Cognitive Simplicity as an Idealization

Appeals to the simplicity of hypotheses about cognitive processes are common in comparative psychology. Much recent work has discussed the role of simplicity in privileging some hypotheses over others. Simpler hypotheses tend to be taken as the default, working hypothesis, so long as there is not any strong evidence against them. Here, I argue that cognitive simplicity also plays a role in hypothesis generation, aiding comparative psychologists to create new hypotheses about behavioural processes. I attempt to justify the role that cognitive simplicity plays here. One approach is to justify that some hypotheses really are simpler than others. Unfortunately, there are several jointly contradictory and individually problematic ways of defining cognitive simplicity that undermine this effort. Instead, I propose that cognitive simplicity is more appropriately interpreted as a family of idealizations about behavioural processes. Idealizations are useful abstractions about phenomena, based on potentially false assumptions, which are justified by serving a purpose for practicing scientists. Idealizations about the properties of behavioural processes help comparative psychologists to creatively generate novel hypotheses about animal behaviour. This is a useful strategy when handling the fact that there are usually several empirically distinct hypotheses that could explain behavioural observations. This view preserves cognitive simplicity as a useful concept for hypothesis generation, while blocking it from involvement in hypothesis selection, in line with previous work.