Does the deflationary account of truth exemplified in the T-sentence: "'p' is true iff p" escape the logic of correspondence? Does it not rely on 'p' corresponding to p? That is, does not the phrase "snow is white" correspond to the actuality snow is white? If not, then how else could it make sense ? — John

Nagase (the excellent dude from the other forum) explained that Tarski's goals were not clear. He may have wanted to somehow redeem Correspondence, but the T-sentence doesn't do that.

For a good explanation of Tarski's definition of truth, see Scott Soames' Understanding Truth.

Deflationary conceptions of truth, or coherentist accounts such as Davidson's theory of meaning, which also centrally depend on Tarsky's T-schema, don't commit one to anything like an idea of correspondence between true sentences and facts. Quite the contrary, it seems to me, since the latter seems to presuppose a form of representationalism -- i.e. the idea that (doxastic) mental states, such as beliefs, are a matter of 'having' internal syntactically structured mental representations that represent the world to be (in some respect) thus and so. Deflationary conceptions don't commit one to representationalism at all. If for some agent to believe that P doesn't entail anything like her being related, in some way, to an 'internal' representation (some type of neural state, say), then there is no issue of correspondence that even arise to start with.

This leaves intact the idea that for a cognitive agent to be disposed to endorses the sentence 'P' as expressing a truth, and this truth being the fact that P (look up Jennifer Hornsby on identity theories of truth), then this agent is indeed representing (some feature of) the world to be thus and so (mamely, P). There being such representational acts doesn't entail that there are physical representations (syntactical items) that 'correspond' to the world. This correspondence notion remains rather obscure and quite dispensable. Beliefs (mental acts), and assertions (speech acts) that are expressive of them are actualizations of rational powers that belong to whole rational animals, and aren't features or items ('tokened sentences' or 'mental representations') somehow located in them. Indeed, Davidsonian 'holistic hermeneuticism' (as I would characterize his theoretical apparatus of radical interpretation) shares with Dennett's idea of the intentional stance a natural enmity to representationalism.

That's because, on such interpretivist accounts, definite propositional attitudes can only be ascribed to individuals on the basis of a global assignment of meanings(*) to all the sub-sentential terms that she employs when she endorses the truth of 'P', and this can only be effected in the context of a whole range of attitudes, background beliefs, intentions, etc. (with essential reliance of the constitutive ideal of rationality, and the principle of charity, in Davidson's account) Hence, in the essential background of her endorsement of the claim that P, as she expresses it with 'P', are involved her understandings of all the meaningful words being used. And those meanings (e.g. how she means what she thinks or says) can only be determined, even by her, within a whole 'conceptual scheme' (a notion that Davidson would of course have some reservations about that aren't important here), within, that is, a conceptually informed, and hence rationally structured, way to engage with the significant world (Umwelt) that human beings inhabit.

(*) Those meaning assignments to meaningful expressions (singular terms and predicates) are the 'axioms' that underscore the derivation of the T-shema (the 'theorems') of a Tarskian theory of truth.

I also wish Nagase would pay us a visit sometimes. That's part of his domain of expertise.

Yep. As it turns out, you'll do. :)

Thanks for both of your responses. It appears to me that you have answered the wrong question, though. You have answered or referred to the question as to how correspondence can be justified whereas I am asking the question as to how it could be eliminated given that for what we say to be about anything just is for it to correspond to that thing.

The logic of correspondence is even presupposed in any attempt to question it; so what could justify thinking that correspondence itself requires or could be susceptible of logical justification? Correspondence seems to be the basis of all justification as well as the basis of all sense, since claims about empirical states would be senseless without the assumption that the claims either correspond or fail to correspond ( the negative function of correspondence) to what they are claims about.

Claims, and the sentences used to make those claims, (or express the content of doxastic states such as beliefs or intentions) are different things. Correspondence theories purport to establish correspondence relations between worldly items such as facts, on the one hand, and linguistic items (or mental items that are somehow structured as linguistic items are) on the other hand. The rejection of representationalism entails that one can dispense entirely with the second relata of the alleged correspondence relation. So anti-representationalism (which is consistent with deflationary accounts of truth) makes nonsense of the very idea of correspondence. There isn't anything for the facts to intelligibly correspond to.

A correspondance theorist could object: If sentences aren't representations (such that facts can correspond or fail to correspond to them), then what are they? Well, in themselves they only are syntactically structured marks on paper, or vocal patterns, etc. Those linguistic items don't have any intrinsic semantic properties. That's a bit of a truism. They are meaningless in themselves but the standard, patterned, use that is made of them in a linguistic community can be interpreted. In that case the sentences inherit meanings from the fact that they are used to anchor intelligible patterns in the behavior (including, but not restricted to, the linguistic behavior) of rational agents.

