Internalism and externalism in epistemology

4. The distinction applied to knowledge

As for knowledge, internalism would seem to be less plausible than in application to justification, while externalism is more plausible. This is because of doubts that justification of belief is a necessary condition of knowledge. The dominant view of knowledge in the twentieth century has been that it consists of justified true belief. Since Gettier posed his famous difficulty for this view, those who take true justified belief to be at least sufficient for knowledge have tried to evade these difficulties by adding a fourth condition, but with indifferent success so far (see Gettier problems §§1–2). If justification were necessary for knowledge, then any constraint on the former would equally be a constraint on the latter, and so accessibility internalism could not be less plausible for knowledge than for justification. But if justification of belief is not necessary for knowledge, we have a different story. More than one prominent externalist has developed a conception of knowledge in which justification is not mentioned at all (Dretske 1981; Nozick 1981). Something like reliability, or probability of the belief on its grounds, is taken as sufficient to turn true belief into knowledge. For example, on Nozick’s view the following four conditions are individually necessary and jointly necessary for S knowing, via method M, that p.

Nozick puts this by saying that a true belief counts as knowledge when it ‘tracks’ the truth, that is, when its being held or not varies with the truth value of the proposition believed. This form of externalism is cut loose from dependence on the right way of thinking of justification, though, of course, its success does depend on the right way of thinking of knowledge. (For some criticisms of Nozick’s proposals see Luper-Foy 1987.)