This is the book about consciousness that I noted a little while ago, and I’ve just got around to reading it. It’s very good.
In the first section, Rosenberg raises most of the problems about consciousness that had occurred to me. In short, no matter how good a physical description you have, there is nothing in there about where consciousness comes from, so something has to be added to physical theories to say “and this gives rise to conscious experiences”. It seems very likely that that is the way to go, since even if global waves of electrical activity across the brain are the physical correlates of consciousness, current theory does not say that they would be conscious, so something must be missing. However, if we add consciousness properties to electrons and the like, it is deeply obscure how they could give rise to the unified conscious experience that we have.
While these problems are not accepted by everyone, they are fairly generally known. However, the fact that Rosenberg was promising to address the problems I see head-on was very promising. The meat of the book, then, is his solution to the problems.
The basic outline of the solution is very simple. He creates a new theory of causation, in which there are both effective and receptive properties, which combine to make individuals. Effective properties are conscious experiences, and a receptive property is a field of conscious experience. These receptive properties exist at different levels, so there is no reason why there cannot be a single receptive property for human minds, which would explain why we have a single unified experience.
Obviously, I’ve cut all the arguments and details from the above. However, it’s a very clever move. There are definitely problems in the theory of causation, and this theory does answer some of them. His argument that Humean theories have no way to define the universe was particularly interesting; I suspect that he may be right. Humean theories say that there are just individual events, with no links between them, and causation is simply the patterns that occur. However, in strict Humeanism you construct space, time, and causation from the events, which means that you can’t define the universe as “causally linked events” or “events in the same space” or “events in the same time”. You need the set of events first. That means that Humeans need a reason to exclude Tolkien’s Middle Earth from the universe. “It’s made up” isn’t enough, because they have no obvious way to pick out the set of “real events”, other than arbitrary stipulation. And the fact that Middle Earth does not exist is not an arbitrary stipulation; it reflects a deep truth about the universe.
So, back to the main argument of the book. The claim that metaphysics needs “real” causation is quite convincing, and Rosenberg’s split between symmetric and asymmetric constraints has the potential to do a lot of useful work. Similarly, the argument that the relational properties of physics (negative charge is just different from positive charge, there is nothing inherent about them) need some sort of categorical basis is prima facie convincing. Using conscious experiences to fill the role is, frankly, a brilliant idea. They are categorical properties that we know exist, and so it is metaphysically parsimonious to do things this way.
This is by far the best attempt to grapple with the hard problem of consciousness that I’ve ever seen, and I’d say it’s essential reading for anyone interested in the issue.
I’m not, however, fully convinced. One problem is that I’m not quite sure how the causal theory will work out in detail. This is simply due to the fact that I haven’t gone through it with a fine tooth comb yet; it’s possible that, when I do so, everything will be fine, or that there will be small changes that solve the problems. (The theory is unlikely to be perfect at this point, even if it is fundamentally right.)
I do have a more philosophical worry, however. The experiential basis of a particular physical property could be anything. There might be good reasons for all the visual experiences being based on the same sort of thing, but none were canvassed in the book. The unity of the physical process is not, by itself, enough; that does not feed down to the level of the experiences. Thus, there doesn’t seem to be a reason why red isn’t a sound rather than a colour.
More fundamentally, it seems entirely possible that pleasure could feel like pain, and vice versa. That is, something could be exactly the experience that pain is, and not be aversive. It is, on this theory, pure good luck that we have evolved to seek out things that feel good. We were bound to evolve to seek out things that feel, since an experiential basis was needed for those causal properties, but the experiential basis could have been agony. If it had, we would seek out agony, say we liked it, and call it “pleasure”, but it would still feel horrible.
I’m not entirely sure that this is even coherent as a thought experiment, but since failure of imagination is a poor philosophical argument I’ll let that pass. The other problem is that it means that the theory means that there is no possible investigation that can tell us anything about conscious experiences. We might be able to determine whether something has them or not, but even that is a bit tricky. We certainly wouldn’t be able to determine what they were like. (Parsimony would let us say that other people had similar ones to ours.) Now, this might be the way the world is; there is no reason why everything should be accessible to investigation. However, given that we are only just starting to investigate consciousness, I think I’d like to try a more optimistic approach first.
Of course, that means coming up with an alternative theory of consciousness, and given how thin on the ground they are, that’s far from a trivial proposition.
So, this book is great. But I think I hope it’s completely wrong.