A Cultural Psychology of Agency: Morality,Motivation,and Reciprocity

Culturally Variable Moralities of Community

A multistudy program of cross-cultural research that we undertook established the existence of qualitative cultural variation in moralities of community. Calling into question the universalistic assumptions that characterize both classic (e.g., Gilligan, 1982; Kohlberg, 1971; Turiel, 1983) and contemporary approaches to morality (e.g., Graham et al., 2011), our work documented the existence of a duty-based approach to community among Hindu Indians and a voluntaristic approach to community among European Americans.

Our studies established that Indians tend to treat helping family and friends as moral duties whereas Americans tend to treat them as matters for personal decision making. In vignette-based research, Miller and Luthar (1989) showed that Indians tended to categorize meeting the needs of family and friends as a moral obligation, whereas Americans tended to categorize meeting these needs as a matter of personal choice. These cultural differences are reflected in the following reactions to a real-life incident involving a son who refused to house his elderly parents. An Indian participant categorized the son’s behavior as a moral breach in violating a son’s duty to his parents: “It’s a son’s duty—birth duty—to take care of his parents. It’s not only money that matters. It’s being near your dear ones which counts more.” ¹ In contrast, an American participant categorized the son’s behavior as a matter of personal choice: “It’s up to the individual to decide. It’s duty to the parents versus one’s own independence and, I guess, one’s self interest. He’s fulfilling the minimal obligations to his parents. It wasn’t a life and death situation and their needs were being taken care of. Beyond that, it’s a personal choice” (p. 253).

In a vignette-based follow-up experiment, Miller, Bersoff, and Harwood (1990) demonstrated that the sense of interpersonal responsibility held by Indians is broader in scope than that held by Americans. Whereas Americans did not differ from Indians in the perceived desirability of helping and in the perceived moral duty to help in life-threatening situations, they more frequently categorized helping as a matter for personal decision making as compared with moral duty than did Indians. Fewer Americans considered helping a moral obligation as the need decreased and the role relationship became more remote, whereas Indians’ moral judgments were less contingent on level of need and role relationship.

In later research, Miller and Bersoff (1998) demonstrated that the form of interpersonal morality observed among Americans but not among Indians is contingent on liking. Both Americans and Indians treated the obligation of a parent to be responsive to their young child as unaffected by the parent’s liking for their child. However, in situations involving siblings and friends, Americans judged that there was a greater moral responsibility to help persons whom they liked than did not like, whereas liking had no impact on Indians’ moral judgments. Challenging claims that caring is purely other-oriented (Gilligan, 1982), this research provided evidence that the voluntaristic approach to interpersonal morality observed among Americans involves a self-serving element and that the presence of such a self-serving element is culturally variable.

Our research also demonstrated that interpersonal responsibilities to be responsive to the needs of family and friends are accorded greater priority than justice obligations by Indians than by Americans. When appraising situations in which meeting justice obligations conflicted with interpersonal responsibilities, Americans tended to give priority to the justice expectations and Indians to the interpersonal responsibilities (Miller & Bersoff, 1992). This work called into question assertions that interpersonal responsibilities are invariably subordinate to justice obligations (Kohlberg, Levine, & Hewer, 1983).

In sum, a major theoretical implication of our research on interpersonal morality is to highlight the existence of cultural variation in moralities of community. Our work not only documented a cultural main effect in the tendency for Americans to treat interpersonal morality in more discretionary terms than did Indians, but provided insight into ways that culturally variable outlooks held by Indians as compared with Americans are associated with differences in the weight given to such situational factors as need, role, liking, and competing justice obligations. As we have argued elsewhere (Miller, 1994), the implication of our program of research is not to point to the existence of only two culturally based approaches to interpersonal morality, with one linked to individualism and the other to collectivism. Rather, it is to establish that interpersonal morality is multiple, not unitary or universal in form. Congruent with this claim, studies by other investigators have provided insight into other types of cultural variation in moralities of community, such as in research highlighting moralities of community that extend to the social collective among Nordic populations (Vainio, 2015) or that are based on a normative sense of communal responsibility rooted in omoiyari among Japanese (Scarr, 1996; Shimizu, 2001).

