Vox Romanica 67 (2008): 100-132 Anglo-French in rural England in the later thirteenth century Walter of Bibbesworth’s Tretiz and the Agricultural Treatises Le Traité de Walter de Bibbesworth a été publié pour la première fois dans la thèse parisienne d’Annie Owen (1929), soutenue en 1930. La thèse elle-même ainsi que les réactions de son jury (et leur compte-rendu) reflètent l’attitude de leur temps, polarisée par des questions de phonologie, d’orthographe et de «grammaire». Or, cette approche a abouti à une évaluation profondément erronée du texte de Bibbesworth: elle négligeait l’importance de son origine insulaire, et l’absence de toute considération sérieuse du lexique du Traité escamotait l’un des aspects les plus frappants de l’ouvrage. En réalité, le texte de Bibbesworth ne peut se comprendre que replacé dans son contexte culturel, social et linguistique, et surtout lorsque les manuscrits divergents sont systématiquement comparés, et leurs gloses anglaises confrontées au texte anglo-français. 1. At the end of January 1930 the soutenance of a thesis entitled Le Traité de Walter de Bibbesworth sur la langue française submitted for the Doctorat d’Université by Miss Annie Owen took place in the Sorbonne, before a jury made up of three of the most eminent French philologists of their day, Mario Roques, Alfred Jeanroy and Antoine Thomas 1 . The thesis was written in French, submitted for a French academic degree, published in Paris and the language of the text which provided the subject-matter of the thesis was medieval French. The thesis had been published in 1929 by the Presses universitaires de France and a review of the edition signed by two of the examiners had appeared that same year in Romania 55 (1929): 575-79, of which the third examiner was editor. Apart from the Middle English glosses which accompanied the French text, everything about this thesis was French, so it was not surprising that it would be assessed according to the criteria applicable to editions of medieval French texts at that time. The language of the text being edited was the widely disfavoured Anglo-French variety from the late thirteenth or early fourteenth century, with variants from thirteen other surviving manuscript versions of the text also in Anglo-French stretching from the fourteenth into the early fifteenth century 2 , so that, given the low esteem in which insular French was held by philologists in France, the editor was perhaps naturally anxious to play down the extent of the undesirable insular features present in her text, affirming in her Introduction that: «Les passages incorrects ne sont pas très nombreux, on ne trouve pas tous les traits qui 1 The Secrétaire Général of the Sorbonne in a personal letter dating from 1982 gives the date of the soutenance as follows: «j’ai l’honneur de vous confirmer que Mademoiselle Annie Owen a soutenu devant la Faculté des Lettres de Paris, le 30 janvier 1930, une thèse pour le doctorat d’université intitulée: Le Traité de Walter de Bibbesworth sur la langue française. » 2 Owen 1929: 31-2. A further fragment identified in 1971 by Dr. Oschinsky will be referred to later. Anglo-French in rural England in the later thirteenth century caractérisent le français de l’Angleterre . . . et la syntaxe . . . est assez française» (28). Be that as it may, the French examiners duly interpreted the author’s reference in his prologue to what he called his «tretiz . . . pur aprise de langage» as promising a grammatical study of the medieval French language, with the various areas of grammar being treated separately, but this traditional approach resulted in their review article in Romania being strangely ambivalent. The first section, written by Jeanroy, begins in very condemnatory vein: «Ce ‹traité› n’est en réalité qu’un Nominale . . . », a not unexpected demotion in line with current philological opinion which put grammar before lexis in the academic scale of values 3 , and he goes on to say that the work contains over a thousand words which are «longuement commentés en vers d’une incroyable maladresse» to such a degree that «le texte est, en maints passages, inintelligible: l’éditrice aurait dû, soit s’efforcer de l’expliquer, soit au moins signaler la difficulté. L’analyse du traité est vraiment trop vague et incomplète, et le glossaire français laisse de côté beaucoup de mots ou de formes intéressantes» (575). In his judgement then, whilst the text itself and Miss Owen’s edition of this mere «nominale» could not be simply dismissed out of hand as being totally worthless, the text was riddled with grammatical errors to the point of unintelligibility and the editor had not provided adequate help in making its lexical content accessible to the reader, so both author and editor were deemed to be grossly at fault. Yet nowhere in the review is there any hint that the examiners sought to mitigate this lack of intelligibility said to be rife in the French text by making use of the abundant Middle English glosses which Bibbesworth states in his prologue to be an integral part of his work, intended to help anglophone readers understand the French vocabulary - «Dounc tut dis troverez vous primes le fraunceis e puis le engleise amount». Consequently, the Tretiz would be judged by the examiners solely as an exercise in medieval French grammar as defined by the philologists of the time, in total disregard of the author’s intentions as clearly expressed in his Prologue and in ignorance of the wider social background obtaining at that period in the area of rural England which had provided the incentive for 101 3 The belief in the primacy of phonology in linguistic studies held sway amongst philologists until the closing years of the twentieth century, when Bernard Cerquiglini produced his work on the origins of the French language in which he writes: «Ne disait-on pas, alors, que la phonétique était la partie ‹matérielle› de la langue, ce qui parvient physiquement à nos oreilles, tout le reste (morphologie, syntaxe, lexique et sémantique) étant un ensemble de relations immatérielles interprété par notre esprit ? Or cette partie ‹matérielle› de la langue évolue selon des lois propres qui ne doivent rien au hasard, sont indifférentes au sujet qui prononce, aveugles au sens dont le signifiant phonique est le vecteur. . . . la loi phonétique est l’aune à laquelle on mesure les langues vernaculaires, elle est l’instrument de leur archéologie, le support des investissements idéologiques et des réinterprétations biologiques . . . En appliquant ainsi les lois de la phonétique, on pouvait dériver, par construction, la plupart des mots français à partir de leur étymon en latin vulgaire. Au plan phonétique, du moins, mais avec la certitude de la science, et selon une écriture quasi mathématique, l’origine du français était prouvée. Il avait fallu plusieurs siècles pour aboutir a la phrase de Brunot: ‹Le français est du latin parlé›.» (Cerquiglini 1991: 23). William Rothwell 102 his work. The Tretiz would be judged as an exercise in medieval French phonology and morphology without reference to either space or time. However, Jeanroy’s strictures sit uneasily alongside other more positive remarks in which he unexpectedly acknowledges the value of the Tretiz in general, outside the confines of grammar, saying that «il est intéressant par sa date et sa richesse» and «Son importance . . . avait été rendue plus évidente par les extraits qu’en avait publiés, en 1877, P. Meyer . . . » He is also complimentary of the editor’s efforts in writing the thesis, referring to it as «ce travail très méritoire» and is presumably speaking also for his colleagues when he says that: «Miss Owen s’est acquittée de cette tâche avec une conscience et une intelligence très digne d’éloges: son édition fondée sur les quatorze mss connus donne tout ce que ceux-ci contiennent d’utile.» This unexpectedly fulsome conclusion to an otherwise very critical review contains the nub of the whole question regarding Miss Owen’s edition of the Bibbesworth text and the examiners’ assessment of her work. If the editor’s application and intelligence had indeed produced a thesis that covered all that was useful in the numerous known manuscripts of a text characterised by its «richesse», notwithstanding her failure to make it intelligible and her inadequate treatment of its lexis in her Glossaire, then responsibility for any remaining deficiencies in the edition would have to be laid at the door of the author and/ or the scribes who copied and adapted his work. Nevertheless, although concrete examples of the lexical richness of this «nominale» did not find their way into the examiners’ review article and so remain unidentified 4 , Jeanroy’s mention of their presence indicates that the maligned author and scribes must be credited with possessing a surprisingly extensive and detailed knowledge of at least one sizeable register of medieval French vocabulary, that of the countryside, despite the grammar of the text being beyond redemption. The admitted richesse of the French terminology found in the Tretiz, supplemented by the Middle English glosses, indicates that the author must somehow have managed to acquire a large body of uncommon and often specialised vocabulary in both French and English from undetermined sources, but yet appeared to lack the elementary grammatical knowledge of French that would have enabled him to set it down intelligibly. Moreover, the irremediably flawed work which resulted from these grammatical failings had inexplicably been preserved for over seven centuries in a considerable number of versions up and down the country, not to mention the numerous cases where extracts from the Tretiz can clearly be seen to have been incorporated from Bibbesworth’s work into texts of a quite different character up into the fifteenth century. If the factors of time and space are added to the author’s intentions set out in his Prologue, however, it may be seen that the examiners’ review did not address the primary purpose of the work. The prologue to the G manuscript chosen by Miss Owen as her base text begins with the statement that it is a «tretiz . . . pur aprise de 4 These matters may have been taken up at the soutenance, but no record of this has apparently been brought into the public domain. Anglo-French in rural England in the later thirteenth century langage», its aim being to provide the offspring of his landed patroness, Dyonise de Mountechensi, with all the vocabulary (tut le fraunceis) that they will need in order to manage their estates in French when they come of age to inherit, the terminology of what he calls «husbondrie e manaungerie» (l. 7-8). These two largely synonymous Anglo-French terms, the one English in origin, the other French, cover the varied procedures involved in estate management which are all set out in the Prologue, so that Bibbesworth’s Tretiz is to be understood as an attempt on the part of an influential landowner living in the East Midlands in the late thirteenth century to preserve the role of French in the agricultural life of her region. In this situation lexis might be expected to take precedence over grammar, but Bibbesworth’s Prologue was in prose and so would appear to have completely escaped the attention of the examiners, whose preoccupation was with the rhyming verses of the main text on account of the phonological information which it was believed they could provide, rather than with its vocabulary.This lapse is evidenced by their failure to comment on Miss Owen’s important mistranslation in her Glossaire of the key term manaungerie by the obsolete and totally aberrant mangerie. «Husbandry» and «feasting» do not naturally go hand in hand, the pleasures of the second being entirely conditional on the success of the first. «Feasting» does, however, occupy the final section of the Tretiz, but its presence in a work ostensibly dealing with the vocabulary of the countryside is explained neither by Miss Owen nor the examiners in their review article. Like the misunderstood manaungerie, it needs to be set correctly in space and time. When the manuscripts of the Tretiz are considered as a whole they demonstrate that their various compilers were capable not only of listing the French terminology for the parts of the human body, items of clothing or the flora and fauna of the English countryside in which they lived, but also of giving the French names for many birds and animals, together with their calls, the names of trees, flowers and plants, sometimes with their medicinal properties, and were even able to provide vocabulary of a more technical kind such as the names for the components of a plough or a cart, the materials necessary for building a house, the terms for weaving cloth and the techniques of making bread or ale. All these areas of vocabulary are far removed from the everyday terminology that makes up the bulk of the lexical content of most literary texts and raise the question as to how such specialised knowledge of French was acquired by English writers who were apparently ignorant of even the most elementary rules of French grammar. Furthermore, if these technical French terms used in the different versions of the Tretiz are checked against the dictionaries of Godefroy and Tobler-Lommatzsch, with their Middle English glosses similarly checked against the Oxford English Dictionary and the Middle English Dictionary 5 , the result shows that in the great majority of cases the 103 5 Whilst T-L and the Middle English Dictionary were not available for consultation at the time of the soutenance, the ten-volume Gdf. and the equally comprehensive Oxford