ABSTRACT

This paper addresses the question: “How is it possible to understand?”. The source of knowledge for what we have to say is a long-standing research effort to produce an understanding program called Merlin. That effort has been one primarily in artificial intelligence and thus this paper will also reflect primarily artificial intelligence concerns. Its appropriateness in a psychological context rests on a view that artificial intelligence is simply part of theoretical psychology (Newell, 1970). Enough evidence for this general view exists that it need not be labored here (e.g., Hunt, 1971; Newell & Simon, 1972).