An Effect Of Verbal Conditioning On Nonverbal Behavior

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content This d issertation has been 64— 9630
m icrofilm ed exactly as received
ZISKIN, Jay H ersell, 1920-
AN E FFE C T OF VERBAL CONDITIONING ON
NONVERBAL BEHAVIOR.
U n iversity of Southern C alifornia, P h.D ., 1962
P sychology, general
University Microfilms, Inc., Ann Arbor, Michigan
AN EFFECT OF VERBAL CONDITIONING
O N NONVERBAL BEHAVIOR
by
Jay H ersell Ziskin
A D isse rta tio n Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In P a rtia l F ulfillm en t of the
Requirements fo r the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
June 1962
UNIVERSITY O F SOUTHERN CALIFORNIA
GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES 7. CALIFORNIA
This dissertation, written by
JAY BEFELL ZISKIN
under the direction of hL&—Dissertation C o m
mittee, and approved by all its members, has
been presented to and accepted by the Dean of
the Graduate School, in partial fulfillment of
requirements for the degree of
D O C T O R O F P H I L O S O P H Y
Dean
D a t e V 1 , / .< .............
DISSERTATION COM M ITTEE
V ^ \
r Chairman
/
DEDICATION
To Father and M y Wife
A C K N O W LED G M EN TS
The completion of a doctoral program f i t s e a sily
in to one of mankind’s c la s s ic a l m etaphors--that of a
t r a v e l e r ’s progression along a road. For the tr a v e le r ,
the road seems arduous, sometimes fraught with doubts
and indecisio n, and often lonely. However, re tro sp e c tio n ,
which becomes p o ssib le near the end of the road, reveals
th a t th ere were some people who accompanied the tra v e le r
a l l the way, and other people who appeared more b r ie f ly
to provide some needed aid or comfort at points where the
road seemed d i f f i c u l t or impassable. While there were
many of the l a t t e r , pure fe elin g d ic ta te s the naming of
a few--Dr. Milton M etfessel, Dr. Herman Harvey and the
p ro fessio n al s t a f f of the Los Angeles S tate College
Counseling Center--whose encouragement came at what seemed
to be c ru c ia l times. The members of my d is s e rta tio n
committee made completion of the la s t p art of the journey
p o ssib le, and I wish to thank them for competent advice
and co n stru c tiv e c r itic is m , and p a r tic u la r ly for th e ir
finding time to help when such time could only unreasonably
be hoped fo r. For me, as for most candidates, the
iv
constant companions were my family, including my fa th e r
and my bro ther, as well as my wife and ch ild ren , a l l of
whom shared the hopes and doubts, the gains and the
fr u s tr a tio n s , and most im portantly, supplied the f a ith
and confidence which nourished the sometimes flagging
w ill to follow the road to i t s term inal. Thus, a t the
end of the road, the tra v e le r becomes aware th a t,
as on any other road, people were always there to make
the journey possible and meaningful.
TABLE O F CONTENTS
Page
DEDICATION ............................................................................ i i i
ACKNOW LEDGM ENTS ................................................................... iv
LIST OF TABLES ................................................................... v i i i
Chapter
I. BACK GROU ND OF THE STUDY ............................... I
The Role of Language in Behavior
Language and Psychotherapy
Language and A ttitudes
Summary of Background
II. THE PROBLEM ..................................................... 8
Survey of the L iteratu re
Summary of Review of L ite ra tu re
Rationale of Present Study
Hypothesis
I I I . M ETH O D .............................................................. 25
Subjects
M aterial
Procedure
IV. RESULTS .............................................................. 37
Conditioning of Meaning
Effect of Conditioned Meaning of Labels
on Choices of Blocks
V. DISCUSSION ......................................................... 57
v i i
Chapter Page
T h eoretical Im plications
Im plications fo r Psychotherapy
Im plications fo r A ttitu d e s
Suggestions fo r Further Research
VI. SU M M A RY AND CONCLUSIONS............................... 81
Method
Results
Conclusions
BIBLIOGRAPHY ......................................................................... 85
APPENDIXES
93
LIST OF TABLES
T able Page
1. Comparisons of Semantic D iffe re n tia l Ratings
and Syllables Receiving Conditioning of
Pleasant (+), Unpleasant (- ) , and Neutral
Meaning........................................................................ 38
2. T heoretical D istrib u tio n of Group Scores
for 12 Groups of 6 Subjects Where
the Expected Score Is 1.0 42
3. Kolmogorov-Smirnov One Sample Test for
Observed and T heoretical D istrib u tio n s of
Choices of "Pleasant" Blocks by 12
Groups of 6 S u b j e c t s ........................................... 44
4. Kolmogorov-Smirnov One Sample Test for
Observed and T heoretical D istrib u tio n s of
Choices of "Unpleasant" Blocks by
12 Groups of 6 Subjects ........................... 45
5. Wilcoxon Matched-Pairs Signed-Ranks Test
for Choices, on F irs t or Second Choice,
of "Pleasant" and "Unpleasant" Named
Blocks by 12 G r o u p s ........................................ 47
6. Chi Square for Choices of "Pleasant" (+),
"Unpleasant" (- ) , and "Neutral" (0)
Blocks by Subjects Having Possible
Overlearning T ria ls (NOL) .............................. 48
7. Chi Square for A ll Subjects for Frequencies
of Subjects Choosing "Pleasant" (+),
"Unpleasant" (-) and "Neutral"
(0) Blocks .......................................................... 50
v i i i
ix
T able Page
8. Chi Square for Choices of "Pleasant"
(+), "Unpleasant" (-) , and "Neutral"
(0) Blocks by Subjects Who Met C rite ria
of Both Conditioned Meaning and Block
Name Learning ................................................. 52
9. Chi Square for Choices of "Pleasant" (+),
"Unpleasant" (- ) , and "Neutral" (0)
Blocks by Males and Females .................... 54
10. Chi Square for Choices of "Pleasant" (+),
"Unpleasant" (- ) , and "Neutral" (0)
Blocks by Subjects Having the Semantic
D iffe re n tia l before Choosing (SDB)
and Having the Semantic D iffe re n tia l
A fter Choosing (SDA) ............................... 55
11. Chi Square for Frequencies with Which
Blocks Were Chosen from Various
Positions in the Bucket ........................... 56
12. Chi Square Analysis for Frequencies with
Which Subjects Chose Blocks Whose Names
They Had Rated Pleasant (+), Unpleasant
( - ) , or Neutral (0) on the Semantic
D iffe re n tia l 61
CHAPTER I
BACKGROUND OF THE STUDY
This study proposes to in v estig ate the e ffe c t of
verbal or semantic conditioning on the voluntary, non
verbal behavior of people toward objects or s itu a tio n s
in the environment. Concepts of m ediational processes,
within a framework of b e h a v io ristic theory, provide a
basis for the b e lie f th at verbal manipulation of the
meaning of verbal labels can be employed to modify non
verbal behavior toward the object sig n ifie d by the lab el.
The study o rig in ates from three following currents of
psychological thought: (1) the re la tio n sh ip between
language and other behavior, (2) attem pts to r e la te
learning theory and psychotherapy, and (3) a ttitu d e s .
The Role of Language in Behavior
A growing in te r e s t in the area of psycholin
g u is tic s is re fle c te d by increasing reference to language
processes in psychological l i t e r a t u r e . For example,
1
2
Krasner (1958) and Salzinger (1959) have re cen tly pub
lished reviews of the expanding body of research in the
area of verbal conditioning. The stu d ies reviewed show
th a t the production of classes of verbal responses is
amenable to m odification through conditioning techniques
employing verbal responses of other persons as the
reinforcem ent. The classe s of verbal responses for which
conditioning has been demonstrated include p lu ra l nouns,
verbs, pronouns, Rorschach responses, acceptance of s e l f ,
persons, anim als, and opinions.
The importance of the study of language is
fu rth e r supported by such th e o ris ts in psychology as
Mowrer, Kendler, H ull, Dollard and M ille r, Razran, and
Osgood among o th e rs .
Mowrer (1954), for example, discusses the issue
of language in psychology p rio r to the 1930's. He
points out,
This neglect of language is esp e c ia lly remarkable
when one considers how much of the waking l i f e
of human beings is spent in ta lk in g , lis te n in g ,
w ritin g , reading or using the subvocal equivalents
thereo f in thinking.
He s ta te s th a t neobehavioristic developments have pro
vided th eo ries and methods for incorporating such
3
subjective events as ’’meaning" into the generally behavior
i s t i c mainstream of psychology, and he emphasizes the
importance of mediation concepts in the development of
the language area.
Kendler (1961) s ta te s ,
Another healthy lin e of research produced by
consideration of the mediating function of
language in concept formation consists of
studying the effect of associating verbal
responses to stim uli th at w ill la t e r be used in
concept formation tasks. The theory is that
verbal responses become covert and as such
serve to mediate overt responses.
Hull (1943) discussing secondary generalization
concludes:
. . . the common sense notion of sim ila rity is
based upon the presence or absence of primary
generalization gradients, whereas the so-called
log ical or ab stract s im ila r itie s and differences
a r is e from secondary, learned, or mediated sim i
l a r i t i e s , p a rtic u la rly those mediated by verbal
reactions.
Dollard and M iller (1950) devote a chapter to the
ro le of words and sentences in arousing drive, mediating
rewards, and producing fo resig h t. They point out the
great economy of the use of words as cue-producing
responses such th at changing one verbal response to
another can a l t e r a large number of complex instrumental
and emotional responses.
4
In a recent review of Soviet psychophysiology,
Razran (1961) demonstrates the emphasis on semantic con
d itio n in g w ithin the Pavlovian th e o re tic a l framework of
second or higher-order signaling systems. The research
provides considerable evidence fo r the conclusion th a t,
in the determ ination of human responses, i t is the semantic
or meaning c h a ra c te r is tic s of language, ra th e r than the
phonetographic or physical c h a r a c te r is tic s , th at is
dominant.
The foregoing references make c le a r the importance
to the advancement of psychology of in v estig atio n s into
the re la tio n sh ip s between symbolic processes and other
types of human behavior.
Osgood (1952, 1953, 1957) has given impetus to
such in v estig atio n s by formulating an approach to meaning
which elaborates on H u ll's concept of the a n ticip ato ry
and cue-producing response and in which verbal mediation
plays the c e n tra l ro le . The development of the semantic
d i f f e r e n ti a l by Osgood, Suci, and Tannenbaum (1957) has
provided an operation for the measurement of established
dimensions of meaning. The ra tio n a le for the present
study is based on Osgood's formulation of meaning as a
re p re se n ta tio n a l m ediational process.
5
Language and Psychotherapy
The major medium in almost a l l current forms of
psychotherapy is language--the verbal productions of the
c lie n t and th e ra p is t. The th e ra p is t, depending on his
p a rtic u la r "school," attempts to e l i c i t various types of
verbal behavior, and to respond verbally to the verbal
behavior of the c lie n t in various ways in the b e lie f th a t,
as a re s u lt of these productions and responses, the
personality and/or behavior of the c lie n t w ill be changed
in some d esirab le d ire c tio n . Attempts to in v e stig a te the
e ffe c ts of psychotherapy have fa ile d to provide convincing
evidence th at such change does in fa c t occur. This
f a ilu r e has led Winder (1957) to summarize one view of
psychotherapy, as follows: "One lin e of thought would
consider psychotherapy as a complex and not always e f f i
cien t conditioning process in which the p atien t learns to
ta lk d iffe re n tly and l i t t l e e ls e ." A demonstration th at
nonverbal behavior can be modified through manipulations
involving only verbal m aterial would provide a vehicle
from which i t might be in ferred th a t the verbal changes
produced in psychotherapy may have broader consequences.
6
Psychotherapy is commonly viewed as a learning
process, although the degree to which learning theory
concepts and data are e x p lic itly or im p lic itly incorpor
ated by d iffe re n t "schools" v aries g re atly . Bandura (1961)
has summarized attempts to bring learning concepts such
as counter-conditioning, e x tin ctio n , and discrim ination
learning, to bear d ire c tly on the psychotherapy process.
Demonstration th a t behavior m odification as a re s u lt of
conditioning of verbal meaning is p redictable from
Osgood's "mediation" hypothesis, would suggest the
p o te n tia l usefulness of e ffo rts to bring therapy into
the framework of systematic learning theory.
Language and A ttitudes
Osgood (1953) s ta te s th a t g eneralizatio n from
verbal sign to the object represented by the sign is of
great so cial importance because reactions (a ttitu d e s )
associated with verbal labels tra n s fe r to the so cial
objects represented. Sim ilarly, S taats and Staats (1958)
have attempted to show how a ttitu d e s may be established
through conditioned meaning attached to verbal lab els.
Direct evidence that conditioning of verbal labels gener
a liz e s to the object represented, which so fa r has been
7
lacking, would lend weight to such conceptualizations
regarding the formation and function of a ttitu d e s .
Summary of Background
Psychology shows increasing in te re s t in the ro le
of language in behavior. One lin e of approach u t i liz e s
concepts of mediated g eneralization in which verbal
responses or im p licit or subvocal or fra c tio n a l components
of verbal responses play a key ro le as cue-producing
responses which a ffe c t behavior.
Language is the p rin c ip le medium of psychotherapy.
Several attempts applying learning theory concepts to
psychotherapy theory and technique have emphasized the
ro le of verbal mediation.
T heoretical and experimental publications have
attempted to apply learning theory and verbal mediation
concepts to the study of a ttitu d e s .
This current and widespread in te re s t in language
suggests the relevance of an in v estig atio n of the e ffe c ts
of conditioning of verbal meaning on nonverbal behavior
toward objects represented by the verbal labels conditioned.
CHAPTER I I
THE PROBLEM
Survey of the L ite r a tu r e
Among the e a rly research es on th e phenomenon of
g e n e ra liz a tio n were those of Pavlov (1927), and Anrep
(1920) dem onstrating th a t a s a liv a ry response conditioned
to s tim u la tio n of a p oint on th e skin of a dog would
occur w ith decreasing magnitude as the point of s tim u la
tio n was f u r th e r removed from the p oin t of o r ig in a l con
d itio n in g . Bass and Hull (1934) reproduced t h i s phenomena
w ith human s u b je c ts . Hovland (1937a, 1937b) was a b le
to show th a t GSR conditioned to a p oint on the p itc h or
loudness continuum would occur to o th er p oints on the
continuum, but w ith decreasing magnitude as th e d iffe re n c e
between the co nditioned and th e g en e ra liz e d stim ulus
in creased . While not d i r e c t l y re le v a n t to the p resent
study, because they deal only w ith p h y sical stim ulus
s i m il a r it y as the b a sis fo r g e n e ra liz a tio n , th ese stu d ie s
re p re se n t the h i s t o r i c a l background fo r in v e s tig a tio n s
8
9
o f g e n e r a liz a t io n b ased on o th e r d im en sion s o f s i m i l a r i t y .
Of more d ire c t relevance to the present study,
are those researches on the phenomena of semantic or
mediated g e n e ra liz a tio n . This research may be c l a s s i f i e d ,
fo r convenience of discussio n, in to stu d ies involving
autonomic responses, verbal responses, and voluntary,
nonverbal responses, su b c la s s ifie d in to attem pts to
demonstrate g e n e ra liz a tio n from object to sign, from sign
to sign, and from sign to o b ject.