English Dictionary would have been readily to hand, the latter under its early name of New English Dictionary. William Rothwell 104 «ignorant» author and scribes were correct in their use of such specialized terminology in both languages 6 , thus emphasising the gulf between the lexical content of the Tretiz and the grammar in which it is couched. The varied spellings in both French and English found in the different manuscripts do not detract from their semantic accuracy, such orthographical variation being a prominent feature of both languages. In contrast to this lexical richness the only phrase in the Tretiz which might be construed as referring to grammar is to be found at the end of the Prologue where Bibbesworth says that he will teach his pupils «le dreit ordre en parler e en respundre» (l.15-16) - a sort of afterthought briefly expanded in just a pair of lines (v. 25-26) of the verse text itself to mean no more than the correct use of the possessive adjectives mon/ ma, etc., hardly sufficient material to provide the substance of a treatise on French grammar, especially when the grammatical «rules» are not observed. The second important feature of the Bibbesworth texts that went unmentioned by the editor and unnoticed by the examiners is the compilation of numerous groups of a whole range of (quasi-)homonyms and synonyms in French which vary from one manuscript to another, accompanied by their counterparts in Middle English, a specialized linguistic exercise that would call for an unusual command of the lexis and semantics of French on the part of an English author and English scribes in an age long before the advent of the printed dictionaries which would facilitate work on such groupings for later generations.Yet nowhere in their review article do the examiners address the important question of how Bibbesworth and the unknown scribes who copied and adapted his work came to acquire this ability to construct lists of homonyms and synonyms in both English and French 7 along with their wide-ranging knowledge of the French vocabulary of the countryside, and yet commit a string of elementary grammatical errors. Only when this dichotomy is addressed, with the manuscripts examined individually and the texts correctly identified as belonging to the society of rural England in the late thirteenth and fourteenth centuries, the only conclusion capable of explaining the presence of the Middle English glosses, can Bibbesworth’s work be properly assessed. A striking example of a failure to appreciate the extent of the author’s command of the French lexis (as shown by his ability to bring together such homonyms) is provided on p. 576 of the Romania review where Jeanroy writes that: «Il (i. e. Bibbesworth) distingue (v. 833s. 8 ) quatre rais: le rai de soleil, le rai de roue, la raie (poisson) et ceux qui ‹ver la feire vount›; aucun ne figure dans l’analyse, deux seulement au Glossaire et le quatrième nous reste mystérieux.» His criticism of the editor for her failure to explain adequately these French terms is justified, especially as regards the fourth homonym, which is passed over without comment 6 Some of the small number of exceptions will be examined later. 7 A selection of these are shown in the Appendix. 8 The correct numbering is v. 835-44; cf. N9. Anglo-French in rural England in the later thirteenth century in the edition, an indication that she, like her examiners, did not understand it, but his admission that he and his colleagues were defeated at this point confirms that they must have worked solely from the printed thesis set before them. Had they consulted the manuscripts or even Miss Owen’s footnote to the word that baffled them in her v. 836 9 , they would have seen that in no fewer than three of the manuscript versions this raie is glossed by forms of burel, a French term that had been taken into Middle English. What is more, on f. 290va of the base G manuscript itself below «le (l. lé) raies» this word bureles (whether French or English) has been crossed out. A further gloss borer clot (i. e. ‘burel cloth’) in the A manuscript (f. 303rb) makes it clear that the word designates a fabric, although it is incorrectly set above the wrong line. The quotation indisputably means ‘the striped cloth goes to the fair’ (i. e. to be sold), burel being amply attested in Gdf. 8.395b-96a as ‘grosse étoffe de laine’ from the twelfth century onwards (Rothwell 1960). The French raie itself is similarly well attested in both continental and insular French, although the only example given in Gdf. under rai 2 (6.558c) is defined as rayure and comes from an Anglo-French source. Under roie T-L 8.1406-7 illustrate the juxtaposition of raie and burel in medieval French from the thirteenth century. The correctness of Bibbesworth’s work here is also verified by official texts on the English side of the Channel around the time of his Tretiz. In a law suit of 1292 a man accused of passing himself off as a cleric in order to claim clerical privilege is told by the judge that his plea is unacceptable because: «Vous estes vestu des draps de Ray», i. e. not clerical attire (Bolland 1925: 25) 10 . Under rai(e (n. 2) (a) the MED gives examples of the meaning ‘A kind of striped cloth; also, a piece of this cloth’ in both Anglo-French and Middle English from the early fourteenth century. This case of a word in the French text being glossed in «English» by another French word is a reminder that the French influence on English has been predominately in the area of lexis rather than grammar. The extensive knowledge of the vocabulary of French and Middle English that would be required to enable the author to assemble such groups of homonyms from memory is nowhere acknowledged in the review article. Although from the examiners’ point of view the Bibbesworth text offered nothing more than a long list of French words set in verses that made little attempt to conform to the accepted norms of medieval French grammar or spelling, being no more than a barbarous perversion of the language that often appeared to make no sense, if accurately transcribed and placed in its correct social 105 9 Numbered verse references will not be made to Miss Owen’s edition of the Tretiz (Owen 1929) except when quoting directly her text or references to it, because she omitted v. 584 «Nos averoms grisil puis qu’il grele» together with the M. E. gloss «hailith», nullifying Bibbesworth’s juxtaposition of grele ‘small’ and grele ‘hail’ and upsetting the correct verse-numbering from that point onwards. See N21 for the exception to this rule. 10 I am indebted to Professor D. A. Trotter for providing me with a copy of this passage. The phrases robes de ray and draps de ray (with varying spellings) are to be found from the late thirteenth century onwards in Harvey 1965: 247, Rotuli Parliamentorum 2,241, and in Riley 1859-62: 676. William Rothwell 106 context, the Tretiz may be seen to occupy a significant position in English history and the history of English. As was mentioned earlier, despite their recognition of the many difficulties of comprehension presented by the text, there is nothing anywhere in the examiners’ review to suggest that they ever turned to the manuscripts themselves in an attempt to resolve these problems, as was shown with the raie difficulty above and will be demonstrated again later, so that their verdict when allocating responsibility for the errors in the edition is based entirely on the sole authority of the text as printed. This already fragile judgement was further clouded by the editor’s convoluted explanation of her choice of G as her base manuscript (32), in which she asserts that: «Walter de Bibbesworth a préparé une édition du traité (représentée par T) pour M me de Munchensy, et une autre (représentée par G) pour le public», that «la seconde édition est un remaniement de la première» and that «l’auteur a préparé la seconde édition avec plus de soin que la première» (33) 11 . No concrete evidence is adduced in support of these assertions, so that the examiners’ evaluation of the various Bibbesworth manuscripts with their many errors is therefore predicated on an unsupported hypothesis enlisting the personal responsibility of the original author himself in order to help establish the authenticity and primacy of the G manuscript said to have been prepared avec plus de soin que la première (T) and destined for an unidentified general «public», whilst, against all the rules of courtesy and class, his noble patroness 12 who instigated the work in the first place would have to be content with his earlier uncorrected T version, all the other versions presumably being deemed to be nothing more than copies of one or other of these two. The examiners’ apparently uncritical acceptance of this uncorroborated ranking of the manuscripts, their attention being focussed on the imperfections in the grammatical forms of the single text submitted to them, and their consequent acceptance of Miss Owen’s contention that her error-strewn G version as printed was the authentic definitive product of Bibbesworth himself, as would be indicated by the absence of any dissent or even doubt expressed in their review, might explain their failure to refer back to the manuscripts when faced with the innumerable problems of meaning or intelligibility presented on the printed pages of the Tretiz. Unsurprisingly, however, Miss Owen’s gratuitous premiss regarding the ranking of the manuscripts is demonstrably untenable once the Middle English glosses are brought into the picture. In the very first section of the supposedly authoritative G text one of the Middle English glosses is blatantly aberrant, whilst other less favoured manuscripts, including the «uncorrected» T, have the correct form, thus demolishing Miss Owen’s assertion of the primacy of G and with it the certainty 11 For ease of reference the manuscripts are referred to in the forms given on p. 30-32 of Miss Owen’s Introduction. John Koch’s use of a completely different set of symbols makes his article (Koch 1934) very difficult to use. 12 Dionyse de Munchensy or Mountechensi, Bibbeworth’s patroness, was from a noble landed family in the Chelmsford area. Anglo-French in rural England in the later thirteenth century of Bibbesworth’s personal responsibility for its compilation, together with the examiners’ faith in the correctness of her text as printed. When illustrating the different senses of the medieval French top/ toup, the G manuscript reads: «E[n] la lute desrenés le toup», incorrectly glossed in Middle English as «wind the yarne» (v. 38). The correct meaning of the French is the very different: «In the wrestling win the ram», as supported by the Middle English glosses from several other «inferior» manuscript versions given in Miss Owen’s footnotes 13 . This is a reference to the well-known wrestling event in centuries-old country sports where a ram was the winner’s prize, as may be seen in the Seneschaucy 14 , written at about the same time as the Tretiz, and later in Chaucer’s Canterbury Tales, where it is said of the brawny Miller that - «At wrastlynge he wolde have alwey the ram» 15 . The scribe of the G manuscript is confusing the English word tup (still to be heard in the north of England) with its French homonym as used two verses above - «E serencez de lin le toup», which is glossed correctly in Middle English as hechele a toppe 16 of flax. A further mistake is to be found later in a list of bird-names, where G incorrectly glosses a form filaundre by star (i. e. ‘starling’). This time the error is in the French, not the Middle English gloss. Filaundre is not a bird-name at all, but means ‘gossamer’ (v. 719) 17 . As was the case with «tup», three other manuscripts have the correct gloss here. Miss Owen was aware of both of these errors, referring to them on p. 30 of her Introduction, but without drawing any negative conclusion from them regarding the primacy of the G manuscript. Similar scribal errors in glossing are to be found elsewhere in the chosen G text. For instance, in v. 93 molet is glossed in G by the Middle English hole instead of ‘lobe (of ear)’ (T-L 6.180), whilst other ms. have lap or dewlap which carry the correct sense of ‘a pendulous piece of skin’; in v. 164 the English term pees (‘thighs’) is used incorrectly in G to gloss the French reynes, even though the editor publishes the correct Middle English lendes (‘loins’) at this point in her footnote references for ms. B and O. In v. 1065 of the G text the Middle English gloss steppes is wrongly used for the French escous instead of laps, but is used correctly for the French esclos only a few verses later (v. 1072). That Bibbesworth himself would be responsible for such obvious 107 13 The correct gloss is to be found in ms. A: «A la lute dereynetz le toup», «Wyn pe ram atte wrestling» (f. 299rb), «at pe wrastlinge win pe rom (l. ram)»; ms. C (f. 2va), «pe ram at wre[st]ly[ng]»; ms. T (f. 120v). T-L 10.384 under top s. gives both quotations from the Bibbesworth ms., but without recognising the difference in meaning between the two and so giving only the one gloss ’Schopf, Büschel’. 14 This agricultural treatise will be referred to again later in this article. In this text la lute, along with the feire, marché and the taverne is mentioned as a distraction forbidden to those in charge of animals (p. 278). 15 General Prologue v. 548 (Benson 1987). 