ABSTRACT
There are, perhaps, considerable gains to be expected from the development of so-called teaching machines and programmed instruction (TMPI, for short). There are surely dangers, however, in a behavioristic rationale that is philosophically naïve or logically confused. For this rationale not only guides the technical work of the laboratory but is offered to the public and profession as educational theory, with general import for practice, and moreover, the stamp of scientific authority. A consideration of two pieces in the Fall 1961 issue of the Harvard Educational Review will, I believe, indicate certain of the defects and dangers involved. I refer to the article by B. F. Skinner and the review by Harlan Lane. 1