Studies Involving Autonomic Responses
G eneralization from object to sign was demonstrated
by Kapustnik (1934) who estab lish ed conditioned sa liv a ry
reactio n s to v isu a l and au ditory stim u li and obtained
s ig n ific a n t amounts of g e n e ra liz a tio n to the verbal signs
for the o rig in a l cues. Traugott (1934) found th a t in h ib i
tio n conditioned to a blue lig h t g eneralized more to the
word “blue" than to a red lig h t or the word "re d ."
Kotliarevsky (1935) conditioned a cardiovasomoter re fle x ,
pulse re ta rd a tio n , to the sound of a b e ll and found gen
e r a liz a tio n to the word " b e l l ." Metzner (1942) obtained
s im ila r r e s u lts with the eyelid clo su re re fle x . Markosyan
(1958; c ite d in Razran, 1961) found th a t blood
10
coagulation responses conditioned to the sound of a
metronome and the flash of a lamp tra n sfe rre d to both
phonetographically and sem antically re la te d words.
Several studies have employed autonomic response
measures in in v estig atio n s of g en eralizatio n from a verbal
sign to sem antically re la te d verbal signs. Razran,
employing the saliv ary response has shown: th a t s a liv a
tion was g re a te st to the v isu al stim ulus, the word
"saliv a" in his nativ e language, Russian, next g re a te st
in his most p ro fic ie n t one, English, and less in three
s lig h tly known languages (1936); th a t there is g reater
gen eralizatio n to sem antically re la te d words than to
phonetographically re la te d words (1939); th at there are
semantic gradients to tr a d itio n a l asso ciatio n categories
such as subordinates, c o n tra sts, part-whole, w hole-part,
coordinate, and supraordinate (1949a, 1949b, 1949d, 1949e);
th at there is more tra n s fe r from the sentence "Poverty is
degrading" to "Wealth is u p liftin g " than to "Poverty is
not degrading" (1949c).
In his review of current Soviet Psychophysiology,
Razran (1961) c ite s the following stu d ies. Volkova (1953)
found th at salivary responses conditioned to the word
11
"well" or "good" generalized to sentences evoking in the
subject a ttitu d e s of approval. Shvarts (1954t 1960) con
d itio n in g v aso co n strictio n found g re a te r tra n s f e r to
sem antically re la te d words of a second known language
than to phonetographically re la te d words of the n ativ e
language, with a re v e rsa l of th is e ffe c t when the "higher"
lev els of functioning were depressed by adm in istration
of the drug, c h lo ra l hydrate. The congruity of these
findings with the previously quoted statem ent of Hull
regarding primary and secondary g e n e ralizatio n seems
apparent. Vinogradova and Eysler (1959) report a number
of findings showing magnitude of vasomotor change as a
function of degree of semantic re la tio n s h ip s of words,
but the number of subjects was small and no s t a t i s t i c a l
treatm ent of the data is given so the findings do not
appear conclusive. Luria and Vinogradova (1959) present
evidence th a t semantic g e n e ralizatio n of vascular
responses is re la te d to in te llig e n c e in a study comparing
normal ch ildren with child ren of varying degrees of
feeblemindedness.
In an e a r lie r study, Reiss (1946) employing the
GSR also found evidence showing a re la tio n s h ip between
12
in te lle c tu a l capacity and g eneralization of semantic con
ditio n in g by comparing subjects of d iffe re n t age groups,
with more g en eralizatio n being shown by the older groups.
Lacey and Smith (1954) conditioned heart ra te
changes to the word "cow" for one group and "paper" for
another, and found th at the group conditioned to "cow"
showed s ig n ific a n tly g reater heart r a te changes to semanti
c a lly re la te d words of ru ra l connotation such as "h arv est,"
"g rain ," and "haystack." They concluded th a t semantic
gen eralization of unconscious anxiety had occurred.
K e lle r's (1943) study is a ra re case of negative
evidence regarding semantic g e n e ralizatio n . He conditioned
GSR to a p ictu re of a boy scout hat and found g en e raliza
tion to a p ictu re of a firem an's hat but not to the word
"h a t." In a la te r study, Branca (1957) found th at GSR
conditioned to the word "hat" generalized to a p ictu re of
a h at. However, his re su lts do not appear conclusive
because a f te r elim inating a l l subjects who did not show
i n i t i a l conditioning, he had only eight subjects l e f t ,
of whom only four showed g re ater gen eralizatio n of GSR
to the t e s t stimulus than to other stim u li.
Lipton and Blanton (1957) found th at meanings of
i n i t i a l l y meaningless words could be manipulated
13
co n sisten tly by experimental procedures whether measured
by GSR or semantic d i f f e r e n t i a l , suggesting th a t the
responses used in the semantic d i f f e r e n t i a l technique are
re la te d to autonomic responses. The above authors also
found th a t differences in conceptual a b i l i t y , as measured
by the Hanfman-Kasanin Test, and lev el of learning were
not re la te d to level of g eneralizatio n measured by e ith e r
method.
P h illip s (1958) established a scale of mediated
verbal s im ila rity by teaching subjects Turkish names for
five d iffe re n t shades of gray which were eq u id istan t on
a color chart by v irtu e of being 100 JND's a p a rt. Subjects
were then conditioned with a loud tone to the word given
as the asso ciate to the dark extreme of the scale. GSR
was then measured to the fiv e words and showed a con sisten t
decrease according to scale position of the words except
for an upturn in the curve a t the word for the lig h te s t
shade which P h illip s explains on the p rin cip le of asso cia
tion of opposites. She in te rp re ts her re s u lts as showing
that semantic s im ila rity can mediate autonomic responses.
She supports Osgood's (1953) position th at study of the
mechanisms determining degrees of s im ila rity among media
tions processes is necessary for f u lle r understanding
14
of semantic g en e ralizatio n .
There is l i t t l e experimental evidence on tra n s fe r
of autonomic responses from verbal sign to object.
Kapustnik (1934) demonstrated th at saliv a ry reactions
conditioned to verbal signs representing auditory and
visu al stim uli tran sferre d to those stim u li themselves.
However, the g en e rality of h is findings is lim ited as
he used only four su b jects, ten years of age.
Studies Involving Verbal Responses
Several studies have investigated verbal mediation
employing verbal responses as the measure. In in v e s ti
gating the ro le of mediated g en eralizatio n in i n t e r - l i s t
tra n s f e r, Foley and Cofer (1943), Cofer, Ja n is, and Rowell
(1943), and Foley and Mathews (1943) employed the r e c a ll
method with homonym, synonym, and antonym l i s t s , and
found s ig n ific a n t tra n s fe r e f fe c ts . Peters (1935),
Bugelski and Sharlock (1952), and Norcross and Spiker
(1958) demonstrated th at verbal mediated association can
produce both f a c i l i t a t i o n and in terfe ren ce in paired
asso ciate learning.
Scott (1957) using a seven point ra tin g scale
found th at subjects sh ifted th e ir a ttitu d e s on
15
co n tro v e rsial Issues from th e ir I n i t i a l position to one
c lo se r to the opposite a f te r receiving so c ia l approval for
verbally presenting arguments contrary to th e ir o rig in a l
p o sitio n. Subjects who did not receive such reinforcement
did not show th is s h i f t .
Eriksen, Kuethe, and Sullivan (1958) attempted
to determine whether a response bias acquired without
verbal awareness would generalize and influence responses
in a s itu a tio n d iffe re n t from the learning s itu a tio n .
They used reinforced and nonreinforced nonsense sy llab les
which the subjects learned as names for d iffe re n t size
blocks and also as responses in an ESP task . Their
an a ly sis, in which both block naming and ESP responses
were pooled, showed th a t the subjects used the reinforced
sy llab les more often than the control s y lla b le s. However,
these re s u lts do not j u s t if y the autho rs' conclusion th at
size judgments had been influenced by the verbal condi
tioning procedure. The fa ct th a t the reinforced verbal
response occurred more often is ju s t what one would expect
from the studies reviewed by Krasner (1958) which demon
s t r a t e th a t the emission of various classes of verbal
responses can be increased by reinforcement of such
responses.
16
Nuthman (1957) was aware of the above type of
lim ita tio n . She conditioned the c la s s "acceptance of s e lf "
responses on a tr u e - f a ls e p e rso n a lity t e s t , and found
th a t verbal reinforcem ent co n sistin g of the word "good"
re su lte d in an increase in "self-accep tan ce" responses.
However, she in te r p r e ts these r e s u lts only as showing an
increase in the c la ss of verbal responses, and not as
evidence of change in a c tu a l self-accep tan ce.
S taats and S taats (1958), attem pting to e s ta b lis h
a ttitu d e s by c la s s ic a l conditioning techniques, paired
n a tio n a l names, "Swedish" and "Dutch," w ith words of
p o s itiv e or negative evalu ative meaning fo r one group,
reversing the valences for the other groups. The p re
d icted d ifferen ces in ra tin g these n a tio n a litie s on the
semantic d i f f e r e n t i a l were found. The authors concluded
th a t conditioning of a ttitu d e s toward n atio n al groups by
a sso cia tio n of the names of the groups w ith words of
p o sitiv e or negative ev aluativ e meaning was demonstrated.
The authors conceded th a t the measure employed does not
d ir e c tly show th a t the s u b je c t's behavior toward the
o b ject, a Swedish or Dutch person, has been changed
because the semantic d i f f e r e n t ia l can only show change
17
to the sign, the n atio nal name. However, they c i t e the
previously mentioned study of Kapustnik as evidence th at
a response previously conditioned to the name of an o b ject,
generalizes to the ob ject. This is tenuous support for
th e ir conclusions because in addition to r e s tr ic te d popu
la tio n of Kapustnik*s study, the finding of g eneralization
of an autonomic response from sign to object, does not
n ecessarily j u s t if y the conclusion th at voluntary, overt
behavior w ill occur as a re s u lt of g eneralization from
sign to o b ject. Another lim ita tio n of the Staats study
derives from the fa ct th at while the p o sitiv e ly conditioned
group showed a s ig n ific a n tly more favorable ra tin g to the
p a rtic u la r n atio n al name than the negatively conditioned
group, both groups rated the name on the p o sitiv e side
of the semantic d if f e r e n t i a l . The re s u lts of the Staats
study should not leg itim ately be generalized beyond a
demonstration for p o sitiv e a ttitu d e s .
Eisman (1955) sought to demonstrate a ttitu d e
formation, the development of a color preference, through
mediated g en eralizatio n . Her re s u lts supported the
hypothesis th at i f individuals are taught to attach the
same label to two stim u li which d if f e r from one another
18
in a l l relev an t ways, and i f these in dividu als subsequently
have rewarding experience with one of these s tim u li, then
a p o s itiv e preference for the other stim ulus w ill be
esta b lish e d . Her formulation involved a chain of mediating
reactio n s going from object to lab el to another object
id e n tifie d by the same la b e l. The g e n e ra lity of her
findings is lim ited by the age range of her su b je c ts,
six to nine years of age. She employed a verbal response
measure, statem ent of preference, so her re s u lts do not
c le a rly demonstrate th a t the verbalized preference would
be re fle c te d in nonverbal approach behavior. There is
a lso the p o s s ib ility th a t in the process of learning
lab els to a c r ite r io n of fiv e successive co rrect t r i a l s ,
with every c o rre c t response rewarded, some lab els could
have received more rewards than o th ers, c reatin g nothing
more than a tendency to respond with or use th a t label
more frequently than others as shown by operant verbal
conditioning s tu d ie s , ra th e r than modifying preferences
as measured by other than verbal c r i t e r i a .
In a s e rie s of stud ies based on Osgood's "media
tion h y p othesis," S taats and S taats (1957, 1958);
S ta a ts , S ta a ts, Heard, and Nims (1959); and S ta a ts , S ta a ts,
and Heard (1959) have demonstrated conditioning of
19
meaning to verbal stim u li. The stu d ies have followed a
b a s ic a lly sim ila r procedure in which nonsense s y lla b le s or
n e u tra l words have been paired with several words which,
though d if f e r e n t, had a common meaning element, such as
high loading in p o sitiv e ev aluativ e component. The r e s u lts
c o n siste n tly showed th a t a previously n e u tra l verbal
stimulus w ill be rated on the semantic d i f f e r e n t i a l in
accordance with the "meaning" loading of the c lass of
words with which i t has been asso cia te d . The studies
provide strong evidence for the th e o re tic a l ra tio n a le of
sign to sign conditioning, but do not deal with nonverbal
behavior or g e n e ra liz a tio n from sign to s ig n if ic a te .
In an early study, Birge (1941) employing children
three and a h a lf to fiv e and a h a lf years of age, reinforced
the response of pushing a handle to a white patch as
stimulus and p u llin g i t to a black patch, with the
responses reversed for a second group. A t e s t on a gray
patch showed no g e n e ra liz a tio n . She then gave m ediational
tra in in g in which one group was tra in e d to respond with
the word "white" to both white and gray while another group
was train ed to respond "black" to both black and gray.
Her re s u lts showed th at following th is labeling tra in in g ,
20
the o rig in a l push or pull response generalized to the gray
patch. The g eneralization was g re a te st in subjects
required to make the mediating verbal response overtly
p rio r to the push-pull response, but i t also appeared in
subjects who did not o vertly verbalize the mediating
response. Birge also found th at when the push-pull
responses were reinforced to a new stimulus o b ject, that
object was more lik e ly to be called by the name th a t was
attached to the previous object to which such response had
been reinforced. It is not c le a r whether these re s u lts
can be in terp re te d as demonstrating sign to object con
d itio n in g , because the conditioned stimulus was the object,
white or black, with the label possibly serving only as
a mediator to e sta b lish the equivalence of white or black,
and gray. However, i t is also q uite possible th at the
subjects were making im p licit verbal responses, "white"
or "black" when the conditioned stim uli were presented,
and th at the reinforcement was attaching to the lab els,
by v irtu e of th e ir contiguous occurrence. In any event,
the study can be distinguished from the present study
because d ire c t verbal conditioning was not involved.
Summary of Review of L ite ra tu re
21
The above survey has shown th a t th ere have been
several in v e stig a tio n s of the ro le of verbal mediating
responses in the g en e ralizatio n of autonomic responses
and verbal responses, including a ttitu d e s and meaning,
from object to sign and from sign to sign. The general
r e s u lts of such stu d ies in d ic a te th a t verbal mediation
can be a s ig n ific a n t v aria b le in g e n e ra liz a tio n . In v e s ti
gations of sign to object conditioning are ra re , as are
stu d ies involving voluntary, nonverbal responses. There
appear to be no in v e stig a tio n s of g en e ralizatio n of the
e ffe c ts of conditioning meaning to verbal signs on volun
ta ry , nonverbal behavior toward objects s ig n ifie d .
