Beyond the Moral Domain: The Normative Sense Among the Chinese

In this paper we report a study on how different types of normatively relevant transgressions are evaluated by Chinese participants. We hypothesized that, given the continuing influences of Confucian worldview on contem­ porary Chinese societies, the Chinese will not make a distinction between moral (daode) and conventional norms of cultured behavior (wenming). Our results indicate that Chinese participants expressed a strong normative conviction not only towards harmful and unfair actions, usually subsumed under the moral domain in Western literature, but also towards violations of what would be widely accepted as conventional (or cultural) norms. Similarly, Chinese participants expressed a strong normative conviction towards violations of the traditional Chinese value of family reverence (xiao), thus further supporting our general thesis. Moreover, results indicate that, overall, explicit consi­ derations of wenming (unculturedness) emerged as the best predictor of a normative conviction response among the Chinese. Though considerations of harm and fairness also emerged as significant predictors of normative conviction response. The results are discussed in the light of recent debates about the moral/conventional distinction and the scope of morality.


Introduction
There are two related debates in contemporary moral psychology. One debate has been centered on the distinction between moral norms and nonmoral norms. Indeed, the attempt to define and delineate the moral domain, as opposed to non-moral conventional domains, can be traced back to much of mid20 th century moral philosophy (Stich, 2018; for review see Gert, 2002). Later, Turiel and his colleagues drew on this philosophical tradition and integrated the distinction into their psychological research (Nucci & Turiel, 1978;Smetana, 1993;Tisak & Turiel, 1984;Turiel, Killen, & Helwig, 1987;Turiel, 1983;Yau & Smetana, 2003). More specifically, according to this tradition, folk intuitions syste matically distinguish between moral transgressions (e.g., hitting other people, stealing or cheating) and conventional transgressions (e.g., eating with your bare hands or dressing inappropriately to school). Moral transgressions have a distinctive normative signatu re -their wrongness is authority-independent, universal, and is justified by appealing to considerations of harm and fairness. Conventional transgressions, on the other hand, don't have this signature. This distinction, it is argued, is a rather robust and universal psychological phenomenon that could be observed across different age groups and across different cultural or religious contexts (e.g., Nucci & Turiel, 1993;Yau & Smetana, 2003).
Another debate is about what constitutes the moral domain, assuming that the distinction between moral and conventional norms is indeed a psychological universal and not a Western cultural construction. In other words, what kinds of normative considerations (contents) are at the core of moral judgment? Do people across different cultural and demographic groups distinguish a particular set of norms as morally relevant norms? As it was mentioned above, Turiel and colleagues explicitly tied the moral domain to consi derations of harm and fairness. Yet others argued that it is possible to conceptually reduce all moral considerations into considerations of harm (Gray, Young, & Waytz, 2012;Gray, Schein, & Ward, 2014;Schein & Gray, 2015). Similarly, Baumard and colleagues argued that it is possible to conceptually reduce all moral considerations into considerations of fairness (Baumard, 2016;Baumard et al., 2013;Sousa & Piazza, 2014). These monistic positions are challenged by an alternative pluralistic position developed by Haidt and his colleagues, who claim that there are at least five morally relevant concerns. Besides harm and fairness, non-WEIRD people (i.e. Western, Educated, Industrialized, Rich, Democratic; see Henrich, Heine, & Norenzayan, 2010) worldwide also moralize respect for authority, ingroup loyalty, and purity violations (e.g., Haidt, 2012;Haidt & Graham, 2007;Graham et al., 2013). Importantly, none of these positions (be they monists or pluralists) deny the distinction between moral and nonmoral norms, all the disagreement is about the scope of the moral domain.
In some sense, the first debate is more fundamental. If it turns out that the moral/conven tional distinction is not a universal feature of human normativity across different cultures, then the question of how many moral domains there are is irrelevant. It is quite possible that this distinction emerged as an artifact of a rather limited class of transgressions used either by Turiel and colleagues or by Haidt and colleagues (see Clifford, Iyengar, Cabeza, & SinnottArmstrong, 2015;Graham et al., 2011) and does not capture all the relevant nonWestern issues (see also O'Neill & Machery, 2019). Arguably, the normative sphere could be parsed and classified in the plurality of ways (O'Neill, 2017). Anthropologists, for instance, showed that many nonWestern smallscale societies strongly condemn, and sometimes punish, various taboo violations, that are not necessarily comparable to Western style of harm and fairness (Boehm, 2008). Thus a more proper question would be to ask what normative considerations are indeed relevant to a particular cultural group? That is, which norm violations evoke a strong normative conviction that an action is "wrong"?
The current paper will address both debates by presenting new evidence from Chinese participants. Following Turiel's tradition, we will characterize the normative conviction as a metacognitive attitude, where wrongness of an action is conceived to be independent of any authority's opinion and universally so. But unlike this tradition, we will disconnect the normative conviction from considerations of harm and fairness (Berniūnas, Dranseika, & Sousa, 2016;Landy, 2016). Also, the scope of normative concerns will not be limited to the list provided by Haidt and colleagues (Clifford et al., 2015;Graham et al., 2011). Chinese participants will also be presented with violations of norms that are related to specific culturally relevant concerns, namely, civilized behavior (wenming) and traditional values xiao (filial piety). Finally, they will be asked to evaluate those norm violations.

The normative conviction
It is reasonable to expect that different cultural traditions and normative systems around the world have different ways of structuring the normative sphere and put emphasis on different normative concerns. At the same time, it is reasonable to expect that at least some normative concerns are shared across cultures. To make some progress in this re gard, though, an independent criterion is required to delineate culturally specific and/or universally recurrent normative contents. It is not sufficient to rely on our own (Western) intuitions about which concerns are moral concerns. Arguably, the contents of those intu itions might not be as representative of non-Western populations (e.g., Berniūnas, 2019;Dranseika et al., 2018). More cross-cultural studies are needed.
To address the point of the independent criterion, consider the notion of normative conviction. Intuitively, it seems that people respond to certain transgressions with a strong conviction that an action is clearly wrong, though what this strong conviction consists of is still a matter of debate and still remains an open methodological question (see Goodwin & Darley, 2008;Nichols & Folds-Bennett, 2003;Skitka, Bauman, & Sargis, 2005;Shweder et al., 1987;Turiel, 1983). For the purposes of current research, following Turiel's tradition and the moral/conventional task, we characterize the normative conviction specifying transgressions in terms of three dimensions: First, given transgressions are seen as (a) wrong. Then these wrong transgressions are seen as (b) authority independent (i.e., their wrongness is not cancelable by the decision of any authority) and (c) general in scope (i.e., their wrongness extends to different places and times) (Berniūnas et al., 2016;Sousa & Piazza, 2014). As a result, if a participant in the study responds with a strong norma tive conviction to a particular action or situation, then this response could be counted as expressing a strong normative conviction -that is, it is strongly normativized. Note, this framework does not presuppose any fixed distinction between moral and non-moral con ventional norms (as it is in Turiel tradition), nor any fixed list of normative concerns (as it is for monists and Haidt's pluralistic Moral Foundations Theory). Again, universality of the distinction and a list of concerns is a matter of crosscultural evidence.

Daode, wenming and xiao: Chinese normative thinking
As a contrast to the Western intellectual tradition that was shaped by Ancient Greek thought and later by Christianity, the Chinese intellectual tradition poses an interesting challenge to various conceptual distinctions in contemporary moral psychology (see Nisbett, 2003, for other challenges to universality claims).
It is a commonplace practice across contemporary Western moral philosophy literature that discusses the Chinese tradition to take the Chinese term daode 道德 as a technical translation of the English term moral. On the other hand, many authors also notice pro blems with equating morality with daode (see Kupperman, 2002;Rosemont, 1976). These critics note that the specifically English term moral reflects a specifically Western ethical tradition, and is onesidedly taken as a reference point in the subsequent crosscultural discussions (Shun, 2009). As a result, nonWestern cultures are forced to answer Western moral questions, even though such questions may not be easily expressed in nonWestern languages (Rosemont, 1988;Goldin, 2005). Unsurprisingly, characterization of universal moral domain (or moral cognition) followed the same Western philosophical heritage.
