In short, anticommercialism and xenophobia won out, and China re-tired from the world scene. The military declined and bureaucrats ranthe show except when the big eunuch establishment, handling surveil-lance and investigation for the emperor, produced from time to time, un-der weak rulers, eunuch dictatorships that terrorized the scholars. Thecontradiction between Ming China’s superior capacity for maritime ex-pansion and conservative Neo-Confucian throttling of it suggests thatMing China almost purposely missed the boat of modern technologicaland economic development.This disparaging judgment comes out of the context of the late twen-tieth century, when technology and growth have created innumerabledisorders in all aspects of life all over the world without disclosing asyet the principles of order that may postpone the destruction of humancivilization. In time the self-contained growth of Ming China with its

I think this has the added benefit of making data warehousing ETL much more organizationally scalable. The classic problem of the data warehouse team is that they are responsible for collecting and cleaning all the data generated by every other team in the organization. The incentives are not aligned: data producers are often not very aware of the use of the data in the data warehouse and end up creating data that is hard to extract or requires heavy, hard to scale transformation to get into usable form. Of course, the central team never quite manages to scale to match the pace of the rest of the organization, so data coverage is always spotty, data flow is fragile, and changes are slow.

I'll give a little bit of the history to provide context. My own involvement in this started around 2008 after we had shipped our key-value store. My next project was to try to get a working Hadoop setup going, and move some of our recommendation processes there. Having little experience in this area, we naturally budgeted a few weeks for getting data in and out, and the rest of our time for implementing fancy prediction algorithms. So began a long slog. We originally planned to just scrape the data out of our existing Oracle data warehouse. The first discovery was that getting data out of Oracle quickly is something of a dark art. Worse, the data warehouse processing was not appropriate for the production batch processing we planned for Hadoop—much of the processing was non-reversable and specific to the reporting being done. We ended up avoiding the data warehouse and going directly to source databases and log files. Finally, we implemented another pipeline to load data into our key-value store for serving results. This mundane data copying ended up being one of the dominate items for the original development. Worse, any time there was a problem in any of the pipelines, the Hadoop system was largely useless—running fancy algorithms on bad data just produces more bad data. Although we had built things in a fairly generic way, each new data source required custom configuration to set up. It also proved to be the source of a huge number of errors and failures. The site features we had implemented on Hadoop became popular and we found ourselves with a long list of interested engineers. Each user had a list of systems they wanted integration with and a long list of new data feeds they wanted. ETL in Ancient Greece. Not much has changed.

Effective use of data follows a kind of Maslow's hierarchy of needs. The base of the pyramid involves capturing all the relevant data, being able to put it together in an applicable processing environment (be that a fancy real-time query system or just text files and python scripts). This data needs to be modeled in a uniform way to make it easy to read and process. Once these basic needs of capturing data in a uniform way are taken care of it is reasonable to work on infrastructure to process this data in various ways—MapReduce, real-time query systems, etc. It's worth noting the obvious: without a reliable and complete data flow, a Hadoop cluster is little more than a very expensive and difficult to assemble space heater. Once data and processing are available, one can move concern on to more refined problems of good data models and consistent well understood semantics. Finally, concentration can shift to more sophisticated processing—better visualization, reporting, and algorithmic processing and prediction. In my experience, most organizations have huge holes in the base of this pyramid—they lack reliable complete data flow—but want to jump directly to advanced data modeling techniques. This is completely backwards. So the question is, how can we build reliable data flow throughout all the data systems in an organization?

You don't hear much about data integration in all the breathless interest and hype around the idea of big data, but nonetheless, I believe this mundane problem of "making the data available" is one of the more valuable things an organization can focus on.

Data integration is making all the data an organization has available in all its services and systems.

Behind thislay the Confucian disdain for the military, which classed them evenlower than merchants. So deep-laid was this dislike that the militarywere excluded from the standard Confucian list of the four occupationalgroups or classes—scholar(shi),farmer(nong),artisan(gong),and mer-chant(shang).Derk Bodde (1991) tells us that this four-part division ofsociety was never put forward by Confucius or Mencius but first ap-peared probably among Legalist writers of late Zhou and early Han. Fortwenty-one centuries since that time, however, the four classes have beenstandard fare in the lore about China.

For the scholar-elite, Zhu Xi promoted academies. He had contactwith about 24 such unofficial institutions and taught 20 students in hisown. The object of this teaching was the individual, who must learn howto achieve his own grasp of morality and bear responsibility for hismoral self-cultivation in his effort to become a sage. Zhu hoped thatproper government, finally, would rest upon “universal self-disciplinebeginning with the ruler’s self-rectification.” This could be aided byscholars’ lectures to him (as part of court ritual) as well as by the subse-quent judgments of the court historians. In discussing moral questions,minister and emperor should talk as equals.

