The claim that psychology is an autonomous science (e.g., Bunge, 1990), is approached from the question of what a proper level of analysis is. Emergentism (Bunge, 1980; Margolis, 1986) and Turing Machine functionalism (Fodor, 1975) are discussed as alternatives to the received view that psychological laws should be reduced to lower, more elementary levels of explanation, such as physiology. This paper attempts to argue that psychology is distinct from lower level sciences due to the emergent nature of its phenomena, and also in its use of functional rather than causal subsumptive explanation (Cummins, 1983). Functionalism is seen as either narrow or wide. "Narrow" (Turing Machine) functionalism adheres to a "narrow" construal of mental content, in which reference to the environment is banished from psychological laws (Fodor, 1991); it is suggested that recent work in the philosophy of mind on so-called "wide functionalism" (e.g., Harman, 1988) where psychological theories are taken to contain (implicit) reference to the environment, is congenial to a functional teleological perspective, and may supplement functional explanation. Additionally, Millikan's (1989) biosemantics, treating mental processes as the proper functions of the mind-brain, is introduced as an example of a biologically based approach to mental content.