Extended Family and Jiagui (‘Family Rules’)

The elderly hutongers often described to me what an ideal siheyuan would look like, which could be summarised as below:

Siheyuan is often laid out along a north–south axis, with its buildings on four sides placed in alignment with the four cardinal points. South is the auspicious direction, which a siheyuan should ideally face. With doors and the main windows of the buildings opening onto the interior yard, the household space created by siheyuan is connected with the outside world by the entrance gate, often preferably located in the south-eastern corner. The square or rectangular shape and south-facing orientation are central to a siheyuan, on which a more elaborate series of spatial and domestic arrangements are made to accommodate an extended family.

I was told this layout of the siheyuan followed the rituals of fengshui (‘wind and water’) and tian yuan di fang (‘round sky and square earth’). It was based on the same rituals that the Forbidden City – the emperor's home and the most splendid siheyuan compound made up of a series of courtyards – was also built. That is to say, the space of the commoners’ home and that of the ruler's home are similarly constructed, albeit at different levels. These, as articulated by the hutongers, reflected the ancient cosmology and state discourses of tianxia (‘all under heaven’) and the ruler as the tianzi (‘Son of Heaven’), to whom all the people should defer.

Particularly, a set of rules were mentioned to apply to room assignment and property inheritance among the family members. For instance, hutonger Grandma Yang told me:

The south-facing zhengfang (‘main rooms’), situated to the north of a siheyuan, were of the highest status and were occupied by jiazhang (‘the household head’, usually the senior male). Rooms on the eastern and western sides of the courtyard were called xiangfang (‘wing rooms’). They were of secondary status and were usually distributed among the sons first and then daughters. Servants and guests usually stayed at the daozuofang (‘back-facing rooms’), the north-facing rooms beside the front gate, located across the yard from the main rooms. This position was least important.

The scheme of room assignment showed the importance of seniority and gender in the household. In fact, Grandma Yang invoked the phrases zhang you you xu (‘There is an order between the senior and the junior’) and zhang xiong wei fu (‘the eldest son acts as father after the father died’) when explaining to me the general family rules. She said that jiazhang had the authority in all respects, and sons enjoyed more privileges than daughters. The birth order mattered.⁵ But primogeniture was not strictly practised in the hutong, and all sons could inherit the siheyuan more or less equally. ‘But daughters had to leave their niangjia (‘natal home’) and join their pojia (‘husband home’) after marriage, she added.

The family rules were also remembered as related to family rituals of ancestor worship by some other hutongers. For instance, Grandpa Zhang recollected that the ritual focus of their old siheyuan was the tangwu (similar to a living room), located in the middle of the row of zhengfang. On special occasions, a few wooden ancestral tablets would be placed on a tiaoan (‘a long narrow table’) in the tangwu. Offerings of fruits and meats would be served. Grandpa Zhang explained that it was mainly about xiao (‘filial piety’), the reverence for and submission to parents and ancestors even after they were dead. Not all hutong families performed this kind of ritual at home. Grandpa Liu insisted that the ancestral tablets should be placed at the communal ancestral hall for the descendants to practise xiao. Ritual and family rules may vary, but the central Confucian idea of xiao is universal.

Described in this way, the siheyuan strongly feels like a space created by and for a certain conception of the family, aspects of which are akin to the house ‘not just of unity but also of various kinds of hierarchy and division’ (Carsten and Hugh-Jones 1995: 12), as well as to other symbolic meanings expounded by observations elsewhere (cf. Bourdieu 1979; Prussin 1995; Sneath 2000). These descriptions and the everyday life in the old siheyuan are usually generalised by the hutongers in terms of ‘Confucian family’, and more generally, ‘Confucianism’. This association with classical learning situated siheyuan in line with a great tradition.

Imperial Kinship and the ‘Confucian State’

There were cases when elderly hutongers spoke of the family rules in disciplinary and punitive terms similar to that of guofa (‘state law’). For instance, Grandpa Zhang explained,

In the old days, it was very important for a grown-up son to practise xiao and look after his parents when they reached old age. Neglect of xiao was a serious breach of the family rules, and sometimes even treated as an offence deserving punishment under state law.