Through discussion in these Community Compact meetings, gooddeeds could be praised, errors corrected, and rites and customs pre-served. Zhu saw this institution as fusing together private and public

To reach the populace, Zhu Xi used the vernacular and also advo-cated using the periodic local residents’ meeting known as the Commu-nity Compact(xiangyue).Although it came into general use only in theMing dynasty after 1368, this institution originated in a prototype estab-lished in 1077 by the Lü family. It consisted of a monthly assemblywhere food was eaten and a record of proceedings kept. One or twoheads were elected, and quite detailed regulations regarding behaviorwere adopted. Zhu Xi produced an amended version of the Lü family’sregulations that was even more detailed. It stressed hierarchy, for exam-ple by establishing five age-grades with rules for the conduct of all mem-bers of different categories. The aim was obviously to tell educated elitefamilies how to behave. Zhu assumed that the ordinary dress and majorrituals would be those of the elite. Seating by seniority would not applyto the non-elite, if they ventured to be present. Zhu Xi’s amendmentsalso contained detailed instructions on the way to greet a fellow compactmember, when to make calls on fellow compact members, how to invitethem to banquets, and how to conduct banquets—what to wear, whatname cards to use, and so on. What an organization man

In recent decades, after the collapse of the traditional society inwhich Neo-Confucianism was the principal faith among the elite, schol-ars of Chinese thought have again “repossessed the Way” through ap-preciation of Zhu Xi’s teachings.

Zhu Xi’s cosmology asserted a dualism, that the great immutableprinciples of form(li)give shape to the material stuff(qi)that, whenshaped byli,creates existent reality. Behind this duality, however, is theDao,the Way, the vast energizing force that pervades the universe and allthings in it. Only through disciplined self-cultivation could a man getsome understanding of the Way and in pursuit of it form his character.Song Neo-Confucianists believed that the true Way for the moral im-provement of the individual and the world had been set forth by Confu-cius and Mencius but had not been transmitted thereafter. Their aim fif-teen hundred years later was therefore to “repossess the Way.”In effect Zhu Xi found a means of smuggling a needed element ofBuddhist transcendentalism into Confucianism. This new philosophy,both eminently rational and humane, was promoted by its adherents as achallenge to the court and the literati to be less selfish and live up to theirConfucian ideals. Through the writing and teaching of this critical-minded minority, Neo-Confucianism became the living faith of China’selite down to the twentieth century, one of the world’s most widespreadand influential systems of ethics.

In effect Zhu Xi found a means of smuggling a needed element ofBuddhist transcendentalism into Confucianism. This new philosophy,both eminently rational and humane,

The established order should be repaired, to be sure, but not trans-formed by a blueprint. Landlords and tenants were a natural result ofthe difference in human abilities. The key function of the ruler was theselection of talent, which was to be found among the Confucian-trainedliterati.

The most famous and controversial Song reformer was Wang Anshi.Though his reforms have been variously regarded, the most recent anal-ysis sees him as a totalitarian-minded man ahead of his time. As a clas-sicist, he regarded China’s ancient sages down to Confucius as modelsof perfection whose intentions at least could still be followed. Wang’sNew Policies aimed to establish a “perfect, self-contained, and self-perpetuating system,” as Peter Bol (1992) phrases it. Backed by the em-peror from 1068, Wang bypassed the bureaucracy by getting his own

As we look at this recurrent aspect of Confu-cianism, two features may be noted: reform-advocating officials nor-mally hoped that the emperor would grant them power to make theirreforms. They assumed that the imperial autocracy was the origin of allpolitical power. They might strengthen it or use it but never sought to gobehind it and consider any other forms of authority in state and society.Second, would-be reformers regarded the great mass of the commonpeople as passive recipients of the benevolent despotism they sought toguide. They assumed merchants were perniciously addicted to greed andmilitary men to violence. The reformer’s job was to keep them in theirplace and secure a wise application of the unified central power repre-sented by the emperor. Viewed this way, reform was a high calling, ameans of preserving the imperial order and benefiting (while controlling)the mass of the people.

Holding office in the civil service had be-come only one factor—and not a necessary one—in establishing elite sta-tus. In other words, the elite had broadened out to include local mag-nates, family heads, and informal public servants as well as ex-officials.The prerequisite for all was a classical education that qualified one cul-turally as a member of the class ofshi—literati or “gentlemen.” Suchmen, by their Confucian training, felt a sense of responsibility to keepthe world materially and morally in order. They were guided by the creedof Neo-Confucianism, a philosophy of life that grew out of the debatesof scholar-officials in the Northern Song.

