Functionalism

Functionalism — A theory in philosophy of mind according to which mental states are defined by their functional role, that is, by the set of causal relations they bear to sensory stimuli (inputs), behaviours (outputs), and other mental states, rather than by the physical substrate that realises them. What makes something a pain is not its being composed of particular neurons, but its being typically caused by tissue damage and its causing avoidance behaviour, expressions of distress, and a search for relief. It is, therefore, a relational characterisation of the mental, in contrast to intrinsic or substantial characterisations. The view was formulated with precision by Hilary Putnam in articles of the 1960s, especially “Psychological Predicates” (1967, later reprinted as “The Nature of Mental States”), drawing on an analogy between the mind and the Turing machine: just as the same programme can run on different hardware, the same mental state could be realised by distinct physical substrates. This is the argument from multiple realisability, directed against type-type identity theory, which had identified each mental state with a specific neural state: human pain might activate C-fibres, but in other creatures, or in hypothetical machines, the same functional role would be played by entirely different structures. Functionalism became the background framework of cognitive science and artificial intelligence, holding that mental processes can be studied as computational processes. Its chief difficulty, pressed by Ned Block and others, concerns qualia: it seems conceivable that a system might fulfil every functional role while lacking any subjective experience — the absent qualia objection.

Origins: Putnam and the Turing Machine

Philosophical functionalism was formulated with precision by Hilary Putnam (1926–2016) in articles of the 1960s, especially “Minds and Machines” (1960) and “Psychological Predicates” (1967, republished as “The Nature of Mental States”). Putnam observed that the notion of the state of a Turing machine — defined by the transitions it performs, not by the physical structure that implements it — offered a powerful analogy for mental states. The mind is to the brain as software is to hardware: the programme can run on completely different physical substrates without losing its functional identity.

This analogy led to the argument of multiple realisability: the same mental function — for example, pain — can be realised by very different physical states in different creatures. In humans, pain may correspond to C-fibre activation; in octopuses or hypothetical robots, the substrates would be entirely different. But in all cases, what unites them is the functional role. Multiple realisability constitutes a powerful argument against type-type identity theory (which identifies pain specifically with C-fibre activation): that theory is excessively chauvinistic in privileging the human substrate.

Functionalism and Cognitive Science

From the 1970s, functionalism became the background framework for cognitive science and artificial intelligence: if mental processes are functional or computational processes, then studying them independently of substrate is methodologically legitimate. Jerry Fodor’s language of thought hypothesis (The Language of Thought, 1975) develops functionalism in this direction: mental states are representations in a formal internal system (mentalese), and cognitive processes are computations over these representations, defined in purely formal (syntactic) terms.

Ned Block’s Distinction: Access and Phenomenal Consciousness

Block is functionalism’s most influential internal critic. In “Troubles with Functionalism” (1978), he presents the absent qualia objection through the Chinese Nation (China Brain) thought experiment: imagine that each inhabitant of China simulates the functional role of a human neuron, communicating with each other in exactly the same causal relations that occur in the brain. The whole system performs the same functions as a human brain — but it intuitively seems to have no conscious experience whatsoever. The precise conceptual distinction underlying this objection was formalised later: in “On a Confusion about a Function of Consciousness” (1995), Block distinguishes two concepts of consciousness:

• Access consciousness: information that is available to be used in reasoning, decision-making, and verbal behavioural control. This is clearly a functional notion.
• Phenomenal consciousness: the intrinsic and subjective quality of experience — the “redness of red”, the “painfulness of pain”, what Thomas Nagel called “what it is like to be” something.

Block argues that functionalism can account for access consciousness, but not for phenomenal consciousness. This suggests that phenomenal experience is not capturable purely in functional terms.

Putnam’s Rejection of Functionalism

Remarkably, Putnam himself abandoned the functionalism he had founded. In Representation and Reality (1988), he argued that functionalism fails on several counts: (1) it does not individuate mental states satisfactorily, since multiple realisability can be extended to absurdity — any sufficiently complex physical system can be described as realising any function; (2) it fails to capture intentionality (the property of mental states to be about something); (3) it leaves the problem of qualia untouched.

Legacy

Despite these objections, functionalism remains the dominant theoretical framework in analytic philosophy of mind and cognitive science. Debates about artificial intelligence, machine consciousness, and the possibility of artificial minds rest, to a large degree, on functionalist premises. The decisive question remains: can there be a system that performs all mental functions without having subjective experience? If so, functionalism is insufficient as a theory of mind; if not, mind is function and can be computationally realised.

← Glossary