R ationale of Present Study
Osgood (1952, 1953, 1957) presents a th e o re tic a l
approach to meaning which may be represented as re stin g
on th is basic form ulation: a sign gets i t s meaning through
asso ciatio n with a s ig n if ic a te by v irtu e of coming to
e l i c i t a re p re se n ta tio n a l mediating process which is some
fra c tio n a l p art of the t o t a l behavior e l ic i t e d by the
s ig n if ic a te . Thus, the word "hammer" is not the same
22
stimulus as the o b ject, hammer. The former is a p attern
of sound waves, the l a t t e r is some set of v isu a l, ta c tu a l,
and other stimulus q u a litie s . N evertheless, the sign,
"hammer," comes to e l i c i t behaviors which are in some
manner relev an t to the s ig n if ic a te , hammer. In ordinary
learning experience, the word, "hammer," and the object
hammer are contiguously experienced. That is to say, and
th is is the crux of Osgood's p o sitio n , the word, "hammer,"
is experienced sim ultaneously with the complex of v isu al,
ta c tu a l, proprioceptive, and other reactions e l i c i t e d by
the o b ject, hammer, and thus becomes associated with at
le a s t some portion of th is complex. Osgood uses the term
"rm" to designate th a t portion of the complex which becomes
asso ciated with the sign. The "rm" may c o n sist of purely
neural events, muscular co n tra c tio n s, or glandular s e c re
tio n s . Through the a sso c ia tio n , the word "hammer," by
i t s e l f comes to e l i c i t some portion of the complex of
re a c tio n s. The component e lic i t e d by the word then acts
as a cue or stim ulatio n, which Osgood terms "sm," mediating
behavior which is in some manner relevant to the o bject.
Osgood re fe rs to the "rm" p art of the mediating
response as decoding, or in other words, in te rp re tin g the
stim ulus, and he re fe rs to the "sm" part as encoding,
23
or in other words, tra n s la tin g the reactio n to the stimulus,
in to cues to behave in some way relev an t to the stim u lu s.
He p o stu lates th a t the components of the complex which
re p resen t, or are responses to , the stimulus object are
detachable, and th e re fo re some of the components may occur
when only the lab el is present and the stimulus object
is absent. In th is manner, the lab el acquires meaning
which is made up, a t le a s t in p a r t, of some portion of
the meaning of the o b ject.
If meaning co n sists of detachable components which
may be shared by object and la b e l, i t seems reasonable
to assume th a t the converse of the above may also o ccu r--
th a t is the detachment of components of meaning from
labels and attachment of these components to the object
represented. It would then also follow th at i t should be
p ossible to p red ict th a t modifying the meaning of the
lab el could r e s u lt in some m odifications of the behavior
toward the ob ject. Therefore, increasing the pleasantness
of the lab el should increase approach behavior toward the
object and increasing the unpleasantness of the label
should increase avoidance behavior toward the o b ject,
since pleasantness and unpleasantness are u sually
24
associated re sp e ctiv ely with approach and avoidance tend
encies .
Hypothesis
The general hypothesis to be te s te d is th at
voluntary, nonverbal behavior toward an object can be
modified as a r e s u lt of modifying the meaning of the
verbal label which represents the o b ject.
The s p e c ific hypotheses to be te s te d are:
1. Conditioning p o s itiv e evaluative meaning to
a verbal lab el w ill re s u lt in g re a te r than
chance se le c tio n of the object s ig n ifie d ,
from among a group of o b jects, where the
se le c tio n is made by ph ysically grasping the
object and removing i t from the group of
o b je c ts .
2. Conditioning negative evaluative meaning to
a verbal lab el w ill re s u lt in less than chance
se le c tio n of the object s ig n ifie d from among
a group of o b je c ts, where the se le c tio n is
made by physically grasping the object and
removing i t from the group of o b jects.
CHAPTER I I I
M ETH O D
Subjects
The subjects were 72 students between the ages
of 17 and 23, drawn from an introductory psychology class
of approximately 130 students at Los Angeles S tate College.
The course is a degree requirement for a l l students except
those majoring in psychology. As part of the requirements
of the course, students are obligated to p a r tic ip a te in
research p ro je c ts, and at the beginning of the semester,
each student indicates on a card three d iffe re n t hours
he has free for such p a rtic ip a tio n . The experiment was
designed to con sist of twelve groups of six su bjects, with
three males and three females in each group, and th ere fo re,
the p a rtic ip a tio n cards were sorted in various arrangements
according to the hours of a v a ila b ility indicated, u n til
twelve groups, consisting of four males and four females
per group were found. Thus, each group was overscheduled
to allow for fa ilu re s to appear at the appointed time.
25
26
The hours for which groups were established were as f o l
lows: Wednesday at ten A.M.; Wednesday at eleven A.M.;
Thursday at nine A.M.; Thursday at ten A.M.; Friday at
one P.M.; Friday at two P.M.; with the same schedule
repeated the following week. It is recognized th a t the
se le c tiv e factors which e x ist in obtaining subjects in
th is manner may impose lim itatio n s on the degree to which
the re s u lts can be generalized. However, due to the
nature of the subject population to be employed such
selectio n could not be avoided e n tire ly . Therefore, the
subjects were run in twelve groups of six , with three
males and three females in each group according to the
above schedule, except for group XI where four males but
only two females appeared at the appointed time. There
fo re, for the e n tire group, there were th irty -sev en males
and th ir t y - f i v e females.
M aterial
The m aterials employed in the study consisted of
the following:
1. P rojector slid e s containing nonsense s y lla b le s,
corresponding to those used by Staats and Staats (1957).
27
These consisted of a tra in in g set--VAF, YIM, XAD, GAH,
t
and VEC, and the experimental set--XEH, YOF, LAJ, W UH,
QUG, and GIW.
2. A l i s t of n eu tra l words such as AT, BRIEF,
WITH, taken from the S taats and Staats (1957) study.
3. A l i s t of asso ciatio n words, used by Staats
and S taats, consistin g of 18 words of p o sitiv e evaluative
meaning, 18 words of negative evaluative meaning, and 72
words of n eu tra l evaluative meaning (see Appendix A).
Most of these words were obtained from Osgood, Suci, and
Tannenbaum (1957) with a few obtained by Staats and S ta a ts,
from other sources.
4. Semantic d if f e r e n tia l booklets (see
Appendix B).
5. One round, black, p la s tic bucket, approximately
15 inches high and approximately twelve inches in diameter.
6. Six two inch square wooden blocks, painted
gray and d iffe re n tia te d by arrangements of red thumb
tacks, as shown below:
/ 2 * 3 W C 6
28
7. A q u estionnaire (see Appendix C) requesting
the subjects to name the block they selected from the
bucket, give th e ir reasons fo r the choice i f any, and
make comments i f they care to.
8. Trooth Booths, a se c u rity device providing
a th ree sided chamber in which a l l w ritte n m aterial could
be kept, insuring th a t no subject could see what any
other subject was w ritin g .
9. Six 3 x 4 plain white cards each containing
one of the experimental sy lla b le s p rinted in two inch
c a p ita l l e t t e r s .
Procedure
O rienting Procedure
1. A ll groups were given an id e n tic a l o rien tin g
procedure. F i r s t , the subjects were informed th a t they
would be asked to perform a v a rie ty of tasks in order to
aid in a study of d iffe re n t kinds of learning processes.
2. Subjects were to ld to watch the screen and
pronounce out loud and then repeat to themselves the
nonsense s y lla b le s th a t would be presented. A fter fiv e
presen tatio n s of tra in in g s y lla b le s , VAF, YIM, XAD, GAH,
and VEC, in d iffe re n t orders, subjects were in stru c te d
29
to w rite down as many of the sy lla b le s as they could
remember. The purpose of th is step was to fa m ilia riz e
the subjects with watching the screen and repeating the
s y lla b le s , allowing an opportunity fo r any confusion
about th is to be cleared up before repeating th is process
with the experimental s y lla b le s .
3. Subjects were in stru c te d th a t E would say
some words which they were to repeat aloud and then to
themselves. E spoke the n e u tra l words (such as AT,
BRIEF, UNDER), and at the conclusion of twelve such words,
asked the subjects to w rite down a l l of them they could
remember. This was to provide p ra c tic e in repeating the
words which would l a t e r be used in conditioning meaning
to the experimental sy llab les
4. E then read p airs of words, one of each pair
being taken from the previous l i s t , for example, BRIEF,
and the other standing in some re la tio n s h ip to i t , fo r
example BRIEF--BRIEFCASE, UNDER--BENEATH. The subjects
were in stru c te d to w rite down which of the p air had been
presented on the previous l i s t . A ll of the foregoing
steps followed the procedure of the S taats and S taats
study (1957).
30
Experimental Procedure
1. E in stru c te d the subjects th a t they were now
to learn some new nonsense s y lla b le s as names for some
blocks. F i r s t , the blocks were held up, one a t a time.
Then E held up the cards with the experimental s y lla b le s ,
XEH, YOF, LAJ, W UH, QUG, and GIW p rinted on them. E
pronounced the sy lla b le s and asked the subjects to repeat
them out loud.
2. E held up the s y lla b le cards and corresponding
blocks to g eth er, pronouncing the s y lla b le as i t was shown,
and in s tru c tin g the subjects th a t the s y lla b le was the
name for the block. Then the sy lla b le s and blocks were
shown tog eth er a second time.
3. The six blocks were presented, one a t a time
in a random order, with the subjects in stru c te d to w rite
down the name of each as i t was presented. E then held
up the card and block together fo r th a t order, so th at
subjects could determine i f they had been c o rre c t. This
procedure was continued, using a d iffe re n t random order
of presen tatio n each time, u n t i l the subjects indicated
by ra is in g th e ir hands th a t they had been able to name
a l l of the blocks c o rre c tly in two successive s e rie s of
p re se n ta tio n s. In a few in stan ces, i t was necessary to
31
go on to the next step with one subject in the group s t i l l
not in dicatin g learning of the names to th is c r ite r io n .
This was made necessary by the fact th at only one hour
had been provided for the to ta l procedure, based on
prelim inary t r i a l which had indicated th at 45 minutes was
s u ffic ie n t to carry out the to ta l procedure. However,
in the actual study, due to la te a r riv a ls and other
unforeseen delays, a time pressure developed because many
of the subjects had classes to attend at the end of the
scheduled hour. Inasmuch as a l l groups had been scheduled
in advance, and there was l i t t l e p ro b ab ility th a t groups
could be established with a l l subjects having two con
secutive hours av a ilab le, the decision was made to ris k
the loss of an occasional subject ra th e r than to lose a
whole group due to lack of time to fin is h . The groups
d iffe red in the number of presentations required before
reaching the c r ite r io n of learning, ranging from as few
as three to as many as eig h t, so th a t some subjects had
overlearning t r i a l s . For example, in the group which had
eight presen tatio n s, some subjects reached the c r ite r io n
in three t r i a l s , but p a rtic ip a te d in the remaining five
t r i a l s , which co n stitu ted overlearning t r i a l s for those
subjects. The names for the blocks varied from group to
32
group, so th at for the twelve groups each block had each
s y lla b le for i t s name twice. The group-block-syllable
arrangement was as follow s:
Group Block S yllable
I. 1-xeh, 2-yo f, 3 - la j, 4 -wuh, 5-giw, 6-qug
II. same
I I I . 1-yof, 2-xeh, 3-wuh, 4 - la j, 5-qug, 6 -la j
IV. same
V. 1-giw, 2-qug, 3-xeh, 4 -y o f, 5“l a j , 6-wuh
VI. same
VII. i-qug, 2-giw, 3-yof, 4-xeh, 5-wuh, 6 -laj
VIII. same
IX. 1 -la j,
2-wuh, 3-giw, 4-qug, 5-xeh, 6-yof
X. same
XI. 1-wuh, 2 - la j, 3-qug, 4-giw, 5-yof, 6-xeh
XII. same
In th is arrangement, each block has each
s y lla b le fo r i t s name an equal number of times.
4. The procedure of Staats and Staats
(1957) for the conditioning of meaning to nonsense
sy lla b le s was followed. The subjects were told th a t the
sy llab les they had ju s t used would be presented on the
33
screen and th a t as they were p resen ted , E would say a
word. The sub jects were in stru c te d to look a t the s y lla b le
on the screen and to repeat the word out loud and then to
themselves u n t i l the next s y lla b le was presented. For
a l l odd-numbered groups, XEH was paired with words of
p o sitiv e ev aluative meaning, such as " p r e tty ," "h e alth y ,"
"ric h " (see Appendix A for f u l l l i s t ) , while YOF was
paired with words of negative ev aluativ e meaning, such
as "ugly," " e v il," "insane" (see Appendix A). The other
four sy lla b le s were paired with n e u tra l words, such as
"sh ip ," "cup," "room" (see Appendix A). The p airings
for XEH and YOF were reversed fo r a l l even-numbered
groups, so th a t fo r them XEH was negative and YOF p o sitiv e .
Thus, XEH p o s itiv e , XEH negative, YOF p o s itiv e , and YOF
negative occurred an equal number of tim es.
5. At th is point, the procedure diverged,
with h a lf of the groups making block choices f i r s t and
then completing the semantic d i f f e r e n t i a l , with th is
order reversed fo r the other h a lf of the groups, who
completed the semantic d if f e r e n t i a l f i r s t and then made
th e ir block choices. This was done to make i t possible
to determine whether the overt d ec la ratio n of feelin g
34
about the lab els on the semantic d i f f e r e n t i a l might
influence the block choices.
P rior to the present study, a prelim inary study
had been done to in v e stig a te the p o s s ib ility th a t some
c h a r a c te r is tic s of the blocks might make some blocks
p referred over o th ers. The subjects in th is study were
paid, volu nteers, drawn from an introductory psychology
c la s s . They were run one a t a time, and were simply
asked to take any block they wished out of the bucket.
The data (see Appendix D) did not req u ire re je c tio n of
the hypothesis of no d iffe ren ce among the blocks. However,
the data did reveal a s ig n ific a n t preference for c e rta in
po sitio n s in the bucket, apparently on the b asis of
closeness of the p osition of the block to the su b ject.
This positio n preference was co n tro lled in the present
study as follows: twelve random orders of blocks were
e sta b lish e d , one for each group, and the blocks were
ro ta te d one p o sitio n for each subject w ithin the group,
so th a t each block appeared once in each p o sition for
each group, but in d iffe re n t orders fo r each group.
6. The block choosing took place in the following
manner. On a ta b le , at the back of the room, behind a
screen, E placed the six blocks in a c ir c u la r arrangement,
35
face up, on Che bottom of the bucket. Subjects were
c a lled to the ta b le , one a t a time, and were handed a
w ritte n in s tru c tio n , "Please look a t a l l of the blocks
and take out the one you wish to use in the remainder of
the experiment." When the subject took out a block, E
took i t from him and placed i t on the ta b le , on top of
the w ritten in s tru c tio n . E then handed the subject a
second w ritte n in s tru c tio n , "Now take an a lte r n a te block
th a t you would lik e to u s e ,1 1 proceeding as with the f i r s t
choice. This second choice was given because the nature
of a l l of the fa cto rs which might en ter in to the f i r s t
choice was unknown, and i t was f e l t th at such fa c to rs , or
ind ividual preferences of the sub jects might determine
the f i r s t choices, while the predicted preference might
appear only i f a second choice was given. A fter making
the second choice, the subject was handed the q u estio n
n aire (see Appendix C) which in stru c te d him to w rite in
the names of the blocks chosen f i r s t and second while
remaining at the ta b le , and then to return to his seat
to complete the re s t of the q u estio n n aire. The r e s t of
the question naire asked him to s t a t e his reasons, i f any,
fo r the choices, describe what he thought the experiment
36
was about, f i l l in his name and sex, and make any comments
he cared to about the experiment.
