Virtue Epistemology

Virtue epistemology utilizes virtues in addressing the prominent problems in epistemology. A distinction is made between moral virtues and epistemic or cognitive virtues. Within the virtue epistemology camp, there is a divide between reliabilist and responsibilist understandings of epistemic virtues. I will focus here on a reliabilist account. Roughly put, an epistemic virtue is stable disposition to achieve certain results (true beliefs) in certain circumstances. More precisely, a mechanism M for generating or maintaining beliefs is an epistemic virtue if and only if M is an ability to believe true propositions and avoid believing false ones within a field of propositions F when one is in a set of circumstances C.

The virtue epistemologist’s claim, then, is that a proposition p is epistemically justified for S if and only if S’s believing p is the result of an epistemic virtue of S. Understood as such, virtue epistemology is a type of process reliabilism. By specifying which type of processes can produce an epistemically justified belief, virtue epistemologists attempt to provide an account of epistemic justification (and often knowledge) that avoids the problems of ‘simple’ reliabilism.

The New Evil Demon Problem was a problem for reliabilism since in a world where one is massively deceived it seems as though one can nonetheless have justified beliefs despite the unreliability of the processes that produce them. This appears to be a problem for the virtue epistemologist as well, since one can believe propositions on the basis of what seems to us to be epistemic virtues (and seem to be epistemically justified in those beliefs), but believing in such a way does not lead to true beliefs in the evil demon world.

Ernest Sosa’s response as a virtue epistemologist is to relativize epistemic justification to an environment. In other words, the individual in the demon world is epistemically justified in her belief since she utilized cognitive faculties that are epistemic virtues in our environment. Since coming to beliefs in such a way would be reliable in our environment, and would be the result of an epistemic virtue, we consider the demon worlder to be justified. Epistemic justification is thus relativized to the actual world.

The above response is unsatisfactory, however. Sosa’s response does not account for all of our intuitions here. To see this we can imagine that we are being deceived by an evil demon as well. In such a scenario, coming to beliefs by way of seeming epistemic virtues is not a reliable way to come to beliefs. According to Sosa, our beliefs are not justified in such a scenario, but we still think that they are. Our intuitions are that such beliefs are epistemically justified regardless of whether one is in a demon world, even if the actual world is a demon world. This problem remains for virtue epistemologists.

A second problem for reliabilism concerns reliable belief forming processes at work in an individual that has reason to doubt that his processes are reliable. In such a scenario reliabilism has it that he is justified in the beliefs produced by the reliable process, but is seems as though the evidence that he has regarding the unreliability of these processes renders the resultant beliefs unjustified. This problem too seems to remain for the virtue epistemology response. A belief could be the product of an epistemic virtue, yet one have evidence against it being the product of such a virtue. Can virtue epistemology get the right result that the resultant belief is unjustified?

Sosa attempts to get this result by making a distinction between animal knowledge and reflective knowledge. This is itself a cost, since it seems that we have only one concept of knowledge. Positing two such concepts seems to be a last resort. According to Sosa, to have animal knowledge, one must believe out of epistemic virtue which makes the resultant belief apt, but to have reflective knowledge one must believe out of epistemic virtue and be aware of so doing which makes the belief justified (ie. one must also believe out of epistemic virtue that her [first-order] belief was produced in an epistemic virtue). To be justified, one must recognize regarding her belief that p that it was produced by an epistemic virtue (ie. she must recognize (i) that p falls into the relevant range of propositions, and (ii) that she is in one of the relevant circumstances for her belief producing mechanism to be reliable).

Applied to the case where one has misleading evidence regarding the reliability of the belief producing process, on this account one has animal knowledge (the belief is apt), but lacks reflective knowledge (the belief is not justified). Although this account gets the right result regarding the case of misleading evidence, it has problematic consequences. The problem is that very few people have any beliefs about their beliefs such as that there belief was formed from epistemic virtue. Whereas people may recognize that there belief was formed on the basis of perception, they do not believe that the relevant proposition falls within a certain range of propositions or that they are in a circumstance among a set of acceptable circumstances such that perception is reliable for such propositions in such circumstances. Such propositions are not typically believed, even dispositionally. As a result, Sosa’s account implies that all such individuals (most individuals) are not epistemically justified in their beliefs. There beliefs may be apt, but they are not justified. However, it seems that most individuals are epistemically justified in at least a good number of beliefs, or minimally, that their epistemic standing to such propositions is better than the aptness required for animal knowledge. Such meta-beliefs simply do not appear to be required for epistemic justification.