Review of “Teleofunctionalism” by W.G. Lycan For Scholarpedia July 11, 2008
I should like to offer two small suggestions for improving this superlatively clear and informative statement of teleofunctionalism in philosophy of mind.
First Suggestion: The opening sentence, “Functionalism in the philosophy of mind is one species of the view that mental states are internal states of the brain,” is potentially misleading (at least for novices) in two ways.
First, it conceals the fact that, according to functionalism, (A) mental state types are decidedly not brain states or hardware states of any kind, (B) mental state tokens are hardware states that are also mental states because they instantiate some functionally characterized type, and (C) mental state tokens in humans and other known forms of living, minded organisms happen to be brain states because they instantiate some functionally characterized type. What is missing is an emphasis on what is most unique about functionalism, namely, that the hardware details are of no relevance to the identity and the ontology of mental states.
Of course, the very next sentence – “mental states are characterized, not neurophysiologically, but in terms of their computational…” – sets out to clarify the opening sentence, but there would be less need for this clarification if the opener did not point the reader in the direction of the brain to begin with.
Second, the first sentence asserts that mental states are internal states of the brain. According to functionalism, however, mental state types are not necessarily internal states even of the organism, let alone its brain, because it is conceptually possible that the hardware unit be removed from the organism but still be the thing that mediates (by radio waves, e.g.) the functionally characterized relations between inputs and outputs. This is true of teleofunctionalism since there are possible worlds in which minded organisms co-evolve with whatever external source orchestrates the relations between inputs and outputs and in which the recurring mental types are selectively efficacious and so acquire their teleofunction.
Second Suggestion: I wish to lobby for a small additional plug for Davies 2001. In the penultimate paragraph, Lycan acknowledges that the notion of ‘function’ in teleofunctionalism is controversial. This helpfully points the reader (especially the novice) toward an unresolved problem. But the strategy of this paragraph results in a slightly skewed view of the dialectic situation. The paragraph suggests that some thinkers (Davies 2001, e.g.) worry that ‘function’ is not suitably scientific because it is metaphorical or Aristotelian or theistic or illegitimately intentional, while other thinkers (Neander 1991, Wright 1973) argue that ‘function’ is scientific because it can be explicated in terms of the process of natural selection. This way of describing the situation is inaccurate because part of the argument in Davies 2001 (here comes the small plug) is that attempts by Neander, Wright, and others to explicate ‘function’ in terms of selection are themselves subject to certain difficulties that prevent them from achieving scientific credibility. One such difficulty is that the attempt to locate ‘function’ amid the concepts and claims of evolutionary theory cannot succeed because the resources within evolutionary theory – in particular, the mechanism and processes that constitute natural selection – cannot underwrite the attribution of the sorts of norms required for malfunctions.
The point – a small point about the dialectic of the debate – is that this paragraph, as it stands, seems to suggest that the criticisms in Davies 2001 fail to engage Neander and others over the alleged powers of natural selection to produce functional norms. Yet the force (such as it is) of the objections in Davies 2001 is that attempts to explicate ‘function’ in terms of the theory of evolution by natural selection – the very grounds employed by Neander and Wright – do not and cannot succeed.
So the scientific credentials of ‘function’ alleged by Neander and others have already been challenged, not simply because ‘function’ is metaphorical or Aristotelian or whatnot, but because evolutionary theory cannot do what Neander and others say it can do. That, at any rate, is where the debate is at present. End of lobbying!