TEXTUAL EDITING AND THE ST

The publication in 1989 of Bernard Cerquiligni’s Éloge de la variante followed hard upon by an issue of Speculum (January 1990) devoted in its entirety to what was termed the “new Philology” has reopened the debate on the methodology of editing medieval texts which had been more or less dormant since the publication of Bédier’s famous treatise in 1928.^[1] Cerquiligni aligns himself with Bédier only in his mistrust of modern editors of medieval texts, for he goes on to attack “le bédiérisme” for delivering no real “image” of medieval writing, but only “des instantanées” (Cerquiligni 101). According to him, Bédier “partage l’idéologie nostalgique et anachronique de l’origine autoriale” (96). There can be no doubt that Bédier believed in the existence of an original author. He rejected the woolly romanticism of the cantilène theory of the origins of the chansons de geste in favor of a fruitful collaboration of clerics and jongleurs along the various pilgrimage routes of France. His reason for rejecting the editorial methods of Lachmann and others was that they tended to produce a composite text unavoidably influenced by the esthetic and linguistic tastes of the editor rather than one demonstrably closer to what had emerged from the hand of the original poet. In Bédier’s view, it was preferable to print one authentic medieval text written by a scribe who knew Old French “presque aussi bien que nous,” and who had the advantage of living in a period less far removed from that of the author. His aim, however, remained that of locating the “best” text, that which logic dictated was closest to an original in which he believed even if he could not reach it. Cerquiligni, on the other hand, appears to deny, not only the possibility of arriving at the original text, but the very notion of an original text or an original author. “L’auteur n’est pas une idée médiévale,” he tells us (25). The medieval scribe appropriates for himself what he reads and copies, and “cette appropriation se traduit par une variance essentielle” (42). “La variance de l’oeuvre médiévale romane est son caractère premier” (62). “[O]n pourrait dire que chaque manuscrit est un remaniement, une version” (62). “[L]e scribe agit avec une liberté qui fait de lui davantage un remanieur qu’un copiste” (78). Instead, Cerquiligni speaks of “une authenticité généralisée” throughout the manuscript corpus of a given text (79). Thus “l’écriture médiévale ne produit pas des variantes, elle est variance (111, emphasis added).

Cerquiligni’s slim volume reads like a series of boutades, each one containing more than a grain of truth; but taken all together they are too dogmatic and too extreme to be of much practical value to anyone planning to edit a medieval text preserved in one or many manuscripts. It is disappointing, therefore, to see him quoted by the editor of the above-mentioned volume of Speculum (Stephen G. Nichols) as a prophet of the supposedly “new philology.” The grains of truth that lie at the root of his boutades are hardly “new” since they have been largely taken into account by most modern editors of medieval French texts, those with which I am best acquainted. What modern editor still adheres to Lachmann’s nineteenth century positivism? Surely that went the way of the “laws” of Grimm and Werner. Bédier demonstrated the flaws inherent in a rigid adherence to the doctrine of the faute commune, but to discard entirely the notion that scribes are capable of mindlessly copying even the inadvertent errors appearing on the page before them is as absurd as that, rightly stressed by Cerquiligni, of believing that scribes are incapable of thinking for themselves or of acting like editors and proofreaders when the mood takes them. Bédier’s own dogmatically univocal approach, “celle que régit un esprit de défiance de soi, de prudence, d’extrême ‘conservatisme’” which requires that an editor not tamper with the text of a manuscript except in cases of “extrême et presque évidente nécessité” (71), has been largely superseded by the acknowledgement that since no two texts are completely alike in their transmission history, general guidelines are of more practical value than ex cathedra style pronouncements intended to cover an entire discipline. From my own limited experience as an editor, I would like to draw attention to a few of the problems I have encountered over the years, and which have tended to combine the excitement of challenge with the frustration of not always finding truly satisfactory solutions.

An example of what Cerquiligni terms “variance” is to be found, I believe, in a thirteenth-century Picard version of the Life of St. Catherine.^[2] In a brief passage describing the probable behavior of birds and dogs towards the statue proposed in her honor by the Emperor Maxentius, Catherine’s scornful remarks go beyond the more decorous language of the Vulgata^[3] to describe in graphic terms the specific acts of the two species. The Vulgata remarks that even if the Emperor succeeds by threats in having her statue venerated by the vulgus, he will have no power over the crows and rooks who will soil her face with the filthy excrement of the carrion they have consumed, and young boys who do not know to venerate her statue will gather there and spit upon it, and dogs too will commemorate her by relieving themselves upon her statue?^[4] The Picard poet translates her remarks as follows:

Et se le gent n’osent passer

Vostre ymage sans saluer,

Que diras tu dont des oisiaus,

Des cornelles et des corbiax

Qui sur sa face kieront,

Ne ja pour ti ne le lairont?

Et tout li kien de la contree

Pisseront sus ganbe levee. (737–44)^[5]

This is perhaps strong language for a virgin saint, a scholar, and a princess (although the Blessed Virgin herself in Le Miracle de Theophile is described as threatening the Devil with a kick in the “pance”! [Rutebeuf 1.585]). One manuscript (U.C. Berkeley 106) offers Qui sa face li blanchiront at l.741, and Li corent sus, ganbe levee at l.744. This more decorous version is recorded by the remaining four complete manuscripts with, however, one further modification. Troubled no doubt at the concept of dogs running towards a statue with leg raised, and aware that in nature this is somewhat anomalous, these versions alter ganbe levee to goule baee thus eliminating the anomaly by substituting a reading which, however, does little to render Catherine’s sarcasm. What, indeed, is so bad about dogs running toward your statue with their jaws agape? It hardly compares to the behavior of the birds even in its bowdlerized version. One could argue for either of the two euphemistic versions as that of the “original” though anonymous poet, but common sense, supported in this case by the Latin source, will surely tell us that a scribe making a copy for his community is more likely to “clean up” the language of the saint than to render it more explicit by the use of four-letter words. On the other hand, it is not impossible that an ancestor of the group of four plus Berkeley 106 already contained the goule baee reading, and that the scribe of this latter manuscript re-substituted the original reading.

Further evidence of scribal “editing” can be found in the three extant manuscripts of Clemence of Barking’s Life of St Catherine (MacBain Life).^[6] One manuscript (P) has been transposed into the Picard dialect for the benefit of continental readers, and appears to have “corrected” many of the Anglo-Norman “aberrations” in the two-case system, and replaced some Anglo-Norman forms with continental equivalents. The oldest of the three manuscripts (A) dating from the end of the twelfth century is the work of a scribe who tended to lose patience with his task and make considerable cuts in the latter part of each of the texts he was transcribing.^[7] A few of the cuts are carelessly done, ascribing following statements to the wrong speaker, but most permit of a smooth transition to the next passage, thus concealing the omitted lines. Scribal editing of this kind, however reprehensible we may find it, is certainly authentically medieval, but it is difficult to imagine that Clemence herself would have approved. A poet like Clemence who is sufficiently proud of her literary accomplishment to put her name to her work is not some shadowy figure rising out of the mists of time and equally present or absent in each and all of the surviving copies of her work. That the scribe of A found her at times tedious and long-winded is his privilege—although he seems to have found her so only in the latter part of the poem, specifically the sequence that interests us most today since it is here that she abandons her source and expands lyrically (and rhetorically) on love, betrayal, and loss. This scribe seems to have had a similar problem with both the Alexis and the Brendan,^[8] which suggests to this critic that it was his task of copying rather than the content of the poems that he found wearisome. The tedium of lengthy transcription is alluded to by the scribe of an early fifteenth-century manuscript who, on completion of his arduous task one Saturday “par nuit bien tart” while others were celebrating the feast at Ath (Hainault), asks for the hearers’ prayers. One wonders if he had perhaps been assigned this duty as a punishment, and was thus prevented from joining in the festivities with his fellow monks. This rather prosaic approach to the task of copying seems more in tune with the humdrum reality of the scribe’s life than Cerquiligni’s somewhat romantic concept of the “excès joyeux” from which “variance” arises,^[9] and which might well be more characteristic of the manuscript tradition of romance than of hagiography. However that may be, the impatient scribe of Clemence’s poem who made so many cuts in his text found himself impelled at one point to add two lines. In a Festschrift article published in 1968 (Geschiere, “Un passage obscur”), I was taken to task for suggesting, in a mere Critical Note, that these two lines appeared to have been interpolated by the scribe in order to cover his tracks, since he had inadvertently completed a line de son propre cru without checking his original. At that time, at the height of my Bédiériste rigor, I had been determined to maintain the manuscript reading, even when I judged it to be of scribal origin, so long as it “might” make sense in its context. The manuscript (A) reads as follows:

