5-fold larger images. We saw the

same regions selectively activated by Faces, Shapes, and Learned symbols irrespective of stimulus size, order, font, or position (Figure S2). Because of their age, we could not scan the juveniles before we commenced Symbol training, so we cannot rule out the unlikely possibility that the four juvenile monkeys might have exhibited Learned symbol-selective cortical domains without training, though the Paclitaxel cost absence of a Learned symbol-selective region in any of the adults makes this unlikely. Four juvenile monkeys learned to recognize symbols faster than six sexually mature adults and showed faster reaction times than the adults in choosing between symbols, even though the reaction times and learning rates of the adults were comparable to the juveniles when choosing between dot arrays. Functional MRI on the juvenile monkeys showed novel domains that were more active when the monkeys viewed the Learned symbols, compared to visually similar but Untrained shapes, and Faces. The same location in the adults responded as strongly to Untrained shapes Volasertib as to Learned symbols. The anatomical

results indicate that intensive early, but not late, experience can cause the formation of a novel specialized cortical domain, or cause an existing domain to become specialized for the trained shapes. The association of aminophylline a specialized domain with faster learning and responding suggests that having a specialized domain bestows a behavioral advantage. These results raise two important

questions: (1) How could intensive early experience cause the formation of a novel functional domain? Our results are completely consistent with the possibility that early symbol learning modifies the tuning properties of cells in an innately specialized domain (Dehaene and Cohen, 2007). We would like, however, to propose an alternative hypothesis: the emergence, only in the juvenile-trained monkeys, of a domain selective for an artificial object category raises the possibility that early experience plays a causal role in the formation or specialization of functional domains. The functional domains for faces and shapes were not in precisely the same location in each monkey, but the paired pattern of face and shape domains within each of the major subdivisions along inferotemporal cortex was similar in all the monkeys and was similar to what has been previously reported (Bell et al., 2009 and Denys et al., 2004). The experience dependence of the novel functional domain, coupled with the pattern of one pair of face and shape functional domains within each major cortical area, suggests a self-organizing Hebbian mechanism.