In documenting cultural variation in moralities of community, our research also highlights conceptual and methodological limitations in moral foundations theory, a model presently dominating research on morality within social psychology (Haidt & Joseph, 2008). Theorists argue that moral foundations theory is sensitive to cultural variation even as it identifies universal evolutionary-based moral foundations that can be assessed through the Moral Foundations Questionnaire (MFQ). Notably, however, the insight that moralities of community are multiple is not taken into account in moral foundations theory. Not only is caring conceptualized theoretically as an aspect of harm rather than as a qualitatively distinct moral foundation, but it is assessed by a single MFQ scale item (i.e., “compassion for those who are suffering is the most crucial virtue”). Like other scale items on the MFQ, this caring item is phrased in such a decontextualized way that it fails to capture the type of variation in moralities of community documented in the cross-cultural research discussed here. In this regard, it may be noted that whereas marked differences in moral outlook on the MFQ have been identified between individuals with liberal versus conservative political views, few cross-cultural differences have been found (Graham et al., 2011). Although a perceived strength of moral foundations theory is its claim to have identified fundamental moral universals, it must be recognized that these universals identified on the MFQ have limited predictive power in tapping cross-cultural variation in moral outlooks in everyday situations and, in this respect, less explanatory force than commonly assumed.

Cultural Variation in Internalization of Social Expectations

The stance observed among Americans of treating interpersonal responsibilities to family and friends as dependent on individual preferences appears to reflect the types of motives considered autonomous within self-determination theory (SDT; Deci & Ryan, 1985, 1987). However, the motivational status of the emphasis on duty observed among Indians depends on how Indians experience such duties. As SDT theorists have noted (e.g., Chirkov, Ryan, Kim, & Kaplan, 2003), duties may be experienced psychologically as autonomous or controlled, depending on the degree to which they have been internalized.

To address this issue, we undertook research examining the motivational status of helping expectations among Indians and Americans. Congruent with claims about Western cultural outlooks embodying a tension between the group and the self (Markus & Kitayama, 1994; Marriott, 1976, 1990), our research showed that among Americans duty to be responsive to the needs of in-group members tends to be experienced in controlling terms that detract from individual inclinations, whereas it tends to be experienced as internalized among Indians. Our studies also support claims made within SDT about the importance of the experience of volition and choice for motivation universally (Ryan & Deci, 2017).

In early experimental research (Miller & Bersoff, 1994), we demonstrated that both Americans and Indians considered helping a neighbor as more socially expected in a condition in which the help was given after receiving aid from the neighbor on a prior occasion (reciprocity condition) as compared with when it was given in the absence of prior reciprocity (baseline condition). Providing evidence that social expectations are more fully internalized among Indians than among Americans, Indians associated greater satisfaction than did Americans with helping in the reciprocity condition, while not differing from Americans in satisfaction in the baseline condition. Also, Americans associated less satisfaction with helping in the reciprocity than baseline conditions whereas no condition differences in satisfaction occurred among Indians.

In research designed to test the ecological validity of such cross-cultural differences, we assessed outlooks on duty in the case of real life family behavior (Goyal, Wice, Adams, Chauhan, & Miller, 2015; Miller & Bersoff, 1995). Results indicated that Indians more frequently than Americans viewed sacrificing to meet the needs of a family member as a moral duty and associated greater satisfaction with such self-sacrifice. These cultural differences may be seen in contrasting appraisals made of a real-life case involving a wife who remained committed to her paralyzed husband who was no longer able to satisfy her needs. An Indian participant noted the satisfaction that she anticipated the wife would experience in being responsive to her husband’s welfare and fulfilling her duty as a wife: “She will have the satisfaction of having fulfilled her duty. She helped her husband during difficulty. If difficulties and happiness are both viewed as equal, only then will the family life be smooth” (Miller & Bersoff, 1995, p. 275). In contrast, reflecting a dualistic cultural outlook that involves a perceived conflict between duty and desire, an American participant focused on the dissatisfaction that she expected the wife would experience in not acting on her personal desires: “She is acting out of obligation—not other reasons like love. She has a sense of duty but little satisfaction for her own happiness” (Miller & Bersoff, 1995, p. 275).

