It boils down to whether you can accept 1) humanity having a central role in what we call real, and 2) your own perception having a creative role.

Accepting that would be like realizing you're in a dream.

Yes :D

Because we are wired to, yes? So it's all Kant by way of Darwin. — Srap Tasmaner

Sometimes. But the point I'm making is really just that it's scalar, not binomial. At one end is something like the very concept of an external world (totally wired in), at the other, mirages (we all know that we've probably mistaken the perception)...

We recognise that our models are guesses, an amalgam of external causes and internal assumptions. The question is over the proportions of each in any given representation. At one end we act as if there are few assumptions and mostly external causes, at the other end we act as if our representations are mostly assumption with a sprinkling of external causes to pin it all on.

If the relative proportions of a matter are in dispute, it's almost certainly some matter which is at neither end, so arguments demonstrating that such ends exist are not really relevant to those matters which are in dispute.

According to anti-realism, there is no guarantee that every declarative sentence is determinately true or false. This means that the realist and the anti-realist support rival systems of logic.

So we do agree, at least on that definition. Then you are right, I have misunderstood Dummet; "Dummett is packing more into the notion of truth than the disquotational properties..."

I'll do some homework.

Edit: https://thephilosophyforum.com/discussion/11929/devitt-dummetts-anti-realism

The point I'm generally making here (and this goes for Hanover as well) is that no-one assumes all of their models are exact representations of an external reality, and no-one assumes none of them are. The choice over which we behave as if were true and which we approach with uncertainty is a psychological issue, not a philosophical one. — Isaac

I think the philosophical question is not as to which of our "models" (beliefs, ideas) are "exact representations of an external reality" (whatever that could even mean), but as to whether our beliefs and ideas can more or less reflect an external reality.

Which you can only say in hindsight, after catching your wife cheating on you. And it is only in hidsight that you will see certain past events etc. as evidence of the cheating, while at the time, you didn't.

To put crudely, a realist would need to maintain that his wife coming home late on a Wednesday is proof that she's having an affair. (For practical reasons, this is generally not feasible.) — baker

I'm not following your objection. It doesn't matter what I only see in hindsight, that is it doesn't matter that I didn't know at the time whether it was true or false that my wife had been having an affair: it was nonetheless either true or false at the time despite my ignorance.Why should the truth depend on my knowing it?

An exact genetic clone is in principle possible so this isn’t sufficient. — Michael

That a copy of you can be (in principle) created using your unique genetic signature does nothing to deny the way that unique signature identifies you as the organism that bears a history from birth to death. A copy of you is not you, because it was not born when you were, or in the way you were,and had not lived the life you had lived up to the time it was created by copying your DNA.

That this physical process maintains token identity isn’t a mind-independent fact. It’s not unreasonable to say that given sufficient physical changes the object is no longer the same, e.g with the ship of Theseus or the grandfather’s axe it can be warranted to assert that the ship and axe at the end of the story are a different ship and axe from the start of the story. Neither conclusion is wrong. — Michael

After I have made the distinction between a self-regulating organism which, if parts are replaced, replaces them itself, and an artificial construction (where if all parts are replaced it is arguable as to whether the construction should be thought to have retained its identity) I don't know why you would return to the artificial examples as if you thought they have any bearing on my argument.

I think the philosophical question is not as to which of our "models" (beliefs, ideas) are "exact representations of an external reality" (whatever that could even mean), but as to whether our beliefs and ideas can more or less reflect an external reality. — Janus

They do, some more than others. That's what I'm saying. We can't but believe in some external reality which our representations reflect, we also would be naïve in the extreme to not even believe we can be mistaken. So the question is already trivially answered.

Thus the only question of import is, given any particular belief, to what extent is is caused by an external reality and to what extent by internal assumptions. That it is, in some proportion, caused by both, is something we can't help but agree to, so it drops out of the conversation (or should). The actual proportions, in each case, are what matter.

Thus the only question of import is, given any particular belief, to what extent is is caused by an external reality and to what extent by internal assumptions. That it is, in some proportion, caused by both, is something we can't help but agree to, so it drops out of the conversation (or should). The actual proportions, in each case, are what matter. — Isaac

I half agree. On the one hand I see no reason to limit realism to semantic realism. Realism is about the existence of a mind-independent reality, not about how we try to describe reality with the use of human language. Dummet's concepts do not reflect the actual, social meaning of the words he is using. All this talk of anti-realism appears to me as artificial and needlessly confusing. Your gradient or scale idea is far clearer and more convincing. We don't need to chose a side for or against realism as if this was some sort of fight with only one possible winner. We can combine realism with the understanding that we can't have a direct access to reality, that truth is often hidden, including by our very ideas and concepts. So I agree by and large on the 'scale' idea.