In this sense, the household is brought under the purview of the state. This reminded me of a well-regarded book, Law and Society in Traditional China (Qu 1947), which lists detailed articles stipulating proper household rituals and individual conduct, and discusses the ways in which ‘Confucianism’ and state law are intertwined to maintain a hierarchical and gendered ‘tradition’ and order. To some extent, the boundaries between family rules and state law were blurred.

More straightforwardly, the phrase jun (‘ruler’) chen (‘minister’) fu (‘father’) zi (‘son’) from Confucius’ Analects prescribes a set of law-like rules for everyone to follow. In John King Fairbank and Merle Goldman's interpretation:

Confucius had said (rather succinctly), ‘jun jun chen chen fu fu zi zi’, which in its context meant ‘Let the ruler rule as he should and the minister be a minister as he should. Let the father act as a father should and the son act as a son should.’ If everyone performed his role, the social order would be sustained. (2006: 52)

Here, something important was lost in translation. The four Chinese characters (jun ‘ruler’, chen ‘minister’, fu ‘father’, zi ‘son’) each repeat themselves (jun jun, chen chen, fu fu, zi zi) to make four noun–verb patterns, suggesting an agreeable superiority of jun (‘ruler’) to chen (‘minister’), and of fu (‘father’) to zi (‘son’) and drawing an analogy between the ruler–minister relationship and father–son relationship. This wording frames social relationships and the state order in a language of kinship, in that ‘the relations between central power and individual subject were intended by the state to be mediated, up to a point, by his kinship ties’ (Freedman 1979: 338). It is worth noting that women are not mentioned in this phrase, because they are excluded from the ‘male-dominated kinship system’ (Watson 1982: 615;).

Studying ‘metaphor of the state as an enlarged family’, Weiming Tu comments that ‘Kinship so conceived extends way beyond the normal confines of the family and substantially fills the perceived space between family and state’ (1994: 1136). Nevertheless, we might go further, since when the state is conceptualised as ‘an enlarged family’, the family does not stand alongside the state as a discrete entity, but is entangled with it. These practices reflect the longstanding notion of jia guo tong gou (literal translation ‘family/lineage and state sharing the same organising principles’). The imperial discourse of jun chen fu zi continued to play a crucial role in organising family and the state even after the imperial political systems collapsed with the fall of the Qing Dynasty in 1911. In the Republican Era, Xiaotong Fei perceives a patriarchal China that ‘government officials are supposed to . . . “Be parents to the people” ’ (1992 [1947]: 118).

It is worth emphasising that the concepts of ‘Confucian family’ and ‘Confucian state’ can themselves be problematic. They risk treating ‘Confucianism’ as a grand narrative, or great tradition. This type of description, similar to the claim of the ‘walled culture’ of China, involves what Michel Foucault (1989: 17) rejects as the problem of ‘cultural totalities’. More specifically, as Stephan Feuchtwang (2009) notes, the model of ‘great traditions’ was employed in a series of books on comparative studies of cultures and civilisations published in the 1950s and 1960s.⁶ On that theme, the two volumes on China both adopt a stance focusing on ‘textual traditions, high status practices’, and so forth (Ibid.). This approach has been largely abandoned in social anthropology as it downplayed ‘everyday practices’ and ‘the work of transmission at all levels ’ (Ibid.).

In this article, I treat the idea of ‘Confucianism’ broadly, not as a particular school of thought or as a fixed system of knowledge and principles, but as an old, powerful, expansive and ever-changing phenomenon that still influences aspects of personal life, kin relations and state discourses. The concept ‘Confucianism’ offers a useful reference framework within which I can better analyse my fieldwork data, because ‘Confucianism’ was the term frequently used by the elderly hutongers when they talked about the ‘everyday practices’ in the past.

Family Stories of Uncle Song

Take the siheyuan that belonged to Uncle Song's maternal family, for example. Uncle Song, born in 1950, is a retired teacher of history from a vocational school on the outskirt of Beijing. When I first visited the siheyuan in 2008, the siheyuan was inhabited by Uncle Song in the inner court, and his aunt Qian (wife of Song's mother's brother) and her son Yu in the outer court. It was run-down and shapeless as a result of higgledy-piggledy extensions and lack of maintenance.

Since then, I have visited the siheyuan many times and had numerous conversations with Uncle Song and his cousin Yu, either separately or together, about their house and family, and what had happened in the hutong and beyond. Qian suffered serious chronic illness, so had difficulty communicating with us. But she was always happy to see me there and to share what she could remember, such as the names of those who had lived in the siheyuan before. Over time, what they told me, either complementary or contrary to, or overlapping with each other, gradually helped me build up a cross-generational picture of their big family.

The siheyuan had been in Uncle Song's family since it was bought about one hundred years ago by Song's maternal great-grandfather, a minor official in the Qing government in charge of grain at the local Jiaodaokou Grain Bureau. Song's grandfather went to Beijing First Middle School, a state school and part of the Qing keju (‘imperial examination’) system. Its graduates got the prestigious xiucai diploma. After graduating, he got a good job with the Bank of China. Uncle Song's mother went to a French-style school attached to Shanghai Zhengdan University. That was during the Sino-Japanese war, when the whole family moved to Shanghai. But she didn't graduate because of the war. They had to move to Hong Kong and then Chongqing. Later she worked as an accountant. She had quite a successful career for a woman at that time, but she didn't have a very happy family life.

It had been a long ‘Confucian’ tradition for a woman to practise sancong (‘three subordinations’): to her father before marriage, her husband during marriage and her son in widowhood. Although these patriarchal norms were declining during Uncle Song's mother's life, patrilocal residence was still common. Following most other women of her generation, Song's mother left her niangjia (‘maiden family’) to live with her pojia (‘husband's family’) when she married Song's father in the 1940s. Unfortunately, for reasons unclear to me, their marriage dissolved, and Song's mother moved back with her young son Song to re-join her parents and siblings in their siheyuan.⁷

The ‘maiden home’ of Song's mother was originally a rectangular siheyuan courtyard house, wider along the north–south axis and narrower on the east–west axis, with main rooms in the north and south of the courtyard, and a kitchen and a storage room in the east and west respectively. This left an open space in the middle. It was designed for one family.

Uncle Song recalled:

My great-grandfather had two sons, my great-uncle and my grandfather. The siheyuan was equally divided into two yuan (‘courts’) after they grew up and married respectively. My great-uncle lived in the outer court in the south. His wife died of tuberculosis and their only son passed away aged only one year. My grandfather, his wife and their two sons [one being Yu's father] and five daughters [one being Song's mother] lived in the inner court in the north. My great-uncle, widowed and childless, adopted a boy [Yu's father] and a girl of my grandfather's. Later on, my great-uncle remarried and had a daughter.

Uncle Song did not mention if his great-grandfather had daughters. Probably because daughters were irrelevant as they did not count as inheritors then. Clearly, however, it was one family who shared the house and hearth when Uncle Song's great-grandfather was jiazhang (‘household head’). After the two sons reached adulthood and set up their own families, not only was the house split in half, but also the hearth. This was significant because the way in which family members were related experienced generational changes, as Carsten and Hugh-Jones note, ‘the hearth is as much a defining feature of the house as eating together is a defining feature of kinship. But the hearth is not just a symbolic centre; it is also instrumental in processes of transformation’ (1995: 42).

Socialist House and Kinship

To show us how the siheyuan originally looked, Uncle Song drew a plan with the two courts split by a pond in the middle. There were trees, flowers and a wooden screen in the yard. Uncle Song especially marked where an almond tree, a date tree and a Chinese mahogany tree had stood. While drawing, he said that, unfortunately, the almond tree died in 1967. He said that he remembered the year 1967 very clearly, because the following year was one of the biggest turning points in his life. He left Beijing as a member of the zhiqing (‘educated youth’) to work in the Heilongjiang Production and Construction Corps in beidahuang (‘the great northern wildness’) in northeastern China. Misfortunes are often attributed to the adverse forces of nature and the death of an old tree is a bad omen.

