Siblings Matter

On a spring morning, MC observed a brother–sister dyad interacting with another child:Footnote ¹

This episode, combining elements of kinship, friendship, and gender, provides a glimpse into young siblings’ moral life in peer context: The young boy Yi-kun “bribed” his peer Yan-yan with the explicit goal of getting Yan-yan to accept his little sister Mei-yu as her playmate. Yan-yan accepted Yi-kun’s gift – a popular play object among young girls in the community, as well as his request. Perhaps benefitting from her seniority in age or her position as the desired playmate, Yan-yan also took the initiative to lead the subsequent interactions, commanding the control of resources upon Mei-yu’s agreement. Above all, we see a little boy helping his younger sister to make friends with peers, in a social world outside the home, through exchange of gifts and favors.

Sibling relations provide a unique lens through which to examine core questions of learning morality. How do young children learn their first lessons with their siblings, about relating to others and asserting oneself, about cooperation and conflict, and about negotiating parental control and love, within and beyond familial contexts? While many in other species never know or meet their siblings, humans have almost always known and lived with their siblings throughout our evolutionary history. Across cultures, instead of having adults supervise and teach them, children spend a lot of time playing with and learning from other children, and siblings are important agents in such peer learning processes (Maynard and Tovote Reference Maynard, Tovote, Lancy, Bock and Gaskins2010). Cross-cultural research on child development has long demonstrated that siblings always matter (Weisner Reference Weisner and Zukow1989).

A special relationship that cuts across the peer world and the family world, sibling relation can shed invaluable light on family dynamics. Despite the fact that Chinese children also spend a lot of time with their siblings, the anthropology of the Chinese family since its inception has obscured the importance of sibling relation. Classic works, shaped by British social anthropology’s lineage studies traditions,Footnote ² rarely focused on sibling relations in childhood, only analyzing brother-brother relations as adults.Footnote ³ Later research, inspired by the new anthropology of kinship that emphasizes the lived experience of “relatedness” (Carsten Reference Carsten2000) instead of formal structures, still paid little attention to sibling experience.Footnote ⁴ Neglecting sibling relation reflects a broader problem in the anthropology of kinship, that is, the bias toward the centrality of parent–child and conjugal relations (Alber, Coe, and Thelen Reference Alber, Coe and Thelen2013).Footnote ⁵

Studying sibling relations also honors and expands from Arthur Wolf’s legacy in biological and evolutionary anthropology. From a bio-evolutionary perspective, sibling experience in childhood is a key component in human kin detection – detecting genetic relatedness (Lieberman, Tooby, and Cosmides Reference Lieberman, Tooby and Cosmides2007). Even today, Arthur Wolf’s research on Taiwanese sim-pu-a (“little daughter-in-law”) (A. Wolf Reference Wolf1970) and communally-reared children in the Israeli Kibbutzim (Shepher Reference Shepher1971) remain two classic cases of “natural experiments” that are often taken to have wider implications for human kinship in general. Drawing from demographic records in the Haishan area, Arthur Wolf found lower fertility rates and higher divorce rates of adult couples who grew up as stepsiblings compared to those who did not. As I mentioned in Chapter 4, he has written extensively to support the Westermarck hypothesis that intimate association in childhood leads to incest avoidance through sexual disgust.Footnote ⁶ Along these lines, later research based on laboratory experiments further established sibling co-residence experience as a key parameter for kin detection and moral sentiments related to incest (Lieberman and Lobel Reference Lieberman and Lobel2012; Lieberman, Tooby, and Cosmides Reference Lieberman, Tooby and Cosmides2003).

But Wolf’s previous research did not explicitly address the linkage between sibling relation and kin-based morality. Sibling-directed morality as altruism, formulated in kin-selection or inclusive fitness theory (Hamilton Reference Hamilton1964), is an important prediction of human behavioral ecology.Footnote ⁷ Help, care and collaboration among siblings are a major focus of this chapter. I will also delve into the dark side of sibling morality. Even though natural selection should lead to kin-based altruism at a very general level and as what’s called “an ultimate explanation” (Scott-Phillips, Dickins, and West Reference Scott-Phillips, Dickins and West2011), in everyday reality kinship is not always “warm and fuzzy.” While classic theories on kinship in sociocultural anthropology tend to focus on dimensions of solidarity, many recent reflections have emphasized the negative sentiments, or ambivalence of kinship (see e.g., Bamford Reference Bamford and Bamford2019; Cartsen 2013).Footnote ⁸ Sibling relations, especially cross-sex siblings, are constructed simultaneously as equal (shared parenthood) and as different (age, gender, birth order) (Thelen, Coe, and Alber Reference Thelen, Coe, Alber, Alber, Coe and Thelen2013). Exploring sibling relations in childhood can help us understand the “double-edged quality of human kinship” (Carsten Reference Carsten2013): love and control, connection and exclusion in everyday practice of relatedness.

