jjuujukal ummAim m m
nilr Q ltiflM

TMiSHtldS !H TtfS

K m m 9f msArn m m

AUTHOR

pm Mom srmi

UNIVERSITY

MICH S ftft Cctt> DATE
PUBLICATION NO

DEGREE

MW

l!!ll

UNIVERSITY MICROFILMS
J y mnm:
/ U

I LUI

t k Ik A

k

k i I A II I A

A

kI

DURATION THRESHOLDS IN THE PERCEPTION
OF UNPLEASANT WORTS

By
David. M. Sterne

A Thesis
Submitted tc the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Psychology

1952

d

ACKNOWLEDGMENT
The writer wishes to express his appreciation to
Dr. Albert I. Rabin, his committee chairman, whose suggest­
ions,

encouragement, cooperation,

and foresight contributed

so significantly to the completion of this dissertation.
Thanks are also due Dr. Milton Rokeach, Dr. M. Ray Denny,
and Dr. C. A. Lawson,

the other committee members, for their

stimulating questions and advice which did much to help the
writer clarify his own thinking in the formulation and treat­
ment of the problem.
Dr. Gerald Hover,

Chief Clinical psychologist at the

Saginaw Veterans Administration Hospital, was instrumental
in providing the writer with time, materials,
support1’ when these were most needed.

and ’’moral

To Mr. Richard Eehan

who was of valued assistance in the statistical treatment of
the data,

the writer feels particularly indebted.

Further

appreciation is also due the members of the Michigan State
College Department of Psychology who permitted the writer to
use members of their classes as experimental subjects, and
to the staff of the Veterans Administration Hospitals at
Ft. Custer, Michigan, and Saginaw, Michigan, who cooperated
in the use of hospital patients as subjects.
however,

Above all,

the writer wishes to make mention of his gratitude
ii

iii
to

t'rlose

patient

withe'ut whose

individuals,

the experimental

subjects,

services this project and the tar.v others like

it wer-uld have been impossible.

4

TABLE OP’ CONTENTS
A CK K 0 V.LEDGI.'E NTS
LIST OF T A BLES
LIST OF A P P E NDICES
INTRODUCTION
General Character!stics of* Studies in
the P r o b l e m Area
Recent Studies
E x p l a n a t o r y Concepts for Motivational
F a c t o r s in P e r c e p t i o n
Studies Supporting the Importance of*
Motivational Determinants in Perception
Differences of Opinion

and

Criticism

Studies Casting Doubt on 1he Generality
of the Results of the Foregoing
Experiment s
Summary
OFF PROBLEM
the

s -l e c t i o n

or s t i m u l u s l o r d s
Met hods Investigated
Summary

PREPARATIONS FOR THE EXPERIMENT
Special Problems
The hospital Group
The College Group
Summa ry
METHODOLOGY
Summary

:e '.t o *-1 t h e d a t a
Tests for Homogeneity and Transfornation of
the Data
Variables Investigated
Sub-categorizing the Variables
The Analyses of Variance
CULTS
The hospital Sample
The College

Sample

SCUS31ON OF THE RESULTS
The Hypotheses
Interpretations and Conclusions
Additional Conclusions and Comments
Comparison of Results with
Studie s
.Summary
General Conclusions
FI IGGRAPHY

those of Related

LIST O? TABLES
Page
No.

Table
No •
Homogeneity of Variance of Duration
Thre shold s

66

II

Frequencies in the Arbitrary Groupings of
Unpleasant Stimulus Worcs

81

III

Summary of Analysis of Variance of the
Number of Exposures to the Point of Word
F.ecognition by Thirty Hospital Patients

98

I

IV

V

Summary of Analysis of Variance of the
Number of Exposures to the Point of Word
Recognition by Thirty-two Hospital Patients
Summary of Analysis of Variemce of the
Number of Pre-reco nition Hypotheses made
tv Hosoital Patients
V

VI

VII

VIII

99

100

—

Summary of the Analysis of
Variance of the
Number of Exposures to the Point cf Word
Recognition ty College Students

101

Summary of the Analysis of

103

A Comparison Summary of Analyses of V a r ­
iance with the College and Hospital Samples

loif

Variance of the
Number of Pre-recognition Hypotheses made
by College Students

IX

Interaction of Unpleasant vs. Control x
Categories of Patients' Pre-recognition
Hypothe se s

106

X

Interaction of Unpleasant vs. Control x
Strongly Disliked vs. Mildly Disliked,
with Students

107

Interaction of Unple asant vs. Control
Common vs. Uncommon, with Students

108

XI
XII

x

Comparison of Mean Duration Thresholds
and Mean Thorndlke-Lorge Frequency Ratings
for Short and Long 'Words
-vi-

128

APPLTIDICES

A.v,pendix
:~c.
1.

Questionnaire - Personal LI leer and Dislikes

2.

Adams Sentence Completion Test

3.

Word List of 952 Unpleasant Lords
Hospital Peer Group Word List

5.

Pleasant and Control Lords List

6.

College Peer Group Word List

7

.

8.

College Subject Questionnaire
Taboo Lord Hating List

-vii-

INTRODUCTION
General Characteristics of Studies In the Problem Area
Since the time of Wundt,

considerable attention has

been focused upon the experimental
processes.

More recently,

exploration of perceptual

those aspects of perception which

have seemed to demonstrate the effects of personal m o t i v a ­
tion have been examined w i t h increasing concern.

Within the

last decade especially, the interest in experimentally i n ­
vestigating motivational determinants of perception has
risen to the point of stimulating many ingenious

studies,

and these in turn have inspired a number of provocative
hypotheses .
One avenue of investigation has employed stimulus
material under a v a r iety of circumstances in which numbers
of relatively plausible interpretations could be given to the
stimulus.
imental

The particular interpretations made b y the exper­

subjects have then been used as bases for making

inferences as to the motivational background of the responses
selected.

The use of ambiguous stimulus material

to provoke

responses from which it is possible to Infer antecedent m o ­
tivational processes in the observer is h a r dly novel;

It

occupies a central position among the techniques familiar to
the clinical psychologist today.

A primary justification

for the use of such ’’projective ’1 personality tests,

as

are the Rorschach and the Thematic Apperception Test,

is

that they frequently reflect motivational factors - often
factors of which the test subject is unaware.

In effect,

such tests capitalize upon certain overt responses

sugges­

tive of other responses occurring below the limen of
consci ousness.
In an article in which he reviewed experimental

studies

of unconscious behavior in relation to perception, I.Iiller
(l 6 } listed investigations carried on as early as 1863 *
These and somewhat later studies covered areas of subliminal
perception,

subliminal learning and conditioned response,

reportable effects of subliminal neural stimulation,
related topics.

In Miller*s own study,

and

in which he presented

his subjects with geometric figures on a transparent mirror,
the subjects were able to discriminate between the various
figures at illumination intensities b e l o w the level at which
the bare presence of the figures could be detected.

His

results have been rather generally accepted as evidence
supporting the existence of discriminatory responses in
which the subject was reacting to subliminal

cues.

This in­

terpretation has been criticized by Lazarus and McCleary
on the grounds that:

(a)

(8 )

the criterion of the subliminal per

ception was the correctness of the s u bject’s verbal

statement

(and thus apparently was not as strictly ’’subliminal" as
would have been an indication of response over which the

-

3-

sufcject had no degree of voluntary control), and

(b) the

determination of each subject*s threshold was a m a t t e r of
statistical treatment combined with subjective judgment.
More recent studies (2, 6 , 9# 1Q> 13» 2L, 2£, and 28)
have made use of a wide variety of ambiguous stimuli, i n ­
cluding patches of light to be equated with the size of
discs bearing various purportedly affectively stimulating
and neutral symbols, chromatic and achromatic figures, and
dimly illuminated or indefinitely outlined pictures, d r a w ­
ings, and words.

The tachistoscopic exposure of words or

symbols for very short periods o f time is a similar approach,
designed like the others to force the observer to depend to
a greater extent upon his own resources by rendering less
obvious the perceptual cues to which he is ordinarily accus­
tomed.

The assumption is that under such ambiguous circum­

stances, whatever hypothetical

interpretation of the stimulus

is aroused will be influenced more forcefully by motivational
factors and less effectively by the nature of the

stimulus

than would be the case with optimal methods of presentation.

d

-kRecent Studies
Explanatory Concepts for Motivational
Factors in Perception
In the attempt to connect various affective

states or

motivational needs in the individual with response phenomena
which have been observed in subjects in such states or with
such needs,
invoked.

a number of explanatory principle s have been

Certain of these have been attacked on the grounds

that they were derived as explanations of the results of
experimental investigations,

rather than having b e e n hypoth­

esized prior to the research and confirmed thereafter.
nature of their derivation, however,

should not be considered

as the Important criterion of their worth.
hypothesis is to be avoided,
inductive thinking,

The

The ”ad h oc”

not because it arises through

but because it explains nothing besides

the phenomenon which it was originally invoked to explain.
Thus,

if suggested explanatory concepts can be demonstrated

to have predictive value,

their contributions c a m o t

legit­

imately be disdained.
In the studies to be summarized in the following pages,
evidence for and against certain of these hypothesized inter­
vening processes will be presented.

Proponents of these

concepts report results of experimentation which suggest
that motivational determinants may play a significant role
In the perception of visual stimuli.

Critics present data

which they interpret as relegating personal motivation to a
negligible or minimal role in perceptual processes.

Certain

investigators have proposed that a more profitable point of
view would be to examine the phenomena in light of various
antecedent conditions,

attempting

to define interactions of

a number of variables which might be presumed to account for
experimental results otherwise difficult to reconcile.

This

is the point of view basic to the treatment of the problem
presented here.
Studies Supporting the Importance of
Motivational Determinants in Perception
Personal values as perceptual determi nants.
Postman,

Bruner and McGinnies

In 19if8,

(19) presented the results of

an experiment in which they administered the A l l p o r t - V e m o n
Scale of Values to a group of college students.

This was

followed by the tachistoscopic presentation of thirty-six
words selected to represent the six value categories devised
by Spranger and utilized by the authors of the scale.

Each

of the words was exposed three times at .01 seconds, repeated
three times again at .02 seconds, and continued in this
fashion with three exposures at each timer setting,

the

settings being increased in steps of .01 seconds until the
word was correctly identified.

This minimum length of

exposure required for the correct identification of the
stimulus word has been termed

the "duration threshold” .

-

6-

Rccords were kept of the exposure duration at which each
word was recognized and also of any "pre-recognition
hypotheses ’1 or guesses made prior to correct identification
of the word.
Pre-recognition hypotheses were categorized into the
following groups:
(1 )

Govaluant responses - those considered to represent
clearly the same value as the stimulus words.

(2)

Contravaluant responses - antonyms to the stimulus
words or responses serving to derogate them.

(3)

Structural responses - incorrect hypotheses prompted
bv
structural similarities to the stimulus word.
v

(i|.)

Nonsense responses.

(3 )

Unrelated responses - any responses not classifiable
into one of the above categories.

It was found that there was a significant negative
relationship between value orientation in terms of AllportVernon score rankings and the length of expcsure needed for
the recognition of the representative words for the values
ranked; hirh value worts were perceived more rapidly by the
group of subjects as a whole than were low value words.
Covaluant hypotheses were reported as occurring more freq­
uently in response to high value words than to low, while
both contravaluant and nonsense hypotheses appeared more
often in response to low value words.

-

7-

This experiment h as been criticized (17) on the
grounds that the statistical treatment of the data was i n ­
adequate,

a nd differences in familiarity with certain

classes of the words may adequately and more parsimoniously
explain the results.
An investlgation of the postulated process of "percept­
ual d e f ense .11

Some time later,

periment by McOinnies

the report of a related ex­

(llf) was released.

In this study,

the

investigator employed the following eleven ostensibly neutral
and seven emotionally toned

(underlined) words:

apple,

dance,

r a p e d , child, b e l l y , glass, river, w h o r e , sleep, k o t e x ,
broom,

stove,

p e n l s , music, trade, f i l t h , clear,

and bitch.

The emotionally toned words were apparently selected on the
a priori assumption that they carried their affective
connotation because of their association with subjects which
were in some degree "socially taboo" and thus apt to be
emotionally "loaded"
McGinnies*

for individuals educated in our culture.

experimental

subjects were eight male and

eight female college undergraduates in elementary psychology.
The stimulus words were exposed t a c h i s t o s c o p i c a l l y , once at

.01 seconds, once at .02 seconds, and so on until the d ura­
tion threshold had been determined.

A galvanic

skin reflex

apparatus was used concurrently w i t h the tachistoscopic
presentation.

McGinnies found a highly significant

relation­

ship between the GSR and word meaning during pre-recogni ti on
(and thus nominally subliminal level)

exposures, with higher

-

8-

GSR's for the emotionally toned words than f or the neutral
words*

In addition, without

exception,

the observers d is­

played significantly higher duration thresholds for the
emotionally toned words than for the neutral words, when
mean thresholds for each class were compared.
McGinnies also attempted a content analysis, with a
classification of the pre-recognition hypotheses into four
group ss
(1)

Structurally similar - resembling in structure the
stimulus words.

(2)

Structurally unlike - with no structural

resemb­

lance to the stimulus word.
(3)

Nonsense - without dictionary

meaning.

(If)

Part - any letters not connected into a group.

He found that his subjects made proportionately more unlike
and nonsense hypotheses to the critical words than they did
to the neutral words,

and fewer similar and part responses.

The chi-square test of Independence between type of hypoth­
esis and type of stimulus words was

significant below the

.01 level of confidence.
McGinnies interpreted his experimental results as dem­
onstrating the presence of ’’perceptual

defense," a concept

which had arisen earlier in connection with the experiment
by Postman, Bruner and McGinnies previously cited (19)Perceptual defense refers to a raising of the duration
threshold displayed generally by Individuals when faced with

-

9-

stimulus objects perceived as threatening.

McGinnies has

also called attention to occasional instances in his data
of responses to taboo words at thresholds significantly
below the mean thresholds for neutral words, and this
phenomenon has been termed "vigilance."

Both concepts are

usually interpreted in terms of subliminal activity involv­
ing the autonomic response levels.

In the case of

the

specific critical words used in his experiment, McGinnies
attributed their threatening connotation to the effect of
early emotional conditioning, with punishment by parents as
the conditioning agent associated vi th the use of the
socially taboo words.
As pointed out by McGinnies (llj.) this experiment has
been criticized by Bruner on the grounds that the critical
words used appear less frequently in print than do the others,
and that therefore the Increased threshold may be explained
on the basis of a difference in familiarity with the words.
In addition,

the heightened GSR, it has been suggested, may

have been an effect attributable primarily to the greater
effort required to recognize the lees familiar words.

In a

similar vein, Howes and Solomon (3>) claimed the following:
(1)

V.hen the effects of differences in frequency of
appearance are extracted from the data,

the

differences in threshold disappear.
(2)

The experimental situation In which McGinnies used
a female assistant would tend to promote Inhibition

-

10-

of overt report of the taboo words eliciting strong
GSR*s.

This would constitute voluntary refusal to

express verbally the nature of the immediate p e r ­
ceptual experience,

rather than the effect of any

genuine subliminal blocking processes.
A corollary possibility is that the heightened G S R fs
with the taboo words may have been due to conflict between
the desire to cooperate with the experimenter* s instructions
and the urge to inhibit the verbal communication of the taboo
word.
Further investigation of dlscrimination without aware­
ness.

An interim report on a study by McCleary and Lazarus

(12 ) was provided by these investigators as further evidence
that observers give discriminatory (GSR) responses when con­
fronted with visual stimuli presented tachistoscopically at
subthreshold durations.

The implied perceptual process has

been termed by them " subception.M

In their ingenious study,

nonsense syllables were utilized as stimuli, with the ten
syllables used with each subject divided into two groups of
five each.

The ten syllables thus paired were equated in

terms of the number of times each word of each pair was used
and the number of times it was recognized during a 100
response equation period of practice for each subject.

One

group of the experimental syllables used with each individual
was conditioned to electric shock, using partial reinforcement,

4

-

11-

to the point where consistent conditioned responses were
obtained.

During test administrations of the ten syllables,

the subjects showed a reliable tendency to greater GSR at
subthreshold durations in response to the
with shock.

syllables paired

Of additional interest was the observation that

when the accuracy of recognition for the two groups of
syllables at each of the five exposure speeds used was
equated for the frequency w i t h vihich the syllables had been
employed,

the shock-condL tioned syllables were recognized

more accurately than the others,
first subjects used.

with four out of the five

However, a later report (8 ) revealed

that additional experimentation did not

substantiate this

trend - that in fact no significant difference was dem­
onstrated between the accuracy of identification of shock
as compared to non-shock paired
posure

syllables at various e x ­

speeds.

Generalization effects w i t h perceptual d e f e n s e .

In a

very recent experiment designed by the senior author as a
tollow-up of his study concerned w it h emotionality and p e r ­
ceptual defense

(11+-), McGinnies and Sherman (l£) have

Investigated further the hypothesis that perceptual defense
represents a genuine interference or temporary repression of
perceptual nature rather than merely suppression of verbal
report.

Eighteen ostensibly neutral words,

each five letters

in length and with approximately the same frequency rating

-

12-

in terms of the Thorndike-Lorge list

(27)* ana f o u r of the

socially taboo words used in the previous experiment were
employed in this study.
cable,

flush,

The y are as follows:

bitch, brand,

frock, glide, grind, hound, legal,

outer, penis, phone,

quest, ranch, raped,

spray,

and whore.

towel, weave,

twenty naive male

rider,

lucky,
scent,

The subjects used were

college undergraduates.

The words were

exposed tachistosc opically, with eight pairs of the words
being shown to each subject.

The first words presented in

four of the pairs were neutral words, whereas in

the remain­

ing pairs,

words.

the initial word was one of the taboo

In

all of the pairs, the second word exposed was a neutral
word.

In each pair,

the first word was

shown for a period

long enough to insure correct identification by all subjects.
The second word of the pair was shown at much more rapid
exposure settings.
The following procedure was observed.

The first word

was exposed at a duration of approximately two seconds,

and

immediately thereafter,

the second word of the pair was ex­

posed for .01

If the second word was not Identified,

seconds.

the first word of the pair was a g a i n presented for the twosecond period,
of the pair,

and immediately followed by the second word

this time exposed for

routine was continued,

.02 seconds.

Thl s

with the exposure period for the

second word of the p a i r being increased in increments of .01
seconds to the point of recognition.

Three pairs of the

-

13-

neutral words were similarly presented before the test runs
for practice purposes.

The pairing was random, with a dif­

ferent order being used for each subject.
In general, the duration thresholds for the neutral
words which followed the exposure of taboo words were greater
than those obtained for neutral words preceded by other
neutral words.

The differences between the thresholds of

these two groups of neutral words were significant at the
.01 level of confidence.

However, the proportion of hypoth­

eses made (number of hypotheses made by the subject to each
stimulus word,

divided by the total number of exposures to

the point of recognition) was approximately the same for both
neutral and " c r i t i c a l p a i r s .

The results with respect to

the duration thresholds were interpreted as supporting the
assumption that the actual perceptual threshold is affected
by embarrassment or anxiety producing stimuli, with the
raised thresholds of the second words of each pair due to
a generalization of perceptual defense effects rather than
mere suppression of verbal report.
Differences of Opinion and Criticism
Not all of the experiments designed to test the
postulated effect of motivational determinants on perception
have lent as much support to the various hypotheses regarding
their characteristics or effects as have those already
mentioned,

nor has there been an absence of pointed criticism.

-

1^ -

On the basis of their own investigations, Klein,

Schlesinger

and Keister (6 ) believe that the discriminatory error found
in their research which involved estimating and matching
physical attributes of stimuli with value connotations,
could not be unequivocally attributed to the influence of
value aspects of the stimuli employed.

They suggest the

advisability of returning to individual variations, p e r ­
ceptual organization,

and ego structures as areas of fund­

amental importance in the study of perception.

Luchins (11)

decries the emphasis on the hypothetical and deductive
approaches to the problem, to the neglect of inductive ob­
servation and investigation.

In addition, he stresses the

importance of field conditions and the desirability of for­
saking the tachistoscope in favor of techniques In closer
conjunction to everyday life and the clinic.
Studies Casting Doubt on the Generality of the
Results of the Foregoing Experiments
Word frequency v s . personal values in visual duration
thresholds.

Recently,

was, in effect,
by Postman,

Solomon and Howes (26) conducted what

a replication of the earlier experiment done

Bruner,

and McGinnies

(19) using the Allport-

Vernon Scale and tachistoscopic presentation of representative
value symbols.

They used words actually selected from the

Allport-Vernon test,

and compared with them the perception of

synonyms or cognate words not differing greatly from them in

-

15-

numbers of letters but considerably discrepant in terms of
frequency - the m a tched cognates were all much less common
words, in terms of Thorndike - Lorge

(27) frequency ratings.

The results bear attention.
Their data showed no indication of systematic variation
of duration thresholds with value

scores, but did reveal the

definite influence of frequency.

The difference between the

mean thresholds for the frequent and infrequent words was
appreciable, with the infrequent words exhibiting considerably
greater variation in mean thresholds as a function of value
ranks than did the frequent words.

While the results do not

bear out those obtained by Postman, Bruner, and McGinnies In
significant degree,

a trend seems in evidence pointing

toward lower thresholds for words representing the high
value ranks,

but the authors point out that the evidence for

the operation of emotional

determinants in visual duration

thresholds Is inconclusive.
Perceptual

selectivity with name s of traits regarded as

desirable or u n d e s i r a b l e .

In another study dealing with

perceptual selectivity as influenced b y value, Postman and
Leytham (2 0 ) employed trait names as stimuli, presenting
these tachistoscopically on a screen.

Not only did their

data provide no evidence for the faster recognition of traits
rated by the subjects as desirable, but these names were in
fact Identified more slowly than others exposed.

The

-

16-

writers pointed out that the cases of heightened sen­
sitivity to positively valued stimuli in psychological
research are matched by reports of lowered thresholds for
negative and threatening objects.

In essence, however,

they contended that it is the intensity rather than the
nature of the consequences which is the important element
in establishing the degree of motivational

support for a

perceptual hypothesis.
Task completion and its effect on perceptual
sitivity to stimuli symbolic of the tas k .

sen­

Postman and

Solomon (21) have endeavored to test the hypothesis that
stimuli symbolic of completed and incompleted tasks will
influence perception differently.

They made no specific pre­

diction as to the direction of the difference,

assuming

that this would be a function of the Individual subject's
attitude toward completion or incompls tion of the task, in
particular when his task performance represented success or
failure to him.

For their group of subjects as a whole, no

significant differences were obtained between symbols of
completed and of incompleted tasks, but the majority of in­
dividual subjects did deviate significantly in the direction
of lowered thresholds for either success or failure words.
A relationship between personality dynamic s and auditory
perception.

Lazarus, Eriksen, and Fonda (7) reported the

results of a study in which they attempted to examine the

-

17-

relationship between performance on a sentence completion
task and the perception of sentences with sexual, aggressive,
and neutral reference, partially masked by noise, on a wire
recording.

They found positive correlations of ,1^.6 to •7U-*

significant beyond the .01 level when evaluated via Fisher*s
z, between the sentence completion test responses and the
perceptual accuracy for sentences concerned with sex and
hostility.
(1)

Two basic types of reactions were observed:
High accuracy in perception, accompanied by ready
verbalization;

(2 )

and

low accuracy with minimal verbalization and block­
ing.

Individuals were found to be consistent with b o t h .
Summary
The recent intensification of interest in experimentally
investigating the role and nature of motivational deter­
minants of perception has

stimulated ingenious experimenta­

tion and provocative hypotheses.

Proponents o f motivational

determinants as Important factors In perception have been
led to the postulation of Intervening processes such as
perceptual defense and subception.

These have been Invoked

in attempts to relate differences observed in the perceptual
thresholds for high and low valued stimuli,
and neutral stimuli,

socially taboo

shock-associated and non shock-associated

stimuli, to differences in motivational connotations these

-

18-

stimuli are assumed to have had for the subjects.

Other

investigators have found no significant differences in per­
ceptual thresholds which they could relate to differences
in value, degree of task completion, and similar motivation­
al variables symbolized by the stimuli.

Instead, their data

have suggested variations in response availability on the
basis of prior familiarity with the stimuli as being one
significant factor in the differences of thresholds which
have been observed.

Uncontrolled variations in experimental

field conditions have also been suggested as responsible for
results obtained.

Thus,

the nature and role of motivational

determinants of visual perception,

and the existence and

character of the intervening variables postulated to account
for the observed phenomena remain controversial.
In short, the investigator in the field of the relation
ship of visual duration thresholds to motivational factors
is confronted with a number of alternative conceptions:
(1)

Duration thresholds of motivationally ’’loaded”
stimuli are different from those of neutral
stimuli.

(2)

Duration thresholds of motivationally loaded
stimuli are not different from those of neutral
stimuli•

(3)

When motivationally loaded stimuli have been ob­
served to have different thresholds from those
observed with neutral stimuli, the difference is

-

19-

no t a function of the motivational element, but
of uncontrolled experimental field conditions
which have biased the results in the predicted
direction,
(I4.)