7. Instructions for completing the semantic
d if f e r e n tia l were given on the f i r s t page of the booklet
(see Appendix B). In th is part of the experiment, sub
je c ts were asked to ra te each of the sy llab les on a
seven-point scale for the dimension pleasant-unpleasant,
with 1 representing the extreme of pleasantness and 7
the extreme of unpleasantness. Order of the sy llab les
was controlled by using six d iffe re n t counterbalanced
orders for the position in the booklet fo r each group,
and each order appeared once for each group.
8. At the conclusion, subjects were asked not
to discuss the study with anyone for two weeks, with the
explanation th at the study would be completed by then,
but th at i f other students who were yet to p a r tic ip a te
knew in advance what was going to be done, th e ir approach
to the tasks would be d iffe re n t and th a t th is could
cause serious erro rs in drawing conclusions from the
r e s u l t s .
CHAPTER IV
RESULTS
Conditioning of Meaning
Table 1 shows the re s u lts of conditioning meaning
to the sy llab les XEH and YOF by asso ciating them with
words of pleasant or unpleasant meaning. In a l l compari
sons of mean ra tin g s on the semantic d if f e r e n t i a l , the
sy llab les receiving conditioning of pleasant meaning
were rated s ig n ific a n tly more pleasant than the next
most pleasantly rated s y lla b le , and the sy lla b le s receiving
conditioning of unpleasant meaning were rated s ig n ific a n tly
more unpleasant than the next most unpleasantly rated
s y lla b le . Subjects having pleasant meaning conditioned
to the s y lla b le s , XEH and YOF, rated these sy llab les
s ig n ific a n tly more pleasant than subjects having unpleasant
meaning conditioned to them. Differences between XEH and
YOF conditioned pleasant and between XEH and YOF condi
tioned unpleasant were not s ig n ific a n t.
37
TABLE 1
COMPARISONS OF SEMANTIC DIFFERENTIAL RATINGS AND SYLLABLES RECEIVING CONDITIONING
OF PLEASANT (+), UNPLEASANT (- ) , AND NEUTRAL (0) M EANING
Groups Syllable 1 Syllable 2
M 1
m2 S
x i- x2
df t
Odd XEH (+) W U H (0) 2.111 3.000 .474 35 1.876®
Q U G (0) YOF (-) 3.889 5.583 .436 35 3.885
Even YOF (+) LAJ (0) 1.806 2.917 .378 35 2.939b
Q U G (0) YEH (-_ 4.250 5.611 .449 35 3.031
All XEH (+) XEH (-) 2.111 5.611 .445 70 7.865b
YOF (+) YOF (-) 1.806 5.583 .392 70 9.635
XEH (+) YOF (+) 2.111 1.806 .387 70 . 787c
XEH (-) YOF (-) 5.611 5.583 .454 70 .062c
Note: In the comparisons involving odd-numbered or even-numbered groups, the mean rating
o£ the experimental sy lla b le is compared with the mean ra tin g of the sy lla b le rated
c lo se st to the conditioned d ire c tio n . For example, in the even groups, excluding XEH,
the pleasant conditioned s y lla b le , the s y lla b le W U H received the most pleasant ra tin g
of the other 5 s y lla b le s . Therefore, i f XEH is rated s ig n ific a n tly more pleasant than
W UH, i t is rated s ig n ific a n tly more pleasant than the other s y lla b le s .
S i g n i f i c a n t a t b e tte r than the .05 le v e l, o n e -ta il t e s t .
^S ignificant at b e tte r than the .005 le v e l, o n e -ta il t e s t . ^
cNot s ig n ific a n t, tw o -ta il t e s t . °°
39
These re s u lts confirm the e a r l i e r findings of S taats
and S taats (1957) and warrant the conclusion th a t the
s y lla b le s , XEH and YOF, had acquired the predicted pleasant
or unpleasant meanings as a r e s u lt of the conditioning
procedure employed.
Effect of Conditioned Meaning of Labels on
Choices of Blocks
The o rig in a l design of the study c a lle d for chi
square an aly sis for the frequencies of subjects se le c tin g
the blocks whose labels had received conditioning of
p leasan t, unpleasant, or n e u tra l meaning. However, a f te r
the data had been c o lle c te d , i t was re a liz e d th a t running
the su b jects in groups created an objection to an analysis
pooling a l l su b jects, in th a t, there may have been v a r i
ables other than the conditioned meaning occurring in some
groups and not in others which may have influenced the
choices of blocks by subjects in such groups. In other
2
words, i t might be possible to obtain a s ig n ific a n t X
due to a large co n trib u tio n of "c o rrec t" choices by the
subjects in one or two groups, which could have occurred
on the b asis of some v aria b le other than conditioned
meaning. The w rite r is not aware of the presence of any
40
such v ariab le, with the exception of the differences among
the groups in the numbers of t r i a l s given in order to
reach the c r ite r io n for learning the names fo r the blocks.
Analysis (see discussion of Table 6, below) indicates
there was no re la tio n sh ip between th is v ariab le and
“correct" block choices. The only other observable d i f
ference among the groups lie s in the behavior of the
su b jects, with subjects in some groups occasionally
joking a l i t t l e at the beginning of the procedure. Once
the presentation of the experimental sy llab les on the
screen began, no fu rth e r v erb alizatio n or in te ra c tio n
occurred, except for one group where a subject asked i f
a sy lla b le and i t s asso ciate could be presented again and
was informed th is was not possible. At the c r ite r io n
task, block choosing, no differences occurred among
groups as the subjects were called up by number and
handed w ritten in s tru c tio n s, with the e n tire process
taking place in sile n c e . However, in view of the remain
ing p o s s ib ility that some v ariable of which the w riter
was unaware may have been operating, i t appeared necessary
to analyze the data tre a tin g the groups as the experi
mental u n its .
41
Assuming there were no systematic influences other
than the experimental ones under in v estig atio n , the sub
je c ts are trea te d as 12 groups, with 6 subjects in each
group. The n u ll hypothesis is th a t the experimental
treatment did not influence the groups in the selection
of blocks. If the n u ll hypothesis is tru e , then the
experimental treatment has fa ile d to e s ta b lis h a response
to a p a rtic u la r block th a t was d iffe re n t from the response
to the other five blocks, and thus the chance of selectin g
th at p a rtic u la r block out of the six blocks present is
ju s t one out of six for each individual. With six in d i
viduals in each group, i f the n u ll hypothesis were tru e ,
then the number of times the p a rtic u la r block should have
been chosen in each of the groups, purely by chance, is
6 (IV6) * 1.0 times. But the block cannot be expected to
be chosen exactly once in each of the groups; one can
expect, purely by chance, that sometimes i t w ill be
chosen 0 or 1 or 2 or more times. For 12 groups, the
th e o re tic a l d is trib u tio n of group scores, where the
expected score is 1.0, can be found, by in terp o latio n in
the tab le of binomial d is trib u tio n (Burrington and May,
1953, Table 1, p. 247). Table 2 shows the th e o re tic a l
d is trib u tio n derived from th is ta b le .
42
TABLE 2
THEORETICAL DISTRIBUTION OF GROUP SCORES
FOR 12 GROUPS OF 6 SUBJECTS W HERE
THE EXPECTED SCORE IS 1.0
n X p pN (N~ 12)
0 .339 4.07
1 .397 4.76
2 .199 2.39
3 .055 . 66
4 .009 .11
5 .001 .01
6 .000 .00
n = the number of t r i a l s w ithin a s e t.
X = the number of occurrences of an event (choice
of the " c o rre c t" block) w ithin the group,
p = the p ro b a b ility of the occurrence of the event.
Source: Burrington and May, 1953, Table 1,
p. 247, by in te rp o la tio n between p * .15
and p * .20 where p = .167 fo r a
" c o rre c t" choice on any one t r i a l .
43
According to the p r o b a b il i tie s s ta te d in Table 2,
i f th e n u ll hypothesis were tr u e , then th e scores fo r
each group should have been d i s t r ib u te d as fo llo w s :
4.07 or 4 groups should have had a score of 0
4.76 or 5 groups should have had a score of 1
2.39 or 2 groups should have had a score of 2
.66 plus .11 plus .01 or 1 group should have
had a score of more than 2.
Comparison of th e t h e o r e t i c a l d i s t r i b u t i o n w ith the
observed d i s t r i b u ti o n of choices of th e blocks whose
la b e ls had received co n d itio n in g of p le a sa n t meaning is
shown in Table 3. The r e s u l t s of the Kolmogorov-Smimov
one sample t e s t y ie ld a p ro b a b ility of le s s than .10 th a t
the observed d i s t r i b u t i o n could have occurred purely by
chance. Inspection of th e data shows th a t the number of
" c o r r e c t’* choices obtained exceeded the number expected
by chance. Thus, i t is p o ssib le to conclude, a t b e t t e r
than a 10% le v e l of confidence, th a t co n d itio n in g p leasan t
meaning to the name fo r a block re s u lte d in establishm ent
of p referen ce or approach behavior toward th a t block.
In the comparison between the t h e o r e ti c a l d i s t r i b u
tio n and observed d i s t r ib u tio n of choices of the blocks
w ith "u npleasant" la b e ls , shown in Table 4 the d iffe re n c e
TABLE 3
KOLM OGOROV-SM IRNOV ONE SAM PLE TEST FOR OBSERVED AND THEORETICAL DISTRIBUTIONS
OF CHOICES OF "PLEASANT" BLOCKS BY 12 GROUPS OF 6 SUBJECTS
Number of Possible "Correct" Choices 0 1 2 3 4 5 6
Observed numbers of "correct"
choices by groups 0 5 5 2 0 0 0
T heoretical cumulative d is tr ib u
tion of choices under H0 .339 .736 .935 .990 .999 1.000 1.000
Cumulative d is trib u tio n of
observed "co rrect" choices
by groups 0 .417 .833 1.000 1.000 1.000 1.000
Deviations .339 .319 .102 .010 .001 0 0
D * .339 N = 12 p < .10, tw o -ta il t e s t
-p -
TABLE 4
KOLM OGOROV-SM IRNOV O N E SAM PLE TEST FOR OBSERVED A N D THEORETICAL DISTRIBUTIONS
OF CHOICES OF "UNPLEASANT" BLOCKS BY 12 GROUPS OF 6 SUBJECTS
Number of P ossible "C orrect"
Choices 0 1 2 3 4 5 6
Observed number of "c o rre c t"
choices by groups 5 5 2 0 0 0 0
T h eo retic al cum ulative d i s t r i b u
tio n of choices under H 0 .339 .736 .935 .990 .999 1.000 1.000
Cumulative d is tr ib u tio n of
observed " c o rre c t" choices
by groups .417 .833 1.000 1.000 1.000 1.000 1.000
D eviations .078 .097 .065 .010 .001 0 0
D - .097 N = 12 p > .20
-o
U i
46
is not s i g n if ic a n t, and i t cannot be concluded th a t con
d itio n in g of u n pleasant meaning to the name of a block
re s u lte d in estab lish m en t of a v e rsiv e or avoidance behavior
toward th a t block.
To determ ine i f the groups showed a p re fere n ce fo r
blocks whose la b e ls had receiv ed co n d itio n in g of p le a sa n t
meaning over blocks whose la b e ls had receiv ed co n d itio n in g
of unpleasant meaning, th e Wilcoxon roatched-pairs signed-
ranks t e s t was employed. To meet the requirem ents fo r
use of th is t e s t , both th e f i r s t and second choices of the
su b je c ts in th e groups were used. Thus, th is a n a ly s is
compares choices of th e same groups under two d if f e r e n t
experim ental c o n d itio n s , c o n d itio n in g of p leasa n t meaning
and co n d itio n in g of unpleasant meaning. The r e s u lt s given
in Table 5, show th a t a s ig n if ic a n t p re fere n ce fo r blocks
w ith " p le a s a n t” names over those w ith "u n p lea san t” names
by the groups was e s ta b lis h e d .
A ch i square a n a ly s is , shown in Table 6, was
performed to ev a lu a te the e f f e c ts on block choices of
p o ssib le o v erle arn in g t r i a l s in th e block name le a rn in g
procedure. To make t h i s a n a ly s is , i t was necessary to
combine a l l su b je c ts who had any t r i a l s beyond the p o in t
where they had met th e c r it e r io n of two su ccessiv e c o rre c t
47
TABLE 5
W ILCOXON MATCHED-PAIRS SIGNED-RANKS TEST FOR CHOICES,
O N FIRST OR SECOND CHOICE, OF "PLEASANT'* A N D
"UNPLEASANT" N A M E D BLOCKS BY 12 GROUPS
Choices of Choices of Rank w ith
"p lea san t" "unpleasant" Rank of le ss f r e
P air blocks blocks d d quent sigi
Group 1 4 2 2 8.5
Group 2 2 3 -1 -3 3
Group 3 3 3 0
Group 4 4 2 2 8.5
Group 5 3 2 1 3
Group 6 4 2 2 8.5
Group 7 3 2 1 3
Group 8 2 0 2 8.5
Group 9 3 4 -1 -3 3
Group 10 3 1 2 8.5
Group 11 2 3 -1 -3 3
Group 12 3 1 2 8.5
N = 11 T - 9
p < .025, one t a i l te s t
48
TABLE 6
CHI SQUARE FOR CHOICES OF ’’PLEASANT" (+), "UNPLEASANT"
( - ) , A N D "NEUTRAL" (0) BLOCKS BY SUBJECTS HAVING
POSSIBLE OVERLEARNING TRIALS (NOL)
*
Fe
OL
F„
NOl£e
o
+
-
0
11.67 5 23.33
12 5 23
9.33 4 18.67
9 4 19
X2 * .031 df * 2 p > .95
ilr
Expected frequencies were derived from m arginal
t o t a l s .
49
namings. T herefore, a su b je c t was defined as having
p o ssib le overlearn in g t r i a l s i f th e number of t r i a l s given
h is group exceeded the number a t which he had met the
c r it e r io n . Forty of the su b jects had p o ssib le o v e rle a rn
ing t r i a l s , the number of which ranged from 1 to 6. The
r e s u lts show no s ig n if ic a n t d iffe re n c e in block choices
between su b jects w ith and w ithout p o ssib le overlearning
t r i a l s , and i t i s , th e re fo re , concluded th a t th is v a ria b le
had no e ffe c t on block choices.
The foregoing r e s u lts provide evidence th a t the
experim ental trea tm en t, a t le a s t the co n d itio n in g of
p leasan t meaning, did have the p red icted e f fe c t on the
response measured. As these r e s u lts cannot be a ttr ib u te d
to unknown v a ria b le s which may have a ffe c te d some groups
and not o th e rs, i t is deemed ap p ro p ria te to re p o rt chi
square analyses pooling a l l of the su b jects in accordance
w ith the o rig in a l design.
