Reference edition:

Aristoteles Latinus I 1-5, Categoriae uel praedicamenta. Translatio Boethii - editio composita, Translatio Guillelmi de Meorbeka, Lemmata e Simplici commentario decerpta, Pseudo-Augustini Paraphrasis Themistiana, edidit L. Minio Paluello, Bruges-Paris 1961, 45-79 (Corpus Philosophorum Medii Aeui).

Nobody has ever questioned whether Boethius actually translated and commented on Aristotle’s Categories around 510 AD: he personally refers to his translation and commentary numerous times, and various authors contemporary with or slightly later than Boethius mention the existence of these two works, sometimes even citing them (Cassiodorus, Variae 1.45.4, Inst. 2.18; Anecdoton Holderi, Usener: 4).  For a long time, however, editors have given Boethius full credit as author of a “composite edition” of the Categoriae; this text, attested by numerous manuscripts postdating the 10^th c., combines passages safely attributed to Boethius with others by a second translator.  This composite translation was printed in the editio princeps of Boethius’ assembled works (Boethii Opera, impressa Venetiis per Ioannem de Forlivio et Gregorium fratres, 1491), then in all subsequent editions.  Included in the latter group was the the 1546 printing of Heinrich Loriti (Glareanus), which was the ultimate source of the text Migne used in compiling Patrologia Latina 64 (Minio Paluello 1943).

Manuscripts containing the Latin translation of the Categories can be divided into two groups: 1) those that include both Boethius’ Commentarius and the fragmented lemmata of Aristotle’s text, and 2) those that present the textus continuus of the translation without any commentary.  It is undisputed that Boethius translated the lemmata included in the Commentarius.  These consist of Latin renderings of about two thirds of Aristotle’s original Greek, namely those passages that Boethius intended to discuss: sections that did not interest him were simply omitted.  The selective nature of these lemmata makes it impossible to reconstruct a complete translation of Aristotle’s text from them.  The manuscripts that report the textus continuus without any commentary can themselves be subdivided into two groups: 1) the large number of codices containing the “composite edition,” a group that enjoyed wide diffusion from the 10^th c. onward; and 2) a group of three manuscripts, all dating to the 11^th c., that contain a Latin translation very similar to the “composite edition” in sections that coincide with the Commentarius’ lemmata, but that differ in the other parts of the text (viz. Einsiedeln 324; Paris Bibl. nat. lat. 1788; Marciano Z.L. 497; Arras 862 [343]).  It seems clear that this latter version was written entirely by Boethius: it coincides with the text cited by Cassiodorus (Inst. 2.10 = Mynors: 114, 6-9) and Isidore of Seville (PL 82: 144), and scholars have observed that its syntax and language mirror the usage found in other translations by Boethius whose authorship is undisputed (Hadot 1960).

It seems likely that the anonymous author of the “composite edition” possessed a copy of Boethius’ Commentarius and a Greek manuscript of Aristotle’s Categories.  Presumably driven by a desire to reconstruct the Latin textus continuus of the work (which he did not possess), he will have first extracted and organized Boethius’ lemmata, then filled in the gaps with a new translation from the Greek; according to this theory, the only sections of the “composite edition” written by Boethius will be those drawn from the lemmata of his commentary (Minio Paluello 1943).  Another theory, however, holds that the anonymous medieval translator possessed a badly damaged codex of Boethius’ textus continuus and that he will have corrected as much as was in his possession, then filled in any gaps with a new translation from the Greek in order to produce a usable copy (Minio Paluello 1961). 

Between the 6^th and 9^th c., the only commonly known sections of Boethius’ Categoriae were those cited by Cassiodorus and Isidore; starting in the 10^th c., however, philosophers once again began to study this Aristotelian work.  It seems possible that their renewed interest in argument, together with the scarce availability of Boethius’ textus continuus, may have been one of the reasons for the new translation: indeed, since we possess no attestations of the “composite edition” prior to the 10^th c. and since the lexical profile of this text is quite similar to the glosses of Boethius’ Opuscula sacra that also are known to have been compiled in that century, it is possible that our anonymous medieval translator was active at that time.

It is also possible that Boethius produced both translations, but at two distinct times. On this model, the rougher sections of the “composite edition” are the product of Boethius’ first draft, which he will later have corrected and improved after consulting another Greek manuscript: indeed, certain terminological differences between the two versions find parallel in Boethius’ two translations of the Topics and the Peri hermeneias. In short, it is largely unclear how or why these two texts were brought together (Brams 2003).

The “composite edition” actually enjoyed a wide diffusion: it was known to and used by Notker of San Gallo, Abelard, Albert the Great, and Peter of Spain (Minio Paluello 1943).  When Lorenzo Minio Paluello published it in the 1961 collection Aristoteles Latinus (1.1-5), he employed three different fonts: romans are used for the sections of the text identified as lemmata of the commentary; majuscule is used for parts added by the hypothetical second translator; and bolded text is used for individual expressions that belong to the lemmata, but that were modified in some way.  Lastly, asterisks indicate points at which there are omissions in the “composite edition” relative to Boethius’ textus continuus.  [M. Ferroni; tr. C. L. Caterine].