Mais un poi t’estuet prendre aie

A parfaire ta folie,

Car pas n’avum fait uncore,

Mais faire vulum des ore

A noz deus lur service

As quels jo faz cest sacrifise. (317–22)^[10]

I have italicized the two “offending” lines. It was (and remains) my contention that the scribe had before him a text which read substantially as follows, the italicized words marking emendations to the base manuscript A suggested by the readings of the two remaining manuscripts and supported by the Latin source:

Mais un poi t’estuet attendre ore,

Car pas n’avum parfait uncore

A noz deus leur deü service

As quels jo faz cest sacrifise.

Geschiere remarks that I had applied “la méthode lachmannienne la plus pure,” and produced “fatalement” the usual “division dichotomique,” admitting somewhat grudgingly that I had “usé de cette liberté avec beaucoup d’acribie (sic! for acuité?) et avec la prudence nécessaire” (343). He then goes on to a defense of the interpolated lines, finding obscure juridical uses for such expressions as prendre aïe, parfaire ta folie, and service, but failing to address the question of poor scansion in an otherwise almost impecccable poem and, worse still, failing to take account of the words of Clemence’s Latin source which indicate beyond all possible doubt that it is indeed W and P rather than A that conserve the original meaning:

While, therefore, we continue with the rites we have begun, you must meanwhile wait for us because you are to go with us to the palace and be honored with gifts if you submit to our commands.^[11]

Rationalizing in a totally anachronistic manner from Clemence’s next statement, Quant ço out dit d’iloc s’en vait (323), Geschiere argues that the entire sacrifice in the temple has now concluded when Clemence had made it clear at the outset that Catherine had interrupted the proceedings when they were still in progress. The Emperor has not gone back to his palace, but has simply moved out of Catherine’s hearing to give orders to his privez clers to summon the fifty orators. Having done so, he returns to complete the sacrifice and then has Catherine escorted to the imperial palace, as the Latin source makes abundantly clear:

The messenger complied with the King’s commands and on completion of the sacrifice, the Emperor gives orders that the maiden be seized and led to the palace.^[12]

Instead of searching for a special sense of the word “service” (l.321) to signify the Emperor’s summoning of the orators in defense of his pagan gods, Geschiere might better have contented himself with the obvious: that the Emperor is referring to the completion of the rites which Catherine had interrupted by her intrusion. Despite his defense of the possible authenticity of the interpolated lines, Geschiere grants that the major fautes communes which I allege “semblent avoir assez de force probante pour emporter la conviction quant à la parenté entre A et P.” Nevertheless, he goes on to adduce possible fautes communes between P and W, although it is surely self-evident that if two of the three manuscripts have a common ancestor not shared by the third, then the third cannot logically share a common ancestor with either member of that pair unless it be with both simultaneously. Any apparently “faulty” reading shared by one member of the pair and the third manuscript must necessarily result from coincidence. In such circumstances one must examine very critically the evidence to see whether these supposed “errors” are of the kind that scribes can so easily make independently of one another. Scribes, like modern secretaries and academics, did not always pay complete attention to the matter in hand, and often transcribed mechanically without due attention to details of content. That medieval scribes read much more slowly than we do today is not seriously open to question, and the effort involved in copying a text must of necessity have slowed down their reading even more. It should not surprise us, then, that the rhythm of a verse text was frequently impaired by adding or subtracting syllables. Contrary to what Cerquiligni suggests, I suspect that the majority of scribes did not attempt to rewrite the text before them. The Picard scribe of B.N. fr. 23112 recording Clemence of Barking’s poem attempts at times not only to change the orthography to suit his Picard readers, but occasionally “corrects” insular declensions and verb forms in line with continental “norms.” Virtually all of his lines are regular octosyllables achieved sometimes by procrustean means. But, unlike his Anglo-Norman counterpart of manuscript A, he neither adds nor subtracts lines. His work is that of transposing into a medium more familiar to his audience, not adapting it to different tastes in the manner of one or more of the bowdlerizing scribes of De Sainte Katerine.

