COMputational Analysis of PEripheral Literatures

ongoing

The ability to attend to literary data in a variety of languages is crucial towards an understanding of literary practices across time and traditions. Yet few are the languages endowed with adequate computational resources for literary analysis. Increasing the resources enabling computational literary analysis in lesser-studied traditions is the goal of COMputational Analysis of PEripheral Literatures (COMPEL), focusing on Galician literature, a rich but peripheral tradition at the crossroads of two major romance literatures, Portuguese and Spanish. This will be an opportunity to study which features become canonical in a peripheral tradition compared to major traditions influencing it, with descriptive and predictive models. The project promotes a more inclusive and precise literary history via approaching a tradition little studied using computational means. By concentrating on Galician, a language covered by the European Charter for Regional or Minority Languages, the project contributes to European linguistic diversity goals. Beyond Galician, results are expected to be applicable to comparing core and peripheral literatures in other language contact settings, e.g. the influence of French and Spanish literature in Catalan or French and German literature in Alsatian. The project will focus on two genres, poetry and narrative, and two main Galician literature movements in the 19th century and at the turn of the 20th century; public domain texts will allow us to release project resources under open licenses. We will develop several methods to compare texts across literary traditions: Among others, a supervised learning method (perspectival modeling) based on versification features and content features, carrying out a cross-lingual comparison unattempted in earlier studies. By comparing a regional tradition with two major European ones, COMPEL helps promote a sense of European identity via smarter cultural heritage acess, a current EU priority.