Dennett on evolution and the determinacy of meaning
And why not? Here, I think, we find as powerful and direct an expression as could be of the intuition that lies behind the belief in original intentionality. This is the doctrine Ruth Millikan calls meaning rationalism, and it is one of the central burdens of her important book, Language, Thought, and Other Biological Categories, to topple it from its traditional pedestal (Millikan, 1984. See also Millikan forthcoming) Something has to give. Either you must abandon meaning rationalism--the idea that you are unlike the fledgling cuckoo not only having access, but in having privileged access to your meanings--or you must abandon the naturalism that insists that you are, after all, just a product of natural selection, whose intentionality is thus derivative and hence potentially indeterminate.
Uh, Dan. If meanings are indeterminate then guess what. It's indeterminate what you mean. No one can possibly determine whether any argument is valid or not, because if, say, it's a categorical syllogism, there's no way to determine whether we've got three terms, four terms, five terms of six terms.
So let's have a look at Dennett's argument.
1. If naturalism is true, then meaning is indeterminate.
2. Naturalism is true.
Therefore, meaning is indeterminate.
And here's mine.
1. If naturalism is true, then meaning is indeterminate.
2. Meaning is determinate. (A presupposition of reason and science).
3. Therefore, naturalism is false.
Labels: Dennett, The Argument from Reason
posted by Victor Reppert @ 10:13 PM
14 Comments
14 Comments:
I like this one better:
1. Naturalism is true.
2. Meaning is determinate.
-----
3. Therefore, it's not the case that (if naturalism is true, then meaning is indeterminate).
If some meanings are indeterminate (or disjunctive), is that such a bad thing? For one, it wouldn't imply that all meanings are indeterminate. (for instance, why think the meanings he is using in the argument aren't determinate)? Even if all meanings were not determinate, they could be disjunctive such that concept X refers to a set of equivalence classes (triangle, three-sided polygon) that are extensionally equivalent such that many arguments are valid anyway, indeterminacy of intension but not extension.
The context would be useful, for instance, what he means by being unlike the fledgling cuckoo.
I don't know. If meaning of "gavagai" is indeterminate between "rabbit" and "undetached rabbit parts," then we have no way of deciding whether the following syllogism is valid.
1. All gavagais are vegetarians.
2. Peter is a gavagai.
3. Therefore, Peter is a vegetarian.
Did you follow the link to the Dennett essay (which is of course pretty well-known).
I guess I'll have to do some reading here. It seems to me that none of this poses a problem for the naturalist -- at least not a Moderate or Liberal Naturalist.
I would say that causal descriptivist accounts like those of Evans, D. Lewis, and Searle (perhaps -- see his book, Intentionality, where he seems to be advancing something close to causal descriptivism in his replies to content externalist arguments). On those accounts, the content of a sentence or statement is descriptive, but the descriptive content includes the causal history of the component concepts.
It seems that neo-Fregean two-dimensionalists like Frank Jackson and David Chalmers can beat back indeterminacy as well. On their account, each sentence is has two components of content: a primary intension and a secondary intension. The primary intension of a sentence is a function from centered worlds to truth-values, and the secondary intension of a sentence is a function from centered worlds plus Kaplan-esque contexts to truth-values.
What's wrong with either of these accounts?
I apologize for jumping into this discussion -- and late at that. So feel free to ignore me if my remarks are obtrusive, or at any rate too far afield.
That argument would work regardless of which extensionally equivalent meaning of gavagai you pick, or even flip between them. That argument actually is a good example of what I'm talking about.
That said, clearly we have concepts with different meanings but identical extensions, e.g., Clark Kent and Superman. My main point was that some degree of indeterminacy isn't the end of the world or science. And some degree of indeterminacy in all concepts may be fine, may not lead to the catastrophe's you are talking about.
No. Undetached rabbit parts cannot be vegetarians. Unless the meaning of gavagai remains invariant throughout the syllogism, it's invalid.
EXAP: I would say that causal descriptivist accounts like those of Evans, D. Lewis, and Searle (perhaps -- see his book, Intentionality, where he seems to be advancing something close to causal descriptivism in his replies to content externalist arguments). On those accounts, the content of a sentence or statement is descriptive, but the descriptive content includes the causal history of the component concepts.
VR: The trouble here is what the ontological price of this kind of theory. I accept Kim-type arguments that at least when you are talking about what can be causally relevant, you really have to be ontologically conservative if you want to stay within the framework on naturalism. If we allow all sorts of supervening properties then we can fit these properties into a naturalistic framework so long as we are willing to live with a mysterious supervenience relationship, but now we've got problems making it causally relevant.
If the parts include the brain, then why not? If they are coextensional, then the argument works, referring to the selfsame classes of entities.
Note I am not saying that any coextensional class supplies a valid argument! Intensional contexts for one will kill them. But there is wiggle room: indeterminate doesn't mean meaningless, it just means there is some slop in some terms.
There has to be enough determinacy to insure sameness of meaning across a reasoning process. When I work a syllogism, I consciously identify the content of a word in one premise with the content of that word in the other. I "see" that I mean the same thing. Can you guarantee that kind of sameness going from the physical up?
We just use the same meaning-atom (a la Fodor) in the different bits of the inference. Fodor 101. The hard part is getting symbols grounded (and we've talked about that ad nauseum with the bees), but once that is done you can recombine and use them in arbitrary ways with various intra and inter proposition rules.
It is an interesting question about introspection and recognition of content-samness over time. Does introspection of same content imply the content is actually the same? I am not sure. For one, two people could have the exact (type) same introspective experience but different contents (twin earth), so we know phenomenonlogy is not sufficient for content fixation. As for the psychological mechanisms of semantic tracking, that is interesting, and I would not be surprised if we had a little bit of phenomenological access to this. An interesting psychological question!
This shows how different the approaches are that we take to content. I always want to bring the conversation back to bees, you always want to bring it back to the deliverances of introspection :)
Hi Victor,
Would you consider this a related problem:
1) Joe desires water
2) Joe believes water is in the fridge
--
3) Joe walks to the fridge.
This practical syllogism assumes that the 'Joe' in premise one is identical with the Joe in premise two, or else there is no reason why Joe would walk to the fridge. For example, if 'Joe' was not identical we get something like this:
1) Joe desires water
2) Sally believes water is in the fridge
--
3) ---
Joe does not go to the fridge, for he does not have the belief that water is in the fridge. In this way, personal identity must be preserved through time as well, or else there is no explanation for our rational actions.
Comments?
That's why I include the argument from the unity of consciousness in my list of arguments from reason. Without the unity of consciousness, reason is impossible.
Huh? What does his Sally Joe example have to do with the unity of consciousness? (Did you ever say exactly what you mean by the unity of conciousness? I think you might have but can't remember.).
I don't see how this 'personal identity' problem is unique to explaining the behavior of people as opposed to explaining the behavior of anything. And I don't see really why it isn't just saying 'Don't equivocate.' Which again isn't unique to arguments about human behavior.
Post a Comment
Subscribe to Post Comments [Atom]
<< Home