Xia Dynasty

← Encyclopedia

The Xia is the first of the dynasties listed in the traditional account of Chinese history, said to have ruled the central plains of the Yellow River before the Shang. The later histories — above all Sima Qian’s Records of the Grand Historian, compiled around the first century BCE — give it a founder, a line of kings, and a span of several centuries placed roughly in the early second millennium BCE. What those histories report and what can be independently established are not the same thing, and the gap between them is the central problem of the Xia. It is the place in the long Chinese dynastic sequence — the sweep itself belongs to China — where the record and the ground first part company, and where the question of how a civilization came to write its own beginning is posed in its sharpest form.

The source and the king-list

Nearly everything the tradition says about the Xia descends from a single channel. Sima Qian, the Grand Astrologer of the Han court, set the dynasty at the head of his basic annals, after the era of the legendary Five Emperors and before the Shang and the Zhou. His chapter gives the Xia a founder in Yu the Great, a succession of kings running from Yu through some seventeen reigns, and a span of several centuries — a continuous line with names, an order, and an ending. The Records was assembled around the close of the second century BCE and finished early in the first, which is to say more than a thousand years after the events it narrates for the Xia. For the Shang, the comparable distance was later partly closed: the Anyang oracle bones preserved a royal genealogy that matched, with striking fidelity, the succession Sima Qian had recorded centuries afterward, vindicating the transmitted memory for that dynasty. The Xia has had no such confirmation. Its king-list rests on the credit the tradition earned elsewhere, not on a check of its own.

A second, older text complicates rather than settles the picture. The Bamboo Annals — a Warring States chronicle of the state of Wei, written on bamboo strips and recovered from a royal tomb when it was opened in 281 CE — preserves its own Xia chronology, and where it can be compared with the Records the two diverge. They differ on dates and, more tellingly, on the manner of the succession at the dynasty’s hinge. The discrepancy is itself instructive: even within the received tradition there was no single agreed account of how the Xia began, only competing reconstructions of a past already remote to the men who wrote them down.

Yu the Great and the water

The dynasty’s defining story is a story about water. The founder, Yu the Great, is remembered for mastering the great floods of the Yellow River where his father, Gun, had failed — and the contrast between the two is the whole moral of the tale. Gun had tried to dam the waters, raising barriers to hold them back, and the barriers gave way; Yu read the river instead of fighting it, dredging channels and cutting outlets so the flood could run to the sea. The labor is told as the measure of the man. He is said to have worked for thirteen years without returning home, passing his own door and not entering, even hearing his infant son cry within and walking on. The water in the story is not mere catastrophe but disorder itself, and Yu’s dredging is an ordering of the world: he traced the courses of the rivers, divided the drained land into nine provinces, and made the realm habitable and legible at once. In one strand of the legend he gathered the metal of those nine provinces and cast nine bronze cauldrons — the jiu ding — whose possession became the palladium of legitimate rule, handed down, the later texts say, from the Xia to the Shang to the Zhou and lost only in the chaos that ended the classical age. The ordering of the waters, the partition of the land, and the casting of the bronze are one continuous act: the founding of order, rendered in the very metal that defines the Chinese Bronze Age.

There is a modern footnote to the flood that has stirred its own debate. In 2016 a team led by the geologist Qinglong Wu published in Science the case for a real catastrophe behind the legend: sediment from a vast lake dammed by a landslide in the Jishi Gorge of the upper Yellow River, breached around 1920 BCE in an outburst flood, dated by the radiocarbon ages of children killed in the same earthquake at the Lajia site downstream. The authors read the event as a physical referent for the Great Flood and as support for a Xia beginning near 1900 BCE. The claim drew immediate published rebuttals in the same journal, disputing the dating, the flood’s scale, and above all the leap from a geological event to a named dynasty. The argument is a clean instance of the whole Xia problem in miniature: a real trace in the ground, a story in the texts, and a contested bridge between them.

