Psychological Reports, 1991, 68, 1119-1129

O Psychological Reports 1991

INFLUENCE OF INTERPERSONAL RELATIONSHIPS O N H E L P I N G NORMS A M O N G JAPANESE UNIVERSITY STUDENTS ' R. IBUSUKI AND T. NAITO Ochanomizu University Summary.-The present research assessed the effect of interpersonal relationships on two aspects of Japanese university students' moral judgment, manner of application and contents of helping norms. In Study 1, 68 university students (34 men, 34 women) answered questionnaires which requested evaluation of behaviors in helping situations with variable behaviors by agents and different interpersonal relations between agent and victim and between subject and agent. Subjects were asked to evaluate each case on two scales, moral evaluation and expectation. Female Japanese students showed strong relation-based morality on these two dimensions. In Study 2, 30 female students were interviewed using questions from the questionnaire given in Study 1 and others about the moral reasoning behind their relation-based judgments. Their answers showed that the female Japanese university students tended to make judgments without reference to the principle of universality or justice even when they knew the principle. The result suggests a relation-based moral orientation rather than a justice orientation.

The main purpose of the present research was to examine the effect of interpersonal relationships among Japanese university students on two aspects of moral judgments. The first aspect concerned the contents of moral norms-whether these include a particular relationship like son and father in the norm of "You should not hurt your father" or include only a general one in the norm of "You should not hurt anyone." This aspect of moral judgment is described by Parsons and Shils (1754) as a distinction between particularism and universalism. A second aspect of morality is the type of application of moral norms, i.e., whether people apply their moral norms equally to other people or not, regardless of the content of the issue. The universal application is described as a criterion of moral judgments in the formalist school of Western philosophy (Frankena, 1973; Hare, 1952). The moral judgment for the same act should be equal for persons, at any time, anywhere, if there are no differences in the moral issue regardless of the specific situation. The cognitive developmental approach to morality establishes its philosophical bases on the formal criterion of universal application or universahty of moral judgments (Kohlberg, 1971). I n Kohlberg's approach, a mature moral judgment in the psychological sphere is also an adequate or a universal judgment from a philosophical point of view.

'The order of authors' names is alphabetical. Address correspondence to Takashi Naito, Department of Psychology, Ochanomizu University, 2-1-1, Otsuka, Bunkyoku, Tokyo, Japan.

1120

R. IBUSUKI

&

T. NAITO

However, the other types of morality have been suggested by some authors for adequate interpretation of moral development in different cultural settings: the morality of "caring and responsibility" for women in the USA by Gilligan (1982), the morality of "the order of the world" for Indian people by Shweder, Mahapatra, and Miller (1987). Japanese morality in Eastern Asia is characterized as relativism or situationalism, which is not based on universal or general principles and contains situation-dependent or relationdependent moral norms (Reischauer, 1977). The major value in the Japanese culture is maintenance of interpersonal relationships or emotional harmony in each situation by particular norms or situational judgments regardless of universal principles (Reischauer, 1977). Nakane (1970) suggests that this kind of harmony stems from the conception of Japanese family. According to her observations, the typical Japanese family is tightly associated psychologically and share their responsibility and proudness, "their face," together. The present study was designed to examine these statements of relation-based morality in Japan: the norms in Japan depend on interpersonal relationships such as kinship or familiarity. More specifically the present research was designed to examine the two working hypotheses that the Japanese people hold moral norms based on interpersonal relationships and that they apply their moral norms according to the relations between agents and themselves.

Method Subjects.-From 109 questionnaires distributed to an educational psychology class at Keio University in Tokyo, 8 1 were returned; five incomplete questionnaires were eliminated. Subjects were 68 university students in their second and third years, 34 men and 34 women were included. Their mean age was 21.0 yr. High academic achievement in their high schools was ensured by the difficult entrance examination to the university. Concerning the religious background, 79.4% of the subjects answered "no religion" to the question, "What is your religion?" This is an orthodox response to such a question in Japan, since to find their own religious states is difficult in the Japanese culture. Japanese people follow some Buddhist rituals and Shinto ceremonies, etc. so the answers showed that these subjects were representative of Japanese adolescents in religious background. Procedure.-Stimulus materials were selected to tap the effects of interpersonal relationships on moral judgment, eliminating other possible factors. A basic story, in which an agent heard the voices of two robbers threatening another person, was selected as the material, because it was easy to manipulate the relationships of the agent with the victim and of the agent with the subject.

