I have attempted, in the foregoing pages, a description of the function of knowledge in its own terms and on its merits--a description which in intention is realistic, if by realistic we are content to mean naturalistic, a description undertaken on the basis of what Mr. Santayana has well called “following the lead of the subject-matter.” Unfortunately at the present time all such undertakings contend with a serious extraneous obstacle. Accomplishing the undertaking has difficulties enough of its own to reckon with; and first attempts are sure to be imperfect, if not radically wrong. But at present the attempts are not, for the most part, even listened to on their own account, they are not examined and criticised as naturalistic attempts. _They are compared with undertakings of a wholly different nature, with an epistemological theory of knowledge, and the assumptions of this extraneous theory are taken as a ready-made standard by which to test their validity._ Literally of course, “epistemology” means only theory of knowledge; the term _might_ therefore have been employed simply as a synonym for a descriptive logic; for a theory that takes knowledge as it finds it and attempts to give the same kind of an account of it that would be given of any other natural function or occurrence. But the mere mention of what _might_ have been only accentuates what is. The things that pass for epistemology all assume that knowledge is not a natural function or event, but a mystery. Epistemology starts from the assumption that certain conditions lie back of knowledge. The mystery would be great enough if knowledge were constituted by non-natural conditions back of knowledge, but the mystery is increased by the fact that the conditions are defined so as to be incompatible with knowledge. Hence the primary problem of epistemology is: How is knowledge _überhaupt_, knowledge at large, _possible_? Because of the incompatibility between the concrete occurrence and function of knowledge and the conditions back of it to which it must conform, a second problem arises: How is knowledge in general, knowledge _überhaupt_, _valid_? Hence the complete divorce in contemporary thought between epistemology as theory of knowledge and logic as an account of the specific ways in which particular beliefs that are better than other alternative beliefs regarding the same matters are formed; and also the complete divorce between a naturalistic, a biological and social psychology, setting forth how the function of knowledge is evolved out of other natural activities, and epistemology as an account of how knowledge is possible anyhow.