본문 바로가기
  • Home

Artificial Religious Intelligence Simulation: Comparative Study of the Stark-Bainbridge Model and Durkheimian Model of Religious Cognition

  • Journal of Humanities, Seoul National University
  • 2019, 76(2), pp.271-312
  • DOI : 10.17326/jhsnu.76.2.201905.271
  • Publisher : Institute of Humanities, Seoul National University
  • Research Area : Humanities > Other Humanities
  • Received : April 10, 2019
  • Accepted : May 22, 2019
  • Published : May 31, 2019

Wook Joo Park 1

1연세대학교

Accredited

ABSTRACT

One of the important contemporary questions about the possibilities and limits of artificial intelligence is concerned with human religiosity. “Can artificial intelligence simulate religious faith?” The present study investigates the implications of a machine learning experiment of William Sims Bainbridge, an outstanding researcher in the field of the sociology of religion and cognitive science. He attempted to simulate human religiosity and faith in supernatural beings in 2006. The focus is on three factors: his social cognitive theory to explain human religiosity and faith, a scenario for the actualization of this theory, and a mathematical-statistical strategy and its principles applied to experiment. In this comparative analysis of the Stark-Bainbridge model of religion and religious cognition and that of Durkheim, it is demonstrated that the insights of Stark and Bainbridge found in Bainbridge’s artificial intelligence simulation would likely be acknowledged as valid from the perspective of the Durkheimian methodology of religious studies, despite some significant differences between them. Most importantly, it seems certain that what Stark-Bainbridge and the Durkheimian model of religious cognition have in common is that they locate the origin of primitive religiosity in society, in continuous social processes. It also seems certain that they both translate the social processes into the categorizing norms of religious cognition so as to illuminate the religious nature of human kind.

Citation status

* References for papers published after 2023 are currently being built.