Abstract

In this work, I defend an empiricist conception of logic. Hartry Field has recently argued that logic must be a priori because we could not gather evidence against certain basic logical rules. Though I agree that logic is probably a priori in Field's sense, I distinguish strongly empirical reasons of the sort that Field considers from modestly empirical reasons. These latter kinds of reasons are those we make when our concern is to revise conservatively over the evidence. To demonstrate my claim, I appeal to Hilary Putnam's famous suggestion that we revise classical logic in favor of quantum logic to help make sense of certain quantum mechanical anomalies. I show that when the evidence doesn't clearly determine how we should revise either way, as in Putnam's example, we are entitled to revise our logic by appeal to pragmatic rules. Since in Putnam's example, the choice is between revising logic and adopting a particular physical interpretation of quantum mechanics, I argue that revising logic is just as empirical as is the more physical option: both seek to explain and make sense of the evidence. ;Next, I must show that pragmatic rules are not, themselves, unrevisable for empirical reasons. To do this, I turn to a recent argument by Jerry Katz that states that pragmatic rules could not be revisable without being self-defeating. I argue that Katz's claim is true assuming that justification is factual, but that this is an assumption that empiricists have independent reasons to reject. I defend a non-factualist conception of justification, and show how non-factualism offers a way out of Katz's problem for the radical empiricist. ;Finally, I show how my version of empiricism can go some way toward solving the long-standing problem of justifying deduction. I argue that since I have shown that logic is revisable for at least modestly empirical reasons, logic is always at risk of being revised in favor of different rules that better fit the evidence. I conclude that empiricism can better account for the justification of logic than can the more rationalist account of justification put forth by Paul Boghossian