they are really what they initially seem. None of the sources quoted considers the point made by Pope (1934: 438) that Anglo-Norman showed attrition of the final schwa vowel in the thirteenth century. This may have been responsible for the use of apparent masculine determiner and adjective forms with feminine nouns in latethirteenth-century texts, e.g.: (3) a. en cel prison    Bolland 1914: 37 (3) b. sicum en acun rivere    Britton 404 (3) c. par certeyn enchesoun    Gippewyz, 22

Conversely, determiner and adjective forms with final –e can be found with masculine nouns, e.g.:

169

AN-1a.indd 179

11/03/2010 12:00:28

The Anglo-Norman Language and Its Contexts (4) a. de ceste trepas    Bolland 1914: 12 (4) b. de ceste escrit    Records of the Borough of Leicester 194 (4) c. en le haute chemin vers Donestaple    Tanquerey 1916: 50

If final –e spellings no longer corresponded to any audible distinction, their use to mark noun gender would have been vulnerable to confusion at the graphemic level. Apparent gender errors such as (3)–(4) may therefore not have been a matter of imperfect grammar, but rather the erosion of particular phonological means of making otherwise intact gender distinctions. This concept is already familiar in medieval French dialect phonology in the case of Picard, which neutralised the distinction between le/la, me/ma, etc. (Pope 1934: 488). In order to determine whether noun gender in Anglo-Norman was impaired independently of phonological factors such as the loss of final schwa, we therefore analysed a number of Anglo-Norman texts for gender agreement on pairs of premodifier forms having very distinct phonology, i.e. mon and ma, son and sa. These two forms, in addition to their differing vowel qualities, were distinguished by the nasal consonant in the masculine, which is absent in the feminine. Consequently, if gender errors cropped up here as well, this would surely form good evidence for imperfect grammar learning. Anglo-Norman gender-marking using these possessive forms was accordingly sampled in a range of texts, principally the PROME database, from which data are reported directly. We first carried out two pilot studies, however, using Bolland (1914) and a number of other texts of a legal character drawn from around 1300 and from early-sixteenth-century Law French. A possessive form was considered to be conventional if gender agreement with the noun it premodified was as indicated in a standard single-volume Old French dictionary (Godefroy 1901). Detailed findings of these initial explorations are reported at http://www.richardingham.com/id11.html. The main outcomes can be briefly summarised here. In Bolland (1914), where 41 tokens of son/sa were found, no gender errors at all occurred among these cases.3 For a text condemned by its editor for the inaccuracy of the French to have avoided gender errors so thoroughly with these forms seemed rather striking. To clarify the matter, further analysis of Anglo-Norman legal writings around 1300 was conducted, and comparison was made with early-sixteenthcentury Law French. A manual search was carried out on five legal French texts from the 1280–1320 period,4 and five texts representing Law French, c. 1500–30, which is known to show heavy interference from English.5 We 3





4



5

We excluded for this purpose natural gender examples, such as soun baron, sa feme, etc. which were also error-free. These were: the Britton (c. 1290), the Brevia Placitata (manuscripts from c. 1280–90), the EELR III (c. 1290), the Novae Narrationes (c. 1310), and manuscript Y of the Year Books (c. 1320), edited from manuscripts attributed to the late thirteenth and early fourteenth century. These were: Spelman’s and Port’s notebooks, Henry VIII Law reports I, and Readings and Moots II, extant in manuscripts of the first half of the sixteenth century and the

170

AN-1a.indd 180

11/03/2010 12:00:28

The Transmission of Later Anglo-Norman wished to see how far in each period the Continental French noun gender norms were observed. A rather simple sampling technique was used, taking the first ten items of each target gender class from each text, giving a total of 50 of each gender from each period. Nouns used only in Insular French were excluded, since they had no Continental French gender target, and also nouns beginning with a vowel letter to eliminate the possibility that a masculine form used here was a reflex of a liaison. Tokens of each noun were only counted once within each work, i.e. only the first mention of a noun type was counted. In the sixteenth-century material, gender errors were rampant, almost all showing the use of mon for ma or son for sa. Only four out of the fifty French feminine nouns (8%) were correctly marked with the feminine possessive; the other forty-six had the masculine possessive form. Occasional errors of ma with masculines were also observed, though most French masculine nouns had mon or son. By contrast, the late-thirteenth- to early-fourteenth-century works showed 100% correct gender-marking, just as we found with the Shopshire Bills in Eyre.6 This pilot exploration indicated that noun gender was very reliably known to Anglo-Norman professional writers around 1300, whereas with users of Law French two hundred years later, noun-gender distinction had almost completely collapsed. At some point along the way, then, the grammatical competence of insular users of French started to diverge from Continental French in favour of a system conforming to English, in which noun gender was not grammatically encoded. In order to gain a clearer idea of the time period within which this development took place, the PROME database was searched for the relevant variables. We began with the issue of gender neutralisation on gender-agreeing forms that were phonologically distinguished in Continental French only by final schwa. The data from the parliaments of Edward I and II in the PROME corpus (from the later thirteenth century up to 1326) show neutralisation of the un/une distinction: un enqueste, un espeye, un meyn, une bon boys, une payement,7 Similarly gender-marking is neutralised with cel/cele, and cest/ceste: cel ferme, cel foiz, cel prison, cele heritage, cele compassement, cest chose, cest fausete, ceste terme, ceste bail. Such examples, which are not uncommon, show that the forms stigmatised by Bolland (1914) as evidence of provincial ignorance of French were also found in the higher echelons of central government administration, beginning at roughly the same period. They show that even the most competent professional users of French seem to have experienced an erosion of the phonological means of keeping noun-gender-marking distinct in many contexts. 6





7

Henry VIII Yearbooks, from a printed edition of the late 1520s. Interestingly, Bibbesworth lines 25 ff. (Rothwell 1990) provides evidence that the learning of noun gender in French took place specifically through possessives mon, ma, son, sa etc., indicating that to do so via determiners or other premodifying forms had become unreliable, presumably because of phonological neutralisation. It should be noted that, as in Picard, neutralisation of the le/la distinction was also common: le guer, le garde, le comunaute, le eglise.