The Confucian hinge

In the traditional reckoning the throne came to Yu as a reward for his service to the people — the last in a chain of sage-rulers who chose worth over blood, the aged Yao having passed power to Shun and Shun to Yu, each selecting the most capable man rather than his own son. With Yu the chain broke. At his death the rule passed not to a chosen successor but to his son, Qi, and with that the age of abdication gave way to the age of dynasty. The texts handle the transition delicately and not in unison. Sima Qian’s version is diplomatic: Yu had designated worthy ministers as his heirs, first Gao Yao and then Yi, but the people, out of love for Yu, turned to his son, and Yi yielded the throne to Qi of their own accord. The Bamboo Annals are blunter, telling of Qi taking the throne and the displaced heir set aside. The divergence preserves the discomfort of the moment itself. This is the point where, in the Confucian retelling, sage-kingship became hereditary monarchy — the loss of an ideal that Confucianism would hold up ever after as a measure against which later rule fell short, the abdications of Yao and Shun figuring in the canon as the model of government by virtue rather than descent.

The dynasty’s end is told as the mirror of its beginning. The last Xia king, Jie, is cast as a tyrant — cruel, dissolute, deaf to remonstrance — whose misrule exhausted the patience of Heaven and the loyalty of his people, until the Shang leader Tang rose against him and overthrew the house. The pattern is the founding instance of a template that would govern Chinese political thought for three thousand years: a dynasty falls because its last ruler has forfeited the right to rule, and the conqueror is not a usurper but an instrument of a sanction withdrawn. The framing of that sanction as the Mandate of Heaven, and its full articulation as a doctrine, belong to the Zhou, who would justify their own conquest of the Shang in exactly these terms and project the logic backward onto Jie. The Xia furnished the first fallen house of the sequence — the original tyrant, the original just rebellion — long before the doctrine that read it was composed.

The silence and the ground

Whether the Xia existed as the texts describe is unsettled, and the difficulty is of a particular kind: it is an absence of writing. No inscription has been recovered that can be assigned to the Xia, and none of the contemporary records of its supposed successor names it. The oracle-bone inscriptions of the Shang, which do survive in tens of thousands and which faithfully preserve the Shang’s own royal ancestors, offer no clear notice of a Xia predecessor — no conquered dynasty, no displaced line. The earliest securely datable Chinese writing is Shang and late; behind it the record is mute. A dynasty known only from texts set down a millennium and more after the fact, and unattested in the writing of the age that should have remembered it, is exactly the kind of object that historiography learns to hold at arm’s length.

Against the silence stands the ground. In 1959 the archaeologist Xu Xusheng, searching expressly for the Xia, located a large Bronze Age settlement at Erlitou, near Yanshi in the Luoyang basin of Henan — in roughly the time and place the texts assign the dynasty. Excavation over the following decades revealed a site of the first rank: rammed-earth palace foundations laid out on a planned axis, the earliest such monumental architecture known in China; a bronze foundry casting the earliest Chinese ritual vessels, the small jue cups and the first ding in metal; workshops in turquoise and jade; and, among the finds, a dragon some seventy centimeters long fashioned from more than two thousand pieces of inlaid turquoise, lifted in 2002 from the tomb of a member of the elite. Radiocarbon places the Erlitou culture at roughly 1750–1530 BCE, and the conventional four-phase scheme spans about 1900–1500 BCE. It is, by any measure, an early state — a centralized polity with a ruling elite, craft monopolies, and a ceremonial center commanding a wide hinterland — and it occupies precisely the slot the Xia is supposed to fill.

