SOME PARAMETERS OF VISUAL APPARENT
ABIENT MOVEMENT

Thesis gov the Degree of DE. D.
MECHZGAN STATE UNIVERSETY
Melvin Levitt
1957

THEIIB

This is to certify that the

thesis entitled

Some Parameters - in
Visual Apparent Ambient Movement

presented by

Melvin Levitt

has been accepted towards fulfillment
of the requirements for

_Eh._D.-.degree in Psychology

W -

Major professor \

Date December 5, 1957.

 

0-169

   
     

  

LIBRARY

Michigan State
University

 

SOME PARAMETERS OF VISUAL APPARENT ABIENT MOVEMENT

By

MELVIN LEVITT

A THE SIS

Submitted to the School for Advanced Graduate Studies of
Michigan State University of Agriculture and Applied
Science in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1957

ABSTR AC T

This investigation was conducted with the purpose of explor-
ing the parameters of stimulation necessary for the perception of
optimal abient movement. Two simple luminous targets were exposed
in succession. each being located at the same optical distance and
with their centers on the principal visual axis. The first target was
of constant area and intensity; for different exposure sequences the
second was made to vary in area or intensity. The temporal rela-
tion between termination of the first exposure and initiation of the
second exposure was serially varied, thereby producing varying de-
grees of unstimulated interval between exposures and varying degrees
of temporal overlap of exposures. Monocular viewing was employed.
and stimulation was localized in the fovea.

Threshold measures for the magnitudes of both the unstimu-
lated interval between exposures and the interval of temporal over-
lap indicate that there is a range of relative differences in both
stimulus area and intensity within which movement is perceptible.
Within these ranges, the absolute values of the threshold measures
of the temporal relation between exposures were found to lawfully
increase as direct functions of both relative difference in area and

ii

intensity. An important provision was the exposure of both targets
above minimal durations.

Collateral observations revealed an orderly variety of related
phenomena. These were: continuity of light (fusion), discontinuity of
light (flicker), intrusion of the smaller second square within the
first. contour obliteration resulting in either of two forms of pure
movement. and the total obliteration of the effect of the first stimu-
lus. Each of these phenomena was semiquantitatively discussed in
relation to the parameters of stimulation.

The interpretation of all phenomena reported is made in
terms of a tentative, primitive conceptual scheme, which not only
embraces these phenomena, but applies in part to aspects of flicker
fusion and contrast phenomena. It is believed that this scheme is

capable of generating testable predictions.

iii

ACKNOWLEDGMENT

The encouragement and guidance of Professor S. Howard
Bartley have significantly contributed to this work. I wish to ex-

press gratitude for this support.

iv

TABLE OF CONTENTS

Page

RE SUI—l TS t o o e o o O a oooooooo o e o e oooooooooooooo e 3 4
DISC IISSION e e e e e e ooooooooooooooooooo e e e e e e e o o 47
SUMMARY 0 o ooooo e e e e e e e e e e ............. e e e e e 5 9

BIBLIOGRAPHY ...... ................. ......... 61

INTRODUC TION

The perception of motion has been a topic of no mean sig—
nificance in the history of psychology. As a psychological event in
need of explanation. it comprised the spearhead of the attack by the
Gestalt school on nineteenth century content psychology. It was
then that Wertheimer (31) convincingly argued that phenomenal move-
ment was a basic datum of perception; immediately given, nonana-
lyzable. and noninferential. Wertheimer's polemic was based
largely upon the visual perception of apparent movement in which
the stimulation was spatially and temporally discontinuous. and the
special case of objectless movement. Prior to the demonstration of
stroboscopic movement, the psychology of motion perception consti-
tuted no source of puzzlement; when there is a relative displacement
between an object and the body of the perceiver, there is a corres-
ponding displacement of the pattern of stimulation on the receptor
surface. Continuity in the space-time field of the surrounds has its
correlated continuity in sense organ excitation; hence, the perception
of moving objects. With the demonstration of stroboscopic move-
ment in the nineteenth century, and the almost simultaneous demon-

strations of negative aftersensations of movement. the topic of motion

perception assumed an enigmatic status. These demonstrations
achieved the ignoble distinction of being classified with numerous
other phenomena under the rubric of "illusions"; the designation
referring to the lack of correspondence between perception and the
surrounds. Traditionally. as with all of the so-called illusions. the
perception of movement of the stroboscopic and aftersensation vari-
eties was rationalized in terms of cognitive or oculo-motor proc-
esses. Numerous researches were conducted. especially at the
beginning of this century. in the attempt to elucidate the matter.
Wertheimer's publication emerged as the most outstanding among
those of others which marshalled against the interpretations of seen
movement in terms of cognitive or oculo~motor processes. Histor-
ical reviews of the research and theoretical issues involved are
presented by Boring (6) and Neff (22). The latter has especially
well elucidated the early issues; hence. no. attempt at review will be
presented here.

Once it was established that movement is immediate in per-
ception, the task became that of describing it. specifying the condi-
tions necessary for its phenomenal occurrence. and establishing the
underlying physiological correlates. A taxonomy of movement is
hardly recognized in the literature; although much research has

been performed on the subject of movement. a systematic effort is

sorely lacking; and, unfortunately. little is known of the physiology
of perceived movement.

Let it be recognized at the outset that the proper study of
perception consists first of all in the description and organization of
the facts of perception. This is not intended to mean that experi-
mentation and theorization are to await a complete phenomenology.
but rather that the object of study should be conceptualized as
clearly as is currently possible. The facts of perception are
phenomenal. and as the objects of inquiry, they should not be con-
founded with abstractive constructions. Secondly. let it be attested
that the task which is set in the study of perception consists in dis-
covering and lawfully relating those conditions--energistic, physio-
logical. and psychologica1--which are causal for particular per—
ceptions. Emphasis is particularly directed against the frequently
overlooked logical fallacy which is committed by attempts to explain
one perceptual event in terms of another (and the literature is re-
plete with examples of this error). Consistent with these tenets. a
tentative conceptual scheme will be proposed for the study of move-
ment. Included within this format will be a taxonomy of phenomenal
movement. Conceptual movement is not the immediate concern of
this discourse. As additional information becomes available. the

scheme will undoubtedly be modified.

I. Conceptual movement
11. Phenomenal movement
A. Taxonomy
1. Pure
a. Phi
b. Gamma
2. Object
a. Two dimensional
b. Rotary
c. Three dimensional
d. Perspective
e. Deformative
f. Disjunctive
3. Self
a. Two dimensional
b. Rotary
c. Three dimensional
d. Perspective
B. Programme
1. Verifiable

a. Consonant

2. Nonverifiable
a. Consonant
b. Dissonant
It is well to distinguish between conceptual movement and
phenomenal movement. As has already been asserted. the latter is
an immediate event of direct experience. On the other hand. con-
ceptual movement is an abstraction. constructed from a history of
phenomenal events. Accordingly. conceptual movement is not to be
confused with direct experience. An example of conceptual move-
ment which is pertinent to the study of sensory processes is the
linear transformation of a rigid distribution of radiant energy in the
space-time field. One might say that this is a physical construct of
movement. The taxonomy of movement will merit little discussion
here. Let it suffice to make two remarks. The identifications cor-
respond in part with the recent treatment by Gibson of the percep-
tion of movement in the visual world (11). In addition. inasmuch as
the perception of objects or the self is of concern. there will be
various combinations and multiples of these types of movement.
We can subsequently programme consideration of phenomenal
movement as consisting of two main modes. (Neglect should not be
made of the fact that schematization of the phenomenal is itself a

conceptual act; taxonomy and programme should be recognized as

such.) We may consider those instances of phenomenal movement
which are capable of verification, and those which are not. By
"verification" we shall mean established constancy of movement in
experience via various receptor orientations. This includes the
general gamut of movements which are present in the ordinary
perceptual world of the individual. The class of phenomenal move-
ment thus considered might generally be referred to as ”verifiable
movement." In contrast, we can consider the class of phenomenal
movements not subject to verification by varied receptor orienta-
tions. It should be obvious that herein are included the varieties of
apparent movement alluded to earlier. as well as other forms of
"subjective movement.” to be described below.

One more conceptual distinction of heuristic value might be
made. There is an abundance of evidence for the fact that phenome-
nal movement which is nonverifiable by varied adjustments of the
receptors can be a consequent of either of two stimulus conditions;
viz.. that in which the conceptual distribution of energy is continu-
ously transformed spatially and temporally on the receptor surface.
and that in which it is not. In the former case. we may simply say
that the stimulus moves or is kinetic; in the latter case we may
say it is stationary or static. Henceforth. phenomenal movement

that is a consequent of kinetic stimulation can be referred to as

”consonant movement." insofar as there is continuous flux in both;
and. accordingly. phenomenal movement which is a consequent of
static stimulation can be signified as "dissonant movement," insofar
as there is continuity only in the phenomenal. Itis to be noted that
consonant movement includes all perceived movement that is veri-
fiable via immediate experience as well as a class of those per-
ceived movements which are not verifiable.

