4/04/2006

Semantics and necessary truth

I have recently been reading, with great profit, Jason Stanley's draft manuscript Philosophy of Language in the Twentieth Century, forthcoming in the Routledge Guide to Twentieth Century Philosophy. The paper's synoptic scope matches the ambitious title.

I'm curious about one relatively small claim Jason makes. On MS p.17, he says:
intuitively an instance of Tarski's schema T [GF-A: '...' is a true sentence if and only iff ...] such as (7) is not a necessary truth at all:
(7) "Bill Clinton is smart" is a true sentence if and only if Bill Clinton is smart.
(7) is not a necessary truth, because "Bill Clinton is smart" could have meant something other than it does. For example, "is smart" could have expressed the property of being from Mars, in which case (7) would be false.
This certainly has (as Stanley says) an "intuitive" ring. But now I'm not sure it's correct.

Here's my worry: as a preliminary, recall (as Tarski taught us) that semantic vocabulary should always be indexed to a particular language -- e.g., we must say 'is a true sentence of English' or 'x refers to y in Farsi' etc. in the full statement of sentences like (7). But then I am not so sure that such sentences are not true in all possible worlds. Is it really the case that, in English, "is smart" could have expressed the property of being from Mars? We specify a particular language (in part) by specifying the semantic values of the words of that language (at least, if we are not proceeding purely formally/ proof-theoretically). Wouldn't we be speaking another language at that point, that was similar to English, but not the same?

My intuitions lean towards saying that this would not be English, but those intuitions aren't firm. I think the question boils down to: "Is 'English' a rigid designator (i.e., does 'English' refer to the same thing(s) in all possible worlds)?", but I'm not sure about that, either. Which way do your intuitions run?

9 comments:

Anonymous said...

I agree with you. Cf. (Q) "'Two plus two equals four' is true if and only if two plus two equals four." (Q) isn't a necessary truth because (I take it the argument would go) 'equals four' might have meant 'eats Big Macs.'

Anonymous said...

in an informal seminar at the university of essen we recently discussed the related question wether instances of the T-scheme are analytic. Though many shared my intuition that they are not necessary(although without having any real argument), most of us thought they are analytic.
One thought: What about languages, which we can not specify my giving their interpretation function because we simply haven't succeeded in interpreting these languages yet? Surely we could only refer to them as "the language spoken in country X" or similar?
Do you know about any further literature on this?

thx in advance
Kim

p.s.:sorry for my probably bad english

Greg Frost-Arnold said...

Hi Kim --

~ The question of whether instances of the T-schema are analytic is one (of many) that divide Carnap and Quine: Carnap says they are (Foundations of Logic and Mathematics, Introduction to Semantics), and Quine does not (in a footnote to "Notes on the Theory of Reference," in From a Logical Point of View). I think this difference is very important for understanding the fundamental differences between Carnap and Quine (see my earlier post: Is 'p is a priori' itself a priori?), but that is not a common opinion.

~ I would've thought that if we don't have an interpretation function for a given set of symbolic strings, then we wouldn't be able to specify which of those strings were true and which false. So the question of the T-scheme wouldn't come up. (Because:
"Qwert yuiop" is true-in-Martian-language iff Qwert yuiop
is a nonsense mixture, not an instance of the T-schema.)

~ And your English is quite good! Probably better than mine... most of the Europeans I know have a better grasp of English grammar better than I do.

Anonymous said...

Hi,

I should have mentioned that I'm familiar with the Quine/Carnap-issue. I just don't know about any _recent_ discussions of that question.

I would've thought that if we don't have an interpretation function for a given set of symbolic strings, then we wouldn't be able to specify which of those strings were true and which false.
Surely we couldn't do that, but we could still specify the language by pointing to the mars and saying "the language spoken on that planet". And we could also ask wether ""Qwert yuiop" is true-in-martian-language iff philosophy is great fun." would be analytic/necessary if it was an instance of the t-scheme for this language.
Does that make any sense? I'm not sure why you think this question "wouldn't come up".

Anonymous said...

I know very little about this subject, but would like to ask a question:
Is there in fact such a thing as a necessarily true statement? The gist of the above seems to imply that these objects do not exist - is this correct?

Greg Frost-Arnold said...

Kim:
I misunderstood your point in the original comment; sorry. And I unfortunately do not know of any recent literature on the subject (if I did, I'd probably be reading it instead of posting about it on my blog).

Anonymous:
I think a large majority of current philosophers believe there are some sentences that are necessarily true. An example that I think most people could agree upon is "Everything is identical to itself" -- any other logical truth would work too. (Since you said you're new to the field, I'll mention the definition of 'necessary': A sentence S is necessarily true (resp. false) if and only if S is true in all (resp. no) possible worlds/ possible circumstances.)

Anonymous said...

A little late here, but just responding to your query for intuitions. My intuition is that "is smart" could have expressed the property of being from Mars in English. English would survive a small number of semantic shifts. (Not wholesale semantic disruption, just small shifts.) This could be so even if "English" is a rigid designator, since (as my intuitions have it) not all of the properties of English are essential to it.

But then this might not address your original worry with Stanley's claim. Maybe the right way to index a T-sentence is not just to English but something more fine-grained: English-with-the-very-semantic-properties-it-actually-has-now or something. Then the T-sentences would come out being necessary. And it seems like if you're doing semantics you have good reason to use a more fine-grained index like this.

Greg Frost-Arnold said...

Hi Geoff --

I think you're definitely right to say that "English would survive a small number of semantic shifts" -- for it has done so in the past, and it will in the future. English (or Turkish or Japanese) is clearly a "historical kind," like a species. English, like humans and frogs and fungi, changes over time and yet remains English nonetheless. (Similarly, just as there is variation among members of a species at any given time, individual speakers have their own idiolects.) But historical kinds are still a type of natural kinds (albeit a funny type). I clearly need to read up on them before I make any further pronouncements on this score.

And, unsurprisingly, I am very sympathetic to your re-formulation of my point about the necessity of instances of the T-schema. Are you thinking that the reason something more "fine-grained" is needed is to eliminate vagueness, ambiguity, etc. in the specification of truth-conditions? -- or did you have some other rationale in mind?

Kevin Winters said...

What if the "vagueness, ambiguity, etc." that is trying to be eliminated is an essential aspect of language? What would that say for attempts to eliminate them? If such is the case then any attempt to speak of English-with-the-very-semantic-properties-it-actually-has-now might itself be ineffectual, trying to crystalize what is in itself ambiguous.