BMCR 2008.06.03

Martianus Capella. Les noces de Philologie et de Mercure. Tome IV. Livre IV. La dialectique. Collection des Universités de France

, , , , , Les noces de Philologie et de Mercure. Collection des universités de France. Série latine, v. 372, 388, 389, 401, 407. Paris: Les Belles Lettres, 2007. volumes 1, 4, 6-7, 9 ; 20 cm.. ISBN 9782251014678. €35.00 (pb).

This is another Budé volume devoted to Martianus Capella’s De Nuptiis Philologiae et Mercurii, along with those devoted to Books 6, Geometria, by Barbara Ferré (2007), and 7, Arithmetica, by Jean-Yves Guillaumin (2003). The present volume is by Michel Ferré, who studied Capella’s dialectic in previous contributions in AC 72 (2003) 167-185 and REA 106 (2004) 147-173. After an Introduction (vii-lxxvii) come a select Bibliography divided into texts and studies (lxxix-lxxxvii), a List of Abbreviations (lxxxix-xc), the edition with critical apparatus and facing French translation (3-60), the Notes (61-124), and Indexes (exemplorum, nominum, and verborum, 125-140). The final Table des Matières (141) is a little confusing, as the page numbers it indicates for the Abbreviations, Notes, and three Indexes do not correspond to the actual numbers. But this is probably not a fault to be ascribed to the author, who has produced a satisfactory work.

The first half of the Introduction (vii-lxiii) is devoted to the work’s contents, whereas the second (lxiii-xc) addresses issues related to the edition of the text: a history of the text is provided on the basis of the extant evidence, and with Willis and Préaux Ferré believes that from Securus Melior Felix’s emended recensio (around A.D. 354) several codices derived, of which one in the Carolingian age was the archetype of our whole manuscript tradition. A survey and discussion of previous editions is provided, which the present one takes into consideration. There is no extensive setting out of the ecdotic criteria of this edition and no stemma is provided — among the editions of the whole De Nuptiis only Willis’ edition offers a stemma.1 Ferré does describe the manuscripts on which his own edition is based: the six defined as indispensable by Préaux, plus three other good codices.2 Moreover, Ferré has employed one representative of the indirect tradition, Remigius of Auxerre’s commentary on Martianus. The decision to take into account Remigius’s lemmata in the constitutio textus is, in my view, felicitous, although it is not always an easy task. However, at least two other ancient commentaries would have been very helpful to compare and use systematically in the collatio, all the more since they are earlier than Remigius’s (end of the ninth century): Eriugena’s Adnotationes in Marcianum, which were composed around A.D. 840-850, and the corpus of glosses formerly ascribed to Dunchad and Martinus of Laon, which are even more ancient.3 These ninth-century commentaries and glosses are contemporary with, and often earlier than, the Martianus manuscripts used by the editor in the constitutio textus. The codices themselves in which they are preserved are contemporary with those of Martianus: e.g. the Corbeiensis (Paris, Bibliothèque Nationale, fonds lat., MS 12960, which includes the commentary ascribed to Dunchad and then Martinus and the one probably due to Eriugena or an abbreviation of it) dates back to the ninth century.4 Thus, they bear full value also as testimonia to Martianus’s text.

In the first part of his introduction, Ferré sets out to contextualize Martianus among other Latin authors of the imperial period who lived in Africa and wrote works related to the ars dialectica : Marius Victorinus, who wrote a work on hypothetical syllogisms; Augustine, who wrote a treatise on dialectic; Tullius Marcellus from Carthago, who, according to Cassiodorus, summarized the doctrine of syllogisms, both categorical and hypothetical; and Apuleius, who wrote a Peri Hermeneias well known to Martianus and Cassiodorus, who partially abridged it in their De Nuptiis and Institutiones respectively. It must be noted that the Apuleian paternity of that treatise is uncertain, but Ferré does not enter into this question. It is also interesting to add that Apuleius, the only Middle-Platonist among the four abovementioned authors — Victorinus and Augustine belong to Neoplatonism — is also a literary model for Martianus, especially with his Metamorphoses, and not only for the first two books of the De Nuptiis but also for the broad narrative sections in the others.5

After summarizing the contents of the De Nuptiis, Ferré points out that the choice of the satiric genre is due to its allowing a mixture of tones and forms (given the way he dwells on the satire, its history, the work’s prosimetric structure, and the separate Nachleben of its first two books in respect to the other seven, it is somewhat surprising that neither in the notes nor in the bibliography is any contribution by Danuta Shanzer mentioned6). After underscoring the ambivalence of dialectic as presented by Martianus in the very description of the woman who is its prosopopoeia, because of its capacity for doing good or severe damage thanks to its demonstrations, Ferré rightly observes that Martianus devotes no word to the philosophical status of this discipline. Basing himself on Ilsetraut Hadot’s Arts libéraux et philosophie dans la pensée antique (Paris: Vrin, 2005), he also highlights the relatively late fixing of the system of the liberal arts, finally organized within the framework of the trivium and the quadrivium. Regarding the latter, at any rate, it may deserve mention that Plato in his Republic, besides dialectic itself, recommended to the “guardians” precisely the study of the disciplines subsequently indicated as quadrivium.

A succinct history of dialectic occupies a good portion of the introduction (xvi-xlix). Two branches are individuated first: the Peripatetic, stemming from Aristotle, and the Stoic, stemming from Chrysippus. Consequently, we find a description of Aristotle’s logical writings, later known as Organon : Categoriae, De Interpretatione, Analytica Priora, Analytica Posteriora, Topica, and Elenchi Sophistici. Then the Megarian school is presented, from which some scholars distinguish the “Dialectic” school, whose existence is denied by others. No writing by the Megarians is extant, and Ferré expresses the not ungrounded suspicion that some of the Megarians’ or Dialecticians’ discoveries may have been ascribed to the Stoics. The Stoics’ logic is then outlined, essentially in conformity with the study of Jean-Baptiste Gourinat, La dialectique des Stoïciens (Paris: Vrin, 2000). Ferré rightly calls attention to the fact that for the Stoics logic was not a tool, an ὄργανον, but one of the three parts of philosophy, along with physics and ethics. Then he studies the introduction of logic into the Latin world, with Aelius Stilo’s first treatise on dialectic in Latin and the work of his disciple Varro. Latin dialectic is inspired by Stoicism and is seen first of all as a support to rhetorical inventio; Cicero transposed into Latin the Greek logical vocabulary. The Aristotelian renaissance in dialectic began in the second century A.D. with commentaries on Aristotle’s logical works, the first entirely extant of which is the Peri Hermeneias ascribed to Apuleius. As for Alcinoos’s Didaskalikos, it is pointed out that it endeavors to prove that Plato in fact invented both the categorical syllogisms of Aristotle and the hypothetical syllogisms of Chrysippus. The main representative of this Aristotelian renaissance is Alexander of Aphrodisias, who, between the second and the beginning of the third century, commented on the whole logical corpus of Aristotle, and Porphyry’s Isagoge will then be placed at the beginning of Aristotle’s Organon. Other important commentaries on various works of the Organon between the fourth and the sixth century were prepared by Themistius, Ammonius, Olympiodorus, Philoponus, Elijah and David.

