Tuesday, August 11, 2009

A Priori Neuroscience

Either Ned Block's Blockhead counts as thinking, or it's an empirical matter what possible configurations of stuff would count as thinking. For, suppose the claim labeled (2), below, were true and knowable a priori. Then you could do a priori neuroscience as follows:
(1)I think (Descartes).
(2)A blockhead (i.e. a big physical system which passes the turing test just by having a big look-up table, with all possible series of questions that could be asked in, say, half an hour) would not count as thinking. (Ned Block)**
(3)Therefore, some part of me (call it my brain) is not a big look-up table.

**in the paper where he introduces the blockhead, Block argues that such a machine would not count as 'intelligent', not that it wouldn't count as 'thinking' but I presume he would also accept a similar claim for 'thinking'?

Now, the blockhead is explicitly cooked up to be a worse-case scenario for what physical stuff might realize a given behavior, so if you admit that the blockhead thinks it's hard to see how you could deny that any other thing with suitable behavior would count as thinking. Thus, rejecting a posteriori necessities, seems to be a direct route to behaviorism!

I wonder if this could have motivated Wittgenstein's near behaviorism? I don't remember if he ever retracts the serious use of a prior=necessary in the Tractatus...and there definitely is that passage about 'what if you opened your skull and there was nothing inside?', so he does seem to have thought about these issues...
Read more!

Friday, March 27, 2009

Pleasure from proofs and pleasure from 'psychologically insightful' novels

Another post about something I don't really know about:

Most people would agree that mathematical proofs can be beautiful and produce aesthetic pleasure, sometimes to a very high degree. From personal experience (though I don't think this is too controversial) I would say that beautiful proofs take statements and inferences that are well-known to you and/or feel obvious and combine them to produce something unexpected. [It may also be important that you can 'mentally survey' *how* familiar elements combine to yield something unexpected - that would be why proofs that split the question up into a large number of cases, or involve lengthy arithmetical computations tend to seem less beautiful.]

Now, I propose what we find pleasing in 'psychologically insightful' passages in novels might be essentially the same as what we find pleasing in proofs.

I mean, merely learning new truths about psychology (e.g. by reading a psychology study) doesn't generally produce aesthetic pleasure. Nor do just any descriptions of people behaving in ways that are psychologically plausible (e.g. people running away from danger, or trying to earn money or whatnot). Rather, the psychological passages in novels which give us aesthetic pleasure are ones where people do things that seem a bit strange or surprising at first, but which we then recognize (on thinking about it more, or considering what the novelist says about this behavior) as in fact being psychologically plausible. When Proust's narrator describes feeling overwhelming excitement over a young woman's momentary nasal tone of voice, this seems strange at first (a nasal voice isn't generally an attractive feature) but then on a little further consideration it seems completely psychologically plausible, and actually very typical of what it feels like to experience erotic obsession.

In these cases like this, we suddenly see how something rather new and surprising follows from principles about human psychology which we already implicitly accept. So the same thing is going on as when we read a beautiful proof - only it's our ability to recognize what's plausible psychologically, rather than our ability to recognize what's a good mathematical argument that's being milked to yield a surprising new conclusion.

p.s. This idea might also explain how novelists would wind up learning things about psychology without making new empirical observations. They aren't coming up with new laws/principles but showing how new facts about the psychological plausibility of certain kinds of repetitive behaviors, neuroses, ambitions etc. are a consequence of things we already accept about (folk) psychology.

Read more!