ABSTRACT
This chapter argues that introspection is a legitimate source of designs in artificial intelligence (AI). Introspection is legitimate as a basis for AI designs because it would be used in the context of discovery, where ideas are born. Herbert Simon continues Watson’s objections to introspection, though his case is a bit more complex. The focus of the literature condemning subjectivity and introspection is in the science of the psychology, that is, the development and assessment of knowledge and models about the natural facts about human behaviour. Introspection as a formal research method in the psychology, of course, had been comprehensively discredited decades earlier. John B. Watson saw himself as pushing for better scientific practise in psychology. His actual position was both more subtle and more strident than that – but that makes little difference to his eventual influence on AI.
