Armstrong’s Causal-Role Central-State Materialism

Philosopher David Armstrong’s central-state materialism combines a functional analysis of mental concepts with an empirical identity theory: a mental state is conceptually identified by the causal role it normally plays among stimuli, behavior, and other mental states, while science determines which physical state occupies that role. In humans, Armstrong argues, the occupants are states and processes of the central nervous system, so sensations, perceptions, beliefs, desires, and conscious experiences are not merely caused by brain states but are brain states. The approach naturalizes mental causation and avoids both Cartesian substance and behaviorist elimination, but its causal-role analysis does not by itself explain why particular physical realizers possess phenomenal character (Armstrong, 1968/1993, 1981a).

Structurally, Armstrong's central-state materialism is a two-stage theory of mind whose account of consciousness, developed in a companion entry, is the very same structure applied reflexively. The first stage is a topic-neutral causal analysis: the concept of any mental state is the concept of whatever state is apt to be caused by certain stimuli and apt to cause certain behavior and further mental states (Armstrong, 1968). The second stage is empirical and contingent: since the states that in fact occupy these causal roles turn out to be states of the central nervous system, mental states are identified with brain states. The view is simultaneously an identity theory and a causal-role analysis close to the functionalism Lewis developed independently (Lewis, 1972), and Armstrong presented it as a synthesis dissolving the opposition between Cartesian dualism and Rylean behaviorism. For phenomenal consciousness, the upshot is deflationary: sensations and experiences are inner states individuated by their causal roles and realized physically, their felt qualities carrying no autonomous ontological weight—the precise point at which critics locate the explanatory gap.

From Behaviorism to Inner Physical States

Armstrong began as a behaviorist and, under the influence of J.J.C. Smart's identification of sensations with brain processes (Smart, 1959), but came to reject the behaviorist reduction of mind to behavior. He accepts behaviorism’s insight that mental concepts are systematically connected to behavior but rejects the thesis that mental states are nothing more than actual or hypothetical behavior. Pain can exist without overt withdrawal, and the behavior produced by a belief depends on interacting desires, perceptions, and other beliefs. Mental states must therefore be inner causes whose complex relations explain behavior rather than behavioral dispositions redescribed (Armstrong, 1968/1993).

Behaviorism, Armstrong held, correctly ties the mental to its behavioral manifestations but wrongly denies the existence of inner causes; dualism correctly makes the mental inner and causally potent but wrongly makes it immaterial. Central-state materialism keeps what each gets right: mental states are inner states of the person, causally responsible for behavior rather than identical with it. The mind is not a pattern of dispositions but the categorical inner ground of those dispositions—the "central state" behind the subsequent action.

A mental concept specifies such a state through its causal role. Pain is characteristically caused by injury, tends to produce aversion and avoidance, and interacts with beliefs and desires; perception is normally caused by environmental objects and guides adaptive action. These role descriptions analyze ordinary mental concepts rather than neural anatomy: they say what a mental state does without yet specifying what it is physically (Armstrong, 1968/1993, 1981a).

Functional Analysis and Physical Realization

As noted, the theory has two stages. First, causal analysis characterizes mental states functionally through inputs, outputs, and relations to other mental states. Second, neuroscience identifies the concrete states that perform those functions. Armstrong predicts that these will prove to be physico-chemical brain states, yielding contingent, a posteriori psychophysical identities (Armstrong, 1968/1993).

Armstrong is consequently a precursor of functionalism but not a pure role functionalist. On standard functionalism, a mental state is the multiply realizable functional role itself; for Armstrong, it is the concrete occupant or realizer of that role. If another species performed an analogous pain role through a different physical organization, its pain could have a different physical realization. Every actual mental episode nevertheless has a fully physical constitution (Jackson, 2001; Levin, 2023).

The Causal Analysis of Mental Concepts

The theory's distinctive move is conceptual rather than empirical. To specify what a mental state is, Armstrong analyzes the concept by its place in a causal network: a state apt to be the effect of certain causes and the cause of certain effects (Armstrong, 1968). Pain is the state typically caused by bodily damage that typically produces the desire that it cease and behavior tending to minimize the damage; a perception is the state apt to be caused by the environment and apt to guide action toward it. These role descriptions analyze ordinary mental concepts rather than neural anatomy. Because the analysis fixes each state by what it does, not what it is made of, it is topic-neutral—silent, at the conceptual level, on whether the occupant is physical or immaterial. This is why Armstrong's causal roles can equally be described as functional roles, and why his position is bracketed with functionalism even though he reached it by a distinct route (Armstrong, 1968/1993, 1981a).