16 OED s. top sb.1, section 2 ‘A tuft or handful of hair, wool, fibre, etc.’ 17 This incorrect filaundre appears again in the Nominale sive Verbale (Skeat 1906): «Arounde esturnel et filandre/ Swalewe and starlinge» (v. 809-10). This text is one of the works deriving from Bibbesworth’s Tretiz. The two French forms esturnel et filandre for the single «starlinge» might well be an echo of the error in the G manuscript. William Rothwell 108 mix-ups in the very manuscript that is claimed to be his personal choice, whilst other less favoured manuscripts carry the correct sense, is simply not credible, so these errors confirm beyond any reasonable doubt the rejection of Miss Owen’s claim that the G manuscript is the original work of Bibbesworth himself. A similar but more complicated situation arises in v. 548 of the G text where a form vert tenail is glossed by the scribe as grene balke, with tenail again glossed by balke in v. 561, and repeated in v. 564 and 567 without being glossed. The printed form tenail in v. 548 is an editorial error, the manuscript reading giving the correct terail, glossed correctly by the Middle English balke, but the erroneous tenail in v. 561, 564 and 567 is scribal, not editorial. Terail and tenail are part of a group of near-homonyms that Bibbesworth wishes to separate for his readers - terail (‘bank’ or ‘slope’), tenail/ tenailles (‘tonge(s)’), tenoun (‘handle’ of a plough). These errors of scribe and editor are compounded by the editor’s further misreading of ber as ver in v. 567, giving the meaningless «E par le tenail passe meinte ver», instead of the correct «E par le terail passe meinte ber» (‘And many a (noble) man passes over the bank/ slope’). The forms tenail (‘tonge(s)’) and ver (whether taken to mean ‘spring’, ‘a boar’ or ‘a worm’), make no sense in this context. Scribe and editor combine to reduce a perfectly intelligible text to nonsense and the hapless author is made to appear an ignoramus. However, these errors in the G manuscript, claimed by the editor to be Bibbesworth’s personal version of choice, do not mean that the «second-best» T version is always correct and can be elevated to the role of base manuscript in place of G. In folio 122v, T has the following verse: «Le clerk soune le dreyn apel», with apel wrongly glossed as a mouster (i. e. not an English word, but French, meaning ‘a monastery/ church’) instead of the correct gloss «The cleric tolls the knell» 18 . This is another example of the mixing of the languages. Other manuscripts, including G, have the correct Middle English knel here. Again, in folio 127r of the T text the French Le cem is incorrectly glossed by the Middle English pe paw (‘the thaw’), but in f. 130r the same word, spelt ceym has the correct gloss sleet, not «thaw». These few random examples of errors in both G and T suffice to show that neither of these manuscripts can be regarded as the original text composed by Bibbesworth himself, but are to be treated along with the other extant versions as being the work of unknown scribes, the original manuscript presumably having been lost 19 . These errors present in both manuscripts would mean that Bibbesworth himself cannot be held responsible for all the manifold errors in the text as 18 As in Gray’s Elegy. 19 A similar error to that of Miss Owen is made regarding the status of the glosses in Koch 1934. Because Bibbesworth states in his Prologue that the reader will find the English glosses above the French text, Koch assumes that the author himself was responsible for them «woraus hervorgeht, dass auch die Glossen, wenn auch nachmals teilweise geändert, von Walter herstammen» (p. 33). If the text itself was altered by later scribes, as it manifestly was, it stands to reason that the glosses too would be liable to be altered. This alone can explain the considerable differences between the glosses in different versions of the Tretiz. Anglo-French in rural England in the later thirteenth century published. The error concerning his personal authorship of G and T led in turn to a lasting failure to assess correctly the role of the Tretiz in the society of later medieval England for which it was destined. Bibbesworth’s treatise was dismissed as nothing more than a gross travesty of the «correct» language of continental France, just a botched attempt to teach French grammar that could not have any bearing on the history, linguistic or social, of its country of origin - England. This derogatory label would remain attached to Bibbesworth’s work for decades to come. Yet the numerous surviving manuscripts of the Tretiz were composed in widely separated areas in England (Koch 1934: 58) over a period of a century or more, thereby attesting to its long-lasting, if unexplained, popularity, and their scribes display far more independence in their treatment of its various sections than is customary with literary works of the period, suggesting that they were not merely copying blindly a «received» text. The length of the Tretiz varies from one manuscript to another as the scribes not only reduce or omit altogether one or more of its sections, but also present them in a different order or vary their content, even adding significant new material, as will be shown later. This independence carries with it another difficulty in arriving at a correct assessment of the Tretiz that is not to be found in similar assessments of literary material. The fact that the scribes of the various Bibbesworth manuscripts are dealing with areas of lexis that are seldom encountered in more conventional texts and differ widely in their treatment of them means that there is often no «standard» set of items of vocabulary by which the accuracy of all the manuscripts may be judged. The amount of glossing also varies from one manuscript to another, being completely absent, for example, from Sloane 513, sparse in the B. N., Nouv. acq. lat. 699 version and very abundant in the later All Souls 182 20 . These factors mean that the G manuscript cannot be taken as being representative of all the other versions of the text. Although these important differences between manuscripts are not readily discernible in the Owen edition, where, as mentioned earlier, they are reduced to the level of footnote variations of individual words in both languages placed below the printed base text without any surrounding context, their presence indicates a widespread and long-lasting interest in the subject matter of the work on the part of educated scribes and, concomitantly, the presence up and down England of a literate readership for a text that in an era dominated by phonology and morphology has been judged to be unacceptably flawed. If the manuscripts of the Tretiz are correctly transcribed and situated in their historical context as individual entities in their own right, with the diversity of their French versions and their Middle English glosses set out in full, the complete work being regarded as an exercise in Anglo-French vocabulary rather than continental French grammar, this will produce a quite different and more appropriate assessment of the text. 109 20 For details of the 13 manuscripts known to Miss Owen, see p. 30-32 of her edition. Cf. Dean/ Boulton 1999: n° 285 (17 ms.). William Rothwell 110 Whilst the readers of the Tretiz in the later Middle Ages, whose interest lay in the lexical content of the text, would have had before them tangible authentic manuscripts, Miss Owen’s examiners, concerned more with sounds and forms, must have had at their disposal nothing more than an unverifiable printed transcription of one single manuscript, accompanied at the foot of the page by variants from other versions. Consequently, they were quite unable to question the accuracy of her readings of even the base G text itself, had they been so minded, and were dependent for the texts of all the other manuscripts on nothing more substantial than the incomplete and often inaccurate individual variants in both languages as given in her footnotes. In concrete terms, if the text and its footnote variants as set out in the edition are compared with the manuscripts themselves, it will be seen that many of the errors in the base text as printed, in the accompanying variant readings and also in the Middle English glosses, which together account for a very considerable proportion of the lack of intelligibility deplored by the examiners, stem not from the ignorance of author or scribes, but from faulty transcription in both languages on the part of the editor (cf. Rothwell 2004). There are, in fact, two quite distinct layers of error in the text as printed, those mistakes which are attributable to the various scribes and those which are the responsibility of the editor herself. The separation of these two categories of error, essential for a correct assessment of Bibbesworth’s competence, would be impossible for the examiners in the absence of all the manuscript evidence. In sum, the examiners were judging not Bibbesworth’s Tretiz, as they imagined, but merely Miss Owen’s faulty transcription of one imperfect version of it. They cannot have set eyes on any of the manuscripts, not even the B text which would have been readily to hand in Paris, having been acquired by the Bibliothèque Nationale in 1908, as stated in the Owen edition (32), and were content to concentrate their attention on just the one text, despite its obvious manifold errors. The following examples of such errors in the printed edition of the Tretiz, the variants and also the Middle English glosses, as revealed by scrutiny of the base manuscript, make no claim to being exhaustive, but are simply intended to illustrate the fact that any judgement of the work based exclusively on Miss Owen’s printed version cannot be reliable and therefore cannot reflect accurately the role played by Bibbesworth in the changing linguistic climate of his day, as French in England became increasingly an acquired rather than a vernacular language, the preserve of the literate, whilst Middle English was spoken at all levels but not yet written in proportionate quantity. In v. 34 of Miss Owen’s edition 21 the French text as printed reads: «Moun toup vous prie estanchez», the verb estanchier in the editor’s Glossaire being glossed as ‘arrêter’, which would give the nonsensical reading: «I pray you to stop/ halt my forelock/ quiff». The Middle English gloss hevese (OED evese ‘to cut short (of hair)’) written by the scribe and printed by the editor over the alleged estanchez 21 In this section the verse numbers quoted must inevitably be those of the printed thesis. Cf. N9. Anglo-French in rural England in the later thirteenth century ought to have warned both editor and examiners that this reading could not possibly be correct. If read as estauchez instead of estanchez the sense is clear. In his entry for estaucier Gdf. 3.603b-c gives the correct meaning ‘tondre, tailler’, although under top (7.746b) he repeats Paul Meyer’s erroneous estancez in this same quotation, a mistake that may be the source of Miss Owen’s error. In v. 234 and 276 cerceles and cercel are not the same word ‘crécerelle, oiseau de proie’ (i. e. English kestrel) as given in the Glossaire. The Middle English gloss teles (i. e. ‘teals’) above cerceles in v. 234 is supported by the similar reading telys in no less than five other manuscripts according to Miss Owen’s own footnote, but was ignored by editor and examiners alike, whilst cercel (v. 276) is preceded by faucoun, so is to be read as tercel, giving faucoun tercel as in Gdf. tiercel 2, tiercelet (7.710a), the English tercel as in OED: ‘The male of any kind of hawk’. On p. 62 of the printed text, in additional material taken from ms. O, where the scribe is making the distinction between the noun forure (‘scabbard’, applying to a sword) and the adjective furé (‘lined’, applying to a garment), the «f’» of furé is misread by the editor as a «long s» giving the reading sure (‘safe’) which makes no sense in that context. On p. 63 the text as edited reads «De la ceinture le pendaunt / Passe par un trespasse le mordaunt» (v. 191-92), another nonsensical statement, but if the minims in un are simply read as mi, the corrected text gives good sense: «De la ceinture le pendaunt / Passe parmi, trespasse le mordaunt» (‘the end of the belt goes through, traverses the buckle’). Again, Bibbesworth rounds off his account of the techniques of brewing by saying that the method of preparation should be worked through until good beer is produced - «Deskes vous eez bone serveise» (G f. 286rb), but Miss Owen prints «Deskes iour cez bon serveise» (v. 493), which makes no sense. On p. 70, in the context of protecting a barbican from attack, the printed text reads: «Pur le barbican defendre / Al assaut ke home vent vendre» (v. 268), but reading veut for vent and rendre for vendre provides the good sense intended by the author. The text of v. 367-68 in the edition runs: «Mes pur plus parfitement / Parler devant une gent» instead of the correct « . . . bone gent». On p. 97 in v. 545 the reading «Puis pas ceo bois . . . Passerez» should obviously be «Puis par ceo bois . . . ». The same mistake occurs a little further on where «Ki pas bost . . . » (v. 605) must be read as «Ki par bost» (i. e. ‘who by a boast . . .’, with the English boast being used as a French term). Again on page 97 (v. 550) streams are said to be «si chers e si beles», when the manuscript reading is clers. A few verses later (v. 554) the printed text reads: «Et par le orail est meint home» instead of the author’s correct «Et par le orail oit . . . », i. e. ‘hears’ (through his ears). The letters f, l and s are again confused in the following examples: in v. 318 of the printed text it is stated that a frog «fist en un reoun» instead of «sist en un reoun» (i. e. ‘sat in a furrow’); in v. 644 it is said that the primrose and the cowslip «le mustrent en tens de veir», instead of «se mustrent» (‘show themselves, appear in the springtime’) as in the manuscript. Verses 683-4 in Miss Owen’s text make no sense as printed: «Mes si ad diverseté grant, / Ki ne sevent une a quaunt» but if the minims in une are read as mie and a quaunt as aquant the corrected sense is ‘ . . . which some people do not know’. When the 111 William Rothwell 112 song of the cuckoo is said to be unpleasant and: «Poyne serreit si riotuse, / Si l’un chaunt fu graciouse» (v. 793-94), the manuscript clearly reads Poynt (the negative ‘not at all’) and sun (‘its’) not lun, restoring the author’s sense. Again with regard to the cuckoo, when the base manuscript reads kokel in the French, glossed in Middle English as kockou (v. 791), with four other manuscripts giving the gloss cuckow, it is perverse to print le cucknel instead of le cuckuel as being the reading in manuscripts A, P and T (117). Moreover, the Middle English gloss in ms. A at this point is misread as a kocken instead of the obvious a kockeu. The meaning of the text at this point is further blurred by its being printed as: «Il ne semble que le cuknel Tent sun chaunt . . . » instead of «Il me semble . . . », the correct reading. On p. 138, v. 1069 reads: «E herbe qe cirst al huis del estable . . .,» when the correct reading is: «E herbe qe crest . . .,» i. e. ‘grass that grows at the stable-door’ (f. 293va). A variant reading from ms. O f. 336va on p. 100 is said to state that in winter «les ees se tiennent en nischez», but bees are more likely to remain in their hives - ruchez - the correct manuscript reading, rather than in niches 22 . On p. 141.5, quoting a difficult passage taken from ms. B, the printed text has the meaningless «Qant le kens onli quistron . . .,» but if u instead of n is read in kens and onli this produces the correct forms «Qant li keus ou li quistron . . . » , i. e. ‘when the cook or the kitchen-boy . . .’. A few verses later, in the context of negligently burning the beef during cooking, the incomprehensible printed text runs: «Si dit homme qant enrisé / Que li evesque ust passé», but if the manuscript is read correctly as: «Si dit homme dunc en risé / Que li evesque i est passé» (f. 106r), meaning ‘then people say as a joke that the bishop has called there’, the sense is perfectly clear, being a jocular reference to the bishop’s power to send his flock to burn in Hell. Similarly, the errors of spelling or sense in the English glosses contain the following: in v. 16 the gloss it lagge him should read bilagge him (‘soil/ dirty himself’); in v. 249 the vache is the ‘cow’ in English, not the low; in v. 253 the printed text says that a cat mewich, instead of mewith (i. e. ‘mews’); in v. 337 dranck should read drauck (modern English ‘drawk’); in v. 502 the gloss for aroé should be hose (‘hoarse’) not hole; in v. 761 ther should read pei (‘thy’); in v. 830 boutes are bontes or bondes (the metal hoops round the wheels of a cart, OED bond sb.1). In v. 754 and 756 the English gloss for the French tresel/ trestel 23 is printed as stonc where other manuscripts are listed as reading stak and shocke (i. e. modern English ‘stack’ and ‘shock’ of corn), so the printed stonc is yet another blatant confusion of n/ u, a misreading for stouc (MED stouk(e, stouc), the modern English ‘stook’. Although this far from complete set of such elementary errors allegedly committed by Bibbesworth and/ or the scribes in their native language would provoke deep scepticism on the part of English readers, they are not commented on by the examiners, who must evidently have ignored them as lying outside their purview, or, if they had understood them, must 22 Gdf. 10.202c and T-L 6.635 record niche only from the very end of the fourteenth century, about a century later than Bibbesworth. 23 See Gdf. 8.48c for similar forms with this sense in modern Norman and other dialects. Anglo-French in rural England in the later thirteenth century have accepted them without question as proof of writers’ utter incompetence in both French and English. The errors in the edition extend into the Glossaire, although on occasion the context in which the Anglo-French terms are used ought to have alerted the editor to the problem. For example, on p. 150 of the Glossaire the noun agaz (f. 291rb) is defined as an independent adjective ‘rusé’, despite its being glossed in the manuscript by the substantival Middle English gloss scorne and forming part of the well-attested locution tenir a gaz ‘to take lightly, scorn’, based on the noun gab/ gas/ gaz, etc. ‘mockery’ (Gdf. 4.196a-b). On the same page of the Glossaire is an infinitive aourser unattested elsewhere and coined from aourt (f. 291vb), an attested present indicative form of the commonly used verb aerdre (Gdf. 1.121c- 123c). The editor’s mistranslation of manaungerie by the obsolete mangerie (‘feast’) in the prologue to the Tretiz, that was mentioned earlier, is not the only case of a failure to recognize different registers of vocabulary. It occurs again in a more difficult context concerning the building of a house, where the roof is said to need a genchour to strengthen it (f. 292ra). This word is translated in the Glossaire as ‘garde-manger’, which would call for the highly unlikely practice of situating a larder at the top of a house as a support for the roof. The error arises from a Middle English gloss given in the edition as pantre, interpreted as modern English ‘pantry’, but this is another, although more difficult, case of the frequent confusion of n/ u in the manuscript. The word must be the Middle English pautre (MED ‘A beam, prob[ably] a wall plate (OF poutre)’) 24 . Incidentally, Bibbesworth’s use of this word in the later thirteenth century could provide the MED with its earliest attestation, its present entries coming from the second half of the fourteenth century and the fifteenth. On p. 96 Miss Owen complicates matters by printing a whole additional passage on fish names with their glosses said to be found in ms. B O 4, but not in the base G text, an example of the scribal independence referred to earlier. Only one version of this passage is printed in the edition, which would indicate a most unusual case of three manuscripts written at different times in different places carrying precisely the same texts and making the same mistakes in both languages. In fact, however, the manuscripts named by the editor as having the same text differ considerably in their listing of the types of fish in French and also in the English glosses which apply to them. The mistakes in both the Anglo-French and Middle English 113 24 The form genchour appears to be unattested elsewhere, but in this context of building a house other Bibbesworth manuscripts support the G version at this point. Ms. C glosses its guenchour by pautreo (v. 1075, f. 12vb); ms. O f. 339va gives the Middle English pautree to gloss trabes (v. 1016), pautre to gloss mesier (‘wall’) (v. 1018) and furst to gloss its gwengeur (v. 1022). For this latter term see OED first ‘The inward roof or ceiling of a chamber, also a ridge-pole’. In the O manuscript this beam is said to be necessary en longure i. e. ‘lengthwise across the house’. Schellenberg’s explanation of genchour as «Wohl zu guenchir im Sinne von ‘obliquer’ = ‘in schräger Richtung gehen’ is difficult to justify, the beam needing to go straight from one end of the house to the other (for Schellenberg’s work, see below). William Rothwell 114 to be found in this passage are difficult to deal with on the sole basis of the printed text without having the manuscripts to hand, their source as given by the editor not always being accurate. For instance, if the manuscripts of B and O are used in order to point up the discrepancies without too much complication, the first verse of the addition is said to read: «E si de heise y peschez», with peschez glossed by fesches, but this is valid only for ms. O (f. 335vb), not for B (f. 100v), where the form heise is replaced by heche glossed by hoke (i. e. ‘hook) and peschez glossed by fisches. The text of v. 6 in this additional piece is said to read: «Le platon, et luy espines», with espines glossed by thurllbak. However, as in the example above, the version given in the printed text is again taken from ms. O (f. 336ra), not B (f. 100v), and the correct form is espinés glossed correctly by thornbak, i. e. ‘stickleback’. The reading in B is: «Le platoun & li espineis», and neither of these nouns is glossed. Yet again, in v. 8 the edition reads: «Oistre creveis et cok hanoun», glossed as oistre, wiwe and rueskill, where the second and third elements are incorrectly transcribed, but this is yet again a form of the O version, in which the correct glosses are oistre, welke and cokil muskill (i. e. ‘oyster’, ‘whelk’ and ‘cockle/ mussel’). At this point B (f. 100v) has «Oister, crevis & kochamon» without any Middle English glosses. Verse 11 as printed reads: «Le goleis e le taupe», with goleis glossed as schripes, but this also is taken from O (f. 336ra). The correct forms would be shri[m]pes (i. e. ‘shrimps’), and tanpé. The correct version is found this time in ms. B (f. 100v), but not used in the printed edition: «Le geleis ne le tauntepé», glossed as wylke & prane (i. e. ‘whelk and prawn’). At the foot of p. 96 Sanguille ought to read L’anguille and goioun should be read as gojoun, i. e. ‘gudgeon’. This inserted passage contains also other fish-names of interest such as the grelet and the carbonel (B 100v), glossed as congur and colfysh respectively. The grelet is recorded only once by Gdf. 4.346b in an Anglo-French text of 1396 25 and defined in French as ‘l’ombre’, modern English ‘grayling’, whilst the carbone is similarly recorded only once in the same text, but left by Gdf. without a proper definition as ‘sorte de poisson’. It is the ‘coal-fish’. The importance of this passage found in the three manuscripts concerned is that it shows clearly the independence of the scribes who adopted and adapted Bibbesworth’s Tretiz for their own requirements, moving away from his stated aim of providing French agricultural vocabulary for future estate owners in the English countryside. The majority of the fish mentioned in these manuscripts are creatures of the sea, whose French forms would be unfamiliar to all but a minority of the educated inhabitants of England in the fourteenth century. In other cases the examiners themselves propose corrections to Miss Owen’s text that are erroneous. For example, Jeanroy rejects the editor’s translation of bulenge in v. 378-79 ‘bluteau à passer la fleur de farine’ in her Glossaire, saying that it means not the modern ‘blutoir’, but ‘blutage’, i. e. not the instrument, but the process of sifting or bolting flour, yet the word is glossed by the Middle English 25 Kristol 1995: 11. Both grelet and carbonel are found there. As will be demonstrated, this text draws on Bibbesworth for some of its material. Anglo-French in rural England in the later thirteenth century bolting cloth in v. 379. Also, in the variants given in the footnotes to the edition, bultingge clot for v. 378 occurs in no less than four other manuscripts, with bultyng, the operation of sifting, being found in a further manuscript referring to v. 379. The presence of both meanings (the process as well as the instrument) in the Middle English glosses would counsel against any dogmatic «correction» on the part of the examiner and demonstrates the danger of not being able or willing to handle the important Middle English content of the text. Again, when attempting to correct Miss Owen’s translation of the Anglo-French drache (ms. G, v. 275) by ‘carouge’ in her Glossaire, Jeanroy would replace it by the modern French ‘gousse de légumes’, saying that she had taken her gloss inaccurately from Godefroy, but Gdf. 2.766a gives the correct meaning ‘rafle du raisin’ as his second meaning, immediately below his ‘carouge’ sense, in the quotation «drasche que les pors mangeoient», a context that matches perfectly the one in Bibbesworth’s quotation: «Troye groundile qaunt drache quert» (the editor’s reading of the Middle English gloss over troye as soune is yet another n/ u error and ought to be read as souue = sowe = modern English ‘sow’), the phrase meaning ‘the sow grunts when seeking draff’. The OED, which would have been available for consultation by the examiners from 1928 onwards, defines draff as: ‘ . . . wash or swill given to swine; hog’s-wash’ 26 . Elsewhere Jeanroy rejects Miss Owen’s translation ‘moustache, favoris’ for Bibbesworth’s gernoun (v. 97), replacing it by giron. However, Gdf. 4.348a-b under grenon 1 has many examples of the word which he glosses as ‘moustache, favoris’, evidently the source of Miss Owen’s translation. The Bibbesworth text refers to the gernoun as being situated below the ear - «desouz le orail», therefore at some distance from Jeanroy’s ‘giron’ (‘lap’). The Middle English gloss thonewonges (i. e. MED thunwonge ‘The temple of the head’) here in ms. G is supported by other ms. In ms. C f. 3ra the French reads as follows: «Desouz le orayl avez gernoun», with the Middle English gloss: the her over the ere (= ‘the hair over the ear’), and ms. O f. 332ra reads: ere of pe hed, so the correct meaning would be ‘sideburns’ or ‘side-whiskers’ growing on the cheeks. Antoine Thomas is more circumspect than Jeanroy in his section of the review, making some useful corrections to Miss Owen’s Glossaire, but when he «corrects» her translation of aillie in the expression «il ne vaut pas un aillie» (: nascie) in v. 1076 as ‘ail’, claiming that its form should be changed to ‘aillade’, he is ignoring the Middle English gloss a pile of garlec (i. e. an object of little worth). This use of ‘ail’ to mean ‘a worthless object’ is frequently attested (Gdf. Complément 8.58a, T- L 1.238) and has also the variant forms aly, ailli and allie in Anglo-French. Thomas again leads the reader into error when he states that Miss Owen’s translation of frenol by ‘jonc’ is wrong, claiming that the Middle English gloss «keiex, aujourdui (sic) kex signifie ‘ciguë’.» The correct meaning is again confirmed by the OED as ‘the dry, usually hollow stem of various herbaceous plants’ (OED s. kex). In the context of a servant lighting a frenole, the key element must be the dryness of the 115 26 See also the entries under drast in OED and MED. William Rothwell 116 plant, making it combustible, hence the general sense ‘rush’, as Miss Owen says, rather than the precise ciguë, ‘hemlock’. 2. Apart from the examiners’ printed review, Miss Owen’s study aroused little interest amongst scholars working in the field of medieval French, giving rise only to a short Berlin thesis by Gerhard Schellenberg entitled Bemerkungen zum Traité des Walter von Bibbesworth (1933) that was based entirely on her work 27 and the article by Koch referred to earlier that concerned itself with the Middle English glosses rather than the French text. Although Schellenberg makes no mention of their review article or to having seen the base manuscript for himself, he followed in the examiners’ footsteps by attempting to analyse the language of the work in accordance with the traditional grammatical divisions applied to Old French at that time and, like them, was content to accept Miss Owen’s version of the text at face value. Not surprisingly, he too was most dissatisfied by what he found. His negative statement «Es ist schwer, im Anglonormannischen festen Boden unter die Füsse zu bekommen» (5) indicates an expectancy of a less than adequate work in conformity with the prevailing view that the firm ground of «rules» as laid down in the manuals of Old French for the different areas of grammar is not to be found in the later Anglo-French across the Channel. He proceeded nonetheless to work through each of these areas, comparing the forms in the text as set down in Miss Owen’s thesis with the «correct» ones given in the grammar books, and contrasting their spelling with «die normalen altfranzösischen Schreibungen» (7-33), a criterion which the abundance of forms given for many of Godefroy’s entries shows to be very elastic indeed 28 . In short, Schellenberg found that there was a good deal of «Unübersichtlichkeit» in the phonology of the text, that the morphology was even worse: «es würde ins Uferlose führen, jede flexivische Ungenauigkeit zu besprechen», and that the syntax was simply hopeless: «Syntaktisch ist das vorliegende Material nicht ernst zu nehmen, ein Beispiel dafür gibt die stellenweise völlige Ausserachtlassung der Beziehungen» (5). In his Einleitung, however, he refers to «die Eigenart dieses oft mehr als Nominale zu bezeichnenden Traités» (5), his demotion of the work from a «treatise» to a mere «wordlist», like Jeanroy before him, unwittingly recognizing the predominance of the lexis in Bibbesworth’s text, yet neither of them drew from this observation any conclusions regarding the author’s purpose in compiling the work. Schellenberg’s thesis has been quoted in some detail in order to show that, schooled in the medieval French grammar of the day, he was broadly in accord with the findings of Miss Owen’s examiners as regards the unsatisfactory nature of the text and their method of approach to the thesis. However, he did at least suspect that the editor herself was responsible for some of the textual errors in the work, 27 «Was über Dichter und Werk wissenswert ist, findet sich in der Einleitung von A. Owen» (p. 5). 28 To give only one example of this, noticed quite by chance when looking for something different, the verb herbergier is recorded in no fewer than 36 forms in Gdf. 4.455a. Anglo-French in rural England in the later thirteenth century although he failed to follow through his suspicions. Directly beneath the title of his final chapter «Zu erklärende Fehler und Besonderheiten» (64-69) he writes in brackets «Zum Teil auf die unsorgfältige Behandlung des Textes zurückzuführen», proof that he must have been aware of the potential danger inherent in equating uncritically Miss Owen’s edition with the manuscript on which it was ostensibly based. Schellenberg’s tentatively sceptical attitude towards Miss Owen’s handling of the manuscript was repeated more strongly in Koch’s article dealing with the Middle English glosses, where his earlier familiarity with the Bibbesworth manuscripts housed in what is now the British Library made him aware of what he termed «die mancherlei Unvollkommenheiten» in respect of both the French and English in her edition, and he rightly observes bluntly that: «Ausserdem sind die Variantenangaben der Herausgeberin . . . weder vollständig, noch immer genau genug, ja, mitunter geradezu irrig» (31), a far cry from the positive report of the examiners, but a judgement based on solid manuscript evidence, which theirs was not. Regrettably, as in the case of Schellenberg, this awareness of the defects in the editor’s handling of both languages of her texts was not set out in detail by Koch and did not become widely known to future readers of the Tretiz. The overall similarity between Schellenberg’s approach to the text and that of the examiners, even though there is no evidence of his being familiar with their review, is indicative of the generally dismissive approach to Anglo-French at the time of its publication. The lack of interest in the Tretiz on the part of the philological community is reflected in the small circulation of the edition compared with that of the standard works on the grammar of medieval French. Nyrop’s orthodox sixvolume Grammaire historique de la Langue française, published in 1930, went through four editions in as many years, and Miss Pope’s From Latin to Modern French, with especial Consideration of Anglo-Norman, Phonology and Morphology of 1934 has been reprinted more than once, has been widely quoted and has consistently figured in the reading lists of British universities for the last seventy years. Miss Owen’s thesis is not included in the Bibliography of the original edition of this work or its reprints. In contrast to the widespread success of these publications, Koch states that he was unable to find a copy of the Owen edition for use in his Anglia article of 1934 and had to resort to asking the Bibliothek des Romanischen Seminars zu Berlin for help 29 . The negative attitude adopted by scholars in the twentieth century towards Bibbesworth’s Tretiz as printed by Miss Owen contrasts markedly with its positive reception by his compatriots in his own day. Its popularity is evident not only from the number of surviving manuscripts but also from the proof of its wider dissemination and influence in the later Middle Ages that emerges from other sources.The early fourteenth-century Anglo-French Nominale sive Verbale (Skeat 1906) and Femina (Rothwell 2005) from the early fifteenth with their complete Middle English translations are not independent works in their own right despite their 117 29 «die . . . Ausgabe, von der ich mir nur mit Mühe ein Exemplar verschaffen konnte» (p. 31). William Rothwell 118 titles, but adaptations of basic material taken from sections of Bibbesworth, the Nominale being a fully glossed copy of a small section of the Tretiz and Femina adding to a similarly fully glossed (but again truncated version) of the text a moralising element based on Urbain le Courtois and the Proverbes de bon enseignement of Bozon, evidence of a breadth of interest present in the educated section of English society in the later medieval period that will be dealt with later. Although the Nominale omits much of the original Bibbesworth text, it greatly expands his list of bird-names. In an article entitled «The Taxonomy of Bird-Naming in Anglo- Norman and in Channel Island Patois» (Evans 1993) the late Dafydd Evans pointed out that: «The Nominale doubles Walter’s total» (111), one of the rarer varieties of bird being the praele glossed as «buntynge» (113; v. 805 in the Nominale) 30 . The same word occurs again twice with the spellings prel and prelle in the later manuscript of the Tretiz from the early fifteenth century housed in All Souls, Oxford (O f. 334rb, v. 407 and 409). This uncommon word is thinly attested in Gdf. 6.363c under the headword praieor, with the forms praiere and praiiere in his quotations, and T-L 7.1691 give the forms praier and praiere, referring the reader to Tilander’s Glanures lexicographiques (Tilander 1932: 209), where evidence is given of the word being attested on both sides of the Channel. It does not figure in Miss Owen’s footnote variants or her Glossaire, but its presence in the Nominale and the Oxford version of Bibbesworth show that other Englishmen a century after him must have been similarly familiar with a far wider range of French vocabulary than might be expected in what has been called the «period of degeneracy» of Anglo- French (Pope 1934: 424). In his article, Evans went on to quote a letter in the collection of Anglo-Norman Letters and Petitions edited by M. D. Legge for the ANTS (Legge 1941), which gave a long list of bird-names, dividing them into three distinct categories. Bibbesworth was clearly not alone in England in being able to provide his compatriots with vocabulary from the less familiar areas of the French lexis more than two centuries after the Conquest. What is more, in a number of other Anglo-French texts dating from around the same later medieval period as the Tretiz and quite distinct from the various Bibbesworth texts themselves may be seen unmistakable «borrowings» from the Tretiz. The editor of the fifteenth-century Orthographica Gallica (Johnston 1987) writes concerning lists of French words found in different manuscripts of his text that: «The lists are plainly compiled with indebtedness to other writers, e. g. Walter of Bibbesworth» (26). These «borrowings» from Bibbesworth are scattered over p. 26 (L. 50) of the variant readings and extend into p. 27. Additionally, in the base text of the work five lines of quasi-homonyms treated by Bibbesworth in different sections of his Tretiz are to be found on p. 15 (L. 50), followed on p. 17 (L. 86) by another example of the same feature: «Item habetur diversitas inter . . . kyvil et kevil» 31 , and then by examples of cases where Bibbesworth shows that one English 30 The form preal is found in the Bibbesworth ms. O v. 406, 407 and 409. 31 Misread by the editor as kynil and kenil. Anglo-French in rural England in the later thirteenth century term needs to be translated by several different words in French: «Item habentur diversa verba gallica pro isto verbo anglico ‹reed›, videlicet rous chivaler, chival; et harang soor; escue de goules, une rose vermaile» (L87). Similarly, on p. 18 (L88) is found: «Item habentur diversa verba gallica pro isto verbo anglico ‹breke›: fruschez chaud payn; debrusez l’os; rumpez la corde; enfraignez covenaunt, depessez la hanap», all of which are copied more or less verbatim from v. 308s. and 1053s. respectively of the Bibbesworth text as set down in the Owen edition.Again, it might be assumed that the fifteenth-century Liber Donati (Merrilees/ Sitarz-Fitzpatrick 1995) would be entirely devoted to the Ars grammatica, but, after only seven pages of the expected «grammar» (7-14), the text moves on to give French equivalents of Latin verbs and adverbs (p. 14-18) and then the names of the days of the week and numbers (p. 18), followed by seven pages given over to so-called «dialogues» (19-26) which contain lists of birds, animals, parts of the body, clothing, etc., very reminiscent of the corresponding lists in Bibbesworth, together with other clearly recognizable «borrowings» amongst the «Variant Readings» (27-31). Similarly, in the Manières de Langage (Kristol 1995), which deal with the vocabulary necessary for the English traveller in France, the 1415 version contains a section on the parts of the body (77s.) that follows in abbreviated form the similar section in Bibbesworth (v. 40s.), and a list of clothing on p. 78 is reminiscent of a similar list on p. 61-62 of the Owen edition. The long list of fish names on p. 11 and 43-44 of the Manières resembles the one given in the B O 4 manuscripts of Bibbesworth referred to earlier. More generally, on p. 145-47 of her thesis Miss Owen herself gives Extraits de Nominalia from colleges in both Oxford and Cambridge which deal with glossing from Latin or French into English and concentrate especially on Bibbesworth’s sets of homonyms and synonyms which the examiners failed to take into account when compiling their review article. All these extraneous pieces of evidence bear witness to the currency and influence of the Tretiz in England during the whole of the fourteenth century and beyond, despite its reprehensible grammar, yet the importance of the work was not recognized. For example, the Tretiz may be shown to be useful in respect of the history of modern English. The OED affirms that the adjective raucous is not found in English before 1769, being belatedly taken from the Latin raucus. The MED in its turn lists only rauc, not the fuller raucus, deriving the word from the «L[atin] raucus & OF rauque», its earliest attestation coming from a fifteenth-century translation into Middle English of a medical treatise by the French surgeon Chauliac, where the form rauc in the Middle English text is glossed by the genuine English hose (i. e. ‘hoarse’) 32 , suggesting that rauc was not generally accepted as being an English term at that time. For French, the form rau (corrected to raus) is given by Gdf. in a quotation from the later thirteenth century meaning enrhumé, enroué (6.622a, correction in 10.489a) and T-L 8.1340-41 under ro provide numerous examples of 119 32 «Rauc (adj.) Chauliac ? a1425 . . . pe lepre, . . . ffoulenez of pe lippez, Rauc, i hose, voyce [L. vox rauca] . . . ». William Rothwell 120 raus/ rowes/ roie, etc. in both continental and insular works from the early twelfth century, but there is no mention of «rauque». However, the form recous is given as a French term in no fewer than three manuscripts of Bibbesworth’s Tretiz with reference to the sound of the cuckoo 33 . In the B and G manuscripts it is not glossed, but the later O manuscript glosses it as hors (i. e. ‘hoarse’), so it would seem that the modern English raucous may have been first attested as a «French» word not found in France itself, long before its earliest attestation in the OED. Remaining with the «cuckoo», the point of Bibbesworth’s play on words - when he says that the sound of the cuckoo is heard more frequently in the lady’s chamber (oriol) than that of the (golden) oriole (oriol) - seems to have been missed 34 . The sound of the cuckoo is a euphemism for ‘cuckolding’. Similarly, the OED’s first attestation of soil as a noun meaning ‘The earth or ground’ is dated as «13..» i. e. some time in the fourteenth century, being derived from «Anglo-French soil, soyl and app[arently] representing the Latin solium», but the Latin dictionaries do not support this derivation. The MED under «soil (n. 1)» with the same meaning as the OED entry again gives the derivation of the word as coming from the Anglo-French soil, a «var[iant] of OF sueil, suil, souil» and dated «c1400». The Trésor de la Langue française states that its sol 1 with this sense dates only from the fifteenth century and comes from «haut bret[on].» The new Petit Robert also dates the word from the fifteenth century, but as coming from the Latin solum. The true situation is simpler, the word not only being used by Bibbesworth in the late thirteenth century «En bace tere ad bon soil», with soil glossed by the M. E. gloss erthe (v. 102 in the Owen edition), but attested even half a century earlier in the Anglo-French Rules of Robert Grosseteste (1240-42), where he gives advice to the landowner regarding the use of corn seed - «enquerez cumbien prent l’acre de semayl de cel soyl de terre», meaning that the landowner ought to find out how much corn seed per acre is needed for land with this type of soil. The correction of just some of the many textual errors in Miss Owen’s edition of Bibbesworth’s Tretiz set out above goes some way towards making possible a more accurate assessment of its contents and purpose in that it can no longer be dismissed out of hand as a failed attempt to teach French grammar. Yet so long as it is viewed in isolation there is no apparent reason for the appearance of a bilingual work of this kind in the England of the later thirteenth century and its subsequent spread through the numerous versions produced in the decades that followed. The link between the Tretiz and the society of rural England which alone 33 «Le chaunt de kokel (M. E. kockou) est recous» Owen 1929: v. 791; «Le chant de cocoil (M. E. cockow) est bien recous (M. E. hors)» ms. O f. 337vb; «Le chaunt de cocoel est ben rotous (l. rocous)» ms. B f. 104r. 34 G here has the erroneous orkoil to mean the bird, thus breaking the repetition of oriol, but B 104r has the correct oriol (tour) and oriol (wodewale), whilst O (f. 338ra) has oriol (tour) and arioule (wodewale). Anglo-French in rural England in the later thirteenth century would permit the correct attribution of the Tretiz to the domain of medieval English history rather than to that of medieval French grammar was provided only in 1971, being an indirect consequence of the publication of Dorothea Oschinsky’s Walter of Henley and other Treatises on Estate Management and Accounting (Oschinsky 1971: 396), the work not of a philologist, but a medieval historian. In her study of the many manuscripts which contain one or other of these agricultural treatises Dr Oschinsky came across copies of the Bibbesworth text set down together with literary works of different kinds. As a historian untrammelled by considerations of sounds and forms to the detriment of content and meaning, her association of Bibbesworth with a number of literary works correctly situated the Tretiz for the first time alongside manuscripts dealing with agriculture in the social context of later thirteenth-century England and enabled the Tretiz to be properly situated in the English society for which it was produced 35 . Her attention was focused principally on four agricultural texts produced in England in the second half of the thirteenth century. The earliest of these was written «in 1240 to 1242» (5) by Robert Grosseteste, bishop of Lincoln, better known to scholars working in the field of Anglo-French literature as the author of religious texts in keeping with his clerical calling. His text begins: «Isci comencent les reules ke le bon eveske de Nichole Robert Grosseteste fist a la contesse de Nichole de garder e governer terres e hostel», putting together the administration of a substantial household and the management of a large country estate. This treatise was followed by a lengthy anonymous Seneschaucy, which sets out in detail the many duties of each of the landowner’s chief officers together with the wide range of skills and knowledge that would be demanded of them, beginning with the seneschal himself (the steward) and working down the hierarchy. The third text is Walter of Henley’s Hosbondrye, its title perhaps an influence on Bibbesworth’s statement in his Prologue that he will deal with the French of «husbondrie» and «manaungerie». Henley’s work is more personal than the Seneschaucy, taking the form of parental advice from father to son on the techniques and difficulties of estate management. The fourth text is another anonymous treatise on husbandry. Of particular interest for present purposes is the Walter of Henley work which is thought to have been written in about 1280, around the same time as Bibbesworth’s Tretiz. When viewed as a whole, these agricultural texts in Anglo-French show the importance of the rural economy in thirteenth-century England and, consequently, of the written Anglo-French language in which its methods of management were expressed. The four treatises had been first published together by Elizabeth Lamond (Lamond 1890), but without the wider literary and social perspective brought in by Dr Oschinsky, and they remained for almost a century confined to the sphere of me- 121 35 The indissoluble link between language and the society which uses it was firmly established in Möhren 1986: 17: «Wortforschung ohne Sachforschung [ist] nicht möglich»; «Geschichte und Philologie erhellen sich gegenseitig. Und beide sind Hilfswissenschaften der Menschheitsgeschichte . . . ». William Rothwell 122 dieval historians without any link being made with other types of text in Anglo- French produced in the same period. In France itself, Louis Lacour had published a Traité d’Économie rurale composé en Angleterre au XIII e siècle as early as 1856 (Lacour 1856) 36 . His first footnote (123) is a quotation from L. Delisle’s preface to his even earlier Études sur la condition de la classe agricole en Normandie in praise of English agriculture: «C’est en Angleterre que les écrivains du moyen âge semblent avoir eu le plus de goût pour l’économie rurale. Ils nous ont laissé sur cette matière des travaux originaux du plus vif intérêt.» Lacour himself opens his edition of the Anglo-French Traité d’Économie rurale with a similar eulogy: «Au treizième siècle, l’agriculture en Angleterre était arrivée à un degré de perfection que nous ne devions pas atteindre de si tôt», with again a similar quotation from Delisle in support: «Le lecteur aurait été surpris de la perfection qu’avait dès lors atteinte en Angleterre l’exploitation des champs» (128), and goes on to praise the organization of the work-force from landowner down to labourer, also its homogeneity and its common interest in the success of the enterprise. The changing balance in thirteenth-century England between the language introduced by the conquerors in 1066 and the rising tide of Middle English must have provided to at least some extent the incentive to set down the French agricultural texts as a defensive measure. Lacour’s publication was known to Dr Oschinsky who lists it in her Bibliography a century or so later 37 , but without referring to it in the body of her study of the agricultural treatises. The end of the Traité d’Économie consists of paragraphs entitled «De faire payn», «De faire cervoise», «Médicine pur breez» and «Médicine pur cervoise rouge» which are very reminiscent of the corresponding sections of Bibbesworth’s Tretiz dealing with the making of bread and ale, thus establishing a significant connection between the highly praised work produced in France and the error-strewn text written in England as published by Miss Owen. Dr Oschinsky’s book not only illustrates the methods of farming in later thirteenth century England and the specialized French vocabulary that goes with them, but also shows the wide range of professional skills brought together by the demands of this successful agriculture. She points out that the texts were «compiled for the legal public» (6), and also that «high farming made efficient organization and accurate methods of accounting and auditing essential» (4), going on to state that «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» (73-4), continuing « . . . the new profession of qualified baillifs put estate management on a scientific basis and intensive farming was encouraged by the support of 36 I am indebted to Professor D. A. Trotter for providing me with a copy of this work. 37 His initials are given as «M. C.» in the bibliography, but there is no doubt about the text involved. Anglo-French in rural England in the later thirteenth century Statutes. Young baillifs, trained on the lines described in the Seneschaucy and Walter . . . needed efficient methods of compiling accounts . . . » (233). The principal language used in these texts on accounting and the law was Anglo-Latin containing many terms «borrowed» from French, together with some French (e. g. 475-78). Along with this account of the different areas of intellectual activity in later thirteenth-century England brought together by the promotion of agriculture, Dr Oschinsky’s book ranges more widely as she examines the diverse works which accompany the treatises in the manuscript collections.An example of the varied areas of knowledge to be found in such manuscripts and consequently in the educated community who collected them at this time may be seen amongst the intellectual writings in a manuscript now in the Cambridge University Library. Dr Oschinsky says that it «is thought to have been the work of John of Longueville, an eminent Northhampton lawyer who lived during the late thirteenth and early fourteenth centuries». One section of this very substantial manuscript is «a collection of texts useful in particular to estate officers», consisting of legal texts, «a treatise on letter writing, a treatise on accounting, the Extenta Manerii, the Seneschaucy, . . . and Walter [of Henley]» (24). She does not indicate the nature of the rest of the contents of this particular collection, but when dealing with the wide dissemination of the agricultural works in which «eighty-four manuscripts include one or more of our treatises» (10), she mentions in respect of one of them, the Bodleian Selden Supra 74 dating from about 1300, that: «The MS. is of interest because its content shows that the texts by Walter of Bibbesworth, Walter of Henley, and Nichole Bozon were considered in some circles as equally ‹good› reading, and because the comparison of the three treatises shows that the authors knew each other’s work.» (44). Later, in a footnote (154 N2), she develops this point about the relationship between the three works: «This text (i. e. Les proverbes de bon enseignement de Nichole Bozon), together with Walter (i. e. Walter of Henley’s Hosbondrye) and the nominale by Walter of Bibbesworth (i. e. the Tretiz) was copied in MS. 68 . . ..» 38 . She might have added that in this manuscript the final verse of the truncated text of Bibbesworth’s Tretiz to which she refers leads directly into the blurred opening line of a version of Walter of Henley’s Hosbondrye, so the two texts could not be closer together physically. A corresponding link between Bibbesworth and Bozon is to be seen in another manuscript, the one that houses Miss Owen’s C version of the Tretiz. This came originally from the Phillipps collection in Cheltenham as ms. 8336 and is now in the British Library as Additional 46919. In the Introduction to his edition of Nine Verse Sermons by Nicolas Bozon, Brian Levy (1981: 1) wrote regarding the location of his material that: «The poems are all found together in a single manuscript (MS British Library Additional 46919)» 39 , so it would be strange indeed if the Bibbesworth text alone in these two compilations were unintelligible. 123 38 This number indicates its position in Dr Oschinsky’s work, not the Bodleian numbering. [74] 39 This is the «Thirlestaine House 8336» ms. later acquired by the British Library as Additional 46919 and given the identity C by Miss Owen. William Rothwell 124 The connection between Bibbesworth and Bozon had been noticed as early as 1884 in an article by Paul Meyer (1884: 499s.) dealing with the contents of the Phillipps 8336 manuscript. Alongside Bibbesworth’s Tretiz, which Meyer apparently did not understand, calling it «ce curieux opuscule» (500), he makes reference to the presence of the poems of Nicholas Bozon, writing: «la portion la plus considérable, celle qui contient les poésies de Bozon et de Gautier de Bibbesworth . . . n’a pas été étudiée jusqu’à ce jour» (499) 40 . On p. 31 of her Introduction to the Tretiz Miss Owen lists the other varied contents of this manuscript: «Les Chastel de Leal Amour, l’Art de Venerie, par Mestre Guillaume Twich, un traité sur les oiseaux, et un autre sur l’amour, des prières à sainte Marie, une Description de Chivalerie par Hue de Tabarie.» This very disparate group of texts accompanying the Bibbesworth work testifies both to its acceptance by his contemporaries and also to the wide range of their literary interests, in sharp contrast to the modern approach which tends to separate the contents of medieval manuscripts into specialized sections to be studied by different groups of scholars working independently in each particular section. Although Meyer did not pursue his linking of Bibbesworth with Bozon, his remark complements the similar linking of Bibbesworth and Walter of Henley referred to above, thus bringing together Bozon’s many religious writings, Henley’s agricultural treatise and Bibbesworth’s essay on the French vocabulary relating to that agriculture, confirming Dr Oschinsky’s statement regarding the eclectic reading of the educated ruling classes in later medieval England and broadening the perspective in which the Tretiz needs to be judged. Bibbesworth and Bozon may again be seen to be linked by an unusual detail common to both their works. In his Contes Moralisés 41 Bozon says that the industrious badger is driven from its newly-made set by the fox which leaves its excrement at the entrance, driving the badger away, and then installs itself in the set when the badger has gone. This tale is said by the editors of the Contes to originate in a version of the Physiologus by Bartolomaeus Anglicanus entitled De proprietatibus and to be found only in that text. However, Bibbesworth mentions this same feature briefly in his section on the names of animals in his Tretiz: «Jeo vi vener un graunt tesschon (M. E. brocke) / Ki ad guerpi sa mansioun / Pur les fens du gopil (M. E. fox) / Ki l’ad mis en exil» 42 . This linking of Bibbesworth’s Tretiz on the one hand with contemporary works well established in the canon of Anglo-French literature and on the other hand with texts dealing with the techniques of medieval English agriculture moves it from the category of an isolated and severely flawed essay on the grammar of me- 40 For details of the present state of publication of the many works by Bozon see AND 2 p. xxxi. 41 Toulmin Smith/ Meyer 1889: Conte 144, p. 179, and Owen 1929: v. 799-802. 42 Since the Contes are attributed by the editors to the early decades of the fourteenth century whilst the Tretiz is commonly said to date from around 1280, it could be argued that Bibbesworth’s work might have provided the story in the first place, but the version in the Contes is fuller. Anglo-French in rural England in the later thirteenth century dieval French into the mainstream of the literary and social history of medieval England. By revealing the juxtaposition in the same manuscript collection of three texts belonging to what have been regarded in modern times as separate areas of knowledge - medieval French language, medieval French literature and medieval English history - Dr Oschinsky drew attention to the need for a more integrated view of the legacy of French in medieval England. Further evidence in support of such breaking of the barriers between works in Anglo-French that have traditionally been considered in modern times to be the preserve of different groups of scholars is provided by the other works of Bibbesworth himself that have received little attention and are far removed from the subject-matter and written style of the Tretiz. As early as 1841 Thomas Wright published in his Reliquiae Antiquae (Wright 1841), an Anglo-French poem which he entitled «Dialogue between Henry de Lacy and Walter Bibbesworth on the Crusade» (134-35). This was published again by Suzanne Thiolier-Méjean (1980) as an appendix to an article entitled «Croisade et registre courtois chez les troubadours» 43 . The poem would seem to be a joint effort by the two friends. It begins: «Ci commence le counte» and Henry de Lacy proceeds to lament that he is torn between his duty as a Crusader - «Ore sui croisee, pur Deu servir» - and his love for his lady - «Cele au cler vys, au ryaunt oil.» He knows that if he does not go to join the Crusaders in the Holy Land everyone apart from his beloved will hate him and he will be dishonoured, but if he leaves her he will lose her love. Bibbesworth replies («Respont sire Gauter») that love is like the honeysuckle clinging to a fine tree and smothering it, in an apparent reference to Marie de France’s Lai du Chievrefeuil. To save the tree it must be cut down near the ground, so that new branches may flourish. Henry de Lacy remains unconvinced and says that he is constrained «Par force d’amour qe tut veint», to which Bibbesworth replies that his friend must not abandon Christ «Qui fust de un glayve au quer enpeint» in favour of a woman «qe vus veut mener Au fu d’enfern qe ja ne esteint» and ends with the harsh admonition: «Cil qi de gré se veut noier N’en doit par raisoun ester pleint». It is noteworthy that this Dialogue, whether the work of one or both parties, does not display any sign of the abundance of grammatical error found in the Owen edition of the Tretiz and its numerous footnote variant readings. The reason for this may perhaps be that, not being of such general interest as the Tretiz with its wide-ranging vocabulary, the more private Dialogue would have been less likely to be copied by unknown scribes of unverifiable knowledge and competence who provided the various versions of the Tretiz and to whose errors were added those of Miss Owen herself. This Dialogue also gives a hint of the place of Anglo-French in the intellectual climate of Europe in the later Middle Ages. The 1980 article shows that the Bibbesworth-Henry de Lacy poem is not to be seen as an isolated piece of writing from 125 43 I am indebted to Professor D. A. Trotter for drawing my attention to this later article and providing me with a copy of it. William Rothwell 126 a distant unimportant and unlettered outpost of French, but is on a par with numerous similar expressions of unease about the Crusades found across the Channel in both troubadour and trouvère circles. The writer quotes a couplet by Conon de Béthune dating from 1189 in which he confesses that, although his body will go on the Third Crusade, his heart will remain behind with his beloved. Two similar confessions are cited from poems by Thibaut de Champagne half a century later, whilst on the troubadour side the names of prominent poets such as Raimbaut de Vaqueiras, Marcabru, Sordel and Gaucelm Faidit are to be found associated with the same kind of text. Furthermore, nearly half a century after Wright’s Reliquiae Antiquae, Paul Meyer (1884: 531-32) included in his article referred to above short extracts taken from two other poems by Bibbesworth found in the same manuscript as the version of his Tretiz mentioned earlier (ms. C, now British Library Additional 46919, Cheltenham 8336 in Miss Owen’s list, 31). The first of these is in honour of the Virgin, the second in honour of women, but Meyer printed only a few verses from what are poems of some considerable length 44 , regarding them as worthless on account of their play on words and lamenting that «le goût des vers équivoqués avait pénétré en Angleterre» (531). This feature not appreciated by Meyer dominates the two poems from beginning to end, the author displaying again a talent for juggling with words that is reminiscent of his play on (quasi-) homonyms in his Tretiz, a linguistic skill that would call for a thorough command of the French language, however its literary quality might be judged by later generations. The seriousness of the poems is made clear by Bibbesworth’s Introduction: «Cy comencent les dytees moun syre Gauter de Bybeswurthe: Regardez, lysez, apernez». Like the Dialogue between Bibbesworth and Henry de Lacy, these poems show no trace of the gross grammatical errors allegedly found in the Tretiz, and so they reinforce the view that the blame attached to the author in this respect ought to be transferred to scribes and editor. When added to the contacts with writers across the Channel referred to above concerning the crusades, these poems by Bibbesworth lead to the other factor which must be taken into account when assessing the importance of the Tretiz in the late thirteenth and fourteenth centuries, namely the social position of the personages with whom he was in regular contact. When identifying the writer in her Introduction, Miss Owen mentions that the Bibbesworth family held land in Essex and that Henry de Lacy, treated as Bibbesworth’s equal in the Dialogue, was not only Count of Lincoln but a counsellor of Edward I (23). Indeed, the Lacy family also had an extensive country estate in Dorset over to the west of the country, in the area now known as Kingston Lacy 45 . Walter of Henley too was a prominent landowner, and Bibbesworth’s patroness herself would be linked later to the de Vere family, counts of Oxford, by the marriage of her grand-daughter. All these landed families were originally of French extraction and would have been most un- 44 They occupy f. 92v to 95v. 45 The present estate, in the care of the National Trust, covers over 8000 acres. Anglo-French in rural England in the later thirteenth century likely to allow themselves to be associated with a corrupt French text littered with all manner of errors as printed by Miss Owen. An understanding of Bibbesworth’s position in society helps to explain the final section of his Tretiz entitled «Ore le fraunceis pur un feste araer», which moves away from the vocabulary of the countryside to set out the French terminology associated with the organization of a feast - «un mangerie» (v. 1100 in the Owen edition of ms. G). This departure from his teaching of «tut le fraunceis . . . de husbondrie e manaungerie» is not explained or even mentioned by the editor. The confusion in Miss Owen’s Glossaire between manaungerie and mangerie (i. e. ‘management’ and ‘feast’) referred to earlier which showed that she did not understand the synonymity of husbondrie e manaungerie as agricultural terms would also have prevented her from appreciating the purpose of this description of a mangerie. When set in the context of the upper classes in English society towards the end of the thirteenth century, this concluding chapter of Bibbesworth’s Tretiz may be seen to be neither aberrant nor isolated. Ample evidence of the presence of a rich Anglo-French culinary vocabulary in use in England as early as the Bibbesworth era is to be found in the «Culinary Collections» edited by Hieatt/ Jones 1986. A later example of this branch of literature from quite different sources has recently been brought into the public domain by Lisa Jefferson 1998. Composed more than a full century after Bibbesworth’s Tretiz, two of these feasts were organized to celebrate the awarding of degrees at Oxford, the third being held in London in honour of the King’s Sergeants. In the light of the Bibbesworth manuscripts discussed above, the use of language in these feast menus at such a late date is of interest for the history of both English and Anglo-French.The article states that: «All three of the languages of medieval England (English, French and Latin) are deployed, and the admixture is such that one cannot always be sure whether one should count a word as English or French (blamange, flampayn, jussell). French, or rather Anglo-French, is the base language, and is employed for almost all culinary terms (endoré, enarmez, en comfyt). English words have been gallicized (rostez, bakez), and French word-order is preserved (chykyn farsé) except in a few instances. The small connecting words of a phrase are also usually French (ové, de, en) and names for animals and birds are purely English . . . Latin appears in the abbreviated ‹Epi. Slez . . . › and in the interjected ‹vel bakemet raylis roiall vel doucetes›.» (246). The list of the linguistically mixed terminology used in these menus given on the three pages which follow bears witness to the extent of this culinary vocabulary in use in fifteenth-century England. Thirty years before the «Three Fifteenth-Century Feast Menus» evidence of this kind of fluctuation between French and English in the area of feasting, but at a much earlier time and at a much lower social level, had been published in «The Guardian» newspaper on Friday, December 22, 1967 (6) under the title of «A fourteenth century feast». The subject of the article is an English poem of 250 lines, regrettably said to be «too long to print here in its entirety», which tells of the antics 127 William Rothwell 128 of the «sweaty Swinkers 46 and true Drinkers of Tottenham» on the occasion of a mock tournament held in the alehouse after «a Sports Meeting, beside the high way . . ., with the hand of the landlord’s daughter as the prize». The menu for «this joyous mockery of a lord’s manorial banquet» held «in ridicule of the foreign imported fashions of the period» is understandably quite different from that of the decorous proceedings described by Bibbesworth and those recorded in the «Three Fifteenth-Century Feast Menus», but the vocabulary of the food enjoyed by the «labourers, bakers, potters, and working fellows» who took part in this farce contains French terms alongside English ones (e. g. capon, mortrews, browet of Almayne, pestils (cf. MED pastil)). These texts, the one serious, the other mocking, show that the ending to Bibbesworth’s Tretiz was not just an inexplicable and irrelevant afterthought. The organisation of feasts is mentioned in one of the agricultural treatises referred to earlier. A section in Grosseteste’s Rules from the mid-thirteenth century sets out in detail the hospitality in hall to be offered to honoured guests, the seating arrangements, the saying of grace, the parading of the baker and butler before the lord carrying their food and drink, the rules for the order in which guests were to be served, etc., a «lord’s manorial banquet» as mentioned in the Tottenham farce (Oschinsky 1971: 402-6). Instruction in the arrangement of this kind of ceremonial feast would have been an essential part of the education received by the rising generation of landlords on large estates many decades before the «Three Fifteenth-Century Feast Menus». The passage of culinary terminology from French into English that may be seen in the final section of Bibbesworth’s Tretiz, in the «Three Fifteenth-Century Menus» and the «Guardian» text contrasts markedly with the very limited adoption into modern English of the rural vocabulary that forms the basis of Bibbesworth’s Tretiz. His prologue in particular lists the terminology of agriculture that he intends to pass on to the offspring of his patroness - arer (‘to plough’), rebingner (‘to turn over ground a second time’), waretter (‘to plough fallow land’), semer (‘to sow’), searcler (‘to hoe, weed’), syer (‘to reap’), fauger (‘to mow’), carier (‘to cart’), muer (‘to stack’), batre (‘to thresh’), ventre (‘to winnow’), mouwere (‘to grind’), pester (‘to knead’), brescer (‘to brew’), bracer (‘to malt’), haute feste araer (‘to arrange a feast’), but of all these terms only carier (‘to cart’), batre (‘to beat’) and feste araer (‘to arrange, organise a feast’) have come through into modern English. Similarly, the English vocabulary of the countryside in general has not been substantially modified by Bibbesworth’s work, only a few of the many French terms it contains being present in the lexis of modern English. The explanation for this may be found in the economic and social condition of medieval England in the later thirteenth and fourteenth centuries. The complimentary comments of Delisle and Lacour mentioned earlier regarding the advanced state of English agriculture as seen in French texts show that there was 46 I. e. ‘labourers’. Anglo-French in rural England in the later thirteenth century no need for the English landowners of French descent to bring the relevant vocabulary over from France, but, whilst they and a handful of their principal officials might regulate the overall activities of their work-force through the medium of French, native or acquired, the all-important direct contact with the soil itself, its products and the implements required for their cultivation would be made by a much larger contingent of monolingual and unlettered English workers whose passage through life would leave little trace and whose silence hinders any detailed understanding of the linguistic situation obtaining in the English countryside in the later Middle Ages. Often working in thinly-populated rural areas, these farm-labourers would have had little immediate contact with their French-speaking masters and many would have spent their lives in one small community 47 where they would be less exposed to French influence from outside than the more cosmopolitan inhabitants of large towns, at least some of whom would be working in national or local organisations where French would remain the language of communication for many years to come, or were more mobile for reasons of trade and commerce. Also, the number of intermediary officials working at different levels in the agricultural community who would have been capable of using both languages with varying degrees of competence from one generation to another and so acting as intermediaries between the francophone masters and their anglophone workers must remain an unknown quantity. Once cleared of the layers of error for which Bibbesworth was not responsible and viewed as a social phenomenon produced at a particular time, in a particular area and for a particular group of readers, the Tretiz may be seen as an important historical document that marked the end of an era. The fact that the work was commissioned by Dyonise de Mountechensi specifically to maintain the presence of the French voice in the practice of English agriculture in the fertile East Midlands based on Lincoln, the diocese of Grosseteste, is an indication that she must have been aware of the threat to its future continuation in that role in the face of the rising tide of English and must have chosen the bilingual Bibbesworth as being best qualified to carry out the task of preservation. Yet although the Tretiz would be copied and modified right through the fourteenth century, with some of its material «borrowed» even in the fifteenth century into texts not concerned in any way with estate management, the work as a whole made little lasting impression on the linguistic history of England, the language of agriculture at the level of the labourers being too deeply rooted to be changed from above in favour of the language of a steadily declining rural aristocracy. The French terminology which transformed Middle English into modern English would only rarely be related to agriculture or the natural world of the country- 129 47 As was mentioned in N14, the Seneschaucy forbids the workers in charge of animals to frequent fairs, markets or taverns and to indulge in wrestling-matches, activities which would reduce the isolation of small communities. William Rothwell 130 side 49 , but rather to more urban concerns such as governance and the law at national and local level, or trade and commerce at home and abroad. Using Chaucer as being representative of English in the later fourteenth century, his work shows little sign of French grammar, but if the French terminology that pervades his writings were to be removed, they would be incomprehensible (cf. Rothwell 1996; 2006; 2007). Swansea William Rothwell Appendix A selection of 1) French words in the Tretiz which have similar forms but different meanings and 2) English words which are to be translated by more than one French word. 1) French cece (‘stumble’) chece (‘fall’) v. 19; greve (‘parting in hair’) grive (‘thrush’) v. 30-32; toup (‘lock of hair’) toup (‘top of flax’) toup (‘spinning top’) toup (M. E. ‘tup’, ‘ram’) v. 34-38; temples (‘temples of head’) temples (‘churches’) v. 41-42; graciose (‘comely’) chaciouse (‘bleary’) v. 43-44; rupie (‘nasal mucus’) rubie (‘ruby’) v. 46-52; chouue (‘chough’) jouwe (‘cheek’); la levere (‘lip’) le levere (‘hare’) livere (‘pound’) livre (‘book’) v. 61-66; ascel (‘armhole’) escel (‘axle’) v. 99-100; coste (‘rib’) costee (‘side’) v. 125; gareters (‘garters’) charetters (‘carters’) v. 141-43; la zure (‘calf’ of leg) ensure (‘on, above’) le assure (‘protects him’) v. 145-47; keviles (‘ankles’) kyviles (‘pegs’, ‘dowels’) v. 151-56; la char (‘flesh’) le char (‘cart’), eschar (‘scorn’) v. 169-74; apel (‘knell’) apel (‘appeal’) v. 179-80; ceintez (‘gird’) enceintez (‘make pregnant’) v. 187-90; jaroile (‘squawkes’) garoil (‘trap’) v. 260-66; baleie (‘bleats’) bale (‘dances’) bale (‘bag’) baal (‘yawns’) baille (‘gives up, hands over’) v. 285-88; se espreche (‘stretches’) presche (‘preaches’) pesche (‘fishes’) hesche (‘hook’) v. 289-92; fresche (‘fallow’) freische (‘fresh’) v. 293-94; lesche (‘píece, slice’) lesche (‘licks’) v. 297-98; reyne (‘queen’), reyne (‘frog’) v. 318-19, rey (‘king’), rey (‘furrow’) v. 320-21; rastel (‘rake’), rastuer (‘a scraper for cleaning a kneading-trough’); v. 387-90; littere (‘hay’/ ‘straw’), litteir (‘vehicle’) v. 397- 98; fusil (‘spindle’ for weaving), fusil (‘fire-iron’) fusil (‘mill-spindle’) v. 435-40; breser brece e bracer cerveise (‘to malt malt and brew beer’) v. 455-93; grele (‘hail’/ ‘small’) v. 584-85 49 ; tonn (‘thunder’ verb/ ‘cask’/ ‘go numb’) v. 581-82; parele (‘red dock’), parel ‘pair’), v. 654-701; varole (‘caterpillar of the cabbage butterfly’), verole (‘small-pox’), virole (‘ferrule’) v. 622-32; coingner (‘quince tree’) coingnier (‘to put a person in the stocks’), coigner (‘wedge’) and coigne[r] (‘coiner’) v. 681-90; naer ( (‘to swim’), noer (‘to drown’) and nager (‘to row’) negger (‘to snow’) v. 733- 39; essel (‘axle’), assel (‘protecting plate under cart-body’), ascel (‘arm-hole’) 867-68; poutre (‘a support beam’/ ‘a filly’) v. 947-48, arable (‘maple tree’/ ‘arable’) v. 699-700; poun ‘peacock’/ ‘pawn’ (at chess); ventrere (‘midwife’), ventrer (‘belly-rope’ (for horse) v. 875-79; etc. 48 The French herbe would supplement the English «grass» and provide a useful distinction between the medicinal and the agricultural; prune would allow a distinction between modern English «plum» and «prune» , etc. 49 The first grele in the sense of «hail» is not found in Miss Owen’s text because she missed out a complete verse - Nos averoms grisil puis q’il grele as was mentioned in N9. Anglo-French in rural England in the later thirteenth century 2) English ‘red’: rous, sor (of a horse or herring), goules (heraldic), rouge, vermaille (of wine) v. 308-14; ‘stack’ as verb: muez (of corn in barn), tassez (of corn in field) v. 347-48; ‘stack’ as noun: moye (M. E. reke), moiloun (M. E. reke), thase (M. E. stake), thas, v. 349-52; terms for ‘to blow’: suffler (of kitchen-boy fanning fire), venter (of wind in bushes), corneer (of hunter) v. 369-73; ‘to break’: Frussés (of bread) 1059, Debrusés (of bone) 1060; Rumpés (of rope) 1061; Enfreinés (of covenant) 1062; Partiez (of herring) 1063. Bibliography Benson, L. D. 1987: The Riverside Chaucer, Oxford Bolland, W. L. 1925: Manual of Year Book Studies, Cambridge Cerquiglini, B. 1991: La naissance du français, Paris Dean, R. J./ Boulton, M. 1999: Anglo-Norman Literature. A Guide to Texts and Manuscripts, London Evans, D. 1993: «The Taxonomy of Bird-Naming in Anglo-Norman and in Channel Island Patois», in: I. Short (ed.), Anglo-Norman Anniversary Essays, London: 105-34 Riley, H. T. 1859-62: Liber Albus. Munimenta Gildhallae Londoniensis, London Harvey, B. F. 1965: Documents Illustrating the Rule of Walter of Wenlok, Abbot of Westminster, 1283-1307, London Hieatt, C. B./ Jones, R. 1986: «Culinary Collections edited from British Library Manuscripts Additional 32085 and Royal 12.C. xii», Sp 61: 859-82 Jefferson, L. 1998: «Neville, Babthorpe and the Serjeants: Three Fifteenth-Century Feast Menus», Oxoniensa 63: 241-49 Johnston, R. C. 1987: Orthographia Gallica, London Koch, J. 1934: «Der anglonormannische Traktat des Walter von Bibbesworth in seiner Bedeutung für die Anglistik», Anglia 58: 30-77 Kristol, A. M. 1990: «L’enseignement du français en Angleterre (XIII e -XV e siècles): les sources manuscrites», R 111: 289-330 Kristol, A. M. 1995: Manières de Langage, London Lacour, L. 1856: «Traité d’économie rurale composé en Angleterre au XIII e siècle», BECh 4 e série 17: 123-41, 367-81. Lamond, E. 1890: Walter of Henley’s Husbandry, London Legge, M. D. 1941: Anglo-Norman Letters and Petitions, Oxford Levy, B. J. 1981: Nine Verse Sermons by Nicolas Bozon, Oxford Merrilees, B./ Sitarz-Fitzpatrick, B. 1993: Liber Donati, London Meyer, P. 1884: «Notice et extraits du ms. 8336 de la bibliothèque de sir Thomas Phillipps à Cheltenham», R 13: 497-541 Möhren, F. 1986: Wort- und sachgeschichtliche Untersuchungen an französischen landwirtschaftlichen Texten, Tübingen Oschinsky, D. 1971: Walter of Henley and other Treatises on Estate Management and Accounting, Oxford Owen, A. 1929: Le Traité de Walter de Bibbesworth sur la langue française, Paris Pope, M. K. 1934: From Latin to Modern French, with especial consideration of Anglo-Norman, Manchester Rothwell, W. 1960, «Medieval French Bureau», MAe 29: 102-14. Rothwell,W. 1996: «The Anglo-French Element in the Vulgar Register of Late Middle English», NM 97: 423-36 Rothwell, W. 2004: «Ignorant Scribe and Learned Editor: Patterns of Textual Error in Editions of Anglo-French Texts», The Anglo-Norman On-Line Hub, http: / / www.anglo-norman.net/ articles/ scribe.xml Rothwell, W. 2005: Femina, Anglo-Norman On-Line Hub, http: / / www.anglo-norman.net/ texts/ femina.pdf 131 William Rothwell 132 Rothwell, W. 2006: «Anglo-French and English Society in Chaucer’s The Reeve’s Tale», English Studies 87: 511-38 Rothwell, W. 2007: «Synonymity and Semantic Variability in Medieval French and Middle English», MLR 102: 363-80 Rotuli Parliamentorum 1767-77: Record Commission, London Schellenberg, G. 1933: Bemerkungen zum Traité des Walter von Bibbesworth, Berlin thesis Skeat, W. W. 1906: «Nominale sive Verbale», Transactions of the Philological Society 1903-1906: 1*-50* Thiolier-Méjean, S. 1980: «Croisade et registre courtois chez les troubadours», in: D’Heur, J.-M./ Chérubini, N. (ed.), Études de philologie romane et d’histoire littéraire offertes à Jules Horrent, Liège: 295-307 Tilander, G. 1932: Glanures lexicographiques, Lund Toulmin Smith, L./ Meyer, P. 1889: Les Contes Moralisés de Nicole Bozon, Paris