Abstract:

According to J. Mackie’s celebrated INUS account of causation, a cause is (at least) an Insufficient but Non-redundant part of an Unnecessary but Sufficient condition for the effect. This account underlies Mackie’s useful reconstruction of J.S. Mill’s “canons of induction”, as well as an influential approach to qualitative analysis and causal inference in social science. In this talk I present a dilemma on the requirement of non-redundancy in the INUS account. I will start with examples suggesting that the non-redundancy requirement is too strong. I then consider possible responses to the apparent counterexamples, and argue that in the light of the responses, the non-redundancy requirement (as well as the stronger requirement used in the Boolean approach to causal inference) are too weak. Either way, therefore, the N in INUS is inadequate.