Duration thresholds of motivationally loaded
stimuli are different from those of neutral
stimuli, but only when the motivational element
acts in conjunction with certain other variables.
It is the interaction of two or more variables,
each accentuating the effects of the others,
which is responsible for the differences, which
do not occur prominently when the motivational
element acts singly.

The other variables thus

limit the range under which the differential
thresholds are exhibited and define more specific
ally the character of the motivational

stimuli

which produce this effect.
The last alternative provides a general frame of
reference for the problem to be presented and for the mode
of approach.

THE PROBLEM
The writer*s main interest in investigating the
phenomenon of raised duration thresholds with certain types
of stimuli involved the conviction that the topic could be
explored fruitfully in somewhat closer association to actual
life situations than was the case w i th the experiments
previously described.

The question was raised, would not

this phenomenon be expected to occur most forcefully and
definitely with symbols representing particularly salient
experiences for the individual,
Further,

if it occurred at all?

if one may accept the conditioning of nonsense

syllables with electric shock as a paradigm of associative
learning with unpleasant or fear-provoking life situations,
could not one also expect an individual to react uncon­
sciously or subliminaly to verbal symbols associated with
actual past experiences under situations of this nature?
Would not one, in fact,

expect perceptual defense phenomena

to be at least as apparent with symbols particularly
relevant to the individual in terms of his own learning
experiences and personal associations as w i t h those chosen
because of the a priori assumption of their culturally
stereotyped characters as symbols of the
-

20-

socially taboo?

-21To demonstrate the operation of the process of
perceptual defense,

one must show that the duration thres­

holds for the critical stimuli are higher than those ob ­
served for otherwise equated stimuli of neutral character.
The nature of the motivation which provides the stimulus
employed with Its response arousing power must be specified,
and its existence determined independently of the prin­
cipal experimental procedure contemplated.
framework,
(l)

Within this

a twofold problem was suggested for investigation:
That of testing the primary hypothesis that
perceptual defense would occur with words of
idiosyncratically unpleasant
the individual involved.

significance for

This involved (a)

exploring the feasibility of using the tachistosccplc technique with idiosyncratically
unpleasant and neutral words,

equated in terms

of frequency of appearance in general reading
matter and in number of letters, to investigate
perceptual processes;

and (b) developing a proced­

ure to determine words which could reasonably
well be assumed to have definitely unpleasant
significance for the Individual

in terms of

whatever they represented to hi m personally as
a function of their associations for him.

-22(2)

That of investigating the secondary hypothesis
that pre-recognition responses made to the
idiosyncratically unpleasant words would differ
from those made to neutral words equated w i t h
them in terms of frequency and n u mber of letters,
w h e n the difference is examined in regard to the
frequencies of structurally similar,

structurally

unlike, nonsense and part responses*
The primary hypothesis, of perceptual defense, wa s
utilized as the focal point of the experimental design.
Specifically it had been hypothesized that for words of
idiosyncratically unpleasant connotation to the group
members,

the duration thresholds w o u l d be significantly

higher than those observed f o r words of ostensibly neutral
emotional reference, matched with the unpleasant words in
terms of number of letters and Thorndike-Lorge frequency
ratings.

’’Duration t h r e s h o l d ” , for reasons to be explained,

is used in this study to apply to the number of exposures
up to and including the point of word identification,
rather than to the timer setting employed at the point of
recognition.
The secondary hypothesis was that there would be signif­
icant differences between the frequencies of structurally
similar,

structurally unlike, nonsense,

and part responses

elicited by the unpleasant words and those elicited b y the
neutral words.

Since it was of interest to deal w i t h

-23interactions or a number of other variables which could
conceivably act to promote or inhibit the phenomena o b ­
served by Bruner, Postman, Mcjinnies, and their fellowv/orkers (II4-, I1
?, 19 ar*d 20), five subsidiary hypotheses
were developed in order that treatment of the data might
be rendered more systematic and effective.
hypotheses,

In these

the differences predicted were in terns of the

relationship between the unpleasant words and neutral o^r
’’control” words equated in number of letters and f r e q u e n c y .
The following
(1)

subsidiary hypotheses w ere proposed:

That words selected as strongly disliked,

on the

basis of composite weighted scores in terms of
peer group ratings, would have g r e a t e r duration
thresholds than those selected on the basis of
the same criteria as comparatively m i l d l y dis­
liked.
(2)

That there would be significant differences
between the duration thresholds fo r unpleasant
words selected as

strongly socially taboo on the

basis of prediction ratings by clinical psycho l­
ogists,

and those for unpleasant words selected

as mildly or not taboo,

the thresholds being

greater with the foraer.
(3)

That unpleasant words rated as least-liked by
individual

subjects would have greater duration

thresholds than w o u l d unpleasant words rated as

i

-2Unext least-liked,
(ij.)

That there would be significant differences
between the duration thresholds for uncommon
unpleasant words and those for common unpleasant
words,
ceived,

with the former being less rapidly per­
On the basis of this hypothesis,

the

differences in threshold between uncommon unple asant words and their matched control words
were predicted to be greater than those between
common unpleasant words and their controls.
(5>)

That there would be significant differences
between the duration thresholds for long u n ­
pleasant words and those for short unpleasant
words,
former.

the thresholds being higher with the
On the basis of this hypothesis,

it was

predicted that the differences in thresholds
between unpleasant and control words would be
greater with the long than with the short words.
The rationale for this and the preceding hypoth­
esis was that with very short or common
unpleasant words, there would be such rapid
recognition that motivational determinants should
be less effective than with longer or relatively
uncommon words, assumed to be of somewhat more
ambiguous appearance at very rapid exposures.

THE SELECTION OP STIMULUS WORDS
Methods Investigated
I n order to investigate the role played by motivational
determinants in the perception of tachistoscopically presented
unpleasant words, the first step was to develop a method for
finding suitable stimuli.

As has been indicated, what were

needed were words w h i c h could reasonably well be assumed to
have definitely unpleasant
subject,

significance for the individual

in terms of whatever they represented to him p e r ­

sonally as a function of their associations for him.

Further,

it was necessary that these words be suitable for presenta­
tion within the framework of the experimental design - they
had to be words for which it was possible to find a measure
or frequency of appearance In the f o r m of reading material
(their recognition was to be a reading task),

since it was

essential that control words equated in terms of number of
letters and frequency be selected for them.

This consid­

eration ruled out the use of any words considered profane or
obscene which did not appear fairly frequently in print and
had thus evaded the compilers of the Thorndike-Lorge word
list

{27), the criterion measure for frequency used through­

out the

study.

The exclusion of such words did not appear

-26to constitute any appreciable handicap in finding suitable
stimuli,

since at no time was the assumption made that the

words used were necessarily the most forceful or most
disagreeable words within the individual s u b ject’s vocab­
ulary.

It was desirable, however, to approximate this

extreme as closely as seemed feasible for each prospective
sub je c t .
In the search for an adequate technique for finding
suitable stimuli, words selected by five different methods
were considered and examined.

In the first technique,

words were abstracted from clinical folders of delusional
neuropsychiatric patients as symbolizing stabilized d e l u ­
sional systems of unple asant nature.

In the

words were selected on the basis of the

second method,

subject’s responses

to specific items in the Minnesota Multiphasic Personality
Inventory'*' - responses suggesting fearful or avoidant
tendencies toward environmental stimuli or stimulus sit­
uations, thus assumed to be unpleasant to him.

Since neither

of these two techniques was finally considered satisfactory
for the purpose under consideration,

they were discarded as

far as the experiment w as concerned,

and no further consid­

eration will be accorded them here.

*^S.
Hathaway and J. C. McKinley, Manual for the
Minnesota Multiphasic Personality Inventory, 195>1 revision,
The Psychological Corporation, New York, 31 pp.

-27Th© three other techniques which were Investigated
were the followings
(1)

Words were chosen by the writer on the basis of
his inferences made to a subject*s responses to
a specially developed questionnaire (Appendix 1).

(2)

Words were

selected by the writer on the basis of

his inferences made to a subject*s responses to

a

sentence completion test (Appendix 2).
(3)

Words were rated by the

subject as those least

liked of a list of words previously indicated as
definitely unpleasant by a peer group (Appendeces
I4. and 6 ) .
Each of these techniques was subsequently en^loyed and
will be considered in greater detail as follows.
The que stlonnalre.

The need for an adequate method of

identifying words of idiosyncretically unpleasant signif­
icance suitable for use as tachistoscopic stimuli still
remained,

and a number of oilier techniques were tentatively

considered.

The possibility of developing a questionnaire

specifically designed to explore

situations and experiences

which had been associated with personal discomfort or trauma
of one kind or another was attractive,
promise a more direct and economical

since it seemed to

solution to the problem

than had either of the techniques already examined.

-28A questionnaire
devised.

(Appendix. I) of nineteen items was

The questionnaire was introduced with the follow­

ing instructions mimeographed at the top:
In requesting you to fill out this questionnaire
as carefully and honestly as possible, we are
asking your cooperation in gathering material to
be used later in a psychological experiment.
We
wish to study certain things which most people
like and others which they dislike, and we hope
to learn more about how they react to these feel­
ings.
Take as much time as you feel is necessary.
All answers will, of course, be held strictly
confidential•
The questionnaire included Items regarding objects and
experiences of both pleasant and disagreeable nature,
although only the latter were of real interest within the
framework of the study.

The questions covered such elements

as lines of work liked and disliked, animals liked or found
disagreeable or repulsive,

and various personal experiences,

actual or potential, which the subject might consider pleas­
ing, irritating,

embarrassing, frightening,

The sentence comple ti on teat.

etc.

Concurrently, the writer

had been involved in another research project employing a
battery of psychological tests, among which was the Adams
Sentence Completion Test

(Appendix 2).

On the basis of the

Inspection of test responses of a number of patients, it was
decided to try out this technique for the purposes of the
study.

-29The sentence completion test was considered in this
connection, because it seemed to have the possibility of
covering a wide range of problems to which the patient was
almost forced to respond in his own terms,
at all.

In this respect,

if he responded

it resembled the Rorschach and

T. A. T., yet the responses obtained were often sufficiently
limited in scope to suggest specific words as representative
symbols for them.

The particular test adopted for trial is

one of sixty-one items,

constructed with a simple vocabulary

so that it is easily comprehended while providing a minimum
of information for the subject to use as a frame of reference
for his responses.

The items themselves were considered of

suitable character for the purpose,

and the comparatively

limited number of non-relevant items was an additional
desirable feature.
The peer group word 11 s t .

The need for so me method to

include more words likely to be idio sync rat ic ally unpleasant
to a major proportion of the experimental sample still
remained, however.

It was felt that the coverage of un ­

pleasant objects, words, and situations could be profitably
enlarged, and this provoked tne development of the fifth
technique examined - the "peer group word l i s t .11
Since it was obvious that any of the unpleasant words
used to test the hypotheis of perceptual defense and related
phenomena In the experiment would be limited to the confines

-30of the Thorndike-Lorge list (27)* the scheme of using this
list as the basic source of stimulus words

seemed not im­

practical.

Accordingly, the word list was examined care­

fully, and

words were selected by the writer as being

in his opinion conceivably unpleasant to a sizable number
of individuals.

These words were roughly divided into four

groups (2ij.O, 2 i|.0 , 239 , and 236 words respectively) on a
random basis, and each subgroup was then mimeographed on
a separate sheet
a small

(Appendix 3).

To each sheet was attached

slip of paper bearing the following instructions:

When people talk to one another, they do so with
words which may mean pleasant or unpleasant things.
When a person talks to a doctor or a psychologist, he
talks about his problems and uses words which often
mean unpleasant things.
We are interested in finding
more about how people react to the unpleasant words
which they use and hear.
In order to do so, we need
to find words which a great many people feel are u n ­
pleasant.
Please read through this list and put a
check in front of any word which you find you definitely
dislike.
The word may be unpleasant because of the way
it looks, the way it sounds, or what it represents.
Do
not mark any words which you do not know or fully under­
stand.
Your opinions will be held confidential and
your name is not required.
The lists of words were then distributed to forty-nine
patients of the Veterans Administration Hospital at Saginaw,
Michigan, a general medical and surgical installation.
Sixteen members of the hospital staff, nine of whom were
attendants, were also employed as judges, with a few of
these individuals making selections from all four of the
sub-group lists, over a period of time.

In all, eighty-four

-31of the lists were inspected and checked, with each subgroup
being examined by at least twenty-one individuals.

The

frequency with which each of the 955 words had been checked
was tabulated,

aid each word selected by five or more

individuals (as definitely disliked) was used to compile
a composite list of 96 words (Appendix I4.).

The composite

list was mimeographed with the following instructions:
Look over the following words aid select the five
words which you like least.
Mark a 1 in front of the
word you like least of all, a 2 in front of the word
you like next least, a 3 in front of the word you like
third least, and so on until you have indicated the
five words you like least in the order of their un­
pleasantness for you.
Make your choices from all the
words listed below, those typed on the list as well as
those mimeographed.
Ignore any word which you do not
recognize or which is unfamiliar to you.
Since no assumption was made that the words to be used
in the tachistoscopic presentation were of any specific
degree of unpleasantness (they were acceptable so long as
they could be assumed to be "definitely unpleasant"), and
since In any event,

only five words were to be rated as

least liked from a list of ninety-six words selected as dis­
liked by members of a fairly uniform cultural group, the use
of rigorous criteria for selection of the 65 judges did not
seem warranted.
The peer group word list csme to be the basic technique
utilized in the study.

The reference in the direction for

rating typed words as well as those which were mimeographed

i

-32was developed in regard to those words which were added
to the list on the basis of inferences made to responses to
the questionnaire and the sentence completion test.
three techniques, the questionnaire,
test, and word list,

These

sentence completion

were used with the first sixteen hospital

subjects prepared for the tachistoscopic procedure.

Inspect­

ion of the individual subjects 1 choices of the words rated
numbers one and two on the augmented word list - the onlyunpleasant words used in the tachistoscope - revealed that
out of the thirty-two cnoices,

thirty had been drawn from

the mimeographed section of the list,

and only two had been

elicited on the basis of questionnaire and test responses.
Because of the rapid turn-over of hospitalized patients, it
was often difficult to maintain contact wi t h subjects long
enough to complete the testing process, and the use of the
questionnaire and sentence completion test necessitated an
additional interview period to those required for the remain­
der of the procedure.

Therefore, it wfas decided to use the

questionnaire and sentence completion test only when time
permitted,

and to rely upon the ninety-six words of the

peer group word, list as the basic technique.

i

-33Summary
In the search for an adequate technique for finding
stimulus words for the contemplated experiment, words
selected by the following methods were

considered and

examined:
(1)

Words abstracted from clinical folders of delu­
sional neuropsychiatric patients as representative
of stabilized delusional

systems of unpleasant

nature.
(2)

Words selected on the

basis of the subject 1s

responses to specific

items In the ivlMPI - re ­

sponses suggesting fearful or avoidant tend­
encies toward environmental
situations,
(3)

stimuli or stimulus

thus assumed to be unpleasant to him.

Words chosen by the experimenter on the basis of
his inferences made to a subject*s responses to a
specially developed questionnaire.

(1+)

Words selected by the

experimenter on the basis of

his inferences made to a subject* s responses to a
sentence completion t e s t .
(5)

Words rated by the subject as those least liked of
a list of words previously indicated as definitely
unpleasant by a peer group.

The last technique was adopted as the basic method
employed In the study.

i

PREPARATIONS FOR THE EXPERIMENT
Special Problems
The primary purpose of the p re - expe r5 mental contacts
with contemplated subjects was to determine the specific
unpleasant words to be used as stimuli with them.

The

decision to use as the criterion of unpleasantness the
subject* s own verbal report raised certain difficulties
which had to be circumvented in one way or another.
problem of major proportions,

A

as has been the case with

so

many psychological experiments, was that concerned with the
necessity of keeping the subject relatively naive as to the
actual purpose of the experiment.

Were he to realize that

the words he selected as least liked would later be exposed
to him tachistoscopically in an attempt to determine the
effect of personal factors in his perception of the word, it
appears highly likely that these words would gain particular
salience for him, apart from that due to their unpleasant
connotation.

One would predict that under such circumstances,

the mention of the tachistoscope alone would tend to rein­
state associations with the word which would tend to lower
the duration threshold for it significantly; the subject
would, in effect, be looking for the word,
-31*.-

and ordinarily

-35mlnlmal cues could have pronounced and unambiguous meaning
for h im when the word was exposed,
A second problem involved no association w i t h the
tachisto scope itself.

It was simply concerned with the

point that under the limitations of time during which each
subject might be available,

especially within the hospital

situation, learning effects generated in the rating of the
words might carry over to the occasion of the tachistoscoplc
administration,

despite the fact that the subject might have

no specific knowledge of the experimental

sequence nor any

set for anything at all to follow the initial contact.

Such

learning would be especially prone to occur when emotionally
loaded material was concerned, and particularly when references
might be made to such material in not one but a number of
psychometric techniques (another consideration against the
use of the questionnaire and sentence completion test, In
addition to the word list).
Even the fact that all contacts with the experimenter,
for purposes of defining the stimuli as well as for the
actual test runs, might take place in the same room, with
the

same administrator, was conducive to the reinstatement

of various cues which could conceivably affect the experiment­
al results.

While It was impossible to control all of these

factors perfectly,

certain measures were adopted in the

attempt to cope with the major difficulties which were
fore seen•

-36The Hospital Group
Initial contacts wl th the sub ject s.

The first group

of subjects employed were patients of the Saginaw Veterans
Administration Hospital.

All of these subjects were seen

in the writer*s office, which was to serve as the exper­
imental laboratory as well.

The subjects could be class­

ified into two groups on the basis of the nature of the
writer*s initial contact with them.

Eight of the thirty-two

had been referred to the writer for psychological evaluation;
six of these had been referred as skin disorder cases in
connection with a research project concerned with dermatological difficulties with possible psychosomatic implica­
tions.

The remaining twenty-four subjects were simply

requested by ward personnel to report to the writer*s office
at scheduled times.

Thus, in the case of the first eight,

psychological rapport had been already establi shed, and the
rating of the words, the use of the questionnaire,

and the

presentation of the sentence completion test were accomplished
against a background of other psychological examinations the Rorschach,

the E*ender-Gestal t, and other standard

psychometric tools.
With the other twenty-four subjects,

the entire proced­

ure would have been meaningless and perhaps even somewhat
threatening,

had not a plausible

rationale been provided.

Since the cooperation of these subjects was important, even

-37after the actual testing was concluded,

the maintenance

of good relations with them w a s g i v e n constant attention;
widespread release of the nature of the actual procedure
would have rendered the results obtained with other, then
non-naive,

subjects suspect and probably unreliable.

these reasons,

For

each non-referral subject w a s acquainted in

a few brief words w i t h the skin disorder research project
concurrently in progress,

and then asked whether he would

volunteer as a control subject.
that his contribution woulc
psychological tests,

It was explained to him

involve taking a few short

extending at the most over a week and

a half and necessitating three or possibly four meetings
with the psychologist.
many cases,

In no cases, was there refusal;

in

the subjects voiced the conviction that this

would help pass the time, which apparently hung rather heavily
on their hands.
Selecting the stimulus w o r d s .

As soon as the coop­

eration of the subject had been enlisted, h e was given a
copy of the peer group word list to rate,

or when it was

considered that circumstances and time would permit,
administered

he was

the Adams Sentence Completion Test and the

specially devised and mimeographed questionnaire

(this was

accomplished with sixteen of the thirty-two subjects).
respect to the word rating,

if the subject expressed his

oifficulty in establi siting a frame of reference for

In

-38dlsilking the words,

it was suggested to him that the

words themselves were merely black marks on a white bac k ­
ground,

and that whatever the words signified to him was

probably a function of his own acquaintance with them.
If the subject had filled out the sentence completion
test and questionnaire during his first meeting with the
psychologist, he was administered the word list,

augmented

by whatever additional words his responses to the first two
instruments seemed to imply to the experimenter.

The

additional words, incidentally, were typed at the bottom of
the mimeographed list, using the same typewriter and spacing
employed in cutting the mimeograph stencil and thus rendering
the difference In appearance between the mimeographed and
typed words

relatively small.

Selecting th<3 control w o r d s .

As soon after the words

on the list had been rated as was feasible,

those words

rated number one and two, the least liked and next least
liked, were recorded,

and control words for each were

selected from the Thorndike-Lorge list

(27).

In each case,

the control words selected were of the

same number of letters

as the unpleasant word with which they were being matched.
i£ach control word was one which appeared to have no partic­
ular possible emotional connotation of any consequence,
the w r i t e r ’s opinion.
of that criterion,

in

Pecause of the recognized fallibility

and because of the possibility of such

-39words possess ing rare but

significant

ive loading for the subject,
for each unpleasant word.

idiosyncratic a f f e c t ­

fbur control w o r d s were selected

Nearly all control words were

m a t c h e d exactly with the critical w o rd in terms of its
numerical Th orndike-Lorge rating.

I n the few cases when it

appeared impo ssible to m a t c h four control w o r d s with the
unple asant w o r d while observing all three criteria (affective
n e u t rality and equality in letters and frequency rating),

the

last criterion was relaxed to the extent where a control word
wit h the next g r e a t e r or smaller numerical frequency rating
w as acceptible.

In such case, however,

this deviation w a s

bal anced by a deviation in the opposing d i r e ction in another
of the control words.

Words w h i c h met all of the criteria

but appeared to have markedly esoteric mean ings were avoided.
T he writer is well aware of the fact that

structural

differences in the matched words w ere not being controlled.
Unfortunately,

as Potter

(23)

and others have pointed out,

while differences distinguished between one word and another
are easily described,

little is understood about the p r o ­

cesses of discrimination w h e r e b y the d i s t i n c t i o n s are drawn.
One gesture in tnis direction was made,

however.

it was possible to do so, while observing
criteria for selection of control words,
with

two identical letters

control words w i t h the
the word,

"nigger,"

as one of four.

Whenever

the o t h e r three
unpleasant words

juxtaposed were matched with

same arrangement.

the control word,

In this wag,

for

"ballet," w a s employed

-4oControls for posslble learning effects.

In an attempt

to equalize for learning effects conceivably occurring in
regard to the two unpleasant words whe n they were rated, the
eight control words were shortly thereafter presented to the
subject, on the day following the rating whenever this was
feasible.

The control words were also administered in the

context of a rating procedure, in which the

subject was r e ­

quired to Indicate the eight words he liked least from a
list of thirty-one (Appendix 5>).

Twenty-three of the words

had been selected under the assumption that they would

have

a relatively pleasing connotation to most people - they in­
cluded such words as:

reward,

considerate, vacation,

attractive, delicious, amusing,
of the list were,

etc.

The other eight words

of course, the control words.

More often

than not, as h ad been the aim behind the design of the list,
the control words were the majority of w ords checked as
least liked.
Following the administration of this list, at least six
days and in most cases more were allowed to elapse.
end of this period,

At the

the subject was used to test out the main

hypotheses of the study, with the tachistoscope.
The College Group
Use of a college g r o u p .

In accordance with the exper­

imental design originally set up f o r the
two similar groups of hospital

study,

the use of

subjects had been contemplated

-1*1as cross valida tion rather than as experimental and control
samples.

It soon became apparent that possible influences

on experimental results of learning and cue reinstatement
arising from the use of the

subject’s verbal indication as

the criterion of unpleasantness of prospective experimental
stimuli,

and from the procedures used to identify such

stimuli, would be difficult to control wit h assurance.

When

the apparent superiority of the peer group word list as a
technique f or defining the idiosyncratically unpleasant
words was demonstrated (thirty out

of the first thirty-two

words chosen were drawn from the standardized mimeographed
word list rather than from the presumably personally relevant
augmentations typed on it),

the possibility of using a group

technique in selecting these stimulus w o rds received a t ten­
tion.

While responses to the sentence completion test and

the questionnaire often d i r ectly involved contemporaneous and
highly personal material,

such was not the case with the word

list, in which the motivational determinants frequently
seemed to function Indirectly.
advisable

Thus,

while it was

considered

to limit the use of the test and the que stionnai re

to the individual

interview situation, no such restriction

was felt necessitated by the rating procedure of the word
list.

Thi s was the essential

consideration which prompted

the decision to use a college group as the second sample•
Certain other advantages,

however,

also seemed to

accrue from the use of the non-hospital population.

It would

provide some indication as to the generality of any
results observed in the hospital sample and would also
provide a more suitable basis f or comparison of findings
with those of other investigations using college subjects.
In addition it made possible the economy of the adminis­
t r a t o r ^ time and energy inherent in the use of the group
technique as compared to that involving individual testing.
Moreover the fact that a college population was a relatively
stable one in terms of availability for testing and retest­
ing over a sizeable period of time appeared to be a distinct
asset.

At the same time,

from the point of view of

mechanics of test administration, it seemed possible in the
college situation to devise procedures whereby the element
of cue similarity between the circumstances of stimulus
determination and those of the tachistoscopic presentation
would be minimal.

Thus,

consideration was g i v e n to develop­

ing an experimental design using college subjects, in which
the rating of the word list would appear to have no connect­
ion with the tachistoscopic procedures, and in which there
would be a sufficient time interval between the two to
provide reasonable assurance against the probability of
learning factors significantly Influencing experimental
re suits.
Developing a. college peer group word list.
probable differences in intelligence, education,

Because of
socio-economic

level, and general cultural attributes which could be

-1*3presumed to exist between the members of the hospital
group and those of the college sample, the use of the same
peer group w o r d list f or b oth populations seemed inadvis­
able.

Accordingly,

the following procedures were instituted.

The 955 words abstracted from the Thorndike-Lorge list
and mimeographed in four portions

(Appendix 3) were admin­

istered to students in six elementary psychology classes
at Michigan State College.

The directions clipped to the

lists of words were identical to those used with the Saginaw
patients,

requesting the

students to check any words which

they definitely disliked for any or a number of reasons.
In this manner, 53 judgments were obtained in regard to the
words on each of three of the four lists,
were obtained on the fourth.

while 5k- judgments

O n the basis of the frequency

with w h i ch the various words had been indicated as definitely
disliked,

a composite list of the 196 words checked by twelve

or more judges was compiled and mimeographed (Appendix 6 ).
It was considered that the intellectual level and reading
skills of the college group could be assumed to be above
that of the hospital patients used previously, and that
therefore,

a longer list of words to be rated, covering as

wide a range of unpleasant words as feasible,

could be used

without the danger of producing excessive fatigue or boredom.
The words themselves were of sufficiently stimulating nature,
it was assumed,

to tend to reduce these effects to a level

-

14

-

considerably lower than that which mic^ht have been observed
with nonsense syllables, for example.
The new rating lists bore the following directions:
Please read these instructions carefully:
Look o v e r the following words and select the five
words you like l e a s t . Mark the number, 1, in front
of the one word you like least of all the words.
Next, mark a 2 in front of the word you like next
least.
Now mark a 3 in front of the word you like
next least, and so on until you have indicated the
five words you like least, in order of their u n ­
pleasantness for you.
You may find them unpleasant
because of their appearance, their sound, what they
represent or remind you of, or for any number of
these or other reasons.
V'.hen you have finished,
you should have five, and only five v/ords marked,
each with a different number.
An explanation for
this procedure will be supplied you later in the
course.
Sele cting the stimulus w o r d .

The wor d list developed

for the college students was then administered to two
classes in elementary psychology w h ich had not participated
in the compilation.

In the case of one of these classes,

the word lists were re-rated after a period of five weeks.
The per cent of agreement in terms of all five of the words
rated as least liked, was
thirty-nine students.

on the' basis of ratings by

The agreement in respect to words

rated numbers one and two only, the words selected as
prospective stimulus words, was

Since the ratings

were made from 196 words, all of which had been selected by
peer groups as definitely unpleasant to at least a quarter
of the judges,

the agreement cannot be regarded in the same

temns applied to common measures of test-retest reliability.

It is felt that the percentages of agreement obtained
represent a significant degree of stability in the choices
of unpleasant words, particularly since many of the shifts
from the first judgment to the second were to synonyms or
cognate words, not taken into account in the percentage
calculations*
It was from the second class to whom the word lists had
been administered that the college subjects for the tachistoscopic procedure were drawn*

The process of recording the

words rated with numbers one and two, and the subsequent
selection of four neutral control words equated with the un­
pleasant words in number of letters and frequency rating was
identical to that followed with the hospital subjects.
After a period of at least two and a half weeks had
elapsed since the administration of the word list for rating,
appointments were scheduled for individual meetings with
prospective subjects for the announced purpose of a "vision
experiment.”

The experimenter was identified by name before

the meetings took place, and the experiment was conducted
in a small room with which none of the subjects had had
previous contact.

Neither the names of the apparatus nor of

the experimenter had been mentioned in connection with the
rating of the words, nor was the experimenter present at any
time during the presentation or col] ection of the lists,
so the possibility of his being connected with the ratings
was remote.

Af ter the tachistoscopic runs had been completed
with all

subjects, a brief mimeographed brochure explaining

the background and purpose of the experiment was passed out
toeach student who had acted as a judge
words compiled into the word list or
the completed word list.

in selection of

the

who had actually rated

A short four item questionnaire

was attached to the back of the brochures given to members
of the class from which participants in the tachistoscopic
tests had teen drawn (Appendix ?)•

This questionnaire con­

sisted of the following items:
(1)
(2)
(3)
(1+)

Did you know that vision experiment was connected
with the ratings, before you saw any of the words
in the tachl sto scope'?
______ If not,
Did you have any idea that the two might be
connected before you saw the words in the tachistoscope? ______ If not,
Did you realize that the two were connected after
seeing the first stimulus (unpleasant) word? _____
If not,
Did you realize that the two were connected after
seeing both stimulus words? ______________
Comments:

Completed questionnaires were received from thirty of
the tnirty-three

students used as subjects.

Twenty-six of

the subjects answered all of the questions in the negative.
One

student answered "yes" to question (3), while another

stated that

she felt

some inclination toward the conclusion

that the rating and tachlsto scope run might have some
connection, after identifying the first unpleasant word.
Three students,

including the individual who had had the

-47inclination, mad© the connection after seeing the second
unpleasant word.
Summary
The decision to use the subject's own verbal report
as the criterion of unpleasantness of the critical test
stimuli raised the necessity of pre-experimental contacts
in each case.

Yli th subjects from the hospital group, the

following procedures were observed:
(1)

Each patient was interviewed and asked whether he
were willing to volunteer for a research project
which would involve his taking a few short
psychological tests and necessitating three or
possibly four meetings with the psychologist.

(2)

The

subject was then asked to rate the peer group

word list, or, when it was

considered that cir­

cumstances and time would perr.it, he was admin­
istered the Adams Sentence Completion Test and
the specially devised questionnaire.
(3)

If the subject had filled out the sentence comple­
tion test and questionnaire during his first
meeting with the psychologist,

during the second

meeting he was administered the word list aug­
mented by additional wards inferred by the
writer on the basis of the subject's responses
to the test and questionnaire given him previously.

i

(I4.)

Pour control words, matched in number of letters
and frequency

rating, were selected by the writer

for each of the two unpleasant words rated
respectively as least liked and next least liked
from the word list by each subject.

The control

words were chosen from the Thorndike-Lorge list
(2 7 ) and were all of ostensibly neutral affective
connotati on.
(5)

The eight control words chosen for each subject
were then administered to him in combination with
twenty-three words assumed to be pleasing to most
people.

Prom this combined list of thirty-one

words, the subject was asked to check the eight
words he liked least.

In this w a y he was exposed

to the control words in a situation similar to that
involved in his rating of the unpleasant words.
(6 )

Six days or more were allowed to elapse and the
tachistoscopic presentation was employed.

With the college

subjects,

the following procedures were

instituted:
(1)

A peer group word list was developed for the college
subjects in a manner similar to that employed with
the patients.

(2)

This word list was administered to two college
elementary psychology classes, and after a period

-ltf-

of five weeks was re-admini stered to one of
these groups for purposes of estimating the
stability of the ratings.
(3 )

For each member of the class not re-administered
the word list,

the words rated least liked and

next least liked were recorded, and four matched
control words were selected for each of these
rated words, meeting the same criteria used for
control words employed with the hospital subjects.
The writer was not present at any of the class
meetings during which the word list was developed
or rated, nor had his name been employed in
connection with any of these procedures.

(I4.)

At least two and one half weeks later, appointments
were made for individual meetings w i t h the students
for whose rated words, the control words had been
selected.

The appointments were made for what was

described as a "vision test."
(5)

The tachistoscopic presentations were instituted at
the time of the appointments.

(6 )

After all the subjects had been run through the
tachistoscopic procedures, a four item question­
naire designed to provide an estimate of the extent
to which the word ratings and the tachistoscopic
presentation had been associated by the students
was administered to each subject.

METHODOLOGY
Subjects.
the study.

Two groups of subjects were selected for

Since no inter-group comparisons were con­

templated and since each individual was to serve as his own
control, no procedures of matching individual to individual
or group to group were instituted.
The first group of subjects w as made up of thirty-two
male patients at the Veterans Administration hospital at
Saginaw, Michigan,

This hospital is a general medical and

surgical institution and the patients selected h ad been
hospitalized for a variety of reasons, with diagnoses cover­
ing such categories as hernias, rheumatoid arithritis, polio
sequelae, fractures,

bums,

skin conditions,

etc.

The only

criteria observed in the selection of these subjects were
the following:
(1)

That they were to be hospitalized for a period of
at least a week and a half after the time of the
initial contact with the experimenter.
necessitated by the design of the study,

This was
so set

up to minimize as much as possible memory effects
carrying over from the initial selection of the
test words to the actual test presentation.
(2)

That

they were ambulatory or at least wheelchair
-

50-

-

cases.

51-

The apparatus to be used was unsuitable

for bed patients nor was it practical to employ
as subjects patients who required assistance from
already heavily burdened hospital personnel,
(3)

That they gave no evidence of being psychotic or
subject to intellectual impairment of the type
found in senile or brain-damaged individuals.
was,

It

of course, desirable that reports of what

was perceived be as reliable as possible.
(I4.)

That they were not afflicted with an obviously
disabling visual handicap.

No attempt was made

to control for differences in visual acuity and
similar physiological factors, since each indi­
vidual was to serve as his own control and since
it was assumed that save for random variations,
such inter-subj ect differences would remain fairly
constant for all test stimuli.
(5)

That they could be assumed to be reasonably familiar
with the words used as test stimuli.

This was

determined at the time of the first exposure of
the control words to each individual by simply
asking him if any of the words were new or un ­
familiar to him^-.

Each subject was specifically

J-None of the words were indicated as be'ing ~unf am ilTar.

-52instructed to ignore a ny words which were
strange or u n f a mi liar to h i m in his unpleasant
word ratings,

and the control w o rds w e r e

selected to meet the criterion of familiarity
i n s ofar as the expei'Imenter could predict and
choose fro m the words available w h i c h met the
other criteria of length and frequency.
The ages of these pa tients ranged from twenty-one to

sixty-

four, w i t h the m a j ority of these below thirty-five.

No

assumptions were made in regard to differences in rate or
character of perception as a function of age,

intelligence,

socio-economic status or similar variables, and no attempt
to control for these was made.
The second group of subjects consisted of tnirty-three
college
these,

students f r o m one elementary psychology class.
seventeen were male and sixteen were female.

assumption w as made

that

Of

The

all of these subjects w o u l d be

familiar w i t h the comparatively common words u s e d as test
stimuli.1

The selection of the college students as subjects,

as has b e e n pointed out, was und ertaken in the attempt to
avoid the influence of learning and m e mor y factors in the
hospital experimental

situation which,

because of practical

-1-With the possible exception of two words,
to be the case.

this proved

-53con si derations , it seemed difficult to control.^
Equipment,

The instrument employed for presentation

of the stimulus words in this study was a Gerbrands Mirror
Tachisto scope similar to that used by Postman and Eruner
(18, 19) and by McGinnies (lit, 13, 19) in their tachis­
toscopic experimentation.

With this instrument, it was

possible to expose the words at durations ranging from ,01
seconds on upwards, with increments in exposure duration
of even .01 seconds.
The words were typed on sheets of white bond paper,
one word to a sheet, approximately centered on the page
alon? the longer axis.

Each word was typed on the same

electric typewriter to minimize differences in legibility
due to changes in pressure and stroke.

The words were

typed in capital letters, with one space intervening between
adjacent letters.
Experimen tal Procedure.

For the purpose of organizing

the experimental design, the hypothesis of perceptual
defense was utilized as a focal point.

Specifically it had

^All of the hospital subjects were questioned immed­
iately after the tachistoscopic presentation in regard to
whether they had remembered any of the words which had been
exposed, as coming from the word lists they had previously
rated, and especially as to whether they felt that they had
been able to anticipate words to any degree prior to their
correct recognition.
Three individuals answered the first
question in the affirmative, and only one, the second.

-5 k -

been hypothesized that for a given group of experimental
subjects, the duration thresholds for words of idiosyncratically unpleasant significance to the group members
would be significantly greater than those for words of
ostensibly neutral affective connotation,

equated w i t h the

unpleasant words in terms of numbers of letters and
Thorndike-Lorge frequency rating.
however,

It seemed desirable,

to utilize three successive exposures of the

stimuli at each duration in order to encourage maximal
development of pre-recognition hypotheses as well as to
minimize the effect of momentary vacillations in physio­
logical efficiency and visual acuity,
fatigue phenomena, blinking, etc.

and of accommodation,

Vvith this in mind,

the

number of responses up to and including the correct iden­
tification of the stimulus word was adopted as the actual
basis for comparison.

To have used merely the timer setting

of the duration threshold would have been to utilize a less
sensitive measure,

since with such a unit,

it w o uld not have

been possible to differentiate between the recognition of
words identified at the first exposure of a g iven duration,
and those identified at the second or the third exposure
with the same setting.

To rephrase the hypothesis in terms

of the number of exposures,

then, it was postulated that for

the group of subjects, the number of exposures required to
identify unpleasant words would be significantly greater

-55than those needed for recognition o f equated neutral words.
The experimental setting was held as uniform as possible
for all the members of each group, and the actual procedure
of the administration was essentially the same.

All members

of the hospital group were tested in the same office,
all the college students were
room.

and

seen in the same experimental

Each Individual was introduced to the tachistoscope

with a few words of explanation in which the instrument was
named and described as simply a machine with w h ich pictures,
designs, or words could be shown for very short periods of
time.

The subject was then Informed that he would be shown

seventeen words,
er whatever he

and he was instructed to tell the experiment­

saw or thought he saw.

guess whenever possible.

He was encouraged to

If he was able to distinguish any

word, he was asked to report this.

In order to maximize

any ’’projectiveM possibilities of the technique, if he could
identify only certain letters but had any vague hunch or
fancy that it might be an intellirible word, he was requested
to follow this hunch and respond with a guess.

He was urged

to give the individual letters only when it was impossible
for him to construct a word from what he saw.

If he could

identify nothing but a flicker, he was to say, "flash".
A slight variation was used with the college students
who had been scheduled for a "vision experiment".

They were

Introduced to the test situation as one which was concerned

-

56-

with reading and vision, more specifically, with the
identification of common words and the letters making them
up, as a function of the shapes of the individual letters
and their arrangements in words.

Then, the

same instruct­

ions for responding were used with them as with the other
subjects.
The ten test words for both groups of subjects comprised
two series of five words each.

One word in each series was

the word previously rated by the subject as least liked or
next least liked of the prepared list of words
unpleasant by members of his peer group.

selected as

In addition, with

certain members of the hospital group, the lists had been
augmented with words chosen on the basis of the

subject’s

responses to the open-end questionnaire (Appendix 1) and the
Adams Sentence Completion Test (Appendix 2).

The remainder

of each five-word series included four "neutral" or "control"
words - words of no particular apparent emotional

signif­

icance, equated with the unpleasant word of the series in
terms of number of letters and frequency rating In the
T h o m d i k e - L o r g e Y.ord List
Each

(27).

series was presented with two of the control words

first, followed by the unpleasant word and then the other
two control words.

The first series was preceded by the

presentation of five "warm-up" words - common words of
apparently neutral emotional connotation, made up of three,

-57four , five, six, and seven letters respectively, and
snown in that order.
subjects:

These five words were used with all

rug, loaf, short,

summer, and opening.

In

addition, to establish a more similar spatial frame of
reference for the test words to follow, as an aid in
focusing,

an additional warm-up word, a common neutral

word which equaled in number of letters the words of the
series to follow it, was shown immediately before each test
series.
The following procedure was observed with each word,
including those administered for warm-up purposes.

The

word was exposed three times at .01 seconds exposure
duration, with as much time between showings as the subject
required for his response.

If the word had not been

correctly Identified, it was re-exposed three times at .02
seconds exposure setting, then at .03 seconds exposure,

and

so on until correct recognition occurred, with the exposure
periods increasing in even units of .01 seconds.

All

responses were noted with the timer setting of the exposure
which preceded each.
At the conclusion of the testing of the hospital patients,
the general purpose of the experiment was explained to each
subject,

Whenever the scheduling of appointments, permitted,

each individual was asked for the reason for his choices
of the two least-liked words.

It wras then requester that

he not disclose any of the words used r or the specific nature

-58of the experiment.-^

When subsequent sii? jects were

questioned as to whether they had any ideas of what

the

experiment was about, prior to their owi initial introduct­
ion, none gave any evidence that the specific procedure or
purpose of the study had ’’leaked out” .

Since the

subjects

were selected from four different hospital wards over a
period of some months,

and since most of them did not

remain in the institution for more than two weeks, this
confidence may have required less effort to maintain than
might at first seem to be the case.
With the college students, no disclosure of the actual
purpose of the experiment was considered desirable until
the entire number of subjects had been run.
was, however,

Each subject

requested not to divulge the specific words

exposed to him,

since, it was explained to him, prior

knowledge would lead to fallacious findings or confusion
on the part of the subject who came in expecting to see
certain words.

All subjects agreed to

and apparently were most coopera tive.

this request freely
In two cases,

sub­

jects stated their convictions that there was a connection
between the rating; lists administered previously and the
tachistoscopic runs.

When they faced the experimenter with

lln all cases the subjects willingly agreed to this
request.

-59their conclusions, the connection was admitted and the
reason for the deception explained.

These subjects readily

acquiesced to the request that they not reveal the connect­
ion until the experiment was over.
Summary
Two groups of subjects,

thirty-two male patients of a

Veterans Administration general hospital, and thirty-three
college elementary psychology students, were used in the
study.

The instrument employed for presentation of the

stimuli was a Gerbrands Mirror Tachisto scope with which it
was possible to expose the typewritten stimulus words at
durations ranging from .01 seconds on upwards, with incre­
ments in exposure duration of even .01 seconds.

Each

individual was shown seventeen words, of which seven were
administered for practice purposes while ten were actual
test wcrds.

Subjects were instructed to report whatever

they saw or thought they saw, and they were encouraged to
guess whenever possible.

The words were shown three times

at each exposure setting, beginning at .01 seconds, with
the settings increased in steps of .01 seconds, until the
words were correctly identified.
All responses were recorded, with the timer setting of
the exposure which preceded each.

The general purpose of

the experiment was then explained to each of the hospital
patients, after which he was requested not to divulge this

-60i n f o r m a t i o n n o r specific particulars of the procedure.
W h e n e v e r time permitted,

each hospital

subject was asked

for the reasons for his c h o ices of the two least liked
words.
up any

He was also

questioned as to whether he h ad picked

significant infoimation about the experiment prior

to his own experiences.
With the coll ege students,

no routine disclosure of

the purpose of the experiment was made, but each individual
was asked not to divulge the specific words used with him.

TREATMENT OF THE DATA
Tests for Homogeneity and Transformation of the Data
While the design of the experiment was of a nature
which would have permitted treatment by several statis­
tical methods, the fact that the interaction of a number of
variables was of particular interest suggested analysis of
variance as an appropriate technique.

This technique is

commonly employed in testing the hypothesis that several
independent samples of data have been drawn at random from
a cominon normal population.

In the case of this particular

study, it was of importance to know whether the variation
between the response of groups of subjects treated under
diverse experimental conditions (in regard to characteristics
of the stimulus material) was significantly greater than the
uncontrolled variations in response between subjects treated
alike.

To rephrase the problem with an appropriate example,

was the variation between the duration thresholds of subjects
presented with unpleasant words and those of subjects present­
ed with neutral v.ords significant ly more pronounced than the
uncontrolled variations which would have been observed with
the subjects when presented with words not differentiated
in temns of the variable concerned?

Could any significant

differences in the means of the groups under the various
-6l-

-62experlmental conditions be legitimately attributed to the
differences in the experimental conditions themselves?
Certain assumptions are involved in the use of analysis
of variance:
(1 )

that the measures are independent;

(2 )

that they have been drawn at random;

(3 )

that they were taken from a common population,
not differing significantly in distribution from
normality.

There were no obvious reasons to suspect that the first
two assumptions were not being met within each of the two
sample s used.

It was, however,

considered advisable to test

the third assumption before attempting to estimate the
normality of the measures when grouped according to the
different experimental conditions.
Following the suggestion of Edwards (I4.), Bartlett* s Test
of Homogeneity of Variance was applied to determine whether
the variances of the data from the college sample, fran the
hospital sample, and from the two combined and treated as a
single sample were sufficiently homogeneous for the use of
the contemplated method.

Actually, there was no particular

reason to presume that the two samples did come f rom the
same population,

since the experimental treatment was not

strictly parallel nor was there any indication that the
populations themselves were matched In respect to such
variables as intelligence, education,

socio-economic level,

-63and general cultural background, any one of which might
logically have been predicted to effect certain of the var­
iables under consideration.

The chi-square obtained in the

test of homogeneity of the duration threshold variance of
the combined populations,

grouped into the four subsanples

of the thresholds for college unpleasant words, college
control words, hospital unpleasant words, and hospital con­
trol words,

was 233*2.

This constituted statistical grounds

at far beyond the .01 level of confidence that the combined
samples were not homogeneous enough to be treated by analysis
of variance, with the data combined in raw form.

The size

of the chi-square and the results of inspection of the data
suggested that much of the variance might be a function of
skewness in the distribution, with the means and variances
tending to be correlated.
As has also been the case with much other data recorded
in experiments involving timed measures, the distribution
appeared to be positively skewed.

That is, in comparison to

the median thresholds for each series of five equated words
(one unpleasant word w i t h four matched controls),

the more

pronounced raw score deviations were predominantly in terns
of raised thresholds.

The timer settings of the tachis-

toscope ranged from .01 seconds to 1 second, and the correct
identification of the words usually took place somewhere
between exposure durations of .02 to DI4. seconds.

Yet

-

61^-

recognitions at durations beyond .07 seconds were not
uncommon, and some were considerably greater,
ation of .80 seconds in one extreme case.

with a dur­

When the effect

of physiological limitations are taken into account, as
well as that resulting from the mechanical limitations of
the timer,
able.

such a skewed distribution is readily understand­

Deviations, in terms of raw scores, treated either as

normally or linearly distributed, would not accord sufficient
significance to the deviations below the mean.
Since inspection of the data suggested that the prin­
cipal source of variation lay in the thresholds observed
for the college group, the test of homogeneity was applied
to this group and the hospital group separately.

In each

case, the thresholds were classified into the two sub­
samples of responses to unpleasant and to control words
which were then compared for homogeneity of variance.

The

chi-square obtained for the college group alone was 129 »95,
a very significant difference again, and one which would
not permit the acceptance of the hypothesis of homogeneous
variance.
ever,

With the hospital group treated separately, how­

the obtained chi-square of .11 was not

significant,

suggesting that the analysis could be applied to this data
in its present form.
The distribution of the raw data was o f a nature
Intimating that the means and standard deviations would tend

-65to be roughly proportional,

thus violating a fundamental

condition of the analysis of variance which depends upon
the independence of the measures of inter-group variation
from those of inter-individual variability.

This suggested

that a transformation be employed in the attempt to convert
the original scale of measurement used to another with which
the analysis of variance wculd more likely be valid, the
variances being more homogeneous.

The t r a n s f o m a t i o n which

appeared appropriate was one of logarithmic character.^
Accordingly, the logarithm for the number of exposures to
the point of recognition for each word was recorded, and
this figure, multiplied by 100 and rounded off to the near­
est whole number, was the one adopted as the score unit to
be used.

This, it was felt, would stabilize the variance and

in addition would decrease the skewness of the raw data dis­
tribution, adequately reflecting the comparative significance
of both positive and negative deviations from the m e a n w h i l e
providing a convenient unit for computation.
The logarithmic transformation was first applied to
the data combined for the two groups,

since it was still of

interest to determine whether or not they could be considered
as coming from similar populations in respect to the var­
iables to be investigated.

The chi-square, computed with

the transformed data for the combined groups,

T^f. Edwards (Tp) p. 199*

divided into

66
four subsamples as before (college unpleasant, college
control, hospital unpleasant, and hospital control words),
was 13 .9 , significant at between the .05 and .01 levels of
confidence.

While this was interpreted as still admitting

the possibility of adequate homogeneity, under the more
rigorous criterion it was considered advisable to treat the
college and hospital groups as separate and different
samples.

Consequently, the test was then applied to these

two groups after transformation.

Chi-squares for the college

and hospital samples under these circumstances were 1.63 and
,0l|_ respectively,

statistically insignificant in both cases

and interpreted as indicating adequate homogeneity for the
analysis.

The results of the tests of homogeneity are

summarized in tabular fora as follows.
TABLE I
HOMOGENEITY OP VARIANCE
OF DURATION THRESHOLDS

Sample

Combined hospital and college
College alone
Hospital alone

Unt ran s fo rme d
_ dat a
X2
p
233.2

.01

129*95 .01
.11 N. Sig.

Logarl thmic
data
X2
p
P
13.9

.05-.01

I .63

N. Sig.

.0I4.

N. Sig.

-67Slnce after conversion to logarithms the variances
were appreciably stabilized with both samples, it was
decided to use the transformation with all the data despite
the fact that none of the basic assumptions for the analysis
would appear to have been violated, had the raw, untreated
data been utilized for the hospital group alone.

In

addition, with all data similarly treated, cross compar­
isons would be more readily grasped,

should they be desired.

The same transformation was applied to the measures of
frequency of the pre-reco^nition hypotheses.
Variables Investigated
V.ith the data transformed logarithmically, it was nov;
Dcssible to set up designs for statistical

treatment.

