Using Digital Editions to Analyze Iliadic Text Reuse and its Poetic Tradition

Since its foundation, the Homer Multitext project has been creating digital diplomatic editions of manuscripts of the Iliad, editions which record every intentional mark on the manuscript, in order to create a resource to explore larger Homeric questions. Our poster presents one aspect of Homer Multitext research. The authors’ original research involves analyzing the Venetus A to look for multiforms, that is, substitutions for Homeric text that fit both metrically and contextually. The Venetus A, a tenth-century manuscript currently housed in Venice, is the oldest complete manuscript of the Iliad existing today. In addition to the intact poem, the Venetus A includes an abundance of scholia, scholarly notes which comment on the poem. It is in these scholia that we find the evidence for multiforms. Our use of the term "multiform" intentionally follows Albert Lord’s of the Iliad as an orally composed and transmitted poem: "the word multiform is more accurate than variant, because it does not give preference or precedence to any one word or set of words to express an idea; instead it acknowledges that the idea may exist in several forms."  (Lord, 2000). Ultimately, in looking for multiforms, we found a definitive correlation between the appearance of multiforms in the scholia and the reference to one of three Homeric textual editors, all of whom served as the head of the library at Alexandria at various times from the third to second centuries BCE. While it has been previously hypothesized that the work of these three Alexandrian editors, preserved in the Iliadic scholia, is responsible for the current standard canon of available multiforms, the research represents our first steps in using statistics to better understand the transition from an oral tradition to today’s critical print editions.

The first task in our research was to analyze how many multiforms existed in a select portion of the Iliad. Books 18 and 19 of the Iliad served as our sample, and we accessed the manuscript through openly-licensed digital photography. From these photographs we created digital diplomatic editions of the poem and its corresponding scholia using an XML editor and following TEI guidelines. Within our editions we used the TEI element “quote” to identify every instance of text reuse found in the scholia, and then generated a separate table in which we classified the types of text reuse.

 Our classification was modelled on work by Monica Berti who defines “text reuse” as “the meaningful reiteration of text, usually beyond the simple repetition of common language” https://wiki.digitalclassicist.org/Text_Reuse). By assigning distinct URNs to the different types of text reuse in the table and not in our XML mark-up, we applied a unique approach that kept our external analysis separate from our digital edition.

We automatically generated a table displaying the occurrence of multiforms in relation to references to Alexandrian editors in preparation for a chi-square test for independence.

This table shows 34 instances of overlap, in which one of the scholars was named with a multiform. For the chi-square test, we compared these two sets of data, beginning with the null hypothesis that they were independent. Then, having set a significance level of 0.01, we obtained a p-value of 1.2 *10^-42. This p-value shows that the probability of randomly obtaining the results we did is negligibly low. Thus, we were able to reject our null hypothesis and conclude that there is a correlation between the occurrence of a multiform and references to one of the Alexandrian grammarians.

Given the correlation between Alexandrian editors and multiforms attested to in the Venetus A, we conclude that the Alexandrian editors were responsible for setting the canon of multiforms which has been transmitted in this manuscript, and from there to the modern world. In respect to further avenues of research, the strength of this correlation remains to be tested.  Additionally, we have not yet attempted text reuse analysis with the majority of the Venetus A; drastically increasing the sample size will likely yield more insights into the nature of the poem’s transmission and production. The ability to statistically support this conclusion exemplifies the great promise statistical, analytical methods hold for the field of Homeric studies, as well as wider areas in other fields of the humanities.