Renee Baillargeon - US grants

Investigators have traditionally claimed that infants do not begin to view objects as permanent until 9 or 10 months of age. However, preliminary studies indicate that infants as young as 5 months of age recognize that objects continue to exist when occluded. The research described in this proposal builds on the results of these initial studies. Three general issues will be addressed. The first concerns infants' ability to represent the physical and spatial properties of objects and to use this information to reason about simple occlusion events. Two types of occlusion events will be investigated. In one, a stationary object will be placed in the path of a moving screen. The proposed experiments will test whether infants represent information about the height and location of the object and use this information to predict when the screen should stop. In the second type of occlusion sequence, an object will travel along a track, part of which will be occluded by a screen. The proposed experiments will test whether infants represent information about the height of the object and use this information to determine whether a portion of the object should remain visible above the screen. The second issue is closely related to the first. It concerns infants' ability to make inferences about properties of occluded objects. One experiment will examine whether infants can infer the distance an object travels behind a screen, based on its speed of movement and the duration of its occlusion. Other experiments will test whether infants can infer where an object is hidden, based on specific types of cues. For instance, if shown a flat cover and a cover with a distinct lump, will infants realize that the object must be hidden under the lumpy cover? The third issue concerns the development of infants' reasoning about occluded objects. At what age do infants understand that a stationary object continues to exist when occluded, or that a moving object continues to exist and pursues its trajectory behind an occluder? At what age do infants become able to represent and use information about the physical and spatial properties of occluded objects? Finally, when do infants become able to make inferences about the physical features and spatial properties of occluded objects? The research described in this proposal has implications for two, hitherto disjoint areas of infancy research: infants' object concept, and infants' knowledge of the physical world.

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 the physical and spatial properties it possessed prior to occlusion; and (c) is subject to the same physical laws as prior to occlusion. Piaget (1954) claimed that infants do not begin to share these assumptions until about 8, 12, and 18 months of age, respectively. Over the past three years, we have conducted an extensive series of experiments designed to investigate Piaget's description of the development of infants' beliefs about occluded objects. The results we have obtained do not support Piaget's description. These results indicate that (a) infants as young as 3 months of age understand that objects continue to exist when occluded; (b) infants as young as 5.5 months of age are aware that objects retain their physical and spatial properties when occluded; and (c) infants as young as 9 months of age make inferences about occluded objects. In general, these findings indicate that young infants' understanding of the physical world is far more advanced than the work of Piaget and his successors had led us to believe. The first aim of the proposed research is to pursue a number of questions raised by our findings, namely: (1) Are infants less than 5.5 months of age able to represent the properties of occluded objects? (2) How sophisticated are 9-month-old infants at inferring the physical and spatial properties of occluded objects, and are younger infants also capable of making such inferences? Finally, (3) How can one account for the discrepancy between Piaget's conclusions and those suggested by our findings? The second aim of this application is to explore other facets of infants' physical world. Two physical domains will be investigated: (1) infants' understanding of support relations and (2) infants' understanding of collision events. Since our research has revealed that infants' understanding of object permanence is surprisingly sophisticated, it is reasonable to expect that investigations of other aspects of infants' physical world will also bring to light hitherto unsuspected competencies.

Recent research on infants' understanding of the physical world indicates that infants share many of adults' beliefs about objects. For example, there is now substantial evidence that even young infants appreciate that objects cannot remain stable without support. The proposed research builds on these findings. It explores infants' ability to use their physical knowledge to make decisions about (a) which surfaces belong to the same object and (b) which objects belong to the same category. Segregation and categorization are both tasks that show clear knowledge effects in adults. Preliminary findings indicate that young infants can use their intuitions about support to segregate and categorize objects. The proposed experiments will attempt to extend these findings by considering other facets of infants' physical knowledge. In addition, the proposed experiments will begin to explore how infants come to integrate their physical knowledge with conflicting information, especially perceptual information, in grouping surfaces or objects. Preliminary results suggest that, in both segregation and categorization tasks, infants go through phases in which they attach greater importance to perceptual than to physical knowledge information. Experiments are planned. to test the generality of these phases, and to specify the experiences that enable infants to progress past these phases.

