That truth has something to do with correspondence to extralinguistic and extramental fact is a deeply entrenched intuition. One could call it the classical intuition about truth inasmuch as one can find formulations of it in Plato and Aristotle. When suppressed, it has a way of reasserting itself. Sent packing through the front door, it returns through the back. Herewith, two brief demonstrations that this is so.
A. Truth as Idealized Rational Acceptability
One way of suppressing the classical intuition is by offering an epistemic definition of 'true.' One attempts to explicate truth in terms of mental states. Thus someone might suggest that a proposition is true just in case it is believed or accepted by someone. But this won't do, since there are truths that are not accepted by anyone. So one proposes that a proposition is true just when it is acceptable. This proposal, too, is defective inasmuch as what is acceptable to one person will not be acceptable to another. This defect can perhaps be handled by identifying truth with rational acceptability. But what it is rational to accept at one time or in one place may be different from what it is rational to accept at another time or in another place. Much of what we find rationally acceptable would not have been found rationally acceptable by the ancient Greeks. (For example, that the same physics holds both for terrestrial and for celestial bodies.) So one advances to the notion that truth is rational acceptability at the ideal limit of inquiry. One can trace this notion back to C. S. Peirce. In Reason, Truth, and History, Hilary Putnam presents a version of it. Let's consider the theory in the following form:
1. *P* is true =df *p* would be accepted in cognitively ideal conditions.
Now we know that
2. Cognitive conditions are not ideal.
From (2) it follows via the trivial equivalence principle *p* is true iff p that
3. *Cognitive conditions are not ideal* is true.
It follows from (3) via (1) that
4. *Cognitive conditions are not ideal* would be accepted in cognitively ideal conditions.
But (4) is self-contradictory, whence it follows that
5. The definition of truth in terms of acceptability in cognitively ideal conditions is incorrect.
What I take this argument to show is that the notion of truth as correspondence to the way things are is primary and irreducible. For surely (2) is true. But its being true cannot be explicated in terms of what anyone would accept or assert under ideal epistemic conditions. Therefore, (2) is true in a sense more basic than the sense spelled out in (1).
This supports the 'truthmaker intuition': some if not all truths require truthmakers. Truths do not 'hang in the air.' What is actually true cannot depend on what some merely possible subject would accept at the ideal limit of inquiry.
B. Truth as Coherence
We get a similar result if we try to construe truth as coherence. Suppose
6. P is true =df p would be accepted by a person whose set of beliefs is maximally consistent and coherent.
But we know that
7. No one's set of beliefs is maximally consistent and coherent.
From (7) it follows via the above equivalence principle that
8. No one's set of beliefs is maximally consistent and coherent is
true.
It follows from (8) via (6) that
9. No one's set of beliefs is maximally consistent and coherent would be accepted by a person whose set of beliefs is maximally consistent and coherent.
But (9) is self-contradictory, so
10. (6) is incorrect.
Comments