Perception of causality

Michotte (1946) carried out a series of studies on perceived causality. In some studies, observers watched as one square moved towards a second square, with the first square stopping and the second square moving off at a slower rate than the first one as they came into contact. According to Michotte (1946), observers perceived that the first square had caused the motion of the second square (the "launching effect"). The perception of causality disappeared when there was a time interval between the contact and the second square moving off, or if the second square moved off in a different direction to that of the first square. Another effect was termed "entraining". This occurred when one object moved towards a second object, and then the two objects moved off together at the same speed until they stopped together. It seemed to the observers as if the first object were carrying the second one or pushing it. The launching and entraining effects did not seem to be affected by the nature of the objects involved. In addition, the effects were observed even when the two objects were very different from each other.

It has proved hard to replicate Michotte's (1946) findings, perhaps because he often relied on rather small numbers of highly practised participants. Beasley (1968) found that only 65% of participants reported the impression of causality with the launching display, and that figure fell to 45% for the entraining effect. In strong contrast to Michotte's findings, Beasley (1968) found that 45% of participants reported causal impressions when the second object moved off at a 90° angle to the direction of motion of the first object. Finally, Beasley (1968) found that the perception of causality was influenced by the nature of the objects used, which is directly contrary to Michotte's findings.

Theoretical accounts

Michotte (1946) put forward a Gestaltist view of perceived causality, according to which it occurs naturally when specific motion sequences are seen. He argued that causality is perceived in a rather direct way which does not rely on inferences or other cognitive processes. In addition, Michotte claimed that the perception of causality is innately determined. However, if Michotte is correct, it is hard to understand why many people fail to show the predicted effects.

If the perception of causality is direct, then we might expect to find it even in infants. Leslie and Keeble (1987) obtained evidence that six-month-old infants could perceive the launching effect. This finding suggests that fairly basic processes are involved in the perception of causality. Oakes (1994) obtained similar findings from seven-month-olds using simple displays, but these infants failed to perceive causality in more complex displays.

Michotte's (1946) assumption that the perception of causality does not involve the use of inferences was tested by Schlottmann and Shanks (1992). They arranged matters so that a change of colour by the second object always predicted its movement, whereas impact of the first object on the second object was less predictive. The participants learned to draw the correct inference that the change of colour in the second object was necessary for its movement of the second object, but this did not influence their causal impressions. However, when the first object collided with the second object, which changed colour and moved off, the observers claimed that it looked as if the first object caused the second one to move.

What do these findings mean? Schlottmann and Shanks (1992, p. 340) concluded as follows: "The results support the distinction that Michotte advocated between causal knowledge that arises from inference and that which is directly given in perception." This conclusion was supported by finding that 85% of the participants regarded their inference judgements and their ratings of perceived causality as independent of each other.

Schlottmann and Anderson (1993) studied the launching effect. They manipulated the gap between the two objects, the time period between the collision and the second object moving, and the ratio of the speeds of the two objects. They identified two successive processes, which they termed "valuation" and "integration". Valuation involves assigning weights to the various aspects of the moving display, and there were substantial individual differences in this form of processing. Integration involves combining or integrating information from these various aspects, and there were great similarities in this process across participants. Schlottmann and Anderson (1993, p. 797) concluded as follows: "The averaging integration model may correspond to the invariant perceptual structure of phenomenal causality, as proposed by Michotte. The valuation operation, on the other hand, can accommodate individual differences that may have experiential components, as suggested by his critics."


Michotte was correct in assuming that causality can be perceived in a fairly direct way owing little to experience or to inferences. However, the perception of causality is more complex than he assumed. The existence of substantial individual differences in the perception of causality suggests that learning and experience play a greater role than was admitted by Michotte. It is hard to disagree with the conclusion of Schlottmann and Anderson (1993, p. 799):

In adult cognition... the perceptual illusion of phenomenal causality must function together with acquired knowledge about causality in the physical world. Thus, ways are needed that can make effective progress on the innate-plus-learned question.

Stop Anxiety Attacks

Stop Anxiety Attacks

Here's How You Could End Anxiety and Panic Attacks For Good Prevent Anxiety in Your Golden Years Without Harmful Prescription Drugs. If You Give Me 15 minutes, I Will Show You a Breakthrough That Will Change The Way You Think About Anxiety and Panic Attacks Forever! If you are still suffering because your doctor can't help you, here's some great news...!

Get My Free Ebook

Post a comment