The Ars versificaria of Gervase of Melkley Structure, Hierarchy, Borrowings

Gervase of Melkley, a younger contemporary of Geoffrey of Vinsauf, writes his De arte versificatoria et modo dictandi at the peak of a revisionary movement that places the discussions of figures and tropes inherited from classical and medieval grammatical and rhetorical traditions in new contexts, creating what we now call the Arts of Poetry and Prose. Gervase’s art draws upon the works of Matthew of Vendome, Geoffrey of Vinsauf, and Bernardus Silvestris for its doctrine and its examples. But how often does Gervase refer to these writers? How does he use their arts in his art? When does he borrow from them? What doctrine and which examples does he borrow? Does he cite his references and, if so, what are his citation practices? This chapter surveys Gervase’s borrowings from the works of Matthew, Geoffrey, and Bernardus by way of a review of the Index nominum and Index scriptorium of Hans Jurgen Graebener’s modern edition of the De arte versificatoria. The review locates Gervase’s borrowings of doctrine and examples with greater precision, and corrects errors in the indices as needed. Charting the precise citation practices of Gervase clarifies the meaning of his hierarchy of the three writers, places his long supposed use of the Poetria nova in serious doubt, and reopens the question of his art’s date. Summary


Introduction
In the early 13th century, the Englishman Gervase of Melkley composed a lengthy art of poetry and prose known as the Ars versificaria 1 -written in prose with examples in verse and prose, organising rules both common and specific to verse and prose -to answer the request of one Johannes Albus for an art to instruct the rudes in polished speech by way of the rhetorical colours and arguments. 2atherine Yodice Giles, in the preface to her translation of the art, affirms that Gervase's "organization is neither arbitrary nor idiosyncratic", and suggests Gervase "was trying to treat the stylistic devices philosophically". 7illiam Purcell, the latest to investigate the structure of Gervase's art, recognises that the three loci of identitas, similitude, and contrarietas "amount to three progressively sophisticated levels of expression and meaning". 8nfortunately the systematic, philosophical and progressively sophisticated whole of Gervase's taxonomy of figures and tropes is very large and difficult to comprehend, and from Faral to Kelly to Purcell, the investigator's frustration with the task is clear.If Faral finds the originality of the art to lie in its interesting system, he nonetheless tires of describing the system before he finishes detailing its first component. 9If Kelly recognises Gervase's art as "the only really deliberate innovation in the classification of tropes and figures", he considers this innovative classification at only its highest level, identitas, similitudo and contrarietas. 10And, if Purcell makes some progress addressing the sophistication of Gervase's similitudo, his primary focus on transumptio leads him to gloss identitas and to misplace some tropes. 11Thus Faral, Kelly, and Purcell all recognise that the interest of Gervase's art lies in the structural whole that organises it, but all dismiss the Ars versificaria before exploring its complexity fully.
We need some means to grasp the whole of Gervase's art and discover its essential structure without losing ourselves in the details.Without a grasp of the whole art, the arrangement of the Ars versificaria cannot be kritische Ausgabe von H.-J. Gräbener.Münster, 1965, XXIX-CX.See also Gräbener's overview of the structure, Gervais von Melkley, Ars poetica, 286-7.9 Faral, Les arts poétiques, 329.Faral does investigate sections of the art further in "Le Manuscrit 511 du Hunterian Museum de Glasgow".Studi medievali, 9, 1936, 62-107, but he refers the reader (63) to his earlier discussion of its structure in Les arts poétiques, downplays the importance of Gervase's system given the developments of modern linguistics, and suggests Gervase was, in any case, more interested in practice than in theory (93-4).
1.This task is beyond the powers of this writer to accomplish, and thus I fear the criticism of the learned; and 2. since I have little hope of successfully completing this task, I am open to the charge of indiscretion for even trying, but 3. I will do the task nonetheless to fulfil the request of my friend. 16rvase's reasoning is nearly identical: 1. To write an art is so great a work, so difficult an undertaking, already done by authorities like Matthew, Geoffrey and Bernardus, that 2. it would have been more prudent for me to remain silent than to pursue or to have rashly promised fulfillment of the task, but 3. faith and obedience to the request of my friend overrule my fear and discretion.
