I think Chalmers' paper on Ontological Anti-Realism is a great companion to Dummett's view. It makes the same sort of distinction between realism and anti-realism:

Ontological realism, at least in its strongest variety, holds that every unproblematic ontological existence assertion has an objective and determinate truth-value.

...

An ontological existence assertion has an objective truth-value if its truth-value does not depend on a context of utterance or a context of assessment: that is, if every ontological utterance of the same sentence has the same truth-value, and if the truth-value of these utterances do not vary with different ontological contexts of assessment.

...

Ontological anti-realism is the denial of ontological realism.

I find this approach to the debate between realism and anti-realism far more fruitful than the traditional interpretation (at least on here), that often leads me to saying "anti-realism isn't un-realism".

It's also why I tend to say that realism requires something like the correspondence theory of truth, and that attempts to use something like the T-schema to defend realism and attack anti-realism miss the point completely.

An ontological existence assertion has an objective truth-value if its truth-value does
not depend on a context of utterance or a context of assessment: that is, if every ontological utterance of the same sentence has the same truth-value, and if the truth-value of these utterances do not vary with different ontological contexts of assessment.

Can you unpack this for me?

An example that comes to mind:

A realist states that the tree blew over in a high wind the night before while nobody was around. The anti-realist will say it is neither true or false that the tree blew over, absent observational justification.

The realist thinks there is a state of affairs in which the tree did or did not blow over while nobody was observing it, while the anti-realist about physical objects does not accept this, since physical objects are what appears in perception, and nothing more.

The realist believes they can use inferences from other observed physical events, such as trees blowing over in high winds, and the weather report from last night, to justify the truthiness of the statement about an unobserved tree falling in the woods. The anti-realist won't accept those inferences.

Isn't it more along the lines that whilst both the anti-realist and realist can accept that weather reports/general facts about tree stability in the face of high winds etc can justify the claim that the tree fell over during the night, for the anti-realist the truth of that claim actually consists in those justifications whilst for the realist, those justifications allow one to infer the existence of a state of affairs that makes the claim true regardless of those justifications?

Incidently, if one is an anti-realist about truth in general, wouldn't that entail anti-realism for all domains about which true statements can be made?

303
Incidently, if one is an anti-realist about truth in general, wouldn't that entail anti-realism for all domains about which true statements can be made? — MetaphysicsNow

Yes, but is that a coherent position to take? Or is it just ancient skepticism? Actually, I'm not sure whether skeptics denied that claims could be true, only that we could know whether they were true or false.

I'm not sure being an anti-realist about truth is to be skeptical - at least, I'd need to see the argument. As far as I'm aware anti-realism about truth boils down to the idea that the truth of a statement consists in its justification (where what counts as justification will vary from domain to domain). I might be wrong about that, and I might be wrong about it not entailing skepticism, but you'd need to help me see the error of my ways if so.

Isn't it more along the lines that whilst both the anti-realist and realist can accept that weather reports/general facts about tree stability in the face of high winds etc can justify the claim that the tree fell over during the night, for the anti-realist the truth of that claim actually consists in those justifications whilst for the realist, those justifications allow one to infer the existence of a state of affairs that makes the claim true regardless of those justifications? — MetaphysicsNow

Maybe so for trees. A better one might be:

Realist: Life exists on Mars.
Idealist (anti-realist stripe): There is no truth to this statement until we observe evidence for or against life on the red planet.

However, the idealist might grant that life could have it's own perceptions, depending on the kind of life, and depending on the kind of idealist (whether they are strictly anti-realist about such claims).

As far as I'm aware anti-realism about truth boils down to the idea that the truth of a statement consists in its justification (where what counts as justification will vary from domain to domain). — MetaphysicsNow

Isn't that the same as deflationist? Which might be anti-realist, however, "the snow is white", isn't denying a state of affairs, it's just saying that this particular statement is made true by whether the snow is white, and nothing else more needs to be said.

I don't think "truth consists in justification" is the same as a deflationary account, although it's been a while since I've thought about these issues so I might be wrong. As far as I remember deflationary accounts just point to the fact that there is no distinction between asserting a statement and asserting that a statement is true, and then claim that this is all there is to say about truth. An anti-realist theory of truth as justification would seem to be offering rather more to say about what truth is than a deflationist would be comfortable with (a realist theory of truth as correspondence would also be offering rather more than a deflationist would be comfortable with as well).

