Printer Friendly

Edited text and medieval artifact: the Auchinleck Bookshop and "Charlemagne and Roland" theories, fifty years later.

It's an interesting time to be a medievalist. For several years now, and in academic venues ranging from print to conference to classroom, textual critics of late medieval literature have demonstrated a multiform alienation from the work of their scholarly predecessors. This condition, which as far back as the later 1970s was expressed in passing as a frustration at the lack of consensus concerning the representation of manuscript textuality, has developed in recent years into a wide-ranging paradigm shift regarding the nature of medieval textuality and the numerous obstacles to its engagement by readers in the postmodern age. (1) While there are several unique causes for this shift in editorial assumptions and goals, its energetic propulsion into new modes of discourse resembles, in many interesting ways, the transition from antiquarian publishing to modern textual criticism in the mid-nineteenth century. This noted shift from amateur to professional editing (and from private to public circulation of edited medieval work) may be now a century and a half in the past, but as we embark today on what can accurately be called the "third age" of editing such texts, we have a tremendous opportunity to take stock of our relationship to our academic forebears.

As the study of Middle English continues to develop in our new millennium, its leading voices are becoming ever more interested in the "messy details," the exceptions and contradictions, of clerical and secular manuscript production. But the work of today's investigators is equally suffused by a fascination with the activities of modern editors, the invaluable teachers and researchers who transform medieval manuscript texts into classroom and scholarly editions. We are now able to see that the methods of textual recovery and editorial presentation in most twentieth-century publications had clear and uninterrupted precedents in the work of nineteenth-century pioneers like F. J. Furnivall and W. W. Skeat; our debt to the armies of contributors to the Early English Text Society and other collective philology projects is clear in their continued availability and use as authoritative texts. Our contemporary poststructural philology embraces, in short, at least as much desire to scrutinize our relationship to our modern predecessors as it does to scrutinize our ostensible medieval subjects--and while several gifted writers have shared truly innovative and thoughtful work with us in response to this disciplinary development, a mature and comprehensive interrogation of our textual role models remains some distance away. (2)

Whoever eventually undertakes the desirable project of narrating the full story of Middle English editing will find the heaviest task upon reaching the singularly fascinating topic of the organic text and its romantically conceived scribal environment. On the one hand, it will be an opportunity for a lot of undiluted, anecdotal fun--the approach-avoidance behaviors displayed by many editors toward the minutiae of scribal variation often seem like a visit to the Cirque du Soleil. But bringing an historical perspective to this phenomenon will also be fraught with danger, and for one clear reason: it will be a retrospective evaluation of a recognizably conflicted paradigm, and whoever accepts the role of evaluator will run the risk of becoming (or merely of becoming perceived as) a desecrator of the lifetime monuments of scholars who, directly or at a mere one or two removes, educated today's generation of specialists. One might be comforted by the fact that at some point in their scholarly careers most Middle English editors have equivocated about the distance between manuscript witnesses and their printed or e-texted counterparts--but putting this equivocation to a constructive and contextualized use is still such a new impulse that its motive may be easily interpreted as dilettantish or even just plain mean.

For example, what we are accustomed to finding in editions before, say, 1975 (when the publication of the Kane-Donaldson B-text of Piers Plowman initiated the now-legendary cascade reaction of public editorial angst) is that in place of a description of the editor's own method and assumptions there is consistent discussion of extrinsic variation (that is, a display of variants only in their relationship to the perceived organic textual tradition of a specific work) without a complementary discussion of intrinsic variation (that is, the nature of the variants in their intimate scribal environment). Only a handful of perspicacious editors of Middle English before 1970 (such as Eugen Kolbing, who did most of his work in the 1870s and 1880s) attempted to engage both forms of variation in their work, and they usually found themselves overwhelmed by the job; more commonly we find textual critics who resolve whole classes of problems by skirting the voice of methodology altogether and who opt instead for a systematically arbitrary synthesis of variant texts based on silent assertions of perceived authorial intention. (3) The nearest question is the toughest one of all: To what degree is the life's work of the hundreds of women and men who created medieval editions under this positivistic paradigm still usable to a poststructural readership?

Just how our hypothetical future historian will handle this difficult problem of perception and presentation will have to wait until ... well, until her history of editing is out in the open and has itself become part of the record of textual studies. The need for diplomacy, so to speak, is patent in such a situation. For the present, we can maintain a bridge to the receding generation with a few basic, open observations. It has been clear for a long time now, for example, that editors are enormously influential through the visible textures of their publications--from the font, to the punctuation, to the handling of scribal variation and decoration, modern editors are simply incapable of "reproducing" medieval textuality because the social and technological dynamics are fundamentally different--instead, the textuality they are creating is a texture, an approximation of the medieval text, and it brings about interpretations of the text that are made necessary or likely by its modern format. We can also now say, conversely, that editors are influenced by the actual or perceived needs of their modern audiences, by the limitations and freedoms of modern publication formats, and by modern standards of judgment (aesthetic, political, sexual) which are naturally different from those through which the artifact originated--each of these three social particles of postmodern philology vastly affect the choices editors make as they move back and forth between textual localities. (4) I believe we can also agree that for more than a century (from, say, 1860 to 1970), Middle English editors perpetuated their practices and methods in a rather un-self-reflective manner. Far from setting into place truly new approaches to textual reconstruction, even the more celebrated practitioners such as George Kane only made a long-standing practice of proscribing conservative editorial activity momentarily more emphatic: his work can, in this historical sense, be seen as simply a corollary form of best-text editing. (5) But as the romantic fascination with poetic ur-texts has declined into invisibility, the once unassailable citadel of the critical edition has revealed to us its shifting foundation, its subjective context, and therefore its finiteness as a scholarly tool. I think that above all we can agree that we must move into the next generation of textual criticism by consciously bringing our past--our own teachers--along with us.

We all know what it means to surpass the methods and awarenesses of our teachers. It involves at once a deeply felt respect and admiration for the value of past scholarship, and a thrilling desire to move beyond the point to which they brought us. Periods of change such as ours are always characterized by the temptation to lose touch with the work of the past, to disengage from the work of previous generations in favor of important new reference points--we need hardly look any farther than our predecessors' attitudes toward their predecessors, the antiquarians of 1780-1860, for an object lesson in this. (6) Just as was the case during the shift from the amateur to the professional, the work of editors of Middle English from 1850 to 1970 now seems to us supremely frustrating at minute levels; their methods of transferring manuscript textuality to printed type now seem to us to be largely acts of a kind of performance art; and the care with which most of the generation of critical editors documented information on dialects and textual analogues now seems to us accompanied by a comparable naivete on issues of scribal and authorial context. But if we wish neither to perpetuate stoically our immediate predecessors' methodology, nor to abandon their work as the product of an anachronistic or "broken" editorial mechanism, we must embrace their legacy in more ambitious ways than we have to date. If we do not, we will instead be deciding that the life's work of respected professionals of the recent past such as Bliss, Ker, and Mills will end up fully archaic, alongside the shelves full of Ritsons, Laings, and Ellises: our teachers will quickly become relegated to the status of curiosities among "serious" medievalists. It has happened before, and the chief reason not to let it happen this time, or at least not to let it happen without a deliberate decision to do so, is that it causes a diminution of the integrity of our common scholarly fugue. I think we are now at a point where we can publicly and fully recognize that the shift in medieval textual criticism is an adventure for all of us--I think that we can become better teachers and students by openly investigating how our moment in time (which itself will inevitably become dated and of questionable direct value) was shaped and predicated by the moments of our predecessors.

The following essay is something of an object lesson in this, a gesture in the direction of this kind of editorial biography. Its subject is the extension and intersection of two specific lines of Middle English textual criticism which originated just over 50 years ago, and in which we can vividly see the features and consequences of modernist editorial theory. The bibliographic legacies I want to share with you are the famous Auchinleck Bookshop Theory, first advocated by Laura Hibbard Loomis in the 1940s and eventually one of the more influential accounts of vernacular English manuscript production, and the "Charlemagne and Roland" Theory, constructed by Ronald N. Walpole and Hamilton Smyser shortly after Loomis's work appeared. Though now established as off-target for a variety of reasons, Loomis's theory, especially in its rhetorical presentation, can tell us much about the relationship between printed editions and the criticism based upon them during the latter half of the twentieth century.

My goal in presenting it is not to exhibit its flaws for the sake of the exhibiting of them, but rather to present both Loomis's insights and her limitations in order to help us better understand and visualize the relationship between editorial theory and literary criticism during this period. And far from poking holes in the hard work of Walpole and Smyser (both of whom I respect tremendously), I wish to forestall what I'm beginning to see as quick dismissal of their kind of bibliographic work in favor of new and exciting paradigms--much as the academic professionals of their own generation themselves dismissed the work of their antiquarian predecessors, to our general disservice.

Loomis's Bookshop Theory is an attractive candidate for a contemporary case study because it is fairly finite. It draws textual data from the contents of a single Middle English manuscript: the fourteenth-century Auchinleck Manuscript (National Library of Scotland Advocates' MS. 19.2.1), one of the largest single volumes of medieval English verse. The manuscript contains many devotional works, as well as texts of the romances "Lay le Freine," "Arthur and Merlin," "Sir Orfeo," "Beves of Hamton," and "Sir Tristrem"; each of the latter has been edited several times since the donation of the volume to the Advocates' Library in Edinburgh by James Boswell's grandfather in 1740. There is thus a small sea of critical commentary surrounding the Auchinleck and its parts, but it is navigable and readily charted. (7)

The chief focus of my interest here are Loomis's suggestions, (1) that a particular kind of London scriptorium (her "Bookshop"), with ascertainable resources and procedures, produced the Auchinleck manuscript around the year 1330, and (2) that Geoffrey Chaucer used the Auchinleck manuscript as a wordmine for some of his tail-rhymes in "The Tale of Sir Thopas." But one cannot engage the subject of the Bookshop without considering a lesser-known but wider-ranging theory concerning the two Charlemagne romances in the manuscript, "Roland and Vernagu" and "Otuel a Knight": the majority of critical use of Loomis's Bookshop Theory was made, in fact, by Ronald Walpole and Hamilton Smyser in their attempts to reconstruct the provenance of the Charlemagne legends in England. The intersection of these two theories can teach us much about the consequences of implicit trust in editorial action, and thus serves well as an example of the uses to which traditionally edited material can be put in the twenty-first century.

