Epistemic: To say that a proposition is true is just to say that it meets a high standard of epistemic warrant, and that we are thereby justified in asserting it. The search for truth and the search for justification are not separate goals. There is no more to truth than sufficiently powerful justification.
I’m honestly confused about this definition. Does it mean that once we define what truth is, we can then assert if a proposition is true or not ( truth and “justification” are in this sense intertwined)?
In this case, isn’t this just a broader case of Correspondence, or better, isn’t Correspondence just a particular subcase of Epistemic?
Perhaps an example of an epistemic theory of truth will help. Suppose you have a scientific community grappling with a certain set of problems. As their inquiry proceeds, they solve some problems and uncover new ones. In the process, they also greatly refine their methods. Imagine they eventually come to a stage where all their problems have been solved according to their own standards of warrant, and no open problems remain. One theory of truth (due to Charles Peirce) says that their beliefs at this stage of inquiry are true by definition. Truth is just what a community will arrive at at the ideal end of inquiry.
How is this different from a correspondence theory? Well, the correspondence theorist would say that even though the scientists have fully resolved their inquiry according to their own standards, and it seems like their beliefs are highly justified, it still might be the case that they got it wrong. It still might be the case that their beliefs don’t correspond to reality, in which case their beliefs are false. For Peirce, this claim makes no sense. According to him, what it means for a belief to conform to reality (or to be true) is for it to be a belief held by a community at the ideal end of inquiry. If we allow that our beliefs can be as justified as we could possibly make them and still fail to be true, then truth becomes a potentially unattainable goal, and for Peirce this would make truth a philosophically useless concept.
The epistemic theory of truth is generally held by philosophers who are skeptical of metaphysics, and who think metaphysical concepts are only valuable to the extent that they make a difference to our lives. If there is a metaphysical distinction that does not correspond to a distinction in what we observe or how we should behave, it should be discarded. This is why they think the theory of truth should be closely tied to our epistemic practices, the stuff we do when we are trying to find the truth.
OK, so… suppose this community then encounters some new experiences at time T that cause them to reject the solution (S1) to a problem previously considered solved, which they then re-close (with a different solution S2).
A correspondence theorist wasn’t sure before T whether S1 was true or false, they’re still not sure after T whether S1 is true or false, and the same goes for S2 (and, for that matter, anything else someone might assert). Presumably they will also make statements like S1 was justified before T and unjustified after T, that S2 is justified after T, and might disagree about whether S2 was justified before T.
An epistemic theorist asserts that S1 was true before T and is false after T, that S2 is false after T, and might disagree about whether S2 was true before T.
A deflationary theorist… um… is not committed to any particular position about the truth, falsehood, or indeterminacy of S1 or S2 at any time.
For Peirce, an end to all problems (and therefore an end to inquiry) includes the claim that there are no reecalcitrant experiences. It isn’t just a temporary end to problems, it’s a permanent end. This is why he refers to this as the ideal limit of inquiry. So the situation in your first paragraph wouldn’t apply. Peirce specifies things this way precisely to avoid the consequence that truth-values can change with time. I should say, though, that this is only one specific epistemic theory, and an early (and therefore kind of unsophisticated) one. I chose it for ease of exposition.
Well, OK, but then it seems Peirce’s conception of truth is just as potentially unattainable (and thus philosophically useless, by his own account) as the correspondence conception. Unless I’ve reached an ideal limit of inquiry—which of course I can’t really know I’ve done, and am unlikely to have done—then it seems I don’t actually have truth, even on Pierce’s account.
That aside, though, point taken about epistemicism != Peirce; presumably if I actually care about the latter I should just read Peirce. I’m just being intrigued by it. I do think I roughly understand the concept now; thanks for the explanation.
Many thanks to you and Vaniver, I have a clearer picture now (though this mostly looks like a debate over definitions, that can basically be solved by tabooing the word “truth”).
The three theories of truth are different attempts to taboo the word “truth”.
Isn’t this the point? Tabooing “truth”, one can see that the theories really speak about (slightly) different concepts. Going back to your previous example, if one theory claims the scientists have reached the truth and the second doesn’t, how does it change the reality? You can easily define some new words to correspond to the different concepts, and refer to the appropriate label under the different circumstances.
Tabooing “truth”, one can see that the theories really speak about (slightly) different concepts.
Then, you would merely choose which of the concepts is the one needed for a particular theoretical purpose. Right?
Wrong! The arguments go to the concepts’ coherence. This is why it’s philosophy, not lexicography.
For example, a correspondence theorist generally argues that the notion of an epistemological limit to which scientific findings converge need not exist and can never be established empirically. If correspondence theory is true, you aren’t allowed to use the Piercian limit. It’s a vacuous concept.
Or, the correspondence theorist argues that the epistemological limit of scientific investigation can’t even be defined without assuming a correspondence variety of truth (which the Piercian, in turn, argues can’t exist). The correspondentist argues that if you define truth at a limit, then you have to define the truth that science is converging as itself the result of a scientific investigation at an endpoint, and similarly for the concepts you use to define scientific investigation, etc. Thus, a Piercian view, it’s contended, produces an infinite regress.
It’s possible that both concepts are coherent, but that too would require a philosophical argument—and it’s an unlikely result here, at least in my opinion: it’s probably more likely that both concepts are incoherent than that both are coherent.
These kinds of conclusions, philosophical and lacking in direct application, help inform the priors one assigns to just about every scientific controversy.
Ok, this starts to sound more interesting, thank you for the reply. I tried to briefly google for “Piercian limit”, though and it didn’t turn out anything relevant. Any quick reference?
If correspondence theory is true, you aren’t allowed to use the Piercian limit. It’s a vacuous concept.
(blink) If I accept acorrespondence theory of truth, it seems that correspondence theory is not the sort of thing that is allowed to have a truth value. And if I reject a correspondence theory of truth, then I ought not believe that correspondence theory is true. So it seems that “correspondence theory is true” is necessarily false. No?
That’s an excellent argument if it’s the case that correspondence theory is not the sort of thing allowed to have truth values under correspondence theory. Why do you say it’s not?
Well, using pragmatist’s cited definition of correspondence theory, a proposition is true if and only if it bears some sort of congruence relation to a state of affairs that obtains.
What state of affairs is “correspondence theory is true” congruent with?
I can’t think of any.
If you can, I’ll happily be convinced my argument doesn’t hold, but basically it seems to me that correspondence theory lays out a framework for thinking about truth, just as governmental constitutions lay out a framework for thinking about law. Correspondence theory itself is no more true (or false) than constitutions are legal (or illegal).
I might still want to know which concept I’m using when I say something is true. I might also want to know which concept you’re using when you say something is true. Sure, if I don’t know that (or am not confident I know that) then we can taboo “truth,” but that gets unwieldy; if we can agree on a shared referent then communication is more efficient. Tabooing key words is sort of like running code under a debugger… a great way of identifying points of failure, but not the way I want to live my life. If I know whether someone’s conception of truth is correspondist or epistemic, they can say “X is true” and I know what they mean about X without having to taboo “truth”.
This is something I understand and I can agree with. But it’s a very practical problem, like making Europeans and Americans agree on the meaning of the word “football”. It’s very likely that I’m still missing something, though (see metaphysicist reply).
This particular story introduces what seems like it should be an extraneous detail: Bob, in saying he doesn’t carry cash, is intending to deceive. But our everyday concept of truth is related somehow to honesty; the truth is what a well-informed and honest person would say. Bob here is being dishonest (he wants the mugger to believe something Bob thinks is false) and misinformed, so his dishonesty fails at his goal of protecting his cash. This goal is already lost. Knowing that, the question “but is his statement true?” seems to be unneeded essentialism.
Knowing that, the question “but is his statement true?” seems to be unneeded essentialism.
My impression is that this is what the whole debate is about. What matters to Corresponders is whether or not statements describe reality. Bob’s statement correctly predicts whether or not he has cash on him- and “correct prediction” is their standard for truth. They wouldn’t care that his dishonesty and misinformation cancel each other out, but would agree that in general dishonesty and misinformation lead to less correct predictions.
