China has the longest continuous tradition of systematic thinking about social order anywhere, a bureaucratic state a thousand years older than any in Europe, and an academic sociology that was abolished by government decision in 1952 and reconstituted in 1979.
It also produced one concept that belongs in general social theory and has never been given its place there.
Two ways of drawing a society.
Ask a Western sociologist to draw social structure and you will get boxes: groups with boundaries, containing members. Class, family, organisation, nation. You are in a group or you are not. The boundary is the thing.
Fei Xiaotong , writing in 1947 for a Chinese newspaper readership, offered a different image, and he took it from something everyone had seen.
Throw a stone into water. Ripples spread outward in concentric circles from the point of impact, each weaker than the last, fading with distance.
Every person, Fei says, stands at the centre of their own set of ripples. The circles are relationships — family first, then kin, then neighbours, then acquaintances, then strangers — and obligation declines with distance from the centre.
Three consequences follow immediately, and they are the whole argument.
Everyone's circles are different. There is no shared boundary. My network centres on me; yours centres on you ; and while they overlap, they are not the same structure seen from two sides. There is no group here that we are both "in."
Obligation is graded, not binary. You do not owe the same to everyone inside a boundary and nothing to those outside. You owe more to those nearer , continuously, without a line.
And the circles expand and contract. In prosperity they widen to include distant kin and associates; in adversity they contract to the immediate family. "The scope of the network is elastic."
Fei called this chaxugeju differential mode of association — and contrasted it with what he called the Western tuantigeju organisational mode : bounded groups with defined membership, in which everyone inside stands in the same relation to the group.
It is a formal contrast between two ways of organising social relations , and it belongs beside Tönnies (4.6.2), Durkheim's two solidarities (4.2.2) and Simmel's forms (4.4.1).
It is almost never taught with them.
The intellectual problem this Topic addresses is different from India's.
India's sociology developed under colonial rule , with the categories of a colonial state (6.1.1), and its central problem has been the book view and the field view.
China was never fully colonised. It was defeated, forced into unequal treaties, partitioned into spheres of influence and humiliated — but it retained a state, a language, a bureaucracy, an examination system and an intellectual tradition of enormous continuity.
Which produces a different question : what happens when a civilisation with its own two-thousand-year tradition of social analysis encounters Western social science as the knowledge of the powers that have just defeated it ?
Three responses were available, and all three were taken.
Wholesale adoption , on the grounds that Western knowledge was the source of Western power.
Rejection , on the grounds that it was foreign and the tradition was adequate.
And translation and transformation — taking the apparatus and rebuilding it with materials from the existing tradition. This is what actually happened , and chaxugeju is its best product.
The classical background
Lesson 3.1.3 covered the classical schools; the point to carry here is what they left as a vocabulary.
Confucianism supplied a relational conception of the person. The five cardinal relations (wulun ) — ruler/subject, father/son, husband/wife, elder/younger brother, friend/friend — define the person by their position in a set of hierarchical, reciprocal bonds. Four of the five are hierarchical and all five carry mutual obligation.
Key concepts : ren (humaneness), li (ritual propriety, which is also the ordering of social conduct), xiao (filial piety), and the rectification of names — the doctrine that social order depends on each role being filled according to its designation. "Let the ruler be a ruler, the subject a subject, the father a father, the son a son."
Legalism supplied the opposite: order through law, punishment and administrative technique, with human nature assumed self-interested and institutions designed accordingly. Its practical influence on the imperial state was far greater than its reputation , and imperial governance is usually characterised as Confucian in ideology and Legalist in method.
Daoism supplied a critique of both: wuwei (non-action), scepticism about ritual and law alike, and an argument that intervention produces the disorder it addresses.
Mohism supplied a rival ethic — jian ai (impartial care), the argument that one should care for others' families as one's own — and it is the sharpest possible contrast with chaxugeju . Mencius attacked it precisely on that ground : impartial care means treating one's father as a stranger, which is "to be without a father."
That dispute — graded obligation against impartial obligation — is a genuine ethical and sociological argument , and it was conducted in the fourth century BCE.
Yan Fu (1854–1921), covered in 4.6.3, is the hinge.
He translated Huxley, Spencer, Mill, Adam Smith and Montesquieu into classical Chinese between 1898 and 1909. He coined qunxue Xunzi's concept of qun : the human capacity to form groups, which Xunzi identified as what distinguishes humans from animals and makes civilisation possible.
That choice of term matters. It attached the new discipline to an existing Chinese concept rather than presenting it as an import.
And he translated with an argument. His rendering of Huxley emphasised that groups , not only individuals, compete and are selected — and that China must strengthen itself or perish. Social Darwinism (3.3.2) entered Chinese thought as an argument for national self-strengthening , and its influence on the reform generation was enormous.
