William Rothwell (1998)

Whilst it is now being increasingly, but far from universally, recognized that the French of medieval England cannot be arbitrarily dismissed as just a corrupt and decadent travesty of the ‘correct’ form traditionally claimed to have been the language of Paris and its environs from at least the thirteenth century onwards,¹ the pervasiveness of the independent development of Anglo-French, often at an early date, in areas of life far removed from the domains of literature and from the whole apparatus of the royal court, the courts of law, Parliament, the Church or other such elevated spheres has yet to be thoroughly explored and catalogued.

The very names ‘grocer’ and ‘greengrocer’ give an indication of the separate semantic development that characterizes Insular French when looked at from the viewpoint of the overall history of the French and English languages. The modern English ‘grocer’ and ‘grocery’ now have nothing in common semantically with the grossiste, or formally with the épicier or the épicerie (either as a product or as the premises where that product is sold). Similarly, there is now no linguistic link of any kind between the ‘greengrocer’ and the fruitier/marchand de fruits et légumes, or between ‘greengrocery’ and fruiterie (the shop) or fruits et légumes (the wares on sale in such a shop). Even though the origin of the English ‘grocer’ lies indisputably in Old French, the diverging paths taken by the two languages over the centuries cannot be adequately reconstructed by simply opening the dictionaries of French and English at the obvious entries and relying on the information given there.²This information can be misleading unless supplemented by further evidence from elsewhere. The modern Tobler-Lommatzsch has an entry grosseor (4.710-11) which just gives the gloss ‘Grosshändler’ and refers the reader to Godefroy, thereby implying that the term calls for neither quotation nor comment. However, the only attested form given in Godefroy’s single quotation under grosseor (4. 367c) is grossour, not the head-word as printed. As so often, Godefroy creates an etymologically ‘correct’ form to suit nineteenth-century philological orthodoxy, and the twentieth-century ^p2 Tobler-Lommatzsch follows blindly, regardless of linguistic reality. More importantly, Godefroy’s quotation is, in fact, Anglo-French and not, as might be assumed by the unwary, Continental French, being taken from a decree of Edward II, a fact not pointed out in either dictionary. Yet, in order to find genuine continental examples of the word the reader need only turn to the entry grossier, one page farther on in both dictionaries and defined as ‘marchand en gros, s’appliquant à plusieurs métiers’ by Godefroy and as ‘e[ine] Art Schmied, e[ine] Art Zimmermann’ by Tobler-Lommatzsch. However, in neither dictionary is the link made between the two terms, grossour and grossier, glaringly obvious though it may be. Disregarding, then, the mythical grosseor, the authentic grossour is simply one of the attested forms of this well documented term grossier in which the continental marker –(i)er denotes an occupation (cf. cuisinier, postier, boulanger etc.) and which becomes the standard -er in Anglo-French and English.

So much for the form of the word: the more important question of its meaning is likewise in need of clarification in the standard dictionaries. Godefroy’s quotations show that, from the middle of the thirteenth century up to the early eighteenth, dealing in bulk was the essential element of the trade carried on by the grossier, whether that trade be in ironmongery, comestibles or whatever. The attempt by Tobler-Lommatzsch to limit the meaning to blacksmiths and carpenters, without any mention of the wholesale nature of the business carried out and in flagrant disregard of Godefroy’s wide range of quotations, reveals a failure to understand this cardinal fact, and so the translations they provide are inaccurate. The nineteenth-century dictionary, not the modern twentieth-century one, is correct at this point. However, the modern French grossiste has not so far been found in print before the late nineteenth century, thus creating an awkward gap of nearly two centuries during which the meaning ‘wholesaler’ was apparently covered by neither the old grossier nor the new grossiste. As in many other cases of similar gaps, at some future date a programme of systematic lexicological research directed specifically at non-literary registers in mainland France may well reveal this hiatus to be illusory.

