So for you, t-sentences are not false, but irrelevant?

Yes.

Rest easy Sam...

Hey? What argument? — Banno

With any theory of truth, you look for certain criteria to determine truth or falsehood. For instance, with correspondence theory, you look for correspondence between an idea and reality. Specifically, you need to determine if it's true that correspondence exists.

This means that in order to make sense of correspondence theory, you'll need to already know what truth is.

* Frege uses the wording "idea" and "reality" in The Thought.

Thanks.

Too many replies, and I'm off to the hardware store. Good to see a consensus developing, even if it is "Banno is wrong..."

Now, if you folk could just agree as to where I am wrong...

Meaningful correspondence to fact is not, and that is where convention has gone completely wrong. The reason:Not having gotten belief(or meaning) right to begin with. Stuck analyzing propositions and attitudes towards them. — creativesoul

There is, as a kind of ground to all our propositions, truths and facts, a pre-linguistic actuality to which they must submit. Analysis and conceptualization cannot gain purchase on that actuality, because to do so is to bring it into the linguistic domain, and there all we have purchase on is our communal perceptions and conceptions of what is the case,

So the necessary submission to pre-linguistic actuality is ever-present and ongoing, but cannot be analyzed in its own pre-linguistic "context": It was really Kant who first pointed this out. I guess this is why truth is unanalyzable, since any analysis presupposes the validity of what is purported to be analyzed.

the famous possibly apocryphal example of someone flipping off Wittgenstein, "What is the logical structure of this gesture?" — fdrake

It was Piero Sraffa, and it's almost certainly true. He was a very original thinker. (I read his book a lifetime ago.) They were friends at Cambridge.

With any theory of truth, you look for certain criteria to determine truth or falsehood. For instance, with correspondence theory, you look for correspondence between an idea and reality. Specifically, you need to determine if it's true that correspondence exists. — Tate

This is very close to what makes the most sense to me regarding falsification/verification. If we nix "it's true that" and swap "idea" with meaningful thought and/or belief it would resemble something very close to what I would be willing to defend.

Too many replies, and I'm off to the hardware store. Good to see a consensus developing, even if it is "Banno is wrong..."

Now, if you folk could just agree as to where I am wrong... — Banno

This presupposes that such an agreement is not already complete in the making aside from making it outwardly known.

:wink:

From where I sit Banno, you equivocate the term "fact" by virtue of vacillating between "fact" as propositions/states of affairs/the case at hand and "fact" as true statements. You also practice rendering all belief in terms of propositional attitudes which has the inevitable logical consequence - on pains of coherency alone - of limiting the very ability to acknowledge the brute fact that some meaningful true belief exists in its entirety prior to common language. You know the drill...

Meaningful true belief exists in its entirety prior to common language creation. Either both meaning and truth emerge prior to common language or meaningful true belief exists without meaning or truth.

That's our differences in a nutshell. Aside from that...

The large bulk of what you argue for, particularly the bits involving direct perception(although you never use those terms) is in perfect accord with my own position which, I believe, dovetails nicely not only with portions of your position, but also many other philosophers with whom you agree. I also nod towards the outright rejection of anything remotely resembling a category of stuff that is totally and completely unknowable but somehow the purveyor of such an approach want to then use this completely unknown empty category of things as a measuring device. In order to know that that is not a tree in and of itself, one must know what a tree in and of itself is. Such an approach is untenable. Things like that are also rejected by us both. Hell, even the rejection of private language is shared. The rejection of the conventional view on language that Davidson was arguing against using Mrs. Malaprop and other intuition pumps(thank you professor Dennett) is also a commonality between our respective viewpoints.

The main flaw I seem to see in your view could be roughly described as placing too much of the wrong kind of value upon language use.

You also seem to want to puke at the mention of anything remotely metaphysical, which is perhaps why you cannot even set aside your own current presuppositions in order to grasp how meaningful true thought and belief exists in its entirety prior to language.