Kang Youwei's Datong Shu — the Book of Great Unity — proposed the abolition of nations, classes, races, the family and private property, drawing on the Confucian concept of datong from the Book of Rites .
Liang Qichao popularised sociological concepts through journalism and analysed the clan system, the relation of family to state, and what he took to be China's deficiency in public morality relative to private.
And the discipline was institutionalised early : sociology taught in Chinese universities from 1913, with Yenching University in Peking becoming the leading centre under Robert Park's visit in 1932 and the influence of Chicago urban sociology.
Fei Xiaotong
Fei Xiaotong (1910–2005) is China's most important sociologist and one of the more remarkable lives in the discipline.
Trained at Yenching, then Tsinghua under the Russian anthropologist Shirokogoroff , then at the London School of Economics under Bronisław Malinowski , who supervised his doctorate and wrote the preface to his first book.
Peasant Life in China Malinowski's preface makes the point explicitly : anthropology had studied "primitive" societies through outsiders; here was a modern society studied by one of its own members.
Its subject is the rural economy under strain : silk production collapsing under competition from Japanese and Italian factory silk, land tenure, debt, and the erosion of the household economy. It is an economic sociology, not a study of exotic custom.
From the Soil Xiangtu Zhongguo , 1947) — a series of newspaper essays, and the source of chaxugeju . Short, accessible, written for a general readership, and the most theoretically important thing he produced.
And Earthbound China (1945), on rural industry, which developed the policy argument he pursued for the rest of his life: that China's rural areas needed local industry rather than only agriculture, to absorb surplus labour without mass urbanisation.
The reason this concept belongs beside Tönnies and Durkheim, stated as an argument rather than as an assertion.
It is a formal contrast (4.4.1), of the kind Simmel would have recognised: two modes of organising relations, abstracted from content.
Tuantigeju — the organisational mode. Bounded groups. Membership is binary. Everyone inside stands in the same relation to the group. The group is prior to and independent of any member ; it has a boundary, a name and a rule of admission.
Fei's illustration is the sheaf of straw: individual stalks bound together into a bundle, bundles into a stack. Each stalk is separate and the bundle is a real thing.
Chaxugeju — the differential mode. Egocentric networks. Obligation graded by distance. No two people share the same structure , and there is no group prior to the relations.
And Fei's crucial claim about ethics follows from the structure.
In the organisational mode, moral standards are universal within the group : the same rules apply to everyone inside. Fei connects this to Christian conceptions of a God standing outside and above all members equally — a structure in which everyone is equidistant from the source of moral authority.
In the differential mode, moral standards are relational : what is right depends on who the other person is to you. There is no general standard, because there is no position outside the network from which one could be issued.
Which produces his most-quoted and most-argued-about claim : that Chinese ethical thought has "no universalist morality," that concepts like ren remain abstract because they must be specified relationally, and that "public" (gong ) and "private" (si ) are relative rather than absolute — the family is public relative to the individual and private relative to the lineage, and so on outward.
This is either a profound observation about a different moral structure or an unfair characterisation, and the argument continues. But note what it is doing: deriving an ethical form from a structural one , which is exactly the move Durkheim makes for mechanical and organic solidarity (4.2.2) and Tönnies for Gemeinschaft and Gesellschaft (4.6.2).
And it generates checkable propositions. If obligation is graded by relational distance, then: recruitment should run through networks rather than open competition; trust should be relationship-specific rather than generalised; enforcement should operate through the network rather than through third-party institutions; and the boundary of legitimate favouritism should be contested rather than fixed. All four are empirically studied, and the guanxi literature below is where.
The abolition, and after
In 1952 sociology was abolished in China , along with political science and several other disciplines, in the reorganisation of higher education on the Soviet model. The stated ground : historical materialism made bourgeois sociology unnecessary.
Departments were closed. Sociologists were reassigned. Fei went to the Central Institute for Nationalities, working on minority classification.
In 1957, during the Anti-Rightist Campaign, Fei was denounced — partly for an essay describing intellectuals' situation — and spent the following two decades unable to publish or teach. During the Cultural Revolution he was made to clean lavatories.
The discipline was restored in 1979. Fei, then sixty-nine, led the reconstruction: the Chinese Sociological Association was re-established with him as its first president, departments were founded, and a generation was trained.
He returned to Kaixiangong , the village of his 1936 fieldwork, repeatedly over the following decades — producing what is one of the longest longitudinal village studies anywhere, spanning roughly sixty years and the entire transformation of rural China.
Two things to take from this.
A discipline can be extinguished by decision , and 4.5.5's mechanisms have a state-level version that operates far faster than institutional neglect.
And the recovery cost a generation. Chinese sociology after 1979 rebuilt with a twenty-seven-year gap in training, publication and continuity — which is a substantial part of why the tradition is less visible internationally than its history warrants.
Go deeper
Guanxi and the empirical literature
Guanxi chaxugeju operating.
Its components , as the literature has specified them: ganqing (affective sentiment), renqing (the obligation created by favours given, which must eventually be reciprocated), and mianzi (face — Goffman's concept, 5.3.2, arrived at independently and elaborated separately in Chinese thought).
The key structural feature : renqing obligations are not settled by equivalent return . An exactly balanced exchange terminates a relationship; leaving a small imbalance sustains it. The debt is the relationship.
And the empirical findings:
Guanxi has real economic effects. Studies of Chinese firms consistently find that connections affect access to credit, contracts, licences and information, particularly where formal institutions are weak.
Its importance is contested and appears to be conditional. The major debate — associated with Yanjie Bian , Douglas Guthrie , Andrew Walder and others — concerns whether guanxi declines as markets and formal institutions develop, or persists and adapts. Guthrie argued for decline; Bian's job-search research found network effects persisting and changing form. The current position is that it does not simply decline but changes function, and that its salience varies enormously by sector, region and period.
And it is normatively contested inside China , not only outside. The boundary between legitimate relationship obligation and corruption is exactly what is argued about , in anti-corruption campaigns and in ordinary life — which is what Fei's framework predicts, because a system of graded obligation has no line where a bounded one would have a rule.
The East Asian argument
The Weber question (4.3.10) has a second act, and it is one of the more instructive reversals in comparative sociology.
Weber's claim : Confucianism's orientation of adjustment to the world, rather than mastery of it, was among the reasons rational capitalism did not develop in China.
Then Japan, South Korea, Taiwan, Singapore, Hong Kong and coastal China industrialised at rates without historical precedent.
And the "post-Confucian hypothesis" appeared — Herman Kahn, Peter Berger, Ezra Vogel and others — arguing that Confucian emphases on education, discipline, deferred gratification, family obligation and respect for authority were advantages for late industrialisation.
Weber inverted, using the same materials.
And this course's position, stated plainly : the second argument has the same defect as the first. Both read economic outcomes back into a cultural essence , and a framework that can explain both stagnation and the fastest growth in history using the same variable is not constraining anything.
The better explanations are institutional and historical : land reform, state-directed credit, export orientation, education investment, Cold War geopolitics and access to American markets, timing relative to the technology frontier, and — in several cases — authoritarian states able to impose high savings rates. These are contested among economists and historians, and they are the right kind of argument.
What survives of the cultural claim : that educational investment in East Asian societies is unusually high and has an identifiable historical relation to the examination tradition. That is a specific, checkable claim, and it is a long way from "Confucian values explain growth."
And what the tradition offers the discipline
Chaxugeju — a formal alternative to the bounded-group model of social structure, with derivable ethical and organisational consequences. It should be in general theory courses and is not.
The renqing/mianzi apparatus — a developed vocabulary for obligation, reciprocity and face that is more differentiated than the Western equivalents and was produced independently.
A very long comparative record. Two millennia of continuous documented bureaucratic administration, examination-based recruitment, land tenure systems and demographic data. Chinese historical sociology has materials no European case can match , and comparative-historical work (4.3.10, 7.6) that ignores it is working with a truncated sample.
And an unusually clear case of the relation between a discipline and a state. Abolished in 1952, restored in 1979, and operating since within constraints that vary by topic and period. Chinese sociologists have produced substantial work on migration, inequality, the household registration system and urbanisation , and the question of what can be researched and published is a live one that the discipline's own sociology of knowledge (5.15) should be examining rather than ignoring.
"Confucianism is a religion." It is a tradition of ethical, political and social thought, with ritual practice; the "religion" question has a long history and no settled answer, and imperial governance combined it with Legalist administrative method.
"Chaxugeju means Chinese people are less moral." It describes a relational rather than universal moral structure. Fei's claim about the absence of universalist morality is contested and is not a claim about moral quality.
"Guanxi is corruption." It is a system of relational obligation, within which the boundary of legitimacy is contested — which is exactly what a graded system produces, and it is argued about inside China.
"Guanxi will disappear as markets develop." The Guthrie/Bian debate found it changing function rather than simply declining, with salience varying by sector and period.
"Confucian values explain East Asian growth." The same variable was used to explain the absence of growth. A cultural essence that explains both outcomes explains neither ; the institutional and historical accounts are the serious ones.
"Chinese sociology began in 1979." It was taught from 1913, produced Fei's major work by 1947, was abolished in 1952 , and was reconstituted in 1979 with a twenty-seven-year gap.
Why it matters
It supplies a formal alternative to the bounded group. Almost all Western social structural analysis assumes groups with boundaries and members. Chaxugeju is a different structure with different consequences
It demonstrates translation as a productive strategy. Yan Fu attaching sociology to qun ; Fei building a formal contrast from a Chinese concept and a European method. This is neither adoption nor rejection but the third option , and it is what Alatas's critique of the captive mind (4.6.3) was calling for.
It shows the Weber question failing twice in the same way. Confucianism explained stagnation, then explained growth. A course that taught only the first would be teaching a discredited argument; one that taught only the second would be teaching the same error with the sign reversed.
And it documents a discipline's abolition and reconstruction. Twenty-seven years, one decree, a generation lost, and a founder cleaning lavatories through the Cultural Revolution before rebuilding the field at sixty-nine. The mechanisms of 4.5.5 have a version that operates in a single afternoon.
Chaxugeju differential mode of association : everyone at the centre of their own concentric ripples, obligation graded by relational distance, no shared boundary, and an elastic scope that expands in prosperity and contracts in adversity. Contrasted with tuantigeju organisational mode of bounded groups with binary membership.
And the ethical consequence he derives : universal standards within a bounded group, against relational standards where there is no position outside the network from which a general rule could issue — with gong (public) and si (private) relative rather than absolute.
The classical vocabulary : Confucianism's five relations , ren , li , xiao and the rectification of names; Legalist administrative method, which shaped the imperial state more than its reputation suggests; Daoist critique; and Mohist impartial care , which Mencius attacked as "being without a father" — a fourth-century BCE version of the graded-versus-impartial obligation argument.
The transformation : Yan Fu coining qunxue qun , translating Huxley, Spencer and Mill with a self-strengthening argument attached; Kang Youwei's Datong Shu ; Liang Qichao; and institutionalisation from 1913.
Fei : LSE under Malinowski; Peasant Life in China (1939) as an economic ethnography of a complex society by one of its members; From the Soil (1947); and a sixty-year longitudinal return to the same village.
The abolition : sociology eliminated by decree in 1952 , Fei denounced in 1957 and silenced for two decades, the discipline restored in 1979 with Fei leading the reconstruction at sixty-nine.
Guanxi chaxugeju operating — ganqing , renqing (where an exactly balanced exchange ends a relationship and an imbalance sustains it) and mianzi — with real economic effects, and a live debate (Guthrie against Bian) on whether it declines or changes function.
And the Weber question failed twice : Confucianism explained the absence of capitalism, then the "post-Confucian hypothesis" explained the fastest growth in history using the same variable. The institutional and historical explanations are the serious ones.
Chaxugeju (差序格局) — the differential mode of association: egocentric networks with obligation graded by distance.
Tuantigeju (团体格局) — the organisational mode: bounded groups with binary membership.
Qun / qunxue — Xunzi's concept of human grouping, and Yan Fu's term for sociology.
Wulun — the five cardinal Confucian relations.
Ren / li / xiao — humaneness, ritual propriety, filial piety.
Rectification of names — the doctrine that order depends on each role being filled according to its designation.
Jian ai — Mohist impartial care, the rival to graded obligation.
Guanxi — personal connections and their obligations.
Renqing — the obligation created by favours, sustained by remaining unbalanced.
Mianzi — face; independently developed and elaborated separately from Goffman's concept.
Post-Confucian hypothesis — the inverted Weber claim that Confucian values explain East Asian growth.
One — draw both structures. Sketch your own social world as bounded groups, then as concentric ripples centred on you. Note which captures more, and which parts each misses.
Two — test the graded obligation claim. List what you owe to a sibling, a colleague, a neighbour and a stranger. If the amounts differ continuously rather than by category, you are operating a differential structure.
Three — find the renqing mechanism. Identify a relationship you maintain through small unreturned favours in both directions. Ask what would happen if it were exactly settled.
Four — check the growth argument. Find one institutional explanation of East Asian industrialisation — land reform, state-directed credit, export orientation. Compare its specificity with the cultural explanation's.
Five — read From the Soil . It is short, translated, written for newspaper readers, and it contains a piece of general social theory that your theory course did not teach you.
The next tradition developed under the most complete colonial disruption of any in this Part, has had to argue about whether its own core concepts are recoverable or inventions, and has produced the sharpest available account of how colonial rule constructed the categories through which its societies are still governed.
Topic 6.4 opens with 6.4.1 — African Social Thought: Ubuntu, the Invented Tribe and the Bifurcated State.