Whilst in France, at least to judge from the evidence available to Godefroy, the grossier would appear to have remained a wholesaler,³ in England the semantic shift to ‘grocer’ is beyond dispute, but there is a period in the latter part of the fourteenth century when it is not easy to determine whether we are dealing with the ‘wholesaler’ or the modern ‘grocer’, since the goods brought into the country by the wholesalers were largely products sold by the grocer. The early examples of the word, found in England slightly earlier than its first attestation on the Continent,⁴ pose no problem of meaning, indicating clearly the original meaning of’ ‘wholesaler’.⁵ Even as late as 1373 the archive of the Grocers’ Company in ^p3 London refers to la compaygnie des grossers, ⁶ and the details of their trade show it to have been wholesale, but by the same time, if not earlier, the appearance of the derivative noun grocerie ⁷ in the sense of ‘spices’ would suggest that the grosser in England might at that time have been synonymous with the earlier (e)spicer ⁸ who dealt in (e)spicerie, ⁹ and who might have been either a wholesaler or a retailer.¹⁰ From a footnote to Godefroy’s entry, in which he gives the plural grosseries as meaning in the dialect of Poitou ‘toutes les céréales excepté le froment’ (4.368a), it would appear that a similar process of derivation may also have taken place across the Channel, but without the word establishing itself in the standard language. This possibility is strengthened by a Middle French quotation given in the FEW under grossus (4.279b), where the plural form grousseries is defined as: ‘tous les effets de laine, de coton, dans une lessive, par opp[osition] au linge’. In England, according to the MED, ‘grocer’ had been taken into English by 1418, part of the merger of Anglo-French with Middle English that has produced the modern English language, containing so many French elements but remaining basically so different from the language spoken across the Channel.

The ‘greengrocer’ must clearly come after the ‘grocer’ had become established in his modern role, since it marks a distinction between the purveyor of fresh, perishable fruit and vegetables on the one hand and the trader in durable products such as sugar or spices on the other, with ‘greengrocery’ following naturally after the pattern of ‘grocer/grocery’. Long before these later developments, the MED has an entry ‘Rob. le Fruter’ as early as 1203, given as a ‘dealer in fruit’, with the more specialized sense of fruter as ‘purveyor of fruit and vegetables in a noble household’ coming as late as ‘c1475’ in the Ordinances of the Household of Edward IV. In fact, however, this specialized sense is found in Anglo-French as early as 1318 in the Ordinances of York.¹¹ The FEW puts the introduction of fruitier in the general sense as ‘seit ca.1300’, a full century after the MED example, and the specialized sense of ‘officier de bouche qui prenait soin des fruits’ as ‘1285-14.jh.’, although it would be more natural to have the general sense coming first and being followed after some time by the more restricted sense denoting an official in one of the noble households. As so often, neglected evidence from Britain would have enabled the FEW to present a more accurate picture of the etymological history of the word. Only in Belgium, it seems, has an occupational term been created from the vegetable side of the business – the légumier, although the FEW (5.246b) sub legumen quotes a form légumiste from 1775 which has not been generally accepted into French.

The generic names for the vegetables themselves have in their turn undergone changes in both France and England since the medieval period, so that there is now little evidence of the common link that once existed between the two countries in the area of what we now call ^p4 greengrocery. Yet in the Middle Ages the picture was quite different. In both Continental and Insular French the Latin legumen survived in the etymological forms leun or leune, ¹²with the modern learned légume being attested no earlier than the sixteenth century.¹³ However, under lëun (5.351-2) Tobler-Lommatzsch give two examples of a learned form legum, the one continental, the other insular. The rarity of this form would explain the lateness of the modern marchand de fruits et légumes. Yet this apparent continuity of etymological transmission from Latin to French represented by cases of leun(e) masks a more important non-etymological reality. In the French of both medieval France and medieval England the concept of ‘vegetable’ was usually rendered not by the etymological leun, but by potage, of which nowadays only a reminder remains in the jardin potager. Potage, regrettably, is another example of a clear and unnecessary dictionary error to set alongside grossier, for here again, Tobler-Lommatzsch perversely ignore Godefroy and even, in this case, their own incontrovertible evidence in order to provide a blatantly incorrect translation: ‘Gemüsesuppe: legumen: potaige, Gl. Lille 37b’ (7.1651).¹⁴ In the temporal perspective, potage in the sense of ‘les légumineuses comestibles’ is given by the FEW (9.268a) no earlier than ‘1378-1570’, but new evidence for a much earlier dating in Anglo-French comes from TLL:

legumen: potage (ii 164); legumina: potage (ii 119), legumina: potajes (ii 96), legumina: potages (ii 165), leguminibus divercis: diverce potagis (ii 100), leguminibus: potages (ii 112 & 121).

This meaning was taken over into Middle English, being attested in the late fourteenth century,¹⁵ but the new evidence would strongly suggest that it did not come directly from Continental French, as stated in the MED, but via Anglo-French, with which Middle English was in daily contact.

In order to set these very commonplace semantic developments in their temporal and cultural context, it is worth while remembering that the stream of literature, scientific and religious as well as imaginative, that announces the arrival on the European scene of a new intellectual force in England, begins in the first part of the twelfth century, and that the earliest legal document surviving in Anglo-Norman, the Leis Willelme, comes from the middle of that century, a precursor of Magna Carta and of the stream of compilations defining English law that start to appear in the later thirteenth century and which make the legal register of Anglo-Norman so much more comprehensive and detailed than anything similar found up to the present in mainland France dating from that time. Bearing this in mind, the study of a number of everyday comestibles on the greengrocer’s stall shows that, roughly contemporaneous with these highly important developments in literature and law in England, some of the English vegetables ^p5 and fruits found on the greengrocer’s stall both then and now were starting to take on names or meanings different from those they had across the Channel. The independent character of Anglo-Norman is not confined to the higher reaches of the mind or to the concerns of the higher social classes. The workers who handled the fruit and vegetables at the docks or in the market, the traders who bought and sold them and the kitchen staff in the houses of the upper classes in medieval England who prepared them for their master’s table must have been familiar with the names of the products they were dealing with on a daily basis. Sir Walter Scott’s lofty notion of a two-tier society in which Anglo-Saxon and Frenchman rigorously maintained their respective social positions through the medium of different languages, without any mingling, will not bear scrutiny. As well as Anglo-Norman in the cloisters there was Anglo-Norman at the greengrocer’s.

For example, there is little doubt that the modern English ‘grape’ goes back to Anglo-French rather than to the French of the mainland. Although the OED refers to: ‘The change of meaning which the word underwent in passing from Fr[ench] to Eng[lish], …’, it would be more accurate to say ‘in passing from France to England’ rather than ‘from French to English’, the modern English meaning being abundantly attested in insular French in both literary and glossarial texts from as early as the middle of the twelfth century. In his Bestiaire,¹⁶ written between 1121 and 1135, one of the earliest Anglo-Norman monuments, Philippe de Thaun (or Thaon) maintains the continental French sense of grape and raisin, although applying them somewhat strangely to an allegorical account of how the hedgehog climbs what appears to be a palm-tree (palmier), strips off the ripest bunch of fruit (grape) and then comes down and proceeds to eat its fill of les raisins (vv. 1746-49). The ostensible palm-tree then becomes a vine¹⁷ and the picture is explained as follows:

Par la vigne entendum Ume … E par la grape entent Anme (vv. 1761-4);

Par le raisin entent Bunté danme ensement’ (vv. 1767-8).

Some three decades later, however, the modern English meaning of ‘grape’ is clear in the Anglo-Norman translation of the Books of Kings:

cent liaz de grapes secchies e dous cenz freels de figes

The grapes secchies here must be ‘dried grapes’, not ‘dried bunches’. At about the same time the Eadwine or Cambridge Psalter twice confirms this shift in meaning:

le sanc de grape (Latin: sanguinem uvae) beust tres cler (p. 274.21); La grape (Latin: uva) de cels, grape (Latin: uva) de fiel, e raisimet (Latin: botri) (p. 277.48).

