Three Problems for the Mutual Manipulability Account of Constitutive Relevance in Mechanisms

In this article, I present two conceptual problems for Craver's mutual manipulability account of constitutive relevance in mechanisms. First, constitutive relevance threatens to imply causal relevance despite Craver (and Bechtel)'s claim that they are strictly distinct. Second, if (as is intuitively appealing) parthood is defined in terms of spatio-temporal inclusion, then the mutual manipulability account is prone to counterexamples, as I show by a case of endosymbiosis. I also present a methodological problem (a case of experimental underdetermination) and formulate two partial, but fallible solutions based on the notions of parthood and synchronicity. 1 Introduction 2 Mechanisms 3 Intralevel Causal Relations and Invariance 4 Interlevel Constitutive Relations and Mutual Manipulability 5 Doesn't Constitutive Relevance Entail Causal Relevance? 6 Parthood and Endosymbiosis 7 Possible Ways Out 8 Interlevel Experiments 9 The Problem of Underdetermination 10 Solving the Problem of Underdetermination: Parthood 11 Synchronicity, the Etiological Nature of Experimental Apparatus, and Causal-Constitutive Propagation 12 Pragmatic Interests and the Boundaries of Mechanisms 13 Concluding Remarks Appendix