Sunday, October 5, 2008

A tidbit

Recall early in the semester we breifly considred a position that identified what is asserted with sentences of a language. The main objection to that position came from the translation principle (the principle that a sentence of a different language can have the same meaning as a sentence in the first language). The objection went as follows:
1) 'There's chaos in my bed' and 'yesh balagan b'mita sheli' assert the same thing
2) A statement of 'yesh balagan b'mita sheli' does not assert 'there's chaos in my bed'
3) A statement of 'there's chaos in my bed' does not assert 'yesh balagan b'mita sheli'
4) (2)&(3)
5) If declarations of sentences assert sentences, then (1)->~(4)
6) ~[(1)->~(4)] (1,4)
7) It's not the case that declarations of sentences assert sentences

The idea behind (5) is that if what is asserted is a sentence, then the most plausible candidate for what is asserted is the sentence used to make the assertion. If this is right, (2) and (3) are plausible. But if (1) is true, and either one of the given sentences must be what is asserted, either (2) or (3) must be false.
One major consideration for accepting (4) is that to choose (2) to be false or (3) to be false would be an arbitrary choice. In other words, there's no principled way to choose what sentence is asserted by a sentence of a given meaning (indeed, what is asserted by ALL sentences with that meaning).
But this is (on the face of it) is just a benaceraf dilema. A proponent of the view could look a few classes forward and say to the proposition theorist that she will face benaceraff dilemas anyway. That is no reason to reject the view at hand.
Obviously the analogy is a bad one, but why?

4 comments:

Wes McPherson said...

Hi Dan,

I was wondering if you could tell me what is the analogy that is obviously wrong? I sorry, but I've missed something.

Maybe we could just hold that a sentence of English does not assert something in English. Maybe no natural languages asserts something that is in a natural language. Maybe Mentalese sentences are asserted by all human languages, such that 'outer speech' sentences assert 'inner speech' sentences.

Perhaps here the Jubien argument gets run: this is just a model. You owe an explanation.——OK, so it is the model: now we face a BD.

Perhaps it is just as appealing to look for a structuralist escape here as for propositions?

Is this your worry, that we get stuck with something like a Quinean account right from the get-go?

Dan said...

Well, I'm not sure why the analogy is obviously wrong, that was the question. The resulting view just seems absurd.
The mentalese approach isn't exactly what the proposal was shooting for. After all, Frege's line could very well be seen as a mentalese approach (as King describes in ch 1). The proposal is one that wants to stop investigation by identifying propositions with something we're already comfortable with. We're not comfortable with mentalese sentences.
A structuralist response to the BD in this case would be something like positing sellarsian functional roles to the sentences. We've already discussed this option at length.

Wes McPherson said...

Hi Dan,

I think the analogy is good. I was wondering if you though we might have to cast the question in terms not of what the sentences mean, and that it is this meaning they have in common, but something else, like Quine's stimulus-meanings, which they would have in common. Then, no need to worry about propositions at all.

Dan said...

no, I had nothing fancy in mind