Miguel Flament
University of Antwerp
In Philosophy of Mind, one way to solve the problem posed by mental causation is to offer a ‘nonreductive materialist’ (NRM) account of how mental properties interact with physical properties. Minimally, this solution holds that (i) mental properties are distinct from physical properties, because (ii) mental properties make a real and independent causal contribution to what happens and yet that (iii) mental properties depend (in a pretty strong and systematic way) on physical properties. In Social Metaphysics, many have argued by analogy that the same solution could be defended to solve the problem posed by social causation (the problem of how social properties can have real causal efficacy). There, a ‘nonreductive individualist’ (NRI) account should “translate” the (NRM) solution and hold that: (i) social properties are distinct from individual properties, because (ii) social properties make a real and independent causal contribution to what happens and yet that (iii) social properties depend (in a pretty strong and systematic way) on individual properties. In that way, or so it is hoped, one can escape in the same fashion the conundrum posed by reductive programs and their epiphenomenal difficulties. This analogy, however, is less innocuous than it seems, and, in this paper, I would like to challenge it and shed some light on a key difference between the two solutions. Both (NRM) and (NRI) are committed to a metaphysical picture of reality according to which different ontological levels occur on a single scale of increasing complexity. I shall argue that if physical, mental, individual, and social properties appear in different positions on a common ontological scale of increasing complexity, then we ought to acknowledge different high order features of each class of properties. A difference between mental and social properties in that scale concerns their locality. It should follow from this that different levels of explanation need to be admitted. By bypassing that key feature, non-reductive accounts of social causation blindly and mistakenly apply a strategy which may be worth pursuing only for mental causation. The structure of my argument is the following. I shall argue that whilst (NRM) has been accused of not being sufficient to handle mental causation, (NRI) is immune to this attack on two grounds: epiphenomenalism of social properties or overdetermination of individual properties can be dealt with more easily. I shall first reconstruct and adapt classic arguments against (NRI) and argue that they fail. I shall then submit that the failure does not rest on a different understanding of what it means for a property to depend on another or of what it means for a property to be causally efficient. Rather, the difference stems from an intrinsic difference between what it takes for a property to be a mental property as opposed to a physical one and what it takes for a property to be a social property as opposed to an individual one.
Chair: Armin Mašala
Time: September 13th, 10:00 – 10:30
Location: SR 1.006, online
SOPhiA 2026 