Table 7 gives the ch i square an a ly sis fo r frequency
of choices of " p le a s a n t," "u n p le a sa n t," and "n e u tra l"
2
blocks by a l l s u b je c ts . The obtained X of 8.25 is s i g
n if ic a n t a t b e tte r than th e .02 le v e l fo r 2 d f. Inspection
of the data shows th a t the choices of blocks whose names
had received cond itio n in g of p leasan t meaning exceeded
50
TABLE 7
CHI SQUARE FOR ALL SUBJECTS FOR FREQUENCIES O F
SUBJECTS CHOOSING "PLEASANT" (+),
"UNPLEASANT" (-) A N D "NEUTRAL"
(0) BLOCKS
+
-
0
f e
12 12 48
Fo
21 9 42
X2 - 8.25 df = 2
P < -02
51
the frequency expected by chance, w hile the choices of
blocks whose names had received co n d itio n in g of unpleasant
or n e u tra l meaning were le ss than th e frequencies expected
by chance. The c a lc u la tio n c a rrie d out re v eals th a t n early
2
a l l of the t o t a l X is co n trib u te d by th e discrepancy
between expected and obtained frequencies fo r choices of
"p leasan t" blocks. I t is concluded th a t the cond itio n in g
of p leasa n t meaning to the name of a block re s u lte d in th e
establishm ent of approach or preference responses to such
b lo c k s.
In a d d itio n to 22 su b jects who fa ile d to show con
d itio n in g of meaning on the semantic d i f f e r e n t i a l , in sp ec
tio n of the s u b je c ts ’ worksheets revealed th a t se v e ra l
su b jects who had in d icated they had met the c r ite r io n of
block name learn in g , in f a c t, had n o t. Of the 72 s u b je c ts ,
only 44 met the c r i t e r i a fo r both conditioned meaning and
block name learn in g . Table 8 shows the chi square a n a ly sis
2
for choices by th ese 44 su b je c ts. The obtained X of
12.74 is s ig n ific a n t a t b e tte r than the .01 le v e l fo r
2 d f. The data and c a lc u la tio n s again in d ic a te th a t i t
is th e excess of the obtained frequency over the expected
frequency for choices of "p leasan t" blocks th a t c o n trib u te s
2
almost the e n tire X . This a n a ly sis provides stro n g er
52
TABLE 8
CHI SQUARE FOR CHOICES OF "PLEASANT" (+ ),
"UNPLEASANT" ( - ) , A N D "NEUTRAL" (0)
BLOCKS BY SUBJECTS W H O M ET CRITERIA
OF BOTH CONDITIONED M EANING A N D
BLOCK N A M E LEARNING
-
0
Fe
7.33 7.33 29.34
Fo
16 4 24
X2 = 12.74 df = 2
P < 01
53
evidence of th e estab lish m en t of a p re fere n ce fo r th e
" p le a sa n t" block .
The r e s u l t s of th e a n a ly s is in Table 9 f a i l to
provide evidence of a d iffe re n c e between males and females
w ith regard to th e v a ria b le under in v e s tig a tio n .
Table 10 gives th e c h i square a n a ly sis fo r p o ssib le
d iffe re n c e s due to th e o rd er in which block choosing and
sem antic d i f f e r e n t i a l ra tin g s were done. The r e s u l t s do
not in d ic a te th a t block choices were a ffe c te d by th e ord er
in which r a tin g and block choosing were done.
Frequencies fo r p o s itio n s in th e bucket from which
th e choices were made a re analyzed in Table 11. The
obtained X^ of 10.00 is not s ig n if ic a n t fo r 5 df although
2
th e p ro b a b ility of o b tain in g a X th is la rg e when th e re is
no r e a l d iffe re n c e among th e p o s itio n s is le s s than .10.
The d iffe re n c e s observed in th i s a n a ly sis a re s im ila r to
those found in th e p relim in ary study , w ith p o s itio n s c lo s e r
to th e su b je c t tending to be chosen most fre q u e n tly w hile
th o se f u r th e s t from th e su b je c t tend to be chosen le a s t
fre q u e n tly . I t , th e re fo re , seems lik e ly th a t th e p o s itio n s
of th e blocks in th e bucket had some e f f e c t on th e c h o ice s,
thus c o n s titu tin g a v a ria b le which could reduce the
p ro b a b ility of o b tain in g th e p re d ic te d outcome.
54
TABLE 9
CHI SQUARE FOR CHOICES OF “PLEASANT" (+),
"UNPLEASANT" ( - ) , A N D "NEUTRAL" (0)
BLOCKS B Y M A X ES A N D FEM ALES
0
Fe* 10.79 4.63 21.58
Male
F 13 5 19
Fe 10.21 4.37 20.42
Female
F0 8 4 23
X2 - 1.629 df * 2 p > .30
" t c
Expected frequencies were derived from
m arginal t o t a l s .
55
TABLE 10
CHI SQUARE FOR CHOICES OF "PLEASANT" (+), "UNPLEASANT"
( - ) , A N D "NEUTRAL" (0) BLOCKS BY SUBJECTS HAVING
THE SEMANTIC DIFFERENTIAL BEFORE CHOOSING
(SDB) A N D HAVING THE SEMANTIC
DIFFERENTIAL AFTER
CHOOSING (SDA)
+
•
0
*
F 10.5 4.5 21
SDA
Fo
11 5 20
F„ 10.5 4 .5 21
SDB
Fo
10 4 22
x 2 = .256 df = 2 p > .80
*Expected freq u en cies were derived from
m arginal t o t a l s .
56
TABLE 11
CHI SQUARE FOR FREQUENCIES WITH W HICH BLOCKS W ERE
CHOSEN FROM VARIOUS POSITIONS IN THE BUCKET
P o s itio n 3
1 2 3 4 5 6
Fe 12 12 12 12 12 12
Number of choices
Fq 7 17 16 14 13 5
X2 * 10.00 df * 5 p < .10
aThe p o s itio n s in the bucket a re as d esignated in
the diagram below.
CHAPTER V
DISCUSSION
The r e s u lts provide evidence supporting the hypothe
s is th a t co n d itio n in g p o s itiv e e v a lu a tiv e meaning to a
v erb al sign can be a determ ining fa c to r in th e e s ta b lis h
ment of preference or approach behavior to the o b ject
s ig n if ie d . The r e s u lts a lso support the more general
hypothesis th a t v o lu n tary , nonverbal behavior toward an
o b ject can be m odified as a r e s u lt of modifying the mean
ing of the verbal la b e l which re p re se n ts the o b je c t.
However, the r e s u l ts , although occurring in the
p re d ic te d d ire c tio n , f a i l to provide s a tis f a c to r y evidence
th a t co n d itio n in g of neg ativ e e v a lu a tiv e meaning to a
v erb al sign is a determ ining fa c to r in th e establishm ent
of av ersiv e or avoidance behavior to the o b ject s ig n ifie d .
Some sp ec u la tio n s may be o ffe red in explanation of the
f a ilu r e of th is aspect of the phenomenon to appear.
I t is p o ssib le th a t the phenomenon of modifying
behavior to o b jects by modifying th e meaning of the la b e l
57
58
fo r th e o b ject is lim ite d to th e e f f e c ts of p o stiv e
m eanings, and th a t th e n e g a tiv e c o n d itio n in g of la b e ls
does not produce avoidance behavior toward th e o b je c t.
There may be a genuine d iffe re n c e in th e estab lish m en t of
p o s itiv e or n e g a tiv e behavior such t h a t, w hile approach
or p o s itiv e behavior can be e f f e c tiv e ly m anipulated by
th e use of secondary re in f o rc e r s such as words, avoidance
behavior may be le s s s u s c e p tib le to t h i s type of r e i n
forcem ent, being d i f f i c u l t to e s ta b lis h w ithout th e use
of prim ary rein fo rcem en t. Common experience lends some
support to t h is n o tio n . For example, a s s o c ia tin g words
connoting danger or in ju ry w ith th e s t r e e t may not be
s u f f i c ie n t to keep a c h ild from going in to the s t r e e t
whereas one or two spankings a s s o c ia te d w ith th e s t r e e t
a re lik e ly to produce th e re q u ired avoidance behavio r.
In c o n tr a s t, c h ild re n tend to re a d ily re p e a t behavior
th a t is a sso c ia te d w ith v erb al expressions of d e s i r a b i l i t y .
The study of S ta a ts and S ta a ts (1958) on th e estab lish m en t
of a t t itu d e s by v erb al co n d itio n in g a ls o supports such
n o tio n . In th a t study involvin g m eaningful m a te ria l, th e
r e s u lt s showed th a t a n a tio n a l name re c e iv in g c o n d itio n in g
of p o s itiv e meaning was ra te d more p leasa n t on th e
sem antic d i f f e r e n t i a l than a n a tio n a l name re c e iv in g
59
c o n d itio n in g of n eg a tiv e meaning. However, w hile th e
d iffe re n c e was s i g n if i c a n t, the mean r a tin g fo r th e
n e g a tiv e ly co n d itio n ed name was a ls o on th e p le a sa n t sid e
of th e s c a le , suggesting th a t n e g a tiv e a t ti tu d e s a re more
d i f f i c u l t to e s ta b lis h by v erb al c o n d itio n in g .
However, a more pragm atic ex p lan atio n may be
a v a ila b le in th e d a ta and in th e inform ation su p p lied in
th e s u b je c ts ' responses to th e q u e s tio n n a ire . There
appear to have been two major f a c to r s , both r e s u ltin g
from th e p a r tic u la r method employed, which could have
reduced or in te r f e r e d w ith th e dem onstration of g e n e ra liz a
tio n of co n d itioned meaning.
The f i r s t of th e se fa c to rs has to do w ith th e
e x iste n c e of meanings fo r th e la b e ls o th e r than th o se
presumably e s ta b lis h e d by c o n d itio n in g . I t was p o s s ib le ,
th a t th e assum ption th a t th e s y lla b le s were n e u tra l on
th e e v a lu a tiv e meaning dimension and would th e re fo re have
no p le a sa n t or u n p leasan t meaning except as m anipulated
by the procedure, was not te n a b le . I f th a t was th e ca se,
then i t was p o s s ib le fo r any given s u b je c t, th a t in
a d d itio n to th e p o s itiv e ly co n ditioned s y lla b le , o th er
s y lla b le s may have had p o s itiv e meaning, and th a t in
a d d itio n to th e n e g a tiv e ly co n ditioned s y lla b le , o th e r
60
s y lla b le s may have had neg ative meaning. The ra tio n a le
underlying th e study was th a t s u b je c ts 1 behavior toward
the o b jects would be determined by the meanings of the
la b e ls . Under th is r a tio n a le , i t is equally lik e ly th a t
a su b je c t would choose a block whose name had p o s itiv e
meaning fo r him independent of th e co n d itio n in g , as i t
is th a t he would choose a block whose name had p o s itiv e
meaning fo r him as a r e s u lt of the co n d itio n in g . I t a lso
follow s, th a t i f names, o th er than th e n e g a tiv e ly con
d itio n e d name, had n eg ativ e meaning fo r the su b je c t, then
the number of a lte r n a tiv e s to choosing a neg ativ e block
a re reduced and the p ro b a b ility of tak in g a negative
block is in creased .
T herefore, an a d d itio n a l a n a ly sis was performed
in terms of the d iscrep an cies between the frequencies w ith
which blocks were chosen whose names had been ra te d
p le a s a n t, un p leasan t, or n e u tra l by the su b ject doing the
choosing, and th e frequencies w ith which blocks whose
names were ra te d p le a s a n t, u n p leasan t, or n e u tra l would
be expected to be chosen by chance on the b a sis of the
propo rtio ns of the t o t a l numbers of ra tin g s which were
p le a sa n t, un p leasan t, or n e u tra l. The r e s u lts of th is
2
a n a ly sis are shown in Table 12. The X of 7.195 is
61
TABLE 12
CHI SQUARE ANALYSIS FOR FREQUENCIES WITH W HICH
SUBJECTS CHOSE BLOCKS W HOSE N A M ES THEY H A D
RATED PLEASANT (+), UNPLEASANT ( - ) , O R
NEUTRAL (0) O N THE SEMANTIC
DIFFERENTIAL
R ating of names
+ - 0
F 33.70 21.17 17.13
e
Blocks chosen ___________________________
F0 44 12 16
X2 - 7.195 df = 2 p < .05
Note: T h e o re tic a l freq u en cies were determ ined by c a lc u
la tin g th e percentage of th e t o t a l ra tin g s of names which
were ra te d as p le a sa n t (3 or l e s s ) , u npleasan t (5 or more)
or n e u tra l (4 ). Of th e t o t a l of 431 ra tin g s (1 su b ject
f a ile d to r a te 1 s y l l a b le ) , 202 or 46.8% were p le a sa n t,
127 or 29.47. were u n p leasa n t, and 102 or 23.87. were n e u tra l.
On th is b a s is , by chance alo n e, 33.70 or 46.8% of 72
blocks chosen could be expected to have names ra te d p le a s
ant by th e s u b je c t, 21.17 or 29.4% names ra te d unpleasant
by the s u b je c t, and 17.13 or 23.8% n e u tra l names.
62
s ig n if ic a n t a t b e tte r than the .05 le v e l fo r 2 d f. In
th is a n a ly s is , the avoidance of the blocks w ith "unpleasant"
names makes a s lig h tly g re a te r c o n trib u tio n to the t o t a l
than the p referen ce fo r blocks w ith "p le a sa n t" names.
I t is a ls o noted th a t th e p ro p o rtio n s of 47% p le a s a n t,
29% u n p leasan t, and 24% n e u tra l ra tin g s suggests th a t the
assumption of o rig in a l n e u tra l meaning fo r the s y lla b le s
i s , in f a c t, u n ten ab le, as th e co n d itio n in g procedure
should account fo r only 17% p leasan t and 17% unpleasant
ra tin g s . I t is c le a r th a t id io s y n c ra tic v a ria b le s , some
of which are revealed in the q u e stio n n a ire , such as
a s s o c ia tin g IAJ w ith L.A .S.C ., the ab b rev iatio n of Los
Angeles S ta te C ollege, or W U H w ith the slang term fo r
romance, put meanings in to the study which were not con
tr o lle d . The conclusion from the foregoing a n a ly sis is
th a t the meanings of the la b e ls were im portant determ iners
of both preference and avoidance behavior toward the
o b je c ts . While th is conclusion does not a l t e r the
neg ativ e fin d in g in regard to the hypothesis fo r negative
co n d itio n in g , i t is f e l t the p o in t should be made in
defense of the underlying th e o r e tic a l form ulatio n, which
is Osgood's. The conclusion supports th e theory of sign
to o b ject g e n e ra liz a tio n , even though the r e s u lts in
63
th is study do not support the hypothesis th a t v erb al con
d itio n in g of n eg ativ e meaning to a v erb al la b e l w ill r e s u lt
in avoidance of th e o b ject s ig n if ie d .
