Mareschal et al. present two well written volumes articulating their version of neuroconstructivism. Specifically, these authors seek to probe the mechanisms and processes whereby the inchoate infant develops cognitively and behaviourally into a sentient, cognizant adult. The authors advocate an integrative, multidisciplinary framework, incorporating data from a range of methodologies, including cognitive studies, computational modelling and neuroimaging.

The authors directly take issue with the Marrian view that levels of analysis can be considered independently at the computational, algorithmic, and implementation levels. Instead, they place much greater weight on the mutual constraints operating between levels of description. Specifically, the authors argue that theorizing about cognitive development should be explicitly informed by considerations of processes occurring at the neural implementation level – a view with which we firmly concur. However, Mareschal et al. apparently neglect to acknowledge the degree of neural rewiring that occurs at later stages of individual development. Specifically, there is accumulating evidence of significant reorganization of the neural substrate occurring during adolescence and even into early adulthood (Casey et al. Reference Casey, Giedd and Thomas2000; Huttenlocher & Dabholkar Reference Huttenlocher and Dabholkar1997; Paus Reference Paus2005; Spear Reference Spear2000). These reorganizational processes take place during a phase of development when fundamental modular neurodevelopmental capacities such as those presented by Mareschal et al. (e.g., visual perception, habituation, phonological development, reading) have already been acquired (Luna et al. Reference Luna, Garver, Urban, Lazar and Sweeney2004). Potentially linked to these neural changes (and notwithstanding issues pertaining to adoption of the maturational versus interactive specialization position; see Foster Reference Foster1997 for a related discussion), there is evidence of development during the teenage years with respect to elements of higher cognitive functioning, particularly regarding the ontogeny of executive functions (Luna & Sweeney Reference Luna and Sweeney2004). Again, this is a domain of the extant literature that is apparently neglected by Mareschal et al. across the two volumes. It is of considerable interest to reflect on how the authors' conceptual framework would address these and related issues, which we are currently investigating empirically in Western Australia (via the Western Australian Pregnancy Cohort Study, evaluating a group of individuals who have been followed developmentally since before birth through to their current age of 16–18 years). Of potential relevance is the consideration that executive functions are not usually regarded as modular in either the weak or strong (i.e., Fodorian) theoretical sense.

In addition to the main conceptual issues discussed above, some more specific assertions made by the authors are also questionable with respect to the nexus between prefrontal cortex and executive functions. Of specific interest is the authors' contention (vol. 1, p. 65) that it is difficult to account for “decreases in the extent of cortical activation in terms of the progressive maturation of prefrontal cortical areas.” This position is not justified by the authors with respect to independently acquired evidence or data, and it represents an interesting contrast regarding other areas of the lifespan developmental literature. For example, in the aging literature it has been argued that overactive cortical networks (as evidenced, for example, by functional neuroimaging investigations) are indicative of less efficient functional capacity, and that such overactivity may portend decline and dissolution of those networks and the cognitive functions they subserve (Dickerson et al. Reference Dickerson, Salat, Greve, Chua, Rand-Giovannetti, Rentz, Bertram, Mullin, Tanzi, Blacker, Albert and Sperling2005; Hamalainen et al. Reference Hamalainen, Pihlajamaki, Tanila, Hanninen, Niskanen, Tervo, Karjalainen, Vanninen and Soininen2007). It has been further suggested that, in older individuals, differences in regional brain integrity may be linked to functional reorganization through changes in processing strategy (Greenwood Reference Greenwood2007). It seems plausible that a converse process may take place during acquisition of cognitive skills, as neural networks become more finely tuned and selective with respect to those external and internal environmental events that are necessary and sufficient for specific networks to be activated. In summarizing this section of text (vol. 1, p. 67) the authors contend that “frontal cortical regions, traditionally believed to be slow to mature, may be active from early in life.” However, Mareschal et al. apparently omit to consider the degree to which such activation patterns may be modified across childhood and into early adulthood, specifically with respect to higher cognitive functions – as previously noted. Considering the multiple cytoarchitectonic regions comprising the frontal cortex (a disproportionately large and highly functionally elaborated brain region in humans relative to other mammalian and primate species), we think it is overly simplistic to adopt a uniform, homogeneous position with respect to the maturation of capacities subserved by this region (see Rabbitt Reference Rabbitt1997 for a review). Moreover, the authors themselves acknowledge (vol. 1, p. 213) that “cognitive development throughout late childhood and the teenage years is a story of increasing abstraction,” and that with respect to the neural substrate underlying this process of increasing abstraction “one candidate may be the prefrontal cortex” (vol. 1, p. 214).