The Englishness of the semantics here is emphasized by the meaning of botrus given in the DMLBS – ‘cluster of grapes, grape or vine-flower’, the ^p6 Anglo-Latin botrus having the same kind of semantic range as the Anglo-French grape, which can now mean the berry of the vine collectively or individually. This is abundantly confirmed in the next century by numerous independent glossators from various parts of Britain, writing at different periods, whose work is gathered together in TLL:

uvam: knoppe et id. grappe (i 142); acinum: marc de grape, marc de le grap (ii 20), acinum: pepin de grape (i 341); acinum: raspe de graps, anglice dref (ii 56-7); uva: grape (ii 30); racemorum: de crapis (ii 99).¹⁹

Yet, even as late as the middle of the fourteenth century, grappe can still be used in Anglo-French to mean ‘bunch’, as in Henry of Lancaster’s Livre de Seyntz Medicines:

une grape de reysins qest issue de la vigne de Jessé.’²⁰

This is not surprising, since semantic change can often take a very long time to become fully realized, with two senses being currently acceptable and distinguished by context over an extended period.

Similar semantic shifts may be seen in ‘prune’ and ‘damson’. The MED and OED, referring to forms rather than semantic content, derive ‘prune’ from Old French, but in French, both medieval and modern, prune has only the meaning of the fresh, as opposed to the dried, plum. The French prune, attested in Anglo-French in the thirteenth century,²¹ was taken into Middle English, despite the fact that the native ‘plum’ had been established for centuries, without the modern distinction in meaning between the two being established for a very long time. ‘Prune’ had certainly taken on its modern sense by the middle of the fourteenth century, both MED and OED giving a Latin quotation from an Ely document of 1345-6 containing ‘prunnes’ as a dried, not a fresh, fruit:

In ij lb. de Prunnes empt.

For some considerable time the noun could either stand alone in Middle English or be clarified by the addition of ‘drie’, but the modern usage was well established by the fifteenth century:

.i. cista prunys;²² Acacia is Iuyse of grene plommez or prunez ²³

Although it is possible to read ‘prunez’ in this second quotation as being synonymous with ‘plommez’, the sense would suggest rather an opposition between ‘fresh’ and ‘dried’.²⁴ In fact, from the thirteenth century onwards medical texts in Anglo-French had been recommending prunes as a laxative, much as they are still advocated in modern medicine. .

lait de amandes … et porcelane quite en ewe et prunes;²⁵ … boille l’en violete, prunes, polipodie en ewe ²⁶ … si devez purgier le malade oveques la decoction de prunes, de violetes ²⁷

In these cases the dried state of the fruit is not made specific, but on occasion the context makes clear that ‘prunes’, not ‘plums’ are meant: ^p7

prunes de outre mer quarante quatre livres ²⁷

Being imported from abroad, these prunes must have been dried. Yet a quotation from 1578 given by the OED shows that ‘prune’ could carry both senses for a long time,²⁹ as was seen above in the case of grap(p)e.

As so often, the limited range of the FEW as a semantic dictionary is evident in its comment under prunum (9.496b): ‘Aus dem fr[anzösischen] übernommen e[nglisch] prune’, giving no indication of the difference in sense between prune and ‘prune’, even though it mentions a form pruneau, glossed as ‘prune séchée’, with the date 1564, thereby suggesting that the French language took an alternative route to acquire the sense of the English ‘prune’. Like Godefroy and Tobler-Lommatzsch, the FEW uses Anglo-French source-material on a non-systematic basis. Whilst, unlike the other dictionaries, it does at least inform the reader whether he is dealing with continental or insular forms, if the large-scale revision now being undertaken is to reflect adequately the new knowledge gained since the inception of the work some three-quarters of a century ago, it will need to take full account of the AND, MED and DMLBS in regard to both forms and meanings. As yet, there is little sign in the new work of any recognition that a very large corpus of medieval French of all registers was built up over centuries in England, influencing and being influenced by Middle English and Anglo-Latin. Happily, the compilers of the Dictionnaire étymologique de l’ancien français ³⁰ currently being actively prepared in Heidelberg have taken cognizance of this situation.

