Rethinking interlevel experiments: no remainder from evidence for causal relations

Abstract This paper examines the transformation of Craver’s (2009) mutual manipulability (MM) account into the matched interlevel experiments (MIE) framework (Craver et al., 2021) and argues that it amounts to a theoretical reduction of mechanistic constitutive relations to causal mediation. While the MIE account successfully resolves the incoherence challenge that plagued MM, it does so by eliminating the distinctive theoretical content that constitutive categories were supposed to provide. The processual reframing that enables this solution replaces hierarchical part-whole relationships with temporal causal sequences, changing what mechanistic explanations are understood to accomplish. Drawing on paradigmatic action potential experiments, I demonstrate that practices satisfying MIE’s formal requirements consistently establish causal mediation relationships without requiring constitutive interpretation. I address several theoretical defenses of constitutive categories—including interpretive objections about two types of constitution, arguments for distinctive explanatory value, and appeals to mechanistic levels—showing that none can rescue constitutive distinctiveness once constitution is explicitly identified with causal betweenness. Rather than undermining mechanistic approaches, this analysis suggests that their explanatory power derives from methodological sophistication in investigating complex, multi-scale causal structures rather than from categorically distinct constitutive relationships.