Forgetting and Heuristic Inference 1 How Forgetting Aids Heuristic Inference

A few theorists, ranging from William James to contemporary psychologists, have argued that forgetting should not be seen as a nuisance but as key for the proper working of human memory. In this paper, we propose the thesis that forgetting may in addition prove beneficial for inference heuristics that exploit mnemonic information such as recognition and retrieval fluency. To explore the mechanisms that link loss of information and heuristic performance, we bring together two research programs that take an ecological approach to studying cognition: fast and frugal heuristics (Gigerenzer, Todd, & the ABC Research Group, 1999) and ACT-R (Anderson & Lebiere, 1998). Using computer simulations, we show that forgetting boosts the accuracy of the recognition heuristic, which relies on systematic failures of recognition to infer which of two objects scores higher on a criterion value. Similarly, our simulations of the fluency heuristic, which arrives at the same inference on the basis of the speed with which objects are recognized, indicate that forgetting maintains the discriminability of recognition speeds. We conclude that the benefits of forgetting are not limited to generic advantages such as uncluttering memory from out-of-date information. Paradoxically, loss of mnemonic information allows these memory based heuristics to work more efficiently. Forgetting and Heuristic Inference 2