Traugott Lawler’s latest contribution to the Dumbarton Oaks Medieval Library series is even more welcome than his previous one. His superb edition and translation of John of Garland’s Parisiana poetria (DOML 65; 2020) updated and improved the already excellent version that he had previously published (Yale University Press, 1974). With the publication of this equally superb edition and translation of Gervase of Melkley's Ars versificatoria, we need no longer depend on the flawed edition by Hans-Jürgen Gräbener (1965) or its English translation by Catherine Yodice Giles (1973). Lawler has made Gervase’s idiosyncratic theory and often obscure writing more comprehensible than I had imagined possible, without minimizing the challenges they present to specialist and non-specialist readers alike.
Like Matthew of Vendôme’s Ars versificatoria (ca. 1165); Geoffrey of Vinsauf’s Summa de coloribus rhetoricis, Documentum de modo et arte dictandi et versificandi, and Poetria nova (late twelfth to early thirteenth century); John of Garland’s Parisiana poetria (ca. 1220; revised 1231-1235); Eberhard the German’s Laborintus (before 1280); and the anonymous Tria sunt (late fourteenth century), Gervase of Melkley’s Ars versificatoria (completed 1220: Lawler, viii) provides instruction in composing Latin texts for students who have already mastered basic Latin grammar. Lawler perceptively observes that Gervase’s specific aim was to teach his students how to compose verse couplets, the form taken by most of his examples (ix). However, there is little in Gervase’s treatise that could not be applied equally well to artful prose. Gervase acknowledges as much by devoting the final two parts to matters that are exclusively relevant to poetry (Part 7) and to prose (Part 8). Like the other works of its genre, Gervase’s “art of making verses” can also be described as an “art of poetry and prose.”
Also like both those earlier and later arts of poetry and prose, Gervase’s instruction chiefly concerns elements of style, above all figurative language. Here the similarities end, as Gervase creates a unique schema for classifying the figures—one that radically departs from the more familiar template ultimately derived from Book 4 of the pseudo-Ciceronian Rhetorica ad Herennium. Blending the grammatical with the rhetorical tradition (citing Donatus more frequently than “Cicero”), Gervase groups all the figures into three large categories that Lawler translates as “sameness” (identitas), “likeness” (similitudo), and “opposition” (contrarietas). These are the subjects of the first three parts, which constitute 236 of the treatise’s 301 pages (counting both Latin text and facing English translation). The very brief Part 4 treats proverbs, subtypes of which fit into each of the three larger categories; Part 5 covers assorted topics associated with “purity of style” (munditia); and Part 6 briefly discusses rhetorical proof based on the attributes of persons and actions.
“Sameness,” the most capacious of Gervase’s three master categories, is also the most idiosyncratic and the one that has proved most difficult for modern scholars to explain. Lawler provides as clear an analysis as one could ask for, but admits that not all the pieces fit together (ix-xi; 352, note to 1.70). Seemingly, figures that do not belong in either of the better-defined categories of “likeness” and “opposition” are assigned to “sameness” by default. Because Gervase puts familiar figures in unfamiliar places, users of Lawler’s edition and translation will be especially grateful for the “Index of Literary and Rhetorical Terms” at the end of the volume (393-96).
Due perhaps to the striking originality, even eccentricity of its approach, Gervase’s treatise seems, at least initially, to have had little impact beyond his own classroom. Only one thirteenth-century copy survives, in a manuscript—Glasgow, Hunterian Museum, MS V.8.14 (formerly 511)—that some believe Gervase himself may have had a role in producing (308). Aside from that copy, which is the base text (H) for both Gräbener’s and Lawler’s editions, the only evidence for this treatise’s influence comes nearly two centuries later: in three additional copies (ABD), closely related to one another and produced at Oxford between the late fourteenth and the mid fifteenth centuries, as well as in several admiring citations of Gervase in the late fourteenth-century Oxford textbook Tria sunt. A more ambiguous example of Gervase’s possible influence is his observation that one may substitute the imperfect indicative for the subjunctive (114/115: 1.77), a practice employed multiple times, but without any reference to Gervase’s authority, in the anonymous early fifteenth-century Oxford treatise on letter writing, Regina sedens Rethorica.
