When in competition against engrained habits, is conscious representation sufficient or is inhibition of the habit also needed?

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd. p < .10 for this contrast.) Finally, even if the U-shaped effect is reliable, its origin need not lie in conscious reflection. A real possibility, and one that Marcovitch and Zelazo consider, is that it results from habituation. That is, infants may be more likely to correctly search at B on the first B trial after many A trials than after an intermediate number of A trials because after many A trials they habituate and, as Marcovitch and Zelazo suggest, start responding randomly. They rule this explanation out because, it is claimed, Marcovitch et al. ’s (2002) participants showed consistency of response on B trials, i.e. ‘once a correct response was produced, all subsequent responses were correct’ (Marcovitch & Zelazo, 2008, p. 12). In fact, while most infants in the 2002 study did show this consistency, 41% did not! Furthermore, the observed level of consistency of response may arise from infants’ attending to the novel event (hiding at location B) following habituation. Thus, the evidence for conscious reflection as a mediator of correct performance remains weak. What, then, can we take from the HCSM? First, despite the above criticisms, Marcovitch and Zelazo deserve praise for using their model to generate a series of novel predictions, and in so doing illustrating the utility of formal methods in theory construction within developmental cognitive science. Whether the authors’ predictions are unique to the HCSM is unclear, but only the development of complete specifications of competing accounts will resolve this issue. More positively, Marcovitch and Zelazo’s survey of A not B models and theories highlights important commonalities across the models – namely that all existing models involve two systems or layers, with the lower being modulated by the higher, and the higher coming on stream later in development. These commonalities represent significant progress in our understanding of development, and clarify the areas of debate. A useful theory of development of EF must, however, do more than posit a single construct at the higher level. It must decompose that construct and relate it more directly to theories of adult EF.