Functionalism
Explanation
Hilary Putnam introduced in the 1960s an idea that would change the philosophy of mind: what matters about mental states is not what they are made of physically, but what they do, what functional role they play in the system. Pain, for example, would be the internal state that is caused by tissue damage and produces the desire to avoid it, screams, requests for help. Any system with that functional role feels pain, whatever its material.
Functionalism's favourite analogy is software and hardware. A chess program can run on a PC, on a Mac, on a phone or on a futuristic quantum machine. What defines the program is not the concrete materiality of the substrate but the logical organization of its operations. Analogously, the mind would be the "software" the brain implements, and could be implemented on other substrates.
This thesis has a revolutionary consequence: multiple realizability. Martians, octopuses, advanced computers, immaterial angels: any system that reproduces the correct functional relations would have genuine mental states. This makes functionalism a liberal position about who can be conscious, distancing it from carbon-centric chauvinism.
There are variants: Turing-machine functionalism identifies mental states with states of an abstract computational machine; causal functionalism characterizes them by their causal relations with inputs, outputs and other states; psychological functionalism appeals to the laws of folk psychology to define them. All share the central intuition: mind is function, not substance.
The best-known objections are the qualia arguments. Searle's Chinese Room asks whether a system that manipulates symbols according to rules has genuine understanding or only simulates understanding. The zombie argument asks whether we could have a system functionally identical to us but without subjective experience. And inverted qualia ask whether two functionally identical systems could see colours in opposite ways.
Despite these critiques, functionalism remains the dominant position, or at least the most influential, in philosophy of mind, and it is the implicit assumption of much of artificial intelligence and cognitive sciences. The question of whether advanced artificial neural networks could have a mind largely depends on whether or not we accept functionalist premises.
Strengths
- Compatible with multiple realizability, avoids biological parochialism.
- Foundational framework for cognitive sciences and AI.
- Naturalist without being eliminative.
- Explains the possibility of comparing minds across substrates.
Main critiques
- Searle's Chinese Room argument: manipulating symbols according to rules does not guarantee understanding.
- Block's nation-brain argument: a coordinated nation could implement the functional role without having experiences.
- Does not explain qualia.
- Integrated information theory criticizes it as insufficient: structures matter, not just functions.