If Gervase’s Ars versificatoria finds a wider readership now, it will be thanks to Lawler’s masterful translation and the extremely helpful notes that accompany it. At points where an accurate translation alone does not suffice to make Gervase’s meaning clear, Lawler is usually able to provide a convincing or at least a plausible explanation; and where a passage has defied his best efforts to explain it, he does not hesitate to acknowledge that (e.g., 350, note to 1.55; 352, note to 1.70; 365, notes to 2.32 and 2.33). In most such cases, I have no better explanation to offer; but occasionally I would be willing to venture a suggestion or at least pose a question. When Gervase says that the figure emphasis “is the perfect thing to teach boys” (107: 1.72), for example, surely it is not “anybody's guess” why he should think so (354, note to 1.71). As Gervase goes on to demonstrate in detail, with two different examples, this figure provides an easily understood method for expanding brief, unadorned statements into longer, more ornamented equivalents, such as the couplets that he would have asked the boys who were his students to compose. Likewise, it is not clear why “Byrrhia” is said to be “Perhaps an error for Pyrrhia” in one place (346, note to 1.41) but is elsewhere identified as the name of a character from Terence’sAndria (361, note to 2.19). I assume that Lawler is poking fun at one of Gervase’s examples of onomatopoeia when he explains: “as for Tartarus, if you trill the r’s it sounds like hell” (360, note to 2.15).
In two other instances, answering questions about what Gervase means and how best to translate his Latin text takes us back to the manuscript witnesses. Since Lawler has produced “essentially an edition of H” (311), he chooses the readings of that manuscript whenever possible. Thus, in the extended metaphor comparing sex to scholarship, he prefers tripusque (H), even though the alternative tempusque (ABD; also, Tria sunt 4.21) is better for the scansion of the hexameter in question: incaustum tripusque / tempusque nigrum, nocturna placere (186: 2.39) and the contrast between “white ink” (album incaustum) and the “dark time” (tempus nigrum) mirrors the subsequent contrast between “night writing” (nocturna...scriptura) and “daylight” (lucis). If a scribe misread an abbreviation for tri- (t‘) as an abbreviation for tem- (t–), rather than vice versa, and the authentic reading is “black stool” (Lawler's translation of tripus...nigrum), then the tenor of this metaphor is likely the male genitalia, as Lawler perhaps implies in his note to the passage (366, note to 2.39). That metaphorical meaning might be clearer to a modern reader if the translation were “dark three-legged stool,” especially if that reader were familiar with Matthew of Vendôme’s comparison of the same three parts to the three syllables—one long and two short—of a dactylic foot (Ars versificatoria 1.53.79–80).
The manuscript witnesses offer only problematic options in a later passage, where Gervase seeks a metaphorical equivalent for “a four-syllable word” (tetrasyllaba dictione; 294/295: 8.3). The reading in H, which I checked on microfilm, is Gräbener’s quadratam cubicam, suggesting that the feminine noun (dictionem?) that these two adjectives modify is missing or implied: “a ‘quadratic,’ ‘cubical’ [word].” (Giles skirts the problem in her translation: “a ‘four-sided cube.’”) Lawler adopts the reading in the other witnesses: quadratam cubitam, which supplies a noun (cubitum) while inexplicably changing its gender. He translates the metaphor as “a squared-off elbow,” which he calls “A desperate guess!” (384, note to 8.3). My less-desperate alternative—“a quadripartite cubit”—takes cubitum in another of its common meanings, as a unit of measure that normally has six parts (“palms”) rather than four.
My other suggestions for improving the translation are much simpler. Two of them concern Part 3.8 (227): “terms opposite to excess” must be a mistake for “opposite terms to excess”; and it’s not clear whymente is translated “spirit” in the example but “mind” in the comment on it. Uniformity is more essential in an example of the figure climax (240/241: 3.11): to translate the verb solatur as “brings comfort” rather than “brings solace” is to obscure its link with the noun solacia (translated “solace”) in the next clause, the feature that defines this figure. Other than these few instances, I find little to criticize and much to praise in Lawler’s translations, which are consistently clear and perceptive and often witty (e.g., the inspired anachronism “teaching assistant” for discipulo scholari; 184/185: 2.39).
Finally, the exceptional accuracy of the text deserves special commendation. Of the half-dozen minor errors I spotted, only three are likely to cause even momentary confusion: “ills” for “illa” in the Latin text (214: 3.2); “ortus, plural (risings, settings)” for “ortus, occasus, plural (risings, settings)” in the English translation (263: 5.7); and “complement” for “conclusive” in the index (393, 394, 395). Traugott Lawler has provided all the tools necessary for studying Gervase of Melkley’s fascinating but heretofore somewhat neglected treatise: an expertly edited Latin text; a faithful yet readable translation; and up-to-date, thorough, and illuminating commentary. Anyone interested in medieval rhetoric and poetics will want to own this book, which has the added merit of a price that should fit any budget.