Neurocognitive Development Unit, London

Our commentary on these four interesting articles will focus on two issues. First, we ask whether the notion of modularity is doing any work for theorists in the 1990s. Despite our theoretical leanings, we strongly prefer Fodor’s (1983) precise definition of a module. Once one weakens Fodor’s criteria, as these four papers do in various ways, then the notion of modularity becomes loosely descriptive rather than retaining the strong theoretical base that leads to precise predictions about what can or cannot interact with what in the human brain. Second, even when one relaxes Fodor’s criteria to create a weaker notion of modularity, we argue that, given everything that is now known about postnatal neocortical development, it is more likely that modularity is the outcome of development rather than its starting state.

[Commentary on: Mike Anderson "Mental retardation, general intelligence, and modularity", Simon Baron-Cohen "Does the study of autism justify minimalist innate modularity", Ianthi-Maria Tsimpli & Neil Smith "Modules and quasi-modules: language and theory of mind in a polyglot savant", and Louise Phillips and Sergio Della Sala "Ageing, intelligence, and anatomical segregation in the frontal lobes", in M. Anderson (Ed.), Modularity and Intelligence (Special Issue of Learning and Individual Differences). JAI Press Inc: New York.]