So, you may want to say that a claim (a linguistic act, e.g. a sort of behavioral episode in the life of a speaker) can correspond or fail to correspond to the way the world is. This is true, in a sense, but it just amounts to saying that claims (or beliefs) can be distanced from the world through being false (that's the way John McDowell puts it). In the case where they are true, the world simply is as it is claimed to be by the person making use of the sentence. But this is just say what the corresponding T-shema already states. It is the deflationary account. This doesn't support any further claim of correspondence between facts and sentences. It merely states a condition for a linguistic acts, suitably interpreted, to be expressing something true.

I highly recommend Jennifer Hornsby's Truth: The Identity Theory, a paper that I found much illuminating when I first read it (as anything else that I read from Hornsby) and that I ought probably to read again.

So, you may want to say that a claim (a linguistic act, e.g. a sort of behavioral episode in the life of a speaker) can correspond or fail to correspond to the way the world is. This is true, in a sense, but it just amounts to saying that claims (or beliefs) can be distanced from the world through being false (that's the way John McDowell puts it). In the case where they are true, the world simply is as it is claimed to be by the person making use of the sentence. But this is just say what the corresponding T-schema already states. It is the deflationary account. This doesn't support any further claim of correspondence between facts and sentences. It merely states a condition for a linguistic acts, suitably interpreted, to be expressing something true. — Pierre-Normand

Yes, it seems self-evident that claims correspond or fail to correspond to the way the world is, and that even false claims (as opposed to nonsense) correspond to the way the world could be. I don't have a feel for what "distanced from the world" could mean; the way I see it, all claims are of or about the world, and its possible states.

When you say that "this is just to say what the corresponding T-schema already states" you seem to be suggesting an opposite relation of dependence to the one I see. As I see it the T-schema simply expresses in minimalist form what is implicit in the ancient self-evident and well established logic of correspondence, which was expressed by Aristotle as "to say of what is, that it is" and is itself thus dependent on that logic of correspondence for its very sense. Agreed, it is a deflationary account, because it says that correspondence is not further analyzable than that expression which is just the schema itself. I also think that there can be no theory of correspondence because correspondence is itself logically presupposed in all theories, and we would be using what we are purporting to explain, producing a mere circularity.

Thanks for link to the Horsnby paper; I'll certainly check it out; but my immediate reaction to the idea of an identity theory of truth is to say that it doesn't seem to make sense that the true things we say are the same as the things we are talking about in the saying.

To explain the deflationary account, here's Frege:

It is worthy of notice that the sentence "I smell the scent of violets" has the same content as the sentence "it is true that I smell the scent of violets". So it seems, then, that nothing is added to the thought by my ascribing to it the property of truth. — Frege, 1918

So what it amounts to is the claim that the sentences "'snow is white' is true" and "snow is white" mean the same thing. It explains the semantic equivalence of predicating a sentence as true and using that sentence. As such it doesn't have anything to do with the correspondence theory.

I am not addressing truth, though, but merely correspondence. I am simply saying that 'snow is white' corresponds to snow being white or else it is senseless; I haven't mentioned truth at all.

I am not addressing truth, though, but merely correspondence. I am simply saying that 'snow is white' corresponds to snow being white or else it is senseless; I haven't mentioned truth at all. — John

The title of the discussion is "Deflationary Truth and Correspondence".

OK, but that title is such as it is only because the T-schema is conventionally understood to be a deflationary account of truth. I am mainly concerned with the logic implicit in the schema and I think it is obvious that it is merely an expression of the logic of correspondence. But I would be happy to hear another account.

I do say in the second paragraph "putting aside the issue of the truth of what we say"...

OK, but that title is such as it is only because the T-schema is conventionally understood to be a deflationary account of truth. I am mainly concerned with the logic implicit in the schema and I think it is obvious that it is merely an expression of the logic of correspondence. But I would be happy to hear another account. — John

I've given another account; Frege's account.

That's an account dealing with truth not with the logic of correspondence.

You're confusing me now. You suggested that the deflationary approach to truth doesn't escape the logic of correspondence. I provided an explanation of the deflationary approach to truth that has nothing to do with correspondence. Saying that "'X' is true" and "X" mean the same thing says nothing about whether or not those sentences 'correspond' to anything, and there's no 'logic of correspondence' implicit in explaining the semantic equivalence of two sentences.

So, leaving aside truth take "'snow is white' iff snow is white". 'Snow is white' corresponds to snow being white. 'Snow' corresponds to snow, 'is' corresponds to being and ;'white' corresponds to white. Anything wrong with this?

So, leaving aside truth take "'snow is white' iff snow is white". — John

That's an ungrammatical sentence. What's it supposed to mean?

Its exactly equivalent to 'P' iff P.

Its exactly equivalent to 'P' iff P. — John

Which is an ungrammatical sentence.

Ungrammatical or not, you said earlier that "'X' is true" is equivalent to "X" so why would "'P' is true" not by the same logic be equivalent to 'P'?

You didn't say that "P" is true iff P. You said that "P" iff P.