The crux that makes this line of inquiry worth sharing is that the problematic nature of their scholarship has amounted to an opaque barrier against its own advancement. Judith Mordkoff and Timothy Shonk, the two most in-depth Auchinleck commentators since Loomis's time, react to the modernist work of Loomis, Smyser, and Walpole in large part by disregarding it. (8) In their work (which dates to after 1980), Mordkoff and Shonk establish a firm foundation of manuscript evidence on which textual and literary criticism may be built in the future--but it is important to note that in large part they have not responded to the internal details of earlier essays written on the subject. Instead they, and Miceal Vaughan and David Burnley, have abandoned those essays and have returned to first principles of their own choosing, a legitimate and obviously effective course of action in this case.

FIXING CHAUCER

Between 1940 and 1951, Laura Hibbard Loomis published four influential articles that display a marked reliance on edited texts for manuscript data, data that Loomis would in turn use to construct the farthest-reaching theory ever made concerning the origins of the Auchinleck manuscript. (9) The first, "Chaucer and the Auchinleck MS: 'Thopas' and 'Guy of Warwick,'" has as its stated goal to prove conclusively that Geoffrey Chaucer had read the Auchinleck versions of several tail-rhyme romances before he composed his "Tale of Sir Thopas." Loomis offers evidence to support her conviction that Chaucer was influenced by the Auchinleck texts of the poems to the extent that phrases and lines from the latter show up verbatim in his own poem (in the context of caricature, of course).

Her investigation involves the assumption that proof of Chaucer's use of the Auchinleck lies in the specific combinations of words shared by "Thopas" and the lengthy Auchinleck romance "Guy of Warwick": while "Guy" does survive in other manuscripts and in other versions, Loomis determined that a "crucial combination of words, phrases, rimes, details, etc." allows us to pinpoint the Auchinleck as the precise "Thopas" source. (10) The supporting evidence consists of passages from "Thopas" and "Guy" printed in parallel columns so that verbal similarities may be observed by readers who may judge the correspondences for themselves. The correspondences are striking, and are made more so by having particularly close words and phrases being printed in italics and footnotes. This evidence continues for eight pages in the essay, at the end of which Loomis closes her discussion by stating that "The long, foregoing list of quotations needs no further commentary; the individual and the cumulative effect can hardly be denied." (11) The presentation format, as well as the strong rhetorical closure in which it is set, compel the reader to understand that Chaucer's use of the Auchinleck Manuscript is a revealed fact.

It was a fabulous prospect for 1940--"a source-seeker's dream," to borrow a phrase from Ronald Walpole, the first respondent to the essay. But upon consideration there arise three difficulties with the notion of textual proof as Loomis has presented it. First is the organic assumption that not only does a text beget a text (the useful, though dangerous, foundation of stemmatics), but that a text can only be begotten by a text--that it would have been unlikely or impossible for Chaucer to have created the phrases in "Sir Thopas" without direct visual scrutiny of a pre-existing exemplar. As anyone who has read half a dozen of them knows, metrical romances are extraordinarily formulaic creations: this is the feature which caused them to be used as target practice by early philologists, and especially by practitioners of the New Criticism. (12) And we must notice that nearly all of the evidence provided by Loomis involves formulaic phrases or end-rhymes belonging to the tail-rhyme repertoire. Loomis acknowledges the formulaic nature of "Guy," but argues that "nearly all the combinations of elements here noted are, in themselves, unique." She continues:

Though we cannot, in view of all the manuscripts known to be lost, give too much importance to the fact that today a manuscript text is unique, still we cannot refuse to credit the converging evidence of unique combination [sic] of texts and of unique readings in those texts. For we have, after all, the positive evidence of all our still extant manuscripts to establish their essential individuality. Manuscripts can be grouped by families, relationships can be traced, but in Middle English at least it is certain that no compilation of romances exists which is an exact copy of another. It is, therefore, fundamentally improbable that a manuscript satisfying not one but all of the exact and peculiar conditions noted above should ever have been duplicated. It must have been the actual manuscript known to the medieval author. (13)

Half a century later, our reaction to this tight enforcement of literate conditions would be that Chaucer, himself a fourteenth-century word-oriented Londoner, would be just as able to pull popular phrases and tail-rhymes from his own head as he would be from a slightly earlier fourteenth-century London source. But the important point for us in this context is that Loomis does not voice any other possibility because the prevailing conditions for proof of textual identification shared by her generation entirely restrict her at this moment:

In other words, if a medieval author shows himself aware of texts found together only in one manuscript, and of texts that are found nowhere else, and of readings that occur nowhere else, we may feel that in all human probability this was the manuscript that he read. (14)

The conditions under which Loomis is writing do not admit her to voice the potential for an unrecoverable variety of linguistic experience on which the Auchinleck texts rest. The conditions likewise do not admit beneficial whimsy and other sorts of unpredictable or untraceable mental processes; they do admit error, according to the principles of Lachmann and his followers, but only those forms of error that are lexically identifiable.

Most importantly to our purpose, however, the textual assumptions Loomis makes are generated from a particularly tenacious sort of bibliographic inquiry. Her original impulse with these textual conditions--to pinpoint a manuscript source for what we, today, can readily allow to be a feature of popular culture--anchors her investigation firmly in the knowable and the structural. In the years since this "Thopas" essay was written, Middle English studies have developed in the direction of weighing the likely gain of such an inquiry: we have learned (partly through the lessons of investigators such as Loomis herself) that we are sometimes able to make such precise linguistic connections, but that our decision to do so must involve estimations of how such connections inform our thinking about art, craft, economy, and so on. (15) Perhaps in our own growing textual promiscuity we have become less enthusiastic at the prospect of finding cause-and-effect linguistic data, but certainly our willingness to be skeptical of the suggestion of structures of all kinds--in this case, to conceive that all verbal constructions are to some degree influenced by other verbal constructions, in Zumthor's model of mouvance--causes us, in spite of all the source-seeker's wishes, to find only a less exciting contextuality. To read Loomis's essay today is thus to look directly at a conflict between structural security and poststructural variation, at the fixed-verbal knocking against the dynamic-verbal. (16) While a scholar of Loomis's generation was prepared to conceive of unknowable dynamic forces being responsible for the inspiration of authors like Chaucer (i.e., the identifiable moving forces behind canonical texts), she was not prepared to treat the actual inscription of texts on parchment by nonauthorities with similar rigor.

The first objection raised by our generation of medieval studies to this thread of investigation, then, is that it relies excessively on literate transmission as the only precedent force to inscription. A second objection is easier to describe, but involves Loomis's rhetorical posture more extensively. The quotations from "The Tale of Sir Thopas" are offered to the reader as extracts in consecutive order. The matching quotations from "Guy of Warwick," however, are imported as needed from all over the poem--from locations 276 stanzas (more than 3300 lines) apart in the case of "Thopas" stanza 2. The asterisks Loomis places next to stanza/line references are intended to suggest particularly close textual similarities, but the lines are in anything but close proximity in the context of "Guy of Warwick." This emphatic presentation of evidence, once again, is based in a willingness to objectify the subject text and view it as a word-mine; Loomis is in effect matching in her essay's own presentation the technique Chaucer "must have" used if he did pull phrases from the Auchinleck.

Now, I do not mean to deny the possibility that Chaucer did once have the Auchinleck manuscript in his hands and read from it--that momentous claim itself is not the issue for us here. Loomis's intention may have been, in part, to offer this comparison of "Thopas" and "Guy" in order to show that the Auchinleck was one of the many physical copies that Chaucer had to have read from in his lifetime. But we must note that this is not what she says--it is not a part of Loomis's rhetorical posture in this essay to allow for a range of probability about her findings. When she is finished presenting her data she presents the case as closed, and offers no qualifications or calls for further inquiry. This syllogistic quality accompanies and reinforces the textual analysis, which treats manuscript distances as inconsequential and which thrusts the two romances together disproportionately.

Finally--and this third objection may contain the most important lesson for our contemporary work--Loomis's quotations from Chaucer and from the Auchinleck poem come entirely from edited sources: from F. N. Robinson's edition of The Complete Works of Geoffrey Chaucer for the "Thopas" selections, and from Julius Zupitza's EETS edition of "Guy of Warwick." Although in one note she does acknowledge "the helpful courtesy of M. R. Dobie, Keeper of MSS [of the Advocates' Library in Edinburgh] in replying to questions and in facilitating my use of the MS," it is clear that Loomis has not seen the Auchinleck itself when she writes this essay. (17) Her choice of Zupitza's edition is not a bad one: it was the standard edition of "Guy" at the time, and in many ways it is a technically reliable critical edition. Her reliance on Robinson is likewise unsurprising, but nowhere does Loomis mention the important fact that what Robinson offers in his Chaucer edition is not the record of a single manuscript. In keeping with a standard practice of Chaucer editors, Robinson's text is composed of the Ellesmere manuscript of the Tales freely emended by a variety of other copies in cases of unsatisfactory readings. (18)

A collation of the "Thopas" text with Robinson's textual notes shows that there is not a great deal of variation among the manuscripts in this particular instance, but there are some readings which do mar the ease of Loomis's findings. Line 805 of "Thopas," for example, which in Robinson's edition reads "That to him durste ride or goon," is noted in the collation to exist only in three manuscripts: Cambridge Dd.4.24, Cardigan, and British Library Royal 17.D.xv. If it were Loomis's project to compare two textual witnesses (Auchinleck and Ellesmere, for example), the line should not be part of the evidence. But she makes the characteristic twentieth-century leap of editorial faith--she equates Robinson with Chaucer--and thus she uses the editor's product in order to quote the line as one element of her textual evidence. When she uses Robinson we see her silently accept the considerable grooming of Chaucer, and with Zupitza we see textual evidence being used without regard to narrative integrity.