In fact, Giles correctly recognises this "protestation of inadequacy" as "an established convention", while Kelly takes it to mean "Gervase states that his treatise is more elementary than the three he extols by Bernardus, Geoffrey, and Matthew".17 The primary source of Kelly's elementary level is most likely Gervase's claim that his art is for the rudes, whom Kelly calls "as yet unformed pupils". 18But rudes has a precise meaning in Gervase's art: the rudes appear to be grammatically informed students who lack rhetorical polish. 19hen Gervase writes that elementares pueri are to be directed to the four distinct parts of diversio, the elementary pueri would seem to be a subset of all the rudes. 20And when Gervase states that the figure emphasis is best for instructing pueris, the third person reference implies he addresses his art to teachers as well as students. 21In any case, Gervase assumes the rudes have mastered the grammatical prohibitions and permissions found in Donatus.The Ars versificaria teaches grammar by way of Priscian's precepts, and rhetoric by way of Cicero's counsels. 22ervase's direction to the dialecticians -"we send the dialecticians to those mentioned above" -might mean that both the dialecticians and the works of Matthew, Geoffrey and Bernardus are "more advanced" than the rudes and the Ars versificaria Gervase writes for them.23 But here we must be careful not to confuse a level of teaching -elementary, intermediate, and advanced -with the level of the subject taught -grammar, rhetoric, and dialectic.Grammar is the first of the liberal arts, but grammar is taught at a variety of levels, from elementary to advanced.When Gervase advises the dialecticians "that they not despise either the Barbarismus of Donatus, the Ars Poetica of Horace, or the Rhetorics of Cicero", he reminds them of the importance of all levels of language study.24 The attitude of Gervase toward the dialecticians is complex; he respects their place in the trivium, but he suggests that they respect grammar, poetics and rhetoric as well.25 In the Ars versificaria, the place of dialectic in the trivium appears to be, in accordance with the traditional hierarchy of Martianus Capella, between grammar and rhetoric.26 For instance, in his discussion of determinatio, Gervase associates grammar with correct phrasings (congrua) and dialec-21 Gervais von Melkley, Ars poetica, 59, 71.Although Woods follows Giles' reading of the rudes as "those who are in the early stages of language study", Giles, Gervais, IX; Woods, Classroom, 51 fn. 4, she also notes that a writer's reference to the rudes in the third person is a sign the commentator is writing for other teachers, Woods, Classroom, 51.

"Consilium tamen est ne contempnant vel Barbarismum Donati vel Poetriam Oracii vel
Rethoricas Ciceronis", Gervais von Melkley, Ars poetica, 2; Giles, Gervais, 2. Kelly reads the sentence as advice to the rudes rather than to the dialecticians, Kelly, The Arts, 62. Giles reads the sentence as advice to the dialecticians and calls them "more sophisticated students", Giles, Gervais, XXXIV.R. Copeland and I. Sluiter, in their selection from the Ars versificaria in Medieval Grammar & Rhetoric.Language Arts and Literary Theory, AD 300-1475.Oxford, 2009, 609, translate dialecticos as "advanced students", but without comment.
25 Notwithstanding his reading of Gervase's directive as advice to the rudes, Kelly gets Gervase's point exactly right: the dialecticians should "know traditional grammar and rhetoric despite their presumed eagerness to learn logic", Kelly, The Arts, 63.
