Several recent theories have emphasized pragmatic features involved in knowledge or in knowledge attributions. The most recent is John Hawthorne’s new book, where the view he seems to like best has the standards for knowledge rising and falling depending on the importance of the issue at hand.
Contextualism has a similar feature, driven by examples such as Cohen’s airport example where the truth of a knowledge attribution about the time of a flight arrival depends on how much it matters that the flight is on time.
Compare these pragmatic encroachments with more traditional outlooks. I like Dick Foley’s account the best, so I’ll use it here, especially his notion of the purely epistemic. Foley has long argued against views that introduce impurities into the notion of epistemic rationality, such as when a belief is judged rational when it will contribute to overall well-being, or when it will contribute to getting to the truth in the long run. The epistemic goal, according to Foley, is to get to the truth and avoid error now. One implication of this would be that if there are various desiderata that drive theory acceptance in science, and these factors are not truth-related, then theory acceptance in science is not a purely epistemic matter. For Foley, what is epistemic is a function of a pure and undefiled interest in the truth.
If you adopt one of the theories that introduces pragmatic encroachment, there is some tension here, since the notion of what is purely epistemic ought to have something to do with knowledge; as many are wont to say, epistemology just is the theory of knowledge ( I don’t agree, but that’s beside the point here). If those allowing pragmatic encroachment are correct, then, if one agrees with Foley on what is purely epistemic, we ought to say that according to such theorists, knowledge is not purely epistemic, in spite of the paradoxical tone of such a remark.
Now, I don’t want to say that knowledge is not purely epistemic, and I don’t think most epistemologists will either. To avoid saying this, one must hold that the Foley account of what is purely epistemic is mistaken.
Maybe such theorists should ask those of the Foley persuasion to talk instead of our purely theoretical interests in the truth, and ask them not to use the label ‘epistemic’ in doing so. What we have been calling the epistemic goal would then be more appropriately termed the theoretical goal. The lesson is, if pragmatic encroachment is unavoidable, that the love of truth can’t play the fundamental role in a theory of knowledge that a view like Foley’s and his ilk suggests.
Is there an alternative to allowing such pragmatic encroachment? Maybe, and that is what I want to explore briefly here. Note first that a crucial feature of knowledge is the way in which an ascription of knowledge involves representing to oneself and one’s audience that inquiry on the matter is legitimately closed. As such, we can expect people to refuse to ascribe knowledge in practical situations where being wrong is very costly, since their awareness of the risks of being wrong makes them wary about whether inquiry is legitimately closed.
But we need not let these pragmatic considerations into our theory directly. We can avoid some of the pragmatic encroachment by endorsing two ideas. The first is about the nature of epistemic justification. The kind of justification needed for knowledge is distinguished from ordinary justification (the kind of justification one has, say, for thinking that one’s lottery ticket will lose) by requiring a body of evidence sufficient not only to justify the belief in question but also sufficient to justify thinking that further inquiry would undermine the belief in question only by discovering misleading information. This point explains why it is strange, awkward, and perhaps paradoxical to say things like, “I know it’s raining, but I ought to go outside to check on it.” The second move concerns internal defeaters, ones found within one’s body of total evidence. We might say that adding a belief or concern that further inquiry would not be a waste of time and effort by itself undermines the aspect of epistemic justification that distinguishes it from ordinary justification. In short, concerns about the possibility of error rob one’s total body of evidence of the power to epistemically justify the belief any further (even if they may leave one justified in the ordinary sense).
Those familiar with the literature on this issue will recall that salience of the possibility of error is a driving consideration in introducing pragmatic features into one’s theory of knowledge (or knowledge ascriptions, for contextualists). The view above does the same: it creates a place for salience of the possibility of error, without allowing any direct encroachment of pragmatic features into the theory of knowledge.
So I think the traditional approach can absorb the motivations for pragmatic encroachment arising from issues connected with the concept of salience. The question remains whether there are other motivations for pragmatic encroachment that need to be considered, but I’ll leave that for another time.