Abstract
What makes some acts immoral? Although Western theories of morality often define harmful behaviors as centrally immoral, whether this is applicable to other cultures is still under debate. In particular, Confucianism emphasizes civility as fundamental to moral excellence. We describe three studies examining how the word immoral is used by Chinese and Westerners. Layperson-generated examples were used to examine cultural differences in which behaviors are called “immoral” (Study 1, n = 609; Study 2, n = 480), and whether “immoral” behaviors were best characterized as particularly harmful versus uncivilized (Study 3, N = 443). Results suggest that Chinese were more likely to use the word immoral for behaviors that were uncivilized, rather than exceptionally harmful, whereas Westerners were more likely to link immorality tightly to harm. More research into lay concepts of morality is needed to inform theories of moral cognition and improve understanding of human conceptualizations of social norms.
When do we judge an act to be immoral? “Moral norms” are proposed to be universal and unique: invoking a special type of norm cognition (Sachdeva, Singh, & Medin, 2011; Skitka, 2010; Turiel, 1983) with distinct social and evolutionary functions (DeScioli & Kurzban, 2013; Haidt & Kesebir, 2010; Krebs, 2008). The definition of moral cognition and function, however, depends on another definitional issue: the content of moral norms—what kinds of behaviors are immoral? The classic moral-conventional distinction—that issues of harm, fairness, and justice distinguish moral from non-moral issues (Turiel, 1983)—has been challenged by the more pluralistic Moral Foundation, Three Ethics, and Relationship Regulation Theory approaches, which argue that different cultural groups may hold a variety of moral concerns (e.g., Graham et al., 2011; Haidt & Graham, 2009; Rai & Fiske, 2011; Shweder, Much, Mahapatra, & Park, 1997). And yet, the search for unified and universal moral principles continues, such as with the recent Moral Dyad proposal that the perception of harm and suffering may be a universal template for all moral judgments (Gray, Schein, & Ward, 2014; Gray, Young, & Waytz, 2012; Janoff-Bulman & Carnes, 2013).
To understand what characterizes morality, especially cross-culturally, it is important to study what humans judge to be morally relevant (Monin, Pizarro, & Beer, 2007; Skitka, Bauman, & Sargis, 2005). In this article, we examine the lay concept of “immoral behavior” in Chinese and Western cultures. In contrast to the West, Chinese morality places unusual emphasis on being civilized and cultured. We ask two questions: First, what behaviors do Chinese and Western laypeople call “immoral”? And second, are the acts they call “immoral” characterized by their harmfulness, or by their incivility? We will argue that the data support a pluralistic approach to morality, that is, the principles that characterize moral issues do differ culturally. But the Chinese lay concept of immoral acts leads to surprising inconsistencies with traditional conceptualizations of morality. This better understanding of the category of “moral violations” in Chinese leads to questions about the psychological definition of morality in general.
Chinese Morality: Being a Cultured Person
China’s Confucian-influenced moral tradition is an important counterpart to the moral traditions of the West. Much post-Enlightenment Western moral philosophy as well as recent moral psychology has focused on “deontological” ethics, in which “moral obligations” are understood to be inescapable, fact-like requirements for behavior that can be generalized to other situations.
In contrast, Confucianism is a form of “virtue ethics” (Angle & Slote, 2013), in which the cultivation of good character is seen as the key source of moral behavior (Hursthouse, 2013; Wong, 2013). In particular, it is through the practice of courteous behavior toward others that one can develop and refine one’s character (Cua, 2007; Rosemont & Ames, 2008; Wong, 2014). One of the central virtues of Confucian ethics, “ritual propriety” (li/礼), is the ability to act with decorum and respect in a variety of situations and social roles (Sarkissian, 2014), describing everyday courtesy, etiquette, and good manners (Chinese-English Dictionary, 1997), and it is commonly perceived as central to moral education (Bakken, 2000). The specific decorous behavior one should observe (such as bowing vs. shaking hands; Fingarette, 1972) changes depending on context; however, this variable behavior is, ideally, consistently appropriate and attentive, motivated by virtuous character (Sarkissian, 2014; Wong, 2014). This philosophical position is echoed in research on East Asian self-concept and personality expression, in which Chinese and Japanese participants have been found to be less consistent across social situations compared with North American participants (Heine & Buchtel, 2009; see also Hwang, 2012; Yang, 2009).
