Alvin Goldman put forward "A Causal Theory of Knowing" in 1967. He said that the knowledge status of a belief depends on the cause (perhaps some event in the external world) that formed the belief.

Ten years later Goldman presented a reliabilist approach to justification. He said:

The theory focused on the truth-ratios of the process types used in belief formation, and generally goes by the name "process reliabilism". In its simplest form, it says that a belief's justificational status hinges on the psychological processes that produce it, e.g., perception, memory, introspection, or various inference patterns. Beliefs formed by highly reliable processes are justified; beliefs formed by insufficiently reliable processes are unjustified. The approach is motivated by examples. Intuitively, beliefs formed by unreliable processes like sheer guesswork or wishful thinking are unjustified even if their propositional contents stand in appropriate relations to evidence beliefs. Process reliabilism contrasts with traditional theories like foundationalism and coherentism in being a "historical" theory rather than a "current time-slice" theory. Foundationalism and coherentism imply that justificational status is fixed by mental states held at the time of believing. According to reliabilism, it's the mental history of a belief that fixes its justificational status.

(A Companion to Epistemology, 2nd ed., p.144)

As David Armstrong pointed out a few years later (1973), Frank Ramsey had in 1929 suggested both "causal" and "reliable process" theories.

I have always said that a belief was knowledge if it was (i) true, (ii) certain, (iii) obtained by a reliable process. But the word 'process' is very unsatisfactory; we can call inference a process, but even then unreliable seems to refer only to a fallacious method not to a false premiss as it is supposed to do. Can we say that a memory is obtained by a reliable process? I think perhaps we can if we mean the causal process connecting what happens with my remembering it. We might then say, a belief obtained by a reliable process must be caused by what are not beliefs in a way or with accompaniments that can be more or less relied on to give true beliefs, and if in this train of causation occur other intermediary beliefs these must all be true ones.

(Foundations of Mathematics and Other Logical Essays , 1931), Last Papers, D. Knowledge, p.258-9)

In 1986, Goldman published Epistemology and Cognition, in which he defended a naturalized version of epistemology. He argued for separate refined treatments of reliabilism, dealing with both knowledge and justification. He also distinguished type reliabilism (he called this "global" reliability) from token-directed model reliabilism ("local reliability).

Goldman argued that epistemology needed help from cognitive science. Years earlier, Willard Van Orman Quine had called for help from the methods of natural science, and thought that epistemology should be "replaced" by psychological science. Goldman says that "historical philosophers never drew a distinction between epistemology and (what is today) called psychology."

In modern times, epistemology should take notice of what cognitive science says about the mechanisms or heuristics of belief formation, both their strengths and their infirmities (Goldman, 2002 a). This way of naturalizing epistemology contrasts with Quine's (1969) way by preserving both the analytic and the normative aspects of epistemology.

(A Companion to Epistemology, 2nd ed., p.144)