Evolutionary psychologists claim that the human mind contains a far greater number of reliably developing (“innate”), domain-specific, computational adaptations than has been standardly assumed. Perhaps the most prominent philosophical criticism of this multimodularity claim (MMC) is due to Richard Samuels, who has challenged the general arguments advanced in its support.

Emphasizing that evolutionary psychologists posit innate, domain-specific (IDS) computational adaptations qua mechanisms, Samuels argues that MMC is “unwarranted and unmotivated”, since the general arguments offered in its favour fail to give us reason to prefer MMC to a supposedly neglected alternative. According to this alternative, the human mind contains many truth-valuable IDS representations operated upon by domain-general computational mechanisms.

I argue that Samuels' criticism fails on several counts.

(1) Samuels frames the debate in terms of a false dichotomy. Pitting a model of the mind involving IDS structure only qua computational mechanisms against one involving IDS structure only qua representational structures, he associates evolutionary psychology with the first option. In fact, evolutionary psychologists consciously posit IDS mechanisms with IDS representations inextricably tied into them.

(2) Evolutionary psychologists have explicitly argued for the need to posit rich IDS representational structures.

(3) Samuels misstates the MMC, describing it as an hypothesis regarding “central cognition”. In doing so, Samuels begs an important architectural question: Evolutionary psychologists reject classical cognitive science's sequestration of a realm of central cognition. However, Samuels' argument applies only to “central cognition”.

(4) Samuels' discussions of MMC lack examples to support his persistent claim of evidential parity for domain-generally processed IDS representations versus IDS mechanisms. Many counterexamples could be given in which Samuels' alternative doesn't seem close to the former in terms of efficiency or evolutionary plausibility.