Yes, but is that a coherent position to take? — Marchesk

I suppose if you take the Wittgensteinian approach to language and argue that meaning is use and then look to how the word "true" is used you'll see that it's used to refer to sentences that satisfy some standard of justification.

I suppose if you take the Wittgensteinian approach to language and argue that meaning is use and then look to how the word "true" is used you'll see that it's used to refer to sentences that satisfy some standard of justification. — Michael

Although actually it isn't quite this. It's really that the meaning (use) of a phrase like "the snow is white" is tied to the empirical (and so not recognition-transcendent) facts that justify our assertion of it. This is how Dummett phrases it. The realist takes the meaning of the phrase to be independent of any standard of justifying its assertion, but then we run into the problem of how to explain how we come to learn what a word or phrase means at all.

I suppose if you take the Wittgensteinian approach to language and argue that meaning is use and then look to how the word "true" is used you'll see that it's used to refer to sentences that satisfy some standard of justification. — Michael

While Dummett agrees with Witty on meaning being use, he disagrees that this makes metaphysical statements meaningless. Thus, his argument for realizing that metaphysical disputes are about the kind of logic one prefers. As such, the way forward for resolving these disputes is finding a way to justify the logic of the realist or anti-realist for a given domain.

An anti-realist theory of truth as justification would seem to be offering rather more to say about what truth is than a deflationist would be comfortable with — MetaphysicsNow

I see. So then on Dummett's account, the anti-realist qua truth rejects bivalence across the board. A statement only has a truth value if you can provide justification for it (as you stated previously).

Does that mean there's (1) no such thing as a proposition whose truth is unknown? Or just propositions where (2) we don't know how to figure out whether they're true or false?

Based on his intuitionist vs platonist example, it would seem it's the first, since the intuitionist denies that any mathematical truth exists beyond constructing it (providing a proof). But that runs strongly counter to common sense (when it comes to ordinary objects at any rate), and ordinary use of language.

Example 1: It's raining outside (where nobody has checked and there is no weather report).

Example 2. The MWI of QM is the correct one.

We know how to verify 1, but nobody has done so yet, so we don't know. Does that mean the truth is indeterminate until somebody looks outside?

For example 2, we don't know how to construct an experiment that would tell us which (if any) interpretation of QM is true.

Does that mean there's (1) no such thing as a proposition whose truth is unknown? Or just propositions where (2) we don't know how figure out whether they're true or false?

Not quite sure what you are getting at here. Anti-realism does allow for there to be sensible questions that we do not know the answer to - at least, it ought to, otherwise it runs not just counter to common sense but to any sense whatsoever. So, to use your example: Is there life on Mars? Now, we could turn this into the form: Is it true that there is life on Mars? Here we seem to have a proposition
P) There is life on Mars
the truth of which we do not know, yet which makes sense. How does an anti-realist account for this fact? I guess the response will be that it has meaning because we at least have some idea (perhaps many) of what would count as providing evidence for accepting it to be true.

In any case, you are correct to indicate that common sense is mired in realism. Berkeley tried to deny that his idealism ran counter to common sense, but was way off target if you ask me.

IBerkeley tried to deny that his idealism ran counter to common sense, but was way off target if you ask me. — MetaphysicsNow

He at least had God keeping things in the quad when nobody was around. Some of the idealist/realist debates on the old forum which didn't typically rely on God got pretty far out there, with talk of post-apocalyptic chairs at the end of the universe and what not.

How does an anti-realist account for this fact? I guess the response will be that it has meaning because we at least have some idea (perhaps many) of what would count as providing evidence for accepting it to be true. — MetaphysicsNow

I understand, but certain kinds of idealists are going to deny there is anything to Mars beyond our perceptions of it. And that definitely came up with the old debates around here. You had some idealists taking a hardline stance against anything existing that's not perceived by us.

I'm not sure the anti-realist can say there is anything determinate that hasn't been verified without committing themselves to realism of some form. Just being able to give an account of how it could be verified is not enough to say that a statement has a truth value, because then you're agreeing with the realist that bivalence is the case.

He at least had God keeping things in the quad when nobody was around

Yes, with God in the picture, Berkeley's idealism becomes a kind of realism, at least insofar as bivalence can hold of propositions that human beings could not even in principle come to know the truth of. It has the merit, then, of allowing for a reality independent of what you or I or our pets might think about it.

The only truly godless idealism that I'm familiar with tends to talk about permanent possibilities of perception being what underlies claims to the effect that unperceived/unknown facts/object etc exist. That just pushes the question back to what grounds those possibilities, of course.

Yes, with God in the picture, Berkeley's idealism becomes a kind of realism, at least insofar as bivalence can hold of propositions that human beings could not even in principle come to know the truth of. — MetaphysicsNow

I just read that Dummett held a similar position:

By means of science, we have made some progress towards understanding the world as it is in itself—we can point to ways in which scientific descriptions of the world are improvements on the description based on our bare perceptions, so our aspiration to know the world as it is in itself cannot be dismissed as an incoherent longing. But insofar as this aspiration is coherent, "in itself" cannot mean "without reference to the perceptions of any being."
https://www.iep.utm.edu/dummett/ — IEP

He then continues on to make a case for a universal perceiver who holds on things together, which would be God. That surprised me.

The only truly godless idealism that I'm familiar with tends to talk about permanent possibilities of perception being what underlies claims to the effect that unperceived/unknown facts/object etc exist. — MetaphysicsNow

People on here who towed the hardcore idealist line sans God or some universal consciousness would say that perception is brute, and there's no explanation to be had for why perception has the structure it does. Or at least we can't know why.

I think Dummett became a Roman Catholic in the end, so that he ends up with a theistic anti-realism isn't a big surprise to some extent. As for the hardcore idealist line, denying that perception has an explanatory structure seems to be going against the grain of the principle of sufficient reason, and that principle is probably what drives them to idealism in the first place.

Dummett made the argument that progress in metaphysics could be made by recognizing that the difference between realists and anti-realists on a subject was the kind of logic they preferred. A realist accepted the principles of bivalence in logic and verification transcendent statements, while an anti-realist rejected both. — Marchesk

I'm inclined to agree that disputes about ontology, especially in the academy, often seem to be mere disputes about logical preferences, about the way we ought to organize our grammar into things we agree to call "entities" and things we agree not to call "entities".

Often there are some criteria in play to motivate the preference, something like "explanatory power" and "elegance".

Can you say more about how the principles of bivalence and verification transcendent statements factor into the dispute between realists and antirealists in general?

In the case of realism versus idealism regarding physical objects, an idealist will reject bivalence and verification transcendent statements about physical objects. — Marchesk

Can you provide examples of the sorts of statements and arguments that your "idealist" rejects here?

They will instead have to present an alternative logic which determines what counts as sufficient evidence for making a statement about the physical true or false. — Marchesk

What kinds of "statement about the physical"?

One needn't play at metaphysics nor present logics in order to determine what counts as sufficient evidence for a statement like "Here is a cat". The materialist, idealist, and logician tend to agree with the rest of us along these lines.

Problems arise when we turn to questions like: Is the statement "Here is a cat" a statement about "the physical"? Do "physical objects" exist qua physical? And so on. I suppose such questions are "verification transcendent". But we don't need them in order to make sense of ordinary "statements about the physical". Much as we don't need any metaphysics at all to make sense of ordinary statements about the physical. We can dispense with materialism and idealism entirely, and so far as disputes about "realism" are concerned we can agree with Rorty that there is no privileged ontology.

Moreover, the idealist is something more than an anti-materialist. The idealist has his own idea about "what is" and "what is real", his own special brand of verification-transcendent statements. His position is no mere critique of the materialist's concept of "the physical".

The debate then turns into how justify either the realist or the anti-realist's logic for that domain. This means justifying the rules of inference used by one side or the other. If one can provide a justification proof for a logic, then the debate is resolved in it's favor, since the structure of the statements for the realist or the anti-realist have been proven to generate true statements, for that domain. — Marchesk

I'm not sure what to make of this conclusion.

What kind of justification is provided when bivalence and verification-dependence are rejected as criteria, both in general and in the special case you've focused on?

Isn't it likely that even after these justifications are provided, the two opposed parties will just remain opposed, each rejecting the other's preferred solution, while many of us in third parties will remain convinced there's no way to settle their dispute, or even convinced that there's nothing really at issue in such language games?