THE BEHAVIORAL CONSEQUENCES OF
ABSTRACT PATTERNS DERIVED FROM;
THREATENING SIGN STIMULUS
DISPLAYS IN HUMANS

Thesis for the Degree of M. A.
MICHIGAN STATE UNIVERSITY
GEORGE ROBERT FLEMING, JR.
(MALIK FUFUKKA)
1972

Ifirh‘

 

 

 
  

LIBRARY

Michigan State
”mm '9'

 
      

 
  

  
  

    

FL

M
USP
.5 (I.

  
  

.‘ , "gm :7
' IIIIAG a SUNS' .
800K BINDERY INC. .

ABSTRACT

THE BEHAVIORAL CONSEQUENCES OF ABSTRACT PATTERNS
DERIVED FROM THREATENING SIGN-STIMULUS
DISPLAYS IN HUMANS

BY
George Robert Fleming, Jr.
(Malik Fufukka)

The experiment tested the effect on approach
behavior of isolated abstract patterns derived from facial
threat displays. It was hypothesized that such abstract
patterns would produce a greater distance response, in that
subjects would not approach the patterns as closely as they
would to control patterns of equal stimulus value.

The experiment followed a 2 x 4 (sex x order)
between-subjects and a 2 x 10 (threatening, nonthreatening
x pattern positions) within-subjects design. Two random
stimulus orders were used to avoid subjects comparing
stimuli pairs. Each random order was also presented in
reverse order, so that four separate random orders were
possible in order to control for sequence effects. Subjects
were given instructions to approach each stimuli pattern
and stop when they felt comfortable.

Five pattern sets produced significant effects in

the direction of the hypothesis, while three pattern sets

George Robert Fleming, Jr.

did not, and two pattern sets produced significant
reversals. The lack of positive results in certain of

the sets also demonstrated the likelihood that there was

a strong relationship between threat displays and specific

areas of the face.

Accepted by the faculty of the Department of
Psychology, College of Social Science, Michigan State
University, in partial fulfillment of the requirements

for the Master of Arts degree.

 

‘44, ’1” . 4’. (1/. -’ I 5

I Directo/Ip o Thesis

Guidance Committee

 

THE BEHAVIORAL CONSEQUENCES OF ABSTRACT PATTERNS
DERIVED FROM THREATENING SIGN-STIMULUS

DISPLAYS IN HUMANS

BY

George Robert Fleming, Jr.
(Malik Fufukka)

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1972

This manuscript is dedicated to the next Black

man, woman, or child who gives their life for

the liberation of people of African descent.
Effective date 2/11/72, 12:01 a.m.

ii

ACKNOWLE DGMEN TS

The author wishes to thank the members of his
thesis committee, Dr. Andrew M. Barclay and Dr. Robert A.
Zucker, for their encouragement and assistance in the
preparation of this manuscript and throughout the past
years in the Psychology Department at Michigan State
University. A very Special note of appreciation is due
to my chairman, Dr. Joel C. Aronoff, for his concern as
an advisor and his understanding as a personal friend.
They're some all right dudes.

The assistance of Ms. Linda Stevenson in the
experiment is gratefully acknowledged. As the experi-
menter, Ms. Stevenson walked 12.5 miles in the presentation
of the stimulus cards and it would have been almost impos-

sible to complete the experiment without her assistance.

iii

ACKNOWLEDGMENTS

LIST OF TABLES

INTRODUCTION

METHOD . .

Subjects
Stimuli
Apparatus
Procedure
Observer

Debriefing

RESULTS .

DISCUSSION

REFERENCES

TABLE

OF CONTENTS

iv

Page

iii

16

16
16
19
20
21
21

23
31

37

Table

LIST OF TABLES

Page

Mean distance in inches by order, sex, and
direction of order the S's stood from
experimental and control patterns . . . . . 23

Between-subjects A x B x C interaction:
(summed over all stimulus patterns) . . . . 24

Summary of multivariate analysis of variance
on sex differences by individual pattern
sets 0 O O O O O I O O O O O O O O O O O O O 26

Summary of mean distance responses in inches
for experimental and control patterns . . . 27

 

INTRODUCTION

The process of intercommunication is certainly of
great service to many animals. To the social animals the
power of intercommunication between members of the same
community, with other species, between the opposite sexes,
and between young and old is of highest importance.
Primitive man is known to have used inarticulate cries,
gestures, and facial expressions long before he "invented"
articulate language. In contemporary man, intercommunica-
tion is generally effected by means of the voice; however
it is certain that his gestures and facial expressions are
to a certain extent mutually intelligible. Facial expres-
sions and gestures also support verbal communication, and
frequently are used in the place of verbal communication,
when appropriate. Although non-linguistic communication
has received little attention in psychological literature,
there is increasing attention being paid to the use of
non-verbal communication throughout the phylogenic scale.

Darwin (1872) was the first to contribute empirical
evidence to support the hypothesis regarding the innateness
and universality of emotional expressions. By comparing
the facial gestures of various mammals, he observed the

continuity of expression in animals and in man, the

functional nature of expression, and its role in the com-

munication of lower and higher animals. Darwin concluded

that in the course of evolution facial expressions changed

in form, through changes in facial muscles, and the original

patterns evolved into displays that conveyed a derivative

meaning, e.g., threat to the other animals. Endowed with

the function of communication they survived from one stage 'fl
of animal evolution to the next and thus were passed along

to man.

Darwin believed reflex actions were due to the
excitement of a peripheral nerve which transmits its influ-
ence to certain nerve cells, and these in turn excite
certain muscles or glands into action. All of this may
take place without any awareness of sensation or con-
sciousness on the part of the individual. These reflex
actions are in all probability subject to slight variations,
as are all instincts. If these variations were beneficial
and of sufficient importance, they would tend to be pre-
served and inherited. Thus reflex actions which were
gained for one purpose might be modified at a later time
independently of will or habit, so as to serve some distinct
purpose. For example, few individuals without practice can
voluntarily act on their grief muscles in the face (Darwin,

1872). These muscles are rarely enacted from bodily pain,

but almost exclusively from the mental distress (upper

eyelids arched, eyebrows raised, with wrinkles on the
forehead) of anticipated anxiety.

In all cases of distress, whether great or small,
Darwin believed the brain sends an order to certain facial
muscles to contract. When infants begin to cry or are
endeavoring to stop crying, they generally control all
other facial muscles more effectively than the distress
muscles. These muscles are the first to contract and the
last to relax. In like manner, those muscles of the face
which are least obedient to the will can alone betray a
slight and passing emotion without the individual being
conscious of their actions. Darwin explains this contrac-
tion under the influence of distress by the principle of
these facial muscles being least under the control of the
will, hypothesizing that they are innate.

To Lorenz (1965) instincts can be characterized by
spontaneity with innate behaviors being genetically organ-
ized responses. This does not mean that these responses
occur without external stimulation. Rather, they require
stimuli or elicitors and may also require a particular set
of antecedent conditions. However, they do not require
Specific prior training such as is true for learned
responses. It is assumed that if a behavior is innate,-
different individuals of a species will react to the same

:stimulus in very much the same way, given similar

antecedent and immediate conditions producing species--
specific behavior (Denny & Ratner, 1970).

A display is the term used to label innate actions
of animals that seem to function as signals to members of
the same Species or other Species. The message of a dis-
play is the information available to an individual as a
result of having received the display. This message speci-
fies a class of activities that the communicator may
perform at about the time of displaying, or predicts a
change in his activities. Because most messages indicate
some selection within the behavioral repetoire of the
communicator, an investigator recognizes each message by
the particular class of behavior associated with the dis-
plays that encode it. For example, aggressive behavior
consists of displays which induce the withdrawal of
intruders. These displays are called threats. A threat
consists of a display of innately understood signals which,
according to the relative strength of the threatened
animal, elicits a counterattack or flight response. A
threat, as a near fight, is common in mammals, but a fight
is rare among the same species. Usually one of the animals
responds with avoidance behavior during the encounter.

This withdrawal from attack provides a relief from tension
and is also desirable for survival.

In agreement with Lorenz, Tinbergen (1952) showed

that some facial expressions are derived from displays

which serve as communication to other animals. Tinbergen
stated that animal facial displays produce social changes
that seem to aim at one goal, the adaptation to the respon-
sive capacities of the reacting individual. He believed
that innate behavior usually responds to Sign stimuli, i.e.,
only few aspects of the total cue situation. Through the
social releasers, i.e., displays, animals seem to produce
Sign stimuli as a result of adaptation. He concluded that
displays seem always to release innate, not conditioned
responses.

Andrew (1963) went further and showed that the
communicative element in a display must be present before
natural selection could begin to shape a response into a
component display. He states that the association of a
particular expression with a high probability of attack,
for example, does not mean that the expression is caused
by a subjective feeling of anger or by the aggressive drive
in the animal. However, the animal acts as though these
internal characteristics are operating in the opponent.

In his review of the literature, Izard (1971)
reports that the study of emotional behavior in the inver-
tebrates and the lower vertebrates is almost nonexistent.
Those researchers who report emotions and emotional be-
havior in lower animals refer only to overt behavior in
two broad classes of responses, approach and avoidance

responses. The approach responses are seen as either

 

social responses and sometimes attacking responses. The
avoidance reaction is typically labeled fear behavior.
Changes in posture and locomotion are the only way in which
the facial muscles system of these lower animals partici-
pates in emotion or emotional behavior. Generally Speaking,
the facial expressions of the lower animals remains in-
capable of being understood by man.

Higher on the phylogenetic scale, are patterned
expressions in birds, 21d3., displays. There is good
evidence that the displays involving head and facial
features are used for social communication and also help
to form social orders and relationships. For example,
Hale, (1948) and Marks, Siegel, and Kramer, (1960) obtained
evidence that the head and facial features of a chicken
play a crucial role in maintaining their social dominance
order. Alteration or removal of facial features produced
a change in the dominance order among the effected chickens.
Hale debeaked white Leghorn hens and removed half or more
of their upper beak leaving a stump of l to 10 mm. The
subjects were alpha (dominant) hens of two organized flocks
of normal birds. Both hens continued to maintain their
dominant position until those hens having most contact with
the alpha bird were the first to ignore pecks delivered by
'the alpha hen. These instances of failure of subordinates
'to react in a subordinate manner to a considerable number

of pecks suggests a weakening of the peck-right relationship,

in the direction of the more loosely organized peck domi-
nance type of social organization. Although the subordinates
did not return the pecks, alteration of the hens facial
display, not its pecking power, forced the alpha hen to
peck more frequently in order to assure the usual degree of
dominance.

Similar findings were obtained by Marks, Siegle, and
Kramer when the removal of combs and wattles from subject
pullets influenced their social position among flocks with
non-dubbed chickens. The dubbed pullets were subordinate
to non-dubbed pullets. After these large flocks had been
housed for 107 days, not a single encounter was won by a
dubbed pullet over a non-dubbed bird. This produced a new
social stratification within each flock with the dubbed
pullets ranking in the lower social positions.

The alteration of facial features as Shown in Hale
(1948) and Marks, Siegel, Kramer, (1960) can limit the
effectiveness of a bird's threat display, producing an
incomplete, non-effective fighting response. It should be
pointed out that the alteration of the facial features did
not limit the subject bird's physical fighting power; how-
ever with the facial features altered, the other birds
could not respond to the subject bird's threat displays.
If dominance in birds and other animals is often determined
lay the outcome of fighting, it can be inferred that in

Iligher species displays correlated with emotions such as

anger and fear play a part in aggressive acts. Another
likely inference is that in evolution, as in the develop-
ment and maintenance of social order, emotional displays
become capable of enhancing or warding off aggression.

The ethological evidence of emotional expression
in lower primates suggests that postural activity remains
an important element of emotional behavior. In Hinde and
Rowell's (1962) careful analysis of emotional expression
in the Rhesus macaque monkey, almost all facial expressions
are described in context with particular postural or other
motor activity. For example, in the threat posture, the
mouth is open with the lower jaw slightly protruding, the
lower teeth are partly exposed, but the upper ones are
mostly covered by the upper lip. The corners of the mouth
come forward. The head iS jerked forward towards the
object being threatened, then returned to its original
position. The body is upright with the neck Slightly
stretched, the eyebrows raised and the eyes are wide open.
Sometimes the eyebrows are moved up and down, and the hair
is often raised. Threat is thus associated with attacking
and fleeing and the precise nature of the posture depends
on the Spatial relationship between the threatening animal
or object.

However, the work of Van Hooff (1967), comparing
facial displays of catarrhine monkeys and apes, concludes

that the facial displays may become somewhat more

independent of posture and locomotion in the anthropoid
apes. Van Hooff notes that facial displays in the primates
consist of a combination of facial elements in the jaws,
the lips, the tongue, the ears, the eyebrows, the upper head
skin, the eyes, and the eyelids. Apart from these are dis-
plays using postures, movements of the body, and vocaliza-
tions. Primate facial displays are related to such basic
and objectively definable social behavior patterns as
attack, approach, and flight. A tendency to attack iS
present in animals performing "the tense mouth face" and
"the Open-mouth face." This also is often the case in a
primate showing "the staring bared-teeth scream face." In
"the tense-mouth face" and sometimes in "the staring open-
mouth face," the upper eyelid is lifted, but the lower
eyelid is drawn upward as well to form a straight line.
This produces great tension in the region surrounding the
eyes, due to a contraction of the muscles controlling the
Openness of the eyes. The same tense eye display can be
seen in man when he is about to attack a rival. In the
ape, attack is displayed by staring eyes, a Slightly
hunched posture, a lowered and forward thrust head, clenched
fists, and tightly closed lips. Reviewing Van Hooff's work
one notes that "the tense-mouth face," "the staring open-
mouth face," "the frowning bared-teeth scream face" are

the displays that have a distance increasing or‘maintaining

function. Other diSplays, e.g., presenting and mounting,

10

are social releasers whose function is to decrease the
distance between individuals. The social releasers Show

a much greater variability, but are not necessarily associ-
ated with particular facial cues.

Inman the highly complex and differentiated facial
musclar system parallels a highly complex and differentiated
emotional system. To man, posture indicates something of
the intensity and nature of the felt emotion (either favor-
able or unfavorable), while a facial pattern may be char-
acteristic of a particular discrete emotion, or communication
some information with regard to an emotion mixture. Ekman
(1965) hypothesized that head cues convey information
primarily concerning the particular affect that is being
experienced (anger, sadness, joy, etc.), but provide rela-
tively little information about the intensity of the affect
or the level of arousal. Body cues, on the other hand,
reverse this pattern, communicating information primarily
about the level of arousal or the degree of intensity of
the affective experience. Ekman also believed that the
body provides few cues to what particular affect is being
experienced. In research testing these hypotheses, Ekman
used photographs of staged interview Situations which were
rated by naive judges on three dimensions: Pleasantness-
unpleasantness; Attention-rejection; Sleep-tension. The
:relative weights of the head and body in conveying infor-

Ination concerning felt emotion was tested by presenting

11

three separate components of each interview situation to

the judges. They consisted of a picture of the stimulus

person's head, the stimulus person's body up to the neck,

and the whole person. The photographs were presented in a

random order with each whole photograph and its components
representing one of the four dimensions under examination.

Each dimension was rated on a 9 point scale. For example, F
in the dimensions of Pleasantness-unpleasantness, a 1
indicates that the person in the photograph is feeling
about as unpleasant or unhappy as imaginable, and a 9
indicates maximum pleasantness.

The results of the head/whole correlations and the
body/whole correlations suggested some qualification to
Ekman's hypothesis. The ratings of the head and the whole
person on the Pleasantness-unpleasantness scale were highly
intercorrelated, while there was no observable relationship
between the judgments of body and whole person cue versions
on this dimension. Thus, the evidence suggests that little
information about Pleasantness-unpleasantness is communicated
by the body, and that the head area, i.e., facial expres-
sions provides man's communication of emotion. Also these
findings supports Van Hooff's conclusions that postural
expressions of emotion phylogenetically stopped at the
monkey, and facial expressions have become the emotional

displays that have developed innately from apes on to man.

12

In early infancy striate muscle activity in emotion
appears as a few distinct facial patterns, mainly as a cry
of distress or a smile of enjoyment. The development of
other discrete emotions comes with the maturation of the
differentiated muscles of the face. These muscles are
sufficiently developed by the end of the first year of
life for most of the discrete emotion patterns to be S
exercised (Goodenough, 1931). For example, Fantz (1961)
has produced evidence that infants can innately perceive
form and prefer the human face to other stimulus configura-
tions. If an infant consistently turns its gaze toward
some forms more often than toward others, it must be able
to perceive and distinguish among different forms. Working
on this premise Fantz developed a visual interest test with
30 infants, ages 1 to 15 weeks. Four pairs of test patterns
were presented in random sequence: horizontal strips and a
bull's eye design, a checker-board and two sizes of a plain
square, a cross and a circle, and two identical triangles.
The total time spent looking at the various pairs differed
sharply and the more complex pairs drew the greater atten-
tion. Moreover, the relative attractiveness of the two
members of a pair depended on the presence of a pattern
difference. There were strong preferences toward the
stripes and the bull's eye and toward the checker-board

and the square. Neither the cross and the circle nor the

two triangles aroused a Significant differential interest.

13

The differential response to patterns was Shown at all ages
tested, indicating that it was not the result of a learning
process. However, this does not answer the question of
whether or not the infant's innate capacity for form per-
ception introduces a measure of order and meaning into his
sensations. For an active selection process it is neces-
sary to sort out these sensations and make use of them in
behavior.

The facial pattern is the most distinctive aspect
of a person, the most reliable form distinguishing a human
being from other stimulus configurations. So a face-like
pattern would be expected to bring out selective perception
in an infant, if anything could. Fantz tested this hypoth-
esis with three flat objects, the size and shape of a head.
One was painted a stylized face in black on a pink back-
ground, on the second he rearranged the features in a
distorted pattern, and on the third was painted a solid
patch of black covering the top half of the pink object.
The features were made large enough to be perceived by the
youngest infant so acuity of vision was not a factor. The
three objects paired in all possible combinations, were
shown to 49 infants from four days to Six months old.

The results were consistent for all age levels.

The infants looked mostly at the "real" face, somewhat less
at the distorted face, and largely ignored the control

I;attern. The degree of preference for the "real" face to

14

The other face was not large, but it was consistent among
individual infants, especially the younger ones. The
experimenter suggested that there is an unlearned, primitive
meaning in the form perception of infants.

Just as evolutionary development created cognitive
functions and abilities in man, so it created in him also
unique emotions and emotional functions. Tomkins (1962,
1963) divided emotions into positive and negative types.
The positive emotions tend to elicit behaviors which sus-
tain them, because they are rewarding experiences. The
negative emotions; i.e., fear, rage, and anxiety, tend to
elicit behaviors which eliminate them because they are
punishing or discomforting experiences. Darwin (1872)
anticipated Tomkin's proposal when he observed that if we
expect to suffer then we are anxious. It can then be
inferred that when the threat of agression is communicated
to an individual through facial displays or expressions,
the innate response is an anxiety reaction, and the indivi-
dual should therefore react accordingly. It is also
assumed that if a behavior is innate, different individuals
of a species will respond to the same stimulus in very much
the same way, because from the point of view of the evolu-
tionary process, the recognition of the threat of aggression
assures survival.

It is therefore plausible to expect that there may

be specific facial sign-stimuli that human beings

15

selectively interpret as threatening, and that these dis-

plays lead to real behavioral consequences. The present

study attempts to test this proposition by examining the
degree of actual approach behavior to specific patterns
that have been noted, (Aronoff and Barclay, 1971) as indi-
cating threat in human beings. This research tests the
hypothesis that:

1. People respond selectively to certain kinds of patterns.
When these patterns are perceived as abstract displays
people will respond to them differentially.

2. People will respond differentially to abstract displays
of threat by standing further back from them than from
a neutral display of equal stimulus value, under test
conditions of approaching these stimuli from a distance
that is beyond the range of effective social inter-

action.

METHOD

Subjects

Subjects consisted of 64 undergraduate students
(30 males, 34 females), enrolled in the introductory psy-
chology course at Michigan State University. They were
recruited by the experimenter who spoke to these classes
and asked students for their cooperation. Those who vol-
unteered to participate were asked to Sign a form giving
their name, address, and phone number. Subjects were
telephoned two days prior to the experiment and asked to
participate. If they agreed they were assigned a Specific

time to participate in the study.

Stimuli

The stimuli used in the study were derived from the
patterns noted by Aronoff and Barclay (1971) that dis-
tinguished between threatening and non-threatening facial
represensations collected from more than two hundred sub-
jects (from Cornell University and Michigan State
University). Their subjects were asked to imagine that
they were New Guinea head hunters and to draw a war and
courtship mask. As it has been frequently maintained

(La Barre, 1947; Birdwhistle, 1970) that the representation

16

 

 

 

17

of emotion is determined by individual cultural norms.

La Barre and Birdwhistle might argue that the subjects'
reproduction of the mask was a culturally determined
American image and, therefore, the results cannot be gen-
eralized to other cultures. However, the characteristics
that distinguish between threatening (Pic. A) and non-
threatening (Pic. B) facial displays have been found to

appear in a Similar way cross-culturally. Stevenson (1972)

examined masks from nine cultures and an additional mixed
sample of nine other cultures and found that threatening
masks possessed Significantly more of these characteristics
than non-threatening masks. Ekman (1971) also reports very
strong evidence that similar facial expressions are uni-
versally associated with the same specific emotions.

The stimuli chosen for this experiment have been
shown by Stevenson to distinguish between threatening and
non-threatening masks. From the arrangement of stimuli
present indicated by Stevenson, the experimenter selected
the stimuli for this study in accordance with the criteria
that: l) the threatening stimuli (Pic. A) were abstract
and not obviously aggressive, e.g., not bared teeth, etc.;
and, 2) a meaningful control pattern of equal stimulus
value existed in the non-threatening display. See Figure 1
for a representation of the stimulus displays that were
used in the experiment. Pictures 1-4 represent the four

warm up cards that preceeded each trial.

 

 

18

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>(I AIS.
V Warm up Stimuli
l 2 3 4
A Set B A Set 2 B
54 00 %¢ QQ‘
A Set B A ‘Set 4 B
\\// be:
\/ U
A Set B A Set6 B
TAA m
WKR 7:9
A Set B A Set 8 B
w. a; m rm
A Set B A Set 10 B
mm m1 <30

 

 

 

 

 

 

 

 

 

 

Figure l

 

 

19

Apparatus

 

The experimental room measured 38 ft. x 17 ft. and
was evenly lit by fluorescent lighting overhead. The
stimuli were placed on a 1 inch Shelf 5 ft. 2 in. above
the floor on the wall at one end of the room. On the floor
extending out from the stimuli was a walkway 20 ft. long
outlined on the left Side with black tape. The Stimuli
were placed at a point on the shelf that centered on the
walkway.

Against the right wall were a row of classroom
tables (5 ft. x 2 ft. 6 in.) lined up lengthwise from the
right front corner of the room, so as to be parallel to
the walkway and touching the front and right walls. Under
the tables on the right wall, the walkway was marked with
80 strips of black tape 3 inches apart and 1-1/2 inches up
from the floor. This series of tape marks formed the
scale of distance from the wall. The tables, which were
the right side boundary of the walkway served the purpose
of concealing the scale of distance marks from the subjects.

At the opposite end of the room near the door was
set up a table and chair where the experimenter sat during

'the experiment.

20

Procedure

 

The experiment followed a 2 x 4 (sex x order)
between-subject and a 2 x 10 (threatening, non-threatening
x pattern position) within-subject design. Two random
orders of presenting the stimuli were used to avoid sub-
jects comparing the stimuli pairs. Each random order was
also presented in reverse order, so that four separate
random orders were possible in order to control for
sequence effects. The subjects were divided on the basis
of sex and assigned an order as they showed up for the
experiment. Within the design (2 x 10) the stimuli were
being divided into a threatening and non-threatening pair
for each of the 10 categories. Thus, eight groups and
eight replications of the same experiment were produced.
The design is such that the experimenter desired equal N's
for each group (per condition) but unfortunately an error
in assignment let to unequal cells. There were an addi-
tional four stimuli cards (See Figure 1), which were used
as a warm up and were always presented in the same order
before each of the four experimental orders.

The subjects entered the room and were told to
stand at the top of the walkway facing the opposite end of
the room from where the stimuli were presented. At this
time the experimenter would ask if the subjects wore
glasses. If the response was yes, the experimenter would

indicate that they may be needed during the experiment.

 

21

The subjects were told this was to be a memory experiment
and required consentration. Next the subject was told:

In this experiment I shall be putting a series of
designs on the far back wall. You will see each one
individually and approach them as you wish. When you
have reached a spot at which you feel comfortable,
stOp and indicate to me that you have found that place
by saying "here" or "now." Then return to the starting
point. You shall always face the opposite end of the
room while I put up the design. When I return you may
turn around to see the next design. Any questions?

. . . Now begin.

After placing each stimuli card on the shelf, the
experimenter, who was unaware of the purpose of the experi-
ment would return to his seat out of view of the subject

and tell the subject to start.

Observer

 

Each stuject was observed through a one-way mirror
during the walking responses by an observer who was trained
in the measuring procedures necessary for the study. The
observer noted the subject's distance from the stimulus
card on each of the subjects' approach reSponses. Distance
was measured from the wall to the back of the subjects'

heel that was furthest from the wall.

Debriefing

 

At the end of the experiment each subject was
debriefed; i.e., given a Short explanation of what had

occurred and told where they could obtain more information

22

if desired. Subjects were then thanked for their partici-
pation and warned to maintain secrecy in order not to

jeopardize the results of the study.

RESULTS

Five pattern sets produced significant effects in
the direction of the hypothesis (1, 2, 3, 5, 7); while
three pattern sets (4, 8, 10) did not, and pattern sets
6 and 9 produced significant reversals. A test of a pos-
sible interaction between order, sex, and direction over
the 10 sets of patterns is presented in Table 1. An
analysis of these data showed that no significant inter-
action was present between these variables. The overall

summary of the analysis of variance is shown in Table 2.

Table 1. Mean distance in inches by order, sex, and
direction of order the S's stood from experi-
mental and control patterns.

 

 

 

 

 

 

 

 

 

Order 1 Order 2
Pattern Up Down Up Down
*
M<6> Fae) M<8> Fm) M<8> F<8> Mm) Fm)
A 101.9 133.3 129.9 117.6 144.3 122.4 117.7 128.5
B 100.2 140.1 132.0 114.9 142.2 124.0 113.9 127.7

 

Note: ( )* = N's for each cell.

23

24

Table 2 Showed a trend in the direction of a 3 way
interaction (F = 3.15, p < .08). An inspection of the
means in Table 1 revealed that most of the variation

occurred for males in order 1.

Table 2. Between-subjects A x B x C interaction:
(summed over all stimulus patterns).

 

 

Source MS F Ratio P Value
Order (A) 254,666.91 .30 .59
Up/down (B) 67,218.37 .07 .78
Sex (C) 69,639.68 .08 .78
A x B 404,250.33 .47 .49
A x C 323,064.43 .38 .54
B x C 127,003.72 .15 .70
A x B x C 2,691,952.90 3.15 .08

 

[F Ratio = (df(l,58))]

Thus since the above analysis indicated that order
and direction can be effectively ignored and since the
small N's used to determine the interaction yield effects,
which should be considered unstable, the small N's were
combined in order to provide more stable effects. These
combined data were then analyzed by a multivariate analysis
of variance for the sex differences and stimulus pattern
effects alone. In this analysis, the cells were thus
collapsed into a one by two within (male vs. female) over

the 10 sets of patterns, each set consisting of an

25

experimental and control stimulus which were treated as
repeated measures. The summary of this analysis iS pre-
sented in Table 3.

The stimuli chosen for Set_1 were placed on the
upper third of the stimulus card and are representative of
eyes (See Figure 1). The right triangles presented in the
experimental stimulus were selected to test high facial
angularity. Subjects stood further from the threatening
(a) stimulus card than they did from the non-threatening
card (B) (A distance = 142.6, B distance = 128.0; See
Table 4). There were no main effects for sex or inter-
action. These findings were Significantly different. All
distances throughout are in inches.

§2E_2 was also placed on the upper third of the
stimulus card, and represented eyebrow configurations. The
experimental stimulus tested the dimension which corresponds
to high facial angularity. The findings are represented by
the distance totals of 130.5 for pattern A and 120.2 for
pattern B, which were significantly different. However, a
significant Sex x Pattern interaction revealed the males on
Set 2 followed the theoretical hypothesis (distance for pat-
tern A = 134.6, while pattern B = 112.7) (F = 5.38, P < .05;
1,61 df), although the female mean values did not (pattern
A = 126.4 and pattern B = 127.7, non-Significant).

The stimulus patterns for §2E_2 were placed on the

middle third of the stimulus card and represented facial

26

Table 3. Summary of multivariate analysis of variance on
sex differences by individual pattern sets.

 

 

Pattern Sex x Pattern Pattern Main
Set Interaction Effect
1 .09 3.09**
2 3.90** 3.06**
3 .46 5.50***
4 .62 .08
5 2.00* .64
6 .66 14.60***(l)
7 .27 1.92*
8 1.65 .01
9 .14 9.86***(l)
10 .17 .29

 

[F Ratio = (df = (1.64))]

Note - (1) This is a Significant effect due to a
:reversal (see text for discussion).

- Individual pattern set comparisons were per
iiormed by computing E values for the main effects, a pro-
cxedure suggested by winer (1962, p. 208).

*p £ .1
**p < .05

***p s .01

27

Table 4. Summary of mean distance responses in inches for
experimental and control patterns.

 

 

Set V Pattern Males Females Total
1 A 134.1 151.1 142.6
B 117.1 138.9 128.0

2 A 134.6 126.4 130.5
B 112.7 127.7 120.2

3 A 131.5 130.7 131.1
B 112.1 120.0 116.0

4 A 126.3 125.2 125.8
B 120.0 128.2 124.1

5 A 123.8 131.8 127.8
B 127.7 117.7 122.7

6 A 117.5 116.6 117.0
B 139.3 150.2 144.7

7 A 138.8 123.5 131.2
B 125.7 117.6 121.7

8 A 94.4 114.8 104.6
B 104.3 106.1 105.2

9 A 122.3 107.2 114.8
B 144.6 135.5 140.1

10 A 125.3 132.1 128.7
B 131.9 132.4

133.0

 

28

tension in the area of the cheeks. The data revealed a
significant effect between the patterns with mean totals
of 131.1 for pattern A and 116.0 for pattern B. There
were no Significant main effects for sex; and there were
no interactions.

§2E_4 symbolized a pointed chin configuration.
The experimental stimulus tested low facial angularity.
The data showed no significant differences between the
distance totals for pattern A (125.8) and pattern B (124.1).
There were also no Significant main effects or sex inter-
action presented.

§e£_§_was another stimulus pattern placed on the
middle third of the stimulus card. The mean distance total
for pattern A was 127.8 and 122.7 for pattern B. The
experimental stimulus tested mid-facial tension, as did
Set 3. These differences were significant overall. How-
ever, a significant Sex x Pattern interaction (F = 2.00,
p < .1) showed a slight non-significant reversal of the
predicted effect (pattern A = 123.8 and pattern B = 127.7)
while females responded in the direction of the hypothesis
(pattern A = 131.8, and pattern B = 117.7; (F = 3.69,

P < .08; 1,61 df) a Significant difference).

 

Set 6 stimulus patterns were placed on the upper
third of the stimulus card. The experimental stimulus, a
configuration of horns, tested a non-human Sign stimulus

for a human threat response. The mean distance totals

29

were 117.0 for pattern A and 144.7 for pattern B. These
findings reversed the predicted hypothesis and Showed no
sex interactions.

The stimulus patterns for Set_z_were placed on the
upper third of the stimulus card and represented an erec-
tion of hair on the head. The experimental stimulus tested
whether erect hair would be perceived differently from an
inversion of the same stimulus. The mean totals were 131.2
for pattern A and 121.7 for pattern B. These means were
Significantly different with no (sex) main effects or inter-
action.

Set 8's stimulus patterns were also represented by
hair on the head. Mean distance totals of 104.6 for pat-
tern A and 105.2 for pattern B indicated these findings to
be non-significant. Here too, males responded differently
(pattern A = 94.4 and pattern B = 104.3) (F = 3.16, P < .09;
1,61 df) than did females (pattern A = 114.8 and pattern
B = 106.1).

Set 9's stimulus patterns were placed on the upper
third of the stimulus card. The experiment stimulus tested
forehead-tension and represented forehead wrinkles. The
mean totals reversed the predicted effect, and were 140.1
for pattern A and 114.8 for pattern B. There were no sex
interactions present in Set 9.

The stimulus patterns chosen for Set 10 were

placed on the upper third of the stimulus card and were

i? m .M‘. (‘5
1”...” _
AI

30

representative of eyes. The mean distance totals computed
128.7 for pattern A and 132.4 for pattern B. The experi-
mental stimulus tested high facial angularity. These
findings were non-significant with no main effects for

sex or interaction.

DISCUSSION

Of the two experimental hypotheses, hypothesis two
was the major hypothesis tested. Hypothesis two stated
that subjects would respond differentially to abstract
displays of threat by standing further back from them than
from a neutral display of equal stimulus value, under test
conditions of approaching these stimuli from a distance
that is beyond the range of effective social interaction.
Built into the design of the experiment was a strong test
of the hypotheses. The use of four different random orders
minimized the possibility that the subjects would compare
the experimental pattern with its control. This procedure
produced significant scores with the use of subnormal
stimuli that represented real facial configurations and
was therefore a powerful and conservative test of the
hypotheses. The results showed that the two hypotheses
were, to a great extent, upheld and thus were, in agreement
with their theoretical expectation.

The results demonstrated that there was a strong,
positive relationship between threat and various areas of
the face, as noted from Aronoff and Barclay (1971). In
general, contractions in the striate muscle system are

present in emotional activity and these contractions form

31

r‘I

32

lines of expression that are peculiar to an individual
emotion. The results also indicated that where angularity
and diagonality were designated to represent striate

muscle activity, i.e., facial tension, subjects produced
significant desired effects in the pattern sets 1, 2, 3,

5, and 7. These main findings are in the upper half of

the face (from the cheek lines up) and are highly congruent
with the early psychological findings of Busby (1924) and
Hanawalt (1944). Busby made an analysis of facial expres-
sions based only on the mouth, eye, and eyebrow and con-
cluded the upper parts of the face are more important for
correct judgment of facial expression than is the mouth.
Hanawalt studied the upper and lower parts of the face in
determining their expressive meaning and found that the
upper half of the face is most effective in eliciting fear
and surprise responses. In this experiment the greatest
significant differences from the pattern sets were found

in patterns 1, 2, 3, 5, and 7 which were representative

of eyes, eyebrows, cheek lines, and erect hair respectively.
Stevenson (1972), in using Aronoff and Barclay's scale with
threatening and non-threatening masks from nine cultures,
found the angularity and diagonality of experimental

(Pic. A) patterns 1, 2, 3, 5, and 7 present in the threat-
‘ening masks. The facial tension that accompanies a threat
response seems to converge and be most visible in the upper

third of the face and extend down to lower areas of the

 

 

 

 

 

 

33

cheeks to produce the angularity and diagonality accom-
paning tension there. Thus, the position of the pattern
is important in determining the effectiveness of the
angularity and diagonality in the lines.

Pattern sets 4, 8, and 10 yielded nonsignificant
differences between the threatening and control patterns.

Pattern 4 was a pointed chin configuration, that tested

.2.;

“——

t2:

low facial angularity. If it was simply a question of
angularity and diagonality versus the absence of these
properties, then pattern set 4 would be the best test of
the strength of these two elements. However, pattern
Set 4 produced nonsignificant scores. Two strong reasons
for these results can be considered; either an inappro-
priate stimulus and/or control was chosen, or the placement
(position) of the stimulus was the determining factor. It
is unlikely that the control pattern was inappropriate,
as the pattern sets discussed above contrast strong angu-
larity and diagonality against curvilinearity and leads
to strong positive results. However, according to the
findings of Busby and Hanawalt, the latter reason would be
true. Stevenson's work also supports the latter reason.
She also obtained a nonsignificant score and a slight
reversal with the pointed chin configuration.

The experimental stimulus in pattern 10 represented

the diagonality of eyes without angularity and was a

34

pattern from the upper third of the face. The nonsignifi-
cant results in this pattern set indicated that position
and diagonality alone did not produce the predicted
results. It seems that the combination of all three
(position, angularity, and diagonality) is necessary for

a facial pattern to be seen and responded to as threat-
ening.

Pattern 8 represented hair on the top of the head

“—4;

I

and tested whether messed and disorderly hair would be I
perceived as threatening. This, too, produced nonsignifi-
cant results. Here, too, it is possible that both the
experimental patterns were inappropriately chosen. While
it is not fully clear why this pattern did not produce the
expected result, it Should be noted that it is the pattern
most easily interpreted and anthropomorphized. It is
likely that when a stimulus pattern is easily anthropo-
morphized it cannot be considered a sign stimulus for
threat, for it may encompass several stimulus properties,
i.e., cues, and therefore not be responded to as if it
were a Single Sign stimulus.

Upon reexamination and analysis of the material in
this experiment, it is concluded that the reversals pres-
ent in pattern set 6 and pattern set 9 were partly due to
errors in the selection of an apprOpriate control. It
appears this may have been a limitation in the experiment.

The configuration of horns (set 6) tested a non-human sign

35

stimulus for a human threat response. The horn configura-
tion was present in the masks that Aronoff and Barclay
studied to obtain their patterns of threat, and StevensOn
reports the presence of horns on the threatening masks of
the nine cultures She studied. However, when tested as a
human facial characteristic it did not yield significant
results. If man did not have horns, then horns should not
have evolved as a threatening stimulus. Therefore, it is
reasonable that horns would not show a significant score
on a test of human Sign stimuli of threat.

Izard (1971) indicates that, on a post hoc basis,
the experimental pattern in set 9 was a wrong selection.
Izard presents evidence that the forehead lines in this
particular pattern are part of the facial configuration of
fear; an obviously non-threatening stimulus. It is unfor-
tunate that the results of Izard's work were not available
for consultation when the study was designed.

At present there is not sufficient information
available to explain the sex interaction present in sets
2, 5, and 8. Future research is needed to clarify this
issue.

The overall results of this experiment have shown
that the subjects chose to selectively focus on certain
patterns randomly presented to them and produce a behavioral
consequences of avoidance, which was the experimenters

predicted effect. In the debriefing sessions not one of

36

the subjects recognized what the desired effect was or how
the stimuli elicited their responses. The subjects left
the experiment assuming they had just participated in a
memory experiment. Such Significant differences from
approach and avoidance behavior where threat from subnormal
Sign stimuli was indicated, together with Stevenson's

data using the criterion of universality, supports the
existence of an innate recognition of threat that is useful
for man's survival.

In man, facial expression and gestures support
verbal communication, and frequently are used in the place
of verbal communication. This experiment was a test of
one aspect of nonverbal communication. Whether communica-
tion is total or partial without its nonverbal element is
a question for future research; however, the existence of
an evolutionary history of lines of tension from muscle
postures on man's face that are out of his awareness was
explored. With the use of his facial tension, man communi-
cates words and meanings that never get verbalized and are
seemingly understood within the complexity of everyday

face to face interaction.

 

REFERENCES

REFERENCES

Altman, S. A. Social communication among primates.
Chicago: University of Chicago Press, 1967.

 

Andrew, R. J. "Evolution of facial expression." Science,
1963, 142, pp. 1034-1041.

Andrew, R. J. "The origins of facial expressions."
Scientific American, 1965, 213, pp. 88-94.

 

Aronoff, J. C., and Barclay, A. M. "A scale for scoring
sign-stimulus characteristics in the representation
of threatening facial expressions." unpublished
technical report, 1971.

Birdwhistell, R. Kinesics and Context. Philadelphia:
University of Pennsylvania Press, 1970.

Busby, D. E. "The interpretation of facial expressions."
American Journal of Psychology, 1924, 32,
pp2—602-604.

Candland, D. K. "Discriminability of facial regions used
by the domestic chicken in maintaining the social
dominance order." Journal of comparative and
Physiological Psychology, 1969, 69, pp. 281-285.

Cline, M. G. "The Influence of Social Context and the
Perception of Faces." Journal of Personality,
1956, 25, p. 142.

Darwin, C. The expression of the emotions in man and
animals. London: John Murray, 1904, (First ed.,
.1872).

 

Ditman, J., and Bywoff, A. "facial and bodily expression."
Psychiatry, 1965, 28, p. 239.

 

Drag, R. M., and Shaw, M. E. "Factors influencing the
communication of emotional intent by facial expres-
sions." Psychonomic Science, 1967, 8(4):

 

pp. 137-1382

37

38

Ekman, P. "Body position, facial expression and verbal
behavior during interviews." Journal of Personality
and Social Psychology, 1964, 68(3), pp. 295-301

 

 

. "Differential communication of affect by head
and body cues." Journal of Personality and Social
Psychology, 1965, 2, pp. 726-735.

 

 

. "Universals and cultural differences in facial

expressions of emotion." In Cole, J. (ed).

Nebraska Symposium on Motivation. University of

Nebraska Press, 1972. r:

 

, and Friesen, W. V. "Head and body cues in the
judgment of emotion: a reformulation." Perceptual
Motor Skills, 1967, 24, pp. 711-724.

 

 

, and Friesen, W. V. "Constants across cultures in
the face of emotion." Journal of Personality and
Social Psychology, 1971, 17(2), pp. 124-129.

 

, Sorenson, E. R., and Friesen, W. V. "Pan-
cultural elements in facial displays of emotion."
Science, 1969, 164, pp. 86-88.

Fantz, R. "The origin of form perception." Scientific
American, 1961, 204, pp. 66-72.

 

Frijda, N. H. "Facial expression and situational cues."
Journal of Abnormal and Social Psychology, 1958,
'52, pp. 149-153.

, and Philipszoon, E. "Dimensions of recognition
of expression." Journal of Abnormal and Social
Psychology, 1963, 66(1), pp. 45-51.

 

 

Frois-Wittman, J. "The judgement of facial expression."
Journal of Experimental Psychology, 1942, 272,
pp. 5-19.

 

Gaydos, H. F. "Intersensory transfer in the discrimination
of form." American Journal of Psychology, 1956,
69, pp. 1073110.

Gibson, J. J. "Pictures, perspective, and perception."
Daedauls, Winter, 1960.

Goodenough, F. L. "The expression of the emotions in
infancy." Child Development, 1931, 2, pp. 96-101.

 

39

Guhl, A. M. "Peck-orders established in spite of debeaking."
American Poultry Journal, 1956, 20, p. 4.

 

Hale, E. B. "Observation on the social behavior of hens
following debeaking." Poultry Science, 1948, 21,
pp. 591-592.

 

Hall, E. T. The Silent Language. New York: Fawcett

4

Publications, Inc., 1959.

. The Hidden Dimension. New York: Anchor Books,
Doubleday & Co., Inc., 1969.

 

Hanawalt, N. G. "The role of the upper and lower parts of
the face as a basis for judging facial expressions:
II In posed expressions and 'candid camera'
pictures." Journal of General Psychology, 1944,
31, pp. 23-36. '2'

 

Harrison, R. P. "Pictic analysis toward a vocabulary and
syntax for the pictoral code." unpublished Ph.D.
dissertation, Michigan State University, 1964.

Haynes, H., White, B. L., and Held, R. "Visual accommoda-
tion in human infants." Science, 1965, 148,

pp. 528-530.

Hinde, R. A., and Rowell, T. E. "Communication by postures
and facial expressions in the rhesus monkey."
Proceedings of the Zoological Society of London,
1962, 138, pp. 1-21.

Izard, C. E. The face of Emotion. New York: Meredith
Corp., 1971.

 

, and Tomkins, S. S. "Affect and behavior:
Anxiety as a negative affect." C. D. Spielberger
(Ed.) Anxiety and behavior. New York: Academic
Press,_1966, pp. 81-125.

 

La Barre, W. "The cultural basis of emotions and gestures."
Journal of Personality, 1947, 16, pp. 49-68.

Lewin, K., Lippitt, R., and White, R. "Patterns of aggres-
sive behavior in experimentally created social
climates." Journal of Social Psychology, 1939,

10, pp. 271-299.

Lorenz, K. Evolution and Modification of Behavior. Chicago:

University of Chicago Press, 1965.

5*:

_.£_A£

‘12::

 

40

. On Agression. New York: Harcourt, Brace &
World, 1966.

 

. "Preface to C. Darwin," The expression of the
emotions in man and animals. Chicago: University
of Chicago Press, 1965.

 

McDougall, W. An Introduction to Social Psychology.
London: Methuen Co., 1923}

Marks, H. L., Siegel, P.B., and Kramer, C. Y. "Effect of
comb and wattle removal on social organization of
mixed flocks of chickens." Animal Behavior, 1960,
8, pp. 192-196.

Marler, P. "Communication in monkeys and apes." In
Devore (Ed.) Primate behavior: Field studies of
monkeys and apes. New York: Holt, Rinehart &
Winston, 1965, pp. 544-584.

 

Mordkoff, C. "A factor analysis study of the judgement of

 

 

emotion from facial expression." Journal of
Experimental Research in Personality, 1967, 2,
pp. 80-85.

Penny, M. A., and Ratner, S. C. Comparative Psychology
Research in Animal Behavior. Homewood, Ill.:
The Dorsey Press, 1970.

Sade, D. C. "Determinants of dominance in a group of free-
ranging rhesus monkeys." In S. A. Altman (Ed.)
Social communication among primates. Chicago:
University of Chicago Press, 1967, pp. 99-113.

 

Smith, W. J. "Messages of vertebrate communication."
Science, 1969, 165, pp. 145-150.

Sommer, R. Studies in personal space." Sociometry, 1959,
22.
. "The distance for comfortable conversation: a

further study." Sociometry, 1962, 25.

Stevenson, L. "A cross-cultural study of threatening sign-
stimulus displays." unpublished M.A. thesis,
Michigan State University, 1972.

Timbergen, N. "Derived activities: Their causation,
biological significance, origin, and emancipation
during evolution." The Quarterly Review of Biology,
1952, 21, pp. 1-26.

 

41

Tomkins, S. S. Affect, imagery, consciousness. Vol. I.
The positiVe affects. New York: Springer Co., 1962.

. Affect, imagery, consciousness. Vol. II.
The negative affects. New Yofk: Springer Co., 1963.

Tormey, A. The concept of expression. Princeton, N.J.:
Princeton University Press, 1971.

Van Hooff, J. A. "Facial expressions in higher primates."
Symposia of the Zoological Society of London, 1963,
10, pp. 103-104.

. "Facial displays of catarrhine monkeys and apes."
D. Morris (Ed.), Primate ethology. London:
Weidenfeld & Nicolson, I967.

Winer, B. J. Statistical Principles in Experimental Design.
New Yofk? McGraw-HilI Co., 1962.

J ' ' F... in 4'