More on Knowledge and Pragmatics

In their "Evidence, Pragmatics, and Justification", Fantl and McGrath argue as follows:

1. [CLOSURE] If S knows that p, and (S knows that if p, then A is, in light of all of her goals, the best thing to do), then S knows that A is, in light of all of her goals, the best thing to do.
2. [KNOWLEDGE & RATIONAL ACTION, or KRA] If S knows that A is, in light of all of her goals, the best thing to do, then S is rational to do A.
Thus, 3. S knows that p only if, for any act A, if S knows that (if p, then A is, in light of all of her goals, the best thing she can do), then S is rational to do A.

Through a series of strengthenings, Fantl and McGrath arrive at:

4. S knows that p only if, for any states of affairs A and B, if S is rational to prefer A to B, given p, then S is rational to prefer A to B, in fact.

The conclusion of this argument is mirrored in Hawthorne's recent assertion that:

4'. "[I]f the question whether p is practically relevant, it is acceptable to use the premise that p in one's deliberations if one knows it" (Hawthorne, 30)

Given this, however, we can argue as follows:

5. If the question whether p is practically relevant, it is acceptable to use the premise that p in one's deliberations only if one's belief is reliably supported to the level required by the context of the action to which the question whether p is relevant.
Thus, 6. If the question whether p is practically relevant, one knows that p only if one's belief is reliably supported to the level required by the context of the action to which the question whether p is relevant.

Of course, an argument with such a radical conclusion invites criticism. Thus, at the 2004 Pacific APA, Leo Iacono, in his "Pragmatics and Epistemic Justification", argues against [KRA] as follows (note the similarity between this argument and one of John Turri's recent arguments -- Argument A -- against Hawthorne's 4'):

A. Some cases of knowledge that A has the best consequences are cases in which you cannot be fully confident that A has the best consequences.
B. In all cases, if you are not fully confident that A has the best consequences, then, if the consequences of being wrong are sufficiently bad, you shouldn't do A.
C. Therefore, in those cases where you know that A has the best consequences but you are not fully confident of this, and in which the consequences of being wrong are sufficiently bad, you shouldn't do A.

If we understand "S is rational to do A" in [KRA] as "it is rationally permissible for S to do A", then, since (C) entails that there are cases in which one knows that A has the best consequences but one shouldn't do A, Iacono argues that A - C disprove [KRA].

I've already argued that B, formulated in terms of confidence, is false. Thus, argument A-C fails. However, I conceded that:

B'. In all cases, if your belief that A has the best consequences is not based on a reliable enough process, given the consequences of being wrong, then you shouldn't do A,

is acceptable. But now, the modified premise:

A'. Some cases of knowledge that A has the best consequences are cases in which your belief that A has the best consequences is not based on a reliable enough process, given the consequences of being wrong,

does not seem to be acceptable. Certainly, it is not intuitively obvious in the way in which premise A seems to be. Thus, the modified argument A'-C' in an attempt to disprove [KRA] fails. The way remains open to accept 1-6. I'll say more about some positive reasons for accepting a link between knowledge and pragmatics in a later post.