What it does not contain is its own name. No writing has been found at Erlitou that would let the site speak for itself, and the identification with the Xia must therefore be argued rather than read. Many Chinese scholars make that argument, equating Erlitou — or its later phases — with the Xia, or with the Xia and the earliest Shang in succession, on the strength of the fit in time and place and the authority of the transmitted record. Others resist. The objection, pressed most influentially in English by the historian Sarah Allan, is methodological: an archaeological culture and a dynasty named in much later texts are categorically different things, and to call the one by the other’s name is to read the Records onto the ground rather than to test them against it. On this view Erlitou is a primary state worth understanding on its own terms, and pinning a textual label to it adds nothing but the illusion of confirmation. The dispute runs along a fault line between scholarly traditions — a culture-historical archaeology accustomed to matching artifacts to named peoples, and a processual one that infers no ethnic or dynastic identity from material remains alone. The state’s Xia–Shang–Zhou Chronology Project, directed by Li Xueqin, lent the conventional dates the weight of official consensus, placing the Xia at roughly 2070–1600 BCE; but consensus on a date is not evidence of identity, and the central question is untouched by it. The matter carries weight beyond the academy, bound up with how far China’s documented history can be said to extend and with a national stake in its antiquity.

The textual and scholarly record

What can be known of the Xia is, in the first place, a problem in the reading of texts. The primary source remains Sima Qian’s Records of the Grand Historian (Shiji), the Han history that gave the dynasty its shape; against it the recovered Bamboo Annals furnish the one substantial independent chronicle, and the divergence between them frames every modern reconstruction. The competing chronologies make the uncertainty concrete: the Han scholar Liu Xin’s calculation placed the Xia at about 2205–1766 BCE; the Bamboo Annals yield roughly 1989–1558; the astronomer David Pankenier and the sinologist David Nivison, arguing from recorded planetary conjunctions, proposed about 1953–1555; and the Xia–Shang–Zhou Chronology Project settled on 2070–1600. That four serious efforts span more than two centuries at each end measures how much the dynasty’s dates are reconstructed rather than recorded.

The archaeological literature is younger and turns on the Erlitou site. Sarah Allan’s Erlitou and the Formation of Chinese Civilization (Journal of Asian Studies, 2007) is the standard English statement of the case for treating the culture on its own terms rather than as the Xia under another name; the dispute between Chinese and overseas scholars over the identification is surveyed at chapter length in recent collected work. The geological turn is represented by Qinglong Wu and colleagues’ Outburst flood at 1920 BCE (Science, 2016) and the rebuttals it provoked in the same journal — a debate conducted in radiocarbon and sediment rather than in philology, but circling the same question. A further line of work approaches the chronology through the astronomy embedded in the chronicles themselves: David Pankenier’s identification of a five-planet conjunction in 1953 BCE, and David Nivison’s reconstruction of the Bamboo Annals around it, are set out and reassessed in recent scholarship that treats the dynasty’s dating as a problem in historical method rather than a fact to be looked up. For the dynasty’s place in the longer arc of early China — the late Neolithic chiefdoms of the central plain that preceded it, and the divinatory and ritual culture of the literate Shang that followed — the fuller treatment lies under China and the Yijing. What sets the Xia apart from every later house of the sequence is starkly simple: it is the one dynasty for which not a single line written in its own time has ever come to light, so that everything said of it is said from the outside, by the ages that remembered it.

The afterlife of Yu

For the traditions that grew up afterward, the historical question mattered less than the example. Yu entered Chinese memory as a model ruler and, in some strands, as a figure of near-cosmic authority — his ordering of the rivers read as an ordering of the cosmos, his draining of the flood as the act that made the world fit to live in. That cosmic Yu passed into ritual. In the liturgies of Taoism, the officiant treads a measured, halting pattern of steps over a sacred diagram — the Pace of Yu (Yubu), and the wider family of bugang paces that walk the stars of the Northern Dipper — and the mythic source of that gait is Yu himself, said to have walked with a limp from the toil of taming the waters. The priest who paces it treads the constellations as Yu trod the rivers, ordering the unseen world by the same bodily measure with which the founder ordered the seen one. The labor by the flood became, a thousand years on, a step in the architecture of the rite. The Xia stands at the head of the dynastic sequence less as a documented period than as the point where Chinese history, as it chose to tell itself, begins.

→ In the library: The Texts of Tâoism (Legge, SBE 39 & 40, 1891)

→ Related: Zhou Dynasty · Shang Dynasty · China · Confucianism · Taoism · I Ching · Han Dynasty · Neolithic · Divination