政治理論における数理モデル分析： 多様性が能力に勝る定理をめぐる論争を題材として

抄録

　Recent epistemic democracy research on political theory has focused on the epistemic effectiveness of the democratic decision-making process, by which the collective intelligence derived is intended to find correct answers. The methodological characteristic of epistemic democracy is that it uses mathematical model analysis as the basis for its normative argument. The problem is that controversy has surrounded the implications of mathematical models built using ideal and normative democratic conditions for real-world political issues.

　The current research presented this problem by introducing the controversy over the "diversity trumps ability" theorem, a collective intelligence model in deliberation. It developed two approaches to alleviating the problem via the theory of epistemic democracy: One is to enhance the internal validity of a model and prepare the necessary conditions for ensuring its external validity through the robustness analysis of mathematical models within the scope of model analysis. The other is to go beyond model analysis to investigate the behaviors of mathematical models in an experimental system and accumulate results as a "library of phenomena." These proposals enable explorations into the range at which model analysis is generalizable to the real world while maintaining the normative nature of political theory.