Intelligence versus rationality: Simon’s dueling reference points

Although Simon is best known for the idea of bounded rationality, he rarely used the term in his research in cognitive science: for instance, it is not in the index of his and Newell’s seminal book, Human Problem Solving (1972). This chapter argues that when he became a cognitive scientist, Simon changed his scientific reference point. In the first part of his career Simon debated economists. Those arguments were structured by his opponents’ theoretical reference point: the fully rational agent. In that context, Simon’s contrasting notion of bounded rationality made sense. In the second and longer part of his career, he used an empirical reference point: an intelligent agent who can learn to solve hard problems, such as playing expert chess. This chapter explores how this change in Simon’s scientific reference point affected important aspects of his research. For example, as a cognitive scientist Simon rarely studied the satisficing heuristic, a backward-looking adaptive rule central to his early work on bounded rationality. To a surprising degree, his research from the 1960s onward was often about how experts plan.