Comments on Philosophy, et cetera: The Logic of Indeterminacy

To clarify: I take it that compositionality fails ...

2008-02-05T10:06:00.000-05:00

To clarify: I take it that compositionality fails not because the semantic value of 'Bob' changes from one sentence to the next (I may have mistakenly written that in my first post), but because the semantic value of the whole sentence cannot be reduced to the SVs of its atomic parts.

That seems more suspicious, insofar as it involves...

2008-02-05T10:01:00.000-05:00

That seems more suspicious, insofar as it involves introducing two distinct meanings (and so, arguably, two distinct words). My example of 'Bob' seems to hold the same meaning throughout, and the stipulation distinguishing it from 'Mirror-Bob' seems more natural/intuitive, and less gerrymandered/ad hoc. (The condition really seems to capture how we would use those names. I can't imagine any natural language term behaving as you describe, by contrast.)

Something worries me about basing an argument agai...

2008-02-05T03:13:00.000-05:00

Something worries me about basing an argument against a semantic theory based on arbitrarily stipulated meanings.

If that were allowable, couldn't you simply stipulate the meaning of a term so as ensure that compositionality was false without the need for the intermediate argument? "I define 'blah' as meaning X when embedded in sentences of kind Y, and meaning F when embedded in sentences of kind G. Therefore strict compositionality is false."

But I fear I'm misunderstanding here, so I'll give it some more thought.

It's the same reason for thinking that the names r...

2008-02-04T10:12:00.000-05:00

It's the same reason for thinking that the names refer to people rather than pumpkins: I just stipulate that as a condition on the terms as I mean to use them. That is, it's analytic that Bob and Mirror-Bob are different people (if they exist at all).

(Again, I might be missing something, but..)It mak...

2008-02-04T06:09:00.000-05:00

(Again, I might be missing something, but..)

It makes sense, but I don't see the justification for thinking that (6) is more justified than the weaker (4). Why think that the only "allowed" scenarious are (b) and (c)?