No Exception for Belief?

Consider the following activities:

(A) Wearing a rain coat

(B) Listening to an Angels and Airwaves album

(C) Playing with one's dog

(D) Working out

How do we decide whether it's rational to engage in these activities? Here are some plausible answers: It is rational to engage in one of the above activities just in case doing so has a high expected value. Or, It is rational to engage in one of the above activities just in case doing so would be an effective means to one's ends. Take this last theory. If my goal is to stay dry while walking outside in the rain, then (A) is an effective way of achieving that goal. Intuitively, then, it is rational for me to engage in (A). 

Susana Rinard argues that, however we understand the rationality of engaging in activities like (A)-(D), the rationality of belief states is no exception. Call this NO EXCEPTION. On this view, it is rational for any agent to believe P just in case doing so is an effective way of achieving her goals, or (alternatively) just in case doing so has a high expected value, or (alternatively) just in case...(fill in your favorite instrumentalist story). Again, determining the rationality of beliefs is no different than determining the rationality of non-believing activities like (A)-(D).

This conflicts with evidentialism, according to which:

• Evidentialism (roughly): it is rational for S to believe that P just in case P is supported by her total evidence.

On evidentialism, even if believing P is rational is some other sense, it won't be rational in the epistemic sense of rationality if it's not supported by one's total evidence. For example, it may be pragramtically (or instrumentally) rational to believe that I'm going to ace my logic exam, since doing so is an effective way of overcoming test anxiety (and I certainly don't want test anxiety), but it isn't epistemically rational to believe I'll ace it, since my evidence suggests otherwise (I won't comment on whether this is a real-life example). On this view, there are two different senses of rationality. There is the epistemic sense of "rationality" (concerned entirely with beliefs) and the instrumental/pragmatic sense of "rationality." Sometimes an agent is rational in one sense and not in another (as in the above example). Sometimes an agent is rational is both senses (e.g., belief in God may be pragmatically advantageous for me and, at the same time, supported by my evidence). Whatever the case, these two sense of "rationality" are different. NO EXCEPTION denies this. There is only one sense of "rationality."

Rinard discusses some of the advantages of NO EXCEPTION over exceptional views like evidentialism (double meaning intended):

• NO EXCEPTION is more simple than exceptional views. NO EXCEPTION says there is only one sense of rationality, not two. 
• NO EXCEPTION can still accommodate our intuition that evidence matters for rational belief. This is because, in most ordinary cases, we can better pursue our goals if we have true beliefs and not false ones. And since evidence is that which points to truth, we are often rational in believing what the evidence supports. Evidence is useful in helping us achieve a large (very large) number of our goals. Nonetheless, it makes room for the intuition that non-evidential factors can also make a difference to rationality (see below).
• Suppose an athlete's evidence isn't strong enough to justify her belief that she'll win a race. Nonetheless, she knows she'll perform better if she believes she's going to win. It seems like she can be rational in believing she'll win. Or think of cases involving medical patients who don't have a high chance of surviving their ailment. Even though their evidence suggests they won't survive, they're still rational in being optimistic (intuitively). 
• It can be rational to promise doing something even if the evidence does not clearly support the proposition that one will keep one's promise. The divorce rate is high. Nonetheless, it is rational for a couple to promise that they'll stay together, even if they're aware of the relevant statistics. NO EXCEPTION can explain why. 
• Scientists often use simplicity to adjudicate between empirically equivalent and competing theories. This seems rational. However, it is notoriously difficult to justify/explain the truth-conduciveness of simplicity. Nonetheless, NO EXCEPTION can explain why it is rational to use simplicity (i.e., simplicity is a good way for scientists to achieve their goals, even if it isn't clearly evidentially significant...and that's okay, cuz rationality is about more than evidential support. Gosh, this one is puzzling. Here's a dilemma for Rinard: either simplicity is truth-conducive or it is not. If it is not, then it can't be a reason to believe NO EXCEPTION is more likely than exceptionalism. If the no-exceptionist claims that simplicity *is* truth-conducive, then she loses the above example as a bit of support for her position. After all, the point is to explain why scientists are rational in using simplicity, even though it isn't clear how it is truth-conducive. So if the no-exceptionist is comfortable appealing to simplicity as evidence even though it's hard to say how it's truth-conducive, then we can explain the scientist's use of it the same way. She regards it as an evidentially significant virtue, even though she realizes that it's hard to say why it is so). 
• Rationality is normally thought of as an ideal which can guide both behavior and belief formation. But it is poorly suited for this role, if exceptionalism is true, since rationality could pull us in different directions, recommending incompatible actions ("it's rational to believe you'll win the race"/"it's not rational to believe you'll win the race"). NO EXCEPTION unifies the recommendations of rationality. 

________________________________________________________

Bibliography

Rinard, "No Exception for Belief."