I rediscovered the once untold tale of a brother–sister dyad, two main characters in the opening vignette. Their family was featured in Margery Wolf’s A Thrice-Told Tale (1992), but what happened to the children remained a mystery. This intrigued me. I recovered these children’s voices in their family drama. I traced their social network and behavioral patterns. I explored how they looked out for each other and united against their peers but competed and maneuvered at home. Finally, reconstructing the story of these two children also led me to ponder the nature of fieldnotes and ethnography.

A “Crazy” Mother and Her Children

The family of Wang Yi-kun and Wang Mei-yu was featured in Margery Wolf’s book A Thrice-Told Tale: Feminism, Postmodernism, and Ethnographic Responsibility (1992), including its shorter, article version (M. Wolf Reference Wolf1990b). This family “had lived in the village for nearly ten years, but by village tradition they were still newcomers.”Footnote ⁹ In a community bonded through kinship, although sharing the same surname, they didn’t belong to the village’s prominent Wang lineage (the Wolfs’ landlord, the Lims of The House of Lim).Footnote ¹⁰ At the beginning of the Wolfs’ fieldwork, this family (see Table 5.1) included father Wang Tian-lai (#47), mother Wang Chen-hsin (#48), and three children: Wang Yi-kun (#49), Wang Mei-yu (#50), and an infant boy Wang Ju-kun (#51).

The protagonist of A Thrice-Told Tale, Wang Chen-hsin, a woman who “went crazy,” was originally from Keelung, and her mother still lived there. Her younger sister Wang Hsiu-chu (#52) somehow lived nearby. Her sister and her mother were both around to look after her, during her mental breakdown, an incident featured in “A Thrice-Told Tale:”

This incident was related to a fight among children. Wang Chen-hsin “had quarreled in recent months with a woman from a Lim household when her young son had been slugged by a Lim boy.”Footnote ¹² The “Lim” household in this quote was actually another Wang household (#15b). From multiple women’s testimony during village gossip, MC learned that when Chen-hsin went to the other mom Wang Lin-hua (Household #15b) to report that fight, Lin-hua scolded her harshly: “If children fight and kill each other it serves them right. If your children get killed, then you come and take your children home and bury them. You don’t need to come and talk to me about it.”Footnote ¹³ During the days of Chen-hsin’s mental breakdown, when she acted like a Tang-ki (spirit medium), confusing many people, Lin-hua’s household did not go to see Chen-hsin as other neighbors did, and Chen-hsin tormented that household.Footnote ¹⁴ Another woman, according to Chen-hsin’s mother, had cursed Chen-hsin due to children’s fights: “You are going to go crazy and take off your clothes.” Village women also told MC that when Chen-hsin fell into the field, she lay there lamenting: “Just because of children’s things other people bully me, other people bully me just because of children’s things. I won’t forget this! I won’t forget this!”Footnote ¹⁵

When I reapproached A Thrice-Told Tale, I looked everywhere to find clues about the role and experience of Chen-hsin’s children. I gleaned sporadic information about her children in the raw fieldnotes section, the second tale of that book.Footnote ¹⁶ It turns out that Chen-hsin’s older boy Wang Yi-kun played a unique part in the development of that story, although Margery Wolf did not write about this child in her final, ethnographic tale. First of all, the boy knew his mom’s psychology quite well and seemed more perceptive than his father Tian-lai. Had his father been as insightful as he was, they might have successfully prevented Chen-hsin from “going crazy” this time.Footnote ¹⁷ On top of a grudge against other mothers accumulated recently, the immediate trigger of Chen-hsin’s erratic behavior, on March 4, 1960, was her losing NT$ 90 and believing that a neighbor had stolen it. When she couldn’t find the money and asked Yi-Kun about it, he told mom that his father took it to gamble. But his father denied it. She then plunged into the rice field. Neighbors criticized the husband who, they said, was aware of his wife’s illness, because he should not have let her worry. He should have said that he had taken the money even if he hadn’t, because she then would not worry about the missing money.Footnote ¹⁸

Yi-kun and his mother were very close. He was her confidant. After she fell ill, she once called her son into the house and lay on the bed with him. Yi-kun told his mother that a medicine his father gave her was made of worms, prescribed by local gods. She called her son in, both to confirm that matter and to protect him from being spanked by his angry father. Upon confirming with Yi-kun, she ran out of the house and her children looked scared.Footnote ¹⁹

As hard as it could be, her children, both Yi-kun and Mei-yu, were seen helping her when everyone else present was frightened by her possession-like behavior. On March 10, 1960, Chen-hsin drew a crowd in her courtyard, dancing like a Tang-ki, making bai-bai motions. She summoned MC, hugged her close and praised her for being kind to all the children, but the chaotic scene frightened MC and children in the crowd. Her own children, however, did not see scared. They rushed to pick her up each time she fell down.Footnote ²⁰