Ferré is skeptical in respect to Minio Paluello’s hypothesis (based on Martianus, 4.335, but not supported by further evidence) that Varro translated the whole Organon into Latin. So, “il reste très vraisemblable que le premier traducteur d’Aristote en latin est Marius Victorinus” (xxxviii). In fact, I find it probable that Martianus’ words in 4.335 ought to be taken in a less specific sense, in reference to the inclusion of dialectic in his Disciplinarum Libri. Augustine is singled out as the author who allowed the passage of dialectic to Christian doctrine, where it was cultivated by Boethius, who translated and commented on Aristotle’s Organon and wrote his own treatises on dialectic. Whereas Boethius’ philosophical works do not contain Christian references, the passage of dialectic to Christianity did transform it into an ancilla theologiae. Ferré finally traces a table of contents of Martianus’s Book 4 and for each section indicates the corresponding work in Aristotle’s Organon and possible intermediate sources, probably the Peri Hermeneias for the syllogistic part and a treatise like the Paraphrasis Themistiana or Categoriae Decem for the part on categories. The influence of both Aristotelian and Stoic logic upon Martianus’s treatment is rightly observed, and the conclusion is “Martianus Capella a donc martyrisé les six livres de l’ Organon pour les faire entrer dans ce lit de Procuste qu’est la classification stoïcienne en trois parties” (lvi).7 Finally, Ferré shows how Martianus’s Book 4, in addition to being handed down together with the others of the De Nuptiis, also had a separate transmission between the ninth and the eleventh century in ten codices, none of which was utilized in the collatio. They all belong to the period of the logica vetus, before the rediscovery of Aristotle’s other logical works from Arabic and Greek, which evidently made Martianus’s synthesis less interesting.

Among the explications and literary and lexical parallels provided in the notes to the edition and translation, I pick out only a few notable points. In n. 142, while commenting on Martianus’ statement in 4.370 that substances cannot be subject to increment or decrease whereas qualities can, Ferré cites, besides Greek and Latin commentators of Aristotle on this issue, also Remigius’ commentary on Martianus: Iustitiae vero perfectio non recipit magis et minus, sed appetentia et voluptas eius. Remigius actually comments on this point at much greater length, and owes much to Eriugena’s (probably abridged) Adnotationes, 172.13 and 172.17.8 Pretty different is the interpretation in the commentary that was ascribed to Dunchad and then to Martinus at 172.11 = 379.121 (Ramelli, Commentari, 641), but from a comparison between their interpretations it is clear that Remigius depends on Eriugena even ad verbum in his observations (much fuller that the section quoted) at 172.11 and 172.16 = 370.121.9

In n. 43 Ferré has recourse to Remigius in support of his textual choice in 4.332, verosae, attested by all nine manuscripts of his collatio apart from B before the correction, instead of virosae, preferred by Dick and Willis. Remigius read verosae and explained it as deriving either from verus or from veru ( verosae id est firmae et veridicae, vel, quod est melius, verosae id est acutae, ab eo quod est veru, 154.4 = 332.108; Ramelli, Commentari, 1130-1131). He would have done better to have recourse also to two other, and earlier, commentaries: Eriugena’s Adnotationes and the commentary ascribed to Dunchad and subsequently to Martinus. At 154.2-3 = 332.109 they both have verosae as their lemma, which they evidently read in their Martianus manuscripts, and they explain it as acutae (Ramelli, Commentari, 267, and nn. 37 and 109; 285, 631), which will then constitute Remigius’s second option. Thus, they both support Ferré’s textual choice.

In n. 38,on the other hand, Ferré does not cite Remigius or any other ancient commentator to buttress his textual emendation in 4.331, venenatorum (“venomous beasts”), in lieu of venenorum (“venoms”), which is the reading of all manuscripts, retained by Dick and Willis. But a look at the commentators here would be very useful. Remigius is the only one who comments on this specific passage, at 153.10 = 331.107: Et amicitia venenorum] subaudis docet hoc quod de Lybia sit quia recognoscitur a venenatis animalibus (Ramelli, Commentari, 1128-1129). As is evident, Remigius’ lemma reads venenorum, but his own paraphrase has venenatis, which implies that he either read precisely venenatorum in Martianus or interpreted venenorum as venenatorum. Of course, he may have had earlier commentaries or glosses at hand which induced him to do so.

