We have a sentence "the cat is on the mat", we have the cat on the mat, and we say that the former is about or describes the latter. Is that mysterious? I don't really think so. So why would it be mysterious to say that the former corresponds to the latter? — Michael

Sentences are not favored as truthbearers outside artificial systems. Propositions work better for that purpose, although they're abstract objects. There's no 'aboutness' to a proposition. It's the content of an uttered sentence, which can take many forms: usually speech or writing.

How would you say a proposition corresponds to a truthmaker? Where do we look to see this relation?

It seems to me that the counterfactual says something about some other possible world. — Michael

Sure, by way of distinguishing it from ours. "If Hitler had not committed suicide, ..." says right up front that in fact he did. You can't be a counterfactual if you don't start with a fact you're countering. And I don't know how else to take "then he would have been executed by the Allies" except as a statement about what the Allies were like, what sort of action they were likely to take. How is any of this not also about our world?

What would it mean for the truth to not be about our world? — Michael

That's a fair question. Some of this is a little odd. If the kettle here hasn't quite come to boil yet, but might have, there is a nearby world where it has. In our world, "The kettle is boiling" is a falsehood, but not so far away it is a truth. Because these come in pairs, you get to say that "The kettle is not boiling" is a truth here. Every falsehood is also "about" our world in this degenerate sense, that its negation is a truth about our world. But this pairing business has another consequence, that you aren't compelled to go theorizing about negative facts and absent truthmakers and such; you only need the positives, because across all possible worlds you have all the positives instantiated -- somewhere. The negatives only duplicate (and then some!) what we already have. Instead of saying there's no truthmaker here for some sentence, you get to say a given sentence does have a truthmaker, it's just that it's somewhere else.

And if you take this positives-only approach, then the question is precisely whether that truthmaker is here, whether it's ours, whether it belongs to this world or another, because it does belong to some world somewhere.

Sentences are not favored as truthbearers outside artificial systems. Propositions work better for that purpose, although they're abstract objects. There's no 'aboutness' to a proposition. It's the content of an uttered sentence, which can take many forms: usually speech or writing.

How would you say a proposition corresponds to a truthmaker? Where do we look to see this relation? — Tate

I think you're making things far too complicated. We use speech and writing to talk about/describe the world. If there's nothing mysterious about this then there's nothing mysterious about correspondence.

I think you're making things far too complicated. — Michael

It's not me. This is a long standing objection to correspondence: that it lacks analytical clarity.

We use speech and writing to talk about/describe the world. If there's nothing mysterious about this then there's nothing mysterious about correspondence. — Michael

And yet, "It's fuzzy" isn't really truth apt until you know the context. At that point, you have an abstract object on your hands.

And yet, "It's fuzzy" isn't really truth apt until you know the context. At that point, you have an abstract object on your hands. — Tate

What abstract object? All I see there is a sentence with no explicit referent.

What abstract object? All I see there is a sentence with no explicit referent. — Michael

And that's not truth apt.

Sure, although I don't understand the relevance of this?

Sure, although I don't understand the relevance of this? — Michael

Me neither

And that's not truth apt. — Tate

So you're using "proposition" to mean something like, fully saturated -- I'm thinking of the way Frege calls predicates "incomplete symbols" or something. "... is red" is a function, and needs a name there or a bound variable to be complete. But it turns out completeness in that sense only works for formal languages, and in everyday usage of natural languages you might need to disambiguate, you might need a certain amount of context or background knowledge, all manner of things before your statement is, as I was putting it, "fully saturated" and ready to be true or false.

I find that general approach reasonable, but how do you deal with the circularity? What I mean is, if asked how much context we need to pull in before a statement is truth-apt, the answer is something like "enough for it to be truth-apt." The initial answer anyway. I guess I'm asking for reams a theory, because I have dim memories of work on this problem. Just wondering if you have any sense of how such a project is faring.

you might need a certain amount of context or background knowledge, all manner of things before your statement is, as I was putting it, "fully saturated" and ready to be true or false. — Srap Tasmaner

Right. We don't assign the truth predicate to strings of words, but to the content of an uttered sentence.

What I mean is, if asked how much context we need to pull in before a statement is truth-apt, the answer is something like "enough for it to be truth-apt." The initial answer anyway. I guess I'm asking for reams a theory, because I have dim memories of work on this problem. Just wondering if you have any sense of how such a project is faring. — Srap Tasmaner

I didn't know anyone was researching that. :grin: I would guess you'd have to go in the direction of Chomsky and provide a theory of language acquisition. It appears that a fair portion of it is innate.

What I mean is, if asked how much context we need to pull in before a statement is truth-apt, the answer is something like "enough for it to be truth-apt." — Srap Tasmaner

Enough for the proposition to be understood, I would think. It would be difficult to understand a proposition without any context.