Finally, her mental breakdown strained her marital relation, and her children were inevitably involved in. When Chen-hsin found out that her husband was going to send her to the hospital again, as he had when she stumbled into the rice paddy, she kept saying that her husband was too dumb and that she wanted a divorce: “I am a woman and you are a man, so I’ll take the one girl (Mei-yu) and you take the boys (Yi-kun and his baby brother). I won’t have to take your boys [the heirs of the husband’s family]. I’m going to take the girl home [to Keelung] with me. I am going to take my girl home with me.” A neighbor said she did bring her daughter with her when she tried to run away from the house.Footnote ²¹

Children’s voices were mostly omitted when Margery Wolf transformed the raw fieldnotes into her ethnographic piece, “The Woman Who Didn’t Become a Shaman” (M. Wolf Reference Wolf1990b, Reference Wolf1992: 93–126). But children’s voices are precious in this ethnography: What were Chen-hsin’s children like? How did these little ones cope with their life, as part of a marginalized family going through a scandalous crisis? One episode in Child Observation (CO)Footnote ²² clearly shows that children were active participants in village gossip. The next day after Chen-hsin made a scene with MC, children from other families, while playing a hopscotch game, teased their “Older Sister” observer.

Two days later, on the way to school, MC overheard an eight-year-old girl, Chen Yu-li, telling other children this story:Footnote ²³

Anthropologists rely on gossip to gather data, for example, MC checking with villagers to understand what happened to Chen-hsin. Children also gossiped among themselves, an important way to establish their own reputation and spread information about other people’s reputation. One can only imagine the kind of teasing, mocking or social exclusion Chen-hsin’s own children would have experienced. One can further speculate that Chen-hsin’s children might have to rely on themselves and support each other at this difficult time. Because Chen-hsin was very shy, Wolf’s team did not conduct concentrated observations or interviews with her as they did with some other mothers. Fortunately, her two older children Yi-kun and Mei-yu were quite visible in the archive.Footnote ²⁴ CO includes sixty-four timed episodes that involve at least one of the two children, of which seventeen episodes involve both children. Besides, eight episodes of other (un-timed) observations involve one child or both; in terms of projective tests, both children participated in TAT, and the sister also participated in Doll Play. Drawing from the core data, CO, the next section explores these two children’s relative positioning in the context of their social networks.

Siblings in Peer Network

Social network analysis of CO reveals patterns of how Yi-kun and his sister Mei-yu interacted with each other and with other children and adults. We have already seen in Chapter 3 that household number is the most important demographic variable for predicting network homophily, more so than gender and age. This pattern holds true in both co-occurrence network, meaning that children from the same household tend to appear together during CO episodes and behavioral-interaction network, meaning that they tend to actually interact together too. Children from the same household are mostly siblings, and some cases include cousins. It is thus clear that sibling relations are significant in children’s overall social network structure.

With regards to the dyad of Yi-kun (#49) and Mei-yu (#50), they two belonged to the same clique out of the four cliques mentioned in Chapter 3, suggesting that Yi-kun and Mei-yu often appeared together in play groups. Figure 5.1 depicts a subset of Child Observation co-occurrence network, the union network of Yi-kun and Mei-yu, a regional network with these two children as the focal nodes. Each node in this union network represents a person. The size of a node is proportional to the person’s betweenness-centrality, which measures how often a node occurs on all shortest paths between two nodes and represents the person’s importance as a “bridge” between other people in the network. Each edge (line) between two nodes represents the co-occurrence of these two people; the thickness of edges is proportional to the frequency of co-occurrence observed.

Figure 5.1 Child observation subset (#49 and #50) union network. The number on each node represents the person’s ID. White nodes represent the two focal children (#49 and #50), dark gray nodes represent other children and light gray nodes represent adults; Square nodes represent females and round nodes represent males

The union network of Yi-kun and Mei-yu, which includes all the nodes and edges connected to Yi-kun or Mei-yu, displays the following features. First of all, similar to the overall co-occurrence network configuration, children occupied a central position and adults were at the periphery. Moreover, a closer look at this brother–sister dyad’s network attributes (Table 5.2), in comparison to the respective maximum numbers observed in the child network, reveals these two children’s different social statuses in their peer world: Brother co-occurred with seventy-nine other children in the network (measured by “degree,” the number of other nodes he is connected to), many repeatedly (total co-occurrence measured by “weighted degree”), and close to the maximum values observed. This means that Brother was a popular figure in the peer network. But in terms of between-ness centrality, brother was far from being the central hub of connections. Sister ranked much lower in all three attributes compared to her brother, suggesting that she was probably a less important figure in the village play groups.

Departing from simple co-occurrence patterns, social network analysis based on behavioral interactions provides a finer-grained look brother–sister similarities and differences in social life. Figure 5.2 (a & b) juxtaposes brother’s and sister’s “ego networks,” in which a node represents a person, an edge represents behavioral interactions between two nodes, and the arrow of an edge represents the direction of behavior between the two connected nodes (from the initiator to the recipient). In Figure 5.2, each ego network depicts the focal node (“ego”)’s “neighborhood,” the collection of ego and all nodes (“alters”) with whom ego has direct interactions. Besides representing ego-alter ties, the “neighborhood” also includes all of the ties among the “alters” (Hanneman and Riddle Reference Hanneman and Riddle2005).

Figure 5.2 (a) #49’s ego network of behavioral interactions, (b) #50’s ego network of behavioral interactions. The number on each node represents the person’s ID. White nodes represent the two focal children (#49 in a and #50 in b), dark gray nodes represent other children and light gray nodes represent adults; Square nodes represent females and round nodes represent males. In these behavioral-networks, the size of a node is proportional to the person’s weighted degree centrality, measured by the sum of frequency of behavioral interactions this person has with all other people. The thickness and darkness of an edge is proportional to its weight, meaning the frequency of behavioral interactions between the two nodes (people) connected by the edge

Comparing these two ego networks reveals the following patterns: (1) These two children have overlapping peer networks and they two often interacted together. But Brother seemed to be a more important character in Sister’s circle than Sister was to Brother: In terms of interaction frequency,Footnote ²⁵ while Sister was not among the top five children that Brother interacted with, Brother was the third most important child Sister interacted with. (2) Brother interacted with a wider circle of children than Sister did, measured by the number of alters in each ego network. (3) Brother was the center of his ego network, ranked #1 in weighted degrees, while Sister was not, ranked only #7, next to Brother, in her own ego network. Taken together, Brother was an active figure in the village peer network, despite his mother and family’s marginal status, but Sister was a more peripheral figure in the observations and was often observed overlapping with her brother in play. Perhaps Sister was used to trailing her brother, as in MC’s observation in the opening vignette of this chapter? To understand what contributed to these patterns – age, gender and/or personality, let’s look at the content of observations through NLP (natural language processing) and granular-level behavioral analysis.

Play, Cooperation, and Conflict

I used NLP techniques to analyze linguistic patterns of those CO episodes,Footnote ²⁶ where Brother and/or Sister appeared, a subset of the cleaned and preprocessed CO corpus. Figures 5.3 and 5.4 depict high-frequency words in this subset of observations: Figure 5.3 presents the top 100 high-frequency words (in wordcloud format) and Figure 5.4 presents the top 50 high-frequency words (as a bar graph).

Figure 5.3 Top 100 high-frequency words in the Child Observation subset (#49 and #50)

Note: Analyses in this figure and Figure 5.4 were performed and visualized in the Python programming language.

Figure 5.4 Top 50 high-frequency words in Child Observation subset (#49 and #50)

Based on these, Figure 5.5 depicts the co-occurrence network of top twenty-two words: Each node represents one high-frequency word, and each edge represents co-occurrence of the two nodes/words. Node size is proportional to the word’s frequency and edge thickness is proportional to degree of co-occurrence between the two words. These analyses identify several notable patterns: The words “say” and “play” ranked #1 and #5, indicating the abundance of verbal interactions and the overall prevalence of play in their interactional contexts; The word “hit” ranked #21, possibly related to children’s fighting and/or pretend play; The words “laugh” and “yell” ranked #20 and #34 respectively, giving us a clue of the emotional texture of children’s play; Finally, co-occurring word-pairs with the highest-frequency rankings, such as hit-smile, hit-laugh, and hit-run, offer insights into how different behaviors intertwined in children’s naturalistic play.

Figure 5.5 Word co-occurrence network in the Child Observation subset (#49 and #50)

Note: The word co-occurrence matrix in this figure was computed in the Python programming language, via word embeddings analysis, and then visualized in Gephi (layout: Force Atlas 2 algorithm).

I also analyzed this subset of CO using a standardized behavioral grading protocol and grouped peer interaction behavioral themes into two main categories, cooperation (positive) and conflict (negative). Table 5.3 summarizes the behavioral analysis results. First of all, in terms of absolute numbers, Brother not only appeared in a lot more observations than Sister did, as co-occurrence social network analysis revealed. He was also involved in many more behavioral interactions than Sister was, as the initiator or recipient, and he interacted with more children than Sister did. Moreover, regarding percentage of positive behavior among all behaviors, Brother’s and Sister’s numbers are similar. Both initiated more cooperative behaviors than receiving cooperative behaviors. Was Brother a more popular person than Sister? Was Brother helping Sister in peer interaction, as the opening vignette portrayed? If both of them were ridiculed or bullied by peers, for what reasons then? To understand what these quantitative patterns mean and what motivated these behaviors, we need to explore the behaviors in ethnographic detail.