Finally, in n. 29 Ferré explains that in 4.330 he follows Boettger in conjecturing urbe instead of the reading of all the manuscripts, rupe ( urbe, it must be said, is preferred by Willis as well). He then cites the twelfth-century mystic and learned teacher Hugh of St. Victor as a witness to the legend of the “rock of Parmenides,” investigated by Klibanski in MRS 1 (1941-43) 178-186. Now, it is precisely the commentators on Martianus who initiated this legend, and an analysis of the three earliest commentaries treating this passage reveals that the first still bears a trace of the original reading, urbe, whereas Remigius has no reference to a city and recounts the legend of the rock on which Parmenides used to meditate, since he did not read Aegyptiorum urbe atque in Parmenidis (exinde gymnasium), but Aegyptiorum rupe atquin Parmenidis, and understood: “on a rock in Egypt, indeed, on Parmenides’ rock.”10 The old commentary ascribed to Dunchad and then Martinus, instead, although it does not include urbe in its lemmata in 153.2 = 330.17, does bear a trace that at a previous stage of the manuscript tradition and its interpretation precisely this reading existed, since in it, after the lemma rupe, explained by the fact that certain philosophers abandoned everything and in isolated places discovered dialectic, and after the lemma atquin, glossed with certe, the third lemma Parmenidis is explained as “a city in Egypt,” pointing exactly to the reading urbs (Ramelli, Commentari, 630 and n. 34). This provides good support for the conjecture at stake. Indeed, Eriugena’s Adnotationes, probably slightly later and here abbreviated, but earlier than Remigius’s commentary, still maintain a vestige of this interpretation in the ‘topographic’ understanding of Parmenides’s name: Atquin] certe. Parmenides] locus est ubi Parmenes philosophus philosophatus est et inicia huius artis repperit (Ramelli, Commentari, 266 and nn. 26 and 99).

Overall, this is a careful11 and useful volume, with a profitable commentary on dialectic as it is expounded by Martianus, and moreover at an affordable price. It is to be hoped that the other books of the De Nuptiis that are still missing in the Budé series will be soon available as well.


1. A discussion of Willis’ stemma, an analysis of the principal manuscripts, and a proposal of a new stemma is to be found in a study that is not mentioned in the book under review: D. Shanzer, “Felix Capella: Minus Sensus Quam Nominis Pecudalis,” CPh 81,1 (1986) 62-81. The proposed stemma, on p. 77, includes twelve manuscripts.

2. London, British Library, Harleianus 2685, ninth century (A). Bamberg, Staatsbibliothek, Bambergensis Class. 39 (M.V. 16), ninth century (B). Paris, Bibliothèque Nationale de France, Parisinus 8670, ninth century — even if this is not recorded in the description, lxxi-lxxii — (D). Vatican City, Biblioteca Apostolica Vaticana, Reg, Lat. 1987, first half of the ninth century (H). This codex was not used in the previous editions of Martianus Capella. Leiden, Bibl. der Rijksuniversiteit, B. P. L. 87 (L). Leiden, Bibl. der Rijksuniversiteit, B. P. L. 88 (N). Karlsruhe, Badische Landesbibliothek, Reichenauensis 73, first half of the ninth century (R). Vatican City, Biblioteca Apostolica Vaticana, Reg. Lat. 1535, third quarter of the ninth century (T). Leiden, Bibl. der Rijksuniversiteit, Voss. Lat. F 48, middle of the ninth century (V).

3. These texts are studied, translated, and commented on together with the other Mediaeval Latin commentaries on Martianus in I. Ramelli, Tutti i commenti a Marziano Capella: Scoto Eriugena, Remigio di Auxerre, Bernardo Silvestre e anonimi (Milan: Bompiani – Istituto Italiano per gli Studi Filosofici, 2006), 263-314 and 571-573 for Eriugena’s Adnotationes, on Book 4; 626-650 and relevant notes for the glosses formerly ascribed to Dunchad and Martinus, on Martianus’s Book 4; 1116-1207 and 1731-1734 for Remigius’s commentary on Martianus’s Book 4. Cf. Ead., Marziano Capella. Nozze di Filologia e di Mercurio (Milan: Bompiani, 2001), xxi-xxiii with relevant notes, xxxix-xlvi, 202-283, 849-867 on Book 4.

