A Methodological Approach to Anonymously Transmitted Medieval Translations of Philosophical and Scientific Texts. The Case of Bartholomew of Messina

The important change in the approach to philosophy in Western Europe from the late 12th century onward was to a large degree caused by the new availability of Aristotle's works and other treatises from antiquity and the early medieval period translated from Greek into Latin. These texts were until that time unknown to the West. The contribution by some translators like Burgundio of Pisa and William of Moerbeke in the 12th and 13th century is well-known and often studied. Yet, many translations were allegedly made by various unknown scholars. The relatively high number of anonymous translators is surprising, since contemporary sources complain that there were very few people able to translate Greek texts.Often the manuscripts that transmit the texts provide very little clues in the titles or colophons to identify their translators. For authoritative texts like those of Aristotle the translator mostly remains invisible. Still, several of them are known because they signed other translations with their names, which allows to stylistically compare the anonymous ones.The translators worked according to a very strict verbum de verbo principle, i.e. that virtually every Greek word should have its counterpart in Latin, even if the result is unidiomatic or difficult to understand. In the middle of last century, Lorenzo Minio-Paluello used this characteristic to develop an identification method based on the standardized rendering of certain particles, conjunctions, and adverbs. He observed that in spite of the similarity in the translators' approach it was possible to distinguish different preferences for each of them. Minio-Paluello's lists of typical renderings of particles, conjunctions, and adverbs have been guidelines for scholars to attribute anonymously transmitted versions to known translators.While Minio-Paluello's method is based on the mechanical renderings of certain words, Fernand Bossier complemented it by giving more attention to the deliberate choices in the translators' vocabulary. He observed that some terms can be considered 'signature translations' for a particular translator, and that the presence of several of these translations can provide additional arguments for an attribution.Although Minio-Paluello and Bossier came to many important results in the field of late-medieval translations of philosophical and technical treatises, their research fundamentally remained intuitive, based on their remarkable knowledge of the subject matter and the close reading of many translations and their Greek originals. It was my aim to look for a more theoretically based approach, and to examine whether it could be useful to ascribe more Latin versions to known translators.I found the main inspiration for my theoretical approach in the study of various forms of equivalence in translation as they are defined by Mona Baker. Modern translation studies often use the concept of 'equivalence', the idea that source text and translation share a relation on some level. I followed Baker's classification of equivalence at various levels and adapted it to the needs of my research. It is my firm belief that the study of the methodology of medieval translators benefits from this more structured approach. In particular, the idea of non-equivalence, i.e. cases where there is no immediate (formal) equivalence between source language and target language, or the existing equivalence is not perceived as such by the translator, was useful for my purpose.Among the unattributed translations, it was conspicuous that the name of Bartholomew of Messina was often considered as the potential translator, but in many cases was rejected. I therefore decided to establish the characteristics of his translation method in greater detail according to Baker's classification of equivalence and to reconsider the arguments brought forward in previous studies.I first examined the Latin version of the Aristotelian De coloribus that was previously thought to be made by Bartholomew. The study of particles confirms the attribution, and several lexical choices in this translation get confirmation from other versions for which there is no doubt that he translated them. As an additional element, the characteristics of Bartholomew's translation can be contrasted with those in a partially preserved contemporary Latin translation made by William of Moerbeke. Although the stylistic features of Bartholomew's and William's works are often claimed to be very similar, there appeared to be enough significant differences to distinguish their translation methods.When I applied my newly established method to other translations that were not attributed or for which the attribution was uncertain, I was confronted with the lack of firmly established conclusions in previous scholarship. It was therefore often necessary to divide the assessment of translations in two stages, a negative one in which I weighed the value of earlier studies and often invalidated their arguments, and a positive one using my additional criteria.Among the translations that were attributed to Bartholomew of Messina, those of the treatises by Hippocrates and Hierocles had previously received little attention. Among the medical works by the former, De humana natura and De natura pueri were certainly translated by Bartholomew. It was also known that the Greek model for his translation is still extant, which provides interesting material for comparison. I suspected that the Latin versions of Hippocrates' Lex and De pharmaciis also came from Bartholomew's pen. To my surprise, Stefania Fortuna had recently reached the same conclusion. The criteria that I had established to identify Bartholomew's style confirmed the existing hypotheses to perfection.Professor Fortuna kindly pointed me to another medical translation, the commentary by John of Alexandria on book VI of Hippocrates' Epidemics. Most of the Greek text was lost, which gives a special status to the Latin translation as the remaining witness for John's complete commentary. Unfortunately, the existing critical edition is not satisfactory and the editor's attribution of the translation to Bartholomew is founded on very shaky arguments. Based on a comparison between all preserved Greek fragments and the Latin translation, I argued that the latter was probably made by Bartholomew of Messina. Some diverging translation choices may point to a stage in his career when his command of the Greek language was less secure.The transmission of Hierocles' hippiatric text poses similar, yet more complex problems. In Greek, his work is fragmentarily preserved in Byzantine collections on horse medicine, none of which is identical to the versions preserved in Latin. Some Latin manuscripts mention Bartholomew as the translator in the incipit of the work. However, it is very likely that this was a secondary addition by later copyists. As for the Latin text of the most widely spread and longer version, most passages show important divergences both with the transmitted Greek text and with Bartholomew's translation method. I hypothesized that this Latin version results from an editorial process, combining material from the Latin veterinary tradition with textual information from Greek sources, potentially translated by Bartholomew of Messina. The find of a passage in another Latin version of Hierocles' text which verbatim mirrors the version ascribed to Bartholomew confirms the hypothesis that the translations also went through an editorial stage with intracultural influence.The following chapters of my dissertation study several translated texts by two important philosophers from Antiquity, viz. Aristotle and Sextus Empiricus.The critical edition of the Rhetorica by Bernd Schneider in the Aristoteles Latinus series contains the translation by William of Moerbeke, and an anonymous translation which is extant in just a few manuscripts. Schneider claims that the anonymous translation is prior to William's, but he remains unclear about its date. Moreover, he hesitates regarding a possible influence from the anonyma on Moerbeke's version.As for the attribution to a translator, Schneider only briefly considers the possibility that it was Bartholomew. I showed that his arguments against Bartholomew are all but sound. Following Baker's classification, I provided several compelling coincidences between Bartholomew's 'signature translations' and the translation method in the Rhetorica anonyma, which must with certainty be ascribed to his activity. However, Schneider's observation that the translator is hesitant and not very competent in rendering the Greek remains valid. It suggests that the Rhetorica must be ranked among Bartholomew's earlier works.Two translations of the pseudo-Aristotelian Rhetorica ad Alexandrum were each preserved in a codex unicus. They are known as the translatio Vaticana and translatio Americana from the locations were these manuscripts are preserved. The translatio Vaticana was studied by Leonhard Dittmeyer in the thirties of last century. He concluded that Bartholomew is not its translator, yet Sicily is its likely place of origin. The translatio Americana shows many characteristics similar to the Vaticana that exclude Bartholomew as its translator. Although scholars claim that the two translations were made independently but based on the same Greek manuscript, there is evidence suggesting that there may have been some mutual influence. Since the Americana also originated in South Italy or Sicily, and this version of the Rhetorica ad Alexandrum can be linked to a Latin translation of Demetrius' De elucutione and possibly also the translatio vetus of the Economica, it suggests that there was a keen interest in Aristotle's practical philosophy in that region, although its influence may have remained reduced.The last Aristotelian translation that I examined is the anonymous translation of De partibus animalium. It is equally preserved in a codex unicus from the Biblioteca Antoniana in Padova (XVII 370) and has not yet been published. Its presence in that particular manuscript must be significant, as the same manuscript is also the unique witness for other translations made by Bartholomew. Pietro Rossi, who studied book I of the translation, consequently explored the possibility that Bartholomew is responsible for the anonymous version of De partibus animalium. His conclusion was negative. However, when I scrutinized his evidence by comparing it to data from books II through IV, the conclusion appeared to be less compelling. Further research yielded a high number of correspondences with Bartholomew's usual vocabulary and his 'signature translations'. The anonyma of De partibus animalium must therefore be counted among Bartholomew of Messina's translations.It is remarkable that Bartholomew apparently translated no other treatise from Aristotle's zoological corpus. However, Pieter De Leemans showed that the anonymous translation of De motu animalium, which is only known through the fragments preserved in Albert the Great's commentary, originated in South Italy. De Leemans also found some correspondences with Bartholomew's vocabulary and with the Rhetorica anonyma. Although the evidence remains inconclusive, it is an attractive hypothesis to attribute the lost version of De motu animalium to Bartholomew.Even more translations of treatises on animals made by Bartholomew might have existed. In an Italian manuscript of Aristotle's zoological corpus in the version by William of Moerbeke, two marginal notes explaining terms that were transliterated from Greek by William coincide with what I have called Bartholomew's 'signature translations'. The origin of the notes cannot be traced, but they must result from the comparison with another source, either with the Greek text, which is hardly likely, or with another - no longer extant - Latin translation. In the latter scenario, Bartholomew potentially also translated the long treatises De historia animalium and De generatione animalium, which were subsequently lost.Finally, the attribution to Bartholomew of Messina of the Latin translation of two works by sceptic author Sextus Empiricus had been considered and rejected by various scholars. Sextus Empiricus' Pyrrhonie informationes is preserved in three manuscripts. Jean Eloy, who edited the text in his MA dissertation on the basis of one manuscript, compared the translations of particles, conjunctions, and adverbs with the characteristics established by Minio-Paluello for several well-known translators from the 12th and the 13th century. He concluded that the translation method is not identical with any of them, although in his classification according to similarities Bartholomew of Messina takes the top spot. I re-examined the translation, both in the edition and in the two manuscripts that Eloy did not use. The number of similarities with Bartholomew's 'signature translations' turned out to be significant, although the Latin version contains an important number of innovations. These may partly be due to the phrasing ant the vocabulary of the Greek original but could also result from an evolving insight into philosophical thought and the Greek language by Bartholomew.Significantly, many of these innovative renderings are also found in the translation of Adversus mathemicos, which is only fragmentarily extant in a badly preserved paper manuscript from the Biblioteca Marciana in Venice. There is little doubt that one and the same translator is responsible for the Latin version of the two treatises by Sextus Empiricus. If the identification with Bartholomew is accepted for the Pyrrhonie informationes, it logically follows that it is also valid for Adversus mathematicos.Although the characteristics of Bartholomew's translation method are clearly recognizable in the Latin version of the sceptic treatises, the translator's interventions are more pronounced than in his other works, especially in the Latin word order. The translator intends to clarify the difficult text to his readers by re-ordering the parts of the Greek sentence or by gathering words that logically or syntactically belong together. The same tendency, which is visible in some of Bartholomew's other translations, is an emphatical feature of his sceptic translations.The divergences in the style and translation method of the latter treatises, which I nevertheless attribute to Bartholomew of Messina, give rise to the methodical question whether these differences still allow the attribution to a single translator, or that two or more should be hypothesized. By comparison with a similar methodological problem and its various conclusions regarding mathematical translations from the 12th century I argued for a balanced view, which does not attribute too much importance to isolated discrepancies and gives room for contingent choices or changes over time in a translator's mind.Finally, I attempted to contextualize the newly attributed translations in the social and political context in which Bartholomew worked, and the apparent impact on his translation project as it was previously conceived, viz. as the translator of pseudo-Aristotelian treatises who filled the space that was left open by the more prestigious and more encompassing project by William of Moerbeke. The greater success of Moerbeke's endeavour was probably supported by the more performant system through which his translations were copied and spread. Accordingly, the last section of my dissertation considers the likelihood that Bartholomew produced translations that have not come down to us. For the sake of comparison, I used probabilistic models from different case studies in book history.My dissertation is completed by (pre-)critical editions of the two Latin versions of Aristotle's De coloribus (by Bartholomew of Messina and by William of Moerbeke), of Bartholomew's translations of the Hippocratic De humana natura and De natura pueri, and of the remaining fragments of Bartholomew's translation of Adversus mathematicos by Sextus Empiricus.

Publication year:2020

Accessibility:Closed