Using Stylometry to Model Transmission of Arabic Wisdom Literature in Medieval Europe: the Case of the Bocados de oro

David Joseph Wrisley
dwrisley@gmail.com

Comunicación larga
Estilística y estilometría


Using Stylometry to Model Transmission of Arabic Wisdom Literature in Medieval Europe: the Case of the Bocados de oro One instance of knowledge transfer from the Islamic world into medieval Europe that remains understudied is a textual compilation of the lives and sayings of mostly Hellenic philosophers composed in Arabic by al-Mubashshir Ibn Fatik, known as the Mukhtār al-Ḥikamwa-maḥāsin al-kalim. Unlike the other Hellenic gnomic collections compiled in Baghdad over the period of the eighth to tenth century (Gutas), al-Mubashshir Ibn Fātik was active in Cairo in the second half of the eleventh century. This means that the text’s genesis is coterminous with the height of Fatimid power and Shi’i Ismaili learning, an influence that emerges prominently in different parts of the compilation. This text circulated in Iberia and was translated into Castilian in the mid-thirteenth century, a fact which facilitated its widespread reception in Europe with subsequent translations into Latin in the thirteenth century, and then into French, Occitan and English all in the fifteenth century.

The significance of studying the tradition should be obvious: it illustrates the transmission of one strand of Greco-Arabica—Mediterranean political and philosophical thought collated by Muslim scholars—into Western Europe. The number of manuscript witnesses of the translations in all languages is considerable—about one hundred twenty at the time of writing—with about seventy in middle French alone. Outside of Hispanist circles, where the MHMK is a well-known influence on thirteenth-century Castilian literature, through the text known as the Bocados de oro, the text’s influence in Italy, the Empire and northwestern Europe is not well documented. This article attempts to show how digital research methods can help us to explore such uncharted domain.

I will argue that computational stylistics facilitate exploration of such a corpus at the formative stages of a larger research project. Using a corpus of some 60+ Castilian texts, 240+ medieval Latin texts and 360+ medieval French texts, consensus-tree stylometric analyses were performed using the Stylo package in R (Eder et al). These analyses were carried out in order to identify the particular linguistic communities bearing a similar lexical imprint, the results of which I claim are heuristically suggestive of the communities in which these translations are likely to have emerged. This research generates new viewpoints from which to consider contexts of textual transmission of works of Arabic origin in medieval Iberia and beyond. Such communities were, however, not monolingual. The bilingual interface with Latin in both the Iberian and French corpus pose certain methodological problems which will be discussed in the paper. I will engage with some of previous hypotheses about authorship, dating and provenance, but I will also point out what we do not yet know about the MHMK tradition and how stylometry points out new paths of inquiry for me.