171

AN-1a.indd 181

11/03/2010 12:00:28

The Anglo-Norman Language and Its Contexts Turning now to the main focus of our enquiry, the preservation or otherwise of gender-marking on possessive forms, the PROME data were searched to reveal whether masculine nouns were used with son, and feminine nouns with sa. We investigated every occurrence of possessive forms son and sa, from the earliest data onwards. Until 1376, very few cases of gender-marking diverging from Continental French practice could be found. The three exceptions are illustrated by: (5) a. Qe la baillie ly soit livre a profit le roi et a son suffisante seurte. PROME 1328 ‘That the bailiwick be delivered to him to the king’s profit and with his sufficient guarantee.’ (5) b. Qe sa dit commune ad eu et soeffert en son temps pur plere et atteigner a son seignurie    PROME 1362 ‘Which his said commonalty has experienced and suffered in his time, in order to please and satisfy his lordship.’

In the late fourteenth century, more numerous gender errors occurred, e.g. the uses of son in the following: (6) a. Come s’il l’eust dit au roi de son propre bouche    PROME 1376 ‘As if he had said it to the king in his own words’ (6) b. Le dit William Elys par son fauxe suggescion fist arrester le dit Johan    PROME 1376 ‘The said William Ellis by his false accusation caused the said John to be arrested’ (6) c. Salvez ce q’al roi et a son coroune d’Engleterre appartint du saak de layne    October 1378 ‘Saving that which pertains to the king and his crown of England from the sacks of wool’ (6) d. Et Maistre Henry retournast en son dit chambre    PROME 1383 ‘And Master Henry went back into his room’

Errors of sa used with a masculine noun also occur in the later fourteenth century: (7) a. . . . estoit empeschez de sa voiage nadgaires fait vers les parties de Bretaigne    PROME 1376 ‘(And he) was . . . impeached concerning his voyage formerly made to parts of Brittany’ (7) b. Q’ils ont sustenu de bon gree en pes et guerre, pur luy et soun honur mayntener, depuis sa coronement,    PROME 1376 ‘That they have sustained willingly since his coronation in peace and war, to maintain him and his honour’

172

AN-1a.indd 182

11/03/2010 12:00:28

The Transmission of Later Anglo-Norman (7) c. Le roi nostre seignour ent eust donez \a vous/ sa congie pur departir.’    PROME 1378 ‘The king our lord has given you permission to depart’

In analyses of linguistic variation, it is generally desirable to provide information about the frequency of the relevant phenomena, and in principle the very extensive nature of the PROME database ought to lend itself very readily to variationist analysis. As regards the particular focus of enquiry in this study, however, there is a problem in that grammatical errors among highly proficient L2 users would not be expected to be at all frequent, so even if the hypothesis is correct that the clerks responsible for writing the petitions were advanced L2 learners rather than bilinguals, the frequency of errors as a proportion of all uses of the traits studied might still be quite small. Conventionally, what distinguishes native speakers from advanced L2 users is that under normal circumstances the former do not make grammatical system errors of the type in question here. On the very rare occasions when they do, such occurrences would be treated as inadvertent slips of the tongue or of the pen. On that assumption, an observable difference between advanced L2 users and bilingual users of Anglo-Norman would amount to a difference between a fairly low incidence of errors, and virtually no errors at all. This is not a scenario which lends itself to numerical analysis. Table 1. Frequency of gender errors with son and sa, PROME 1311–1400 Date of parliament*

Sa for son Son for sa

Total errors

No. of sessions

1311–1320 1321–1330 1331–1340 1341–1350 1351–1360 1361–1370 1371–1380 1381–1390 1391–1400







total

11

0 1 0 0 0 0 7 2 1



0 1 0 0 0 1 12 5 7

26

0 2 0 0 0 1 19 7 8

37

6 6 11 6 5 6 10 12 7

69

*The number of parliamentary sessions in a given decade for which PROME provides texts

Nevertheless, we recognise that it is useful to give some indication of the scale of the phenomenon being investigated. Accordingly, Table 1 shows the frequencies of the gender errors in PROME by decade, beginning in the early fourteenth century, where error is defined as above. In order to give a very rough indication of the comparative sizes of the data samples per decade, we have also shown the number of parliamentary sessions recorded in PROME;

173

AN-1a.indd 183

11/03/2010 12:00:28

The Anglo-Norman Language and Its Contexts the amount of text per session varied very considerably. The total number of uses of son8 and sa in the relevant period comes to around 5,000 tokens. It can be seen that in PROME gender-marking errors with son and sa are extremely rare events indeed until the last few decades of the fourteenth century, averaging just one in every twelve sessions. Thereafter, they become more common, with thirty-four errors occurring in the twenty-nine sessions between 1371 and 1399.