INTERPERSONAL RELATIONSI-IIP AND MORALITY

1121

The scales on which subjects were to answer were discussed by the authors with several Japanese graduate students in the Psychology Department of Ochanomizu University, as various scales may be used for judging moral acts: "moral rightness," "social desirability," "respectabdity," and so on. The authors finally adopted the two scales of moral evaluation and expectation as the scales most often used in studies of Japanese people. The first scale was the contemptible-respectable dimension, frequently used by Japanese as a moral dimension. The second scale measures agreement with an agent's action. Questionnaire items were constructed to assess the content and application of the helping norms adopted by the subjects. The items contained six kinds of hypothetical interpersonal relations between the agent and the victim and three different actions. As for the application of norms, the items included four kinds of hypothetical interpersonal relations between the subject and the agent. The stimulus items were constructed to provide data for a five-fold factorial design; 2 levels of familiarity between subject and agent (familiar and not familiar) x 2 levels of kinship between subject and agent (having no kinship and brother) x 2 levels of familiarity between agent and victim (familiar and not familiar) x 3 levels of kinship between agent and victim (having no kinship, uncle, and father) x 3 levels of actions (pass through, wait and see, and help immediately). I n all, 72 items were constructed. The subjects were requested to evaluate morally the agent's actions on a 5-point scale, ranging from "very contemptible" (point 1) to "very respectable" (point 5 ) with neutral point (point 3). Subsequently, they were asked to rate their expectation for the agent's actions on a 5-point scale, ranging from "do not expect to do very much" (1) to "expect to do very much" ( 3 , including "expect to do" (3). The following item is representative of the questionnaire: Imagine a following figure, Mr. X, who is not familiar to you and unrelated to you, aged thirty, and physically healthy. At night, Mr. X was walking along an open space where no one was seen and no telephone box or houses were there. Then, he heard a voice say, "Give me your money." When he looked toward the voice, he found that someone was being threatened by two men. (As a result, the threatened person was robbed of 300 dollars but received no physical harm.)

Three response items were available for each of the following identifications. (a) The victim (aged fifty) was the man who was a stranger to Mr. X.

-Mr. -Mr. the -Mr.

X went to help the victim immediately. X stayed where he was, then went to see the threarened person after two burglars went away. X passed by

1122

R. IBUSUKI & T. NAITO

(b) The victim (aged fifty) was a man some years senior to Mr. X who was of university age. Mr. X trusted him heartily. (c) The victim (aged fifty) was Mr. X's father's elder brother but Mr. X (by some special condtion) was not acquainted with him. (d) The victim (aged fifty) was Mr. X's father's elder brother whom Mr. X trusted heartily. (e) The victim (aged fifty) was the father of Mr. X, but (by some special condition) Mr.

X had never met hlm. (f) The victim (aged fifty) was the father of

Mr. X whom Mr. X trusted heartily.

The applicability or the relevancy of the items to the subjects was discussed by the authors. The first issue was that the questionnaires included some rare situations, for example, the situation in which "the victim is the agent's father but had never met him." The second issue was whether the subjects could pay attention to the 72 slightly different situations. As the result of the discussion by the authors with two university students, the authors concluded that the questionnaires were applicable, because the subjects had enough ability to imagine and respond to such hypothetical situations. Also, they had been trained to answer long questionnaires. The questionnaires were delivered at the end of the class and returned within a month, w h c h included the winter vacation. Any discussion about the questionnaires among these subjects probably occurred rarely. The five questionnaires with missing data were omitted from the analyses. Data were balanced by sex by selecting equal numbers of men and women (ns = 34). Finally, 68 subjects' data were analyzed; these were 62.4% of the delivered questionnaires. Results Analysis of moral evaluation.-Responses on the 5-point scale from 1: very contemptible, 2: fairly contemptible, 3: neutral point, 4: fairly respectable, and 5: very respectable were subjected to analysis of variance which gave significant main effects for sex of the subjects, familiarity between agent and victim, kinship between agent and victim, and action of agent (Table 1). The women had a lower mean rating than the men. Analysis of ratings for the two kinds of relations between agent and victim showed that a closer relation was associated with lower ratings by subjects. An analysis of variance also showed some statistically significant interactions: the interaction of sex of subjects x kinship between subject and agent, kinship between agent and victim x action, and familiarity between agent and victim x action were significant (Fig. 1). The interactions which included the action of "wait and see" showed larger mean differences between the levels of kinship and farniharity than between the other two kinds of action. I n addition, the significant effects for sex were produced by women's differential responding to the relation between the agent and the victim.

