It is becoming increasingly clear that little in development is predetermined or permanently fixed. Rather, gene expression is activity dependent, and epigenesis is probabilistic. So, the study of genetic disorders needs to change from the still widely held view that developmental disorders can be accounted for in terms of intact versus impaired modules, to one which takes serious account of the fact that the infant cortex passes from an initial state of high regional interconnectivity to a subsequent state of increasing specialization and localization of function. With such early interconnectivity in mind, developmental neuroscientists must consider the possibility that an early deficit in one part of the brain may have subtle effects on other parts of the developing brain, even when scores fall 'in the normal range'. In studying developmental disorders, it is thus crucial to examine not only domains of clear-cut deficit, but also domains of behavioural proficiency. Atypical epigenesis may often involve a lack of specialization and localization of brain function over developmental time, even in cases of behavioural proficiency.