3b Object pronouns in later Anglo-Norman We next considered contexts for clitic object pronouns in PROME. The apparent divergence of Anglo-Norman from Continental French as regards the grammar of object pronouns is well-known, and was among the features noted by Brereton (1939): in later Anglo-Norman strong-form pronouns such as luy/ ly were often used preverbally, where the clitic le would be expected, e.g.: (8) Qi ly tuerent et les soenes    Scalacr. 10 ‘Who killed him and his men’

However, this contrast in the 3sg. was clearly vulnerable to phonological attrition in Anglo-Norman, since confusion also arose between le and lui/ly as definite articles, e.g.: (9) a. De cel hour an auaunt . . . fust ly roy counsaillez de William de Mountagow    Scalacr. 1069 (9) b. Le roy repaira a Andwerp    Scalacr. 126

Similarly, the pronunciation of pronoun forms spelt me, se etc. versus those spelt moi/mei, soi/sei, etc. may have been less than fully distinct in Anglo-Norman, which would again have affected the clitic/strong pronoun distinction. As with gender-marking, then, we have to be wary of taking as a grammatical problem what could well have been a matter of phonological neutralisation. For the purpose of the present analysis, therefore, we used only the 3pl. pronoun forms les and eux (the latter also spelt eus euls eaux), where no phonological assimilation could plausibly have occurred. If these two forms were nevertheless confused by Anglo-Norman users, this would unquestionably constitute a grammatical divergence from Continental French. Results were as follows. Until the last quarter of the fourteenth century no use of eux in a clitic context was found, but thereafter several such cases cropped up. The first of these featured eux used for indirect object lour:10

8 9

10



As well as tokens of the variant spelling soen. Ly here is not plausibly to be treated as an Old French nominative case article, at this late date, especially since Old French case distinctions were lost early in Anglo-Norman. Use of les for leur as an indirect object pronoun was not uncommon in Anglo-

174

AN-1a.indd 184

11/03/2010 12:00:28

The Transmission of Later Anglo-Norman (10) Et la dite commissioun eux serroit delivere voluntiers. PROME 1376 ‘And the said commission would be delivered to them willingly.’

The others however, showed eux as a direct object clitic, e.g.: (11) a. Les ditz burgeys et comminaltee ensemblerent en grantz routes . . . et eux amesnerent au dite ville  PROME 1381 ‘The said burgesses and commonalty assembled in great bands . . . and they led them into the said town’. (11) b. Et le roi les resceut, et eux deschargea des ditz offices PROME 1390 ‘And the king received them and discharged them from the said offices’ (11) c. Et eux ajourna a Salopesbury en mesme le parlement PROME 1397 ‘And (he) adjourned them to Shrewsbury in the same parliament.’

Eux was also used preverbally in place of the reflexive clitic se: (12) a. Les mair, baillifs, burgeys, et comminaltee du dite ville, entour .x. de la clokke le noet ensuant, eux assemblerent al tolbothe   PROME 1381 ‘The mayor, bailiffs, burgesses, and commonalty of the said town assembled at the tollbooth around 10 o’clock the following night’ (12) b. Sibien le dit Johan come le dit suppliant eux submistrent en haut et bas en la tresdreite ordinance    PROME 1391 ‘Both the said John and the said supplicant submitted themselves entirely to the most rightful ordinance.’

These instances clearly foreshadowed the strong-pronoun forms noted above in fifteenth-century Law French, at least in morphological terms, though not syntactically, in that the pronoun was still positioned preverbally. Taking these findings together with the Law French late-fifteenth-century data, the process involved seems thus to have been first a decliticisation of object pronouns, and then a shift in position to postverbal placement. Incipient decliticisation of object pronouns in the late fourteenth century is further attested in PROME by the separation of a clitic pronoun from the verb, e.g.: (13) a. Et de les sauvement remesner devant les ditz seignours en le dit parlement.    PROME 1377 ‘And to bring them safely back before the said lords in the said parliament’ (13) b. Quele chose les ditz justices lui outrement ount denyez. PROME 1384 ‘Which the said justices have utterly denied him.’ Norman, as in Northern Old French.

175

AN-1a.indd 185

11/03/2010 12:00:29

The Anglo-Norman Language and Its Contexts (13) c. Qe le meir, viscontz, et aldermen . . . facent redresser et corriger les errours . . . et les duement punir de temps en temps    PROME 1394 ‘That the mayor, and aldermen should have these errors redressed and corrected . . . and duly punish them from time to time.’

Such uses of a clitic pronoun were completely impossible in any other medieval French variety, to the best of our knowledge. Correspondingly, prepositional datives in main clauses are now encountered placed preverbally, as if they were clitics, e.g.: (14) Et par especial il demandast suretee de la paix de monsire Michel de la Pole, chanceller d’Engleterre; et celle requeste a lui fuist grantez. PROME 1384 ‘And in particular he demanded surety of the peace from Sir Michael de la Pole, chancellor of England; and this request was granted him’

This surface order was possible in Old French subordinate clauses where subject pronoun cliticised to a complementiser, and the verb stood in a late position in the clause (Vance 1997), but here it is found in a main clause, where the subject and verb could not normally be separated by an intervening constituent in Old French prose. In several respects, therefore, 3pl. pronoun usage by the late fourteenth century was showing signs that fundamental distinctions made by the grammar of French were being ignored, in areas where phonological neutralisation cannot have been a factor.

4. Discussion This examination of pronoun forms and gender agreement in the PROME database has been deliberately restricted to forms where a shift away from the norms of Old French would be observable independently of phonological factors. It has found that a qualitative shift took place later on in the fourteenth century, showing decliticisation of preverbal object pronouns, and the spread of noun-gender errors, even with phonologically quite distinct forms which had earlier respected masculine and feminine gender-marking. It seems significant that non-conventional 3pl. pronoun uses were not found until the same period as when gender errors with possessives started to proliferate, that is, the last decades of the fourteenth century. The simultaneous appearance of these two error types is hard to dismiss as coincidence. It is true that what appear to be gender-marking errors with possessives were very occasionally found in the Rolls before the later fourteenth century. However, it must be borne in mind that variation in noun gender is found in Old French as well; compare these thirteenth-century examples from the Miracles de Notre Dame taken from the Textes de français ancien database:

176

AN-1a.indd 186

11/03/2010 12:00:29

The Transmission of Later Anglo-Norman (14) a. Ainc nus ne t’eut en sa memoire  Miracles ND, p. 112 (14) b. Et tant l’ avoit en son memoire  Miracles ND, p. 148

Sometimes the same sources show what may just be a slip, cf.: (15) . . . mainte riche ensaigne en son lance levee.  Buevon, p. 55

This author otherwise uses lance as a feminine noun, e.g.: (16) Ascuns des chevaliers tint sa lance en sa main.  Buevon, p. 44

Presumably a very low level of non-standard gender-marking in medieval texts must be allowed for without supposing that the writer was a second-language speaker of French. But the sudden expansion of non-standard gender-marking in PROME in the 1370s seems to rise above this baseline. These findings have implications for the timing of a change in the status of French in medieval England. They suggest that, until the later fourteenth century, French was transmitted to Francophone clerks in a milieu in which use of the relevant grammatical features did not significantly depart from Continental French norms. In psycholinguistic terms, their language knowledge would have been stored and processed in such a way that it was not influenced by English. Their competence in French was thus that of a balanced bilingual, at least in the respects investigated in this study. By the late fourteenth century, that was plainly no longer the case: the grammar of gender-marking and object-pronoun use was showing signs of convergence with English, which did not make distinctions between masculine and feminine noun gender, and clitic versus strong-form pronouns. We interpret this as indicating that their syntactic competence in French appears to have been more at the level of an advanced, but not native-like, L2 learner who was to an extent influenced by English. Up to that period, the most expert users of French, at least, seem to have been balanced bilinguals, whereas after it they were not. It is readily acknowledged that considerable variation in proficiency of Anglo-Norman users must have existed, depending on the professional social status of the speaker. Accordingly, caution is in order in generalising the results from PROME beyond the professional community of clerks who produced the texts examined. Nevertheless, it is the persistence of French into the fourteenth century precisely among this community, long after the native language of all Englishmen outside the royal family had become English, that gives rise to the paradox mentioned earlier as regards the long survival of French. So our findings clearly address the problem in the domain where it arises. There might be a temptation to represent central-government clerks using Anglo-Norman as a cadre of language specialists, perhaps comparable to modern-day interpreters and translators, and thus to question their significance in terms of the nature of French used in England more generally. Granted, the existence in any society of a body of professionals whose job it is to learn and use a foreign language near-perfectly will tell us rather little about

177

AN-1a.indd 187

11/03/2010 12:00:29

The Anglo-Norman Language and Its Contexts ordinary language use in that society. However, it is hard to see the French of the Anglo-Norman petitions as a matter of foreign-language expertise, given that it is in numerous respects (e.g. morphologically and lexically) very distinct from Continental French and abounds in forms that were peculiar to the insular context. As Trotter (2003a) has pointed out, Anglo-Norman clerks were well aware of differences between Anglo-Norman and Continental French usage, and were able to modify their language to suit the provenance of continental addressees. They represented the educated literate class of their day, who received their training largely with regard to Latin (Orme 1973). So to consider the Anglo-Norman used by the clerks responsible for the documents in PROME as representing nothing beyond the work of an elite of professional linguists specialised in French would in our view be quite misconceived.

5. Conclusions Our main findings may be summarised by the observation that, in the grammatical respects studied here, the standard of French in the Rolls data shows no evidence of divergence from Continental French norms until the last quarter of the fourteenth century. Up to that point, Anglo-Norman petitions respected Continental French gender distinctions and object-pronominal distinctions in a virtually error-free fashion, where grammatical markers were not vulnerable to phonological attrition. This finding casts doubt on the view that, long before that point, Insular French was essentially different from Continental French in that it was an instructed second language. The grammatical accuracy of the petition writers on the variables studied here was such that French seems more likely to have been acquired naturally in a milieu where it had the status of a spoken vernacular, i.e. ‘absorbed’ rather than taught, to quote Legge (1980) again. Many areas of grammar could no doubt be considered as indicative of linguistic competence, but the two we chose were selected as areas that have been shown to distinguish bilinguals from advanced L2 learners. They allow us to sustain the position that writers of the material analysed had acquired French as bilinguals. In psycholinguistic terms, their grammatical knowledge would have been stored and processed in such a way that it was not influenced by English. In the respects investigated in this study, their competence in French was thus that of an early-childhood bilingual. The numerous non-continental features observable in their French were thus dialect features and were thus not errors, but were conventional in the variety in question. The position we take is controversial, we recognise, in the light of earlier claims to the effect that French was by the later thirteenth century not a vernacular but an instructed second language. However, no previous study to our knowledge has undertaken a close study of grammatical features that would allow that hypothesis to be convincingly tested. We have found that the data analysed here do indeed provide support for it, but they do not do so until

