Koffka (1955: 381) suggests that causality is one of the relationships encountered in the phenomenal world that can be studied if the reductionist stance on perception is rejected, because it is unlikely that the perceived causality corresponds to a single stimulation or sensation. Earlier Stumpf (1930) had claimed that assuming that the fundamental concepts, like causality, are a priori categories is cutting corners, because it is necessary to find the phenomena that provide their perceptual foundation. He claimed that Hume’s analysis was not sufficiently deep to recognize the cases where a real inherent connection between two objects is experienced, rather than only a succession thereof. It is possible that Koffka capitalizes on Stumpf’s claim. Yet he goes farther by claiming that this question could be decided experimentally. He argues that even if the long-established model of causality as the contact between two billiard balls is admitted, it must be possible to design some specifiable conditions to test whether or not the movement of the second ball is differently perceived given the contact with the first moving ball. Moreover, if the constancy hypothesis is rejected, it is conceivable that there could be a distribution in the perceptual space-time of the stimulation that acts as external constraint on the rules of perceived causality.
Albeit quite independently, Michotte puts Stumpf’s claim and Koffka’s hypothesis to experimental test. Michotte (1941: 290, 1952: 228) claims that even though many philosophers and psychologists have claimed that causes and actions are not perceivable, cutting the bread with a knife or pushing a weight with force and fatigue are properly perceived facts of experience. Not all facts of this sort provide a model to study the perception of causality. Yet suitable systems of stimulation can be constructed to find the features and structures of the perceived causality. The experimental construct consists of two rectangles A and B at rest or in motion, whose speed, direction and displacement are varied through different kinds of devices such as spirals drawn on rotating discs that are seen through a slit as rectangles in motion or spots of geometric forms projected on a screen.
In a first system of stimulation, A is shown to begin to move at the constant speed of 30 cm/sec., B to be at rest at a distance of 4 cm. Then A reaches and touches B so that they both continue moving along the same direction and at the same speed while remaining in contact with each other. In a second system of stimulation, A and B are displayed in the same initial state as in the first system, then A reaches B but soon after the contact A stops and B begins to move at a slower speed than A (6-10 cm/sec.) until it stops a little farther along. From subjects’ descriptions a “phenomenal scission” of B movement emerges as a feature of the perceptual responses for both patterns. B appears to move but this movement is also a displacement, since it appears more similar in the first case to a change of place, by which B merely occupies successive spatial positions, and in the second to an extension of A’s movement, namely as a movement in reality due to A and not to B. Michotte emphasizes the phenomenal nature of this distinction between movement and displacement for a physical motion. Actually, subjects’ descriptions differ from those that would have been satisfied if
- the movement of A ended at contact with B;
- from the outset of the pattern A and B were in contact and either
- they began to move together or
- only B moved while A remained still in its place.
For example, subjects agree in reporting that “B is in motion, but A has the movement” or that “B partakes of the movement of A.” This is a perceptual difference that brings about a “duplication” of the movement of B: B is perceived at the same time as moving and as being displaced.
According to Michotte, this is the fundamental feature of the phenomenal structures that arise given the competing perceptual properties with which A and B are endowed for the conflicting systems of stimulation. In the first case, the phenomenal structure of “entraining” emerges in the particular form of the “push.” One perceives the unitary movement of A and B but this is not such that A and B are seen as a whole new object (A + B) in motion, like a unitary grouping of A and B or a rectangle with A and B as parts. The unitary movement appears only as a functional unit, because it regards A and B, but the movement still appears to belong to A since at the contact B appeared to be at rest. From the contact onward, the system of stimulation is such that one sees a longer rectangle (A + B), but nonetheless A and B continue to appear as two adjacent rectangles for the earlier perception of their distinct contours. The perception of A and B as two distinct objects preserves their opposed kinetic properties that compete with the functional unit of the unitary movement. The phenomenal scission settles this perceptual conflict because A appears to move while B appears more properly to be displaced by A. As a consequence, A appears to push B. In the second case, the phenomenal structure of “launching” emerges. One sees a unitary movement that preserves its direction from the beginning to its end, but this unity contrasts with the fact that the bearer of movement is seen to change from A to B. From the contact onward, the system of stimuli is such that one sees A and B as distinct objects that yet have new kinetic properties in contrast with the perception of the unitary movement: A becomes immobile and B is set in motion. The phenomenal scission settles this perceptual conflict because the movement of A appears to be continued through the displacement of B: A appears to launch B.
Once the distinguishing feature of these phenomenal structures has been found, Michotte investigates the factors that account for the appearance, the change and the disappearance of these forms of perceptual causality. In both cases the systems of stimulation change so little from the first to the later phases that they preserve the former perceptual properties of A and B, but they also change enough to bring about new perceptual properties. The compromise between former and new competing properties is realized in the phenomenal structure whose constitution is due to what Michotte calls “ampliation of movement.” For both patterns, this notion designates the phenomenal extension of the movement from A to B, which accounts for the fact that A and B appear integrated in a new structure despite their distinguishing properties. In the push pattern, the extension of movement allows the movement and the displacement to appear as one and the same after the contact between A and B. In the launch pattern, it allows the movement to appear to belong to A, even if it stops at the contact with B, and B to preserve the identity of its earlier inertial state in spite of its movement. The ampliation of movement is attributed to two well-known grouping factors (Wertheimer, 1923: 316, 320). The factor that accounts for the ampliation of movement in the structure of push is the “common fate,” for which perceptual elements that move simultaneously at the same speed with the least change of direction are grouped in a unitary object. The factor that accounts for the ampliation of movement in the structure of launch is the “good continuation,” for which collinear perceptual elements are grouped in a unitary object. These factors underlie the integration of A and
B in the structure and allow the perceived causality to tradeoff the competing perceptual properties of A and B.
It is interesting that a deeper analysis of the phenomenal structure provides additional evidence of the independence of perception. Small changes in the speed of movement of A and B and their time interval have effects on the phenomenal structures that may be modified or disappear. As regards the launch, for example, if B begins to move, after being hit by A, even slightly faster than A, the launching is replaced by the appearance of the “release effect” (declenche- ment): B appears to be brought into motion by an internal mechanism, acting like a spring, that is triggered by the impact with A. If B begins to move after a certain time has elapsed after the contact, two independent movements are perceived. If B continues its movement beyond a measurable definite spatial extent, the phenomenal scission no longer holds, so that the movement loses the character of displacement and B’s movement becomes as autonomous as A’s. On the grounds of this kind of observation, Michotte argues that were the perception of causality dependent upon inferential or other psychic functions, there would hardly be a reason for the fact that it changes into another definite phenomenal structure for only a small change in the speed of B, the duration of the contact or the extent of the movement of B.
Source: Calì Carmelo (2017), Phenomenology of Perception: Theories and Experimental Evidence, Brill.