ABSTRACT
In discussing papers presented to the Carnegie Symposium on visual per ception a half dozen years ago, Allen Newell (1973) argued that the speak ers were playing 20 Questions with nature. He appeared to feel the number of admissable queries in that game was being exceeded, and it was time to establish a different strategy for the analysis of mind. The strategy he proposed was the development of detailed informationprocessing models appropriate to complex tasks such as reading or listen ing. These models could be used to integrate the findings of experiments and would themselves be tested by comparison to human performance.