178

AN-1a.indd 188

11/03/2010 12:00:29

The Transmission of Later Anglo-Norman the late fourteenth century, when the competence in French of the petition writers appears to have been that of advanced, but not native-like, L2 learners: they were not reliably acquiring grammatical gender as an inherent feature of a noun, and were showing some difficulty with the grammatical distinction between clitic and strong-form object pronouns. In both respects we may see the L1 grammar as having exerted an influence in that Middle English did not categorise nouns for grammatical gender, and did not have the clitic/strongform distinction with object pronouns. This outcome suggests that a significant change took place in the circumstances in which French was transmitted in England in the second half of the fourteenth century, such that balanced bilingual English–French users became extinct. This sort of date would fit well with the dates of a number of manuscripts studied by Kristol (1990) containing materials used for the teaching of French, and with the evidence of the activities from the 1370s onwards of the teachers of the ars dictaminis at Oxford (Richardson 1942). It accords also with the recent study by Lusignan (2006) of the use of French in diplomatic treaties between France and England. At the treaty of Brétigny in the 1360s the English used French for purposes of negotiation without demur. In 1393, at the peace of Leulinghem, however, they felt that they were at a disadvantage when French was so used. A further conclusion that the results on the maintenance of noun gender seem to point to is that French was maintained in the fourteenth century as a spoken language. The syncretism between phonologically close gender exponents, and non-syncretism between phonologically more distant ones, would be hard to account for if French had been confined to a largely written mode. We do not draw any conclusion from these findings as regards the social spread of the use of French in England in the fourteenth century. To the extent that the users of Anglo-Norman were represented by the Rolls, the picture we get is one of surprising accuracy in grammatical variables – surprising, at least, in terms of certain pronouncements in the earlier literature. Very little work has been done on the everyday French of the fourteenth century, though Wright (2002, this volume) and Trotter (2006, this volume) have made some inroads into this problem, with regard to accounts and similar documents. More work is required, with respect to a range of genres of later Anglo-Norman, before a more definitive assessment of the position can be established.

Primary sources insular french Bolland, W. (ed.), Select Bills in Eyre 1292–1333. Selden Society 30. London, 1914. Britton: The French Text Carefully Revised with an English Translation, Introduction and Notes, ed. by F. Nichol. Oxford, 1865.

179

AN-1a.indd 189

11/03/2010 12:00:29

The Anglo-Norman Language and Its Contexts Gippewyz: Le Domesday de Gippewyz, in The Black Book of the Admiralty, ed. T. Twiss, vol. I. Rolls Series. London, 1871. Henry VIII law reports I: Reports of Cases from the Time of King Henry VIII, vol. I, ed. J. Baker. Selden Society 120. London, 2003. Henry VIII year books: Year Books of Henry VIII, 12–14 Henry VIII, ed. J. Baker. Selden Society 119. London, 2002. Novae Narrationes, ed. E. Shanks, completed by S. Milsom. Selden Society 80. London, 1963. Port’s Notebook, ed. J. Baker. Selden Society 102. London, 1986 Readings and Moots at the Inns of Court in the Fifteenth Century, vol. II, ed. S. Thorne and J. Baker. Selden Society 105. London, 1990. Records of the Borough of Leicester, ed. M. Bateson, vol. I. London, 1899. Reports of Sir John Spelman, vol I. Selden Society 93. London, 1977. Scalacr.: Sir Thomas Grey, Scalacronica 1272–1363, ed. A. King. Surtees Society 209. Cambridge, 2005. SCEC: Select Cases in the Exchequer Chamber 1377–1461, ed. M. Hemmant. Selden Society 51. London, 1933. Year Books 1320 Ms Y: Year Books of Edward II, 14 Edward II Michaelmas 1320, ed. G. Turner and W. Bolland. Selden Society 104. London, 1988.

continental french Buevon: Adenet le Roi, Buevon de Conmarchis, in Les oeuvres d’Adenet le Roi, ed. Albert Henry, vol. II. Bruges, 1953. Miracles ND: Treize miracles de Notre-Dame, ed. Pierre Kunstmann. Ottawa, 1981. Textes de français ancien: online database at http://www.uottawa.ca/academic/ arts/lfa/

Bibliography Bartning, I. (2000), ‘Gender agreement in L2 French: pre-advanced versus advanced learners’, Studia Linguistica 54/2, 224–37. Berndt, R. (1972), ‘The period of the final decline of French in Medieval England (fourteenth and early fifteenth centuries)’, Zeitschrift für Anglistik und Amerikanistik 20, 341–69. Birdsong, D. (2004), ‘Second language acquisition and ultimate attainment’, in A. Davies and C. Elder (eds.), Handbook of Applied Linguistics, Oxford, pp. 82–105. Brereton, G. (1939), ‘Some grammatical changes made by two revisers of the Anglo-Norman version of Des grantz geantz’, in Studies in French Language and Mediæval Literature Presented to Mildred K. Pope by Pupils, Colleagues and Friends. Manchester. Bruneau, C. (1955), Petite histoire de la langue française, vol 1: Des origines à la Révolution. Paris. Clanchy, M. (1993), From Memory to Written Record. England 1066–1307. Cambridge, Mass. Doughty, C., and M. Long (eds.) (2003), Handbook of Second Language Acquisition, Oxford.

180

AN-1a.indd 190

11/03/2010 12:00:29

The Transmission of Later Anglo-Norman Godefroy, F. (1901), Lexique de l’ancien français. Paris. Granfeldt, J. (2005), ‘The development of gender attribution and gender agreement in French’, in J.-M. Dewaele (ed.), Focus on French as a Foreign Language: Multidisciplinary Approaches. Clevedon. De Houwer, A. (1990), The Acquisition of Two Languages from Birth: A Case Study. Cambridge. Ingham, R. (2006a), ‘Syntactic change in Anglo-Norman and Continental French chronicles: was there a ‘Middle’ Anglo-Norman?’, Journal of French Language Studies 16/1, 25–49. ——(2006b), ‘The status of French in medieval England: evidence from the use of object pronoun syntax’, Vox Romanica 65, 1–22. Johnson, J., and E. Newport (1989), ‘Critical period effects in second language learning: the influence of maturational state on the acquisition of English as a second language’, Cognitive Psychology 21/1, 60–99. Kibbee, D. (1996), ‘Emigrant languages and acculturation: the case of AngloFrench’, in H. Nielsen and L. Schøsler (eds.), The Origins and Development of Emigrant Languages, RASK Supplement 6, Odense, pp. 1–20. Kristol, A. (1990), ‘L’enseignement du français en Angleterre (XIIIème–XVème siècles): les sources manuscrites’, Romania 111, 289–330. Legge, M.-D. (1980), ‘Anglo-Norman as a spoken language’, in R. Brown. (ed.), Anglo-Norman Studies 2, Woodbridge, pp. 108–17. de Lettenhove, K. (ed.) (1863), Le premier livre des Chroniques. Texte inédit publié d’après un manuscrit de la Bibliothèque du Vatican. Paris. Long, M. (2003). ‘Stabilization and fossilization in interlanguage development’, in Doughty and Long (2003), pp. 487-536. Lusignan, S. (2004), La langue des rois au moyen âge: le français en France et en Angle­ terre. Paris. ——(2006), ‘Parler en français: altérité et identité des anglais et des français à la fin du moyen âge’. Talk given Nov. 2006, Germany. Meisel, J. (2004), ‘The bilingual child’, in T. Bhatia and W. Ritchie (eds.), The Handbook of Bilingualism, Oxford, pp. 91–113. Orme, N. (1973), English Schools in the Middle Ages. London. Pope, M. (1934), From Latin to Modern French, with Especial Consideration of AngloNorman: Phonology and Morphology. Manchester. Prévost, P. (2003), ‘Truncation and missing inflection in initial child L2 German’, Studies in Second Language Acquisition 25, 65–97. Richardson, H. (1942), Letters of the Oxford Dictatores, Oxford History Society, ns 5, 360–416. Richter, M. (1979), Sprache und Gesellschaft im Mittelalter. Stuttgart. Rothweiler, M. (2006), ‘Early successive acquisition of German’, in C. Lleó (ed.), Interfaces in Multilingualism, Amsterdam, pp. 91–113. Rothwell, W. (1990), Walter de Bibbesworth, Le tretiz. Anglo-Norman Text Society, Plain Texts Series 6. London. ——(1996), ‘Playing follow my leader in Anglo-Norman studies’, Journal of French Language Studies 6, 177–210. ——(2001), ‘English and French in England after 1362’, English Studies 82, 539–59. Sorace, A. (2003), ‘Near-nativeness’, in Doughty and Long (2003), pp. 130–51. Tanquerey, F. (1916), Lettres anglo-françaises. Paris.

181

AN-1a.indd 191

11/03/2010 12:00:29

The Anglo-Norman Language and Its Contexts Togeby, K. (1974), Précis historique de grammaire française. Copenhagen. Trotter, D. (2003a), ‘Not as eccentric as it looks: Anglo-French and French French’, Forum for Modern Language Studies 39, 427–38. ——(2003b), ‘L’anglo-normand: variété insulaire ou variété isolée?’, Médiévales 45, 43–54. Valenzuela, E. (2006), ‘L2 end-state grammars and incomplete acquisition of Spanish CLLD constructions’, in R. Slabakova, S. Montrul and P. Prévost (eds.), Inquiries in Linguistic Development: In Honour of Lydia White, Amsterdam, pp. 283–304. Vance, B. (1997), Syntactic Change in Medieval French. Dordrecht. Vising, J. (1923), Anglo-Norman Language and Literature. London. Wright, L. (2002), ‘Code-intermediate phenomena in medieval mixed-language business texts’, Language Sciences 24, 471–89.

182

AN-1a.indd 192

11/03/2010 12:00:29

INDEX

Agriculture 44 Benskin, M. 106–7, 110 Bergounioux, G. 30 Berndt, R. 156–7, 164 Bibbesworth, W. de 47–50 50–1, 118, 166, 171 Bilingual, bilingualism 9–10, 22, 24, 50, 58–60, 88–9, 116, 128, 150, 158, 162–6, 173, 177–8, 181 Bolland, W. 169–71 Borrowing 11, 55, 59–60, 68 Bozon, Nicole 47 Brand, P. 6, 94, 97, 101 Brereton, G. 167, 174 Bruneau, C. 21, 164 Brunot, F. 29 Buridant, C. 10, 18 Cerquiglini, B. 30–2 Charles VI, king of France 77 Chaucer, G. 109, 158 Chrétien de Troyes 69 Clark, C. 120 Clyne, M. 12 Conjugation 16–17 Conon de Béthune 33 Contact 9ff., 34, 42, 53, 61, 67, 75, 78, 88, 94, 118, 139–41, 143, 145, 150, 155, 159, 162 Convergence 12 Correctness 49 Coutant, Y. 52–3 Curry, A. 6, 76, 77, 81, 83, 85, 89 Dees, A. 33 Definite article (French) 60, 129–40 Dialect variation 31, 86, 162 Dictionaries 50, 63–7, 113, 135, 139, 141, 170

