1252 Journal of Cross-Cultural Psychology 42(7) (Peng & Nisbett, 1999), in predicting cross-role and within-role self-concept consistency. Also, I test whether dialecticism moderates the relationship between self-consistency and both (subjec) tive well-being and self-concept certainty. Finally, I explore how felt authenticity, or the extent to which one believes his or her behavior is freely chosen (or self-determined Ryan & Deci, 2000), interacts with dialecticism, consistency, and well-being. The Importance of Consistency in Western Psychology Consistency means different things to different theorists, but generally it refers to inner consis- tency (i.e., coherence, congruence, freedom from conflict and/or ambivalence), cross-situational consistency (i.e., consistency across relationships, roles, contexts, etc.), and temporal stability for present purposes, I limit discussion to the first two meanings of the term. However defined, the need for consistent psychological experiences is the basis for many influential theories in social and personality psychology. In social psychology, awareness that one has acted inconsis- tently with one’s attitudes or other self-views gives rise to cognitive dissonance, an aversive state that can be reduced by either changing one’s behavior or one’s attitudes (Festinger, 1957). For example, the possibility of having made a bad choice between close alternatives, and thus acting inconsistently with a self-view as competent, leads to the spreading of alternatives (i.e., bolster- ing one’s opinion of the chosen alternative and disparaging the unchosen one Brehm, 1956). In attribution research, consistency plays a major part in the mental calculus people enact when determining whether a dispositional or situational attribution is warranted (Kelley, 1967). In personality psychology, the weight placed on the trait concept is a testament to the impor- tance of consistency, as traits provide ready explanations for consistencies in behavior. Knowl- edge of someone’s traits is assumed to afford predictability of that person’s behavior both across situations and over time (Allport, 1937). Self researchers also rely on notions of consistency: For example, Swann and colleagues argue persuasively in self-verification research that consistency provides a powerful sense of coherence, serving both epistemic (i.e., a sense of knowing oneself) and pragmatic (i.e., smooth social interactions) needs. They show that people will go to great lengths to have others see them as they see themselves, employing strategies as varied as choos- ing verifying interaction partners to selectively remembering confirmatory feedback (for a review, see Swann, Rentfrow, & Guinn, 2003). Others have written about the importance of consistency for well-being. James (1929) dis- cussed at length “sick” or “divided” souls, where there is “a certain discordancy or heterogeneity in the native temperament of the subject, an incompletely unified moral and intellectual constitu- tion” that, when extreme, “may make havoc of the subject’s life” (pp. 164-166). Lecky (1945) argued that self-integrity results when “we seek those experiences which support our values, and avoid, resist, or if necessary forcibly reject those which are inconsistent with them” (p. 99). Psy- chologists in the humanist tradition also discussed the dangers associated with incongruity among self-perceptions, and inconsistency between behavior and one’s self-concept, in terms of anxiety, defensiveness, and so on (Jourard, 1965 Maslow, 1954 Rogers, 1951). These theoreti- cal treatments are supported by a great deal of empirical evidence: Regarding inner consistency, the humanistic school was influential in the development of self-discrepancy theory, which doc- uments that anxiety and depression ensue from discrepancies between one’s actual self-views and one’s ought and ideal selves, respectively (Higgins, 1987). As for cross-situational consis- tency, although the self-concept is multifaceted and dynamic, in that people express different self- aspects in various roles, relationships, and situations (Chen, Boucher, & Tapias, 2006 Markus & Wurf, 1987 Sheldon et al., 1997), a great deal of differentiation among roles is associated with outcomes such as lower self-esteem, higher depression, anxiety, and psychosomatic prob- lems (Block, 1961 Donahue et al., 1993 Sheldon et al., 1997). at Bibliotheek fac Psych en on January 4, 2012 jcc.sagepub.com Downloaded from

Boucher 1253 Challenges to the Universality of the Consistency Motive Until fairly recently, researchers assumed that the need for consistency was universal and that both the expression of consistency and the deleterious effects of inconsistency would be mani- fested similarly everywhere. However, it is notable that all the research cited above was con- ducted in prototypically Western countries (usually involving participants of European descent) and there is now considerable evidence suggesting that consistency is not as prevalent in East Asia as it is in the West. For example, while European Canadians showed the spreading-of- alternatives effect in response to dissonance aroused from making a choice between two simi- larly appraised items, Japanese did not (Heine & Lehman, 1997). Even more compelling is evidence that looks directly at national or ethnic differences in self-consistency. Regarding internal consistency, Koreans, Chinese, and Japanese were more likely to indicate that contradictory traits (e.g., mature, immature) were self-descriptive, and Chinese and Chinese Americans were more likely to agree with items reflecting both positive and negative self-esteem than were European Americans (Boucher, Peng, Shi, & Wang, 2009 Choi & Choi, 2002 Kim, Peng, & Chiu, 2008 Spencer-Rodgers, Peng, Wang, & Hou, 2004 Spencer-Rodgers, Boucher, Mori et al., 2009). Inconsistency in self-views also extends to implicit self-concept and self-esteem (Boucher et al., 2009 Spencer-Rodgers, Boucher, Mori et al., 2009). Regarding cross-situational consistency, Koreans’ self-descriptions across roles were more inconsistent than those of Americans (Suh, 2002). In a six-nation study, Church et al. (2008) reported that Japanese expressed the most cross-role inconsistency, followed by Australians, who in turn were followed by American, Mexican, Filipino, and Malaysian participants. Also, the association between two relationship- specific self-views (e.g., self as friend vs. self as roommate) was less strong for East Asian Americans than for European Americans (English & Chen, 2007). Finally, inconsistency seems to be