The open core of the strengthened exclusion argument: can role functionalism vindicate mental causation against Parthood, Nonredundancy, and the core/total realizer demarcation?

Rellihan's strengthened exclusion argument (Synthese 198(7):6631–6659, online 2019) reconstructs Kim's causal exclusion argument so that, restricted to functionalist forms of nonreductive physicalism, it requires neither Kim's contested Exclusion principle nor a blanket ban on higher-level causation. Two premises suffice: Parthood — for a role-functionally realized mental event $M$ with total physical realizer $P_M$ , the sufficient physical cause $P_C$ of a characteristic effect is a proper part of $P_M$ — and Nonredundancy — no proper part of a causally sufficient event, or of its physical supervenience base, is itself causally sufficient for the same effect at the same time. From these, the epiphenomenality of functional mental properties follows under both production-based and dependence-based (counterfactual, interventionist) conceptions of causation. The argument leaves open a precise, connected family of questions that jointly decide the fate of role functionalism: whether Parthood admits principled or actual counterexamples; whether the hidden realizer identity (2.5) $P_M=P_C$ can be defended without abandoning the core/total realizer distinction; whether dependence-based accounts of causation can be refined to vindicate functional causation in the edge cases; which higher-level properties admit the core/total realization structure the argument presupposes; and whether Nonredundancy can be refined or rejected. No resolution of any of these questions appears in the subsequent literature; the problem is open.

Background

Mental causation — the question of how mental states can cause physical behavior — generates the exclusion problem when combined with physicalism. In Kim's exclusion argument (Kim 1998), physicalism, the causal closure of the physical (every physical event that has a cause has a sufficient physical cause), the distinctness of mental from physical properties, and an Exclusion principle forbidding systematic overdetermination jointly threaten to make mental properties epiphenomenal, that is, causally inert: the physical cause always preempts the mental one. A key step is often left implicit: identifying the mental event's minimal physical supervenience base $P_M$ (the smallest physical condition minimally sufficient for the mental event $M$ to occur) with the sufficient physical cause $P_C$ of the downstream physical effect $P^{*}$ . Rellihan labels this hidden identity premise (2.5). Most nonreductive physicalists adopt role functionalism: a mental property is a second-order property — the property of having some first-order state that occupies a given causal role (say, the pain role) — which is what makes mental properties multiply realizable by different physical states. Such properties are realized by physical states. Shoemaker's realization theory (Shoemaker 2007) distinguishes, for a role-functional property, the core realizer (the first-order state that occupies the role, e.g. C-fiber firing) from the total realizer (the core realizer plus all the background conditions that enable the core to occupy that role). Because a characteristic effect is typically produced by the core realizer together with only some of those background conditions, the total realizer contains conjuncts that are causally redundant relative to a given effect. Rellihan exploits exactly this structure. Parthood says: if $M$ , together with background conditions, is causally sufficient for a characteristic physical effect $P^{*}$ , then some physical event $P_C$ that is a proper part of the total realizer $P_M$ is causally sufficient for $P^{*}$ . Nonredundancy says: if an event is causally sufficient for an effect, then no proper part of it, and no proper part of its physical supervenience base, is itself causally sufficient for that effect at that time. Together they yield a short reductio of the assumption that the functionally realized $M$ is causally sufficient for $P^{*}$ , without any appeal to Kim's Exclusion principle. The argument is deliberately adapted to both production-based conceptions of causation (on which causes produce or bring about effects) and dependence-based conceptions (on which causes are what their effects counterfactually or interventionistically depend on): Rellihan argues that Lewis-style counterfactual dependence (David Lewis's theory of causation as chains of counterfactual dependence between distinct events), Loewer-style anti-preemption conditions (Barry Loewer's proposal to supplement the counterfactual test with a condition ensuring that mental events are not preempted as causes of their effects by their own physical realizers), and Woodwardian interventionist tests (Woodward 2015) all select the proper part $P_C$ rather than $M$ as the cause once $P_C\subset P_M$ . Unlike Kim's original argument, this one does not generalize to all higher-level properties: it applies only to properties whose realization admits the core/total decomposition with causally redundant conjuncts, leaving intrinsic or dispositional properties such as fragility exempt — a restriction whose principled basis is itself unsettled. If the argument stands, the functionalist variety of nonreductive physicalism — adopted largely on the strength of multiple-realizability arguments — faces epiphenomenalism for mental properties; if it can be answered at any of its load-bearing points, role-functional mental causation is vindicated.

