interest (the sucking wanes). But when presented with the new phoneme “pa,” sucking again becomes vigorous, signaling recognition of the change (Eimas et al., 1971). Similar research has indicated attentiveness in the form of shifting gaze in infants as young as 5 months in response to changes in number (Canfield and Smith, 1996). And bodily actions such as kicking and arm waving have, in controlled experiments, indicated the object recognition capacity of infants (Rovee-Collier, 1989).
As they have crafted procedures suited for toddlers and preschool-age children, researchers have documented early cases of abstract reasoning and self-motivated efforts to learn (National Research Council, 1999; Goswami, 1995). In a study by Karmiloff-Smith and Inhelder (1974–1975), preschool children were offered repeated opportunities (over several sessions) to balance each one of a set of blocks on top of another block. Some of the blocks had concealed weights and therefore did not balance at their geometric centers. Initially children succeeded at balancing each block, no matter what kind it was, by using a combination of trial and error and brute force. When a trick block fell, a child moved it around while pushing down on it until balance was achieved. After repeated opportunities to play with the blocks, children started to behave as if they were systematically applying a “blocks balance at their midpoint” hypothesis. As a result, they now started to make errors with the trick blocks. These they placed aside, saying things like “don’t work.”
Of special interest is the fact that there came a time when children changed their strategy again, in a seeming effort to find an all-inclusive solution. On their own they came to realize that they would have to use a third block as a counterweight for blocks that they could feel would not balance at their geometric center. Therefore, they were once again able to balance all the blocks. It is important to highlight the fact that these children gave up the trial-and-error solution that worked in favor of a more systematic one. Had they not done so, they would not have generated the negative data that eventually encouraged them to find a yet more advanced solution.
This view of the child as having ideas or theories about how things work, and as actively engaged in the construction of knowledge stands in stark contrast to earlier views of develop-