On the other hand, to me the relation between our sense of realities and our ideas is perhaps more like a dance performed by two dancers (where each of them stays separate from the other but tightly collaborates with the other) than like a mixture between two ingredients. In the latter metaphor, it's as you said a question of dose. "The only question of import is, given any particular belief, to what extent it is caused by an external reality and to what extent by internal assumptions." (how much realism do you take with your idealism, or vice versa?). With the metaphor of the dancers, it's a question of how well these two collaborated to produce or affirm a belief. A question of precision of fit between our sense of reality and our ideas.

the relation between our sense of realities and our ideas is perhaps more like a dance performed by two dancers — Olivier5

I like this very much right up to the word "precision" in the last sentence. Science might be like a kind of competitive ballroom dancing, where you lose points for having your hand a few centimeters too high on your partner's back (I'm making this up, just by analogy to gymnastics scoring, which I used to have to try to understand), but in everyday life I'm with LW: "stand roughly there" is not by definition inexact given the circumstances.

Herbie Hancock used to tell a story about when he was playing with Miles: they were playing some tune and Herbie, as he puts it, played something that was just wrong, he knew it immediately and was mortified (he was after all pretty young) but, he goes on, Miles played something on top of it that made it work. His point was that Miles was missing some typical preconceptions and could allow almost anything to find its way into his music. That's also a kind of realism, right? Allowing the other to lead. Reality's going to do what it's going to do, so our dance cannot be completely choreographed but must also be at least partly improvised. @isaac points out that such surprise is expensive, so we try to make good predictions that will minimize the expense of revising our model. (And you get institutional momentum there that can lead you to throw out outliers in the data that you should have updated on -- you continue to follow the choreographed dance despite your partner's deviation.)

I have mixed feelings about the "separateness" of the dancers too. On the one hand, we have to distinguish between organism and environment; on the other, they are tightly, physically and chemically connected. When I eat part of my environment, I get raw material to maintain myself with, but also information that goes into the model. The relation between my environment and my model of it looks at least a little causal, it's just that the those causes hook into some pretty complex predicting machinery rather than doing something like knocking me down.

The word "representation" has to be handled really carefully if it's going to be anywhere close to adequate as a description of all of this.

Thus the only question of import is, given any particular belief, to what extent is is caused by an external reality and to what extent by internal assumptions. That it is, in some proportion, caused by both, is something we can't help but agree to, so it drops out of the conversation (or should). The actual proportions, in each case, are what matter. — Isaac

This sounds so reasonable, but I'm not sure I understand how it's supposed to work. If we took this quite literally, is a judgment with a higher external-to-internal ratio supposed to be better? Then you might be saying something like Hume did, that a wise man proportions his belief to the evidence. (Which is helpful, because in my mind I immediately connected this to a Ramsey-ish partial-belief model.) In practical terms, we compare theories sometimes on how much of the evidence they account for, more being better. But we also have to measure the assumption side, and it's true we like theories that require fewer assumptions, but sometimes the key step in theory development is adding an assumption so that you can account for a whole lot more of the evidence, so it is indeed a question of bang for the buck, of the ratio.

That, again, all sounds very reasonable as a practical matter, but there is a problem if we follow Quine and do not expect to be able to separate the internal and external bits of a given proposition. If we can't even do that, we can't sort our propositions into observations and theory. If it's holism all the way down, we can't even sort internal-ish from external-ish based on ratios because we can't ever determine those ratios. I expect your answer here is that we really don't: the ones that we deem most internal-heavy are wired in, and next come those that align with convention, in some sense, with a script or a narrative. Then the measure (what places a proposition near the center of the web rather than on the periphery) is not the internal-to-external ratio but this other thing, conformance to our narratives (and below that, narratives we're born with).

