ABSTRACT
SUBCEPTION AS A VERIDICAL PROCESS AND PERCLPTUAL DnEENSE
AS AVOIDANT AND VIGILANT BEHAVIOR

by Raymond J. Clausman

This study focused upon two major questions that have arisen from
the "New Look“ approach to perception. First, whether subception as a
veridical process can occur when there is no awareness of the stimulus
presentations, and second, how to account for both avoidant and vigilant
behavior as a response to threatening stimuli. The two major objectives
were: (a) to test for subception using a verbal accuracy indicator of
discrimination when the stimuli are exposed too briefly to permit a
phenomenological report of awareness, (b) to extend the concept of per-
ceptual defense to include both avoidant and vigilant behavior.

The experiment was divided into three phases, the defense assess-
ment, subception, and perceptual defense phase. In the defense assess—
ment phase, 55 were administered a ranking task used in conjunction with
the Blacky Pictures, from which defenses to various conflict dimensions
were inferred° Four or more first place ranks of either Avoidance or
Projection as a defense plus total rank consistency served as the two
criteria in selecting 88 for the perceptual phases. On the basis of the
selection criteria, 22 Ss were placed in an Avoidance Group and 18 Ss in
a Projection Group. A random sample of 25 85 who showed little defense
consistency were placed in a No Defense Group.

A two-choice discrimination task was used in the perception phases

Raymond J. Clausman

of the study. The perceptual stimuli were arranged in three pairs and con-
sisted of a sexmneutral, aggression-neutral, and neutral-neutral pair.

In the subception phase, it was hypothesized that all 55 would show a
greater response accuracy to the conflict stimuli as compared to the neutral
stimuli. The exposure settings for presenting the stimuli at a level below
phenomenological awareness were determined individually for each S. 83 were
given a DO trial discrimination series using the conflictwneutral pairs.

In the perceptual defense phase, it was hypothesized that the Avoid-
ance Group would show an avoidant response to die conflict stimuli; the
Projection Group would show a vigilant response; and the No Defense Group
would show no differential response to the conflict and neutral stimuli.
In this phase, the exposure setting was increased to a level, previously
determined by a pilot study9 at which approximately 50% correct discrimi-
nations of the neutralmneutral pair had been made. This exposure setting
remained constant for all 83. A 20 trial acuity test series was given,
using the neutral pair. This was followed by a hO trial perceptual defense
series using the conflict-neutral stimuli.

The data were subjected to both an inter-category and interugroup
analysis. The results failed to support the subception hypothesis. It was
concluded that perception as a veridical process does not operate when
the stimulus exposure is to brief to permit a phenomenological report of
awareness. The results of the perceptual defense phase supported the
view that avoidance and vigilance both serve as a defense in perception,
and that the type of defensive behavior utilized can be predicted from

a knowledge of 85‘ inferred ego defense mechanisms.

Raymond J. Clausman

It was suggested that there was a greater need for a more precise
definition of awareness threshold, for further study of the autcnomic
response as an indicator of perception, and for a study of the percep-

tual learning process within the experimental setting.

 

//// “31.79%le

SUBCEPTION AS A VhRIDICAL PROCESS AND PLRCRPTUAL

DEFEKSE AS AVOIDANT AND VIGILANT BEHAVIOR

By

{‘2‘
v

Raymond J. Clausman

A THESIS

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

DOCTOR OF PHILOSOPHY
Department of Psychology

1960

G15”?
lulu/W

A C K N O W L E D G M E N T

I am pleased to acknowledge my indebtedness to Dr. Albert I.
Rabin, chairman of my doctoral committee, for his guidance in this
study, and to the other members of the committee, Dr. M. Ray Denny,
Dr. Charles C. Hanley, and Dr. Gerald F. King, for their helpful
comments and suggestions.

I am also indebted to Dr. S. H. Bartley for taking the illumim
nosity readings in the perceptual phase of due experiment and to my
wife, Norma Clausman, for making the free hand drawings of the per~
ceptual stimuli. I must also express my appreciation to the teaching
assisstants who made their students available as subjects and to the

students themselves for their cooperation.

ABLE OF CONTENTS

LIST OF TABIES AND FIGURES. . . . . .

Chapter

I. INTRODUCTION . . . . . .

Perceptual sensitization .
Tabled word frequencies.
Discussion . . . .

Subception . . . . . .
Discussion . . . . . .

Perceptual defense . . . . .

Recent trends in perceptual research . . . . . . . .
Review of the experimental literature. . . . .

O O 0 e 0 0 O O

O O O O O O O O O O

O O O O O O O O O

Personality-centered approach. . . . . . . . . .
Psychoanalytics concepts of defense. . . . . . .
Perceptual defense as evoidant and vigilant

behavior . . . . . .
Discussion . . . . . . .

Purpose of present investigation

Statement of hypotheses. .

II 0 mmOD O Q I O O O O O O O O O 0

Selection of defense measure
A pilot study

Subjects

materials and apparatus
measures

Procedure

III. RESULTS. . . . . . . . . . . . .

IV. DISCUSSION . . . . . . . . . .

V. SUNLARY AND CONCLUSIONS. . . . .

REFERENCES .

APPENDICES.

O O O O C O O O O a O O O 0

iii

and

perceptual stimuli

101.
1.15
119

126

LIST OF TABLES AND FIGURES
Table
1. Summary of Operational Design. . . . . . . . . . . . . .

2. Error hhans and mean Error Differences between Stimulus
Categories for all Groups in the Subception Series . . .

3. Error means and mean Error Differences to the Sex and

Aggression Stimuli in the Subception Series. . . . . . .

h. Total Correct Response Thane for the Three Groups in

the Subception Series. . . . . . . . . . . . . . . . . .

5. Analysis of Covariance of Correct Responses for the
Three Groups in the Perceptual Defense Series. . . . . .

6. Comparison of Avoidance and Projection Group Maan
Accurary Scores on the Perceptual Dafense Series with
the No Defense Group 0 O O O O O O O O O O O O O O O O O

7. Analysis of Variance of Error Scores to the Neutral
Stimulus Presentations in the Perceptual Defense
Series . . . . . . . . . . . . . . . . . . . . . . . . .

8. Error means and Mean Error Differences to the Neutral
and Conflict Stimuli for the Three Groups in the
Perceptual Defense Series. . . . . . . . . . . . . . . .

9. Error Means and Mean Error Differences in Response to
the Sex and Aggression Stimuli in the Perceptual
Defense Series . . . . . . . . . . . . . . . . . . . . .

lO. Correct Response Means and Mean Differences between
male and Female Subjects in the Perceptual Defense
Series . . . . . . . . . . . . . . . . . . . . . . .

Figure

1. Group Error means for Successive Trial Blocks in the

Perceptual Defense Series. . . . . . . . . . . . . . . . .

Page
8h

86

88

89

92

93

95

97

99

103

101

INTRODUCTION

RECENT {ENDS IN PERCEPT’AL RESEARCH

Traditionally, the approaches used in the study of perception have
been of two kinds. One approach seeks to explain perceptual experience
through the properties of the stimulus field. In the second approach,
the emphasis has been upon the operations of the sense organs and
associated neural structures which mediate the impinging physical
stimuli. Within the last decade, however, there has emerged an ever
increasing trend to treat perception in a broader behavioral contex+.
This so called "New look" or "Functionalist" approach to perception is
summarized by Bruner and Postman (l9h9) in the following manner:

"...its primary concern is with the manner in which porn

ceptual functioning is imbedded in and interacts with other

forms of psychological functioning...for a full understand-

ing of the perceptual process it is necessary to vary not

only the physical stimulus and the sensory conditions of

the organism, but also the central conditions-~motives,pre~

dispositions, past learnings—awhich have remained outside the

formal limits of the perceptual system."

The initial experimental studies using the "New Look" approach
stressed the importance of motivational states created by such factors
as physiological needs or induced stresses as determinates in percep~
tion. Critical reviews of these early experiments have been presented
elsewhere by Bruner (1951a). Allport (1955), and Bartley (1957). The
initial experiments dealt with a broad motivational state which was

assumed to be more or less uniform for all individuals.

' l

more recently, the trend has been to consider the idiosyncratic
motives or needs of subjects in the study of perception. As Jenkins
(1957) in a recent review of the literature so aptly states:

"This enterprising step Opened up a new path for research

in which the focus on perception was directed through the

lens of personality rather than through the broadly dew

fined motivational state."
The personality-centered approach created a new and surprising amount
of interest, as well as considerable controversy, in the area of perm
ception. Most of the experimentation and debate centered upon three

concepts which grew out of the personality-centered approach. These

concepts were: perceptual sensitization, perceptual defense, and

 

subception. It was especially the concept of perceptual defense that

 

engendered the most research and argumentation.

Postman, Bruner, and McGinnies (lQhS), and Bruner and Postman
(19h?) introduced the concepts of perceptual defense and perceptual
sensitization to account for the variations in recognition th esholds
for tachistoscopicallr presented words of idiosyncratic va no or threat.
Perceptual sensitization was introduced as a principle to describe the
lower recognition thresholds for words related to areas of high value
held by subjects, such as art, religion, or esthetics. The authors con~
cluded that an active, selective, perceptual process was pres-nt which
permitted the organism to recognize stimuli related to his own value
system more readily than the recognition of neutral stimuli. Percepw
tual vigilance has been used synonymously with the term perceptual

sensitization. Vigilance was originally introduced by Bruner and

Postman (19h?) to describe the behavior of certai. subjects who had con~
sistently lower thresholds for threatening stimuli than for neutral
stimuli. However, subsequent writers also used this term to describe
lower recognition thresholds to positively valued or non—threatening
stimuli, thereby, equating this behavior with perceptual sensitization.

Because of this trend, little effort was made to account for the
differential results found in many of the perceptual studies. One such
attempt has been to treat perceptual vigilance as a second manifestation
of perceptual defense. Within this framework, perceptual vigilance is
considered to represent a tendency to recognize threatening stimuli
quickly so that the threatening aspects of such stimuli may be altered
or vitiated. Although perceptual vigilance and perceptual sensitizam
tion both represent a lowered recognition threshold response, the stim-
uli eliciting this response are quite different. Consequently, a dif—
ferentiation between these two terms is necessary. In this paper, the
use of the term perceptual.vigilance will refer only to lowered recog-
nition thresholds that are elicited by threatening stimuli.

The second and most controversial concept introduced was that of
perceptual defense. This concept was postulated to account for the
raised recognition thresholds for words representing threat to the
individual. By using such a concept, Postman, Bruner, and MbGinnies
were implicitly drawing from one of the key concepts in psychoanalytic
theory, that of the defense mechanisms. The idea of defense has proven
invaluable in the theory of adjustment and maladjustment. Briefly, den

fense refers to the various operations by which the individual deals

with or binds anxiety or guilt feelings related to the expression of
certain instincts. Certain external objects or situations are also

defended against because they represent an inner instinctual demand.
As Fenichel (l9h5) states:

"There are defensive attitudes against painful perceptions

just as there are defenses against any pain. Nevertheless...

defenses against perceptions (and affects) seem to be per—

formed first and formost in the service of defenses against

instincts."

It is this defense against painful.perceptions that is implied in the
concept of perceptual defense.

The concept of perceptual defense raised the crucial question of
how the organism is able to know what to defend against and led to the
third concept, that of subception. Bruner and Postman (19h?) dealt
with this problem in a theoretical manner by postulating a hierarchy
of perceptual thresholds, one of these being voridical report. Other
types of perception are probably affective in nature, and have thresh—
olds below voridical recognition. Thus, they suggested two perceptual
processes, voridical perception and affective perception. Some experi-
mental evidence for such a View has been presented by McGinnics (l9h9),
and Lazarus and MbCleary (1991). They found that subjects could make
an autonomic nervous system response to threatening stimuli even though
the stimuli were presented too fast for conscious identification.
MbGinnies hypothesized that anxiety which was manifested in the autono-
mic response had a disrupting effect upon the voridical perceptual pro-

cess, enabling the individual to avoid recognition of the stimulus for

at least a time. Lazarus and thleary labeled this autonomic discrim-

ination, subception. Postman (1953a) later revised his affective—

 

veridical view and restricted perception to a single process, that of
veridical response. His revised view will be considered in greater
detail when the exrerimental studies related to perception are reviewed.

The concept of perceptual defense drew the interest of many exper-
imentally oriented clinicians to the area of perception. It was well
recognized that what an individual does in a perception task may reveal
pertinent information about his personality, conflicts, and defenses.
The opportunity to subject certain defens mechanisms to a more rigorous
experimental scrutiny now seemed to be available. Psychoanalytic con~
cepts have always been very difficult to test experimentally, and if it
could be shown that perceptual behavior could be predicted by the
psychoanalytic concepts of ego defense, some much needed validation of
these concepts would be provided. The basic experimental design used
in the study of motivational determinark3in perception also had some
elements in common with the familiar clinical tests, such as the
Rorschach test, in that they both used relatively ambiguous stimuli to
insure the maximum play of subjective factors in the individual’s re—
sponse. The experimentally oriented clinician also had a broader back_
ground in the psychoanalytic theory of defense than many experimenters
studying perception, so he could at least feel that he was on some
familiar ground as he invaded the rigorous and formal atmosphere of per—
ceptual research.

It was the utilisation of ambiguous stimuli which drew early prom

tests from many critics of the personality-centered approach to percep~

tion. They argued that when subjects are given fuller 'nformation as
in the usual pg: chophy sical experiments, the percept is stimulus bound.
The use of ambiguouss stimuli was de; elded by its proponents who stated
that only when there is a relative degree of axiicuitv pres out can
there be opportunity for subjective factors to enter into t1 1e FCwquuC.
There is a point as ambiguity decreases where all normal subjects must
finally come to grips with the stimulus and deal with it. Only the
psychotic individual would be capaole of alterine the stimulus under
such conditions. druner (1951a) further justifies the use of ambiguity
in the experimental Si tiati on by su resting that "Most complex percepw
tion in our daily social lives is dependent upon the i Itegration of far
less reliable information than is given in a tachistoscope at rapid
exposure."

Experimenters whose approach to perception was rooted in the psy~
chophysical methods could not help but view with considerable skepticism
the modification of perception by idiosyncratic motivational processes.
Some critics (Howie, 19§2; Prentice, 10§63 Luchins, 1951) directed their
arguments more to the ones- tion of why motivation should or should not
influence perception rather than to.ard the belaiioral data. Others
(Pratt, 19503 Wallach, 19h9) ouestioned whether the Functionalist apu
preach was actually dealing with perception. Pratt (1950) admitted that
motivational factors may affect judgrxent and motor r spo nses, but he
stated that they do not affect perception. He felt that whenever past
experience does seem to affect perception we are dealing with a motor

['7‘

esponse. inis view leads one to the difficult task of disti r.guishing

between perception pgg‘gg and its response components.

Similarily, Wallach (19h?) attempted to make a distinction between
sensorily determined perceptual experience and a recalled trace complex
which gives identity and meaning to the experience. He suggested that
motives operate upon the recalled trace complex rather than upon the
perceptual experience. Recent research by Bruner and Minturn (1955)
placed this view in question. They demonstrated that identification,
which is considered by Prattzznd'Wallach to be a response component,
can modify the primitive perceptual organization of the field. It ap-
pears that a clear and acceptable distinction between a pure percep-
tual process and its response components is often impossible to make.

If this be so, then relationships between needs and perception can hard~
ly be refuted by such argumentation.

Other critics (Solomon a Howes, 1951; Freeman, l95h3 Goldiamond,
1958) focused more directly upon the experimental data and attempted to
reduce the concepts of perceptual sensitization and defense to more
general principles such as set, word frequency, and response bias. They
argued that the results in most instances could be explained in terms of
non-perceptual variables. Postman (1953b) now entertains a similar view,
but he does not go so far as to deny the importance of motivational facm
tors in perception. He feels that motivation plays its role by influenc»
ing such variables as frequency, recency, and set.

The proponents of the concepts of perceptual defense and sensitiza-
tion reply that these are not phenomena pointing to some underlying gen-

eral law. General laws may influence their manifestations, but with

such factors as frequency and set controlled, the concepts can still be
demonstrated.

In spite of prolific research, the effect of motivational and perm
sonality variables upon perception is little understood. Many of the
experimenters have reported contradictory results. The results can be
explained in many instances in terms of non-perceptual variables. Mist
of the experimental studies havelieen designed merely to demonstrate
the validity of motivational concepts, with the result that there have
been fewer attempts to study a variety of conditions under which motiva~
tional variables may or may not have a demonstrable effect. Attempts to
reduce motivational concepts to simpler principles are quite plausible,
but they do not appear adequate to explain all of the available data.

In order to evaluate some of the previous experimentation and to give
clarity to directions recent studies have taken, a review of the literaw
ture will be presented. This rGViOW'Will be divided arbitrarily into

studies dealing with each of the three concepts: perceptual sensitizaw

 

tion erceotual defense. and subception. Although the studies dealinr
J l . - -

 

 

with each concept are reviewed independentlysit must be emphasized that
there is a close relationship among these studies. Some experiments
deal with two of the concepts simultaneously, others find their roots
in a related concept. Consequently, a particular experiment may be re_

viewed in part under one concept and in part under a second concept.

A REVIEW OF THE EKPERILBNTAI.IJTLRATURE

Perceptual sensitization. Postman, bruner, and McGinnies (l9h8)

 

introduced this concept in a study relating personal values to per~
ceptual behavior. Measures on the Allport—Vernon Scale of Values were
compared with subjects' recognition thresholds for tachistoscopically
presented words related to these value areas. An inverse relationship
was found between value rank and recognition thresholds of value related
words. Of interest was the finding that some subjects had lower recog-
nition thresholds for words related to low value areas as compared to
thresholds for words related to high value areas. This type of behaVior
was often ignored by subsequent experimenters but has recently acquired
new significance and will be discussed in greater detail when dealing
with perceptual defense.

The results of the Postman, Bruner, and McGinnies (PBE) study have
been confirmed in several experiments. Haigh and Fisk (1952) repeated
the PBM study using an Analysis of Variance to treat their results.

They also had subjects rank the word stimuli of the Allport—Vernon Scale
according to personal preference. This ranking was made four weeks afw
ter the perceptual experiment. An inverse relationship was found be~
tween rank value and recognition thresholds, supporting the hypothesis
that positive values tend to be associated with perceptual sensitization.
Vanderplas and Blake (19h?) found the same results using auditory stimuli.
They noted that a small number of subjects showed an opposite trend, in
that they had lower recognition thresholds for stimuli of low value.

MCOinnies and Bowles, (19h9) used portraits rather than stimulus
words. A value was attached to each picture by telling subjects that

the pictures represented a scientist, minister, etc. These occupations

lO

corresponded to the value areas of the Allport~Verren Scales. Recogni-
tion thresholds and scores on the Allport—Vcrnon Scales were then corre—
lated. Fifteen subjects had negative correlations, while the remaining
nine had positive ones. For the majori y of subjects, then, the recog~
nition thresholds were lower for portraits corresponding to areas of
high value on the Allport-Vernon Scales. The nine remaining subjects
had lower thresholds for portraits related to low value areas. The au—
thors questioned whether some subjects might not show a consistent use

("I

of perceptual vigilance to low value stimuli. lhey suggested this as
an hypothesis for future research.

thlelland and Libernan (l9h9) obtained measures of p-achicvement
for individual subjects using the T;T and other tests. Subjects were
divided into three groups, high, middle, nd low on p—achievement. In
the perceptual task which was administered three months later, subjects
were presented with achievement and security related words. The high
p—achievement group had lower recognfljon thresholds for achievement~
related words than did the middle or low groups. Both the high and
middle groups had lower:recognition thresholds for the security—related
words as compared to the low group. Ifrom these results it was concluded
that subjects tend to have lower thresholds for strong but acceptable
needLrelated.stimuli.

A number of experiments have been reported vhich have attempted to
place in question the role of value in perception as was indicated by

Postman, Bruner and McGinnies. Mausner and Siegel (1950) criticized

the use of the Allport—Vernon Scales as a test of values. In their

ll

GXperiment, amateur stamp collectors were required to learn the monetary
values of a set of stamps. Recognition thresholds for the various
stamps were then determined. No significant relationship was found bem
tween recognition thresholds and respective stamp value. The results
were interpreted as contradictory evidence to the findings reported in
the FEM study. The term value as used in the Mausner—Siegel study was
an attribute given to the stimuli by an external authority, and as such,
may or may not be related the individual's needs. Secondly, we have no
data related to the learning sitration which could have produced indivi—
dual differences i stimulus familiarity, and third, we have no knowledge
of the effect of color on recognition thresholds.

The studies demonstrating perceptual sensitization had a more seri—
ous shortcoming as was shown by Howes and Solomon (1951), who demon—
strated a relationship between frequency of prior exposure to the stim-
uli and recognition thresholds for such stimuli. In a further experi-
ment based on the FEM design, Solomon and Howes (1951) attempted to
answer the question of whether selective sensitivity to words from
different value areas could be reduced to differences in the frequency
of usage of such words. 'Word frequency was determined by reference to

the ‘horndike-Lorge'fiord Count (lfihh). They found that when stimulus

words were of high frequency, value had no signi’icant effect on recog
nition thresholds. A significant relationship did exist between low

frequency words and value rank. Their results indicated that although
word frequency accounted for a substantial part of the variance in re»

cognition thresholds, it did not account for it all. Solomon and Howes

attributed this residual variance to idiosyncratic word frequencies.
They stated that the Thorndike-lorge'Word Count was only a population
index and was not precise enough to reflect individual differences.
They also attempted to reduce the Allport—Vernon Scale to a more meas—
ure of word frequency.

Postman and Schneider (1951) in a similar study obtained essen-
tially the same results as Solomon and Howes. However, they disagreed
that the residual variance in subject's thresholds could be attributed
to idiosyncratic word frequencies not measured by the Thorndike-Lorge
WOrd Count. Postman and Schneider suggested that different degrees of
emphasis or reinforcement connected with some words might exist over
and beyond sheer frequency and may influence the strength of response
dispositions. They preferred to think of frequency and recognition
thresholds as dependent variables both of which are manifestations of a
more fundamental property such as habits, hypotheses, or perhaps per-
sonal values. The authors emphasized the need for further research in
which the frequency variable could be more precisely controlled.

Subsequently, Solomon ani Postman (1952) performed an experiment
in which the frequencies of various nonsense syllables were experimen—
tally controlled. Subjects were required to read and pronounce lists
of nonsense syllables with varying frequency. The variations in recog-
nition thresholds were then measured. The authors found a significant
decrease in recognition thresholds as frequency increased. The results
confirmed the effect of the frequency variable under controlled condi-

tions, but at the same time suggested‘that its effect may be limited.

13

Solomon and Postman concluded: "It will be well to exercise caution in
invoking frequency as a mediator of all observed motivational selectivity
although the contribution of this variable may be considered as estab-
lished.“

A rebuttal of Solomon and Howes' attempt to treat the Allport—
Vernon test as a measure of word frequency was presented by Brown and
Adams (195h). They revised the Allport—Vernon test so that the alter—
natives in all value areas except one were expressed in synonyms of low
frequency as determined by the Thorndike-Lorge Word Count. The remainw
ing value area had alternatives of high word frequency. Six forms of
the test were constructed, each one favoring a different value area in
terms of synonym frequency. According to the Solomon and Howes hypo—
thesis, the subjects should choose the high frequency words and thus,
achieve a high rank for the corresponding value area. Six groups of
subjects answered one form of the revised test as well as the standard
Allport—Vernon test. Subjects failed to show consistent changes in
scores for the favored value area when compared with the scores of the
other five groups to that value area. Correlations of the revised
tests with the standard Allport-Vernon test remained significantly
positive. It was concluded that word frequency E3222 does not determine
the performance on the Allport4Vernon test.

A study which appeared to pose problems for both the word frequency
hypothesis and the concept of perceptual sensitization was reportel by
Gilchrist, ludeman, and Lysak (195h). Students scoring high and low on

an anti—Semitism scale were used.as subjects. Positively valued, negaw

M1

tively valued, and neutral words were used as stimuli in the per,eptual
task. The positive and negative value words were equated for frequency
bv use of the Thorndike—lorge Word Count. The neutral_wcrds were of
higher frequency than either the positive or negative value stimuli.
Each stimulus word was resented twi 0, once with the word "ink” writm
ten above and below it, nd once with the word "Jew" written above and
below it. The additional.words were called the surrounds. Subjects
were required to recognize and report only the stimulus words. Recog~
nition thresholds for the combined comiitions were found to be lower
for both the positive and negative value stimuli as compared to the
neutral stimuli. When the recognition thresholds for the two surrounds
were compared, the results indicated that the thresholds for both the
positive and negative value stimuli in the "Jew" surround were signifi—
cantly higher, while thresholds for the neutral stimuli were signifi~
cantly lower than thresholds for the “ink" surround.

The results cannot be explained by the frequency hypothesis since
the neutral words were of higher frequency than he positive and nega-
tive words. The question was also raised as to wly the emotionally
loaded surround did not result in higher recognition thresholds for the
neutral words. The authors also stated that the results cannot be ex—
plained by the concept of perceptual sensitization, since thresholds
for the combined conditions were lower ibr the negative as well as the
positive stimuli. They concluded that their results posed problems for
all the existing concepts attemptirg to acount for differential thresh»

olds in perception. The authors seemed to have overlooked the fact

that Bruner and Postman (19h7), and HeGinnies and Bowles (lQhQ) found
that some subjects had lower thresholds for stimuli of low value, a
behavior which they termed perceptual vigilance. They also suggested

that some subjects might show a consistent use of vigilance to such

 

stimuli. In the Gilchrist, Ludemai, and lysak experiment, a signifi-
cant variance within prejudiced subjects was found. This could indi—
cate that there were consistent differences between subjects which were
independent of prejudice classification and that categorization should
be based, perhaps, on some other variable. Experiments such as this

one point to the need for a better understandi g of the conditions under

which perceptual sensitization may or may not operate.

Tabled word frequencies. Some important questions have been raised

 

concerning the use of tabled word frequencies as presented in the Thorndike—
Lorge'flord Count (l9hh). These word frequencies are based on the language
behavior of various magazine writers in the 1920's and 1930‘s. It seems
plausible that many changes in language behavior have occurred in the
last twenty—thirty years. Many of the so called "taboo” words now apm
pear in print, especially in the pocket size novels which are read by
many college students. A new word count would be needed to indicate
what changes have taken place in language behavior, a task both labori—
ous and time consuming. ‘We also have the recent medium of television
which undoubtedly exerts a significant influence on word frequency.

A variable which would seem to be closely related to word frem
quency, especially words of low frequency is that of intelligence. 'he

more intelligent students are likely to have had greater amounts of

16

exposure to the low frequency words in terms of both interest and read-
ing experience. Au yet, no study has been reported which describes the
influence of the intelligence variable on recognition thresholds of
word stimuli.

Variations in general word familiarity are also considered to exu
ist within a given frequency range. For example, Solomon and Howes
(1951) used such low frequency words as "percipience, erudition, un—
coerced, and vignette" in their study, while Postman and Schneider
(1951) used such words as "conception, logic, dominant, and litera-
ture", which fall.within the same frequency ranges as the words above.
The latter group of words is considered to be more familiar to students
than the former group. Lazarus (l95h) has suggested that the experi-
mental results could be made to go in any direction by an appropriate
selection of the word stimuli within a given frequency range. He also
felt that giving l'w frequency ranks to the "taboo" words and then
equating them with low frequency neutral words greatly‘underestimated
the familiarity subjects had for such words.

Howes (lQSh) investigated the relationship between tabled word
frequencies and student familarity with words of such frequency. Stu—
dents from two college populations were required to rank the infrequent
words taken from the earlier Solomon and howes (1951) study in order of
frequency of usage among college undergraduates in general as well as
frequency of personal usage. Correlations between student's rankings
and ranks based on the tabled word frequencies were .80. The results
indicated a significant relationship between word count data and stu—

dent familiarity.

17

Some Contradictory results were subsequently reported by Cowen and
Christ (1958). The neutral words used in their study were ranked by
the subjects according to individual familiarity along with the neutral
words used in the Postman, Bronson and Cropper OXpGTiant (1953). Both
groups of words fell within the same frequency ranges. It was found
that the neutral words used by Cowen and Christ were consistently given
higher ranks by the subjects. The authors suggested that this result
might be due to idiosyncracies of the students at the University of
Rochester, or they could be related to differences in familiarity with
words falling within the same frequency ranges. The familiarity prob»
lem is far from solved, especially when dealing with low frequency
words.

Some years ago, Postman (1953b) began a research program to study
the effects of the frequency variable in detail. One of the first stud—
ies reported dealt with the question of whether it was frequency of
prior eXposure, or frequency of prior word usage that was crucial vari—
able for word recognition. To study these variables, Postman and Conger
(195h) used three letter stimuli in the perceptual task. They obtained
a frequency measure of the three letters as a word, and a measure of the
three letters as a trigram. The tabled frequencies of the trigrams were
significantly different from the word frequencies since the trigrams
were not only words in themselves, but also three letter combinations
used in other words. It was found that sheer frequency of stimulation
is not significantly related to speed of word recognition, unless such

stimulation is accompanied by specific responses from the subject. On

18

the basis of these results one might hypothesize that most students have
made fewer responses to words such as "percipience" or "vignette" than
to words such as "bitch", "penis", or Ilwhore", which are often equated
with uncommon words like the above in terms of frequency.

Since there are so many problems connected with tabled frequency
data, a recent trend has been to avoid words and to use other stimuli
with which subjects have had no previous experience. Postman (1953a)
questions the use of such stimuli unless it can be shown that they have
been equated in terms of frequency of previous experience. If pictures
were used as stimuli, Postman would argue that some subjects have had
more prior experience with pictures and would, therefore, have lower
recognition thresholds. Such an argument can be applied to any stimuli
used in a perceptual eXperiment. Individual differences undoubtedly
exist in previous exposure to all classes of stimuli. It would have to
be shown that previous experience with a class of stimuli such as pie-
tures would have an effect upon the recognition of a specific picture
stimulus before such as argument would be appropriate.

Discussion in spite of the criticism based on the frequency vari-

 

able, the concept of perceptual sensitization appears to be a valid

and useful one, since much of the evidence suggests and active, selec-
tive process in perception. most of the experiments have related indi—
vidual differences as measured by the AllportAVernon Scale of Values

to differential thresholds in perception. There has been little attempt
to study the relationship of other acceptable needs to perceptual be-

havior. In fact, interest and research dealing with this concept has

19

declined rapidlv within recent years. Druncr (1951) had stated that
such data does not serve to explain perception but merely to indicate
its problems. The current feeling may he that such studies are of
little value, or that the frequency principle is sufficient to account
for such behavior. however, it would appear that there is still con—
sider able room for exploratO‘y work in this area.

One current theory treats percegtion as an affective—voridical
process. The affective response has lower thresholds than veridieal
response and tends to facilitate or disr nit the vericical process.
This theory suggests some interesting problems for further research.

irst, it may he iuesti oned w Heiue r both positive and negative stimuli
occasion an a fective response, and if so, how do the responses differ.

Second, how does an affective response which facilitates voridical

 

response differ from one which ricrrlts voridical perception? The
answers to such questions would increase our understanding of the relau
tionship of the various concepts such as sensitization and vigilance
which now exist to describe perceptual behavior.

Subception. The concept of subception grew out of a basic problem

 

encountered in the study of perceptual.defense. This problem concerned
the ability of the organism to detect threatening stimuli so hat it
might avoid them. Howie (1952) in his criticism of perceptual defense
argued that such a concept suggested the presence of an homunculus withu:
the organism who directs tie organism toward or away from certain stim~
uli. The problem of stimulus detection immediately became a crucial one.

Bruner and Postman (19h ,) attempted to deal with this problem in a

theoretical manner by positing a hier arclgr of response thresholds. Perw
ceptual discrimination need not be restricted to one t;'pe of response
such as veridical report. Rather, there may be a number of discrimina»
tion responses s, each having its own tl1reshold. The authors sug
these other responses are affective in nature. In some situations an
affective discrimination response would have thresholds below veridical
report and would lead to avoidance behavior at t11e voridical level.
Under this theoretical view, two t;.'pes of perception were suggested,
veridical perception and affective perception. No attempt was made to
explain how these perceptual responses were mediated.

Some subsequent experimental data tended to confirm the Lirner—
Postman view. McGinnies (lQhQ), in a much criticized study, obtained
an autonomic response measure prior to recognition of both taboo and
neutral words. He found that the GSR preceding recognition of the taboo
words was si "1 dificantly greater than that preceding neutral words. He
suggested that the presence of a taboo stimulus sets off an anxiety
signal manifested in the autonomic response, which in turn leads to
avoidance behavior in perception. McGinnies indulged in some speculau
tion concerning the neurophvs iolo: ical processes mediating such behav-
ior. He hypothesized that the autonomic response is initiated as the
visual impulses reach the optic thalamus. He further stated that:

"A rerouting of the afferent activity might then take

place in the visual centers so that cortical integration

would be effectively modified in the direction of phe—
nomenological distortion."

This speculation stemmed from his analysiso i the pre e—recognition

lypotheses to the taboo and neutral stimuli. he found a significantly

21

greater number of nonsense and unlike prerecognition hypotheses or
guesses to the taboo presentations than to the neutral presentations.
Recently, Reyher (1960) has exoressed a similar view concerning
the disrupting effects of anxiety—producing stimuli. He created in—
tense conflict situations under hypnosis and then studied perceptual
recognition of relevant anxiety—related stimuli. He found that not
all subjects show subception, but when such behavior occurs, it repre~
sents a pathological condition accompanying a highly conflictual emo—
tional process. This differential.response among subjects led him to
postulate two processes in perception. First, for certain subjects,

the anxiety response is activated at veridical thresholds and intern

 

feres with the cognitive process of identification. Second, in other
subjects, the anxiety response activates levels of cortical integra—
tion not associated with conscious awareness. This occurs only under
pathological conditions. Reyher's results indicate that subception is
a function of the degree of threat.

Howes and Solomon (1950) in their criticism of the experiment by
McGinnies (19h?) stated that the simple factor of response suppression
could account for the greater autonomic response to the taboo stimuli.
Subjects could recognize the taboo word and register a high GSR, but
they would then delay report of the taboo word until they were certain
that it was such a word.

To control for the variable of response suppression, Lazarus and
LhCleary (1951) used nonsense syllables as stimuli. They conditioned

an autonomic response to half of the nonsense syllables by using shock

as the unconditioned stimulus. All of the syllables were then presented
at below recognition thresholds. The average G83 following the prescnm
tation of the shock syllables was significantly greater than that foi~
lowing the non—shock syllables. Their results confirmed McGinnies

findings. The authors introduced the term subception to describe this

 

autonomic discrimination. Subception was deaned as:

"A process by which some kind of discrimination is made

when a subject is unable to make a conscious discriminw

ation."

Support for the subception concept has been provided by a number
of other experimental studies (Iowenfeld.et_alg 19563 Blum, l9§h3 Nelson,
1955‘ The results reported by lowenfcld, Rubenfeld, and Guthrie (1996)
are of particular interest. They replicated the Lazarus and MbCleary
experiment and then attempted to extinguish the autonomic response pre~
viously conditioned to the shock stimuli. They found that responses
to subliminal stimuli may follow a very different pattern of extinction
than responses to stimuli presented at veridical thresholds. They also
found a marked decrement in the GSR to shock stimuli presented supram
liminally, when subjects were assured of no more shock. The GSR was not
reduced in spite of assurance that shocx would no longer occur, when
such stimuli were presented too quickly for veridical identification.
An autonomic reaction persisted even though subjects were no longer
punished. These results suggested that the veridical perceptual pro_
cess has an inhibitory effect upon the affective process.

A criticism of the concept of subception was presented by Bricker

and Chapanis (19g3), who demonstrated that even when the subject’s

23

first verbal response is incorrect, the stimulus has conveyed some use~
ful information to him. Subjects were given printed lists of the stimw
uli which were to be presented tachistoscopically. After a sunuthreshold
presentation of one of the stimuli, the SibjectS'were instructed to use
the word lists and guess which word had been presented. Subjects were
able to select the correct stimulus with fewer guesses than would be
predicted by chance. It was concluded that submthreshold presentations
of the stimuli did give subjects some cue information concerning the
stimuli"

Bricker and Chapanis' results were subsequently confirmed by Lysak
(l95h) and Murdock (lQSh). These experimenters suggested that the sub—
ception effect could be eXplained in terms of the partial recognition
of the shock syllables which then led to an anticipation of shock and
produced the GSR. They concluded that the GSR is merely a sensitive
indicator of this partial recognition of the stimuli.

The above studies did not, however, obtain a measure of the GSR.
In a later experiment, Voor (1956) investigated the relationship bem
tween veridical information obtained from the stimulus and the GSR
discrimination of the stimulus. He hypothesized that the subception
effect would not be found when the subjects could not guess signifi»
cantly above chance which stimulus had been presented. By using three
levels of illumination, he was able to vary the amount of cue informa~
tion given to his subjects. He found that as veridical information is
lessened, the ability to respond affectively is also weakened. He in_
terprets his results to mean that the GSR is dependent upon veridical

cue information.

2h

The results presented by Bricheerhapanis, and others confirmed
Postman's (1953) revised view of the perceptual process. In this rem
vised view, he restricts the perceptual response to veridical report
only. He then posits a hierarchy of resuonse predispositions within
veridical perception, each requiring a different quantity and perhais
a different pattern of stimuli before veridical response occurs. Under
this view, the perceptual process becomes one of hypothesis confirmaw
tion or nonwconfirmation. In any given situation the o ganism is set
for a limited range of events. The stimulus input serves either to
broaden or narrow the range of possible hypotheses, and as the stimulus
input increases the individual tends to confirm or deny specific hypo~
thesis. The strengti of hypotheses is measured in terms of the appro—
priate information needed to confirm or deny them. The stronger the
hypotheses, the less information is needed to confirm them. Conse~
quently, ambiguous stimuli are necessary to study hypothesis confirm
nation. The strength of hypotheses is related to past experience, such
as frequency of previous confirmations. Motivational.su,port for an
hypothesis may increase or decrease its strength. Hypotheses whose
confirmation signifies punishment or threat may be quite weak and rem
quire more information to confirm them. however, Strong hypotheses

may also be related to negative stimuli, then the Opposite of defense

or perceptual vigilance will appear to operate. This latter postulate

 

seems to have been introduced to account for contradictory results
occurring in perceptual defense research. Postman does not indicate

under what conditions the hypotheses related to inimical stimuli may

be strong or weak. This would appear to be a crucial factor in predic-
ing perceptual behavior, and one that will be dealt with in this
present study.

Eriksen (1956) presents a similar view of subception. he critim
cizes the lazarus and McCleary design in that it makes subception a
statistical artifact of the stimulus conditions. He states that in
this design, a restriction was placed upon the subject's verbal report
since his response was limited to being either correct or incorrect,
while the USE was free to detect any partial discriminations the subject
might make. A GSR could occur in the absence of a verbal discriminam
tion as measured by a correCt—incorrect indicator.

Recently, Dixon (1957) designed an experiment to meet the criti-
cisms of both Bricker and Chapanis and Eriksen. He ensured total
unawareness of the stimuli by using stimulus magnitudes well below the
awareness thresholds for each subject. The subjects knew that a stim—
ulus was being presented only because they were told so by the experi~
mentor. They were unable to report seeing anything in the visual
field. Dixon also allowed his subjects a potentially infinite range
of verbal response categories. A signal was given before each stimulus
presentation, and the subject was instructed to respond with any word
that came to mind. The stimuli used were threat and neutral words. A
GSR measure was taken at each stimulus presentation. The results indi~
cated that subject's verbal responses bore no structural similarity to
the word stimuli. All subjects had greater GSR response to the threat

stimuli as compared to the neutral stimuli.

20

Following tEe perception phase of the experiment, Dixon presented
the subjects with frintcd lists of the stimuli and verbal Ies penses.
The accuracy of the stimuluswr e spouse mate dzinCS’was significantly better
than would be predicted by chance. On the basis of his resu ts, Dixon
concludes that the ability to verbalise a recofnition of
cannot be taken as t1e sole or i erion of awar ness. Both au
discrimination and re: sal awareness riajr rave thresholds below a verbal.
recognition response. D:;on'sr es*l‘ s are based upon seven subjects
onlv and need, therefore, lurther substantiation.

'l

Discussion. The preceding review 01 the experimental litera ure

 

has indicated two theoretical vicml to acco int for the subception
phenomenon. One view considers the autonomic or affec ive discriminaw
tion to be based upon a direct response to the threatening aspects of
the stimulus. Such a response either acts independently of tigher
levels of cortical functioning or tends to disrupt them. In the second
view, the autonomic response is thoupht to be mecifted bf a conscious,
tlough partial recognition of the stimulus rather than by unconscious
processes. This latter view has received the most experimental support.
f

However, a recent experiment has placed the theory 0 »ar'tial recogi'm

97“

tion in Question. Dixon (1057 was able to demonstmr to both an autow
nomic discrimination and a verem ] “soriv‘naoion at stimulus magnitudes
below verbal awareness. his r sults indicate the need for Iurther re—
search dealing with so called unconscious processes in perception.

Re ardless of which theory ultimately proves to be correct, the

concept of subception retains a plausible explanation of how the on~

2?

anism 1Hl0\8 what Stimuli to avoid. Such a concept does not explain
the “eater responsiveness shown by certain subjects to threateninr
stimuli. This might be done by subsuming the co ncept of perceptual
vigilance under the general rubric of defense. The treatment of per_
cept aal vigilance as def er sive behavior will be discussed in greater
detail in the next section on perceptual defense.

Perceptual defense. In their study of personal values, Postman,

 

Bruner, and McGinnies (lQhS) al: 0 introduced the term perceptual de-
fense to describe t} Ie elevated recognition thresholds for words related

0 areas of low value as determined by the Allport—Rv erIon Scale of

Ct

Values. It was not the higher recognition thresholds that prompted

the introduction of this concept, but rather, the type of pre~recognition
hypotheses or guesses prior to recognition that suggested an active
avoidan process for such words. The pro—recognition lr'poth ses were
categorized in terms of the relat :wx degree of similarity to the actual
stimulus. Analysis of the pre~xecognit ion hjpotIm ses revealed that more
nonsense words or words of opposite value occurred as pre-recognition
hypotheses to low value areas than to high value areas. From this
analysis the autlors concluded that subjects made an active effort to
avoid recognition of low value stimuli.

A su oseouent experiment bv Bruner and Postman (1% 7) added support
to th e concept of perceptual def nse. However, this study was published
first and was ignored bv later erzperimenters whose interest wa sin the
above "value 5" experiment. In tneir 1017 stwwd,, Bruner and Postman

used stimulus words considered to have threat value for individual

subjects because if long "cactien times to such words on a word associa-
tion test. A positive relationshi was found between reaction tine en
the word association test and perceptual.recognitien thresholds. 'While
most subjects had hifher recognit tion thresholds for threatening stimuli,
some subjects again showed consistently loner thresholds 01 peicep tual
vigilance for such inimical stimuli. This result was als e ignoredf

some time as interest beet 1e cexsrwee on an independent study by
McGinnies.

In a much criticized experiment, McGirnies (10’?) used the so
called "taboo" words as stimuli. SileCCtS were found to lave signifiw
cantly higher recognition thres Iolds for the taboo words than for neum
tral words. This experiment did much to give to the concept of perm
ceptual defense an existence as a special pc: ceptual mechanism and at
the same time, it precipitated uI‘1erous criticisms asserting that a
concept of perceptual defense was unwarranted.

Hewes and Solomon (1950) suggested that two simple recesses could
account for the results of the McGinnies eXperiment. The first was the
factor of word frequency. The authors subsequently demonstrated the
effect of this variable on recognition thresholds. The second factor
was that of response suppression. It was sugges ied that responses to

taboo words are not delayed in reco nition but ale we elf delayed in

verbal report. A number of experiments followed sn stantin‘ing tLe
proposed existence of reapons e suppression to th -e se~called taboo

stimuli.

Lacey, Lewinger, and Pdamsen (qu3) divided subjects into three

29

groups on the basis of the amount of information received in the in~
structions concerning taboo words. They found that taboo words are re-
cognized more quickly when subjects are forewarned that taboo material
will be used in the experiment. An habituation effect was also noted.

that is, with succgssive wresentatiens of the taboo words the thresh-~

5
olds tend to become lower. It was conc.uded that instructions change
the relative dominance of certain sets in the subjects.

.,

Freeman (lQSh) in his first study performed two experiments. In
the first, he varied the threat expectancy of his subjects. He found
that the GXPOCttnt group had lower thresholds for both neutral and
taboo words. An habituation effect was also noted. In the second ex~
periment, he changed the first letter of the taboo words making them
taboo-similar. Subjects were then forewarned to eXpect taboo words.
He found that a type of vigilance seemed to develop to all taboo-
similar words following the incorrect perception of a taboo~similar
word as a taboo word. This finding gave confirmation to an interprem
tation of the results in terms of dominance of alternative sets which

predispose the subject tenarcs certain types of stimuli.

.f‘
A.

In a further experiment, Freeman (1055) used separate groups 0
males and females. Threat echetancy was varied, and subjects were
also given ego involving instructions. They were led to believe that
the perceptual.task was related to academic success. Raised thresh~
olds for taboo words were found in the uninformed group9 but there was

no significant difference between taboo and neutral thresholds in the

informed groups. Informed females showed less reduction of thresholds

30

for taboo words than did males. Ego involv ment had the effect of rem
ducing all thresholds, neutral as well as taboo. This effect was more
pronounced for the females than for the males. These studies again
confirmed the hypothesis that recognition thresholds are a function of
the dominance of alternative sets.

Postman, Bronson, and Gropper (1953), in a carefully controlled
study used four different sets of instructions to manipulate the sub~
ject's readiness to report taboo material. Sex of the experimenter
and the subject were also systematically varied. The thresholds for
taboo words were found to be lower than for neutral words for all four
experimental groups. Both the taboo and neutral words had been equated
for frequency. The authors hypothesized that the lower thresholds for
the taboo words might be due to an underestimation of the frequency
counts of such words. The relative thresholds varied significantly
with the type of instructions. The group receiving instructions to
facilitate the report of taboo words had the lowest thresholds for such
words, while the uninformed group had the highest thresholds. The rem
lative thresholds were not significant y different in the presence of
an examiner of the opposite sex as compared with an examiner of the
same sex. An analysis of the prewrecognition hypotheses failed to con-
firm earlier findings of differences in subject's guesses to taboo and
neutral words. The authors suggest that even if differences in prew
recognition hypotheses are found, these differences can be as equally
well ascribed to the uncontrolled *ar'ations in word frequency as to a

defense concept. Because of the unexnected direction of their findings,

they were forced to conclude that perc eptual defense has at best, the
status of an unconfirmed hwypot esis.

Stern (1952) used personally relevant threat words of a non-taboo
variety to avoid possible re mio use suppression. The threat words were
equated with neutral words for frequency. A comparis;on of recogzition
tmlre holds for emotional and neutral words revealed no evidence to supm
port the concept of perceptual defense. Analysis of pre-recognition
hypotheses did show that subjects used more nonsense and contravaluant
words as guesses to emotional stimuli than to neutral Sb ‘2uli. Similar
findings were reported by'Uylie (1957). Threat was associated Wl
certain numerical stimuli bx placing subjects in a failure situation
involvinga ritM1ietic tasks. No significant differences were noted in
recognition thresholds for the threat and neutral stimuli. A differ

ence was found between pre-recegnition guesses to the threat and neu~

tral stimuli.

..

In general, the analysis of pre—recognition hypotheses has been
presented in addition to threshold data to lend more support to the
concept of perceptual defense. This factor has been ignored by most
experimenters attempting to account for variations in recognition
thresholds by the perceptual principles of word frequency or response
set. The concept of perceptual defense could oe posited on the basis
of pre-recognition hypotheses, but it would be in a rather weak posi-
tion when no eifiere nccs are found in recognition thresholds. This

i

would indicate that the defensive technique is unsuccessful in its

efforts to avoid certain stimuli. On the other hand, if it can be

32

shown that pre-recojnition hypotheses to threatening and neutral stimuli
are not different, this finding would still not be crucial to the dew
fense concept. It could be argued thzt pre-recog ition hypotheses simw
ilar to the actual stimulus still permit the individual to avoid the
real stimulus. As yet, there has been little systematic invistigation
of the prewreeognition factor as a major variable in perceptual defense.
Some CXperiments have been reported which attempt to control.for
one or both of the response suppression and frequency variables while
claiming to demonstrate perceptual defense. Cowen and Beier (1950,
1953, lQSh) used essentially the same design in three eXperiments in
an effort to demonstrate the existence of such a concept. Subjects
were required to examine mimeographed booklets in which increasingly
clear versions of a word appeared as each page was turned. The book-
lets contained taboo and neutral words. One group was informed that
taboo words would be presented, the other group remained uninformed.
In the first two studies, the authors failed to equate for the frequency
variable. They also went to considerable effort to interpret their re“
sults in favor of the defense concept when the data seemed more clearly
to support the response suppression hypothesis. One interesting findm
ing did result from the first two studies. The individual and group
variance for the informed subjects was significantly greater for the
taboo words than for neutral words. The variance was also signifi~
cantLy greater when compared with the variance of the uninformed group
for the taboo words. This result suggests that alerting subjects to

erpect taboo words has the effect of lowering thresholds to such words

3}

for some subjects but not others, or leads to a differential intra~
subject response to the taboo stimulus category. It is possible that
response suppression still has an effect even though subjects are
alerted to expect taboo words, or perhaps the frequency variable has a
differential effect on the various taboo words.

In their third experiment, Cowen and Beier (lQSh) controlled for
both response suppression and the frequency v riable. They found that
more time and trials were required to report the taboo words than the
neutral words for both the informed and uninformed groups. The greater
time spent in reporting taboo words again suggests the operation of
response suppression. However, the writers feel that an explanation
of their results in terms of conscious inhibition is inadequate. At
best, their results remain equivocal.but show what can be done with
little data when an experimenter is really trying to support a parti-
cular hypothesis.

Newton (1955) attempted to avoid the problem of response suppres-
sion by using pleasant and unpleasant words of a non-taboo variety.
The words were also equated for frequency. He found that signifi—
cantly less recognition errors were made to the pleasant as compared
to the unpleasant words. Wiener (1055) controlled for word frequency
and response suppression by using non—taboo words in a threat and nonw
threat context. The same words were used in both contexts. Selective
set was controlled by also placing "neutral" stimuli within the threat
context. If selective set was operating for the critical words in the

threat context, then raised thresholds would be expected for the

'11

neutral worls. leis did not occur. he threat group had significantly
lower thresholds for the critical words than the non—threat group.

This result was opposite of the direction predicted. However, since
the frequency hypotheses would predict no differences in thresholds

for the critical stimuli while a motivational hypotheseS'would, the
author interprets this as evidence in favor of a motivational determi-
nant of perception. The results reported in this study are s milar to
those found by Postman gt a}. (1953) and suggest a vigilan.e effect to
threatening stimuli.

De Lucia and Stagner (lQGh) replicated the earlier Bruner and
Postman (19h?) study of emotional selectivity in perception. They used
an analysis of variance technique which permitted an analysis of the
main effects of the variables of word frequency and degree of threat
as measured by reaction times on a word association test. hey found
that both word frequency and degree of threat had a significant effect
on recognition thresholds, a result supporting both the Solomon a Howes
(l951),and Bruner—Postman (19h?) experiments. Again, some subjects
had lower thresholds for words eliciting long reaction times on the
word association test.

De Lucia and Stagner posit a linear function for recognition
thresholds in relation to frequency, which probably represents the
trend for all words. Deviations from this trend occur when a word of
defined frequency is found to have recognition times either in excess
of or shorter than that predicted on a frequency basis. De Lucia and

Stagner preferred to discard the mechanism concept of perceptual defense

35

and speak of the homeostatic aspects of human behavior. In some in—
stances the homeostatic relationship is furthered by the rapid percep-
tion of certain stimuli, in others, the percept may arouse other asso-
ciative responses than the mere naming of the word which would tend to
interfere with the verbal response. These writers would still be faced
with the problem of predicting under what conditions either vigilant
or avoidant behavior would occur.

Osler and Lewisohn (lQSh) used words classified as "acceptable"
or "unacceptable" to test recognition thresholds. The words were equa-
ted for frequency and structure, except for one letter. Examples of the
words used are: tit —— tat, and bitch —- botch. high and low anxiety
groups as measured by the Taylor Anxiety Scale were used as subjects.
The low anxiety group showed no significant differences in recognition
thresholds between acceptable and unacceptable words. In the high n—
xiety group, the recognition thresholds for the unacceptable words was
found to be significantly lower the for acceptable words. It was conm
cluded that anxiety is related to perceptual vigilance.

McGinnies and Sherman (1952) stated that what was needed was an
experimental demonstration of behavior that was clearly a reflection
of differential perceptual organization rather than selective verbal
report. Eighteen neutral words equated for frequency were used as the
task stimuli. Four taboo and four different neutral words were pre»
sented to subjects at easiLy recognizable exposures. One of the four
taboo or neutral words was first presented at an exposure speed which

ermitted veridical report. .mediatelv followin: this resentation
& U D

\JJ
I
V

a task word was flashed at a speed below varidical threshold. Th sew

quence was repeated, gradually increasing the exposure time of the task

.J

word until the subject was able to identify it. hecogzition thresholcs

L

for the task words following the supra iminal presentation of the taboo
words were significantly higher than the thresholds for the task words
following the presentation of neutra1.words. The writers concluded

q

that the disrupting effect of anxiety in relation to the taboo w rd is

‘

still present when a second stimulus is immediately presented and rem
sults in a raised threshold for the succeeding stimulus. The concept

of perceptual defense is then treated in terms of reinforcement theory.

Verbal.responses involving taboo symbols have for most subjects been

\

['1 ' '

punished by external figures. laboo words then become a negati“e rein—
forcing agent. Uhen presented as stimuli they signal an approaching
state of punishment and, thereby, become cues for eliciting the anxiety
associated with actual punishment. It had previously been shown that
the threshold for an anxiety response is lower than the threshold of
veridical recognition. The autonomic cues are apparently adequate to
initiate a perceptual avoidance reaction. This voidance reaction is
accomplisned by distortion of the stimulus as evidenced by the pre-
recognition hypotheses and by elevated recognition thresholds. The
avoidance response reduces anxiety and is thereby reinforced.

Eriksen and Browne (1966) have recently pr sented a similar
theoretical analysis of perceptzal defense in terms of principles dew

rived from punishment and .voidance conditioning. One factor that

early studies dealing with perceptual defense had in common was the

assumption that all subjects would react to threatenin ng stimuli in per—

1.

ception with avoidance behavior. bxplanatiens in terms of learning
principles would adequately account for such avoidant behavior. However,
in many experiments the opposite of avoidance occurred. Subjects showed

I

an increased responsiveness to threatening stimuli. Theory based on
conditioning principles did not account for his behavior. Erikscn and
Browne (1? Se) \HC 0 aware of this and limited the application of their
analysis to those subjects known to deal with threat by avoidance be~
havior. The contradictory exper‘rent 1 results created the necessity
for determining more adeou a 0 methods of predicting perceptual response.

.

One approach to this problem has r zsul ted in an emphasis upon person—v

ality factors in perception.

The personality-centered approach. lasarus (lQSh) and Lrikscn

 

(lQEh), while favoring the perceptual defense concept, criticized many
of the previously cited experiments as being inadequate to demonstrate
the concept. Drawing more heavi y from the concept of defense as prem
sented in ps;wcho aly ‘ c tEleory, they presented what they felt were the
necessa y methodo cg: cal requirezWen s to adequately test r such a conw
cept. It is necessary to show, first that the stimuli ar eanxiety prom
ducing for all subjects, and second, that the subjects have learned to
deal with anxiety from this particul; r source by repressive or avoidance
defenses. By us in; taboo owords, the cxper'xenter is making the gratui-
tous as sunption that these words are anxiety producing for all subjects
It is also impossible to determine if the performance is due to the

emotional characteristics of the stimulus or to other factors such as

word faliliariUx, conte: t respcnse set, or values. The above authors
have conducted a nrmbei‘ of wiperiments using the 13010 pr rho: 1l:"t.:,v'—~

centered approach to pc mrception.

Eriks en (1051), usii" t1e TAT to detect area

0)
1.1
H
L—J
(— L-
(D
*U
C.)
"S
C‘)
9
E13
H
H
C f
“‘1

1..

.here defense might occur, found thaU suojects who blocked or showed

emotional di st area nce to the TAT cards Isually eliciting are: cssive

themes had higher recognition thresholds for aggr»ssive scenes when

presented tachistoscopically, than cid subjects who expressed aggres~

Vi

sive themes freely. In 2orther experiment, Erihsc en (1951) used t1 iree

pg
}L

l

psycniatric populations as on jects. Two groups, schizophrenicsw
paranoid type, and chronic alcoholics were chosen because they are conw
sidered to have one cceptaole succorant and homosexual needs. A third
group was made up of other diagnoses to represent a miscellany of other
unacceptable needs. Subjects were then administered a word association
test. He found that a disturbance on the word association testx as
Lositively related to recognition thresholds for scenes depicting perm
sons in the act of expressing or gratifying the co1respond1ng needs.

An analysis of th e pre—recoL nition hypotheses rev calcd a tendeIuCI for
subjects who showed disturb ance in a particular need to distort scenes
corresponding to this need area so as to remove the threat. This dism
tortion did not oec1r for all sun cts nor for all need scenes for a
given subject.

*1

nrikscn's results suL"es 'r of avoidance defense from

. W
J .‘ ' a

c+

verbal behavior to perceptual behavior. This finding dii;cred from t1 e

1 u o

arlier studies in which disturbance on the word association test was

39

related to de.iati ous in “one“ iUcon thresholds either above or below
the thr-sholds for neutral stimuli.
Sup W:>ort for defense consistencv can from another experincnt by

Eriksen (1952) in which he first obtained a measure of recall o1 crmw

pleted and inconpleted tasks in a failureuinduced situation. he :ouno
that those subjects who reacted to induced ego threat by f rgetting

their failur's had higher recognition thresholds for 1ailur: r‘lated
perceptual stimuli. Only those subjects who showed an avoidance type

1,

of defense in the memory tmi 1: showed perceptual defense. In both sit”
ations the respoz1se was to avo o.d recognition of the minulus.. to keep
it from awareness.

A number of experiments have been performed using nosological eat-
egories as a basis for predicting perceptual behavior. Iazarus.
Eriksen, and Fonda (1951) used hysterie and obsess ve—compulsive pa-
tients as subjects. According to psve ehoanalvtic theory, the major dew
fense of the hysteric group is repression. Consequently? it was prem
oieted that this group would show perceptual defense. The obsessivew

compulsive group is considered to use intellectualizing defenses in

dealing with conflictual stimuli. Since this latter group is able to

deal more freely with CCltC aspects of impulse CIpIGS sion, it was
redicted that this *rouo would not show avoidance in erce tion. The
P .

predictions were confirmed. Subjects using the intellectualizing dew

fenses perceived the stimulus material Signiiieantl more accurately

P

than did subjects using1eprcs sive de enses. The authors suthcsted

Ct

hat individuals may have a generalized method of responding to th eat.

to

A disturbance in one area does not necessarily mean a disturbance in
another area, but when this does occur, the same defense appears to be
utilized.

Daston (1956) found that paranoid patients had lower thresholds
for homosexual'words than did other types of schizophrenics. Lindner
(1953) reported that sexual offenders had lower recognition thresholds
for sexual pictures as compared to other types of criminals. Stein
(1953) using neurotic patients was able to classify them as either de—
fenders or sensitizers on the basis of deviations in recognition thresh~
olds for conflictual material as compared to neutral material. Subjects
then showed a consistent use of either perceptual defense or sensitiza-
tion to a subsequent series of pictures depict'ng aggressive and neutral
scenes.

The nosologieal approach to perception did not go unchallenged.
{urland (lQSh) failed to find significant differences in recognition
‘hresholds for threat words between hysteric and obsessive-compulsive
patients. As a possible explanation of his rusults, Kurland suggested
that many f his subjects in the patient groups might have been sensi~
tized to the emotional stimuli as a result of psychotherapy. The data
was further examined for a differential response related to the type
of conflictual stimuli. No significant differences were found.

Bitterman and Kniffin (1953) studied the perceptual behavior of
high and low anxiety subjects ascietermined by the Taylor Manifest
Anxiety Scale. No significant relationship between anxiety level and

recognition thresholds was found. A significant difference in threshold

‘l ‘ V

between neutral and taboo w rds was Lound and was interpreted in terms
of dii ferenti a1 re adines: to report rather than in terms of perceptual

distortion. Spence (1957) also attempted to relate anxiet; level to

3 found that subjects tended

>-—c
5“
,a

avoidance and vigilance in perception.
to to use both types of behavior and could not be labeled specifically
as either avoid rs 7r (lsitizers.

An experiment by Chodorkoff (1956) also failed to support a view
of defense consistency. Hiourd that t1 10 degree of threat or word
stimuli as measured by reaction tlthS on a word association test was

‘ q a n n A o [.0

not related to aw»o dance oeravior in perception. n signi-Jcant relam

.L

tionship was feind between the degree of threat to certain words and
the degree of deviation of recognition thresholds for such words 0; ther
above or below recognition threshold for neutral words. Some subjects
who had long reaction times to CL rtain words on a word association test
showed greater responsivity to those words perceptually than to neutral
words. Other subjects had ligher recognition thresholds for threatenu
ing words han for neutral words. A subje::t's perceptual behavior
could not be predicted from a hnonle(h e of his perfor :ance on a word
association test.

A measure 0; selfnadjustment we 8 also outaiizcd using the Q tech~

c ntly related to the degree of

H:
P.)

nique. Level of adjustment was signii
threshold deviation for the threatening words, disregarding direction:
as compared with the threshold for the neutral words. Chodorkoff conm

sidered level.of adjustment to be the variable which determined whether

 

 

perceptual defense or vigilance would occur. The setter adjus )th sub-

h2

jects tended to use vigilance to deal with threatening stimuli. The
question might be raised, however, as to why the better adjusted sub—
jects showed long reaction times to certain words on the word associa—
tion test. Is the test inadequate as a measure of subjects' defensesg
or do subjects vary in behavior from one situation to the next? To
answer these questions, other instruments were devised to determine
subjects' defense mechanisms in dealing with threatening stimuli.

A study which exerted a significant influence on later experi—
ments dealing with perceptual defense was reported by Goldstein (1952).
He devised an instrument called the Defense reference Inquiry which he
used in conjunction with the Blacky Pictures (19h9) to obtain a measure
of the subject's defense methods in dealing with material related to
the various stages of psychosexual development as presented in psycho-
analytic theori. Goldstein found that the majority of his subjects
tended to choose a variety of defenses across the psychosexual dimen-
sions. However, a significant minority tended to choose the same dew
fense, regardless of conflict dimensions. The former group he labeled
"specific defenders", and the latter group "general defenders". The
general defenders also showed more disturbance in their spontaneous
stories for the Blacky Pictures.

Goldstein's finding offered a possible explanation for the contra-
dictory results reported in the various experiments on perceptual de-
fense. Some of the subjects in these experiments may have fallen within
a specific defender group while others may have belonged to the general

defender group. Subjects in the former group might then show perceptual

1L3

defense for certain stimuli and vigilance for others. Different types
of stimuli although falling within the "threatening" or "emotional"
category might have, on this basis, led to differential results.

A few studies have been reported in which the Defvnse Preference
Inquiry was used to determine subjects defenses to psychosexual con—
flicts. The hypotheses tested were based specifically upon psychoana—
lytic concepts. Blum (19Sb) investigated both vigilance and defense in
perception. He stated that since wardcd—off instincts continue to
strive for discharge, they tend to utilize derivatives in an effort to
gain indirect expression. The individual is, therefore, responsive to
environmental cues relevant to impulse expression. This responsiveness

he termed perceptual vigilance. his use of the term differs from pre—

 

vious definitions in that it refers to a discrimination of conflictual
stimuli below the level of verbal awareness. Blum's definition of vigi—

lance corresponds more closely with subception as has previously been

 

discussed. Vigilance, in its more accepted meaning, refers to lower

veridical thresholds for inimical stimuli. Blum further stated, that

at the more conscious level actual impulse expression or its derivatives

are subject to ego defenses, because of the anxiety evoked. By con-

trolling the subject's verbal awareness to anxiety producing stimuli,

demonstrations of vigilance (subception) and defense were attempted.
Four reproductions of the Blacky Pictures were used as stimuli.

Two of the pictures were used as distractors; the other two consisted

of one conflict—related picture and one neutral picture. The stimuli

were arranged on cards so that each picture appeared in the same posi—

hit

tion (top, bottom, right, and left) an equal number of times. The ex»
periment was then divided into a subception phase and a defense phase.
In the subception phase, subjects were given a series of trials both be-
fore and after a situation considered to arouse fee ings of anxiety in
relation to certain psychosexual conflicts. Four pictures were pre-
sentcd simultaneously at each trial, and the subjects were required to
call which position stood out the most or was clearest. It was found
that the conflict-related stimulus was called significantly more often
in the second series of trials although both had been presented below
the level of "conscious" awareness. This was interpreted as vigilance,
or more aptly, subception to cues relevant to the threatening impulses.
In the defense phase of the experiment, the exposure time was in-
creased, and the subjects were required to locate both the neutral and
conflict-related stimulus. This exposure time was assumed to represent
a high level of awareness. It was hypothesized that at this level the
ego defense would operate. Blum found that subjects made significantly
more errors in locating the conflict—related stimulus than were made
in locating the neutral stimulus. The results were interpreted in
favor of the concept of perceptual defense. In a second xperiment,
Blum (1955) used essentially the same design but controlled more care-
fully for selective verbal report, familiarity, set, and antecedent
conditions. A variant in this experimental desifn permitted him to
show more clearly that defensive behavior could be traced to the per-
ceptual process itself. Subjects were led to expect that all eleven

Blacky Pictures would be presented tachistoscopically, but actually only

145

four of them were shown repeatedly. In the defense phase of the experi—
ment, subjects were asked to name which Blachy Pictures appeared in two
of the four positions indicated by the experimenter. Responses were
classified as to whether the pictures called by the subjects were prom
sent or absent and whether the subject used avoidance as a defense in
relation to the particular psychosexual dimension. Subjects avoided
calling the names of pictures related to their own conflicts, but only
when these pictures were actually presented. No avoidance occurred for
the neutral pictures, when they were actually presented. Blum concluded

resentation of the conflictual stimuli which

-‘~‘ J- r. J— r (w 4- . '. s“ .
that it tau the Sloliminal r

k
.-

k

led to perceptual defense.

Kelson (1955) using Blum's design, studied the perceptual behavior
of subjects classified as having high or low conflict to the psycho~
sexual dimensions represented by the Blacky test. Each subject was
.given a high or low conflict score for each psychosexual dimension as
well as a Specific defense preference for each dimension. Subjects
classified as high conflictmrepressers showed significantly more avoidm
ance behavior in perception than subjects of low conflictwother defense
preferences. He noted that subjects favoring projection as a defense
to a particular dimension tended to call thc stimulus related to that
dimension more often. On the basis of this behavior, Nelson suggested
that projection may take quite the opposite direction of perceptual dew
fcnse. He concluded that the perceptual process could be related to

1 x

specific personality dynamics of the individual.and emphasized the value

7

of psychoanalytic theory as a oasis for research in perception. In most

of the experiments dealing with perceptual defense, no explicit refer—
ence was made to psychoanalytic theory as a basis for hypothesis formu-
lation. Even Blum (1955) and Nelson (1955) who did make specific reform
ence to psychoanalytic concepts considered vigilance to be the anti_
thesis of defense. Chodorkoff (1956) had suggested that the better
adjusted individuals tend to show perceptual vigilance to threatening
stimuli. Neel (195h) found that subjects who had little conflict in
areas of sex and aggression showed vigilance to stimuli related to mild
sexual behavior but avoidance to stronger or more definitely sexual

stimuli. The opposite reactions occurred to mild and strong aggressive

 

stimuli. The high conflict group was less consistent but tended to
use avoidance to all types of sex and aggressive stimuli. These con~
tradictory results indicate the need for a better understanding of the
defense mechanisms.

Psychoanalytic concepts of defense. This discussion will, for the

 

most part, be based on the psychoanalytic theory of defense as presented
by Fenichel (l9h5). Defense, as has previously been stated, refers to
the various operations by which the individual attempts to avoid or deal
with the anxiety associated with instinctual expression. As an infant,
the organism is biologically helpless and, therefore, requires external
objects to satisfy instinctual demands. The external objects are not
always present so that the child is placed in a helpless position in
which he experiences pain. Experiences of this kind are considered to
lead to the first impression that instinctual impulses are also a

source of danger. This impression is further reinforced by threats and

h?

punishments from parental figures during the child's development.
These threats are effective because the child is also dependent upon
parental figures for maintenance of self—esteem. In order to avoid
feelings of anxiety, guilt, and the subsequent loss of self—
esteem, the individual blocks the discharge of certain instinctual
impulses. The warded—off instincts continue to strive for expression
and often become associated with other objects, attitudes or impulses
in an attempt to gain indirect discharge. These substitute feelings
or impulses are termed derivatives. If instinctual expression via the
derivatives also becomes anxiety producing then the derivatives are
defended against. Certain perceptual stimuli may constitute a part of
the anxiety producing derivatives. Then, defenses are established
against painful perceptions just as there are defenses agains anxiety—
producing instincts. *

Freud (1959) first held repr ssion to be synonymous with defense.
He distinguished between two types of repression. The more basic, he
called primal repreSsion, which referred to the denial into conscious—
ness of an instinct. The second was termed repression proper, which is
the pushing into the unconscious the derivatives associated with primal
repression. Fenichel (l9h5) prefers to call the first "repression
proper" which represents a continued counter—cathexis against the in—
stinct, and the second, "secondary repression" which represents a den
fense against derivatives. Freud later changed his view concerning re-
pression and considered it to be but one of a number of techniques

which might be utilized by the ego to protect itself from threat.

Anna Freud (l9h6) emphasizes this latter view. Defense is a gen-
eral designation for all the techniques used by the individual, whereas
repression refers to a specific method of defense. She further states:

"At particular periods in life and according to its own spe-

cific stricture, the individual ego selects now one defen—

sive method and now another....these it can employ in its

conflicts with instincts and in its defense against the 1i~

beration of affects".

This statement implies that the individual may use various methods
of defense against instincts and their derivatives. Some of these other
techniques of defense are termed: projection, reaction formation, isow
lation and intellectualization, which is a particular form of isolation.
These methods of defense permit the individual to avoid the oftmrepeated
acts of secondary repression. Of all the various techniques of defense,
repression, projection and intellectualization would seem most likely
to produce effects in perceptual.behavior.

In repression, the external stimulus is wardcd off just as the
threatening instinct is warded off. In projection, the offensive imw

I'Y

pulse is perceived in another person rather than in ones self. 'he in—
dividual is sensitized to an external threat in place of anxiety or
guilt associated with instinct expression. It is this sensitivity to
the external world that might be reflected in perceptual behavior.
Isolation ani its variant intellectualization could possibly have two
different effects upon perceptual behavior. Through the use of this
defense, the derivatives are freely admitted, but the emotional or afw
fective components have lost their connection. It becomes possible to

tolerate many aspects of ones impulses that would be extremely dis“

tressing otherwise. On the one hand, isolation may lea d to a disinm
terest in common ezzternal objects or stimuli in fay r of more abstract
pursuits. On the other hand, it may lead to a heightened sensitivity
to external stimuli because it may freely deal with such stimuli due
to the emotional detachment. Psychoanalytic theory would certainly
suggest that maladjusted individuals could show vigilance to threatenm
ing stimuli rather than avoid mzce behavior. It also provides some
hypotheses concernino the type of individual who may be expected to
show v igila nce in perceptual behavior. syehoana lytic principles have
led to a consideration of avoidanczz and vigilance behavior as methods
of defensa in dealing with ccni l.ctual stimuli.

Perceptual defense as avoidant and vigilant behavior. The view

”F..- '—

 

 

 

of vigilance as defensive behavior was pres ented by Carpenter,

.5

U

and Carpenter (1956) in a studv attenptin» to devise a method for dew

‘-

termining subjectsg oei ense mechanisms to various areas of conflic b.
Vigilance was termed a method of defense and defined as a heigiler ei
waroness which operates when the environment suggests the presence of
a threatening stimulus. The individual is sensitive to the conflictual
stimulus but to decrease the anxiety aroused by it, he Changes the
significance of it through intellectualizing, undoing, or projection.
The authors were also influenced by Goldstein's (1952) findings that
there were intramindividual differences in defense when dealing with
various areas of conflict.

On the basis of a sentence completion test containing sexual and

aggressive stems, Wiener E} 31, (I956) classified subjects into four

groups: hostility repressers, hostility sensitiz rs, sex repressers,

and sex sensitizers. An individual subject, for example, might be

0

lassified as a sex re resser and a hostility sensitizer, or he might
so either repression or sen itization for both areas of conflict. A
large minority did tend to use the sane defense for both areas. It was
concluded that ther would be little improvement over chance in atm
tempting to predict any one sub ject's mode of deiense in a pa: ti cular

1

area from knowledge oi his defense in another area of conflict.

In a further study, Carpenter, Wiener, and Carpenter (1956) pre~
dicted perceptual behavior from he subject's method of defense used
on the sentence completion test. Using the blurred booklet technique
in wl ich successivel" clearer presentations of words appear as the
pages are turned, it ”as iOILnd that subjects classified as six senSi-
tizers recognized words related to sex significantly faster than those
subjects clas ' fied as sex repressers. The same results were found for
the area of hostility. The results of this st wd empha size the need
for an accurate and independent meas re of defenses before predicting
perceptual behavior.

A further question concerning perceptual defense had been intro-
duced by Binder (195 8), calling for a renewed consideration of response
suppression. The CXperiments just reviewed, all used stimuli for wh ch
there was no reason to assume that subjects would delay report. Binder’s
findings suggest that this assumption may not always be correct. He
used subjects scoring high on various scales of the XKPI. He then

tested for the relationship between the amount of cue information accumu

ulated prior to tie recognition response and the poisonality variables
as measured by the MMPI. For the recognition task, he used a sequence
of test figures which pr vided more cue information with each succes—
sive exposure. The figures carried no emotional or threatening compo~
nents. Each figure was labeled, and subjects were given label to fi—
gure paired associate learning to criterion. The figures were then pre—
sented tachistoscopically, each presentation giving more cue informa—
tion concerning its identity. The longer a subject waited before label—
ing the figure, the higher was his recognition score. A significant
relationship was found between high scores on the Paranoia Scale (Pa)
of the MKPI and recognition scores on the perceptual task. Binder con—
cluded that the increasing distrustfulness and suspiciousness of high
23 individuals carries with it a tendency to give recognition responses
only after a relatively large number of ones are accumulated.

Binder's findings would certainly suggest that a delay in report
to ambiguous stimuli is related to certain personality characteristics
even though such stimuli are not of the socially disapproved variety.
This would apply particularly to subjects classified as obsessive-
compulsive or sensitizers. The former group is known to have conflicts
over self-expression and would be expected to show hesitancy in commit~
ting themselves in ambiguous situations. Subjects classed as sensiti_
zers often show obsessive or compulsive characteristics and might also
show hesitant behavior. Yet, in some eXperimonts just reviewed, these
subjects were found to have lower recognition thresholds for conflictual

stimuli. Perhaps, it is the emotional or threatening elements in the

I‘u

\‘f K

stimuli that constitute the crucial variable leading to a lower threshold
in ambiguous situations. A replication of Binder's >>erihcnt using
figm res which also had emoti nal connotations might Support such an
hypothesis. Bind r's study does suggest that the response suppre tssion
variable may not be as easily controlled as has previouer been thetlght.

In a recer t critique,f oldiar end (1958) has focused attention upon
the indicators of perceptual rcsponse. he questions the validity of

suen response measures as phenomenal report and the forced—choice ac-

 

 

curacy indicator, and precedes to illustrate the types of response bias

 

O
}--I

that can occur when using each type “ measure. his criticism is quite
pertinent, nd there is little doubt that the re m‘lts in many eizper i-
ments dealing with perceptual defense can be explained in t rms of rem
Sponse bias related to the particular measure of perception being utim
lized. Such criticism indicates *he need for greater methodological
rigor and an increased emphasis upon the use of psychophysical research
in suudvin7 motivational variables in perception.

Discussion of percent” 1 defense. ln Spite of efforts to minimize

 

the concept of defense as a perceptual problem or to subsume such beham
vi or under m re general principles of perception, a substantial body of
experimental data still exists to supportt the platw bility of such a
concept. Criticism has brought to light some very crucial variable<
and has led to much improvement in experimental methodology.

The most promising trend appears to be the one in which person~
ality variables are related to the perceptual process. The proponents

of the personalitymcentered approach have limited their view of percepw

tual defense to avoidant behavior. Although Lazarus (l95h) and Eriksen
(l95h) have explicitly stated that subjects may use ego defenses other
than repression when dealing with trreat, they also limited their view
of perceptual defense to avoidance behavior. Such behavior did not
always occur in response to threatening stimuli. Many subjects showed
a vigilance response to threat. Chodorkoff (1956) had reported that
the better adjusted subjects tended to use vigilance in dealing with
threatening stimuli. Nelson (1955 suggested that subjects using

projection as a defense might respond to the perceptual task in quite

 

the opposite manner from subjects using repression as an ego defense.
These differential results indicated the need fzr some external criteria
from which specific perceptual response could be predicted.

Carpenter et al. (1956) attempted to treat both avoidance and
vigilance in perception as a defense. however, these authors did not
investigate the perceptual behavior of a group in which little defense
to threat was infe‘red. Although they were able to predict avoidanee
and vigilant behavior in perception, they had little basis for concludm
ing that both types of perceptual response represented a defense, since
they failed to include a group which might be expected to show little
defense in perception. This is especially important in View of
Chodorkoff's finding that better adjusted subjects use vigilance in
dealing with threatening stimuli. however, a defense approach does
provide a plausible emplanation for the occurrence of both types of

perceptual behavior to threatening stimuli.

In the present investigation, a more specific study of the ego

\II
C“

defenses and their relationship to perceptual behavior was attempted.
It has been suggested by some critics that continued experimentation
dealing with motivational variables merely adds to the existing ple~
there of eXperimental data, but does little to increase our knowledge
or theory of perception. Since perceptual theory faces the problem of
accounting for individual differences, whether they be included or

,-,, + a; ' -- -, -" » ,. «
a systematic StudJ 01 the LOU“

minimized in perceptual experimentation,
ditions under which such differences occur is still of major importance.
This type of research can lead to the resolution of existing contradicm

tions and provide the groundwork for dealing with thorny problem of the

mediating mechanisms of motivational variables in perception.

THE PURPOSE OF THE PRESENT II'ESTIGATIOh

«”1 .

The present study had a twofold purpose. rirst, it was designed

to investigate the subception concept under conditions which were conm

 

sidered to be a more adequate test of so called "unconscious” discrimm
inations. Secondly, an attempt was made to eatend the concept of 235”
c’ptual defense to include both avoidant and vigilant behavior.

From the foregoing review of the literature, it was noted that two
theories of subception had evolved, with supporting experimental.data
for each. The first theory recognizes two processes in perception, the
affective and veridical. The second theory accepts only the veridical
process in perception. roponents of the dual process theory of sub”
ception first postulated that affective perceptual discriminations ocm

cur at "unconscious" levels and are more precise than veridical discrim~

inations. t was further hypothesized that affectite discriminations
involve lower centers of cortical integration and function either in-
dependently or tend to disrupt higher levels of cortical integration.

In order to avoid the assumption of "unconscious" processes, a
variation of the dual process theory was presented: which accounts for
subception in terms of the partial information afforded by the stimulus.
This partial information elicits the affective response which in turn
either facilitates or isrupts the categorization or integrating mech-
anisms in veridical discrimination. The most recent variation of dual
process theory considers affective and veridical discriminations to be
concurrent response systems, neither of which is more accurate than
the other.

The theory limiting perception to a veridical process states that
perception is dependent upon the individual's particular hierarchy of
rOSponse hypotheses which are based upon his past experiences. The
more dominant an hypothesis in the individual’s hierarchy, the less
stimulus information is needed to confirm it, and subception seems to
occur. 'Nhen the stimuli are related to less dominant hypotheses, more
stimulus information is needed to produce shifts in the hierarchy. rem
sulting in raised veridical thresholds or perceptual defense. Liotivaw
tional variables serve in determining the individual's particular rem
sponse hierarchy.

Discrcpencies in research data and subsequent differences in them
ory are related in a large part to the difficulty of defining percep~

tual thresholds. "Awareness" or veridical thresholds have been defined

generally in terms of the stimulus magnitude which permitted subjects

to make total identification of the stimuli. Stimulus magnitudes below
this level were considered to be beltw awareness and haie also been
given the term "below consciousness". however, it has been demonstra,ed
that awareness is a continuous variable and not a discrete "information~
no information" category.

A consideration of awareness as a continuous variable poses cer—
tain probl‘ms. Host subception experiments have used the degree of
asynchrony between a verbal accuracy measure and an autonomic response
measure as evidence for the subception phenomenon. This autonomic re“
sponse has been termed the affeCLive discriminatory process. blur
(l9Eh) and Dixon (1958) have also reported 'he subception effect, using
a verbal accuracy measure. Their results suggest that subception might
be more than an affective process and could involve higher centers of
cortical integration. The question arises as to which is the most ad-
equate indicator of subception. Also, is there more than one type of
response involved, and where along the awareness continuum does subm
ception begin or cease to operate? Blum and Dixon employed their verw
bal indicator at stimulus magnitudes at the extreme, low end of the
awareness continuum. On the other hand, Voor (1956) found that as the
lower end of the continuum was approacied, the autonomic response was
no longer elicited. Using an autonomic indicator at stimulus magniw

tudes permitting from 50% to total identification, subception has been

 

demonstrated, but when a verbal accuracy indicator is used at such

stimulus magnitudes, the result is perceptual defense.

-—-.—. - .._.

 

It is apparent that a more systematic approach to the study of
perception is needed, with explicit emphasis placed upon the awareness
variable and the types of indicators used to measure subception. A
starting point would be to test for the subception phenomenon under
stimulus conditions representing the lowest extreme of the awareness
continuum, using both a verbal and an autonomic measure. Such a test
would appear to be crucial to a theory of "unconscio s" perceptual
discrimination. Dixon (1958) conducted such a study and obtained sig~
nificant results. However, his results were based upon data obtained
from only seven subjects, and his experimental design is subject to
certain methodological criticisms. he does not indicate whether the
conflict and neutral word stimuli were equated for length, or fre—
quency of usage in the English language. The possibility also exists
that his verbal accuracy indicator was subject to response bias. If
Dixon's results could he substantiated under more adequately controlled
experimental conditions, this would tend to place other theoretical
views which explain subception in terms of partial awareness or strict
veridical processes in serious question.

The first part of the present experiment was designed, therefore,
to test for the subception phenomenon under stimulus conditions similar
to those used by Dixon. The stimulus conditions represent the extreme
lower end of the awareness continuum at which no information is avail—
able to the subject as measured by phenomenological report. A more
direct and immediate verbal accuracy indicator was used to test for

o r71

perceptual discrimination. ihis p rt of the ixperiment was also de_

58

o
*1

signed to control for the usual sources of response bias such as stimw
ulus frequency, differences in visual acuity, and response suppression.
The second purpose of this investigation was to broaden the con~
cept of perceptual defense to include both a heightened responsivity
as well as avoidapt behavior to threatening stimuli, when such stimuli
are presented at magnitudes permitting a phenomenal report of awareness
but below the level of total identification. The extension of the con~
cept of perceptual defense to include both types of behavior could acw
count in part for the contradictory results reported in the literature.
Perceptual defense and perceptual sensitization have been consi~
dered most often to be antithetical processes. This seems quite logi-
cal when sensitization refers to a heightened responsivity to stimuli
related to values, interests, or acceptable needs. However, a height-
ened responsivity to threatening stimuli has also been demonstrated.
There has been little attempt to differentiate this increased respon—
sivity in terms of the stimuli eliciting such behavior. An initial
effort was made to term the response to values or interests, percep-

tual sensitization, and to call the response to threatening stimuli,

 

perceptual.vigilance. However, these two terms continue to be used
interchangeably, and the term vigilance has also been used in refer~
ring to subception.

The lack of differentiation between the two responses has led to
a certain amount of confusion in the literature and to stated contram

ictions that actually do not exist. A distinction of response in

terms of the eliciting stimuli would appear quite necessary} An active,

59

selective process mav be operating to permit the indi. idual to recognize
value-related stimuli more quick y, but one could harlly infcr'such a
process in relation to threatening s in uli unless such behavior were
viewed within a framework of perceptual ceic use.

It has been emphasized that the concept of perceptual defense had
its roots, at least implicitly, in psychoanalytic t heorj Although it

theory supports the 'iew that individuals may

(.0

is recognim ed that thi
use other mechanisms besides ieiiession in dealing with threat, never~
theless, defense in perception has been restricted to avoidance behavw
ior. Within the framework of psychoanalytic theory, the assumption that
because an individual does not show avoidance in perception he is not
showing defense is certainly questionable. From the previous discussion
of the concepts of ego defense, it was sugges ted that persons using pro-
jection as a major defense, since they are considered to be more sensiw
ivc to the external enviro nment, would show an increased responsivity
to threatening stimuli in pcrciptual behavior . the oeiensi‘ve aspect of
such be ehavior lies in the fact that it permits the individual to locate
the source of threat as being external to the self, thereby, decreasing
feelings of anxiety or g uilt and maintaining self esteem. It was fu -
ther suggested that persons using other moo} anisms of defense, such as
intellec ualiz ation or isolation might also show an increased responu
sivity in perception, although this need not necessarily be so. The
present investigation represents an initial step toward a more system
matic study of specific ego defenses in perception. It is focus ed upon
subjects preferring avoidance and projection as the major defense techm

niques.

60

A study was reported by Carpenter, Wiener, and Carpenter, (1956)
which attempted to treat both avoidance and vigilance ir perception as
a defense. Subjects were classified as either repressers or sensiti—
zers on the basis of their responses to a sentence completion test, and
the subsequent perceptual behavior was predicted. Although between
group differences were significant in the predicted direction, the au—
thors had no way of actually substantiating the defense hypothesis
since they id not include a standard or control based upon the per-
formance of a group showing little or no defense. Such a group would
also be expected to have lower recognition thresholds than the represser
group. Not all subjects taking the sentence comnletion test were judged
as having conflict and showing defense. A group of such subjects could
have served as a control from vhich perceptual defense for both the re-
presser and sensitizer groups might have been inferred. Also, combining
subjects using defense mechanisms other than repression would appear
questionable in spite of the significant results. Such a combination
assumes that the subjects compose a homogeneous group. Such an assumpw
tion seems unwarranted at present, especially in view of the fact that
Goldstein, (1952) found an affinity between the use of avoidance and
reaction formation defense. Subjects using the latter defense were

placed in the sensitizer group in the Carpenter et al. study.

 

The position often taken is that all persons show some defense.
However, if this position is broadened to the extent that it makes dif—
ferentiation among individuals using the same defense impossible, then

it becomes meaningless. In dealing with a so called "normal" college

61

population, it would seem justifiable to assume at least a phenomeno-
logical continuum of defensiveness, ranging from little or no defense
to pathological defensiveness. Subjects at the lower end of this con—
tinuum might be eipected to show less defense in perception and more
flexibility in their approach to various types of threatening situa—
tions. Individuals at the more extreme and of the continuum should be
more rigid in their use of defense, and the effects of such defensive—
ness should be demonstrable in numerous facets of the individual's be-
havior. Consequently, these subjects at the more extreme end of the
defense continuum who used Avoidance and Projection as major defenses
were selected for study. Subjects at the lower end of the continuum
were selected to serve as a control from which defensiveness in percep-
tion might be inferred.

To select subjects on the basis of degree of threat and type of
defense utilized, the Defense Preference Inquiry (Blum, 1956) was
chosen (See method chapter). Using this assessment instrument, sub~
jects may be classified as either "general" or "specific" defenders.
The "general" defenders are considered to be more rigidly defensive and
more maladjusted than the "specific" defenders. Consequent y, the
"general" defender would be expected to show defense in perception.

The term "specific" defender is a somewhat misleading one because
it refers to the majority of college subjects on whom the instrument
was used. Since the Defense Preference Inquiry is a forced—choice
ranking technique, subjects must select a particular type of defensive

behavior for each conflict dimension presented or else refuse the task.

62

It can hardly be inferred under these forced—choice conditions that the
subject has conflict on each area presented and utilizes the type of de—
fense chosen. The "specific" defender group actually represents a less
defensive and better adjusted group than the ”general" defenders and
would be expected to show little defense in perception. From these con-
siderations, the specific hypotheses were formulated to test for sub-
ception and perceptual defense.

here than a decade of experimentation using the "New Look" approach
to perception has clearly demonstrated the multiplicitly of variables
affecting recognition thresholds. Geldiamond (l?§8) in an excellent re-
view of the "New Look" methodology states that in almost every case, the
experimental results can be accounted for in terms of response bias
which is extraneous to perception. His criticism, in most instances,
seems well taken.

It must be recognized that the individual enters the perceptual
situation with certain built~in response biases as the result of past
events and conditioning. It is not possible nor even desirable to elim-
inate these. In a perceptual experiment, the effects of response bias
must be systematically examined. To do this, the experimenter must in—
corporate inte his experimental design specific measures to test the

effect of response bias. The present experiment was so designed.

STATEIENT OF HYPOTHESES

The specific hypotheses formulated to test for Subception and

Perceptual Defense are as follows:

63

SUBCBPTION

Hriethesis I. All subjects will make si nificantl'r more discrim~
a _ a J

 

inations of the conflict stimuli as compared to the neutral stimuli

when such stimuli are presented below the level of awareness as measm

ured by phenomenal report.

PERCEPTUAL DEFENSE

Hypothesis

IF!
0

Subjects using av-idance as a "general defense"

 

 

against psychesexual impulses will show perceptual avoidance when the
conflict stimuli are presented at a stimulus magnitude permitting
awareness, but below veridical threshold.

Hypothesis II. Subjects using projection as a "general deferse"

 

against psychosexual impulses will show perceptual vigilance when the
conflict stimuli are presented at a stimulus magnitude permitting
awareness, but below veridical threshold.

hypothesis III. Subiects classified as "specific defenders"
against psychosexual impulses will not show perceptual defense to the
conflict stimuli when such stimuli are presented at a stimulus tagm

nitude permitting awareness, but below veridical threshold.

eh
LETHOD

The experiment consisted of three phases: The Defense Assessment
the Subception, and the Perceptual Defense phase. The Defense Assess—
ment phase was used to select 88 for the three experimental groups. The
Subception phase tested for perceptual iscriminatiens when stimuli were
presented below levels of verbal awareness, and the Perceptual Defense
phase was designed to test for both avoidance and vigilance defenses in
perception.

The initial problem confronting E was the selection of an adequate
instrument to assess SS defensive behavior and the selection of percepm
tual stimuli that could be equated for ease of recognition based upon
their structural characteristics. Several defense assessment instru~
ments were considered and a pilot study was conducted to equate the pore

ceptual stimuli.
SELECTION OF A DEFENSE MEASURE AND PERCEPTUAL STIMULI

Three different instruments were tested for their applicability in
determining the methods of defense used by subjects to handle material
related to aggression and sex. The first of these is a Phrase Lssocian
tion Test devised by Heath (1956), who presented a convincing argument
for the use of more highly structured tests in the study of ego defenses.
This test is similar to the word association tests but uses phrases in—
stead of words and requires phrases or sentences as responses. The sec—
ond instrument tested is a highly structured sentence completion test
containing sexual and aggressive stems. These stems were taken from a

larger test devised by Forer (1950). While broad category classifica-

tion was possible on the basis of these two instruments, considexable
difficulty was encountered in attempting to classify subjectst rem
spenses into specific types of defenses. Consequently, a third measure
'was selected for use in the present experiment. This measure is the
Defense Preference IDQuir‘ (KS3) as revised by Blum (1956).

The Defense Preference Inquiry (DPI) provides an objective method
of determining subjects‘ preferred ego defenses in dealing with various
psychesexual conflicts in terms of five defense mechanisms as presented
in psychoanalytic theory. The DPI is used in conjunction with the
Blacky Pictures (thQ). In the DPI, the subjects are asked to rank a
series of five statements, each describing some feeling or behavior
judged to represent a particular tgpe of defense, in terms of new Blacky
may be feeling or acting in a particular picture. The instructions are
formulated to encourage subjects' identification with Blacky, thereby,
revealing their own personal reactions. Spontaneity is further faCllim
tatcd by very short time limits which preclude the possibility of dew
liberation in assigning the ranks. The serial.pesitions of the five
defenses are rotated in a systematic manner throughout the inquiry.

The five alternative defenses are: avoidance (the generic term ceincd
for the repressienmdenial family), regression, reaction formation, prom
jection, and intellectualization.

An illustrative set of the DPI items relating to Oral Sadism are
presented here. The complete set appear in Appendix B of this paper.

The subject is first instructed to write a two minute story about the

particular Blacky Picture. he is then asked to rank the corresponding

DPI items. For esahfle:

Rank all of the following statements according to how
well.they seem to fit the situation.

1 ' fits best h = fits fourth best
2 L fits second best 5 3 fits worst
3 2 fits third best

(Rg) A. When Blacky gets angry, he often throws a temper
tantrum like he did in his earlier days.

(RF) B. Blacky tries to pretend that he‘s ferocious. but
when Hamma is around he is sure to be overly
gentle, calm, and wellmbehaved.

(Int) C. Blacky is a firm believer in the idea of releasm
ing one's aggressions, so he feels justified in

ripping Hanna's colar here.

(AV) D. Blaeky is so intent on chewing the collar to pieces
that he doesn‘t even realize it belongs to karma“

(2) E. In Blacky's own way of thinking, his family has

been treating him so unfairly that he feels en-

titled to chew up the collar.
Each group of DPI items is related to a particular psych0wsexual
dimension.

The issues of reliability and validity must be considered when
using such a technique. The DPI has been used in an extensive rem
search program at the University of Michigan (1956): and most of the
reliability and validity data comes from this experimentation. The
test-retest measures resulted in product moment correlations of .LG
and .hé. Of the first choices on one administration, 73 per cent ocw
curred as either first or second choices on the second administration.
last choices proved to be almost as stable as first choices.

Construct validity has been demonstrated in a number of studies.

Blum (195h, 1955) and Nelson (1955 both found that avoidance preform

ences on the DPl were signifies ntl" as ' ' ated with perceptual defense
in a task involving the presentation of the Blacl;y Pic ures. ln the

first fraternity study at the University of Kiohigan, avoid; n39 p11.
ferenee was significantlya associated tit “.fo1ge tirg of pictxres 1n 1
recall taskgs “10 ion of neutral rath r +:han conflict «relevant solu~
tions in word completion and anagram experiments.
Significant differences between the "general" deiender who tend;
to choose the sare defense ior all psr hosexual dimension sand the mrre
,. .f‘

common "specific" defender xh tends to choose different d01 awn Cm 1or

each psychosexual dimension have been reported. Shi1e (i9'fl1) demon“
strated that tne general defender was signijieantly more malaIijust ted r
two genotypic measures~~the Munroe Inspection Technique for the Rorschach;
and the spontaneous Stories on the Blacky Pictures. Segal (1)5h) diVid d
female subjects into the general and specific defender categories. Bo1h
groups were equated for strength of hostile and dependent impulses on
basis of their ”AT and Blacky stories. Her results indicated that sub“
jects in the general defender category were less able to eXpress fee1~
ings of he ility or dependency towards their mothers in a personal
interview situation than subjects in the speci ie defender category.

All of the validity studies just reviewed haVe dealt onl; with

avoidance as a defense reference. As yet, little evidence has been

 

 

reported for U G Validitv of the other defense items. A study by Cohen
(1958) using the DPI has indicated that individuals sha aring a preform
once for projection as a defense show m re negative inter-personal re»

actions than paired individuals preferring different delcnse mechanism.

68

The present study also served as a test of construct validity of the

projection defense items.

 

Since the DPI was administered to what may be considered a normal
college population, it is reasonable to assume that subjects will.range
from little or no conflict in relation to sexual.or aggressive impulses
to rather intense conflict in these areas. On the basis of previous
research, subjects having little conflict should fall within the specim
fic defender category. Subjects with more intense conflicts should
fall within the general defense category. It was hypothesized that
this latter group would be more likely to show defensive behavior in
perception. Consequently, these subjects who were classified as gen“
eral defenders and who preferred either avoidance or projection as a
defense were selected for this experiment.

To place subjects in the two defense categories, eight Blacky
Pictures and the corresponding DPI items were used. Three of the pic—
tures were selected because they related to sexual.behavior3 four picm
tures because they were related to hostile impulses; and an eighth
picture was chosen to complete the series. It was felt that eight pic—
tures were sufficient to give an adequate sample of subjects defense
preferences, as well as an adequate measure of defense consistency. The

eight pictures chosen, represent the following psychosexual dimensions:

Ora1.Eroticism I masturbation Guilt V
Oral Sadism II Overt Aggression VII
Anal Sadism III Sibling Rivalry VIII
Oedipal Intensity IV Guilt Feelings IX

The spontaneous stories to the eight pictures served to sensitize sub“

jects to the various situations depicted. It was assumed that in

69

composing stories to the stimuli that the subjects's own conflicts and
feelings would be aroused and would influence his iefeise rankings.

Perceptual stimuli. In selecting stimuli for the perceptual task,

 

the variables of stimulus familiarity and physical similarity 11* to be
considered. Familiarity was controlled by using stimuli with which
subjects had had no previous experience. matching the stimuli in terms
of their physical properties for ease of recognition was a more diffiw
cult task. This was done by trial.and error matching of the neutral
and conflict stimuli and by altering some of the characteristics of the
neutral.stimulus.

Small black and white drawings of P‘acky as depicted in the Blackv
Pictures (Blum, 19h9) were used as perceptual.stimuli. There were sevw

"lie

eral advantages in using this tyse of stimulus material- First,
Blacky Pictures have been used in previous perceptual studies, and
avoidance behavior to such stimuli has been demonstrated. Second, such

drawings are easily altered to permit matching for =ase of recognition.

Pilot study. A pilot study was conducted to equate the perieptual

 

stimuli for ease of recognition. To do this, a test group was needed

in which the motivational variable was minimized. It had been shown by
Shire (lQSh), and Segal (l95h) that 85 classified as "specific defenders”
on the Defense Preference Inquiry were less maladjusted than "general

def nders". It was hypothesized that the specific defenders would 810”
less defensive behavior, and that any differential response to the per“
ceptual stimuli would be due mainly to the physical characteristics of

he stimulus material. The Defense Preference Inquiry was administered

70

to the pilot group, and a random sample of "specific defenders" was
selected for the perceptual experiment.

Six drawings of Blacky were selected as the perceptual sthmzli.
These drawings were arranged in three pairs. The first pair was law
beled Sex-Neutral, the second pair, Aggression-Neutral, and the third
pair, which consisted of two neutral drawings, was labeled the Acuity
Test pair. The experimental design used to equate each member of a
pair with the other is a four stimulus configuration. This‘design aid
procedure has been described in detail by Blum (IOSh). Four Blacky
Pictures were flashed simultaneously in a tachistoscope. Two of the
pictures were termed the ritical stimuli, which were one of the above
mentioned pairs. The other two pictures served as distractors. 55
were instructed to locate (top, bottom, right, or left) one of the

ritical pictures for a series of ten trials and then the other critiw
cal picture for a series of ten trials. The same procedure was repeated
for each critical pair.

The euposure speed was varied until a setting was found at which
55 could locate one or both of the critical stimuli with approximately
75% accuracy. Errors in locating each critical stimulus were tabulan
ted, and the neutral stimulus of each pair was changed or altered until
no significant error difference between critical stimuli was found.

During the test series, 53 were questioned about any stimulus
characteristics that seemed to help them in recognition. It was found
that 85 reports were quite unreliable. They would describe certain

characteristics as aiding recognition of a stimulus, when in fact, they

were making more errors in trying to locate that ,articular stimulus.
55 would also describe characteristics of a stimulus tia' actually did
not exist. Consequentl3, a trial and error method was used in maith;g
the stimuli. Although the general position of the figures had to be
similar, the figures had to be dis tinctl3 diif eren t to permit discrim~
ination. The three pairs of critical stimuli selected are presented in
AppendL" C of this paper.

A second sample of Ss taken from the " .3pe cL ic dei ‘endei" category,
as determined by the Dei‘em: e Faei'erer ce Inquir3, were used to QLTP“*1U8
and exposure setting at which an initial measure of visual acuity could
be taken for use in the perceptual defense analysis. The Acuity Test
stimuli which consisted of two neutral pictures of Blacky in a standing
position were reproduced on cars} so that e ch memaer of a pair apm
peared once on the right and once on the left side of a card. A blank

space of equal area was outlined on each card between the Acuitv Test

stimuli. Each member of the Acuit“ Test pair wa:3 also reproducei iri1~

v

\

vidually. When these individual.stimuli were flushed, they appc cared in
the blank space between the Acuity Test stimuli. The Acuity Test pair
remained constant in Ss' visual field, except that the rightwleft side
positions were reversed on alternate trials. The individual stirnuli
were exposed in a random m nner for a series of 20 trials, and 53 were
asked to state which of the con tan stimuli had been flashed on each
trial.

The exposure time for the individual stixnulus presentta ions was

varied until.a setting was found that enabled Ss to make appro: initel'

72

50% correct discriminations of the Acuity Test stimuli. This exposure
setting remained constant for all 83 ‘n the Acuity Lest series and Per~
'ceptual Defense series oi the main experiment. ?he exposure setting
selected was .15 seconds with illuminosity readings as follows:
Adapting fiel; .OLZ ft. lamherts
Adapting field

plus .UBl “t. lamherts
Test field

The second illuminosity reading represents the simultaneous exposure of
both adapting and test fields. Illumination in the adapting iield rem
mained constant during the test field exposure.

A third sample of Se taken from the "specific defender“ group was
used to test for adaptation to repeated exposures of the stimuli when
these are presented at stimulus magnitudes below verbal awareness.

Each stimulus comprising the Sex-Neutral and Aggressionuyeutral pairs

was first presented to S at a .10 second exposure. Exposure time was

decreased .01 second decrements until S reported no longer seeing any~

’1

thing in the visual field. The shortest exposure time for the h stimuli

was used as the exposure setting for the series. If thresholds tended
V

to become lower with succeeding trials, then it was expected that $3
would show more accuracy in discrimination near the end of the series.
Responses were divided into 5 blocks of trials with h trials in each
block. Group error means were computed for each trial block. No sysw
tematic increase in response accuracy'with succeeding blocks of trials

was noted. 53 did not report seeing the stimuli during the 20 trial

series.

73

On the basis of tLe pilot study, the pores; itual stimuli to he used
in the m in experiment were CQL ated i‘or nlysical chara:t ;ristics which
might influence recognition thresholds, and an eXposure time was estar~
lished which rem aincd constant for all SS in the Perceptual Defense
Phase of the main experiment. A test of the adecuany of phenomenal.reu
port as an indicator of below awareicss thresholds was also coniu" ed.
Recognition thresholds did not vary with repeated presentations of the

stimuli.

SUBJECTS

A group of 31h students from introductor3 classes in psychology
took part in the defense assessment plase of the e} periment. 85 were
taken from both morning and afternoon classes, nd from classes meeting
on each day of the school week. From this larger pool, three experiu
mental groups were selected on the twa is of the type of defense used in
dealing'with a number of psychosexual con 'licts. Using the Defense
Preference Inquiry as a measure of defense, 88 were required to meet
two criteria for placer~ent nto the experimental groups. The first
criterion used was the number of first place rankings of a particular
defense mechanism to the eig ht psychose11ual dimea nsions. SS'WhO sel eted

the same defense mechanism four or more times as their first chOice were

 

classified as "general" defenders . A second criterion measure was then
applied to Ss in the "gen ral” defender group who had selected avoidance
and projection as the preferred defenses. The second criterion dealt

with the overall consistency of rankings to the eight psycl‘zosexual.

dimensions. To measure consistency among the eight sets of ranks for
each S, the Coefficient of Concordance (Edwards, lQSD) was used. The
cut off point for inclusion in one of the two groups was a value of W
that was significant at the .20 levels

Twenty-six 83 who met both selection criteria were placed in the
Avoidance group. Three of these 53 indicated that they did not wish
to participate in the perceptual.phases of the experiment and were
dropped. A fourth S would not make a differential response during the
perceptual phases of the experiment, even when the exoosure time was
increased to a 2 sec. exposure. Situational anxiety was quite eVident.
Because of her refusal or inability to follow instructions, this S was
not included. The final Avoidance group contained 22 SS, 10 male and
12 female.

Twenty—one 53 were originally placed in the Projection group.

Two 53 did not appear for the perceptual phases of the experiment, and
one S stated that he did not wish to participate further in the eXperim
ment. The final Projection group was composed of 18 Ss, ll_male and

7 female.

The remainder of the 55 who took part in the Defense Assessment
phase were classified as "specific" defenders. A random sample of 30
85 from this group was selected for placement in the No Defense experi»
mental group. This group was given the he defgpsg label since it was
hypothesized that the specific defenders", as determined by the DPI
would not show defense in perception. Five 85 were excluded from the

original selection, either because they did not appear or did not wish

75

to participate in the perceptual phases of the experiment. The final
No Defense group consisted of 25 85, ll male and lh female. A total of
65 53 completed all phases of the experiment. See Appendix A for the

list of experimental Ss.

LL'JI'ERLKIS AND APPARATUS

To measure 55 preferred defense mechanisms, eight Blacky Pictures
and the corresponding DPI items were used (See Appendix B). The DPI
items were presented in booklet form, and blank pages were provided
between each set of DPI items for recording the spontaneous stories to
the Blacky Pictures. Black and white slides of the eight Blacky Pic»
tures were projected upon a screen for group administration.

Six 1" by 2" black and white dravings of Blacky were used as perm
ceptual stimuli (See Appendix C). The drawings were arranged in three
pairs and reproduced so that each member of a particular pair appeared
both on the right and on the left side of a card. The stimulus pairs
were mounted upon light gray posterboard cards. The first pair of
drawings were designated as neutral, and depicted Blacky in two dif~
ferent standing positions. This pair was labeled the Acuity Test stim~
uli. The second pair was labeled Sexuheutral and consisted of a drawing
of Blacky as pictured on Card V (insturbation Guilt) of the Blacky Picw
tures and a picture of Blacky in a neutral, reclining position. The
third pair was labeled Aggression—Neutral and consisted of a drawing of
Blackv as depicted on card II (Oral Sadism) and a neutral picture of

I.

Blacky in a standing position.

76

A blank space of equal area was outlined between the two drawings
comprising a stimulus pair. A fixation point was located in the center
of each outlined space. The stimulus pairs were exposed to S from the
adapting field of the tachistoscope and remained a constant stimulus for
each trial.

The 6 drawings constituting the 3 stimulus pairs were reproduced
individually on gray postcrboard cards. These individual stimuli were
exposed from the test field of the tachistoscope and appeared in the
outlined space between the constant stimulus pairs. Because of the
mirror arrangement in the tachistoscope, the constant stimulus pairs
were reproduced in reverse position to the individual reproductions, so
that upon ex,osure in the tachistoscope all pictures would appear in
the same direction.

A Gerbrands tachistoscope was used in the perceptual phases of the
experiment. A slight modification of the instrument was necessary to
permit a constant power supply to the lamps in the adapting field. This
was done by wiring around the micro—switch so that the contacts with the
adapting field remained constant during the test field exposure. Illum-
ination was reduced by the use of gray art paper filters and remained
constant for all subjects. Illuminosity was measured by a Macbeth
illuminometerl, with readings as follows:

Adapting field .Oh2 ft. lamberts

Adapting plus test field .031 ft. lamberts

 

1Grateful acknowledgement is made to Dr. S. H. Bartley for taking
the illuminosity measures.

2"?

Th- second reading represents the illuminosi y conditions during the
simultaneous exposure of both the adapting and test fields. Exposure
settings for the tachistoscopic presentation of stimuli in the subcepw
tion series were determined individually for each S. The exposure time

of .15 sec., which was determined in the pilot study, was used for all

55 in the Perceptual Defence phase of the experiment.
THE KEASURES

“he measures used in this investigation were termed "indicators"
after Goldiamond (1958). The indicator used to determine below awarew
ness thresholds for the subception phase was labeled a phenomenal report
indicator. This indicator represents a judgment or opinion. 55 stated
when they could no longer see anything flashed in the visual field.

The indicator used to test for subception and perceptual defense
effects was termed a forceduchoice accuracy indicator. 85' responses
were either correct or incorr ct. In the present study, this indicator
included both an accuracy and a category dimension.

Subception was measured in terms of the asynchrony between phenom
menal report and the forced-choice accuracy indicator. Perceptual dew
fense was measured in terms of inter—group and inter-category response

accuracy.
PROCEDURE

DEFENSE ASSESSMEIT

The method of administering the Blacky Pictures ani the correspondw

ing DPI items was essentially that described by Blum (1956). 53 were
seen in groups and the Blacky Pictures were projected on a s reen. Ss
wrote their spontaneous stories in the test booklets provided by E, and
then ranked the DPI items. The following verbal instructions were
given:

I am going to show you a series of pictures about a dog
named Blacky. These are like the cartoons you see in th
paper. I will present one picture at a time, and I want
you to make up a short story about each one. Tell what is
happening in the picture any why it is happening. Since
this is a test of how good your imagination can be, try to
write vividly about how the characters feel. You will have
two minutes for each stony. It is desirable to write as
much as possible within the time limit. I will signal when
there are only thirty seconds left to finish a story.

 

After the two minutes for the story “e up, you will be

asked to turn the page, where you will find a series of

five statements describing some possible feelings or re-
actions of Blacky in the cartoon you just saw. I want you

to rank these statements as to how well you think they repre-
sent the way Blacky seems to be feeling or acting in that
situation. Naturally, there are no right or wrong answers
involved. Just write a "l" alongside the statement that fits
best, a "2" alongside the second best, a "3“ for third best,
a "h" for the one that fits fourth best, and a "S" for the
one that fits worst. Regardless of how well or poorly the
statements seem to fit, be sure to rank them all "1" through
"5". Never leave a statement unranked. You will have 55
seconds for each set of rankings so you will have to work
rapidly. I will indicate when there are only ten seconds
left.

 

 

Remember then, for each picture you will have two tasksmu
first to write a two-minute, imaginative story about how
Blacky is feeling or acting, and second, to rank order the
five statements about the picture. Never turn a page until

I give the signal and never look back at what is already done.

85' rankings for the eight sets of DPI items were recorded. On the basis
of these rankings, 85 were classified as either "general" or "specific"

defenders. Ss classified as general defenders who selected avoidance

and projection as the preferred defenses and who showed consistency

[.1

I1

their rankings were placed into two experimental groups, designat d s

O
93

the Avoidance Defense and Projection Defense groups. A random sample
was taken from the specific defender group and was designated as the Ho

mg.
i

Defense experimental group. 11080 53 took part in the perceptual phases

of the experiment.

SUBCEPTION PHASE

To test the subception hypothesis, it was nee ssary to determine
an exposure level at which 85 could no longer report seeing anything in
the visual field. This exposure time was termed the below awareness
threshold exposure. below threshold exposure was determined for each
stimulus of the S x-Neutral and Aggression—Neutral pairs. Each stimulus
was first presented at a .10 sec. exposure and was decreased by .01 sec.
decrements until S no longer reported seeing anything in the visual field.
The order of presentation varied systematically for each S. The shortest
exposure time recorded for the four stimuli was used as the exposure
setting for the discrimination task in the subception series. 53 were
seated at the tachistoscope and were given the following instructions:

Do you see the black dot in the center of the screen? When

I say ready, focus on the black dot. I will then flash a

picture. The picture will be flashed quite rapidly so you

probably won't know what it was. This doesn't matter. I

will continue flashing this picture but each time the expo—

sure will be a little shorter. I want you to tell me when

you no longer see anything flashed. Be sure that you don't

see anything. I will always say ready right before I flash

the picture.

The procedure was repeated for all four of the stimuli comprising the

conflict—neutral pairs. 83 were then given a one minute rest period

80

before starting the subception task.

After the rest period, Ss were presented with the discrimination
task. An A — B — x type of experimental design was utilized. The Sex—
Eeutral and Aggression~heutral pairs were labeled either A or B. The
stimulus pairs were presented on alternate trials and remaired a een~
stant stimulus for that trial. Then either A or b of the constant
stimulus pair appearing on that particular trial was flashed in the
blank space between the constant stimuli. 5 indicated which picture
had been presented. The right~left side positions of each constant
stimulus pair was also reversed on alt‘rnate trials. Forty trials were
given so that each individual stimulus of the Conflict-Neutral pairs
was presented 10 times. Two set random orders for presentation of the
individual stimuli were used. The exposure time varied for each S.
However, the range was quite small, varying from .01 sec. to .03 see.
with illumination remaining constant for all 55. The following verbal
instructions were given:

Do you see the two pictures labeled A and B? They are small

drawings of the Blacky Pictures that I showed you in class.

Picture A shows Blacky discovering sex. B is a picture of

Blacky lying down. All right, now look at this pair. (E

removed the Sex—Heutral pair and presented the Aggressionw

Neutral pair). A is a picture of Blacky in a standing posim

tion and B shows Blackv chewin an ri " on Mamma's collar.

3 a

New notice the outlined space between the pictures, with the

small dot in the center. I am going to flash some pictures

which will appear in that blank space. Each time, the pic—

ture flashed will be either A or B of the pair that you see.

I want you to tell me which picture was flashed. Just call

out A or B.

The two pairs of pictures I have just show} you will be pre—

sented on alternate trials, but in each case the picture

flashed will be either A or B of the pair that you see on the

screen for that trial. Also, the positions of the pictures
making up a pair will be reversed on alternate trials, so

81

that if A aple

e'ar on the right for one trial, it will
appear on the l

s
eft on the next trial.

The pictures will be flashed so rapidly that you won't be

aware of anything happening on the screen, but each time

take a guess as to whether it was A or 3 that was flashed.

The pictures will not appear in any set order, and one may

appear more often than the other, so do not make vour next

choice on the basis of the previous guess. It has been

reported that people can still see things even when they

are not aware of it, and I want to see if this is so. Each

time then, take an independent guess as to which picture

was flashed. When I say ready, focus on the small dot in

the center of the outlined space. It will then flash the

picture.

Although the instructions appear onite lengthy, the task was actually
quite simple and 85 had no difficulty in comprehending what was exm
pected of them.

The order of presentation of the constant stimulus pairs was varied
among 58. Half of the 85 were presented with the Sex~heutral pair first
and then alternately with the Aggression—Neutral pair for the series.
The other half were presented with the Aggression—Neutral pair initially
and then alternately with the Sex—Neutral pair. 53 were given hO trials,

with a thirty second rest period after each set of 10 trials, followed

by a one minute rest period before continuing with the defense phase.

DEFENSE PHASE

In the defense phase, the stimulus magnitude was increased to a
level permitting some awareness of the stimulus presentations, but below
the threshold for veridical identification. A measure of individualvisw
ual acuity was first taken, using the Acuity Test stimuli. The exposure
time used in this phase of the experiment was .15 see. This exposure time

was determined by the pilot study and represents a level at which approxw

82

imately 50% correct discriminations of the Acuity Test stimuli had been
made. 83 were again required to make an A — B — x type discrimination.
The two acuity test stimuli were labeled A and B and remained constant in
Ss' visual field. Either A or B was then flashed individually in the
outlined space between the constant stimuli. The right-left side posi»
tions were reversed on alt rnate trials for each constant stimulus. 85
were given 20 trials with a 30 sec. rest period after the first ten
trials. The verbal instructions were similar to those used in the subcep~
tion discrimination task.

Now we are going to do about the same thing we just finished,

using these pictures. (E presented the Acuity Test stimulus

pair). Only this time the pictures will be flashed much

slower. Again, I will flash either A or B that you now see.

You tell me if it was picture A or B that was flashed. ‘When

I say ready, focus on the small dot in the center of the out~

lined space. I will then flash the picture. If you are un~

certain as to which picture was flashed, always take a guess.
Individual differences in subjective experience became apparent on this
task. Some 88 stated that they had to guess on most of the trials.
Others felt that they were able to make discriminations based upon stimn
ulus characteristics after a few trials. After the one linute rest
period, 35 were administered the defense series.

The Sex-Neutral and Aggr ssionmheutral stimulus pairs were again
used as constant stimuli in the defense series. The exposure setting
for the individual stimulus presentation was again .15 see. with illum—
ination also remaining constant for all Ss. The same A - B — x type of
discrimination task was used, and.Ss were given a series of hO trials.

A thirty second rest period was provided after each set of ten trials.

85 received the following instructions:

83

.1.

Now lets try the same thing again vith UhC pictures that

you have already seen. (E presented the Sex—Neutral and

then the Aggression—Neutral pairs). Again, you are to

tell me if it was picture A or B that was flashed. Re-

member to focus on the black dot in the center of the

outlined seace when I say ready. I will then flash the

picture.
The Sex—deutral and Aggression—Neutral constant stimuli were presented
alternate y for the hO trial perceptual defense series, and the right
left side positions of each member of the constant stimulus pairs were
reversed on alternate trials.

Table 1 presents a summary of the general operating procedure.
85 were first selected for defense preference and placement into the
experimental groups. Four measures were then taken for each S. First,
a measure of stimulus magnitude was taken for the below awareness ex—
posure. Second, 83' responses on the subception series were recorded.
The third measure was Ss' responses to the Acuity Test stimuli, and
fourth, 55' responses to the Perceptual Defense series were recorded.

The last three measures were used in the statistical treatment of the

results.

8b

TABLE 1

SUKLARY OF OPERATIONAL DESIGN

 

 

 

 

Phase Operations
Defense Administration of DPI, and selecm
Assessment tion of 53 based on two criteria.

h or more lst place rankings
of Avoidance or Projection as

preferred defenSe.

FJ
O

2. Total rank consistency as measw
ured by Coefficient of Concordw

(NICO o

 

MM~_~~ .-

 

Subception 1. Determination of awareness
thresholds for the h stimuli
constituting the Sexmneutral
and Aggressionmneutral pairs.

2. AD trial discrimination task.

 

Perceptual defense 1. 20 trial visual acuity test
series 1sing neutral stimuli.

2. hO trial defense series using
Sex~neutral and Aggrzss10n~
neutral stimulus pairs.

 

.—..-~ ~--—. h-

 

85
RESLLTS

SUBCEPTION FHA T

It was hypothesized that if the subception principle were to operate,
it would manifest itself in a greater response accuracg'by all groups to
the conflict stimulus category as compared to the neuxral stimulus eaten
gory. To test for the subception effect, the numser oi ”slates errors
to each stimulus category was tabulated for all 85 (Sce.Appenuix D).

Group error mears for each stir ulus categor were con 1puted and the
group mean error differences between stimu l 5 categories were tested for
significance. One tailed t tests for 1e*e‘.cd measures on the same Si

(Edwards, 1950) were used to test for significance of the category mean

error differences. F tests for homogeneity of variance oetwe en categew
ries were not significant. [no ortained mean error differenceS'were

tested against an hypothesized mean error difference of zero. Table 9

presents a summary of the statistical analysis. None of the oi itain ed

t values irovcd significant. The mean error difference in the n‘Qldl"“
group was in the unCXpected direction and was treated as equivalent to

the arsence of difference froza 0. There was no evidence ta 8; ppoit tre
h'PoThesi of a differential response accuracy'in favor of the conflict
stimulus category.

Since the conflict category'was composed of both a Sex and an Ag:
gression related stimulus, it was possible for 55 to have made a differw
ential intraucategory response, an occurrence which might not have been
reflected in the interwcategory co mr} arisen. To tes t for such an even'

the number of errors to each stimulL of the conflict category was tabuw

86

TflBLE 2

ERROR HEARS AND ELAN ERROR DIFFEREQCES BETHLEH

STIMULUS CATEGORIES FOR ALL GROUPS

1N THE SUBCEFTION BER ”5*

_.__ -. c. .,..a ..._. o - ,
_ - —.-,_ ......- -. _.. - ‘7 , , i

—--—————.- -_ .
- ~--- “0-. ~—.—.. — ..

~“—.—.—.M ~-—. - —

M errors LI errors V .3
Group Heutral Conflict M diff. t“" p

. 4

-“-u.-*—‘---—-o----x- -- ~_-c.. .--- a-u . ~

9.36 9072 m 03

9.56 .11 o.ui n3

Av.
PTOj. 9067

No Def. 10.2h 9.32 .92 1.u5

-—-.-_.--. y...\.\- ~- 1 .. ~r....--.--- —- -uno—u—

—- _..-.__.-\~..u_.--.. ‘- _ ~--.~—.._. -__-_.- —. _.A -_...

 

3+ , .,. . _. .
Mean diiferences tested against an hypotheSized mean difference
of O.

** One tailed t test for repeated ”assures on same 85.

AA‘VQ

I ..‘ a " A 1 f‘ x / '-:‘ . “y . H‘ " " ‘
latiri Its all gas a.) u-1~3 a 3.3;).
mt ,...,. . ,.- I. M. ‘ .~ L4. ..
lne group e;ro1 Jean: Ior each sriaulu

Y‘"

INUaI-l "

were W“Ufcl and the arro

Ma

r diifcrence

of

' ' ' ' ~*I." -. ‘- v'”. ‘ ~ .1 r r". \V ‘— 1
for Sign canoe in all grctps. b Tests iol IoIoe
be tween Stihflll were le Jig.ifieart. Cease rent

for

of the mean ezror d14fcrtn‘e Ic‘nccn sillali. LL
ences were again tcs,e against an Lrbothesised w
“able 3 presents a sunaar, of these comparisons.
dlfiknwunces Exitneer sf.fieul‘ i1 tlm!<.onfliCT;cgatcg
The cessibilitj also ciis;od that 85 Light A
total response accuracy to 1L3 quIGrtion series
inter—categorJ differenCcs 'cre non mgniricant.
corr ct responses LC the ouncepoion sezies was

If 55 were more

\ ‘54,“ -

to ;u ted xean number of cor rect

aé.‘..'

and

" ‘ "I “ ' VI L ‘ ‘\
. ‘I e oi tainei 3o;reet res

\'

the analysi Again, none

used in

~4

,_ - n-.- ,. a
Jill Let;

U}
}_J I

-
I

lie results 01 410 scacoption analyses
1““. r NV M' r‘ ~. "i- f '71 C 11’"\ “.fl')" " ‘JBIC
"le 8-;L4 (It; l‘lnl - ‘ Ukaax L1 LO 3 l /n ‘c 22.1 a K24.

“1‘0 1‘ go

1(I’T'JL31‘t l

w c
.0...)
“r Y\--.
3. 313.172)

of the mean dif

.. c-
‘
Q.

ti

I;C)LVN3er;

1'"\,1' 7
“L .
”7"!
(A‘LJ

LAC

7
94A

“V
L)

ti I1

Ill; 0"

“c;

. ,-‘_. I" _' .
1U (‘k/iifi AL-\-'

1

- +

o
.1 <-
9 .L
’f‘
I.‘

H

l V - ‘ r.
Ine total

C
.1
ct ‘
#
LI

t.

I" 77 ‘ flrfr.) 0'
r‘A ‘ ~41.

" {a

I
“J ' O
A-..‘ \l. .
‘ Y“.‘v1 ‘v
.4 IJ.LA ’ L! J". [fili‘
.fi.
3 ~. a. 4 p .A»
- -l‘ ’ .‘IV
-21.; tL;l'-’..AI
"7 rvww- '4' ‘ fifiv- 4
‘6-‘l‘L-‘u ~A.»
. .
JF‘“C ““1
.J LL; (A "1.x: 1-

:l‘

L..I."I

. (-
OI.

n\'

TABLE ,3
EQIOR LLEJRLJS AIID LEJJI ElfiiOd DIIVb‘ELEIICES III 111351301359

TO THE SEX .‘ZI‘ID AGGILESSION SILLULI III THE bUbLLE‘t’TIOl-I

c "1'; Two
Qhowo

._...____.- _-s-.... -. ._ - .__- - - .. -.- .7-.. s -. _. - - .. - _. _ - .
._ __.-...-v....4-.._. ----. - - -__....- .- ._ _ .-._,... - - .. - - .. .. - --..- - . _. - H... -s
‘3?

Group M Sex lingg. M Blff. T

A---.._.-4.-_-_-.-——-,A

 

Avoid. b./O 5.:9
PYOJ- Itoé7 £1.93 ~~ .2

No Def. 5. s L.~5 .32 0.53

A. -- ‘.——a-..‘.-..M- ,,__. U. ---_....I... .m......-..«.....~...-_—___-.... _. a. ..,..A-__-.‘_-..‘- .——... ._.

 

* Two tailed t tests for repeated measures on same 83.

(If:

16

.- .. —o -__...-w..._— _.-_-.--.._—... W-...——_ _

89

QKBIE h
TOTAL CORRECT RESPONSE MEANS

F011 ‘I‘I:E 111111319. mom-13 011 THE SU‘JL‘EPTION StiR'IESSI

w...~._..«_.- - - - .-.- ~.-— ~ - .A . .
—-.-- -. - u.-. a ‘_____..M -. — -.._.-..«.- -._...._ - .. a- «an- «

 

Group Evan SD t df p

.....——..u— .g.—-..—.--m..u..—- - a. . .. H.-- h-_- --

 

Av. 20.91 2.99 l.h3 21 ns

Proj. 20.78 3.06 1.08 17 us

INTO Def. 20.11h 3.5/53 0061 Eli 1):.)

m ---.- -.-- - ———-—- -_~ o.

* Obtained means tested against any hypothesized mean of 20.

** One tailed t tests.

90

whether repeated rescueations of tre conflict stimuli might not result
in an adaptation or satiation effect, so that 53 would no lozgtr show an
increased sensivitity to the conflict stimuli. Failure to find signlw
ficant results could he attriLuted to such a varia'le. The significant
differential categoiy res;»onse otiained in the Pereeptua 1 De few1 “ series
arr gues against the presence of an adaptation effect. The same stimuli
were used in the Perceptual Defense series, the only difference being

an increase in stimulus Inagnitude. If adaptation or satiation were
present, a differential response would not have occurred in this series.

It was concluded that when a verbal accuracy indicator is used as +he

 

measure of subception, and when the sti1nuli are presented at Lia agni»

tudes below the level of verbal awareness as measured by phenome1o~

(D

Te
fl,

logical revert, iflunxa is no ev «twang to indicate innit 83 are ahl.

make a perceptual discrimination.

PEdCErmUAL DEFEISB PHASE

The experimentald esign used in this phase of the invesrifation
permitted both an interwgroup and an interwstimulus category analrsis
in testing forp Cb‘lltual dei ens e. In the intermg1o11p com1; risen, an
Analysis of Covariance was fi.st employed to test for overall differ—
ences in discrimination ”WO‘” the three groups. 1h: initial measure
used in this analysis was the number of correct responses for each
group on the 20 trial Acuity Test seriesfi and the final measure was the
number of corre cct responses on the hO trial Perceptual Defense series

(See Appendix G & H). Bartlett's test for homogeneity of variance w.

w
U)

applied to be h the Acqu and Defznse measures. The resulting Chiw

squz res were not significant.

91

1

Table 5 precepts a summ"*y of tne Analysis of Covariance. The ob-
tained F value of 15.67 was significant beyond the .U1 point for 2 and
61 df. Differences in discrimination among the three groups cannot he
accounted for by initial differences in visual acuity since an adjust~
ment is made for any such initial differences. In computing the Analy—
sis of Covariance, it was found that the groups actually did not differ
significantly on initial level of visual acuity.

Since overall differences in discrimination pr ved to te signifim
cant, between group comparisons were performed to determine more pre-
cisely the source of difference among the three groups. The measure
used in this analysis was a mean difference score which was derived by
subtracting 53' correct responses on the Acuity Test series from the
number of correct responses on the Perceptual Defense series (see
Appendix I). From the Perceptual Defense hypotheses it was predicted
that the Avoidance group would make significantly fewer correct re~
Sponses than the Ho Defense group, and that the Projection group would
make significantly more correct responses than the No Defense group.

Table 6 shows the mean difference comparisons of the Avoidance and
Projection groups with the No Defense group. An F test for homogeneity
of variance was not significant for either the Avoidance~ho Defense or
the Projection—ho Defense comoarisons. Consequently, one tailed t tests
were used to test for the significance of the difference between means.
The obtained t value of h.25 for the Projection—No Defense comparison
was significant beyond the .01 point for hl df. The t value of 0.66

obtained for th Avoidance—No Defense comparison was not significant at

TIRE LE [3
ANALYSIS OF COVARIANLE OF CORRLCT RESPONSES FOR THE

THREE GROUPS IN THE PERCBHTU£1.D£FENSE SERIES

.- -.--. -~ .- _ .o—_. . -4 o.-- -......--_. _.—— ~..--

 

 

TO 1 a :1. '5 1 U1 1 H Be My C e r.-

 

 

 

 

351.52 23.80

\A)
‘J
\ r1
0

J..."
N

1. Sum of products

{2‘
1'“
O

M
L“

2. Sum of SQUCIUS X
(Acuity Test)

3. Sum of squares Y 1573.1b 1117.2h h55.93
(Defense series)

u. (11" 6h 62 2

 

5. Correlation .Lm .522 .519

5a. df for r 63 él 1
6. k3w¢ value .9151 .8508
7 . Adj :15th if 1229.80 812 .15 h1.;7.;35

8. df 63 61 2

—- -——s w~ ~.n.——_...-_—-.——-—- -n‘afi~ q“... _.‘_., -r... q. ,. é.» __.. H...»..~.q._.._‘u .-._

 

TABLE 6
COMPARISON OF AVOIDANCE AND PROJECTION GROUP KERN
ACCURACY SCORES OH TRE PERCEPTUAL DEFENSE SERIES

WITH THE NO DEFENSE GROUP

 

 

 

 

 

Group khan SD t p
Avoid. 11.86 2.90
0066 US
No Def. 12.32 h.53
Mean Diff. .bé
Proj. 17.73 6.36
L. .25 < .01

Ho Def. 12.32 b-53

Idean Diff o 5 all].

 

* One tailed t tests

9L:

J

k

F‘;
g

the .05 point r he df. The results indicated that the overall differw
ences in discrin nation as fouId in the A’IIal l"sis of Colaiiunce could he
attributed only to diff ere' nces fornd in the Projection grOLp. in re wire
no significant differences between the Avoidance and No Defense groans.
ince it was fonmrd f at tie three groups actually did not differ
significantly on initial level of visua acuity.9 an inteI g1 OHp compari-
son of responses to the neutral stirulus category was then conducted.
An Analysis of Variance was employed, using the total error scores to
the neutral stimulus category as a moasxuce of discrimination. (See
Appendix J). mahle 7 presents a s ImuI” of this analysis. The obtained
P value of 2.6h was not significant at the .05 point for 2 and 6? df.

There were no significant differences among the three groups in response

to the neutral st timulus category. iron this analysisJ it was concluded

V

that the Projection group differed significantly from the other grougs
only in response to the conflict stimulus category.

There are two possible conditions which could have reenlted in the
failure to find significant differences betwee I the Avoidance and No
Defense groups. First, the possibility exists that the Avoidance group
did not show avoidant behavior in perception, in whicu case} the first
Perceptual Defense hypot thesis would not be supported. A second possible
condition is that both the Avoidance and he Defense groups might have
shown avoidant behavior in perception, in which case. tlie third PGIC cep~
tual Defense hvpothe H1 would not have been supported. To determine

which condition existed, a stimulus category ana vsis was conducted.

For the stimulus category analysis: total reaponse errors to the

THEE LE 7

ARAIISIS OF VARIANCE OF ERROR SCORES TO ThE NEUTRAL

ST RULES PRESENTATIONS IN THE DEFERSR SERIES

 

Source of variation souares df square F

Between grorps 35.81 2 17,90 2,6h“

.‘ ' 4" ' —y‘. ‘5 '0 :fi ()1" K «C "7 I
Karl. bhin £1. ULLflJ hL‘) .111 (22.) (R o I 5

——-..-M -—

To tal use . 9’" (in

 

* Not significant (P ;"OS)'

96

conflict and neutral categories for the ho trial Perceptual Defense
series were tabulated for all 85 (See Appendix J). From the Perceptual
Defense hypotheses it was predicted that the group using avoidaflgg as

a general defense would make significantly more errors in response to
the conflict related stimuli than to the neutral stimuli, and that the

group using projection as a general defense would make significantly

 

fewer errors to the conflict stimuli than to the neutral stimuli. Fiw
nally, the £2 defense group would show no significant error difference
in response to the stimulus categories.

Table 8 shows the group_error means to each stimulus category and
the significance of the mean error differences between categories for
each group. One tailed t tests for repeated measures on the same 53 were
used to test for significance of the mean category error differences. F
tests for homogeneity of variance between categories were not significant.
All of the obtained mean error differences were tested against an hypom
thesized mean difference of O. The t value of 2.88 for the Avoidance
group was significant beyond the .01 point for 21 df. and the t value
of 1.86 for the Projection group was significant at the .05 point for 1?
df. The t value of 1.73 for the No Defense group was not significant at
the .05 level for 2h df. These results supported all three of the Perm
ceptual Defense hypotheses.

The questioned still remained as to why the Avoidance and No Defense
groups failed to show significant inter—group differences in response to
the conflict category. A further examination of the interucategory an»

alysis for the No Defense group revealed that although the t value of

9

T” '2 "‘ «'1
_ v u ~
‘5‘) (I

ERROR MEANS AND TEE LEAN ERROR DIFFHR’KCB SLOflES TO

5

Tm I‘IEUTRAL mm oomtlcr smmu FOR TV“ TlfltEE

la-J

; Had

GROUPS IN THE PERCEPTUAL DEVVVSE SERIES

 

f‘ .T w - 1"
urcup I lnaut. l£(aonf. Ill)-fi. t p

.— ~ \ 3 ~ . ' r» ‘3’
V-Uu " c.09 :.CB <:.Lu

No Def. 6.92 6.0h m 1.12 1.73 :> i~ww

I U;

* A ‘ ' " 4. . V . ,i “
One tailed t tests :01 repeated measures on name ‘

** .i a
Two tailed t test ior

repeamwinmasurescx13ane Se.

98

1.73 was not significant at the .05 level, it did reach significance at
the .10 level. This indicated a tendency for the No Defense group to
make more errors in response to the conflict stimuli—ma resporse similar
to the Avoidance group. secause of this tendency, interwgroup differ~
ences proved non~significant.

A further analysis was conducted to determine wheth>r a differential
response to the Sex and Aggrission stimuli comprising the conflict stim~
ulus category existed (See Appendix K). This analysis provides a test
of the construct validity of the Defense Preference Inquiry which was
used to classify 55. The construct tested is that the so called "gen~

ense in dealing with all areas of

5“;

eral defender" will use the same de

\

conflict. This would be substantiated bv a lack of significant differ_
ences in response to the two conflict stimuli used in this study.

Table 9 shows the group error me ns to each stimulus of the conw
flict category and the significance of the mean error difference in
each group. Two tailed t tests for repeated measures on the same 55
were used in the analysis. The obtained mean error difference scores
were tested against an hypothesized mean error difference of 0. None
of the resulting t values proved significant. The groups showed no
differential response based on stimulus type.

Some additional analyses were performed to insure that the results
could not be attributed to extraneous sources of variance which might
have influenced the verbal accuracy indicator independent of the porn
ceptual variables. It has been reported in previous research (Howes &

Solomon, lSSl; Freeman, l95h) that 83 show an adaptation effect to

99

TABLE 9
ERROR MIMNS AND THE MEAN ERROR DIFFERENCES IN RESPONSE
TO THE SEX AI-JD AGGRESSION STILJLI IN THE PERCEPTUAL

DEFENSE SERIES

 

m<-.——-———.-~_.—o _”,.__.‘- «.V - —. ....-—.»-‘ ——.——..~_._ .—._ -‘-.o- W.- .._..-- ~—— .- um,-

 

 

.- -— Aim — - —- -.-._.,__,.,___- h .a— i-—¢-.--—~--- “- .. ~_ —--...——_—-.--—-——-

'1:-

 

Group M Sex M Agg. LI Diff. t p
Avoid. h.h5 LL59 -— .111 0.27 ns
No Def. 3.80 b.2h — .hh 1.07 ns

 

 

 

 

m...fi---.-¢.-—---—-~-—. ..- ... .. m-—<——_——n

* Two tailed t tests for repeated measures on same 85.

threatening stimuli. Thresholds for such stimuli become lower with
succeeding presentations. If adaptation occurred in the present study,
55 should have shown a consistent error decrement on succeeding trials
in the Perceptual Defense Series. To test for such an effect, the hO
trial series was divided into 10 trial blocks with h trials in each
block. Error means for each trial block were computed for all groups.

Figure 1 shows the performance curves for the three groups conn
structed from the error means for each trial block. All.groups show a
sharp error decrement from the first to the second trial block. This
is most likely a function of continued adaptation to the lower illumm
ination in the tachistoscope. There is no apparent systematic trend
in error performance following the initial decrement in all groups. Of
particular interest is the much lower error mean for the Projection
group on the initial trial block, especially since the three groups did
not differ significantly in visual acuity. The Projection group shows
an immediate vigilance response, which it maintains throughout the den
fense series. This initial error difference cannot be accounted for in
terms of adaptation to repeated stimulus presentations.

A second variable which could have led to differential response
accuracy independent of perception is that of response bias. 55 might
have shown a response disposition for one or the other of the stimulus
categories, especially since both a neutral.and conflict stimulus re—
mained constant in Ss' visual field on each trial. It might be argued,
for example, that the Avoidance group could have avoided calling the

conflict category, and consequently, would have made fewer errors to

107'

 

 

—-—-- Avoidance group
H Projection group
~---- No Defense group
2 .00 t—
l.75 ___
1.50
Error [——
hhans
1.25 __
1.00 __
.75 __
J l I l l l I l I

 

 

 

Blocks of 1; Trials

Figure l. GROUP ERROR AIEAI‘JS FOR SUCCESSIYE TRIAL BLOCKS IN THE PERCEPTUAL

DEFENSE SERIES

l0?

the neutral stimulus presentations. A response disposition in the Pro_
jection group might have led to overcalls of the conflict stimuli result~
ing in more errors to the neutral stimulus presentations. Response bias
would have resulted then in a differential response between groups to

the neutral category. Table 7 shows the Analysis of Variance of responses
to the neutral category for the three groups. The obtained F value of
2.6' was not significant. The results indicate that the differential
intermcategory response is not a function of response dispositions cperu
ating independently of the discriminative stimuli.

A final.analvsis was conducted to determine whether differences in
discrimination might be related to sex of S. Correct responses were
tabulated for the two sexes in all groups (See Appendix L). Male and
Female response means were computed for all groups and the mean differ»
ences were tested against an hypothesized mean difference of 0. Table
10 presents a summary of the analysis. Two tailed t tests were used in
making the mean difference comparisons. None of the obtained t values
proved significant. There was no evidence to indicate that sex of S
was related to differential accuracy in perceptual.discrimination.

The results of the additional analyses failed to provide any evi«
dence to support the view that differential response in the perception
task can best be explained as a function of »xtraneous variables inde»
pendent of actual perceptual processes. The differences in response
would appear, therefore, to involve perceptual discrimination, indi~
eating a relationship between personality variables and perceptual

processes.

TABLE 10

SIGNIFICANCE OF CORRECT RESPONSE MEAN DIFFERENCES

BETWEEN MALE AND FEMALE SUBJECTS IN.ALL GROUPS

IN THE PERCEPTUAL DEFENSE SERIES

103

 

 

 

 

Group M Male u Fem. M Diff. t* p

ltvoid 2h.50 23.58 .92 0.68 ns

Proj. 31.09 29.28 21.81 0.81 ns

No Def. 23.91 25.92 .. 2.01 1.03 ns
*

Two tailed t test.

10b
DISCUSSION

SUBCEPTION PHASE
The results of the subception phase indicated that 85 were unm
able to make a perceptual diserim nation as measured by a verbal accuw

d below

(I)

racy indicator when the discriminative st'nuli were present
the level of awareness. The stimulus magnitudes representing "below
awareness" were those .xiosure speeds at which s could no longer
report seeing anything in the visual field. It is believed, therefore,
that the results represent contradictory evidence to this one aspect
of the findings reported by Dixon (1958). Since the present study is
not a replication of Dixon's experiment, the existence of a true conm
tradiction assumes, of course, the comparability of the indices of peru
ception as well as the nature and relevance of the perceptual stimuli
used.

In Dixon's study, taboo words tere used as the threatening stimu
uli. An implicit assumption in using this type of stimulus material
is that such words will be threatening to all 55, and that they will
react to such threat in the same manner. Previous research in percepw
tual defense has shown this to be a rather gratuitous assumption.
There are differences in the degree of threat, and not all.Ss deal
with threat in the same way.

In the present study, 85 particular response to threatening stimm
uli was first determined. Two criteria were used to select those Ss
most threatened by the stimuli used and who would show a significant

consistency in their method of dealing with this threat. Consequently.

I

\. 1‘!

v M

the stimuli used in the present investigation were considered to be
even more relevant and uniformly threatening to Ss than those used
in the Dixon study.

In both the Dixon study and the present investigation, the subm
ception effect was meas*red in terms of a discrepancy between a
phenomenological indicator and a forcedmchoice verbal accuracy indim
cater of perception. fhat is, S reports seeing nothing in the visual
field and then makes a verbal discriminatory response. Although
phenomenal report has been subjected to criticism as an indicator of

nense variance extraneous to

..

U)

perception, because it readily admits re
perception, such criticism would not seem as pertinent to either
Dixon's or the present study since neither employed this indicator as
a forcedwchoice measure. 8 did not have to make a judgment until he
felt absolutely sure of it. In the present investigation, S was
encouraged to see, rather than not see the stimulus presentation.
Although both experiments employed a verbal accuracy indicator,
there was a marked difference in the application of such a measure.
In the present study, S was required to make an immediate discriminaw
tory response in a two~choice situation. In Dixon‘s experiment, a
time interval of considerable length occurred between the stimulus
presentation and the actual verbal accuracy measure. The effect of
this time interval upon subsequent response is certainly open to
question. In a recent experiment, Goldstein and Barthel (l960) re»
ported that the presentation of a subliminal stimulus has an immediate

effect upon subsequent responses but no sustained effect over a time

106

interval. In any event, one would expect the responses to be more ac“
curate in the immediate and original stimulus situation. The effect,
if any, of the stimulus presentations should have been more pronounced
in the present experiment.

Dixon's results had dealt a blow at the current affective-
veridical theory of perception. According to this theory, the affec~
tive response in subception is occasioned by partial information
afforded by the discriminative stimulus. If the affective response is
positive it facilitates veridical response, with perceptual sensitiza-
tion as the result. If negative, the affective response tends to dis»
rupt higher levels of integration, producing perceptual defense.
Dixon's results did not support the view of affective-veridical thresh-
old differences or the disrupting effect of the effective response,
but rather, they indicated that both affective and veridical processes
may be activated at stimulus magnitudes below phenomenal report. The
results of the present investigation failed to support Dixon's find»
ings and, therefore, raise some question concerning the function of
veridical processes at such low stimulus magnitudes.

Although it was not a point of emphasis in their experiments,
Blum (l95h) and Nelson (1955) have also shown that veridical thresholds
may be quite low. Using stimulus magnitudes described as "a low level
of awareness below recognition thresholds", they demonstrated the sub_
ception phenomenon as measured by a verbal response indicator. At
these stimulus magnitudes, S was aware of something being presented in

the visual field to the point of being able to specify the general

class of stimulus material, but could not identify it. 'Hhen the stimw
ulus magnitudes were increased to a point described as "a high level

of awareness below recognition thresholds", avoidant behavior occurred
as measured by a verbal acc ‘acy indicator. These latter stimulus magu
nitudes permitted a more specific identification of the stimulus but
did not afford total recognition.

Besides indicating low thresholds for veridical processes, the
results reported by Blum and Nelson suggest that at low stimulus magw
nitudes the veridical process may not be disrupted by an affective
response. These results along with the work on perceptual_sensitizam
tion raise some interesting questions concerning the affective response
system. First, the question may be raised as to whether there is a
direct relationship between stimulus magnitude and the strength of the
affective response as measured by an autonomic indicator. Second, how
do the affective responses to positive and negative stimuli differ,
and can this difference be measured by an autonomic indicator? Third,
does the disruption of higher integrative processes which is attributed
to the negative affective response occur only when the affective rem
sponse reaches a certain magnitude? These questions suggest directions
for further research.

Eriksen gt a}. (1959) also opposed the view that affective rem
sponses occur temporally prior to, and are more accurate than, veridim
cal response. They suggested that affective and veridical processes
represent concurrent response systems, neither of which is more accurate

than the other. Their results indicated that thresholds for the affegm

tive response say even be higher than veridical response thresholds.

Although the above results have created some interesting problems:
they do not render untenable a theory of subception as a so called
”unconscious" process. By "unconscious" it is meant that S is unable
to verbalize an awareness of the stimulus being presented in the
visual field. It may be that stimuli presented at magnitudes below
phenomenal report can elicit an autonomic response. Dixon‘s results:
using an autonomic indicator of 1.:erce1‘.vt¢'L0n.9 support such a view.
Previous experiments which have employed the GSR as a measure of suhw
ception have not tested tnis hypothesis: since stimulus magnitudes
were used which permitted a partial awareness of the stimuli. Much
of the contradictory data can be related to differences in meaning of
awareness thresholds and to differences in response indicators used to
measure perception. Awareness thresholds must be more precisely dcw
fined, and the validity of the various measures used to indicate perm
ceptual response must be examined. In relation to this latter point,
it may be asked whether verbal judgments, forcedwchoice verbal.accu~
racy measures, and total verbal identification are all manifestations
of higher level or veridical.processes. Future research will require
greater recognition and integration of the available data as well as
an increased emphasis upon psychophysical methodology if we are to

show systematic progress in theory and understanding.

PERCEPTUAL DEFENSE mess

The results of the inter-category analysis in this phase of the

109

investigation sup; sort an extension of the concept Oi perceptual defense
to include both avoidant and vigilant behavior. This concept of per-
ceptual defense differs from that previously espoused by Blum (195h)
who considered avoidance and vigilance to be two opposing processes in
perception. According to Blum’s view, vigilance occurs when conflietw
related stizauli. are presented at magnitudes below aims eness. When the
stimuli are presented at magnitudes approaching conscious recognition,
a second process occurs by whic a 83 may delay or avoid recognition of
the coni lict stimuli. This latter process is terzfled we ceptual defezn
Similar formulations within a behavioral framework also have been preu
sented. Stated briefly, the vi w expressed is that the conflictw
related stimuli elicit an anxiety response which in turn serves as dis“
criminatixe stimulus for the emission of previously conditioned avoidant
behavior.
The thesis receiving support in this present study is that when

threatening stimuli are presemi ed at stimulus ma agn itudes ap p1 eaching

conscious recognition, a differential response occurs. The rOSponse

 

may be either avoidance or vi ilance and is a function of previous con;

ditioning to threatening stimuli. This view is supported by both psy

choanalytic theory which recognizes defer ses othc r than repression in

dealing with threat, and by behavioral theory. In learning terms, it

may be stated that an anxiety response could serve as a discriminative

stimulus for emitting either a conditioned avoidant or vigilant response.
In most experiments on perceptual defense, the method of dealing

with threat is assumed to have been acquired in the past life of the

llO

individual. The problem then becomes one of making inferences about
this past experience and of predicting perceptual behavior in terms of
these inferences. The general assumption has been that Ss will avoid
threatening stimuli in perception. Often, the opposite or vigilant be~
havior was found. Further attempts to predict perceptual behavior from
the degree of threat or level of anxiety have also led to inconsistent
results. The results of the present study indicate that an approach to
prediction in terms of 55' inferred defense mechanisms is a fruitful
one and also suggest that not only the degree of threat but a knowledge
of the particular method of dealing with threat is necessary for prew
dieting perceptual behavior.

The present findings are in agreement with those reported by
Carpenter 33.32' (1956) who first treated avoidant and vigilant be~
havior within a framework of perceptual defense. In the present study,
a third group, the No Defense group, was included to permit a more defm
inite inference that vigilant behavior was actually serving as a defense.
When only avoidant and nonwavoidant groups are compared, differences in
response to threatening stimuli are to be expected. It is difficult to
infer that because recognitions thresholds for the non—avoidant group
are lower, this greater sensitivity is a defense. It must be demonm
strated that a group showing vigilance as a defense differs from a No
Defense group. The inter—category and inter»group ana yses in this
phase of the invesoigation supported such an inference.

The failure to find significant inter—group differences between

the Avoidance and No Defense groups was unexpected and demands explanau

lll

tion. One explanation may be made in terms of the classification crim
teria. The Defense Preference Inquiry (DPI), as with most instruments
used in classification is undoubtedly far from being 100% efficient.
The criteria used for placement into the "general" defender category
may have failed to include a significant number of Ss which belonged in
that category. An examination of the individual difference scores in
the No Defense group based upon response to the Acuity and Perceptual
Defense series revealed that 8 55 had as low or lower scores than the
lowest score cluster found in the Lvoidance group. These 88 certainly
appeared to be responding to the perceptual task in a manner similar

to the Avoidance group. The need for further examination of the SGlCCw
tion criteria is indicated.

In his original presentation of the DPI, Goldstein (1952) dis~
covered a relationship between choices of avoidance and reaction for—
mation for the "general" defender category. 55 who selected these
defenses tended to give them first or second choice rank on all the
psychosexual dimensions presented. The rankings for the 8 low Ss in
the No Defense group were examined for such a relationship. Although
no S selected either defense h or more times as the first choice, it
was found that most first and second choices were distributed between
these two defenses. Extending the selection criteria to include such
an affinity of preference might increase the efficiency of the DPI.

A second explanation for the non-significant differences between
the Avoidance and No Defense groups is possible. It may be that 85

who show little consistency in their rankings on the DPI, and who may

actually experience little conflict on the psychoeehual dimensions prew
sented, nevertheless, tend to show avoidant behavior in perception. If
such a tendency exists, it stands in contradiction to the findings prem
sented by Chodorkoff (1956) who found that the better adjusted Ss tend
to show vigilance in perception. In any event, the results obtained
call for further investigation of the so called "specific" defender
group.

A mayor criticism of the experiments dealing with subception and
perceptual defense is that the substantive contributions of such studies
can generally be explained as a function of such variables as frequency
of prior exposure to the stimuli, response suppression, bias, set, or
learning. It becomes necessary, therefore, for E to control these
sources of variance, or in those conditions in which control is not
possible, to be able to test for the effects of extraneous variance.
The present experiment provided such controls. Factors such as visual
acuity, adaptation, response bias, and sex of S did not affect the re”
sults. The perceptual response obtained would certainly appear to in“
volve a motivational variable.

Goldiamond (1958) in a recent critique of methodology in studies
on perceptual defense discusses the perceptual response as gperapt bem
havior. By an operant, he means that the reSponse is not elicited by
the discriminative stimulus but will be emitted by the discriminative
stimulus if there has been previous reinforcement in conjunction with
the stimulus presentations. Consequently, the response can co—vary with

all the variables that affect operant behavior. Goldiamond emphasizes

(L)

he point that cash variables are typically not perceutual on;s. To
extend his point, it ma“r be added that a reinforcing stimulus is not
on y operating in relation to an immediately prior response, but also
is serving as a discriminative stimulus for a subsequent response.
hhen the stimuli presented by E are of too low a magnitude to serve as
discriminative stimuli, a particular response pattern may be occasioned
by irrelevant stimuli such as expectancy, utility, or partial recognim
tion of the discriminative stimulus. The response pattern then may be
sustained independently of the actual stimuli provided by E. Analysis
of the data did not support the presence of this form of response bias
in the present study.

The position that Goldiamond seems to have taken is that perception
is sensorily determined and is not affected by learning or other variam
bles that modify response. Under such a view, defenses, needs, and
affect would play no role in perception. Such a position has been
attacked by Hebb (1952) who has attempted to show that perception of
even simple objects such as a triangle is, in part, a learning process.
Bruner and Minturn (1955) have demonstrated that the perceptual organiw
zation of the field is not a pure stimulus process but operates differw
entially with respect to identification given to the stimulus. A disw
tinction between perception £33.29 and its response indicator, while
often proposed, may be theoretically untenable.

If perception can be considered to involve response, then it is
subject to modification by variables such as learning. Many of the

principles that underly learning in general should also apply to per~

11h

ception. Previous research has been little concerned with such princi~
ples since learning was assumed to have occurred in the past. The conm
troversial nature of this research has pointed up the need for more
specific knowledge of the individual’s past ex>erience in order to avoid
ad_hog explanations nd to permit prediction. A study of SS3 defense
mechanisms in dealing with threat provides one effective means of in»
ferring past experience and for making predictions. However, it is not
only important to know that defensive behavior and other personality
variables affect perception, but also to increase our understanding of
the ways in which perceptual learning is modified.

A direction for future research would be to investigate perceptual
learning in the orperimental setting. In this manner, we may be able
to determine more directly the conditions under which reinforcement can
alter perceptions; what effects past experience and motivational faew
tors have upon new learning, as well as the role of autonomic and pro”
prioceptive stimulation upon perception. Such a direction would appear
necessary, if we are to increase our knowledge of how we perceive ace

curately or inaccurately.

115

SUMMARY AND CONCLUSIONS

The two major objectives in the present investigation were: (a)
to test for the subception phenomenon using a verbal accuracy indicator
at stimulus magnitudes too low to permit a phenomenal resport of awarem
ness, (b) to extend the concept of perceptual defense to include both
avoidance and zigilangg in perception.

The experiment was divided into three phases, the defense assess-
ment, subception, and perceptual defense phase. In the defense assess-
ment phase, 31h 85 were administered the Defense Preference Inquiry, a
ranking task used in conjunction with the Blacky Pictures from which S's
defenses on a number of psychosexual dimensions may be inferred. Two
criteria were used to select 55 for the experimental groups. These
criteria were: (I) h or more first place ranks of the same defense
mechanisms, (2) total rank consistency. On the basis of the selection
criteria, 22 85 were placed in an Avoidance group, and 18 85 were placed
in a Projection group. A random sample of 25 58 was chosen from the
group showing little defense consistency and were placed in a No Defense
group. The three groups then took part in the perception phases of the
experiment.

A two-choice discrimination task was used in the perception phases.
A pair of stimuli labeled A and B were presented on each discrimination
trial and remained constant stimuli for that trial. Then either A or B
was exposed briefly between the constant stimuli, and S was required to
state which stimulus had been exposed. The stimuli consisted of sex-

neutral, aggression-neutral, and a neutral—neutral pair.

In the subception phase, the stimulus magnitudes rc presenting he1ew

awareness thresholds we1e cote1m1ncd 101 ea< h S individually. 53 were
mien given a LO trial discrimination series using the sexmneutral and
aggression—neut1al pairs

In the perceptual defense phase, the stimulus magnitude was irn
creased to a level, previously determined in a pilot stat; , at which
approximately 50$ correct discriminations of the neutralnneutral stimum
lus pair could be made. The stimulus mag1itude remained constant 1or
all 85 in this phase. 53 were given a 20 trial acuity test series
using the neutral stimulus pair, and this was followed by a hO trial
discrimination ser es usirg the conflictmneutral stimulus pairs.

Total response accuracy and conflictuneutia 1 category errors veie
recorded for Se in toth the subception and pe1'ceptual. do fens ephases.
The results of the subception phase revealed that

1. There 1018 no sigm1_ant ca tego1y error differences foi the
three groups.

2. There were no significant diffe1er ces in response to the sex
and aggression stimuli co.a11tuoixg t11e conilic t cat egory'for the three
groups.

3. Total response acc m:y was not significantly greater than
would be expected by cha1ce for all groups.

The results provided no support for the hypothesis that all 55 would
show subc eption to the conflict stimuli DTCS mited at stimulus maLu1fud32
below awareness thresholds.

The data obtained in the perceptual.defense phase were subjected to

ll?

both an interwgrOLp and an interwcategory aral;sis. The results may he
sr.narized briefly as follows:

1. There were no significant differences among the three groups
on the visual acuity test or in response to the neutral stimulus eaten
gory.

2. There were no significant differences for the three groups in
response to the sex and aggression stimuli constituting the conflict
category.

3. Analyses of the data to determine the presezce of extraneous
variance such as might be related to adaptation effects, set, and sex
of 5 yielded non—significant results.

h. All three perceptual defense hypotheses were supported hf the
inter-category analysis. The Avoidance group made significantly more
errors in response to the conflict category as compared to the neutral
category, while the reverse was true of the Projection group. The No
Defense gr up show no significant interwcategory response differences.

5. The Projection group was significantly more accurate in rew
sponse to the conflict category than the other two groups. There was
no significant difference between the Avoidance and No Defense groups
in conflict category response.

On the basis of these findings and those of recent studies on
subception and perceptual defense, it was concluded that:

l. uhception as measured by a verbal indicator does not occur
when the stimuli are presented at magnitudes too low to permit a phem

nomenal.report of awareness.

118

2. Se respcnd differentially to threat in perception, and this
differential responsivity can be predicted from a knowledge of 85‘
characteristic ego defenses.

3. The differential reSponse in perception cannot be explained
as a function of response bias.

b. 83 who utilize one particular mechanism as a general ego dew
fense show a similar generality in perceptual response.

The results support the plausibility of viewing perceptual avoid~
ance and vigilance within a framework of defense, and further suggest
that avoidance and vigilance in perception are directly analogous to

the ego defenses of repression and projection.

 

 

Several questions for further research mere raised. It was sugw
gested that there is a need for: (a) a more precise definitions of
awareness threshold; (b) further study of the autonomic or affective
response as it relates to perceptual vigilance and sensitization; (c)
and investigation of the perceptual learning process per se within the

experimental setting.

119

REFERENCES

ALLPORT, F. H. Theories of perception and the concept gf strugtgrg.
New York; Wiley, 1955:

 

HARTLEY, S. H. Principles 3f perception. New York: Harper, 1957.

 

BEER, E. 0., & Cowen, E. L. A further investigation of the influence
of "threat—expectancy" on perception. g. Pers., 1953, 22, 25b,-
257.

BINDER, A. Personality variables and recognition response level.
,1. abnorm. soc. psychol. 1958, 57, 136—lh2.

 

BITTDRMAN, M. E., a KNIFFIN, c. w. manifest anxiety and perceptual
defense. ,1. abnorm. soc. Psychol., 1953, h8, 2h8—252.

 

BLUIJ, G. S. A study of psychoanalytic theory of psychosexual develop»
ment. Genet. Psyche}. yonggn, 1919, 39, 3—99

 

LL'LE, G. S. An experimental reunion of psychoanalytic theory with
perceptual vigilance and defense. ,1. abnorm. soc. Psychgl.,

l9Sh, h9, 9u-98.

BLULI, G. S. Perceptual defense revisited. g. abnorm. soc. Psychol.,

1955, 51, 2h—29.

BLUM, G. S. Defense preferences in four countries. ,1. pr 1'. 3%.,
1956, 20, 33-hl.

 

 

 

 

BRICKER, P. D., & CHAPANIS, A. Do incorrectly perceived tachistoscopic
stimuli convey some information? Psychol. Rev., 195.3, 60, 181-
188.

 

BRO‘JJI‘I, D. R. , 8r. ADAMS, J. Word frequency and the measurement of value
areas. :1. abnorm. soc. Psychol., 1951;, h9, h27~h30.

 

BRUNER, J. S. Personality dynamics and the process of perceiving. In
R. R. Blane & G. V. Ramsey (Eds.), Perception: an approach in
personality. New York: Ronald, 1951; Pp. 1214lfi7. (a) '

 

 

BRUNER, J. S. One’kind of perception: A reply to Professor Luchins.
Psychol. Rev., 1951, 58, 306-312. (b)

 

BRUIKIER, J. 5., & POSTL’IAN, L. Emotional selectivity in perception and
reaction. ,1. Pers., 191:7, 16, 69—77.

BRUNER, J. S., 8: POSCIMAN, L. Perception cognition and behavior. _._I.
Pers., 19h9, 18, lh-3l.

120

BRUNER, J. 8., & KINTURN, A. L. Perceptual identification and percepw
'tual organization. J. gen. Psychol., 1955, 53, 21~28.

 

CARPENTER, D., WIENER, m., & CARPENTER, JANETH T. Predictability of

_..._ _

380-383.

CHODORKOFF, B. Anxiety, threat, and defensive reactions. .J. gen.
Psychol., 1956, 5h, 191—196.

COHEN, A. R. Experimental effects of egowdefense preference on inter—
personal relations. ‘i. abnorm. soc. Psvchol., 1956, 52, 19—27.

COHEN, E. L., & BEIER, E. G. The influence of "threat_expectancy" on
perception. 'J. Pers., 1950, 19, 85—9b.

CCMEN, E. L., & BEIER, E. G. Threat~expectancy, word frequencies and
perceptual prerecognition hypotheses. g. abnorm. soc. Psychol.,

l95u, h9, 178-182.

 

 

COHEN, E. L., & OBRIST, P. A. Perceptual reactivity to threat and
neutra1.words under varying experimental conditions. .i' abnorm.
SOC. PsyChOlc, 1958, 56, 3OS"3100

 

DASTON, P. G. Perception of homosexual words in paranoid schizophrenia.
Percept. mot. Skills, 1956, 6, h5-55.

 

DE LUCIA, J. L., & STAGNER, R. Emotiona1.vs. frequency factors in word
recognition time and association time. 'J. Pers., 195h, 22, 299—

309.

DIXON, N. F. The effect of subliminal stimdlation upon autonomic and
verbal behavior. “J. abnorm. soc. Psychol., 1958, 57, 29~36.

 

DULANEY, D. E. Avoidance learning of perceptual defense and vigilance.
g. abnorm. soc. Psychgl., 1957, 55, 333—338.

 

EDWARDS, A. L. Experimental design ig_psychological research. New
York: Rinehart, 1950.

 

 

EDWARDS, A. L. Statistica1.methods for the behavioral sciences. New
York: Rinehart, 19Sh.

ERIKSEN, C. W. Some implications for TAT interpretation arising from
need and perception experiments. .J. Pers., 1951, 19, 282-288.

IERIKSEN, C. W. Perceptual defense as a function of unacceptable needs.
J. abnorm. soc. Psychol., 1951, h6, 557—56h.

 

121

ERIKSEN, C. N. Defense against ego threat in memory and perception.
J. abnorm. soc. Psychol., 1952, h7, 230—285.

 

 

ERIKSEN, C. W} The case for perceptual defense. Psychol. Rev., 195A,
61, 175—182 .

 

ERIKSEN, C. W. Subception: fact or artifact? Psychol. Rev., 1956,
63, 714—80 0

ERIKSEN, C. W}, & BRONNE, C. T. An experimental and theoretical analy—
sis of perceptual defense. lg. abnorm. soc. Psychol., 1956, 52,

22h-230.

ERIKSEN, C. W., AZUMA, H., &.HICKS, R. Verbal discrimination of pleas—
ant and unpleasant stimuli prior to specific identification. pg.
abnorm. soc. Psychol., 1959, 59, 11h-119.

 

 

 

 

FENICHEL, O. The_psychoanalytic theory_gf neurosis. New York: Norton,
19h5.

 

FORER, B. A. A structured sentence completion test. J. 2393' Tech.,

FREEMAN, J. T. Set or perceptual defense? J. Exp. Psychol., 195b, h8,

FREEMAN, J. T. Set versus perceptual defense: A confirmation. J.
abnorm. soc. Psychol., 1955, 51, 710-712.

 

FREUD, A. The ego and the mechanisms_gf defense. New York: Interna-
tional Universities Press, 19H6:

 

FREUD, S. Repression. In Collected_papers. Vol. IV. New York:
Basic Books, 1959, pp. 88-97.

 

GIICHRIST, J. 0., LUDEMAN, J. F., & IYSAK,‘W. Values as determinants
of word recognition thresholds. .J. abnorm. soc. Psychol. l95h,
’49, 1423“”)426 o

 

GOLDIAMOND, I. Indicators of perception: I. Subliminal perception,
subception, unconscious perception: An analysis in terms of
psychophysical indicator methodology. Psychol. Bull., 1958, 55,
373”bllo

 

GOLDSTEIN, S. A projective study of psychoanalytic mechanisms of
defense. Unpublished Ph. D. dissertation, Univ. of Michigan,
1952.

GOLDSTEIN, M. J., & BARTROIu R. P. Fantasy responses to subliminal
stimuli. J. abnorm. soc. Psychol., 1960, 60, 22~26.

 

 

122

HAIGH, G. V., & FISKE, D. W. Corroboration of personal values as
selective factors in perception. ‘J. abnorm. soc. Psychol., 1952,

A7. 39h-398.

 

HEATH, D. Projective tests as measures of defensive activity. _J.
proj. Tech, 1958, 22, 28h-292.

 

HOUES, D. H. On the interpretation word frequency as a variable
affection speed of recognition. 'J. exp. Psychol., l95h, h8,

 

HONES, D. H., & SOLOEON, R. L. A note on MCGinnies' "Emotionality and
perceptual defense." Psychol. Rev., 1950, 57, 229—238.

 

HOWES, D. H., & SOLOMON, R. L. Visua1.duration threshold as a function
of word—probability. _J. exp. Psychol., 1951, Al, hOlrhlo.

 

HOWIE, D., Perceptual defense. .EZEREE'.EEX'J 1952, 59, 308u315.

JENKIN, N. Affective processes in perception. Psychol. Bull., 1957,
Sh, 100—127. '

 

KURLAND, S. H. The lack of generality in defense mechanisms as indi—
cated in auditory perception. .J. abnorm. soc. Psychol., 195A,

89, 173-177.

 

 

LACY, O. N., LEWINGER, N., & ADAMSON, J. F. Foreknowledge as a factor
affecting perceptual defense and alertness. .J. exp. Psychgl.,

1953, AS, 169-17h.

LAZARUS, R. S. Is there a mechanism of perceptual defense? A reply
to Postman, Bronson and Gropper. .J. abnorm. soc. Psychol., 1958,

A9. 396-398-

 

 

LAZARUS, R. 8., ERIKSEN, C. W. & FONDA, C. P. Personality dynamics
and auditory perceptual recognition. J. Pers., 1951, 19, h71~
h82. —

LAZARUS, R. S., & NhCIEARY, R. A. Autonomic discrimination without
awareness: A study of subception. Psychol. Rev., 1951, 58, 113~
122.

 

LINDNER, H. Sexual responsiveness to perceptual tests in a group of
sexual offenders. .J. Pers., 1953, 21, 36h—37h.

IONENFELD, J., RUBENFELD, 5., & GUTHRIES, c. M; Verbal inhibition in
subception. 'J. gen. Psychol., 1956, 5A, 171-176.

 

LECHINS, A. S. An evaluation of some current criticisms of Gestalt
work on perception. PsycholJ Rev., 1951, 58, 69-95.

 

LYSAK,'N. The effects of punishment upon syllable recognition threshw
olds. .J. exp. Psychol., 195b, A7, 3h3—350.

 

HCCIEIlAND, D. C., & TIBERRAN, A. M. The effect of need for achievement
on recognition of need—related words. ‘J. Pers., 19h9, 18, 236~251.

McGINNIES, E. Emotionality and perceptual defense. Psychol.Rev., 19A9,
56, 2hh~251.

McGINNIES, E., & BOULES, N. Personal values as determinants of para
ceptual fixation. .E' Pers., l9h9, 18, 22h~235.

 

McGINNIES, B., e SHERNAN, H. Generalization of perceptual defense.
:1- aelara- £329.: .Psv'cbslu 1952: 147: 81—85-

LAUSNER, B., & SIEGEL, A. The effect of variation in "value" on porn
ceptual thresholds,_J. abnorm. soc. Psychgl., 1950, AS, 760r763.

 

ANRDOCK, B. B., JR. Perceptual defense and threshold measurements. .J.

NEEL, ANN F. Conflict, recognition time and defensive behavior. Amer.
Psychologist, 195A, 9, b37. (Abstract)

 

NELSON, S. H. Psychosexual conflicts and defenses in visual perception.
g. abnorm. soc. Psychol., 1955, 51, h27—h33.

 

 

NEWTON, K. R.‘ A note on visual recognition thresholds. J abnorm. soc.
Psychol., 1955, 51, 709~710.

OSLER, S. F., & IENINSONN, P. The relation between manifest anxiety
and perceptual defense. Amer. Psychologist, 195A, 9, hh6. (Abm
stract)

POSTNAN, L. The experimental analysis of motivational factors in perm
ception. In J. S. BROWN, H. F. HARLOW, L. POSTNAN, et al., Cur-

rent theory and research in motivation: s‘ osium. Lincoln:
Univer. of Nebraska Press, 1953. Pp. 59— 10

 

 

POSTMAN, L. On the problem of perceptual defense. Ps ycholm Rev.,
1953, 60, 298~306. (b)

POSTNAN, L., BRUNER, J. 8., & McGINNIES, E. Personal values as selecu
tive factors in perception. _J. abnorm. soc. Psychol., 19h8, h3

1112-1511 .

 

POSTNAN, L., & SCHNEIDER, B. Personal values, visual recognition and
recall. Psychol. Rev., 1951, 58, 271M288.

 

12h

POSTMAN, L., BRONSON, W. C., & GROPPER, G. L. Is there a mechanism of
perceptual defense? .E' abnorm. soc. Psychol., 1953, A8, 215—225.

 

 

POSTMAN, I”, & CONGER, BEVERLY. Verbal habits and the visual recogniw
tion of words. Scieppg, 195A, 119, 671—673.

PRATT, C. C. The role of past experience in visual perception. .J.
Psychol., 1950, 30, 85—107.

PRENTICE, W} C. H. "Functionalism" in perception. Psychol.Rev.,

1956; 63) 29‘38 0

 

REYHER, J. The posthypnooic stimulation of hypnotically induced con"
flict: III Perceptual Vigilance and Perceptual Defense. In
press.

SOLOMON, R. L., & HOMES, D. S. Nerd frequency, personal values, and
visual duration thresholds. Psychol. Rev., 1951, 58, 256—270.

 

SOLOMON, R. L. & POSTMAN, L. Frequency of usage as a determinant of
recognition threshold for words. J. exp. Psychol., 1952, A3,
195—202. "

 

SECAL, A. B. Prediction of expressed attitudes toward the mother.
Unpublished Ph. D. dissertation, Univ. of Michigan, 195h.

SHIRE, A. Personality correlates of defense preferences. Unpublished
Ph. D. dissertation, Univ. of Michigan, 195h.

SPENCE, D. P.‘ A new look at vigilance and defense. J. abnorm. soc.
Psycppl., 1957, 5A, 103—108.

 

STEIN, K. B. Perceptual defense and perceptual sensitization under
neutral and involyed conditions. ‘J. Pers., 1953, 21, h67—h78.

STERNE, D. N. Duration thresholds in the perception of unpleasant words.
Unpublished Ph. D. dissertation, Michigan State Univ., 1952.

THORNDIKE, E. L., & IORGE, I. The teacher’s word book pf 30,000 words.
New York: Columbia Univ. Press, 19Hh.

 

 

VANDERPLAS, J. N., & BLAKE, R. R. Selective sensitization in auditory
perception. .J. Pers., 19h9, 18, 252-266.

VOOR, J. H. Subliminal perception and subception. J. Psychol., 1956,
Al, b37-b58-

WALLACN, H. Some considerations concerning the relation between per~
ception and cognition. .E° Pers., 19h9, 18, 6—13.

12.5

NIENER, M. Nerd frequency or motivation in perceptual defense. _._J_.
abnorm. soc. Psychol., 1955, 51, 21h—218.

 

 

WYLIE, R. C. Cognitive set and motivational factors in the perception
of neutral and threat stimuli. 11' abnorm. soc. Psychol., 1957,
SS, 227*2330

 

APPENDIX A

DISTRIBUTION OF SUBJECTS

 

 

 

---——‘—.~.- -‘——.—.———.——--.~.— .~ . --........_..__.._.. ‘- ..-... -..

-~A~'I--t~ .»_...“_ -.—.——.—-.- -un— -.... n ..

 

-.—. ~_ —.'.- _- -—- - -.-~...-— u...—

Avoidance Group Projection Group

~..-—.. -..--_... ,. . . .3..-

No Defense Group

 

--——- - .5.— u——_'--_—._—-.—v—. --_ . 4..

.—-—- ~~~u.-.. - H

 

-___.. “-~-.—.——o—-—~— .. -—---—._._..-_._ --. —--.-

 

22 Subjects 18 Subjects
10 male 11 Male

12 Female 7 Female

Total Subjects = 65

 

.m‘. _-....‘~/-

25 Subjects
11 Male

11; Female

127

APPENDIX B

m: DEFENSE PREFERENCE mourn

Oral eroticism

Rank all of the statements according to how well they seem to fit the

I
situation.
(Int) A.

(Rs)

II

(Re)

(Rf )

(Int)

(A)

(P)

Oral

A.

B.

O
o

1 2 fits best A 2 fits fourth best

2 = fits second best 5 2 fits worst

3 2 fits third best

Blacky makes sure he eats heartily to facilitate the growth

of healthy bodily tissues which will.fortify him for acti—
vities which might lie ahead.

Blacky tends to act in the same helpless infantile way as
when he was first born, often stuffing himself more than is
good for him.

Blacky is busy getting his dinner here, but he thinks it is
really'lhrmm.who makes him eat so much all the time.

AS Blacky eats, he thinks to himself that it's time he went
after his own food rather than having to depend on Hanna.

When Blacky is busy with other things, he often forgets to
come to eat, but here he's making up for lost time.

sadism

‘Nhen Blacky gets angry, he often throws a temper tantrum
like he did in his earlier days.

Blacky tries to pretend that he's ferocious, but when Namma
is around he is sure to be overly gentle, calm, and well~
behaved.

Blacky is a firm believer in the idea of releasing one's
aggressions, so he feels justified in ripping Namma's
collar here.

Blacky is so intent on chewing the collar to pieces that he
doesn't even realize that it belongs to Namma.

In Blacky's own way of thinking, his family has been treating
him so unfairly that he feels entitled to chew up the collar.

128

THE DEFEI'.‘SE PREFERENCE II‘MUIRY

III Anal sadism

(A) A.
(Int) B.
(Re) C .
(P) D.
(Rf) E.

It may look like Blacky is relieving himself between his
parents' houses, but possibly he's just digging a hole to
bury a bone.

Black; knows Lhnra.and Papa are not going to like the spot
he chose, but to his way of thinking a dog's physical well—
being is extremely important.

The consequences of Blacky releiving himself there might
very well make him wish he were a young puppy again.

Blacky thinks his behavior is perfectly all right because
he senses that Mamma and Papa have been unreasonably irri-
tated with him, even though their actions didn't show it.

Blacky figures that Lemma and Papa will be pleased to find
him so clean and neat about covering his mess.

IV Oedipal intensity

(Rf) A.
(P) B.
(A) C.
(Re) Do
(Int) E.

Blacky is feeling irritable, but he'll cheer up when he
realizes that Namma and Papa love each other so much.

Though Blacky is the one behind the bushes, he's still up»
set by the thought that his parents are hiding their love-
making from him in order to keep him out of the group.

At the moment Blacky is upset watching his parents to—
gether, but he'll soon forget his anger as he starts play-
ing again.

As Blacky watches, he works himself into a fit of anger and
helpless rage which will force Mamma to take care of him
again.

Blacky feels justified in getting angry here because he
wants his parents to enjoy the other activities that he had
planned.

129
THE DEFENSE PREFERENCE INQUIRY

V masturbation guilt

(Int) A. Blacky has come to believe that frequent explorations of this
sort are necessary to learn more about the role of his sexual
anatomy in the functioning of his body.

(A) B. Though licking himself, Blacky isn't affected by sexual senm
sations and'will soon move on to other parts of his body.

(Rf) C. Blacky'will soon give up this childish practice and will
devote his time to more constructive activities.

(P) D. Blacky is enjoying his discovery, but he knows that others
often get very upset and guilty over such actions.

(Rg) E. ‘When Blacky gets puzzled by a strange new experience like
this, he naturally thinks back to the "good old days" before
such problems existed.

VII Overt Aggression:1

(Int) A. Blacky is so perplexed and frustrated by the toy that he
may lose control of his temper the way he did when he was a

pup.

(P) B. Blacky wants to mind his own business but he figures the toy
dog is trying to start a fight by blocking his path.

(A) C. Blacky is eagerly calling the family's attention to his new
toy dog, which he is very proud of.

(Int) D. Blacky feels justified in this furious outburst against the
toy dog because, after all, discipline is vital to the devel~
opment of a wellprounded personality.

(Rf) E. Blacky is delighted to have this little companion to whom he
can give advice, love, and affection.

 

1The defense alternatives on Picture VII concern reactions to the
overt expression of aggression rather than the identification
process, which the picture is intended to convey in the standard
Blacky.

l30

THE DEFENSE PREFERENCE INQUIRY

VIII Sibling rivalry

(A) A.

(Rf) B.

(Int) C.
(Rg) D-
(P) E.
IX Guilt
(P) A.
(A) B.
(Int) C.
(Rf) D.
(Rg) E.

Standing off at a distance like this, Blacky is impressed
primarily by the fact that his family is such an intimate
group 0

Blacky is pleased to see Lhnmm.and Papa being affectionate
to Tippy, since he feels that Tippy deserves a turn at
getting attention.

Blacky believes that insight into his own jealousy of Tippy
will enable him to handle himself better in competitive
situations later on.

As Blacky watches the rest of the family, he‘ll act like a
helpless infant so they will have to treat him like a baby
the way they once did.

Blacky suspects that Tippy has been trying to win over Mamma
and Papa because Tippy is envious of their feelings toward
Blacky.

feelings

Blacky feels that he wouldn't be in the spot he's now in
if others hadn't led him astray.

Blacky's unhappiness will force him to drift into other
thoughts which don't bother him as much.

Though he's suffering new, Blacky will come to realize that
his code of ethics is needlessly strict and confining.

After an experience like this, Blacky will become a model
of virtue and will scrupulously avoid any wrongdoing.

This experience proves so disturbing to Blacky that it will
be a long time before he is able to act his age again.

131

APPENDIX C

PERCEPTUAL STINULI

 

Sex~neutral stimulus pair

   

Aggressionuneutral stimulus pair

   

Neutral—neutral stimulus pair

132

APPENDIX D
TOTAL.ERRORS TO THE CONFLICT AND NEUTRAL STIEULUS

PRESENTATIONS FOR ALL SUBJECTS IN THE SUBCEPTION

 

 

 

SERIES
Avoidance Group Projection Group No Defense Group
Conf. Neut. Conf. Neut. Conf. Neut.

1 8 2 12 1h 10 9
2 10 11 7 10 12 8
3 l 8 9 ll 11 10
h 11 9 8 10 11 ll
5 8 1h 9 10 8 11
6 9 10 9 13 9 9
7 1A 10 12 lo 12 7
8 ll 9 9 9 h 6
9 ll 8 12 8 10 1h
10 6 12 12 ll 8

9 9 9 6 7 8

10 9 9 10 9 11

5 18 9 12 13 6

8 o 11 6 13 1A

8 10 5 9 lb 9

l3 8 ll. 7 9 1.1

9 7 8 12 1A 5

8 10 ll 5 11 9

9 ll 10 10

8 9 10 8

10 9 7 8

10 10 1h 13

10 11

9 9

8 8

TOTAI.ERRORS FOR ALL SUBJECTS TO THE SEX AND AGGRESSION

APPENDIX E

STIMULI IN THE SUBCEPTION SERIES

F.)

 

“-—

 

 

S Avoidance Group Projection Group No Defense Group
Sex Agg. Sex Agg. Sex Agg.

l 2 6 S 7 5 S
2 A 6 A 3 7 5
3 6 9 5 A 5 6
A 7 A 3 5 6 5
5 3 S A 5 3 S
6 S A S S 5 A
7 7 7 6 6 6 6
8 6 5 5 3 0 A
9 A 7 6 6 5 5
10 A 6 5 7 6 5
11 S A 5 A A 3
1.2 5 S A S S A
13 A l 2 7 5 8
1A 3 S 5 6 8 S
15 A A A l 9 E
16 6 7 A 7 A 5
l7 3 6 5 3 9 5
18 S 3 6 S A '7
19 A '5 ,3 7
20 3 S 7 .3
21 6 A 2 5
22 3 7 9 .5
23 6 A
2A 6 3
3 S

APPENDIX F

13A

NUMBER OF CORRECT DISCRIMINATIONS FOR ALL SUBJECTS

ON THE A0 TRIAL SUBCEPTION SERIES

 

_~_.. -._ ~-. -

 

———.—- .“wwm

 

S Avoidance Projection No Defense
Group Group Group
1 30 1A 21
2 19 23 2O
3 17 2O 19
A 20 22 18
S 18 21 21
6 21 18 22
7 L6 18 21
8 2O 22 30
9 21 20 16
10 2A 16 21
11 22 25 25
12 21 21 2O
13 17 19 21
1A 23 23 . 13
15 22 26 17
16 19 22 20
17 2A 20 21
18 22 2A 20
19 2O 2O
2O 23 22
21 21 25
22 20 13
23 19
2A 22
25 2A

APPENDIX G

TOTAL CORRECT DISCRIEIHATIONS FOR ALL SUBJECTS

ON THE ACUITY TEST SERIES

-..-

H.—.- — -_—---_ ._...__.—._. _. 1.. -_

 

 

 

...-o_---

 

S Avoidance Projection
Group Group
1 8 1A
2 13 13
3 10 17
A 11 1A
5’ 15 10
6 1A 16
7 1A 1A
8 11 9
9 10 12
10 12 10
11 16 11
12 12 l3
l3 l3 8
1A 11 15
15 15 15
16 10 12
17 12 10
18 15 15
19 10
2O 11
21 12
22 12
23
2A

—-.-

No Defense

Group

 

18
1.2
13
10
18
15
16
15
11

9

12
IO
10
17
16
16
ll
10
1A
13
11

8
12
11
IO

“ha—o...

C’

APPENDIX H

136

TOTAL NUMBER CORRECT DISCRIMINATIONS FOR,ALL SUBJECTS

ON THE PERCEPTUAL DEFENSE SERIES

 

 

.— —.._.

 

 

 

S Avoidance Projection No Defense
Group Group Group

1 21 33 2A
2 21 29 17
3 23 33 20
A 23 32 27
S 27 23 29
6 2A 37 29
7 23 3A 31
8 22 26 35
9 22 2 S 26
10 20 23 23
11 32 25 30
12 28 32 31
13 22 30 20
1A 27 36 3o
15 22 3A 31
16 28 32 22
17 25 27 23
18 27 36 28
19 19 22
2o 23 23
21 2A 18
22 25 21
23 22
2A 25
2S 19

APPENDIX I

137

DIFFERENCE SCORES FOR‘ALL SUBJECTS BASED UPON TOTAL

CORRECT RESPONSES TO THE ACUITY AND DEFENSE SERIES

 

S Avoidance Projection No Defense
Group Group Group
1 13 19 6
2 8 16 5
3 13 16 7
A 12 18 17
5 12 13 11
6 1o 21 1A
7 9 2o 15'
8 11 1.7 20
9 12 13 15
10 8 13 1A
11 16 1A 1.8
12 16 19 21
13 9 22 10
1A 16 21 13
15 7 19 15
16 18 2o 6
17 13 17 12
18 12 21. 18
19 9 8
20 12 10
21 12 7
22 13 13
23 1.0
2A 1A
25 9

138

APPENDIX J
NUMBER OF ERRORS TO THE CONFLICT AND NEUTRAL STIEULUS

PRESENTATIONS FOR ALL SUBJECTS IN THE DEFENSE SERIES

—--—-"._‘__

 

Avoidance Group Projection Group No Defense Group
Conf. Neut. Conf. Neut. Conf. Neut.

 

Hi—J
fiF-ommwmmtwmp

SEES

.Nmmmmmt—IHH
th-Camp—acwoow

12 7 1 6 8 8
9 1o 7 A 1.3 10
12 5 2 5 7 13
7 1o 2 6 A 9
10 3 8 9 6 5
1o 6 2 1 8 3
10 7 3 3 A S
9 9 S 9 3 2
11 7 S O 6 8
9 11 9 8 9 8
A A 7 8 6 A
8 A S 3 2 7
13 5 7 3 12 8
7 6 1 3 6 A
11 7 2 A S A
S 7 A A 10 8
9 6 5 8 12 S
6 7 l 3 7 5
12 9 11 7
11 6 10 7
6 1o 10 12
8 7 1.1 8
9 9

10 5

1.2 9

 

APPENDIX K
TOTAL ERRORS FOR ALL SUBJECTS TO THE SEX AND AGGRESSION

STIHULI IN THE PERCEPTUAL DEFENSE SERIES

.— - —.————_— .-......._ - m.---—.—.._‘_-~— ..—.....-.—...~ g..-_ -_- .-

 

.. .._- . -... ~

Avoidance Group Projection Group No Defense

Sex Agg. Sex Agg. Sex

._- ...—_.___~...-._--— . ..-

 

 

\o CON O\\n.t"w N y—J

wrt'mmz'm-qc'oowpomt‘moxmt‘mt‘w
\nN-xl-xufvlot'wmm'ww OWLFLC'CEW-flm'v'l
HOWHHOWNWC’NOOl—JC’HHWH
AJFJAJCJFJFJt-Conavieaviba+J£r+J+J£*C>
‘mCNUIt‘O\O~£'1T"mmmeIr—J mmF—v L—l mrmL—awmm

139

-\I+':'£T-\)C'P-\J‘w~\1\nNIT-\H—Jt’lr‘vl N mr‘w-s’wt‘co'w

1A0

APPENDIX L
NUMBER OF CORRECT DISCRIMINATIONS FOR KATE AND FEEAIE

SUBJECTS IN THE DEFENSE SERIES

 

~--_ -_-

 

S Avoidance Group Projection Group No Defense Group
Dela Fem. Male Fem. Male Fem.
1 23 21 33 29 17 2A
2 22 21 33 32 2O 27
3 2O 23 23 3A 35 29
A 32 27 37 26 23 29
S 28 2A 23 25 31 31
6 28 23 25 32 28 26
7 28 22 30 27 22 3O
25 22 36 21 31
9 19 27 3A 22 2o
10 23 22 32 25 3O
11 25 27 36 19 22
12 2A 23
13 23

"3434’.“

NOV 4; ‘81-‘72 '33

‘Ap. ‘
361%,, ,"jrw. -.a -
.‘av ‘ p' 1 - ' .’ r '
w‘v/i'u_.._.; ‘ l‘fll “