The attendant question for us is not one of motivation: Laura Loomis has not taken advantage of her editors for the sake of pursuing a thesis that is ambitiously out of step with prevailing critical practice. Instead, she is accepting, as nearly every professional medievalist of her age did, that the editorial mechanism has done its job correctly and that there is no reason why the edition should not be attributed the authority necessary to make this kind of comparison. To avoid condescension, the perspective I believe we in the twenty-first century must adopt regarding this kind of use of edited material is that Loomis's paradigm of editorial authority at once caused her to profess manuscript fidelity while not in fact using manuscript data. In nearly every paragraph of the essay she makes earnest claims of explicitness in her use of what she calls "manuscript evidence"--in fact, however, when we read Loomis's article we are witnessing a stage in Middle English studies when the phrase "manuscript evidence" is earnestly being used to describe textual data that may have undergone any of a number of editorial transformations.

WITHIN THE WALLS OF THE AUCHINLECK BOOKSHOP

I have presented her methodology in this rather protracted fashion because Laura Loomis again uses the same approach in a much more influential article, "The Auchinleck Manuscript and a Possible London Bookshop of 1330-1340," published in 1942. There is a great deal about which to applaud this article: it represents the first scholarly suggestion of comprehensive pre-planning on the part of the Auchinleck's principal scribe, and Loomis is responsible for bringing the actions of the scribes themselves into the foreground of Auchinleck criticism. Her general conclusion, in fact--that the principal scribe held the role of a sort of "general editor" of the manuscript--is still compelling, and has been corroborated by subsequent scholarship using more extensive manuscript data. But her reliance on edited publications as sufficient witnesses of manuscript texts in this second essay again brings about misleadingly forceful conclusions.

Starting from the observation that lines of the Auchinleck poems penned by different scribes repeat with accuracy, Loomis posited that the only explanation for such borrowings was that the five scribes who wrote out the manuscript were gathered in one place at one time. (19) Adding her opinion that the Auchinleck as a book is not sufficiently decorative to be considered a work of art (she compares the extant miniatures to the work of the illustrator of Cotton Nero A.x. as examples of poor-quality illumination) she concludes that the manuscript is the result of a commercial enterprise, a "bookshop" under the direction of an organizing leader who was in charge of handing out copying assignments to the men working for him. She suggests that a purchaser would order a copy of this or that text, and if it were available it would be translated into English verse (if necessary) and copied into gatherings that would later be bound with others to complete the special order.

After printing parallel quotations from "Guy of Warwick" and "Reinbrun," and from "Guy of Warwick" and "Amis and Amiloun" in the format noted above in the study of "Sir Thopas," Loomis offers the following summary of her findings:

For all but five romances (Orfeo, Degarg, Otuel, Horn Childe, and the King of Tars) in the Auchinleck MS, we have still extant French texts. Though no one of these texts may be the precise original from which the English translators made their versions, it must have been on French texts close to these that they worked; by these they meant the source book or geste to which they not infrequently referred. These Auchinleck romances were copied from the texts of translators, or workers with texts, not with tradition or invention. The "authors" were in no wise original poets, and did not, as it is generally admitted, achieve distinction of style, though a few poems, like Orfeo, have genuine charm. (20)

We cannot disagree with the fundamental wisdom of this observation: there are ubiquitous verbal parallels among the Auchinleck poems, some number of which are no doubt due to precisely the direct scribal copying that Loomis characterizes. But the strictly literate paradigm applied to the copying of the manuscript, as well as the lack of other types of physical evidence with which our generation of medieval studies complements verbal data (catchphrases, quire arrangement, rubrication, variations in script suggestive of scribal stints, etc.) is once again accompanied by a rhetorical forcefulness like that in the "Thopas" article. After declaring that it is impossible to ascribe the verbal parallels between "Amis" and "Guy" to anything but direct textual borrowing, Loomis goes on to dismiss those instances in which the scribes have not conformed to her expectations by downplaying scribal accuracy in vernacular texts of the fourteenth century. (21)

Laura Loomis is marking an important point in the history of work on the Auchinleck manuscript: she has created in the Bookshop a setting where any of a number of imaginable verbal relationships can take place. Scribes' accuracy in the copying process can alternately be "slavishly" close or can vary through the result of "inspiration" by the texts they copy; these same scribes most often plod along in a middle-ground of "hackwork." Scribes can make separate copies of the same poem in various meters and rhymes, and then choose one and discard the others; they can also forget what they are doing and splice the copies together, resulting in the kind of enigmatic metrical shifts that so vexed Eugen Kolbing in Beves of Hamtoun. Literally anything can happen to manuscripts within the walls of the Auchinleck Bookshop because the prevailing academic ethos to connect texts allowed twentieth-century critics to feel that they were looking at The Answer to the Manuscript. The conviction of positively identified verbal transmission, in this case, carries with it no invitation to future scholarly dialogue on the subject.

There is a good deal of irony in the point that such a detailed explanation of the genesis of a medieval book should be made by a person who had not actually seen it--I doubt there can be any more symbolic point made about the textual studies of Loomis's generation than that a reliance on printed transcripts predicated such a situation. Faith in critical editing made the Auchinleck Bookshop theory possible; it was the logical product of the romantic textual vision inherited from nineteenth-century philologists, reinforced by a strong desire to see order existing in the midst of chaos and the vocalization of that order in a commanding tone. (22)

A MIDDLE ENGLISH CHARLEMAGNE CYCLE?

The little tail-rhyme romance "Roland and Vernagu" (item 36 in the Auchinleck manuscript) might seem an unlikely candidate to be the subject of any major textual theory. Just 879 lines long, the poem tells of various adventures of Charlemagne and Roland during the former's apocryphal wars in Spain. Its narrative is episodic, a series of scenes selected from the Pseudo-Turpin Chronicle; its meter and rhyme are not unusual for the tail-rhyme genre; there are no plot elements (apart from the simultaneous duel and theological debate between Roland and the Saracen giant Vernagu) which strike the reader as extraordinary. Although it is a unique copy and has been edited three times, the poem has never been considered important to scholars or students by virtue of its poetic content. (23) Surprisingly, however, this obscure romance lies at the center of an impressive theory of a lost Charlemagne epic--a theory which gained most of its eminence in the later 1940s because, interestingly enough, it was perfectly compatible with Laura Loomis's Bookshop Theory.

Long before Loomis's time, scholars who looked at the ten extant Middle English Charlemagne verse romances busied themselves with documenting the ways in which two of them, "Roland and Vernagu" and "Otuel and Roland" (the latter of which is in "The Fillingham Manuscript," now British Library Add. MS 37492), could be seen to be remnants of an older and longer English work given the name "Charlemagne and Roland" by Gaston Paris in 1865. Paris called this hypothetical ur-text a "cyclic poem," a momentous collection of English verse based on French sources that contained material now extant only in the shorter poems. Despite the casual nature of his suggestion in its original context, this hypothetical "Charlemagne and Roland" has since Paris's time attracted an increasing number of attributions normally associated only with existing literary works. Its existence is promoted without qualification in Hamilton Smyser's entry on Charlemagne legends in the first fascicle of the revised Manual of the Writings in Middle English, and subsequent published scholarship has failed to question seriously any of the claims made for it there. The concept of the cyclic poem "Charlemagne and Roland" is therefore still, in 2003, an unchallenged explanation of the transmission of matiere de France into England. (24)

Gaston Paris offered the first comprehensive scholarly discussion of the Middle English Charlemagne romances in his Histoire poetique de Charlemagne, where he sought to argue that the English poems showed the same general filial relationship to common sources as did the Charlemagne works in French, German, Italian, Spanish, and Norse. His method was not necessarily to point out gaps in contemporary knowledge, but rather to reconstruct decisively with available evidence the spread of the Charlemagne legends throughout vernacular cultures in the later Middle Ages. Paris introduces "Charlemagne and Roland" to his readers as a fait accompli--as if he were only providing a name for a somewhat neglected, but definitely recognizable, work: "This is the title that I believe should be restored to a poem which has not yet been considered as it should be." Stating further that the poem is "a type of resume of Charlemagne's wars against the Saracens," he lists its parts as coming from three sources: the Pseudo-Turpin Chronicle, the Chanson d'Otinel, and the Latin account of Charlemagne's journey to Jerusalem commonly referred to as the Descriptio. Paris believed that the poem was "a kind of cyclic work," similar to the German Karl Meinet and the Norse/Icelandic Karlamagnus saga, though of inferior quality. (25)

The claims are very impressive, and Paris's reasoning is quite direct. He knew from published descriptions that "Roland and Vernagu" and "Otuel and Roland" were both in twelve-line tail-rhyme stanzas, and that, apart from the Otuel material with which the latter poem begins, they both contained episodes derived from the Pseudo-Turpin Chronicle. Further, he knew that there was very little overlap in the episodes, and that "Roland and Vernagu" generally drew from the first half of the Pseudo-Turpin while "Otuel and Roland" drew from the second half. The two poems thus appeared complementary to him, and to explain their estrangement he stated that their author, presumably once he had finished copying "Roland and Vernagu" into the Auchinleck manuscript, decided that he preferred a couplet version of the Otuel story and "spliced" the latter in so that it became Auchinleck item 37, titled in the manuscript "Otuel a Knight." "Charlemagne and Roland," as Paris introduced it, must therefore have immediately predated the Auchinleck manuscript and must have been composed of an ur-"Roland and Vernagu" followed by an ur-"Otuel and Roland."

As is the case with the Bookshop Theory, we can identify several rhetorical devices as being largely responsible for the impression of unity in this proposed solution. Gaston Paris was not concerned with the details of verse forms or of manuscript filiation, but rather with larger structural features: he expresses the elements of his "Charlemagne and Roland" as numbered sections of a larger corpus and does not discuss aspects interior to those sections. Moreover, he offers three venerable sources from which a fourth unknown one seems to derive naturally--the weight of tradition is thus on his side, and the notion of a cyclic work which brings together diverse strands of a legend is always attractive and compelling. But we must realize the implications of this macroscopic approach: in comparing "Charlemagne and Roland" to collections like the Icelandic sagas, Paris intended the word "cyclic" to be synonymous with "collective," like an anthology, and the result of a single directing hand.