China was under high socialism then. State discourses took a sharp turn to ‘smash the old world and to create a new one’. An all-out attack on the pre-revolutionary sijiu (‘four olds’) was launched during the early stage of the Cultural Revolution in 1966. The ‘four olds’ comprised old ideas, old culture, old customs and old habits, with the once highly valued Confucianism at the core. This seemed to bring about positive changes to aspects of women's rights, including receiving equal inheritance shares with men. It did not make any real sense, however, as private houses were largely nationalised and privatisation was illegal.

Uncle Song mentioned that the people's commune emerged in the city during the Great Leap Forward period (1958–1960), which also affected everyday life in the hutong. In fact, in 1960, the Jiaodaokou Street Office, which oversaw the South Gong neighbourhood, was renamed ‘Jiaodaokou People's Commune’. Uncle Song remembered that they had been organised to eat at a small communal shitang (‘dining hall’) set up by the community centres in the hutong. This did not last as long as it did in the countryside, he said, but it changed the hutongers’ ideas about food-sharing. Several other hutongers also told me that at that time the Soviet word ‘comrade’ was introduced as a universal term to address one another, both in public and in private, even replacing many of the kin terms, to such an extent that it became common for husband and wife to call each other ‘comrade’. These changes indicated the emergence of a socialist form of interpersonal relatedness.

Uncle Song spent six years driving a tractor on the state farm on the Sino-Soviet border in Heilongjiang province. He said that, during those times, their siheyuan was requisitioned and strangers moved in. Their garden was cleared and the trees removed to allow space for extensions for a larger number of dwellers. This was said to have happened to nearly all the private siheyuan owned by the ordinary hutongers, who were only allowed to keep one or two rooms, depending on the size of the family. It is worth noting that the state-owned siheyuan were also overcrowded due to policy-caused massive immigration of both Communist cadres and proletariat workers.

As can be seen, under socialism, siheyuan, the assumed locus of uninterrupted ‘Confucian home’, was stripped, at least in part, of its ‘Confucian’ identity and was used like the socialist danwei (‘work unit’) compound to accommodate the new proletarian subjects. Siheyuan occupants were a heterogeneous group from all walks of life. Some of them belonged to local district-level danwei. Whatever their backgrounds, as mentioned above, they were now related by a common name ‘comrade’, a new kinship-like terminology symbolising membership of a vigorously promulgated concept of ‘big socialist family’.

Eventually, Uncle Song was fortunate to secure an opportunity to return to Beijing and was allocated a job at a semi-conductor factory in 1974. For Uncle Song, another important thing was that higher education re-opened in 1977 after a ten-year shutdown.

The Reform Era

In 1979, Uncle Song started a university correspondence course in history. In 1985, he was offered a position as a history teacher at a vocational school in a suburb of Beijing. In 1995, when post-Mao China became increasingly open, Uncle Song sought an opportunity to study world civilisation at Helsinki University. Uncle Song remarked:

I spent ten years in Finland, and only returned to visit China once in 1998. During the ten years, tremendous changes took place in China. I found it difficult to adapt when I finally returned to Beijing in 2005.

I could tell that Uncle Song himself has changed greatly over time. The only university graduate and also the only person with overseas experience in his generation of hutongers whom I have met, Uncle Song did not seem to be satisfied with his achievements, but implied his material marginality and loneliness. Partly because of his long absence from his danwei, the vocational school, he was made to retire early after he returned from Finland. His family have gone their separate ways, and three of them were now living in different countries – himself in China, his son in Finland and his ex-wife has remarried in Japan. His mother died in 2007.