Funny that almost a decade ago you and I participated in a much better debate than our recent one:Truth is prior to language. I argued in the affirmative. My view has evolved a bit since then as has your own.

It was Piero Sraffa, and it's almost certainly true. He was a very original thinker. (I read his book a lifetime ago.) They were friends at Cambridge. — Srap Tasmaner

Very cool. Thank you!

There is, as a kind of ground to all our propositions, truths and facts, a pre-linguistic actuality to which they must submit. Analysis and conceptualization cannot gain purchase on that actuality, because to do so is to bring it into the linguistic domain, and there all we have purchase on is our communal perceptions and conceptions of what is the case, — Janus

Submission of all our propositions to an unknowable actuality?

Analysis and conceptualization cannot gain purchase on that actuality... — Janus

That pulls the rug out from under our own analysis, does it not?

1. "p" is X iff p

Does (1) tell us the meaning of "X"? If not then the T-schema doesn't tell us the meaning of "true". It sets out the condition under which "p" is true, but nothing more.

This, perhaps, is the point @Sam26 makes when he says that the T-schema is irrelevant?

Think it's Davidson rather than Tarski. Tarski's work came out of considerations for formal languages right, — fdrake

From what I've read, it's Tarski, in "On the Concept of Truth in Formal Languages" (1935), where he's trying to resolve the liar paradox. Also, in the course of his thinking he uses Godel's incompleteness theorem as a model for his theory. This includes a meta-language to talk about our everyday language, in terms of truth, that's my understanding.

This, perhaps, is the point Sam26 makes when he says that the T-schema is irrelevant? — Michael

My point was, that it doesn't help us understand the meaning of truth in the object language. The object language is fine without it. It just adds a layer of confusion to the nature or meaning of truth.

In On the Concept of Truth in Formal Languages he says:

In § 1 colloquial Ianguage is the object of our investigations. The results are entirely negative. With respect to this language not only does the definition of truth seem to be impossible, but even the consistent use of this concept in conformity with the laws of logic.

...

If these observations are correct, then the very possibility of a consistent use of the expression 'true sentence' which is in harmony with the laws of logic and the spirit of everyday language seems to be very questionable, and consequently the same doubt attaches to the possibility of constructing a correct definition of this expression.

...

For the reasons given in the preceding section I now abandon the attempt to solve our problem for the language of everyday life and restrict myself henceforth entirely to formalized languages.

The object language is a formalized language, specifically the calculus of classes in his example.

I haven't spent much time on his works. I've mostly been reading secondary sources, and even there, it's sporadic. My main point, is that it doesn't help to define truth for everyday usage. I don't think it advances much in terms of that.

You've also got the weirdness that comes from convention T working for factual, declarative language and using it to, generically, set out the meaning of non-factual, non-declarative language through how the sentence somehow 'pictures' the relevant state of affairs. EG, like you can elucidate the speech act of flipping someone off through ""fdrake flipped someone off" is true if and only if fdrake flipped someone off". — fdrake

Just a side note, but Convention T and the T schema are different things.

Convention T. A formally correct definition of the symbol 'Tr', formulated in the metalanguage, will be called an adequate definition of truth if it has the following consequences:

1. all sentences which are obtained from the expression 'x E Tr if and only if p' by substituting for the symbol 'x' a structural-descriptive name of any sentence of the language in question and for the symbol 'p' the expression which forms the translation of this sentence into the metalanguage;
2. the sentence 'for any x, if x E Tr then x E S'.

So Convention T is the claim that an adequate definition of "true" will entail the T-schema for all sentences.

And, as I've mention before, this highlights the fact that Tarski didn't offer the T-schema as a definition of truth, but as a consequence of a correct definition. As I mentioned here, we still need an actual definition of "true".

A propositional attitude is a mental state held by an agent toward a proposition. — Wikipedia

What would have been wrong with calling such an attitude a sentential attitude? And making it a mental state held by an agent toward a sentence?

