Striking a Blow for Sanity in Theories of Rationality

I took the title of this chapter from an email Herbert A. Simon sent me in May 1999. In this email, he wrote a statement for the back cover of Simple Heuristics That Make Us Smart in which he commented: “I think the book strikes a great blow for sanity in the approach to rationality [and shows] why more than minor tampering with existing optimization theory is called for.” But Herb wouldn’t be Herb if he hadn’t added “and you wouldn’t believe I had ever skimmed the volume if I didn’t fi nd SOMETHING to disagree with.” And so he continued, pointing out that he hadn’t found the expert/novice topic treated, that scientifi c discovery would have been a great example for ill-structured domains ... Bringing sanity into theories of rationality was a major guideline in Herbert Simon’s scientifi c life. However, as he himself was prepared to admit, sanity in rationality entered his thinking as a negatively defi ned concept, a kind of black box that contained everything that was not optimization. What he opposed has various names: full rationality, substantial rationality, maximization of expected utility, Homo economicus, or simply optimization. What he proposed had its seeds in his revised dissertation, Administrative Behavior (1947), and eventually became termed bounded rationality, satisfi cing, or procedural rationality. Because of its initial vague defi nition, bounded rationality, however, came to mean many things to many people.