Effects of Motion and Figural Goodness onHaptic Object Perception in InfancyArletle StreriVniversite de Paris VElizabeth S. SpeikeCornell UniversitySTREHI, ARLETTE, and SPELKE, ELIZABETH, S. Effects of Motion and Figural Goodness on HapticObject Perception in Infancy. CHILD DEVELOPMENT, 1989, 60, 1111-1125. 4-month-old infants held2 rings, 1 in each hand, out of view. Tbe rings moved rigidly together and were either the same(Experiment 1) or difFerent (Experiment 2) in substance, weight, texture, and shape. After baptichabituation to a ring display, patterns of preferential looking to visibly connected vs. separated ringsprovided evidence that tbe infants perceived tbe rings in both experiments as parts of one connectedobject. This perception was no weaker when the rings differed in shape and substance, even thoughinfants were shown (Experiment 3) to detect that difference. In tbe haptic mode, as in the visualmode, infants appear to perceive object unity by analyzing motion but not by analyzing figuralgoodness. The findings suggest that an amodal mechanism underlies object perception.Human adults perceive the surrounding a larger object that moves in a difFerent direc-environment as a layout of unitary, bounded tion, for example, infants perceive the objectsobjects. This ability is remarkable because as distinct units, even if their images overlapthe stimulus information for object bound- in the visual field (Hofsten & Spelke, 1985;aries is highly incomplete. In the visual see also Spelke, Hofsten, & Ke.stenbaum,mode, every opaque object is partly hidden: 1989). When a moving object is presented be-lts back is occluded by its front, and its front hind a central occluder such that its two vis-is usually partly occluded by other objects, ible ends are displaced together, infants per-Most objects, nevertheless, are perceived as ceive the object as a complete unit thatcomplete units rather than as collections continues behind the occluder (Kcllman,of visible fragments (Michotte, Thines, & Gleitman, & Spelke, 1987; Kellman & Spelke,Crabbe, 1964). In the haptic mode,^ only 1983; Kellman, Spelke, & Short, 1986).small regions of the surface lavout are en- , „ ,countered at any given time as a perceiver , *" contrast, young mfants do not appearexplores by touching. Objects are again per- ^^ Perceive object unity arid boundaries in ac-ceived as continuous wholes, however, not as ?°J^ ^'* ^^ general tendency to maximizethe patches of surfaces in contact with the fig»^al goodness. Experiments provide evi-fingers (Gibson, 1962). "^^^"^^ ?^f* !"f^"t^ perceive two stationary ob-jects of different colors, patterns, and shapesSome recent attempts to understand as a single unit if the objects stand adjacent tothese abilities have focused on their early de- each other in any dimension, including adja-velopment (see Spelke, 1988, for a review), cency in depth (Kestenbaum, Termine, &Studies of object perception in the visual Spelke, 1987; see also Hofsten & Spelke,mode provide evidence that 3-5-month-old 1985; Prather & Spelke, 1982; Spelke et al.,infants perceive object boundaries and object 1989). Furthermore, infants do not appear tounity by detecting patterns of surface motion, perceive the unity of a stationary, center-When a moving object is presented in fTont of occluded object of a uniform color and pat-Supported by grants to Elizabeth Spelke from NSF (BNS-8613390) and NIH (HD23103) and bya grant to Arlette Streri fVom INSERM (859012). We thank Sylvle Michet, Isabelle Dauphin, andLiesette Brunson for assistance with the experiments. Address correspondence to Arlette Streri,Laboratoire de Psyehologie du D^veloppement et de I'Edueation de l'Enfant, 46 rue St. Jacques,75005 Paris, France, or to Elizabeth Spelke, Cornell University, Department of Psychology, UrisHall, Ithaca, NY 14853-7601.Following Gibson (1962), the present paper will use "haptic" to refer to all perception that isderived from aetive touch, regardless of the sensory pathways through which the perception arises.[Child Development, 1989. 60, 1111-1125. © 1989 by the Society for Research in Child Development IncAll rights reserved. 0009-392(1/89/6005-0017$01.00]1112 Child Development - ^teming and a simple and regular form (Kell-man & Spelke, 1983; Schmidt & Spelke,1988; Schmidt, Spelke, & LaMorte, 1986;Schwartz, 1982; Termine, Hrynick, Kesten-baum, Gleitman, & Spelke, 1987). Infants evi-deudy perceive objects by analyzing the mo-tions of surfaces so as to form units that moveas wholes, but not by analyzing the colors,textures, and shapes of surfaces so as to formunits of maximal simplicity and regularity.To our knowledge, only one series ofstudies has focused on infants' haptic percep-tion of object unity (Streri & Spelke, 1988).Since it provides the basis for the present ex-periments, it will be described in some detail.Streri and Spelke (1988) presented 4-month-old infants with two rings, one in each hand,under a cover that blocked the infants' view ofthe rings and of their own bodies. In one con-dition, the rings could only be moved to-gether rigidly. In the other condition, therings could be pushed together, pulled apart,and displaced independently.A series of experiments, using a haptichabituation and visual transfer method (Streri,1987; Streri & Pecheux, 1986a, 1986b), inves-tigated infants' exploration and perception ofthe ring displays. Infants were found to ex-plore the rings actively by displacing them.Whereas the rigidly connected rings couldonly be moved together, the infants tended tomove the independently movable rings in op-posite directions. Infants also were found todiscriminate these motion patterns and totransfer the motion discrimination from touchto vision. After habituating to the rigidly mov-able rings, infants looked longer at a visualdisplay of the same rings moving indepen-dently than at a visual display of the ringsmoving rigidly; after habituating to the inde-pendently movable rings, infants showed thereverse looking preference. Habituation gen-eralized, to some extent, from the hapticmode to the visual mode. , , .,Most important, these experiments pro-vided evidence that the felt motion patternsof the rings influenced infants' perception ofobject unity and boundaries. After habituat-ing to the rigidly movable rings, infantslooked longer at a visual display in which thetwo rings appeared as distinct objects sepa-rated by a gap than at a visual display inwhich the two rings constituted the ends