A proposition would be no less such an attitude than belief, fear, assertion, doubt etc. The proposition that snow is white would be (e.g.) the proposal that "snow is white" be accepted, or that "snow is white" correspond to reality, or that "snow is white" be true etc.

Not solving much, of course, as such attitudes generally don't.

But folks might be less prone to confuse sentence with reality.

And, as I've mention before, this highlights the fact that Tarski didn't offer the T-schema as a definition of truth, but as a consequence of a correct definition. As I mentioned here, we still need an actual definition of "true". — Michael

Makes sense! Thank you for the clarification.

For myself, at least, this is what iteration was supposed to address (and answer in the affirmative) -- by iterating actual sentences we can come to understand meta-lingual predicates through the comparisons of those sentences.

Truth takes on a meaning, then, but only through our using a natural language to analyze itself -- through our shared, in this case English, language. (and the sentences we choose to compare)

EDIT: Iteration not in some abstract space of reasoning, but rather, the iteration takes place dialogically. Just to be clear on that.

And yet, they are correct. I mean, you would not disagree that ('p' is true IFF p), would you?

So their use might be in providing some sort of grounding in relating meaning to truth. — Banno

I've already explained quite clearly why '"p' is true if p" says nothing about the meaning of truth, or even anything about the relationship between meaning and truth. To state it briefly, that is a simple repetition. Context is an essential aspect of meaning, and contexts are not repeated. They are each and every one of them, unique and particular. The expression you've given removes "p" from any and every context, so it effectively renders "p" as void of meaning.

What is important to note is that meaning is a feature of content. Content is material, in the sense of "subject matter", and it is dependent on interpretation. Formalizations (or formalisms) are intended to remove all content, to provide valid logic without the aspect of uncertainty inherent within content. But modern formalism attempts to apply formalizing tactics to content itself. This is a mistake, because we cannot remove the uncertainty from content, and the result is formalizations which are tainted with the uncertainty of the content which infiltrates. In other words, content, (which contains uncertainty), is imported into the formalization allowing uncertainty to contaminate the entire structure. The T-sentence proposed by Banno creates the illusion that we can have certainty with respect to meaning, or content, through exact repetition. But exact repetition is not a real aspect of meaning.

Now, if you folk could just agree as to where I am wrong... — Banno

I think consensus has been reached. The T-sentence does not do what you say it does.

1. "p" is X iff p

Does (1) tell us the meaning of "X"? If not then the T-schema doesn't tell us the meaning of "true". It sets out the condition under which "p" is true, but nothing more.

This, perhaps, is the point Sam26 makes when he says that the T-schema is irrelevant? — Michael

And, as I've mention before, this highlights the fact that Tarski didn't offer the T-schema as a definition of truth, but as a consequence of a correct definition. As I mentioned here, we still need an actual definition of "true". — Michael

OK, so let's say that, where Banno is wrong, is with interpretation. The T-schema doesn't say anything about the meaning of "truth", as a definition would. Nor does it provide a relationship between truth and meaning, because it removes truth from the context of meaning, thereby denying any meaning for truth. That is Banno's faulty interpretation.


It may be worthwhile to consider the relationship between "the correct definition of truth", and the T-schema. We may be able to produce as a valid conclusion, that "the correct definition of truth" is not possible, due to the relationship between truth and the particulars of the circumstances.

Very little of that chimes with what I understand of deflation. — Banno

Deflation implies that truth is relative, right?

My guess is that Tarski is basing his theory on false premises. For example, the liars paradox. — Sam26

Odd, since it's clear he explicitly deals with the liar by introducing levels of language. It's certainly not a premise in his argument, obviously.

So let me ask you again, are t-sentences correct? Even if they are irrelevant, is it false that ('p' is true IFF p)?

I'm asking because there is a substantive body of work, by the strongest logicians of the last hundred years, that depends on t-sentences. It would be odd if that were irrelevant. Worse if they were wrong.

With any theory of truth, you look for certain criteria to determine truth or falsehood. For instance, with correspondence theory, you look for correspondence between an idea and reality. Specifically, you need to determine if it's true that correspondence exists.