If "'X is true' is equivalent to 'X' then why can't "'X' is true' iff X" be written as "'X' iff X". I am not particularly familiar with the conventions of predicate logic, so if I am missing something then I'm happy to be corrected. Although, all this is a bit of a distraction anyway because my main point is that 'snow is white' corresponds to snow being white; the argument is that if what we say does not correspond to anything then we cannot be saying anything about anything; we would be merely spinning in the void, so to speak.

If "'X is true' is equivalent to 'X' then why can't "'X' is true' iff X" be written as "'X' iff X". — John

The bit in bold doesn't make sense. That "X" is true iff X is not that "'X' is true" iff X.

Although, all this is a bit of a distraction anyway because my main point is that 'snow is white' corresponds to snow being white; the argument is that if what we say does not correspond to anything then we cannot be saying anything about anything; we would be merely spinning in the void, so to speak.

Does "the present king of France is bald" correspond to something? Or what about "Obama is a senator"?

OK, I see what you mean, now. But anyway, as I said above my point was about the sense of the statement 'snow is white' not about what would make it true. If it doesn't correspond to snow being white, then no sense could be made of the T-schema and, as a corollary incidental to what I am primarily concerned with, 'snow is white' would then not be true iff snow is white.

Does "the present king of France is bald" correspond to something? Or what about "Obama is a senator"? — Michael

Yes the words correspond to a possible state of affairs. There could be a king of France and he could be bald. Same with "Obama is a senator". You still seem to be focusing on puzzles about truth; I want to consider correspondence independently of truth.

If 'snow is white is true' is equivalent to 'snow is white' then why bother with the "is true" since that is the one part of the sentence that we don't know what it corresponds to?

If 'snow is white is true' is equivalent to 'snow is white' then why bother with the "is true" since that is the one part of the sentence that we don't know what it corresponds to? — John

Why bother indeed. The redundancy of it is the very thing that deflationary approaches to truth try to show.

You still seem to be focusing on puzzles about truth

Only because you brought up the deflationary theory of truth. It's strange that you brought it up but don't want to talk about truth. It seems to me to be a red herring. What you really want to ask about is the sensibility and correctness of a correspondence account of meaning.

I want to consider correspondence independently of truth. — John

You aren't the only person who thinks correspondence is fundamental. It's assumed in the representational theory of mind (RTM).

To understand why people would abandon that view in favor of something like knowledge externalism, I think you have to zero in on the challenges to representationism.

If you want to do a sort of group reading of this SEP article, we could.

Well, could I have referred to Deflationary Correspondence and expected anyone to know what I was taking about? I explained early on that I am not concerned with the problem of truth but with the question as to whether we can do without the logic of correspondence, and with the question as to whether even the Deflationary account relies on it for its own intelligibility.

In any case, you're right; I am interested in the viability, or even more strongly, the question of the indispensability, of the notion of correspondence for making sense of any account of meaning, and only secondarily to, and derivatively of, that primary concern, am I concerned with making sense of accounts of truth or of the correctness of meaningful statements.

So, since you have had a bit to say about what I am not primarily interested in, do you have anything to say about what I am primarily interested in?

As you say, I am tending to think that the logic of correspondence is fundamental to all discourse and hence to meaning itself. The article you linked is about Metaphysical Realism; I think it could certainly be worthwhile to do a group reading of that, and it is another issue, apart from what I have been trying to address here, that I am certainly interested in.

Do you think my position in relation to what I have been trying to deal with here entails a commitment to metaphysical realism? I had thought I was being cautious in only claiming that what we say must be able to correspond to what we experience, and leaving aside the further questions of whether, or how, what we say could correspond to a mind independent world.

I had thought I was being cautious in only claiming that what we say must be able to correspond to what we experience, — John

If I say I want an apple, I don't mean that I want a mental state (which is usually what we mean by experience.)

If you mean something other than that... then yea, you're probably committed to metaphysical realism

Actually I don't think of experiences as 'mental states'. In the Hornsby paper linked by Pierre there is an interesting distinction between 'thought' differently considered as 'acts of thinking' and as 'the content of acts of thinking'; and that reminds me of the distinction I make between 'experience' as 'acts of experiencing' and 'experience' as 'the content of those acts'. Interestingly neither of those seem to be neatly characterizeable as 'mental states'. What if I say that I want to experience eating an apple. By, and in, itself that would not seem to entail that there be any mind-independent apple to be eaten.

I'll be upfront and say that my preferred ontology is something like a form of ontic structural realism; I believe it is most plausible that there is some real, that is independently existent (to my experience), structure, whatever that structure might be, that gives rise to the experience of eating an apple. But, would it make sense to call that posited structure an apple, apart from within our everyday talk about shared formal identities? In any case, it is not such ontological or metaphysical questions and the like that I have been concerned with here, and I still can't see how what I have been concerned with necessarily entails such questions (but nonetheless I am interested in them).