In manuscripts of this latter poem, another tendency can be observed which at first sight appears to be an example of “variance,” but which may well have resulted from inadvertence. This is the occasional repetition of two or more lines in a different context. In no case do the lines jar in any way in the new context in which they find themselves, but a glance at the Latin source makes it clear that, whereas in their proper context they are merely a rendering of the Latin text in its customary sequence, their repetition in a second context reveals the involvement of the scribe. The question arises then as to whether this is a conscious intervention of the scribe, and thus an example of “variance,” or a purely unconscious saut du même au même. An analysis of the specific examples occurring in several manuscripts of De Sainte Katerine (MacBain “Remembrance”) reveals a pattern of apparently unconscious interpolations triggered by a word or phrase in the immediately preceding line, and which has strong associations with the interpolated passage in its original context. What is more striking is the fact that in some cases the “original” context occurs later in the text rather than earlier as one might expect. A case in point occurs in an early fifteenth-century manuscript of this poem. Catherine suffers two major scourgings in the course of the poem, the first prior to being thrown into the dungeon, and the second after she has languished in the dungeon for twelve days without food or water. On both occasions she cries out defiantly to the Emperor warning him that his time of punishment will soon come. The Vulgata presents both warnings, but differentiates them as do the other manuscripts of this text. On the first occasion, Catherine warns him that the time will soon come when the Devil will exercise power over him, and he will suffer eternal torment for the pain he now inflicts on the servants of Christ. On the occasion of the second scourging, however, she adds a further warning that Christ will raise up a Christian adversary who will cut off his head, and consign him to Hell. The manuscript in question (Bibliothèque Royale de Belgique 10295–304) interpolates into the first passage the four lines from the later passage dealing with the beheading of the Emperor. What is astonishing is that this process of anticipation is found three more times in this same manuscript, and occurs several times in two more manuscripts of this same text (Arsenal 3516 and the Carlisle Cathedral manuscript). In all cases but one (lines 348/658) the phenomenon is the same while the interpolations themselves differ from manuscript to manuscript.

Such anticipated repetitions can only be the result of complete or partial memorization of the entire text by dint of repeated copying, or repeated recitation, or even repeated hearing of the text in question. These “versions,” therefore, cannot possibly have the same authorial validity as those which follow the sequence of the Latin source. Indeed it is the discovery of the Latin source which the hagiographic author claims to be rendering that permits the modern editor to determine, as far as humanly possible, the preferred reading in a given context. In no way do I mean to imply that objective certainty can be achieved. A poet can abandon his source from time to time, add details, explanations, and commentaries as does Clemence in her nonetheless faithful rendering of the Vulgata, but a hagiographic writer, almost invariably dependent on existing written sources, is less likely than a writer of romance to stray far from the written record. Together with other criteria, consulting the very source that the writer acknowledges using will help the editor to avoid the kind of error to which Geschiere left himself open.

