Internalism restricts justifiers to what is "within" the subject. two main forms of internalism are (1) perspectival internalism (pi), which restricts justifiers to what the subject knows or justifiably believes, and (2) access internalism (ai), which restricts justifiers to what is directly accessible to the subject. the two forms are analyzed and interrelated, and the grounds for each are examined. it is concluded that although pi is both unacceptable and without adequate support, a modest form of ai might be defended.

In current epistemology, there are two different conceptions of epistemic justification. According to the first genetic conception, a justified belief is a well-formed belief. According to the second defense conception, how the belief is formed is irrelevant to the epistemic justification of the belief. What is important for the justification of the belief is whether the cognitive agent has a defense of the belief in question. ;I construct my own defense account of epistemic justification on the basis of criticizing current (...) defense accounts of epistemic justification. The gist of my account is that I am justified in believing that p only if my belief that p is stored in my memory with a proper causal connection with other psychological states of mine that are adequate grounds for the belief that p. This illuminates an interesting symmetry between the genetic conception of epistemic justification and the defense conception of epistemic justification: in the genetic conception, justified belief is well-formed belief, whereas in a defense conception justified belief is well-stored belief. ;My account is different from current defense theories of epistemic justification in two respects. First, in my account some type of causal relation between the belief and its adequate grounds is crucial for epistemic justification, while other theories deny the relevance of any causal consideration to the justification of beliefs. Their denial is a manifestation of a dogma--the Psychologistic Assumption--deeply rooted in traditional epistemology, such that epistemic justification is wholly a function of psychological states of a cognitive agent. I show that this dogma has to be abandoned. Second, other theories require the explicit representation of the belief about the evidential relation between the belief in question and its adequate grounds. In them, this explicitly represented higher-level belief constitutes the defense of the belief in question. I show that, due to this requirement of explicitly represented higher-level beliefs, current defense accounts run into numerous insuperable problems. Meanwhile, my theory does not face these problems because it does not require higher-level beliefs. (shrink)