Interest in the study of early memory development has increased over the last 20 years. A key step in this progress has been the development of several nonverbal tasks that allow assessment of memory in human infants and in nonhuman primates who cannot express their memories through language. These tasks include delayed non-match to sample (Diamond, 1995), mobile conjugate reinforcement (Rovee-Collier, 1997), deferred imitation (Bauer, 1996; Collie & Hayne, 1999; Meltzoff, 1990), visual paired comparison (Fagan, 1973; Fantz, 1964; Pascalis, de Haan, Nelson, & de Schonen, 1998), habituation-dishabituation (Martin, 1975), and recording of event-related potentials (ERPs; see Nelson, 1994, for a review). Of these tasks, visual paired comparisons and habituation-dishabituation were among the first used in the scientific study of infant memory and are perhaps still the most common. Both techniques infer memory for familiar stimuli based on longer looking at novel compared to familiar stimuli. There has been renewed interest in these tasks due to recent studies of the neurobiological bases of novelty preferences. The aim of this chapter is to review the data and recent theories on the meaning of novelty preferences and evaluate the merits and limitations of these models in light of recent ERP and other neuroscientific data (for comprehensive reviews of the de-
velopment of infant memory more generally, see Diamond, 1990; Nelson, 1995; Rovee-Collier, 1997).