BolVis: Visualization for Text-based Research in Philosophy

By Pauline van Wierst, Steven Hofstede, Yvette Oortwijn, Thom Castermans, Rob Koopman, Shenghui Wang, Michel A. Westenberg, and Arianna Betti

Traditional research in philosophy consists for a large part in conceptual analysis and close reading of texts. This is a precise but time-consuming approach, in which the researcher focuses on one particular text passage or one philosophical concept from one or more works of an author. In this paper, authors present BolVis, a visualization tool for text-based research in philosophy. BolVis allows researchers to determine quickly which parts of a text corpus are most relevant for their research by performing a semantic similarity search on words, sentences, and passages. It supports activities such as filtering, exploring the semantic context, comparing, performing close reading on selected passages, et cetera. The approach enables in-depth analysis of texts at a significantly greater scale than is possible by traditional means, thereby allowing researchers to gain in speed without compromising on precision. The authors demonstrate the usefulness of BolVis by applying it to a corpus consisting of about 11,000 pages of the writings of the Bohemian polymath Bernard Bolzano (1781–1848). Their use case addresses an open question about Bolzano’s ideas concerning size equality for sets of natural numbers, and show that the use of BolVis enabled them to find (at least a significant part of) the reason why he came to accept one-to-one correspondence as a sufficient criterion for size equality.

Read the paper