[unreadable] DESCRIPTION (provided by applicant): As they look about them, infants routinely observe many different physical events: for example, they may see a parent pour juice into a cup, stack dishes on a table, or store groceries in a cupboard. Over the past 10 years, my collaborators and I have been working on developing an account of how infants reason about simple physical events, and how this reasoning becomes gradually more sophisticated with experience. In the Progress Report, I summarize the results of 15 projects conducted during the last grant period that helped us test and extend our account. As a result of this research, we now have a much clearer idea of how infants' physical-reasoning system operates: what specific information infants are likely to represent when watching a physical event, and how they are likely to use this information to interpret and predict the event's outcome. Over the past few years, we have also begun to explore how infants who do not spontaneously represent key information about an event may be induced to do so through various contextual manipulations. Some of these effects appear to depend on subtle interplays between the physical-reasoning system and two other systems suggested by findings in the adult and infant visual cognition literature: the object-tracking system and another system my collaborators and I term the object-representation system. This is a truly exciting era in the field of infant cognition, as developments in different subfields are coming together to paint a much more detailed picture of the computational architecture that underlies infants' responses to objects and events. The first project in the next grant period (PROJECT 1) will continue our investigation of the physical-reasoning system, and will attempt to shed light on striking findings from the last grant period having to do with event category effects and dicalages in infants' acquisition of their physical knowledge. The other projects (PROJECTS 2-11) will explore possible links between the physical- reasoning system and the two other systems mentioned above. In particular, these projects should shed light on how infants respond not only to single events, but also to sequences of events involving the same or different objects. In all, 11 projects (comprising 30 experiments) are planned that will allow us to dramatically expand our understanding of infants' physical reasoning, and to establish new and productive connections to other subfields of infant cognition. As in the previous grant period, the projects will use violation-of-expectation and action tasks. In addition, we are in the process of developing one additional method anticipatory-looking tasks that should provide further converging evidence for our conclusions. PUBLIC HEALTH RELEVANCE: During the first year of life, infants normally acquire a great deal of knowledge about the physical world. How do they do so? The present research will help shed light on the cognitive architecture that allows infants, from a very early age, to represent, to reason, and to learn about physical events. [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Traditionally, researchers assumed that infants understand very little about the physical world. With the advent of more sensitive methods, however, investigators have come to realize that even young infants possess expectations about physical events. What is the nature of these early expectations, and how do they develop over time? These questions have been at the core of Baillargeon's research program for the past 20 years. This application seeks to test and extend two accounts recently developed by Baillargeon and her collaborators: the first focuses on how infants use their current physical knowledge to reason about physical events and predict their outcomes (reasoning account); the other account examines how infants attain new knowledge about physical events (learning account). Each account makes several testable predictions that will be investigated in the next grant period. The proposed research builds on experiments conducted in the previous grant periods, and also introduces several new research directions. The proposed experiments will make use of two different methods, the violation-of-expectation (VOE) and the object-manipulation (OM) method. In all, 18 projects are planned, organized into seven, inter-related lines of research. These lines examine: (1) the formation and use of event categories; (2) the acquisition of variables in individual event categories; (3) early competencies in infants' reasoning about events from different categories; (4) cueing infants to reason about a new variable in an event category, through exposure to an event from a different category in which this variable has already been identified; (5) effects of event category knowledge on infants' ability to detect surreptitious changes in variable information; (6) teaching infants (in or out of the laboratory) a new variable in an event category, through exposure to appropriate events from the category; and finally (7) the formation and use of abstract object categories, namely, inert and self-moving objects. The proposed research will help us better understand how infants reason and learn about physical events, and as such will give us a conceptually richer and more detailed picture of this facet of cognitive development in infancy.

As they look about them, infants routinely observe many different physical events: for example, they may see a parent pour juice into a cup, stack dishes on a table, or store groceries in a cupboard. Over the past 10 years, my collaborators and I have been working on developing an account of how infants reason about simple physical events, and how this reasoning becomes gradually more sophisticated with experience. In the Progress Report, I summarize the results of 15 projects conducted during the last grant period that helped us test and extend our account. As a result of this research, we now have a much clearer idea of how infants' physical-reasoning system operates: what specific information infants are likely to represent when watching a physical event, and how they are likely to use this information to interpret and predict the event's outcome. Over the past few years, we have also begun to explore how infants who do not spontaneously represent key information about an event may be induced to do so through various contextual manipulations. Some of these effects appear to depend on subtle interplays between the physical-reasoning system and two other systems suggested by findings in the adult and infant visual cognition literature: the object-tracking system and another system my collaborators and I term the object-representation system. This is a truly exciting era in the field of infant cognition, as developments in different subfields are coming together to paint a much more detailed picture of the computational architecture that underlies infants' responses to objects and events. The first project in the next grant period (PROJECT 1) will continue our investigation of the physical-reasoning system, and will attempt to shed light on striking findings from the last grant period having to do with event category effects and d¿calages in infants' acquisition of their physical knowledge. The other projects (PROJECTS 2-11) will explore possible links between the physical- reasoning system and the two other systems mentioned above. In particular, these projects should shed light on how infants respond not only to single events, but also to sequences of events involving the same or different objects. In all, 11 projects (comprising 30 experiments) are planned that will allow us to dramatically expand our understanding of infants' physical reasoning, and to establish new and productive connections to other subfields of infant cognition. As in the previous grant period, the projects will use violation-of-expectation and action tasks. In addition, we are in the process of developing one additional method anticipatory-looking tasks that should provide further converging evidence for our conclusions. Project Narrative During the first year of life, infants normally acquire a great deal of knowledge about the physical world. How do they do so? The present research will help shed light on the cognitive architecture that allows infants, from a very early age, to represent, to reason, and to learn about physical events.