Problem Statement

The problem left open by Rellihan's strengthened exclusion argument is whether role functionalism — the view that a mental property is the second-order property of having some first-order physical state that occupies a given causal role — can be vindicated against it. Throughout, $M$ is a role-functionally realized mental event; $P_M$ is its total physical realizer, that is, the core realizer (the first-order physical state that occupies the role, such as C-fiber firing) together with the background conditions that enable that state to occupy the role; $P_C$ is a sufficient physical cause of one of $M$ 's characteristic downstream physical effects; and (2.5) is the identity premise $P_M=P_C$ , the hidden step in Kim's original reconstruction of the exclusion argument. At the level of the load-bearing premises, the argument raises the following questions, which the paper leaves open or contested: (1) Parthood — must role-functional realizers, in principle or as a matter of actual cognitive and neural architecture, typically admit a proper-part sufficient physical cause $P_C$ for characteristic effects, or are there principled or actual counterexamples in which $P_C=P_M$ or no proper-part sufficient cause exists? (2) Realizer identity — can defenders of role functionalism supply principled grounds for the identity (2.5) $P_M=P_C$ in relevant cases, without abandoning the core/total realizer distinction or conceding other functionalist commitments? (3) Dependence-based rescue — can counterfactual-dependence or interventionist accounts of causation be refined (concerning comparative similarity in counterfactual analysis — which counterfactual worlds count as closest to the actual world when a counterfactual is evaluated — the admissibility of interventions that hold complex supervenience bases fixed, and the treatment of redundancy ranges, i.e. ranges over which a candidate cause can vary without any change in the effect; any one of these concerns may be engaged independently, but a full resolution of this sub-question must settle each of the three) so as to vindicate the causal efficacy of functional properties precisely in the edge cases where $P_C\subset P_M$ ? (4) Demarcation — precisely which higher-level properties are analyzable into core and total realizers and thus susceptible to the functional exclusion, and which are intrinsic or dispositional (like fragility) and exempt, on what principled basis? (5) Nonredundancy — can Nonredundancy be refined (regarding the nomological contingency of background conditions, preemption — one cause's getting in ahead of another that was independently poised to produce the same effect — and complex joint causation; any one of these concerns may be engaged independently, but a full resolution of this sub-question must settle each of the three) or rejected, so as to block the reductio without reinstating the premises the strengthened argument was designed to avoid? These questions are load-bearing points of a single reductio; a complete answer to any one of them that settles whether functional mental properties are causally efficacious — for example, a refutation of a load-bearing premise — resolves the task, and an answer that leaves that question open is a partial contribution toward it.

References

Matthew Rellihan, 'Strengthening the exclusion argument', Synthese 198(7):6631–6659 (2021; first online 2019), DOI: 10.1007/s11229-019-02481-6, https://doi.org/10.1007/s11229-019-02481-6
Matthew Rellihan, 'Functional Properties are Epiphenomenal', Philosophia 48(3):1171–1195 (2020), DOI: 10.1007/s11406-019-00118-z, https://doi.org/10.1007/s11406-019-00118-z
Matthew Rellihan, 'An equivocation in the simple argument for downward causation', Thought: A Journal of Philosophy 10(4):249–256 (2021), DOI: 10.1002/tht3.502, https://doi.org/10.1002/tht3.502
Thomas Krödel, 'Removing Realizers', Thought: A Journal of Philosophy 11(3):150–156 (2022), DOI: 10.5840/tht202392720, https://doi.org/10.5840/tht202392720
Matthew Rellihan, 'A familiar dilemma for the subset theory of realization', Analytic Philosophy 64(1):68–90 (online 2021; issue 2023), DOI: 10.1111/phib.12246, https://doi.org/10.1111/phib.12246
Matthew Rellihan, 'Functionalism, interventionism, and higher-order causation', Synthese 203(3):91 (2024), DOI: 10.1007/s11229-024-04500-7, https://doi.org/10.1007/s11229-024-04500-7
Jaegwon Kim, 'Mind in a Physical World: An Essay on the Mind-Body Problem and Mental Causation', MIT Press (1998), DOI: 10.7551/mitpress/4629.001.0001, https://doi.org/10.7551/mitpress/4629.001.0001
Sydney Shoemaker, 'Physical Realization', Oxford University Press (2007), DOI: 10.1093/acprof:oso/9780199214396.001.0001, https://doi.org/10.1093/acprof:oso/9780199214396.001.0001
James Woodward, 'Interventionism and Causal Exclusion', Philosophy and Phenomenological Research 91(2):303–347 (2015; first online 2014), DOI: 10.1111/phpr.12095, https://doi.org/10.1111/phpr.12095
Jessica Wilson, 'Non-reductive Realization and the Powers-based Subset Strategy', The Monist 94(1):121–154 (2011), DOI: 10.5840/monist20119417, https://doi.org/10.5840/monist20119417
Vera Hoffmann-Kolss, 'Interventionist Causal Exclusion and the Challenge of Mixed Models', in Levels of Explanation, Oxford University Press, pp. 136–152 (2024), DOI: 10.1093/oso/9780192862945.003.0006, https://doi.org/10.1093/oso/9780192862945.003.0006
Haicheng Zhao, 'Overdetermination, Causal Exclusion, and the Insufficiency of Mental Causation', Philosophia 52(4):1165–1184 (2024), DOI: 10.1007/s11406-024-00772-y, https://doi.org/10.1007/s11406-024-00772-y