Evolutionary psychology is sometimes proposed in response to the shortcomings of idealized models of cognition:
Distributed and bottom-up AI shows how incremental changes (as in genotype - eg. A-life simulation of evolution) might give rise to apparently qualitative differences in functioning. Perhaps complex psychological functions evolved in this way?
Selective processes have been suggested at scales ranging from the neural to the cultural. Some ubiquitous, systematic cognitive biases, for example, in probabilistic and formal reasoning, in judgement under uncertainty and in memory (seen also in computing of functions requiring holding in mind of large amounts of information), have been attributed to selection pressures of the hunting-gathering lifestyle whose demands included tool-making and using, escaping from predators, mating, and child-rearing.
At the same time innate strengths of the human nervous system in the visual and language modalities have also been ``explained'' thus. These include parsing of visual information, especially information coded by color and/or motion, language acquisition and grammars. Some domain-specific innate cognitive strengths have been suggested, like naive physics (eg. Spelke), biology and psychology (Wellman & Gelman (1998) Knowledge Acquisition in Foundational Domains).
In the context of situated cognition it has been proposed that there might be an evolutionary selection of context-sensitive mechanisms. (And/ or that in the appropriate context the presumed evolutionary cognitive limits may not be detectable?)
The field of evolutionary psychology has however been marked by bitter controversy. See Rose, Hilary and Steven Rose (Eds.) (2001) Alas Poor Darwin: Arguments Against Evolutionary Psychology. Vintage (Random House), London, U.K.
Psychology Tutorial (1-11-02)