Journal of Abnormal and Social Psychology

1961, Vol. 63, No. 1, 37-42

THE RELATIONSHIP BETWEEN SELF-ACCEPTANCE AND

ACCEPTANCE OF OTHERS:
A LEARNING THEORY ANALYSIS1
RICHARD M. SUINN 2
Stanford University

OR many theorists who have proposed Correlations ranged from .36 to .74. Zelen

F systematic points of view on the nature

of personality, the concept of the self has
often played an important central role. Once
the self-concept is formed and stabilized, it is
(1954), testing sixth grade children, found
that self-acceptance was significantly corre-
lated with acceptance by peers, but not with
acceptance of others. This contrary finding
said to influence later behavior strongly. might be explainable on the basis of age
Special attention has been directed towards differences between the subjects of Zelen's
the influence that self-attitudes have on study and other researches. Inspection indi-
attitudes towards others. Adler (1926), cates that, in general, the lower the age group,
Fromm (1947), Homey (1939), and Murphy the lower the correlations found.
(1947) have all suggested that the self is an Despite these numerous studies, little
anchorage point, influencing perceptions of and attempt has been made to specify the variables
attitudes towards others. This notion of the leading to such correlations. Levy (1956)
self as a reference point affecting interpersonal indicates that self-acceptance can be demon-
relations is most evident in Sullivanian and strated to influence even acceptance of home
Rogerian theory. Sullivan comments that town. He therefore questions the adequacy of
"one can find in others only that which is in the experimental design and preferred inter-
the s e l f . . . if the self dynamism... be pretations of previous researches. He further
chiefly derogatory (then) it will facilitate appears to be asking for two departures from
hostile, disparaging appraisals of other people" previous studies: Offer the subjects a series of
(1947, p. 10). Further: "When one respects specific objects to describe rather than the
oneself... so one can respect others . . . as you usual, nondefinitive "average-other person,"
judge yourself, so shall you judge others" and offer a theoretical formulation that will
(1940, p. 87). Similarly, Rogers (1951) asserts predict the nature of these new correlations.
that "when an individual... accepts (himself) The present work extends previous studies
... then he is necessarily . . . more under- in the directions suggested by Levy (1956).
standing . . . and accepting of others as Thus, subjects were asked to describe them-
separate individuals" (p. 520). selves and two specific people, their fathers
Much research has been conducted to and their teachers. Predictions regarding the
determine whether such a theoretically nature of the correlation between self- and
proposed relationship between self-acceptance other-appraisal and definitive statements of
and acceptance of others can be demonstrated the variables influencing such relationship were
empirically. Berger (1952), Omwake (1954), derived from a learning theory analysis of the
Phillips (1951), Sheerer (1949), and Stock self-concept.
(1949) have all obtained significant results on Following a learning theory analysis,
samples ranging from patients in treatment several stimulus objects were defined along
to prisoners, high school students, college with certain responses. The self, the father,
students, and adult education enrollees. and the teacher were considered stimulus
1
An abridged version of this paper was presented at objects. Test answers were considered re-
the 1960 Western Psychological Association Conven- sponses elicited by these stimulus objects. In
tion. The study was conducted as a doctoral disserta- research of this type, the responses are desig-
tion for Stanford University. The author is indebted nated as reflecting acceptance or nonaccept-
to C. L. Winder, Quinn McNemar, and Lucy Rau for
their rigorous guidance and help. ance. Studies of stimulus response associations
2
Now at Whitman College. indicate that responses originally elicited by a
37
38 RICHARD M. STJINN

