347 INFANTS’ PERCEPTION OF CAUSAL CHAINS Leslie B. Cohen, Leslie J. Rundell, Cara H. Cashon, and Barbara A. Spellman Department of Psychology, University of Texas, Austin, TX 78712 Previous research indicates that infants not only perceive causal relationships between objects (e.g., Oakes & Cohen, 1990; Oakes, 1994) but also perceive the role of agency in simple launching events (Leslie & Keeble, 1987). In fact, Cohen and Oakes (1993) found that for infants 10 to 12 months of age, agents but not patients are afforded special status. Our next logical step was to examine infants’ perception of a more complex causal event, a causal chain in which multiple events follow one after another each seemingly caused by the preceding event and resulting in a final outcome. Ten and 15month-old infants were habituated to a causal chain event, in which one toy rolled across the screen and made contact with a second toy that then rolled into a house “causing” a puppy to pop up from the roof. Jn the direct condition, the first toy made spatial contact with the second toy which immediately began to move. According to adult ratings, the frost toy would be the causative agent of the outcome; and the second toy would be an instrument. In the delay condition, there was a 2 second pause from the time the first toy made contact until the second toy began to move. According to adult ratings, the two rolling toys would not be not causally related, and the second toy would be the causative agent of the outcome. Following habituation infants were tested on events similar to the habituation event (direct or delay) but with a novel toy replacing one of the familiar toys. If infants were perceiving these events in the same way as adults, infants in the direct condition should dishabituate to the test event in which the first toy had been replaced by a novel toy, since infants seem to attend more to agents and their actions than patients (Cohen & Oakes, 1993). Using the same rationale, it was also predicted that infants in the delay condition should dishabituate to the test event in which the second toy had been replaced by a novel toy. Results with 15month-old infants confiied these predictions. Infants in the direct condition dishabituated only to a replacement of the first toy in the direct condition and only to a replacement of the second toy in the delay condition. Results with lo-month-old infants were quite different. Jn both direct and delay conditions, they only dishabituated to a replacement of the first toy. One possible explanation for the lo-month-olds is that they perceive only that the first object is an agent and do not make a link to the cause of the final outcome. Another possibility is that the events are too complicated for the lO-month-olds to process as a whole, so they only process the initial portion, the action of the first object. Whichever explanation is correct, these results clearly indicate a developmental shift in causal perception between 10 and 15 months of age, with 15month-olds responding much more like adults.