To examine processes of internalization associated with duty, Miller, Das, and Chakravarthy (2011) conducted experimental research that asked participants to rate the extent to which their reasons for helping reflected a range of different motives. These motives included both the standard types of reasons assessed on SDT scale measures as well as duty. Utilizing a bipolar Likert-type scale, the investigation also assessed the degree to which participants experienced helping as controlling or as involving choice. We manipulated social expectations in Study 1 by varying whether real-life help given to friends was either strongly or only weakly socially expected, and in Study 2 by using vignettes that portrayed help being given to family as compared with to strangers. Providing evidence that duty is less fully internalized among Americans than Indians, Americans associated less satisfaction and less subjective sense of choice with cases in which helping was strongly as compared with weakly socially expected. In contrast, Indians associated equal satisfaction and subjective sense of choice with both situations. The investigation also showed that Americans experienced a controlling sense with acting out of duty, whereas Indians experienced a sense of autonomy with acting out of duty: Whereas Americans rated helping subjectively as controlling in the socially expected condition and as freely chosen in the condition involving lesser social expectations, Indians rated helping subjectively as involving a sense of choice in both conditions. Also, duty positively correlated with satisfaction and sense of choice only among Indians and not among Americans. Congruent with the claims of SDT, the investigation also provided evidence for the universal importance of a sense of autonomy, with a sense of choice associated with satisfaction in both cultural groups.

Evidence for these trends was also found in a recent three-study investigation (Goyal, Wice, Aladro, Kohlberg-Shroff, & Miller, 2017) that we conducted among children and parents from the United States and India. Manipulating the presence of social expectations using vignette-based measures, we demonstrated that whereas young American and Indian children do not differ in the satisfaction they associate with fulfilling parental expectations, older American children and their parents associate lesser satisfaction with fulfilling parental expectations than with acting more spontaneously. This research highlights culturally variable age trends, as children over development come to be increasingly influenced by the outlooks on role expectations emphasized in their culture.

Notably, the present evidence for duty to help family and friends as being more fully internalized among Indians than among Americans is congruent with claims made about contextual influences within SDT—that is, the assertion that internalization occurs more fully in contexts in which social norms are more integrated and less conflicting (e.g., Williams & Deci, 1996). Applied to understanding the impact of cultural environments on internalization, the more ambivalent stance held toward social expectations in U.S. than Indian cultural contexts (e.g., Markus & Kitayama, 1994; O’Flaherty & Derrett, 1978) appears to lead to social expectations being less fully internalized among Americans than among Indians.

In work closely related to the present claims, Bao and Lam (2008) point to the role of relatedness as facilitating internalization of social expectations and as explaining why Chinese children experience the absence of objective freedom of choice more positively than do American children (e.g., Iyengar & Lepper, 1999). Bao and Lam demonstrated in a series of experimental studies that although Chinese children with objective freedom of choice evidenced greater motivation to comply with decisions made by their mothers or teachers than did children without objective freedom of choice, the effect of objective choice was moderated by children’s socioemotional relatedness. When children had positive relatedness with the people who had made choices for them, their motivation was as internalized as if they had made their own choices. Bao and Lam maintain that compliance with social expectations is internalized among Chinese children through the children becoming highly motivated to please the individuals with whom they are closely attached. As Lepper, Corpus, and Iyengar (2005) likewise conclude in pointing to cultural variation in internalization of social expectations, “where children in the United States may see pressure from parents or teachers as externally imposed constraints, children in more interdependent contexts may see useful supports that serve the needs of the family and society” (p. 193).

Whereas cultural work on motivation is thus providing evidence that internalization of social expectations occurs more fully in collectivist cultures than in individualistic cultures, this work also highlights the need to update SDT scale measures to take into account the culturally variable meanings associated with obligation to family and friends and with wanting to please authority figures. Presently, when such items involving obligation or relational commitments appear on self-regulation scales, they are scored as either introjected or external motives (see http://www.psych.rochester.edu/SDT/questionnaires.php). For example, on the Academic Self-Regulation and Friendship Self-Regulation questionnaires, introjected or external scoring is applied both to the items tapping felt obligation (e.g., “Because that’s what I’m supposed to do”; “Because I think it is what friends are supposed to do”) and to the items tapping relational concerns (e.g., “Because I want the teacher to think I’m a good student”; “Because it would threaten our friendship if I were not trustworthy)”. Furthermore, in the scoring of SDT scales, a relative autonomy index (RAI) is computed by assigning positive weights to the intrinsic (+2) and identified (+1) scale items and negative weights to the external (−2) and introjected (−1) items.

The tendency for collectivist cultural populations to endorse duty and relational motives more frequently than individualist cultural populations then leads to such populations being scored as maintaining controlling motivational orientations on SDT scales. To give an example, in research involving Chinese adolescents (Pan, Gauvain, & Schwarz, 2013), results showed that endorsement of collectivist outlooks involving the child protecting and upholding honor for their parents (e.g., “Is the child supposed to do things that parents can take pride in?”) negatively predicted autonomous motivation on the RAI. However, the cross-cultural research reviewed here implies that this conclusion is questionable in that it fails to consider that SDT scale items tapping duty and relational commitments have internalized meanings among collectivist populations. In sum, the conclusion drawn by SDT theorists about the need to take into account the internalized meanings that may be associated with duty and relational commitments among collectivist populations (Chirkov et al., 2003; Ryan & Deci, 2017) has to date not been incorporated in the coding of items assessing duty and relational concerns on SDT scales. Cultural work on motivation not only identifies cultural variation in internalization of social expectations but also highlights the importance of more fully incorporating culturally variable meanings in the scoring of scale measures bearing on social expectations in research on motivation.

Cultural Variability in Norms of Reciprocity

Whereas the findings on motivation explain ways that collectivist populations may experience agency even when complying with social expectations, questions arise about the social experiences that support such motivation. Our results showed that Indians not only considered social responsibilities to family and friends as more obligatory than did Americans but associated greater satisfaction and sense of choice with such responsibilities than did Americans. Whereas this type of motivational outlook is explicable in terms of internalization (SDT), concerns might be raised that it represents a highly selfless stance that is incompatible with the self-interested aspects of human motivation. In particular, the present account of motivational orientations held among Indians and various other collectivist populations might be questioned as presenting an overly socialized view of the person in which individuals happily accept whatever is required of them, without expecting or receiving anything in return.

In part to explore this issue, we have undertaken a program of cross-cultural research examining norms of reciprocity in the context of family and friend relationships. Challenging earlier models which assumed that individuals are exclusively motivated by a rigid tit-for-tat form of reciprocity, theorists in behavioral economics and evolutionary theory argue that altruistic forms of reciprocity form a key basis for morality and human cooperation (e.g., Gintis, Henrich, Bowles, Boyd, & Fehr, 2008; Rand & Nowak, 2013). This suggests that norms of reciprocity may be implicated in supporting the present types of culturally variable moral and motivational outlooks.

Two forms of non-market-based reciprocity have been identified: communal norms and reciprocal exchange norms. As Clark and Aragón (2013) note, communal norms involve a form of reciprocity based on mutuality: “people do expect reciprocity, but it is reciprocity of a particular type—it is reciprocity of concerns for needs and responsiveness to needs and desires, if and when they arise” (p. 266; see also Clark & Mills, 2012). In turn, reciprocal exchange is a relaxed form of exchange in which the terms of the exchange have not been specified in advance and no requirement exists that returns be comparable to benefits received (Molm, 2010; Molm, Schaefer, & Collett, 2007). In the case of reciprocal exchange, uncertainty exists over reciprocation since it is unknown when, how, or even whether a return will be made. Acts of reciprocation reduce this uncertainty in demonstrating that the partner is committed to preserving the relationship, thereby enhancing trust and social solidarity. Whereas theorists have tended to associate expression of appreciation and gratitude with communal norms (e.g., Algoe, 2012; Algoe, Haidt, & Gable, 2008), exchange theorists maintain that appreciation and gratitude are also experienced in the context of reciprocal exchange, with individuals reciprocating in ways that convey appreciation—that is, that involve “acknowledging and conveying appreciation for benefits received” (Molm et al., 2007, p. 201).

In a three-study cross-cultural investigation, we demonstrated that Indians relied more frequently on communal norms in helping friends whereas Americans relied more frequently on reciprocal exchange (Miller et al., 2014). This was seen in Indians reciprocating by being responsive to the friend’s needs when and if such needs arose, even years into the future, and Americans reciprocating through expressive gestures given soon after receiving help, such as by sending flowers (see also Goyal & Miller, 2017). We provided direct evidence that Americans were motivated by exchange and not solely by gratitude or by politeness norms in a follow-up vignette-based experiment demonstrating that only Americans and not Indians were less inclined to help a friend who had helped them if they had already reciprocated by taking the friend out for dinner (and thus had at least partially repaid their debt) than if they had not reciprocated (and thus were still fully in debt). These findings are compatible with long-standing assertions linking collectivism to communal norms and individualism to exchange norms (e.g., Triandis, 1989).

In subsequent work, we provided direct evidence that reciprocity norms form a background for social support as well as addressed contradictory claims made about culture and social support. As noted, we anticipated that in relying on communal norms based on mutuality, Indians would feel not only an obligation to be responsive to the needs of family and friends but social assurance that their family and friends would, in turn, be responsive to their needs. This led us to predict that Indians would experience greater comfort in social support than did Americans, with the reciprocal exchange relied on among Americans associated with greater uncertainty about reciprocation and thus lesser social assurance. However, based on a series of studies that they conducted comparing East Asian and European American populations, Kim and her colleagues argue that the greater sense of social obligation to be responsive to other’s needs associated with collectivism leads Asians to develop relational concerns about not burdening others in their social network and thus to be more, rather than less, hesitant to ask for help than are Americans (Kim, Sherman, & Taylor, 2008; Taylor et al., 2004).

To address these contrasting theoretical claims, we examined the relationship between reciprocity and social support among Americans, Indians, and Japanese (Miller, Akiyama, & Kapadia, 2017). We sampled Japanese to represent the type of East Asian population assessed in the Kim et al., research and as a collectivist population who do not show the felt obligation to in-group members found among Indians (Oyserman, Coon, & Kemmelmeier, 2002). Our results revealed that, as we had predicted, Indians relied more frequently on communal norms and showed greater comfort in asking for help, judged appropriateness of getting help, and commonness of providing help to family and friends than did both Americans and Japanese, with reliance on exchange fully mediating the relationship between culture and social support in the comparison between Americans and Indians and partially mediating it in the comparison between Japanese and Indians. The research also demonstrated that, congruent with the findings of Kim and her colleagues (e.g., Kim et al., 2008), Japanese reported greater hesitation in asking for help, lesser judged appropriateness of getting help, and lesser commonness of providing help than did Americans as well as mentioned greater relational concerns, with these relational concerns partially mediating the relationship between culture and social support in the comparison between Japanese and Americans.