The second fa c to r in the f a ilu r e to confirm th e
avoidance hypothesis has to do w ith the " s e t" given the
su b je c ts . They were in s tru c te d th a t the experiment was
a study of le a rn in g . Examination of the q u e stio n n a ire s
re v eals th a t the reason most fre q u e n tly given fo r the
choice involved knowledge of th e name of the block, fo r
example, " I knew the name b e s t" or " i t was the e a s ie s t
to remember." I t seems th a t many of the su b je c ts f e l t
they would be asked th e name of the block they chose and
th e re fo re , chose the one whose name they f e l t most secure
about. This was tru e even of the 44 su b jects who met the
c r i t e r i a of co n d itio n in g and block name le a rn in g . D espite
the fa c t th a t they had been ab le to name a l l of the blocks
c o rre c tly in two consecutive s e r ie s , 15 of the 44 gave
knowledge of the name as the reason fo r t h e i r choice.
A pparently, the o b je c tiv e evidence th a t they knew a l l of
the names was subordinate to t h e ir fe e lin g s th a t they
knew some b e tte r than o th e rs. While th is may have m ini
mized, to some degree, th e r e s u lts w ith regard to the
approach h y p o th esis, th e p o s s ib ility of a g re a te r
64
minimizing e f fe c t on avoidance behavior is in d ic a te d by
th e fa c t th a t of the conditioned and name-knowing s u b je c ts ,
only 3 of th e 16 who chose th e p o s itiv e block gave knowl
edge as th e ir reason, w hile 2 of th e 4 who chose the
n eg ativ e block gave knowledge of th e name as the reason
fo r t h e i r choice.
T herefore, i t appears p o ssib le th a t the n eg ativ e
r e s u l ts regarding th e avoidance hypothesis may have been
due to f a il u r e to c o n tro l im portant v a ria b le s , p a r tic u la r ly
a ffe c tin g avoidance behavior.
T h e o re tic a l Im plications
Osgood (1953) has s ta te d th a t w hile phenomena of
o b ject to sign and sign to sign g e n e ra liz a tio n have been
dem onstrated, experim ental evidence fo r sign to o b ject
g e n e ra liz a tio n has been lack in g , although the theory of
semantic g e n e ra liz a tio n makes th is type of g e n e ra liz a tio n
p re d ic ta b le . The r e s u lts of th e p resen t study re p re se n t
a ste p in the clo sin g of th is research gap and support
of Osgood's conclusion.
The r e s u lts are o ffered in support of Osgood's
form ulation of the m ediation pro cess, p a r tic u la r ly w ith
regard to detachable meaning components. The evidence
65
ap p ears to p ro v id e s tro n g su p p o rt f o r t h i s fo rm u la tio n
because a t no tim e d id th e s u b je c ts e x p e rie n c e th e o b je c t
in c o n ju n c tio n w ith any sy ste m a tic m an ip u latio n o f m o tiv a
t i o n a l p r o p e r tie s nor were o v e rt approach o r avoidance
re sp o n se s s y s te m a tic a lly r e in f o rc e d a t any tim e . N ever
t h e l e s s , th e o b je c ts d id appear to a c q u ire to some d e g re e ,
th e m o tiv a tio n a l p r o p e r tie s t h a t were a s s o c ia te d only w ith
th e v e rb a l l a b e l s . Any attem p t to e x p la in t h i s phenomenon
seems to r e q u ir e p o s tu la tio n of some ty p e of p ro cess
whereby th e stim u lu s " o b je c t" e l i c i t s an e x p l i c i t or
im p lic it m ed iatin g re sp o n se " la b e l" or some p o rtio n
th e re o f in c lu d in g th e m o tiv a tio n a l p ro c ess c o n s t i t u ti n g
some p o rtio n or component of t h i s m ed iatin g re sp o n se . The
m o tiv a tio n a l component, fo r example " p le a s a n tn e s s " in
t h i s stu d y , e l i c i t e d now by th e o b je c t becomes th e
stim u lu s or cue fo r th e s u b je c t to behave in a manner
a p p ro p ria te to such m o tiv a tio n , fo r example by p r e f e r r in g
o r approaching th a t p a r t i c u l a r o b je c t in t h i s stu d y .
Osgood (1957) uses th e term s "decoding" and "encoding"
to d e s c rib e t h i s p ro c e s s . The f i r s t or "d eco d in g "stag e
ap pears to in v o lv e p ro c e sse s of re c o g n iz in g , i d e n tif y in g ,
and i n t e r p r e tin g th e e x te r n a l s tim u lu s , w h ile th e second
o r "encoding" s ta g e appears to in v o lv e the tra n s m is sio n
66
of th e se processes to th e en erg izin g and/or d ire c tin g
mechanisms of the ov ert response system . The follow ing
schem atization is intended to i l l u s t r a t e t h i s co n c ep tu al
iz a tio n of the process u n derlying th e r e s u l ts of th i s
study. The symbol 0 re p re se n ts th e o b je c t, L th e la b e l,
(1 (1
R th e observed response, and rm (2 and sm (2 th e
(: (:
(n (n
m ediating p ro c esses.
• 0
|7>
F'l
j
i
*
s
7
i
«
*
When th e su b je c t observes 0, the block, a s e t of
in te r n a l re a c tio n s , rm, is s e t o f f , which would re p re se n t
th e variou s p h y sical asp ects of 0, fo r example, rm
(1 -sq u are, rm (2-g ray, rm (3 -sm all, rm (4 -ta ck arrangem ent,
and so on. These become in te r n a l s tim u li sm to make some
response RA (in clu d in g the response o f no a c tio n ).
. 0 1
AH'
n
r n
%
i
)
9
V
a
#
0
♦
#
#
F U
67
Here th e su b je c t experiences th e la b e l and block
U .J ■ *
to g e th e r, w ith th e complex of s tim u la tio n s from th e block
p re se n t, and presum ably, w ith rm and sm, which m ediate
th e response to the block, o cc u rrin g .
3.
L-*
T ')
T 'N
V
\
\
\
t
%
ZrJ
By v irtu e of th e preceeding a s s o c ia tio n , th e la b e l,
when presen ted alone e l i c i t s some of th e m ediating reactions
to th e block. The symbol RA' is used to in d ic a te th a t
the response w hile re le v a n t to the o b je c t, is not n e c e s
s a r il y the same response as to the o b je c t.
L
j
3
r
r
\
1
%
i
t
S ')
R a 'b
This sta g e re p re s e n ts th e a d d itio n of a new media-
tio n a l component to the la b e l by v ir tu e of i t s a s s o c ia tio n
w ith some rein forcem ent not r e la te d to the o b je c t. In
the p resen t study th is is th e a d d itio n of th e meaning
" p le a sa n t" to the la b e l, leading to response m o d ificatio n
68
B added Co RA1.
i
isrJ
This stag e is s im ila r to stag e 3 but in re v e rse ,
and re p re se n ts the in clu sio n of th e new component in the
meaning of the o b ject because of the e x is tin g a ss o c ia tio n
of o b ject and la b e l. The block is now decoded square,
gray, sm all, tack arrangem ent, e t c . , and p le a sa n t.
Pleasantness is encoded as a cue to p re fe r or approach
th is block.
While the present study was not intended to be an
in v e s tig a tio n of the p h y sio lo g ic al bases of behavior,
one can sp ecu late regarding the congruence of the fin dings
w ith c u rre n t neuroph ysiolog ical theory and d a ta . The
notion of n eu ral c i r c u i t s , both in n a te ly given and the
product of le a rn in g , u n d e rlie s most c u rre n t th e o rie s of
behavior. Hebb's concepts of c e l l assem blies exem plifies
th is type of th e o riz in g . The re cen t work of Olds (1958)
in d ic a tin g the ex isten ce of s p e c ific lo c i in the CNS fo r
"p le a sa n t" and "unpleasant" sen satio n could be in te g ra te d
69
w ith n e u ra l c i r c u i t th eo ry , to ex p lain th e r e s u l ts of th i s
study in th e framework of th e m ediation h y p o th e sis. One
can s p e c u la te , on th is b a s is , th a t th e name and o b je c t
i n i t i a l l y re p re s e n t independent n e u ra l c i r c u i t s . By
v ir tu e of being co n tig u o u sly experienced, th a t i s , both
c i r c u i t s being brought in to sim ultaneous o p e ra tio n , th e
c a p a c ity of e i th e r stim ulus to f i r e a l l , or p a rt of the
c i r c u i t re p re s e n tin g th e o th e r is e s ta b lis h e d . For
example, th e stim u lu s, " l a b e l," s e ts o ff th e n eu ral
c i r c u i t , " la b e l" (Osgood’s "rm") which serves as th e
stim ulus "sm" fo r a response a p p ro p ria te to th e la b e l.
Now, by v ir tu e of th e a s s o c ia tio n of la b e l and o b je c t,
th e stim u lu s, " o b je c t," a ls o s e ts o ff th e n e u ra l c i r c u i t
" la b e l" (or p o rtio n th e re o f) which becomes th e "sm" fo r
behavior a p p ro p ria te to the l a b e l . I f one can assume from
O ld s's d a ta , th a t in th e process of c o n d itio n in g " p le a sa n t"
meaning to th e la b e l, c e r ta in neurons or c e n te rs giving
r i s e to p le a sa n t se n sa tio n become p a rt of th e n e u ra l
c i r c u i t involved in th e "rm" to the stim u lu s, " l a b e l,"
then one can a ls o sp e c u la te th a t th e stim u lu s, " o b je c t,"
because i t a ls o s e ts o ff th e c i r c u i t "rm" is capable of
giv in g r i s e to p le a sa n t se n s a tio n , which o rd in a r ily would
lead su b je c ts to seek or p re fe r the o b je c t to o th er
70
o b je c ts which la c k t h i s c a p a c ity . Of c o u rse , as Osgood
has p o in te d o u t, th e e x p la n a to ry and p r e d ic tiv e v alu e of
th e m ed iatio n h y p o th e sis need n o t be t i e d to any p a r t i c u l a r
n e u ro p h y sio lo g ic a l th e o ry .
Im p lic a tio n s f o r Psychotherapy
I f , as i s g e n e ra lly ag re e d , psychotherapy i s a
le a rn in g p ro c e s s , th e fin d in g s of t h i s study have im p lic a
tio n s fo r tech n iq u es o f psy ch o th erap y . Bandura (1961)
s t a t e s ,
I f one s e r io u s ly s u b s c rib e s to th e view th a t
p sychotherapy is a le a rn in g p ro c e s s , th e methods
o f tre a tm e n t should be d e riv e d from our knowledge
o f le a rn in g and m o tiv a tio n . Such an o r ie n ta tio n
i s l i k e l y to y ie ld new te c h n iq u e s of tre a tm e n t
w hich, in many r e s p e c ts , may d i f f e r m arkedly from
th e p ro ced u res c u r r e n tly in u se .
The p o s itio n to be taken in t h i s d is c u s s io n r e s t s on th e
c o n c e p tu a liz a tio n of psy chotherapy as a le a rn in g p ro c e s s ,
w ith th e le a rn in g of meanings in v o lv ed in th e ra p e u tic
ch an g e.
Bandura (1961) c l a s s i f i e s v a rio u s e f f o r t s to apply
le a rn in g th e o ry co n cep ts d i r e c t l y to th e th eo ry and
p r a c t ic e o f p sy ch o th erap y , in th e fo llo w in g manner:
c o u n te r-c o n d itio n in g ap p ro ach es, as e la b o ra te d by Shoben
(1949) and Wolpe (1958), based on th e p r i n c i p l e t h a t i f
71
strong responses which are incom patible w ith anxiety
re a c tio n s can be made to occur in the presence of an x iety
provoking cues, the incom patible responses w ill become
attach ed to th ese cues and thereby rep lace and elim in ate
th e anxiety responses; e x tin c tio n approaches, proposed by
D ollard and M iller (1950) and Dunlap (1932), based on th e
p rin c ip le th a t when a learned response is repeated w ithout
reinforcem ent, the stre n g th of the tendency to perform
th a t response undergoes a pro g ressiv e decrease; d isc rim in a
tio n learn in g (D ollard and M ille r, 1950), Fenichel (1941),
Rogers (1951), and S ullivan (1953), based on the p rin c ip le
th a t i f a p a tie n t is aware of the cues producing h is
behavior, of the responses he is making, and of th e
reasons he responds as he does, h is behavior w ill become
more su sc e p tib le to v e rb a lly mediated c o n tro l, and
v o lu n ta rily guided d isc rim in a tiv e behavior w ill re p la ce
the autom atic, o v er-g en eralized responses; reward
approaches, suggested by experim ental stu d ie s c ite d by
Bandura, and which appear to be based on the p rin c ip le
th a t new, more adaptive responses w ill be increased by
a s s o c ia tio n w ith rew ards.
I t is p o ssib le to view th ese d iv erse techniques
as re s tin g on a common b a se --th e m odification of meaning,
72
or more s p e c ific a lly in Osgood's terms, a change in some
portion of the re p re se n ta tio n a l mediation process. In
counter-conditioning, the change would occur in the "sm"
portion. Stimulus 1 is s t i l l decoded 1 in the "rm,"
process, but "sm 1" is replaced by "sm 2" due to the
g re a te r habit strength of "sm 2" as a re s u lt of the con
d itio n in g process. On th is basis R 2 occurs to S 1,
ra th e r than R 1, presumably providing a more adaptive or
more comfortable response and thus c o n stitu tin g "cure"
or "improvement." For example, one of Wolpe's techniques
is to tr a in the p a tien t to make "a sse rtiv e " responses in
c e rta in so cial s itu a tio n s as an an tid o te to anxiety or
to some maladaptive defense to anxiety previously employed
by the p a tie n t in these s itu a tio n s . In e x tin c tio n , the
locus of change would also be in the "sm" or encoding
stage, d iffe rin g from counter-conditioning in th at the
th e ra p is t does not tra in a sp ec ific new encoding, but
r e lie s on an ex istin g re p e rto ire of adaptive responses to
take over when the o rig in a l "sm" drops out as a re s u lt of
nonreinforcement. Osgood points out th a t a l l the con
ceptual machinery of the Hullian formulation can be
transposed to the "rm-sm" mediation paradigm, so th at
the process in counter-conditioning can be viewed as
73
changing the habit hierarchy in the mediation process by
increasing the habit streng th of R 2 to the point where
i t is stronger and thus more lik e ly to accur than R 1.
In e x tin ctio n , R 1 is weakened to the point where i t is
lower in habit strength than some other response, R X,
so th a t now R X is more lik e ly to occur to S 1, and
presumably (hopefully?) R X w ill be more adaptive or
comfortable than R 1.
The process in discrim ination learning d iffe rs in
that the locus of change is the "rm" or decoding p ro cess.
That is , where both S 1 and S 2 are decoded "rra 1," S 2,
as a re s u lt of train in g (usually in labeling) becomes
decoded "rm 2." Whereas "rm 1" was linked to "sm 1,"
"rm 2" is not so linked, but is fre e to become attached
to some other "sm x" av ailab le in the organism, again
presumably more adaptive or comfortable.
In reward methods, a t le a st as viewed by Bandura,
the process appears to involve the development of some
previously nonexistent or unused encoding process. The
s itu a tio n s described by Bandura for the use of th is
method involve delinquents or schizophrenics in whom the
encoding processes involved in normal learning are
inoperative. However, there is no reason for lim iting
74
th is method to such individuals or c la s se s. The th e ra p ist
may be in a position to observe th a t the c lie n t haa fa ile d
to develop a p a rtic u la rly e ffe c tiv e response mode for
nonpathological reasons (for example, lack of knowledge
or lack of opportunity). Thus, where S 1 has been decoded
"rm 1" and encoded "sm 1" leading to R 1, which is not
maladaptive or d istu rb in g , i t is possible th a t changing
the encoding to "sm 2" may lead to R 2 which may be a
more e ffe c tiv e response and one which may ultim ately
enable the individual to secure more g r a tif ic a tio n s in
his t o t a l l i f e s itu a tio n . Or i t may be th a t the in d i
vidual is simply unresponsive to S 1 but th a t the tra in in g
of "sm 2" can lead to g reater effectiv en ess or possibly
to g re a te r enrichment of the in d iv id u a l's l i f e , increasing
his happiness and thus reducing o v e r-a ll d is s a tis fa c tio n
or tension lev el.
The evidence of the present study in dicating th at
behavioral change, beyond mere change in verbal h a b its ,
can be achieved through manipulation of the meanings of
verbal symbols lends support to the above conceptual
approach. A dditionally, the findings suggest th a t, in
many cases at le a s t, "improvement" may be produced by
the d ire c t application of conditioning techniques.
75
Obviously, a g reat deal more research is needed to
d e lin e a te the parameters of p e rso n a lity , co g n itiv e, and
s itu a tio n a l v aria b le s w ithin which such techniques w ill
be e f fe c tiv e , but the findings in the r e la ti v e l y simple
s itu a tio n of the present study suggest th a t fu rth e r
p u rsu it of th is approach would be worthwhile.
Extending upon the findings of Nutthman (1957) th a t
"p o sitiv e " s e l f statem ents can be increased by verbal
conditioning methods, the findings of the present study
should have im plications for "se lf-co n ce p t" th eo ries of
psychotherapy. The present study suggests th a t g re a te r
frequency of p o s itiv e " s e lf " v erb aliz atio n s should r e s u lt
in more frequent p o sitiv e or approach behavior toward the
s e l f which sounds simply lik e another way of saying the
c lie n t would a c tu a lly a tta in g re a te r s e l f acceptance or
p o sitiv e se lf-re g a rd with concomitant in crease in s e l f -
confidence, freedom to function, and general feelin g s of
w ell-being. I t may well be th a t the long-abandoned
approach of Coue should be re-examined in the lig h t of
present knowledge.
One a d d itio n a l im plication of the present findings
for psychotherapy remains. There is the p o s s ib ility th a t
verbal conditioning methods may be valuable as an adjunct
76
in more tr a d itio n a l th erap ies. It seems lik e ly that these
methods could be u seful in promoting p a rtic u la r types of
behavior required of the c lie n t in various types of
therapy. For example, i t may be possible to increase free
associatio n by strengthening the c l i e n t s ' approach tendency
to th is type of behavior by means of d ire c t verbal con
ditio n in g methods. I t also seems possible th a t, ra th e r
than relying on the p o s s ib ility of "p o sitiv e growth
forces" or dormant ad ju stiv e responses to spring in to
function when old p attern s have been broken down, the
th e ra p is t could ac tiv e ly supply these ad ju stiv e responses
%
by conditioning th e ir verbal symbols. I t can be noted
here that the objections, voiced by psychoanalysts in
p a r tic u la r , to so -called "symptomatic" treatm ents, re fe r
to instances in which elim ination of the presenting
symptom was followed by the emergence of another, and
sometimes worse symptom. I t is possible to view these
instances as cases in which the next response in the
c l i e n t 's hierarchy was more maladaptive or disturbing
than the one being tre a te d . In th is view, i t may be
important for the th e ra p is t to insure th at the replacement
response would be an adaptive one by u ti li z i n g verbal
conditioning methods. In fa c t, th is point would appear
77
equally ap plicab le to " in s ig h t" therapy where being aware
of the cues producing h is behavior, of the responses he
is making, and of the reasons he responds as he does may
not r e s u lt in more adaptive behavior unless the verbally
mediated co n tro l to which h is behavior w ill become more
su scep tib le is based on verbal m ediators leading to more
adaptive behavior. The present study suggests th a t th is
is a m atter which could be brought under the c o n tro l of
the th e ra p is t.
Im plications for A ttitu d es
In th e ir study of the establishm ent of a ttitu d e s
by c la s s ic a l verbal conditioning, S taats and S taats (1958)
re ly on the previously c ite d study of Kapustnik (1934) to
imply th a t behavior toward the o bjects s ig n ifie d by the
conditioned verbal lab els would correspond to the meanings
of the labels as measured by verbal techniques. I t is
suggested th a t the r e s u lts of the present study supply
more relevan t support for t h e i r p o sitio n . These r e s u l ts ,
and p a r tic u la r ly those expressed in Table 12, in d ic a te
th a t nonverbal behavior toward objects does correspond
to the meanings of the verbal lab els for the o b jects.
This should be of considerable importance to research
78
on a ttitu d e s by providing evidence of the v a lid ity of
verbal measuring instrum ents, such as the semantic d i f
f e r e n t i a l , in such research.
The present r e s u l t s , along with the work of S taats
and S ta a ts , and Eisman, also suggest th a t meaning e sta b
lish ed through verbal asso cia tio n s are an important
v aria b le in the development of a ttitu d e s .
Suggestions fo r Further Research
The present study presents a prelim inary demonstra
tio n of a phenomenon of g e n e ralizatio n from meaning con
d itioned to verbal signs to nonverbal behavior toward
objects and suggests c e rta in im plications fo r both
th e o r e tic a l and applied psychology. Therefore, i t
appears th a t the fu rth e r in v e stig a tio n of the phenomenon
would be worthwhile, with the following representing some
of the d ire c tio n s such research might take.
1. In view of the methodological d i f f i c u l t i e s of
the present study, i t would be d e sira b le to reproduce
the phenomenon under conditions supplying b e tte r c o n tro ls.
One suggestion would be to s e le c t experimental subjects
by p re te stin g fo r conditioning of meaning and to take
79
steps to insure th a t knowledge of object names is ade
quately esta b lish e d .
2. In v estig atio n of the dem onstrability of th is
phenomenon with more complex stim u li, p a r tic u la r ly stim u li
with p re -e x is tin g meanings, such as classe s of persons and
s i t u a t i o n s .
3. Attempt to discover c h a r a c te r is tic s which d i s
tin g u ish people who show the g e n e ra liz a tio n e f fe c t from
those who do not. P o te n tia lly important v aria b le s in
th is connection might be in te llig e n c e , a b s tra c tin g a b i l i t y ,
o v e r-a ll p erso n a lity type in terms of conventional c a te
g o ries, and sp e c ific p e rso n a lity t r a i t s .
4. In v estig atio n of the e ffe c ts of varying se t
and other s itu a tio n a l co nd itio n s, such as physical con
d itio n s or c h a r a c te r is tic s of the re in fo rcin g person, on
the phenomenon.
5. In v estig atio n of the e ffe c t on g e n e ra liz a tio n ,
of d iffe re n t amounts of reinforcem ent, d iffe re n t kinds of
verbal re in fo rc e rs , such as words, sentences, paragraphs,
and e f fe c ts of c o n flic tin g reinforcem ents.
6. In v estig atio n of the duration of the g e n e ra liz a
tio n and the p o s s ib ility of and minimum requirements fo r
m aintaining i t by in te rm itte n t verbal reinforcem ent.
80
7. Determination of the e ffe c ts of s tre s s on the
establishm ent and persisten ce of the generalized behavior.
8. Relation of the phenomenon to physiological
v a ria b le s, such as tis s u e damage or psychopharmacological
ag en ts.
CHAPTER VI
SU M M A R Y A N D CONCLUSIONS
There is current in te re s t in verbal processes and
th e ir possible ro le in the determination of various kinds
of behavior. While there have been several studies
dealing with g eneralization from object to sign and from
sign to sign in re la tio n to autonomic and verbal responses,
there is a paucity of research l i te r a t u r e in v estig atin g
sign to object g e n e ra liz a tio n , p a rtic u la rly in terms of
voluntary, nonverbal behavior.
It was the purpose of th is study to in v estig ate
whether meaning conditioned to a verbal sign would a ffe c t
voluntary, nonverbal behavior toward the object sig n ifie d .
The study was based on Osgood's in te rp re ta tio n of meaning
i
in which meaning consists of rep resen tatio n al mediation
processes, components of which are detachable. The
general hypothesis was th at voluntary, nonverbal behavior
toward an object would be affected by the meaning asso
ciated with the label for the object. The sp ecific
81
82
hypotheses were as fo llo w s:
1. Conditioning p o s itiv e ev aluative meaning to
a verbal lab e l would r e s u lt In establishm ent of approach
or preference behavior toward the object s ig n ifie d .
2. Conditioning negative evaluative meaning to
a verbal lab e l would r e s u lt in establishm ent of avoidance
or aversion behavior toward the object s ig n ifie d .
Method
Subjects were 72 students in an introductory
psychology c la ss who were required to p a r tic ip a te in
research. The subjects were selec te d on the b asis of
hours a v a ila b le fo r p a r tic ip a tio n , and were run in 12
groups of 6 su b jects each, h a lf male and h a lf female.
P ositive or negative meaning was conditioned to
various nonsense s y lla b le s by the method of S taats and
S taats (1957). These s y lla b le s were learned by subjects
as names for blocks d if f e r e n tia te d by the d iffe re n t manner
in which th ree thumb tacks were arranged on th e i r su rfaces.
Following the conditioning and name learn in g , su bjects
were asked to take any block they wished to out of a
co n tain er.
Results
83
1. Analysis of ra tin g s on the semantic d i f f e r e n t i a l
showed th a t conditioning of p o s itiv e or negative meaning
for the s y lla b le s had been esta b lish e d .
2. Analysis of block choices showed th a t a s i g
n if ic a n t preference or approach response to blocks whose
names had received conditioning of p o s itiv e meaning was
estab lish ed .
3. The analy sis of block choices did not show
th a t a s ig n ific a n t avoidance or aversion response to
blocks whose names had received conditioning of negative
meaning was e stab lish ed although th ere was a trend in
th is d ire c tio n .
4. The variab les of sex, overlearning of block
names, and order of semantic d i f f e r e n t i a l ra tin g and
block choosing, were not s ig n ific a n t variab les in the
determ ination of block choices.
5. Disregarding the conditioned meaning, the
re s u lts showed th a t subjects preferred blocks whose names
had p o sitiv e meaning fo r them, for undetermined reasons,
and avoided blocks whose names had negative meanings for
them, for undetermined reasons.
C o n clu sio n s
84
The findings were regarded as o ffe rin g evidence
in support of a phenomenon of g e n e ra liz a tio n from verbal
sign to object s ig n ifie d . This phenomenon was in te rp re te d
in terms of Osgood's mediation hypothesis of meaning,
p a r tic u la r ly with regard to detachable components of
meaning. The findings were f e l t to have s ig n ific a n t
im plications for behavior theory, theory and techniques
of psychotherapy, and the development and m odification
of a ttitu d e s . Some evidence of the v a lid ity of the
semantic d i f f e r e n t i a l , as a measuring instrum ent, was
offered.
The p o s s ib ility th a t methodological considerations
might have been responsible fo r the f a ilu r e to demonstrate
g en e ralizatio n of negative conditioning was discussed.
Future research was suggested, including a
re p lic a tio n of the present study with d iffe re n t methodology
to elim inate uncontrolled v a ria b le s . Other research
p o s s i b i l i t i e s were also discussed.
B I B L I O G R A P H Y
85
BIBLIOGRAPHY
Anrep, G. V. Pitch d iscrim in atio n in the dog. J . Physiol
1920, 53, 367-385.
Bandura, A. Psychotherapy as a learning process. Psychol
B u ll.. 1961, 58, 143-159.
Bass, M . J . , and H ull, C. L. The ir r a d ia tio n of a t a c t i l e
conditioned r e fle x in roan. J . Comp. Psychol..
1934, 17, 47-65.
Birge, Janet S. The ro le of verbal response in tr a n s f e r.
Unpublished d octoral d is s e r ta tio n , Yale U niversity,
1941, c ite d by J e ffre y , W . E. The e ffe c ts of verbal
and nonverbal responses in mediating an instrum ental
a c t. J . Exp. Psvchol.. 1953, 45, 327-333.
Branca, A. A . Semantic g e n e ralizatio n a t the lev e l of
the conditioning experiment. Am. J . Psychol..
1957, 50, 541-549.
Bugelski, B. R ., and Scharlock, D. P. An experimental
demonstration of an unconscious mediated a sso c ia tio n .
J. Exp. Psychol., 1952, 44, 334-338.
Burrington, R. S ., and May, D. C. Handbook of p ro b a b ility
and s t a t i s t i c s with ta b le s . Sandusky, Ohio:
Handbook P ublishers, In c ., 1953.
Cofer, C. N ., J a n is , M . G ., and Rowell, M . M . Mediated
g en e ralizatio n and the in te rp re ta tio n of verbal
behavior; I I I . Experimental study of antonym
g ra d ie n ts. J . Exp. Psychol.. 1943, 43, 266-269.
D ollard, J . , and M iller, N. E. P ersonality and psycho
therapy. New York: McGraw-Hill, 1950.
87
Dunlap, K. H abits, th e i r making and unmaking. New York:
L iv erig h t, 1932.
Edwards, A. L. S t a t i s t i c a l methods fo r the behavioral
sciences. New York: R inehart, 1954.
Eisman, Bernice, S. A ttitu d e form ation: the development
of a co lo r-p referen ce response through mediated
g en e ra liz a tio n . J . Abnorm. Soc. Psychol.. 1955,
50, 321-326.
Eriksen, C. W., Kuethe, J . W., and S ullivan, D. F. Some
p erso n a lity c o rre la te s of learning without verbal
awareness. J . P ers.. 1958, 26, 216-228.
Fenichel, 0. Problems of psychoanalytic technique.
(Trans, by D. Brunswick.) New York: Psychoanalytic
Q uarterly, 1941.
Foley, J . P ., and Cofer, C. N. Mediated g en e ralizatio n
and the in te rp re ta tio n of verbal behavior: II .
Experimental study of c e rta in homophone and synonym
g ra d ie n ts. J . Exp. Psychol.. 1943, 32, 266-269.
Foley, J. P ., and Mathews, M . Mediated g en e ralizatio n
and the in te rp re ta tio n of verbal behavior: IV.
Experimental study of the development of i n te r -
lin g u is tic synonym g ra d ie n ts. J . Exp. Psychol..
1943, 33, 188-200.
Hovland, C. I. The g en e ralizatio n of conditioned responses:
I. The sensory g e n e ra liz a tio n of conditioned
responses with varying frequencies of tone. J . Gen.
Psychol. . 1937a, 17, 125-148.
________. The g e n e ralizatio n of conditioned responses:
I I. The sensory g e n e ra liz a tio n of conditioned
responses with varying in te n s itie s of tone. J . Genet.
Psychol. . 1937b, 51, 279-291.
Hull, C. L. P rin cip les of behavior. An intro duction to
behavior theory. New York: A ppleton-C entury-C rofts,
1943.
