Li lox, lililions and their compaig: exemplary error in the fables of BN MS, f. fr. 19152

Li lox, lililions and their compaig: exemplary error in the fables of B N M S , f. fr. 19152 In writing out one of the fables in a thirteenth-century manuscript, now listed as British Library (BL), M S Harley 978, the scribe made a mistake, a dittology, repeating a word already written on the line above. Instead of scraping the letters off the vellum as he might have done, the scribe simply 'removed' the offending word by expunctuating it. Then the rubricator made doubly sure of the suppression by drawing a pretty red line through the word. So instead of disappearing, the mistake is made doubly visible: Par ceste fable del peissun Nus mustre essample del felun Que par agueit e par felun engin Mes cunseille sun bon Ueisin (f. 43v ) This would be simply curious were it not for the fact that the 'error' is echoed later in the same set of fables, in a passage again involvingfelun: De la cumpafnie as-tale feluns (f. 61v ) Furthermore, there is another pair of 'errors', similarly 'eliminated', in the collection of lais preserved in the same codex: Li reis loi souent numer loer (Equitan 38, f. 125V) Cheurefoil lapelent le nument en franceis (Cheu. 116, f. 15iv) lust asfelun occurs in both of thefirsttwo passages, so is numer involved in the second two; and in both cases the terms written 'in error' themselves form a matching pah, not only in visual appearance but in meaning as well: for as uile and felun evoke the low cunning of scoundrels, numer and lapelent both refer to naming. The very consistency in these relationships proves that none of these mistakes is arbitrary or accidental. To treat each as an isolated error, on the other hand, and to 'correct' it, as editors in the past have done without even mentioning the special features of the case, is to miss the supplementary level of meaning which relies on the echoing technique for its effectiveness. Within the medieval tradition from which these sets of lais and fables draw their material and justification, error and its visual impact clearly have a well-defined function. The validity of this statement is to be tested in the following, where we will focus largely on another set of fables from the same century, a relative of the London collection, in Bibliotheque Nationale (BN) M S , f.fr.19152. PARERGON ns 13.2, January 1996—Text, Scribe, Artefact 204 B. A. Masters This set of fables has not been edited.1 Nor is it likely, in present circumstances, to enjoy the special attention of editors or literary critics. For although it is one of the twenty-three versions of the Ysopet ('iEsop') tradition generally believed to have originated with one 'Marie de France', the fact that it contains only sixty-six tales while the largest collection holds one hundred and three2 means that this set is regarded as less than representative of this tradition in its supposed original state. Furthermore, the text of the fables in this set is studded with errors in its literal and narrative construction. It is therefore condemned in advance, on both counts, to playing at best a secondary role in recuperating the vernacular fables from the manuscripts in which they have been preserved. The purpose of this essay is to show that the criteria according to which manuscripted texts are selected for edition, based as they are on the notion that medieval literary traditions would surely have remained stable and fault-free had the scribes paid closer attention to their task, are themselves flawed. T o this end, a small number of fables in B N M S , f. fr. 19152 (ff. 15r -24v ) will be compared with versions of the same fables as inscribed in BL, Harley M S 978 (ff. 40r -67v ), the set judged till n o w as the most nearly authentic among the twenty-three surviving collections. As w e shall see, each of the two texts represents the central tradition in a way specific to itself. Each uses error as an integral part of its o w n expressive agenda, the difference between the two at this level being largely...