way into the main text, adding another layer of textual difficulty to Walter’s work. Overall, I suggest that the De nugis curialium as we have it is best understood as five separate works in various stages of completion that have been bound together, almost certainly after Walter’s death. Seen in this light, it is clear that Walter does not deserve his reputation as a scatterbrained author. It is hardly his fault that the only surviving copy of his work has been taken as the definitive testament of his literary talents. Not only does this reevaluation render Walter’s presence as an auctoritas for the Lancelot-Grail Cycle less incongruous—he had the patience and focus needed for such a work—it also bears directly on his reputation as a writer who worked in the Matter of Britain. As Chapter 4 shows, understanding Walter’s practice of revision sheds new light on one method medieval authors used to write stories set in ancient Britain.
Evidence of Revision in the De nugis curialium
James Hinton was the first scholar to examine the structure and plan of the De nugis curialium in depth.16 Although he recognized that Walter’s text survives in an unedited state, he proceeded to reconstruct the text in the order in which he believed it had been composed. Identifying as many termini a quo and termini ad quem as possible, Hinton distinguished twenty separate “fragments,” which he thought gave little evidence of a larger design: “From what has been noticed of the casual manner in which Map wanders from one topic to another even while he is writing straight ahead, it is clear that he was not restrained by a definite plan; he wrote willingly upon whatever occurred to his mind, careless of the drift of his discourse.”17 This is the Walter Map familiar to scholarship. Indeed, I will concede that dividing the De nugis up into small pieces and ordering them on the basis of chronology makes Walter’s text even less coherent, but it must also be admitted that splitting up almost any literary work into the chronological order of its composition would result in disorganization, too. The Canterbury Tales would certainly look the poorer for it. And everyone, it seems, has followed Hinton in claiming that Walter all but announces his lackadaisical style of composition when he writes, “Hunc in curia regis Henrici libellum raptim annotaui scedulis” (I have swiftly [raptim] noted this little book down in pages of parchment in the court of King Henry).18 Hinton takes raptim with its etymological force “by snatches,” which lends credence to the belief that Walter’s literary activity occurred at intermittent stages. However, this meaning is not attested in Medieval Latin (nor in Classical Latin for that matter).19 Instead, it is best to take raptim here with its normal meaning of “swiftly” or “hurriedly”—a subtle, but important, distinction. Although Hinton’s contributions remain valuable, especially his observations that the De nugis curialium is an unfinished work and that the chapter titles are the product of a later scribe or compiler, this chronological arrangement is unnecessarily complicated and relies on a rather constrained view of literary composition.20
Brooke and Mynors, Walter’s most recent editors, accept many of Hinton’s arguments regarding dating. However, they propose that the De nugis curialium has more structure than Hinton allows. Instead of a series of fragments thrown together by Walter or a later scribe with little attempt at order, they suggest that the work “was composed more or less as a single book, into which additions small and large were later inserted.”21 They show that apart from eight interpolations, the work belongs mostly to the early 1180s. In their view, the manuscript’s current disarray results overwhelmingly from Walter’s subsequent tinkering and erratic insertions.
The bulk of it was drafted in 1181 and 1182, and it lay for a number of years in loose quires, roughly arranged in the order dist. iv, v, i, ii, iii. It was still a draft, not a finished work, and included two versions of the satire on the court; some chapters were never completed. From time to time the author added insertions small and large on slips of vellum; in 1183 he provided the whole work with a prologue. At some date unknown, he decided to make the satire on the court the opening of the book, and so cut his loose quires like a pack of cards, arranging the material in approximately its present order.22
This explanation has the apparent benefit of originally placing the two versions of the satire on the court in succession, with the more polished version immediately following the earlier draft version (though exactly why this is preferable is left unexplained). Additionally, in the original order proposed by Brooke and Mynors the book begins with the Dissuasio Valerii, Walter’s most popular work, which alone of the contents of the De nugis curialium circulated widely. Since it first circulated under a pseudonym, it would have been a good marketing ploy to open a work of some considerable size with the surprising revelation that Walter himself had composed the popular Dissuasio Valerii. But, as Brooke and Mynors admit, this account does not solve all the infelicities of the De nugis curialium. Here, Walter comes in for yet more criticism. Since they argue that the manuscript’s current form results more or less from Walter’s own meddling, the blame for all the faults of the De nugis curialium lies squarely on his shoulders. Brooke and Mynors, for example, are more inclined to believe the subpar rubricated chapter headings are Walter’s own invention: they are “untidy in every possible way, and with an untidiness which clearly reflects in part the mind of Master Walter.”23 And what of the internal epilogue, which even in their reconstructed form still sits oddly in the middle of the work? They suggest that “since it was evidently written on a loose slip or bifolium, it is possible that Map, finding his prologue unhappily sandwiched in the middle of the book, with gay abandon attached the epilogue to it.”24 Just to be clear, what Brooke and Mynors propose is that Walter wrote a coherent book, cut it in half so that it began with the satire on the court, neglected to discard his first draft of said satire on the court, perhaps placed an epilogue in the now middle of the work because that is where his prologue lay, and afterward inserted a few stories here and there. This scenario, as they readily admit, is conjectural. Nonetheless, in my opinion it relies too heavily on the supposition that Walter Map is a flighty writer, unable or unwilling to write an orderly narrative—only thus could an author demonstrate such carelessness with his text. However, the only evidence for Walter’s mental “untidiness” is itself the manuscript of the De nugis curialium. This is a significant problem.
There is, however, a way around the tautological explanation that the De nugis curialium is disorganized because Walter Map is disorganized, a fact that is in turn proven by the disorganization of the De nugis curialium. In an astute review, A. G. Rigg suggests that previous editors and scholars have confused “the order of composition with the final intended arrangement, as though only scribal incompetence could account for a nonchronological order.”25 In other words, writers do not work in a strict chronological fashion, starting a work with page one and completing it neatly with the final sentence; this scenario neglects the messy business of drafts, omissions and additions, and innumerable starts and restarts that are familiar to any writer. Walter is no exception. Indeed, Rigg seems to have been the first to grasp the importance that Walter was rearranging and revising previous material, and, as any good reviser will do, he moved sections about, cut some passages, and expanded others, while retaining some phrases verbatim.26 Either Walter never finished this process of revision or the only surviving text reflects an earlier state of affairs. We should therefore view the doublets present in the De nugis not as two versions of the same tale, nor as the handiwork of a particularly inept scribe, but as Walter’s earlier and later revisions of the same episode. Rigg claims that distinctiones 1–3 “are in nearly final state,” while distinctiones 4 and 5 consist largely of outdated drafts and material that Walter either had not yet reworked or had not yet decided where to place.27 While in the middle of reworking the De nugis, Rigg supposes that Walter “took the whole pile of material with him to Oxford in 1197, where it lay until a fourteenth-century editor copied it all out.”28 Walter’s only surviving work lies frozen in the midst of revision.
While I do not agree with all of Rigg’s brief suggestions—I remain unconvinced that Walter was writing a single, unified work and it is certain that the copyist of Bodley 851 was not working from an authorial copy—they do provide a valuable point of departure for a new investigation