This study focused on two questions: (1) is there a difference between social and mechanical non-social causality in infancy? and (2) do infants use information about contact between objects in inanimate, mechanical causal sequences more than they use it in situations where a social event is to be caused? Forty-eight 10- to 10.9-month-old infants and their mothers participated in the study. Infants were given the opportunity to produce three events selected to present a continuum from inanimate and mechanical to animate and social. The events were (a) the appearance of a picture in a window; (b) the movement of a mechanical dog; and (c) social responses from a female experimenter. The events were activated by operating a manipulandum (a knob). Two conditions were employed: 24 of the infants could trigger each event by moving the knob toward the event, thereby creating contact with it, while the other 24 could activate each event by moving the knob away from the event. Mothers demonstrated to their children how to use the knob. Infants were given a five minute period in which they could use the knob to trigger the events. Each session was videotaped from one camera located behind a one-way mirror and a second camera located in the room. After each toy session the mother and one other observer rated the infant's level of understanding of the cause-effect relationship between moving the knob and the occurrence of the event. An analysis of variance performed on the number of pushes of the knob toward and away from the events and on the two qualitative ratings of the infant's understanding revealed that infants most frequently used the knob to produce pictures in the window, and that they understood that event best. The results indicate that infants use different causal strategies in social and non-social situations. The role of contact between cause and effect is discussed. (Author/MP)