If justification of this schematization of a programme for
phenomenal movement be needed, then it might be found in at least
two sources; viz.. the research programs which have been conducted
by various investigators and the theoretical implications arising
herein. Allport has contributed an extensive systematic treatment
of the significance of phenomenal verification in his general consider-
ation of perception (1. pp. 14-46).. Nonverifiable (traditionally labelled
nonveridical) perception illustrates the important fact that the
organism contributes to the sensory input. Historically. this. fact
was of especial theoretical import with regard to what is here called
dissonant movement. as discussed earlier. Although the theoretical
furor may have dwindled. experimentalists are still disposed to
treat dissonant movement as unique. Those investigators who are
Particularly attendant to the psychophysiolog of sensory processes

aSpire to reveal some of the characteristics of the mechanisms

underlying these events. Dissonant movement (and its attendant
phenomena) is useful for the study of the spatial and temporal
properties of the sensory apparatus. Furthermore. in his research
program dealing with the psychophysics of the visual world. Gibson
has recently (11) made an analysis of the stimuli for what is here
called consonant movement. However. Gibson is avowedly con-
cerned mainly with verifiable phenomena. In recent years. the
events which are here designated as nonverifiable consonant move-
ments have been highlighted. along with other comparable perceptual
events. by a relatively new theoretical trend in the psychology of
perception (16; 30). Here again. emphasis is made of the organ-
ism's constructions from the phenomenal. Actually. in the terms of
this discourse. it is largely the process and results of phenomenal
verification which is underscored by these investigators. Surely.
phenomenal movement which is nonverifiable by varied receptor ad-
justments has filled and still occupies a distinctive position in theo-
retical and experimental psychology.

Having made this somewhat tedious excursion in the interest
of systematization. we can now devote the remainder of this discus-
sion to the. findings of research. Particular attention will be devoted
to phenomenal movement dissonantly produced. for the main intent of

this discourse is the presentation of some experimental findings of

that sort. First. however. a brief resume of some of the previously
obtained information will be offered.

After Wertheimer's classical publication. the problem of ap-
parent movement (referring specifically to movement perceived sub-
sequent to the presentation of dissonant stimulation) was broadened
to include movement in the cutaneous modalities as well as the
visual domain. Interestingly. the interaction of the sense modalities
in the occurrence of phenomenal movement has also been demon-
strated (see Neff, 22), as well as a special case of intrasensory ef-
fects (see Gibson. 11). It is generally recognized that the perception
of movement subsequent to dissonant stimulation is governed by cer-
tain contributions by the organism to the stimulus imput; i.e.. atti-
tudinal and cognitive processes (11; 22) are variously cited as
determinants. The likelihood of movement being perceived is said
to be facilitated when the observer adopts a set to perceive it. and
when the targets employed produce phenomenal objects which are
mobile in the experiential context of the observer. Aside from the
demonstration of these facts. precise relationships have yet to be
determined.

Experimentation has yielded descriptions of several "types"
0f apparent movement. dissonantly produced. The types are dis-

tinguished on the basis of the approximate stimulus conditions for

10

their production. Some of these have been investigated more fre-
quently than others. and typically in the visual modality. As a fur-
ther concession to programme. we shall make two broad identifica-
tions here; we shall separately consider movement perceived
utilizing steady dissonant stimulation. and movement perceived as

a function of the onset and offset of discrete stimuli.

By steady dissonant stimulation. reference is made to an as-
sortment of movement experiences including: the familiar expansion
and contraction of Plateau's spiral; the negative aftereffects of this
and the oft-described waterfall illusion (6). whereby subsequently
inspected objects are seen to move in an opposite direction; the
seen movement induced relative to a concurrent figure in the same
sensory field; the erratic autokinetic movement of a visual object in
an undifferentiated sensory field; and the total movement of phenom-
enal space characteristic of the vertigo. A systematic treatment of
these examples of apparent movement is not available. It is worth
adding that in some cases of induced movement as well as in the
case of vertigo observer himself may have the experience of mo-
tion. Effects such as vertigo. autokinetic movement. and induced
relative movement are special instances of the influence of concom-

itaut stimulation on movement in phenomenal space. The negative

11

aftereffects of Plateau's spiral and the waterfall illusion exemplify
the operation of temporal processes in perceived movement.
Attention is now directed to apparent movement as a conse-
quent of the onset and offset of discrete stationary stimuli. Origi-
nally. objectless movement was of the greatest theoretical interest.
This has been described as motion without phenomenal reference to
that which is moving. Wertheimer called this pure phenomenal
movement. or ”phi movement." Descriptions of "phi" tend to be
somewhat weak in communicative value. and not very numerous in
count. The stimulus conditions are the appropriate spatial. temporal.
and intensive combinations of at least two discrete luminance distri-
butions. "Gamma movement" is analogous to phi inasmuch as it
may be described as the centrifugal or centripetal movement of
light (or darkness) occurring during abrupt brightness changes in the
visual modality. Some movement in depth is reported to sometimes
accompany the production of gamma movement (23). The indepen-
dence of gamma from a contoured figure (3. pp. 156-165) places it in
the class of pure movement. The stimulus for gamma is a rela-
tively abrupt temporal transition of luminance of an appropriate
magnitude. Gamma movement is the only one of these presently
under consideration in which the perceived movement is a consequent

0f a temporal transition of a single distribution of impinging energy.

12

Every other type is crucially dependent upon a spatial dimension be-
tween at least two stimulus distributions. "Part movement" is the
general case in which an object appears to jump partially toward a
successively appearing and spatially distinct object; the number of
objects may vary with the number of stimuli. A complex form of
part movement between the sense modalities of vision. audition. and
touch has been described (see 22. p. 28). A visual. an auditory. and
a tactual object appear to move toward one another in various orders
dependent upon the order of modal stimulation. The type of dis-
sonantly produced apparent movement which has perhaps received
the greatest experimental attention is referred to as "optimal" or
"beta movement." This type of movement is equivalent with the
common experience of an object moving across the visual field.
Complications of beta have been described in which more than one
object are seen in motion. traveling in different directions and
crossing paths. A special case of beta has been demonstrated in
which an object is seen to move in a circular path. the diameter of
which alters with the phenomenal velocity of movement. The next
type. "delta movement." is phenomenally akin to beta insofar as an
Object is seen to move across the visual field. though decreasing in
brightness; however. an important distinction exists between the two

concerning the energistic antecedants. The distinctive stimulus

13

condition for delta is that the weaker of two spatially and intensively
different stimuli has temporal priority. The curved path of move-
ment taken by an object around the locus of a currently or pre-
viously perceived object has been called "bow movement." This too
is akin to beta; the distinction is in terms of the particular opera-
tions of stimulation necessary for induction of a curvilinear path.

In addition. there are the two general cases in which an object ap-
pears to divide into (or emit) other objects which move laterally
(one may remain in the original locus); and the converse phenomena
in which a plurality of objects move into a unitary merger. Of these.
only the case of "split movement" has been labelled. Then. we have
the type of dissonantly produced movement known as "alpha move-
ment." This designates the continuous change in size of'a phenome-
nally moving object under the unique condition in which the discrete
stimulus distributions are of equal size. The targets in this case
are of the complex type known as size "illusions." The Muller-
Lyer figure is frequently employed. The dissonant production of
movement in the third dimension is least well knowm This is
achieved primarily by the successive exposure of targets differing
in area. but not necessarily at different viewing distances. Under
appropriate conditions. one may see in depth: part movement. pure

Phi movement. and optimal movement (7). The path of movement

14

may be unidirectional or to and fro. depending upon the stimulating
conditions.

The main point of interest concerning these modes of ap-
parent movement is the fact of phenomenal flux as a consequent of
discrete static stimulation. Excepting for the case of gamma. two
separate distributions of energ. effectively impinging upon the re-
ceptor surface (and under certain conditions). are sufficient for the
experience of movement in phenomenal space. Two receptor proc-
esses. spatially and temporally distinct. can be the physiological
initiators of a phenomenal movement in space. The psychophysical
and psychophysiological problems are salient; what are the parame-
ters of stimulation which are causal. and what are the physiological
mechanisms which are operative in the dissonant production of ap-
parent movement?

Unfortunately. the writer has been able to find only partial
attempts to deal with the first of these problems. and these exclu-
Sively in the visual modality. A clarification of the stimulus condi-

tions for gamma movement has been presented by Bartley (3. pp.
156-165). His considerations emphasize the necessity of a proper
temporal course taken by a gradient of illumination on the retina.
H e further asserts that the discretion of the gradient is of no con-

sequence; this is implied to be true also with respect to the other

15

forms of movement to be considered. Using two stationary visible
radiant energy sources. exposed simultaneously or separated by a
variable interval. Wertheimer produced part moVement. beta move-
ment. and phi movement. He concluded that the interspace and
temporal interval between exposures are of critical significance.
Wertheimer's study set the style for subsequent investigations: the
exposure of luminous targets in a dark field. varying mainly their
intensity. size. exposure duration. interspace. and interperiod (hue
and shape have also been manipulated). The experiment of Korte
yielded his classical generalizations of the interrelationships of in-

tensity. exposure time. interspace. and interperiod which proposedly

delimit the stimulus conditions for beta. However. his procedure

.has been criticized and his conclusions have been partly contradicted

on the grounds that the ranges of his variables were too narrow.

and he confused absolute intensity with total luminous flux. Neuhaus
repeated the study. correcting these limitations. and similarly pre-
8tented his findings in the form of generalizations about the interre-

lationships of exposure time. interperiod, and interspace. He re-

pOrted no influence of intensity and hue on these relationships. (An
English summary of the generalizations of both Korte and Neuhaus

is Presented by Graham [12].) A survey of the numerous researches

of this problem (see 12 and 22) indicates that a coherent set of

16

principles is still wanting. The perception of movement under these
conditions of stimulation is with certainty related to the duration of
stimulations. the time gap between stimulations. the distance be-
tween stimulations. and the form of stimulus distributions; intensity
and size of stimulus distributions remain as probable variables. as
well as retinal locus. similarity, and complexity. No attempt will
be made to reiterate the detailed findings inasmuch as systematic
quantitative results are unavailable. The misleading incomparability
of experimental conditions and parameters of stimulation do not lend
themselves for precise quantitative statements. Too frequently the
investigator's limited concern with the general abstract concepts of
intensity. time. space. and form has caused an inadequate specifica-
tion and limited application of the energy distributions impinging
Upon the receptor surface. There is obvious neglect concerning the
comparability of seen movement in various regions of the sensory
fields.

There have been too few and meager attempts to relate phe-
nomenal movement to physiological processes. Bartley (3) has con-
Sidered the role of neuroretinal function in his interpretation of
gamma movement. The latency of differential neural response is
implicated. Numerous considerations enter into the neural basis of

apparent movement as response to multiple static stimulation.

17

Insofar as there is. within limits. an interrelationship of the stimu-
lus parameters of intensity. area. and time for threshold response
and the varieties of interactions in the optic system (see 3. pp. 19-
30 and Chapter IX). and visual processes are organized in accordance
with successive stimulation (4; 32; 33). it is to be eXpected that the
various concepts of nerve discharge and interaction are applicable
here. Concerning the apparent movement of objects having distinct
borders. current neurophysiological concepts have proved resistance
to a satisfactory understanding of contour processes and their rela-
tion to visual movement (3. Chapter X). One author (24. pp. 245-
248) has recently conjectured about the role of shifting peaks of
summating excitation distributions in the primary visual cortex; this
discussion is admittedly not applicable to objectless movement.
Furthermore. this conjecture is derived from an assumption of strict
isomorphism which presupposes a space~time copy of perceptual
events in the neural substrate. This bias was set by Wertheimer
and his Gestalt school in their approach to the problem (17; 18). A
physiological field theory was proposed to offset the inadequacy of
existing neurophysiological concepts for the solution of the space and
time gaps. These theorists asserted the need for postulating a con-
tinuous dynamic brain field operating in accordance with electro-

chemical principles. such that phenomenal continuity initiated by

18

physical discontinuity would have its correlated spatial-temporal
continuity in the hypothetical substrate. Although the reasoning may
seem to have some plausibility. it rests upon the questionable as-
sumption of rigid isomorphism. and its proponents are typically not
deeply steeped in neurophysiology. Saucer has recently proposed a
revision of this principle (25; 26).

The question of the critical anatomical loci for the physio-
logical integration of movement processes has been the object of a
few inquiries. Bartley (3) stresses the properties of the neuroretina.
Some new techniques employed by Motokawa have provided results
which allow him to also deduce the involvement of retinal processes
(20; 21). On the other hand. Shipley £131. (27) obtained evidence
that interaction in the neuroretina was not an essential factor for
the production of apparent movement. These investigators obtained
positive results using interocular stimulation. They also found that
binocular stimulation was more effective than monocular. Interocular
stimulation was similarly employed by Smith (2 8) with the derivation
of positive results. This investigator alternately stimulated both
nasal retinas. and on the basis of the fact that the cortical projec-
tions are so widely separated. he rejected cortical field theory.

The available information does little to clarify the psycho-

physiology of apparent movement. We can hardly expect much more

19

at present. The nervous system as a whole represents a vast and
insufficiently explored territory. It is likely that an understanding
of the optic mechanisms involved in apparent movement will be
greatly enhanced when further insight is gained with regard to the
processes underlying spatial and temporal brightness discriminations.
contour formations. contour effects. and sensory aftereffects--to
name a few basic visual phenomena. Notwithstanding the old argu-
ment for the uniqueness of perceived movement (see Neff. 22). it is
assumed that the visual response is organized and therefore each of
these processes are functionally interrelated.

With the intent of substantiating this contention as well as
reporting some newly acquired information bearing upon apparent
movement. some experimental results will be presented concerning
the parameters of stimulation essential for abient movement in the
line of visual regard.

Bartley and Miller (5) have recently reported some experi-
mental observations in which they cursorily relate some variables
of stimulation to the occurrence of visual movement. oscillating in
the line of regard. They label this form of movement "adab move-
ment"; signifying the adient and abient directions of motion. The
procedure consisted in the fixation of three square luminous targets

exposed briefly in immediate succession on a dark field. The first

20

and third target were identical in all respects; the second was
varied in duration. decreasing intensity and area. Three facts of
especial pertinence are revealed by the findings of these authors:
Apparent movement can be produced by the abrupt shifts in area
and intensity of a retinal stimulus distribution produced by targets
equally distant from the viewer; contour processes are intimately
involved in the occurrence of this form of movement; and the stimu-
lus parameters of duration. intensity. and area are critical variables.
A significant observation here is the necessity for the con-
tour process.which is correlated with the smaller square. to be
initiated in order for movement to be visible. As the exposure time
of the second target was lengthened. starting from zero. a duration
was reached at which the border of the second. smaller square first
emerged. It was only when the second target was presented for
periods above this duration threshold that "adab movement" was
seen. The indications were that the duration threshold for border
emergence is inversely related to the intensity of the second target.
within limits; the authors hypothesize an absolute intensity threshold.
The inference to be made is that the initiation of contour processes
and border emergence is dependent upon the total flux of the
smaller distribution being above threshold. and the occurrence of

adab movement is dependent upon exposure durations above threshold.

21

Movement was difficult to observe when the areal decrement was
greatest and intensity decrement least. A complex relationship was
suggested for varying degrees of areal decrement.

In an earlier and little known study. Calavrezo explored the
relation between apparent movement in depth and successive stimu-
lation by patterns differing in area (7). Besides varying the area
difference between successive targets. he simultaneously varied the
exposure durations of the targets. varied the period between ex-
posures. and varied the configuration of the targets as well as degree
of lateral displacement. Judgments were made of the quality and ex-
tent of movement. Calavrezo reports that optimal movement in depth
occurs when the exposure durations are relatively long and the
interperiod is relatively short. When the exposure durations were
relatively brief. pure phi movement was observed; with still briefer
exposures. part movement was seen. Above an exposure duration
threshold for optimal movement. further increases in duration had
no effect upon movement in depth. The extent of movement is re-
ported to rapidly increase with increasing differences between the
areas of successive stimuli; however. very large differences did not
produce movement.

The present study is intended to further elucidate the circum-

stances relevant to apparent movement in the line of regard. The

22

plan of experimentation was comprised of two parts. In the first
group of observations. the attempt was made to approximate for the
second of two stimulus distributions. the lower limits of duration.
intensity. and area. under present conditions. for the visual appear-
ance of movement in the line of regard. In the interest of sim-
plicity. only two targets entered into a given sequence throughout the
experiment. When optimal movement was perceived. it was only of
abient character. Previous evidence indicates that under certain
conditions successive stimulations have a complex interactive effect
on the visual result (4; 8; 10; 29; 32; 33); hence. a third stimulus
pulse would probably obscure the relations of the stimulus parame-
ters for the two-pulse sequence. Consequently. the general problem
was limited to the case of the apparent visually receding movement
of a luminous object in a dark field. With the aim of more fully
exploring the parameters of stimulation pertinent to the interaction
of the two successive pulses and the emergence of apparent move-
ment. two additional stimulus variables were systematically manipu-
lated; viz.. the duration of the first of the two pulses. and the
temporal relations between the termination of the first pulse and the
onset of the second pulse. The area and intensity of the first stim-
ulus distribution were held constant throughout the entire experiment.

These initial observations were largely exploratory in character.

23

serving a twofold purpose. They provided approximations of the
relevance and ranges of the above variables in various combinations
within which optimal abient movement occurred. thereby facilitating
more precise psychophysical measurement in the second portion of
the investigation. Secondly. they revealed in a semiquantitative
manner the various phenomena which occur. under these general
conditions of stimulation. just beyond the thresholds for movement.
Inasmuch as all of these phenomena are related to the extent that
they occur as gradual emergences which are consequences of sys-
tematic variation of basically the same parameters of stimulation.
the)? will all be reported. It is most likely that there is some com-
monality of processes underlying these phenomena. such that an
understanding of any one or more of these phenomena may contribute
to the explanation of them all.

The second portion of the investigation was designed with the
intent of quantitatively determining the relationship of the magnitude
0f the temporal interval between the two stimuli. within which op-
timal abient movement is seen. with the relative area and intensity
difference of the two patterns of stimulation. In the classical in-
vestigations of apparent movement. particular emphasis has been
traditionally directed toward this interval. The preliminary observa-

tions from the first part of this general investigation indicated that

24

this temporal interval and the relative area and intensity decrement
between the two stimulus patterns are lawfully interrelated for the

occurrence of optimal abient movement.

ME THOD

Stated in terms of direct experimental manipulations. the pre-
liminary observations were monocularly obtained by using targets of
the following character. Two luminous squares (presented in the
orientation of a diamond). equally distant from the observer and in
the line of regard. were exposed in succession. The temporal rela‘
tion between the two was varied so as to produce: an interval of
varied length between exposures. the immediate succession of ex-
posures. and varying degrees of temporal overlap. The observer
was set fo fixate the center of the targets. maintaining steady fixa-
tion throughout a sequence. The first target. of fixed area and in-
tensity. was made to vary in duration of exposure. The second
target was made to vary in area. intensity. and duration of exposure.
Each of these variables was independently manipulated. the tem-
poral relation between exposures being the main experimental vari-
able. utilizing a modified method of serial exploration. Notations
were made of the perceptual event which occurred for each of the
many combinations of stimulus parameters employed. Because this
portion of the inquiry was exploratory. a rigorous method of obser-
vation was bypassed. The number of exposures to each combination

25

26

of parameters was free to vary at the discretion of the observer.
permitting response definition. The lack of prior expectations pre-
cluded a fully systematic exploration of all conditions.

The viewing distance between cornea and target was set at
70 cm. The constant area of the first target was given by a 1.22
cm. side. or a retinal subtense of approximately 60 minutes of visual
angle; the fixed intensity of the first target was 76.1 c/ftz. In these
preliminary observations. the area of the second target was varied
twice. 1.04 cm. and .61 cm. side dimensions. corresponding to
retinal subtenses of 51 and 30 minutes respectively. Two intensity
levels were employed for the second target. 46.4 and 12.1 c/ftz.
Durations of exposure for each of the two targets varied stepwise
over a range of from 10 to 600 msec. The varied interval of time
between the termination of the exposure of TargetI and the initiation
of the exposure of Target II ranged from 100 msec. through zero
and to negative values as large as 200 msec. (A negative value
designates temporal overlap of exposures; i.e.. exposure of Target 11
before Target 1 is withdrawn.)

The second major portion of the experiment was addressed
toward the assessment of optimal abient movement thresholds. with
the temporal relation between the termination of the exposure of

Target land the initiation of Target 11 as the primary experimental

27

variable and relative area and intensity decrement of Target 11 as
secondary independent variables. Insofar as this temporal relation
was varied between positive values and negative values. a positive
and a negative threshold are not inconceivable. These thresholds
were assessed employing six intensity levels and six areal magni-
tudes for the second target. The area and intensity of the first
target remained fixed throughout all measurements. The exposure
durations for both targets were held constant and equal at 280 msec.
The intensity of the first target was 76.1 c/ftz; its area was that of
a 1.22 cm. square. giving a subtense of 60 angular minutes. Inten-
sity levels for the second target were 46.4. 28.2, 18.2. 12.1. 7.6. and
4.9 c/ftz. The six areal magnitudes utilized for the second target
were given by 1.10. 1.04. .92, .61. .49. and .31 cm. squares. sub-
tending 54. 51. 45. 30. 24, and 15 minutes of angle respectively.
Both targets remained equally distant from the cornea. 70 cm. The
temporal interval. OFF-ON. was varied as before. through positive
and negative ranges in 10 msec. steps.

The two arms of a Harvard tachistoscope had been extended.
Light sources were stationed at the far end of each arm. These
were each two four-watt fluorescent tubes. Opal glass filters were
Placed in front of these light sources to diffuse the energy and con-

trol the intensity levels. An opaque screen with a square aperture

28

was placed in each arm of the tachistoscope between viewing apera-
ture and filters at a transmitting distance of 70 cm. from the cornea
of the right eye. The aperture in Arm II was continuously variable;
the other was fixed. These constituted the targets.

The tachistoscope was activated by an electrical timer which
provided the control of the three temporal parameters. independently
variable. Each of the three scales on the timing mechanism was
graduated in 10 msec. units. When the three scales were set. the
depression of one button closed a circuit and initiated the desired
stimuli sequence. A very low level of illumination was continually
transmitted through the aperturein Arm 1. When observed through

the viewing aperture of the tachistoscope Eh a dark adapted eye.

 

this continual low energy level. after passing through the opal glass
filters and the target aperture. was just barely noticeable as a
vague amorphous glow. In effect. a threshold stimulus for the dark
adapted eye served as a fixation point in the locus of the to be
presented target. This proved to be an invaluable operational aid.
An eyeshield around the viewing apertures of the tachistoscope pro-
vided head stability.

All observations were made monocularly. and in a dark room.
A 40 watt shielded red lamp was positioned over the timing mech-

anism to permit the necessary manipulations and readings. The eyes

29

were dark adapted for about fifteen minutes before observation be-
gan. An observational session lasted from one to two hours. and
all were staggered throughout a period of months. Although this
was not controlled. each pair of threshold measurements usually
occupied a full observational session.

The OFF-ON thresholds were measured by the method of
limits; ten series were run using alternating series for every de-
termination in the second portion of the study. Three general cate-
gories of response were employed. yielding estimates of two thresh-
olds. For example. the targets were presented in succession sepa-
rated by a noticeable time gap. The physical time gap was then
decreased by 10 msec. decrements for each subsequent presentation
of the two targets. until the visual response changed to abient move-
ment. A reading was then taken from the scale on the timing
mechanism which designated. in milliseconds. an estimate of the
largest separation between the two exposures with which optimal
abient movement could be seen. Subsequent presentations of the two
targets were made with the temporal relation between termination of
the first exposure and initiation of the second exposure decreased 10
msec. each time. This was continued until the visual response
changed to one other than movement; a reading was again taken. It

generally occurred that this reading was negative. signifying that the

30

onset of the second target took place before the offset of the first;
the two overlapped in physical time. Hence. this threshold estimate
indicated the largest temporal overlap with which movement was
just not noticeable. This description pertains to one series with
the parameter varied in a given direction; i.e.. decreasing time gaps
through zero and into the overlap range. The next series was run
with the parameter varied in the other direction. In other words.
the targets were successively presented with an amount of temporal
overlap which was calculated to produce a nonmovement percept.
For each subsequent presentation. the temporal relation between off-
set of Target I and onset of Target II was varied. by 10 msec. incre-
ments effecting a physical transition from temporal overlap. through
immediate succession. and into the range of delayed succession.
This series similarly yielded estimates of two thresholds. To repeat.
ten series were run in alternation. The ten estimates for each of
the two thresholds were averaged. thereby giving two empirical
thresholds for apparent abient movement under a given set of condi—
tions.

Such pairs of thresholds were determined for each of the con-
ditions described earlier. All of the intensity levels for a given
area of Target II were explored. in random order. before the area

was altered.

31

Because this study was largely exploratory in character and
the innumerable observations in addition to the threshold determina-
tions occupied a countless number of total hours. the findings re-
ported hereunder are based upon the observations of one observer;
viz.. the author. Although this procedure leaves much to be desired,
it is a mode for obtaining data which would otherwise remain in
want. The writer is aware of no preconceptions which may have in-
fluenced the relationships revealed by the results; therefore. if the
data comply in some way with relevant information gleaned from
other sources. their value can be accordingly assessed.

It may be pertinent to declare that these data were obtained
from an experimentally sophisticated observer. much practiced and
operating with strict. relatively constant criteria. After innumerable
pilot observations. two important observational stances developed;
viz.. the stability of fixation and the definition of the response cate-
gories. The naive observer found it difficult to maintain steady
fixation on a threshold point; and when the luminous target suddenly
and briefly emerged in a dark field. the gaze almost invariably
wandered in an indeterminable manner. For purposes of this sort.
at the moment of onset and during stimulation a steady fixation is a
significant factor. Furthermore. with the very brief exposures em-

ployed in this investigation. it was found that an event which to the

32.

uninitiated looked generally like a brief flash of light having more
or less definite form. became a variety of rather definite and

fairly well articulated visual events of some complexity to the ex-
perienced observer. This last fact is especially of some theoretical
import. Very brief perceptual events may become articulated with
repetition. However. further consideration of this point is tempered
by recognition of the need to clarify whether or not this increased
definition of perceptual categories is independent of the acquired
stability of fixation. The fact remains-~the observational data of the
experienced observer are richer than those of the inexperienced ob-
server.

It is also of consequence at this point to speak of the criterion
employed in recording data. Although the observations were made
with complete cognizance of the stimulating conditions. a very delib-
erate attempt was constantly exercised to avoid making a stimulus
error; i.e.. attention was focused conscientiously upon the phenomenal
event as such. The various perceptual response categories. when
finally defined in experience, were each capable of complete verbal
description. independent of the stimulating conditions. During the
determination of thresholds for optimal abient movement. a rigid
criterion was maintained; movement was recorded only when an ob-

ject was seen to recede into the distance. As is usually the case in

33

psychophysical measurements. thresholds were not phenomenally dis-
crete and the brevity of phenomena incurred difficulty in the response
categorizations here. This difficulty was mainly existent at the
negative threshold (temporally overlapping exposures) under certain
stimulus conditions; i.e.. when the intensity of Target 11 was rather

low.

RESULTS

Figure l is presented as a schematic representation of the
approximate relations of the various observed phenomena with the
temporal parameters of stimulation. The relationships depicted in
this diagram can be understood as generally true for those combina-
tions of stimulus area and intensity revealed adequate for the per-
ception of optimal abient movement (as discussed below). An im-
portant qualification for this illustration is that the phenomena
pertain only to the circumstances in which Target II is exposed for
durations which are at least approximately 100 msec.

The zone bounded by the two curves on the left is intended
to represent variations in the exposure duration of Target I. t1.
ranging from 10 msec. durations at the bottom to 600 msec. at the
top. The zone bounded by the two verticals at the far right repre-

sents exposure durations of Target 11. t . over approximately 100

Z
msec. Between these two zones lies a zone representing the maxi-
mum durations of the unstimulated interval between exposures.

OFF-ON. over which Target 11 has a perceptual influence on the ef-
fect of Target 1. This interval will henceforth be referred to as the

"critical interval." For clarity of illustration. this latter zone has

34

35

.wmQSK uxxoacx 33 \o wzngmx «exoo‘Eou
m3 $2.203. \0 5.3309 \czulmuxml .\ .90..

.1
I
1

 

tr-------"""""
l
l
L

mxozmn «to

  
    

wtmfioon oxiuusfwtnvu
-m>.~mxuo\~m .8 .zVo‘nzmfis

mtmmeoS \sE.~.~o\0
.T I II. I. II I. .23 to.~n§\

 

0
§
G
K
‘s
8
\o

36

been given values applicable to the situations in which the second
target was a 1.04 cm. square with an intensity of 12.1 c/ftz. This
latter zone is seen to vary from approximately 70 msec. for very
brief durations of Target I to 25 msec. for exposure durations
greater than approximately 100 msec. For very long exposures of
Target I. the critical interval remains constant. The evidence sug-
gests that the precise magnitude of this parameter varies with the
particular set of relative area and intensity differences employed
for the two targets. This will be considered below.

Optimal abient movement is seen only when Target I is ex-
Posed at durations greater than approximately 100 msec. and is
separated in time from Target II no more than a particular constant
critical interval. There need be no temporal separation between the
EXPOSures of the two targets for optimal abient movement to occur;
in fact it will occur with some degree of temporal overlap (to be
considered later).

As the duration of Target I is decreased below that necessary
for Optimal movement. different phenomena occur and the critical
interval during which they occur increases. The first to appear as

the EXposure duration. t . of Target I is decreased is what shall be

1

called ”adab phi movement." This phenomenon appears as a flash

of light which first sometimes looms slightly forward and then

37

suddenly recedes to a greater distance. taking contoured form only
at its distant terminus. Movement in the abient direction is the
more predominant as a perceptual characteristic and was somewhat
more frequent in occurrence. Occasionally. with these temporal
parameters. a form of movement similar to gamma was perceived;
i-e-, an inarticulate flash of light would expand on sudden appearance;
then it would quickly contract--finally taking contoured form. This
is not precisely the same as the more familiar gamma movement
which is produced by the sudden presentation and withdrawal of a
single target. The expansive-contractive phenomenon reported herein
is very much more compelling than is usually the case when one
target is employed. In any case. utilizing these particular temporal
Parameters. the initial phases of these phenomena-~during move-
ment-"were without definitive contours.

When the duration of exposure of Target 1. t1. was very
brief. less than about 40 msec.. the observation was that of a
unitary. stationary object; vis.. the second lighted square. It was
dUPing these circumstances that the critical interval was the
greatest for a set of targets of given intensity and area; for the
case illustrated it was 70 msec. There is no doubt that the seen
Square was exclusively the perceptual effect of the second stimulus

and not the first. By starting with the long exposures of Target I

38

and continually shortening its duration. it was possible to trace the
gradual transitions of the phenomenauin the order described--
whereby first the contour effect. and then the complete perceptual

effect of Target I was obliterated by the processes initiated as a

 

consequence of the second stimulus. Further clarification is offered
by the fact that when Target I was presented alone for durations at
least as brief as 10 msec.. it was readily visible as a luminous
square object.

The phenomena described above were observed under the
special conditions in which the two targets of appropriate intensity
and area difference were separated in time by some value less than
the limit of the respective critical intervals. and when the exposure
durations. t2. of Target 11 were greater than approximately 100
msec. When tz was less than this value. the second square was

seen to intrude within the first square. providing the two target ex-

 

Posures were separated in time by no more than the limit of the
appropriate critical interval. For a given pair of targets the mag-
nitude of the critical interval was found to vary with the magnitude
0f 1:1 in a manner similar to the situation described above and illus-
trated in Figure l. The primary difference in result is that move-

ments were not seen. The phenomenon of intrusion takes the form

of a lighted square which is surrounded by a black frame. which in

39

turn is surrounded by a lighted frame. When tl is long. the first
square appears. then to be intruded by the second smaller square.
As tl decreases. the relative latency of intrusion decreased until the
phenomenon appears immediately as the smaller square surrounded
by the two frames. The two frames may vary in width. depending
upon the area difference between the two targets. With a large dif-
ference. the black frame and the lighted frame were broader. and
were segregated by a gradual brightness gradient. When the areas
of the two targets differed to a lesser degree. both frames appeared
narrower and possessed a sharper separating edge. Another factor
which influenced the width of these frames was the exposure duration
of Target I. Intrusion appeared to get stronger when tl was de-
creased to small values; i.e.. the black frame became wider and
more pronounced while the lighted frame became narrower. Also.
similar to the preceding conditions. as t1 was made brief the outer
bounds of the lighted frame lost its sharp edgelike character; and

when t was decreased below approximately 40 msec.. the frames

1
vanished and only the second square was visible. providing t2 was
not smaller than t1. If both t1 and t2 were made very small. the
frames persisted around the small square although they were some-

what less salient. Unfortunately. the observations were insufficient

for a reasonable assessment of the influence which stimulus intensity

40

and area may exert on the fates of these phenomena. A precise
statement of the relationships of relative differences in area as well
as intensity with the target exposure durations. insofar as these
phenomena are concerned. would require different experimental con-
ditions. The exposure durations would have to be varied over more
and smaller steps than those employed here.

These phenomena of intrusion and obliteration could be dupli-
cated under conditions. If tl was not too small and t2 was greater.
the orderly occurrence of intrusion and obliteration could be pro-
duced by causing the exposures of the two targets to overlap suffi-
ciently in time; on I--on II--off I--off II. A critical factor concern-
ing obliteration seems to be the time between onsets. Although the
observations of these events were not systematically precise enough
to permit an exact statement. there were indications that the thresh-
old magnitudes of the ON-ON intervals would be comparable whether
the exposure duration of Target I were decreased without temporal
Overlap of a stimuli. or whether the stimuli overlapped in increasing
amount. In the latter case. it was noted that if the intensity differ-
ence between the two targets was too great. intrusion never took
place; two squares of different brightness were seen in immediate

succession. Obliteration took place when the relative intensity dif-

ference was as great as .96.

41

Two additional facts may be mentioned. All of the phenomena
described above were observed when the critical intervals had nega-
tive values; i.e.. when the onset of the second stimulus just preceded
the offset of the first stimulus by some appropriate magnitude.
thereby effecting a temporal overlap of stimuli. This may not be
surprising in the cases of perceptual intrusion and obliteration; how-
ever. it is to be noted that the varieties of movement were observed
without the existence of an unstimulated interval between exposures.
Some quantitative data are presented below.

Almost needless to say. in every case when the positive limit
of the critical interval was exceeded. all phenomena gave way to the
perception of successive duality. When the area difference was not
great. this took the form of flicker; the dark phase was predomin-
antly in the center. and became more general as the unstimulated
interval was increased. With larger area differences. exceeding the
limit of the critical interval gave rise initially to the immediate
succession of two different squares with no noticeable dark phase
as in flicker. This phenomenon existed throughout a measurable
range before flicker became apparent.

Inasmuch as the primary phenomenon intended for investigation
was movement. rigorous measurement was attempted only of the

limits within which optimal abient movement was seen. The

42

preliminary observations discussed above indicated that above mini-
mum exposure durations for the two targets. which were thresholds
for movement. further increases had no effect on optimal abient
movement and the magnitudes of its critical interval. Hence. all
subsequent measures were made with t1 and t2 each constant at 280
msec. The relative decrements in area and intensity. as well as the
temporal relation between offset of Target I and onset of Target 11.
were found to be determining parameters above the exposure thresh-
olds for the two targets.

Figure 2 is designed as a three-dimensional graph which
yields an optimal abient movement solid depicting the interrelation-
ships between the three independent variables. for these conditions
of experimentation. The two horizontal axes represent continua for
relative differences of area. (Al-AZ)/Al. and intensity. (ll-Iz)/Il.
These data are plotted in terms of relative difference because it is
likely that the obtained relationships have some generality to abso-
lute magnitudes other than those employed. The temporal relation
between termination of Target I and initiation of Target 11. OFFI-ON.
is represented on the vertical axis. Negative values on this axis

indicate the extent to which the onset of Target II precedes the off-

set of Target 1. Positive values specify the magnitude of the

OFF- ON INTERVAL

'201

43

 

’IO 1
.?o

.84I

 

 

'30

 

 

Fiy. 2. fielaz‘fonship: between OFF-
ON threshold for optimal-abl‘ent-moyement ana’
relative intensity and area differences
between two targets. Frontal and ,oer—
Speotive axes represent relative
differences in :I'nz‘enszt)’ and area.

44

unstimulated interval between the two exposures. The plotted data
are also contained in Table 1.

Within the volume bounded by the curves. any set of coordi-
nates comprise parameters of stimulation adequate for the perception
of optimal abient movement. A bright square is seen to briefly ap-
pear in a dark field. and quickly and smoothly recede to a distant
position-”then suddenly it disappears. No attempt was made to
assess the extent or velocity of phenomenal movement. Beyond the
bounds of the solid. movement is not seen. When the relative differ-
ences of area or intensity are too great. movement is replaced by
the perception of the immediate succession of two squares-~differing
in area. brightness. or both. These transitions are relatively
abrupt. A very slight difference in area gives rise to one square
VVhich may decrease in brightness. Experimental conditions did not
Permit observations with a very small difference in relative intensity.
Beyond the positive OFF—ON threshold. movement is replaced by
either the immediate succession or the flickering succession of two
Squares differing in area. brightness. or both. When the area dif-
f'l'l‘I‘ence is small. flicker is seen. As the area difference increases.
the category of immediate succession becomes interpolated between
fliclker and movement. The range of this category increases pro-

greasively with increasing relative area difference. In every case.

OFF-ON THRESHOLDS FOR OPTIMAL ABIENT MOVEMENT

TABLE 1

FOR VARIOUS RELATIVE DIFFERENCES IN
STIMULUS AREA AND INTENSITY a

#—

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

,r
.93 - -
(15') - -
.84 5 (1.8) 13 (1.7) 16 (2.1) 2;(2.1) -
(24') 2 (0.6) 8 (2.6) 9 (1.6) 24 (2.4) -
.75 25 (1.5) 32 (1.5) 39 (1.6) 4; (3:) 28 (2.1) f
Al-AZ (30') 13 (1.3) 21 (2.2) 28 (2.1) 37 (3.3) 31 (2.2) -
A1 ”T43 13 (2.3) 18‘ (1.5) 35 (2.6) 25 (2.1) -
(45') 7 (2.0) 11 (2.2) 25 (1.5) 31 (2.2) -
:7 11 (2.1) 14 (1.8) 25 (2.6) -
(51') 13 (2.0) 11 (2.2) 15 (2.1) -
19 5 (0.5) - -
(54') 5 (0.5) — -
WW -
.39 .63 .76 .84 .90 .94
(46.4 (28.2 (18.2 (12.1 (7.6 (4.92
c/ftz‘) c/ftz) c/ftz) c/ftz) c/ftz) c/ft )
11'12
1

ard errors in parentheses.

aData from one subject.

Upper and lower numbers in each
0'311 designate positive and negative thresholds. respectively. Stand-

absolute dimensions of second stimulus.
square at 76.1 c/ftz.

Marginal parenthetical values represent

First stimulus was a 60'

46

when the negative OFF-ON threshold is crossed. the second square
is seen to intrude within the first.

The movement solid indicates that the temporal interval be-
tween termination of Target I and initiation of Target II. within
which optimal abient movement is seen. lawfully varies in magni—
tude--both positively and negatively--with manipulations of both the
relative area difference and the relative intensity difference. Posi-
tive and negative OFF-ON thresholds tend to vary together in the
same manner (although their values differ). so that the negative
portion of the solid resembles a reflection of the positive portion.
Qualitatively. it might be said that movement is more impressive
under the conditions in which the absolute values of the critical
intervals on the OFF-ON dimension are largest; e.g.. when the rela-
tive differences in areas and intensities are .75 and .84. respectively.
This is not necessarily intended to imply that movement is more

Compelling than at other combinations of parameters. but rather that

it appears to be more extensive in space or time.

DISC USSION

The varied facts of apparent movement. dissonantly produced.
have invariably prompted the invocation of hypothetical physiological
processes to account for the continuity of perception which may re-
sult from such stimulation. In this respect. the problem is analo-
gous to that of flicker fusion. The necessity of postulating continuity
of physiological events remains a moot point if one seeks a correla-
tion between perception and neural action. in view of the fundamentally
discrete character of the nerve discharge. Nonetheless. the facts
indicate that the processes associated with these perceptual events.
whatever their true nature may be. must continue to be active for a
brief time after stimulation has ceased--in order .to give rise to the
perceptual after-effect of stimulation. Evidence for both neural and
perceptual after—effects of stimulation is not lacking in a variety of
sources. The problem concerning spatial continuity (movement) as
well as temporal continuity (fusion) in perception vastly increases
the challenge of the task. Spatial interaction at both the perceptual
and neural levels is varied in form. and continues to hold a some-
what enigmatic status in both areas of inquiry. The following dis-
course will attempt to interpret the obtained experimental findings

47

48

in terms of hypothetical neural correlates. Speculation will not be
offered for the precise character or locus of these processes.
Almost needless to say. temporal continuity or fusion in per-
ception is an essential condition for apparent movement. When
movement was seen. the brightness of the object was steady. This
was true throughout the range of an unstimulated critical interval.
The range for fusion was usually somewhat greater than that for
movement. In any event. the percept lingered at least as long as a
period measured by the critical interval. The maximum of this in‘
terval. 43 msec.. was obtained when the relative differences in area
and intensity were .75 and .84 respectively. Manipulations of these
latter two variables systematically altered the magnitude of the
critical interval for movement; they similarly appeared to alter the
interval for fusion. Furthermore. fusion of the effect of this type
of stimulation was found to obtain only when the first pulse was de-
livered for durations. as long as approximately 40 msec.; shorter
durations resulted only in the visual effect of the second pulse. Ex-
amination of Figure 1 suggests that a minimum interval between
pulse onsets is a necessary determinant for fusion to result from
two successive pulses. It should be noted that above this ON-ON
threshold for fusion. increasing the ratio of the stimulated to the

unstimulated interval is of no apparent consequence. Providing the

49

interval between pulse onsets is not decreased. the exposure dura-
tion of Target I may be increased and the unstimulated interval may
be decreased--fusion always obtains. This finding is consistent with
that of Bartley (2). He similarly reported that within the limits of
the ON-ON threshold for fusion of intermittent stimulation the ratio
of the stimulated to the unstimulated intervals may vary consider-
ably. Bartley also notes that when the pulse duration is made too
brief. flicker may result. The results of this investigation imply
that when the pulse duration is decreased while stimulating at a fre-
quency rate equal to the ON-ON threshold for fusion. flashes are
seen in succession and hence. flicker would result
The time course of obliteration can be traced by varying the
length of the ON-ON interval. This was effected in the preliminary
observations by independently manipulating the magnitudes of the
exposure duration for Target I and the duration of the unstimulated
interval. (Similar observations were noted when targets overlapped
in time.) The perceptual result of shortening the ON-ON interval
below approximately 143 msec. was: first the contour of the first
square failed to appear; and as the interval was further abbreviated.
the total effect of the first pulse failed to appear. This perceptual
Obliteration took place continuously. from periphery to center. as

t1’18 ON~ON interval was shortened. In perceptual terms. this is as

50

though the dark surrounds of the second square began to emerge be-
fore the square itself appeared. One may hypothesize that the neural
process correlated with the emergence of a figure has associated
with it in its surrounds a neural process of opposite character

which antedates the figural process in the form of a temporal-spatial
gradient of activity. The activity of this surround process is initi-
ated first in the periphery and then spreads toward the bounds of

the figure process during the course of time. The time course of
this centripetally spreading surround activity is given by the course
of obliteration of contour and flash as the ON—ON interval is
diminished (see Fig. l). The surround activity is characterized as
inhibitive of ongoing activity differing in strength. When the intensity
difference between Target 11 and its field is very large. the ante-
cedent inhibitory activity is strong; this cuts short the persisting
after-effect of the first exposure--resulting in flicker at short OFF-
ON intervals. As the intensitydifference between Target II and its
field is lowered. the preceding inhibitory effect becomes weaker and.
accordingly. the figural process of the first exposure persists longer;
this yields a larger OFFeON interval for fusion. Within the limits
of the critical interval for movement. the centripetal spread of in-
hibition is terminated by the figural processes and movement is

seen. If the inhibitory effect is thereby weakened to a large extent.

51

it ceases to antedate its figural process and the first figural process
therefore persists throughout its unimpeded limit-~resulting in the
immediate succession of two squares within the limits of a critical
unstimulated interval for fusion. A similar interpretation is ap-
plicable to the effects of variations of the area of Target 11. Large
areas are preceded by strong inhibitory activity. and give rise to
brief critical intervals. Reduction in the area of the target is ac-
companied by a corresponding reduction in the strength of the in-
hibitory activity and a subsequent increase in the critical intervals.
If the inhibitory activity is weakened too much by area reduction.
the centripetal spread of inhibition fails to occur and two different
sized squares are seen in immediate succession. The combined
effects of area and intensity difference are maximal.

The foregoing interpretation of the interaction of successive
figural processes conforms with all observations made in this in.
vestigation. Although thresholds for fusion were not measured. a
trend was noticeable suggesting an increased OFF-ON interval for
fusion as both relative difference in area and intensity were in-
creased. In conjunction with this. the measured. OFF-ON thresh-
olds for movement were found to similarly vary. The fact that the
negative OFFsON thresholds varied in the same manner as did the

positive is further support for the thesis that the inhibitory activity

52

of the surround process associated with a figural process is weak-
ened as area and intensity are decreased. Furthermore. it is
additionally significant that whenever intrusion was noted. the salience
of the dark frame (representing the immediate surround of the sec-
ond square) was a function of both the intensity and area of Target
II. At large intensity and area. the frame was very dark and
sharply contoured. The opposite was true in the reverse case.

Numerous observations of the course of obliteration as the
OFF-ON interval was decreased convincingly illustrate the centrip-
etal spread of the dark surround. In conjunction with the facts of
fusion. and the fact that the phenomena of obliteration are dependent
upon successive stimulation. it is highly likely that the surround
activity is attributable to the later pulse.

The behavior of the second process when exposures are very
brief is perhaps of secondary interest. Under these conditions.
movements were never seen. If the exposures of Target I and
Target 11 were. respectively. over and under approximately 100
msec.. the first square was seen to be intruded by the second
throughout a measurable critical unstimulated interval. When the
'exposure duration of Target I was made shorter than that of Target
II. the first square was no longer seen. The intrusion by the brief

second square. suggests that very brief figural processes are

53

accompanied by very weak antedating surround processes and very
weak contemporary surround processes. Therefore. the persistence
of the processes initiated by the first exposure would be prolonged
to its near maximum limit. only to be slightly impeded by the
contemporary immediate surround process associated with the sec-
ond figural process. Only when the first exposure was itself very
brief did evidence for the inhibitory activity of antedating surround
processes emerge. If the exposure duration of Target I was made
less than that of the second. the first square did not appear even
though the second exposure was less than approximately 100 msec.
Seemingly. the weak antedating surround process of the second fig-
ural process could inhibit the first figural process (within the limits
of the critical interval) only if the latter was itself much weakened
by a very brief exposure. It is worth noting that the fact that the
critical interval for the total obliteration of the first square is
greater than that for movement is evidence for the greater magni-
tude of the critical OFF-ON interval for fusion.

In the experiment of Bartley and Miller (5) the presentation
of targets was of the order I-II-l; Target II was varied in duration.
decreasing intensity and area. All exposures were in immediate
succession. These authors reported adab movement. the recession

and approach in the line of regard of a bright square in a dark

54

field. Their findings indicate that an essential determinant for the
occurrence of adab movement was exposure of the second target at
durations somewhat greater than that necessary for the emergence
of the contour of the second square. They assert that the duration
increment necessary for movement is a function of the total energy
level of the second target employed to elicit the perception of the
second square; weak second targets required long exposures for
their contours to emerge in perception. A subsequent increase in
exposure produced movement (providing other parameters were ad-
justed accordingly). The present conclusions concerning the effect
of a later stimulus on the processes initiated by a previous one are
consistent with the observations reported by these authors. It may
well have been the case that their second and third exposures were
interacting in a manner like that of the first and second exposures
described herein. They report that the total emergence of the
former is a function of an inverse relationship between intensity and
duration; the occurrence of movement resulted from a further in-
crease of duration. The intensity of the first exposure was not
varied in this experiment. but minimal durations were observed
above which first the figure and then movement emerged. Had the
intensity of the first exposure been manipulated in this investigation.

the rationale presented above would predict the function reported by

55

Bartley and Miller; i.e.. the duration threshold below which total
obliteration occurs would be inversely related to intensity. This is
precisely the result obtained independently by Cheatham (8) who in-
vestigated only the phenomena of obliteration. under similar condi-
tions of stimulation.

Bartley and Miller also suggested a complex role played by
the area of their second target. Large targets are said to involve
two opposing effects; Viz... increased total energy and the closer prox-
imity of contours. Although the former .would tend to lower the
duration threshold. the latter would presumably raise it The authors
reported that small targets required briefer durations for their con-
tours to emerge. Here again. although the area of the first target
was not manipulated in this study. the principles derived would make
similar predictions. Inasmuch as the antedating inhibitory activity of
surround processes spread centripetally. it is to be expected that
contour processes representing a small area would be affected later
than those representing a large area; hence. a lower duration thresh~
old would exist for the obliteration of contours.

At this point it is necessary to direct consideration to the
phenomena arising when the target exposures overlap in time. ON-
ON-OFF-OFF. The contemporary surround process associated with

a figure process is characterized by inhibitory activity. the strength

56

of which is maximal at the boundary between the two processes--
decreasing toward the periphery. This applies to the phenomena of
intrusion. when a graded black frame was seen to separate the con-
tour of the smaller square from that of the larger. It is also con—
sistent with the more familiar manifestations of "simultaneous
contrast" (24. pp. 229-236). The strength of this activity is directly
related to the steepness of the intensity gradient at the border of the
stimulus; and. accordingly, to the duration of exposure. When the in-
tensity of Target 11 was relatively high. intrusion was observed to
occur with minimal temporal overlap of stimulation and the black
frame was very dark. The fact that movement was observed with
small amounts of temporal overlap suggests that the centripetal
spread of inhibitory activity can be contemporaneous with the onset
of stimulation. and an appreciable amount of time is required for
the strength gradient of inhibition to develop. If the intensity
gradient at the border of the stimulus was made too shallow. the
strength gradient of inhibitory activity never developed sufficiently
for intrusion to take placeo—despite very large temporal overlap of
stimulation As before. similar interpretations apply to the effects
of area. Small areas have weak surround processes associated
with them; hence. more time is required for the strength gradient

of inhibitory activity to develop. and accordingly. a greater amount

57

of temporal overlap is necessary for intrusion to occur. Although
very small areas were not employed. it would be predicted that in—
trusion would never occur if the relative differences in area as well
as intensity were made too large.

As mentioned earlier. no attempt will be made here to ground
these concepts in principles of neurophysiology. The facts of neuro‘
physiology which can be successfully applied to these findings are
not presently available. The various familiar forms of spatial and
temporal interaction in perception have not as yet been adequately
related to concepts of neural organization. It is believed that the
results of this study pose a particular problem for neurophysiolog
which has not yet received its due recognition; viz.. the fact that a
pulse of stimulation may have considerable influence upon the effect
of a preceding pulse. Traditionally. attention has been directed
toward the reverse relation. Psychologically. this fact was first
recognized by Werner (Z9) and has since been explored by
Cheatham (8).

Concerning the phenomenal fact of movement in three dimen-
sional space. conclusions are to be made with caution. It is to be
emphasized that observations were made with a definite set to see
one object moving into the distance. The appropriate relative dif-

ferences in area and intensity between two successive exposures.

58

under proper temporal conditions. were found to favor this set.
However. this does not preclude the possibility that another observer
(assuming sophistication in observational technique) might perceive
one stationary object continuously decreasing in size and bright-
ness. were he operating with a different set. This is an empirical
matter. There have been numerous demonstrations of the effect
which observational sets have on the perceptual outcome (13; 16).
The likelihood remains that the centripetal spread of the dark sur-
round closing in on a figure which decreases in brightness (and this
is a post hoc construct) may be a sufficient condition for the re-
spective sets to result in either outcome. The relationships between
perceived distance and both size and brightness have been amply

demonstrated (9; 14; 15; 19).

SUMMARY

This investigation was conducted with the purpose of exploring
the parameters. of stimulation necessary for the perception of op-
timal abient movement. Two simple luminous targets were exposed
in succession. each being located at the same optical distance and
with their centers on the principle visual axis. The first target was
of constant area and intensity; for different exposure sequences the
second was made to vary in area or intensity. The temporal rela-
tion between termination of the first exposure and initiation of the
second exposure was serially varied. thereby producing varying de-
grees of unstimulated interval between exposures and varying degrees
of temporal overlap of exposures. Monocular viewing was employed.
and stimulation was localized in the fovea. Data reported are from
one trained observer.

Threshold measures for the magnitudes of both the unstimu-
lated interval between exposures and the interval of temporal over-
lap indicate that there is a range of relative differences in both
stimulus area and intensity within which movement is perceptible.
Within these ranges. the absolute values of the threshold measures

of the temporal relation between exposures were found to lawfully

59

60

increase as direct functions of both relative difference in area and
intensity. An important provision was the exposure of both targets
above minimal durations.

Collateral observations revealed an orderly variety of related
phenomena. These were: continuity of light (fusion). discontinuity
of light (flicker). intrusion of the smaller second square within the
first. contour obliteration resulting in either of two forms of pure
movement. and the total obliteration of the effect of the first stim-
ulus. Each of these phenomena were semiquantitatively discussed in
relation to the parameters of stimulation.

The interpretation of all phenomena reported is made in terms
of a tentative. primitive conceptual scheme. which not only embraces
these phenomena. but applies in part to aspects of flicker fusion and
contrast phenomena. It is believed that this scheme is capable of

generating testable predictions.

7.

10.

ll.

BIBLIOGRAPHY

Allport. F. H. Theories of Perception. New York; John. Wiley.
1955. pp. 14-46. 'T m

 

Bartley. S. H. "The neural determination of critical flicker
frequency." 3. exp. Psychol.. 1937. 21. 6787686.

 

 

Bartley. S. H. Vision: A Study of Its Basis. New York: Van
Nostrand. 194l.fi fi fi i

Bartley. S. H. ”The features of the optic nerve discharge
underlying recurrent vision." J. exp. Psychol.. 1942. 30. 125‘
135. fi h“

 

Bartley. S. H.. and Miller. J. W. "Some circumstances sur-
rounding apparent movement in the line of regard." J. Psychol..
1954, 38. 453-456. T “T

 

Boring. E. G. Sensation and Perception in thefiHistory of Ex-
perimental Psycholog. New York: Appleton-Century-Crofts.

 

 

T942.

Calavrezo. C. "Uber den Einfluss von Grossenanderungen auf
die scheinbare Tiefe." Psychol. Forsghu 1934. 19. 311-365.

 

Cheatham. P. G. "Visual perceptual latency as a function of
stimulus brightness and contour shape." J. exp. P_sychol.. 1952.
43, 369-380. R V

 

Coules. J. "Effect of photometric brightness on judgments of
distance." J. exp. Psychol.. 1955. 50. 19-25.

 

Fry. G. A. "Depression of the activity aroused by a flash of
light by applying a second immediately afterwards to adjacent
areas of the retina." Am. J. PhysioL. 1934. 108. 701-707.

 

Gibson. J. J. "The visual perception of objective motion and
subjective motion." Psychol. Rev” 1954. 61. 304-314.

 

 

61

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

62

Graham. C. H. "Visual perception." In S. S. Stevens (Ed.).
Handbook of Experimental Psycholog. New York: John Wiley.
1951. Pp. 887-901 fl

 

Hastorf. A. H. "The influence of suggestion on the relationship
between stimulus sizes and perceived distances." J. Psychol..
1950, 29, 195-217. T ‘ “T

 

Ittelson. W. H. "Size as a cue to distance: static localization."
Amer. J. Psychol.. 1951. 64. 54-67.

 

 

Ittelson. W. H. "Size as a cue to distance; radial motion."
Amer. J. Psychol.. 1951. 64, 188-202.

 

Kilpatrick. F. P.. and Ittelson. W. H. "Three demonstrations
involving the visual perception of movemen ." J. exp. Psychol..
1951. 42, 394-402. ' T“ T T

 

Koffka. K. Principles of Gestalt Psychology. New York: Har-
court. Brace Co.. 1935.

 

Kohler. W. Gestalt Psycholog. New York: Liveright. 1947.

 

Lloyd. V. "The interaction of stimulus area and intensity as
cues in the perception of distance." J. gen. Psychol.. 1953. 49.
167-183. T I "YT

 

Motokawa. K. "Retinal traces and visual perception of move-
ment." J. exp. Psychol.. 1953. 45. 369-377.

 

Motokawa. K.. and Ebe. M. "The physiological mechanism of
apparent movemen ." Lexp. Psyctgln 1953. 45. 378-386.

 

Neff. W. S. "A critical investigation of the visual apprehension
of movement." Amer. J.fslchol.. 1936. 48. 1-42.

 

 

Newman. E. B. ”Versuche iiber des Gamma-Phenomen."
Psychol. Forsch.. 1934, 19. 102-121.

 

Osgood. C. E. Method and Theory in Experimental fsycholoa.
New York: Oxford Univ. Press. 1953, 245-248.

25.

26.

27.

28.

29.

30.

31.

32.

33.

63

Saucer. R. T. "The nature of perceptual processes." Science.
1953. 117. 556-558. “'“f'”

Saucer. R. T. "Processes of motion perception." Science.
1954. 120. 806-807.

Shipley. W. C.. Kenney. F. A.. and King. M. E. "Beta apparent
movement under binocular. monocular. and interocular stimula-
tion." Amer. J. Psychol.. 1945. 58. 545-549.

 

Smith. K. R. ”Visual apparent movement in the absence of
neural interaction." Amer. J. Psychol.. 1948. 61. 73-78.

 

Werner. H. "Studies of contour: I. Qualitative analysis."
Amer J. Psychol.. 1935. 47. 40-64.

 

Werner. H. "Motion and motion perception: a study on vicari-
ous functioning." J. Psychol.. 1945. 19. 317-327.

 

Wertheimer. M. "Experimentelle studien uber das sehen von
bewegung." E. Psychol.. 1912. 61. 161-265.

 

Wilkinson. F. R. "The organization of the visual response."
Unpublished Ph.D. dissertation. Michigan State University. 1955.

Wilkinson. F. R. "The initial perception of pulses of light."
J. Psychol.. 1957. 43. 265-268.

 

4-.

I111. Kalli. .vfllkhii

l‘. T." .~ '5“; ‘3)"