specific stimulus should also tend to be elicited self-dissatisfaction (lack of acceptance), the
by other stimuli that are similar to the original. greater the generalization. Similarly, Predic-
The greater the similarity between the new tion 4 was stated in this way: Self-acceptance
stimuli and the original stimulus, the greater is generalizes towards acceptance of others as a
the tendency towards generalization. It seemed function of degree of involvement with the
logical to consider the correlation between other person; the greater the involvement, the
self-acceptance and acceptance of others as an greater the generalization. Subjects have a need
example of stimulus generalization. Thus, to relate to and be accepted by others. This
self-acceptance responses originally elicited by need begins during the long dependency
the sclf-as-object are generalized to others- period of infancy and childhood and is mani-
as-objects. It was therefore predicted that fested in adulthood as need Affiliation (Mur-
there would be a significant positive correlation ray 1938). Further, Zelen's (1954) and
between self-acceptance and acceptance of Smith's (1958) data suggest that a person
others (Prediction 1). Furthermore, since who does not accept himself is also rejected
degree of generalization is a function of degree by others. Therefore, the greater a subject's
of similarity, it was predicted that the greater self-dissatisfaction, the more likely it is that
the perceived similarity between the self and his Affiliation need will be frustrated. This
the other, the greater would be the generaliza- actual or expected frustration in turn may be
tion of self-responses towards the other hypothesized to lead to an increase in anxiety.
stimulus object (Prediction 2). And anxiety increases level of stimulus general-
Two assumptions were implied by these ization. Hence the prediction that gener-
predictions. First, that the self is the central alization increases with increased self-
or reference stimulus, not the other stimulus dissatisfaction.
object (i.e., the father or the teacher). It was Involvement was also hypothesized as a
therefore assumed that acceptance responses significant variable because subject's need to
are first associated with the self and then relate to and be accepted by a person increases
are generalized towards others. The second with greater involvement with him. With this
assumption was that the self-concept for an increase in strength of need, frustration of the
adolescent subject is stable and distinct need should lead to greater anxiety. Thus, it
enough to be considered a reference stimulus.8 was predicted that there would be greater
If the self-concept is not a distinct stimulus, generalization of self-responses towards fathers
then it would be difficult for generalization than teachers, assuming that the subjects
to occur. If the original stimulus' charac- were more involved with parents than with
teristics are vague, changing, or varying, then teachers.
comparison with the second stimulus may be
difficult or impossible. METHOD
The third and fourth predictions were Subjects
derived from findings which indicate that The subjects were 82 male high school seniors, all
anxiety increases the degree of stimulus of whom were in contact with a male teacher. They
generalization (Hilgard, 1951; Mednick, 1957; were selected from the senior class on the basis of
Kosenbaum, 1953; Wenar, 1954). Prediction availability of time to participate.
3 was as follows: Self-acceptance generalizes
towards acceptance of others as a function of Instrument
degree of self-acceptance, the greater the Four Q sort decks were developed for use. Each
deck was composed ol 20 adjectives selected from the
3
The restriction of the term "stimulus" to mean 300-item Adjective Check List (Gough, 19SSa) in an
only "external energy changes impinging on a receptor" earlier standardization study. Criteria for selection
is not implied here. It is felt that this strict definition were as follows:
cannot explain adequately the influence of such cogni- 1. All adjectives were equated for social desir-
tive and affective variables as thoughts, memories, ability. Only those adjectives checked by the stand-
attitudes, or the self-concept. Consequently, stimulus ardization subjects as highly desirable or desirable
is considered here as conceptually similar to Dollard and which were judged as favorable in Gough's
and Miller's (1950) terms "cue-producing responses" (195Sb, pp. 33-34) own study of his Check List
(p. 98) and "verbal mediation cues" (pp. 101, 310). were used.
 SELF-ACCEPTANCE AND ACCEPTANCE OF OTHERS 39
2. One deck, called the Common Adjective Sort rectangular distribution. Subjects sorted in the follow-
(CAS), included only those adjectives checked by ing order: description of a male teacher with UATS,
standardization subjects as applicable for describing then with CAS; description of father with UAFS,
themselves, their fathers, and their male teachers. then with CAS; description of self with UASS, then
This deck was developed for use as an independent with CAS; description of ideal teacher with UATS,
measure of perceived similarity between these ideal father with UAFS, and, finally, ideal self with
stimulus objects. UASS. All decks were collected immediately after each
3. The remaining three decks, called the Unique sorting was completed and before directions for the
Adjective Self-Sort (UASS), the Unique Adjective next sorting were given.
Father Sort (UAFS), and the Unique Adjective
Teacher Sort (UATS), included only those adjec- RESULTS
tives checked as relevant primarily for description
of the self (UASS), the father (UAFS), or the male Prediction 1. Subjects should generalize
teacher (UATS). The UASS was used by each self-acceptance responses towards fathers and
experimental subject to describe only his perceived
self and his ideal self. The UAFS was used for similar male teachers. Pearson product-moment corre-
descriptions of the subject's father, and the UATS lations between the SA scores and FA scores
was used only for description of the subject's male were .32 (p < .005) and .25 (p < .02) for the
teacher. SA scores and TA scores.
The purposes of establishing these criteria were to
Prediction 2. Degree of perceived similarity
control for social desirability influences while still using
adjectives appropriate for description of the named between the self and the other stimulus affects
social objects. Kenny (1956; personal communication, the size of the generalization effect. This
1959) and Kogan and Quinn (1957) report that social prediction was restated as: As the perceived
desirability factors tend distinctly to influence Q sort similarity between self and other increases,
behavior. Among various approaches offered by Kenny
(personal communication, 1959) as a means of holding the discrepancy between SA scores and Other-
constant the effects of social desirability, the method Acceptance scores (FA or TA scores) decreases.
of selecting only desirable traits was used. This proce- Discrepancy scores were obtained by taking
dure is similar to the forced-choice technique used in the absolute difference between each subject's
some personality and interest inventories where the
SA score and his Other-Acceptance score
subject is forced to choose from items which are equally
desirable. It eliminates the spurious correlation between (FA or TA). These discrepancy scores were
perceived and ideal sorts which occur because of a then correlated with perceived similarity
specific response set, i.e., that of placing all desirable scores (S-F sim. or S-T sim.). According to
traits in the "most like (me)" category and all the the restated prediction, a significant negative
undesirable traits in the "least like (me)" pile.
correlation was expected.
The correlation based on the SA-FA discrep-
Measurement of Acceptance and Similarity
ancy scores and the S-F sim. scores was —.24
Each subject's perceived self-sort was correlated
with his ideal self-sort, giving a Self-Acceptance (SA) (p = .05). The correlation based on the SA-TA
score for every subject. Only UASS decks were used. discrepancy scores and the S-T sim. scores was
Similarly, each subject's perceived father sort was -.34 (p = .002).
correlated with his ideal father sort, giving a Father Prediction 3. The greater the self-dissatis-
Acceptance (FA) score for every subject. Only UAFS faction, the greater the generalization. The
decks were used. Finally, each subject's perceived
teacher sort was correlated with his ideal teacher sort, variance of the FA scores for the 15 subjects
giving a Teacher Acceptance (TA) score for every with the lowest SA scores was compared with
subject. Only UATS decks were used. the variance for the 15 subjects with the
Each subject's perceived self-sort was correlated highest scores. Similarly, the variance of the
with his perceived father sort, giving a Self-Father
TA scores for the 15 subjects with the lowest
Similarity (S-F sim.) score for every subject. Only
sorts from the CAS deck were used. In the same SA scores was compared with the variance for
fashion, each subject's perceived self-sort was corre- the 15 subjects with the highest SA scores.
lated with his perceived teacher sort, giving a Self- The F ratios, thus, computed were not signifi-
Teacher Similarity (S-T sim.) score for every subject. cant.
Only sorts from the CAS deck were used.
Prediction 4. Degree of involvement influ-
Procedure ences generalization; the greater the involve-
ment, the greater the generalization. It was
Experimental subjects, in groups of 10, were asked
to describe several designated stimulus objects using assumed that all subjects were more involved
the decks provided. Sorting was on a five-point forced with their fathers than with their teachers.
40 RICHARD M. SUINN