4. See also the detailed picture drawn by M. Teeuwen, “The Study of Martianus Capella’s De Nuptiis in the Ninth Century,” in Learned Antiquity, ed. A.A. MacDonald, MW. Twomey, G.J. Reinink (Leuven: Peeters, 2003), 185-194.

5. A methodical study of this literary relationship is still missing; a preliminary investigation has been offered in my “Materiali per lo studio della presenza di Apuleio in Marziano Capella: tipologie e significati,” Stylos 11 (2002) 97-114.

6. At least her book is worth mentioning: A Philosophical and Literary Commentary on Martianus Capella’s De Nuptiis Philologiae et Mercurii Book 1 (Berkeley: University of California Press, 1986), which by no means limits itself to commenting on Book 1 but offers an investigation into the date and historical context of the De Nuptiis (1-28) and a whole study of this work as a Menippean Satire (29-44). On Menippean Satire and prosimetrum in Martianus and the late antique Latin tradition I also refer to P. Dronke, Verse with Prose from Petronius to Dante (Cambridge, Mass. – London: Harvard University Press, 1994).

7. On this I add a reference to F. La Vecchia, “The Logic in Martianus Capella’s De Nuptiis Philologiae et Mercurii,” Metalogicon 12 (1999) 41-70.

8. Magis iustus] iustus, quando de perfecto iusto dicitur, non recipit magis et minus, quando autem de voluptate, tunc recipit magis et minus. Plus enim adipiscitur iustus fieri unus quam alter. Similiter etiam de omnibus nominibus quae a qualitate sumuntur, ut est doctus. Nam quando perfectum doctum significant, non recipit magis et minus; potest enim et doctus et doctior fieri, hoc est potest perfectior appetitus esse unius quam alterius. In appetere enim comparatio est … Si consideraveris substantiam individuam, dico quoniam perfecta in semet ipsa est, non recipit magis et minus. Si vero participationem qualitatis per substantias consideraveris, invenies magis ac minus. Potest enim fieri quaedam substantia quae plus participat eandem qualitatem quam alia (see Ramelli, Commentari, 278 and nn. 67-68 and 240-241).

9. Iustitiae vero perfectio non recipit magis et minus, sed appetentia et voluptas eius. Iustus ergo, cum de perfecto iusto dicitur, magis et minus non recipit; quando autem de appetitu iustitiae iustus dicitur, tunc recipit magis et minus. Potest enim unus iustus magis adipisci iustitiam, minus alter. Similiter de omnibus quae a qualitate sumuntur intelligendum, ut verbi gratia doctus. Nam, quando perfectum significat, magis et minus non recipit, quando autem appetitionem doctrinae, recipit magis et minus. Potest enim et doctus et doctior fieri, hoc est potest perfectior appetitus esse unius quam alterius. In appetendo enim fit comparatio. … Si consideraveris substantiam individuam, dico quoniam perfecta est et individua, non recipit magis et minus. Si vero partitionem qualitatis per substantias consideraveris, invenies magis et minus. Potest enim quaedam substantia esse quae plus participat eandem qualitatem quam alia (see Ramelli, Commentari, 1166-1167).

10. 153.1 = 330.107: in rupe Aegyptiorum, et quasi explanans in qua rupe, nam multae rupes sunt aput Aegyptum: atquin, id est certe, in rupe Parmenidis … hic philosophus fuit et primus apud Aegyptios artem dialecticam repperit. Erat autem solitus deserere civitates … et in hac rupe solus residere ut liberius posset dialecticam meditari, unde et a Parmenide rupes Parmenidis vocata est (see Ramelli, Commentari, 1166-1167).

11. This is generally the case on the editorial level, too. I found only a few misprints, mainly in Greek accents, e.g. p. 76: μή τὴν instead of μὴ τὴν; 82: γὶνεται for γίνεται and ἀπογινεται for ἀπογίνεται; 84: ἐννοίαν for ἔννοιαν and προληψιν for πρόληψιν; 92: δῆλον ὃτι for δῆλον ὅτι; 97: ἕχειν instead of ἔχειν.