Confucianism can thus be seen as a form of interpersonal duty-based, rather than rights-based, morality (Chiu, Dweck, Tong, & Fu, 1997; Dworkin, 1978; Miller & Bersoff, 1994; Miller, Bersoff, & Harwood, 1990) with an important caveat: Rather than focusing on absolute and abstract obligations, it elaborates on the importance of a cultured, cultivated character, sublimely attentive to the social requirements of the situation, as the guide of virtuous and appropriate behavior. A corollary to this moral focus on character cultivation is that Chinese law has historically separated morality from law, seeing criminal law (and not morality) as a motivational tool applicable only to deviants whose behavior falls well beyond the bounds of civility and who are unmotivated by the desire for virtue (Bakken, 2000; A. H.-Y. Chen, 2011). Virtuous character is thus shown in small daily actions rather than avoidance of serious crimes. Uncivilized behavior, such as not showing good manners or following appropriate etiquette, may show a person to be lacking in Confucian-style virtue—but does not well-describe a thief or murderer.
Chinese moral education has been effectively synonymous with Confucian moral education for more than 2000 years (Li, Zhong, Lin, & Zhang, 2004), and even now Confucian values influence the mandatory moral education courses in both primary and secondary schools in Mainland China (Reed, 1995; Zhan & Ning, 2004), both in inculcating specific virtues such as filial piety and also in a general emphasis on cultivation of moral character and appropriateness. For example, contemporary policies on moral education in Mainland China emphasize cultivation and civility: “The fundamental task of . . . moral education work is to cultivate students as citizens who . . . have public morality and civilized behavior and customs” (Article 5, Ministry of Education of the People’s Republic of China, 2006, pp. 37-38).
Civility Versus Harm as Morally Relevant
Influenced by a culturally based moral emphasis on character cultivation, Chinese lay moral cognition may be closely tied to judgments of civility and politeness. In particular, issues of harm, rights, and freedom infringement may not be closely associated with lay Chinese “moral” considerations. This would be in contrast to both classic and recent psychology theories of morality, which argue that moral judgment is universally based on perceptions of harm and suffering, justice, and fairness (Gray et al., 2014; Janoff-Bulman & Carnes, 2013; Royzman, Leeman, & Baron, 2009; Turiel, 1983). For example, Gray et al. (2014) have recently argued that accounts of moral instincts apparently founded on other principles (purity, loyalty, etc.) can be ultimately explained as concerns about harm.
Culture-minded psychologists have argued that these theories reflect a Western cultural bias; the “thinner and less binding morality” of modern Gesellshaft cultures (Haidt & Kesebir, 2010, p. 800), where communal and hierarchical relationships are less emphasized (Rai & Fiske, 2011), may have led to a moral emphasis on harm and unjust infringement of others’ basic individual rights. In non-Western cultures, other issues might be perceived as more relevant to moral judgment (Graham et al., 2011; Haidt & Joseph, 2007; Rai & Fiske, 2011; Shweder et al., 1997). Might immorality and harm be most strongly linked in the minds of those influenced by Western cultures, whereas those influenced by Chinese cultures have stronger non-harm-based conceptions of morality?
Moral Versus Non-Moral Norms: A Prototype Approach
The moralization of a behavior is assumed to be the first step in evoking the specific mental processes of moral cognition (Monin et al., 2007; Rozin, 1999). Moral norms may be distinguishable from norms about “aesthetic, religious, legal, or economic judgments” (Sinnott-Armstrong & Wheatley, 2013, p. 2), from conventions or issues of personal choice (Turiel, 1983), or behavior prohibited because it is imprudent or irrational (Machery & Mallon, 2010). Psychologists concerned with the identification of moral norms have used different theoretical models to guide their search. Deontologically oriented researchers have searched for acts perceived as “objectively obligated” (DeScioli & Kurzban, 2013; Miller et al., 1990; Shweder, 1990; Shweder, Mahapatra, & Miller, 1987) or those additionally justified by “concepts of welfare, justice, and rights” (Turiel, Killen, & Helwig, 1987, pp. 169-170). A social-functional perspective looks for evolutionarily based social and psychological mechanisms that produce prosocial outcomes (Haidt & Kesebir, 2010), especially those associated with automatic affective evaluations (Graham et al., 2013). Others have emphasized the communicative and meaning-making aspects of morality, both intra- and inter-individual (e.g., Haste, 2013; Kesebir & Pyszczynski, 2011; Park, Edmondson, & Hale-Smith, 2013; van den Bos, 2009).
How do laypeople distinguish between “moral” and “non-moral” norm violations? Laypeople may note a behavior’s similarity to prototypical immoral acts, rather than to an abstract definition (Rosch, 1975). The prototype approach proposes that lay, natural language concepts are defined by a mental network of prototypical and less-prototypical examples, rather than being defined by abstract, clear-cut criteria (Rosch, 1975). No published studies have yet reported lay prototypes of immoral acts, though prototypes of other concepts have been shown to affect cognition and reveal dimensions not predicted by theory (e.g., Fehr, 1988; Fontaine, Poortinga, Setiadi, & Markam, 2002; Kearns & Fincham, 2004; Smith, Smith, & Christopher, 2007; Walker & Pitts, 1998).
Given the ongoing debate about how to theoretically define moral norms, and especially the potential inapplicability of deontological definitions to Chinese virtue ethics, a prototype approach can provide new data on what Chinese and Western people judge to be a “moral issue.” In this case, we are interested in what defines the natural language category “immoral.” The current studies thus use Western and Chinese layperson-generated examples of “immoral” behaviors to examine what is most typical of this lay concept, and what characterizes immorality, in the minds of these participants.
The Current Studies
Three studies test the hypotheses that (a) compared with Chinese, Western laypeople are more likely to identify harmful behaviors as typically immoral, whereas Chinese are more likely to identify uncivilized behaviors as typically immoral, and (b) a higher degree of harmfulness distinguishes Western immoral behaviors from other kinds of unacceptable behaviors, but Chinese immoral behaviors are distinguished from non-moral wrongs by a higher degree of incivility. To test these hypotheses, in Study 1, we asked Chinese and Western participants to list examples of immoral behavior, and in Study 2, the most typical behaviors were presented to new participants who labeled them as either “immoral,” “wrong, but not immoral,” or “not wrong at all.” Cultural differences in how often behaviors were called “immoral” are described. Finally, in Study 3, we asked participants to judge the harmfulness and incivility of the selected behaviors, and compared behaviors labeled “immoral” with those that were “wrong, but not immoral” to determine whether harmfulness or incivility better characterized the “immoral” category.
Asking participants for examples of what they label immoral, rather than asking them to respond to a theory-based definition of immorality, allows us to cast a wide net for moral concerns and may have the best chance of capturing unpredicted lay-cultural concepts. However, this means that the choice of equivalent words in Chinese and English is the definition of comparability. Translation of the key word immoral (“没道德/mei dao de” or “不道德/bu dao de”) was thus given special attention, translated first by the bilingual authors, and verified by reference to multiple dictionaries (e.g., English-Chinese Dictionary, 1997; Oxford Chinese Dictionary, 2010) and official usage (e.g., Ministry of Education of the People’s Republic of China, 2006, and Ministry of Education Republic of China [Taiwan], 2012, using the word “moral” [dao de] in the context of “moral education”). In addition, we carried out a survey of naive bilingual informants. Forty-seven bilingual Chinese undergraduate students (26 Mainland Chinese, 21 Hong Kong Chinese; 85% female) attending a university psychology course in Hong Kong were surveyed in two formats, open-ended (16 participants; “What is the Chinese translation for the English word ‘immoral’?”) and closed-ended (31 participants; “What is a better Chinese translation for the English word ‘immoral’: ‘不道德’ or ‘邪恶的’?” with the second option being another common dictionary translation that can be also glossed as “evil”). In the open-ended format, 15 of the 16 participants wrote “不道德” (bu dao de) as the translation of “immoral” (the one exception wrote an idiosyncratic variation of “bu dao de”/不道德). In the closed-ended format, all of the 31 participants chose “bu dao de”/不道德. There were no differences between Mainland and Hong Kong Chinese.
Participants from a variety of English-speaking Western countries (Australia, Canada, and United States) and culturally distinct areas of China (Beijing, Hong Kong, and Shanghai) were recruited. Numbers were not pre-determined by power analysis; we attempted to reach around 100 to 200 participants per city/culture group, within practical limitations. In all studies, English was used for Western participants and Chinese for Hong Kong and Mainland Chinese participants. Translation of all materials was carried out by the first two authors (both English/Mandarin speakers) and trilingual (English, Cantonese, and Mandarin 1 ) research assistants, through back translation and discussion.
Study 1: Examples of Immoral Behaviors
Method
Participants
Participants were university students born or resident for >7 years in Canada (Vancouver; n = 83, 23% male, Mage = 22; 70 Caucasian, 6 East Asian, 7 Other), Australia (Melbourne; n = 88, 30% male, Mage = 25; 79 Caucasian [including 26 “Australian”], 9 Other), Hong Kong SAR (n = 160, 35% male, Mage = 22), or Mainland China (Shanghai, n = 182, 15% male, Mage = 21, all born in Mainland China; Beijing, n = 99, 54% male, Mage = 22). 2
Materials and procedures
Participants were asked to complete the sentence, “that person is really immoral. For example, he/she . . .”/“这个人真没道德,比如说他/她 . . .” by writing examples of behaviors. This brief question was placed as the last page of various unrelated questionnaires, the content of which were not predicted to affect answers.