"How would you say a proposition corresponds to a truthmaker? Where do we look to see this relation?"


"I think you're making things far too complicated. We use speech and writing to talk about/describe the world. If there's nothing mysterious about this then there's nothing mysterious about correspondence."


"It's not me. This is a long standing objection to correspondence: that it lacks analytical clarity."


"We use speech and writing to talk about/describe the world. If there's nothing mysterious about this then there's nothing mysterious about correspondence."


I think sometimes we expect more from certain concepts than they give us, or we over analyze certain concepts in search of a some phantom that will answer our intellectual itch. Philosophers have a tendency to take concepts out of their natural habitat, and place them in an unnatural one.

I think sometimes we expect more from certain concepts than they give us, or we over analyze certain concepts in search of a some phantom that will answer our intellectual itch. Philosophers have a tendency to take concepts out of their natural habitat, and place them in an unnatural one. — Sam26

I think this is one of the things that Wittgenstein got right in the Philosophical Investigations. I'm not entirely convinced that meaning is as simple as use, but at the very least I think it's a good approach to dissolve some of the problems that philosophers effectively invent by injecting undue significance into a word (like "truth").

I agree.

I didn't know anyone was researching that. — Tate

I was thinking back ages ago, Barwise and Perry, situation semantics.

Enough for the proposition to be understood, I would think. It would be difficult to understand a proposition without any context. — Luke

Indeed, the point should have been made already that the only reason to go questing after non-circular criteria for the "saturation" of a statement is so you can formalize it properly. You have the option of fiddling with logic to do some of that, but if your target is classical logic, that's tenseless and contextless, which means all the statements have to be too.

A simplistic analogy for coders: if you write in a standard imperative style where there's state laying around all over the place, your functions may only need to take an argument or two and pull everything else they need from whatever's in scope; if you write in a functional style -- or, next door, Prolog -- then your functions might have to take a zillion arguments, or one or two plus a big fat one bundling a bunch of others together, because you have to carry the state around with you.

Real life is like the imperative example -- state is laying around and accessible, more state is implicated by your utterance, so you do understand things without maybe even knowing quite how you do, though you might be able to work out a lot of it. Logic isn't like that. Formal semantics isn't like that. If it's not explicit somewhere, it's no help at all.

So yeah in real life, it should be said again, we either don't face these problems or resolve them easily. Disambiguation, for instance, is the easiest thing in the world. This stuff is only challenging when you try to formalize it.

That's a fair question. Some of this is a little odd. If the kettle here hasn't quite come to boil yet, but might have, there is a nearby world where it has. In our world, "The kettle is boiling" is a falsehood, but not so far away it is a truth. Because these come in pairs, you get to say that "The kettle is not boiling" is a truth here. — Srap Tasmaner

@Michael - isn't this reasonably straightforward? You can say a counterfactual is about world x if it is evaluated in world x. For example, so "Hitler would have been executed if he had not committed suicide" consists of a comparison of two worlds, this one (in which Hitler committed suicide) and a 'nearest possible' one y in which Hitler had not committed suicide. We find that "Hitler has been executed" is true in y, so the counterfactual evaluates as true in x.

I suppose it's that it seems like it means something, but the attempts as specifying a meaning are not universal to the use of "... is true", and the correspondence theory creates unnecessary entities (facts in addition to true sentences, and the problem of non-referring names though I think that's been adequately answered by Luke for me for now), and even if we were to grant facts the division of which fact is important cannot be specified. The number of facts in the world are innumerable, like sentences. So, without an ability to spell out correspondence, we could substitute one fact for another and still claim truth.

Those have been the three arguments I've offered against correspondence so far.

Did you mean to quote stuff about Hitler instead of the kettle?

We find that "Hitler has been executed" is true in y, so the counterfactual evaluates as true in x. — fdrake

This is more or less the standard view right? But I have to say, "We find that ..." sure looks odd to me.

I guess if I'm going to keep wading into these waters, I'll have to study up some. Presumably you can proceed by defining worlds in which the Allies did and worlds in which they didn't execute Hitler, and then you'll have to defend some way of determining how far each sort is from us, which might be closest, which is close enough, and so on. I'll do my homework.

So my intent is to comment on each of the replies to @Sam26's request.

draws attention to the distinction between Truth and truth. Well worth keeping at the back of one's mind, since there are folk who hold to there being a statable Truth, usually religious, but also often philosophical. I'll take here a Wittgensteinian approach, that such things are ineffable, or they achieve only for aporia; hence, silence.

