reliabilism a type of theory in epistemology that holds that what qualifies a belief as knowledge or as epistemically justified is its reliable linkage to the truth. David Armstrong motivates reliabilism with an analogy between a thermometer that reliably indicates the temperature and a belief that reliably indicates the truth. A belief qualifies as knowledge, he says, if there is a lawlike connection in nature that guarantees that the belief is true. A cousin of the nomic sufficiency account is the counterfactual approach, proposed by Dretske, Goldman, and Nozick. A typical formulation of this approach says that a belief qualifies as knowledge if the belief is true and the cognizer has reasons for believing it that would not obtain unless it were true. For example, someone knows that the telephone is ringing if he believes this, it is true, and he has a specific auditory experience that would not occur unless the telephone were ringing. In a slightly different formulation, someone knows a proposition if he believes it, it is true, and if it were not true he would not believe it. In the example, if the telephone were not ringing, he would not believe that it is, because he would not have the same auditory experience. These accounts are guided by the idea that to know a proposition it is not sufficient that the belief be ‘accidentally’ true. Rather, the belief, or its mode of acquisition, must ‘track,’ ‘hook up with,’ or ‘indicate’ the truth.
Unlike knowledge, justified belief need not guarantee or be ‘hooked up’ with the truth, for a justified belief need not itself be true. Nonetheless, reliabilists insist that the concept of justified belief also has a connection with truth acquisition. According to Goldman’s reliable process account, a belief’s justificational status depends on the psychological processes that produce or sustain it. Justified beliefs are produced by appropriate psychological processes, unjustified beliefs by inappropriate processes. For example, beliefs produced or preserved by perception, memory, introspection, and ‘good’ reasoning are justified, whereas beliefs produced by hunch, wishful thinking, or ‘bad’ reasoning are unjustified. Why are the first group of processes appropriate and the second inappropriate? The difference appears to lie in their reliability. Among the beliefs produced by perception, introspection, or ‘good’ reasoning, a high proportion are true; but only a low proportion of beliefs produced by hunch, wishful thinking, or ‘bad’ reasoning are true. Thus, what qualifies a belief as justified is its being the outcome of a sequence of reliable belief-forming processes.
Reliabilism is a species of epistemological externalism, because it makes knowledge or justification depend on factors such as truth connections or truth ratios that are outside the cognizer’s mind and not necessarily accessible to him. Yet reliabilism typically emphasizes internal factors as well, e.g., the cognitive processes responsible for a belief. Process reliabilism is a form of naturalistic epistemology because it centers on cognitive operations and thereby paves the way for cognitive psychology to play a role in epistemology.
See also EPISTEMOLOGY , NATURALISTIC EPISTEMOLOGY , PERCEPTIO. A.I.G.