Wednesday, March 29, 2006

Bill Vallicella attacks eliminativism

7 comments:

Blue Devil Knight said...

His post is neither original nor convincing. He doesn't address the basic concerns and claims of eliminative materialism.

For the San Diego connectionists, and the Churchlands, public linguistic expressions are a linear string of symbols that serve to make public some of the rich nonpropositional, nonsentential contents that brains use to represent the world. Is your description of your experience of a rich visual scene really doing justice to the experience? Probably not. This is because our public, ordinary language descriptions are not in the right representation format to do them justice. To do them justice, we need to go under the hood and look at the formal structure of neural representations.

Eliminativists merely believe that it would be miraculous if the crude belief-desire psychology, based on anemic public expressions, actually got it right. Folk physics, which is roughly Aristotelian, doesn't even get it right. The early mind theorists used the only example of content-bearing states they had at their disposal (public language) as their model of internal representational states. I think it is likely they got it wrong.

Eliminative materialists can preserve many features of beliefs, but not as things that are internal representations, but public linguistic structures, which indeed have a sentential structure, which belief-desire psychology has extended into the head via analogical extension.

Churchland's A Neurocomputational Perspective addresses all this stuff in great detail. He got the idea from Wilfrid Sellars while he was a grad student with Sellars at Pittsburgh.

Time will tell. Dogmatism right now is foolish about all these debates, given our state of ignorance of the biology of the situation. People who claim they can divine the future, how they will feel once the neuroscience is "complete", are being naive. As are the people who dogmatically claim that neuroscience proves that materialism is true.

Blue Devil Knight said...

Just to point out what an eliminativist wrt beliefs would say is the problem with his argument, Vallicella says:
C. EM is true.

He really means "The proposition 'EM is true' is true."

But this is not what the eliminative materialist claims! (He, characteristically, assumes that his interlocutors are idiots). Eliminativists would claim that 'The sentence 'EM is true' is true,' where truth is a property of linguistic structures, or even a property of internal, nonpropositional, neuronal representations.

His argument is a straw man. Again, if he had just read some Churchland, who discusses such obvious and superficial counterarguments to his view.

Blue Devil Knight said...

One more thing. Vallicella, in his previous post "defining" EM, defines it incorrectly as the claim that there are no mental states. This is not true. The Churchlands believe that there are mental states, mental representations, as they have made clear since their first publications. They are eliminativists only wrt propositional attitude psychology, not to psychology in general.

For instance, see their article on neural representations.

It would be a lifetime job responding to critics of the Churchlands who don't know what they are talking about. I am glad I am a scientist where we deal with data!

Anonymous said...

"Dogmatism right now is foolish about all these debates, given our state of ignorance of the biology of the situation. People who claim they can divine the future, how they will feel once the neuroscience is "complete", are being naive. As are the people who dogmatically claim that neuroscience proves that materialism is true."



Excellent point. I really wish more people who engage in these discussions about the mind were more interested in understanding the mind than using what little knowledge or lack of knowledge we have about the mind to push their philosophical agendas.
I’m just starting to delve into some of the writings by the Churchlands and my initial impression agrees with your basic comments. Appears that many of those attacking them don’t really understand their position. The strawman fallacy inevitably follows.
t.

Blue Devil Knight said...

Obviously I was writing in broad strokes, to give a more charitable reading of EM that Vallicella's vitriolic obfuscation.

Dave said:
"Then they [public linguistic expressions] cannot preserve the features of beliefs which make them beliefs."

Like what? They have all the semantic properties that beliefs have (they can be true or false, they have meaning, they have evidenciary and inferential relations to one another). What is it leaving out, other than the hypothesis that similar things are going on in our brain?

"It seems obvious that it's the other way around, that the pattern of thinking comes first."

Beware of the "it seems obvious" operator! Yes, neural representations, are required for l inguistic representation. That doesn't imply that belief-desire psychology is true.

