Date

4-30-2020

Description

A current debate in the philosophy of religion ("PoR") is about future routes for scholarship (see Bilimoria and Irvine 2010; Crockett, Putt, and Robbins 2014; Wildman 2010). Yet, while many scholars have assessed where the field should expand, few have addressed the disciplines' modern history. Making use of new methods in the digital humanities, this project proposes a history of the field based on empirical evidence, a meaningful foundation for future growth. "Mapping the History" utilizes various computational techniques to process twenty-thousand research articles from fifteen journals in PoR and religious studies, spanning a period of over ninety years (from 1921 to 2015). Machine-learning algorithms (namely, Latent Dirichlet Allocation and k-means clustering) create groups of related articles. These groups exhibit topics that have historically comprised PoR. This project then links these topics to associated bibliographic data, establishing authorial networks, patterns of semantic change (through word-vectoring), and even yearly shifts in the discipline's focus. The aim of "Mapping the History" is two-fold. This project's first iteration, presented at the American Academy of Religion (under the title "Analyzing Philosophy of Religion Journals via Digital Humanities"), sought in large part to demonstrate a successful application of digital methods to philosophical source material. Building upon that, this second iteration is admittedly more ambitious: No longer merely methodological, it describes the field's recent history and suggests that analytic approaches to Christian source texts have become the discipline's established orthodoxy.

Share

COinS
 
Apr 30th, 12:00 AM

Mapping the Modern History of Philosophy of Religion with Machine Learning

A current debate in the philosophy of religion ("PoR") is about future routes for scholarship (see Bilimoria and Irvine 2010; Crockett, Putt, and Robbins 2014; Wildman 2010). Yet, while many scholars have assessed where the field should expand, few have addressed the disciplines' modern history. Making use of new methods in the digital humanities, this project proposes a history of the field based on empirical evidence, a meaningful foundation for future growth. "Mapping the History" utilizes various computational techniques to process twenty-thousand research articles from fifteen journals in PoR and religious studies, spanning a period of over ninety years (from 1921 to 2015). Machine-learning algorithms (namely, Latent Dirichlet Allocation and k-means clustering) create groups of related articles. These groups exhibit topics that have historically comprised PoR. This project then links these topics to associated bibliographic data, establishing authorial networks, patterns of semantic change (through word-vectoring), and even yearly shifts in the discipline's focus. The aim of "Mapping the History" is two-fold. This project's first iteration, presented at the American Academy of Religion (under the title "Analyzing Philosophy of Religion Journals via Digital Humanities"), sought in large part to demonstrate a successful application of digital methods to philosophical source material. Building upon that, this second iteration is admittedly more ambitious: No longer merely methodological, it describes the field's recent history and suggests that analytic approaches to Christian source texts have become the discipline's established orthodoxy.