THE METAPHYSICS OF SIMILARITY AND ANALOGICAL REASONING

Abstract

This work introduces the importance of similarity and analogy to philosophy, argues that analogy should be seen as “similarity based reasoning,” overviews different philosophical discussions to illustrate the scope of similarity-based reasoning, and introduces the assumptions for similarity-based reasoning that form the central topics of the present work. It demonstrates that approaches that reduce or identify relations to non-relational ontological categories fail primarily through the strategy of seeking truthmakers for relational claims. It takes up the related problems of co-mannered relations, substitution instances, individuating relations and similarities. It attacks the notion that substitution instances provide a non-relational account of relations and provides a proof that similarity is a more fundamental concept than substitution. Given, however, that similarity is relational, it argues relations are non-reducible to non-relational entities, but that if given the notion of relating in general and of similarity, one can construct a non-reductive theory of relations that can individuate all relations, including similarity relations themselves. This provides a workable theory of relations, but does not solve problems related to the epistemology of relations and similarity. Sanskritic debates concerning the metaphysical nature of similarity and its knowability are explored through contrasting the views of four Indian schools of thought: the Buddhist, Nyāya-Vaiśeṣika, Bhaṭṭa Mīmāṃsāka, and Prābhākara Mīmāṃsā. Within the Islamic tradition, analogy is rejected as a valid tool for legal reasoning by the Ẓāhirī school, and this is contrasted with the very tempered defense by the Shāfiʿī school. The Islamic debates bring out more clearly the hermeneutical challenges, but it is argued that these challenges of interpretation are bound up in the challenges of the epistemology of relations and similarities. The work concludes that an epistemic virtue-theoretic account can help us better understand how analogical arguments can be true and non-vacuous, and argues we should cultivate the virtue of similarity and relation sensitivity.