Open-ended responses observed in each cultural group illustrate the reliance placed on communal norms by Indians and the qualitatively distinct approaches to exchange found among Americans as compared with Japanese. In a sample response, an Indian participant conveys the sense of social assurance that his needs will be met that is associated with communal norms: “Between siblings, this kind of give and take is not required . . . my brother cares a lot about me but he is not expressive but I know that when the time comes and I require his support he will always be there” (Miller et al., 2017, p. 87). Reflecting his reliance on exchange norms that communicate appreciation and gratitude, an American participant described the positive feelings involved in reciprocating in ways that have personal meaning for the helper: “She should either write her a thank you letter and make it personal or buy her flowers or something small. It doesn’t have to be big, but just to let her know that she’s grateful. When someone helps you, you feel good to respond by doing something for them” (Miller et al., 2017, p. 89). In contrast, as seen in the following response of a Japanese participant, Japanese treated debt as normatively expected and underscored the need for reciprocations to be equivalent to the benefits received to preserve good relations and compensate for any trouble incurred by the helper: “Since she helped you by cancelling all her weekend plans and using the whole weekend for you, you’d need to give back for the time and energy in some way by giving her a little snack or taking her out for a meal. You kind of created a debt toward Kaori. You’d probably worry that it might eventually affect their future relationship” (Miller et al., 2017, pp. 88–89).

The present research provides insight into how agency is realized in each of the cultural communities sampled. In relying on communal norms, Indians not only feel a duty to promote the well-being of family and friends but can expect to benefit from these individuals in a noncontingent way should they be in need. In such a social system, a sense of moral obligation to family and friends is not experienced as oppressive in that it is associated with a strong sense of social assurance. Likewise, the tendency for Americans to communicate appreciation through symbolic gestures is consonant with their voluntaristic approach to interpersonal responsibilities. In such a social system, expressive gestures serve as signs that help is appreciated and that one is committed to the relationship and ready voluntarily to be responsive to the other’s needs—something that builds good feelings and strengthens relationships. In maintaining a normatively based approach to exchange in which reciprocation is expected, in Japanese social relations the individual can anticipate getting their needs met in ways that preserve good relations and that involve the individual acting with a sensitivity to contextual considerations.

In addition to uncovering cultural variability in the reliance placed on communal as compared with reciprocal exchange norms in family and friend relationships and documenting a relationship between reciprocal exchange and social support, our research has broader implications for social psychological theory. Our findings imply that researchers need to be sensitive to the multiple forms of exchange that exist and to avoid portraying exchange in the case of family, friend, and other non-market-based relationships in terms of a rigid tit-for-tat orientation that involves explicit expectations of return. Not only have forms of exchange involving explicit expectations of return been shown to involve external motivational orientations (e.g., Oarga, Stavrova, & Fetchenhauer, 2015), but they fail to capture the more internalized and personalized form of exchange that characterizes everyday social interaction. Our work also highlights the need to go beyond the individualism/collectivism dichotomy and give greater attention to the distinctive cultural outlooks held by different cultural populations.

Conclusions

In its focus on how agency is constructed in culturally variable ways that draw on moral outlooks, motivational orientations, and norms of reciprocity, the present review makes clear ways that cultural views on the relationship of the self to the group lend coherence to psychological theories of social development (Miller, 2004). The present considerations also underscore respects in which cultural work in psychology provides insight into the perspective of different cultural populations and to asking new theoretical questions and gaining new insight into existing psychological claims. The justification for cultural work is not only to gain insight into diversity in psychological functioning but to contribute to theory and knowledge in psychology more generally. An ongoing theoretical challenge, however, is to acknowledge the cultural grounding of many existing psychological claims and to work toward expanding psychological theories and their associated methodologies to be more culturally inclusive.

Regardless of whether cultural research makes use of quantitative or qualitative methods or cross-cultural or single cultural studies, such research needs to be held to the same rigorous theoretical and methodological standards of other approaches in psychology. It also must be recognized that whereas cultural research has the potential to identify cultural variation in basic psychological processes, there is no reason to expect that all psychological claims are culturally variable. Rather, the present vision is one of recognizing that culture, like biology, is integral to all theory building and research efforts in psychology. Cultural psychology does not represent a subfield of psychology as much as a perspective that must be taken into account in all theorizing and data collection in the discipline.