One of the major developments in recent philosophy is the movement towards a naturalized epistemology.² Although there is the usual disagreement in detail among proponents of this movement, one proposition is central: that human knowledge is a natural phenomenon. This, in turn, suggests an evolutionary epistemology based on the thesis that the human species appeared in this world without any privileged access to its nature, and without any privileged access to the best methodology for learning about the world. Through time we have learned much on both fronts, and the development of science in particular has increased both our knowledge of the world we live in, and our knowledge of how to learn about that world. Still, in both cases, our knowledge remains tentative and uncertain, and we have every reason to assume that continued research will continue to bring surprises. An immediate consequence of this is that we should not expect a higher degree or certainty in epistemology than in science, and one of the epistemologist's major concerns is to examine the best available examples of human knowledge as a basis for reassessing our current understanding of the way in which the knowledge enterprise proceeds. In this paper I want to consider scientific observation from a naturalistic point of view.