26 See Fredborg, K.M. "The Unity of the Trivium".Sprachtheorien in Spätantike und Mittelalter, 3 Bd., Geschichte der Sprachtheorie.Hrsg.von S. Ebbesen.Tübingen, 1995, 325-38  (328-30), for Capella's order, and for the rise of dialectic to and fall of rhetoric from the top position of the trivium in the twelfth and thirteenth centuries.tic with true phrasings (vera); he then shows how dialectical truth grounds the appropriate and ornate phrasings (competens et ornata) of rhetoric. 27rammar can explore rhetorically useless areas: certain peculiarities of epithet "may be left to the grammarians". 28But some of the intricacies of dialectic are rhetorically useless, too.The transumptio or transformation of conjunctions, prepositions, and interjections belongs to the "subtlety of dialectic rather than to the usefulness or beauty of rhetoric". 29Moreover, it is clear from some of his phrasings that Gervase does not consider himself to be a grammarian or a logician.Some forms of apposition are "inappropriate for the grammarian, more inappropriate for us". 30Some of the details pertaining to the consignification of verbs belong "to the logicians rather than to us". 31 It appears Gervase considers himself to be a rhetorician, one who desires the equivocation that the logician disdains. 32f Gervase's humility is feigned, his audience of rudes more intermediate than elementary, and his treatise more rhetorical than dialectical or grammatical, there is little ground for Kelly to construct an elementary level below the three Gervase explicitly states.Certainly it is illogical to place the Ars versificaria at an elementary level when the treatise borrows copiously from and comments incisively upon the writings of Matthew, Geoffrey and Bernardus supposed to be more advanced than it.
Still, the hierarchy of authorities Gervase presents might provide the means to relate the whole of the Ars versificaria to the three of them in some other way.

Borrowing and Citation Practices in the Ars versificaria
This investigation surveys Gervase's borrowings from the works of Matthew, Geoffrey, and Bernardus by way of a review of Gräbener's Index  nominum and Index scriptorum. 33How often does Gervase refer to the three writers?How does he use their arts in his art?When does Gervase borrow from them?Does he borrow doctrine, examples, or explanations?How does Gervase cite his references?The review seeks to clarify Gervase's borrowings of doctrine, examples and a few explanations, and to correct errors in Gräbener's indices as needed.
The two charts collect borrowings of names and doctrine (d), examples (x), and explanations (X) from identitas (Chart 1) and from similitudo, contrarietas and the remaining sections of the Ars versificaria (Chart 2). 34nspection of the charts reveals the following three main points: Point  49 Litus arat and laterem lavat also appear in logic manuals contemporary with Geoffrey and Gervase. 50elly assumes Gervase knows the Poetria nova and relies on Gräbener's indices to support his claim. 52Glendinning also depends upon Gräbener's index scriptorum to assert "it is certain that Gervase of Melkley […] was a beneficiary of Geoffrey's Poetria nova". 53Woods recently restates the critical consensus: when Gervase arranges the hierarchy of Matthew and Geoffrey and Bernardus, he "is referring here first to Matthew of Vendome's Ars versificatoria and then to Geoffrey of Vinsauf's Poetria nova". 54ut a close review of Gräbener's indices reveals that Gervase -although he borrows doctrine from Matthew's Ars, Geoffrey's Summa and Geoffrey's Documentum -never borrows doctrine from Geoffrey's Poetria nova.Furthermore, the examples Gräbener claims Gervase borrows from the Poetria nova refer either to Geoffrey's independent occasional poetry or to literary commonplaces easily found elsewhere. 55In short, the Ars versificaria shows no sign that Gervase borrows from the Poetria nova at all. 56he investigation also reveals that Gervase is remarkably consistent in his borrowing and citation practices, especially as they relate to Matthew, Geoffrey, and Bernardus.Gervase never names these authors as sources of doctrine or of examples not their own.He only credits them for examples of their poetry: for the opening lines of Matthew's Christe, tibi sit honor, for passages from Geoffrey's Richard poems and for his epitaph for Henry II, for excerpts from Bernardus' Cosmographia and Mathematicus or Paricidali.These practices are too consistent to be unintentional.55 Camargo believes "the lament for King Richard [...] definitely circulated on its own from an early date" (Camargo, M. "Geoffrey of Vinsauf's Memorial Verses".Nottingham Medieval Studies, 56, 2012, 81-119 (98).Camargo also believes it likely that Geoffrey circulated "all of his occasional poems', Camargo, From Liber versuum, 13.