The examination system became an enormous and intricate institu-tion central to upper-class life. During a thousand years from the Tang to1905 it played many roles connected with thought, society, administra-tion, and politics.

Printed books gave a great impetus to the education carried on inBuddhist monasteries as well as within families. The government had atfirst tried to control all printing, which was widespread. But by the1020s it was encouraging the establishment of schools by awarding landendowments as well as books. The aim was to have a government schoolin every prefecture. The schools enrolled candidates, conducted Confu-cian rituals, and offered lectures. John W. Chaffee (1985) tells us that bythe early 1100s the state school system had 1.5 million acres of land thatcould provide a living for some 200,000 students.

By 1078 North China was producing an-nually more than 114,000 tons of pig iron (700 years later Englandwould produce only half that amount).

Theexhaustion of forest cover by aboutad1000 obliged iron smelters to usecoal instead of charcoal in coke-burning blast furnaces.

By the eighth century it appears that the holding of office had becomethe main criterion of family status, and the pedigrees of the great clanswere less important. Everything now depended upon the official rank ofthe person listed, not on his family origins. Legally, officials were nolonger regarded as a special elite. Though sons of officials were given aminor rank in the Tang legal code, there was no longer an upper-classstatus recognized in the code that gave special claims for appointment tooffice. The imperial institution had won out over the social interests ofthe great clans.

The lack of primogeniture in China meant that equal division ofproperty among males was the common practice when the head of afamily died. The imperial code of laws required partible inheritance andso prevented the rise of a landed nobility such as occurred in Europe. Ifa member of the family did not become an official for two or three gen-erations, the family would sooner or later disintegrate. Each generationwas potentially insecure and had to prove itself in official life.

As David Johnson (1977) says, “Unlike Englandor France, where a man could rise to a position of high social statusthrough a career in law, medicine, commerce, the church or the military,in China there was only one significant occupational hierarchy: the civilservice.”

The Buddha, who lived probably during the sixth centurybcin Nepal,began life as an aristocrat. After renouncing his palace and its haremand luxuries, he achieved through meditation an illumination in whichhe realized the great principle of the wheel of the law or the wheel of theBuddha. This may be defined as a theory of the “dependent origination”of life: that everything is conditioned by something else in a closed se-quence, so that in effect the misery of life is dependent upon certainconditions, and by eliminating these conditions it is possible to eliminatethe misery itself. Thus desire—which ultimately leads to misery—origi-nates in dependence upon sensation, which in turn originates in depen-dence upon contact and the six senses, and so on. The Buddhist objec-tive therefore becomes to cut the chain of conditions that bind one into

This line of thought raises a major question, the relationship be-tweenwenandwu. Wenmeans basically the written word and so byextension its influence in thought, morality, persuasion, and culture. Letus call it in the most general terms “the civil order.”Wuconnotes theuse of violence and so stands for the military order in general. TheConfucian-trained scholar class went to great lengths to exaltwenanddisparagewu.Yet I wonder ifwu(including the founding of dynasties,extermination of rebels and evildoers, and punishing of officials) shouldnot be considered the stronger andwenthe weaker element in thewen—wucombination. For example, was the virtue of loyalty (an aspect ofwen) as powerful as the practice of intimidation (an aspect ofwu)?Often it seemed that when he wanted to control a situation, the emper-or’s principal tactic was intimidation. Take for example the case of Chi-na’s greatest historian Sima Qian.

rated in 124bcand continued into the Southern Song) the “NationalUniversity,” or to call theGuozijian(from Song to Qing) the “Director-ate of Education.” Focused on the classics, these institutions mightequally well be called indoctrination centers. The fact remains that impe-rial power, books, and scholars were all seen as integrally related aspectsof government

Several assumptions seem implicit here. First is the emperor’s role asa source of spontaneous, irrational, or unpredictable acts, as opposed tothe routinized, predictable action (or inaction) of bureaucrats. The of-ficials sought order. The emperor could shake them up with disorder.Second, the emperor was considered to have an arbitrary and unbridledpower of life and death.

nized in gradations of inferiority and superiority. This hierarchic princi-ple in turn was the basis for a stress on duties rather than rights, on theevident assumption that if everyone did his duty everyone would getwhat he deserved. Thus, the filial son obedient to his parent would baskin the parent’s approval. With all duties performed, society would be inorder to everyone’s benefit.

