Virtue epistemology

3. Advantages and disadvantages of virtue epistemology

When Sosa introduced the concept of intellectual virtue into the contemporary literature, he thought that the shift of focus from properties of beliefs to properties of persons should make it possible to bypass the dispute between foundationalists and coherentists over the logical and evidential relations between beliefs needed for proper epistemic structure. In addition, two possible advantages of reliabilism have already been mentioned. First, its requirements for knowledge are easier to satisfy than internalist theories and, hence, it is said to be closer to common sense than are internalist theories. Second, its requirements for knowledge do not include the possession by the knower of an answer to the sceptical challenge. For example, they need not be able to answer the question ‘But what if there is an evil demon?’, even though the presence of an evil demon would deprive them of knowledge. Since having an answer to this question is usually thought to be at best difficult, and probably impossible for the average knower, it is an advantage of a theory of knowledge if its formulation permits us to ignore such worries (see Scepticism).

These advantages of reliabilism may be offset by certain disadvantages, however. Consider first the requirement that reliability is a necessary condition for knowledge. Most versions of virtue epistemology have this requirement. A common objection to the reliability requirement for knowledge or justification is the so-called ‘generality problem’. This is the problem that a particular belief can be described as an instance of many different belief-forming processes, some broader (believing whatever you read), some narrower (believing what you read in a particular journal), some so narrow as to apply only to a single belief (believing what you read in a certain journal on a certain subject by a certain writer). The degree of truth-conduciveness will vary greatly, depending upon the way in which the belief-forming process is described. This problem requires refinements to the theory rather than outright rejection. Ways of handling the generality problem have been suggested by Goldman (1986), Greco (1993), Sosa (1991) and Zagzebski (1996). Zagzebski suggests that the class of cases against which truth-conduciveness is to be measured should be empirically determined by the way in which people naturally generalize their belief-forming processes. So if a person who believes what they read in a reliable journal then goes on to believe whatever they read in any periodical, their belief would not count as deriving from a reliable belief-forming process, whereas it would be reliable if they do not tend to generalize their belief-forming tendency to unreliable sources.

Simple reliabilism maintains that reliability is not only necessary, but is sufficient for knowledge. Zagzebski denies that simple reliabilism ought to be considered a form of virtue epistemology, but it is worth mentioning the most serious objection to it since reflection on it was one of the motivations leading to the development of virtue epistemology. The problem is that simple reliabilism does not rule out of the category of knowledge beliefs formed by processes or faculties that are reliable by accident (Plantinga 1993a; Zagzebski 1996). This suggests that any theory that includes reliability as a requirement for knowledge should include an account of what makes a process, trait or faculty reliable in the right way. Virtue epistemologists, then, characteristically add requirements for knowledge other than reliability to eliminate this problem by showing how reliability arises out of something else – in the case of Sosa, the believer’s inner nature; in the case of Plantinga, faculties whose proper function is determined by design; for Greco, the norms the believer countenances; for Zagzebski, intellectual virtues. These conditions involve activity on the part of the knower that exhibits epistemic responsibility or features praiseworthy from an internal perspective.

Requirements for knowledge in addition to reliability are seen by their adherents as improving upon simple reliabilism, but these requirements can also be perceived as detracting from one of the advantages of simple reliabilism. This is its generosity in spreading the domain of knowledge to include true beliefs formed in typical circumstances by young children and unreflective adults. But the more we add responsibility and internal awareness to the conditions for knowledge, the more we shrink the resulting class of knowing states. Can children and unreflective adults have perceptual and memory knowledge on these more demanding theories? Sosa’s way of handling the problem is to distinguish animal knowledge from reflective knowledge. Zagzebski prefers not to make it a condition for knowledge that the believer actually possesses the full virtue but, rather, that the believer has virtuous intellectual motivations and acts cognitively the way virtuous persons do in the same circumstances. Given that virtuous persons formulate perceptual and memory beliefs in many circumstances with little or no reflection or investigation, it is not difficult for children to imitate their behaviour and, hence, to have knowledge in these cases. Greco’s way of handling the problem (1990) is to say that while a belief has positive epistemic status only if it conforms with the norms that the believer countenances, this does not require that the believer should have cognitive access either to the norms themselves or to the fact that their beliefs conform with them.

It can be argued that virtue epistemology has the potential to deal effectively with two other problems of contemporary epistemology. First, there are the problems surrounding the concept of justification which have led to the impasse between internalism and externalism. Since justification is a property of a belief, it is very difficult to adjudicate disputes over this concept if the belief is treated as the bottom-level object of evaluation. If we focus instead on the deeper concept of an intellectual virtue and treat the justifiability of a belief as derivative, it may turn out that justifiability is only one normative property of beliefs among others and the competing intuitions of internalists and externalists require the analysis of more than one property of beliefs, all of which are based in some way on the concept of virtue. Second, focusing epistemology on belief has led to the neglect of two epistemic values that have been important for long periods in the past – understanding and wisdom. An approach focused on intellectual virtues, on the other hand, naturally leads to an account of these values, since understanding and wisdom are either virtues themselves or are very closely connected with virtues (see Wisdom).