56 Which leads one to wonder why the critical consensus has remained so firm for so long.Perhaps the order of Glasgow, Hunter V.8.14 is somewhat to blame.Since the order of authors is likely to be chronological (Matthew, Geoffrey, Gervase), and the order of the works' composition arguably so (Ars versificatoria, Summa, Documentum, Poetria nova), and Gervase clearly borrows from the first three treatises, why not assume the Ars versificaria, assumed to be written after them all, borrows from the Poetria nova as well?
57 The references Gervase makes to John of Hanville's Architrenius follow a contrary, but equally consistent practice.Gervase refers to John by name twelve times, always when discussing doctrine.Gervase refers to the Architrenius by name fifteen times; in all but two instances he is discussing examples from John's poem.

The Hierarchy of Matthew, Geoffrey and Bernardus Reconsidered
That Gervase does not refer to the Poetria nova should come as a great relief to those who would like Gervase to recognise Geoffrey's treatise as a masterpiece.For if Gervase does view the Poetria nova as less than first-rate, either his judgment is faulty, or he sees a flaw in the treatise unseen by his contemporaries and successors. 58But Gervase does not deny that the Poetria nova is a masterpiece; he simply does not include it in his assessment of Geoffrey's' work.Gervase does not place Geoffrey's poetic accomplishments at the level of Bernardus' Cosmographia because he bases his evaluation upon Geoffrey's Summa, Documentum and early occasional poems.The investigation of citation practices demonstrates that the hierarchy of Matthew, Geoffrey and Bernardus is primarily a ranking of versifiers not theorists: that the three authorities "scripserunt autem hanc artem" more or less art-fully. 59Of course we first consider the hierarchy of Matthew, Geoffrey and Bernardus to be an evaluation of their treatises not their verse, even if Bernardus appears to lack a treatise to evaluate. 60But Gervase knows that rhetorical practices occur spontaneously to subtle spirits, even to those ignorant of theory, and such a spirit Bernardus might well have been. 61If so, someone else would need to provide a fitting art (de arte) to explicate Bernardus' artful practices (ex arte), and who better than Gervase, who cites Bernardus more than any other author? 62Gervase's subtle yet robust self-praise neatly balances the affected modesty of his introductory remarks.
Is it possible for us to follow Kelly's version of the hierarchy a bit further and describe Matthew's Ars versificatoria as grammatical and Geof-58 Woods notes how uncommon it is for Gervase to place Geoffrey's text below the highest level, Woods, Classroom, 48.Woods generously suggests Gervase was an exceptionally gifted teacher.
60 Gräbener reviews the literature that discusses Bernardus' supposedly missing art, Gervais von Melkley, Ars poetica, XXV-XXVII.Kelly suggests Gervase's reference to Bernardus, "in prosaico psitacus, in metrico philomena", Gervais von Melkley, Ars poetica, 1, pertains to just one work, the Cosmographia, a prosimetrum, Kelly, The Arts, 58-9.Kelly sees no reason to assume Bernardus also wrote an art, nor do I.Le poetriae del medioevo latino, 205-224 frey's Poetria nova as rhetorical?Yes, but only if we see the authorities themselves struggling to understand the distinction.Matthew, Geoffrey in his Summa and Documentum, and Gervase, all try to reconcile two disparate traditions of names, definitions, and examples for the figures and tropes.The Barbarismus of Donatus transmits a grammatical tradition that treats schemes and tropes with Greek names and Latin translations and excludes figures of thought as proper to the more advanced study of rhetoric.Rhetorica ad Herennium Book IV transmits a rhetorical tradition that treats all the colores by way of Latin names lacking explicit Greek equivalents including the figures of thought.In terms of the scope of their treatment of the figures, the Ars versificatoria, Summa, Documentum and Ars versificaria are all grammatical treatises.Only the Poetria nova is a rhetorical treatise because it alone includes the names, definitions and examples for figures of thought.