This means that in order to make sense of correspondence theory, you'll need to already know what truth is. — Tate

Sounds right. As mentioned, the substantive theories of truth try to tell us which sentences are true, and not what truth is.

That's a neat potted summation of that part of Davidson's early work. Well done and thank you. It's gratifying to be talking to someone with a bit of background.

And I agree with the sentiment of what you have said. ""fdrake flipped someone off" is true if and only if fdrake flipped someone off" looks like a trick. The action is taken and reduced to an extensional equivalence that sets out the time, place, and truth function of each element in the gesture; everythgn is in plain view, and yet something seems to be missing...

I get caught by a question from my old lecturer, something like "You are looking for the meaning of some utterance. If you have set out an extensional equivalence that shows exactly what is needed for the utterance to be true, what more could you need?"

It took me a while to realise that this was not a rhetorical question, but a challenge.

Because whatever you say in answer to this question can itself be incorporated into the T-sentence.

Hence I agree that the T-sentence "sets out everything to someone who already knows the meaning, but nothing at all to those who don't". Nicely phrased. The T-sentence presents an interpretation, if you prefer, and so already presumes that speaker and interpreter share both the world and their beliefs about it.

The result seems to be that whatever is missing from the analysis performed by the T-sentence is stuff that cannot be said.

Hence, in making use of T-sentence and the other components of Davidson's machine, we can explicate what is being said, what is meant. And in so doing we clarify what is not being said, what is ineffable. Like Wittgenstein's ladder, use the T-sentence to climb beyond language, then throw it away.

I get caught by a question from my old lecturer, something like "You are looking for the meaning of some utterance. If you have set out an extensional equivalence that shows exactly what is needed for the utterance to be true, what more could you need?" — Banno

Knowing how to map the extensional equivalence itself to an intended interpretation. When a given person writes a correct extensional equivalence, they've provided evidence that they understand how to form them. An account which sets out how to form such equivalences; namely, which language elements are in which sets; would be an account which sets out the meaning of language items. What it's actually doing is leveraging already known language items to form extensional equivalences without telling you the mechanism that maps the already known language items to the extensional equivalence classes. Only fleshing out the latter is a theory of meaning of natural language sentences.

At best, then, the T-sentence construction places a constraint on the space of possible semantics for (declarative) sentences of natural language. Rather than any particular semantic account.

I get caught by a question from my old lecturer, something like "You are looking for the meaning of some utterance. If you have set out an extensional equivalence that shows exactly what is needed for the utterance to be true, what more could you need?" — Banno

Lol. By the looks of it you've been inflicting that rhetorical gesture on people you've been debating with ever since! If you want to challenge it, recognise it for the rhetorical gesture that it is, and reframe the discussion. It's asking someone to go on a hunt for a "never before seen creature", whenever you come back with a creature, you just add it to the list of seen creatures and send them back out.

IE you challenge it by demanding a positive argument for why a theory which, it is imagined provides an extensional equivalence for every true statement specifies a pragmatic mechanism by which sentences obtain their meanings. What more could you want? A specification of the mapping of natural language statements to extensional equivalence classes; that would at least explain something.

The result seems to be that whatever is missing from the analysis performed by the T-sentence is stuff that cannot be said. — Banno

On the assumption that the meaning of an arbitrary sentence can be set out by the collection of other sentences which are true when it is.

There is also an issue of natural language not consisting entirely, or even mostly, of declarative sentences. Like "Your partner's love" - a perfectly valid fragment of language, you know exactly what it means, it inspires feeling, memory etc.

"Your partner's love" is true if and only if... Not even the poets could answer that one.

That's a neat potted summation of that part of Davidson's early work. Well done and thank you. It's gratifying to be talking to someone with a bit of background. — Banno

Also thanks. Over the years we've read and annotated decent chunk of Davidson in threads like this one. Wouldn't've studied him if you weren't a fan!