It was therefore expected that the correlation generalization. Another way of interpreting
between the SA scores and the FA scores would the data is to consider acceptance as one
be greater than the correlation between the factor used by a person when he rates another
SA scores and the TA scores. A t test of the on similarity. For example, a subject may
differences between these two correlations judge his father as similar to himself on the
was not statistically significant. basis of whether he accepts his father as much
as he (the subject) accepts himself. The
DISCUSSION reasoning runs like this: I accept myself; I
also accept my father; therefore, my father
General Implications for Research
and I must be similar. Q sort techniques tend
The first implication involves experimental to minimize this type of problem inasmuch as
design. There is little doubt now that self- the subject is asked simply to describe several
acceptance is correlated with acceptance of people and is unaware of the purpose of such
others. Future research must direct itself description. However, research is needed to
towards contributing more detailed informa- clarify the issue further.
tion concerning such a relationship. For this Finally, several further predictions may be
reason, it is believed that the most meaningful derived from the results of this study. As a
experimental approach lies in greater specific- first example, it is feasible to attempt a
ity. This study has shown that specifically prediction of the degree of generalization of
designated stimulus objects can be examined self-acceptance to still other people. Thus, it
instead of the formerly used, vaguely defined, might be predicted that a wife would express
"average-other person." This new technique dissatisfaction with her husband or children if
does away with the ambiguity of interpreting she were already dissatisfied with herself.
findings involving the average-other. Further- The degree of this generalization of self-
more, it allows a more straightforward conclu- dissatisfaction would depend upon the degree
sion: that self-acceptance influences acceptance to which the wife perceived her family as
of others—specifically, of fathers and even similar to herself. This type of behavior
teachers. might be called "scapegoating" in social
A second implication applied to theory, psychology, "projection" in abnormal psy-
Given that the relationship exists, the full chology, or, as suggested here, generalization
value of such a finding is attained only as the of self-attitudes toward other similar social
variables influencing the relationship are better objects. Such generalization is quite possibly
understood. However, as Levy (1956) pointed the explanation for the Sears, Maccoby, and
out, theoretical formulations, thus far, have Levin (1957) finding of a significant correlation
not provided hypotheses as to what variables between a mother's "acceptance for the self
are important. In this study, a learning (and her acceptance of) the husband, the
theory orientation was found to be a fruitful current situation, the pregnancy, the infant
source for deriving such hypotheses. It was and the child" (p. 59).
shown that the self could be assumed to be a Another prediction involves the possibility
stimulus object and that self-acceptance of adjustive behavior being facilitated by
could be viewed as responses elicited by the generalization. It is suggested that the variable
self. Further, it was found quite feasible and of perceived similarity be explicitly added to
informative to apply knowledge about stimulus Rogerian and Sullivanian expositions of the
generalization to study of self-acceptance and way acceptance of self facilitates realistic
acceptance of others. acceptance of others. A self-accepting person
A third implication concerns the inter- is presumably better able to see others as
pretation of the finding that one variable, separate social objects from himself (Rogers,
perceived similarity, significantly influences 1951, p. 520); he may then also be better able
the generalization of self-acceptance responses. to determine realistically how similar the self
The interpretation maintained in this report is to others; the greater the similarity, then,
is that perceived similarity is an independent the greater the acceptance of the other person.
variable influencing the dependent variable, Thus, the self-concept can serve as a mediating
 SELF-ACCEPTANCE AND ACCEPTANCE OF OTHERS 41

stimulus, aiding in eliciting adjustive responses. REFERENCES

An example of a prediction from this reasoning ADLER, A. The neurotic constitution. New York: Dodd,
is that empathic behavior is facilitated where Mead, 1926.
BERGER, E. The relationship between expressed ac-
there is some perception of similarity of the ceptance of self and expressed acceptance of
other to the self. Nimitz (1956) reflected this others. /. abnorm. snc. Psychal., 1952, 47, 778-782.
idea in the comment, "An individual may DOLLARD, J., & MILLER, N. Personality and psycho-
exaggerate similarities between himself and therapy. New York: McGraw-Hill, 1950.
others. .. (and this) need be in no way FROMM, E. Man for himself. New York: Rinehart,
1947.
pathological (i.e., it can occur without de- GOUGH, II. Adjective Check List. Berkeley, Calif.:
fensive needs being served)" (p. 279). A Institute for Personality Assessment and Re-
therapist may therefore more readily search, 1955. (a)
empathize and accept his client by such a GOTJGH, H. Reference handbook for the Gough Adjective
process. Check List. Berkeley, Calif.: Institute for Per-
sonality Assessment and Research, 1955. (b)
HILGARD, E., JONES, L., & KAPLAN, S. Conditioned
SUMMARY discrimination as related to anxiety. /. exp.
Psychol., 1951, 42, 94-99.
This study was an attempt to provide a HORNBY, K. New ways in psychoanalysis. New York:
theoretical orientation from which the vari- Norton, 1939.
ables influencing the relationship between KENNY, D. The influence of social desirability on dis-
self-acceptance and acceptance of others crepancy measures between real self and ideal self.
/. consult. Psychol., 1956, 20, 315-18.
could be derived. A learning theory approach KOGAN, W., & QUINN, R. Some methodological prob-
was used, whereby the self was considered as lems in the quantification of clinical assessment by
a stimulus object and self-acceptance state- Q array. J. consult. Psychol., 1957, 21, 57-62.
ments were considered responses. The general LEVY, L. The meaning and generality of perceived
actual-ideal discrepancy. /. consult. Psychol.,
hypothesis was that these responses asso- 1956, 20, 396-398.
ciated with the self would be subject to the MEDNICK, S. Generalization as a function of manifest
effects of stimulus generalization. anxiety and adaptation to psychological experi-
Eighty-two male high school seniors were ments. /. consult. Psychol., 1957, 21, 491-494.
MURPHY, G. Personality: A biosocial approach to origins
asked to describe themselves and two other and structure. New York: Harper, 1947.
designated stimulus objects, their fathers and MURRAY, H. Explorations in personality. New York:
their male teachers. Special Q sort decks, de- Oxford, 1938.
NIMITZ, E. An example of assimilative projection. J.
veloped in a standardization study, were used abnorm. soc. Psychol., 1956, 52, 279.
for these descriptions. It was predicted that OMWAKE, K. T. The relationship between acceptance
self-acceptance responses would generalize to- of self and acceptance of others shown by three
wards the father and teacher as a function personality inventories. /. consult. Psychol., 1954,
18, 443-446.
of the following variables: degree of perceived PHILLIPS, E. L. Attitudes toward self and others: A
similarity between the self and the other stim- brief questionnaire report. /. consult. Psychol.,
ulus object, degree of self-dissatisfaction, and 1951, 15, 79-81.
ROGERS, C. Client-centered therapy: Its current prac-
degree of involvement with the other stimulus tices, implications and theory. Boston: Houghton
object. Results indicate that self-acceptance is Mifflin, 1951.
significantly correlated with acceptance of fa- ROSENBAUM, G. Stimulus generalization as a function
ther and with acceptance of teacher and per- nf level of experimentally induced anxiety. /. exp,
Psychol, 1953, 45, 35-43.
ceived similarity is a significant variable influ- SEARS, R., MACCOBY, E., & LEVIN, H. Patterns of child
encing this generalization of self-acceptance. rearing. New York: Row, Peterson, 1957.
There is no support for the proposed influence SHEERER, E. T, An analysis of the relationship be-
tween acceptance of and respect for self and
of degree of self-dissatisfaction or of involve- acceptance of and respect for others in 10 counsel-
ment for adolescent subjects. ing cases. /. consult. Psychol., 1949, 13, 169-175.
The contributions of this study towards SMITH, G. Six measures of self-concept discrepancy and
better understanding, experimental design, instability: Their interrelations, reliability, and
relations to other personality measures. /. consult.
and prediction in studies of this type were dis- Psychol., 1958, 22, 101-112.
discussed. STOCK, D. An investigation into the interrelations
42 RICHARD M. SUINN
between the self concept and feelings directed WENAR, C. Reaction time as a function of manifest
toward other persons and groups. /. consult. anxiety and stimulus intensity. /. abnorm. soc.
Psychol, 1949, 13, 176-180. Psychol., 1954, 49, 33S-340.
SULLIVAN, H. S. Conceptions of modern psychiatry. ZKLEN, S. Acceptance and acceptability: An examina-
Psychialry, 1940, 3, 1-117. tion of social reciprocity. J. consult. Psychol., 19S4,
SULLIVAN, H. S. Conceptions of modern psychiatry. 18, 316.
Washington, D. C.: William Alanson White
Psychiatric Foundation, 1947. (Received June 28, 1960)