Contingent Identity with the Central Nervous System

Topic-neutrality is what allows the identification to be both substantive and contingent. Having defined mental states functionally, Armstrong appeals to the verdict of natural science that the occupants of these roles are neurophysiological states, and concludes that mental states are, as a matter of contingent fact, states of the brain (Armstrong, 1968). The identity is not conceptually necessary—the concept leaves the occupant open—but it is established a posteriori, in the way the gene was identified with DNA or water with H₂O. This structure lets Armstrong claim the explanatory economy of physicalism without asserting that "mental" and "physical" are synonymous, distinguishing his central-state theory from cruder type-identity claims.

Identity Rather than Correlation

Mind and brain are not two correlated domains, and the brain does not produce an additional mental substance or property. The mental occurrence and the neural occurrence are one event described through different concepts. Because mental concepts specify causal roles while neuroscientific concepts specify physical structure, the identity is not analytic but empirically discovered (Armstrong, 1968/1993).

This identity secures mental causation. A desire causes action because the desire is the neural state participating in the relevant physical causal sequence. No interaction between immaterial and material substances is required, and no physical event competes with a separate mental cause. Armstrong also argues that brains, minds, and mental capacities can arise gradually through evolution and development rather than appearing at an ontological threshold (Armstrong, 1968/1993).

Perception, Sensation, and Phenomenal Character

Armstrong interprets perception as the non-inferential acquisition of information or belief about the environment through the senses. Bodily sensations are similarly perception-like: warmth presents part of the body as warm, while pain presents something as amiss and includes an immediate negative attitude (Armstrong, 1962, 1968/1993). He rejects private, nonphysical sense-data as intermediaries between subject and world.

For phenomenal consciousness, however, an explanatory gap remains. Causal roles distinguish pain from belief or visual perception, and identity theory locates each state physically, but neither derivation shows why the realizer feels painful or has a particular visual appearance. Armstrong’s response is that phenomenal qualities are not extra properties awaiting a second explanation; they are features of relevant physical states or, in his later engagement with representationalism, of what experience presents in the world (Armstrong, 1995). Critics answer that this establishes an ontology more clearly than it explains subjectivity.

The Synthesis and Australian Materialism

Armstrong's book became, with Smart's and Place's work, a founding text of Australian materialism and helped launch the physicalist turn in analytic philosophy of mind. Its lasting influence is twofold. First, the two-stage template—functional definition followed by empirical identification—became the standard form of reductive argument in the field, later given its most systematic statement by Lewis. Second, by covering the whole mind under a single schema—perception, belief, sensation, will, and introspection alike—Armstrong showed that a materialist could offer not merely a slogan but a detailed, comprehensive theory of mental functioning. That single schema is precisely why his theory of consciousness is not a separate doctrine: introspection is one more mental state to be analyzed by the same causal method.

Sensation, Qualia, and the Explanatory Gap

Applied to phenomenal consciousness, the causal analysis treats the felt qualities of experience as properties of first-order sensory states, individuated by their role in discriminating and guiding behavior and realized in the brain. Armstrong grants qualia no special, introspectively revealed nature; introspection is itself analyzed by the same method, as an inner sense that scans the mind's own states—central-state materialism turned reflexively upon itself, and the subject of the companion entry on his higher-order theory of consciousness.

Critics press that a role-plus-realizer account states the conditions under which experiences occur and are detected while leaving untouched why occupying such a role should feel like anything at all, so that the knowledge argument and the explanatory gap survive the reduction (Jackson, 1982). Whether Armstrong's materialism narrows that gap or merely reformulates it remains the central question in recent reassessments (Anstey & Braddon-Mitchell, 2024).

References

Anstey, P. R., & Braddon-Mitchell, D. (Eds.). (2024). Armstrong's Materialist Theory of Mind. Oxford: Oxford University Press.

Armstrong, D. M. (1962). Bodily Sensations. London: Routledge & Kegan Paul.

Armstrong, D. M. (1968/1993). A Materialist Theory of the Mind (revised ed.). London: Routledge.

Armstrong, D. M. (1981a). The causal theory of the mind. In The Nature of Mind and Other Essays (pp. 16–31). Ithaca, NY: Cornell University Press.

Armstrong, D. M. (1995). Qualia ain’t in the head: Review of Michael Tye’s Ten Problems of Consciousness. Psyche, 2, 31–34.

Jackson, F. (1982). Epiphenomenal qualia. Philosophical Quarterly, 32(127), 127–136.

Jackson, F. (2001). David M. Armstrong. In A. P. Martinich and D. Sosa (Eds.), A Companion to Analytic Philosophy (pp. 413–418). Oxford: Blackwell.

Levin, J. (2023). Functionalism. In E. N. Zalta and U. Nodelman (Eds.), The Stanford Encyclopedia of Philosophy. Stanford: Metaphysics Research Lab, Stanford University.

Lewis, D. (1972). Psychophysical and theoretical identifications. Australasian Journal of Philosophy, 50(3), 249–258.

Smart, J. J. C. (1959). Sensations and brain processes. Philosophical Review, 68(2), 141–156.