A more complex problem faces the editor of the so-called Franco-Veronese version of the Life of St. Catherine (ed. H. Breuer).^[13] This version which details not only Catherine’s passion, but also her youthful conversion to Christianity, her visit to Heaven, and her mystical marriage to Christ exists in a single manuscript dated 1251. The language of the author appears to be northern or north-eastern in origin, but shows a heavy overlay of northern Italian forms and spellings, many of which are probably attributable to the Veronese scribe. The author bases his poem largely on a text he claims to have seen in St. Sylvestre in Rome but which he found too detailed and too long for his purposes, and partly on accounts of eyewitnesses who had visited the tomb of St. Catherine on Mt. Sinai. Two later northern Italian poems in manuscripts of the fourteenth century (see editions by Mussafia, and Renier), are so close to one another and to the text of this Franco-Veronese version as to be at times almost word for word translations. But these striking parallels end with the “mystic marriage” section. The two Italian versions, one evidently a somewhat abbreviated adaptation of the other, then proceed to follow the traditional Vulgata sequence of Catherine’s passion and remain close to one another, while the French poem diverges into a completely fanciful version of the passion, telescoping certain traditional sequences, and omitting others such as the debate with the fifty orators and the episode of the wheel which is so important in the iconography of St. Catherine. Here one can truly speak of “variance,” but these three texts do not purport to be copies of one original. Each is a separate entity with its own rationale and its own audience.

The Anglo-Norman scribe of Additional 40143 in the British Library apears to combine an imperfect knowledge of the language he was copying with a possible tendency to dyslexia. While his hand is fairly regular and the individual letters remarkably clear if occasionally accompanied by ornate flourishes, he appears at times to invert the order of letters even at the end of the line thereby producing some alarming non-rhymes. Thus respoundre : confundure (fol. 2a),^[14] balasice (balaunce!) : creaunce (fol. 6b), sey : coueri (conrei! fol. 5b), and alorr (aloer!) : pouer (fol. 6^va). Within the line we find matire for “maitre” (fol. 2^va), amaunt for “avant” (fol. 1^va), vertoy for “verroy” (fol. 4a), fernite for “fermeté” (fol. 6b), mon for “son” (fol. 1^va), il for “li” or “lui” (fol. 4^va), and arrifices for “artifices” (fol. 4^vb). When Catherine makes a vow to her mother which she promises never to break, it is expressed by the scribe of Additional 40143 in terms of a doun (“gift”) which makes no sense in the context, especially in view of the fact that it is later referred to as a vou (“vow” fol. 5b), and is certainly no “gift” from the mother’s point of view since she disapproves of it entirely and sees it as the height of foolishness. “A vous, en fey, un doun dorray / Qe ja en fyn ne fauceray, . . .” (fol. 2^va, emphasis added).

To these curious phenomena we must add the fact that no less than 53% of the lines deviate from the traditional octosyllable by a considerable margin (from -2 to +3 syllables) suggesting that the scribe has at best a rudimentary understanding of scansion, or is transcribing a language with which he is unfamiliar.

These are only a few of the many “problems” of this manuscript, regrettably the only extant copy of a version of St. Catherine’s “mystic marriage” with Christ which appears to be of supreme importance in the development of the legend. The vagaries of the Anglo-Norman dialect, especially in its later stages when, to all practical purposes, it had ceased to be a spoken language and had become merely a convenient literary, commercial, and legal language for speakers of Anglo-Saxon dialects (see Kibbee), are too well known to need repeating here, but the evidence presented above goes far beyond the limits of what might be considered normal in Anglo-Norman linguistic “variance.” Is this highly imperfect witness, then, to be accorded the same respect as a version which came from the hand of the author, were we ever lucky enough to find it, or even that of some non-dyslexic scribe still capable of understanding what he was copying, and of transcribing it with at least moderate accuracy? Are the dyslexic non-rhymes detailed above, and the six-, seven-, nine-, ten-, and even eleven-syllable lines which occur with no apparent rhyme or reason to be seen simply as an example of this same “variance” or “excès joyeux” that allegedly characterizes medieval as opposed to post-medieval works? Such an approach confers both literary intention and discrimination on a copyist who, far from consciously editing the text before him, was in all probability incompetent for the task to which he had been assigned.

A different kind of problem is presented by a fifteenth-century narrative poem dealing with the “mystic marriage” and passion of St. Catherine in monorhymed alexandrine quatrains. The text is preserved in only one manuscript (Paris, Bib. Nat. n.a.fr. 14313) and in no less than four separate printings dating from 1485 (Lyon) to 1491 (Paris).^[15] The problems involved in editing a text preserved largely in incunabula are daunting indeed, since in such a case the individual scribe may well be replaced by a group of typesetters and supervisors each with his own perspective on the form and content of the book in question. Cerquiligni remarks: “Le texte imprimé est rien moins que sûr” (20), and tells how Erasmus spent eight months in Venice in order to supervise the printing of his Adages. The manuscript version of the St. Catherine text which is almost certainly earlier in date than any of the printed books differs from all of these in certain respects, and does not appear to be the ancestor of any of them. At the same time, the three printed versions that I have been able to study closely, share with the manuscript version certain “unsatisfactory” readings which would lead one to postulate a “faulty” archetype. I enclose these adjectives in quotes merely in order to spare the sensibilities of those to whom such value judgments are anathema, but I am referring to passages which jar in logic and scansion with the remainder of the narrative in such a way as to constitute evidence of attempts to reconstruct a part of the text which the individual scribes may have found difficult to decipher.

A case in point is presented by a hemistich in the second quatrain where the manuscript version and all of the printed versions appear to be a syllable short. The poet is explaining God’s purpose in the Incarnation. It was, he says, to bring back a tender ear which had been lost (“condemned” in the printed versions) through wrongdoing: “Ce fut pour recouvrer une oreille tendre / qui estoit esdiree (condamnee) pour cause de mesprendre.” In the next quatrain the oreille appears again, but this time in a hemistich of the full six syllables:

L’oreille que je dy est humayne lignie

qui par le pechié d’Ayve fut si fort exillee

que home ne trespassoit tant fust de bonne vie

que le deable d’enfer n’eüst sur luy maistrie. (9-12)

It would seem, therefore, that the word “oreille” is “correct,” however bizarre the image may appear to the modern reader. What is more troubling is the hypometric hemistich. Are we to understand that “une” could have a disyllabic pronunciation before a following vowel? Or was the medieval ear not shocked by a limping hemistich? What confounds the problem even more is that the three printed versions to which I have had access offer a slightly different version of these lines, and one which is every bit as convincing as that printed above from the Paris manuscript. The three printed versions read as follows:

L’oreille que je dis est humaine lignee

qui par le peché d’Eve fut si fort exillee

et d’Adam qui pecherent mengeans le fruit de vie

dont le dyable d’enfer eut sur eulx la maistrie. (9-12)

There is little to choose between the two versions. The first stresses the theological concept of the condemnation of all men, good or bad, to Hell as a result of the Fall. The second takes pains to include Adam on equal terms with Eve in precipitating the Fall, and speaks of the Devil’s power over them, rather than over all of humanity. The second version may perhaps be a revision, carried out by a nun, or by a less misogynistic cleric, seeking to emphasize that it was the sin of Adam, not of Eve, that brought about the Fall.

Other passages in this text show evidence of the telescoping of some quatrains, an understandable occurrence in view of their monorhyme nature, the similarity of many of the rhymes and above all the fact that the text is presented, both in the Paris manuscript and in the printed versions, in hemistiches, not full alexandrine lines. I am hopeful that the 1485 Lyon printing when it is located will help to shed some light on the interrelationship of the printed texts and in turn on their relationship to the only recently discovered Paris manuscript.

Cerquiligni’s answer to the problems presented by multiple manuscripts of a given text is the computer.

Car l’ordinateur, par son écran dialogique et multi-dimensionnel, simule la mobilité incessante et joyeuse de l’écriture médiévale. (114)

Would that the necessary data bases were available right now, but I fear they lie very far in the future. This would indeed be an excellent tool for the preparation of an edition, but it is hardly a practical answer for those not interested in textual editing per se, but rather in finding reliable editions on which to base their literary or linguistic research, or even their own reading pleasure. But the issue of practicality is not my principal quarrel with Cerquiligni’s formulations. Rather, it is his refusal to acknowledge the validity of the search for the author, or rather for the author’s artistic intent. He appears to make no distinction between poet and scribe. His concept of “authenticité généralisée” smacks rather too much of the Romantic notion of “le peuple” as poet. The mere fact that many medieval works are anonymous, or their named authors untraceable in rolls and charters, does not mean that they never existed as individual creative spirits. Malherbe is said to have characterized poets as mere “arrangeurs de syllabes.” He undoubtedly had a point, just as composers might be termed “arrangeurs de sons.” But some arrange their syllables and their sounds considerably better than others. Those who, like the authors of the Roland and the B Minor Mass continue to impress us centuries after their death are sometimes termed geniuses in a manner that in their own day would have been thought absurd, and possibly even outrageous. The fact that we may not be able to identify the owner of a particular gift for arranging syllables in such a harmonious, dramatic, and moving way as we find in the Roland does not mean that there never was such a person. The esthetic weaknesses of the later versions of the Roland (or rather the characteristics that appear to many of us as esthetic weaknesses) are the clearest indication that these examples of variance need to be distinguished from the work of the Oxford poet, whoever he (or she?) may have been.

The task of editing medieval vernacular texts is much less a science than an art as Bedier himself made clear by the title of his famous essay. Editors cannot pretend to a kind of pseudo-scientific objectivity in establishing their text, but they must neverthless establish one, and that means making choices and substantiating them. Cerquiligni himself proclaims at one point: “L’édition est un choix; il faut trancher, et savoir les raisons du geste qui récuse” (43). Despite the often intractable problems posed by the relatively few manuscripts that have survived to the present day, editors must continue to use their computers and also their wits and their judgment to provide dependable copies of the texts we so badly need in order to better understand and appreciate the wisdom and the beauty of these so-called Middle Ages. A slavish adherence to a given manuscript, taking no account of what the original poet was in fact attempting to say, may result in a completely erroneous interpretation of the text. Editors owe it to their readers to provide not the most conservative reading of the text, but rather the one which most closely corresponds, as far as can be judged by careful analysis of the facts at our disposal, to the version that issued from the quill of the original author. But we should also heed Bédier’s admonition to relegate to the Critical Notes any emendations not fully supported by the evidence at hand, taking care to give readers all of the information needed to reconstruct the text as they see fit. Only in this way can we exercise our own judgment without obstructing that of other scholars whose intuitions may prove closer to the mark than our own.

Works Cited

Bédier, Joseph. La Tradition manuscrite du “Lai de l’ombre”; réflexions sur l’art d’éditer les anciens textes. Paris: Champion, 1929.

Breuer, Hermann. Eine gereimte altfranzösisch-veronesische Fassung der Legende der h. Katharina von Alexandrien. Beihefte zur Zeitschrift für Romanische Philologie 53. Halle: Max Niemeyer, 1919.

Cerquiligni, Bernard. Eloge de la variante: histoire critique de la philologie. Paris: Éditions du Seuil, 1989.

Geschiere, Lein. “Un passage obscur de la Vie de Sainte Catherine.” Festschrift Walther von Wartburg zum 80. Geburtstag. Ed. Kurt Baldinger. Tübingen: Max Niemeyer, 1968. 1: 343–58.

Kibbee, Douglas A. For to Speke Frenche Trewely: The French Language in England 1000–1600. Amsterdam: John Benjamins, 1991.

Knust, Hermann. Geschichte der Legenden der h. Katharina von Alexandrien und der h. Maria Aegyptiaca nebst unedirten Texten. Halle: Max Niemeyer, 1889.

MacBain, William. De Sainte Katerine: an Anonymous Picard Version of the Life of St. Catherine of Alexandria. Fairfax, VA: George Mason UP, 1987.

___. The Life of St. Catherine by Clemence of Barking. Anglo-Norman Texts 18. Oxford: Basil Blackwell, 1964.

___. “Remembrance of Things Still to Come: Scribal Memory and Manuscript Tradition.” Manuscripta 31.2 (1987): 77–87.

Mussafia, Adolfo. Zur Katharinenlegende. Vienna: K. Gerold’s Sohn, 1874.

Nichols, Stephen G. “Introduction: Philology in a Manuscript Culture.” Speculum 65.1 (January 1990): 1–10.

Renier, Rodolfo. “Una redazione tosco-veneto-lombarda della leggenda versificata di Santa Caterina d’Alessandria.” Studj di Filologia Romanza 7 (1894): 1–83.

Rutebeuf. Le Miracle de Théophile. Ed. Grace Frank. 2nd edition. Paris: Champion, 1967.

Waters, E. G. R. The Anglo-Norman Voyage of St. Brendan by Benedeit. Oxford: Clarendon, 1928.

^[1]Bédier’s essay originally appeared in Romania 54 (1928) before being published separately. Page references are to the 1929 edition.

^[2]De Sainte Katerine: An Anonymous Picard Version of the Life of St. Catherine of Alexandria, ed. William MacBain.

^[3]Edited by Hermann Knust (Geschichte 231–314), and reprinted in my De Sainte Katerine (177–216).

^[4]Veruntamen si sit hec interim inanis recordatio quantumlibet in honorem dum regio metu pressus hanc mihi vulgus venerationem exhibeat, sed aves celi haudquaquam istud verebuntur dum milvius et corvus, undecunque avolantes, sedem in me sibi usurpabunt et immunda digesti cadaveris proluvie faciem meam innotabunt. Quid interim pueri facturi sunt qui divinum cause misterium venerari nescientes huc simul egesturi convenient? [Other manuscripts add: aut quid canes commincturos in me ego commemorem?] (Knust 271; MacBain, De Sainte Katerine 198. Translation of quotations from the Vulgata are my own.)

^[5]Line references are to my edition of De Sainte Katerine.

^[6]A Paris, BN n. acq. fr. 4503, P Paris, BN fr. 23112, W London, Brit. Lib. Welbeck IC1.

^[7]E. G. R. Waters (Brendan xii-xiii) noted also that it was mainly in the latter part of the text that the scribe tended to make his cuts. This would seem to preclude the possibility that these were attempts to remove longueurs. See MacBain Life (xv-xvi).

^[9]“Cette appropriation se traduit par une variance essentielle, dans laquelle la philologie, pensée moderne du texte, n’a vu que maladie infantile, désinvolture coupable ou déficience première de la culture scribale, et qui est seulement un excès joyeux” (Cerquiligni 42).

^[10]Quotations are from my 1964 edition of Clemence’s poem (Life). A new edition of this text is currently in preparation.

^[11]Dum ergo nos incepta sacra peragimus te interim nos opperiri oportet, quia nobiscum itura es ad palatium et regis honoranda muneribus si nostris adquiescis jussionibus. (Knust 244; MacBain, De sainte Katerine 184.)

^[12]Paruit nuntius regis mandatis. Et expletis imperator sacrilegis officiis virginem jubet comprehendi et ad palatium duci (Knust 245; MacBain, De Sainte Katerine 185).

^[13]A new edition by Peter J. Burgess, doctoral candidate Univ. of Maryland is in preparation.

^[14]The scribe’s u is frequently written n, seldom the reverse. Both forms of r (following o, and in all other positions) are clearly differentiated from all other letters.

^[15]Two copies are in the Réserve in the Bibliothèque Nationale in Paris, and the third is housed in the National Library of Wales at Aberystwyth. The present whereabouts of the 1485 Lyon printing are unknown. It was listed in the 1910 catalogue of the library of C. Fairfax Murray, but did not figure in the sale of this collection which occurred at Sotheby’s in 1919 following Fairfax Murray’s death. I am currently attempting to locate this copy for my projected critical edition of this text.