Location memory in 8-month-old infants in a non-search AB task: Further evidence

When an object is hidden in a location A and then in a location B, 8-month-old infants tend to search in A if forced to wait 3 s before retrieving the object, and to search randomly in A or B if forced to wait 6 s before retrieving the object (Diamond, 1985). Most investigators have attributed infants' perseverative and random search errors to some immature memory mechanism (e.g., Bjork & Cummings, 1984; Harris, in press; Kagan, 1974; Schacter, Moscovitch, Tulving, McLachlan, & Freedman, 1986; Sophian & Wellman, 1983; Wellman, Cross, & Bartsch, 1987). Baillargeon and Graber (1988) recently tested this hypothesis. They reasoned that if infants' search errors reflect memory difficulties, infants should perform poorly in any task requiring them to keep track of changes in an object's hiding place. The task Baillargeon and Graber devised was a non-search task. In this task, an object was hidden behind one of two screens; after 15 s, a hand retrieved the object from behind the correct screen (possible event) or from behind the incorrect screen (impossible event). The results indicated that the infants still remembered the object's location after the 15-s delay. The present experiments were similar to the experiment carried out by Baillargeon and Graber (1988) except that longer delays were used. In Experiment 1, the object remained hidden for 30 s, and in Experiment 2, for 70 s. The results of the experiments yielded evidence that the infants still remembered the object's hiding place after the delays. Such results point to a remarkable gap between search and non-search assessments of 8-month-old infants' location memory. Like the findings of Baillargeon and Graber (1988), the present findings cast serious doubts on accounts that attribute infants' search errors to inadequate memory mechanisms. In the conclusion section, we speculate on alternative explanations for these errors.