I’m honestly confused about this definition. Does it mean that once we define what truth is, we can then assert if a proposition is true or not ( truth and “justification” are in this sense intertwined)? In this case, isn’t this just a broader case of Correspondence, or better, isn’t Correspondence just a particular subcase of Epistemic?
Perhaps an example of an epistemic theory of truth will help. Suppose you have a scientific community grappling with a certain set of problems. As their inquiry proceeds, they solve some problems and uncover new ones. In the process, they also greatly refine their methods. Imagine they eventually come to a stage where all their problems have been solved according to their own standards of warrant, and no open problems remain. One theory of truth (due to Charles Peirce) says that their beliefs at this stage of inquiry are true by definition. Truth is just what a community will arrive at at the ideal end of inquiry.
How is this different from a correspondence theory? Well, the correspondence theorist would say that even though the scientists have fully resolved their inquiry according to their own standards, and it seems like their beliefs are highly justified, it still might be the case that they got it wrong. It still might be the case that their beliefs don’t correspond to reality, in which case their beliefs are false. For Peirce, this claim makes no sense. According to him, what it means for a belief to conform to reality (or to be true) is for it to be a belief held by a community at the ideal end of inquiry. If we allow that our beliefs can be as justified as we could possibly make them and still fail to be true, then truth becomes a potentially unattainable goal, and for Peirce this would make truth a philosophically useless concept.
The epistemic theory of truth is generally held by philosophers who are skeptical of metaphysics, and who think metaphysical concepts are only valuable to the extent that they make a difference to our lives. If there is a metaphysical distinction that does not correspond to a distinction in what we observe or how we should behave, it should be discarded. This is why they think the theory of truth should be closely tied to our epistemic practices, the stuff we do when we are trying to find the truth.
OK, so… suppose this community then encounters some new experiences at time T that cause them to reject the solution (S1) to a problem previously considered solved, which they then re-close (with a different solution S2).
A correspondence theorist wasn’t sure before T whether S1 was true or false, they’re still not sure after T whether S1 is true or false, and the same goes for S2 (and, for that matter, anything else someone might assert). Presumably they will also make statements like S1 was justified before T and unjustified after T, that S2 is justified after T, and might disagree about whether S2 was justified before T.
An epistemic theorist asserts that S1 was true before T and is false after T, that S2 is false after T, and might disagree about whether S2 was true before T.
A deflationary theorist… um… is not committed to any particular position about the truth, falsehood, or indeterminacy of S1 or S2 at any time.
Yes?
For Peirce, an end to all problems (and therefore an end to inquiry) includes the claim that there are no reecalcitrant experiences. It isn’t just a temporary end to problems, it’s a permanent end. This is why he refers to this as the ideal limit of inquiry. So the situation in your first paragraph wouldn’t apply. Peirce specifies things this way precisely to avoid the consequence that truth-values can change with time. I should say, though, that this is only one specific epistemic theory, and an early (and therefore kind of unsophisticated) one. I chose it for ease of exposition.
Well, OK, but then it seems Peirce’s conception of truth is just as potentially unattainable (and thus philosophically useless, by his own account) as the correspondence conception. Unless I’ve reached an ideal limit of inquiry—which of course I can’t really know I’ve done, and am unlikely to have done—then it seems I don’t actually have truth, even on Pierce’s account.
That aside, though, point taken about epistemicism != Peirce; presumably if I actually care about the latter I should just read Peirce. I’m just being intrigued by it. I do think I roughly understand the concept now; thanks for the explanation.
Many thanks to you and Vaniver, I have a clearer picture now (though this mostly looks like a debate over definitions, that can basically be solved by tabooing the word “truth”).
When you’ve reached this point, you understand a philosophical debate, for most 20th-century philosophical debates.
Not sure what you mean by this. The three theories of truth are different attempts to taboo the word “truth”.
Isn’t this the point? Tabooing “truth”, one can see that the theories really speak about (slightly) different concepts. Going back to your previous example, if one theory claims the scientists have reached the truth and the second doesn’t, how does it change the reality? You can easily define some new words to correspond to the different concepts, and refer to the appropriate label under the different circumstances.
Then, you would merely choose which of the concepts is the one needed for a particular theoretical purpose. Right?
