Aristoteles Latinus III 1-4, Analytica Priora. Translatio Boethii (recensiones duae), Translatio Anonyma, Pseudo-Philoponi Aliorumque Scholia, Specimina Translationum Recentiorum, edidit L. Minio-Paluello, Bruges-Paris 1962, 5-139 (Corpus Philosophorum Medii Aeui).
The Prior Analytics aim to describe a useful means of guaranteeing the truth of a conclusion that is based on the assumption of known premises; in a strict sense, “analytical” functions are to be identified with the mechanism for reducing any deductive process to a syllogism. The work comprises two books. Book 1 is divided into three sections: the first (1.1-26) treats the structure of syllogisms; the second (1.27-30) explains what sort of facts must be assumed so that one can come to a syllogistic conclusion about the problem that has been posed; the third (1.32-54) discusses the problem of reducing any deductive process to a syllogistic form, taking account of the errors that may arise from using ordinary language in the assumption of one’s premises. Book 2, while presenting a less homogenous structure, can likewise be divided into three parts: the first (2.1-15) deals with the function of syllogistic figures that deduce new conclusions from premises that are based on a set consequence; the second (2.16-21) contains a treatment of certain fallacies, namely those concerned with begging the question, the assumption of false causes, and certain dialectic processes; the third (2.23-7) addresses the problem of reducing certain types of argument to a syllogistic form, e.g. induction, example, the apagogic process, objection, and enthymeme. It seems that nobody had translated the Prior Analytics into Latin before Boethius: indeed, the version of Vettius Praetextatus was not a translation of Aristotle, but actually of Themistius (Meiser 1877). Boethius twice makes explicit reference to his translation of the Prior Analytics: first in his commentary on Cicero’s Topica (PL 64 1047B), then in the De differentiis topicis (PL LXIV 1185A). Even so, Cassiodorus does not include it when he lists the works of Boethius that are concerned with the logical disciplines, and it seems that Boethius’ translation was practically unknown before the beginning of the 12th c. (Minio-Paluello 1962). Analysis of the codices of ‘Aristoteles Latinus’ has brought to light the existence of two remarkably different redactions of at least one part of the Prior Analytics: the text offered by the majority of the codices has been called “common” or “vulgate;” the oldest two codices of the “common” redaction are the Florentinus (Florence, Bibl. Nat. Centr. Conv. Soppr. I.VI.34; from the middle of the 12th c.) and the Mediolanensis (Milan, Bibl. Ambros. I.195.inf; from the second half of the 12th c.). Another codex, Carnutensis 397, which was destroyed in 1944, but survives as a microfilm in numerous libraries, is the only version that contains a text of the Prior Analytics markedly different from the common transmission (viz. at folios 296r.-318v.); many of the differences between the two redactions can be explained by the existence of two separate Greek originals (Minio-Paluello 1954). Both redactions were already in circulation by the middle of the 12th c., and many codices from later periods offer versions of the text that are contaminated by both; this has produced an intricate textual tradition that is hard to reconstruct. Despite their differences, however, it is clear that the two texts were not independent translations: it is highly likely that the “common” version postdates that of Carnutensis 397 and is the result of a revision to that text undertaken on the basis of a Greek version that was at various points different from the source of this latter translation. The “common” version is consequently the more accurate: it is closer to the original Greek, more careful in its rendering of particulars, and almost always better from an interpretive point of view (Minio-Paluello 1954). [M. Ferroni; tr. C. L. Caterine].