less troubling for East Asian groups. For example, while cross-role consistency was strongly subjective well-being and positive affect among Americans, this was less so for Koreans (Suh, 2002). Church et al. (2008) found a similar pattern of results, in that cross-role consistency was associated with the outcomes above in addition to greater self-esteem and less social anxiety for all groups except Japanese. Naïve Dialecticism and Self-Inconsistency Researchers, drawing inspiration from national/ethnic differences in self-consistency, turned to understanding what it is about nationality/ethnicity that causes these differences in the first place (for a discussion, see Peng, Ames, & Knowles, 2001). In this article, I explore naïve dialecticism (or dialectical thinking) as a potent mechanism responsible for self-consistency. Dialectical thinking is an example of the beliefs or theories tradition invoked to explain cross-national dif- ferences and is a belief system thought to characterize East Asian groups. It is based in religious (especially Taoist), philosophical, and epistemological traditions, institutions, and practices and is composed primarily of three concepts: the theory of holism, in which everything in the uni- verse is interconnected and interpenetrating the theory of change, in which the universe and everything in it are dynamic and ever in flux and the theory of contradiction, in which both sides of an apparent contradiction hold some piece of the truth. When faced with a contradiction, dia- lectical thinkers adopt a compromise approach, trying to retain elements of both sides (Peng & Nisbett, 1999). A great deal of evidence supports the proposition that these theories are operative to a greater extent among East Asian groups than prototypical Western ones (for a review, see Nisbett, Peng, Choi, & Norenzayan, 2001). Theoretically, it seems plausible that dialecticism would be related to self-inconsistency. A holistic outlook necessitates an awareness of one’s part in the surrounding context. This context at Bibliotheek fac Psych en on January 4, 2012 jcc.sagepub.com Downloaded from

1254 Journal of Cross-Cultural Psychology 42(7) may be restricted to the immediate situation (e.g., me with my professor) or it may be so broad as to encompass the whole universe and everything in it. Individuals from groups thought to be dialectical spontaneously refer to themselves located in both kinds of contexts for example, they are more likely to list self-statements that reflect immediate situational variation (e.g., “I am stressed out at school”) and one’s place in a system that is much larger than the individual self than are members of relatively nondialectical ones (e.g., “I am an earthling” Cousins, 1989 Spencer-Rodgers, Boucher, Mori et al., 2009). Because all things are interconnected, holistic beliefs give rise to an acknowledgment that as the context changes, so must the elements within that context, including the self. Thus, changing across contexts, roles, and relationships is both necessary and desirable in fact, this may be the sign of a self that is well-adjusted, mature, and able to feel attuned to and connected to important others (Kitayama & Markus, 1999 Morling, Kitayama, & Miyamoto, 2001). Consistent with this, Japanese self-descriptions changed more with shifts in the immediate context than did those of Americans (Kanagawa, Cross, & Markus, 2001). Also, Chinese were more likely to spontaneously describe themselves as changing, grow- ing, and evolving (Spencer-Rodgers, Boucher, Mori et al., 2009). Finally, because change is constant, the self in one context may be contradictory to the self in another, different one. More- over, even within a given role or context, it might be necessary to be flexible in one’s behavior, adapting different kinds of behavior as the situation requires. Instead of this kind of variability and flexibility being a sign of confusion, superficiality, or neurosis, however, it is simply the inevitable result of a complex reality of interconnection. Indeed, the self, according to this set of beliefs, is nuanced and balanced. There is now growing empirical evidence that dialecticism plays an important role in national/ ethnic differences in self-consistency. For example, cross-role inconsistency was more evident and less maladaptive in Japan (believed to be collectivistic and dialectical) than in countries thought to be collectivistic but nondialectical (e.g., Mexico, the Philippines Church et al., 2008). Other evidence comes from studies that use instruments designed to measure individual differences in dialectical thinking, such as the Dialectical Self Scale (DSS Spencer-Rodgers, Srivastava et al., 2009). Here, dialecticism mediates the relationship between nationality/ethnicity and both self- concept and self-esteem inconsistency (Boucher et al., 2009 English & Chen, 2007 Spencer- Rodgers et al., 2004 Spencer-Rodgers, Boucher, Mori et al., 2009). In addition, those scoring high on dialectical thinking were more likely to prefer verifying over nonverifying feedback about a context-specific self-view than those scoring low (Chen, English, & Peng, 2006). Finally, Chinese and Asian Americans were more likely to alter their self-views when presented with inconsistent self-relevant feedback, relative to European Americans, and this effect was accounted for by dialectical thinking (Spencer-Rodgers, Boucher, Peng, & Wang, 2009). The Present Research I had several goals in mind for the present study, which builds from previous research examining the effect of dialecticism on the self-concept. One was to examine whether dialectical thinking would be associated with less consistency in self-views across roles. While there is ample evi- dence that dialectical thinkers express more inconsistency within their global self-views (Choi & Choi, 2002 Spencer-Rodgers, Boucher, Mori et al., 2009), there is little evidence that this holds true across roles (English & Chen, 2007). To examine this, participants completed a measure of dialectical thinking and then described themselves in general (i.e., global self-concept) and in two roles (as a son/daughter and as a friend). From this, I calculated an index of cross-role con- sistency much like past researchers have done (e.g., Suh, 2002). Given English and Chen’s (2007) evidence that dialectical thinkers are more likely to elaborate upon and clearly demarcate the selves they are in specific close relationship contexts, and previous demonstrations of differences at Bibliotheek fac Psych en on January 4, 2012 jcc.sagepub.com Downloaded from