Divergence (of Anglo-Norman from continental French) 8, 102, 167, 171, 176 Edward I, king of England 74, 78, 94–5 Edward II, king of England 97 Edward III, king of England 75, 83 Fondet C. 30 Foulet, L. 18 Froissart, J. 75–6, 166 Gender 13, 131, 168–77 Godefroy, F. 50, 70, 170 Gossen, C.-T. 31–2 Grosseteste, R. 44, 47 Haeberli, E. 5, 145, 152 Haugen, E. 28–9, 33 Hawkwood, Sir J. 85 Hemricourt, J. de 14 Henry III, king of England 94 Henry IV, king of England 85–6 Henry V, king of England 77, 82, 84 Henry VI, king of England 77, 86 Higden, R. 108, 120, 157 Hunt, T. 50, 102, 117 n. 36 Imperfect tense 17 Indentures 80 Ingham, R. 5, 7–8, 143, 149–51, 153, 160, 162, 164 Insular French lexis and English 66 Intermarriage 11, 88 Isoglosses 28 Jefferson, L. 65, 68 Jones, M. 12 Kibbee, D. 8–9, 164–5, 169 Koineisation 35

183

AN-1a.indd 193

11/03/2010 12:00:29

The Anglo-Norman Language and Its Contexts Kristol, A. 76, 149 Kroch, A. 151, 155, 158 Lacy, H. de, 47, 49, 118 Laing, M. 112, 115 Language acquisition 88–9, 166–8, 173, 177 Latin 15, 21, 35. 45ff., 50, 53ff., 68–9, 75, 78–81, 84, 89, 95, 100, 104, 106, 128, 130, 132, 178 Law reports 96 Legge, M.-D. 108, 165, 178 Letters 85 Lexis 22, 46, 64–5, 67, 98, 105, 120 Loanwords 58, 63–4, 118–19 Lodge, A. 2, 26, 33 Lusignan, S. 77, n. 13, 81, n. 31, 86, 165, 167, 179 Mannyng, R. 70 Meyer, P. 49–50 Middle English 1, 22, 54, 63, 103, 133– 49, 157, 160 Mixed language 4, 53, 70, 130–42 Modal perfect 20 Molencki, R. 20 Musters 79 National identity 75, 84 Old English 65, 137, 143–4, 148, 151–2, 154, 160, 169 Orr, J. 22 Oschinsky, D. 44–6, 48 Owen, A. 48–50 Owen-Crocker, G. 63, 64 n. 1 Paris , G. 28–30 Penny, R. 39 Philip VI, king of France 75 Pope, M. 8, 14, 23, 115, 164, 167, 169 Pouzet, J.-P. 3, 102 Prestige 1–2, 11, 20–2, 26, 33, 55 Price, G. 8

Pronouns 5, 10, 14–16, 144, 148, 150–1, 155–6, 159, 167–8, 174–6, 178–9 Purvey, J. 156 Richard II, king of England 84 Rothwell, W. 2–3, 9, 44, 64–5, 67, 109–10, 114 n. 31, 149, 158, 164, 166 Scandinavian 145, 155 Shift, language 11 Short, I. 2, 13 Standard language, standardisation 26–8, 30–1, 41, 48, 81, 90, 102, 108, 159, 169, 177 Stevens, J. 85–6 Suggett, H. 77, 81 Syntax 2, 5, 10, 14, 143–5, 150, 152–5, 159, 160–1, 164 166–8 Tanquerey, F. 169 Taxation 83–4 Taylor, A. 151, 155, 158 Thiolier, J.-C. 118 Thomas, duke of Clarence 83 Tiptoft, Sir J. 81 Trevisa, J. 76, 86, 98, 120, 157 Trotter, D. 3–4, 9, 52, 64 n. 1, 65, 118 n. 38, 132 n. 1, 140, 165, 178–9 Vance, B. 151, 153, 160, 166, 176 Vernacular 4, 11, 28, 30, 32–4, 39, 50, 60, 71, 76, 103, 106, 108–11, 114, 118, 131, 156–8, 178 Vising, J. 105 n. 11, 111, 164 Völker, H. 34 Walloon 14–15 Walter of Henley 44–51 Warner, A. 158 William the Conqueror 77 Winford, D. 11 Wright, L. 3–4, 53, 130, 179 Writing 26–7, 29–34, 39, 41, 57, 80, 83, 112–14, 130 Wyclif, J. 156–8