Yeah, your guess is right. I don't think the fact that we can theorise a scale of increasing external causes and concomitant decreasing internal assumptions means that we necessarily judge beliefs by their position on that scale, though it may well be a factor. The problem is, all we have with which to judge a belief's position on that scale is verification, and peer agreement. If I believe the pub is at the end of the road, that may well be partly made up of external causes (pub>perception>memory>belief) and partly of internal assumptions (pubs are usually at the ends of roads>belief). When I get to the end of the road, I'll find the pub there, but all I've really got is one more belief about the pub being at the end of the road, I haven't somehow gathered something of a different kind to my untested belief, just more perception/expectation combinations.

So yes, we might want to judge our beliefs by such a position on the scale, but we can no more access that data directly than we can the original data that informed the belief in the first place. I don't know if you recall a conversation we had with phforrest a while back covering the same ground. He was sure there was some solid ground he could reach about theories being less 'clashy' than others and I was arguing that the extent to which a belief clashes with other beliefs is itself a belief, we don't get out of the rabbit hole that way.

We can, however, develop habits of thinking which tend to give less surprising results when followed as means of building and sorting beliefs. We can theorise that this might be because they are further toward the end of the scale with a higher proportion of external causes, but we can't test them by that measure - it's a post hoc theory as to why they yield less surprising results, not a means by which we find them in the first place.

But we need to take care when applying the principle of surprise, it's only (in this case) a measure of the match between our priors and the incoming data. It's not to be confused with a sort of correspondence theory, surprise is not generated by a lack of correspondence between belief and external word, it's caused by a mismatch between priors and input data - that input data might well be a perceptive feature which itself is the output from some other model whose degree of external causality is completely unknown to us. It links up to the external causes only on the assumption that the chain of models is more influenced by forward acting updates (of priors) than backward acting suppression of surprising data. Clearly with our interaction with the physical world, this is going to be the case, but the more complex the beliefs get, the less we can be sure of such a favourable balance, until, with our most conceptualised beliefs, the backward acting suppression of surprising data might rule the roost and surprise reduction is going to make matters worse, not better (if correspondence with some external reality is our goal).

I think this is where my kind of Ramsey-Quine-Friston monstrosity gets us. Our beliefs are models which produce tendencies to act, but those actions might be to update priors to better match the data, or to suppress the data to better match the priors, or to reach out to the external causes to make them yield less surprise-inducing data in the first place, and by and large, the choice of tactic in each individual case is a largely a psychological habit resulting from embedded narratives.

That's also a kind of realism, right? Allowing the other to lead. Reality's going to do what it's going to do, so our dance cannot be completely choreographed but must also be at least partly improvised. — Srap Tasmaner

Yes of course, it's not perfect. Keeping up with the tempo is important, in real time. More important than perfection.

. isaac points out that such surprise is expensive, so we try to make good predictions that will minimize the expense of revising our model. (And you get institutional momentum there that can lead you to throw out outliers in the data that you should have updated on -- you continue to follow the choreographed dance despite your partner's deviation.) — Srap Tasmaner

Deviation is also creative though, as in your jazz example. As is darwinism. By definition, all surprises are not bad. Some surprises are good so we need to keep our mind and our societies open to novelty and change. A balance between the expected, the classical, and the novel. Something like that.

They do, some more than others. That's what I'm saying. We can't but believe in some external reality which our representations reflect, we also would be naïve in the extreme to not even believe we can be mistaken. So the question is already trivially answered.

Thus the only question of import is, given any particular belief, to what extent is is caused by an external reality and to what extent by internal assumptions. That it is, in some proportion, caused by both, is something we can't help but agree to, so it drops out of the conversation (or should). The actual proportions, in each case, are what matter. — Isaac

I agree with this. I'm not sure we could work out "proportions" very definitely, but in principle I think it's a good way of looking at it. But it would seem to be incompatible with any form of anti-realism.

After considering the Devitt article, we might proceed.

According to anti-realism, there is no guarantee that every declarative sentence is determinately true or false. This means that the realist and the anti-realist support rival systems of logic.

Realism is about there being stuff. Whether our statements about that stuff are true or false is incidental to realism. Whether we understand things about that stuff is also incidental to realism.

A realist might well adopt a three-valued logic with regard to statements. Nothing in realism locks the realist into a particular logical system.

That is, it seems what is loosely called semantic realism, the view that realism must make use of a correspondence theory of truth, is a bit of a straw man.

Or if you prefer, antirealism is a theory about belief, and has little to do with truth.

