THE “LAUNCHING EFFECT” DEPENDS ON SIZE OF COLLIDING OBJECTS Michele Vicovaro, Luigi Burigana Department of General Psychology, University of Padua, Italy michele.vicovaro@studenti.unipd.it luigi.burigana@unipd.it Abstract Suppose that two squares are aligned horizontally on a computer screen. At a point in time one square (A) starts moving towards the other (B), which remains stationary. When A and B come into contact, the latter starts moving with the same velocity as A, while A comes to a stop. Albert Michotte (1946) demonstrated that, under these stimulus conditions, observers usually have the impression that A “launches” or “pushes” B, a perceptual effect he called the “Launching Effect”. Michotte claimed that features of objects A and B (e.g., their size) exert only a slight influence on the phenomenon, however to our knowledge no research has systematically tested this claim. In our experiment we manipulated the size of two simulated spheres, and found that the velocity ratio most favouring the perception of the Launching Effect actually depends on the size of both spheres. Many philosophers and psychologists agree that the concept of causality is the “cement” of the universe. As soon as one grasps the concept of causality, events in the world start to be conceived as relations between causes and effects. This makes an evolutionary advantage: cause-effect connections may be generalized in order to predict future events. The Belgian researcher Albert Michotte (1946) was a pioneer in demonstrating that causality can be directly perceived. He performed novel experiments showing how the perception of causality directly depends on stimulus conditions. In one of his experiments, he presented observers with two small squares aligned horizontally. At a moment one square (A) started moving towards the other (B), which remained stationary. When A and B came into contact, the latter started moving with the same velocity as A, while A came to a stop (see Figure 1 for a 3D version of Michotte’s stimuli). Observers reported the impression that A “launched” or “pushed” B, a perceptual effect called the “Launching Effect”. Michotte maintained that the Launching Effect is a purely perceptual phenomenon, that it may be explained in terms of Gestalt-theoretic principles, and is not influenced by observers’ experience with real mechanical collisions. Scholl and Tremoulet (2000) reformulated Michotte’s theorization in terms of perceptual module, thus emphasizing the “cognitive impenetrability” of the phenomenon. A compelling argument for this thesis is that the Launching Effect occurs even when A and B are fuzzy coloured shadows (Michotte’s Experiment 27), or when object A is a real wooden sphere, and object B is just a shadow (Michotte’s Experiment 28). This suggests that physical plausibility of collisions is not a necessary condition for the perception of the phenomenon, thus discrediting the possible role of past experience with real collisions. Recent experimental findings have further supported this claim: given appropriate contextual stimuli, the Launching Effect may occur even when spatiotemporal relations between A and B should exclude real collisions (Choi & Scholl, 2004; Bae & Flombaum, 2010). In addition, Michotte (1946, p. 78) reported the qualitative observation that shape and dimension of A and B may influence the “vividness” of the Launching Effect, but classified this influence as slight and marginal. He did not elaborate this qualitative observation in detail. In partial contrast with Michotte’s view of “purely perceptual” character of the effect we are discussing, the possible influence of objects’ shape and dimension on its perception