The second advantage that the problem-solving deficiency explanation has over alternative accounts of infants' AB errors is that it can be integrated relatively easily with the explanation, discussed in the previous section, of young infants' failure to search for hidden objects. Briefly, it is assumed that infants fail to search because they are unable to plan means-end sequences of actions; and that they search perseveratively, once they begin to search, because they are overly inclined (for reasons that are still unspecified) to rely on previously computed means-end sequences, rather than recompute or replan new ones. Furthermore, in both cases, infants show themselves better able to identify than to generate correct action sequences: as shown earlier, infants identify sequences that can result in the recovery of objects placed under obstacles or at the far end of supports long before they produce these sequences themselves ( Baillargeon et al., 1990, 1992); in addition, infants identify context-appropriate searches after delays of 15, 30, and even 70 seconds long before they search correctly with similar delays ( Baillargeon & Graber, 1988; Baillargeon et al., 1989).
The research summarized in this chapter has implications for at least three areas of infant development: object permanence, physical reasoning, and search. They are discussed in turn.
When adults see an object occlude another object, they typically assume that the occluded object (a) continues to exist behind the occluding object; (b) retains its physical and spatial properties; and (c) remains subject to physical laws. Piaget ( 1954) proposed that infants initially do not share adults' beliefs about occlusion events, and adopt these beliefs one by one over the first two years of life.
The findings reported in this chapter clearly contradict Piaget's proposal. Consider the many experiments that obtained positive results with infants aged 3 1/2 to 5 1/2 months: the rotating screen experiments ( Baillargeon, 1987a, 1991, 1992; Baillargeon et al., 1985), the rolling car experiments ( Baillargeon & DeVos, 1991), the sliding rabbit experiments ( Baillargeon & Graber, 1987; Baillargeon & DeVos, 1991), and the searching hand experiments ( Baillargeon et al., 1990). The infants in these experiments seemed to have no difficulty representing the existence of one, two, and even three hidden objects. Furthermore, the infants represented many of the properties of the objects, such as their height, location, and trajectory. Finally, the infants expected the objects to behave not in capricious and arbitrary ways but in the same regular and