Student Number : 0311005K - MA dissertation - School of Social Sciences - Faculty of Humanities / In this paper I track the work of Alvin Goldman, the American epistemologist, from 1979 to 1992 to assess its adequacy as a theory of justified belief. Many philosophers have pointed out that the theory faces problems, the three most important of which I consider. The first is the “clairvoyance problem.” In this case we intuitively deny the status of “justified” to certain beliefs produced by the reliable process of clairvoyance. This indicates that reliable belief formation is not sufficient for justification. The “generality problem,” the second problem, concerns the correct identification and description of the process forming each belief. If the process cannot be identified, then no assessment can be made of a belief’s epistemic status. Moreover, if the process is described too narrowly such that each process only has one output belief, then all true beliefs will be “justified” and all false beliefs “unjustified,” an unacceptable result. If the process is described too broadly then all output beliefs of that process will share an equal epistemic status, also an unacceptable result. Finally, it is possible to challenge the necessity of reliable formation for justification using the case of a cognizer in an evil demon world such that his unreliable visual beliefs are intuitively “justified” since those beliefs are produced by the same reliable processes in our world where they produce justified beliefs. I defend Goldman against these challenges by elucidating subtleties in Goldman’s work that answer these problems and by adding three necessary conditions to his theory. I argue that by modifying Goldman’s early work and rejecting parts of his later work, we can formulate a version of his theory that counts as an adequate theory of justified belief immune to the foregoing problems.