“I’m so bored. I hate my life.” - Britney Spears

Das Langweilige ist interessant geworden, weil das Interessante angefangen hat langweilig zu werden. – Thomas Mann

"Never for money/always for love" - The Talking Heads

Thursday, October 09, 2003


We hate zombies. We�ve recently been copyediting some philosophy papers. One of the papers we had the privilege of marking up makes a plausible argument about consciousness as an emergent property. But the paper�s worth, we thought, was vitiated by its easy acceptance of the so-called zombie argument. That argument, which is associated with David Chalmers, goes approximately like this: a molecular level simulacrum of a human being without consciousness is imaginable. This human being, in other words, has no inner feels. Chalmers calls this imagined creature a zombie. He uses the logical possibility of zombies to argue that mental properties � feels, qualia, whatever you want to call them � can�t be explained by the physicalist agenda. In other words, consciousness is something over and beyond the physical properties of the brain, or whatever we take to be the material site of cognition, sensing, etc.

We thought that argument was bogus at the time � imagining the logical possibility of the unicorn doesn�t tell us squat about the ethology of the mountain goat � but having to confront it in a paper on emergence, it occurred to us that complexity theory gives a rather unique refutation of the zombie thesis.

Complexity theory is, of course, a huge topic in itself. However, one of its principles is emergence. Jaegwon Kim, who is a pretty heavy hitter in the world of event ontology, has an on-line paper about emergence that I�d urge LI readers to go to. The introduction is succinct and, for my purposes, sufficient:

� In trying to make emergence intelligible, it is useful to divide the ideas usually associated with the concept into two groups. One group of ideas are manifest in the statement that emergent properties are "novel" and "unpredictable" from the knowledge of their lower-level bases, and that they are not "explainable" or "mechanistically reducible" in terms of their underlying properties. The second group of ideas I have in mind comprises the specific emergentist doctrines concerning emergent properties, and, in particular, claims about the causal powers of the emergents. Prominent among them is the claim that the emergents bring into the world new causal powers of their own, and, in particular, that they have powers to influence and control the direction of the lower-level processes from which they emerge. This is a fundamental tenet of emergentism, not only in the classic emergentism of Samuel Alexander, Lloyd Morgan, and others but also in its various modern versions. Emergentists often contrast their position with epiphenomenalism, dismissing the latter with open scorn. On their view, emergents have causal/explanatory powers in their own right, introducing novel, and hitherto unknown, causal structures into the world.�

From Kim�s outline, it is obvious that emergence is a probabilistic property. That is, we need to scale our search for emergent phenomena to sets that are big enough to allow statistical analysis. Chalmers, and most writers on zombies that I�ve seen, seem oblivious to this approach. Chalmer�s thought experiment involves a one to one relationship between Chalmers and a zombie Chalmers. Yet Chalmers notion of consciousness seems to be at least consonant with the idea that consciousness is an emergent property.

Now if consciousness is truly emergent, than there is a big discrepancy between Chalmers example of a zombie and the emergent paradigm. That paradigm would take into account the fact that there is an exact molecular correlation between humans and zombies and it would predict that in any statistically significant set of zombies, consciousness would emerge in one of them. Nor would there be any way of culling the zombies to prevent this happening. Why? This is where the non-reductionism of emergence kicks in. There is no single physical feature, or single causative factor, that accounts for the emergent property. That is, indeed, the whole import of emergentism. So Chalmers zombies, which have been accepted as an imaginable, would have to have something else within them blocking consciousness � in other words, something distinguishing them, on the molecular level, from human beings � in order for one to be able to pick out, with 100% certainty, from a set of zombies the zombie without consciousness.

This leaves us with two possibilities: we can rescue zombies as a counter-factual by abandoning emergentism -- in which case we would have to explain why emergence doesn't happen given circumstances in which the emergent paradigm seems to apply -- or we can abandon zombies, and seek another counter-factual to talk about consciousness.

I�m rather surprised more people haven�t brandished the emergent program against the zombie thought experiment. Perhaps this is because the seduction of thinking of the zombie as a one to one relationship blinds philosophers to the fact that, in dealing with living creatures, they are dealing with biology. And biology works with populations.

No comments: