Friday, February 24, 2012

Colin McGinn on consciousness

Philosopher Colin McGinn writes an interesting discussion of consciousness in the February 20th issue of New Statesman magazine, arguing in favour of his 'mysterian' approach to the mind-brain relationship.

McGinn begins, however, by rejecting five other common positions in the philosophy of mind: (i) eliminativism (there is no mind); (ii) reductionism (the mind reduces to the brain); (iii) dualism (the mind and the brain interact, but are fundamentally distinct modes of existence); (ii) idealism (only mental entities exist); and panpsychism (every bit of matter has a bit of mentality in it).

After rejecting these approaches, McGinn advocates his mysterian contention that "we just don't have the faculties of comprehension that would enable us to remove the sense of mystery [to the mind-brain relationship]. Ontologically, matter and consciousness are woven intelligibly together but epistemologically we are precluded from seeing how."

Whilst McGinn doesn't use the term 'unification', when he speaks of matter and mind being woven together, he's clearly advocating the notion that the mind and the brain are ontologically unified. His position, then, is to accept ontological unification, but reject epistemological unification.

Now, the distinction between ontological and epistemological unification is a useful one; more's the pity, then, that McGinn doesn't acknowledge that this distinction can be equally applied to reductionist accounts of the mind-brain relationship. Thus, whilst the mind may be ontologically reducible to the brain, it might be impossible to comprehend that reduction relationship; one might endorse ontological reductionism but reject epistemological reductionism.

In response, the question might be posed: why prefer ontological reductionism to ontological unification? The answer is that the relationship between the mind and the brain does not merely reside in correspondence, but consists of a supervenience relationship. This means that there is a many-one mapping between brain states and mental states. Multiple brain states correspond to single mental states; change a brain state slightly by changing the firing frequency of a few neurons, and the mental state remains invariant. But change a mental state, and the brain state must change.

This entails that brain states determine mental states, not vice-versa. The relationship between the mind and the brain is not bilateral or symmetrical, it is top-down. Hence, ontological reductionism is required rather than ontological unification.

McGinn is a fascinating writer, but his reasons for rejecting reductionism here are badly misconceived:

"The brain processes held to constitute conscious experience consist of physical events that can exist in the absence of consciousness. Electricity in the brain correlates with mental activity but electricity in your TV presumably does not - so how can electrical processes be the essence of conscious experience?"

This is like arguing that the information processing performed by a computer program cannot ultimately be reduced to what's going on in the electrical circuits of the computer, because a standard lamp also uses electricity, and there's no information processing going on there!

On the contrary, it's the pattern of information processing, the representation of the external world, and the partial representation of the self, which identifies humans as cognitive systems, and distinguishes them from televisions and standard lamps.

McGinn claims that "reduction is tantamount to's like maintaining that people called 'witches' are nothing but harmless old ladies - which is tantamount to saying that there are no witches." Again, however, this is a false analogy. Imagine a comparable statement about the biological cell:

It's like maintaining that biological cells are nothing but systems of interacting molecules - which is tantamount to saying that there are no biological cells.

This clearly doesn't follow. It's perfectly coherent to argue that biological cells are nothing more than interacting systems of molecules, but still exist as objective patterns on that level of structure.

Similarly, it's perfectly coherent to argue that minds are ontologically reducible to brains, but exist as objective information processing patterns on that level of structure.

No comments: