The Mischievous, the Naughty, and the Violent in a Taiwanese Village: Peer Aggression Narratives in Arthur P. Wolf’s “Child Interview” (1959)
On June 15, 1958, Arthur P. Wolf (hereafter, APW), then an anthropology graduate student from Cornell University, arrived at a village south of Taipei, Taiwan, for his dissertation fieldwork. His wife Margery Wolf soon joined him there. More than half a century later, the eminent anthropologist documented the forever-fresh memory of the village and its children in the opening chapter of a draft of his final book just a few months before he passed away:
In 1958 Lower Ch’i-chou was home to thirty-four families with a total membership of 567 persons. They occupied twenty-three houses strung out along the west bank of the Ta-k’e-ken River in the midst of banyan and bamboo. With the river in back and flooded paddy fields in front, the village was like an island for the children living there. The younger ones were not allowed to play near the river for fear of drowning, and no child was allowed to play in the paddy for fear or trampling the rice or breaking down the bunds. (A. Wolf 2015, 8–9)
Under the pseudonym Peihotien, the village of Lower Xizhou has become an iconic landmark in scholars’ map of Taiwan studies and Sinological anthropology (Freedman 1968). It was here that APW and Margery Wolf launched their distinguished careers in anthropology, and this research played a significant role in many of their influential publications on kinship, family, marriage, gender, and sexuality in Taiwan (Wolf and Huang 1981; A. Wolf 1966, 1968, 1970, 1995; M. Wolf 1968, 1972, 1978, 1990, 1992). The research of APW, Margery Wolf, and other anthropologists in the Haishan area, including Lower Xizhou and neighboring villages and towns, has made important contributions to the study of Taiwan and Chinese culture (Wolf and Huang 1981, x).
Imagine, as the young APW walked into this village on the path from the nearby town of Shulin (where the train station connecting to Taipei was located), “on the left hand side of the path, stood the great banyan tree pictured on the cover of Margery Wolf’s The House of Lim, and, on the right-hand side of the path, the home of the Lim family, the largest and finest in the village” (A. Wolf 2015, 8). Students of Taiwan studies might well remember the House of Lim and various women and men from Margery Wolf’s classic ethnographies about the village (M. Wolf 1968, 1972), a Hoklo village with Han people whose ancestors migrated from Quanzhou, Fujian Province, Southeast China, during the eighteenth and nineteenth centuries (A. Wolf and Huang 1980 35–56). Unknown to many, however, the main research topic of the first fieldwork performed by APW and Margery Wolf was neither the Lims nor the women and men, but the village children. The bulk of materials these researchers collected on children and childrearing remain unpublished, because the research stimulated other interests for APW, and also because the analysis of this massive collection of data was too challenging. APW’s work in Lower Xizhou was the first ethnographic and mixed-methods project on children in Taiwan, and even on children in the entire ethnic Chinese population. He understood the unique potential value of this research: “We should emerge from analyzing these data with dramatically greater systematic knowledge about childhood in China than we have ever had before” (A. Wolf 1982, 4).
One interesting subset of this unique data archive (hereafter, the Wolf Archive) is a portion of interviews with seventy-nine children ages 3 through 10. Focusing on this subset, this article aims to tell a heretofore lesser-known story about children and childhood. In particular, it illuminates a dark side of moral development that has been obscured in historical and anthropological studies of Han Chinese childhood by analyzing children’s own narratives of peer aggression, and situates these materials in larger conversations in the anthropology of childhood.
The Wolf Archive, the Six Cultures Study, and the Dark Side of Moral Development
The imagery of “the child” has assumed a significant and evolving role in Han Chinese culture, due to its significance in various branches of philosophical and moral thoughts (Hsiung 2005, xi), its place in cosmological order (Topley 1974), its connection to educational desire and political governance (Bakken 2000; Kipnis 2011), its centrality in the project of nationalistic modernization in the late nineteenth to early twentieth century (Jones 2011) and postwar Taiwan (Stafford 1992), and its importance in Taiwan’s positioning in today’s global political economy (Lan 2018). Among the first few ethnographic studies of Taiwan, the 1958–1960 fieldwork of APW and Margery Wolf not only is a milestone in the “Golden Age” of Sinological ethnography (Harrell 1999) but also marks a historically significant moment when Chinese studies intersected with the anthropology of children and childhood. APW’s dissertation research was designed as an improved replication of the classic Six Cultures Study of Child Socialization (SCS), under the supervision of Cornell psychologist William Lambert, one of the leading figures in SCS, and anthropologist Lauriston Sharp, APW’s dissertation advisor.
Based on comparative fieldwork in six diverse cultural settings in the mid-twentieth century, SCS was a landmark study in the anthropology of children (LeVine 2010). That project focused on children ages 3 through 11 (with a total sample of 136 children from six sites), utilized a series of standardized anthropological and psychological methods, and produced theoretical, ethnographic, and methodological publications on culture and child development (Whiting, Whiting, and Longabaugh 1975; B. Whiting 1963; J. Whiting 1966). APW’s research went beyond SCS in several aspects: he targeted a much larger sample (sixty-four children of primary focus, half male and half female, ages 3 through 10 when the fieldwork started in 1958, in contrast to twenty-four per site in SCS). He ensured the collection of valid data for all of the methods prescribed in SCS, including Child Observation in naturalistic contexts, Child Interview, and Mother Interview; psychological tests such as Doll Play and Thematic Apperception Test (TAT); and general observations of the community. In addition, he added a School Questionnaire for elementary school students in Shulin, observations of mother-child interactions (Mother Observation), a survey of infant care and mother-infant interactions (Baby Survey), and physiological measures (children’s urine samples).
Notably, APW’s own thinking about the nature of this research evolved from scientific hypothesis testing to documenting the children’s world of the past:
Had I written in the 1960’s as intended, I would have focused on testing the hypotheses formulated by the Six Cultures Study. I now pay more attention to reporting as accurately as possible the data I collected. The reason is simply that I now know that my observations in Lower Ch’i-chou can never be replicated. … The path I walked from the train station to the village is now a paved thoroughfare lined with high-rise apartment buildings. The school my subjects attended has been demolished and replaced with a modern multi-storied facility. Three of the village houses still stand but only because a strip of land along the river has been designated a green area. They are unoccupied. A few of the villagers live in apartment buildings near their old homes, but the majority have moved away. What I knew as a village is now an ill defined urban neighborhood. (A. Wolf 2015, 36)
In the 1990s, thirty-five years after APW’s original research, Maria Duryea, then an anthropology doctoral student from the University of Washington, revisited the original field site with him. Her dissertation focused on social transformations of that community, including its drastic urbanization and economic development and the impacts of those changes on childrearing practices (Duryea 1999). However, Duryea was not able to do systematic observations and interviews with children as APW and Margery Wolf had done in 1958–1960. In Duryea’s words, children’s social lives were not as readily accessible, “increasingly removed [as they were] from the interstices of the residential neighborhood” (1999, 105).
Given the rare duality of the Wolf Archive––its significance in anthropology and its unique historical nature—its analysis makes it possible to illuminate children’s lived experience in its historical and cultural context and to bridge the anthropology of children across cultures and historical studies of critical issues in child development. This article specifically presents analyses of the Child Interview data, one of the types of data in the Wolf Archive, with a focus on peer aggression narratives, a darker side of childhood.
These archival records are important because they bring to light a dark side of childhood that remains understudied in the Sinological research of childhood. In the wake of debates about “the discovery of childhood” in European history by Philippe Ariès (1965) and his critics (e.g., Pollock 1983), Sinologists in recent decades have turned their attention to children (Kinney 1995, 2004; Fernsebner 2003; Bai 2005; Hsiung 2005; Tillman 2018), the arguably most overlooked population in historiography (Hsiung 2005, 261). Collectively, these scholars have made an important contribution to reconstructing historical childhood. However, historical studies largely rely on materials pertaining to discourses, representations, and memories about childhood, and those materials rarely include children’s own narratives or experiences. Especially obscured in historical studies is the dark side of children’s own social world—the mischievous, the naughty, and even the violent interactions among children, especially children from non-elite families. Historical representations of Chinese childhood tend to fixate on the “good” and “innocent” (see Hsiung 2005, xi; Bai 2005, 1–20) and the importance of moral cultivation in Chinese views of childhood (Bai 2005; Kinney 2004). Even in the few existing studies about “bad-behaving” children––studies that mainly rely on representations of children or memories of childhood––the focus has been on vertical relationships between youth and their families, schools, or political authorities, especially during turbulent times (Mather 1995; Knapp 2017; Saari 1990; Chan 1985; Unger 1982; Lupher 2005; Solomon 1971; Wilson 1974), rather than young children’s own narratives of peer aggression in everyday contexts. By contrast, although the anthropological research of children in Taiwan has paid attention to children’s “bad behavior,” the literature nonetheless focuses on adult discipline of children’s transgressions (Stafford 1995; Fung 1999), rather than on children’s peer interactions.
The very task of reanalyzing these historical data on peer aggression has theoretical implications in light of larger paradigm shifts in studies of Taiwan as well as in studies of culture and child development. These data occupy a unique niche in connection to SCS. The SCS teams’ analyses of aggression have mostly focused on observational materials (see Lambert and Tan 1979), but Child Interview in those teams failed to yield good data. In contrast to SCS, not only did the Wolf team manage to interview many more children, but Child Interview data in the Wolf Archive are distinctly more valid, hence publishable, which I will explain in more detail in the next section. The study of Taiwan has shifted from a provincial part of Sinology to a research area in its own right (see Shih, Thompson, and Tremlett 2008). Potential cross-cultural comparisons between the Child Interview data in the Wolf Archive and observational data in SCS can shed valuable light on Taiwanese childhood and morality in a global, comparative framework.
Moreover, my reanalysis follows new theoretical frameworks on culture, morality, and child development. Following the postwar trend in American anthropology’s “Culture and Personality School”—understanding aggressive/nonaggressive societies through studying child socialization (see Brown and Schuster 1986)—the SCS project saw aggression as a focal topic. Aggression was also a central theme in APW’s earlier writings about this archive––for example, in his unpublished conference presentation (A. Wolf 1964) and proposal for the National Science Foundation (A. Wolf 1982). Significant changes have taken place since the SCS era (LeVine 2010), most prominently the paradigm shift from behaviorism to cognitive science (Miller 2003). Whereas the SCS project treated the human mind as a black box and theorized learning as stimulus-response processes with the aid of reward-punishment mechanisms, today the mind is assumed to have a central, active role in any meaningful understanding of behavior. Correspondingly, at the intersection of anthropology and cognitive science, synergies between evolutionary, developmental, and cognitive approaches have led to new theoretical understanding of culture and child development. This new paradigm sees children as active and creative learners in acquiring and transforming culture, and goes beyond the old nature-nurture dichotomy to examine interconnection mechanisms (see Boyer 1998; Hirschfeld 2002; Tomasello 2019). The origins and development of cooperation and morality in childhood are among the most vibrant areas under this new paradigm, and they form a promising direction in the anthropology of morality (see Xu 2019).
Specifically, whereas SCS foregrounded the ecology-behavior relationship (Weisner 2010) and APW (A. Wolf 1964) highlighted the mothering style-children’s physiology linkage in aggression, my aim in this article, through analyzing children’s narratives in inter-subjective encounters (interviews), is to illuminate children’s developing knowledge in the context of their socio-moral life. In line with my previous research on culture and moral development (Xu 2014, 2017, 2019), this article combines qualitative and quantitative analyses to examine children’s understandings of peer aggression in relation to adult ideologies—the Chinese cultural value, he, or social harmony—on the one hand and their own developing cognitive capacities underpinning human cooperation and conflict (i.e., negative reciprocity) on the other.
In the next two sections of this article, I introduce the Child Interview methodology and data coding and analysis and report my main findings on a complex spectrum of reciprocity in peer aggression. The conclusion articulates broader theoretical implications of these findings and future directions for research.
Child Interview on Peer Aggression: Methodology, Themes, and Coding
Child Interview: Methodology
Child Interview was an important method used in APW’s fieldwork, administered to children ages 3 through 10 in the year 1959. It was almost an exact replication of the “Child Interview” protocol from SCS, designed as an important individual measure of children’s lives. But Child Interview in SCS failed to yield as reliable data as the data produced by naturalistic observations that have left a long-lasting legacy in cross-cultural research on children (J. Whiting 1966, 120–124; LeVine 2010). This nonsuccess of Child Interview in SCS can be attributed to children’s low response rates, their less-than-meaningful responses to standardized questions, the length of fieldwork, and the limited sample sizes in individual sites (LeVine 2010). The average Child Interview sample size is about twenty, and Child Interview data were barely included in SCS publications. By contrast, APW’s team managed to collect valid interview responses from seventy-nine children (median age 7 years, minimum age 3.08 years, maximum age 10.5 years; 43 girls and 36 boys), which makes this Taiwan data set unique and valuable.
APW’s team translated SCS’s interview protocol into Chinese and administered Child Interview in Hoklo, the target children’s native tongue. The English transcripts of all seventy-nine interviews are intact in the Wolf Archive. Although the SCS researchers reported that children under 6 did not respond well to interview questions or the questions were too abstract for them to understand (J. Whiting 1966, 124), in Lower Xizhou even the younger children responded reasonably well to the interviews. A few factors contributed to APW’s successful collection of Child Interview data. First, his fieldwork was twice as long (two years) as the typical SCS fieldwork; second, children were interviewed by APW’s two excellent Taiwanese research assistants, young women who blended well into the local community, got along with children in the village, and dutifully collected the data. Children even called one research assistant “older sister.” These seventy-nine interviews include complete responses from sixty-two children in APW’s primary sample (sixty-four focal children of naturalistic observations and other methods). Therefore, Child Interview data can be further analyzed together with other types of data in this archive.
Peer Aggression: “Insult,” “Dominance,” and “Assault”
Based on SCS project design, the rationale of Child Interview was to present children with hypothetical but familiar situations and elicit their reactions or feelings. The interview protocol combined standard questions with subsequent probes to allow a certain degree of flexibility (Whiting et al. 1953, 124–126). In line with the SCS project protocol, APW’s Child Interview consisted of fifteen main questions, including various subsequent prompts, which function as probes, under each main question. This article focuses on the three main questions related to aggression in children’s peer interactions, defined in the interview protocol as “insult,” “dominance,” and “assault.” I group these three questions together under the category of aggression because, from a bottom-up perspective, children’s responses to these three questions displayed some degree of overlap, especially regarding verbal and physical aggression and tattling, even though the narratives elicited by each question showed distinct patterns. In fact, narratives about aggression-related behavior and feelings even surfaced in children’s responses to some other questions in Child Interview, suggesting the saliency of aggression in these children’s everyday life.
Data Coding and Analysis
The theoretical orientation and framework of SCS received heavy criticism in both anthropology and psychology (LeVine 2010). Hence, this article does not follow SCS’s categorization and analytic approach. After carefully reading and organizing all the interview notes, I gained insights about the materials at a holistic level and designed the coding scheme accordingly. I coded children’s responses to all main questions and their subsequent prompts. I analyzed individual prompts within every question, the relationships between prompts under a certain question, as well as the relationships across different sets of questions. In the first step of coding, for most questions that could be categorized in a “Yes/No” format (similarly, “Positive/Negative,” “Negative/Neutral,” and so forth), I coded the answers in a binary manner, and coded the rare, outlier answers as “contingent,” “confusing,” or “compromise,” so that they could be analyzed and interpreted both separately and together with the binary answers. For some questions or prompts concerning feelings or emotional states, I coded the answers using a tripartite system (i.e., “Positive/Neutral/Negative”). I then extracted central features from children’s answers and classified them under succinct categories to reveal their underlying rationales and justifications. The first step allowed me to perform statistical analyses in which children’s responses within each set of main questions and prompts were coded as categorical data. The second step, intermediate coding, allowed me to explore meanings underneath the quantitative analyses. These two steps were interconnected, and I constantly went back to the original answers in relation to patterns gleaned both from other children’s answers to the same question and from the target child’s answers to other questions. This analytic approach offered a comprehensive understanding of children’s narratives.
The Mischievous, the Naughty, and the Violent: Child Interview Results
In the analysis of the data from Child Interview, the children’s narratives reveal a complex spectrum of reciprocity in their understandings of aggression. They provide glimpses into a world of mischievous children, naughty children, and sometimes even violent children. The children distinguished serious from milder forms of aggression in perceived negativity and they reacted differentially. These perceptions and reactions reflect important concerns and strategies in local socio-moral life, some of which run counter to adult ideologies.
In this section, I first present analyses of the children’s answers to each of the questions related to the themes of “assault,” “insult,” and “dominance.” I start with “assault,” because physical aggression is the most severe and least ambiguous of the three forms of aggression. Under each theme, I list the specific interview question and prompts and then analyze the children’s answers to each prompt, first qualitatively and then quantitatively, articulating meaningful associations of answers across prompts, and also investigating the impacts of age and gender. After examining all three themes independently, I offer tentative explanations of these findings, compare and analyze the interconnections between these questions, and interpret what they reveal about aggression in peer interaction and socialization more broadly. Quantitative analyses are based on inferential statistics about proportions of children’s answers: N refers to sample size, n refers to sub-sample size, and p refers to calculated probability of a certain proportion (a p-value less than 0.05 is statistically significant).
“Assault”: Prompts and Answers
The English version of the interview guide followed the standard protocol in SCS, using the letter “P” to refer to the target person (interviewee) and the letter “O” to refer to “the other person” in the hypothetical scenario. For the situation or theme of “assault” (O assaults P), the main and subsequent prompts were as follows: “(A) Suppose another boy your age comes up and hits you: What would you do? (B) Suppose he hit you easy? (C) Suppose he really socked you?” In the Chinese version, the prompts were framed in a gender-neutral manner, using the generic designation “another child” instead of “another boy.” The Chinese version of prompt B also inserted the adverb “just” before “hit you easy.”
I coded all three sub-questions as binary variables, whether or not P seeks revenge/intervention, including both directly and indirectly (Yes/No). Specifically, “Yes” answers include direct revenge, mostly physical aggression of various, detailed forms, and indirect revenge––for example, tattling and enlisting help from authority figures (parents and teachers) and older brothers to potentially punish O (scold or hit him). “No” answers mainly include avoidance (“Run back home”) and ignoring (“It doesn’t matter”). Their concrete responses show nuanced contingency to the severity of the scenario, and I call these responses “reciprocal physical aggression”: in the situation of being hit “easy,” many answers express the idea of hitting reciprocally (e.g., “Hitting him easy”) with details; in the situation of being hit hard, many endorse hitting back hard, with rich details. Here are a few examples of the children’s answers: “Hit him dead,” “Hit him with a bench,” “Hit him with my fist,” “Slap him,” “Other children will call the teacher,” “Call older brother to hit him,” “Slap him, tickle him, hit him with a rock,” “Sock him hard back,” “Take a rock and hit him,” “Hit him hard under the ribs,” “Slap him and run back home,” “Everybody is afraid of me because I can throw him down,” and “Tell my father to buy a gun and shoot him.”
Quantitative analysis corroborates this finding about reciprocal aggression (see Table 1). First, children would prefer revenge/intervention in the severe assault (A) and generic assault (C) scenarios but would prefer non-revenge in the mild assault scenario (B). Furthermore, across all three prompts, their expressed choices for revenge or not are contingent on the severity of the scenario (Cochran’s Q test, N = 75, p < .001). Specifically, more children would prefer revenge under prompt A (O hits P) than prompt B (O just hits P easy) (McNemar test, N = 75, p < .001); similarly, more children would prefer revenge under prompt C (O hits P hard) than prompt B (O hits P easy) (McNemar Test, N = 75, p < .001). The children’s answers are not contingent on gender or age, in general, with one exception: under prompt B (O hits P easy), there is a subtle pattern that the older a child is, the more likely s/he would prefer non-revenge (p = .005, binomial logistic regression).
Table 1. Children’s answers to the “assault” scenario: statistical analyses of primary variable (revenge/intervention)
“Insult”: Prompts and Answer
For the theme of “insult” (O insults P), the main and subsequent four prompts were as follows: “(A) Suppose another boy (girl) your age makes fun of you: What would you do? What if he (she) says you are stupid? (B) How would you feel? (C) Suppose it was a boy (girl) much bigger than you? (D) Suppose it was a boy (girl) much littler than you? (E) Suppose it was a girl (boy)?” O refers to a child from the same gender in prompts A–D and the opposite gender in prompt E.
Across all prompts (except B), the primary coding criterion was whether or not P intends to seek revenge/intervene (Yes/No) in the situation when some other child makes fun of P. “Yes” answers include physical aggression (e.g., “I will hit him” or “I will hit him later when I get the chance”), verbal aggression (e.g., “I will scold him/laugh at him back”), social exclusion (“Not make friends with him), tattling and asking an authority to punish (e.g., “Ask my mom/the teacher to hit him”). “No” answers include ignoring (“It doesn’t matter” or “Not listen”), avoidance (e.g., “I will run away”), and no punishment (e.g., “Not hit him”).
As Table 2 shows, none of the behavioral prompts (ACDE) yielded statistically significant results. In other words, children did not express a preference for revenge/intervention in the situation of being made fun of by another child, regardless of the age (older or younger) or gender (same or opposite gender) of that other child.
Table 2. Children’s answers to the “insult” scenario: statistical analyses of primary variable (revenge/intervention)
Notably, more children expressed the intention of revenge/intervention when the other child present was the same age (prompt A) than when the other child was much younger (prompt D) (McNemar test, N = 72, p = .007). A reasonable interpretation is a concern of social propriety, e.g., yielding to younger children, a trend Margery Wolf observed in this community (1978, 245). By contrast, there were no significant differences in their answers between prompt C (when the other child is much older) and prompt D (when the other child is much younger), or between prompts C and A (when the other child is the same age). Although numerically there are more “Yes” than “No” answers in prompt C but more “No” than “Yes” answers in prompt D, the difference in proportions is rather small (p values larger than .05). The fear of offending older children might make some interviewees less willing to say that they would want to intervene/seek revenge––for example, in answering prompt C, some children said, “I want to make friends with him” or “I will laugh at his younger brother.”
For prompt B (emotional reaction), the primary coding criterion was whether P’s response was negative feeling (Yes) or neutral feeling (No). The finding is that more children reported neutral feelings (63 percent, 48 out of 76) than negative feelings (binomial test, p = .029). Some children repeated their answers to prompt A (revenge or not), thus the same categories as presented above; other children mentioned feelings, either neutral (e.g., “Nothing”) or negative (e.g., “Feeling mad,” “angry,” or “embarrassed”). This pattern of neutral rather than negative feeling might shed some light on why children did not express a preference for revenge under the behavioral prompts (ACDE).
Intriguingly, even though on an aggregate level children did not express a preference for revenge/intervention, the child’s age, but not gender, made a difference on an individual level: across all four prompts, the older the child was, the more likely s/he would prefer no revenge/intervention (binomial logistic regression, ps < .05).
“Dominance”: Prompts and Answers
For the theme of “dominance” (O attempts to dominate P), the main and subsequent prompts were as follows: “(A) Suppose you are doing something interesting, and another boy your age tries to make you do something else: What would you do? Suppose you were playing ball, and another child called you to do something else: What would you do? (If no answer) Would you go with him or not? (B) How would you feel? Would you be mad? Would you be willing to go? Would you be happy? (C) (If the child says he would not go) What if it was something more interesting? (If the child would go) What if it were something less interesting? (D) (If the child says he would not go a second time) What if he kept trying to make you do it?”
Across all prompts (except B, emotional reaction), the primary coding criterion was whether or not P submits to O’s dominating attempt (Yes/No). The finding is that children, under the three prompts (ACD), did not express a preference for whether or not to submit to domination (see Table 3). Further analysis comparing prompts A and C reveals that making O’s proposal more interesting (when the answer to A is “No”) or less interesting (when the answer to A is “Yes”) did not change the pattern of children’s responses (McNemar test, N = 71, p = .256). However, children’s answers to prompt B (emotional reaction), coded into three categories—positive (n = 37), neutral (n = 10), and negative (n = 26)—revealed a significant pattern (chi-square test, p = .001): more children expressed a non-neutral (positive or negative) feeling than a neutral feeling.
Table 3. Children’s answers to the “domination” scenario, statistical analyses of primary variable (rejecting domination/intervention).
Examining the children’s concrete answers can explain the apparent gap between “feeling” (non-neutral) and “behavioral choice” (no preferences under prompts ACD): P can be compelled to submit to O’s domineering attempt out of negative feelings, e.g., embarrassment (of not going with O), or positive feelings, e.g., feeling happy to go with O. Likewise, P can reject O’s domination out of negative feelings, e.g., anger, or positive feelings, e.g., feeling happy to do one’s own thing. Taken together, children’s answers regarding both behaviors and feelings reflect a variety of concerns, such as autonomy or rejecting domination (e.g., “You should do things by yourself,” “Not listen to him,” “Not go, because he purposely, knowing that I was doing something else, called me to do something else”), the nature of the activity (e.g., “If it is for play, then I’ll go; if work, then I’ll not go), authority rule or punishment (e.g., “My mom would tell me not to go,” “My mom would hit me if I go”), friendship, time contingency (e.g., “If I had time, I would go, otherwise I would not go”), avoidance (e.g., “Run away”), politeness when facing persistent requests (e.g., “Because he keeps asking/not going would be embarrassing”), as well as normative expressions (e.g., “I ought to stay home and not go”).
Children’s answers across prompts are not contingent upon their age or gender, with one exception: among those who persistently answered “No” to previous prompts, older children preferred to submit to O’s second domineering attempt (prompt D). A plausible interpretation is that, with age, these children become more concerned about social propriety—the embarrassment of not cooperating upon persistent requests—or at least they become more aware of this norm.
Aggression Narratives: Negative Reciprocity, Ideology, and Reality
Comparing children’s narrative responses across the three themes of “insult,” “dominance,” and “assault” offers a comprehensive picture of their expressions, perceptions, and reactions related to peer aggression. Quantitative analyses of children’s answers to the main prompt (A) across the three themes show that the same child was more likely to express a preference for revenge/intervention in the scenario of assault than in that of insult (McNemar test, N = 68, p = .014) or dominance (McNemar test, N = 70, p = .007); but in a child’s expressed preferences, there was no difference between the scenario of insult and that of dominance (McNemar test, N = 69, p = .248). In other words, among the three categories, “assault” stands out as the one theme eliciting a clear preference for revenge.
A close look at the children’s narratives illuminates the contrast between “assault” and the other two themes, the similarities in children’s justifications across themes, the subtle differences between “insult” and “dominance,” and how such patterns reflect the social meanings of these scenarios in the local context. First, their rich narratives of physical revenge in response to generic and severe assault scenarios suggest a more negative and reciprocal reaction to physical aggression than other, perhaps milder, forms of aggression.
Second, with regard to “insult,” some children, in responding to the prompt about their feelings or emotional states, reported their behavioral reactions instead of feelings, and a majority of children reported neutral feelings. One contributing factor might be a reluctance to engage in introspection about emotions in interview settings in village life (Potter 1988), especially about hypothetical scenarios. But their colorful narratives about feelings in response to the other theme, “dominance,” contradict this hypothesis to some extent. A more plausible explanation is that the prevalence of neutral feeling under “insult” aligns with the pattern of neutral behavior in their narrative responses, suggesting that children do not perceive “being made fun of” in a truly negative way. As APW stated in the draft of his final book, “Insults laughingly is not the same as insults angrily” (2015, 33). Studies in other parts of rural Taiwan at a later period also observed that teasing––in verbal and mild, physical forms––is usually a playful game; the purpose is not really to pose harm to children but to teach them important lessons about desirable socialization goals—for example, to toughen up (Stafford 1995, 51–53, 179–180). This idea of mild physical aggression as playful teasing also helps explain why children express a preference for non-revenge in the scenario “O only hit P easy” (prompt B in “assault”), contrary to their reactions to more severe types of aggression.
Third, although children’s reactions to “insult” and “dominance” do not differ statistically at the level of binary codes, their concrete narratives reveal nuanced concerns. In contrast to “insult,” the theme of “dominance” elicited a richer set of emotional statements, including contrastive justifications such as autonomy and social propriety (“embarrassment”). Therefore children’s perceptions of “being asked by another child to do something else” seem more ambiguous and diverse than their attitudes toward physical aggression or teasing.
Fourth, although “assault” and “insult” carry different social implications in the local context, there are some common categories in their responses to these two situations, such as physical revenge, avoidance, and tattling. Notably, their responses reveal three types of tattling strategies: reporting to their own parents (mostly mothers), reporting to their teachers, and reporting to the transgressor’s parents (mostly mothers). This last strategy, which Margery Wolf called an “ingenious technique of retaliation,” seems to be “both safe and rewarding,” because their own parents might get mad about any conflicts in which they are involved (1978, 239).
Fifth, children’s responses to more serious types of aggression––generic assault and heavy assault––are not associated with age or gender, suggesting the robustness of reciprocal aggression attitude to physical assault. By contrast, age, but not gender, is associated with their responses to some milder, or less negative, forms of aggression scenarios: their preferences for non-intervention increase with age, in response to all prompts under “insult,” to the last prompt (O’s persistent attempt) under “dominance,” as well as to the mild assault scenario. This contrast lends further support to the conclusion that children distinguish between serious and milder forms of aggression in perceived negativity, and they react accordingly. It also suggests that as children grow older, they might be more aware of, and they might even identify with, the moral norms of non-retaliation, but not to the point where they will not want to intervene in serious physical conflicts (see Harrell 1982, 134–149).
Lastly, such patterns and narratives can shed light on a fundamental question concerning the nature of these materials: do children’s answers reflect the moral norms about what they should do, or the reality, what they would do? Even if children only reported what they ought to do, the answers would still give us important insights into local moral norms and their early acquisition. I argue, however, that these answers not only tell us about local ideology regarding peer aggression but also reflect, at least partially, what children want to say or do in real life.
Take physical aggression as an example. Fighting among children was very common in this village. Conflicts among children could be so severe as to estrange families and destroy adults’ relationships in this close-knit community (M. Wolf 1990, 1992; Duryea 1999, 93–95). Parents, whose paramount concern is maintaining good relationships with their fellow villagers, see such conflicts as very severe, and the prevalent parenting ideology at the time that APW and Margery Wolf performed their fieldwork was to teach children to avoid conflicts with peers and punish children at home if they are involved in fights (Duryea 1999, 105). One way to achieve this goal is to tell children “that other children will beat up on them if they go out to play.” Yet the children, old and young, were not reluctant to describe vengeful actions in splendid detail when presented with the hypothetical physical aggression scenario.
These narratives in Child Interview pose a striking contrast to children’s responses in another standardized measure, the School Questionnaire used in elementary schools in nearby townships. When asked about reactions to physical assault, children responding to the School Questionnaire tended to give socially desirable answers (“Do nothing”), rather than saying what they would really do (“Hit others”) (A. Wolf 2015, 21–22). Such a contradiction has to do with methodological differences between these two types of data. Child Interview was conducted in local dialect in a familiar, informal setting by APW’s research assistants, young Taiwanese women whom these children saw as “older sisters,” whereas the School Questionnaire was administered by APW himself, a foreign man who was much more senior in age and status than the schoolchildren. When APW administered the questionnaire in the classroom, it made no difference to children’s responses whether or not the teacher was present (A. Wolf 2015, 22). Perhaps a more important explanation for the variance between the two sets of data is the classroom setting itself. Mandarin was the only (official) language allowed at school; teachers never would have dared to speak Hoklo, and under the Kuomintang (KMT) regime’s nationalistic language policy between the mid-1950s and 1979, children were punished for speaking it (Klöter 2004). The settings and interviewers of Child Interview could explain why the children might have been more comfortable disclosing what they would want to do, instead of merely what they should do.
Conclusion: Aggression, Negative Reciprocity and Moral Development
Child Interview reveals a rich spectrum of “negative reciprocity”––defined as responding to a negative action with a negative action––in children’s understandings about peer aggression. First, physical aggression, especially generic and more severe assault, elicits the strongest preferences for revenge (most prevalently physical revenge, followed by tattling), and children’s relative preferences for revenge corresponded to the severity of assault (easy, generic, severe), revealing the most robust pattern of negative reciprocity. By contrast, negative reciprocity is much more ambiguous in children’s answers to questions about “insult” and “dominance”; their responses to these two scenarios were less negative and less uniform. Moreover, despite the apparent negative valence of all three abstract labels (insult, dominance, and assault), children’s narratives demonstrate a rich set of strategies in handling aggression, such as physical revenge, verbal revenge, tattling, and avoidance. These narratives reflect locally specific understandings, such as the playfulness of teasing and mild physical aggression, and they reveal diverse, sometimes contradictory, concerns (such as the contrast between social propriety and “tit-for-tat” revenge) in their social life. Divergent from the earlier behaviorist paradigm in SCS but congruent with new, cognitive approaches, these strategies, understandings, and concerns illustrate how children’s developing social cognition—such as intention evaluation (Vaish, Hepach, and Tomasello 2018)—in the process by which natural predispositions are culturally “mediated” (Wertsch 1994) influence reciprocity in children’s peer aggression.
Bringing to light the dark side of moral development, the findings of this article contribute to bridging the studies of Han Chinese childhood in historical times and the anthropology of children across cultures. They highlight children’s narratives of peer aggression that would otherwise remain obscured in the historical literature. Children’s varied responses to peer aggression scenarios illustrate the complexity and ambiguity of he, roughly translated as “harmonization” or “social harmony.” He has long been considered a core ideal in traditional Chinese philosophy (Li 2006) and childrearing beliefs––for example, reconciling mother-child cosmic antagonism (Topley 1974). However, close examination of adults’ actual interpersonal communication in Taiwan has revealed turbulent “currents” of aggression and manipulation underneath the surface of performed “social harmony” (Chang 2001). Margery Wolf’s House of Lim (1968) has shown us the paradoxical coexistence of harmony and conflict within and between Lower Xizhou families. Zooming into the understudied world of children’s peer interactions, this article sheds light on children’s complex motivations, emotions, and understandings in response to conflict scenarios, despite the parental ideology of maintaining harmony.
Moreover, reciprocity is an important foundation of human morality (Curry, Mullins, and Whitehouse 2019), yet the vital issue of negative reciprocity, which might be psychologically distinct from positive reciprocity (Yamagishi et al. 2012), remains under-theorized in anthropology (Narotzky and Moreno 2002). In particular, the origins of reciprocity in childhood have not received much attention in cultural anthropology compared with psychology (Olson and Spelke 2008; Vaish, Hepach, and Tomasello 2018). Offering important insights into the development of negative reciprocity, this article joins forces with the recent move in cultural anthropology to take children’s learning seriously (Blum 2019) and to critically engage with cognitive science in understanding morality (Xu 2019).
The Child Interview data pose intriguing questions for future research and call for extending the analysis to the vast data of Child Observation in the Wolf Archive. One important question is why Child Interview data did not show significant influences of age and gender on peer aggression. Gender and age differences in children’s aggressive acts have been found in observational research on Han children in contemporary China (Jankowiak, Joiner, and Khatib 2011) as well as other cultures, such as those studied in SCS (Whiting and Edwards 1973) and its extension (Whiting and Edwards 1988). According to Margery Wolf (1978), gender and age were important factors in parental discipline in this community: local people saw ages 6 and 7 (about the median age of this Child Interview sample), when parental discipline becomes stricter, as a time of abrupt change, with parents clearly favoring boys over girls. APW’s preliminary analysis of Child Observation data presented mixed findings: the average rate of younger Taiwanese children’s retaliatory aggression was similar to the pan-cultural rate in SCS (Lambert and Tan 1979), but the average rate was higher in older Taiwanese children. Also, girls were as likely as boys to respond to aggression in kind. The prevalence of revenge in physical conflicts among children might explain why the impact of gender and age on Child Interview responses is relatively trivial. Further examination of Child Observation is needed to illuminate the impact of age and gender on actual peer aggression interactions.
Child Interview data also present a paradox about children’s aggression in Lower Xizhou, namely, the prevalence of children’s fights despite the parental ideology demanding that children avoid conflict (Duryea 1999, 105). This puzzle invites us to carefully analyze how, and to what extent, parental discipline shapes child development. For one thing, APW found a positive correlation between the “harshness” of maternal discipline in Mother Interview data collected at the beginning of fieldwork—a standardized measure adopted from SCS—and children’s adrenaline level, as determined by urine samples collected at the end of fieldwork (A. Wolf 1964; Duryea 1999, 139). This intriguing finding, defying SCS’s behaviorist, reward-punishment learning theory, makes it imperative to further investigate the relationships between Child Interview and Mother Interview, and compare these interview data in the Wolf Archive with Mother Interview data in SCS (Minturn and Lambert 1964, 162) to illuminate cross-cultural similarity and variability in maternal attitudes toward children’s aggression. Moreover, it is important to look beyond direct parental discipline, because in addition to well-known socialization strategies in Taiwan, such as shaming and punishment by parents (Fung 1999), social life can shape child development through multiple pathways, such as instigation by peers’ aggressive acts (Hall and Cairns 1984), teaching from other adults (Stafford 1995), and observational learning/modeling, a learning strategy that has assumed a central place in traditional Chinese socialization repertoire (Munro 1975). For example, Margery Wolf reported, “On several occasions I have heard a three or four-year-old imperiously warn his mother to stop interfering with his (usually dangerous) activity lest he summon his father to beat her” (1978, 226).
Child Interview is only the tip of the iceberg. Looking ahead, I plan to combine Child Interview with Child Observation, Mother Interview, and other types of data in the Wolf Archive to examine these children’s experiences across different relationship contexts, such as peer, sibling, and adult-child relationships. This comprehensive approach can shed invaluable light on the various ways by which children develop reciprocal aggression, with its different types and meanings in this community. Connecting Wolf’s Child Interview to SCS and the anthropology of childhood literature on these learning mechanisms will help situate the study of Taiwanese childhood in a global, comparative framework. Positioned at the rare intersection of Sinological anthropology, the anthropology of children, and historical studies of childhood, reanalyzing the Wolf Archive offers a unique contribution to understanding moral development across time and space, within and beyond Taiwan.
Ariès, Philippe. 1965. Centuries of Childhood: A Social History of Family Life. Oxford: Vintage Books.
Bai, Limin. 2005. Shaping the Ideal Child: Children and Their Primers in Late Imperial China. Hong Kong: Chinese University Press.
Bakken, Børge. 2000. The Exemplary Society: Human Improvement, Social Control, and the Dangers of Modernity. Oxford: Oxford University Press.
Blum, Susan D. 2019. “Why Don’t Anthropologists Care about Learning (or Education or School)? An Immodest Proposal for an Integrative Anthropology of Learning Whose Time Has Finally Come.” American Anthropologist 121 (3): 641–654. https://doi.org/10.1111/aman.13268.
Boyer, Pascal. 1998. “Cognitive Tracks of Cultural Inheritance: How Evolved Intuitive Ontology Governs Cultural Transmission.” American Anthropologist 100 (4): 876–889.
Brown, Paula, and Ilsa Schuster. 1986. “Introduction: Culture and Aggression.” Anthropological Quarterly 59 (4): 155–159, 200–204.
Chan, Anita. 1985. Children of Mao: Personality Development and Political Activism in the Red Guard Generation. London: Macmillan.
Chang, Hui-Ching. 2001. “Harmony as Performance: The Turbulence under Chinese Interpersonal Communication.” Discourse Studies 3 (2): 155–179.
Curry, Oliver Scott, Daniel Austin Mullins, and Harvey Whitehouse. 2019. “Is It Good to Cooperate? Testing the Theory of Morality-as-Cooperation in 60 Societies.” Current Anthropology 60 (1): 47–69. https://doi.org/10.1086/701478.
Duryea, Maria. 1999. “Changing Lives and Life Changes on Taipei’s Urban Border 1959–1994.” PhD diss., University of Washington.
Fernsebner, Susan R. 2003. “A People’s Playthings: Toys, Childhood, and Chinese Identity, 1909–1933.” Postcolonial Studies 6 (3): 269–293.
Freedman, Maurice. 1968. “Foreword.” In The House of Lim: A Study of a Chinese Farm Family, by Margery Wolf, vii–xii. Upper Saddle River, NJ: Prentice Hall.
Fung, Heidi. 1999. “Becoming a Moral Child: The Socialization of Shame among Young Chinese Children.” Ethos 27 (2): 180–209.
Hall, W. M., and R. B. Cairns. 1984. “Aggressive Behavior in Children: An Outcome of modeling or Social Reciprocity?” Developmental Psychology 20 (5): 739–745.
Harrell, Stevan. 1982. Ploughshare Village: Culture and Context in Taiwan. Seattle: University of Washington Press.
———. 1999. “Lessons from the Golden Age of ‘China Ethnography.’” In Anthropological Studies in Taiwan: Retrospect and Prospect, edited by Hsu Cheng-kuang and Huang Ying-Kuei, 211–240. Taipei: Academia Sinica.
Hirschfeld, Lawrence A. 2002. Why Don’t Anthropologists Like Children? American Anthropologist 104 (2): 611–627.
Hsiung, Ping-chen. 2005. A Tender Voyage: Children and Childhood in Late Imperial China. Stanford, CA: Stanford University Press.
Jankowiak, William, Amber Joiner, and Cynthia Khatib. 2011. “What Observation Studies Can Tell Us about Single Child Play Patterns, Gender, and Changes in Chinese Society.” Cross-Cultural Research 45 (2): 155–157.
Jones, Andrew F. 2011. Developmental Fairy Tales: Evolutionary Thinking and Modern Chinese Culture. Cambridge, MA: Harvard University Press.
Kinney, Anne Behnke, ed. 1995. Chinese Views of Childhood. Honolulu: University of Hawai‘i Press.
———. 2004. Representations of Childhood and Youth in Early China. Stanford, CA: Stanford University Press.
Kipnis, Andrew B. 2011. Governing Educational Desire: Culture, Politics, and Schooling in China. Chicago, IL: University of Chicago Press.
Klöter, Henning. 2004. “Language Policy in the KMT and DPP Eras.” China Perspectives [online], 56. https://journals.openedition.org/chinaperspectives/442.
Knapp Keith N. 2017. “There Are Maggots in My Soup! Medieval Accounts of Unfilial Children.” In Behaving Badly in Early and Medieval China, edited by N. Harry Rothschild and Leslie V. Wallace, 19–38. Honolulu: University of Hawai‘i Press.
Lambert, William, and Allen L. Tan. 1979. “Expressive Styles and Strategies in the Aggressive Actions of Children of Six Cultures.” Ethos 7 (1): 19–36.
Lan, Pei-chia. 2018. Raising Global Families: Parenting, Immigration, and Class in Taiwan and the U.S. Stanford, CA: Stanford University Press.
LeVine, Robert A. 2010. “The Six Cultures Study: Prologue to a History of a Landmark Project.” Journal of Cross-Cultural Psychology 41 (4): 513–521. https://doi:10.1177/0022022110362567.
Li, Chenyang. 2006. “The Confucian Ideal of Harmony.” Philosophy East and West 56 (4): 583–603.
Lupher, Mark. 1995. “Revolutionary Little Red Devils: The Social Psychology of Rebel Youth, 1966–1967.” In Chinese Views of Childhood, edited by Anne Behnke Kinney, 321–344. Honolulu: University of Hawai‘i Press.
Mather, Richard B. 1995. “Filial Paragons and Spoiled Brats: A Glimpse of Medieval Chinese Children in the Shishuo xinyu.” In Chinese Views of Childhood, edited by Anne Behnke Kinney, 111–128. Honolulu: University of Hawai‘i Press.
Miller, George A. 2003. “The Cognitive Revolution: A Historical Perspective.” Trends in Cognitive Sciences 7 (3): 141–144.
Minturn, L., and W. W. Lambert. 1964. Mothers of Six Cultures. New York: Wiley.
Munro, Donald J. 1975. “The Chinese View of Modeling.” Human Development 18 (5): 333–352. https://doi.org/10.1159/000271495.
Narotzky, Susana, and Paz Moreno. 2002. “Reciprocity’s Dark Side: Negative Reciprocity, Morality and Social Reproduction.” Anthropological Theory 2 (3): 281–305.
Olson, Kristina, and Elizabeth Spelke. 2008. “Foundations of Cooperation in Young Children.” Cognition 108 (1): 222–231.
Pollock, Linda A. 1983. Forgotten Children: Parent-Child Relations from 1500 to 1900. Cambridge: Cambridge University Press.
Potter, Sulamith Heins. 1988. “The Cultural Construction of Emotion in Rural Chinese Social Life.” Ethos 16 (2): 181–208.
Saari, Jon L. 1990. Legacies of Childhood: Growing up Chinese in a Time of Crisis, 1890–1920. Cambridge, MA: Council on East Asian Studies, Harvard University.
Shih, Fang-long, Stuart Thompson, and Paul-François Tremlett. 2008. Re-writing Culture in Taiwan. London: Routledge.
Soloman, Richard H. 1971. Mao’s Revolution and the Chinese Political Culture. Berkeley: University of California Press.
Stafford, Charles. 1992. “Good Sons and Virtuous Mothers: Kinship and Chinese Nationalism in Taiwan.” Man 27 (2): 363–378.
———. 1995. The Roads of Chinese Childhood: Learning and Identification in Angang. Cambridge: Cambridge University Press.
Tillman, Margaret Mih. 2018. Raising China’s Revolutionaries: Modernizing Childhood for Cosmopolitan Nationalists and Liberated Comrades, 1920s–1950s. New York: University of Columbia Press.
Tomasello, Michael. 2019. Becoming Human: A Theory of Ontogeny. Cambridge, MA: Belknap Press.
Topley, Majorie. 1974. “Cosmic Antagonisms: A Mother-Child Syndrome.” In Religion and Ritual in Chinese Society, edited by Arthur P. Wolf, 233–244. Stanford, CA: Stanford University Press.
Unger, Jonathan. 1982. Education under Mao: Class and Competition in Canton Schools, 1960–1980. New York: Columbia University Press.
Vaish, Amrisha, Robert Hepach, and Michael Tomasello. 2018. “The Specificity of Reciprocity: Young Children Reciprocate More Generously to Those Who Intentionally Benefit Them.” Journal of Experimental Child Psychology 167: 336–353.
Watson, James L. 1976. “Anthropological Analyses of Chinese Religion.” Review of Religion and Ritual in Chinese Society, edited by Arthur P. Wolf. China Quarterly 66: 355–364.
Weisner, Thomas S. 2010. “John and Beatrice Whiting’s Contributions to the Cross-Cultural Study of Human Development: Their Values, Goals, Norms, and Practices.” Journal of Cross-Cultural Psychology 4 (41): 499–509.
Wertsch, James V. 1994. “The Primacy of Mediated Action in Sociocultural Studies.” Mind, Culture, and Activity 1 (4): 202–208.
Whiting, Beatrice B., ed. 1963. Six Cultures: Studies of Child Rearing. Oxford: Wiley.
Whiting, Beatrice Blyth, and Carolyn Pope Edwards. 1973. “A Cross-Cultural Analysis of Sex Differences in the Behavior of Children Aged 3–11.” Journal of Social Psychology 91 (2): 171–188.
———. 1988. Children of Different Worlds: The Formation of Social Behavior. Cambridge, MA: Harvard University Press.
Whiting, Beatrice Blyth, John W. M. Whiting, and Richard Longabaugh. 1975. Children of Six Cultures: A Psycho-Cultural Analysis. Cambridge, MA: Harvard University Press.
Whiting, John W. M. 1966. Field Guide for a Study of Socialization. Oxford: Wiley.
Whiting, John W. M., B. C. Ayres, H. Geertz, G. W. Goethals, C. Holzinger, E. Lowell, and E. E. Maccoby. 1953. Field Manual for the Cross-Cultural Study of Child Rearing. New York: Social Science Research Council.
Whiting, John W. M., and Beatrice B. Whiting. 1973. “Altruistic and Egoistic Behavior in Six Cultures.” In Cultural Illness and Health: Essays in Human Adaptation, edited by Laura Nader and Thomast W. Maretzki, 56–68. Arlington, VA: American Anthropological Association.
Wilson, Richard W. 1974. The Moral State: A Study of the Political Socialization of Chinese and American Children. New York: Free Press.
Wolf, Arthur, P. 1964. “Aggression in a Hokkien Village: A Preliminary Description.” Unpublished mimeograph presented at the Seminar on Personality and Motivation in Chinese Society, Bermuda, January 26–28.
———. 1966. “Childhood Association, Sexual Attraction, and the Incest Taboo: A Chinese Case.” American Anthropologist 68 (4): 883–898.
———. 1968. “Adopt a Daughter-in-Law, Marry a Sister: A Chinese Solution to the Problem of the Incest Taboo.” American Anthropologist 70 (5): 864–874.
———. 1970. “Childhood Association and Sexual Attraction: A Further Test of the Westermarck Hypothesis.” American Anthropologist 72 (3): 503–515.
———. 1982. “Project Summary for National Science Foundation: Chinese Children and Their Mothers: Cultural and Psychological Aspects of Socialization in a Taiwanese Village.” Unpublished document, A. P. Wolf’s private library.
———. 1995. Sexual Attraction and Childhood Association: A Chinese Brief for Edward Westermarck. Stanford, CA: Stanford University Press.
———. 2015. “Chinese Children and Their Mothers.” Unpublished book manuscript , A. P. Wolf’s private library.
Wolf, Arthur P., and Chieh-Shang Huang. 1980. Marriage and Adoption in China, 1845–1945. Stanford, CA: Stanford University Press.
Wolf, Margery. 1968. The House of Lim: A Study of a Chinese Family. Upper Saddle River, NJ: Prentice Hall.
———. 1972. Women and the Family in Rural Taiwan. Stanford, CA: Stanford University Press.
———. 1978. “Child Training and the Chinese Family.” In Studies in Chinese Society, edited by Arthur P. Wolf, 221–246. Stanford, CA: Stanford University Press.
———. 1990. “The Woman Who Didn’t Become a Shaman.” American Ethnologist 17 (3): 419–430.
———. 1992. A Thrice-Told Tale: Feminism, Postmodernism, and Ethnographic Responsibility. Stanford, CA: Stanford University Press.
Xu, Jing. 2014. “Becoming a Moral Child amidst China’s Moral Crisis: Preschool Discourse and Practices of Sharing in Shanghai.” Ethos 42 (2): 222–242.
———. 2017. The Good Child: Moral Development in a Chinese Preschool. Stanford, CA: Stanford University Press.
———. 2019. “Learning ‘Merit’ in a Chinese Preschool: Bringing the Anthropological Perspective to Understanding Moral Development.” American Anthropologist 121 (3): 655–666.
Yamagishi, Toshio, Yutaka Horita, Nobuhiro Mifune, Hirofumi Hashimoto, Yang Li, Mizuho Shinada, Arisa Miura, Keigo Inukai, Haruto Takagishi, and Dora Simunovic. 2012. “Rejection of Unfair Offers in the Ultimatum Game Is No Evidence of Strong Reciprocity.” PNAS 109 (50): 20,364–20,368.
About the Author
Jing Xu is Affiliate Assistant Professor of Anthropology at the University of Washington. The author would like to thank the two anonymous Cross-Currents reviewers and managing editor Keila Diehl for the time and effort they put into her manuscript. Their feedback greatly helped improve her work. She is grateful to Hill Gates for generously hosting her at the Wolf Ranch through the Rockpile Foundation, and for providing guidance in this project. The author also thanks Stevan Harrell for introducing her to this unique project and for his support in the writing process. The author appreciates the insightful feedback on an earlier draft of this paper from participants in the Global Taiwan Workshop at the University of Washington, 2018, and from John Shepherd. Lastly, she is thankful to the organizers of this special Cross-Currents issue: Peter Thilly, Graeme Reed, and James Lin.
 With the exception of direct quotes (e.g., “Ch’i-Chou” in APW’s writings), I use pinyin without hyphenation (e.g., “Xizhou”) to transliterate Chinese characters.
 “Families” here refers to extended families.
 See James L. Watson’s explanation of this term: “‘Sinological anthropology’ is a term of convenience; it is generally used to designate all anthropologists who work in the field of Chinese studies” (1976, 355).
 For a list of research and publications about the Haishan area, see Wolf and Huang (1981, appendix B).
 At the time, the term Taiwan studies was synonymous with China studies.
 By “China,” Wolf meant the Chinese cultural region.
 Only a small part of the Wolf Archive—e.g., some data from Mother Interview and General Observations (two types of methods used in the fieldwork)—was ever used in the published works of APW and Margery Wolf.
 The six groups studied were the Nyansongo, a Gusii community in Kenya; the Rajputs of Khalapur, India; Taira, a village in Okinawa; the Mixtecans of Juxtlahuaca, Mexico; the Tarong in the Philippines; and New Englanders in Orchard Town in the United States (all pseudonyms).
 APW selected sixty-four children as his primary target sample according to SCS guidelines, but when conducting Child Interview in the beginning phase of his fieldwork, he enlarged the scope and interviewed more children in the village; therefore, the sample number for Child Interview is seventy-nine.
 Some studies have examined texts produced and circulated among rebellious youth during China’s Cultural Revolution—“big posters” (dazibao)—but these texts are highly formulaic and served specific political purposes (Lupher 1995), giving them a very different nature than the Child Interview responses analyzed in this article. Also, as high school students, these rebellious youth are older than APW’s research subjects.
 These “outliers” are excluded from statistical tests but included in qualitative analyses.
 The total number of answers is exceptionally low under prompt D due to the nature of this question. Whether or not a child is presented with prompt D is contingent on his/her responses to previous prompts.
 By contrast, in a nearby village in the 1970s, Harrell observed that conflicts between children or women were seen as not important, with no substantial consequences for interfamily relations (1982, 138–143).
 Part of the incident titled “The woman who did not become a shaman” (M. Wolf 1990, 1992) is a mother’s estranged relationship with the Lim household because of a fight between children. Maria Duryea’s dissertation (1999) also documents several cases of children’s fights from the Wolf Archive’s general observation notes.
 From a field note dated March 14, 1959, regarding an interview with psychiatrist Dr. Tsung-yi Lin, who agreed with a local research assistant’s observation on this matter.
 According to Stevan Harrell (personal communication), APW talked in his Stanford classroom in the 1970s about the discrepancy that when children filled out the school questionnaire, they checked the box for “Do nothing” while mumbling “Hit him,” “Kill him,” or a similar response under their breath.
 One of the two research assistants, in particular, “had become everyone’s confidante, everyone’s friend” in the village (M. Wolf 1990, 422), and she was accorded a central role in Margery Wolf’s ethnographic accounts of an important village incident (1990, 1992).
 This definition of negative reciprocity conforms to the convention in research on child development.
 This is not to say that Lower Xizhou children’s narratives speak only to the darker side. In fact, Child Interview contained many questions on children’s prosocial development, such as helping others in need.
 A draft of APW’s 1982 proposal to the National Science Foundation documents this finding (A. Wolf 1982, 5).
 The “other child” in Child Interview meant a child from another family. This interpretation can be discerned from the interview questions as well as the children’s answers.