184

AN-1a.indd 194

11/03/2010 12:00:29

YORK MEDIEVAL PRESS: PUBLICATIONS God’s Words, Women’s Voices: The Discernment of Spirits in the Writing of Late-Medieval Women Visionaries, Rosalyn Voaden (1999) Pilgrimage Explored, ed. J. Stopford (1999) Piety, Fraternity and Power: Religious Gilds in Late Medieval Yorkshire 1389–1547, David J. F. Crouch (2000) Courts and Regions in Medieval Europe, ed. Sarah Rees Jones, Richard Marks and A. J. Minnis (2000) Treasure in the Medieval West, ed. Elizabeth M. Tyler (2000) Nunneries, Learning and Spirituality in Late Medieval English Society: The Dominican Priory of Dartford, Paul Lee (2000) Prophecy and Public Affairs in Later Medieval England, Lesley A. Coote (2000) The Problem of Labour in Fourteenth-Century England, ed. James Bothwell, P. J. P. Goldberg and W. M. Ormrod (2000) New Directions in Later Medieval Manuscript Studies: Essays from the 1998 Harvard Conference, ed. Derek Pearsall (2000) Cistercians, Heresy and Crusade in Occitania, 1145–1229: Preaching in the Lord’s Vineyard, Beverly Mayne Kienzle (2001) Guilds and the Parish Community in Late Medieval East Anglia, c. 1470–1550, Ken Farnhill (2001) The Age of Edward III, ed. J. S. Bothwell (2001) Time in the Medieval World, ed. Chris Humphrey and W. M. Ormrod (2001) The Cross Goes North: Processes of Conversion in Northern Europe, AD 300–1300, ed. Martin Carver (2002) Henry IV: The Establishment of the Regime, 1399–1406, ed. Gwilym Dodd and Douglas Biggs (2003) Youth in the Middle Ages, ed. P. J. P. Goldberg and Felicity Riddy (2004) The Idea of the Castle in Medieval England, Abigail Wheatley (2004) Rites of Passage: Cultures of Transition in the Fourteenth Century, ed. Nicola F. McDonald and W. M. Ormrod (2004) Creating the Monastic Past in Medieval Flanders, Karine Ugé (2005) St William of York, Christopher Norton (2006) Medieval Obscenities, ed. Nicola F. McDonald (2006) The Reign of Edward II: New Perspectives, ed. Gwilym Dodd and Anthony Musson (2006) Old English Poetics: The Aesthetics of the Familiar in Anglo-Saxon England, Elizabeth M. Tyler (2006) The Late Medieval Interlude: The Drama of Youth and Aristocratic Masculinity, Fiona S. Dunlop (2007) The Late Medieval English College and its Context, ed. Clive Burgess and Martin Heale (2008) The Reign of Henry IV: Rebellion and Survival, 1403–1413, ed. Gwilym Dodd and Douglas Biggs (2008)

185

AN-1a.indd 195

11/03/2010 12:00:29

Medieval Petitions: Grace and Grievance, ed. W. Mark Ormrod, Gwilym Dodd and Anthony Musson (2009) St Edmund, King and Martyr: Changing Images of a Medieval Saint, ed. Anthony Bale (2009) Language and Culture in Medieval Britain: The French of England c. 1100–c. 1500, ed. Jocelyn Wogan-Browne et al. (2009) The Royal Pardon: Access to Mercy in Fourteenth-Century England, Helen Lacey (2009)

York Studies in Medieval Theology I Medieval Theology and the Natural Body, ed. Peter Biller and A. J. Minnis (1997) II Handling Sin: Confession in the Middle Ages, ed. Peter Biller and A. J. Minnis (1998) III Religion and Medicine in the Middle Ages, ed. Peter Biller and Joseph Ziegler (2001) IV Texts and the Repression of Medieval Heresy, ed. Caterina Bruschi and Peter Biller (2002)

York Manuscripts Conference Manuscripts and Readers in Fifteenth-Century England: The Literary Implications of Manuscript Study, ed. Derek Pearsall (1983) [Proceedings of the 1981 York Manu­scripts Conference] Manuscripts and Texts: Editorial Problems in Later Middle English Literature, ed. Derek Pearsall (1987) [Proceedings of the 1985 York Manuscripts Conference] Latin and Vernacular: Studies in Late-Medieval Texts and Manuscripts, ed. A. J. Minnis (1989) [Proceedings of the 1987 York Manuscripts Conference] Regionalism in Late-Medieval Manuscripts and Texts: Essays celebrating the publication of ‘A Linguistic Atlas of Late Mediaeval English’, ed. Felicity Riddy (1991) [Proceedings of the 1989 York Manuscripts Conference] Late-Medieval Religious Texts and their Transmission: Essays in Honour of A. I. Doyle, ed. A. J. Minnis (1994) [Proceedings of the 1991 York Manuscripts Conference] Prestige, Authority and Power in Late Medieval Manuscripts and Texts, ed. Felicity Riddy (2000) [Proceedings of the 1994 York Manuscripts Conference] Middle English Poetry: Texts and Traditions. Essays in Honour of Derek Pearsall, ed. A. J. Minnis (2001) [Proceedings of the 1996 York Manuscripts Conference]

Manuscript Culture in the British Isles Design and Distribution of Late Medieval Manuscripts in England, ed. Margaret Connolly and Linne R. Mooney (2008)

186

AN-1a.indd 196

11/03/2010 12:00:30

spine 20mm 14 jan 10

Chapman, Anne Curry, Eric Haeberli, Richard Ingham, Andy King, Anthony Lodge, Jean-Pascal Pouzet, William Rothwell, David Simpkin, Louise Sylvester, David Trotter, Laura Wright. Cover: Part of a thirteenth century letter in Anglo-Norman from Gillian de Tregoz to W. de Merton, from National Archives Ancient Correspondence SC1/7, no. 198. Photograph by Emma Cavell; reproduced by kind permission of the National Archives.

Boydell & Brewer Ltd PO Box 9, Woodbridge IP12 3DF (GB) and 668 Mt Hope Ave, Rochester NY 14620-2731 (US) www.boydell.co.uk / www.boydellandbrewer.com

INGHAM (ed.)

YORK MEDIEVAL PRESS

The Anglo-Norman Language amd its Contexts

The development of Anglo-Norman (the variety of medieval French used in the British Isles), together with the role it played in the life of the medieval English kingdom, is currently a major topic of scholarly debate. The essays in this volume examine Anglo-Norman from different perspectives and in different contexts, though with a concentration on the theme of linguistic contact between Anglo-Norman and English, seeking to situate such contact more precisely in space and time than has hitherto been the case. Overall they show how Anglo-Norman retained a strong presence in the linguistic life of England until a strikingly late date, and how it constitutes a rich and highly valuable record of the French language in the middle ages. Contributors: Adrian Bell, Paul Brand, Mark Chambers, Adam

YORK MEDIEVAL PRESS

The

Anglo-Norman Language and its Contexts Edited by RICHARD INGHAM