Monday 21 November 2011

Epistemology and the Ought-Can Principle

In certain contexts, it can be fairly natural to give voice to an assessment of somebody's belief by using the word 'ought'. For instance, suppose that your friend is engaged in wishful thinking: this friend believes that some guy will call her when (unfortunately) that is pretty clearly fantasy on her part. You might say to her (if you were trying to wake her up to the cruel reality), "That's not what you ought to believe. We both know that this guy rarely calls anybody the day after." She might respond, "That is what I ought to believe. My case is different because ... ."

Now, as natural as these assessments might be, there is some room for discomfort as to whether we can ultimately make sense of them. One concern revolves around the alleged entailment from 'ought' to 'can'. The idea is that in order for it to be the case that one ought to bring about a particular outcome (e.g. a belief or action), one must have the (appropriate kind of) ability to bring about that outcome. In other words, if one cannot control the outcome, then it is not possible that one ought to bring it about. Where the outcome in question is belief, it boils down to this: if it is not possible for the subject to exert control in order to have better beliefs (i.e. beliefs that fit the evidence better), then it cannot be that the subject ought to have better beliefs.

As this suggestion applies to our fictitious example, it appears that what you've said to your friend may not be strictly correct. Supposing that your friend is just incorrigibly prone to wishful thinking (as many of us are), it may be that she does not have the cognitive resources to take a position that respects her evidence. Assuming so, the suggestion is that it cannot be that she ought to believe other than how she does on this matter. Perhaps we can assess her belief as bad, but not as a belief she ought not have.

I find this result fairly unsatisfactory although the issues here are complex. One consideration that concerns me is that this way of thinking appears to let people off the hook for bad decisions merely because they are incorrigibly bad at managing their beliefs. Think about somebody who you consider to (1) have terrible beliefs about public policy or religion (whether for or against), and (2) be so entrenched in these terrible beliefs and so bad at thinking that there is no way that they could give them up. Presumably, people of this sort are a bad influence on society; they make bad decisions about, for instance, whom to vote for. It strikes me as fairly reasonable that we should hold them responsible for this bad influence, even if they sincerely take themselves to be doing what is best. However, if it really is the case that the prospects for them to improve their beliefs are hopeless, then the suggestion last paragraph appears to have the result that it cannot be that they ought to be believing differently. On what grounds, then, can we hold them responsible?

Maybe we should we give up the idea that the 'ought' in question entails 'can' of the relevant sort. What do you think?

1 comment:

  1. Hi Ben, an interesting post. One way to reconcile the tension between endorsing the view that some doxastic ought claims are true (claims of the form 'X ought to believe that p" alongside the view that doxastic voluntarism is false (and ought implies can is true) is to construe doxastic oughts as of the 'ought-to-be' claims. Matthew Chrisman defends such a view in the Journal of Philosophy 2008: http://www.philosophy.ed.ac.uk/people/documents/Chrisman-oughttobelieve-jan28.pdf

    ReplyDelete