Wrong! The arguments go to the concepts’ coherence. This is why it’s philosophy, not lexicography.
For example, a correspondence theorist generally argues that the notion of an epistemological limit to which scientific findings converge need not exist and can never be established empirically. If correspondence theory is true, you aren’t allowed to use the Piercian limit. It’s a vacuous concept.
Or, the correspondence theorist argues that the epistemological limit of scientific investigation can’t even be defined without assuming a correspondence variety of truth (which the Piercian, in turn, argues can’t exist). The correspondentist argues that if you define truth at a limit, then you have to define the truth that science is converging as itself the result of a scientific investigation at an endpoint, and similarly for the concepts you use to define scientific investigation, etc. Thus, a Piercian view, it’s contended, produces an infinite regress.
It’s possible that both concepts are coherent, but that too would require a philosophical argument—and it’s an unlikely result here, at least in my opinion: it’s probably more likely that both concepts are incoherent than that both are coherent.
These kinds of conclusions, philosophical and lacking in direct application, help inform the priors one assigns to just about every scientific controversy.
Ok, this starts to sound more interesting, thank you for the reply. I tried to briefly google for “Piercian limit”, though and it didn’t turn out anything relevant. Any quick reference?
Theories using Piercian concepts are today usually termed antirealist or instrumentalist.
Thank you, this is turning out a lot of material that I will definitely read.
(blink) If I accept acorrespondence theory of truth, it seems that correspondence theory is not the sort of thing that is allowed to have a truth value. And if I reject a correspondence theory of truth, then I ought not believe that correspondence theory is true. So it seems that “correspondence theory is true” is necessarily false. No?
That’s an excellent argument if it’s the case that correspondence theory is not the sort of thing allowed to have truth values under correspondence theory. Why do you say it’s not?
Well, using pragmatist’s cited definition of correspondence theory, a proposition is true if and only if it bears some sort of congruence relation to a state of affairs that obtains.
What state of affairs is “correspondence theory is true” congruent with?
I can’t think of any.
If you can, I’ll happily be convinced my argument doesn’t hold, but basically it seems to me that correspondence theory lays out a framework for thinking about truth, just as governmental constitutions lay out a framework for thinking about law. Correspondence theory itself is no more true (or false) than constitutions are legal (or illegal).
The concept of scientific truth—the concept used by scientists—is the state of affairs some correspondence theories purport to be congruent with.
I might still want to know which concept I’m using when I say something is true.
I might also want to know which concept you’re using when you say something is true.
Sure, if I don’t know that (or am not confident I know that) then we can taboo “truth,” but that gets unwieldy; if we can agree on a shared referent then communication is more efficient. Tabooing key words is sort of like running code under a debugger… a great way of identifying points of failure, but not the way I want to live my life.
If I know whether someone’s conception of truth is correspondist or epistemic, they can say “X is true” and I know what they mean about X without having to taboo “truth”.
This is something I understand and I can agree with. But it’s a very practical problem, like making Europeans and Americans agree on the meaning of the word “football”. It’s very likely that I’m still missing something, though (see metaphysicist reply).
Typical examples of the difference between correspondence and epistemic have to do with people being right on accident. For example:
Is Bob’s statement true? Is it a justified true belief?
Correspondence would say that Bob did tell the truth, but epistemic would say that Bob’s statement, though true, was unjustified.
The philosophers call these Gettier problems.
This particular story introduces what seems like it should be an extraneous detail: Bob, in saying he doesn’t carry cash, is intending to deceive. But our everyday concept of truth is related somehow to honesty; the truth is what a well-informed and honest person would say. Bob here is being dishonest (he wants the mugger to believe something Bob thinks is false) and misinformed, so his dishonesty fails at his goal of protecting his cash. This goal is already lost. Knowing that, the question “but is his statement true?” seems to be unneeded essentialism.
My impression is that this is what the whole debate is about. What matters to Corresponders is whether or not statements describe reality. Bob’s statement correctly predicts whether or not he has cash on him- and “correct prediction” is their standard for truth. They wouldn’t care that his dishonesty and misinformation cancel each other out, but would agree that in general dishonesty and misinformation lead to less correct predictions.