Data-driven Methods in Philosophy

For more information, see: https://www.ozsw.nl/activity/data-driven-methods-in-philosophy-2/

Computational methods have transformed academic research, including in the humanities. Philosophers have been comparatively slow to adopt them, but as contemporary language modelling techniques now enable much more sophisticated analyses, they are seeing increasing interest. We want to explore techniques from the digital humanities, linguistics and AI research (Betz 2022) that can support the study of philosophical and scientific corpora, with applications for philosophy of science (Lean, Rivelli & Pence 2021; Noichl 2023. See also the contributions to Pence & Rivelli 2022), the history of philosophy (Petrovich, Verhaegh, Bös et al. 2024; Verhaegh, Petrovich & Bös forthcoming), and metaphilosophy (Petrovich, 2022).

This activity is built around a ‘hackathon’ – an extended period of collaborative programming and discussion. During the three-day in-person event, the participants develop their own projects, either individually or in small groups. The first two days start with keynote lectures that present state-of-the-art research. In the lead-up to the event, we organize two hybrid seminars, in which participants present recent research to each other, to get an idea of what’s possible in this space. During the seminars, participants brainstorm research ideas and discuss with seminar leaders how to apply digital methods, identify appropriate data sources, and determine which digital skills to develop. During the event, our keynote speakers Charles Pence and Gregor Betz contribute their expertise in argument representation, LLMs, digital methods for history and philosophy of science. They will also be available during the event to discuss research ideas, share practical knowledge, and support the seminar participants.

As an additional help for participants without programming experience or who have not yet used data-driven methods in their research, the organizers prepare coding templates and assist in using LLMs for writing code. More experienced participants can focus on exchanging ideas and developing their own projects. A few weeks after the hackathon, we reconvene in a hybrid event to discuss the results of the projects and avenues for further work.

The aim of the course is to offer an introduction to data-driven methods for philosophy and focuses on participant-designed research projects. At the end of the course, participants:
a) Know examples of state-of-the-art data-driven research methods in philosophy and are in a good position to apply them.
b) Have gained experience in starting their own computational philosophy project, in the best case, the hackathon can be the starting point for a research project in the participants’ domain of expertise.