But the treatises prior to the Poetria nova are not simply grammatical.Matthew primarily draws upon the grammatical tradition, as his Greek names for, and numbers of, schemata (17) and tropes (13) make clear. 63till, Matthew is aware of the distinction between grammatical schemes and rhetorical colours, and he offers eight correspondences between the two. 64Matthew goes on to list the Latin names for twenty-nine of the thirty rhetorical colours treated by Marbod. 65In the Summa, Geoffrey borrows from Marbod to fill out Matthew's list of names.Although he does not suggest any correspondences with the grammatical figures, Geoffrey discusses twenty of Marbod's rhetorical colours. 66When he turns to tropes, though, Geoffrey attempts to identify the colour circuitio with the figure emphasis, showing he is somewhat aware of the parallel grammatical tradition. 67In the Documentum, Geoffrey repeats the identification of circuitio with emphasis, and he discusses at some length the relation between synecdoche (figure) and intellectio (colour), and metonymy (figure) and denominatio (colour). 68Geoffrey's further identification of zeugma a medio with conjunctum and zeugma a superiori/ab inferiori with adjunctum con- 68 Geoffrey, Documentum, II 3 30, II 3 35, II 3 45-46.Geoffrey's discussion of synecdoche and metonymy is especially unfortunate as he seeks to create independent tropes of their species, and he uses the different grammatical and rhetorical names to make spurious distinctions.
nects the grammatical figures Matthew discusses at the beginning of his art with the proper rhetorical colours. 69eoffrey's next step will be to leave the grammatical tradition behind and simply adopt the figures of speech and thought from the rhetorical tradition. 70In the Poetria nova, Geoffrey will treat all the figures in the exact order of Rhetorica ad Herennium Book IV, deploying the figures of speech in a poem on the fall and redemption of man, naming and defining the figures of thought, and using the figures of thought in a long poem on papal duties and other matters. 71Only the tropes will remain in a different order and place. 72

The Structure of the Ars versificaria at its Highest Levels
We know Geoffrey's next step, but Gervase does not.Clearly, Gervase has a strong dialectical and theoretical interest in organising the colores. 73The Ars versificaria is largely devoted to classifying them. 74How, then, does Gervase handle the figures and colours he finds in the Ars versificatoria, Summa, and Documentum?
The early portions of the Ars versificaria neatly blend and coherently organise the two traditions of figures and colours.Not much of the doctrine is new, but the arrangement of the material deserves the attention of scholars interested in the history of semiotics.
The Ars versificaria builds upon the earlier attempts by Matthew and Geoffrey to relate the grammatical and rhetorical traditions.Gervase folalows the trajectory of Geoffrey away from grammar and toward rhetoric.He borrows names, doctrine and examples from both traditions, but he prefers the rhetorical terminology he borrows from Geoffrey's Summa and Documentum.Gervase usually restricts his references to Donatus to identifying the Greek names of grammatical figures with the Latin names of rhetorical colours. 75n the Ars versificaria, figures of speech (marked + on the charts) take up much of identitas, all of conservantia and a good deal of mutatio. 76The arrangement of the figures is based on deviations from a degree zero of expression, rudis identitas, expression lacking any fault or figure or color.Gervase defines conservantia as decoration in itself, without mutatio, which involves subtraction from or addition to or disorder of the expression.The arrangement of figures in the Ars versificaria may be inspired by Geoffrey's distinction between figures of amplification and abbreviation in the Documentum, but Gervase places many of Geoffrey's figures differently. 77larifying the precise differences between the arrangements of Matthew, Geoffrey, and Gervase requires a closer reading than I can pursue here.Gervase so often borrows from Matthew, Geoffrey and Bernard throughout the Ars versificaria that the charts generated by the investigation map out the whole work.The charts provide a clear and comprehensive outline of the highest levels of the Ars versificaria.The outline is not fully detailed.Borrowings from the three authorities only begin to cover the lower levels of the work.The charts provide, at best, a solid overview of the major divisions of the Ars versificaria. 78he following remarks draw upon the view of higher levels the charts provide to suggest three areas worthy of more detailed investigations.
1. John Ward has written, in regard to Boncompagno's Ars dictaminis, that "the work is ostensibly designed [...] to make up for the prac- tical deficiencies of classical rhetorical theory, particularly in the area of thirteenth-century legal applications". 79The borrowing and citation practices of Gervase suggest the Ars versificaria is designed to supplement the traditional treatment of grammatical figures and rhetorical colours in the area of twelfth and thirteenth century verse and prose composition.Twelfth century writings like the Cosmographia and Architrenius appear to require the creation of new theory.Gervase borrows examples from both the grammatical and rhetorical traditions to discuss figures of speech, but when he treats determinatio, he draws many examples from Bernardus. 80Gervase cites Bernardus often in other non-traditional areas: in equalitas and transmutatio, in assumptio, in munditia. 81Gervase cites John of Hanville in non-traditional areas, too: in equalitas, frequently in assumptio, in the special rules for verse and prose. 82Further study of the placement and explication of twelfth and thirteenth century examples may help us understand why new areas of theory appear in the Ars versificaria. 832. If the sequence of identitas, similitudo, and contrarietas represents progressively sophisticated levels of expression, Gervase, placing his tropes across the three loci in about the same order the tropes are presented in Rhetorica ad Herennium Book IV, shows that he understands the role the tropes play as a bridge between figures of speech and thought.Under identitas-mutatio-diversio, Gervase begins to lay out the tropes: under diversio-digressio, (2) antonomasia, (3) metonomy, (4) circuitio, and (7) synecdoche; under diversio-transcensus, (5) hyperbaton and (6) hyperbole.Gervase continues to lay out the tropes under similitudo: under assumptio, (1) onomatopoeia and (8) cat- achresis; under transumptio, (9) metaphor and (10) allegory.Only synecdoche and onomatopoeia deviate from the order presented in Rhetorica ad Herennium Book IV: synecdoche most likely for its traditional connection with metonymy, and onomatopoeia for its traditional connection with catachresis. 84his presentation simplifies a much more complex situation.Areas that intervene in the sequence are excluded, and the division of allegory into species belonging to similitudo and contrarietas is neglected.Still, the movement from vox to dictio to oratio that spans the discussion of similitudo is intriguing, and further study of the arrangement of tropes in the Ars versificaria is warranted.3. Unlike Geoffrey's Poetria nova, which quickly became a fixture in the medieval classroom, the Ars versificaria seems to have had no immediate impact at all.In fact, Gervase's art does not appear to have generated much interest before the late 14th century Oxford renaissance of Anglo-Latin rhetoric described by Martin Camargo. 85At that end of the 14th century, the Ars versificaria is copied in its entirety, and the Tria sunt, whose eighth chapter is "On the Functional Categories into Which All of the Colors Can Be Sorted and How the Figures Correspond to the Colors", quotes Gervase frequently. 86Further investigation of how the Ars versificaria interacts with the Tria sunt must await a critical edition of the latter, but we are likely to learn much from a study of its borrowing and citation practices.
For now, we need to revise our understanding of the time and place of the Ars versificaria.The work does not refer to the Poetria nova, and its date does not depend on the date of the Poetria nova.But how can it be that Gervase borrows from Geoffrey's other works, and not from the Poetria nova?Perhaps it is that the two works are being composed simultaneously, but Gervase is one of the clerics with Stephen Langton in France, separated from Geoffrey during the interdict.But this, too, is a matter for further investigation.

63
Matthew, Ars, III 3 for schemata, III, 18 for tropes.Matthew discusses only the schemes and tropes he considers most useful to the writer of verse.64 Matthew, Ars, III 45. 65 Matthew, Ars, III 47. Marbod's figures are a subset of the thirty-five figures of speech treated in Rhetorica ad Herennium Book IV. 66 Geoffrey, Summa, 321-5.67 Faral, Les arts poétiques, 325.
(1): Gervase never cites by name of author or work his borrowing of doctrine from the arts of Matthew and Geoffrey.35d.Because Matthew borrows doctrine heavily from Donatus and Isidore, but he mentions Isidore only three times and the Barbarismus just once, Gervase's explicit borrowings from the Ars versificatoria might be difficult to pin down.Gervase uses examples from Bernardus' works to make theoretical points, but he makes only one such point without providing an example.This occurs at the very end of the De Arte versificatoria, in the section dictamen prosaicum, where Gervase notes that Bernardus did not follow Isidore's advice in his prose and avoid eliding the letter ʻmʼ with vowels, Gervais von Melkley, Ars poetica, 222.names Matthew or his art as the source of the doctrine he borrows from him or it.e.Similarly, Gervase's borrowings of doctrine from the Summa de coloribus rethoricis and Documentum de modo et arte dictandi et versificandi never name Geoffrey or his works, and Gervase does not borrow doctrine from the Poetria nova at all.Point (2): Gervase cites by name of the author only the examples he borrows from the poetic works of Matthew, Geoffrey, and Bernardus.a. Gervase may introduce the classical examples he shares with Matthew by the name of the author (Ovid, Lucan, Statius), or he may not (Virgil, Juvenal), but whether or not he mentions an author, Gervase never names Matthew as the source of the quotation. 39In fact, after the initial mention of Matthew in his hierarchy of authorities, Gervase names him just once in the Ars versificaria, as Vindocinensis, when he presents, as an example of singula singulis, the first lines of the poem that closes the Ars versificatoria, Matthew's Christe, tibi sit honor. 40b.Gervase's borrowings from Geoffrey of Vinsauf are easier to assess.Hector | viribus" -appears to be as close in thought if not in meter. 47b.Gräbener suggests that Gervase's example of metaphor lilia faciei relates to the Poetria nova's lilia frontis, but Gervase explicitly cites the Anticlaudianus for frons lilia, and he provides his own variation, menti lilia, too. 48c.Gräbener suggests that Gervase's examples of allegory litus arat and laterem lavat relate to the Poetria nova's "Litus arat, laterem lavat, auram verberat", but both expressions are commonplaces that can be traced back to Ovid and Terence respectively.
36Fortunately, when Gervase borrows from Donatus, he cites him or his work by name in 21 of 24 instances.For example, Matthew borrows from Donatus in 32 instances, but Gervase shares only two of these, and he labels both clearly a Donato.37Matthewalsoborrowsfrom Isidore in 35 instances, and Gervase shares three of these, but only the discussion of paronomoeon, one of the three times Matthew mentions Isidore, clearly is borrowed from Matthew's Ars versificatoria.38Inshort, whether or not Matthew borrows doctrine from Donatus or Isidore, Gervase 33 Index nominum, Gervais von Melkley, Ars poetica, 250-7; Index scriptorum, 279-85.34 Citations introduced by name of the author or the work are highlighted.Passages in the Ars versificaria that Gräbener notes as similar to, rather than identical with passages in other works are italicised.Bolded entries mark points of interest mentioned in the following discussion.35 38 Gervais von Melkley, Ars poetica, 12 for paronomoeon; also 113 citing Virgil, 61 modifying Virgil, not charted.never In his Summa, Geoffrey borrows most of his examples from Marbod of Rennes' De ornamentis verborum, and he cites less than twenty classical examples from Ovid, Lucan, Statius, Virgil, and Juvenal in the entire Documentum. 41Gervase does not appear to use Marbod directly.His three uncited borrowings of Marbod's verses for ratiocinatio, occupatio, and articulus (noted mx on Chart 1) reduce the examples in Geoffrey's treatises to no more than two lines, and treat all three figures in Geoffrey's terms. 42moribus,