This point is well illustrated by the history of the small plum, the ‘damson’, literally ‘Damascus plum’. The FEW, sub damascene (3.8b) claims that the modern English ‘damson’ comes from a Middle French form damaisine recorded by the Englishman Cotgrave in 1611 (3.8b). However, the DMLBS, AND, MED and Tony Hunt’s researches³¹ are unanimous in showing that the compounds prunes damascenis, prune damascene and ‘prunes damysyns’ are found in all three languages of medieval England centuries before the first attestation for French given in the FEW. To judge from the evidence at present available, damacene does not appear to have been used in Anglo-French as an independent noun with its modern English meaning, but in Middle English the adjective ‘damascene’ (‘of Damascus’) takes on the role of a noun, with a period of hesitation, during which it can either stand alone or be preceded by ‘prune’, as is shown in the following quotations taken from the MED:

The date with the Damesene (c1390); With raysons, corance & prunes damysyns (a1399); … and of that tree is many maner of kynde; but the damacene is the beste (a1398); Take xxti damacyns (c1425), Damysyns (1432)

Given the Middle Eastern origin of the damson and the realities of Mediterranean trade in the thirteenth and fourteenth centuries, it is simply not feasible to suggest that the fruit was unknown in France until ^p8 the early seventeenth century. This is yet another example of the pressing need for a large-scale investigation of all the surviving commercial and business documents in the archives of municipalities and mercantile corporations up and down France, with special attention being paid to the records of the major ports.

The modern English ‘currant’ is another case in point, presenting a striking problem for the dictionaries of both French and English, because the indispensable evidence from Anglo-French sources was simply not used, although some of it was available at the time when these dictionaries were compiled. Whilst it is generally agreed that the term is an abbreviation of the French raisin de Corinthe, given by the FEW under Corinthus (2, ii 1184b-1185a) and defined as: ‘ “esp[èce] de petit raisin” (Cotg[rave] 1611-Trév[oux] 1771) …’, evidence for its existence in early Continental French is strangely elusive, neither Godefroy nor Tobler-Lommatzsch providing any examples. The FEW is forced to admit its inability to account satisfactorily for the term, explaining in its entry that: ‘Auch e[nglish] currant … wird von OxfDict auf eine afr. raisin de Corauntz zurückgeführt, für das ich allerdings keine belege habe’ (2, ii 1185a). Yet the abundance of dialectal forms given in the dictionary in phonetic script for regions all over France, including corinte in Jersey (2, ii 1184b), shows beyond any shadow of doubt that the word must have been well known on the French mainland. Moreover, it has been in print for over a century as the Anglo-French reseyns corences in the Archive of the Grocers’ Company for 1381 (p. 55).³² The quotations given in both the MED and OED would support the latter’s view that the expression was reduced to courant etc. ‘before 1500’ in England, but Anglo-Latin and Anglo-French evidence shows that this happened perhaps as much as two centuries before that. The Customs records printed long ago by Gras and mentioned earlier in this article (Note 22) refer to the import of ‘iiijxx libratis ficorum, racemorum, dattylorum et coriorum’ as early as 1303 (p. 268). As was seen above in the case of ‘.i. cista prunys’, very often in this type of document the Latin used is no more than the conventional medium chosen to house information thought out in French or English,³³ so the form coriorum would strongly suggest the existence of a corresponding form such as corint/corant in Anglo-French. This is now confirmed by the fourteenth-century Anglo-French version of the Old Testament recently edited by Pierre Nobel:

Il cuillent des herbes … . Eles sunt ameres, …, Corintes (var. Corinthes) les apelat l’um a l’ure (v. 15727).³⁴

Currants turn up again in the early fifteenth century in a PRO manuscript:

Item liveré a lui en xx lb canell’ … en xx lb de sugre apuré … en x lb corences. ³⁵

Currants, then, were being imported into England from the beginning of the fourteenth century at the latest and must surely have been likewise ^p9 imported into France. It would not be sensible to deny this on the grounds that the form is so far unattested on the mainland in written French. The history of the currant is a modest illustration of Tony Hunt’s statement at the beginning of TLL: ‘the reliance of the standard dictionaries on edited, literary materials has rendered us oblivious of the alternative, untapped sources such as lexical works, technical treatises and glosses’ (i, p. vii). Thanks to the efforts of a small number of like-minded scholars working on Anglo-French, we are now in the paradoxical situation of having access to a far wider range of medieval source materials stemming from insular French than from mainstream Continental French. There is no point in producing new dictionaries of Old or Middle French unless and until this position is rectified by a large-scale attack on all forms of non-literary documents still awaiting attention in France. Nor should this new initiative be limited to the langue d’oil: any unpublished documents relating to the trade carried on through the Mediterranean ports, whether couched in the Latin of convention, with the Occitan masked by made-up Latin forms, or in some form of the langue d’oc, will be invaluable in enabling a revised FEW to present a vastly more comprehensive picture of the development of the French language. A century and a half devoted to an apparently endless succession of detailed studies on the time-honoured but tiny band of troubadours has fostered the false impression that Old Provençal may safely be equated with Courtly Love, ignoring the fact that science and learning in its broadest sense came into France through the southern ports in the medieval period along with the more tangible products of the Near East. Until attention is switched from troubadours to traders and scholars, there can be no adequate dictionary of Old Provençal and, consequently, no truly adequate FEW.³⁶

Confirmation of this contention may be found without leaving the semantic area of fruits. At the beginning of the thirteenth century an Occitan version of the Chirurgia of Roger Frugard (or Roger of Salerno or of Parma) was produced by Raimon of Avignon. By the middle of the thirteenth century an Anglo-Norman version had appeared, ‘as if coming from nowhere’ as its modern editor says in his Preface.³⁷ In this treatise pierre is used in the sense of ‘stone’ in a fruit: ‘[P]ernez … les pierres des dates …’ (p. 58); a few pages later in the same work (p. 62) we find another example: ‘les nuals des peires de peschier e des cerices’; the term occurs yet again about the same time in another medical work: ‘De cerises pernez les pieres … En .i. morter bien les triblez’,³⁸ and also in a text of the same period which repeats this recipe almost word for word.³⁹ Another example is found somewhat later in a collection of culinary recipes thought to come from about 1320-40. A recent American edition of this text⁴⁰ does not include piere in its glossary, its editors apparently being unaware that there is anything significant about it, but a new edition prepared by Stephanie Wolf for her Heidelberg thesis picks it up and describes it in a subsequent ^p10 article as Erstbeleg,⁴¹ commenting that the FEW has no examples of this sense before the seventeenth century, although it is suspected that the usage must be much older. Wolf believes that the term is unlikely to have been taken over as a Lehnübersetzung from English, because the OED has no examples of ‘stone’ in this sense before the sixteenth century. The MED, however, under ston 10 ‘the pit of a fruit’, dates the sense as ‘a1325’, thus bringing the Anglo-French and Middle English datings much closer together. Since none of the other Romance languages appears to have developed this sense from the Latin petra, it is at least possible that we are dealing yet again with the product of the mixing of Anglo-French and Middle English.

Other cases in this area of comestibles where Anglo-French and Middle English can be shown to have combined to produce modern English usage concern the English interpretation of ‘s’ as a plural marker in medieval French cerise and peis, leading to the modern singular forms ‘cherry’ and ‘pea’, whilst, on the other hand, the medieval French plural laitues/letues and coins are interpreted as singular forms to make the modern singular ‘lettuce’ and ‘quince’.⁴² In all these cases Anglo-French provides forms that will become the modern English standard, obviously of French origin but, equally obviously, not taken directly from the French of the mainland:

Coynz, cornaile et cirie (M.E. Coyns, pekede and chirie)⁴³

cerasa: cyries, cereses, ciriis, chireberie, cerisses; cerasa: ceriez, cirises, cerisses (TLL ii 131 and ii 142)

coctanum: coign; coctana: coyns, coinx and quuyns, (ibid. i 427, ii 143, i 53 and ii 153 respectively) cf. coctana: quinces (ibid. ii 143); coctanus: coynere, quincer, coyner (ibid. ii 142-3)

lactucas: letusez, letusers, letuse, letue (ibid. ii 131)

Sauge, letuse e persylle ⁴⁴

This Anglo-French evidence shows that the modern English ‘lettuce’ developed in England, so that the MED‘s derivation of letuse (c.1300) from ‘[OF letues, pl. of laitue]’ needs to be amended. Similarly, it is difficult to accept the FEW’s assertion under pisum (8.608b) that the English ‘pea(s)’ represents a borrowing from Latin.⁴⁵ Once again, the wealth of glossarial material now available in TLL makes abundantly clear that the Anglo-French pais/peis/pois must have been in regular use in England since at least the thirteenth century,⁴⁶ providing an obvious intermediary between the Latin and the English.

The modern English ‘almond’ is said by the dictionaries to be an adaptation of the Old French almande/amande, without any reference being made to the possible influence of Anglo-French, However, although modern standard English pronunciation does not sound the ‘l’, it can still be heard quite widely in northern speech even today. The explanation that immediately springs to mind is that this is due to the influence of the ^p11 printed word. Unfortunately, this will not hold water, because the ‘l’ was regularly sounded by people born around the middle of the nineteenth century who had no schooling, could neither read nor write, and were, therefore, unaffected by the appearance of the word on the page.⁴⁷ Once again, a reading of Anglo-French non-literary material will show many examples of spellings such as alemand(e), alemundes, almaundes, almandele, but only very seldom amande, ⁴⁸ a situation paralleled in the forms listed in both the MED and OED.

A good example of the need for this type of basic lexicological research to be carried out in mainland France as well as in England is provided by the English ‘radish’. The word is derived ultimately from the Latin radix, and the MED, under radich(e), claims that: ‘Forms in ish-e may have been influenced by OF radise, var. of radice ‘. This information comes from Godefroy’s entry radice (6.545a), where he mistranslates two of his own examples, both of which do indeed mean ‘radish’, but which he unaccountably glosses as racinage. The FEW , under radix (10.27a), dates the earliest attestation of radice as 1507, with the modern radis being introduced by the Englishman Cotgrave as late as 1611, although a dialectal form radis is dated as c.1300. This 1611 dating would make the English ‘radish’ very late. Clearly at a loss for a coherent explanation of the late appearance of the word in French, the Petit Robert claims that the modern French radis is a ‘mot it[alien]’ The English dictionaries make it abundantly clear that the radish was at home in England during the Old English period, so it is odd that such a non-courtly word, rooted – if one may be permitted the term – in the English countryside, should have been remodelled on French, which, in turn, is supposed to have been modelled on Italian. If it was common in medieval England, the radish may confidently be assumed to have been equally at home in France from early times. Godefroy gives one single example of radace (6.544a) from the fourteenth century, but it is is highly implausible that the French would have felt either the need or the inclination to have recourse to Italy some two centuries later to find a more suitable form for a vegetable that they had been eating since time immemorial: after all, as radace or radis it tasted the same.Yet again, the difficulty melts away as soon as non-literary texts are consulted. As early as the thirteenth century radich and radice are found in Anglo-French,⁴⁹ so the English had no need to copy from the French, nor the French to seek a remodelling of their word in Italy.