Personalities in Coalition

Close reading of naturalistic observation records can flesh out these two young characters and demonstrate the moral lessons these youngsters learned. Brother and Sister had distinct personalities. Brother played with many more children and was more active in their peer network. Sister was not popular and less savvy in navigating peer network. On many occasions the sibling dyad often acted in solidarity, like a small coalition facing the world outside their home. Together they learned to coordinate with peers, negotiating fairness and ownership, adapting to respect game rules and submitting to peer leaders.

Figure 5.6 A boy carrying his baby brother.

Source: Photo by Arthur Wolf

As behavioral grading results suggest, Brother Wang Yi-kun, although accused by his mother and other adults of being a troublemaker (fighting with other children),Footnote ²⁷ was more of a cooperator than an aggressor. Although sibling care was mainly girls’ responsibility, Yi-kun also helped take care of his little brother (#51). In one scenario his sister Mei-yu called the observer MS: “See? Big brother can take care of little brother.”Footnote ²⁸ Of course, this scenario alludes to her assumption, the local norm that girls, instead of boys, were supposed to take care of younger siblings.

Beyond sibling care, by the age of seven, Yi-kun had already mastered clever skills in negotiating rules of games and access to group play with other kids. He used the principle of reciprocity to his own advantage, reminding playmates of his previous favors to them. For example, when he saw an older girl holding sticks with feathers, he asked: “Give me one of yours, will you? You have two in your pocket.” Rejected by that girl, he insisted: “I, I, I did something for you.”Footnote ²⁹ He joined girls playing house and helped them with what they needed.Footnote ³⁰ Across several occasions Yi-kun was found controlling his anger and yielding to other children: One girl Chang Ah-yin (#172, a year and half older than he), a leader in their small group, bullied him, but he did not fight back. That’s why a bystander boy teased him: “A boy is afraid of a girl!” At another time, chased, cursed and hit by two naughty boys much younger than he, Yi-kun appeared scared. Those boys’ older brother laughed: “Ha ha, the older one is afraid of the younger one.”Footnote ³¹

The following observation, about Yi-kun’s unintended conflict with a slightly younger boy Wang Chen-jin (#508), is a vivid illustration of his personality: Chen-jin was quite a dominant character, bulky, strong and popular among boys. Initially they two were playing tag, and Yi-kun accidentally hit Chen-jin on the face. Chen-jin cried. Another older boy came by, scolded Yi-kun and pushed him down. Yi-kun did not respond or avenge. After crying for a couple minutes, Chen-jin walked over to Yi-kun and hit him four times. Yi-kun looked angry, but he still didn’t hit back or say anything. Chen-jin repeated that, hitting him three times. Yi-kun just held his emotions and walked away.Footnote ³²

Mei-yu, on the other hand, had a different personality, which helps to explain why her status in their peer network was more peripheral compared to that of her brother. She did not have many friends, even before the onset of her mother’s scandalous insanity. Unlike her brother, she was not easy to get along with. She interrupted other children’s play or bullied them, including her big brother. Her brother’s friends saw her as a troublemaker, reluctant to include her during their play. She did want to play with other children, though. Once two boys were playing a marble game, using little rocks marbles. She asked to join but was bluntly rejected by one of them. She turned to the other boy, her classmate, attempting to manipulate him: “He (your playmate) doesn’t want our Jung (zhong) class to play!”Footnote ³³ The boy who rejected her stayed firm: “Yes, I do. I just don’t want you to play.” During a jumping rope game, one boy explicitly asked the group leader not to let Mei-yu play: “She hit my little brother.” Mei-yu walked off, looking unhappy.Footnote ³⁴ She was not popular among girls either. She really wanted to play with Pai Yan-yan (#138), the girl who appeared in this chapter’s opening scenario. Yan-yan manipulated her: “Give it to me and I’ll be your friend. Otherwise I won’t play with you.” And Mei-yu gave in.Footnote ³⁵ Girl leader Ah-yin (#172) did not allow her to join their tree game, which is, to make “boats” out of bamboo leaves.Footnote ³⁶

One day, Mei-yu was excluded by her peers again. Three children were playing house. She walked up to watch, but a girl shouted at her (angrily with a threatening look): “Get out of here! I don’t want you to look.” She looked sad and walked away. But her brother Yi-kun happened to be there. He tattled, yelling to his mother, trying to protect his sister.Footnote ³⁷ Yi-kun appeared as a brother with a caring heart: In dyadic interactions, he yielded to Mei-yu many times even when she grew aggressive. In multi-person interactions, he used his social skills to help her. The opening vignette is a great example: He “bribed” a girl with face-powder so that she would agree to play with his little sister.

Sometimes help was mutual and cooperation was fruitful: Playing house with other children, they partnered in crime, taking away a small boy’s stones and teasing him.Footnote ³⁸ But seeing another child cheating a younger one when playing tops, Sister echoed her brother to mock and condemn the cheater.Footnote ³⁹ In hide-and-seek games, they two coordinated together to cheat, sister on the watch for brother who was the “ghost” trying to catch hiding children.Footnote ⁴⁰

But since neither of them was a leader in their peer group, negotiating with peers could be difficult. In the following observation, their negotiations failed.Footnote ⁴¹

Note that Yan-yan, the girl quite skilled at manipulating Mei-yu, was not persuaded by Yi-kun’s reasoning. Disagreement continued, until their playgroup leader Ah-yin (#172), a quite dominant girl, intervened and reinstated the importance of turn-taking rules. Brother attempted various methods to help his sister skip the line, improvising fictive parent–child kinship (apparently the real sibling tie was not enough) and asserting his ownership. But in the end, they had to submit to more “powerful” children, even though they were playing on his tricycle. In this quite serious pretend play scenario, we not only see children’s rich imaginative world – riding tricycles as cars and using leaves as tickets, but also how they learned to coordinate with peers and comply with rules.

Sibling Rivalry: Storytelling in Projective Tests

What about life inside their family? As I mentioned earlier, the Wolfs’ team didn’t manage to interview their mother or observe their family interactions at home, because their mother was very shy. But fortunately, these children participated in projective tests. Transcripts from Doll Play and TAT, based on verbal communication records, offer a precious window into children’s thoughts and emotions. In both tests, children were asked what they thought people in the scenarios were doing. Doll Play explicitly focused on family life, but some TAT drawings were also interpreted as interactions between family members. Narratives in these two projective tests shine light on children’s own perspectives about their family relations, especially sibling rivalry at home, which poses a contrast to sibling solidarity during peer interactions in the outside world.

In Mei-yu’s Doll Play test, a salient theme is brother dominating and bullying sister. At the first sight of dolls, she looked carefully, checked the dolls’ clothes, and started to tell this story spontaneouslyFootnote ⁴²:

Mei-yu was tested in two sessions. In both sessions, she spontaneously pivoted the conversation to the same topic: Brother is mean to sister, and he deserves to be punished. Unlike “brother eating sister’s candy” or “brother hitting sister” in the earlier discussed excerpt, in some other segments she didn’t even articulate a specific explanation. For example, she just repeated: “He (brother) is so bad.” This deviates from the imagery of a protective brother and the pattern of brother–sister solidarity that emerged in peer play outside the home setting.

One might wonder, though, if this had anything to do with Mei-yu’s real experience, or it was just purely imaginary. Yes, these utterances were a product of children’s storytelling exercise, most likely reflecting Mei-yu’s self-serving bias, that Brother was always the villain at home. But there are several reasons to doubt that stories she told were mere fantasies: In Mei-yu’s two sessions, the names she gave the dolls mapped exactly onto her family’s real situation. For example, she called the boy doll “big brother,” the girl doll “big sister” and the baby doll “little brother.”Footnote ⁴³ Moreover, the scenario of brother–sister fighting, especially over candy, appeared not only in Doll Play. It also surfaced in Mei-yu’s TAT session conducted several months earlier, and in surprisingly similar ways. Compare the following two excerpts: The first one was from her Doll Play transcript and the second from her TAT transcript.

Doll Play excerpt:

The following TAT excerptFootnote ⁴⁴ is especially intriguing. The research assistant Mr. Huang showed Mei-yu the fifth drawing in the set,Footnote ⁴⁵ which featured a girl, a broken bowl on the floor, and an adult woman. Mei-yu interpreted the drawing as about the girl breaking a bowl and her mom reacting to it, including beating her. When the research assistant asked what the girl in the picture would feel when her mom hit her for breaking the bowl, Mei-yu went silent. Then, the research assistant repeated his question: “What will the girl feel?” Surprisingly, Mei-yu pivoted her answer to an apparently irrelevant topic, big brother taking her candy:

Figure 5.7 TAT Drawing #5.

Source: The Wolf Archive. Photo by Jing Xu

Another detail in the story Mei-yu told, about the girl character breaking a bowl, alludes to her own family’s experience. Right before the earlier segment in her TAT transcript, Mei-yu mentioned that the girl’s mother found out she broke a bowl. Asked what would happen then, Mei-yu answered: “When mom found out, she didn’t feel well.” When asked why mom didn’t feel well, she pivoted, again, to the topic of fighting, the girl fighting with her big brother. Children fighting and mom feeling unwell was not something portrayed in that particular TAT drawing at all, but it encapsulated what happened to Mei-yu’s mother in real life, as shown in A Thrice-Told Tale: Her estranged relation with certain neighbors, due to her boy Yi-kun’s fight with neighbors’ children, was a major contributing factor to her mental illness. Coincidentally, Mei-yu’s TAT session was conducted in summer 1960, just a few months after her mother’s “erratic behavior.” Pressed by the research assistant twice to explain why, in the drawing, the mom would feel unwell upon her children’s fighting, Mei-yu kept silent. This little girl’s nuanced reactions, her utterances and hesitancy, hint at the lingering impacts of the chaos in the adult world on these young children’s emotional and moral experience.

Mei-yu’s reactions in projective tests share a similar quality to what I found in many other children’s projective tests transcripts, that is, the quality of “reality-based fantasy.” Sibling conflict is one of the most common themes across all Doll Play transcripts (nearly fifty children), which probably reflects its prevalence in this community. Moreover, the stories children told are closely related to, or mapped from, the reality of sibling rivalry at home. CO abounds with stories of siblings fighting for fairness in resource distribution and tattling to parents about sibling-inflicted injustices. Readers might feel sympathetic for the young mother in the following scenario, who had to settle disputes between her children over some peanuts or ten cents:Footnote ⁴⁶

Mirroring observational records, sibling rivalry emerged as a main theme during children’s spontaneous narratives in projective testing. Some children in their answers even explicitly inferred about parental favoritism. This sheds light on a key mediator of sibling disputes, that is, parental interference. One drawing in TAT featured one big boy (B1) standing behind a tree and watching a smaller boy (B3) playing with other kids in a ball game (see Figure 5.8). When asked to describe what happened in that scene, ten-year-old Cheng Ling-hui answered that the big boy (B1) was about to sneak out and hit the younger one (B3), and that they two were brothers.

Figure 5.8 TAT Drawing #2.

Source: The Wolf Archive. Photo by Jing Xu

Asked why they would get into a fight, Ling-hui told this story:Footnote ⁴⁷

Proceeding to another drawing in TAT (see Figure 5.9), Ling-hui again used the same lens to interpret a different scene: Three boys (B1, B2, and B3) were fighting for some fruits and their mother (F1) came out to see what happened. Mother went back home and told their dad, and dad would tie the oldest child up and beat him. Given that Ling-hui was indeed the oldest child in his family, it is reasonable to infer that this imaginative narrative was partly inspired by his real-life experience. Such storytelling experience probably provided many children like Mei-yu and Yi-kun an opportunity to express their negative sentiments toward their siblings and parents.

Figure 5.9 TAT Drawing #8.

Source: The Wolf Archive. Photo by Jing Xu

“A Fiercely Protective Mother”

The tragic story of Wang Chen-hsin – the mother of Yi-kun and Mei-yu, illustrates the extent to which children can disrupt adults’ relationships with neighbors. In their projective tests these two children painted a picture of their mother as a ruthless punisher at home. Mother’s punishment appeared in both Mei-yu and Yi-kun’s TAT transcripts: For example, when shown the drawing of a girl character, a broken bowl and an adult woman character, Yi-kun mentioned the same scenario as Mei-yu did, that mother would beat up the girl; Presented with other drawings featuring multiple children, both Brother and Sister inferred that children were fighting and would be punished by the mother figure.Footnote ⁴⁸ When Yi-kun or Mei-yu were playing with other children outside, however, their mother appeared more often as their defender. The following is an excerpt of one episode observed right in front of their house.Footnote ⁴⁹ In this scenario a little boy attempted to tattle to Yi-kun’s mother about Yi-kun’s misbehavior, stealing flowers. We see Yi-kun’s naughty disobedience: He ignored mom’s threat and punishment. But we also see why Margery Wolf described Chen-hsin as “a fiercely protective” motherFootnote ⁵⁰ – she cursed the tattler and gave the flower back to Yi-kun:

This episode was observed in summer 1959, nine-months before Chen-hsin’ mental breakdown recorded in A Thrice-Told Tale. The kinds of quarrels Chen-hsin got herself into, with other mothers, probably all evolved from children’s conflicts over trivial things, like this argument about a flower and a dragonfly. A Thrice-Told Tale only presented fieldnotes up to March 15, 1960.Footnote ⁵¹ That ethnography did not tell us what happened to Chen-hsin after her outbursts of “insanity,” but I was curious about it. The last episode of CO where Chen-hsin appeared in was March 30, 1960, twenty-four days after she plunged into the rice paddy. In this episode,Footnote ⁵² Wang Yi-kun was playing with Wang Ching-fu and some other children again, climbing the guava tree. Chen-hsin was commanding Yi-kun to get down, come home and do his homework. Shaking a stick, she threatened to beat him up. But when her son got caught in the tree, she laughed and came up to hold him until he could get down. A couple of weeks ago she was still jumping around and doing bai-bai motions like a spirit-medium, or like a crazy woman who needed to be tied up in a mental hospital.Footnote ⁵³ Now she seemed to have restored her sanity. Wolf’s “general observations” (type “G” notes), focused on adult life, also include further traces of her, although Margery Wolf did not report these in her ethnography. For example, on June 13, 1960, Chen-hsin was gossiping with a group of women about other adults and children, even teasing her own daughter Mei-yu lightheartedly.Footnote ⁵⁴

On September 20, Yi-kun climbed up a neighbor’s fence to get some leaves for the girls who were playing house nearby. The neighbor, a thirty-two-year-old woman, scolded Yi-kun and reported to his mother, but Chen-hsin did not do anything about it:Footnote ⁵⁵

Chen-hsin, “the woman who did not become a shaman” (M. Wolf Reference Wolf1990a), was still around in the village by the end of the Wolfs’ fieldwork. She did not divorce her husband. She did not leave her boys behind and run away to her natal home with her daughter. She continued to gossip with her neighbors and sometimes quarreled with her neighbors over children’s issues. She was still known as a protective mother, even by the village children.

An Untold Tale

My motivation to pursue the story of this brother–sister dyad partly originates from personal experience, growing up as a Han Chinese daughter. Being an only child, among the first generation of singletons born under China’s One-Child Policy, I always wondered what it would be like to have a sibling. I asked my mother why she did not give me an older brother. I was also terribly naïve about sibling relations. To me, having an older brother meant having someone to protect you and rely on. But my mother always jokingly warned me: “You are lucky not to have an older brother! Otherwise, your father would have favored him.” My young mind could not fathom the profound sentiment of sibling rivalry: Not only did I not have the first-person experience, other children in my peer world were mostly singletons too, so I did not even have opportunities to observe sibling dynamics or hear relevant testimonies. Now my son, an only child living in America where most of his friends do have siblings, told me: “My friends all say that I am lucky not to have any siblings.” From the viewpoint of this case study, both my mother and my son were half right: Many if not all families in Xia Xizhou favored a son over a daughter. In Mei-yu’s storytelling, her brother was quite annoying at home. Yet when playing with other children, Yi-kun seemed like a caring elder brother and they two often supported each other. Going beyond general patterns of gender, this might reflect individual children’s personalities and circumstances, but it also points to the ambiguous nature of sibling relation, tension at home and solidarity in the outside world.

This personal curiosity aligns with my scholarly interests: Prior research on “the Chinese family” or Chinese moralities rarely focused on sibling dynamics in childhood. The duality of care and rivalry in the early shape of sibling relation contributes to children’s developing moral knowledge and emotions. Siblings learn to care for each other and build coalitions. Through constant negotiations of how resources and responsibilities should be distributed, they also gain intimate understanding of fairness and justice, authority and hierarchy. The contradictory dimensions of love and power are particularly pronounced in cross-sex sibling relation in patriarchal societies due to differential statuses of boys and girls (see e.g., Joseph Reference Joseph1994). Therefore, the ambivalence of sibling relation in this case study can help us understand the complex, double-edged quality of human kinship more generally.

Besides weaving together various aspects of learning morality, including parent–child dynamics, peer interactions, gender, and siblings, this case study also accentuates critical reflections on ethnographic epistemology. The case recovered what was obscured from a classic ethnography, A Thrice-Told Tale, which was already infused with meta-reflections on ethnographic authority. In that book Margery Wolf reflected on the role of the anthropologist as writer and the position of ethnography as a genre of writing. Every ethnography has its omissions and silences (Silverman Reference Silverman2020). A Thrice-Told Tale is no exception. Not to mention that the anthropologist relied on her intermediary MC, the confidant of village women, to gather evidence concerning a scandal. To Margery Wolf, all the important characters were adults. She distinguished her ethnographic tale from both fiction and raw fieldnotes, reconstructing a plot based on selected materials from what was called “general observations” in their fieldwork, the only type of fieldnotes focused on adults. But even in these raw fieldnotes listed in A Thrice-Told Tale, I discovered traces of children: The protagonist’s children played a role in the unfolding of their family drama. Other children witnessed that drama and participated in gossip. These traces were perhaps less relevant to Margery Wolf’s question, whether the woman in question became a shaman or went crazy. But they are central to my question.

To put children under the spotlight, I systematically searched for and analyzed the voices and behaviors of Yi-kun and Mei-yu in the Wolf Archive. My narrative centered on children’s agency in learning morality: The brother and sister helped each other in peer interactions, longing for social inclusion; they told stories of sibling conflicts, stories laden with moral judgments; they cared for their mother when she was going through a difficult time; they also maneuvered against her punishment. My narrative reflects my intentional choices and implicit biases, just as A Thrice-Told Tale reflects the choices and biases of my predecessor. This untold tale therefore bears witness to the limitations, possibilities, and charm of ethnography.