Journal of Experimental Psychology: Human Perception and Performance 1976, Vol. 2, No. 1, 130-138

Irrelevance of Figural Identity for Resolving Ambiguities in Apparent Motion David Navon University of California, San Diego In order to examine the degree to which form perception affects the formation of apparent-motion experience, subjects were presented with nine ambiguous apparent-motion situations, where the elements of each single flash were various figures. One of the interpretations of each situation preserved the figural identity of the figures across flashes. It was found that figural identity does not have any effect on determining the type of motion experienced, except when the figural analysis involved is relatively simple.

In the vast literature about apparent motion, only little attention is given to the effect of figural identity on perceived motion. The question is whether a subject may experience a sensation of motion when presented with a flash of a certain figure followed by another flash of a different figure in a different location. Early studies (mentioned in Kolers, 1972, p. 43) show that figural identity is by no means a necessary condition for perceiving motion. However, Orlansky (1940) concluded from the results of his experiments that although disparity of shape does not rule out illusory motion, it may make it less compelling and narrow the range of its appearance. The notion, implied by Orlansky's conclusions, that sensation of motion is determined to some degree by the visual similarity between the figures in the alternating flashes, was strongly opposed by Kolers (1972). On the basis of the results of Kolers and Pomerantz (1971) and Kolers (1972, pp. 51-53), he argued that disparity in shape cannot be expected to affect illusory motion, because figural analysis is completed only after the impression of motion is either created or not, so that the perceived motion can inter-

vene in the figural analysis but not vice versa. In Kolers' words, "the visual system responds to locations of stimulation and infers or creates changes of figure to resolve that disparity" (Kolers, 1972, p. 57). Most of the evidence that Kolers relies on are demonstrations that "given the choice of retaining a figure but moving it to a new location, or retaining a location and building a new figure within it, the visual system performs the latter task" (Kolers, 1972, p. 54). For example, when he showed (Figure 46, Array 5) in succession the stimuli displayed here in A and B of Figure 1, no motion was perceived; rather, the extreme figures changed shape and the innermost ones blinked on and off. Kolers' demonstrations indicate that the visual system exhibits location inertia rather than preservation of figural identity. To complete this line of argumentation, one should show that preservation of identity does not affect the perception of motion even when it does not conflict with maintenance of locations. For this purpose I designed several situations of ambiguous apparent motion, each with several interpretations. Each interpretation involves the same degree of change of locations. One interpretation of each This project was supported in part by Grant situation satisfies the principle of preservaNS07454 from the National Institutes of Health. tion of figural identity. The others do not. I thank Donald Norman very much for his careful If figural identity has any effect at all on review and his useful comments. Requests for reprints should be sent to David the formation of motion impression, then the interpretation in which figural identity is Navon, who is now at the Department of Psycholpreserved should be the one most likely to ogy, University of Haifa, Haifa 31999, Israel. 130

FIGURAL IDENTITY IN APPARENT MOTION

o B FIGURE 1. The displays alternated by Kolers (adapted from Kolers, 1972, Figure 46, Array 5).

be perceived. A failure of figural identity to bias perception would be another indication of its irrelevance in determining perceived motion. METHOD Each different situation of apparent motion can be called an episode. An episode is characterized by the appearance of a cycle of two or six consecutive different displays, and it consists of a number of iterations of this cycle. Each display was presented for ISO msec with a 30-msec time interval between any two consecutive displays as well as between any two consecutive cycles. Each of the individual displays will be called an array. An array consisted of two or six elements taken out of either of the sets presented in Figure 2. Each set of elements is called a vocabulary. The vocabularies differ in the degree of difficulty for the subject to identify (or, more generally, to perform figural analysis on) their members. Presumably the simplest are the line segments and the most difficult for an American student are the Hebrew letters.

131

The arrays in one cycle of each of the nine episodes used in this experiment and possible interpretations for the episodes are shown in Figure 3. (The elements shown in the figure are taken from the English-letters vocabulary.) The names of the episodes follow the type of motion that would have been perceived had the identity of elements been preserved across flashes. The "crossing slides" episode. The identity-preserving interpretation of this episode is that of crossed diagonal motion. The other likely interpretation is that of simultaneous vertical movement in both sides of the display. It was already found (Kolers, 1972, pp. 76-79) that subjects do not see simultaneous motion in two diagonal crossing paths, even when the intepretation of crossed motion preserves the identity of the elements while the alternative one does not. This episode is meant to replicate Kolers' finding with the vocabularies used in this experiment. The "seesaw" episodes. In these episodes the most likely interpretations are that of two simultaneous motions along one dimension in opposite directions and that of a seesaw oscillation about either the vertical or horizontal axis. The identity-preserving interpretations in the "vertical seesaw" episode are the ones which involve horizontal motion of the elements, whereas in the "horizontal seesaw" episode they are the ones which involve vertical motion of the elements. The "cha-cha-cha" episodes. The locations of the elements of the second array in these episodes are halfway between two adjacent locations of the elements of the first array on the circumference of the imaginary circle connecting the elements. The intercycle time interval was the same as the interarray time interval. Thus there are two possible interpretations: either that of a circular motion (of the configuration as a whole or of the six elements individually), which can be either clockwise or counterclockwise, or that of a 30° back-andforth tilting motion (again of the configuration as a whole or of the six elements individually). The "identical element" version of these episodes enables the testing of the prior bias toward each of these possible interpretations, because each of them would preserve the identity of the elements. The "one diamond cha-cha-cha" episode introduces a different aspect of disparity between elements,

vocabulary

figures

English Letters

E

I

M

T

L

H

0

Hebrew Letters

LU

1

3

1

n

LJ

0

Line Segments



1

\

0

X

/

X"

FIGURE 2. The elements used to construct arrays tabulated according to vocabularies.

DAVID NAVON

132

INTEPRETATION

NUMBER OF ARRAY

EPISODE

1OENTITY

NUMBER

1

1

NAME

[

CROSSING SLIDES

Z

3

2

4

5

6

PFIESERVING

- E

OTHERS

0 0

E Z

^

2

VERTICAL SEESAW

L

3

HORIZONTAL SEESAW w

L DIFFERENT I ELEMENTS

4

5

CHA-CHA-CHA

ONE DIAMOND

2

i

L

T

^

Z

M

T

M

H

E

Z

-

[ LM H

F

Z

' T

T

T

T

DIFFERENT L T ELEMENTS

M

Z

L

H

E

FIVE DIAMONDS

[o o o

ONE DIAMOND

[T

0"

0

J

-V T T - T

M

Z

0

E

M

T

2

z M

E

0 °0

M T

oO

T

- T , H -

-*

£^D

T

0 o E « E°-

L

o n

ZJ

L E

»

na

E E

H

*

-E

Z

IDENTICAL L T ELEMENT ~

AROUND THE CLOCK

Z E

z

E

L0

E T

H L

E.

-

Z

H

L

M

T

E

0 «E

0

0 00

o Oo

.z - M

E T

(A)

0 L -

0

E

Irrelevance of figural identity for resolving ambiguities in apparent motion.

In order to examine the degree to which form perception affects the formation of apparent-motion experience, subjects were presented with nine ambiguo...
744KB Sizes 0 Downloads 0 Views