Marie-Claire Beaulieu, J. Matthew Harrington, and Bridget Almas
(additional presenters: Marie-Claire Beaulieu, Bridget Almas)
Syntactic structures are a key element of the rhetorical force and nuance of texts, and the objective interpretation and comparison of these structures presents a crucial opportunity for evolving the discipline. Freed from the limitations of print media and connoisseurship, new digital tools, methods and vocabularies enable us not only to accumulate analytical data on the texts but also to systematically collect and describe provenance information about the data underlying a scholars’ arguments. This metadata ensures the reproducibility of the scholarly analyses, down to the smallest detail.
The Satires of Juvenal illustrate the value of a structural analysis of massively parallel constructions as an integral aspect of the author's argumentation: a correlation of syntax with sense. Such a literary argument relies on digital tools for its demonstration and its evaluation. Further, a syntactic analysis of a text can be linked to the precise version of the digital transcription on which it was based, which in turn can be linked to a digital image of the physical manuscript or inscription. We can also record any decisions made and algorithms used during tokenization of the individual words and identification of the syntactic units that might impact the outcome of the analysis. Application of linked data best practices and standard vocabularies and data models, such as Open Annotation and PROV facilitate the shareability of our data, allowing for aggregation of distributed data sets, and for new knowledge to emerge as we can now view analyses that were once isolated within a larger context and compare them objectively.
One such analysis is currently being performed at Tufts University, as teams of students and scholars collaborate on producing digital transcriptions, editions, and annotations of Greek funerary inscriptions. Each text is mapped to the monument it is inscribed on in order to justify editorial choices and provide data on the layout of the text. In addition, a full morphological annotation of each text is performed using the Treebanking module. The resulting data gives unprecedented insight into the inner workings of each text and supports broad corpus analysis. We can now justify or disprove scholarly claims about the poetics of funerary inscriptions and compare this data with the remainder of the manuscript tradition.
This model of scholarship enables those with varying degrees of expertise to engage the materials and evaluate the validity of the claims, while collaborating in the creation of a new level of textual analysis.