Abstract
Considers critical analysis, when departing from concrete problems and generating testable propositions, as a legitimate aspect of scientific development. Examination of the limitations of behavioristic theory in its passage to therapeutic practice and the nonexplicit supplementary mechanisms inherent in the latter generates several such propositions. 4 of these are presented. Arguments justifying circular definitions of reinforcement and extrapolating from social reinforcement studies in animals to human processes are critically examined. (77 ref.)