¹ See, for example, Ward, John O., ‘From Antiquity to the Renaissance: Glosses and Commentaries on Cicero's Rhetorica,’ in Medieval Eloquence (ed. Murphy, James J.; Berkeley and Los Angeles 1978) 25–67. No question was raised more frequently in the discussion that followed the special sessions on medieval rhetoric at the 1982 meeting of the Medieval Academy of America, and many papers confronted it directly. Practical utility is an important theme in Ronald Witt's ‘Medieval Ars Dictaminis and the Beginnings of Humanism: A New Construction of the Problem,’ Renaissance Quarterly 35 (1982) 1–35.Google Scholar

The initial research for this study was undertaken in George Kennedy's seminar at the Southeastern Institute for Medieval and Renaissance Studies, Chapel Hill, North Carolina, Summer, 1979. A shorter version of the commentary was read at the Conference on Twelfth-Century Genres at Davis, California, March, 1981. In preparing the edition for publication, I was aided by a summer fellowship and a travel grant from the Research Council, University of Missouri-Columbia. I owe a special debt of gratitude to Charles Vulliez of Paris, who shared freely his unsurpassed knowledge of French artes dictandi manuscripts. grammatica and ars rhetorica, of Horace and Cicero, preserved in Matthew of Vendôme's Ars versificatoria (ca. 1175).

² For other descriptions of this manuscript, see Martin Camargo, ‘The English Manuscripts of Bernard of Meung's Flores dictaminum,’ Viator 12 (1981) 200–203, and Hampe, Karl, ‘Reise nach England vom Juli 1895 bis Februar 1896, III, xii: Formelbücher und Briefsteller in englischen Hss.,’ Neues Archiv der Gesellschaft für ältere deutsche Geschichtskunde 22 (1897) 610–11.Google Scholar

³ Langlois, Charles-Victor, ‘Questions d'histoire littéraire: Maître Bernard,’ Bibliothèque de l'École des Chartes 54 (1893) 237 n. 2. Other copies of the Cognito exist, for example, in Bruges, Bibliothèque Publique, MS 549, and Brussels, Bibliothèque Royale, MS 2070. Charles Vulliez is preparing an edition of the Cognito. CrossRefGoogle Scholar

⁴ For example, see fol. 10^r: ‘Medeam namque mater induit que in uiscera sua odio se armauit’ (cf. line 83) or fol. 47^r: ‘Me mihi furatum esse conqueror postquam te tetigi’ (cf. lines 218–219).Google Scholar

⁵ Cambridge University Library, MS Dd.9.38, fol. 115^ra. I discuss and document the reasons for identifying the Flores rhetorici as the most important of the ‘summulae turonenses’ cited in the Libellus’ prologue in ‘The Libellus de arte dictandi rhetorice Attributed to Peter of Blois,’ Speculum 59 (1984) 26–38.Google Scholar

⁶ Though dictamen may have reached France, from Bologna, as early as ca. 1130, it was at first taught exclusively through the treatises of the Bolognese dictatores. See, e.g., Haskins, Charles H., ‘The Early Artes Dictandi in Italy,’ in Studies in Mediaeval Culture (Oxford 1929; repr. New York 1965), 190–91, and Klaes, Monika, ‘Die “Summa” des Magister Bernardus: Zu Überlieferung und Textgeschichte einer zentralen Ars dictandi des 12. Jahrhunderts,’ Frühmittelalterliche Studien 24 (1990) 234 n. 154a.Google Scholar

⁷ We cannot be sure of the full context in which this copy originally appeared, however, since P is a composite manuscript and the copy of the Flores occurs at the beginning of a gathering that is preceded by a gathering from a different source. The scribe of Basel, Universitätsbibliothek, MS F.VI.15 (B) treats the Flores rhetorici as the first part of a work, the second part of which is a collection of model letters, many if not all of them Bernard of Meung's.Google Scholar

⁸ The ensemble formed by this ars dictandi and the various collections of proverbs, letters, and documents composed or collected by Bernard of Meung is sometimes given the title Flores dictaminum. Depending on whether one chooses to stress the first or the second word, one can cite this title as evidence for linking or contrasting the work with the Flores rhetorici. Google Scholar

⁹ See Gervase of Melkley, Ars versificatoria, ed. Hans-Jürgen Gräbener, Gervais von Melkley, Ars Poetica (Forschungen zur romanischen Philologie 17; Münster 1965) 1.9–11, and Matthew of Vendôme, Epistule 1.3.69–70, ed. Franco Munari, Mathei Vindocinensis Opera II (Storia e Letteratura 152; Rome 1982) 90. On the correct title of Gervase's treatise, see Franz-Josef Worstbrock's review of Gräbener's edition, in Anzeiger für deutsches Altertum und deutsche Literatur 78 (1967) 99 n. 1.Google Scholar

¹⁰ Silvestris, Bernardus, Cosmographia (ed. Peter Dronke; Leiden and Cologne 1978) 6. The unsuccessful attempts to ascribe to Bernard Silvestris various artes dictandi associated with an Italian master named Bernard have been evaluated most recently by Monika Klaes, ‘ “Summa” des Magister Bernardus’ (n. 6 above) 198–234, esp. 199–202, 232–33.Google Scholar

¹¹ Ed. Dronke, Cosmographia 22–23. On the extensive study of Claudian in the twelfth century, see esp. The Commentary of Geoffrey of Vitry on Claudian ‘De raptu Proserpinae’ (ed. Clarke, A. K. and Giles, P. M., Mittellateinische Studien und Texte 7; Leiden and Cologne 1973).CrossRefGoogle Scholar

¹² Ed. Gräbener 6.6–9 (see also notes to lines 143–144). Gräbener seems to feel that Bernard Silvestris is at least the best precedent for Gervase's ambitious attempt to represent rhetoric (or poetics) as a sharply articulated system (p. XLI). Gervase's tripartite schema receives detailed analysis in Purcell, William M., ‘Identitas, Similitudo, and Contrarietas in Gervasius of Melkley's Ars poetica: A Stasis of Style,’ Rhetorica 9 (1991) 67–91.CrossRefGoogle Scholar

¹³ Faral, Edmond, Les arts poétiques du xii^e et du xiii^e siècle (Paris 1924) 1–3; ed. Munari, Opera II 23–25; and Swietek, Francis R., ‘Three Poems after Matthew of Vendôme,’ Speculum 58 (1983) 917–19. On the rivalry between Matthew and Arnulf of Orléans, also see the introduction to Arnulfi Aurelianensis Glosule super Lucanum (ed. Berthe Marti, Papers and Monographs of the American Academy in Rome 18; Rome 1958), esp. xviii-xxii. Most of what we know about Matthew's life comes from statements made in his poetry, some of which may be pure inventions created for a fictional persona. Are we justified, for example, in assuming that the student author of Epistule 1.3, is supposed to represent Matthew? For independent confirmation of Matthew's debt to Bernard Silvestris, see, e.g., ed. Dronke, Cosmographia 10.Google Scholar

¹⁴ For example, both authors are inordinately fond of antitheses. See below, 227–273, and cf. Milo 101–102; Epistule 2.1.93–102; 2.2.105–116; 2.3.73–80; 2.4.35–40. Matthew seems to prefer what he calls ‘antithetum per verba,’ whereas the Flores rhetorici favors ‘antithetum per constructiones’ (see Ars versificatoria 3.25–29).Google Scholar

¹⁵ Matthew's use of epithets as the basis of description, which is in turn the subject matter of poetry, is echoed in the material that concludes the Flores rhetorici (lines 334–393) and in the excerpts from Claudian and Horace selected to illustrate ‘proprietas’ (lines 164–186). Matthew's emphasis on brevity as the standard of good style (also on Horace's authority) is shared by the Flores rhetorici; both works, in fact, often invoke this standard in the course of effecting a transition. (Cf. Ars vers., Prol., 1.5; I.11.118, and Flores rhetorici, lines 34–35, 187–191, 332–333. The habitual use in such transitions of the negative conditional, especially ‘Ne videar …,’ is another point of correspondence.) It is even possible that the Flores’ failure to consider the salutatio is somehow connected with the author's experience in teaching verse composition. See Matthew's, remarks on beginning a poem with a proverb: Ars vers. I.16.Google Scholar

¹⁶ An especially striking instance is Flores rhetorici, line 386. In quoting Statius, Thebaid 6.106, the Flores supplies the same variant (uadis for fretis) as Matthew's Ars versificatoria (see Explanatory Notes). The heavy reliance on Horace for standards of correctness is typical among twelfth-century teachers of grammar and rhetoric (see Douglas Kelly, ‘The Scope of the Treatment of Composition in the Twelfth- and Thirteenth-Century Arts of Poetry,’ Speculum 41 [1966] 276–77). Though Matthew shows no particular fondness for Claudian in his extant works, he claims to have composed a poem on the Rape of Proserpine (see Faral, Arts poétiques [n. 13 above] 7, 10). On the other hand, the Flores rhetorici never quotes Lucan, one of Matthew's favorite sources of examples. Cicero, besides Horace (and possibly Claudian: see note to line 170) the only author named in the Flores rhetorici, is invoked with some frequency in both works, though not always accurately (see Flores, lines 127–128).Google Scholar

¹⁷ Cf. Quintilian, , Inst. orat. 8.6.4–18; Donatus, Ars gram. 3.6; Isidore, Etymol. 1.37.2–5; Bede, De schem. et trop. (ed. Karl Halm, Rhetores latini minores [Leipzig 1863; repr. Frankfurt, 1964]) 611–12; Alberic of Monte Cassino, Dictaminum radii 6.1; etc.Google Scholar

¹⁸ In employing translatio, the Flores rhetorici belongs to the Ciceronian tradition. Cf. Cicero, De oratore 3.155–166; Rhetorica ad Herennium 4.45; Martianus Capella, De nuptiis 512; Alcuin, Rhetorica (ed. Halm [n. 17 above]) 794; etc. Quintilian uses both translatio and metaphora. Google Scholar

¹⁹ Gräbener (p. XLI), Worstbrock (review cited in n. 9 above) 101, and Hardison, O. B., Jr. (Classical and Medieval Literary Criticism [New York 1974] 294) attribute the same division, as it appears in Gervase's Ars versificatoria, to the influence of dialectic. Of course, Gervase could also have taken the triad indirectly from the dialecticians, through an intermediary such as Bernard Silvestris.Google Scholar

²⁰ The citation from the Epistule poses problems of another sort, since the student claims that his teacher was Bernard Silvestris (1.3.69–70). If Bernard's teaching is accurately reflected here, then the case for ascribing to him the tripartite division of metaphor is weakened, though it is certainly possible that the Flores rhetorici derived its theory of metaphor from Bernard and the material that specifically concerns letters from other sources.Google Scholar

²¹ Southern, R. W., ‘Peter of Blois: A Twelfth-Century Humanist,’ in Medieval Humanism and Other Studies (Oxford 1970) 110; ed. Dronke, Cosmographia 9.Google Scholar

²² Dronke, Peter, ‘Peter of Blois and Poetry at the Court of Henry II,’ Mediaeval Studies 38 (1976) 193. The bibliography of Peter's poems is on pp. 215–32.CrossRefGoogle Scholar

²³ Cf. n. 5 above. Tore Janson rejects the attribution in Prose Rhythm in Medieval Latin from the Ninth to the Thirteenth Century (Studia Latina Stockholmiensia 20; Stockholm 1975) 97–98.Google Scholar

²⁴ In designating the entity to which his rules apply, the author uses the terms materia (7 times), oratio (3 times), and the classical littere (twice). Even in the passages borrowed directly from the Flores rhetorici, the Libellus consistently uses epistola. The failure to discuss salutatio could be a throwback to classical rhetoric, in which the first part of an oration is the exordium, the part that most closely resembles the Flores‘ prouerbium. Google Scholar

²⁵ Minus the references to meter, Matthew of Vendôme's definition of verse (Ars vers. 1.1) could easily have been used in the Flores rhetorici to define good epistolary prose: ‘oratio succincte … progrediens venusto verborum matrimonio et flosculis sententiarum picturata, que nichil diminutum, nichil in se continet ociosum…. elegans iunctura diccionum, expressio proprietatum et observatum uniuscuiusque rei epytetum.’ The similarity with the diction used in the Flores to describe stylistic ideals is especially striking.Google Scholar

The interrelationship between the ars dictaminis and the ars poetriae in the schools of northern France and England is a complex topic, which I have addressed in a separate essay, ‘Toward a Comprehensive Art of Written Discourse: Geoffrey of Vinsauf and the Ars Dictaminis,’ Rhetorica 6 (1988) 167–94. One even sees the influence of this crossfertilization in some of the dictatores of thirteenth-century Bologna. Guido Faba's Summa dictaminis, for example, contains a long section dealing with colores rhetorici (ed. Augusto Gaudenzi, Il Propugnatore, n.s. 3, pt. 2 [1890] 356–70); and Books 2 and 7 of Bene of Florence's Candelabrum treat stylistic matters covered in the arts of poetry. Also see Witt Ronald G., ‘Boncompagno and the Defense of Rhetoric,’ The Journal of Medieval and Renaissance Studies 16 (1986) 1–31 (especially 5ff.).

²⁶ The emphasis on stylistic variation in the remainder of the Flores rhetorici is already anticipated in the warning that sacietas and fastidium result from similitudo (line 19). Matthew of Vendôme shares this interest in changing wording without changing meaning in order to give variety. See Ars vers. 4.20–31.Google Scholar

²⁷ Black, Max, ‘Metaphor,’ Proceedings of the Aristotelian Society 55 (1954–1955) 276.CrossRefGoogle Scholar

²⁸ At one time there existed another copy of the Libellus, in a manuscript at Dover Priory. See James, M. R., Ancient Libraries of Canterbury and Dover (Oxford 1904) 479 (#327, fol. 111^r).Google Scholar

²⁹ On Arsegino's Quadriga, see Paolo Marangon, ‘La Quadriga e i Prouerbi di maestro Arsegino: Cultura e scuole a Padova prima del 1222,’ Quaderni per la storia dell'Università di Padova 9–10 (1976–1977) 1–44; on Bene's Candelabrum, see Bene Florentini Candelabrum (ed. Gian Carlo Alessio, Thesaurus Mundi 23; Padua 1983), esp. 378–80.Google Scholar

³⁰ See especially Documentum de modo et arte dictandi et versificandi 2.3.7–22. The terms of this discussion closely parallel the Flores rhetorici, but the tripartite division is lacking. On medieval treatments of metaphor, see especially Ulrich Krewitt, Metapher und tropische Rede in der Auffassung des Mittelalters (Beihefte zum Mittellateinischen Jahrbuch 7; Ratingen, Kastellaun, and Wuppertal 1971), and Nims, Margaret F., ‘Translatio: “Difficult Statement” in Medieval Poetic Theory,’ University of Toronto Quarterly 43 (1974) 215–30.CrossRefGoogle Scholar

³¹ Known as the Codex Villarensis (formerly MS 225), Zaragoza 41 consists of 84 folios, the first 42 devoted chiefly to legal and historical materials (some of them in Spanish) and the remainder to the ars dictaminis. The Flores rhetorici (fols. 69^ra–70^va) is preceded by Geoffrey of Bologna (Geoffrey of Vinsauf?), Summa de arte dictandi, with the first chapter at the end and two passages interpolated (fols. 43^ra–49^ra); a ‘Tractatus de regimine parcium’ (fols. 49^va–51^ra); a collection of exordia (fols. 51^ra–53^vb); Introductiones prosaici dictaminis (fols. 54^ra–64^rb); an anonymous thirteenth-century ars dictandi (fols. 64^va–65^vb); and a letter collection (fols. 65^vb–68^vb); and is followed by a brief extract from Henricus Francigena's Aurea gemma (fol. 70^va–vb) and by Bernard of Meung's letter collection (fols. 71^ra–84^rb). Because I have seen Z only in a microfiche of poor quality, a fuller description is not possible. For more details about the second part of the manuscript and for further bibliography, see Klaes, ‘ “Summa” des Magister Bernardus’ (see n. 6 above) 204–205.Google Scholar