Early China’s cosmology (her theory of the universe as an orderedwhole) shows striking points of difference with Western thought. Forexample, the early Chinese had no creation myth and no creator-lawgiver out of this world, no first cause, not even a Big Bang. As JosephNeedham says, they assumed “a philosophy of organism, an orderedharmony of wills without an ordainer.” This view contrasts with the in-veterate tendency elsewhere in the world to assume a supernatural deity.Westerners looking at China have continually imposed their own pre-conceptions on the Chinese scene, not least because the Chinese, thoughthey generally regarded Heaven as the supreme cosmic power, saw it asimmanent in nature, not as transcendent. Without wading further intothis deep water, let us note simply that Han thought as recorded in classi-cal writings built upon the concept of mankind as part of nature andupon the special relationship between the ruler and his ancestors, con-cepts that were already important in Shang thought over a millenniumearlier.

Confucius was neither out to become a ruler himself nor to educatethe masses directly. His priorities put proper ritual first, humaneness sec-ond, and learning only third.

The limitations of the Confucians’ status had been plain from thestart. Confucius had aimed to train an elite who would become superiormen, able both to secure the people’s respect and guide the ruler’s con-duct. Confucius was neither out to become a ruler himself nor to educatethe masses directly. His priorities put proper ritual first, humaneness sec-ond, and learning only third. By his example he showed the way for hisown kind, who would later be the scholar-officials of the imperial era.China’s social structure, in short, was already in place and the philoso-pher’s task in his Chinese form of prophecy was not to arouse the massesbut only to guide the rulers. As W. T. de Bary (1991) points out, the Con-fucians did not try to establish “any power base of their own...theyfaced the state, and whoever controlled it in the imperial court, as indi-vidual scholars...this institutional weakness, highly dependent condi-tion, and extreme insecurity . . . marked the Confucians asju[ru] (‘soft-ies’) in the politics of imperial China.” They had to find patrons whocould protect them. It was not easy to have an independent voice sepa-rate from the imperial establishment.

The Han rulers’ daily regimen of ceremonies and rites required the guid-ance of learned men at court. Han Wudi in particular fostered learningas one channel (in addition to recommendation) for recruitment of of-ficials. He saw education as a way to strengthen his new upper classagainst the older aristocratic families, and he accepted Confucianism asthe ideology in which the state’s officials should be trained. To the des-potic statecraft of Qin Legalism the Han added a monumental structureof ideas of largely Confucian origin that provided an all-encompassingstate philosophy. This Legalist-Confucian amalgam we call ImperialConfucianism, to distinguish it both from the original teaching of Con-fucius, Mencius, et al. and from the secular and personal Confucian phi-losophy that arose during Song times and has since then guided so manylives in the East Asian countries of the old Chinese culture area—China,Korea, Vietnam, and Japan.

At the capital arose a similar problem for the Han ruler: how toavoid domination of the court by the family of an empress. When a Hanemperor died, power resided in his widow, the empress dowager, to ap-point her husband’s successor from the Liu clan (the clan of the Han em-perors). She might appoint a minor of the Liu clan as emperor, and astrong-man regent of her own clan to rule for him. Half a dozen familiesof empresses played this game. An emperor, however, could rely withinthe palace on the staff of eunuchs, whose castration fitted them to lookafter the women selected for the emperor’s harem. There, by having sev-eral sons, the emperor hoped to find one worth selecting as his successor.Eunuchs, being entirely dependent on a young emperor as his servantsand companions, might be his only reliable supporters against an em-press’s family. The palace was a center of intrigue.

Along with this military—liturgical basis for state power and socialorder, the Warring States also fostered, rather paradoxically, an age ofphilosophers who sought theoretical bases for those same things. Dur-ing this time of rivalry and warfare, there was a widespread yearningfor peace and order. Many people idealized a golden age of earliertimes when according to legend all China had lived peacefully under oneruler. Violence inspired the late Zhou philosophers, who acted as whatwe now call consultants, advising rulers on how to get back to thegolden age.Confucius (551–479bc)and his major disciple, Mencius (372–289bc),were members of a considerable group of seminal thinkers in thisera.

Yet among Euro-pean dynasties such as the Capetian kings of France (987–1328), theNorman and Plantagenet kings of England (1066–1485), the Hapsburgs(1273–1919), or the Romanovs (1613–1917), none ruled as large a stateas China or maintained such a monopoly of central power. As institu-tions of government, the major Chinese dynasties are in a class by them-selves. Neither Japan, India, nor Persia produced regimes comparable inscope and power. The Liu clan of Earlier Han provided 13 emperors andin Later Han 14 emperors, the Li clan of the Tang dynasty 23 emperors,the Zhu clan of the Ming dynasty 17 emperors, and the Aisin Gioro lin-eage of the Manchus 9 emperors (see Table 1).

A major Confucian principle was that man was perfectible. In the eraof Warring States, Chinese thinkers of the major schools had turnedagainst the principle of hereditary privilege, invoked by the rulers ofmany family-states, and stressed the natural equality of men at birth.Mencius’ claim that men are by nature good and have an innate moralsense won general acceptance. They can be led in the right path througheducation, especially through their own efforts at self-cultivation, butalso through the emulation of models. The individual, in his own effortto do the right thing, can be influenced by the example of the sages andsuperior men who have succeeded in putting right conduct ahead of allother considerations. This ancient Chinese stress on the moral educabil-ity of man has persisted down to the present and still inspires the gov-ernment to do the moral educating.The Confucian code also stressed the idea of “proper behavior ac-cording to status”(li).The Confucian gentleman (“the superior man,”“the noble man”) was guided byli,the precepts of which were written inthe ancient records that became the classics. Although this code did notoriginally apply to the common people, whose conduct was to be regu-lated by rewards and punishments (stressed by the Legalist school),rather than by moral principles, it was absolutely essential for govern-ment among the elite. This was the rationale of Confucius’ emphasis onright conduct on the part of the ruler—an emphasis so different fromanything in the West. The main point of this theory of government bygood example was the idea of the virtue that was attached to right con-duct. To conduct oneself according to the rules of propriety orliin itselfgave one a moral status or prestige. This moral prestige in turn gave oneinfluence over the people. “The people are like grass, the ruler like thewind”; as the wind blew, so the grass was inclined. Right conduct gavethe ruler power.

fined by authority guided the individual’s conduct along lines of properceremonial behavior. Confucius had said (rather succinctly), “jun junchen chen fu fu zi zi,” which in its context meant “Let the ruler rule as heshould and the minister be a minister as he should. Let the father act as afather should and the son act as a son should.” If everyone performed hisrole, the social order would be sustained. Being thus known to others bytheir observable conduct, the elite were dependent upon the opinion andmoral judgment of the collectivity around them. To be disesteemed bythe group meant a disastrous loss of face and self-esteem, for which oneremedy was suicide.

Confucianism’s rationale for organizing society began with the cosmicorder and its hierarchy of superior-inferior relationships. Parents weresuperior to children, men to women, rulers to subjects. Each persontherefore had a role to perform, “a conventionally fixed set of social ex-pectations to which individual behavior should conform,” as ThomasMetzger (in Cohen and Goldman, 1990) puts it. These expectations de-

Overstated though these considerations may be, they represent agreat fact emerging from Chinese archaeology—that by the beginning ofthe era of written history, the Chinese people had already achieved a de-gree of cultural homogeneity and isolated continuity hard to match else-where in the world. They had begun to create a society dominated bystate power. To it all other activities—agricultural, technological, com-mercial, military, literary, religious, artistic—would make their contribu-tions as subordinate parts of the whole. Yet it would be an error for ustoday, so long accustomed to the modern sentiment of nationalism, toimagine ancient China as an embryonic nation-state. We would do betterto apply the idea of culturalism and see ancient China as a completecivilization comparable to Western Christendom, within which nation-states like France and England became political subunits that sharedtheir common European culture. Again, because we are so aware of theall-encompassing power of the totalitarian states of the twentieth cen-tury, we would do well to avoid an anachronistic leap to judgment thatthe Shang and Zhou kings’ prerogatives led inevitably to a sort of totali-tarianism. We might better follow Etienne Balazs (1964), who called it agovernment by “officialism.” As summarized by Stuart Schram (1987),“The state was the central power in Chinese society from the start, andexemplary behavior, rites, morality and indoctrinations have alwaysbeen considered in China as means of government.” We need only addthat in addition to these liturgical functions the ruler monopolized theuse of military violence.

China of coursewas not alone in idealizing this kind of unity, which was sought in manyof the ancient empires. But China’s geographic isolation made the ideal

We must not over-look the ancient Chinese assumption of a symbiosis between culture(wenhua)and temporal power. Subservience to the dynastic state re-quired acceptance of its rituals and cosmology that gave it Heaven’smandate to rule over mankind. Nonacceptance of this politicized cultureleft one outside of Zhongguo. Yet if one’s language was Chinese, accep-tance was already partway assured by the very terms imbedded in theclassics and in the spoken tongue itself. An identifiably similar way of lifewas widespread throughout late Neolithic China. The task of state-building during the Three Dynasties of the Bronze Age was to gain everwider submission to or acceptance of the central dynastic ruling house. Itfunctioned as the capstone of the social structure, the high priesthood ofthe ancestor cult, the arbiter of punishments, and the leader in publicworks, war, and literature. Among these omnicompetent functions K. C.Chang stresses the ruler’s “exclusive access to heaven and heavenly spir-its.” The result was that the ruler engineered a unity of culture that wasthe basis for political unity in a single universal state.

Finally, the ruler’s primacy rested on his monopoly of leadership notonly in ritual and warfare but also in oracle-bone writing and the histor-ical learning it recorded.

For example, take the character for east , which in the Beijingdialect has the sound “dong” (pronounced “doong,” as in Mao Ze-dong’s name). Since a Chinese character is read aloud as a single syllableand since spoken Chinese is also rather short of sounds (there are onlyabout four hundred different syllables in the whole language), it hasbeen plagued with homophones, words that sound like other words, like“soul” and “sole” or “all” and “awl” in English. It happened that thespoken word meaning “freeze” had the sound “dong.” So did a spokenword meaning a roof beam. When the Chinese went to write down thecharacter for freeze, they took the character for east and put beside itthe symbol of ice , which makes the character (“dong,” to freeze).To write down the word sounding “dong” which meant roof beam, theywrote the character east and put before it the symbol for wood mak-ing (“dong,” a roof beam).These are simple examples. Indeed, any part of the Chinese languageis simple in itself. It becomes difficult because there is so much of it to beremembered, so many meanings and allusions. When the lexicographersof later times wanted to arrange thousands of Chinese characters in adictionary, for instance, the best they could do in the absence of an al-phabet was to work out a list of 214 classifiers or “radicals,” one ofwhich was sure to be in each character in the language. These 214classifiers, for dictionary purposes, correspond to the 26 letters of our al-phabet, but are more ambiguous and less efficient. Shang writing was al-ready using “radicals” like wood, mouth, heart, hand, that indicatedcategories of meaning. From the start the governmental power of theChinese writing system was at the ruler’s disposal. Writing seems to haveemerged more in the service of lineage organization and governmentthan in the service of trade.

Here lies one source of China’s “culturalism”—that is, the devotionof the Chinese people to their way of life, an across-the-board sentimentas strong as the political nationalism of recent centuries in Europe.Where European nationalism arose through the example of and contactwith other nation-states, Chinese culturalism arose from the differencein culture between China and the Inner Asian “barbarians.” Because theInner Asian invaders became more powerful as warriors, the Chinesefound their refuge in social institutions and feelings of cultural and aes-thetic superiority—something that alien conquest could not take away.

The contrast between Inner Asia and China proper is a striking onein nearly every respect. On the steppe, population is thinly scattered; to-day there are only a few million Mongols and hardly more than thatnumber of Tibetans in the arid plateau regions that more than equal thearea occupied by over a billion Chinese who trace their ancestry to theHan dynasty (see Table 1). The thinness of population in Inner Asia initself makes the life of the steppe nomads vastly different from thecrowded life of the Han Chinese.

Contrary to a common myth, a large family with several children hasnot been the norm among Chinese peasants. The scarcity of land, as wellas disease and famine, set a limit to the number of people likely to sur-vive in each family unit. The large joint family of several married sonswith many children all within one compound, which has often been re-garded as typical of China, appears to have been the ideal exception, aluxury that only the well-to-do could afford. The average peasant familywas limited to four, five, or six persons in total. Division of the landamong the sons constantly checked the accumulation of property andsavings, and the typical family had little opportunity to rise on the socialscale. Peasants were bound to the soil not by law and custom so much asby their own numbers.

While the family headship passes intact from father to eldest son, thefamily property does not. Early in their history the Chinese abandonedprimogeniture, by which the eldest son inherits all the father’s propertywhile the younger sons seek their fortunes elsewhere. The enormous sig-nificance of this institutional change can be seen by comparing Chinawith a country like England or Japan, where younger sons who have notshared their father’s estate have provided the personnel for government,business, and overseas empire and where a local nobility might grow upto challenge the central power. In China, the equal division of landamong the sons of the family allowed the eldest son to retain only certainceremonial duties, to acknowledge his position, and sometimes an extrashare of property. The consequent parcelization of the land tended toweaken the continuity of family land-holding, forestall the growth oflanded power among officials, and keep peasant families on the marginof subsistence. The prime duty of each married couple was to produce ason to maintain the family line, yet the birth of more than one son mightmean impoverishment.

The traditional family system was highly successful at preparing theChinese to accept similar patterns of status in other institutions, includ-ing the official hierarchy of the government. The German sociologistMax Weber characterized China as a “familistic state.” One advantageof a system of status is that a man knows automatically where he standsin his family or society. He can have security in the knowledge that if hedoes his prescribed part, he may expect reciprocal action from others inthe system.Within the extended family, every child from birth was involved in ahighly ordered system of kinship relations with elder brothers, sisters,maternal elder brothers’ wives, and other kinds of aunts, uncles, cousins,grandparents, and in-laws too numerous for a Westerner to keep trackof. These relationships were not only more clearly named and differenti-ated than in the West but also carried with them more compelling rightsand duties dependent upon status. Family members expected to be calledby the correct term indicating their relationship to the person addressingthem.In South China the pioneer anthropologist Maurice Freedman(1971) found family lineages to be the major social institutions—eachone a community of families claiming descent from a founding ancestor,holding ancestral estates, and joining in periodic rituals at graves and inancestral halls. Buttressed by genealogies, lineage members might sharecommon interests both economic and political in the local society. InNorth China, however, anthropologists have found lineages organizedon different bases. Chinese kinship organization varies by region. Familypractices of property-holding, marriage dowries, burial or cremation,and the like also have had a complex history that is just beginning to bemapped out.

In addition to this common bond of loyalty to family, the old Chinawas knit together by the common experience of a highly educated localelite, who were committed from childhood to studying and following theclassical texts and teachings. Motherly nurture and fatherly disciplinecombined to concentrate the young scholar’s effort on self-control andon the suppression of sexual and frivolous impulses. Instead, as JonSaari’s (1990) study of upper-class childhood in the late nineteenth cen-tury reiterates, the training of youth was in obedience above all. Once aboy entered his adolescent years, open affection from parents gave wayto intensive training aimed at proper character formation.

The inferior social status of women was merely one manifestation ofthe hierarchic nature of China’s entire social code and cosmology. An-cient China had viewed the world as the product of two interacting com-plementary elements,yinandyang. Yinwas the attribute of all things fe-male, dark, weak, and passive.Yangwas the attribute of all things male,bright, strong, and active. While male and female were both necessaryand complementary, one was by nature passive toward the other. Build-ing on such ideological foundations, an endless succession of Chinesemale moralists worked out the behavior pattern of obedience and passiv-ity that was expected of women. These patterns subordinated girls toboys from infancy and kept the wife subordinate to her husband and themother to her grown son. Forceful women, whom China has neverlacked, usually controlled their families by indirection, not by fiat.Status within the family was codified in the famous “three bonds”emphasized by the Confucian philosophers: the bond of loyalty on thepart of subject to ruler (minister to prince), of filial obedience on the partof son to father (children to parents), and of chastity on the part of wivesbut not of husbands. To an egalitarian Westerner the most striking thingabout this doctrine is that two of the three relationships were within thefamily, and all were between superior and subordinate. The relationshipof mother and son, which in Western life often allows matriarchal domi-nation, was not stressed in theory, though it was naturally important infact.When a father saw the beginning of individuality and independencein his son, he might fear that selfish personal indulgence would disruptthe family. Strong bonds of intimacy between mother and son or son andwife threatened the vertical lines of loyalty and respect that maintainedthe family and the father’s authority. In Jonathan Ocko’s summary (inKwang-Ching Liu, 1990), wives were “ineluctably destabilizing ele-ments,” promising descendants, yet always threatening the bond of obe-dience between parents and sons.

The trembling bride left herown family behind and became at once a daughter-in-law under the con-trol of her husband’s mother. She might see secondary wives or concu-bines brought into the household, particularly if she did not bear a maleheir. She could be repudiated by her husband for various reasons. If hedied, she could not easily remarry. All this reflected the fact that awoman had no economic independence. Her labor was absorbed inhousehold tasks and brought her no income. Farm women were almostuniversally illiterate. They had few or no property rights.

To Americans and Europeans with their higher material standard ofliving, the amazing thing about the Chinese farming people has beentheir ability to maintain a highly civilized life under these poor condi-tions. The answer lies in their social institutions, which have carried theindividuals of each family through the phases and vicissitudes of humanexistence according to deeply ingrained patterns of behavior. These insti-tutions and behavior patterns are among the oldest and most persistentsocial phenomena in the world. China has been a stronghold of the fam-ily system and has derived both strength and inertia from it.Until very recently the Chinese family has been a microcosm, thestate in miniature. The family, not the individual, was the social unit andthe responsible element in the political life of its locality. The filial pietyand obedience inculcated in family life were the training ground for loy-alty to the ruler and obedience to the constituted authority in the state.This function of the family to raise filial sons who would becomeloyal subjects can be seen by a glance at the pattern of authority withinthe traditional family group. The father was a supreme autocrat, withcontrol over the use of all family property and income and a decisivevoice in arranging the marriages of the children. The mixed love, fear,and awe of children for their father was strengthened by the great respectpaid to age. An old man’s loss of vigor was more than offset by hisgrowth in wisdom. As long as he lived in possession of his faculties, thepatriarch had every sanction to enable him to dominate the family scene.According to the law, he could sell his children into slavery or even exe-cute them for improper conduct. In fact, Chinese parents were by customas well as by nature particularly loving toward small children, and theywere also bound by a reciprocal code of responsibility for their childrenas family members. But law and custom provided little check on paternaltyranny if a father chose to exercise it.The domination of age over youth within the old-style family wasmatched by the domination of male over female. Even today, Chinesebaby girls seem more likely than baby boys to suffer infanticide.

This different relation of human beings to nature in the West andEast has been one of the salient contrasts between the two civilizations.Man has been at the center of the Western stage. The rest of nature hasserved as either neutral background or as an adversary. Thus Western re-ligion is anthropomorphic, and early Western painting anthropocentric.To see how great this gulf is, we have only to compare Christianity withthe relative impersonality of Buddhism, or compare a Song landscape, itstiny human figures dwarfed by crags and rivers, with an Italian primi-tive, in which nature is an afterthought.Living so closely involved with family members and neighbors hasaccustomed the Chinese people to a collective life in which the groupnormally dominates the individual. In this respect the Chinese experi-ence until recently hardly differed from that of other farming peopleslong settled on the land. It is the modern individualist, be he seafarer, pi-oneer, or city entrepreneur, who is the exception. A room of one’s own,more readily available in the New World than in the crowded East, hassymbolized a higher standard of living. Thus, one generalization in thelore about China is the absorption of the individual not only in the worldof nature but also in the social collectivity.

Relatedly, Castenada et al (2003) develop a very rich overlapping-generation modelwith life-cycle features, borrowing constraints as in Aiyagari-Bewley models, constant re-tirement and death probabilities independent of age, perpetual youth demographics withaccidental inheritances, pensions, income and estate taxes, and persistent stochastic la-bor endowments. They carefully calibrate their model to the US economy to generatean excellent match to wealth distributions. Their 4-state labor endowment stochasticprocess is however highly skewed, so that at the stationary distribution for labor en-dowments, the top 0.039% earners have 1000 times the average labor endowment of thebottom 61%. Thus to attain a ratio of a 1000, if the bottom 61% earn $32,000 a year onaverage37, the top 0.0389% would have to have earnings of $32,000,000, quite excessiveaccording to the WWID. For top incomes from all sources the WWID gives, excludingcapital gains, an average income for the top 0.01% in 2014 of $17,180,000 and earnings(wages, pensions and salaries) of about $6,000,000, which of course would be lower forthe top 0.0389%. Life cycle/consumption-smoothing considerations arise naturally, soagents at the rare and somewhat persistent highest labor endowment state, sometimescalled the "awesome" state, save at higher rates and accumulate wealth faster. Theseagents decumulate during retirement, but invariably some fraction die early into theirretirement, leaving large accidental bequests. A similar mechanism is at work in Diazet al (2003) who use a standard Aiyagari model with inÖnitely lived agents and threeearnings states. At the top "awesome" state, the top 6% of the population earn 46 timesthe labor earnings of the median, a highly excessive skew relative to the earnings data.According again to the WWID, the 5% of top incomes average $367,100 in 2013, of whichonly 69.44% or $255,000 are labor earnings (wages, salaries plus pensions). The medianincomes however are about $50,000, a factor of 5.1, not 46.

For example Krueger and Kindermann (2014) show that a version of the Aiyagari-Bewley model with su¢ ciently skewed earnings can be made to match the empiricalwealth distribution. But this requires excessive and empirically unrealistic tails for thedistribution of earnings. Krueger and Kindermann (2014) use a seven state Markov chainmodel for earnings that implies, in the stationary distribution, that the average top 0.25%earn somewhere between 400 to 600 times the median income, implying earnings for thetop 0.25% of at least $20,000,000

All this implies that realistic distributions of earnings by themselves, without othercomplimentary mechanisms, have di¢ culty in generating the skewed wealth distributionswe observe