I said:
"Folk physics, which is roughly Aristotelian, doesn't even get it right."

Dave responded:
This is a non-sequitur.

Folk physics doesn't even get classical mechanics, where things are linear (F=ma), right. Nervous systems are highly nonlinear, much more complicated systems. It would be very surprising if we got that right. This isn't a knock-down argument, but suggestive.

As for the Cartesian claim that propositional attitude psychology is based on "first person experience", that is interesting. I don't experience propositions. I experience lots of things (pains, tickles, rich visual scenes), but not propositions. (Conscious contents are not the same as propositional contents in the Churchland's view, a point, again, which the dilettant will not know. This is fine, but someone publicly attacking their view should be familiar with it.

As far as the question "How do brains represent?" my point is largely that we don't know enough neuroscience to answer it. However, as I stated, to state at the outset that they are propositional in structure, is premature and dogmatic. I also listed one of the Churchland's article on this above. And Paul's book 'A Neurocomputational Perspective' and Pat's 'The Computational Brain' both give detailed and empirically motivated stabs at an answer, an alternative to propositional attitude psychology.

Where was I uncharitable to Vallicella? What have I erroneously contributed to his argument?

I don't believe all of the Churchland's stuff, but I at least am familiar with it, and it is annoying to see so many people confidently rejecting a straw man, while claiming to have slain the real deal.

I think their choice of terminology, eliminative materialism, is unfortunate, as it invites people to react (as people think it means minds will be eliminated: a cursory glance at their work, even chapter and paper headings(!), shows this is not the case).

OK, I'm gonna bow because I don't want to spend my weekend with more of these posts.

I'll respond, though, to (cogent and specific) claims that I have been uncharitable to Vallicella. What is the problem with my second post above, criticizing his cute formal argument (which, incidentally, is not unique: there are four references to the "EM is self-refuting" argument at the Standord entry, and Paul has replied in 'A Neurocomputational Perspective' better than I could in this forum).

Anonymous said...

I think it takes a bit more than a refutation of representationalism to eliminate beliefs from the furniture of the mind. The notion that beliefs are representations in the head, after all, isn't something that ever occurs to most people who have no difficulty engaging in discourse about beliefs. Even more pointedly, the notion of representation employed by eliminativists might itself be inadequate -- so even if a particular account of belief as representation is debunked, there might well be other, more adequate notions of representation that could underwrite the belief as representation thesis.

In other words, there might be a lot of straw men sacrificed before we get anything like a definitive account of representation and it's relation (or not) to beliefs.

Blue Devil Knight said...

Dave,

I stand by my claim about Vallicella. On his blog, he acts like a supercilious prig. Such character judgments, though, aren't relevant to any of my claims.

You say that we know propositional attitude psychology is true because of first-person experience, but also admit you don't observe or experience propositions. So, why, based on your phenomenology, which doesn't reveal propositions, are you so attached to propositional attitude psychology?

Both schools, the propositionalists, and the state-space semantics people (Churchland's semantic theory), are offering speculative psychological theories which we don't presently have the data to confirm/refute.

Thanks for the discussion, Dave. You made some good points: especially that linguistic structures, by themselves, do not have semantic features. It is the fact that they were produced by cognitive systems with intent to express certain contents that gives them their semantic features. I should have been more clear about that. (To steal from Hilary Putnam, a picture of Reagan drawn in the sand by an artist is a representation of Reagan. An identical scrawl on the sand done by an ant crawling around is not a representation of Reagan. Being from New Hampshire, the Old Man in the Mountain may be a counterexample :))

For anyone who is serious about understanding eliminative materialism, i.e., the claims that its proponents actually make, I recommend reading some of the original sources rather than superficial criticisms (and expositions) from people on the internet. The best source is Paul Churchland's 'A Neurocomputational Perspective'. There, he criticizes, and provides an alternative, to propositional attitude psychology. It is his best book.