As it was indicated in the introduction, one of the fundamental features of Western moral philosophy and moral psychology is an agreement that the moral/conventional distinction is an important element of ordinary normative discourse. By contrast, the term daode could be understood as a unique Chinese way to categorize norms (Rosemont, 1976). In the Confucian tradition the dao and de were closely associated with refined, cultured, and civilized (wen 文) life and actions, which were subsumed within the notion of li 禮, the ri tualized proper order of human interactions (see the Chinese classical texts Liji: Quli I.8, Xunzi 1.12). An important thing to note here is that the meaning of li -and by extension of daode -in early Confucian texts not only partially covers what in English would be easily conceptualized as moral, but is also inseparable from all the other connotations, including aesthetical norms of decorum, cultural norms of civilized behavior, etc. (Rose mont, 1976). Thus, given the continuing cultural and social influences of the Confucian worldview on contemporary Chinese society, one would expect that conventional cultural norms and regulations, or civilized behavior (wenhua 文化 / wenming 文明 in contempo rary Chinese), would be included in the notion of daode. If that is the case, then it could be hypothesized that transgressions of wenming norms would evoke strong normative conviction and emotional response similar to one that is observed in Western respondents reacting to moral transgressions that involve harm or injustice.
Recent studies in comparative moral psychology provide some empirical evidence showing that indeed there is some discrepancy between the English moral and Chinese daode. In particular, Buchtel and colleagues (2015) asked Chinese participants to gene rate a list of actions that are bu daode (an opposition to daode), and Englishspeaking participants to generate a list of actions that are immoral. Interestingly, Chinese every day usage of bu daode (allegedly the Chinese equivalent of the term immoral) is more applicable to spitting on the street than killing people, a markedly different application from the English immoral. That is, Chinese were more likely to use the word bu daode for behaviors that were uncultured/uncivilized (or bu wenming 不文明 in contemporary Chinese), whereas Westerners were more likely to link immorality tightly to harm and fairness (ibid., p. 1382). Importantly, Dranseika et al. (2018) replicated Buchtel et al.'s (2015) results, but besides asking to provide a list of immoral/bu daode actions, they also asked Chinese and American participants to provide a list of uncultured/bu wenming actions. In Mandarin Chinese, lists of bu daode and bu wenming behaviors appeared to be very similar, while American lists of immoral and uncultured behavior were markedly different. This implies that "the Chinese tend to think about "immoral" and "uncivilized/ uncultured" as tightly interconnected concepts, whereas the Westerners tend to conceive of these two categories as rather different, and perhaps opposing, as in a moral/conventional distinction of the Turiel tradition in moral psychology" (2018, p. 76). Following this line of research, in the current study we will apply an independent criterion to scenarios that involve not only typical Western moral transgressions, but also to scenarios that involve typical Chinese daode/wenming transgressions.
For this purpose we chose one of the highest ranking values within the traditional Chinese value system, which is called xiao 孝, or the ability to act with one's parents in reverential ways. The term has been most commonly translated in English as "filial pie ty," but we are following Rosemont and Ames (2009) in their critique of such translation and their suggested translation as "family reverence." The Confucian classic Lunyu 論 語 depicts xiao as the basis of the ultimate humane conduct (ren 仁; Lunyu 1.2). Xiao is a quality present in the children's interactions and, even more importantly, their attitudes towards their parents (Lunyu 2.7,2.8). The importance of xiao grows as the children reach maturity and the parents grow old and weak, and the responsibility to uphold xiao does not stop even after the death of parents (Lunyu 1.11,2.5). Although xiao is prima rily based in and is learned to be exercised within the family (jia 家), its importance and relevance is extended within the Confucian worldview to the political realm of zheng 政 (Lunyu 2.21). Contemporary Western scholars, acknowledging a centrality of family values for the Confucian worldview, are paying increasingly more attention to the ways of how xiao functions both in traditional ethical political thought, and in the contempo rary Confucian-influenced societies (Rosenlee, 2014;Ivanhoe, 2007;Rosemont & Ames 2016;Nuyen, 2004;Jordan, 1998). Xiao has also caught the attention of psychologists (see, for instance, Hwang, 1999).
Thus, bringing all these strands together, we hypothesize that Chinese participants (a) will have a strong normative conviction towards situations that depict standard Turiel type or Haidttype of moral transgressions. Moreover, given the long-standing influence of the Confucian ethical tradition on historical and contemporary China, we hypothesize that participants (b) will also have a strong normative conviction towards situations that depict the overlapping lists of bu daode (usually translated as "moral") and bu wenming (uncultured) behavior, which corresponds to Turiel's violation of conventional norms. We also hypothesize that participants (c) will have a strong normative conviction towards violations of traditional values of xiao (filial piety). Finally, we predict that Chinese par ticipants will express not only five Haidt-type normative considerations, but also will be concerned whether an action is generally an instance of bu wenming (unculturedness).

Participants
With assistance from local colleagues, we recruited 566 participants from Liaoning uni versity, Shenyang, in the Northeast of China. Most of the participants were students from different study programs, they have been recruited before or after the classes in groups. There were 84% female participants, mean age = 20, range from 18 to 26.

Materials and procedure
Scenarios. For purposes of this study we selected scenarios that depict a transgression of only one type but not another. Following the Moral Foundations Theory, Clifford, Iyengar, Cabeza, and Sinnott-Armstrong (2015) provide a list of examples from each type of transgressions. Note, however, that in their study participants received an expli cit request to rate how morally wrong the behavior is on a 5-point scale, and then were asked "Why is the action morally wrong?". In this study we choose not to use an explicit term moral and its usual Chinese translation daode (see Buchtel et al., 2015;Dranseika et al., 2018;and Silius et al., 2017 for a more detailed reasoning of why such translation is problematic). For our purposes, we use an indirect measure to delineate the normative conviction by employing an independent criteria -a metacognitive task that evokes a tacit conviction that an action is wrong. Also, Clifford et al. compiled the list of the transgressions in each domain from US participants only, so a cross-cultural validity of listed examples is not warranted. For this reason, we've chosen particular instances in each type of transgression that were considered in Clifford et al.'s study as best instances and are culturally recognizable not only to American participants, but also to the Chinese. Additionally, we constructed three scenarios that are typical examples of Chinese daode/ wenming transgressions and three scenarios that explicitly address violation of traditional value of xiao (filial piety).
To cover a wider area of situations, there were three scenarios in each domain describing what an agent did in a specific situation. There were seven groups, where each separate group received three scenarios of the same type (Harm, N = 81; Fairness, N = 85; Loyalty, N = 77, Authority, N = 83; Purity, N = 80, Uncultured, N = 82; Tradition, N = 78). In total there were 21 scenarios (see Appendix).
Normative conviction. After each scenario, participants received the following ques tions, in fixed order. First, participants were asked whether they consider an action a transgression: "in your personal opinion, is it wrong that [agent] did [action] in this situ ation?" ("Yes/No"). Second, participants were presented with two normative conviction questions: (i) "suppose that the common view of your society and of people you respect is that the above described action of agent X is not wrong"; (ii) "suppose that agent X lived in a country where everyone thinks that agent X's action is not wrong". For each normative conviction question, participants had to indicate whether they agreed that the action X would still be wrong under such circumstances. An indication that the action X would still be wrong in both circumstances constitutes a strong normative conviction.
Note, this is a rather strict criteria that combines attitudes of authorityindependence and universality. Arguably, authorityindependence could be understood here as a form of folk objectivism (for a more detailed discussion of folk objectivism and other methods applied to investigate it see Beebe, Qiaoan, Wysocki, & Endara, 2015;Goodwin & Dar ley, 2008;Nichols & Folds-Bennett, 2003;Wright, Grandjean, & McWhite, 2013). More specifically, the fact that authorities' opinion doesn't change the wrongness of the action points to the obligation to comply with the norm as somehow being imposed "externally" (Stanford, 2018). As O'Neill and Machery noted, "every population 'externalizes' some of its norms" (2019, p. 43). Moreover, they argue that people would "tend to think that any suitably similar person in roughly the same situation would be subject to the same sort of obligation" (ibid.). As such, authorityindependence judgment could be construed as indicating a separate, albeit weaker, form of normative conviction, where universality judgment is not necessary. In other words, one can express a weaker normative con viction towards culturally relevant norm violations, without universalizing it to people from other cultural groups. Thus, positive response to first two questions will constitute a weak normative conviction -indicating an attitude that "any suitably similar person in roughly the same situation would be subject to the same sort of obligation" (O'Neill & Machery, 2019, p. 43).
Normative considerations. After completing the questions pertaining to the normative conviction, participants were presented with six explicit questions about six possible nor mative considerations that would provide explanations for why a particular transgression is wrong. Participants had to indicate the degree of agreement or disagreement with six statements about the action on the 7point scale, where 1 indicated strong disagreement, and 7 indicated strong agreement. The normative statements were as follows: Agent's action causes psychological or physical suffering. Agent's action disregards the interests of others and is unfair. Agent's action ignores the will of the respected people in the community. Agent's action expresses disloyalty to one's own people, relatives, and friends. Agent's action is one that makes people feel averse and disgusted. Agent's action is uncultured/uncivilized behavior. All material was in the Chinese language. One of the authors, who is an expert in Chinese language and culture, made translations from English into Chinese and faci litated the selection of particular scenarios that are culturally most relevant to Chinese participants. In addition, Chinese assistants orally backtranslated and commented on the comprehensibility and cultural relevance of the stories and questions.

Results
There are two types of responses that are relevant to our hypotheses. First, there is a pat tern of responses to three normative conviction questions in each type of scenario. Then there are responses to explicit scales indicating disagreement/agreement to six statements that provide explanations for why a particular transgression is wrong (normative consi derations). Each type of response will be presented in succession.
Normative conviction responses. There were three questions constituting the norma tive conviction: (a) a transgression question (is this action wrong?), (b) an authority independence question (is it wrong independently of any authorities' opinion?), and (c) universality question (is it wrong in other places?). Subsequently, there were two relevant combinations of responses. Strong normative sense response: transgression plus positive answers to authority independence and universality questions. Weak normative conviction response: transgression plus authorityindependence, but not universality. Table 1 provides frequencies of normative conviction (NC) responses in each scenario. It appears that Chinese participants expressed strong normative conviction (9 out of 21 scenarios) less often than a weak normative conviction (14 out of 21 scenarios) to wards the presented transgressions. Scenarios related to authority disrespect, loyalty to a group and purity (except for the hands washing case) did not elicit a strong normative response. Note, these type of scenarios were taken from the lists that were generated with the American sample (Clifford et al., 2015). However, scenarios that were independently constructed to convey violations of wenming and xiao norms elicited a strong normative response in four out of six scenarios. Then, in line with Turiel and other monists, four out of six harm/fairness scenarios also elicited a strong normative conviction. Furthermore, a nonuniversalized weak normative conviction has been elicited in all harmbased scenarios and two fairnessbased scenarios. Also, two authority disrespect scenarios also elicited weak normative conviction. Finally, all three violations of wenming and two out of three xiao violations elicited a weak normative conviction.
Responses to questions about normative considerations. In the second part, parti cipants had to indicate the degree of agreement (on the 7point scale) to six statements that explicitly articulated these normative considerations. See Table 2 for total means of normative considerations across all domains and scenarios. On the face of it, it is difficult to determine which of the normative considerations underlie wrongness (NC) judgments. To see specific patterns in the ratings and to test for the effects of participants' norma tive considerations on the NC evocations, we conducted a series of regression analyzes. However, given the fact that in some cases participants almost unanimously agreed about the action's wrongness (see Table 1, especially Harm scenarios), a binary logistic regression, with NC as an outcome variable (0 -No, 1 -Yes), is not a feasible analysis. Besides, there would be too many (21+21) regression analyses. For these reasons, we adopted a data reduction strategy. That is, each participant received three scenarios and three pairs of NC responses; this way there can be four possible responses -from 0 to 3 NC responses. Thus, we created a new ordinal variable, where each participant's response to NC questions was summed (as a response to a 4point ordinal scale). As a result, only 14 regression analyzes have been performed. More specifically, using all six conside rations (harm, fairness, authority respect, ingroup loyalty, purity and unculturedness) as predictors and sums of strong/weak normative convictions as the outcome variables, ordinal regressions were conducted for each domain separately. Table 3 shows that, overall, only few normative considerations were significant predic tors of NC responses. In the case of strong NC, harm consideration, unsurprisingly, was the only significant predictor (  From these results (where regressions were conducted for each domain separately), it is difficult to discern the foundational normative concerns that might underlie all types of transgressions. Recall, monists argued that harm (Gray and colleagues) or fairness (Baumard and colleagues), or both (Turiel and colleagues) are foundational normative considerations underlying all "morally" relevant transgressions. For these reasons, we ran ordinal regression analyses across all scenarios (N = 566) to determine which normative considerations are foundational for Chinese participants. Table 4 shows that, across all scenarios, normative consideration of unculturedness (wenming) was a significant predictor of a strong (b = 0.513, 95% CI [0.341, 0.684], p = 0.000) as well as a weak (b = 0.507,95% CI [0.329,0.686], p = 0.000) NC responses. Then, having a somewhat smaller overall contribution, consideration of harm emerged a second significant predictor of a strong (b = 0.310,95% CI [0.157,0.463], p = 0.000) and weak (b = 0.201, 95% CI [0.039, 0.363], p = 0.015) NC responses. While the normative consideration of fairness emerged as a significant predictor only in the case of weak NC (b = 0.367, 95% CI [0.166, 0.568], p = 0.000). Interestingly, loyalty consideration had a negative effect (on weak NC), that is, the more Chinese participants evoke normative conviction, the less relevant loyalty consideration becomes.

Discussion and conclusions
In this paper, we suggested a tentative characterization of strong and weak normative convictions that can be preferentially elicited by various transgressions. Of course, there might be other (and perhaps better) ways to characterize normative convictions, but for the purposes of our study, we choose to adopt Turiel's criteria. This way, the studies could be more or less comparable. As it was indicated in the methods section, there are two related, but conceptually separable attitudes -authorityindependence and universality. We suggested that it is possible to have a weaker nonuniversalistic normative conviction, which implies an obligation to act normatively in a given situation only for suitably re lated people, i.e., from one's own cultural group. Or, as it is the case when a foreigner visits Chinese homes, one is obliged to follow some norms of conduct as a guest, where some violations made by foreign visitors, and in contexts of cultural significance, are met with strong disapproval. Incidentally, the demarcation line between strong and weak normative conviction is rather malleable and sensitive to different contexts. One thing is clear, though -current data indicates that strong and weak convictions overlap. Thus, in what follows, we will first discuss results related to weak, nonuniversalized, normative conviction and then check them against strong, universalizing, normative conviction. First, we hypothesized that Chinese participants would have a strong normative conviction towards situations that depict standard Turieltype or Haidttype of moral transgressions. Indeed, all prototypical scenarios of harmful acts and two unfair acts (except cheating on cards) elicited both weak and strong normative convictions (hitting with ruler elicited only the weak NC). This indicates that, overall, these types of trans gressions, widely explored in the Turiel tradition, are also morally (in the Western sense) relevant issues for Chinese. Even if the Chinese term daode (as a translation of morality) is not readily associated with acts of harm and/or unfairness (see Buchtel et al., 2015;Dranseika et al., 2018). Now, in the case of three other domains (authority disrespect, loyalty and purity), there was a considerable variation in respect to weak normative con viction. Some scenarios elicited NC (laughing with rivals/loyalty domain; interrupting teacher and dress code/authority domain; hands after toilet/purity domain), others did not. Whereas some scenarios were significantly below the mid-point (coach and rivals/ loyalty domain or revealing clothes/purity domain), indicating that these situations are not morally relevant issues whatsoever.
It seems that these latter scenarios, contrary to Clifford et al. results (2015), are bad examples of domain-specific transgressions. Our results show the opposite -it is overwhel mingly a nonmoral issue. It should be noted that these transgressions have not been generated by Chinese participants themselves. As a result, the presented scenarios might be culturally biased and therefore not touch on the concerns of the Chinese young adult population. Similar issues have been encountered in the Moral Foundations Questionnaire (Graham et al., 2011), where a recent replication of the MFQ did not succeed to uncover a five-factor model across different cultures, casting some doubt on the cross-cultural gene ralizability of five moral foundations (Iurino & Saucier, 2019; though see Doğruyol, Alper, & Yilmaz, 2019 for an alternative analyses). Again, items on the MFQ might be culturally biased, this way creating measurement variance. Indeed, recent attempt to validate MFQ with Chinese sample showed that "concrete statements were less representative of the corresponding moral dimensions" (Zhang & Li, 2015, p. 162), pointing to some external validity issues. So, in the future studies, it is advisable to extract relevant items from the population under study. This way MFQ, and other questionnaires, will only benefit and enhance internal and external validity.
Second, given the long-standing influence of the Confucian ethical tradition on Chinese society, we hypothesized that participants will also elicit a strong normative conviction to wards situations that depict wenming (uncultured) behavior, which corresponds to Turiel's violation of conventional norms. This was the case. All three scenarios in the wenming domain elicited weak NC, and two of them elicited strong NC (cursing and spitting). Importantly, when prompted, all three actions have been freely generated by Chinese participants as the most prototypical instances of both bu wenming (uncultured) or bu daode (immoral) behavior (Dranseika et al., 2018). Taking previous and current results together, we could argue that bu wenming transgressions are not only cognitively salient, but also elicit a strong normative response. Similarly, we hypothesized that participants would elicit a strong normative conviction towards violations of traditional values of xiao (filial piety). This was also the case. Except for the requirement to have kids, an obliga tion for children to visit parents and for parents to provide proper education emerged as important norms. Their violation is met with strong normative response.
Finally, we predicted that Chinese participants would express not only five normative considerations, but also will be concerned whether an action is generally an instance of bu wenming (unculturedness). Table 3 shows that regression models in each domain did not provide a clearcut picture of foundational normative considerations, though considerations of unculturedness emerged more often as a significant predictor of strong and weak NC. On the other hand, Table 4 shows that, overall, considerations of whether an action violates wenming norms had the biggest contribution on the NC judgments. This could be interpreted as an additional evidence for the importance of specific cul tural norms in generating "moral" judgment. The fact that Chinese participants elicited strong normative response to wenming transgressions and later evoked considerations of wenming when appraising all the scenarios, yet again points to the cognitive salience of culturally acquired wenming (and daode) norms. However, as Table 4 indicates, this does not mean that considerations of harm and fairness do not play any role. On the contrary, these normative considerations emerged as other, albeit weaker, predictors of the NC responses. Unlike other recent studies with Western samples (American and Greek), that used slightly different but comparable methodology (Piazza et al., 2019), current study with Chinese participants does not provide strong evidence for the foundational status of fairness considerations, though acknowledging its importance.
Taken all together, current results have some implications for the debates on the moral/conventional distinction and the scope of moral domain. Research in the Turiel tradition made it clear that Westerners make a distinction between moral and conventional transgressions. In the same vein, Yau and Smetana (2003), in their study with Hong Kong preschoolers, found that Chinese children make a similar distinction, which would support a claim about the universality of moral/conventional distinction. In this paper, however, we argued that in the Chinese ethical tradition, conventional cultural norms and regulations of civilized behavior (wenhua 文化 / wenming 文明 in contemporary Chinese) are tightly connected to the notion of daode. If Confucian ethics had any lasting influence on Chinese society as a whole, then transgressions of wenming norms should evoke strong conviction, similar to one that is observed in Western respondents reacting to typical moral transgressions. Current results support this hypothesis. Thus, a clearcut moral/conventional distinction might be a cultural product of the Western ethical tradition and the English language (Buchtel et al., 2015;Dranseika et al., 2018; see also Machery, 2018;Machery & Mallon, 2010;Sachdeva, Singh, & Medin, 2011;Silius et al., 2017;Sripada & Stich, 2006;Stich, 2018;. We suggest that any kind of normative transgression (not necessarily fixed by Haidt's list) that evokes strong normative conviction can be counted as a culturally circumscribed normative issue. It is a flexible type of moral pluralism, where the term "moral" could be dropped in favor of "normative," in order to include nonWestern samples that lack such a term, and to avoid unnecessary connotations that often bias this term.