³² See Vulliez, Charles, ‘Un nouveau manuscrit “parisien” de la Summa dictaminis de Bernard de Meung et sa place dans la tradition manuscrite du texte,’ Revue d'histoire des textes 7 (1977) 133–51, for a description and analysis of NAL 757.Google Scholar

³³ The appendix in P consists of 116 proverbs and a certain amount of commentary, including an introductory and a concluding paragraph (lines 398–411 and 537–540, respectively). The collection in L comprises 57 proverbs, without commentary. Of these, only the first six and the last seven are not also found in P. The 44 shared proverbs occur in the same order in both manuscripts, with only one significant exception: the proverb about adulterous priests (lines 523–524) is #7 in L but #103 in P.Google Scholar

The group of 38 proverbs that T shares with P, including 4 not found in L, comprises #121–180 (fols. 117^va–118^ra), corresponding to lines 457–509 of the edited text based on P. All but one of the 22 proverbs within this cluster that are not in P occur immediately following the first shared proverb (#122–142, 165). The wording of the proverbs in T is very close to that of P and L, but the order differs somewhat. In P, the 38 proverbs occur in the following order: T, #121, 160–162, 157–159, 163–164, 166–171, 173, 172, 143–150, 152, 151, 153–156, 174–180. Immediately following the cluster of proverbs found also in P and L, are the first and the sixth proverbs from the group of seven that ends the collection in L (T, #181, 183; fol. 118^ra). The fourth and fifth proverbs from that same group of seven are also found in T, but nearer to the beginning of the collection (#53–54, fol. 117^rb), in the midst of proverbs found in neither P nor L.

³⁴ Also see Collectio Salernitana V (ed. Henschel, G. E. T., Daremberg, C., DeRenzi, E. S.; Naples 1859) 128.Google Scholar

³⁵ On this section, see Charles Samaran and Robert Marichal, Catalogue des manuscrits en écriture latine portant des indications de date, de lieu ou de copiste III (Paris 1974) 31.Google Scholar

³⁶ Sedulii opera omnia (ed. Johannes Huemer, CSEL 10; Vienna 1885) 14–146, 155–62, 163–68.Google Scholar

³⁷ Lynn Thorndike and Pearl Kibre, A Catalogue of Incipits of Mediaeval Scientific Writings in Latin (rev. ed., Cambridge, Mass. 1963) col. 610.Google Scholar

³⁸ See Thurot, Charles, ‘Notices et extraits de divers manuscrits latins pour servir à l'histoire des doctrines grammaticales au moyen âge,’ Notices et extraits 22, pt. 2 (1868) 50, and Bursill-Hall, G. L., ‘A Check-List of Incipits of Medieval Latin Grammatical Treatises: A-G,’ Traditio 34 (1978) 446, col. 1.Google Scholar

³⁹ The last three items are treated as one in the description by Enrique Bayerri y Bertomeu, Los Codices Medievales de la Catedral de Tortosa: Novísimo inventario descriptivo (Barcelona 1962) 380.Google Scholar

⁴⁰ For example, I would not have supplied the phrase ‘in usus necessarios’ in line 72 were it not for the later occurrence in line 89, which clearly indicates a scribal slip in the earlier passage. In lines 81–84, to cite just one more example, there is clear evidence of scribal confusion: at least two proverbs were omitted, then later inserted in the left margin, where the binder trimmed much of them away. It seems likely that the same confusion caused the scribe to jump from ‘matrem’ in line 82 to the same word in line 83, and so I have inserted in the appropriate place two proverbs found in L, Z, B, and T but not in P. Where the addition of material not in P only slightly improves an already acceptable reading, I have included it only if it is also present in all or most of the other copies (for example, ‘doctoris’ [line 503] and ‘Ille’ [line 504]).Google Scholar

Title] T Incipiunt alie epistole compendiose B om. PLZ 1 depinguntur] compinguntur B; prius] primum T 2 in] om. B; refloret] reflorescat L refloreret T 3 elocutio] LZ eloquacio B locutio T om. P 4 facili] uel facile marg. L; combibatis] uel collibetis marg. L 5 teneatis] affigatis LZT commendetis B