HistoricalChristian.Faith

Does the Jishi Gorge outburst flood (c. 1920 BC) supply the historical kernel of China's Great Flood?

Type: supporting argument (off-timeline) · Assessment: weak — as the identification of a particular geological event with a particular legend. The argument is the most ambitious attempt yet made to give China's Great Flood a dated cause, it was published in Science by a serious team, and its central datum is real: Lajia is a genuine catastrophe site with skeletons in collapsed dwellings radiocarbon-dated to 1922 ± 28 BC. The identification nonetheless fails at both ends. The geological chronology of the Jishi Gorge dammed lakes, worked by several independent teams before and since, puts them thousands of years earlier; and the earliest Chinese texts do not describe Yu controlling a flood of the Yellow River at all. What survives is the smaller and still substantial result: China has an ancient flood-tradition of its own, independent of the Near East, and it is a datum for the convergent-narratives prediction rather than for a dated event.

Related arguments: corroborating the Flood — see also the flood corpus as corroboration · Woolley's Ur silt · the Black Sea deluge as a kernel · the King List's antediluvian reigns · Göbekli Tepe and a Flood horizon · the Wadi al-Hitan desert whales · the Alaskan muck and frozen mammoths

The claim

On 5 August 2016 Wu Qinglong and sixteen collaborators published in Science a paper whose title states the thesis: "Outburst flood at 1920 BCE supports historicity of China's Great Flood and the Xia dynasty." Sarah Allan, who answered it the following year, summarises the argument fairly and at first hand:

A great lake was formed after an earthquake caused a landslide in the upper reaches of the Yellow River. The eventual breach of the barrier resulted in a flood originating at Jishi 積石 Gorge in present-day Qinghai Province, in about 1920 BCE. This flood breached the natural levees of the Yellow River causing rare, extensive flooding, possibly even an avulsion in the lower reaches of the river. The flood was so catastrophic that it was preserved in the collective memory and became the basis of the accounts of a great flood mentioned in early texts, such as the Book of Documents [Shujing 書經] and Records of the Grand Historian [Shiji 史記]. Since Yu was also the founder of the Xia dynasty, the date of the flood provides evidence for the beginning of that dynasty.

— Sarah Allan, Journal of Chinese Humanities 3 (2017), 24, link. Two claims are bundled here and should be kept apart: that a real, dated, exceptionally large flood stands behind the Yu legend, and that the same date fixes the founding of the Xia. This page weighs the first; the second stands or falls with it.

Who advances it

The proponents are the authors themselves — Wu Qinglong of Nanjing Normal University with Zhao Zhijun, Liu Li, Darryl Granger, David Cohen, the late Ofer Bar-Yosef and others — and the claim reached a wide public through the accompanying Science Perspective by David Montgomery and through general-press coverage that reported the Chinese Great Flood as confirmed. It matters on this timeline because the same argumentative shape recurs everywhere on it: a real geological event of the right kind, in roughly the right place, offered as the seed of a flood legend. Here the case is unusually strong on the geology and unusually weak on the philology, which is the reverse of the Black Sea pattern and worth having on the shelf for that reason.¹

The evidence for

The Lajia catastrophe is real, dated, and horrifying. Lajia, in the Guanting Basin some 25 km downstream of the gorge, is a Qijia-culture settlement whose cave-dwellings collapsed on their occupants; excavators found skeletons in postures suggesting adults sheltering children. Radiocarbon on those remains gives 1922 ± 28 BC. The site is buried under flood sediments, and Wu and colleagues report those sediments not merely over the ground surface but inside the collapsed dwellings and even within pottery vessels — which is the strongest single piece of the case, because it dates the inundation to the destruction rather than to some later episode.²

The scale proposed is of the right order to be remembered. On their reconstruction the breach released some 12–16 km³ at a peak discharge on the order of 480,000 m³/s, which would rank among the largest freshwater floods of the Holocene and could plausibly have caused an avulsion in the lower Yellow River. A memory-seed needs an event that is survivable, locally total, and unrepeated; this one qualifies on all three.²

And the date lands near a real cultural transition. Erlitou, the early Bronze Age culture usually (though contentiously) identified with the Xia, begins in the same general window. The coincidence is what gives the argument its force, and even Allan, its most careful critic, calls it "intriguing" and offers her own account of it.³

Adversarial assessment

What holds. A large outburst flood in the upper Yellow River is not a fringe proposal, and the Lajia destruction layer is not in question. Something catastrophic happened there in the early second millennium BC. The paper's methodological instinct — date the event, then ask what the texts do with it — is exactly right, and is the instinct this whole timeline is built on.

The dammed lake will not date to 1920 BC. This is the load-bearing failure, and it comes from sedimentology rather than from any prior commitment about legends. Three technical Comments appeared together in Science in March 2017. Chun Chang Huang and colleagues reported that the landslide-dammed lake terminated around 5,650 years BP while the Lajia ruins were destroyed around 3,950 years BP by earthquake and accompanying mudflows — two events some seventeen centuries apart. Jian-Chiu Han argued that the reconstruction of the lake's scale, its dating, and the peak discharge at dam failure could not be validated from the published data. A third comment, by Wu Wenxiang, Dai Junhu, Zhou Yang and Ge Quansheng, attacked the specific assumption that the onset of dammed-lake sediment deposition approximates the onset of the outburst. Wu and colleagues replied that the commenters had overlooked or misread the outburst-flood sediments themselves, and the dispute has not been settled by concession.⁴

It has, however, been settled by accumulation. Dong Guanghui and colleagues concluded from multiple lines of evidence that the Lajia disaster had no relation to any outburst flood in the upper valley. And a full re-study of the gorges by Zhang Yuzhu, Chun Chang Huang, James Shulmeister and others found not one dammed lake but two, neither of them available: the Maerpo landslide dam formed a lake at about 8.30 ka BP which was gone by about 6.30 ka, and the Gelongbu landslide formed a second at about 6.10 ka which was gone by about 5.60 ka. Their title states the finding: "No relation to China's Great Flood and the Xia Dynasty." Prior to the 2016 paper, seven of eight published dating sets — by radiocarbon and by optically stimulated luminescence, two methods with different failure modes — had already placed the lake between roughly 10,500 and 3,400 BC. The convergence is what makes this decisive: the older date is not one team's result being defended, it is the result several independent teams keep getting.⁵

And the legend is not about a Yellow River flood. The philological objection is independent of the geology and would bite even if the dating held. Allan's statement of it is blunt:

The most critical problem in linking the archaeological evidence of a catastrophic flood in the upper reaches of the Yellow River with that of the story of Yu is that the water controlled by Yu in the ancient texts is never described in the ancient texts as a flood of the Yellow River. Yu's accomplishment was not in dredging the Yellow River—or any other particular river. He dug all the riverbeds in the world, so that the waters could drain into the sea.

The oldest datable attestation says so plainly. The Bin Gong xu, a middle Western Zhou bronze of the ninth century BC, opens: "Sky/heaven commanded Yu to spread out the earth, follow along the mountains, and make channels for the rivers" — a commission from Heaven to lay out the world, not a relief operation on a river. The "Tribute of Yu" chapter of the Book of Documents has him traversing all nine provinces, the Yellow River merely one item in a topographical inventory; the Shanghai Museum bamboo manuscript Rongchengshi makes the cause of the flooding the absence of channels rather than any excess of water. This is a cosmogony — the earth made habitable and fit for agriculture, with Hou Ji appointed to farming immediately afterwards — and cosmogonies do not need a flood.⁶

The Xia half of the claim is weaker still. The earliest text to mention Yu and the flood together, the "Chang fa" hymn of the Classic of Poetry — "Vast were the flooding waters, Yu arranged the lands and regions below" — celebrates the Shang ancestors and says nothing of the Xia; there are no Xia hymns in the collection at all. In the transmitted narrative the dynasty Yu founds is lost immediately, by his grandson Tai Kang, and what follows for three generations is a list of names. Meanwhile the chronological coincidence that the argument leans on has been eroded from the other side: refined radiocarbon work places Erlitou at roughly 1750–1530 BC rather than from 1900, which removes the synchrony, though not every specialist accepts the compressed range.⁷

A better use of the same datum, offered by the critic. Allan does not simply reject the flood; she proposes that it mattered indirectly. Jishi Gorge lies in Qijia territory, and the Qijia were China's earliest bronze-casting culture. Flooding that displaced Qijia people toward the central plains would have carried metallurgical knowledge into the region where Erlitou then rose — a real historical effect for a real flood, with no legend required. She also notes, citing Sun Hua, that the pure silt layers that mark flooding are relatively easy to identify and have not been found at Erlitou itself. That is a testable alternative, and it is the shape the surviving claim should take.³

A note on scope. Nothing here turns on whether a flood tradition can be historical; this timeline holds that it can, and says so elsewhere. Nor does anything turn on excluding the miraculous: Yu's supernatural features in the Zhou and Han texts are reported here as evidence about the genre of the material, not as grounds for dismissing it. The failure is specific and mundane — a lake that drained sixteen centuries too early, and a legend about digging riverbeds rather than about surviving a flood.

Assessment

Weak — as the identification of the Jishi Gorge outburst flood with the flood of Yu. Two independent failures, either of which would be sufficient. The dating fails: independent radiocarbon and OSL work by several teams, before and after the 2016 paper, puts the Jishi Gorge dammed lakes in the eighth-to-sixth millennia BC and finds them already drained by about 5.60 ka BP, with the Lajia destruction a separate seismic-and-mudflow catastrophe some seventeen centuries later. And the target fails: the flood Yu controls in the earliest texts is not a Yellow River flood, is not a rainfall or breach event, and belongs to a cosmogony in which the world is made drainable and fit for farming — a story attested from the ninth century BC in the Bin Gong xu, and one that the earliest source associating Yu with water, a Shang hymn, does not connect to the Xia at all.

Two things are owed to the argument. The Lajia catastrophe is real and precisely dated, and the paper is right that the upper Yellow River suffered an exceptional early-second-millennium disaster; the objection is to where it is pointed, not to whether it happened. And the argument's failure leaves standing the datum this timeline actually wants from it: China has a flood-tradition of great antiquity that developed outside the Near Eastern transmission routes, which is a genuine independent case for the convergent-narratives prediction — and, on the archaeological side, so is the dated flood-collapse of Liangzhu a thousand miles and fourteen centuries away. Independence is what that argument was short of; East Asia supplies it.

Sources