In the book symposia on Insensitive Semantics Herman Cappelen and Ernie Lepore argue that Kent Bach's position collapses into radical contextualism. Bach's position is roughly that many sentences are semantically incomplete, and that they are in need of completion for a full proposition to be pragmatically conveyed. Consider, for instance:
(1) Mary has had enough.
(1) is grammatical. Yet on Bach's view, it fails to semantically express a proposition relative to context. Relative to context, it expresses only a propositional radical, namely the radical that Mary has had enough. The speaker may now fill in the missing constituents and by doing that she may succeed in conveying a complete proposition.
Cappelen and Lepore argue that this view collapses into radical contextualism. For whatever your reason is for thinking that (1) is semantically incomplete, this will also be a reason for you to think that (2) below is semantically incomplete:
(2) Mary has had enough pasta.
The main reason to think that (1) is semantically incomplete is that whenever it is true that Mary has had enough of something or other (e.g. pasta), it is usually also true that Mary has not had enough of something or other (e.g., compliments). But suppose now that Mary owns a pasta company and has just ordered 10 million bags of pasta. Then it would be true to say that Mary has had enough pasta. But if I utter (2) during my pasta dinner with Mary and her family, and Mary wants more pasta, then it is not true to say that Mary has had enough pasta. Since both of these scenarios could obtain at the same time, it may be that it is true that Mary has had enough pasta and true that Mary has not had enough pasta. According to Cappelen and Lepore, it is simply impossible to draw a principled distinction between sentences that semantically express propositions and sentences that semantically express propositional radicals.
What to do? Well, you could join Cappelen and Lepore in saying that (1) and (2) both semantically express full propositions relative to context. Or you could revise Bach's view. Bach says that it is sentences that are truth-evaluable or non-truth-evaluable. But we could also just say that it is occurrences or utterances of sentences that are truth-evaluable or non-truth-evaluable. In the pasta-dinner/pasta-company situation, then, an occurrence of (2) is not truth-evaluable. But other occurrences of (2) may be truth-evaluable. In other words, it is possible to draw a principled distinction between occurrences of sentences that are truth-evaluable and occurrences of sentences that are not truth-evaluable, even if it is not possible to draw a principled distinction between sentences that are truth-evaluable and sentences that are not truth-evaluable.
Tuesday, September 05, 2006
Semantic Minimalism
Subscribe to:
Post Comments (Atom)
2 comments:
Contrary to what Cappelen and Lepore suggest, I think that like (1) sentence (2) is semantically incomplete too. It just takes less to complete (2).
(1) Mary has had enough.
(2) Mary has had enough pasta.
Brit suggests that since both of her scenarios could obtain at the same time, "it may be that it is true that Mary has had enough pasta and true that Mary has not had enough pasta." But not if (2) and its negation are semantically incomplete, hence neither true nor false. The things that are true here have not been fully specified.
If (2) and its negation were semantically complete, then that it could be true that Mary has had enough pasta and true that Mary has not had enough pasta. Since it couldn't be, (2) and its negation are not semantically complete. So, in effect, Brit has implicitly hit on a nice test for semantic incompleteness, contrary to Cappelen and Lepore's insistence that it is impossible to draw a principled distinction between sentences that semantically express propositions and sentences that semantically express propositional radicals.
I'm not sure what Brit has in mind when she suggests that "it is occurrences or utterances of sentences that are truth-evaluable or non-truth-evaluable." What I would say that when a sentence is semantically incomplete, it is what a speaker means in uttering the sentence, something underdetermined by the semantic content of the sentence, that is true or false. It's not that (2) is true in one scenario and its negation is true in the other. So I don't see any need for fussing about occurrences or utterances of sentences, unless the above is all Brit has in mind by this.
"When a sentence is semantically incomplete, it is what a speaker means in uttering the sentence, something underdetermined by the semantic content of the sentence, that is true or false"
But, Kent, what you are saying here seems to imply that it is the occurrence or utterance of the sentence that is semantically incomplete (not the sentence itself). For example, (2) is sometimes incomplete and sometimes not (or if you prefer: (2) semantically expresses a proposition on some occasions of use but does not semantically express a proposition on other occasions of use).
The test then is this:
An occurrence U of a sentence S semantically expresses a truth-evaluable minimal proposition iff it is not the case that the narrow semantic content of U and the negation of the narrow semantic content of U are both true at the circumstance of evaluation.
Post a Comment