Despite its casual nature, Paris's declaration of the existence of a Middle English Charlemagne cycle was enthusiastically taken as fact: "Charlemagne and Roland" was invoked in passing as an extant title in the decades surrounding the turn of the twentieth century. Although the cyclic poem theory was long unverifiable because the Fillingham manuscript containing "Otuel and Roland" was apparently misshelved for most of the nineteenth century, between 1865 and 1950 it was widely acknowledged to be the authoritative explanation for these two romances' existence. (26) The claims for the cyclic work are undermined, however, by the simple fact that there is no other objective evidence of a single comprehensive work such as "Charlemagne and Roland" in English before the time of Caxton--and certainly not one in verse. On both the general and specific levels the theory does not sustain close inspection; it reveals itself, like the Auchinleck Bookshop theory, to be in large part the result of preconceptions about the actions of scribes set out in particularly aggressive rhetoric. And, significantly for our present concern, it shares Laura Loomis's approach in her Auchinleck essays by relying wholly on printed editions for its evidentiary foundation.

Without exception, the men and women who responded to Paris's views on the lost cyclic poem did so from the assumption that "Roland and Vernagu" and its would-be complement, "Otuel and Roland," could not have been written intentionally in their surviving forms. When compared structurally to their eventual source, the Pseudo-Turpin, the romances seemed at best confused and naive readings; at worst, they seemed inept and embarrassing examples of medieval versification. Shuffled folios and scriptorium intrigue were suggested to explain the more stubborn interpretive points in what was repeatedly described as a decay from Latin prose to English verse, and various kinds of textual evidence (similar phrases, and identical or similar spellings of proper nouns) were brought to bear to buttress the broader theoretical discussion. Moreover, each of the contributors to scholarship on the Middle English Charlemagne cycle approached his work on the vernacular forms from a previously existing interaction with the Latin history of the Pseudo-Turpin: their scholarly ideologies, in other words, promoted consideration of the English texts almost exclusively in terms of those texts' perceived stemmatic descent from Latin. (Gaston Paris, for example, very likely chose the title "Charlemagne and Roland" in the first place because it reinforced his a priori assumptions about the influence in England of the Pseudo-Turpin Chronicle--the formal Latin title of the latter is Historia Karoli Magni et Rotholandi--concerning which, in 1865, he had just written his dissertation for the Ecole des Chartes.)

The cyclic poem theory and its dependence on the Auchinleck text of "Roland and Vernagu" was not evaluated at all until 1935, when Mary Isabelle O'Sullivan edited the newly rediscovered Fillingham manuscript's text of "Otuel and Roland" for the EETS; she enthusiastically agreed with Paris that "Otuel and Roland" and "Roland and Vernagu" are the remnants of a cyclic work. (27) O'Sullivan has the distinction of being the first person to bring any kind of detailed manuscript evidence into the discussion of the Middle English Charlemagne corpus, and her pronouncement of the viability of the theory came at just the time that Laura Loomis was writing her essays about Chaucer and the Auchinleck Bookshop. A singular convergence is visible here: the tandem influence of O'Sullivan's edition (1935) and Loomis's articles (1940 and 1942) prompted a subsequent series of articles in the later 1940s and 1950s by Ronald Walpole and Hamilton Smyser that cemented the cyclic poem theory into Middle English literary criticism once and for all. Gaston Paris's seventy-year-old notion of a "Charlemagne and Roland" was about to find a new home--in the Auchinleck Bookshop.

Between them, Ronald Walpole and Hamilton Smyser detailed the circumstances whereby a thirteenth-century French manuscript of the Pseudo-Turpin Chronicle "became" the two extant English tail-rhyme poems. In their essays they build upon each other's work, and eventually they offer a complete conjectural reconstruction of "Charlemagne and Roland" self-professedly free from major omissions or errors. (28) The articles for the most part elaborate on the cruces introduced by O'Sullivan, and attempt to reconstruct an ancestry for the hypothetical cyclic poem and thus to offer evidence for assigning an immediate manuscript source to "Roland and Vernagu."

There are four distinct settings to which Walpole and Smyser point when they describe the textual prehistory of "Charlemagne and Roland." First, at about the turn of the thirteenth century, Pierre de Beauvais translated into French the Latin Descriptio story and recommended it in an introduction as a neglected episode in the story of Charlemagne's Spanish campaigns. At the same time, a Latin C-text of the Pseudo-Turpin Chronicle was translated into French by a scribe identifying himself as "Maitre Jehans" (copies of this particular French version of the Turpin version are thus collectively called "the Johannes Turpin"). Soon afterwards, an unknown scribe (possibly Pierre de Beauvais himself) copied these two prose texts back-to-back in one manuscript, with the French Descriptio followed immediately by French Johannes Turpin.

Second, soon after the above took place, a scribe working for Count Renaud de Boulogne "fused" (to use Walpole's word) the texts together--he copied them without distinguishing between Descriptio and Turpin--and shortened the Descriptio so as to make it a smoother introduction to the more extensive Turpin material. The result is alternately called by Walpole a "redacted Johannes Descriptio-Turpin" or an "Estoire de Charlemagne," since it is in French and contains the journeys to both Jerusalem and Spain. Some 32 copies of the Johannes Turpin are presently known to exist, but the redaction that interested Walpole is British Library Ms. Add. 40142, made in England around 1250. (29)

Thus far the manuscript record may be pieced together with some certainty, and a number of other Turpin experts have corroborated the high value of Walpole's work in the Old French copies. (30) But the third and fourth settings for the textual history of "Charlemagne and Roland" move into the peculiarly speculative, as follows: in the last years of the thirteenth century, slightly later than the activity just described, a French manuscript of the Chanson d'Otinel was translated into English couplets in the same London bookshop (Loomis's Bookshop) that was to have produced the Auchinleck manuscript. This couplet Otinel does not survive, but according to Smyser it must have been redacted twice, once in tail-rhyme form (a copy of which survives as the first half of the Fillingham "Otuel and Roland") and once again in couplet form (surviving as the Auchinleck "Otuel a Knight"). (31) Finally, Walpole believed that at about the same time (i.e., around 1300), the Johannes Pseudo-Turpin contained in Add. 40142 was changed into English tail-rhyme verse in Loomis's bookshop as part of the Auchinleck project. The only surviving portion of this translation, he contends, is "Roland and Vernagu," which he copiously demonstrates to have been directly copied--with no possibility of an intermediate manuscript--from BL Add. 40142. (32)

Although neither Walpole nor Smyser offers a stemma of "Charlemagne and Roland" per se, we can piece together from their statements what one would look like. (33) Figure 1 manifests the necessary stemma to produce the pre-history of the two Auchinleck Charlemagne romances, a plausible and sensible one, despite the presence of several texts or textual groups for which Walpole could not yet identify specific manuscripts (these are indicated with asterisks: *). Figure 2, however, manifests the stemma necessary to reflect the immediate provenance of the Otuel romances (again condensed from the complex prose descriptions of Walpole and Smyser): notice that it is much more problematic since we have made a form change into poetry, and notice that the speculating of these Middle English copies is treated in a much looser manner than the Old French prose.

[FIGURES 1-2 OMITTED]

If we further unroll the skein of complex claims made for this group of texts, we are confronted with an inescapable contradiction. If "Charlemagne and Roland" were actually present as a single entity in the Auchinleck manuscript, with BL Add. 40142 as its direct source, and if the Fillingham "Otuel and Roland" is a remnant of the same "Charlemagne and Roland," then "Otuel and Roland" must also be at least in part a translation of one of the Old French texts in BL Add. 40142. It definitely is not. Moreover, we are still left with the question of why the principal scribe of the Auchinleck would have put the couplet "Otuel a Knight" in his work if a tailorhyme version of the same story had been made at the same time from the same source. The "Charlemagne and Roland" envisioned by Walpole and Smyser would require the nearly paradoxical filiation represented in Figure 3.

[FIGURE 3 OMITTED]

The stemmatic mechanism obviously self-destructs when imposed upon these vernacular texts. The idea would be that the four (or more) works on the second line of Figure 3 would necessarily have been made at the same time and place--Loomis's Auchinleck Bookshop--and toward the common goal of presenting a single, comprehensive collection of the foreign campaigns of Charlemagne. "Charlemagne and Roland," according to Walpole and Smyser's descriptions, must not only have been created in the Auchinleck bookshop, but must have existed in the Auchinleck manuscript in the following order: (1) "Roland and Vernagu"; (2) "Otuel a Knight"; and (3) an exemplar of the second half of "Otuel and Roland." The tail-rhyme "Otuel" must have been discarded "for some reason" once it was completed, and must have disappeared after a copy was made in Norwich 150 years later; to this copy was suffixed a copy of the Turpin-based tail-rhyme episodes originally following "Otuel a Knight" in the Auchinleck, which have also subsequently disappeared. The strict requirement of written exemplars in each and every stage of these changes is clearly unfeasible.

Let me include one final detail of the assumed complementary relationship between "Roland and Vernagu" and "Otuel and Roland," which will complete what my reader is no doubt finding to be an already-tedious overview of the cyclic poem theory. In his 1946 article, Smyser sought to prove that the verse "table of contents" present in "Otuel and Roland" (lines 1-44) was once present at the beginning of "Roland and Vernagu." The general nature of the lines (which introduce characters and situations found in the Pseudo-Turpin), as well as the similarity between the first four extant lines of "Roland and Vernagu" and lines 41-44 of "Otuel and Roland," led Smyser to believe that the entire introduction was present verbatim in the earlier poem. He offers as evidence the six isolated letter-forms that remain visible on the margin stub of folio 262a in the Auchinleck, and in a chart he matches them up to corresponding lines in "Otuel and Roland." Although the result is hardly a close correspondence (and Smyser himself admits his reliance on Eugen Kolbing's descriptions of the margin stub--like Loomis, he himself has not seen the manuscript), he feels secure in claiming "the authority and antiquity of a quondam appearance in the Auchinleck MS" for the "Otuel and Roland" introduction. (34)

My attention in Hamilton Smyser's analysis comes to rest once again on the basic assumption he makes about the literal manner in which the poems were transmitted and on the rhetoric in which he couches his entire discussion of them. The "Otuel and Roland" introduction contains references to many episodes found in that poem, as well as to the duel between Roland and Vernagu and another episode not present in it. Like the rest of the poem, the introduction seems to be an altered version of an earlier English text made by persons not interested in verbal accuracy: several syntactic units make no logical sense, and the stanzas are of varying lengths. Only a highly restrictive (and wishful) characterization of scribal activity would necessitate that the same promises of the same episodes be made at the beginning of "Roland and Vernagu," a comparatively sound and finite work, and that six letters might supply a clinching proof of the contents of a lost text. These are the pragmatic consequences of what Walpole and Smyser describe: a complex series of textual circumstances relying on the confusion or extravagance of the Auchinleck scribes and the momentous re-emergence and subsequent disappearance of written exemplars.

THE BOOKSHOP AS A STATIONARY INSTITUTION

I hope I have amply demonstrated that Laura Loomis's Bookshop Theory was designed to enable tremendous textual interconnections. It offered a textual Land of Cokayne in which even the most puzzling circumstances could be explained, and as investigators of the Charlemagne legends Ronald Walpole and Hamilton Smyser seized on it to describe phenomena which they would have been hard pressed to explain otherwise. They presented the genesis of the Charlemagne romances to us in the same plausible scribal setting that Loomis had described: a group of men, certainly working under one roof in a lay scriptorium, directed by a general editor whose function it was to oversee the assembly of a finished product. Each of the Auchinleck scribes (Loomis's "literary hacks") would translate (as necessary), versify (as necessary), and copy down the piece, relying continually on the written exemplars within his reach for finishing rhymes or apt epithets--and in this recoverable manner they created the intertextual nature of many Middle English romances.

It is certainly the case that the Bookshop setting attracted Walpole and Smyser because it suggested either the presence of a large stationary library or regular access to manuscripts by Bookshop scribes. The critical approaches of both men, like that of Loomis herself, required the availability of an indefinite pool of texts. Walpole's own research on the Pseudo-Turpin brought him to the Old French manuscript BL Add. 40142, and he was convinced that it literally sat on the shelves in the Bookshop and must have been the source for all but the Otuel episode of "Charlemagne and Roland." He repeatedly voiced his conviction that "Roland and Vernagu" was written from this particular manuscript of the Turpin, using stemmatic representations as his visual proof much as Loomis had used edited texts to prove that Chaucer must have used the Auchinleck to write "Sir Thopas."

Our object lesson on the ways in which positivistic editorial postures can create and reinforce a theory of provenance with tremendous verisimilitude is nearly complete--but the accretion of literate characteristics in the case of the Auchinleck Bookshop and the Charlemagne Cycle does not end with Walpole's last article on the subject in 1951. In 1961, two decades after Loomis published her first article on the Auchinleck, French medievalist Andre de Mandach used the work on the Middle English Charlemagne cycle to support his expansive notion of an international carolingian poetic consciousness. Mandach increased the legitimacy of the Walpole-Smyser theory by renaming "Charlemagne and Roland" as "La Geste de Charlemagne Auchin-Fillingham de l'Angleterre." And he took the extraordinary step of offering to name its author: at the conclusion of "Otuel and Roland" in the Fillingham manuscript there is a faint colophon which seems to read "Amen quod I Gage." Mandach suggested that this reference reveals the name of the creator of "Charlemagne and Roland"--that his name was probably John Gage, "'Joseph' being an unpopular name in anti-Semitic thirteenth-century England," and that this "*John Gage" was the single directing hand responsible for the cyclic poem. (35)

Apart from the satisfyingly concrete addition of the name, the pages in Mandach's Naissance et developpement devoted to the English cyclic poem contain the same information as the final major publication which has been influenced by Loomis's Auchinleck Bookshop Theory, the entry on Charlemagne legends in Burke Severs's 1967 revision of the Manual of the Writings in Middle English. As I said in my introduction to this essay, the importance I find in all this textual history lies not just in its ability to help us understand the conditions and paradigms which guided past interpretations, but also in its ability to reveal the extent to which those interpretations are still alive in reference materials we ourselves may own and use. The entry in the revised Manual (by none other than Hamilton Smyser himself) further validates both the "Charlemagne and Roland" and Auchinleck Bookshop theories by assigning them their own bibliographic reference numbers and (once again) avoiding any tentativeness in the presentation of the theories. Smyser chose to keep John Edwin Wells's arbitrary classification system of three "groups" of Charlemagne romances, changing only the order in which the members of the Otuel group are given--he rearranges the section on "Otuel and Roland" so that it follows "Roland and Vernagu," mirroring his conviction of their quondam manuscript unity with a new and permanent proximity in the Manual. (36) The conviction itself does not stand up to the scrutiny of a new generation of investigators--but in this case a considerable amount of publishing on the subject is referenced, and titles are listed, and authors and dates of hypothetical composition are available.

BACK TO THE FUTURE

Let me summarize for you the caveats, or more accurately the "ongoing challenges," I have found in my work with this thread of criticism. We might style them as "rules of thumb," distinct attitudes toward the nature of pre-1970 scholarship which I am finding sound and useful, and which my reader may find so as well. First, while some of our critical predecessors' observations and technical apparatus will be useful to us, their larger conclusive frames rarely will be. In the present case, this means that we can still find value in the perceptions of proximity of texts, but the ultimate reason for their presentation in academic writing--usually, to prove a verbatim textual relationship--will almost never be feasible in its explicit form. This is caused by the nearly universal negative assumptions about scribal activity and the concomitant elevation of the intentionality of named or definable authors. Frankly, I don't see how we can resuscitate scholarly work that is too closely involved with those assumptions.

Second, we must feel invited to re-inspect directly and with great energy two specific kinds of textual situations which our predecessors regularly engaged: those in which the writer has directed our attention to a likely proximity of manuscript sources (though without, as I have just said, the conviction of necessary literal borrowing), and conversely to those places where the writer clearly avoids giving manuscript details. The conspicuous absence of data in a context of otherwise data-richness is usually a foghorn to us that a form of activity that does not fit the investigational paradigm is at work, and it is exactly those marginalized activities we should seek out. We must therefore accustom ourselves to expect, and we must equip ourselves to see through, a variety of forms of rhetorical expression that claim clear proof, that close down lines of inquiry to further investigation, or that seal a scribal situation away from messy (unpredictable, unquantifiable) cultural and social influences. If I could point to any single characteristic of late nineteenth- and early twentieth-century medievalist writing that I would deem so misleading as to actually verge on the pernicious, it is this rigorous vocalization of closure--but it is only a rhetorical posture, and we are rhetorically sensitive readers.

Finally, in instances where actual manuscript details are brought up (i.e., when edited texts are not being relied upon to stand as textual evidence), it is in our best interest to double-check the work of our predecessors using hard-copy or digital facsimile images, or by examining the artifacts ourselves. This might seem like a simple-minded statement, but in the cases of both Loomis and Smyser we have witnessed visual features of the Auchinleck and Cotton Nero A.x. being used second-hand within the weave of an argument--verbal descriptions of manuscript features, in other words, are used because time, technology, and/ or travel may not allow the writer access to images or to the artifacts themselves. Within a few years, not only will the Auchinleck be available in digitized form with linked transcriptions and commentary, but so will the other copies involved in the two theories (the distinct scribal versions of the various Auchinleck items, the copies of Chaucer's "Thopas," and even less celebrated artifacts like the Old French Descriptio-Turpin in BL Add. 40142). Such replications will never replace the artifacts, of course, but promiscuity with visual images is one of our distinct uniquenesses as twenty-first-century medievalists. Just where this last rule of thumb leaves the entire corpus of nineteenth- and twentieth-century critical editions, I really do not know.

When I first read Bernard Cerquiglini's Eloge de la variant some years ago, I nearly laughed myself limp at the title of one of its chapters: Gaston Paris et les dinosaurs. (37) The wit and elegance of that move--backed up, of course, by Cerquiglini's keen discussion of the Franco-Prussian generation and its place in the story of philology--has inspired and lifted my private thinking on the matter of scribal variation more than any other single image. But as this essay obviously demonstrates, I've come to feel that the paleological metaphor doesn't entirely fit comfortably. Not only are we not nearly so far removed in time and space from our predecessors (we might call them our bibliographic grandparents, or perhaps great-grandparents), but more importantly their work is not fossilized: it is lively, it is still very much among us in viva voce despite our new ideological distance from it, and this is especially the case in our more comprehensive reference sources. I should, perhaps, not invoke the easy Jurassic Park philosophy in an august venue like this journal, but instead of occasional bibliographic fossil-diggers we might think of ourselves more as ecologists. We have the ability to keep alive some beautiful and wise creations from a not-too-distant past, though it is up to us to develop ethical consensus so we can make our actions as healthy as they can be--to avoid condescension and sensational freaking, to more deeply understand the professional responsibilities of having fantastic technological advantages not available just a few decades ago, and above all to respect the fact that by embracing the textual (and social, and intellectual, and ecological) assumptions of the previous generation we are doing our best to find the path of wisdom.

Since this essay is intended as much as a revisitation of the useful features of two now-dated theories of scribal activity as it is a historical review of them, let me provide in closing some ways in which Loomis, Walpole, and Smyser's observations have affected the trajectory of comparable new interrogations. The Auchinleck Bookshop theory itself has now been advanced upon in ways which conform to my first rule of thumb: though Loomis's overall idea of a bookshop environment is good, and her perception of the activities of the scribes brought into possibility the very good work of Shonk, Mordkoff, and others, her desire to enforce a verbatim transcription setting has now been contextualized by a broader notion of scribal environment. In other words, the social elements of her observations are in large part being re-spoken in their conceptual form, divorced from their creator's aggressive rhetorical style, and the work that is being undertaken on the Auchinleck now naturally focuses on the artifact itself and the relationship of its contents to its construction. This kind of work, in both published and electronic forms, started up only after a long period of relative scholarly inaction (i.e., from 1950 to 1975).

If the positivistic nature of Loomis's theory caused a slowing of work on the Auchinleck, the obfuscating nature of Walpole's work caused a wholesale gridlock on the subject of the Middle English Charlemagne romances. No one, with the exception of Stephen H. A. Shepherd, has written extensively on them in the fifty years since the Walpole-Smyser essays, probably in large part because of the seemingly impenetrable nature of their conclusions. At last, however, exciting new discoveries such as the Huntington copy of the Middle English Pseudo-Turpin, and powerful new historicist interrogations of the Turpin tradition such as those of Gabrielle Spiegel, are making possible a clear and constructive use of (for example) the valuable elements of Walpole's great facility for the taxonomy of Latin and French prose historical manuscripts. (38) But since all Walpole's publications exist explicitly or implicitly to reinforce his a priori assumption of a master narrative of international Charlemagne cycles, it is unlikely that Spiegel or Shepherd, or any of the other Charlemagne scholars writing today will be able to find value in his Middle English work.

And following my own advice, let me here take one of Hamilton Smyser's passing observations and broaden it a bit. Smyser's observation that the first lines of "Roland and Vernagu" are similar to lines 41-44 of "Otuel and Roland" caused me to wonder if we might find a nonverbal solution to the problem. I'm pleased to be able to say that Smyser was indeed on to something when he drew the attention of the readers of his 1946 essay to the fragmentary Auchinleck leaf preceding the opening of "Roland and Vernagu." The latter begins in progress on the recto of folio 263, after "Lai le Freine" has broken off on the verso of 262. Between them is a missing page, one of the many instances of damage done to the manuscript for the sake of its illuminations in the seventeenth or eighteenth century. In some cases (folios 259 and 268, for example) only the illuminations have been removed and a square or rectangular patch remains. But in other cases, the illuminations are removed along with half or nearly all of the leaf itself--this is the case with the ending of "Lai le Freine" and the beginning of "Roland and Vernagu." The marginal stub Smyser refers to in his 1946 essay (now bearing the designation "folio 262a") allows us to know for certain that the juncture of the poems must have taken place somewhere within the space of two columns, either column b on the recto of this stub or column a on its verso. The space taken up by the text still remaining demands that the juncture was in one of these two columns--that is, in one of the outer columns of the leaf, away from the binding.

Smyser was right: it is very likely that we are missing an entire column of text (forty-four lines, in the regular ruling of this part of the manuscript) from the beginning of "Roland and Vernagu"--but it is a correlative comparison of the condition of the manuscript, and not hypothesis based on structural comparison with other romances in other manuscripts, which suggests this. Three-fourths of the way down the b column on f. 263, beginning at around line 30 and seeming to extend downward, there is a clear, blue, patterned ink stain likely to have been transmitted to the page from the illumination which was once on the verso opposite it. (We can observe similar ink traces throughout the manuscript, opposite more clearly localizable lost miniatures--for example, on folios 15-16, 20-21, 77-78, 200-1,267-68, and 317-18, as well as much darker stains where miniatures or large decorated initials still exist: folios 71-72,303-4, and 325-26.)

If the "Roland and Vernagu" miniature was where the stain indicates--that is, near the bottom of the a column on the verso of the now-missing folio 262a--the number of lines of Auchinleck Scribe I's verse missing would be one entire column (forty-four lines). The poem is in twelve-line stanzas, and if we figure in the isolated four lines with which the extant portion of the poem begins at the top of f. 263, we are left with three stanzas of twelve lines and eight lines of a fourth stanza (making a tidy, Auchinleckian forty-four) in the b column of f. 262a verso. The miniature, in other words, probably occupied the space of about twelve lines at the bottom of the a column, and the text of "Roland and Vernagu" would have begun at the top of b.

I would not have been motivated to make this investigation without Hamilton Smyser's keen attempt to calculate structurally the missing portion of the poem. He was attempting to suggest that the Fillingham "Otuel and Roland" shared verbatim an opening narrative formula with "Roland and Vernagu" as part of his belief that they reflected a pre-existing Middle English exemplar; using only the description of Auchinleck 262a given by Kolbing in 1884, he was searching for a way to match up forty-four lines of lost text between poems written down 150 years apart. Now, half a century later and propelled by his work, I am prepared to corroborate his belief--to the extent that the requisite amount of space was at the right place in the Auchinleck.

That is where I will stop, though. Like anyone in a similar situation I will always wonder about the content of those forty-four lines--but my awareness of the flexibility and fluidity of scribal activity within this manuscript in particular, and my awareness of the role played by performance and memory in the composition of tail-rhyme narratives, will not let me find any constructive reason to try and visually match formulaic openings with another text.

Such is the current condition of one discrete stream of Middle English textual scholarship from the preceding generation: a stagnation that was not able to inspire constructive or profitable engagement, and thus now finds itself on the brink of imminent anachronism. As our new century proceeds, investigators in Middle English will continue to be challenged by the existence of scores (perhaps hundreds) of similar dialogues from the pre-1970 period, and we may witness the summary exclusion from "active" scholarship of more and more names from Middle English scholarship of the period 1870-1970. Our decision about how to deal with this mutability--whether we will try to pull particular elements of profit or delight from these works or will instead relegate whole volumes to acknowledged archival status without much regard to their internal workings--this, I believe, will be a barometer of our true embrace of our New Philology.

Denison University

NOTES

(1) While this essay provides a case study involving Middle English texts in particular, post-structuralist criticism is abundant all across the spectrum of European philology. Some of the best advocates of studying English philology within its historical context, and of the danger of a lack of continuity in philological study, are Ralph Hanna III in Pursuing History: Middle English Manuscripts and their Texts (Stanford U. Press, 1996); Tim William Machan in Textual Criticism and Middle English Texts (U. Press of Virginia, 1994); A.J. Minnis and Charlotte Brewer in Crux and Controversy in Middle English Textual Criticism (Cambridge U. Press, 1992); and Anne Middleton in her essay "Medieval Studies" in Stephen Greenblatt and Giles Gunn's Redrawing the Boundaries: The Transformation of English and American Literary Studies (New York: MLA, 1992), 12-40. Their work is influenced ideologically by, among others, Bernard Cerquiglini's excellent Eloge de la variante: histoire critique de la philologie (Paris: Seuil, 1989); translated by Betsy Wing as In Praise of the Variant: A Critical History of Philology (Johns Hopkins U. Press, 1999), and "The Logic of Textual Criticism and the Way of Genius," the third chapter of Lee Patterson's Negotiating the Past: The Historical Understanding of Medieval Literature (U. of Wisconsin Press, 1987).

The collection of articles in a special edition of Speculum in 1990 had direct influence on the historical frame of the present essay--see especially those by Suzanne Fleischman ("Philology, Linguistics, and the Discourse of the Medieval Text, "Speculum 65 (1990): 19-37), Stephen G. Nichols ("Philology in a Manuscript Culture," 1-10), and Siegfried Wenzel ("Reflections on [New] Philology," 11-18). This Speculum issue has helped to usher in exciting textual and historicist publications on topical issues, including Thomas Prendergast and Barbara Kline's Rewriting Chaucer: Culture, Authority, and the Idea of the Authentic Text, 1400-1602 (Ohio State U. Press, 1999) and Elizabeth Bryan's Collaborative Meaning in Medieval Scribal Culture: The Otho Lazamon [sic] (U. of Michigan Press, 1999).

(2) As I will discuss later in this essay, the vexing issue of conjectural emendation in the mid-1970s seems to me to have been the crucial catalyst for wholesale reevaluation of editorial method, but even before the wave of discussions such as those listed above, we can recognize impulses to collect and evaluate the English editorial past: Hans Aarsleff's The Study of Language in England, 1780-1860 (Princeton U. Press, 1967) is still enormously valuable internally as well as being a milestone of twentieth-century retrospective bibliography. Medieval Manuscripts and Textual Criticism, ed. Christopher Kleinhenz (U. of North Carolina Department of Romance Languages, 1976), also deserves a special place in the fabric of poststructural harbingers.

The earliest clear, graceful, and productive request for a broad survey of Middle English editing in particular is A. S. G. Edwards's "Observations on the History of Middle English Editing" in Manuscripts and Texts: Editorial Problems in Later Middle English Literature, ed. Derek Pearsall (Cambridge: D. S. Brewer, 1987), pp. 34-48. The survey of Chaucer editors in Paul Ruggiers's Editing Chaucer: The Great Tradition (Norman, OK: Pilgrim Books, 1984) is of course excellent but focuses on issues of authority and provenance that are not widely transferable beyond the orbit of Chaucer studies.

(3) Kolbing's edition of Sir Beues of Hamtoun (London: EETS, o.s. 46 [1885], 48 [1886], and 65 [1894]) is a true monument to editorial self-reflection. Frustrated by the inability of critical editing to reflect the complexities of the "Beues" versions, Kolbing several times admits frankly to his readers that he thinks it best to abandon editorial intervention: "As I have supplied the reader with the whole of the materials which I could get hold of in the course of seventeen years, the field is open for any one who takes an interest in these things, and thinks himself able to draw a perfect picture instead of this rough draught.... I do not believe that anybody will ever be able to make a thoroughly critical edition of the poem, without introducing a lot of arbitrary changes. I quite agree with Dr. Furnivall that, in a case like this, even trying to edit the work critically would be only making it far more difficult for the reader to form a judgment of his own.... I have therefore simply restricted myself to laying before the members of the Early English Text Society the whole of the materials" (xli).

At the other end of the spectrum of editorial explicitness is the version of the political satire "The Simonie" presented by Thomas W. Ross as "A Satire of Edward II's England" (Colorado College Studies 8 [1966]), which glosses over editorial difficulties from start to finish and presents a hybrid textual product. My ongoing examination of this kind of editorial documentation (from whatever period of our academic past) causes me to find less and less practical difference between the two kinds of publication: if manuscript textuality is not represented or representable in either format (i.e., neither the fully enunciated nor the invisibly collated), what difference does explicit enunciation of methodology really make?