88
Kapustnik, 0. P. The in te rre la tio n between d ire c t con
ditioned stim uli and th e ir verbal symbols. (Trans,
from Russian t i t l e . ) Psychol. A bstr.. 1934, 8,
No. 153
K eller, M . Mediated g en eralizatio n : the generalization
of a conditioned galvanic skin response established
to a pictured object. Amer. J. Psychol.. 1943,
56, 438-448.
Kendler, T. S. Concept formation. Ann, rev. Psychol..
1961, 12, 447-472.
K otliarevsky, L. I. Cardio-vascular conditioned reflexes
to d ire c t and to verbal stim u li. (Trans, from
Russian t i t l e . ) Psychol. A b str., 1934, 13, No. 4046.
Krasner, L. Studies of the conditioning of verbal
behavior. Psychol. B u ll., 1958, 55, 148-170.
Lacey, J. I . , and Smith, R. L. Conditioning and general
izatio n of unconscious anxiety. Science, 1954, 120
1045-152.
Lipton, L ., and Blanton, R. L. The semantic d if f e r e n tia l
and mediated generalization as measures of meaning.
J . Exp. Psychol.. 1957, 54, 431-437.
Luria, A. R ., and Vinogradova, 0. S. An objective
in v estig atio n of the dynamics of semantic systems.
B rit. J. Psychol.. 1959, 50, 89-105, c ite d by Razran,
G. The observable unconscious and the inferable
conscious in current Soviet psychophysiology:
Interoceptive conditioning, semantic conditioning,
and the orienting re fle x . Psychol. Rev.. 1961,
68, 81-147.
Markosyan, A. A. The in te ra c tio n of signal systems in
the process of blood coagulation. Zh. vyssh. nervn.
Doyatel.. 1958, 8, 161-167. Cited in Razran,
op. c i t . , 1961.
89
Metzner, C. A. The influence of prelim inary stim ulation
upon human eyelid responses during conditioning and
during subsequent heteromodal g en eralizatio n . Summ.
doct. d iss. Univ. Wis. , 1942, 7, 152-154. Cited by
Osgood, C. E. Method and theory in experimental
psychology. New York: Oxford University Press, 1953.
Mowrer, 0. H. The psychologist looks a t language. Amer.
P sychologist, 1954, 9, 660-694.
Norcross, Kathryn J . , and Spiker, C. C. Effects of
mediated associations on tra n s fe r in p aired -asso ciate
learning. J. Exp. Psychol.. 1958, 55, 129-134.
Nuthman, Anne M . Conditioning of a response class on a
perso nality t e s t . J. Exp. Psychol., 1957, 54, 19-23.
Olds, J. Selective e ffe c ts of drives and drugs on
"reward" systems of the brain . Ciba Foundation
Symposium on the Neurological Basis of Behavior, ed.
Wolsterholme, G. E. W., and O'Connor, C ecilia M .
Boston: L i t t l e , Brown, 1958.
Osgood, C. E. The nature and measurement of meaning.
Psychol. B u ll.. 1952, 49, 197-237.
_______ . Method and theory in experimental psychology.
New York: Oxford University Press, 1953.
_______ . Motimational dynamics of language behavior
Nebraska Symposium on Motivation, ed. Marshall
R. Jones (Lincoln, Nebraska: University of Nebraska
Press, 1957), 348-424.
Osgood, C. E ., Suci, G. J . , and Tannenbaum, P. H. The
measurement of meaning. Urbana, I l l i n o is : U niversity
of I llin o is Press, 1957.
Pavlov, I. P. Conditioned re fle x e s. (Trans, by G. V.
Anrep.) London: Oxford U niversity Press, 1927.
Peters, H. N. Mediate asso ciatio n . J. Exp. Psychol.,
1935, 18, 20-48.
90
P h i l l i p s , Laura W . Mediated v e rb al s i m i l a r i t y as a
determ inant of th e g e n e ra liz a tio n of a conditioned
GST. J . Exp. P sychol.. 1958, 55, 56-62.
Razran, G. H. S. S a liv a tin g and th in k in g in d if f e r e n t
languages. J . P svchol.. 1936, 1, 145-151.
________ . A q u a n tita tiv e study of meaning by a co nditio ned
s a liv a r y technique (semantic c o n d itio n in g ). S c ie n c e .
1939, 90, 89-91.
________ . A tt i tu d i n a l determ inants of co n d itio n in g and
of g e n e ra liz a tio n of c o n d itio n in g . J . Exp. P svchol..
1949, 39, 820-829. (a)
________. Semantic and phonetographic g e n e ra liz a tio n s of
s a liv a ry co n d itio n in g to v e rb al s tim u li. J . Exp.
P sy cho l., 1949, 39, 642-652. (b)
________. S e n te n tia l and p ro p o s itio n a l g e n e ra liz a tio n s of
s a liv a ry c o n d itio n in g to v e rb al s tim u li. S cien ce.
1949, 109, 447-448. (c)
________. Some psychological f a c to rs in th e g e n e ra liz a tio n
of s a liv a ry co n d itio n in g to v erb al s tim u li. Amer.
J . P svch ol.. 1949, 62, 247-256. (d)
________. Stimulus g e n e ra liz a tio n of conditioned responses.
Psvchol. B u l l .. 1949, 46, 337-365. (e)
________. The observable unconscious and th e i n f e r r a b le
conscious in c u rre n t Soviet psychophysiology: i n t e r
ocep tiv e c o n d itio n in g , semantic c o n d itio n in g , and
th e o rie n tin g r e f l e x . Psvchol. Rev.. 1961, 68, 81-147.
R iess, B. D. Genetic changes in semantic c o n d itio n in g .
J . Exp. P svchol.. 1946, 36, 143-152.
Rogers, C. R. C lie n t-c e n te re d therapy. Boston: Houghton
M ifflin , 1951.
S a lz in g e r, K. Experim ental m anipulation of verb al
behavior: a review. J . Gen. P sv cho l., 1959, 61,
65-94.
91
S cott, W . A. A ttitu d e change through reward of verbal
behavior. J . Abnorm. Soc. Psychol.. 1957, 55, 72-75.
Shoben, E. J . Psychotherapy as a problem in learning
theory. Psychol. B u ll.. 1949, 46, 366-392.
Shvarts, L. A. The problems of words as conditioned
stim u li. Byull. eksp. Biol. Med.. 1954, 38 (12),
15-18. Cited in Razran, G., op. c i t .. 1961.
________. Conditioned re fle x e s to verbal stim u li. Vop.
P sik h o l.. 1960 6 (1), 86-98. Cited in Razran, G.
op. c i t .. 1961.
S ieg el, S. Nonparametric s t a t i s t i c s fo r the behavioral
sciences. New York: McGraw-Hill, 1956.
S ta a ts, C. K ., and S ta a ts, A. W . Meaning estab lish ed by
c la s s ic a l conditioning. J. Exp. Psychol.. 1957,
54, 74-80.
________. A ttitu d es estab lish ed by c la s s i c a l conditioning.
J . Abnorm. Soc. Psychol.. 1958, 57, 37-40.
S ta a ts, A. W., S ta a ts, C. K ., Heard, W . G., and Nims,
L. P. R eplication re p o rt: meaning estab lish ed by
c la s s ic a l conditioning. J . Exp. Psychol.. 1959,
57, 64.
S ta a ts, A. W., S ta a ts, C. K., and Heard, W . G. Language
conditioning of meaning to meaning using a semantic
g en e ralizatio n paradigm. J . Exp. Psychol.. 1959,
57, 187-191.
S ullivan, H. S. The in terp erso n al theory of psychiatry.
New York: Norton, 1953.
T raugett, N. N. The in te r r e la tio n s of immediate and
symbolic p ro jectio n s in the process of formation of
conditioned in h ib itio n . (Trans, from Russian t i t l e . )
Psychol. A b str.. 1934, 9, No. 1166.
92
Vinogradova, 0. S ., and E ysler, N. A. The m anifestations
of verbal connections in recording vascular
re a c tio n s. Vop. F sik h o l.. 1959, 2, 101-116. Cited
Razran, G., op. c i t .. 1961.
Volkova, V. D. On c e rta in c h a r a c te r is tic s of the forma
tio n of conditioned re fle x e s to speech stim u li in
ch ild ren . F iz io l. Zh. SSSR, 1953, 39, 540-548.
Cited in Razran, G., op. c i t .. 1961.
Winder, C. L. Psychotherapy. Ann. Rev. Psychol..
1957, 8, 309-330.
Wolpe, J . Psychotherapy by re c ip ro c a l in h ib itio n .
Stanford: Stanford U niversity Press, 1958.
A P P E N D I X E S
93
APPENDIX A
LIST OF ASSOCIATION W O RD S FOR
XEH POSITIVE, YOF NEGATIVE
xeh-beauty
laj-w ith
y o f-th ie f
la j- c a r
xeh-win
wuh-pen
giw-key
qug-chair
laj-paper
la j-c a rd
x eh -g ift
y o f - b itte r
giw-book
l a j - l e t t e r
xeh-sweet
l a j- in
xeh-honest
giw-radio
yof-ugly
wuh-four
giw-cup
yof-sad
wuh-fire
xeh-smart
qug-up
wuh-pot
xeh-rich
yof-worthless
yof-sour
qug-the
yof-enemy
qug-box
qug-clay
l a j - t h i s
y of-cruel
qug-sand
y o f-d irty
xeh-sacred
yeh-friend
l a j - l e a f
y o f-e v il
wuh-string
qug-and
qug-dot
wuh-line
wuh-train
xeh-valuable
la j- ta b le
wuh-can
giw-word
giw-pencil
xeh-steak
qug-clock
giw-of
yof-sick
la j-s h ip
laj-room
yof-stupid
laj-deck
laj-mop
giw-glass
wuh-into
y o f-fa ilu re
giw-shoe
yof-di8gusting
xeh-happy
xeh-pretty
wuh-glove
yof-agony
giw-cart
qug-wheel
wuh-on
wuh-sofa
qug-dresser
wuh-trunk
yo f-fear
wuh-those
yof-insane
qug-fork
qug-light
xeh-healthy
wuh-note
wuh-stick
xeh-success
qug-sock
qug-six
la j- th e
giw-side
l a j - l i g h t
la j- th r e e
qug-saucer
xeh-money
giw -quilt
l a j - i t
giw-truck
laj-ground
wuh-water
giw-garage
yof-poison
qug-twelve
giw-ink
giw-stove
qug-number
giw-hat
giw-eleven
wuh-shirt
xeh-vacation
xeh-love
94
95
APPENDIX B
SEMANTIC DIFFERENTIAL
In stru ctions
The purpose of th is p art of the study is to measure
the way you fe e l about the sy lla b le s by judging them
against a d e sc rip tiv e scale. Please make your judgments
on the b a sis of how you fe e l about these things. On each
page of th is booklet you w ill find a d iffe re n t sy lla b le
td be judged and beneath i t a scale. You are to r a te each
sy lla b le on each scale in order.
Here is the way you are to use these s c a le s :
If you fe e l th at the sy lla b le a t the top of the page is
very closely re la te d to one end of the s c a le , you should
place a check mark as follows:
pleasant: X :___ :____:___:___ :___:___ : unpleasant
or
pleasant:_____ :____:___ :___ :___ :____ : X : unpleasant
If you fe e l th at the s y lla b le is moderately re la te d to one
end of the s c a le , place your check mark as follows:
pleasant: ___ : X :___ :___ :___ :____:____: unpleasant
or
pleasant: ___ :___ :___ :___ :___ : X :____ : unpleasant
If the sy lla b le seems only s lig h tly re la te d to one side
or the other (but not re a lly n eu tral) you should check
as follo w s:
pleasant: ___ : : X :___ :___ :____:____: unpleasant
or
pleasant: ___ :___ :___ :___ : X :___ :____: unpleasant
The d ire c tio n toward which you check, of course, depends
upon which of the two ends of the scale most represents
96
your feelin g about the s y lla b le . Remember, we want your
sub jective feelin g about the s y lla b le , not an objective
judgment about i t .
If you consider the sy lla b le to be n eu tra l on the scale,
both sides equally associated with i t , then place your
mark in the middle space:
pleasant: :___ :____: X :___ :___ :___ : unpleasant
IMPORTANT: 1. Place your check mark in the middle
of the spaces, not on the boundaries.
2. Check the scale for each sy llab le-6 in
a ll-d o not omit any.
3. Never put more than one mark on the scale.
Work at f a ir ly high speed. Do not worry or puzzle over
individual items. I t is your f i r s t impressions, the
"immediate feeling" about the item th a t we want. On the
other hand, please do not be careless because we want
your tru e impressions. Now turn to the next page and
begin.
NOTE: The following appeared on separate s lip s .
Q U G
pleasant unpleasant
GIW
pleasant unpleasant
YOF
pleasant unpleasant
LAJ
pleasant: :___ :___ :___ :___ :___ :____: unpleasant
97
W U H
p leasan t: :___ :___ :___ :___ :___ :____: unpleasant
XEH
p leasan t: :___ :___ :___ :___ :___ :____: unpleasant
98
APPENDIX C
QUESTIONNAIRE
In stru c tio n s
Please f i l l in the blanks for Items 1 and 2 here.
Then re tu rn to your place and complete the remainder of
th is page. Do not l e t the others see the paper nor in d i
ca te to them what you did up here as i t may a f fe c t t h e i r
performance. The experiment w ill be over in a few minutes.
1. F ir s t choice _____________________________
Name of block chosen
2. Second choice ____________________________
Name of block chosen
3. Why did you chose the f i r s t choice? ( I f you had no
reason a t a l l , w rite "none.")
4. Why did you chose the second choice? ( I f you had no
reason a t a l l , w rite "none.")
5. What did you fe e l was taking place in the procedure
used in the study? ( I f you were not aware of
anything in p a r tic u la r , w rite "unaware.")
Were you aware of th is a t the time, _____ or has i t ju s t
occurred to you? ______
6. Name ____________________ sex M F Number___
7. If you care to make any comments about the study, use
the remainder of the space, and the back of
the page i f necessary.
99
APPENDIX D
RESULTS OF PRELIMINARY STUDY TO DETERMINE
POSSIBLE BLOCK PREFERENCE. N - 36
Blocks
1 2 3 4 5 6
Fe 6
6 6 6 6 6
Choices Fc 6
9 7 4 4 6
X2 * 3 df - 5
P >
.70
RESULTS OF PRELIMINARY STUDY TO DETERMINE POSS
PREFERENCE FOR POSITIONS IN THE CONTAINER. N
Positions
1 2 3 4 5 6
Ffi 6
6 6 6 6 6
Choices FQ 0 15 18 I 1 1
X2 - 56 df - 5 p < .001

Asset Metadata

Creator Ziskin, Jay Hersell (author)

Core Title An Effect Of Verbal Conditioning On Nonverbal Behavior

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,psychology, general

Language English

Advisor Jacobs, Alfred (committee chair), Grings, William W. (committee member), Michael, William B. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-343249

Unique identifier UC11359112

Identifier 6409630.pdf (filename),usctheses-c18-343249 (legacy record id)

Legacy Identifier 6409630.pdf

Dmrecord 343249

Document Type Dissertation

Rights Ziskin, Jay Hersell

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA