In the SEP article on truth, the author remarks

The contrast [of the coherence theory of truth] with the correspondence theory of truth is clear. Far from being a matter of whether the world provides a suitable object to mirror a proposition, truth is a matter of how beliefs are related to each-other.

When I stand back and read that, it seems like the the two conceptions of truth are distinct, but compatible.

In a way, it seems the correspondence camp is saying

'We want to use the word true to refer to propositions that participate in a certain relation with facts'

While the coherence camp is saying 'We want to use the word true to refer to beliefs that participate in a certain relation with other beliefs.'

If we acknowledge that they're using different senses of true, then we don't even have a substantive contest between camps.

The question I mean to ask isn't just about theories of truth. I see this sort of thing often enough.

For another example, a math major and friend of mine was describing a debate in his discipline about whether a type of algebraic object (rings, if I recall correctly) contained 0. He remarked. "Why is there even a debate? They're different things. Let's just have Ring_1 and Ring_0 and be done with it."

What are these debates about?

Are these kinds of so called debates about which metaphysical object (or abstract object or concept etc.) gets to wear a given label (e.g. 'Truth' and 'Ring') really debates if, right from the start, they aren't talking about the same thing?

  • I'm tempted to say that these debates are just attempts to describe how people use the concept and the accompanying term. But that doesn't seem to be it.They don't seem to be looking at just how people use it.It often sounds like they are talking about some sort of fact that is in a way independent of our use of the term.For evidence, consider the number of topic summaries that begin with something like 'In common parlance people use the term X to mean 'x' but philosophers have a different meaning in mind.' So the debates are often not even about how the concept or term is in fact being used
    – Hal
    Commented Dec 16, 2023 at 19:39
  • Quine had a slogan in this connection, something like, "Change the logic, change the subject." But perhaps disagreement can be formulated if the evaluation is for, "Which analysis best captures the spirit of the pre-theoretic meaning of the term?" (for whichever term is being analyzed). Commented Dec 16, 2023 at 19:41
  • Regarding truth and correspondence vs. coherence specifically, I asked a relevant question some time ago, but aside from comments there wasn't an answer given... Commented Dec 16, 2023 at 19:44

1 Answer 1


This has been a contentious issue in the history of philosophy. Wittgenstein and the Logical Positivists claimed that metaphysical arguments are just linguistic. That can't be true, of course--not in the usual sense of "linguistic" because you can translate the arguments from one language to another. A theory of truth could first be formulated in Polish and then translated to English, and the arguments against it that were first formulated in Polish could then be translated into English as well. If the argument were just an argument about how to use a language, how would that be possible?

However, there is a related position that focuses on concepts rather than on language. Concepts are language-independent. The same concept can be expressed in multiple languages. It makes sense to say, for example, that there are two different concepts of truth, regardless of which language you express those concepts in.

The question then comes down to whether there is one concept that is right and the others are wrong. I like to use the analogy of sets and relations. A set is the normal mathematical construct that everyone knows. A relation is a bit more obscure, but it is easy to describe: basically, a 2-ary relation R is an extensional association between two domains X and Y. If x represents a member of X and y represents a member of Y, then R(x,y) is a proposition that is possibly true of some x and y, and possibly false of others. A 3-ary relation has three domains R(x,y,z), and in general an n-ary relation has n domains.

Now we get to the example. It is common in set theory to say that a relation is nothing but a set of tuples. That is, if R(x,y) is true, then this means the pair <x,y> is in the set R. So there are only sets; relations are reducible to sets. On the other hand, you could say that a set is just a 1-ary relation. That is, if x is an element of S, then this just means that S(x) is true. So there are only relations; sets are reducible to relations.

Which is true? Are relations just sets or are sets just relations? Both sides of the argument are making a different claim. This isn't just a case of people using words differently. It's not a mere linguistic disagreement. But it is simply not the case that one of the positions is in any absolute sense correct and the other position is incorrect. They are both correct.

Does the same apply to all metaphysics? It's plausible that it does apply to a lot of metaphysics, but there are cases where it seems not to. One example where it doesn't seem to apply is your example of definitions of truth. This is because making a claim relies on your notion of truth. If we are to take the coherence camp at their word, then all they are saying is

(1) It is coherent to believe that truth is what it is coherent to believe.

This is much different from what the correspondence theory of truth implies. In the correspondence theory of truth, the coherence theory of truth is

(2) The fact of the matter is that truth is what it is coherent to believe.

By contrast, when the coherentist expresses the correspondence theory of truth, he is saying

(3) It is coherent to believe that the truth is what corresponds to the facts.

but what the correspondence advocate is saying is

(4) The fact of the matter is that truth is what corresponds to the facts.

For myself, I don't find (1) or (3) to be very interesting claims. So it's coherent? So what? I want to know whether (2) and (4) are true, not (1) and (3). So at least in this case, there seems to be a difference that goes beyond conceptualization.

  • Thank you. I'll note that examples (2) and (3) are curious when expanded. For example, consider the expanded (2): 'the fact of the matter is that propositions that correspond to facts are what are coherent to believe.' (1) and (4) seem trivial when expanded. In any case, that was a helpful answer. But I'm not seeing how the four examples show that the matter goes beyond conceptualization. Could you elaborate?
    – Hal
    Commented Dec 17, 2023 at 4:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .