This is a topic that's been bothering me off and on for a while, I've written a few posts on it and then deleted them. Even before I declared that I was leaving message boards I was trying to get a post done on this. Anyway, since I always change my mind on how to approach it, let me just cut and paste this email I sent to Elliot Sober yesterday:
Dr. Sober,
I've read a number of your papers online and your arguments against Design Inference resonate with me more than others I've read. I'm also quite interested in phil. mind. For some reason, "I", keep thinking eliminativism is my only option if the argument works that we can't (in theory) know anything about the content of alien (or computer) minds or even if they have minds/intentions. It seems to me that if there is some kind of ultimate structure to mind, whether it's neuro-chemical reactions or functional relations, then an alien with that would have mental content translatable, ultimately, in human terms. In other words, if humans could come up with a geneneral case model of the mind, we'd be able to reduce all langauges to little thought-bit representations. We'd then have a general theory of language. And if we had a general theory of language/signification, we'd have a prior pattern to compare space signals to. And then we'd be able to test for "specified complexity". Ok in practice I'm sure never, but as a thought exercise, it seems that way to me. That any atomistic theory of mind would at least weakly validate, in principle, the idea of "specified complexity".
thanks for your part in bringing sanity to the world.
--
Given I didn't ask permission to post his response and that his reply was very thoughtful, I won't post it but I'll just give my interpretation of what he said. First of all, it might be helpful to get Sober's take on the "other minds" problem generally, not specifically related to DI:
http://philosophy.elte.hu/colloquium/2000/Marcius/Sober-paper.htm
In his response to me he left the ID stuff out completely and concentrated on elimitavism. If I understood him correctly, he thinks eliminativism is too extreme because we could say the same thing about life, throw out the concept of life because we might not one day be able to identify whether something is alive or not. The problem I gather, in his view, a matter of epistemology only. But I can't quite figure it out whether he means there is no way to interpret mind from behavior in extremely foreign cases or that it's possible in theory, but we need some ingenious invention. The second seems to me to still allow DI as valid in principle, the first may or may not. I'm sure his view is consistent and I just don't get it but I don't want to keep bugging him. That would be like asking God for the same toy over and over again, he has better things to do.
Other solutions may be in assuming too much in my argument. For instance, part of my argument is rooted in Quine's observations and his move from language non-translatability to elimitavism. Perhaps there is a way that thought atoms exist but don't give rise to a universal language. Well, there could be questions like that at every step of my argument.
I had another thought just today on the matter though. Maybe externalism would help. If that sounds obscure, take a look at Putnam's Twin Earth Experiment: http://artsci.wustl.edu/~philos/MindDict/twinearth.html
I probably should have thought of this before, but anyway, in externalism, mental content depends on things external to the individual. So it would be impossible for humans through introspection to reverse engineer mental content and derive a general case, the secret to language (if it makes sense to even call it that) would forever be part of the Ontic world.
After a couple quick searches I can't figure out if Sober is an externalist or not. Another possibility is I'm reading too much into Sober. Maybe he believes DI is a theoretical possibility but well beyond our abilities, and that Dembski's problem is making too many ridiculous assumptions about what we know in order to make a science out of it that gives us the results IDers claim.