The need, firstly, to make the etymological information given in the second edition of the OED accurate in late nineteenth century terms and then, eventually, to bring it into line with current knowledge at the end of the twentieth century is well illustrated by its treatment of ‘chive’. The reader is told that: ‘[The form chive prob[ably] represents a North Fr[ench] chive …]’. ‘The form chive is not recorded by Littré, but its existence in O[ld] N[orthern] Fr[ench] may be inferred from the ^p12 derivatives chivon, chivot (Godefroy).’ The MED similarly derives its lemma cive from ‘[CF (i.e. Central French) cive & AF chive]’, yet its first attestation from c1390 has the form ‘chyve’. Both dictionaries are the victims of the inability of their compilers, past and present, to handle the dictionaries of medieval French. The existence of chive in Old Northern French does not need to be inferred at all: the word is recorded twice on the same page in Godefroy (9.103a), but the editors of the OED and MED have failed to take account of the two thousand pages or so of Godefroy’s highly important Complément (from vol.8 after p. 364 to the end of vol.10). Chive is, however, attested for Anglo-Norman shortly after the middle of the twelfth century – earlier than Godefroy’s examples – and again at the beginning of the thirteenth, with cive being recorded also in the thirteenth century.⁵⁰ The editions in which they are found have all been available for between one hundred and fifteen years and eighty. Further new evidence giving the forms civé, civee, civié and sevé in the same semantic area is provided by the seminal TLL.⁵¹. Elementary common sense would suggest that the chive, like so many of these common plants, would have been brought into England – if, indeed, it was not here already – along with the other baggage of the Norman invaders. Similarly, the OED tells its readers that ‘onion’ is ‘a[dapted from] Fr[ench] oignon ‘, with the first attestation being as late as 1356-7. Likewise, the MED derives its oinyon from ‘[OF oignon … & L unio]’, although it gives a quotation in Latin and Anglo-French as early as 1130: ‘Et in Harengs & Vngeons …’ and another from 1225: ‘De nave … carcata weida & allece & oynoins’. These attestations are much earlier than any of those provided in either Godefroy (10.226c) or Tobler-Lommatzsch (6.1015), where, in both cases, the earliest evidence given is actually Insular French, not Continental. As might be expected, TLL gives no less than twenty-four examples of the word in different spellings recorded in thirteenth-century Anglo-French.⁵² It is simply not sensible to assume that the French who made such an indelible mark on the whole administrative, legal and cultural life of Britain after 1066 would have left behind on the beach in France their botanical, horticultural and agricultural vocabulary, bringing with them only the ‘important’ literary baggage.

Finally, the mushroom: predictably enough, given their general failure to take account of Anglo-French in establishing derivations, the OED and FEW (under *mussario, ‘blätterpilz’, 6, ii 267b-268a) claim that the English word is borrowed from Old (i.e. Continental) French, but yet again, the evidence from TLL shows conclusively that the word came into English through Insular French no later than the thirteenth century.⁵³

This brief outline of the semantic journey of some of our common fruits and vegetables through time has lessons for historians of both the English and French languages. On the one hand it shows how untenable is the explanation still being offered for the great enrichment that took place in ^p13 the lexis of later Middle English, namely, that lexical items were ‘borrowed’ wholesale from Continental French.⁵⁴ It has to be accepted that the civilization of medieval England was trilingual, that it was all of a piece, and that its linguistic history must be viewed in that light, not as three disparate bits functioning in isolation. On the other hand, the examination of these common fruits and vegetables may serve to question from yet another angle the venerable nineteenth-century Vising-Pope vision of Anglo-French after about 1230 as being no more than a debased imitation of the mythical ‘pure’ French of mainland France, an increasingly chaotic jargon used only by a rapidly dwindling section of the population of Britain, therefore of scant importance in the overall history of the French language and of the civilization to which French contributed so much in the Middle Ages. Anglo-French is at one and the same time an indispensable element in the linguistic history of Britain and also of France itself. Sadly, in the teeth of all the evidence, the traditional view is still being perpetuated even today: it is depressing to have to read in a serious journal published as late as March 1996 that: ‘à partir du milieu du 13e siècle … l’anglo-normand tend à devenir une langue morte’.⁵⁵ Only if considered as nothing more than fodder for the sterile exercise of juggling consonants and vowels to support a hypothetical nineteenth-century reconstructionist model of an early standard francien divorced from textual reality can this view of Anglo-Norman be sustained. And there always remains the incontrovertible fact that, but for the wholesale absorption of this ‘dead’ Anglo-Norman into Middle English, we would have the literary splendours of neither Chaucer nor Shakespeare.⁵⁶

W. Rothwell

---