1. Introduction

If an observer views an object undergoing implied or apparent motion, and the object suddenly disappears, then remembering the objects’ final position will be shifted forward in the direction of motion evolution (e.g., Freyd, 1987; Freyd and Finke, 1984; Finke, Freyd and Shyi, 1986). It has been concluded that this memory distortion results from dynamic visual representations that reflect an internalization of the inertial properties of real-world object motion (Freyd, 1987; 1992; Hubbard, 1995).

This conclusion, however, stands in a striking conflict with "naive" physics research, according to which many people hold erroneous beliefs about motion similar to a medieval impetus theory (e.g., Clement, 1982; Clement, 1983; McCloskey, 1983; McCloskey, Caramazza, and Green, 1980; McCloskey, Washburn and Felch, 1983). What led representational momentum and "naive" physics researchers to such opposite conclusions? And, what kind of knowledge is, in fact, incorporated by our perceptual and cognitive systems?

Most naive physics research examines subjects’ beliefs about motion that are verbalizable and consciously accessible, and may be considered to be explicit. Representational momentum studies examine subjects’ automatic responses on a memory task, which is accomplished by unconscious processes and based on implicit knowledge. The hypothesis of the proposed research is that our mental representations regarding objects’ motion are composed of (a) implicit perceptually-based knowledge about objects motion that is not consistent with physics principles, but rather reflects impetus ideas, and (b) explicit conceptual knowledge about motion that primarily depends on the person’s physics background. In addition, we suggest that there is a dissociation between observers’ perceptually based knowledge about objects’ motion and their conceptual knowledge of physical principles.

2. Research Design

To investigate the above hypothesis, we will consider the case of ascending and descending motion. Table 1 shows predictions of Newtonian physics and impetus theory predictions for the case of ascending and descending motions. Will representational momentum for ascending targets be consistent with impetus or physical principles?

2.1. Experiment 1

The goal of Experiment 1 was to investigate whether people, in fact, exhibit greater representational momentum for more massive ascending targets than for less massive ones, as follows from Newtonian principles.

Stimuli and Apparatus

The stimuli were displayed on Apple Macintosh with a viewing distance of approximately 60cm (monitorinch). The target stimuli were filled black squares of two sizes, 0.4 and 0.8 inches, presented on a white background. (The assumption that visually larger physical objects are perceived to be more massive than visually smaller objects is based on research on size-weight illusion of Jones, 1987). In each trial, a vertically moving square was displayed on the computer screen by a sequence of six successive frames with a 150-ms stimulus duration and 150-ms interstimulus interval. During some trials the square moved downward, during the others it moved upward. A target accelerated while descending and decelerated while ascending. Target acceleration or deceleration was displaying by varying the distance between successive presentations of the target. The final frame in the sequence was followed immediately by a test frame (seventh frame). The square depicted in the test frame was either in the same position as the final square, or shifted slightly backward in the direction of motion, or shifted slightly forward in the direction of motion.

The participants were 9 undergraduate psychology students recruited from the Psychology Subject Pool at the University of California, Santa Barbara. They were instructed to compare the square’s final position to the test square’s position and indicate whether the two positions were the same. Each participant received 240 trials.

Results and Discussion

The representational momentum (RM) effect was determined by subtracting the proportion of errors for the 5-pixel displacement of the test frame inconsistent with the implied motion from the proportion of errors for the corresponding consistent displacement (similar to the method of Reed and Vinson, 1996). The RM effect was calculated for each participant and the mean RN effect over all subjects was 0.34 for small ascending targets and 0.01 for large targets. As predicted, a paired-sample t-test revealed a significant effect of target size: t=2.48, p<0.03, df =8. The small ascending target produced more RM than the large ascending target. Figure 1(a) illustrates the RM effect for ascending targets of different sizes. For descending motion, a significant effect of target size was also found t = 3.056, p < 0.02, df =8, with the larger targets producing more representational momentum (RM = 0.21) than the smaller ones (RM = 0.10). (The smaller representational momentum effect for descending motion might be explained by a lower average velocity for descending than for ascending motion). This pattern is entirely consistent with impetus ideas and incompatible with physical principles.

2.2 Experiment 2

The goal of Experiment 3 was to examine subjects’ explicit conceptual knowledge about objects’ motion under the influence of gravity. The participants were 93 undergraduate psychology students who had not received any physics instruction at the college level. The subjects were presented with a questionnaire aimed to assess their conceptions about ascending and descending motion. The analysis of students’ questionnaires showed that a large proportion of students indicated correctly that, in the absence of friction, all physical objects, regardless of mass, fall with the same acceleration. About 50% of subjects indicated that for ascending motion, even in the absence of friction, a more massive object decelerates faster. Ninety-nine percent of all subjects believed that, in the presence of friction, a more massive object decelerates faster.

2.3 Experiment 3

The goal of Experiment 2 was to compare the representational momentum for ascending targets of different sizes between expert and novice physicists.

Stimuli and Apparatus

The stimuli and apparatus used in Experiment 2 were similar to those used in Experiment 1, except that only ascending motion was tested in this experiment. The participants were 18 undergraduate psychology students recruited from the Psychology Subject Pool at the University of California, Santa Barbara. Nine of these students, who took extensive mechanics courses at college level were considered as physics experts. The other nine students were novices and had not taken any physics courses. All experts explicitly stated that the more massive target would accelerate faster while rising; all novices believed that the smaller target would accelerate faster.

Results and Discussion

No significant difference was found between the representational momentum of the experts and novices: F(1, 15) = 0.078, p = 0.783. The larger ascending target produced significantly less RM effect than the smaller ascending target F(1, 15) = 7.30, p < 0.016. The above results provide evidence that knowledge of physics laws about objects’ motion does not influence the representational momentum effect.

3. General Discussion

The results of the above experiments support the idea that our perceptually based knowledge about motion is more similar to impetus than to physical principles. After receiving formal physics instructions, we begin to revise our impetus principles and change them to physical ones. However, this revision takes place only on a conceptual level. Knowledge of physics laws, however, does not change our implicit knowledge about motion, which is based on our everyday experience and consistent with impetus predictions. This explains both why students’ misconceptions about motion are so "resistant" and "not easily changed by classroom instructions". It also explains the painful evolution of the concepts of motion from the medieval ages to Galileo in order to overcome the powerful impetus misconceptions.

References