This paper begins with a summary and analysis of Robert Glaser's arguments, presented in his paper "Education and Thinking: The Role of Knowledge," which contend there are strong interactions between structures of knowledge and cognitive processes, and that problem solving and reasoning are taught best in the context of the acquisition of domain-specific knowledge. Glaser's arguments are considered in light of evidence, cited in his 1984 paper, from three sources: developmental psychology, studies of expert and novice problem solving, and process analyses of intelligence and aptitude tasks. The paper then seeks to translate six of Glaser's claims about expertise in problem solving into testable hypotheses. Evidence in support of these hypotheses is then cited from three curricular domains: physics, mathematics, and social studies. In each of these areas, the differences between novices and experts point to a relationship between structures of knowledge and cognitive processes. It is concluded that although Glaser's hypotheses have received tentative support from novice-expert studies in three curricular areas, more research is needed to investigate the validity of Glaser's assertions. Four pages of references are included. (TE)