Inheritance and Changing Kin Relations

In theory, Uncle Song was not entitled to live in the siheyuan following his mother's death. Because his name is not on the recently issued housing certificate, according to which Yu, I was told, his cousin, is the sole legal owner of the house. This certificate was seen by the locals as a symbol of China's integration into a global capitalist system. The post-Mao housing privatisation was piloted in the late 1980s and early 1990s in Beijing. It was not until the promulgation of Property Law in 2007 that private property ownership rights were formally endorsed. This ownership only applies to the house above the land, because all urban land belongs to the state. Nevertheless, Uncle Song told me, ‘There was no formal written agreement drawn up dividing the property when my grandfather and great-uncle were sharing the courtyard. But following tradition, there was tacit understanding that equal shares of the family estate were divided between brothers.’ As discussed, this reflects the pre-revolutionary inheritance practices.

Uncle Song commented:

Things have changed since 1949, when women started to enjoy equal rights as men and patrilineal inheritance was banned as feudal practice. Although the pond served as a natural border for the two households living in the inner and outer courts respectively, as a property the residence was equally owned by my mother, her six siblings and one cousin, who were all residing in the courtyard when I was a little boy. But many of them moved out over time. In 1983, private houses were returned to their previous owners who were able to produce their old house deed. We got our siheyuan back, and all seven siblings and one cousin had the right to claim their share. My mother was the traditional-thinking eldest daughter of the family. She preferred the courtyard to remain in the paternal name Yu, and so gave up her rights. Her sisters, one brother and the cousin, like most people in the 1980s, had little interest in a rather dilapidated and worthless house in a run-down area, and accepted a small amount of monetary compensation. So the remaining younger brother became the sole owner of the property. After his death, his son and only child Yu, my cousin, inherited the courtyard.⁸

Born and raised under high socialism, Uncle Song invoked socialist values of equality and egalitarianism. He had taken it for granted that he would be one of the inheritors of the family estate, only to find himself disadvantaged by the rapidly changing state discourses, now embracing neoliberal privatisation and property law.

Nobody expected that (in 2008) the value of their siheyuan would soar to as high as seven million RMB (about £700,000) since 2003, when siheyuan started to sell for good prices. Yu, unmarried and in his early forties, has become a multi-millionaire on paper. Since he was the only person named on the property certificate, there seemed no legal dispute over Yu's ownership. Considering himself lucky in the lottery of life, Yu often carefully tried to avoid directly mentioning his kinsmen.

It was not until 2017, when I revisited Yu's siheyuan again and found Uncle Song had gone from there, that the other side of the story started to emerge. It gradually turned out that, rather than the entire siheyuan, Yu only inherited the outer court that had belonged to his father. Uncle Song's two aunts inherited the inner court. Their other siblings, including Uncle Song's mother, chose to give up their share of the siheyuan in return for an entitlement to danwei housing. There was no formal agreement, but the names of the final inheritors were registered at the local Housing Bureau. This left the ownership issue somewhat controversial. However, these two aunts of Uncle Song's were assigned jobs elsewhere and left Beijing. So the inner court was unoccupied. At the time, a housing sale and rental market did not exist and it was fine for Uncle Song's mother and Song to stay at the inner court for free.

Nevertheless, the situation started to change with the emergence of a housing market and the unexpected strong growth of house prices that followed. In particular, after the death of one of these two aunts of Uncle Song's, her daughter, Uncle Song's cousin, travelled to Beijing to claim her share of the property, part of the inner court now occupied by Uncle Song. Uncle Song insisted on his right to stay there: after all, there was no written agreement on inheritance. Besides, the house had little value when it was divided, and Uncle Song's claiming a share of it now as an equal member of the family was seen as a form of compensation for his mother's financial loss then. Their disputes resulted in a lawsuit which lasted for about ten years. Both Yu and another co-owner of the inner court were involved. The siheyuan was re-measured to ensure that the size of the inner court and that of the outer court were exactly the same, and that the inner court was further equally divided between Song's two aunts who had jointly inherited it. In the end, Uncle Song lost the case and had to move away from the siheyuan, which had witnessed the de-kinning processes.

Actually Existing Socialism

Uncle Song did not become homeless. It turned out that he had a danwei flat as backup. This on-campus old flat was allocated to Uncle Song by his former danwei, a vocational school. It was only leased to him at a very low rent during the socialist era, when all property belonged to the state. In the 1990s, however, with the development of post-Maot marketisation, much danwei housing was gradually privatised and sold off to their members at greatly subsidised prices. It was then that Uncle Song, following many of his colleagues, bought the flat from his danwei. When I last spoke to him, Uncle Song said he had lived in his flat since he left the siheyuan. He also expressed that, although he had actually enjoyed living in the flat, he still had strong personal attachment to the siheyuan.

For Uncle Song, the post-Mao practices invoked ‘actually existing socialism’ as more than political rhetoric, but a stark reality manifest in the basic form of shelter. Uncle Song's ten-year overseas experience contributed to his transformation from a socialist ‘educated youth’ into a self-declared liberal-thinking westerniser. He thought he had made his farewell to the ‘socialist other’ (borrowing Chris Hann's terms; 2002: 1), but would eventually have to end up in the seemingly outmoded, yet functioning, danwei system from which he had made an effort to get away. Uncle Song's experience shows post-socialism's ‘twists and unexpected reversals’ (Humphrey 2002: 13). His invocation of socialist notions of equality in the face of the privatisation of the siheyuan indicates the importance of this continuity too.

Neoliberal Practices

Yu was now actively developing a side-career as a landlord alongside his danwei-like job in one of the six residents’ committees at South Gong. In fact, Yu did not have the money to refurbish his run-down siheyuan, but this did not prevent him from making money out of it. In 2008, he rented out extra rooms for about 2,100 Yuan (about £210) per month, which was more than his lower-than-average monthly salary of just over 1,000 Yuan (about £100) from the easy job at the residents’ committee. Besides, while he probably paid income tax on his salary, the rent, paid in cash, was not taxed. This was an instance of a growing trend to engage in business activity alongside formal jobs in China. It was not uncommon, as in Yu's case, that the additional incomes were higher than the salary. Yu once said with satisfaction that his residents’ committee job was a means of safeguard, but money needed to be made in the market. Yu's income-generating exercise of room renting can be understood in terms of what Aiwa Ong and Zhang Li (2008) see as discursive and pervasive neoliberal practices in post-Mao China. They argue that ‘In China, market-driven practices are inextricably linked to state policies’ (2008: 9–10). Obviously, Yu seized the opportunity brought about by the housing privatisation policy and the Property Law.

Yu made it clear that he had no intention of selling the siheyuan, even though it would enable him to become a multi-millionaire overnight. He was more interested in establishing a secure and reliable lifetime income, like those guaranteed by the good danwei in former times.

De-kinning and Continuity

Chinese kinship has always been of ‘philosophical and political’ significance (Ebrey and Watson 1986: 1). One of the best ways to approach it is to treat it as an ‘empirical question’ in Schneider's (1984) sense, through ‘ethnographic particularities of being related in a specific cultural context’ (Carsten 2000: 4). As discussed, the state seems omnipresent in the cultural context of modern China, whether in the form of ‘Confucianism’, socialism or a post-Mao mixture of socialism and neoliberalism. The empirical study of the siheyuan illustrates that property and how it is inherited on the basis of changing state discourses play a significant role in kinning and de-kinning family members. If taking Fei's view that ‘Structurally, Chinese families are lineages’ (1992: 83), the lineage of Yu has gradually fallen apart during the historical processes of socialist construction and post-Mao privatisation. Extended family, a pre-revolutionary ‘ideal type’ of kinship deriving from a ‘Confucian’ state, does not seem to endure through time. After socialism, the ‘marketization of social relations’ (Shue 1994: 75) serves as the last straw to dissolve the presumably indissoluble kinship ties. Or perhaps such forces were always present, and it was just that the state's emphasis on lineage and family had changed.

Subtle continuities are identifiable, however. In the case of Yu's family, the transformation of the siheyuan has witnessed the decline of the residential patterns of si shi tong tang (‘four generations living under the same roof’). Distrusting their relatives, the then unmarried Yu and his aged mother kept each other company in the siheyuan, to some extent living a life typical of the old era. Yu did not distinguish, for example, his income (salary and rent) from his mother's pension, and referred to them as their mutual money. As well as sharing income and spending a large part on his mother's medical bills, Yu looked after her. Yu was widely praised as a xiaozi (‘dutiful son’) by those who knew him in the hutong. In fact, he was said to have been offered his current job at one of the local residents’ committees due to his display of xiao (‘filial piety’). The context has changed, but the notion of ‘filial piety’ persisted in one form or another. As Caroline Humphrey notes from Soviet and Eastern European experience, ‘There never can be a sudden and total emptying out of all social phenomena and their replacement by other ways of life’ (2002: 12).

New Forms of Relatedness

Concurrently, new forms of relatedness are emerging. When I spoke to Uncle Song last time, he seemed happy with his danwei way of life. He said he often met other former teachers for lunch at the school canteen, and enjoyed chatting with them and their company. Enthusiastic about his specialisation in history and world civilisation, Uncle Song has found his audience and the opportunity to mix with like-minded people. Suffice it to say that the danwei has offered him a community in which he finds his place, even long after his retirement.

Yu's life has also changed greatly. When I visited their siheyuan again in 2013, five years after my first period of fieldwork from 2007 to 2008, Yu was married with a young daughter. It was now a family of three generations living under the same roof. Yu valued his job even more than before as a stable source of income, while still actively engaging in the rental market, to support his daughter as he mentioned. His wife, a freelance maths tutor, often worked irregular hours to make more money. There seemed a consensus between the couple on building financial security for their daughter. Yu's mother was frail and had to spend much time resting in bed. Yu's mother-in-law moved in to look after her granddaughter full-time. This division of labour became common in the hutong and beyond, as grandchildren took priority over grandparents. It could be said that, rather than the siheyuan, it is the grandchildren who now unite the family.

There are many reasons contributing to the grandchildren-first phenomenon, and two of them are most frequently mentioned. First, the one-child policy, initiated by the state in the 1970s and 1980s, makes the child, no matter a boy or a girl, more precious. Second, the pressure on the family from a quickly modernising state to raise competitive children. This means concentration of all possible resources on children. For the former, although the one-child policy has been relaxed and two (even three now) children have been encouraged over the past few years, urbanites are used to bringing up just one ‘quality’ child as the state once vigorously promulgated and required. Besides, the cost of child-rearing has become incredibly high, and many families cannot afford to have a second child. For the latter, a child is now often seen as the hope of the entire family. The collapse of the cradle-to-grave socialist welfare system and the economic bubbles have left parents with little choice but to invest in their children. Therefore, it is not difficult to understand what Yunxiang Yan terms ‘descending familism’, ‘the shift of foci of family life . . . from ancestors to grandchildren’ (2019).

We might say that dramatic ruptures and fragmented continuities have intertwined to redefine Chinese kinship. Starting with the pre-revolutionary image of the ‘traditional’ extended family, we have examined the emergence of new forms of relatedness in the high socialist and the Reform periods, from the ‘big socialist family’ to Uncle Song's reconnection with socialist danwei, Yu's participation in the neoliberal housing market, and the shift of family attention onto grandchildren. The thread running through all of them is the theme of house and inheritance, and the underlying deep structure of the changing state discourses. However, it would be an oversimplification to see siheyuan as a symbol of submission to the state. On the contrary, confronted with the turbulent history and conflicts between kinsmen over property and inheritance, the siheyuan residents, such as Uncle Song and his cousin Yu, have demonstrated a high degree of resilience and adaptability, and even the ability to use policies to their own advantage. The processes of ‘de-kinning’⁹ present dynamic, diverse and complex arenas of kinship studies in modern China.¹⁰