Reference edition:

Manlii Seuerini Boetii Opera omnia, accurante J.-P. Migne, tomus posterior, Parisiis 1891, 711-762 (Patrologiae Latinae, 64)

All of Boethius’ works are contained in Patrologia Latina 64.  Included in this collection are Latin versions of the Greek essays of the Organon, comprising the translation of the Posterior Analytics.  The texts published by J. P. Migne ultimately derive from the 1546 edition of Heinrich Loriti (Glareanus), who assembled Boethius’ translations of Aristotle’s logical works from the editio princeps of Jacques Lefèvre d’Étaples (Jacobus Fabrus Stapulensis); this latter volume was published in 1503 as the first Latin version of the Organon, containing Boethius’ (certainly authentic) translations of the Isagoge, the Categories, and De interpretatione, as well as an anonymous version of the Posterior Analytics.  Subsequent publications of Stapulensis’ edition erroneously attached Boethius’ name to the entire collection, however, with the result that he came to be treated as the translator of the whole Organon.  By 1862, nobody doubted that he had composed the “vulgate” version of the Posterior Analytics (Minio-Paluello 1968) even though this text cannot safely be attributed to him.  Indeed, comparison of this text with Boethian translations whose authorship is not doubted (e.g. those of Porphyrius’ Isagoge or Aristotle’s Categories and De interpretatione) has revealed a significant number of lexical, syntactic, and stylistic differences in the handling of the original Greek.  This same analysis has, moreover, allowed us to attribute the translation to James of Venice, an educated, bilingual prose writer of the 12^th c., who - like Boethius - worked on a Latin translation of Aristotle’s corpus.  Indeed, this attribution is also supported by a marginal note in the “vulgate” translation of the Posterior Analytics reported by cod. Oxford, Balliol College 253 (f. 242r), from the second quarter of the 13^th c. (Minio-Paluello 1952).

Although there are no extant manuscripts of Boethius’ Latin edition of the Posterior Analytics, he - after 510 AD -  claimed to have composed one (In Cic. Top. 1 = PL 64: 1051B: “ab Aristotele transtulimus;” De Rijk 1964).  On their own, however, Boethius’ words are insufficient to guarantee the real existence of a translation, since on another occasion he contradicted himself in declaring both that he had already translated the Prior Analytics and - later - that he intended to do so (De syll. Cat. = PL 64: 822B and 829D-830D) (Minio-Paluello 1968).  We are similarly unable to be certain, however, that a Boethian version of the Posterior Analytics did not exist: perhaps it has not come down to us because the treatise’s complexity led it to be transcribed very rarely (it was the most advanced and difficult of the logica nova) or because the high Middle Ages had little interest in the ars demonstrandi.  What we can say for certain is that the oldest extant Latin translations of the Posterior Analytics date to the 12^th c.: the first is that of James of Venice, which was probably written in the second quarter of the 12^th c.; second is Gerard of Cremona’s translation of an Arabic version; and last is the work of an anonymous translator attested by Toledo’s Bibl. Capit. 17.14 (Minio-Paluello 1952). 

Cassiodorus and Isidore never refer to a Boethian translation of the Posterior Analytics, and no one before the middle of the 12^th c. seems to be familiar with any Latin version of this Aristotelian work. In 1159 John of Salisbury, however, who was a student of Theodoric of Chartres and his successor as bishop of that city, records the existence of a Latin translation of the Posterior Analytics in his Metalogicon: there he claims the contents of the treatise were hard to understand and of little interest since only a few philosophers are concerned with mathematical proofs; for these reasons, the text had become corrupted and riddled with errors of transcription, while the examples given were obsolete and thus useless (4.6, 171 Webb = PL 199: 919).  This passage proves that a Latin version of the Posterior Analytics decidedly older than the Metalogicon once existed, but none of its elements allow us to identify the work’s author.  Robert de Torigny, abbot of Mont-Saint-Michel, after mentioning the version of James of Venice, mentions another older Latin version (c. 1157-1169, Chron., 114 Howlett, 177 Delisle, 489 Bethmann = PL 160: 443-4), but makes no explicit reference to Boethius (Haskins 1914).

Toledo’s Biblioteca Capitular 17.14, a manuscript from Gaul dating to the start of the 13^th c., contains three Latin translations of the Posterior Analytics and a Latin translation of Themistius’ commentary on the same work.  The second of these is the one by James of Venice, while the third is Gerard of Cremona’s rendering of the Arabic version; the first is anonymous, but is preceded by a preface in which the translator claims to have undertaken the work because Boethius’ Latin translation was incomplete and the fragments he discovered were barely comprehensible owing to a corrupt textual transmission.  Some Latin codices of the Posterior Analytics do, in fact, report the text in a partial or seriously defective manner: in places where the source text left out a Greek word, they sometimes insert the designation “Grecum” (e.g. Avranches 227, 12^th c. ex.); elsewhere, they attempt Latin transliterations that are utterly incomprehensible (e.g. Ambrosiana R. 55 sup., 13^th c.).  The evidence nevertheless remains insufficient to attribute these fragments of translation to Boethius (Ebbesen 1973).

The copyist of Toledo’s ms. 17.14 is not completely right: the Latin versions presented by our three translators contain frequent errors and omissions.  The codex includes many interlinear and marginal glosses from later hands that correct, augment, or occasionally clarify the sense of the text.  Within these notes, two variants are expressly attributed to Boethius: one is attached to the anonymous translation, the other to that of Gerard of Cremona.  Since these two variants offer a version of the text identical to that of James of Venice, however, it seems likely that the author of the corrections has erroneously credited Boethius with the other translator’s words (Minio-Paluello 1968). [M. Ferroni; tr. C. L. Caterine]