Coding
After deleting 14 uninformative responses (e.g., “do/say something that is morally wrong”), 1,259 examples of immoral behaviors were coded into 46 categories developed through iterative discussions among the first author and trilingual research assistants. Examples were coded by two trilingual research assistants (α = .87), with disagreements resolved by discussion.
Results and Discussion
Table 1 shows the most frequently mentioned categories from each of the five cities (for all 46 categories, see Table S1 in the online supplemental materials). Although uncivil behaviors were frequently mentioned by Mainland Chinese, Western examples were more often of criminally harmful behavior. It is notable, for example, that “killing” was among the top 10 most frequently mentioned immoral behaviors by both Canadians and Australians, but mentioned by only 1 of the 281 Mainland Chinese participants; the converse applies to “spitting on the street,” mentioned by 31 Mainland Chinese and 0 Westerners (see Table S1).
Study 1: Top 10 Most Frequently Mentioned Behavior Categories (by Percentage of Examples per City).
To illustrate cultural similarities and differences, Table 2 shows correlations across category frequency between the five cities. It should be noted that Hong Kong response patterns were significantly correlated with Western response patterns, although they were writing in Chinese in response to the same Chinese survey question as Mainland Chinese. This likely reflects the influence of Hong Kong’s bicultural heritage, especially in the realm of moral education (Cheng, 2004), and shows that the examples generated were not solely controlled by the specific translation used for “immoral” (see Bond & King, 1985).
Study 1: Correlations of Category Frequency Between Cities Across 46 Categories.
Note. Numbers in brackets are 95% confidence intervals.
p < .01. **p < .001.
Study 2: Explicit Categorization: Which Behaviors Are Called “Immoral?”
When Study 1 participants listed the first immoral behaviors to come to mind, results suggested that Westerners use the word immoral to describe behaviors that are more harmful than Chinese-listed behaviors (especially Mainland Chinese-listed). In Study 2, we sought to replicate this finding with a stronger test to see whether Chinese and Western cultural differences would persist even when exceptionally harmful behaviors were brought to mind. In Study 2, all participants were presented with the same list of behaviors, representing each of the most frequently mentioned behavior categories per city (Study 1), and were asked whether or not these behaviors should be called “immoral.”
Method
Participants
Participants were university students who had been born or resident for more than 7 years in Canada (Vancouver, recruiting students who were born in Canada and whose first language was English; n = 123, 20% male, Mage = 21, all Caucasian), Australia (Melbourne, recruiting students of Western cultural background; n = 68, 32% male, Mage = 23, 63 Caucasian, 5 Other), Hong Kong SAR (recruiting Hong Kong permanent residents fluent in Chinese; n = 188, 3 41% male, Mage = 21, 181 ethnic Chinese, 1 Indian, 6 no ethnicity information), or Mainland China (Beijing, n = 101, 51% male, Mage = 22, all ethnic Chinese). Beijing participants were 87% non-religious and 8% Christian/Catholic, whereas Canadian, Australian, and Hong Kong participants were mostly non-religious (60%, 53%, and 67%, respectively) or Christian/Catholic (37%, 36%, and 29%, respectively).
Materials
Twenty-six 4 behaviors were selected from Study 1 to represent the top 10 categories from any city or those that included 4% or more of examples within a city (a common cut-off point for prototype studies; see Table 1). The list was presented in the same order for all participants.
Procedures
Participants filled out a survey online (Canada, Australia, and Hong Kong) or on paper in class (Beijing), for partial course credit (Canada, Beijing) or pay (Australia, AUS$10, about US$8; Hong Kong, HK$50, about US$6.50), in English (Canada, Australia) or Chinese (Hong Kong, Beijing). Participants read that the research concerned how people use the word immoral. They were presented with the list of 26 behaviors and asked to select which category was the best description of each behavior: “immoral”/“不道德的,” “wrong, but immoral is not the best word”/“错的/不好的,不过“不道德”一词并不贴切,” or “not wrong at all”/“根本没有错.” This question was placed as the first page of a longer questionnaire.
Results and Discussion
Australian and Canadian samples were combined to serve as a “Western” sample. 5 To test for cultural differences between Western, Hong Kong, and Beijing samples in the use of the word immoral, the cultural differences in the proportion of participants who chose “immoral” were computed and confidence intervals (CIs) were calculated (with Exploratory Software for Confidence Intervals [ESCI], Cummings, 2012, which uses the method recommended by Newcombe & Altman, 2000; see Table 3). In addition, χ2 tests of cultural differences in the proportion of participants who chose “immoral” versus either of the other two categories (“wrong, but immoral is not the best word” and “not wrong at all” combined) were assessed for each of the 26 behaviors with a Bonferroni corrected p value of .05 / 26 = .0019 for each culture pair (see Table 3 for effect size of the χ2 tests [phi/φ]; see online supplemental materials Table S2 for χ2 values and exact p values).
Studies 2 and 3: Categorization of Behaviors as “Immoral”; Percentages and Cultural Comparisons.
Note. Behaviors ordered from largest to smallest average Western–Mainland China differences (average phi of studies 2 and 3). CI = confidence interval. φ = phi (effect size, interpreted similarly to correlation coefficient r).
For “12 To infringe . . ,” N = 276 (see Note 3).
For “12 To infringe . . ,” N = 187 (see Note 3).
The most frequent category was “Wrong, but immoral isn’t the right word.”
The most frequent category was “Not wrong at all.”
p < .0019; exact value, 2-sided (see online supplementary Table S2 for χ2 values).
Overall, cultural differences were similar to Study 1: Western and Beijing participants showed the largest differences (average absolute-value % difference across the 26 behaviors = 26.0%; 15 behaviors were categorized significantly differently; average effect size phi = .27), whereas Hong Kong was in the middle, somewhat different from both Beijing (15.4% average absolute-value difference; 9 behaviors were significantly different; average phi = .16) and Western participants (14.3% average absolute-value difference; 8 behaviors were significantly different; average phi = .16), a moderate position that is consistent with its mixed Chinese/Western heritage (Cheng, 2004).
Again, the behaviors that were categorized most differently by Western and Beijing samples were characterized by extremes of harmfulness, as shown in Table 3. Behaviors such as spitting, cursing, and littering were more likely to be called immoral by Beijing than Western participants, whereas criminal behaviors such as killing, stealing, and hurting others were more likely to be called immoral by Western participants. For example, it is notable that although 70% of Beijing participants called to spit on the public street “immoral” (11% of Westerners), only 42% of Beijing participants called to kill a person “immoral” (81% of Westerners). 6
Study 3: Are Immoral Behaviors Particularly Harmful, or Particularly Uncivilized?
Studies 1 and 2 provide evidence of cultural differences in the behaviors that are most likely to be termed immoral and suggest that the greatest cultural contrasts occur on behaviors that are very harmful (termed immoral by Westerners) versus uncivilized (termed immoral by Mainland Chinese). Perhaps for Westerners, immoral behaviors are typically seen to cause serious harm, whereas for Chinese, a different prototype—such as incivility—may be used. However, alternative explanations exist; for example, perhaps Chinese are also considering harm, but they consider spitting or littering to be harmful to society as whole, and thus more serious than harm caused against individuals. Which is a more important element in Chinese judgments of immorality, incivility or harm?
In Study 3, we addressed this question explicitly by asking monocultural Chinese and American participants to rate both the harmfulness and incivility of behaviors, and then tested which (harm or incivility ratings) better differentiated behaviors called immoral from those that were called wrong for other reasons. We hypothesized that among Chinese, behaviors would be called immoral if they were judged as uncivilized, whereas harmfulness would distinguish between the categories less well. For Americans, however, the immoral versus wrong-but-not-immoral distinction should be driven more by perceptions of harmfulness than incivility.
Method
Participants
Paid online participants from across the United States (paid US$1) and Mainland China (paid 9 Chinese Yuan, about US$1) were recruited through MTurk.com and SoJump.com, respectively, to fill out an online survey, and a university student sample from Beijing completed an on-paper survey for partial course credit. U.S. participants (n = 128, 53% male, Mage = 33; 97 Caucasian, 7 African American, 15 Asian, 7 Latino/a, 1 American Indian/Alaskan Native) were 52% non-religious, 42% Protestant or Catholic, and 6% Other. Mainland China Internet participants (n = 165, 33% male, Mage = 30, all Chinese) were 74% non-religious, 21% Buddhist, 3% Protestant or Catholic, and 3% Other. Beijing university participants (n = 170, 30% male, Mage = 21, all Chinese) were 92% non-religious, 5% Buddhist, 1% Protestant or Catholic, and 2% Other. 7
Materials
Participants rated Study 2’s 26 behaviors 3 times. As in Study 2, they first selected which word best described each behavior: “immoral,” “wrong, but immoral is not the best word,” or “not wrong at all.” New to this study, they then rated each behavior on harmfulness (from 1 = extremely harmful [非常有害] to 6 = not harmful at all [完全没有害]) and incivility (from 1 = extremely uncivilized [非常不文明] to 6 = has nothing to do with being civilized or uncivilized [跟文明/不文明没关系] 8 ). In addition to translation/back translation, translations of the focal terms harmful and uncivilized were corroborated by dictionary definitions (e.g., Chinese-English Dictionary, 1997).
Results
Categorizations and comparison with Study 2
The two Chinese samples were combined, 9 and American/Chinese cultural differences in labeling behaviors as “immoral” or otherwise were tested as in Study 2 (see Table 3; see online Table S2 for additional comparisons). Results replicated Study 2 10 : Sixteen behaviors were categorized significantly differently by American and Chinese participants (average absolute-value % different = 27.1%; average phi = .25), and the largest cultural differences occurred for behaviors that seemed most and least harmful (e.g., spitting and killing).
Does harm, or incivility, best differentiate between immoral versus wrong behaviors? Within-person ratings
Calculations
Participants marked an average of 13.6 of the 26 behaviors as “immoral” and 11.1 behaviors as “wrong, but immoral is not the right word” (henceforth “wrong”). To compare whether incivility or harmfulness best differentiated between these behavior categories, harmfulness and incivility ratings were reverse-scored for each behavior and two kinds of difference scores were calculated: For each participant, the average harmfulness rating of behaviors that he or she had categorized as wrong was subtracted from the average harmfulness rating of behaviors he or she had categorized as immoral (harmfulness-difference), and the average incivility rating of wrong behaviors was subtracted from the average incivility rating of immoral behaviors (incivility-difference). Larger numbers indicate a larger difference between “immoral” and “wrong” categories on that criterion.
Comparisons of the two behavior types
Figure 1 shows that both Chinese and Americans rated “immoral” behaviors as both more harmful and more uncivil than “wrong” behaviors. However, as predicted, for Americans, the difference between immoral and wrong behaviors was greater on “harmfulness” than “incivility,” whereas for Chinese, the difference was greater on “incivility” than “harmfulness.” Specifically, a repeated-measures ANOVA of the difference between the immoral and wrong behaviors on harmfulness versus incivility ratings (i.e., within-subjects ratings of harmfulness-difference and incivility-difference) from participant culture (China vs. West) resulted in a significant interaction, F(1, 459) = 90.78, p < .001, η2 = .17, such that for Westerners, the difference between immoral versus wrong behaviors was larger in harmfulness ratings (M = 1.65, SD = 0.85) than in incivility ratings (M = 0.69, SD = 1.19); paired t-test Mdiff = −.95 [−1.16, −0.74], t(126) = −8.96, p < .001, d = .93, while for Chinese, the incivility-difference (M = 0.84, SD = 1.10) was larger than the harmfulness-difference (M = 0.36, SD = 1.22); paired t-test M diff = .48 [0.31, 0.64], t(333) = 5.73, p < .001, d = .41. In other words, compared with other wrong behaviors, Chinese considered immoral behaviors to be better characterized by incivility than harmfulness and vice versa for Westerners.

Study 3: Immoral versus wrong: Mean difference between behaviors called “immoral” versus those called “wrong, but immoral is not the right word” on harmfulness and incivility ratings.
Does harm, or incivility, correlate with naming a behavior immoral? Within-culture average ratings
The above findings can be further illuminated by a correlational comparison across the 26 behaviors of how the cultural groups, on average, rate the behaviors (see Table 4). For each behavior, we calculated the cultural group’s average incivility ratings (reverse-scored) and average harmfulness ratings (reverse-scored) and correlated these with the percentage of participants within that cultural group who categorized the behavior as immoral. Correlations of these values across the 26 behaviors (see Table 4) re-illustrate the above findings: The percentage of Americans who rated a behavior immoral was most highly correlated with the behaviors’ average harmfulness ratings (r = .92, vs. r = .75 for incivility ratings), whereas the percentage of Chinese who rated a behavior immoral was most highly correlated with their average incivility ratings (r = .83, vs. r = .27 for harmfulness ratings). It is particularly informative to note that Chinese and Americans had quite high agreement on which behaviors were harmful (r = .80), showing that the main cultural difference is in the definition of immorality, not harmfulness. In Table 5, which shows which behaviors were rated above-average on a given item, we can see that the low correlation between harm and immorality in the Chinese sample is because the Chinese immorality category includes both harmful and less harmful behaviors; and in particular, some exceptionally harmful behaviors (killing, hurting others) were not frequently categorized as immoral.
Study 3: Correlations of Within-Culture Average Ratings Across 26 Behaviors.
Note. Numbers in brackets are 95% confidence intervals (N = 26).
p < .01. **p < .001.
Study 3: Within-Culture Average Ratings of Behaviors. Sorted by USA Harmfulness Ratings.
Note. Above-average ratings shaded.
These findings support our hypothesis that the category of immoral behaviors is better characterized by incivility than harm for Chinese, while the opposite is true of Westerners.
General Discussion
In the above three studies, we found that Western and Chinese participants use the word immoral quite differently. In both open-ended responses (Study 1) and ratings of layperson-generated examples (Studies 2 and 3), Mainland Chinese were more likely to describe uncivilized behaviors as “immoral” compared with harmful behaviors, whereas Westerners did the opposite. The replication of these differences across methods (open-ended and closed-ended) and samples (across student and Internet-recruited samples, across different Western countries, and across Mainland China) suggests that this finding is robust. Finally, the division between “immoral” and “otherwise wrong” behaviors was determined more by the behavior’s harmfulness than its incivility for Americans, but for Mainland Chinese, the behavior’s incivility was a better guide than its harmfulness. Responses of bicultural Hong Kong Chinese used the same Chinese word to describe behaviors called immoral by both Western and Mainland Chinese participants. This suggests that culturally based differences in the conceptualization of morality, and not merely which language or Chinese word was used, affected participants’ responses.
These studies indicate that although Western participants think of immorality as primarily about serious harm, Chinese participants are more likely to focus on the uncivilized nature of behaviors when making moral judgments. Although the above data show that harmful behaviors may be considered “wrong” by both Chinese and Americans, Mainland Chinese most typically use “immoral” to refer to incivility. The diversity in behaviors mentioned by Beijing and Shanghai participants in Study 1, especially compared with the similarity of behaviors mentioned by Canadian and Australian participants, can now be more easily understood: As civility requires cultured attention to the social situation, what is uncivilized in different social contexts may be more different than what is exceedingly harmful.
This evidence suggests that “moral” judgments cannot be described as universally and particularly characterized by the perception of harm and suffering (as suggested by Gray et al., 2014) and is more congruent with an interpretation of moral judgments’ focal concerns being plural and shifting as guided by culture (Graham et al., 2013). It is consistent with a description of Chinese morality as primarily a form of virtue ethics in which immoral behaviors are those that show one’s character to be insufficiently polished; criminal behaviors, however, are too extreme to be relevant to virtue (Bakken, 2000; A. H.-Y. Chen, 2011; Rosemont & Ames, 2009).
Three Interpretations and Future Research
The Chinese and English concepts of “immoral” are obviously dissimilar, both in content (different specific behaviors described as “immoral”) and in underlying concerns leading to the “immoral” label (harmful vs. uncivilized). Are they even the same concept at all? Despite the above documented agreement between Chinese dictionaries, sociologists, laypeople, and government bureaus that “immoral” is correctly translated to “bu dao de,” why is this regarded as the correct translation? To answer this question, we may return to the “cognitive” and “functional” definitions of morality.
One interpretation is that the word immoral in Chinese, despite being different in content, still evokes the same cognitive consequences as the English word immoral. For example, it may be that among Chinese, moralized-uncivilized behaviors (in comparison with less moralized behaviors, such as extremely harmful ones) are more likely to be perceived as violating “sacred” values (Tetlock, Kristel, Elson, Green, & Lerner, 2000), be universally and/or categorically impermissible (Miller et al., 1990; Turiel, 1983), result in social ostracism (Wright, Cullum, & Schwab, 2008), and evoke emotional responses (Cannon, Schnall, & White, 2010; Haidt, 2007; Krebs, 2008). This would be consistent with Mainland China’s current and historical emphasis on the morality of personal cultivation and appropriate behavior (Cua, 2007; Ministry of Education of the People’s Republic of China, 2006) and with research showing that a sense of “moral conviction” predicts social ostracism of dissenters similarly in Mainland China and the United States (assessed by using the same translation of “moral” as in this article; Skitka et al., 2013). It may also illuminate current conflicts within China, such as why anti-Mainland Chinese sentiment in Hong Kong often focuses on the perceived “uncivilized behavior” of Mainland Chinese tourists (Luo, 2014, para. 23; see also T.-P. Chen, 2014; Yung, 2014). And yet, it seems strange if, for example, a murderer would be less socially ostracized than someone who spits on the street. More research is needed to assess the similarity in cognition around moral judgment in Chinese and English.
It is also difficult to fit the seeming overlap between “moral” and “conventional” norms in Chinese (and the peripheralization of some very harmful behaviors within the “immoral” category) into some definitions of the evolutionary and social function of moral norms, especially those that prioritize the prosocial function of morality (e.g., Haidt & Kesebir, 2010; Krebs, 2008). Even if killing is peripheralized as a moral concern, Chinese society does not permit random murders, as evidenced by the serious “social sanction” of the death penalty. Norms against criminal behavior may not be strongly relevant to moral norms; instead, the realm of law is seen as more appropriate for preventing and punishing crimes. If extremely harmful behavior is not prototypically immoral, but is still impermissible and sanctioned by another form of social regulation (law), Chinese lay morality may be more consistent with evolutionary theories focusing on morality’s intrapsychic or moral-circle functions (e.g., Kesebir & Pyszczynski, 2011; Sachdeva et al., 2011).
This leads us to a final implication: that Chinese may not categorize social norms in the same way as Westerners, and thus that a unified and universally identifiable “moral” category of norms may, in fact, not exist at all (Machery & Mallon, 2010; Sachdeva et al., 2011). Especially if Chinese prohibitions are divided into “criminal” and “immoral” rather than the Western “immoral” and “unconventional,” this may support the argument that more scientific traction on “norm cognition” could be gained by dropping the psychological term moral and studying multiple categories of injunctive norms characterized by different concerns and cognitions (Rai & Fiske, 2011; Sachdeva et al., 2011; Sinnott-Armstrong & Wheatley, 2013). If we wish to “carve nature at the joints,” it may be that the joints in social norms are different in China than they are in the United States, and if the joints are placed differently, so may be the cognitive ligaments and social-structural bones that support them.
The conceptual divide of social rules into “moral” (absolutely required) and “conventional” (permissible depending on social context) may be an accident of Western culture and philosophy, not universal. 11 Is there anything universal about the cognition that arises when we call something a “moral” issue? How many aspects of the “moral cognition” that has been documented in Western societies, where harm is a central reference point for moral judgments, can generalize to cultures where harm is not a central focus of moral judgments? More research is needed on the cognitive consequences of such lay concept differences, especially in multilingual contexts such as Hong Kong (Bond, 1983; S. X. Chen, Benet-Martínez, & Ng, 2014).
Conclusion
A Chinese lay concept of “immorality” that is more applicable to spitting on the street than killing people has implications for how we study and define moral norms. Research on moral cognition that uses harmful and unjust behavior as stimuli (such as the classic “trolley problem”) may be inappropriately influenced by Western lay instincts about morality, and theories about the unique evolutionary and social purposes of moral cognition may need to encompass moral systems in which behaviors acknowledged to be exceptionally destructive and legally sanctioned are relatively peripheral to morality. Finally, the potential of differently “jointed” categories of social norms in Chinese and English suggests that more research is needed to define how or whether “moral cognition” is universally distinctive from “norm cognition” in general (Machery & Mallon, 2010; Sinnott-Armstrong & Wheatley, 2013; Sripada & Stich, 2007).
Footnotes
Acknowledgements
Grateful acknowledgment for assistance with data collection, entry, and coding is given to Paul Wagland, Wajma Ebrahimi, and Colin Wastell of Macquarie University; Taylor Davis and Ara Norenzayan of the University of British Columbia; Claire Wai Ting Au and Candy Ho Yee Tang, Hong Kong Polytechnic University; Cindy Xiaocen Liu, Ludan Zhang, Mingwei Ma, and Yang Lu, East China Normal University; and Jane Yee Han Li, Qiyun Zhang, Huixin Ma, Ka Po Yu, Wen Lei, Xinyi Shen, Jingwen Xing, Jessica Ka Hing Wong, Mena Ning Wang, Zhou Lin, and Jiaxin Liu, Hong Kong Institute of Education.
Declaration of Conflict of Interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work described in this article was partially supported by Grant G-YX2U from the Hong Kong Polytechnic University, Grants RG55/2010-2011 and RG19/2011-2012R from the Hong Kong Institute of Education, and a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (HKIEd ECS 859813).
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