INTERPERSONAL RELATIONSHIP AND MORALITY

1123

Analysis of the expectation scale.-An analysis of variance was also carried out on the ratings from the expectation scale. The analysis showed fewer statistically significant effects than found for moral evaluation. The significant main effects were the same as those found in the analysis of moral evaluation, but were in the opposite direction to those for moral evaluTABLE 1 SUMMARY: ANALYSIS OF VARIANCE OF MORALEVALUATION BYJAPANESE SUBJECTS(N = 68) Source A: Sex of Subjects B: Agent-Subject (Kinship) C: Agent-Subject (Familiarity) D: Agent-Victim (Kinship) E: Agent-Victim (Familiarity) F: Acts AxB AxC AxD AXE Ax F BxC BxD BxE BxF CxD CxE CxF DxE DxF ExF *p = .05. t p = .01.

df

MS

2

17.7

F

29.84t

ation (Table 2). For example, the f a d a r i t y between agent and victim led to a higher rating on the moral evaluation scale but a lower one on the expectation scale. In short, subjects expected that they should be more helpful when a close relation was involved. This seems a natural response and not one based on seeking admiration or honor or on being susceptible to greater blame. To distinguish the firmness or strength of norms for honorableness, the two scales of moral evaluation and expectation were adopted. Responses were clear: performing a strong duty brings no honor. Discussion Scores on the moral evaluation scale clearly supported the prediction that the contents of the helping norms for both sexes depend on the relationship between the agent and the victim in the helping situation; women

R. IBUSUKI & T. NAITO

I

No Acquaintance

Trust

MORAL EVALUATION RATING

Help

Wait and See

Pass By

ACT

FIG.1. Interaction between familiarity o l agent and victim and the act

held to particularism in helping. Also, the significant interactions of kinship and familiarity between the agent and the victim suggested the complex nature of effects of such relations on moral evaluations, for example, when the victim was the uncle of the agent but the two were unfamiliar with each other, kinship affected the moral evaluations by a subject more than when the agent and the victim were familiar with each other. Whether relation-based norms are applied to everyone equally by the Japanese adolescents or not is another issue. The nonsignificant main effects of subject-agent relationships supported the notion that subjects apply these norms independently of the relation to the agent. However, the significant interaction between the sex of the subjects and subject-agent kinship called for closer analysis. Scheffi's test for multiple comparisons indicated a signifi-

1125

INTERPERSONAL RELATIONSHIP AND MORALITY TABLE 2 MEANSCORESAND STANDARD DEVIATIONSFOREACHLEVELOF MAIN FACTORS FORMORAL EVALUATION AND EXPECTATION Factors

Level

A. Sex of Subject

Men Women

B. Actor-Subject (Kinship)

Brother No Kinship

C. Actor-Subject (Acquaintance)

No Acquaintance Trust

D. Actor-Victim (Kinship)

No Kinship Uncle Rther

E. Actor-Victim (Acquaintance)

No Acquaintance Trust

F. Acts

G o To Help Wait and Approach Pass By

Moral Evaluation M SD 2.69 2.91

1.54 1.48

Expectation M SD 3.83 3.71

1.17 1.02

cant difference between the two levels of subject-agent kinships for the women (Fig. 2). This result implies that Japanese women may not apply their particularistic helping norms equally to everyone. STUDY2 The results of the first study suggested that female students tended to apply helping norms differently according to the relationships between agents and themselves. This result did not necessarily mean that they rejected the principle of universality willingly. For the purpose of examining this point, some structured interviews were conducted with the female students in a second or follow-up study. Method Subjects.-Subjects were 30 female students in the Department of Psychology at Ochanomizu University, located in Tokyo. They were enrolled in first- to third-year courses, and their mean age was 20.4 yr. They had shown high academic achievement in high school to attain entrance to the university. Procedures.-Interviews with subjects were conducted by the first author. The interviews were recorded by an audiorecorder and the written protocols were subjected to analysis. Interviews required two steps. I n the first step, subjects were instructed to read a short story about the four possible actions by an agent and to evaluate them morally on the five-point scale used in the first study. The story

R. IBUSUKI

& T. NAITO

Men

MORAL EVALUATION RATING

3

No h n s h i p

Brother

RELATIONSHIP FIG. 2. Interaction between sex of subject and kinship of subject and agent

was the same as that used in the first study. The questions were selected from those of the first study as follows. For the agent-victim relationships, a stranger was selected for the agent. For the subject-agent relationships, all four of the relations used in the first study were adopted. For the action, passing by was selected. I n the second step, the interviewer asked subjects to give the reasons for their evaluations and to provide anything they considered during the process of moral reasoning. Queries in a typical interview follow: (a) You have evaluated the actions of the four agents. Your ratings were the same (or different) for the agents. What things d ~ you d consider in the process of your evaluation? (b) Did you apply some criterion or principle of evaluation when you rated the four agents? (c) How do you evaluate the same action of another person? For example, in the case of your father?

INTERPERSONAL RELATIONSHIP AND MORALITY

1127

(d) D o you usually use the kind of evaluation just mentioned? What do you think about your method of evaluation? (e) Someone has asserted that people must rate equally in this case (or someone has insisted that people must rate differently in this case). What d o you think about the opinion?

Data coding.-The authors examined subjects' ratings and reasoning in terms of the following points: (1) whether their ratings for the four agents presented and for their father were equal or not and (2) the type of reasoning involved in their equal or unequal ratings. A seven-category coding scheme was constructed to code subjects' ratings and protocols. Its categories are as follows: (l-a) The first category, unequal ratings, indicated subjects perceived differences or unequalness and included positive statements or justification for those evaluations (e.g., "An action of a man cannot be evaluated without considering the relationship.") or consisted of the same evaluations but included references to the possibility of unequal ratings when rating on more than a five-point scale (e.g., "I judged differently according to kinship, but I could not express the difference on the >-point scale. It was too rough."). (l-b) The second category, unequal ratings without intention, consisted of different ratings but did not include references to either the affirmation or the negation of the differentiation (e.g., "I know that some make unequal judgments intentionally, but I did not have an intention to d o that."). (l-c) The third category, unequal ratings contrary to own expectation, indicated differences among ratings but such ratings expressed the subject's objection or lack of satisfaction (e.g., "I want to evaluate as objectively (equally) as possible usually, but I was not able to do so. I feel strange."). (l-d) The fourth category, ratings contrary to expectation, were equal but included references to an unwillingness to give such evaluations. (l-e) The fifth category, ratings without intention, were equal but did not include references to either the affirmation or the negation of the equal ratings (e.g., "On this problem, my ratings happened to be 'the same for different agents, but this is not a result of my objective thinking."). (l-f) The sixth category, equal ratings with willingness, consisted of equivalent values made with references to the affirmation of an identical evaluation of the same action (e.g., "I think the same actions should be evaluated equally. The evaluation should be objective."). (l-g) The seventh category, different or equal ratings according to the situation, consisted of either different or equal ratings and included references to alternative or similar evaluations on a case by case basis (e.g., "I think that the problem is one of two kinds, that is, the problem of -fairly evaluating and the problem of answering according to feelings."). While the first coding step focused on the ratings and their reasoning, the second coding step focused directly on the attitude toward the universal-

1128

R. IBUSUKI & T. NAITO

ity principle or justice in general. I n this second step, responses were simply coded into two categories by whether or not a subject acknowledged the principle of universality. That is, a two-category coding scheme was constructed to code subjects' protocols. (2-a) The first category, acknowledgement of the universality principle, contained the protocols which included acknowledgement of the principle (e.g., "Usually I evaluate as equally as possible," "I wish to have evaluated equally," "If I am requested to evaluate objectively, I will evaluate equally," "Whether I judge the act according to relation or universally depends on the case .") . (2-b) The second category, nonacknowledgement of the universality principle, contained all protocols referring to negation of or ignorance of the principle. Interjudge reliabilities of the two coding schemes were obtained by comparing the independent scoring by the two authors for 30 protocols. Agreement was 97%.

Results and Discussion The analysis of data for the first and the second codings are shown in Table 3 . The table showed that 40% of the women applied the helping norm according to their relation to the agent and also said that the norm should be applied according to such relationship. Only 10% of the subjects applied the TABLE 3 NUMBERAND PERCENT OF SUBJECTS CLASSIFIEDINTO EACHMORAL TYPE( N = 30) Ratings

Type of Rating

Acknowledgment of

n

a

Universality Principle, n

12 4

40.0 16.7 13.3

3 2 4

Equal Ratings Contrary to Expectation Without Intention With Willingness

0 1 3

0.0 3.3 10.0

0 0 3

According to the Situation

5

16.7

5

Unequal Ratings With Willingness Without Intention Contrary to Expectation

5

norm equally to the four agents and also expressed the notion that the moral norm should be applied equally or in a just way. However, the responses of the remainder were not clear-cut concerning application of their helping norm. They did not show concern about either equal application or relation-based application. One of them answered as follows:

INTERPERSONAL RELATIONSHIP AND MORALITY

1129

I did not concern myself with whether the ratings were equal or not. I judged each case independently. My ratings are equal, but I did not intend to make them so. The equal ratings just occurred.

Of the subjects interviewed 56.7% admitted that there is a principle of universality, while only 10% of subjects expressed a willingness to apply that principle. These results suggest that about half of the subjects who apply moral norms according to their relation to the agents may still acknowledge the existence of the principle. Why did the subjects not apply the principle of universality, although they knew it? The first explanation is that there is cognitive difficulty in applying the principle to a concrete situation. The interpretation is not adequate because some subjects expressed some doubt about the universal application. One subject answered as follows: I understand, but as a human being, I feel it's natural to make unequal ratings.

This answer suggests that there are subjects who acknowledge the principle of universality or justice but d o not apply it intentionally. That is, in moral reasoning, there are at least two components: one is acknowledging the principle and the other is applying the principle. We may argue that subjects make their moral ratings by concatenating two components. The results should be interpreted as confirmation of various types of morality dependent upon the effects of interpersonal relationships on the contents and applications of helping norms among Japanese adolescents. Analyses of what happened in each instance of moral reasoning, just as in the present study, should elucidate the internal organization of judgments. REFERENCES FRANKENA, W. K. (1973) Ethics. (2nd ed.) Englewood Cliffs, NJ: Prentice-Hall. GILLIGAN,C. (1982) In a different voice. Cambridge, MA: Harvard Univer. Press. HARE,R. M. (1952) The language of morals. London: Clarendon. KOHLBERG, L. (1971) From is to ought. In T. Mischel (Ed.), Cognitive deuelopmenf and epistemology. New York: Academic Press. Pp. 151-235. NAKANE, C. (1970) Japanese society. London: Penguin Books. PARSONS, T., & SHILS, E. A. (1954) Toward a general theory of action. Cambridge, MA: Harvard Univer. Press. Pp. 151-235. RUSCHAUER, E. 0. (1977) The Japanese. Tokyo: Tuttle. SHWEDER, R. S., MAHAPATRA, M., & MILLER,J. G . (1987) Culture and morality. In J. Kagan & S. Lamb (Eds.), The emergence of morality in young children. Chicago, IL: Univer of Chicago Press. Pp. 1-83. Accepted May 30, 1991.