(4) Both of these kinds of editorial "texturing," internal and external, are inescapable, and as determinants they hold cumulative force. While the most celebrated cases usually involve judgments based on extrinsic variants between multiple-copy texts, we must put under this heading anything from expurgation based on socially sensitive material (as in Bennett and Smithers's decision to replace the scatalogical passages of the "Owl and the Nightingale" with lacunae (see Early Middle English Verse and Prose [Oxford: Clarendon Press, 1966]), to the introduction of modern punctuation, ubiquitous in modern editions of Middle English even when the Middle English text faces a modern translation (see for example William Vantuono's omnibus edition of the Cotton Nero A.x. manuscript, The Poems of the Pearl Manuscript [New York: Garland, 1988]), to the discarding of nearly every kind of nonverbal manuscript feature. The general result is an inevitable alterity of syntax and meaning. We are probably most accustomed to associating wholesale manipulation to much older (pre-1860) antiquarian publications of medieval verse in particular, but modern editions display texturing impulses which are strikingly similar.

(5) Kane and E. Talbot Donaldson's dictum that any edition that stops short of a fully reconstructed original is "a poor-spirited and slothful undertaking" comes immediately to mind (Piers Plowman: The B Version [London: Athlone Press, 1975], 129). Kane's editorial function has been described, and powerfully defended, by Lee Patterson (80-93), but for balance see the nicely tart summary of positivist editing by Dan Embree and Elizabeth Urquhart in "The Simonie: The Case for a Parallel-Text Edition" (in Pearsall's Manuscripts and Texts, 49-59). A very interesting practical example of contesting approaches can be found in the Hult/Uitti debate revitalizing Joseph Bedier's struggle with Chretien's Lancelot (David F. Hult, "Lancelot's Two Steps: A Problem in Textual Criticism," Speculum 61 [1986]: 836-58; Karl D. Uitti, with Alfred Foulet, "On Editing Chretien de Troyes: Lancelot's Two Steps and Their Context," Speculum 63 [1988]: 271-92; and Hult again, "Steps Forward and Steps Backward: More on Chretien's Lancelot," Speculum 64 [1989]: 307-16).

(6) Though I use it strictly as a time-reference in this essay, the word "antiquarian" had already gained its intensely negative connotations by the middle of the eighteenth century; the diminution of antiquarian historians and editors is fairly complete by 1860 despite a tremendous amount of active and innovative work done by the diverse amateur enthusiasts which the label attempts to homogenize. It is seldom that one actually finds an enlightening specific reason, divorced from hyperbole and venom, for the pejorative against pre-professional medievalists. Since we ourselves don't yet have a guide to help us use the work of the publications of the myriad medievalist book clubs of the 1830s and 1840s such as the Bannatyne, Abbotsford, Maitland, AElfric, Percy, and Warton societies, any modern reader of antiquarian commentary grows frustrated and all too often abandons whatever line of inquiry brought her to these pre-professional resources in the first place. As a result, reams of heavily detailed, labyrinthine antiquarian essays and comments on virtually every genre and subject of historical interest gather dust on library shelves--resistant to index and, to the modern scholarly sentiment, unusable.

(7) Important older descriptions of the Auchinleck Manuscript are those of Walter Scott in the preface to Sir Tristrem (Edinburgh: Constable, 1804), Eugen Kolbing ("Vier romanzen-handschriften," Englische Studien 7 [1884]: 177-201); A.J. Bliss ("Notes on the Auchinleck Manuscript," Speculum 26 [1951]: 652-58); and I. C. Cunningham ("Notes on the Auchinleck Manuscript," Speculum 47 [1972]: 96-8). Most, but not all, of the details of these older descriptions are present in the facsimile of the manuscript, The Auchinleck Manuscript, National Library of Scotland Advocates' MS. 19.2.1, ed. Derek Pearsall and I. C. Cunningham (London: Scolar Press, 1977). Two ongoing websites devoted to the Auchinleck have appeared in the past two years: Miceal Vaughan at the U. of Washington has developed "The Auchinleck Manuscript Page," a site devoted to detailed bibliographical information on the artifact (available online at <http://faculty.washington.edu/miceal/ auchinleck/>), and David Burnley and others at the University of Sheffield and the National Library of Scotland are developing "The Auchinleck Manuscript Project," which pairs digital images and transcriptions of the manuscript's contents (available online at <http://www.shef.ac.uk/auchinleck/>).

(8) The dissertation of Judith Mordkoff ("The Making of the Auchinleck Manuscript: The Scribes at Work," [U. of Connecticut, 1980]) contains excellent evaluations of the intricacies of Loomis's textual proofs. Mordkoff and I. C. Cunningham (co-editor of the Auchinleck facsimile) continue their paleographic inventory in such work as "New Light on the Signatures in the Auchinleck Manuscript," Scriptorium 36 (1982): 280-92. For an exciting and extensive reconstruction of the manuscript's genesis, see Timothy Shonk's studies of the methodical pre-planning of the Auchinleck, "The Scribe as Editor: The Primary Scribe of Auchinleck," Manuscripta 27, no. 1 (1983): 19-20, and "A Study of the Auchinleck Manuscript: Bookmen and Bookmaking in the Early Fourteenth Century," Speculum 60 (1985): 71-91. Their work informs the more recent essay of Ralph Hanna III, "Reconsidering the Auchinleck Manuscript" (New Directions in Later Medieval Manuscript Studies: Essays from the 1998 Harvard Conference, ed. Derek Pearsall [York, 2000]: 91-102).

(9) The two articles discussed in depth in the present essay are "Chaucer and the Auchinleck MS: 'Thopas' and 'Guy of Warwick,'" in Essays and Studies in Honor of Carleton Brown (New York U. Press, 1940) and "The Auchinleck Manuscript and a Possible London Bookshop of 1330-1340," PMLA 57 (1942): 595-627. Both were reprinted in Adventures in the Middle Ages: A Memorial Collection of Essays and Studies (New York: B. Franklin, 1962): 131-49 and 150-87 respectively; all page references are to the latter printing. Two other studies by Loomis which share the methodology discussed in this essay but are not discussed here are "Chaucer and the Breton Lays of the Auchinleck MS" (SP38 [1941]: 14-33), and "The Auchinleck Roland and Vernagu and the Short Chronicle" (MLN 60 [1945]: 94-7).

(10) Loomis, "Chaucer and the Auchinleck MS," 138-39.

(11) Ibid., 148.

(12) See A. M. Trounce's articles collectively entitled' "The English Tail-Rhyme Romances" in the three premiere volumes of Medium AEvum (1932-34) for an older but compelling indictment of the way in which Chaucer has been used "as a stick with which to beat the tail-rhyme romances."

(13) Loomis, "Chaucer and the Auchinleck MS," 135.

(14) Ibid., 140 (italics original).

(15) For an overview of twentieth-century textual criticism, see the one provided by D. C. Greetham in Textual Scholarship: An Introduction (Garland, 1994), 323-35; especially relevant is the assessment of W. W. Greg (327-28).

(16) Excellent new explorations of the fluid nature of tail-rhyme romances can be found in A. N. Doane and Carol Braun Pasternack, Vox intexta: Orality and Textuality in the Middle Ages (U. of Wisconsin Press, 1991), Susanna Greer Fein, "Twelve-Line Stanza Forms in Middle English and the Date of Pearl," (Speculum 72 [1997]: 367-98), and Andrew Taylor, "Fragmentation, Corruption, and Minstrel Narration: The Question of the Middle English Romances" (The Yearbook of English Studies 22 [1992]: 38-62).

(17) F.N. Robinson, ed., The Complete Works of Geoffrey Chaucer (Boston: Houghton Mifflin, 1933);Julius Zupitza, ed. Guy of Warwick (London: EETS, 1888). Loomis's information on the Auchinleck scribes comes entirely from the "old but outstanding" description by Eugen Kolbing, as well as supplementary information from editors such as Ewald Zettl (The Anonymous Short Metrical Chronicle [London: EETS, 1935]) and Karl Brunner (The Seven Sages of Rome [Southern Version] [London: EETS, 1933]).

(18) Robinson, 1004.

(19) Following Kolbing, Loomis's count of the scribes is in error: there are six hands evident in the manuscript. (The error is corrected in Bliss's "Notes," 652-53.) The miscount does not directly influence her arguments.

(20) Loomis, "Bookshop," 163-64 (italics original). Loomis's remarks on the artistic quality of the Auchinleck miniatures is in "Chaucer and the Auchinleck MS," 133: when she comments on the poor quality of the surviving miniatures, for example (132), her notes show that the observation is based on the lithograph facsimile (made from an 1838 hand-drawn copy of an Auchinleck page) prefixed to William Turnbull's edition of' Beves of Hamton for the Maitland Club in Glasgow. It can also be noted that in her 1942 Bookshop article (discussed hereafter) she refers to the illuminations in Cotton Nero A.x. by way of Israel Gollancz's monochrome facsimile for the EETS in 1923.

(21) Ibid., 181-82.

(22) See R. Howard Bloch, "New Philology and Old French," Speculum 65 (1990): 38-58, for a relevant discussion of European and American anxiety during this period and its relationship to the production of medieval scholarship.

(23) William B. D. D. Turnbull, The Romances of Rouland and Vernagu, and Otuel. From the Auchinleck Manuscript (Edinburgh: Abbotsford Club, 1836); Sidney J. H. Herrtage, The Taill of Rauf Coilyear with the Fragments of Roland and Vernagu and Otuel (London: EETS, 1882); Allen D. Grant, "An Edition of the Middle English Romances Roland and Vernagu, The Sege off Melayne, and Duke Rowlande and Sir Ottuell of Spayne" (Diss. U. of London, 1968).

(24) Hamilton Smyser, "Charlemagne Legends," in J. Burke Severs, ed., A Manual of the Writings in Middle English, Vol. 1 (New Haven: Connecticut Academy of Arts and Sciences, 1967). Smyser's classification system, and therefore the cyclic poem theory, is invoked without comment in Alan Lupack's TEAMS edition of Three Middle English Charlemagne Romances (Kalamazoo: Medieval Institute Publications, 1990), and Stephen H. A. Shepherd mentions it in passing in his announcement of the discovery of the fragmentary Huntington copy of a Middle English Pseudo-Turpin Chronicle ("The Middle English Pseudo-Turpin Chronicle," Medium AEvum 65.1 [1996]: 19-34; see esp. pages 21-22). Shepherd s imminent edition of this copy may well revolutionize our understanding of the whole set of romances.

(25) Gaston Paris, Histoire poetique de Charlemagne (Paris: A. Franck, 1865), 156. I have silently translated Paris's French in these brief phrases.

(26) In addition to those discussed in this essay, affirmative citations of or allusions to Paris's hypothesis can be found in most of the standard reference sources from the first half of the century. For example: B. A. K. Ten Brink, Geschichte der Englischen Literatur (London: G. Bell, 1883), 1:245; Leon Gautier, Bibliographie des chansons de geste (Paris: H. Welter, 1897), 155; Anna Hunt Billings, Guide to the Middle English Metrical Romances (New York: Holt, 1901), 49-84; William Schofield, English Literature from the Norman Conquest to Chaucer (London: Macmillan, 1931), 153-55; and A. C. Baugh, A Literary History of England (New York: Appleton-Century-Crofts, 1948), 187.

(27) Mary Isabelle O'Sullivan, ed., Firumbras and Otuel and Roland (Oxford U. Press, EETS o.s. 198, 1935). It is difficult to see how O'Sullivan arrives at her conclusion. She distills her conviction into three points: first, the common stanza form of the two poems; second, their combined use of the Pseudo-Turpin Chronicle, the legendary account of the voyage to Jerusalem, and the French chansons de geste of Otinel and Roland; and third, various specific verbal similarities (lxiv-lxvii).

The first two of these groups do not, of course, constitute evidence of direct textual descent, though they can help express similarities between texts. Both "Otuel and Roland" and "Roland and Vernagu" are in tail-rhyme stanzas, but the types of rhymes and meters used are different enough to preclude the possibility of an observable common origin for them as texts. "Otuel and Roland" generally rhymes aabaabccbddb for the duration of lines 1-1697, with generally four-stress couplet lines and three-stress tail lines. At line 1698 the rhyme shifts to the less demanding pattern aabccbddbeeb, and the stresses in the couplet lines lose any kind of systematization.

This shift in rhyme takes place at the point where the Otuel story ends and the Pseudo-Turpin-based material begins, and is in part what led O'Sullivan to her conviction that the Otuel section was "foreign" to the rest of the work. According to the "Charlemagne and Roland" theory, the rhyme and meter of this second half of"Otuel and Roland" should match up with the rhyme and meter of "Roland and Vernagu"--this would show that they are indeed estranged halves of a single work. "Roland and Vernagu," however, does not compare even with the simple analysis provided above: its first 424 lines rhyme aabccbddbeeb with very regular four-stress couplets and three-stress tails; the remainder (lines 425-879) are in the same rhyme scheme but in universal three-stress lines. The only technical similarity is that of the first half of"Roland and Vernagu" to occasional sections of the second half of "Otuel and Roland."

For further observations on the volume, see the reviews by G. V. Smithers (RES 14 [1938]: 461-66) and G. L. Brook (Medium AEvum 8 [1939]: 67-71).

(28) The central works on the "Charlemagne and Roland" legend include four by Ronald N. Walpole: "Charlemagne and Roland: A Study of the Source of Two Middle English Metrical Romances, Roland and Vernagu and Otuel and Roland' (U. of California Publications in Modern Philology 21.6 [1944]: 385-451); "The Source MS of "Charlemagne and Roland and the Auchinleck Bookshop" (MLN 60 [1945]: 22-26); "'Syr Bertram the Baner' in the Middle English Romance Otuel and Roland" (MLN62 [1947]: 179-84); and "Stanzas 26 and 27 of the Middle English Romance Roland and Vernagu" (Medium AEvum 20 [1951]: 40-47). Walpole briefly reinforces his findings in his much later edition of an Old French text of the Pseudo-Turpin: The Old French Johannes Translation of the Pseudo-Turpin Chronicle (U. of California Press, 1976). Hamilton Smyser's contributions include "Charlemagne and Roland and the Auchinleck MS." (Speculum 21 [1946]: 275-88), and the "Charlemagne Legends" entry in J. Burke Severs's Manual of the Writings in Middle English (New Haven: Connecticut Academy of Arts and Sciences, 1967), 80-100.

Of additional interest are two brief publications by Laura Loomis which enthusiastically support Walpole's findings: her review of Walpole's 1944 monograph in Modern Language Quarterly 6 (1945): 349-50, and her response to his views on "Roland and Vernagu" in "The Auchinleck Roland and Vernagu and the Short Chronicle" (MLN60 (1945): 94-97).

(29) The reconstructed history presented in this and the following paragraph is condensed from Walpole, "Charlemagne and Roland," 396-400; it utilizes the classification of Turpin manuscripts first made by Cyril Meredith-Jones in his Historia Karoli Magni et Rotholandi, ou Chronique du Pseudo-Turpin (Paris: Droz, 1936). Walpole published an edition of the Johannes Turpin in 1976, in the supplement to which he thoroughly describes 32 manuscripts of the Johannes text; Add. 40142 is there referred to as "L3" (164-82). He was aware of only 18 Johannes manuscripts at the time of the 1944 "Charlemagne and Roland" monograph (398), but his discovery of 14 more in subsequent years does not seem to have affected his conviction that Add. 40142 is the source of "Roland and Vernagu"--see his review of de Mandach's Naissance et developpement de la chanson de geste in MLR 60 (1965): 617-18, and Walpole, The Old French Johannes Translation, Supplement, 169-70.

(30) Walpole greatly expanded this research in later years (see The Old French Johannes Translation of the Pseudo-Turpin Chronicle [U. of California Press, 1976]; "La Traduction du Pseudo-Turpin du manuscrit Vatican Regina 624: a propos d'un livre recent," Romania 99 [1978]: 133-42; An Anonymous Old French Translation of the Pseudo Turpin Chronicle [Cambridge, MA: Medieval Academy of America, 1979]; and Le Turpin francais, dit le Turpin I [U. of Toronto Press, 1985]).

Wapole's careful foundational work on thirteenth-century prose historiography has enjoyed much expansion lately: the Johannes Turpin in particular has been the subject of exciting work by Pierre Botineau ("L'histoire de France en francais de Charlemagne a Philippe Auguste: la compilation du manuscrit 624 du fonds de la Reine a la Bibliotheque Vaticane," Romania 90 [1969], 79-99), Gilette Labory ("Essai d'une histoire nationale au XII[I.sup.e] siecle: la chronique de l'anonyme de Chantilly-Vatican," Bibliotheque de l'ecole des chartes 148 [1990], 30154), and Gabrielle Spiegel (The Chronicle Tradition of Saint-Denis: A Survey [Brookline, MA: Classical Folia Editions, 1978]; "Medieval Canon Formation and the Rise of Royal Historiography in Old French Prose," MLN 108 (1993): 638-58; and Romancing the Past: The Rise of Vernacular Prose Historiography in Thirteenth-Century France [U. of California Press, 1993], esp. 55-97), all of which pays explicit homage to Walpole's legacy.

(31) Smyser, "Charlemagne and Roland," 275-79 and 281-88.

(32) Walpole, "Charlemagne and Roland," 396-409, esp. 408-9.

(33) Walpole does offer his readers a large family tree of the manuscripts of the Johannes Turpin in the "Charlemagne and Roland" essay (between pages 400 and 401), prefiguring the work later in his career that would be of such value to specialists in Old French historiography. But the family tree stops short of actually providing a placement of the Middle English poems described in the article itself--a curious circumstance, since the source manuscript he has identified (BL Add. 40142) is positioned front and center in the graphic.

(34) Smyser, "Charlemagne and Roland," 281.

(35) Andre de Mandach, Naissance et developpement de la chanson de geste en Europe (Paris: Droz, 1961), 121. In keeping with the caveats offered by most scholars who reference Professor de Mandach, even in passing, I will add my own very strong reservations on the reliability of his work. A good number of his assertions, large ones as well as small ones, are so capricious and overstated as to be downright wrong. He has a chronic impulse to name or rename practically everything he comes across--scribes, or parts of manuscripts, or manuscript copies, or hypothetical manuscript copies, or whole families of hypothetical manuscript copies. This in itself is simple eccentricity, but de Mandach compounds the challenges for his readers by subsequently endowing his own coined referents with a very vigorous verisimilitude; his rhetorical claims are often so declamatory and absolute that the reader is given no sense of their tentative or ad hoc nature.

(36) Smyser, "Charlemagne Legends," 81.

(37) Cerquiglini, Eloge de la variante.

(38) On Shepherd see n. 24 above, as well as his essays "'I Haue Gone for Thi Sak Wonderfull Wais': The Middle English Fragment of The Song of Roland" (Olifant 11 (1986): 219-36); "The Ashmole Sir Ferumbras: Translation in Holograph" (in The Medieval Translator, ed. Roger Ellis [Woodbridge: D. S. Brewer, 1989], 10321); and "'This Grete Journee': The Sege off Melayne" (in Romance in Medieval England, ed. Maldwyn Mills, Jennifer Fellows, and Carol M. Meale [Cambridge: D. S. Brewer, 1991], 113-31.

Spiegel has nearly single-handedly re-defined the course of scholarship on the Pseudo-Turpin Chronicle. In addition to the works cited above in n. 30, see "History, Historicism, and the Social Logic of the Text in the Middle Ages," Speculum 65 (1985): 59-86; and The Past as Text: The Theory and Practice of Medieval Historiography (Johns Hopkins U. Press, 1997).
COPYRIGHT 2001 University of Iowa
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Porcheddu, Fred
Publication:Philological Quarterly
Date:Sep 22, 2001
Words:15093
Previous Article:Ovid's medieval metamorphosis: techniques of persuasion in Chretien de Troyes' Philomena.
Next Article:The ruined abbey in the eighteenth century.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters