Macroevolution of European Literature

forthcoming

The MELT project will study the macroevolution of European literature: the overarching patterns of intertextual influences in literary fiction spanning the years 1800–2000 and including all major European languages. This project will rely on the full-text digital library of unprecedented size – HathiTrust, with over 10 million full-text books – as well as on advanced methods of computational text analysis and on novel mathematical models of cultural evolution. My interdisciplinary team, consisting of scholars of digital humanities, computational linguistics, and mathematics, will construct massive dynamic networks of book history, capturing the thematic influences between athors, the emergence of new literary communities (e.g., genres), and their decay. In this way, the MELT will essentially build a gigantic “map” of European literary history, having two dimensions: geographic and temporal. In parallel, we will simulate the evolution of books, creating artificial “authors” powered by large language models. Such simulations will allow testing the robustness of our network construction methods. The study of this massive size and scope, employing a powerful computer cluster, will allow answering many fundamental questions about literature – but also about society and politics, since literature often reflects moods and aspirations of the day. For example, what is the role of demographic turnover in the change of literary topics? How do disruptive events – wars, pandemics, and revolutions – influence the fabric of writing? What makes certain books bestselling and, as a consequence, hyper-influential? Do literary genres gradually loose their appeal due to repetetiveness? Literature occupies a central part of culture, shaping our identities and visions of the future, but our understanding of its dynamics is limited. My expertise in both comparative literature, large-scale text mining, and mathematical modeling will allow leading a team able to answer such questions.