Loading field map
Physicalism holds that everything that exists is physical. Yet our ordinary experience insists that mental states—beliefs, desires, intentions—make a difference to what we do. A headache causes me to take aspirin; a belief that it will rain causes me to carry an umbrella. The tension between these two commitments is the problem of mental causation: if the physical world is causally closed—if every physical event has a sufficient physical cause—then how can mental events cause anything without overdetermining their effects? This question has driven a half-century of intense philosophical debate, producing a sequence of frameworks that each try to reconcile mental efficacy with a physicalist ontology.
The first systematic physicalist answer was the Type Identity Theory, defended by J. J. C. Smart and U. T. Place in the 1950s and 1960s. It claimed that mental states are identical to brain states, not just correlated with them. A pain just is the firing of C-fibers, in the same way that water is H₂O. If mental events are physical events, then mental causation is simply a species of physical causation. The causal-closure problem dissolves: there is no separate mental realm that needs to intrude into the physical order.
Type Identity Theory offered a clean solution, but it faced a powerful objection from Hilary Putnam in the 1960s: multiple realizability. A creature with a radically different physical makeup—a Martian, a robot—could still feel pain, even if its neural hardware was nothing like human C-fibers. If pain can be realized by many different physical states, then pain cannot be identical to any one of them. The identity seemed too narrow. This objection did not refute physicalism, but it forced a search for a more flexible framework.
While Type Identity Theory tried to preserve mental states by identifying them with brain states, Eliminative Materialism, championed by Paul Churchland and Patricia Churchland from the 1960s onward, took a more radical path. It argued that the mental states posited by folk psychology—beliefs, desires, intentions—are theoretical posits of a false theory. Just as we eliminated phlogiston and caloric fluid when better theories came along, we should expect to eliminate mental states as neuroscience matures. If there are no mental states, then there is no problem of mental causation to solve.
Eliminative Materialism differs from Type Identity Theory by denying that mental kinds have any genuine ontological standing. It does not try to fit mental causation into a physical world; it rejects the premise that mental causation is real. This position remains a minority view, but it pressures every other framework to justify why mental states deserve a place in our ontology. The challenge is especially acute for frameworks that treat mental properties as irreducible: if mental properties are not identical to physical properties, why think they are real at all?
Functionalism, developed by Putnam, Jerry Fodor, and others in the 1960s and 1970s, offered a way to preserve mental states while avoiding the narrowness of Type Identity Theory. A mental state is defined by its causal role—its typical causes and effects—rather than by its physical makeup. Pain is whatever state, in any system, that is typically caused by tissue damage and typically produces avoidance behavior. This abstraction accommodates multiple realizability: different physical states can occupy the same functional role in different organisms.
Functionalism seemed to rescue mental causation: mental states cause behavior because their causal roles are defined in terms of physical inputs and outputs. But a new worry emerged. If a mental property is just a functional property, and the real causal work is done by the underlying physical realizer, then the mental property itself may be epiphenomenal—a causally idle label for what the physical hardware does. This is the epiphenomenalism worry that would haunt functionalist and non-reductive frameworks alike. Functionalism coexists with later frameworks as a background assumption about the nature of mental states, but it did not by itself solve the causal-closure problem.
Donald Davidson's Anomalous Monism, introduced in 1970, offered a distinctive synthesis. It accepted token identity—every particular mental event is identical to some particular physical event—but denied type identity: there are no strict psychophysical laws connecting mental and physical types. Mental events are anomalous: they resist capture under the kind of deterministic laws that govern the physical domain. Causation, Davidson argued, requires only token identity, not type-level laws. A mental event can cause a physical event because it is, under its physical description, a physical cause.
Critics objected that Anomalous Monism makes mental properties causally irrelevant. If the causal power of a mental event comes entirely from its physical description, then the mental description—the belief, the desire—does no causal work. The framework preserves mental causation in name but drains it of content. This objection pushed later theorists to look for a way to give mental properties genuine causal relevance without reducing them to physical properties. Anomalous Monism remains a landmark attempt to reconcile mental causation with physicalism, but its type-anomaly thesis is now widely seen as too weak to secure mental efficacy.
A different strategy emerged from the metaphysics of causation itself. Counterfactual Theories of Causation, developed by David Lewis and others from the 1970s onward, define causation in terms of counterfactual dependence: an event c causes an event e if, had c not occurred, e would not have occurred. This approach does not require causation to be backed by strict laws. It opens the door for mental causation even if mental events are not covered by physical laws.
For the mental causation debate, counterfactual theories offer a way to bypass the law-based assumptions that drive the Exclusion Argument. If mental properties can make a counterfactual difference to physical outcomes—if, had the mental state been different, the behavior would have been different—then mental causation is real regardless of whether mental types reduce to physical types. This framework coexists with Non-Reductive Physicalism as a complementary strategy: it provides a metaphysics of causation that does not require reduction. However, it faces its own challenges, including the problem of distinguishing genuine causal dependence from mere correlation and the difficulty of grounding counterfactuals in a physicalist world.
Non-Reductive Physicalism, which became the dominant framework from the 1970s onward, tries to have it both ways: mental properties are real and causally efficacious, yet they are not identical to physical properties. The core idea is supervenience: mental properties supervene on physical properties, meaning that any change in mental properties requires a change in physical properties. Mental properties are realized by physical states but are not reducible to them. This allows multiple realizability while preserving physicalism.
The most serious challenge to Non-Reductive Physicalism is Jaegwon Kim's Exclusion Argument, developed from the 1980s onward. The argument runs: if a mental event M causes a physical event P, and P already has a sufficient physical cause P, then P is overdetermined unless M is identical to P or M is reducible to P*. Overdetermination is implausible for ordinary cases of mental causation—we do not think headaches and C-fiber firings separately cause aspirin-taking. Therefore, mental causation requires either identity or reduction. Non-Reductive Physicalism, which denies both, seems to leave mental properties epiphenomenal.
Responses to the Exclusion Argument have taken several forms. Some philosophers, like Fodor, argue that mental properties are causally relevant because they are realized by physical properties that do the causing; the mental property is not an extra cause but a higher-level description of the same causal process. Others, like Stephen Yablo, appeal to the idea of proportionality: a mental property may be a more proportional cause of an effect than its physical realizer, because it captures the relevant difference-making features. Still others, drawing on counterfactual theories, argue that mental properties make a counterfactual difference even if they are not fundamental causes. The debate remains unresolved, and the Exclusion Argument continues to shape the field.
Today, Non-Reductive Physicalism remains the default position in philosophy of mind, but it is a contested default. Most philosophers accept that mental properties are realized by physical properties and that mental causation is real, but there is no consensus on how to answer the Exclusion Argument. Functionalism is widely assumed as a theory of mental state types, but its epiphenomenalism worry persists. Counterfactual Theories of Causation offer a promising alternative metaphysics, but they have not yet won universal acceptance. Eliminative Materialism remains a live but minority position, valued more for its clarifying pressure than for its empirical predictions.
The leading frameworks agree on one thing: mental causation is a genuine problem that cannot be dissolved by simple identity claims or by ignoring the causal-closure principle. They disagree on whether mental properties are genuinely distinct from physical properties, whether causation requires laws, and whether overdetermination is as implausible as Kim assumes. The field is marked by a productive pluralism: different frameworks capture different aspects of the problem, and no single framework has decisively defeated its rivals. The question of how mental events can cause physical events in a closed physical world remains one of the deepest and most active areas of philosophical inquiry.