Tag Archives: animacy

Causality and Animacy

Introduction to Causality

Functional relations are fast, automatic, subconscious and driven by a stimulus. They allow us to perceive properties of simple displays that in themselves objectively do not exist (Scholl and Tremoulet, 2000). In other words, we perceive the objects but the relationships between them are interpreted without a physical basis. Types of relationships that exist due to functional relations include animacy and causality. Both relations were complex and require higher-level cognitive processing, but the stimuli involves are basic. Because these very basic stimuli induce high-level perception, it suggests that our visual system not only determines physical structure but also the properties of the stimulus (ibid).

Michotte and “Illusions of Causality”

A type of functional relation called kinetic depth effect produces causality and animacy. As the name suggests, movement of stimuli at specific intervals causes the perception of a relationship. Michotte conducted one of the best and earliest demonstrations of perceptual causality in 1946 including the launching effect, the entraining effect, the launching effect with a temporal gap, the triggering effect, the launching effect with a spatial gap and the tool effect. His strategy was to show observers various ways that simple stimuli could interact to influence perception of causality. He called it the “illusion of causality” and described it in terms of A causing B’s motion or B’s motion as a continuation of A’s motion. Animacy and causality, Michotte believed, have these qualities because simple motion cues are the foundation for social perception in general (gestures, facial expressions, speech, etc.). Babies rely on these simple motion cues for survival, which explains why it’s a necessary skill from birth. Despite critics’ disapproval of Michotte testing these perceptions on himself and his colleges, these percepts have been proved culturally universal (Morris and Peng, 1994; Rime et al. 1985).

In 1963, Michotte outlined factors that influence our perception of causality, which are vital for the effect to occur: priority, timing, proportionality and exclusivity. If any of these factors are violated, the perception of causality can be broken. Timing, as determined below, has the greatest effect on causality. However, it is important to understand the significance of the other factors as well. First, priority: for A to move B, A must move first; second, proportionality: the action of B must be proportional to the movement of A, this includes consistent velocity, trajectory, etc.; thirdly, exclusivity, for A to cause B’s movement, A must be the only possible factor to have done so.

To further the work on causality, White and Milne (1997) created another simple display based on Michotte’s work. They created another simple display where one stimulus appeared to pull another. Again, White and Milne found the phenomenon of causality to be salient, immediate and irresistible. Further research suggests that the phenomenon is so engrained that even as babies causality is perceived. Leslie and colleges (Leslie et al., 1982, 1984, 1987) researched this principle on six-month-old infants. In the infants were habituated with a short film based on Michotte’s original displays. After habituation, infants found a reversal of the film more interesting. Leslie (ibid.) argues that six month olds looked longer at the reversed displays because it involved in an additional change in the causal roles.

Heider and Simmel

In, Heider and Simmel (1944) created a film showing three geometric figures – a large triangle, a small triangle and a small circle moving around a rectangle with a small opening. Although static clips of film show very little information about the motion properties of the shapes, after viewing the animation observers were consistent when describing the shapes. Observers attributed personality traits and emotions to the geometric figures, regardless of the instructions they were given. Temporal contiguity and spatial proximity created relationships among the figures, such as the large triangle chasing the other two figures. Just as with Michotte’s experiment, several researchers replicated the study and confirmed that they are universal (Morris and Peng, 1994; Rime et al. 1985; Hashimoto, 1966). Berry and Springer (1993) found that even three and four year olds attributed personality traits (f.eg. desires and emotions) to the geometric shapes. 52, replaced the geometric shapes with other simplified objects; this had no effect. The effect of the Heider and Simmel movie was only significantly reduced when the movie was temporally tampered with.

Stewart et al. (1982) proposed that we perceive animacy because when we see movement without a source, we subconsciously attribute the movement to a hidden energy source, in this case animacy.  This energy violations hypothesis delineates three conditions were animacy is assumed:

1. An object starting from rest

2. Direct movement towards a goal

3. Sharp changes in direction to avoid obstacles 

The Importance of Continuity and Contingency 

Temporal tampering has such a significant effect because perceiving animacy depends on the interaction between target and goals and the impression of intentionality in the movement (Dittrich and Lea, 1994). Dittrich and Lea also found that participants reported more intentional perceptions when the trajectory and the speed of the target object was greater than that of the distractors.  Wasserman and Neunaber (1986) also argue that people are sensitive not just to the pairings of cause and effect but also to the contingency or temporal correlation between them (relative continguity). Causal judgments are reduced when the action postpones the occurrence of the outcome; relative contiguity and thus causality will decrease as the delay between action and outcome increases.

Another approach assumes that contingency and contiguity effects are mediated by independent mechanisms (Shanks and Dickinson, 1987). Causality judgments based on contiguity-sensitive mechanisms the result of associative conditioning. The occurrence of an outcome increases the associative strength of stimuli and actions. Their approach is consistent with the dP theory which argues that judgments of contiguity are based upon the differences in the perceived probabilities of conjunction and disjunction of the outcome with the action. A study carried out by Williams (1976) offers support for the dP theory. Wililams’ studied was carried on pigeons, with the dependent variable of performance of pigeons on simple variable-interval schedules with unsignalled delays of reinforcement. He found that the imposition of programmed delay of between 3-5 seconds produced reductions in response time.