It

may be recalled that the primary hypothesis being inves­
tigated was that which dealt with perceptual

defense - the

raising, of duration thresholds in the perception of u n ­
pleasant words.

The secondary hypothesis was concerned with

differences In the frequencies of structurally similar,
structurally unlike, nonsense,

and part resjonses with the

pre-recounition hypotheses to unpleasant as compared to
neutral words.

Five subsidiary hypotheses had also been

advanced as a framework for studying the interaction of the
main variables with others considered of nossibla consequence
in explaining the conflicting experimental results obtained

-68by various investigators in the field.

The five subsidiary

hypotheses concerned the following variables:
(1)

The intensity with which the unpleasant word was
disliked by members of a peer group.

(2)

The degree to which an unpleasant word was

socially

taboo•
(3)

The degree to which the unpleasant word wa s con­
sidered unpleasant by tbe individual subject.

(If)

The relative familiarity with the word, control
words as well as unpleasant.

(5)

The relative length of the word,

control words as

well as unpleasant.
An additional variable was also considered to be of some
possible interest in relation to the college sample - that
of sex (the hospital group was entirely male).
It was in its interaction with each of the subsidiary
variables that the primary variable,

the unpleasantness of

the critical word, was o f maximum interest in respect to the
experiment;

these were the relationships therefore explored

by the analysis of variance used.
hypotheses,

With the pre-recognition

the interaction of the primary variable,

word

unpleasantness, was explored In Its relation to the fre­
quencies of the four categories of pre-recognition responses
which had been recorded, and in relation to the intensity
with which the unpleasant word was disliked by members of a

-69peer group.

The analysis of variance of the pre-recognition

hypotheses was designed after that of the duration thresholds
had been computed, and was developed to investigate the
effects of word unpleasantness for the individual and for
the peer group.
Sub-categorizing the Variables
The primary and the secondary variables.
treat the variables in their interaction,

In order to

it was necessary to

sub-categorize each into two cr more values or intensities
by employing arbitrary cutting points.

The division of the

primary variable of word unpleasantness in terms of u n ­
pleasant and control or neutral words was explicit in the
formulation of the basic hypothesis of the study, and the
cutting point here was determined essentially by the subject's
ratings of the unpleasant words and by the writer's select­
ion of the controls.

In regard to the secondary hypothesis

concerning: the frequencies of the various types of pre­
recognition responses,

the mere adoption of McGinnies*

(li|)

fourfolc system of classification with his published
definitions of the criteria for each category was not found
sufficiently discriminating in all cases.

Consequently it

was considered advisable to re-define the criteria for three
of the groupings.

-

70

-

As was the case with McGinnies* original use of this
categorization, the criterion of a nonsense response was
simply that it could not be found in a dictionary (29).
Under part responses were included any disconnected groups
of letters or any responses which the subject clearly iden­
tified as incomplete or fractional representations of the
stimulus as he perceived it.

In order to differentiate

between structurally similar and structurally unlike
responses,

it was considered expedient to establish a

uniform criterion which could be applied to all the responses
which did not fall into part or nonsense groupings.
A criterion was needed which would take into account
the differential cue strength of letters as a function of
their position in the word as well as of their structural
characteristics.

It was not at all rare to find a subject

making a whole series of pre-recognition responses which had
little in common with the

stimulus word or with each other

save for the first letter in each, yet in such a case the
evidence for an element of structural similarity was too
consistent to deny.

On the other hand, simply to make the

assumption that because two words have one letter in common
they are genuinely similar in structure, especially whe n
both words are long words,

seems to leave more up to chance

than would be warranted or desired in the classification
contemplated.

O n these grounds, the following alternative

arbitrary standards for characterizing words as

structurally

-

71

-

similar to the stimulus word were adopted:
(1)

The first letter or the last letter o f the
hypothesis and the stimulus word were common to
both •

(2)

In the case of hypotheses of three or fewer
letters, one or more of the letters were common
both to the hypothesis and the stimulus.

(3)

With hypotheses of four or more letters, two or
more consecutive letters of the hypothesis were
found in the same order in the stimulus.

(I4.)

With hypotheses of four or more letters, at least
half of the letters of the hypothesis were found,
in any order, in the stimulus.

Words found unpleasant by peer g r o u p s .

The first of

the subsidiary hypotheses was concerned with the intensity
with which words were disliked by members of a peer group.
Essentially the same procedure was followed with both the
college and hospital samples in securing a rough estimate
of the relative degree of unpleasantness w i t h which each
of the unpleasant words used as stimuli were regarded by
the group members.

The method of securing this estimate was

as follows.
With the college group, the five words rated as most
unpleasant by each of the 93 students administered the peer
group word list were recorded in the numerical order of the

ratings (a rating of number 1 represented the word designated
as most unpleasant or least liked, while a number 5 indicated
the fifth least liked).

With the hospital group,

the same

process was used with the ratings of the I4B individuals
prepared for testing.

All choices of the college group came

from the peer group word list, and all but 22 out of 2 l|.0
selections of the hospital
list

subjects came from t h e i r word

(the 22 were chosen fro m words added to the list on the

basis of responses to the questionnaire and the sentence
completion test).
Each word was then given an arbitrary weighted score
from 1 to 5 in inverse order to its numerical rating - words
rated number 1 or least liked were given a weight of 5>, while
words rated number £ were given a weight of 1.

Total weighted

scores for each of the words were then computed, and i n each
sample the unpleasant words used as stimuli were divided
into two groups with the cutting score being approximately
the median of the weighted score values in each distribution.
Thus, in the college group, the unpleasant w o rds having a
composite weighted

score of 18 or more were characterized as

"strongly disliked," while those with a score of 17 or less
were termed "mildly disliked," the assumption being that the
frequency with which the words were chosen as the five least
liked and the order in which they were

selected for this

dubious distinction were functions of their unpleasantness

-73for sample members as a group.

With the smaller number of

raters in the hospital group, scores of 13 and above were
considered indications of relatively strong dislike, while
scores of 12 and below were considered representative of a
milder degree of disliking.

The resulting frequencies in

these and the related groupings which follow are summarized
in Table II, page 81,
Words consldered socially t a b o o .

In two of the more

important experiments in the area of this study,

significant

differences in response were attributed to the effect of
motivational determinants.

These experiments were concerned

with critical words described as "socially taboo" (llg, 15 )*
Consequently it was considered important to explore the
possible effects of this variable within the framework of
the analysis;

this consideration was the basis for the second

subsidiary hypothesis.

It thus seemed desirable to diff­

erentiate among the various unpleasant words used as stimuli,
those which might arbitrarily be teimed "strongly taboo"
and those which could be compared as "mildly taboo."
such a distinction was made,

Once

it appeared feasible to exam­

ine the responses to the two categories of stimuli for
evidence of significant differences.
Rather than depend upon the judgment of a single in­
dividual in the proposed classification, the writer felt
that judgments of a number of clinically trained individuals

-7l+should be pooled to provide a more reliable basis for the
grouping.

A rating scale for all of the words used as

unpleasant stimuli was developed (Appendix 8 ), which bore
the following introduction:
It has been suggested that most individuals learn
early in life that certain words carry with them a
"socially taboo” connotation.
When these words are
used by a child, this use generally results In chas­
tisement by parents, with the consequent establishment
of a conditioned reaction to the verbal symbol.
This
pattern of conditioned emotional response may be con­
sidered to be one of anxiety or f e a r aroused by
symbols having sexual, excretory, or otherwise u n ­
pleasant or "immoral" connotations.
(McGInnies,
Elliott, "Emotionality and Perceptual Defense,"
Psychologlcal R e v i e w , 56, 191+9* P» 2 l4lf. )
Please rate the following words in terms of the
extent to which each word seems "socially taboo" by
putting a check in the appropriate space after it.
The words were then listed In a vertical column at the lefthand edge of the page, and after each word, four spaces were
provided for the r a t e r ’s checked indication of his opinion
in respect to the word.

The categories of rating were:

highly taboo, moderately taboo, slightly taboo,

taboo

and not

.

This rating scale was administered to sixteen college
faculty members and graduate students in clinical psychology,
each of whom had had a minimum of a year of clinical expe­
rience.

The raters were asked to predict how people In

general w o u ld have reacted to the words,

in t e m s

definition of socially taboo used on the scale.
of highly taboo was given an arbitrary weighted
while moderately

of the
Each rating

score of 3 ,

taboo, slightly taboo and not taboo were

-75accorded scores of 2, 1, and 0 respectively.

Composite

weighted scores for each word were computed, and the
approximate median score value of 13.5 was used as the
cutting point, with all words

scoring above it termed

’’strongly taboo” and those b e l o w termed "mildly tabo o . ”
The subject♦s dislike for hi s own selections.

Since

the selection of the unpleasant stimulus words used w it h
each subject was at least in part a function of his own
choosing,

it seemed desirable to Investigate whether

differences in the rating of these words by the subject
would be reflected in a differential response to them as
compared to their control words*

That

is, would differences

if any, in the strength of the effective response to a least
liked word as compared w ith that to a next least liked word
be reflected in the duration thresholds of the two words in
relation to the thresholds of their controls? This was the
basis for the third subsidiary hypothesis.
In order to provide some information in regard to this
question, the duration thresholds for the unpleasant and
control words were analyzed in interaction with a classifica
tion of series of w ords presented first and series presented
second.

Since in the college experimental situation there

seemed little reason to anticipate significant learning
effects carrying over from the rating task to that of the
actual test situation, for matters of convenience the first

-76series shown to each individual included the unpleasant word
selected as ^ ast liked.

In the hospital group, where the

probability of* learning effects and cue reinstatement,
especially after the recognition of the first unpleasant
word,

seemed greater, 15 of 32 subjects were

liked word in the first

shown the least

series administered, while 15 other

subjects saw the least liked word in the second series (2 of
the subjects had not been able to discriminate in their r a t ­
ings of the least liked and next least liked words, and thus
their thresholds could not be employed in this particular
context).

V.ith this split treatment used in the hospital

sample, it was thus possible to avoid confounding the effect
of the variable under consideration with the next one to be
regarded.
Order of presentati on.
of a subsidiary hypothesis,

Although n o t specified in terns
the order of presentation of the

stimuli was felt to be of sufficient interest to be concom­
itantly investigated with the others.
words had b een presented before the
test series,

and

since the

Since six practice

first word in the initial

results of inspection of the data

divulged no consistent changes suggestive of improvement
through learning, practice effects were not felt to be of
great consequence in the test task.
however,

It was of more importance,

to secure some indication as to whether or not the

recognition of the first unpleasant word seemed to have

-

77

-

markedly influenced the identification of the second ,1
Since such effects, if consistently present,
distort the data in half of the cases,

could markedly

some appraisal of

possible differences attributable to this factor was thought
prudent,

especially in the hospital

sample.

With this in

mind, the data was again classified in two subgroupings words presented in the first series,

and words presented in

the second.
Familiar!ty wlth the word.

The effect of stimulus

familiarity has long been recognized in studies concerned
with learning and discrimination, and it was hardly surpris­
ing that criticism of certain of the experiments previously
cited here was focused on the lack or inadequacy of controls
over this variable.

The attempt having been made to exert

as rigorous control of frequency in the selection of the
stimulus material as seemed feasible, efforts were then
turned to measuring effects attributable to controlled
statistical manipulation of this factor, the consideration
basic to the fourth subsidiary hypothesis,

lit was anticipated that the words most likely to be
fixed and thus later to be recalled were the unpleasant words
rather than the controls, because of the subject's own
participation in their selection and possibly because of
their stronger emotional connotation.
This proved to be
the case.

-78In order to do this conveniently within the framework
of the experimental design, all of the words, unpleasant and
control, were divided into two groups on the basis of the
criterion of frequency used throughout - the Thorndike-Lorge
ratings (27).

These ratings are In numerical foim and, for

the first and more common approximately 20,000 words,
range from l|-9 to 1.

1'ach number represents

they

the number of

timer words v:i th such ratings have tee;, observed to occur
per million words in a variety of printed matter.

In

addition to the numerical ratings, the symbols "A” and "AA"
are also employed, representing frequencies of observation
of from 50 to 99* and 1°° o r more respectively, per million
words.
The Thorndike-Lorge frequency ratings f or all of the
unpleasant words (and consequently for their matched control
words) were recorded and tallied for each sample s e p a r a t e ^ .
In this way it was possible by Inspection to find convenient
cuttinr scores (frequency ratings) between rough groupings
which had thus occurred as a function of the words selected
by the subjects.

For the college subjects,

words occurring

six or more times per million, as reflected in their frequency
ratings, were designated as "common” words, while those appearinr w i t h a frequency of five or fewer per million were termed
"uncommon."

For the hospital

subjects, a frequency of

seventeen times per million was adopted as the cutting score,

-79with words above that frequency considered as common,

and

words below as uncommon.
Wo rd length *

The fifth subsidiary hypothesis presented

involved the variable of wor d length,

including the a s s u m p t ­

ion that duration thresholds for long words would be greater
than for short words.

While

s\ich an assumption required no

astute prophetic capacities and in fact appeared quite
logical, it nevertheless

seemed appropriate to include the

variable concerned as one of those to be examined alone and
in interaction with

the variable

of primary interest.

In order to examine the effect of word length most
conveniently and investigate whether it might contribute a
differential

tendency In response to motivational factors,

it was desirable once again to split the samples Into two
sub-groups - In this case, long words and short words.
Accordingly,

the number of letters In each of the unpleasant

words (and thus also the control words equated In this
respect) was recorded and tallied w i t h others o f equal size.
Each sample was arbitrarily divided into groups of approx­
imately equal

size - "long" words and "snort” words.

same cutting line was used with both samples;
more letters were designated as long,
fewer letters were termed short.

The

words of 6 or

while those of 5 or

-50Sex d l f f erences.

No specific hypotheses h ad been made

In regard to the effect of sex differences on the verbally
reported responses to the unpleasant
experimental situation.

words w i t h i n the

M an y of the words used carried

colloquial connotations of sexual character in addition to
their dictionary definitions.

Therefore,

h°lf of the college subjects were female,

since approximately
in lieht of the

crLticism made by Howes and Solomon (>) of M c G i n n i e s 1 p r o ­
cedure (iJ-i) using sexually loaded stimulus words with male
subjects in the presence of a female assistant,

it w a s of

some interest to investigate the possibility that experimental
results might have been influenced by the reactions of female
subjects in the presence of a male experimenter.

With this

in mind, comparisons were made of the variances in the d ur­
ation thresholds of the male and female subjects of the
college group, in interaction with the primary variable of
the study.
Grouping the p r e -recogni tion hypothe ses.
recognition hypotheses,
iables were examined:

V.'ith the pre-

the effects of the following var­
categorization of the hypotheses,

unpleasantness of the stimulus words prompting the hypotheses,
and the intensity with which the unpleasant words were dis­
liked by members of peer groups.

In each case the pr e ­

recognition responses were grouped on the basis

of the stimulus

words, and the criteria for these groupings were similar to

-

01

-

thoss employed in the treatment of the duration threshold
data.
TABLE II

FREQUENCIES IK THE ARBITRARY GROUPINGS
OF UNPLEASANT STIMULUS V.ORDS

Hospital

College

Strongly disliked

33

ks

Mildly disliked

27

21

Strongly taboc

22

26

Mildly taboo

38

1*0

Series presented first

30

33

Series presented second

30

33

Least liked words

30

33

Next least liked words

30

33

Common words

23

13

Uncommon words

kl

53

Lon.,’; words

32

29

Short words

32

37

Categories

Male

32

Female

3k

-

82-

The Analyses of Variance
With each classification of the various samples into
sub-groupings n ow achieved, it was possible to begin the
analyses proper.

The following procedure was observed.

Four separate designs were developed for each sample, with
the operations essentially parallel for the two samples
with the following exceptions:
(1)

No comparison

on the basis of sex differences

could be made for the all-male hospital group.
(2)

The variables
the individual

of relative degree of dislike b y
subject for the two unpleasant

stimuli and that of order of presentation were
treated together in the college sample but
separately in the hospital group.

(Only in the

hospital sample did there appear reason to suspect
that order of presentation might be a significant
influence on the duration thresholds.)
(3) For the college group,

the number of threshold

measures employed throughout the analysis remained
constant (an N of 33 subjects, permitting 330
threshold determinations of which 66 were for
unpleasant words and the remainder for controls.
With the hospital group, the thresholds of all of
the 32 subjects were employed in three of the four
designs, but it was necessary to omit those of two

-83of the subjects from that design involving the
individual designations of least-liked and next
least-liked unpleasant words.

Since inspection

of the data of the two individuals thus omitted
suggested that neither deviated markedly from
their peer groups in respect to the variables of
chief interest,

and especially in virtue of the

relatively homogeneous nature of the hospital
sample implied by the results of the tests for
homogeneity,

this omission seemed to provide

little cause for concern.
The designs having been developed,
formulae were

the following

applied:

2
Total sum of squares:

X X ^ - (SX)

Sum of squares between groups:

(IX!)2
nl

♦

(SX2 > 2

OfcXn )2

_ (XX)2

n2

with n equivalent to the number of observations in each sub­
group represented by the subscript numbers.

Interaction

sums of squares were obtained by subtracting the sums of
squares already calculated for the variables concerned
from the total between groups stun of squares of the groups
involved in the interaction:
r)
J

-

81*.-

v.ith the sums of squares of variables represented b y the
alphabetical designations.
The sum of squares within groups was obtained by
consolidating all of the data from those designs involving
the same numbers of observations from the same samples,
subtracting the sums of squares between groups so obtained
from the total sums of squares,
the mean

F w a s computed by dividing

squares between groups by t he mean squares within

group s.
Sums of squares were computed for the following group­
ings studied in the duration thresholds data:
(1)

Unpleasant vs. control words.

(2)

The

series in which the unpleasant words were

categorized as strongly cislite d by the peer
group vs, the
designated

(3)

series with the unpleasant words

as m i l d l y

di sliked.

The series with the unpleasant words classified
as strongly taboo vs. those series with the
unpleasant words termed mildly taboo.

(q.)

The series of words presented first vs.

the series

of words presented second.
(5)

The series in which the unpleasant words had been
rated as least liked b" the subject vs.

those in

which the unpleasant word was rated next least
liked.

-85(6 )

Common va. u n c omm on words#

(7)

Short va. long words.

(8 )

Responses of male vs.

those of female subjects.

Interactions were obtained for the first grouping with each
of the others,

since these were the relationships of primary

interest.
The following groupings wer e used for computing

sums

of squares with pre-recognition hypotheses;
(1)

The four categories of pre-recognition hypotheses;
structurally similar, structurally unlike,

nonsense

and part responses.
(2)

Hypotheses made to unpleasant vs.

those made to

control words.
(3)

Hypotheses made to those series of w o r d s in which
the unpleasant word was designated as strongly
disliked vs. those in w h i c h the unpleasant word
was classified as m i l d l y disliked.

Interactions were obtained b e t w e e n the unpleasant vs. control
word grouping and each of the other two.

RESULTS
Presentation of the result s.

The results to be

presented here are also summarized in Tables III,
VI, and VII, on pages 98 to 103.

In addition,

IV, V,

for purposes

of comparison, the results of the analyses for both the
college and hospital samples are given in abbreviated form
in Table VIII, page 1 OJLp.

Those results considered to be of

particular significance for the study are included in
greater detail in Tables IX, X, and XI, found on pages 106,
107 and 108.

The means presented below and in the tables

are means of the duration thresholds and frequencies of p r e ­
recognition hypotheses after the application of the log­
arithmic transformation described earlier.
The Hospital Sample
Durati on thresholds in terns of number of exposure s to
the point of word reco; n i t i o n .
and with 300 thresnold values

Wit h one degree of freedom
(thresholds for each of ten

test words exposed to 30 subjects)
data,

shown in Table III,

in one treatment of the

and with 320 threshold values in

the other, presented in Table IV, unpleasant and control
-

86-

A

-

87-

words are indicated as not significantly different.^-

With

means of 83.2 and 86.8 for the unpleasant and control words
in the first treatment,

and 83.6 and 88.0 in the second,

P*s of less than 1 were found in both instances.
The following data are

summarized in Table III and were

gathered from 300 threshold determinations.

No significant

differences were demonstrated between the series of words
(one unpleasant and four control words matched for length
and frequency) in which the unpleasant word had been class­
ified as strongly disliked by the peer group and the

series

in which the unpleasant word had b een considered mildly d i s ­
liked.

With means of 86.3 and 8^.8 respectively,

one degree of freedom,

and with

the *r' found was less than 1.

The

groupings here did not provide controls for the effects of
possible differences in word length or frequency;

this was

however accomplished later through the calculation of the
interaction with the unpleasant-control dichotomy.
In comparing the series with the unpleasant word
designated as strongly taboo with the

series with the u n ­

pleasant word ce si-mated, mildly taboo, the thresholds for

iThe two separate treatments indicated here were neces­
sitated because two of the thirty-tv.o hospital patients were
unable to specify as least liked and next least liked the two
words used as unpleasant stimuli.
Consequently, it was
necessary to omit tneir responses from part of the analysis.
This resulted in one analysis with an N of 320 (thresholds
for ten words with each of 32 individuals), and another with
an N of 30^ (ten words with each of 30 subjects).

-88the former were found longer.
and 82.5-

The means here were 92.2

V.'ith one degree of freedom, the F of 5.381 was

significant at between the .05 and J01 levels of confidence.
In this form again,

the comparison does not take into

account differences of word length or frequency.
This is also the case with the differences found
between the series of words presented first compared with
the series of words presented second, where the means were
90.7 and 81.if.

In this instance, with the same number of

degrees of freedom as the previous analysis,

the obtained

P of 5.312 is significant at between the .05 and .01 levels
of confidence.
Again with one degree of freedom, the comparison between
series in which the unpleasant word was

selected as least

liked by the subject and series in which the unpleasant word

7/as selected as next least liked yields no indication of
significant differences.

With means of 81f. 6 and 87.5

respectively, the F was less than 1.

here, as well,

since

the unpleasant and control groups were not separated and
compared, no controls over word length or frequency were
being exerted.
In Table IV are summarized the data used In the compar­
ison of common and uncommon v/ords, and short and lone words,
drawn from 320 threshold values and making use of one degree
of freedom in each case.

The means for the common and u n ­

common words were 77.8 and 92.3 respectively, while the means

-89for short and long words were 78.5 and 95*7 •

Very

significant differences were obtained for b o t h groupings,
with P's of 13.5ij-3 and 20.670 respectively.

Unpleasant

and control words were combined for both of these analyses.
Interactions.

The following results are summarized in

Table III, with the N of 300.

Investigation of the inter­

action between word unpleasantness and the degree to which
the unpleasant word of the series was Indicated as disliked
by the peer group revealed that this effect was not signif­
icant.

With one degree of freedom, the P computed was 1.772.

The Interaction between word unpleasantness and the degree
to which the unpleasant word of the series had been desig­
nated as socially taboo,

with a similar number of degrees

of freedom, was also not significant, with an P of less than
1.

With one degree of freedom, no significant interactions

were found between v.orc unpleasantness and the order of
presentation of the series,

and between word unpleasantness

and the degree to which the unpleasant word of the series
had been rated as disliked by the individual subject.

F's

in both instances were less than 1 ,
Table IV reveals that w i t h an N of 320 and

one degree

of freedom, interactions between word unpleasantness

and

frequency ("coramoness"), and between word unpleasantness
and length, were not significant.
less than 1 respectively.

P's here were 2.101 and

-90Pre-recognition hypotheses w ith the hospital p a t l e n t s .
For the results of this analysis,

lp5l responses were recorded;

the data are summarized in Table V.
freedom,

W i t h three degrees of

the F of 5.708 obtained between the frequencies in

four categories of pre-recognition hypotheses was v e r y
significant.
similar,
and part,

The m e a n s were as follows:

3^. 0 ; structurally unlike,
36.If.

structurally

1 7 .if; nonsense, 33 *7 ;

This grouping does not differentiate

between the hypotheses made to unpleasant

and control words.

Y.hen these latter two groupings of pre-recognition responses
are compared, with one degree of freedom,
less than 1 and thus not

significant.

were 31 .b. and 33.5 respectively.

the obtained F is

The means obtained

The comparison of h y p o t h ­

eses to words of series in which the unpleasant word had
been categorized as strongly disliked on the b a s i s of peer
choices, with

those to words of

series in which the u n ­

pleasant word had b een classified as m i l d l y disliked
no significant differences.
35-6 respectively,

showed

With means here of 31.2 and

the F, with one degree of freedom, was

2.576.
Interactio ns.

In the investigation of interaction

between w o r d unpleasantness and the categorization of prerecognition hypotheses into the four types of responses,
very

significant differences were revealed with controls for

length and frequency differences here in effect.

With three

-91degrees of freedom,

the obtained F was 56,828.

erence to Table IX discloses,
between the unpleasant
uted to the

the significant

As r e f ­
differences

and the control group m ay b e a t t r i b ­

structurally unlike group,

in which m ore

responses were recorded to the unpleasart

words.

The

basis

for the significance o f the i nteraction itself appears to
reside in the r eversal of trend in the

structurally unlike

and nonsense groups as c o m p a r e d with that e x h i bited b y the
structurally similar and part groups.

I n t h i s regard, wh ile

the structurally similar and part h y p o t h e s e s tend
in great er num ber in response
pleasant,

to control w o r d s

to appear

than to u n ­

structu rally unlike and nonsense hypotheses tend

to appear more profusely in response to unpleasant than to
control words.
Retur ning to Table V, the interaction b e t w e e n stimulus
word unpleasantness and the degree to which

the unpleasant

stimulus word of the series was indicated as disliked by a
peer group was insignificant
hypothesis frequencies.

in respect to pre-rec ognition

Vd.th one degree of freedom, the

obtained F was 1.6914-.
The College Sample
Duration th r e s h o l d s .

The following results,

presented

in Table VI,

were obtained f r o m an N of 330 duration t h r e s ­

hold values,

with one degree of f r eedom applicable to the

-92interpretation of each P value obtained.

ittien the duration

thresholds for the unpleasant words were compared to those
recorded for their matched control words, a n F value of
less than 1 and thus insignificant differences were disclosed.
The means were 98.5 and 96.14- respectively.
Those series of five words (one unpleasant and four con­
trols) in which the unpleasant words had been designated as
strongly disliked on the basis of peer selection were not
perceived differently in respect to threshold than were
those

series in which the unpleasant words had been desig­

nated as mildly disliked.
and 100.7 respectively.

The means

obtained here were 95>.0

Since this preliminary grouping did

not distinguish between thresholds obtained for unpleasant
and for control words, the results here, an F of 3*000 which
was not significant, were o f little particular consequence
per se.
An F of 2.366, also not significant, was provided by
the analysis of the series of words in which the unpleasant
words had been classified as strongly taboo, as opposed to
the series in which the unpleasant words had been classified
as mildly taboo.
respectively.

The means obtained here were 99*7 and 9^*9

This too was a computation preliminary to the

interaction with

the primary variable and was of minor

pertinence by itself.
In another preliminary calculation,

the series of words

In which the unpleasant words had b e e n selected as least liked

-93of those with w nich the individual

subject h ad been confront­

ed, were identified at thresholds n o t different from those
observed with series of words in w h i c h the unpleasant
had been selected as next least liked.

words

With means of 98.0

and 9 5 . 6 , the F here w as less t h a n 1 and the refore not
icant.

The

signif­

series with the least liked words were also the

series which had been presented first to the students.
When the thresholds for common words were compared with
those for uncommon words, unpleasant and controls combined,
a very significant difference with an F o f 7.201 was d i s ­
closed, w it h means of 88.5 and 9 8 .8 .

This grouping diet not

control for possible effects of length differences.
uncommon words,

as had been anticipated, were

The

correctly

identified at higher thresholds.
Whe n the thresholds for snort and l ong w o rds were
compared, a non-significant F of 3.29^1- was obtained.
grouping did not control for possible
differences.

This

effects of frequency

The means were 9^*3 and 99*9.

Very significant differences were found between thres­
holds for words, unpleasant

and control combined,

a d min­

istered to male students and those administered to female
students.

Tnis grouping did not control for differences in

both word length or frequency.
respectively.

The means were 92.9 and 1 0 0 .9

The F obtained was 6 .8 l 6 .

-9^Interactlons.
word unpleasantness,

The Interaction of the primary variable,
with that of peer group dislike of the

unpleasant word of the series was significant at between the
.05 and .01 levels of confidence with an F of 5*l89»

The

sis-nific«n ce of the interaction is attributed to the two
groupings of unpleasant words (strongly disliked vs. mildly
disliked) which were perceived at appreciably differing
thresholds not manifested by their respective controls.
£ords classified as strongly disliked were perceived correct­
ly by the students as a group at lower thresholds t ha n were
those words designated mildly disliked.
in Table X.

This is demonstrated

While tne mildly disliked words appeared to be

perceived significantly more slowly than their controls as
well, the differences between the strongly disliked words
and their controls did not seem significant.
Returning to Table VI, one may observe that the inter­
action between word unpleasantness and the degree to which
the unpleasant word of the series had been rated socially
taboo was not significant, with an F of 2.003.

The inter­

action of word unpleasantness and Individual choice of the
unpleasant word of the

series as least liked as opposed to

the next least liked of the word list was also not signif­
icant, with an F of less than 1.

As was indicated before,

the least liked word series were also those administered
first

.

i

-95I n t e r a c t i o n of word unpleasantness w i t h word frequency
or "commonness” was significant
levels of confidence,

at between the

w i t h an P of 5.018.

.0£ and

.01

As indicated in

Table XI, while the students did not react different ly to
the control common and uncommon w o rds respectively,
did respond differently to the unpleasant
to the unpleasant u n c o m m o n words.

they

common as compared

The common unpleasant

words were per ceived at lower thresholds than were the u n ­
common unpleasant words.
erence to the table,

Further,

it was not

as may be seen by ref­

the u n c o m m o n words which

were the most prominently deviant;
erences between the unpleasant

the

significant

d i ff­

and the control w o r d s lie in

the common words only.
A gain returning

to Table VI,

one may discern that no

significant degree of interaction was d isclosed between the
variables of w o r d unpleasantness and wor d lepgth,
between word unpleasantness and the

nor

sex of the subject, with

F's of 2.666 a n d less than 1 respectively.
Pre-recognition hypothe se s wi th the colle ge student s .
The following data are summarized in Table VII and were drawn
from an N of 630 responses.

Vihen the frequencies of pre-

recognition hynotheses falline- into the four categories were
ru’, jected to analysis,
them were indicated.
urally similar,

very significant differences between
The m e ans were as follows:

struct­

1^5 .1 ; structurally unlike, 2 2 .ip; nonsense,

-

38.8;

and part,

31*3.

96-

With three degrees of freedom here, the

obtained P of 9*814-5 was very significant.
hospital subjects,

As with the

the grouping involved here in no way

discrlminates between unpleasant and control words,

and

consequently it is in itself of little meaning to the study.
When the number o f hypotheses in response to unpleasant
words is compared to that of those prompted b y t he control
words, with one degree of freedom,

the P of less than 1

indicates no significant differences.

The means were 38.1

and 3 5 •8 respectively.
When the number of hypotheses generated in response to
series of words with the unpleasant w o r d strongly disliked
by the peer group is compared to the n u mber of responses
prompted by series in which the unpleasant word was mildly
disliked, the differences are significant at b e t w e e n the
.05 and

.01 levels of confidence.

The means were 3^4-.0 and

1*1.0 respectively.

Ylith one degree of freedom,

5.138*

since this comparison does not involve

Once again,

any control over frequency or word length,
such to the

the F was

its importance as

study is minor.

Interactions.

When the interaction of w o r d unpleasant­

ness with the categorization of the pre-recognition hypotheses
is examined, with an v of 1.801 and three degrees of freedom,
the results are not significant.

The interaction of word

unpleasantness and the degree of peer group dislike for the

-97unpleasant word of the series with relation to the resulting
frequencies of pre-recognition responses also indicates no
significant effects of one variable upon the other.
one degree of freedom,

the F obtained was 2 .If30 .

With

-9 8 table

III
SUMMARY OF ANALYSIS OF VARIANCE OF THE NUMBER
OF EXPOSURES TO THE POINT OF WORD RECOGNITION
BY THIRTY HOSPITAL PATIENTS

Source of Variance

d. f.

Total

Mean
Square

F

P

299

Unpleasant vs. control words

1

6o6

< 1

N.

Series with unpl. word strongly
disliked by peer group vs.
series w3 th unpl. word mildly
di silked

1

26

< 1

N. slg.

Series v;ith unpl. word strongly
taboo vs. those w i t h unpl.
word mildly taboo

1

6,532

5.381

.05

Order of presentation (first
series presented vs. second
serie s )

1

6,449

5.312

.05

Series with u n p l . word least
liked vs. series with unpl.
word next least liked

1

613

< 1

N. sig.

Unpleasant vs. control words x
strongly disliked vs. mildly
disliked words

1

2,151

Unpleasant vs. control words x
strongly taboo vs. mildly taboo

1

608

< 1

N.

Unpleasant vs. control words x
order of presentation

1

129

< 1

N. sig.

Unpleasant vs. control words x
least liked vs. next least
liked words

1

529

< 1

N.

H

.
O

1

H

.oi

sig.

Interact!ons

E rro r

290

1 ,2 1 ^

1.772 N. sig.

sig.

sig.

-9 9 -

TAKLE IV
SUMMARY OF ANALYSIS OF VARIANCE OF T H E NUMBER OF
EXPOSURES TO THE POINT OF YvORD RECOGNITION
EY T IiIP T Y -T V .'0 hOS 'I T A L PATIENTS

Source of Variance
Total

d. f.

Me an
square

F

319

Unpleasant vr.

1

992 < 1

N. sig.

Common vs. uncommon words

1

l5Jt39 13.5^1-3

V. sig.

Short vs.

1

23»$61j- 20.6?0

V . sig.

long

control words

words
Interactions

Unpleasant vs. control words x
common vs. uncommon words

1

Unpleasant vs. control words x
short vs. long words

1

Error

31 I4-

2,395 2.101 N.
273 < 1

l,ll4-0

sig.

N. sig.

-100TABLE V SUMMARY OF ANALYSIS OF VARIANCE OF THE NUMBER OF
PRE-RECOGNITION HYPOTHESES MADE EY HOSPITAL PATIENTS

Source of Variance

d . f.

Total

kSo

Mean
Square

F

J4., 8i |_6

5 .7 0 3

V. sig.

< 1

N. sig.

P

Categories of hypotheses

3

Unpleasant words vs. control
words

1

293

Series with u n p l . words
strongly disliked by peer
group vs. series with unpl,»
word mildly disliked

1

2 ,1 8 7

2 .5 7 6

N. sig.

V. sig.

Interactions
Unpleasant vs. control words
x categories of hypotheses

3

If8,2ip7

56.828

Unpleasant vs. control words
x strongly disliked vs.
mildly disliked words

1

1 ,4 3 8

1 .6 9 4

Error

VA

849

K.

sig.

-

101-

TABLE VI
SUMMARY OP THE A NALYSIS OP VARIANCE O P THE N U M B E R
OF EXPOSURE'S TO THE POINT 0^ V.ORD RECOGNITION
BY COLLEGE STUDENTS

Source of Variance

d. f.

Total

329

Mean
Square

2I4.O

F

p

Unpleasant vs, control words

1

Series with unpl, word strongly
disliked by peer group vs.
series with unpl, word mildly
di sliked

1

2,298 3.000

N. sig.

Series with unpl, word strongly
taboo vs. those w i th unpl.
word mildly taboo

1

1,812 2.366

N.

sig.

Series with unpl. wore least
liked vs. series wit h unpl.
word next least liked (con­
founded with order of
presentation)

1

K.

sig.

Common vs. uncommon words

1

5,5l6 7.201

V. sig.

Short vs. long words

1

2,523 3.29)^

X't

Male vs. female

1

5,221 6.816

V. sig.

Unpl. vs. control x strongly
disliked vs. mildly disliked

1

3,975 5.189

.05 -. 01

Unpl. vs. control x strongly
taboo vs. mildly taboo

1

i,53it 2.003

N. sig.

Unpl. vs. control x least liked
vs next least liked (con­
founded with order of
presentation)

1

1+70

< 1

< 1

N. sig.

•

sig.

Interactions

572

< 1

?
itT•

sig.

-102TABLE VI

continued

Source of Variance

d. f.

Mean
Squ are

U n p l . vs. control x common
v r . uncom m o n : rds

1

Unpl • vs. control x short vs.
Ion words

1

2,01+2

U n p l . vc. control words x male
vs. ferial0

1

31

316

766

Error

P

P

$.018

.05-.01

2.666

N. sig,

< 1

N . si 'T

-103TABLE VII

SUMMARY OF T H E ANALYSIS OF VARIANCE OF T H E NUMBER
OF PRE-RECOGNITION HYPOTHESES
OF COLLEGE STUDENTS

Source of Variance

d. f.

Total

629

Mean
Square

3

13,064

Unpleasant vs. control words

1

522

Series with unpl. word strongly
disliked by peer group vs.
series with unpl. word mildly
di sliked

1

Unpl. vs. control x categories
of hypotheses
Unpl. vs. control x strongly
disliked vs. mildly disliked

P

< 1

N. sig.

6,816

5.138

3

2,390

1.801

N. sig.

l

3,229

2.430

N. sig.

620

1,327

I

V. sig.

1

9.81*5

H
0
.
r
O•

Categories of hypotheses

F

Interactions

Error

-10I4-TAP.LE VIII
A COMPARISON SUI/WAFT OF ANALYSES OF VARIANCE
WITH THE COLLEGE AND HOSPITAL SAMPLES

Source of Variance

Ho spital
F
p

College
F
p

Number of exposures to the
point of recognition:
Unpleasant vs. control words < 1

N.

< 1
Series with unpleasant word
strongly disliked by peer
group vs. series with
unpl. word mildLy disliked
Series with unpleasant word
strongly taboo vs. series
with unpl. word mildly
taboo
Order of presentation (first
series vs. second series)

sig.

< 1

N. sig.

N. sig.

3.000

N. sig.

5.381

.05-.01

2.366

N.

5.312

.05-.01

sig.

K. sig.
N. sig.

Series with unpl. word least < 1
liked vs. series with
unpl. wore: next least liked

N . sig.

< 1'

Common vs. uncommon words

13.5ii3

V . sit*.

7.201

V.

Short vs. long words

20.670

V . si g .

3.291+

N. si g.

6.816

V. si g.

Male vs. female

SiR.

1.772

N . sig.

5.189

Unpleasant vs. control x
strongly taboo vs.
mildly taboo

< 1

N . si g .

2.003

N. sig.

Unpleasant vs. control x
order of presentation

< 1

N. sig.

H

Unpleasant vs. control x
strongly disliked vs.
mildly disliked

•

0
1
.
0

Interactions:

N. si g •
< x 1

-105-

TARLE

vni

Source of Variance

continued

Hospital
F
P

Colie ge
F
p

Unpleasant vs. control x
least liked vs. next
least liked

< 1

N . sig.

Unpleasant vs. control x
common vs. uncommon

2.101

N . sig.

5.016

.06-.01

Unpleasant vs. control x
short vs. long

< 1

>• s ig .

2.660

N . sig.

< 1

N. sig.

Unpleasant vs. control x
male vs. female

N. sig.

Number of pre-reccgnition
hypotheses:
Categories of hypotheses

5.706

V. sig.

9.824-5

V. sig.

Unpleasant vs. control
words

< 1

iv• si £ .

< 1

N . si g .

Series with unpl. word
strongly disliked by
peer group vs. series
with unpl. word mildly
dislik ed

2.576

N . si g.

5.133

.05-.01

56.826

V. sir:.

1.601

sig.

1.695

Is• sr r.

2.530

N. sir.

Interactions:
Unpl. vs. control x
categories of prere cog. hypotheses
Unpl. vs. control x
strongly disliked vs.
nildlv disliked

d

-106table

IX

INTERACTION OF UNPLEASANT VS.
CATEGORIES OF PATIENTS*
PRE-RECOGNITION HYPOTHESES

CONTROL x

Mean Lor. of Number of Pre-recognition Hypotheses
Categories

Unpleasant

Control

Total

Structurally
similar

30.9

314-7

3-V.O

Structural ly
unlike

23.1

16.1

17.it

Nonsense

37. b-

32.6

33.7

Part

31.1|

37. 14.

36.14

31.U

33.1)6

Totals

The differences here between the unpleasant and control
words lie with respect to the structurally unlike group.

No

simple relationship between SS, SU, N, and P categories with
respect to the unpleasant and control words is suggested.
The reversal of the tendency in the structurally unlike and
nonsense groups as compared with that exhibited by the struct­
urally similar and part groups is the basis for the
interaction significance.

A

-107TABLE X INTERACTION OF UNPLEASANT VS. CONTROL x
STRONGLY DISLIKED VS. WILDLY DISLIKED
WITH STUDENTS

Mean Log. of Number of Exposures to Point of Recognition
Categorie s

Strongly Disliked

Mildly Disliked

Unpleasant

92.0

112.5

Control

95.5

97.7

Total

95.0

100.7

Tot al
98.5
96. Ij.

The significance of this interaction can be traced to
the unpleasant words.

In relationship to the control words,

the unpleasant words demonstrate opposing tendencies in
respect to their classification as strongly disliked by a
peer group or as mildly disliked.
demonstrated by the controls.

This difference is not

- 108 TABLE XI

INTERACTION OP UNPLEASANT VS.
COMMON VS. UNCOMMON,
WITH STUDENTS

CONTROL x

Mean Log. of Number of Exposures to Point of Recognition
Categories

Unpleasant

Common
Uncommon
Total

Control

Total

76.5

91.6

88.$

103.9

97.5

98.8

98.5

96.1^

This interaction is significant because while the
students did not react to the control common and uncommon
words differently, they did react differently to the un­
pleasant common and uncommon vc rds respectively, as shown
in the table.

Further,

it may be seen that it is not the

uncommon words which are perceived in different fashion.
The significant differences in duration threshold between
unpleasant and control words lie in the common words only.

1)1SCUSSIOU OF THE RESULTS
The Hypotheses
Since the results of the study can perhaps be most
meaningfully examined in relation to the hypotheses orig­
inally advanced, a reconsideration of these seems
appropriate.
The primary hypothesis:

perceptual defense.

This

hypothesis was used as the focal point of the experimental
design.

It predicted that the duration thresholds (the

number of exposures required for word recognition) for
words of idiosyncratically unpleasant connotation would be
significantly higher than those for words of neutral
connotation, matched with the unpleasant words in number
of letters and frequency.
The secondary hypothe sis.

This hypothesis predicted

that there would be significant differences between the
frequencies of structurally similar, structurally unlike,
nonsense, and part responses elicited by the unpleasant
words and those elicited by the neutral words.
Subsidiary hypotheses.

Five subsidiary hypotheses were

as follows, with the differences predicted in terms of the

-109-

-110relationships between the unpleasant words and their control
words matched in word length and frequency:
(1)

’
.lords selected as strongly disliked on the basis
of composite weighted scores in terms of peer group
ratings would have greater duration thresholds
than those selected as mildly disliked on the
basis of the same criteria.

(2)

There would be significant differences between
the duration thresholds for unpleasant words
selected as strongly socially taboo on the basis
of prediction ratings by clinical psychologists,
and those for unpleasant words selected as mildly
or not taboo, the thresholds being greater with
the former.

(3)

Unpleasant words rated as least liked by individ­
ual subjects would have greater duration thresholds
for them than would unpleasant words rated as next
least liked.

(Ip)

There would be significant differences between the
duration thresholds for uncommon unpleasant words
and those for common unpleasant words, with the
former being higher.

On the basis of this hypoth­

esis, it was predicted that the differences in
threshold between unpleasant and control words
would be greater with the uncommon than with the

A

-

111

-

common words,
(5)

There would be significant differences between
the duration thresholds for long unpleasant
words and those for short unpleasant words, with
the former being higher.

The prediction made

here was that the differences in threshold
between unpleasant and control words would be
greater with long than with short words.
Interpretations and Conclusions
Perceptual defense.

The primary hypothesis involving

the concept of perceptual defense was not supported by the
experimental results from either the hospital or college
sample.

In each case,

the differences obtained between

thresholds to unpleasant and to control words were statis­
tically insignificant.

This is interpreted as indicating

that neither sample of subjects responded differently to
the total group of idiosyncratically unpleasant words than
they did to the ostensibly neutral words, in terms of the
number of exposures required for correct word identification.
These findings taken by themselves lend support to the
contention of Howes and Solomon (5) that the influence of
affective determinants on duration thresholds is minimal.
Categorics of 2 re-recognition hypotheses.

With respect

to the secondary hypothesis involving differences in the

-112frequencies of the four categories of pre-recognition
responses, the hypothesis was supported by the responses
of the hospital group, but was not supported by those of
the college students.

This may be concluded through ref­

erence Lo the interaction obtained between word
unpleasantness and categories of responses (Table VIII,
Results).

The frequencies of responses falling into the

four categories were significantly different
groups,

for both

but only in the hospital sample did differentiat­

ing these categories in terms of unpleasant and neutral
words produce any significant changes in the frequencies.
In this case, as may be seen in Table IX, there were
significantly more structurally unlike hypotheses In
response to the unpleasant words than to the neutral words.
In addition,

the reversal of trends responsible for the

significant interaction term can be attributed to the
greater number of structurally unlike and nonsense responses
produced after exposure of the unpleasant words, and the
greater number of structurally similar and part responses
produced after the neutral words had been exposed.
The fact that the hypotnesis was upheld for one sample
but not for the other suggests that attributes peculiar to
the samples or to the words chosen for each were involved.
This demands tne.t care must be used in generalizing from the
results obtained.

However, the fact remains that within

-

113

-

the hospital g r o u p , *.n spite of insignificant deviations in
terns of deration thresholds, there apparently were diff­
erences in the way in which unpleasant words were perceived
prior to the actual recognition from the way in which the
neutral words were perceived.

This in turn intimates that

the individuals manifesting such behavior were particularly
sensitive to the connotations of the words selected by them
as especially disliked, as if set to avoid verbal corrob­
oration of subliminal pre-recognition hypotheses carrying
such unpleasant meanings.

One may thus suggest as did

McOinnies in explanation of similar results (iIf), that the
relative profusion of unlike and nonsense responses to the
critical words compared to those elicited by neutral words
may represent a type of defensive reaction or blocking.
A further comment in this connection is stimulated by
the fact that the differences in category frequencies, when
these are examined in interaction with unpleasant as opposed
to control words, were significant in the hospital group,
while there were no significant differences in the total
numbers of pre-recognition hypotheses to unpleasant as com­
pared to control words.

This implies that in few cases was

there any total blocking of responses in reaction to the
unpleasant words, but rather,

a substitution o f nonsense or

structurally unlike responses fo r a portion o f the part and
structurally similar responses made to the neutral words.

-lliiThis In turn suggests the possibility that had McSinnies
and Sherman (15) categorized their pre-recognition responses
as did Mclinnies in his earlier experiment (lip)* they might
have found air Terences in respect to the hypotheses generated
by taboo as opposed to neutral words, differences not evident
from the mere comparison of proportions of hypotheses made
to the two classes of stimulus words.
Peer group d i sllke .

The first subsidiary hypothesis,

that words designated as strongly disliked on the basis of
peer group ratings would have higher thresholds than words
classified as mildly disliked, appears not only not to be
substantiated by the results secured with the college
students, but to be contradicted by these data.

The inter­

action term computed between word unpleasantness and the
degree of peer group disliking (Table VI) shows differences
significant at between the .05 and

.01 levels of confidence,

in the direction opposite to that predicted, when the
unpleasant words are considered apart fran their controls.
The more detailed presentation of the results in Table X
reveals that the strongly disliked words were correctly
p e r c e i ved

sooiv-r than those selected as mildly disliked.

This may in part have been a function of t he frequency of
the words so designated.

Nevertheless, the thresholds of the

strongly disliked words and their matched controls combined,
compared with those of li e mildly disliked wo rds with their

-115controls, were not significantly different.

The thresholds

for the unpleasant and controls words were likewise not
si? dficantly different.

The fact that

significant rever­

sal 5 of trend were demonstrated by the strongly disliked
vvor.-s, In respect to thresholds for unpleasant and control
wor is, a: compared to the mildly disliked words,

seems to

indicate that the strongly disliked words were really
perceived differently from those mildly disliked.

However,

it Is no!. She T o m e r which appear to be the center of the
deviation, but it is the mildly disliked words.

These

ucviated prominently, not only from the stronrly disliked
words, but also from their own controls and those of the
ftica<rly disliked words as well.
This may involve idiosyncratic connotations of the
ni'dly disliked words which were not involved in the ratings
*y the peer group, or the latter may simply not have been
fi -eliable indi cation of actual intensity of disliking. At
an/ rate, the results suggest that the concept of perceptual
de 'ense may have application to responses to certain un­
pl-meant stimuli made by certain individuals.

It is

difficult to explain away the fact that, in spite of careful
controls Cor variations which mizht otherwise have been
attributed to discrepancies in lensth and frequency between
the critics! and control words, there remained differences
statistically significant between the mildly disliked words

-n6and their matched controls.
y.irth'3r-nore, the differences obtained cannot be ex­
plained by reference to the accentuation of a n effect which
in itself wa.<

sufficiently potent to provoke significant

threshold differences.

In the case of the mildly disliked

words, it was apparently only in interaction that the two
variables involved were
significant effects with

sufficiently powerful to produce
a portion of the stimuli, and only

with those administered to one of the two samples.

This In

turn, however, suggests that the effect is by no means such
a general one as that of word frequency; with both samples,
threshold differences presumably due to the effects of
frequency were very significant.
Mcf-innies1 earlier experiment using socially taboo words
was criticized by Hov.es and Solomon (5) on the grounds that
no controls for differences in word frequency were employed,
and that the s:* gni r’eance of the results might be more
parsimoniously attributed to these frequency differences than
to any effects of tabooness per se.

The later experiment by

Hcfrinni es end Hho r nan (l£) did employ controls for frequency,
yet results significant at the .01 level were reported,
differentiat *ng between thresholds f o r words following taboo
words and those following neutral words.

In view of this and

of the careful equation of frequencies of the unpleasant and
control words used in the present experiment,

it seems

-117somewhat premature to dismiss the concept of perceptual
defense entirely, yet there certainly appear to be grounds
for doubting the generality of its application.
seems little basis, however,

There

for according to it the

importance credited to word frequency as a determinant of
dura t3 on tb re shold s .
The socially t a b o o .

The second subsidiary hypothesis

involved the prediction that words of strong socially taboo
character would, ,-d, ‘relate higher duration thresholds than
would those of relatively mild taboo nature.

This hypothesis

war not upheld by the data secured with either of the
sampler, v.dth insignificant interaction terms obtained
between word unpleasantness and degree of taboo in each case.
It is of some interest in this connection to compare
the socially taboo words employed by McGinnies in his ex ­
periment (lK)

and in that conducted by him and Sherman (15),

v/ith some of those designated as strongly taboo in this
stuay.

In the experiment by McGinnies alone,

taboo words were as follows:

raped,

the socially

belly, whore, kotex,

penis,

filth, and bitch.

above,

the following of the taboo words were employed:

bitch, penis, raped,

In the second experiment cited

and whore.

The words designated as stronrly taboo

in the present

study, It ma y be recalled, were classified o n the basis of
clinical psychologists’ predictions of how people in general

-118would react to them.

The ten words wi t h the highe st weighted

composite ratings of "tabooness" are presented here in
descending order of their numerical scores?

whore, bitch,

ass, bastard, ejaculate, nigger, prostitute, syphilis, rape,
and vomit.

The overlapping of the three selections of

taboo words is evident.

On the basis

of the data collected

from both, zhe college and hospital samples, no significant
differences were noted in the thresholds of unpleasant as
compared to control words, as a function of tabooness.

It

is suggested that while socially taboo words may serve as
adequate critical stimuli in experiments such as those of
McGinnies', their effectiveness m a y

stem fran other qual­

ities than their character as socially taboo words per se.
The difficulties experienced by this and other Inves­
tigators in securing significant evidence for perceptual
defense in respect to duration thresholds suggests that
McGInnies' results may be due to the effects of a number
of unidentified variables consistently confounded with the
critical variable in his experimental procedure.

Among

these may be uncontrolled factors I n the experimental field
affecting subjects' "set” toward the test procedure or
toward subliminal pre-recoynltion responses to critical
words, especially those after the Initial subliminal reaction
to the first critical word.

It m a y be recalled that a

definite set was Instilled i n members of both experimental

-

119-

samples usee1 in this study.

The hospital subjects were

either referred directly for psychological evaluation or were
introduced to the experiment as ’’control subjects for a
research project in connection with skin disorders.”

To

the college, subjects, the t achi stoscopic presentation was
introduced as a ’’vision test.”

In both instances, it seems

plausible that the words used as stimuli were perceived
within the framework of the introduction used, but in
neither case v/ere there evidences for perceptual defense
for the unpleasant words as a whole or for those designated
as strongly taboo.
Under the conditions of testing employed by McGinnies,
no definite set appears to have been given the college
subjects.

If subliminal resnonses to McGinnies’ stimuli

did occur as he suggests, then such responses would involve
early discrimination of the critical socially taboo words
as such, and even the initial subliminal reaction could
have instilled a set to perceive the critical word being
re-exposed end the critical words to follow, In such a
manner as to avoid corroborating the subliminal discrim­
ination at a verbal level.

Equally probable Is the

possibility that a specific set was unwittingly given to
his subjects.
Indlvidnal dlsllke.

The third subsidiary hypothesis

was concerned with the question of whether or not differences

-120in the unpleasantness of words rated numerically as least
liked and next least liked by the individual subjects would
be sufficiently pronounced and stable in this attribute to
influence significant threshold differences.

The prediction,

that the least liked words would be perceived less rapidly,
was not upheld w ith either group.

The test of this hypothesis

was confounded in the college group with the effect of possible
differences due to order of presentation,

since all of the

least liked words were exposed in the first test series shown
to these subjects.

However,

since no differences were found

in the college group, as anticipated,

it may be assumed that

there were no differences attributable to either of the
variables confounded.
Word frequency.

The fourth subsidiary hypothesis was

that there would be a significant difference in duration
thresholds for uncommon unpleasant words as compared with
common unpleasant woras,
rapicly.

with the former recognized less

Differences between uncommon and common words,

unpleasant and controls together, were
samples,

significant in both

thus apparently justifying- the importance accorded

to word frequency differences and their control in the
experimental design.

These findings were

similar to those

emphasized by Solomon and Howes (5# 26).
It v/as thus conceivable that these data might possibly
have reflected only the effects of differences in word

-121familiarity , and in such case, interaction terms with
word unpleasantness would have been insignificant.

However,

when the unpleasant and control words were treated sep­
arately by means of the interaction computation, significant
differences in threshold to common and uncommon words were
also revealed between the unpleasant and control categories
with the students but not with the patients.

Reference to

Table XI demonstrates that the interaction was significant
because the students did react differently to the unpleasant
words, common and uncommon, but not to the controls.

More­

over, it was the common v.ords which departed most signif­
icantly from the controls, being recognised more rapidly
than the uncommon words.

These same common unpleasant words

were also recognized more rapidly than the common control
words.

Thus, the hypothesis itself seems to be supported by

the data from the college sample but not from the patients.
The prediction based on the hypothesis, anticipating greater
differences in threshold between unpleasant and control words
for the uncommon rather than the common words, was, of course,
not verified.
Once again, the difference between the samples with
respect to the significance of the results obtained here
suggests reservations in the application of the conclusions
to other groups.

It is possible that one reason for the

observed difference between the samples may be attributed
to probable discrepancies in intelligence, vocabulary, and

-122reading skills of the two groups.

One might rationalize

the discrepancy bp postulating that with the patients, all
of the unpleasant words were relatively equally unfamiliar
in terms of reading material, while this was not the case
with college students accustomed to study and to reading as
a recreation.
The fact that the discrepancy between the common un­
pleasant words and their controls was significant suggests
a number of possible alternative explanations.

The first

is that there were genuine tendencies for certain common
unpleasant words to influence perceptual processes subliminally, altering the thresholds normally manifested to
neutral words.

That is, that certain of the subjects were

hypersensitive to the connotations of the critical words
chosen by them as least liked and next least liked, and
these individuals reacted to such words more rapidly or
"vigilantly,” either because of consistent idiosyncratic
response tendencies toward such words or because of specific
salience of the particular words.
A second possibility is that either the common un­
pleasant words were often encountered in sources not included
in the Thorndike-Lorge list; e.g., used in casual conversa­
tion between members of the same sex or read from restroom
walls, so that the matching of these words with their
controls was based on false premises; or, more likely, that

I

-123g e n e rail za ti on

in terms of familiarity with the words

occurs through transfer from spoken vocabulary skills to
those employed in reading.
A third possibility is that the common unpleasant
words appear so often in print and over the air that,
partly because of their repetition and partly because of
their emotional salience, certain subjects develop a set to
perceive these words when they are presented through any
sense modality.

Consequently these words tend to be more

often employed in relatively accurate pre-recognition
hypotheses, verbalized or not, to relatively ambiguous
stimuli.

Were this the case, pre-recognition hypotheses

reported might be expected to show a larger number of
structurally similar and part responses and a smaller number
of nonsense and structurally unlike responses, than those
observed with less familiar or salient unpleasant words.
This rationalization receives some degree of indirect
support from the fact that with the patients, where the
common unpleasant words were not perceived at significantly
different thresholds from the common controls (although
common and uncommon words as a whole were differentially
perceived), there were more structurally unlike and nonsense
responses and fewer similar and part responses to the un­
pleasant words.

This was not the case with the students

whose pre-recognition hypotheses were not different for

-

unpleasant

as c o m p a r e d

the u n p l e a s a n t
the c o n t r o l

icant

c o m m o n wo r d s .

differences

The a v a i l a b l e

i n b o t h groups

wo r d u n p l e a s a n t n e s s ,

This

given t o the
of

that

were

the

p o t e n t a nd g e n e r a l l y

.01 level

present

to

observed

variable

as w e l l .

study w a s that

ful i n itself,

What
the

factors,

o n l y with t h e
.0^ and

here

stu d ents,

.01 levels.

of f r e q u e n c y is a m o r e
in d u r a t i o n t h r e s h o l d s

and perhaps
seems

ef f e c t

words,

to

than other affective

to h a v e o c c u r r e d i n the
of word frequency,

a significant extent with

s i g n i f i c a n t p r o p o r t i o n of the
and a p p a r e n t l y

college

subje c t s .

only in these

power­

n e u t r a l c o n t rols.

The f a c t

i n c o n j u n c t i o n w i t h the

that this

common words

a

I n these

instances,

unpleasant words more r apidly perceived

vigilance

respect

was p a r t i c u l a r l y a c c e n t u a t e d w i t h i d i o s y n c r a t -

ically unpleasant

instances

signif­

in both instances,

affective

effective agent

t h a n is w o r d u n p l e a s a n t n e s s ,
determinants

very

subjects w ith

significance was between the

suggests

only

c o m m o n as c o m p a r e d to u n c o m m o n w o r d s .

d i f f e r e n c e s w e r e b e y o n d the

the

data permit

actual e x p l a n a t i o n .

whereas differences attributable

and

but w h o p e r c e i v e d

significantly more rapidly than

consideration should be

to t h r e s h o l d s f o r
The

to c o n t r o l words,

common words

s p e c u l a t i o n as to the
Due

12^ -

w e r e the

than t h e i r m a t c h e d

effect

occurred only

r e n d e r s the c o n c e p t

of

s o m e w h a t b e c l o u d e d b y the p o s s i b l e c o n n e c t i o n w d t h

a s p e c t s of f r e q u e n c y and

set i n v o l v e d

in the t h r e e

e x p l a n a t i o n s p r e v i o u s l y p r e s e n t e d b y the w r i t e r .

alternative

- 12^Word

length.

The firth subsidiary hypothesis predicted

that d u r a t i o n t h r e s h o l d s f o r l o n g unplea s a n t

w o r d s wo u l d be

s i g n i f i c a n t l y h i g h e r than t h o s e f o r short u n p l e a s a n t words,
in r e l a t i o n to t h e i r r e s p e c t i v e m a t c h e d contr o l s .
effects of d i f f e r e n c e s
controlled,

in w o r d l e n g t h and f r e q u e n c y thus

this h y p o t h e s i s recei v e d no

data r e c o r d e d f o r

either of the two

extent

support

samples;

d i f f e r e n ces i n w o r d le n g t h do not have
upon the

With the

fran the

apparently

si g n i f i c a n t effect

to w h i c h w o r d u n p l e a s a n t n e s s

in f l u e n c e s

du r a t i o n thresholds.
,,ith the patients,

very significant

o b s e r v e d in the t h r e s h o l d s f o r
and neutral,

as

d i f f e r e n c e s were

short words,

both unpleasant

c o m p a r e d to long w o r d s o f b o t h a f f e c t i v e

groupings, w i t h the short w o r d s p e r c e i v e d more rapidly.
However,

when unpleasant

and neu t r a l

words were e x a m i n e d

separa t e ly in r e s p e c t

to le n g t h

and t h e i r t h r e s h o l d s c o m ­

p a r e d in interaction,

the r e s u l t i n g n o n - s ignificant

term

p r o v i d e s no support for the c o n t e n t i o n that w o r d u n p l e a s a n t n -ss
has a d i f f e r e n t i a l e f f e c t f o r short as c o m p a r e d to l o n g e r
words.

That

is, the r e l a t i o n s h i p b e t w e e n l o n g and

words was n o t

s i g n i f i c a n t l y different

short

with u n p l e a s a n t

words

than with control words.
Hot only w e r e

there no

significant

differences obtained

for long u n p l e a s a n t wor ds as comp a r e d to short u n p l e a s a n t
words w i t h the

college sample,

but none w e r e

co m p a r i s o n s of all of the l o n g w o r d s with the

obtained f r o m
short w o r d s

- 126-

uscd,

or the l o n g control w o r d s with t h e short control

words.

Thus,

w i t h both samples, d i f f e r e n c e s in the length

of those words u s e d a s stimuli, v a r y i n g f r o m t h r e e to ten
le tters in the college group,
ne g l i g i b l e

seemed to b e a r e l a t i v e l y

f a c t o r in the e f f e c t of word u n p l e a s a n t n e s s on

du r a t i o n thresholds.

Conversely,

the effect of differ e n c e s

in w o r d u n p l e a s a n t n e s s , u n p l e a s a n t w o r d s o p p o s e d to the
neutral

control words,

the effect

s e e m e d to be of

no

significance in

of w o r d le n g t h o n duration thresholds.

The likely

e x p l a n a t i o n for the d i f f e r e n c e s o b s e r v e d

in r e s p e c t to the e f f e c t s of word le n g t h

on

for all the w o r d s

to the

combined, w h e n

subjects as c o m p a r e d
a p p r o p r i a te here.

to the

exposed

c o l l e g e group,

students were
short

while

those for the

It is the differ e n c e s

of these d i f f e rences w h i c h are important

and not the d i f f e r e n c e s b e t w e e n

samples p e r se.
the

*3 and 99*9*

short and long

and l o n g w o r d s o b s e r v e d i n each sample

ana the s i g n i f i c a n c e s
here,

hospit a l

is perh a p s

subjects f o r the

w o r d s were 7 8 . f? and 95*7 respectively,

b e t w e e n the

thresholds

The m e a n l o g a r i t h m i c a l l y t r a n s forme d

t h r e s h o l d s of the hospital

college

the

I n respect to the

d i f f e rences b e t w e e n the

the m e a n s of the two

latter,

however, while

samples may a p p e a r of consequence,

they can l a r g e l y be e x p l a i n e d w i t h r e f e r e n c e to the fr e q uency
of the w o r d s chosen in the
and next least liked words,

two

samples as the least liked

since this also d e t e r m i n e d the

f r e q u e n c ies of the m a t c h e d control words.

It may be r emembered

-

that

127-

the t h r e s h o l d d i f f e r e n c e s

wcras were highly

between common and uncommon

significant f o r both

samples.

m e a n T h o r n d i k e - L o r g e f r e q u e n c y r a t i n g s are
short and lcnr words

in b o t h

groups,^

p ared w i t h the m e a n s

of the thresholds,

When

the

c o m p u t e d for

and t h e s e

are c o m ­

they a p p e a r

as

shown in T a b l e XII, p r e s e n t e d o n the f o l l o w i n g page.
Note

that w i t h i n e a c h

sample t h e r e

relationship between magnitude
frequency

rating.

As expected,

is an inverse

of t h r e s h o l d a n d that
the

of

c o m m o n words, t h o s e

w i t h the h i g h e r m e a n f r e q u e n c y r a t i n g s were seen at l o w e r
t h r e s h o l d s t h a n the u n c o m m o n w o r d s with the l o w e r m e a n
f r e q u e n c y ratings.
The w r i t e r a t t r i b u t e s the r e l a t i v e l y
likelihood insignificant

d i f f e r e n c e b e t w e e n the

of the long wo r d s f o r the h o s p i t a l
w o r d s of

slight

the college g r o u p

sample

and

thresholds

and the short

(which are i n d i r e c t r a t h e r than

i n v e r s e r e l a t i o n s h i p to the f r e q u e n c y d i f f e r e n c e s )

to

effect of u n c o n t r o l l e d d i s c r e p a n c i e s b e t w e e n the two
Since the e x p e r i m e n t was not
san.ple c o m p a r i s o n s

i n all

the
samples.

d e s i g n e d to include inter-

ex c e p t as i n c i d e n t a l a s p e c t s o f p o s s i b l e

^ T h o r n d i k e - L o r g e n o n - n u m e r i c a l ratings of A a n d AA
w e r e given the m i n i m u m n u m e r i c a l values of the r a n g e s t h e y
represent.
Thus, A, r e p r e s e n t i n g words a p p e a r i n g b e t w e e n
99 an<3 SO times in 1,000,000, was a s s i g n e d a v a l u e of SO,
w h i l e AA, r e p r e s e n t i n g w o r d s o c c u r r i n g 100 or m o r e times
p e r m i l l i o n , was a s s i g n e d the v a l u e of 100.

-

TABLE XII

C O M P A R I S O N OP1 MEAN D U R A T I O N T H R E S H O L D S AND M E A N
T H O R N I I K E - L O R G E F R E Q U E N C Y R A T I N G S FOR
S N O R T AND L O N G V*ORDS

Sample

H o s p I t al

Category
M e a n log.

128-

thre fhol U

Been "”iorndike -Lorse
frequency

College

Short

Long

Short

Long

78.52

95.68

9^.3^

99.92

36.53

12.00

6.73

k*79

129-

-

Interest,

no a t t e m p t w a s m a d e

variables which
was

felt

that

could

visual

acuity,

and

constant
most

allied

not a t t r i b u t a b l e

likely

With

its

processes

to the

sh o r t w o r d s ,

p a t i e n t s but

not for the

the h i g h e r

the p a t i e n t s
r a t h e r t h an
T a b l e XII

the

- the

variations

to e a c h

subject .

The

thresnold differences
appear

to be t h a t

r e t a r d a t i o n of
patients

were

It

in

seens quite

r e l a t i v e l y slow.

differences between thresholds
differences significant for the

students,

it

reflected

differences

is

suggested then

differences

Indicates

that

in length.
the
of

Reference

short words

essentially,

again

to

p e r c e i v e d at

7 3 . 5 2 b y the p a t i e n t s h a d

rating

c e i v e d at

a significantly higher mean
a mean

o f 36.53»

in f r e q u e n c y

frequency

had

subject

thresholds observed for longer words with

mean duration threshold

95»6Q,

the

the s t u d e n t s , and

their reactions were

f o r l o n g and

that

for

concomitant

older than

reference

such

efficiency,

to u s e e a c h

to f r e q u e n c y here, w o u l d

somewhat

It

results would be relatively

accounting

difference with

that

was

c o n t r o l , ” assuming that

physiological response
general

fact o r s ,

stimuli admini s t e r e d

likely factor

or age

of physiological

the e x p e r i m e n t a l
for a l l

these discrepancies.

e f f e c t i v e m e t h o d of c o n t r o l l i n g

as t h o s e

as h i s o w n ’’m a t c h e d
affecting

account fo r

the m o s t

complex variables

to m a t c h f o r a n u m b e r of

a
a mean

w h e r e a s the l o n g e r w o r d s p e r ­
threshold,

f r e q u e n c y ratine; of o n l y

12.00.

t h a t of
Thus,

-

in the process
liked words,
w o r d list,
which

-

of s e l e c t i o n of t h e

and/or

the p atients

less frequently

but

long

Vn'ith t h e

comparatively

with

the

next least

liked words,

patients'

where no prominent

would

seem

thresnold

differences

that w h e n the e f f e c t s

frequency are

t a k e n into

is

suggested

in

the

that w i t h

st u d y ,

once

the

the

the

be

presumably

both long

m a y be p e r c e i v e d
of

the

with

subject's eye

and

s h o r t wo r d s ,
Thus,

effects

for ccntrol

and u n p l e a s a n t

of s u b j e c t s .

reached,

as r a p i d l y a s are

and

span.

employed
long w o r d s

eye and

as a r e s u l t of w h i c h

short w o r d s of

a si n g l e

It

s h o r t words.

d i s tance b e t w e e n the

back,

it

of d i f f e r e n c e s

threshold

has b e e n

no

in

procedure

tachistoscope exposure

Here

differences existed

the t a c h i s t o s c o p i c

a f u n c t i o n o f the

and

s h o r t words.

with b o t h g r o u p s

are r e c o g n i z e d a p p r o x i m a t e l y
This may

l o n g words,

of differences

account,

combined are minimal

other

both being

were f o u n d .

in l e n g t h u p o n d u r a t i o n t h r e s h o l d s
words

considerably

short words which they

b e t w e e n t h e m e a n f r e q u e n c i e s of l o n g
signifleant

were

in the

c o m m o n than the patients'

students,

short w o r d s

small differences

and

markedly less

group

on t h e

s e l e c t e d as l e a s t

c o m m o n t h a n the

least

students,

of the l o n g and

less

chosen

words which

frequency ratings

somewhat

of the p e e r

a p p e a r to h a v e

encountered.

there were

l e a s t a n d next

i n the c o m p i l i n g

were v e r y common,

hand,

130

the l e n g t h

employed

fixation, w i t h i n t h e limits

-

Additional

131

-

Conclusions and Comments

O r d e r of p r e s e n t a t l o n .

It was of p a r t i c u l a r interest

w i t h the h o s p i t a l group to investigate w h e t h e r or not the
i d e n t i f i c a t i o n of the first unpleasant w o r d had an
a pprec i a ble effect on the d u r a t i o n t h r e s h o l d of the
unpleasant word.

This

compar i s o n was made,

di'Ter e n ces significant at between the

second

and while

.05 and

.01 levels

of confidence w e r e obta i n e d w h e n the series of five words
presented first were
second,

this seems

compared with the

to have b e e n largely a f u n c t i o n of

differences in frequency,
W h e n the

series presented

u n c o n t r o l l e d in this grouping.

duration thresholds of each series were d i f f e r e n t ­

iated in terms of unpleasant

and control words,

thus i n ­

v o lving controls over the factors of frequency and wor d
length,

no significant

seemed to indicate

dif f e r e n c e s were

that

there was

revealed.

no appreciable

This
carry- o v e r

from the i d e n t i f i c a t i o n of the first unpleasant word to the
r e c o g n i t i o n of the second,
threshold of the latter.

with respect to the dura t i o n
As alre a d y Indicated, no si g n i f ­

icant d i f f e r e n c e s had been obtained with t h e college group,
in w h i c h this variable had been c o n f o u n d e d with that of
the d e gree of the
unpleasant

subject's

for his

choices of

stimulus words.

Sex differences.
Howes and

dislike

Mclinnies (lJg) was criticized by

Solomon (5) for using a female assistant in his

-

G x p e r l m e n t w h i c h m a d e use
s u b j e c t s of both
affecting

sexes.

132of

sexually l o a d e d w o r d s w i t h

This w a s c o n s t r u e d a s p o s s i b l y

the i n t e r p r e t a t i o n of his

observations,

to c o nfuse the e f f e c t of an i n h i b i t i o n of verbal

tending
report

w i t h what h a d b e e n c o n s i d e r e d as actual p e r c e p t u a l b l o c k ­
ing.

Since

the pre s e n t

study also

frequently interpreted with

included wards

sexual connotations,

and since

a p p r o x i m a t e l y h a l f of the m e m b e r s of the c o l l e g e
were w o m e n

and thus m i g h t

conceivably feel

in reporting their perceptions
was of i n t e r e s t to exp l o r e
have a f f e c t e d

the

While v e r y

to a m a l e

some

constraint

e x p e r imenter,

the p o s s i b i l i t y that

experimental

sample

it

this m i g h t

results.

significant differences were

obtained for

the d u r a t i o n t h r e s h o l d s of the t e n w o r d s r e c o r d e d f o r the
female

as comparec

group,

the point

were

to the m a l e

of i m p o r t a n c e is

e x a m i n e d in r e g a r d

Thus,

the

that when these r e s u l t s

to d i f f e r e n c e s o b s e r v e d b e t w e e n the

u n p l e a s a n t and c o n t r o l words,
revealed.

subjects in t h e college

no

si g n i f i c a n t

i n t e r a c t i o n was

apparent d i f f e r e n c e s n o t e d b e t w e e n

the m e n and w o m e n in r e s p e c t

to the t h r e s h o l d s

seems to have

b e e n a f u n c t i o n of the v a r i a b l e s u n c o n t r o l l e d in c o m p a r i n g
the d u r a t i o n t h r e s h o l d s of v a r i o u s i n d i v i d u a l s
w o r d s but

c o n t r o l l e d in the inte r a c t i o n ,

f r e q u e n c y and w o r d length,

especially

for different

the v a r i a b l e s of

the former.

-1 3 3 C o m p a r i s o n of R e s u l t s w i t h t h o s e o f R e l a t e d

So m e i m p l i c a t i o n s
and

their connections

experiments have

to c o n c l u s i o n s

already

relevant

to

related,

p o i n t i n g out

Eefore

extend this

doing

nique used

of the r e s u l t s

this

question.
so m e w h a t

to

significant

of r e s u l t s

I n the first
different

place,

in nat ure;
words

ti o n e d

syllables,

le ngth.

I n the

hold was

re-defined

the f a c t

The

propriety

and c o n c l u s i o n s

for use

(1)

would

use of

b e e n done.

stimuli

sug g e s t

that

a non-college

assumed

stimuli

the m e a s u r e

i n the

of

the t y p i c a l

sa mple of

not

procedure

makes

condi­
sent e n c e

employed.

strongly

have been

this m i g h t b e t h e

to b e

of d u r a t i o n t h r e s ­

experiment by Elake

group

used were

they c o n s i s t e d o f i d i o s y n c r a t -

the s u b j e c t s w e r e

The

frorn

o p e n to c o n s i d e r a b l e

critical

or auditory

sec o n d place,

that

the

drawn

s y m b o l i c of p e r s o n a l v a l u e s ,

differences which would

this not

tech­

of m a k i n g

u r g e d to

m a k e p r e - r e c o viiticn h y p o t h e s e s m a y h a v e p r o d u c e d
uated

be

s i m i l a r i t i e s and d i f f e r e n c e s .

rather than words

words

nonsense

however,

c e r t a i n p e c u l i a r i t i e s I n the

s o c i a l l y t aboo,

Third,

I t may,

o t h e r s w h i c h a r e r e l a t e d Is

ically unpleasant

certain other

study should be made explicit.

comparisons

study

drawn from

study

d i s c u s s i o n to s t u d i e s l e s s I n t i m a t e l y

C o m p a r l sons \vi th thi s s t u d y .
direct

o b t a i n e d I n tliis

been m e n t i o n e d .

so, h o w e v e r ,

In this

Studies

or a c c e n t ­

obvious,

had

and Vanderplas

case.

Fourth,

the

certain comparisons with

sophomore psychology students used a

3

-1 3 1 1 -

subjects in m a n y of the r e l a t e d studies

R e l ated
Eriksen,

studies.

and Fonda

present study,

somewhat inappropriate.

The r e s u l t s reported by Lazarus,

(7), in common w i t h the f i n d i n g s of the

seem to point toward the defi n i t e p o s s i b i l i t y

of o b t a ining either r a i s e d o r l o w e r e d perceptual
as results of the emotional c o n n o t a t i o n s
specified character.
study,

W i t h college

thresholds

of stimuli of

students i n t h e present

r a ised thresholds w e r e obse r v e d for w o r d s classi f i e d

as m i ldly disliked by p e e r groups, while

l o w e r e d thresholds

were n oted for common unplea s a n t words.

Lazarus,

and Fonda,

dealing w i t h a u d i t o r y perception,

end i n d i v idually consis t e n t

Eriks e n

f o u n d two basic

reactions to th r e a t e n i n g stimuli:

(1)

high percep t u a l accuracy

(2)

low perceptual accuracy and m i n i m i z e d v e r b a l i z a t i o n and

blocking.

Me jinnies

and ready verbalization;

(ill) and M c G i n n i e s and

Sherman

and

(15)

found predorninantly raised thresh o l d s to t h e i r critical
words, while

Postman

and S o l o m o n (21)

L e y t h a m (20)

seem to have

and Postman and

concluded from t h e i r results

lowered thresholds m a y occur with

that

stimuli of either r e w a r d ­

ing or punishing connotation.
The significant thres h o l d d i f f e rences f o u n d in our
college

sample with the

common unpleasant w o r d s as c o m p a r e d

to the u n c o m m o n u n p l e a s a n t words is somewhat difficult

to

reconcile with conclusions drawn from certain other i n v e s ­
tigations.

It m a y be that the u s e of nonsense

syllables

i

-135by Lazarus and McCleary
failure

to f i n d

perception of
pared

In

factor

In their

differences

i n the

accuracy

so c o n d i t i o n e d .

threshold

quite possible
to

one

shock-associated nonsense

occurred with the

learned

12) w a s

significant

t o t h o s e not

significant

(8,

that

the nonsense

the

achieved

exact nature

words,

itself,

is n o t

ingenious

Also,
i ngs,

in

it w a s ,

use of c o m m o n words.

of

e f f e c t of

the

Allport-Vernon value
words,

not

and

is o p e n

(26)

ranks

effect of v a l u e

ranks,

the

which

c r i t e r i a of

Schneider

the

(22),

to

the

too,

observed

the

the

of o s t e n s i b l y m o t i v a t i o n a l

analogue

question.
writer's find­

greater
of

P o s t m a n and

same t e n d e n c y In a n o t h e r
It may be

In differences

stimuli

infrequent

intensity

stimuli.

the A l l p o r t - V e r n o n v a l u e s .
discrepancies lies

the

e x h i b i t e d the

their

key

to s o m e

condi­

to r e a l i t y

that i t w a s the

motivational connotation with

study u s i n g

thus

the

In their experiment using

noted

the c o m m o n w o r d s ,

experiment

i n the

a p p a r e n t c o n t r a d i c t i o n to

Solomon and Kowes

th e

syllables were never

fully understood,
as

com­

a n d it s e e m s

t i o n i n g p r o c e d u r e u s e d b y t h e m a s an a n a l o g u e
situations

as

indicated,

in the present

common unpleasant

the p o i n t

addition,

differences

As

syllables

of

used.

that

In the

the

types

-136Summary
I n summary, then,
in r e s p e c t to this

the f o l l o w i n g f i n d i n g s were n o t e d

e x p e r i m e n t w h i c h empl o y e d a sample of

general h o s p i t a l p a t i e n t s and a sample of college students:
(1)

U n p l e a s a n t w o r d s w h i c h had b e e n d e s i g n a t e d as
m i l d l y d i s l i k e d o n the basis of p e e r group r a t i n g s
were p e r c e i v e d at s i g n i f i c a n t l y hig h e r d u r a t i o n
t h r e sholds by

college students t h a n were neutral

’’c o n t r o l " w o r d s mat c h e d w i t h them in terms of
n u m b e r of letters
d i f f e r e n c e s were

and frequency.

No

significant

o b t a i n e d w i t h the hospital

p a t i e n t s w i t h respect to this

variable.

d i f f e r e n c e s were obser v e d with e i t h e r
regard

No

sample

in

to the t h r e s h o l d s c o m p a r e d for u n p l e a s a n t

words d e s i g n a t e d as strongly d i s l i k e d and t h e i r
matched
(2)

control words.

C o m m o n words,

unpleasant

cei v e d at very

and control, were

significantly lov/er dura t i o n t h r e s ­

h o l d s than w e r e u n c o m m o n words,
(3)

per­

in both samples.

C o m m o n u n p l e a s a n t w o r d s were p e r c e i v e d by co l l e g e
students at significantly l o w e r t h r e s h o l d s than
were m a t c h e d c o m m o n control words.

Uncommon u n ­

pleasant w o r d s w e r e not p e r c e i v e d d i f f e r e n t l y from
their

controls.

'.Vi th the patients, c o m m o n u n ­

p l e a s a n t words were p e r c e i v e d no dif f e r e n t l y
r e l a t i o n to

their m a t c h e d

common

in

control words

-137than w e r e u n c o m m o n u n p l e a s a n t w o r d s
to t h e i r m a t c h e d u n c o m m o n
n e i t h e r c a s e were t h e r e
between unpleasant
(J|)

and

cont r o l words;

significant

in

differences

control words.

Short w o r d s w e r e p e r c e i v e d at very
lower

In r e l a t i o n

significantly

d u r a t i o n t h r e s h o l d s by t h e p a t i e n t s

were long words.
e r e n c e s were

W i t h the

in the

students,

th a n

the d i f f ­

same d i r e c t i o n but w e r e not

si g n i f i c a n t .
(t>)

There were

significant

frequencies

of pre-recognition hypotheses within

the c a t e g o r i e s d e f i n e d
s t r u c t u r a l l y unlike,
made

d i f f e r e n c e s in t h e

as

nonse n s e ,

and part r e s p o n s e s

by h o s p i t a l p a t i e n t s to u n p l e a s a n t a s

p a r e d to c o n t r o l w o r d s .
f o u n d to be

significant

r e s p o n s e s of
No

s t r u c t u r a l l y similar,

the

com­

No s u c h d i f f e r e n c e s w e r e
with the

college

pre-recognition

students.

significant differences in duration thresholds were

n o t e d I n ei t h e r

sample w i t h

respect

(1)

word u n p l e a s a n t n e s s ;

(2)

the d e g r e e

to the f o l l o w i n g :

to v h i c h un p l e a s a n t

w o r d s were rated

s o c i a l l y taboo;
(3)

the o r d e r i n w h i c h t h e u n p l e a s a n t
(in the f i r s t

(Lj.)

The

degree

f i v e - w o r d series

to w h i c h the

n e x t l e a s t liked);

or the second);

u n p l e a s a n t word had

rated as l e a s t l i k e d by the
vs.

word was presented

subject

been

(least liked

-138(5 )

the influence of word l e n g t h u p o n the
yjd r d

(6 )

effect o f

unplea santness ; and

sex of the

subject

(college group alone).

I n r e gard to the f r e q u e n c i e s of pre-re cognition
hypotheses
citec,

no

falling into the f o u r categories p r e v i o u s l y
significant dif ferences w e r e n o t e d for either

sample w h e n responses to u n p l e a s a n t and neutral
words were c o m p a r e d for

matched

series of w o r d s in w h i c h the u n ­

pleasant w o r d s had been I n d i c a t e d as stro n g l y disl i k e d b y
a peer group as opposed to series in w h i c h the unpleasan t
words h a d been indicated as m i l d l y disliked.

G e n e r a l Conclusions

It is c o n c l u d e d o n the basis of the r e s u l t s obtained
in this experiment,

that the

concept of perceptual defense

is not s u p p o r t e d b y the response to unpleasant w o r d s as
such.

No evidence

for this concept w a s found w h e n the

d u r a t i o n thresholds f o r unpleasant w o r d s were
those o b s e r v e d for wo r d s of o s t e n s i b l y neutral

compared w i t h
affective

c o n n o t a t i o n m a t c h e d with the unpleasant words in number of
letters and frequency.
observed with

college

pleasant words, here

Raised durat i o n thresholds were
students,

but only with

certain u n ­

classified as m i l d l y disliked in terms

of p e e r group ratings.
holds were observed with

O n the other hand,
college

lowered t h r e s ­

students for common

-139unpleasant

words.

c o n c l u s i o n is

O n the b a s i s o f these

d r a w n that p e r c e p t u a l d e f e n s e and

vigilance may perhaps

operate for

groups of individuals,
pleasant words,

but

the

evidence

seems at

two c o n c e p t s

Lower perceptual

words

but

observed for common

also

Differences

v a r i a b l e s , w o r d f r e q u e n c y and length,

to g e n e r a l

stimuli.

with b o t h

thresholds were

than f o r long wo r d s .

significant,

or

best i n c o n c l u s i v e

in r e f e r e n c e

thresholds were

than f o r u n c o m m o n w o r d s
lower

perceptual

certain individuals

m o d e s of r e a c t i o n to u n p l e a s a n t v i s u a l

the p a t i e n t s ,

tbe

i n r e s p o n s e to c e r t a i n t y p e s of u n ­

Tor p o s t u l a t i n g t h e s e

words

results,

samples,

and w i t h

r e c o r d e d for s h o r t
in r e s p e c t

to these

appeared highly

c o m p u t a t i o n of the m e a n f r e q u e n c y r a t i n g s

of the l o n g and

short w o r d s

the d i f f e r e n c e s

attributed

f u n c t i o n of f r e q u e n c y .

The

indicated

that i n all

to w o r d l e n g t h w e r e
actual

d u r a t i o n t h r e s h o l d s for the w o r d s

likelihood

largely a

e f f e c t of w o r d l e n g t h u p o n
e m p l o y e d in the s t u d y

appeared minimal.
It is a l s o

c o n c l u d e d t h a t w i t h c e r t a i n i n d i v i d u a l s or

g r o u p s o f individuals,

d i f f e r e n c e s in p r e - r e c o g n i t i o n h y ­

p o t h e s e s to u n p l e a s a n t

as c o m p a r e d

occur,

with

the m o s t

the g r e a t e r number

of

prominent

to p l e a s a n t

differences

wo r d s

do

a t t r i b u t a b l e to

structurally unlike hypotheses occurr­

ing to u n p l e a s a n t words.

I n v i e w of the

fact that

such

-In­
differences were
employed,
this

to b e

stimuli

is

o b s e r v e d only with

the hospital

patients,

a general m o d e

the

one of the

two s a m p l e s

evidence for

assuming

of r e a c t i o n to u n p l e a s a n t

considered inconclusive.

visual

BIBLIOGRAPHY
1.

Blake, R. R., and J. 1.1. Vanderplas.
r e c o g n i t i o n h y p o t h e s e s on v e r i d i c a l
t h r e s h o l d s in a u d i t o r y p e r c e p t i o n .
1950, 19, 95-115.

2.

Fruner, J. S., a n d G. C. Goodman.
o r g a n i z i n g f a c t o r in pe r c e p t i o n .
I 9 I+S, 2 7 / 2 0 3 - 2 0 8 .

3.

Bruner, J. S., and D. Krech, Editors.
P e r s o n a l i t y . Durham, N o r t h Carolina,
P r e s s , 1 9 I+9 » 266 pp.

The effect of prerecognition
J. Personality,

Symbolic value as an
J. Soc. Psychol.,

P e r c e p t i o n and
Duke U n i v e r s i t y

Edwards, A. L.
E x p e r i m e n t a l D e s i g n in P s y c h o l o g l c a l
R e s e a r c h . New York, R i n e h a r t b Company, 1950.
l+lj-b pp.
5.

6.

Howes, D. H., a n d R. L. Solomon.
A Note on McGinnies*
"B-notlonality and oerceotual defense.'1 P s y c h o l . R e v .,
1950. 57, 229-231)..
Klein, (>. S., H. J. Schlesinger, a n d D. S. Meister.
The
effect of p e r s o n a l values on p e r c e ption:
an exp e r i m e n t a l
critique.
P s y c h o l . R e v ., 1951, 5 / 95-112.

7.

Lazarus, R. S., C. V.'. Eriksen, and C. P. Eonda.
Per­
s o n a l i t y d y n a m i c s and a u d i t o r y p e r c e p t u a l recognition.
«J. P ersonality, 1991, 19, 1+71-1+82.

8.

Lazarus,
L ., and P. A. McCleary.
Autonomic d i s ­
c r i m i n a t i o n w i t h o u t awareness:
a study of subception.
P s y c h o l . Pev_., 1951, 5 q , 2, 113-122.

9.

Levine, P . , I. Chain, and G. Murphy.
The r e l a t i o n of the
i n t e n s i t y of a n e e d to the amount of perceptual d i s t o r ­
tion.
J. P s y c h o l ., 19/2, 13, 287-293.

10.

Luchins, A . S.
Social i n f l u e n c e s o n p e r c e p t i o n of c o m ­
plex drawings.
J. S o c . PsychoJL., 19+-5, 20, 257-273*

11.

Luchins, A. S.
A n e v a l u a t i o n of some current crit i c i s m s
of G e s t a l t p s v c h o l o g i c a l w o r k on perception.
P s y c h o l.
Rev., 1951, 5$, 69-95.

12.

McCleary, R. A., and R. S. Lazarus.
Autonomic D i s ­
c r i m i n a t i o n W i t h o u t Awareness:
An I n t e r i m Report.
Pp. 171-179.
Bruner, J. S., and D. Krech, Editors.
See r e f e r e n c e No. 8.

t,

i

- 114.2 13.

M c C l e l l a n d , D. C., 2. A. Clark, T. D. Roby, and J. W.
Atkinson.
P r o j e c t i v e e x p r e s s i o n o f needs; the effect
of n e e d f o r a c h i e v e m e n t on thematic appe r c e p t i o n .
J.
E x p e r . P s y c h o l ., 19^4-9, 39, 2l|.2-255.

llq.

M c G i n nies, E.
Emotionality and perceptual defense.
P s y c h o l . R e v . , 19l4-9, 56, 2ipa.-25l.

15.

F c G i n n i e s , E . , and K.
p e r c e p t u a l defense.

Sherman.
G-ene ral iz at ion of
J. Abnorm. Soc. Psychol., 1952,

Lp7, 61-85.
16 .

F i l ler, 2. C.
i iscri-ri n a t i o n wit h o u t
1- P s y c h o l ., 1939, 52, 562- 5 7 8 .

17.

Pastore, N.
N e e d as a d e t e r m i n a n t
P s y c h o l., 19149, 28, )
4 57-476.

18.

P o s tman, L., and J. S. Bruner.
M u l t i p l i c i t y of set as
a d e t e r m i n a n t of o e r c e o t u a l b e h a v i o r .
^J. E x p e r . P s y c h o l . ,
19J-!9, 39, 369-377*.

19.

Postman, L., J. S. Fruner, and 3. McGinnies.
v a l u e s as s e l e c t i v e f a c t o r s in p e r c e p t i o n ,
S o c . P s y c h o l ., I 9 I+-8 , I_3, ll-i-2—15h*.

20.

Postman,

L., and G.

Leytham.

ambivalence of stimuli.

awareness.

of p e r c e p t i o n .

Pe r c e p t u a l

Amer.

_J.

Personal
J. A b n o r m .

selectivity and

J_. Per sonality, 1951, 19, 390-

1^05.
21.

P o s tman, L., and R. L. Solomon.
Perceptual sensitivity
to c o m p l e t e d and i n c o m o l e t e d tasks.
J. P e r s o n a l i t y ,
1950, 1°, 347-356.

22.

Postman, L . , a n d P. Schneider.
P e r s o n a l v a l u e s , visual
r e c c r n i t i o n , a n d recall.
P s y c h o l . R e v . , 1951, 56, 2712 8 1 |-.'

23.

Potter, I . 0.
Percept: on n P Gymbol Orientatl o n and Early
R e a d i n g S u c c e ss.
New York, T e a c h e r s College, C o l u m b i a
U n i v e r s i t y , 1 7f9 • 6 9 pp.

2 I4..

P r o s h a n s k y , H. and G. Murphy.
The e f f e c t s of r e w a r d and
p u n i s h m e n t on pe r c e p t i o n . J. P s y c h o l ., 19^2, 13, 295-395*

25.

Shafer, P., and G. Ivlurphy.
Role of a u t i s m in visual
figure-ground relationships.
J. Exp e r . P s y c h o l . , 19^3,
3 2 , 335-3J43-

26.

Solomon, R. L . , and D. H. Howes.
kora f r e q u e n c y , p e r s o n a l
values, and v i s u a l d u r a t i o n t h r e s h o l a s . Psy c h o l . Rev.,
1951, 58, 256-270.

27.

Thorndike, 2. L., and I. Lorge.
The T e a c h e r 1 s W o r d b o o k
of 10 ,000 words*
New York, Teachers College, Columbia
Un i v e r sity, 19V+* 27^- PP*

28.

V 9 rville, E.
The effect of emotional a n d motivational
sets on the p e r c e p t i o n of incomplete pictures.
J. d e n .
P s y c h o l ., 194-0 * & 9 1 133-li|-5*

29.

V.ebster’s Collegiate Dictionary.
F i f t h Edition.
Springfield, Mass., G . C. i»erriam Company, 19^3.
1275 PP*

In requesting you to fill out this questionnaire as carefully and honestly
ssible, we are asking your cooperation in gathering material tc be used
in a psychological experiment.
We wish to study certain things which most
e like and others which they dislike, and we hope to learn more about how
react to these feelings.
Take as much time as you feel is necessary,
-all
rs will, of course, be held strictly confidential.
______________________________________

aGE

waRD _____________

.rv there any lines of work which you admire g r e a t l y ? ______________________
that occupation would you admire most? _______________________________________
re there any kinds of work which you would aislike? _______________________
«hat line of work would you dislike most of a l l ? ____________________________
*re you fond of a n i m a l s ? ________________________which animal would you like
T.ost as a p e t ? _________________________________________________________________
trc there any animals which you dislike or which bother you? _______________
_______________..hich animal bothers you m o s t ? ________________________________
Do you find any insects disagreeable or repulsive? _________________________
<(hat insect do you find most disagreeable?__________________________________
*re there any words in the English language which you find particularly
p l e a s a n t ? ___________________ Name three : _____________________ , ______________

-(hat three words are the most unpleasant ones you can think o f ? ___________
------------------------ 1

______________________________________________ j

_________________________________________ —

— —

•

If you were being -chased with something that could be used as a weapcn against
you, what weapon would frighten you most? ___________________________________
chat way of dying would seem most terrifying to you? _______________________

*re there any kinds of people whom you dislike? _____________________________
.(hat kind of person do you dislike m o s t ? ____________________________________
(.hat kind of person do you like b e s t ? _______________________________________
Have any experiences in ycur life been particularly pleasing? ______________
what experience was the most pleasing? ______________________________________

Have you had any experiences which were particularly irritating to you?
_______________________________

What was it that was the most irritating?

Have you ever gone through any particularly frightening experiences?
____________

what was it that frightened you most af all? __________

Do you remember ever having been extremely embarrassed?
..hat was it that caused your embarrassment?___________

have you ever been severely ashamed of y o u r s e l f ? __________________________
'..hat was it that was responsible for t h i s ? ________________________________
Has the possibility of having any illness or disease ever been disturbing
to y o u ? _________________________ Of what illness or disease have you been
most frightened? ____________________________________________________________
Doos any illness or disease condition appear particularly disagreeable or
repulsive to you?

t____________ ..hich illness or disease condition appears

moot disagreeable? ___________________________________________________________
.(hen you think of all the things in the world that could possibly happen
to you:
(a) v.hat do you think would make you the most a n g r y ? _____________________

(b) ..hat would make you the hapoieot?

U)

what would make you the most frightened?

(d) .(hat would make you the most unhappy?

SENTENCE COMPLETION TEST (A d a m *)
NUMBER

TI ENT

DATE

L

00 D

tOUBLE

TO K N O W

ED TIME

A N N O Y S ME

EL A S H A M E D W H E N

FOME

TGRET

BEST

ER P E O P L E

USUALLY

MY M O T H E R

VT

ME

PLI7ZLF. S

I HAD MY W A Y

ST

ROSSES

FEET

'

Nl

HIM

HURT

WHEN

U\FS

I

TUf H

HI

V i

'r

...

WA V .

A

( H I Li)

I ' [ OF-1 F

, 1

F«. ' *. S

■*. '

■ ! Art

WA' - .

VAF.'RiED

II TO
Wr(> Aj

'v i

P 'FtKl

'9 rj I

WORKED

10-2362

Af - STUND

v A DC

128

8. A WIFE
'9. THE HAPPIEST TIME
JO. THE ONLY TROUBLE
31. MY GREATEST HOPE
32.

I HATE

33.

I AM VERY

34. MOST DOCTORS
35. THE WAR
36. A MOTHER-IN-LAW
37. THE FUTURE
38. THE STRANGEST THING
39. MY MIND
40.

I FAILED

41. MY EDUCATION
42. THE BEST JOB
43.

I'M AFRAID

44.

I CANNOT UNDERSTAND WHAT MAKES ME

45. THE WORST THING
46. MOST GIRLS
47. MY FAMILY NEVER
48. MY MOST IMPORTANT DECISION WAS
49. MY GREATEST WORRY IS
50.

I OBJECT

51. A FATHER-IN-LAW
52. TODAY.

I

53. MY SCHOOL WORK
BECOME EMBARASSED

54.

|

55.

I GET

56.

MY

J OB

ANGRY

WHEN

>1.1.

MINER

4*\K

l <.ople. talk "to one another, they qo so with words which i:.ay mean plenrnnt
r
casunt things. When a person talks to a doctor or a psychologist, he talks
t his problems ^nd uses woras which often mean unpleasant things. VR; are la­
sted m finding ir.ore about hew people react to the unpleasant words which tr.« v
and near.
In order to do so, we need to find words which a groat many peopleare unpleasant,
Please read through this list and put a check in front of
word which you find ycu definitely dislike. The woru may be unpleasant bee of the way it looks, the way it sounds, or what it represents. Do net mark
weras which you do not know or fully understand. Ycur opinions will be held
i c i e r tial and your name is not required.

FjiGxN
iLrhCh
i\ion'll..jihL

.....LiGNxNT

onei.eUO
ij_l.ijCij-i.iiL
1;.ai< U ix
Ga NGr EIE

Ihr.oTTLL
FjiienlV iO U ij

SiEGL
NUPT iiiL
KEPUi-SE
TOILET
FF.ISONER
SIN
..EIa NCHOLY
r£TREa T
o jiT h
SC

uRCH

•.

r j

r

i .i‘
.ijjiiC

ij i l L
SKELETON
PnGF a IE
SKE.ER.
...ISEn
T jivlUnE
Fix i i.U;. Xj-Li

SCOi-.PIGN
iiiX T G r x D

xx vnNUE
t i i ’OO 1 I l j . 1 L
o C i X jlj».

FLNlTLNTLnxY
l i jJ,iP
in 1 0 il».iS.ii
SCUTTLL
i l i i . !Ji«.jii ink

FLY
e-LivF

i.iGox
j-.Gfi.

raia

TEST
RECKLESS
NOOSE
SHRIkP
kaNGLE
OCTOPUS
iEPEL
A-ilR

THRa SH
Pa n i c
SICK
NUDL
rE PROa CH
TILID
PhEJUDICL
SILLY
tdiUL
RESENT
OBSCENE
SCa V u NGEr
MENACE

THEFT
PECOXL
NohfriitL
SHRIVEL
Iva N Ia C
OB ST i Na TE
REPRIMAND
IvaSH
THPEa TEN
Pa NSY
blCKNESS
NU ISa NCE
1EPT ILL
TITTEx
PREY
SIMPER

fvEjiGEx
IE STRICT
OBNOXIOUS
SCCFF
I.ENTa L

TOI v'iB
Pa r a SITE

t orient
Pa SSION

SINISTER
IwlLDE'w
PLOP ITLEE
SKIMP
LISERY
TOTTE a
Pa UNCH
SCOUNDREL
WRECK

SINUS
M L I T a HY
PROMISCUOUS
SKULK
IviiNGE
TPlKGIC
Pa UPEH

revolting
F R O T a UDE

SCREECH
PERJUnY
TRAMPLE
PEST
SEDUCE
RIB a LD
PSYCHIxTxioT
SLa SH
ilIu T

PUTHID

scoukge

iiGrE
xEV GLUT ION
PnOVOKE
SCi'En
PERSECUTE
TR a P
PESTILENCE
SEIZE
P.IDICUiE
PSYCHonOGY
SLa T
FlISK
PUSSY

TnlEF
REMORSE
NOTOhiOUS
SHROUD
IvaN IPULxib
OBSESSION
IEPEISa L
Iv-a SS a Ci E

TilHOB
P a Ra L Y Z E

SIDLE
NUkB
KEPUGNa NT
TO a D

PHICK
SIMPLE
NED a L
F E T a L I a TE

OBESITY
SCOLD
IwExClLESn
T Ci X a DC
Pj i S S IV E

SISSY
LEE
PEOP jxGa NDa

SKULL
T OKIE NT
Pa t h e t i c

SCORN
WRITHE
REVENGE
PROST IT U T 'ij
oCOnv L
Pa LN
T Ia IT O u

PERVEa T
SCUM
i-EV jLVEa
Pi.G’wL
oLa u DER
ilIG ld
^Uj.KE
PURGE

OwiJ
Cii
ILY

S a N I T a A IL i L

MURDER
POUCH
SC a B
PR a T T L E
Na K ED

a GE

LiiTY
IL a IE

K
NT
.PON
E
'a GE
R OUS
HT.Ri ioN
iLP
.j.nX)J USI i^ENT
Ia N
ECa
ijUKL
-A 1Xli
l
a

lili

IG N a NT
.v u s

i.LGa IE
.1* Uiix<
.NGRE 1*E
E l IT E E
I i a LC’Ua

iuiT
tavern

NEASEL
06 LE
iviiV1SH
NEUROTIC
SC a L Y
Ka L a D Y

S C a ND a L
H-a L A L I a
pallor

i,a LICE

PAIN
TEST
RECKLESS
NOoSE
SHRIMP
kaNGLE
OCTOPUS
itEPEL
iuR
IH R a SH
Pa n i c

a

TIL ID
PREJUDICE
SILLY

I L a NCHOLY
ITRa a T
a TH

ivLsENT
OBSCENE

ia E T C a a D

Pu Ol T a R IE

OGiA

Ai*.

F E N IT E N T I a a Y
I a a LP

PL o i j lL .1 S*w
SCUTTLE
1 -iii iJil.A-i I a Iv.

IRY
O LAp

i.iG'-E

IG G E R

a

TERRIFY

-PULSE
,1LET
\lSONEPi.
LN

:L vi.NUE

I va W

TEDIOUS
WELT
CFFAL
REBEL

N IG G a R D

£GE

r'iiiiiGiiI I I
iCORPlOK

DEEP
OFFENSIVE

SC a LD
P a E G N a NT
Na U S L a
a a UCOUS

TERRI bLE

JFT i a L

'a GF a I E
■KEhE R
.I S L a
.
'w/KIUaE

SC a F F O L D
P R E a CR
Na S T Y
Ra T T L E
Tah 'DRY

LUSS
POUT

T E N T a CLE

SICK
NUDE

CuRCR
I J Sw
iu EDO
a !E
KELLI ON

OAiUiiUU
MURKY
POUNCE

E P R u a CH

La U L

SC a V a NGL a
IOENa CE

TOLB
Pa Ra S ITE
SINISTER
MILDEW
PROP IIEEh
SKIMP
MISERY
T UTTER
PAUNCH
SCOUNDREL
A RECK
REVOLTING
FHGTi'iUDfcj
SCREECH
PERJURY
TRAMPLE
PEST
SEDUCE
I B a LD
P S Y C H Ia T a I S T
S L a SH
a IC T

a

■^
q

PUTRID

THEFT
i-ECOIE
N u RNia L

SHRIVEL
Pa N I a C
ObS'i iN a T E

REPRIMAND
Ii a S H
THREa TEN
Pa NSY

SICKNESS
NU ISa NGE
iEPTILE
T IT T E a

PREY
SILPER
JWEa GE a
E S T a IC T

SC a I E
Pa LSY
T l IE F

RELCRSE
NOTORIOUS
SHROUD
ItuAKIPU LitXJu
OBo&SoIOu
XEPa IS a L
M aS S aC al .

THROB
Pa a a L Y Z E

SIDLE
NUMB
R EPUG N a NT
TO a D

PRICK
SIMPLE
P«EDAL

i

R E T a L I a TE

OBNOXIOUS
SCOFF

OBESITY
SCOID

RENTa L

IvR a C IL E G w-'
T Ca Aa DO

TORMENT
P a S S ION
SINUS
L I L I T a RY

PROMISCUOUS
SKULK
Iu lN GE
TRAGIC
P a U PER

SCOURGE
u Ga L
rEVCLUlION
P a OV OXE
SGiE <i

PERSECUTE
TRa P

PESTILENCE
SEIZE
RIDICULE
PSYCHOLOGY
s

La T

RISK
PUSSY

PASSIVE
SISSY
MIRE
PROP a Ga NDa

SKULL
TuPwRENT’
P a T H L T 1C

SCORN
WRITHE
REVENGE
PROST I’
lUIE
oCO.. L
Pa .iN
T I va I T O a
IE R V E a T

SCUL
aE

V j LVLa
P i.0 . i’ L
c L a - i Dl R

AiGID
^UAKE
PURGE

p
ICIT
- a M IT Y
b r is

BEST
i Ga

SS

aT
IEx\IQ R a T ION
MP
IPE
a LECT
F , RN'a L

CKIE
Sa I*E
.NAT 1C
.SaPF 0 1 NT IViEivT
UTEi.
lSLrl^.GE
.Ix .Y

JB
JaBLE
ISI.-r».R
ia

BBY

a nous

XLOUY ION
OOT
OOS'iRx.
VOLUTION
ISTOL
ISTx\EsS
CNOPQLY
flj.
.
iOL I N a T E
jx\.DEr

tOnDY
/PILLPSY
HICRE
BHa P E L L S S
BivE a I F U L

..ORBID
uiGE
DRUNK
ILUG
SJa C U LA TE

OUST
SHIFTLESS
Lili-Jf
LUSH
RUMP
MOULDY
r a NKLE
LuiD
ROISON
SHOOT
*U ID

DE a RIE
IMBECILE
callous

LIFELESS
DEBT
INDECENT
Ca SKET
CRATE
DETEST
LIQUIDATE
GROSS
DI a FER
INJURY
DIE
LIZAHD
F a MINE
DIST a STE
LONELY
DISGUST
FAlLUhE
MOCK
iuiC Ia L
DISOBEY
SH a CK
OOZE
EXO m ET luN
LOUD
ROTTEN
EVIL
PITH
DIST r UST
MONOTONOUS
SH a MBL a S
DOSE
ORGAN
RUBBISH
ENVY
LUMP
SHARK
DREa M

IViORGUE
xuiGGED
DULL
PLUMP
EGOTISM
OUTCAST
SHIRK
DYNAMITE
LUSTY
RUPTUiuE
MOURN
RAPE
Iva D a M
POLICE
SHr a PNEL

^UEER

i.-n 1M
R U IH L r SS

MUCK

uj-iLi.Vjj.

OUTRAGE

S a C x iIF IC E

u&Ain

LMLJiJAW Vii

IMPa IR
Ca NNIB a L
LILY
DESTRUCTIVE
INFECTION
Ca TH a RTIC
GREEDY
DEVIL
LISP
GRUDGE
DIAPHRAGM
INJUSTICE
DINGY
F a Ta L
DIRTY
LOATHE
DISEASE
F a LSE
MISTRESS
SiUIT

IMPOTENT
CAT
LIMB
DETECTIVE
INFER IOxv
CHE a T
GRIe F
DIAGNOSIS
LIVER
GRUESOME
DICT a TE
INSECT
DIPHTilExaA
F a TTY
D ISa G r E r ^iLIE
LOCUST
DISCIi LINE

Ra c k e t
disorder

SHADOW
OPERATION
EXCOMiv.UiaCATE
LOUSY
ROUGH
E V a CU a T E

PITY
DIVORCE
MONSTRR
S liA i.E

Dr a GON
ORG a NIC
RUDE
EMOTION a L
LUN a TIC
SHATTER
DRE a RY
MORT a R
R a ID
DUMB
BLUNDER
EFFEMIN a TE
OUT r AW
SHOCK
DYSENTEnY
LYMPH
RUT
MOUSE
r a SC a L
M a DNESS
xrOLLUTE
SHREw
^U a VER
MUCOUS
Sa D
OVAi'.Y

fake

MOAN
iuiBBI
DISHONEST
SEXU a L
OINTMENT
EXTRAV a Ga NT
LOON
ROOK
EX a GGE1\a TE
PI xlaTE •
DISTORT
MONGREL
SH a KY
DOG
OPPRESS
xiOUTINE
ESPION a GE
LOUT
SHa NTY
DR a M a
MOR a SS
Ra D ICa L
DROWN
PLa GUE
EMB xxRR a SS
ORGY
SHERIFF
DUNG
LURCH
RUIN
MORTG a GE
ra

KE

LYNCH
Ma GGOT

POMPOUS
SHRIEK

QUMiDL
MUD
Sa G

OYSTExv

a GONY

Cobweb
DECOMPOSE
FERTILIZE
GUILT
IilSULT
ASLASSINATE
DECREEIT
FIEND
GUT
INTERFERE
IRRxTa BLE
a TROCITY
CONCEIT
FOOL
ca IL

DEFORMITY
iijiuGHTY
RuDLIT
'^;.SPiivx*CY
F J1 OjIGN
J.jRK
EEGRa DL

HEiiRTLESS
BinST
DEMa ND

vUNK
KNIFE
CORPSE
FRANTIC
blCTEn
DEPIiaVE
HIND
GP-mB
FRIGID
BLa ME
hOlUiUK
Ch-Pi'LE
xA'OE
DERISION
FUSS
Ga c H
CiiCOD
HURT
BOMB
Lim
CUCKOO
HYSTERICAL
DESPISE
BRIBE
ILLEGITIMATE
DxR ukGE

La WYEii.
GiiUDY
BULGE
lEi'iiOSY
BuiRK
GLUTTON
BUTT OCKS

CHOKE
DECAY
FAULT
GUILLOTINE
GURGLE
INTERCOURSE
COLLISION
DEFEAT
FIGHT
HAG
INTESTINE
APOLOGIZE
COMMAND
FLABBY
ITCH
DEFIANCE
HARLOT
AWKWARD
Ba RBa RIC
CONTAMINATE
FORGERY
JILT
DEL a Y
HELL
BEGGa R
DEMON
KIDN a P
CONTEMPTIBLE
F ORMa L
BELLY
DENOUNCE
DEPRESSED
HISS
CRASH
FROZEN
SPITE
HOSTILE
CRITICISM
larceny

DESERTER
GANGSTER
CROa K
HUMILIATE
BLIND
La RD
CRUCIFY
HUSSY
DESOLATE
BOWEL
IDIOT
CURSE
DxJViN
lazy

GEivM
BULL
LEWD
DEa D
GOODY
BUZZa RD

CLINIC
DECL a SE
FEAR
GUILTY
INSIST .
ARROGANT
COLON
DEFECT
FILTHY
HAIRY
INTIMIDATE
ASYLUM
COMMUNIST
FLATTER
JACKASS
DEFICIENCY
Ha RSH
Ba BOON
CONDEMN
FORCE
JEALOUS
DEFK a UD
DELINQUENT
HERETIC
BELCH
DENIZEN
KIDNEY
CONTROL
FOUL
BIGOT
DENY
HICK
CORRUPT
FR a UD
BL a CKMAIL
HOG
Ca a ZY
labor

DEPRIVE
FUNER a L
Ga PE
CROOK
HUMP
BLOOD
La s c i v i o u s
CRUDE
HYEN a
despair

Ba a G
IGNGa a NT
CYNICAL
DANGEu
LEECH
GHOST
BUa N
LIa R
gizzard

GOSSIP
Ca CKLE

OL1JMA.
DECEIVE
FEEBLE
GULCH
INSOLENT
ASS
COMBAT
DEFENSELESS
FIRE
HANG
IM
ATHEIST
COMPULSORY
FLESH
JAGGED
DEFILE
HATE
BALD
CONFESS
FORECLOSE
JEER
DEGENERATE
KEa RSE
Ba s t a r d
DELUSION
JOWL
KILL
CONVULSION
FR a ME
BILIOUS
DEPLORABLE
HIDEOUS
C OWa RD
FRIGHT
BLADDER
HONK
CRIMINAL
La GGa RD
DERELICT
FUNGUS
GARBAGE
CROTCH
HUNGER
BOa ST
lash

CRUEL
HYPOCRISY
DESPERATE
BIvaT
ILLEGa L
Da g g e r
La WLESS
Ga SP
BivUTAL
LEER
DANK
GLOOMY
BUTT

blFISH
ITUa L
EW
ERUM
LASH
HAN'TOM
•La VERY
’1LE
JL1KE
:a p
;ipe
[RiCKERY
jPURT
.1CKED
[UBERCULuSIS
rringy
JNL^La a CED
:>Au GG1SH

..HIMFER
TYPHOID
S^Ua NDER
UNFRIENDLY
STRUT
WHEEZE
STUPID
UNKIND
SLUSH
UNSTABLE
SMACK
USUKP
SMEAR
VERMIN
SQUEAL
VILLAIN
SUBMIT
VIRGIN
SUCCOR
SMUT
VU LGAR
SUCK
SNa KL
i»iiLACii*
STALE
SULK
SNOOP
«a 6TE
SOB
SURGERY
STERILIZE
SUSPICION
S»ja DDLE
STINGY
SOUR
SHELTER
STOOP
STORM
STRUGGLE
SPINSTER
Ta INT

SENSUAL
ROa CH
PULP
SEVERE
TREACHEROUS
PHYSIC
SLa Y
PILL
SLINK
WOMB
TREMOR
SPLINTEn
WINCE
ThIFE
STRICKEN
UDDER
SLOUCH
WHORE
TUMOR
SPUTTER
UNBEARABLE
tyrant

STUD
WENCH
SQUASH
UNJUST
SQUa T
UNTIDY
SQUAWK
VEIN
SQUEAK
VICIOUS
SUBMARINE
VIOLATE
SMOTHER
VIRTUOUS
SMUDGE
STa B
VULTURE
SNAKE
STAIN
WANTON
SUICIDE
SNOB
STa RVE
SNUB
SOCIALIST
STENCH
SURRENDER
SORDID
STING
SWa GGExv
SOUSE
STOOL
SN INDIE
SPERM
SWOLLEN
SPIT
tantrum

o&nuEtAMii
ROB
PUCKER
SEWER
TREASON
PIAZZA
SLEET
PIMPLE
SLIPPERY
WOLF
TRENCH
SPOIL
WILT
TROUBLE
STRICT
UGLY
SLOVENLY
WHIPPING
TWIST
Sq Ua BBLE
UNDERTAKER
TYPHUS
STUFFY
WEAK
SLUM
UNSAVORY
STUPOR
URINE
STUTTER
VENGEFUL
SUBDUE
VICTIM
SMIRK
VIOLENCE
SQUIRM
VOID
SQUIRT
SUCCOMB
SNa IL
STa GNANT
SUFFOCATE
warning

SNICKER
STiiRK
Wa RP
STEa L
STEALTHY
SOCKET
SOGGY
STIFLE
SORE
SWEAR
STOLEN
SPa SM
SWINE
SPIDER
SYPHILIS
STRa NGE
TARNISH

UUAU iiiU
RODENT
PEWTER
SLAUGHTER
TREMBLE
PIG
SLICE
PINCH
SLOPPY
WITCH
TRESPASS
SPUHN
WILDLY
TRUSS
STRIKE
ULCER
SLUG
WHINE
TWITCH
SQUa LQR
UNEASY
STRUGGLE
WHELP
STUN
UNGRa TEFUL
SLUMP
UNSCRUPULOUS
SLY
URN
SMASH
VENOM
SNELL
VICTUAL
SQUEAMISH
VIPER
SUCKER
SUCCULENT
VOMIT
STa GGER
SUFFER
WADDLE
SNEER
STAMMER
SULLEN
WASP
SUPERIOR
SUPPRESS
SODDEN
STICKY
SWa B
SORROW
STINK
SWEAT
SPATTER
STRADDLE
SPINE
STRa IN
strangle

STRETCHER

look over the f o l l o w i n g words a n d select the five words wh i c h
ke l e a s t .
M a r k a 1 in front of the w o r d you like least of all*
2 in f r o n t of the w o r d y o u l i k e n e x t le a s t , a 3 in f r o n t of t h e
?rd y o u
ie f i v e
5r y o u *
fped o n

l i k e t h i r d least,, a n d so o n
words y o u like least in the
Liake y o u r c h o i c e s f r o m a l i
t h e l i s t a s w e l l a s t h o s e ’"

DVERTT
;n g r e n e
IGG^RD
ROSTITUTE
IEND
uS'h*RD
ILL
RCTCH
YPCCRISY
iiFiCikSS
n s ^n e
iPILLPSY
:c r g u e
;KEu.PNEL
/ikLr;Ii-rkL
;REEDY
JUDGE
jCUSY
JISii.GiiEEjvDLE
uYKCH
SLAVERY
USURP
SURE
SLUI.I

OGRE
HELANCHOLY
NASTY
OBESITY
GUT
BRAT
BRUTAL
ILLICIT
LICE
GRIPE
DISGRACE
DRUNK
R<*.PE
DEATH
DESTRUCTIVE
DEVIL
Fi.TAL
LUNATIC.
DISHONEST
SHRIEK
YAP
TAINT
SPIT
SYPHILIS

until you have indicated
order of their unpleasantness
the words listed below, those
i^eographed* Ignore any word

RAVAGE
TORTURE
PUTRID
NIGGER
CORPSE
LL.G
BELLY
HUSSY
LuZY
CRti-ZY
hob

POISON
SHOOT
ROTTEN
INJURY
DISASTER
DIRTY
POLLUTE
ORGY
SELFISH
SLIKE
SPITE
TREASON
VOMIT

MALIGNANT
SICK
MAN GE
F.GONY
LEPROSY
FLABBY
CROAK
IDIOT
ARROGANT
ASS
EXECUTION
INDECENT
HOCK
CASKET
DIE
EVIL
DIVORCE
CHEAT
DUNG
TR0.SH
TRICKERY
wkope

SLOVENLY
STINK

you

Appendix 5
(Reproduction of a Checked List)
Indicate the eight words you like 1 east from this list
by putting a check in front of each one you select:

^

REV,API)

BIDLER

VACATION

ATTRACTIVE

CHUBBY

HELPFUL

ADMIRE

SPORTS

PCX'! BALL

______
FE :XER

INTERESTING

PRIZE

cc.;sx p r a t e

gaye

SKILLFUL

HUEOROUS

G HU HP CCP
A PPL,-.;rD
C* I F ’ O

>- P'

G ' ' T*

*>+ ^ -w

^ GOGGLE

Ai usr ng

SINCERE

A EIGHTY

1/ TA-.ISH

APPROVAL

ENTHRTAI KING

? ICE

:o r i i a l

' h * ’ " T

. , 1.

L

j

I1-LI CIGUS

XERCAT

(The underlinings have been employed here to indicate which
of the words were the control words).

Lest neme

Doe

,

Sex (M or F)

SLSMS

'

Please reed these instructions carefully:
Look over the following*words and select the five words you like
least. Mark the number; 1, in front of the one word you like least
of all the words. Next, mark a 2 in front of the one word you like*
next least. Now mark a 3 in front of the word you like third least,
and so on until you have indicated the five words you like least, in
order of their unpleasantness for you. You may find them unpleasant
because of their appearance, their sound, whet they represent or
remind you of, or*for any number of these or other reasons. When
you have finished, you should have five, and only five words marked,
each with a different number. An explanation for this procedure
will be supplied you later in the course.
AGONY
GASH
BELLY
FILTHY
ADULTERY
POVERTY
MALIGNANT
SCORPION
MANGLE
SHRIVEL
NAUSEA
WRITHE
INCEST
INSANE
RABBLE
EPILEPSY
MAIM
OGZE
RUPTURE
CANNIBAL
DIRTY
SHAME
DYSENTERY
GRUESOME
DUNG
TRASH
TUBERCULOSIS
TYPHOID
VERMIN
SNOOP
WENCH
SPIT
TYPHUS
SYPHILIS
SLOPPY
VENOM
WALLOW
TRIPE
GAUDY
HAG
TOILET
PANIC
EJACULATE

FIEND
LEPROSY
HUSSY
■■■i
ASS
MUTILATE
MANURE
MURDER
MAUL
MANIAC
•v NIGGER
PERVERT
CARCASS
FANATIC
DISMAL
MORBID
DEARIE
EXCRETION
RAPE
INFECTION
LOATHE
LUNATIC
POLLUTE
FATTY
MAGGOT
SLAVERY
STRINGY
WHEEZE
SMUT
ROACH
SUICIDE
PUNY
SLUM
PUNK
WITCH
.SUCKER
SNOB
UGLY
BUTTOCKS
BILIOUS
MESS
NASTY
RUTHLESS

GUT
ABORTION
BO’VEL
BELCH
CROTCH
OGRE
GANGRENE
SCAB
PUTRID
OBNOXIOUS
MASSACRE
SCUM
DETERIORATION
DISGRACE
SHABBY
RUMP
CASKET
ROTTEN
OUEER
CATHARTIC
DISEASE
DREARY
MUCOUS
LOUT
PURGATORY
SLIME
SLUGGISH
STUPID
VITLGAR
SLINK
STENCH
SEWER
URINE
RODENT
’T U NE
- VOMIT
SOUAT
STAGNANT
CROAK
SALOON
'TRETCHE D
DEBRIS
SALIVA

ATROCITY
FLABBY
LEWD
LEECH
BRAT
TERROR
TORTURE
SCALY
NIGGARD
PUSSY
SISSY
LICF
GRIPE
FAIRY
ROWDY
MOULDY
MOCK
MORGUE
DEATH
GRFEDY
LOUSY
EFFEMINATE
CHEAT
^RGY
SELFISH
WICKED
WHIMPER
USURP
SUCK
WHORE
SOUSE
SLOVENLY
SMIRK
PIG
WHELP
PEW
SLOUCH
ULCER
HELL
ODOROUS
QUACK
DISTRESS
FAILURE

Last Uame

Please do not fill this slip out unless you took part in
"vision experiment" in which you Identified 1? words for
of the Psychology Building on South Campus, and also had
list of unpleasant words, with the numbers from 1 to 5.
in loth parts of the experiment, please answer yeB or no
tions as honestly and carefully as possible,

, First Name

tho tachistoscopic
Mr. Sterne in Room 18
previously rated the
If you did take part
to the following ques­

1, ?id you know that the vision experiment was connected with the ratings, before
you saw any of the words in the tachistoscope? ______________
ifnot,
2, Lid you have any idea that the two might be connected before you saw the v/ords
in the tachistoscope?_________________ If not,
3, Lid you realize that the two were connected after seeing the first stimulus
(unpleasant) word? ____________ If not,
•d. lid you realize that the two were connected after seeing both stimulus words?

Ccm-entsi

REMOVE TRIG FACrP A*L HA>tL BALK TO YOUR IM3 II UCTCR, IF YOU HAVE AlTShrZREL TIL.
M U E S L I O 'T S ,

It has bet. n suggest, a that most individuals learn early in life that cer­
tain wcras carry with then, a "socially taboo*' connetaticn. When these words
are used by a child, this use generally results in chastisement by parents, with
the consequent establishment of a conditioned reaction tc the verbal symbol.
Inis pattern of conditioned emotional response n.ay be considered to be crte of
anxiety or fear aroused by symbols having sexual, excretory, or otherwise un­
pleasant or "immoral” connotations. (lv.cGinnies, Elliott, "Emotionality and lercoptual Defense", Psychological heview. 56, 1949, p.244.)
Please rate the fallowing words in terms of the <xtcnt to which ».ach word
s< e.m.s "socially taboo" by putting a check in the appropriate space afti i it.
I

Word
bastard
Fatty

. . .h -

Fairy
"queer

kiggur

^

Dearie

teW
Tung"
Ooze

Lffeminate
Suck
oir.ut
ASS
belly

Syphilis
vjIime
*-

aelch

uiIT
Stench
’aausea’
ScaB—
Tripe
Slavery
Lepr#sy
Suicide
Putrid
"babble

~T

Highly
tab#*

Moderately
taboo

SIightly
tabo#

Sot
labo#

i
-t

Highly
taboo

Word
i
foiangle
Ljaculatc
otTnk
Vornit
bilious
Lynch
Divorce

Moderately
tabo*

Slightly
taboo

N«t
Tabo#

j

--------------

_

..

.

..

------------- --

Ei tch

Sn*b
Tuberculosis
tlahc
ka i m
Whore
Prostitute
War
fciVil
Hate
Shrapnel

|

KiH

;

Disaster
Lice
Sluiri

!

j
!'
I
1
j

1

D eath

Morgue
Lotion
L'-isgrace
1 orture
Lie
Cannibal
1 le ncT
Dishonest
Lrutul
Poverty
"Lxe cution
Pcison
Disagreeable

j

1

i
I
i

:
--------- ----- ---- ---

- - ... ..... --------■ -----