Perception of Object Persistence: The Origins of Object Permanence in Infancy

A dominant account of object knowledge in infancy is based on the assumption that infants possess innate core knowledge of objects through which they reason about events and look longer at those that violate their expectations on the basis of this knowledge. In this article, we propose a perceptual model in which younger infants’ perception of object persistence is subject to greater perceptual constraints compared with infants a few months older, and in which young infants require a combination of cues to perceive object persistence across occlusion. Young infants perceive object persistence under limited conditions and over the early months, perception of persistence becomes more robust. The same analysis may be applied to cases in which stationary objects are occluded, including tasks assessing infants’ numerical competence. We argue that these perceptual developments within the first 6 months likely underpin the later development of cognitive principles, including object permanence. KEYWORDS—object persistence; trajectory continuity; object unity; object permanence; object perception Investigation of the development of object permanence in infancy started with Piaget’s (1) the constructionist account, according to which infants progressively construct an objective knowledge of the world, with permanence beginning to emerge around 8–9 months, when infants begin to search for hidden objects. More recently, investigators developed ways to test young infants’ responses to hidden objects that did not rely on searching for objects, and these experiments led to the view that knowledge of the physical world is innate. Specifically, infants possess innate core knowledge and reason about the events that they encounter (2–4). A key aspect of innate knowledge is the principle of object persistence (5). This account is based largely on experiments that measure the time infants look at objects, with longer looks at events that break physical laws interpreted as evidence that infants’ expectations have been violated (the Violation of Expectancy Technique [VoE]). These trials generally follow familiarization to a lawful event, and in this sense, the VoE technique is structurally similar to Habituation-Recovery and Habituation-Novelty techniques that investigate perceptual abilities in infants from birth onwards (6). However, the similarity of the measures leads to interpretative problems for the VoE technique, because experiments must rule out lower-level perceptual bases for longer looking at the violation test trial (7). Investigators have been ingenious in this respect. Nevertheless, we argue that it is often difficult to identify the lower-level principles that may drive infants’ looking. Additionally, we contend that relatively low-level perceptual principles form the developmental origins of later knowledge, and longer looking in VoE tasks reflects detection of a perceptual discrepancy rather than reasoning about the event. Much of the evidence on which claims about innate knowledge are based is from experiments with infants from 3 or 4 months onwards. Stronger evidence for innateness would involve measuring newborn ability, but this kind of work is rare. This leaves a gap of several months during which infants could develop awareness of the world. In this review, we demonstrate that important development occurs during the first 6 months that likely forms a perceptual basis for the knowledge that nativists identify as innate. We are not denying the presence of innate ability, but rather believe that a primarily perceptual ability provides a platform for conceptual knowledge emerging later in infancy. Gavin Bremner, Lancaster University; Alan Slater, Exeter University; Scott Johnson, University of California, Los Angeles. The research in this article was supported by grants from the U.K. Economic and Social Research Council (ESRC): R000239979, RES-000-23-1341, and RES-062-23-1757, to the first and second authors. Correspondence concerning this article should be addressed to Gavin Bremner, Psychology Department, Lancaster University, Lancaster LA1 4YF, UK; e-mail: [email protected].