ABSTRACT
In recent years, cognitivism has come under attack. The computer metaphor has been criticized as fundamentally inadequate, and workers on behavior-based artificial intelligence (AI) have suggested that the cognitive-science approach through representation has failed to achieve real intelligence. Lower-level, nonsymbolic tasks, they argue, are more fundamental to biological intelligence, evolved earlier, and are more difficult to re-create by artificial methods. At the same time, a new theoretical behaviorism is emerging that shares the behavior-based AI emphasis on intelligent behavior as the outcome of self-organizing interactions among independent, unintelligent processes.
