The Normative Skeptical Paradox

1. If you know p, it is rational for you to do A.

2. It is not rational for you to do A.

3. So you do not know p.

 

And similarly for various substitutes of "know" (justifiedly believing and so on), "rational" (justified, appropriate and so on) and "doing A" (including having certain attitudes). While seldom discussed, these arguments have been influential. They underlie the widespread view that we should not be certain of anything and they are at the heart of the recent discussion on "pragmatic encroachment" in epistemology (the idea that stakes affect whether you know or justifiedly believe).

 

Most philosophers who do discuss these paradoxes assume they must receive a solution of a certain type, which I call Possibilist. Possibilists think that in the interesting cases, either it is in fact rational to do A or it is in some relevant sense possible for you that not-p. For instance, it is not rational for you to bet your life against one cent on what you had for lunch, because there is some chance that you didn't have for lunch what you think you had. Few think that it is rational ("Imprudents"), so most think that there is this possibility. Some think that the possibility is incompatible with knowing, others ("Fallibilists") that it is, others that it may or may not be, depending on stakes ("Impurists"). Most think that it is not compatible with rational certainty.

 

In the paper I highlight some questions for Possibilists and I sketch another type of solution to the paradoxes, which I call Bad Habit views. The rough idea is this. Even if you are rationally certain of what you had for lunch (and that the bet is offered in all honesty and will be paid out and so on), it is not rational for you to bet, because you may thereby encourage a bad habit to take bets of that kind. The views preserve ordinary intuitions about the paradoxes without requiring any form of Skepticism, Fallibilism, Impurism or Imprudence.