Nothing in realism locks the realist into a particular logical system. — Banno

:up:

Or if you prefer, antirealism is a theory about belief, and has little to do with truth. — Banno

Dummett's antirealism is the position that truth isn't recognition-transcendent or bivalent. It's very much about truth.

Realism is about there being stuff. — Banno

It's about more than that. From Devitt's paper:

I seek first a statement of the doctrine of Realism that captures its traditional opposition to idealism and phenomenalism about common-sense entities. There are two dimensions to this doctrine: first, a claim about what exists; second, a claim about the nature of
that existence. To capture the first dimension we can say that it is common-sense physical entities that exist. Words that frequently occur in attempts to capture the second are 'independent', 'external', and 'objective'. The entities must be independent of the mental; they must be external to the mind; they must exist objectively in that they exist whatever anyone's opinions. We can capture both these dimensions well enough in the following doctrine:
Common-sense physical entities objectively exist independently of the mental

You tend to avoid committing to this "second dimension" of realism, despite the fact that it is central to the position. As I have repeatedly said, antirealism isn't unrealism. Antirealists accept that "there is stuff", and can even accept that "common-sense physical entities exist." What they deny is that "stuff" (or "common-sense physical entities") objectively exist independently of the mental.

Whether our statements about that stuff are true or false is incidental to realism. Whether we understand things about that stuff is also incidental to realism.

A realist might well adopt a three-valued logic with regard to statements. Nothing in realism locks the realist into a particular logical system.

That is, it seems what is loosely called semantic realism, the view that realism must make use of a correspondence theory of truth, is a bit of a straw man. — Banno

As I mentioned in the other topic:

Regardless of whether or not the metaphysical realist explicitly claims to be a semantic realist when he argues for the “objective independent existence of common-sense physical entities,” it may be, as Dummett believes, that this position entails semantic realism.

If these independent entities are the things we talk about and the things that determine our statements to be either true or false then is this not recognition-transcendent and bivalent truth-conditions?

Or would you say that metaphysical realism as Devitt describes it is compatible with semantic antirealism as Dummett describes it? Is the “objective independent existence of common-sense physical entities” compatible with something like a verificationist account of meaning and truth? If not then proving the latter disproves the former.

As I have repeatedly said, antirealism isn't unrealism. — Michael

Then is seems we are both realists, except that you call yourself an antirealist.

Odd.

Then is seems we are both realists, except that you call yourself an antirealist. — Banno

Good one.

Maybe antirealists are only realist about antimatter, but not about the other stuff? ;-)

:smirk:

Then is seems we are both realists, except that you call yourself an antirealist.

Odd. — Banno

Try addressing the actual points raised in my post instead of deflecting with a non sequitur.

I wonder also if Anscombe's direction of fit works here. It's the difference between the list you take with you to remind yourself of what you want to buy and the list the register produces listing the things you actually purchased. The intent of the first list is to collect the things listed; of the second, to list the things collected. The first seeks to make the world fit the list, the second, to make the list to fit the world. Is it that anti-realism applies to ethics and aesthetics because we seek to make the world as we say, while realism applies to ontology and epistemology because we seek to make what we say fit the world? ↪Tom Storm, this is were I'm up to. — Banno

Try addressing the actual points raised in my post instead of deflecting with a non sequitur. — Michael

Hmm. Let's check. Did I address your last post?

Dummett's antirealism is the position that truth isn't recognition-transcendent or bivalent. It's very much about truth. — Michael

Sure, that's Dummet's anti realism. It takes realism to be the view that truth is recognition - transcendent. The Devitt article questions this, sets out realism

...as the view that physical entities exist independently of the mental. — Banno

That is, that realism is independent of theories of truth. Hence, Dummett is not criticising realism as it is generally understood - he is criticising a new invention of his own, semantic realism. So it is open for someone - you - to accept Dummett's argument and yet be a realist.

Realism is about there being stuff.
— Banno

It's about more than that. From Devitt's paper: — Michael

What more? Here's the stuff you bolded:

...a claim about the nature of that existence.

Words that frequently occur in attempts to capture the second are 'independent', 'external', and 'objective'. The entities must be independent of the mental; they must be external to the mind; they must exist objectively in that they exist whatever anyone's opinions.

Common-sense physical entities objectively exist independently of the mental — Michael

...all of which reinforces the point made above, that Dummett's antirealism is not addressing realism per se.

You tend to avoid committing to this "second dimension" of realism, despite the fact that it is central to the position. As I have repeatedly said, antirealism isn't unrealism. Antirealists accept that "there is stuff", and can even accept that "common-sense physical entities exist." What they deny is that "stuff" (or "common-sense physical entities") objectively exist independently of the mental. — Michael

So both realists and what you call antirealists accept that there is stuff, and the difference comes down to the phrase: "objectively exist independently of the mental".

As argued previously, "objectively" is either a weasel word, or it adds nothing to the statement. So the difference comes down to antirealists accepting that there is stuff but that this stuff is not "independent of the mental". Hence the antirealist is in the contrary position of declaring that there is stuff while insisting that it is always part of our mental world, which is Berkeley's idealism, or declaring that there is stuff but it is not part of our mental word, which is transcendental idealism.

Hence the antirealist is an idealist. But then you write:

Regardless of whether or not the metaphysical realist explicitly claims to be a semantic realist when he argues for the “objective independent existence of common-sense physical entities,” it may be, as Dummett believes, that this position entails semantic realism.

It may be that realism entails semantic realism. But as just argued, semantic realism is just idealism. So this would become the argument that realists are all idealists.

But realists are not idealists, hence not all realism is semantic realism.

But if you are not an idealist but an antirealist, and antirealism is realism, then is seems we are both realists, except that you call yourself an antirealist.

So we are left with this:

If these independent entities are the things we talk about and the things that determine our statements to be either true or false then is this not recognition-transcendent and bivalent truth-conditions?

(my bolding)
...and with mind-independent entities that we supposedly talk about; my opinion on the issue is I hope clear, see the private language thread.

Or would you say that metaphysical realism as Devitt describes it is compatible with semantic antirealism as Dummett describes it? Is the “objective independent existence of common-sense physical entities” compatible with something like a verificationist account of meaning and truth? If not then proving the latter disproves the former.

...and the substantive issue of how truth and realism relate. What I've long advocated is Tarski's analysis of true statements, together with a roughly deflationary view of truth; hence my approach does not depend on correspondence per se.

As for verificationism, is seems to me that it relies on an unsustainable differentiation between theory and observation - I've mentiond that a few times in this thread, it's a common criticism, and I think the answers that have been priovded are inadequate.

So I think

Then is seems we are both realists, except that you call yourself an antirealist. — Banno

...a cogent summation of our relative positions.

I have had what seemed like profound experiences when under the influence of hallucinogens, when listening to music, during sex and also when writing, when painting and when playing piano, both when under the influence of hallucinogens and when not under their influence. I've had even profounder experiences (although very brief in duration) when mediating. I draw no conclusions from those experiences. — Janus

Why not if you don’t mind sharing? Healthy skepticism? Unreliably replicable? Since humans are able to have these “profound” experiences, do you think that says anything or holds any merit about human nature/purpose, removed from the actual substance of the experience itself?

We're stuck playing tennis in our heads between 'object' and 'subject' with an invisible, or non-existent, ball called 'metaphysics'. There is no clear winner and it just might be that there actually isn't even a game being played at all. — I like sushi

There definitely is not one for the Advaitin, or a sternly committed monist.

Why not if you don’t mind sharing? Healthy skepticism? Unreliably replicable? Since humans are able to have these “profound” experiences, do you think that says anything or holds any merit about human nature/purpose, removed from the actual substance of the experience itself? — Ignance

I draw no conclusions from altered states, however profound they may seem, because I see no reason to. As for "purpose", I suppose the attractive nature of such states could lead me to seek them again. I certainly wouldn't want to infer any general human purpose on account of them, if that was what you were alluding to. What I've said is probably inadequate, but I hope it answers your question.

I draw no conclusions from altered states, however profound they may seem, because I see no reason to. As for "purpose", I suppose the attractive nature of such states could lead me to seek them again. — Janus

You and me both on the seeking them again part… :joke:

I certainly wouldn't want to infer any general human purpose on account of them, if that was what you were alluding to. — Janus

I didn’t mean like a universal purpose that you may come up with that seems to make sense within that experience, I meant the actual nature of such “profound” experiences being able to be had in the first place, do you think it says anything, or is it just a feature of consciousness in a way? (that’s what i meant by removed from the actual substance of the experience)

As argued previously, "objectively" is either a weasel word, or it adds nothing to the statement. — Banno

Those were Devitt's own words about what it means for him to be a realist. He clearly believes that it adds something to the statement. And it's not just him:

https://plato.stanford.edu/entries/realism/

On what it means to be a realist:

There are two general aspects to realism, illustrated by looking at realism about the everyday world of macroscopic objects and their properties. First, there is a claim about existence. Tables, rocks, the moon, and so on, all exist, as do the following facts: the table’s being square, the rock’s being made of granite, and the moon’s being spherical and yellow. The second aspect of realism about the everyday world of macroscopic objects and their properties concerns independence. The fact that the moon exists and is spherical is independent of anything anyone happens to say or think about the matter.

...

In general, where the distinctive objects of a subject-matter are a, b, c, … , and the distinctive properties are F-ness, G-ness, H-ness and so on, realism about that subject matter will typically take the form of a claim like the following:

Generic Realism:
a, b, and c and so on exist, and the fact that they exist and have properties such as F-ness, G-ness, and H-ness is ... independent of anyone’s beliefs, linguistic practices, conceptual schemes, and so on.

On what it means to be an anti-realist:

Non-realism can take many forms, depending on whether or not it is the existence or independence dimension of realism that is questioned or rejected. The forms of non-realism can vary dramatically from subject-matter to subject-matter, but error-theories, non-cognitivism, instrumentalism, nominalism, relativism, certain styles of reductionism, and eliminativism typically reject realism by rejecting the existence dimension, while idealism, subjectivism, and anti-realism typically concede the existence dimension but reject the independence dimension.

The notion of independence is very much central to the disagreement. You appear to have recognised as such here, where you admit to being a mathematical anti-realist. Numbers, although real, are not independent in the same way that you might say chairs are. That's why you're a realist about chairs but not about numbers.

What I've long advocated is Tarski's analysis of true statements, together with a roughly deflationary view of truth; hence my approach does not depend on correspondence per se. — Banno

This view is insufficient. Regarding the fact that "murder is a crime" is true iff murder is a crime, does murder being a crime satisfy the existence and independence conditions of realism as described above?

Tarski's analysis of true statements ("X" is true iff X) and the deflationary view of truth (to assert that "X" is true is to assert that X) are consistent with more substantive accounts of truth (or if you prefer, meaning), such as correspondence, coherence, or verificationism. What does it mean for murder to be a crime? Unless you're arguing for something like quietism, one of these more substantive accounts needs to be made.

According to realists, our assertions (attempt to) refer to things and properties that have an independent existence ("the cat is on the mat"), and these things and properties determine our assertions to be either true or false. How is this anything other than a correspondence theory of truth?

Hence the antirealist is in the contrary position of declaring that there is stuff while insisting that it is always part of our mental world, which is Berkeley's idealism, or declaring that there is stuff but it is not part of our mental word, which is transcendental idealism.

Anti-realism is more complex than that. Returning to the ship of Theseus, that the Theseus and the Perseus are the same ship is to be understood according to anti-realism (being the same ship isn't an independent property of some independent object but a way we think and talk about the ship). That's not the same as saying that the ship is made of some mental stuff. Related to this is the view of ontological reductionists, who might claim that something like the fundamental particles of the Standard Model have an independent existence, but something like being a chair "is an epistemological phenomenon that only exists through analysis or description of a system, and does not exist fundamentally,"¹ which would be an anti-realist view of chairs although not necessarily idealist.

¹ https://en.wikipedia.org/wiki/Reductionism#Ontological_reductionism

The T-sentence can be taken in different ways. It can represent truth deflation, which means the truth predicate has a social function, but no meaning beyond that.

For a deflationist, which Banno claims to be, it doesn't make sense to say there are unknown truths, because there can be no unstated truths. We may speak of unknown truths, but this kind of talk must be deflated.

So how can one be a realist without correspondence theory? With a large dose of Davidson, it's possible, but realism is just dogma added on at the end.

And of course: what is asserted dogmatically can be rejected dogmatically.

So how can one be a realist without correspondence theory? With a large dose of Davidson, it's possible... — frank

From The Structure and Content of Truth:

The realist view of truth, if it has any content, must be based on the idea of correspondence, correspondence as applied to sentences or beliefs or utterances - entities that are propositional in character; and such correspondence cannot be made intelligible. — Davidson

He's talking about truth realism.