ANGLO-NORMAN FABLIAUX AND CHAUCER'S MERCHANT'S TALE.

In 1996 I published the text of an Anglo-Norman version of Le Cuvier, a fabliau which also survives in a well-known exemplar in continental French.(1) The story as it appears in AN is brief, banal, and indifferently written. Its intrinsic interest, even for a fabliau specialist, is moderate. Its historical interest, by contrast, is considerable and twofold.

The AN Le Cuvier appears in an AN remaniement of the AN B-text of the Disciplina clericalis of Petrus Alphonsi,(2) an extremely popular collection of tales from which Per Nykrog appropriated five texts for inclusion in the fabliau corpus.(3) The presence of an AN version of a continental fabliau in this context attests to the close association which exists between the fabliaux and a select number of tales in AN collections, tales which exhibit fabliau features and conform to the general characteristics of the genre.(4) Since Le Cuvier, unlike most of the tales containing fabliau elements, has a close analogue in a continental fabliau, its existence calls into question the validity of the arbitrary decision by the editors of the most recent complete edition of fabliau materials(5) to exclude from the fabliau canon tales occurring within more extensive compilations. The fact that two celebrated tale collections, those attributed to Marie and to the author of the B-text of Disciplina, both circulated in England in AN versions(6) means that the editors' decision dramatically reduces the potential number of AN fabliaux, and is detrimental to the status of AN fabliaux as a distinctive group of tales.

The second important significance of the AN Le Cuvier involves its relationship with the continental version of the story (NRCF, V, 135-44). A number of other AN fabliaux with continental analogues are dismissed by Jean Rychner as `versions orales degradees'.(7) Rychner was not familiar with the AN Le Cuvier, of course, but on the basis of the criteria he applied in judging other AN fabliaux it seems likely that he would have reached the same conclusion about this text. But the identified Latin sources of Le Cuvier clearly indicate that some version of the AN fabliau antedates its northern French counterpart. Rychner's views, which seem to have had a pervasive influence on those of the NRCF editors, are also evidently prejudicial to any positive assessment of the historical role played by AN texts in fabliau development, and warrant re-examination in the light of this new evidence.

Joseph Bedier compiled his inventory of fabliaux from the monumental Recueil general et complet des fabliaux,(8) which completed publication just three years before the appearance of his own critical study.(9) He identified five texts as originating in `Angleterre'. Of these two are known only in their AN versions: Le Chevalier a la corbeille (NRCT, IX, 263-78), and La Gageure (NRCF, X, 1-10). Another, Un chevalier et sa dame et un clerk (NRCF, X, 115-42), resembles in its final episode the continental La Borgoise d'Orliens, but prefaces this segment of the story with otherwise unknown introductory material of sufficient extent and significance to create an essentially independent work. Two other AN texts are insular versions of fabliaux which also survive in one or more continental versions, Le Chevalier qui fist parler les cons (NRCF, III, 45-173) and Les Trois Dames qui troverent un vit (NRCF, VIII, 269-81). In addition to those he abstracted from collections of tales, Per Nykrog added two further texts to Bedier's inventory, an AN version of Les Quatre Sohais saint Martin (NRCF, IV, 189-216) from Oxford, Bodleian Library, MS Digby 86, and an AN version of Cele qui fu foutue et desfoutue.(10) Apart from Le Cuvier, the only proposed addition to the AN fabliau inventory since Nykrog published his study is a version of La Housse partie included in NRCF.(11) The fabliau status of this most recent addition to the inventory is, however, disputable, since the continental version, although accepted by Bedier, was rejected by Nykrog. My immediate concern is with those AN fabliaux which are represented by continental versions, and particularly those dismissed by Rychner as inept imitations. Such an assessment obviously diminishes the importance of insular fabliau literature, but Rychner's arguments, although based on a detailed and aesthetically astute examination of the texts, are open to serious objections as regards the conclusions drawn from the evidence examined.

My intention is to submit to re-examination Rychner's view of the posteriority and inferiority of AN fabliaux by comparison with continental versions; to attempt, in the light of this investigation, to establish some characteristic features of AN fabliaux and their distinction from northern French representatives of the genre; and finally to suggest that Geoffrey Chaucer's Merchant's Tale, unlike Chaucer's other fabliau-tales, was derived from an insular fabliau.

I

Le Chevalier qui fist parler les cons survives in five complete continental versions (one of which, I, has been considerably amplified in an extensive remaniement) and a fragment, A. The four complete and unamplified continental versions are of comparable length and exhibit close textual affiliation. The AN version of this fabliau is shorter than its continental counterparts (292 lines compared to 618 in the critical text of the northern French versions) and its surface texture is consistently at variance with theirs, but the insular version is superior in structure to those from the Continent, and differences in its rhetorical practices, while real, do not justify the conclusion that it distorts and degrades the continental text which an AN remanieur has been supposed to have followed. Both of these assertions can be validated by examining a key passage in the AN text, the episode where the knight restores to three bathing fays the clothes stolen by his squire, and is rewarded by each in turn with a gift (lines 81-108):

 
   La eyne dit sire chiualer 
   Vn doun vous vueil ie doner 
   En tous lyus ou vous vendrez 
   De tous honore serrez 
   E molt chery e molt amez 
   Taunt come ileque demorrez 
   E ceux qe vous encounterount 
   De vous grant ioie demerrount 
 
   La puisne dit erraument 
   Vn doun te doynz ie ensement 
   Ne est dame ne damoisele 
   Ne seit ele ia si bele 
   Si sa amour desirrez 
   E de vous amer la prierez 
   Qe s'amour ne vous grantera 
   E tous vos pleysirs en fra 
 
   Dont dit la tierce damoysele 
   Qe s'estuit en la russhele 
   Bel sire cheualer 
   Qe estes si cortois e si fer 
   Vn doun te vueil ie doner 
   Dont meint se doit enmeruiller 
   Je vous dorroi le poer 
   De fere cul e coun parler 
   A vostre requeste communement 
   Derere e deuant la gent 
   De quanque vous lur demaunderez 
   Certeyn respounz auerez 

As a rhetorical performance the virtues of this passage are manifest. It constitutes an effective tripartite structure, neatly balancing repetition and innovation. The descriptions of the gifts of the first two fays are closely parallel, the first line identifying the speaker, the second beginning, `Vn doun', and the next six lines describing the gift and assetting its universal application. The first fay addresses the knight formally as `vous', and retains this respectful approach while bestowing on him the decorous gift of guaranteed hospitable reception. The second fay, whose similarly presented gift has the naughtier effect of promising a happy outcome in all romantic adventures, begins by addressing the knight familiarly as `tu', but then reverts to the more formal `vous' throughout the test of her statement. The passage which identifies the third (and presumably youngest) fay is expanded to four lines, and allows for an address to the knight praising him as `cortois' and `fer'. If this is not sufficient to alert the audience to expect the unexpected, the announcement of the `doun', which also uses the familiar form of address, is expanded by just one line, but indicates that the gift about to be granted will create some astonishment. The usual six-line description which follows does not disappoint, completing the climactic series with a precipitous descent into scurrility.(12)

The sense of ordered purpose evident in the competent rhetorical treatment of this episode is matched by the subsequent dramatic development. The knight and his squire arrive at the castle of a count, where they are made welcome and lavishly entertained. At supper the knight propositions his dining companion, a maid-in-waiting to the countess, and she readily agrees to spend the night with him. Once in bed, however, and after an initial embrace, the knight asks the girl's con whether its mistress is a virgin, and receives an emphatic denial, whereupon the young lady, totally abashed, withdraws. Her account of events to her mistress the countess prompts the latter's wager of one hundred pounds that the knight cannot effectively exercise his powers over her, and sets the scene for the climactic episode when, after the countess has stifled one response by stuffing herself with cotton, the knight employs the second part of the third fay's gift to reveal the countess's duplicity, confound her, and win the wager.

The continental French versions, by comparison, are structurally muddled. The first fay makes the same gift of hospitable reception as her AN counterpart, but the gift of the second AN fay is lost, and replaced, much to the knight's surprise and consternation, by a jump straight to the gift of empowering the knight with the ability to make tons talk. The third continental fay, whose gift has been pre-empted of any surprisingly bizarre effect, is left adding to the second fay's statement a codicil which makes sense only by anticipation of the story's concluding episode (lines 232-7):

 
   ... se li cons par aventure 
   Avoit aucun enconbrement 
   Qu'il ne respondist maintenant, 
   Li cus si respondroit pot lui. 
   Qui qu'an eust duel ne ennui 
   Si l'apelissiez, sanz aloigne. 

Despite the French author's skill as a facile versifier, his treatment of this episode is tentative and unsatisfying by comparison with the rhetorical development in the AN version of the fabliau, and this impression is confirmed by the subsequent lack of any clear correlation between the fays' gifts and the narrative confirmation of their efficacy. In the continental versions the maid-in-waiting visits the knight's bed because she is dispatched there (in all cases overtly against her will) by the countess, presumably in partial fulfilment of the first gift of hospitality. Her visit is necessary, of course, as the means by which the countess is to discover the knight's extraordinary powers. It prepares for the climactic episode utilizing the gifts of both the second and third continental fays, but by comparison with the AN version the maid's excursion has no direct rationale as illustration of one of the magic powers granted to the knight, and it further reflects the tendency towards disintegration of a previously clear tripartite structure and its replacement by a confused bipartite substitute.(13)

If the AN fableor was adapting some version of the continental Le Chevalier qui fist parler les cons, he would need to be credited with a remarkable achievement in perceiving behind the rhetorically amplified surface texture of his source the possibility of a more systematically ordered narrative.(14) It is not on the basis of his structural modifications, however, that Jean Rychner characterizes the AN author's text as a degraded version of its continental source, but rather from a detailed comparison of the rhetorical practices of the two fableors. His account of the versification of the AN text (M) is dismissive:

 
   M se situe a un niveau extremement bas; sa langue (car de style il n'en est 
   pas question) est rudimentaire, l'expression genee, la versification 
   inexacte. Les vers sont souvent trop longs ou trop courts, les rimes 
   dialectales. Grammaire et rimes sont d'accord pour donner au couplet une 
   relativement forte consistance, signe d'une versification elementaire. 
   Bref, c'est une version tres degradee, qui ne peut guere se rattacher a la 
   version originale que par une transmission memorielle.(15) 

There is some truth in each observation that Rychner makes, but his interpretation of the evidence he cites follows from his assumption about the anteriority of the continental versions, rather than demonstrating its validity. Dialectal rhymes, and a versification willing to accommodate hypo- and hypermetrical lines are conventional features of AN verse, so their presence tells us nothing about the relationship between AN and continental versions of a tale. If the versification is elementary, the natural conclusion to draw is that the text is early, rather than that a late AN remanieur struggled to recast a more sophisticated continental version according to the dictates of an outmoded narrative style.

It is obvious from a comparison of the AN and continental texts that we are dealing with two quite distinct rhetorical traditions. The mere fact that the AN text tells very much the same story as the continental version but at a little under half its length is sufficient to establish this difference. But Rychner is wrong to claim that M lacks style altogether. It is, in its own way, quite stylish, as I have tried to demonstrate when examining the episode of the fays' gifts. By contrast with the continental versions of Le Chevalier qui fist parler les tons, M is written in a presentational rather than a representational style. This is particularly apparent in matters of enjambement and transitions on the same rhyme, features which Rychner correctly identifies as imparting souplesse and aisance to the versification, by comparison with `la progression elementaire par couplets d'octosyllabes'.(16) Transitions on the same rhyme appear only twice in M, once in a passage where the poet continues the same rhyme over more than one couplet, a practice he indulges in a dozen or so times in violation of a rule more strictly observed by his fellows on the Continent. Instances of enjambement are almost equally rare. By contrast, the continental versions show a predilection for both devices.(17) What needs to be recognized, however, is that all of the features noted above as characteristic of the AN Le Chevalier qui fist parler les tons are apparent also in the AN Le Cuvier by comparison with the continental version of this fabliau. It too is only half the length of its northern French counterpart, has dialectal rhymes and irregular lines, and shows the same rigidity in its use of the couplet form. There is a clear correlation between the rhetorical styles of the two AN texts and those of the two continental versions. What is true for Le Cuvier, that it is earlier than, and not a debased reworking of, a more polished and sophisticated continental version, could be true also for M's relationship to the continental versions of Le Chevalier qui fist parler les cons.

Both M and the continental versions mix dialogue and narration in typical fabliau fashion, each consigning to dialogue around forty per cent of the text. Dialogue in M, however, is confined mostly to events at the castle, and serves the purpose of advancing the narrative, the maid reporting at length to the countess what transpired in the knight's bedchamber, and the countess announcing what she has learned of the knight's extraordinary powers to her husband. Dialogue in the continental versions bas an ancillary function, as we would expect from a more mimetically oriented treatment, and is used to develop characterization and to articulate certain thematic concerns left unexplored in M. This is particularly apparent in the relationship between the knight and his squire Huet. M devotes only fourteen lines of dialogue to their exchanges, but this total is raised to eighty-three in the continental versions, where Huet plays the role of Sancho Panza to the knight's Don Quixote.(18) The squire is the pragmatist in the partnership, taking the practical if painful step of selling the knight's palfrey to repair their depleted finances, and reminding the knight of his power to invoke the cul to speak when failure to prompt a response from the con of the countess threatens loss of the wager and another descent into impecuniousness.

Pragmatism, however, may lapse somewhat too readily into vilenie, as it does…