And I agree with Moliere that the small-t truth is embedded in language, and add that the discussion of T-sentences gives us an outline of how that works, in terms of the relation between meaning and truth,

talks of sentences which prevail. His analysis is in terms only of the illocutionary force of true statements, which is only part of the story.

writes of truth as "at least a relation between a sentence and something else", a theme played with in our recent discussion of arrows and targets. It seems to me that the "something else" here is roughly meaning; or if you prefer, use. Meaning is "why the LHS tracks the truth-value of the RHS". The holism Srap hints at would then be the holism of meaning, in a way familiar from Davidson Rorty and others.

poses the truth is unanalysable, but what counts as a simple depends on what one is doing, hence what is analysable or not analysable is a question of choice. Tarski presents us with an analysis of truth in terms of meaning.

equates truth and honesty, which is to mistake the logic of truth for its illocutionary force.

talks of the relation between truth and meaning: "It fits and makes fit language to world and world to language." Interestingly he introduces interpretation, Of course the rhs of a true T-sentence is an interpretation of the sentence mentioned on the left, after Davidson. The holism is there, with the addition of an aspect of truth as public, shared or communal, somethign that might be worth further work.

repeats mention of the social aspect of language, combined with what looks like the picture theory of meaning. Truth is defined negatively, "If the proposition misses the mark, or does not accomplish its goal, as a picture or a correlate of reality, then it is false".

refuses to let truth get a grip on the world, restricting it to "what is perceived or conceived to be the case", and so is answering the question "what is belief" rather than "what is truth". The oblique references to the communal and utilitarian nature of language remain.

continues the rejection of truth in favour of belief. Something to do with a correspondence between a mouse going behind a tree and biological machinery. I had difficulty following the discussion.

points to the issue of the incompleteness of coherence theories of truth.

also sees this incompleteness, seeing correspondence as explaining only empirical truths, but thinks we might remedy this by treating all truths as empirical. He sees deflation as too conservative, for reasons I was unable to follow.

So a few folk continue to rely on correspondence to explain truth, while a few have come to see it as inadequate. A few folk point to the social nature of language, all good, but not helping with the specific issue of the nature of truth. It's apparent that all language is conventional, except when it isn't, that all language is public, that all language involves interactions with the world as part of a community. None of this helps to isolate what it is that is true of truth...

I'll note that the minimal logical structure of truth displayed in a T-sentence is compatible with almost all the views expressed here, and suggest it as a consensus.

Also, we might agree that there is a close relation between meaning and truth.

talks of the relation between truth and meaning: "It fits and makes fit language to world and world to language." Interestingly he introduces interpretation, Of course the rhs of a true T-sentence is an interpretation of the sentence mentioned on the left, after Davidson. The holism is there, with the addition of an aspect of truth as public, shared or communal, somethign that might be worth further work. — Banno

I'd imagined an interpretation as a relation, rather than as the RHS ralata. It's what maps the LHS to the RHS rather than the RHS; the arrow itself, not its point. To speak of 'an interpretation' and make it the RHS of a T-sentence very narrowly circumscribes the notion of interpretation and gets you in the wrong frame of mind for tackling interpretation as a topic close to truth IMO. Can elaborate more on the relationship of interpretation to truth from that perspective if required.

Edit: agree with the rest of what you said though.

My take on interpretation is pretty much along the lines of Davidson's radical interpretation, which was modified by the man himself over time. It's that triangulation of my belief, your belief and the T-sentence.

The basic problem that radical interpretation must address is that one cannot assign meanings to a speaker’s utterances without knowing what the speaker believes, while one cannot identify beliefs without knowing what the speaker’s utterances mean. It seems that we must provide both a theory of belief and a theory of meaning at one and the same time. Davidson claims that the way to achieve this is through the application of the so-called ‘principle of charity’... In Davidson’s work this principle, which admits of various formulations and cannot be rendered in any completely precise form, often appears in terms of the injunction to optimise agreement between ourselves and those we interpret, that is, it counsels us to interpret speakers as holding true beliefs (true by our lights at least) wherever it is plausible to do (see ‘Radical Interpretation’ [1973]). In fact the principle can be seen as combining two notions: a holistic assumption of rationality in belief (‘coherence’) and an assumption of causal relatedness between beliefs – especially perceptual beliefs – and the objects of belief.

Luke also sees this incompleteness, seeing correspondence as explaining only empirical truths, but thinks we might remedy this by treating all truths as empirical. He sees deflation as too conservative, for reasons I was unable to follow.

So a few folk continue to rely on correspondence to explain truth, while a few have come to see it as inadequate. A few folk point to the social nature of language, all good, but not helping with the specific issue of the nature of truth. It's apparent that all language is conventional, except when it isn't, that all language is public, that all language involves interactions with the world as part of a community. None of this helps to isolate what it that is true of truth...

I'll note that the minimal logical structure of truth displayed in a T-sentence is compatible with almost all the views expressed here, and suggest it as a consensus.

Also, we might agree that there is a close relation between meaning and truth. — Banno

I was trying to point out that deflationism is also incomplete. If deflationism is only consensus then what is the point of science and testing new theories? What prompts the consensus to change if not some lack of correspondence that we discover between that consensus and the world?

Presumably scientific processes provide us with a justification for believing this or that.

It's a commonplace that science avoids labelling its theories true.

But I'm not following what you mean by "deflationism is only consensus".

poses the truth is unanalysable, but what counts as a simple depends on what one is doing, hence what is analysable or not analysable is a question of choice. — Banno

I think truth is the exception, per Frege's argument.

Tarski presents us with an analysis of truth in terms of meaning. — Banno

If it's an analysis, it's not a particularly informative one.

But I'm not following what you mean by "deflationism is only consensus". — Banno

According to deflationism, truth is no more than an endorsement of what is commonly believed to be true. If meaning is use, as per deflationism, then what is true is whatever most people consider to be true at a particular time. But what most people consider to be true changes over time, and I don’t think this can be accounted for by deflationism. The fact that we test new theories and observe the results of those tests and try and gain a better understanding is not a matter of endorsing what is commonly believed to be true. The inconsistency of some of those results with our current understanding indicates a correspondence and/or coherence element of truth.

↪Metaphysician Undercover equates truth and honesty, which is to mistake the logic of truth for its illocutionary force. — Banno

There is no logic of truth, you ought to realize that by now. If there was a logic of truth, then truth would just be a form of justification.

And, the fact that we often cannot distinguish between when a person is being honest, and when the same person is being dishonest, is clear evidence that "illocutionary force" is irrelevant here. Telling a truth, or telling a lie may have the very same illocutionary force, so the difference between the two lies somewhere else.

I think truth is the exception, per Frege's argument. — Tate

Hey? What argument?

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

The connection between meaning and use is not central to deflation. But it is central to my approach.

And the connection between meaning and use is not just taking a popular vote for meanings.

The main idea of the deflationary approach is (a) that all that can be significantly said about truth is exhausted by an account of the role of the expression ‘true’ or of the concept of truth in our talk and thought, and (b) that, by contrast with what traditional views assume, this role is neither metaphysically substantive nor explanatory. — SEP: Deflationism About Truth

Hence Tate is quite correct, where he talking about deflation:

If it's an analysis, it's not a particularly informative one. — Tate

The basic problem that radical interpretation must address is that one cannot assign meanings to a speaker’s utterances without knowing what the speaker believes, while one cannot identify beliefs without knowing what the speaker’s utterances mean. It seems that we must provide both a theory of belief and a theory of meaning at one and the same time. Davidson claims that the way to achieve this is through the application of the so-called ‘principle of charity’... In Davidson’s work this principle, which admits of various formulations and cannot be rendered in any completely precise form, often appears in terms of the injunction to optimise agreement between ourselves and those we interpret, that is, it counsels us to interpret speakers as holding true beliefs (true by our lights at least) wherever it is plausible to do (see ‘Radical Interpretation’ [1973]). In fact the principle can be seen as combining two notions: a holistic assumption of rationality in belief (‘coherence’) and an assumption of causal relatedness between beliefs – especially perceptual beliefs – and the objects of belief. — Banno

There's much to like in it! It's very language and statement focussed though. I think we've been through this difference in intuitions regarding perceptual belief formation a lot of times over the years, don't think we have to go through it again here.

From the Heidegger angle I've been taking in this thread though, T-sentences, causal relatedness and belief networks are all methods of statement adequation, by which a statement is assigned a place in a web of other statements through causal relationships and conditions of satisfaction. That takes the web as a given. In the same way that I've been harping on about treating an interpretation as a statement being in the wrong frame of mind to get at truth, I'd say exactly the same thing about basing an account of truth of statements as if it could stand by itself. It needs a joint account of how meaning is tied up with a sense of connection to the world; some of that is perceptual, some of that is conceptual, and some of that is practical. You'd doubtless agree to those things regarding the use of language, Heidegger points out that it applies to the connections themselves between statements; they've got their sense fleshed out by something much different.

It reads like the philosopher of language trope where they talk about statements and truth as a way of getting at our connection to the world, without thinking about how the focus on statements and their truth is a distorted picture. What Wittgenstein criticised about the pictorial theory of meaning (statement -> logical facts) also applies to a sentential one (sentence -> world); making a model of the world and forgetting it's a model.

I think we've had that discussion before too though. The years are long.

points to the issue of the incompleteness of coherence theories of truth. — Banno

Si, si señor!