Progress has been made within the general field of educational evaluation in developing more expansive views of inquiry and practice. By contrast, and despite notable exceptions (i.e., Eisner, 1974; Garman, 1983), supervision and evaluation of teaching remains a more restrained field trapped in a web of primitive scientism. The basic issues raised in this paper are not therefore new (Gieryn, 1982) but have been neglected with respect to supervision and evaluation. Presently, normal science in supervision and evaluation can be described as scientific-technical. The cognitive organizers for thinking about this normal science are defining competencies, measuring performance, developing appraisal systems, techniques, and procedures, and conceiving of teachers as instructional treatments (see, e.g., Borich, 1977; Popham, 1975). This highly rational, objective, and technical view of reality reflects a larger world view often referred to as Cartesian and attributed to the philosophers Rene Descartes and Francis Bacon and the theoretical mathematician Issac Newton. Essential to this world view is that matter is the basis of all existence and that mind and matter are separate, with the latter providing the legitimate subject matter for scientific inquiry. Further, this material world is viewed as a multitude of separate objects assembled into a complex cosmic machine. The task of science is to understand this machine by reducing phenomenon to its elementary building blocks and looking for parts that interact. This is a mechanical process known as reductionism. Examples of reductionism in the supervision and evaluation of teaching would be the reducing of complex patterns of human interaction to tallies on a data collection schedule and reducing the phenomenon of student response to teaching to the timing of specific student task behaviors. This paper does not argue against the scientific-technical view of supervision and evaluation per se but against its omnipresent tendencies, its presumed infallibility, and its naive simplicity. Instead, an expansive view is sought, one that provides a comprehensive theory of rationality sensitive to practice. Thomas McCarthy (1978), referring to the philosopher Jurgen Habermas, states the case nicely: