Xu Ziping: The Method That Outlived Its Man

Search for the origins of BaZi and you get a clean story. A Song dynasty master named Xu Ziping 徐子平 reformed an older method, shifted the reading from the year pillar to the day stem, and wrote the Yuanhai Ziping — the foundational text of the system still in use today.

Almost none of that survives checking.

The method is real, traceable, and genuinely a turning point. The man is barely documented, may not have existed as described, and the book bearing his name was compiled roughly three centuries after his supposed lifetime by people whose names we do know.

That gap is worth understanding, because what actually happened is more interesting than the legend.

What is claimed

The standard account: given name Juyi 居易, courtesy name Ziping; from Donghai; active in the transition from the Five Dynasties to the early Northern Song, roughly 907–960. Lived in seclusion on Mount Hua alongside the Daoist figures Chen Tuan, Lü Dongbin and Mayi Daozhe. Reformed the older method and founded the system that bears his name.

What is documented

The Siku Quanshu editors, compiling the imperial library in the eighteenth century, reached a verdict on him that is worth quoting because it is unusually blunt for a reference work:

子平事跡無可考，獨命學為世所宗

His deeds cannot be verified; only his fate-study is honoured by the world.

That is the position of the most authoritative Chinese bibliographic project ever undertaken, and modern scholarship has not improved on it. The standard Chinese reference entry on him opens with 生平事蹟不詳 — life details unknown.

Two specific problems make it worse than vague.

The dating doesn't hold. The scholar Zou Wenyao examined the commentary on the Luoluzi attributed to Xu Ziping and dated it by its internal case examples to around 1135 — Southern Song, two centuries after the traditional lifespan. That puts the "Xu Ziping" commentary later than three other commentators on the same text, when the tradition places him earlier than all of them.

And a serious case has been made that he didn't exist. Zhang Xinzhi, in a thesis on the textual sources of Song and Tang fate-study, argued that Xu Ziping may be a 偽託人物 — a fabricated attribution. A separate Chinese textual study reaches the harder conclusion that the biography is a 層累地造成的學術謊言, a layered scholarly fabrication.

These are minority critical positions rather than settled consensus, and they should be reported as such. But they are serious, and no one has produced the contemporaneous record that would answer them.

"Ziping" as a label

There is a reading that fits the evidence better than either "he existed" or "he didn't."

The Yuan-dynasty writer Liu Yu records, in a way that is almost offhand:

江湖談命者有子平…相傳宋有徐子平，精於星學，後世術士宗之，故稱子平

Among those who discuss fate there is Ziping… it is transmitted that in the Song there was a Xu Ziping, expert in star-study, whom later practitioners honoured, and so it is called Ziping.

Note 相傳 — it is transmitted, hearsay. Already in the Yuan, "Ziping" is being used as the name of a school and the man behind it is something people say rather than something recorded.

That is the most defensible position: 子平 names a method and a lineage. Whether one person stood at its origin is unresolved and may be unresolvable.

Li Xuzhong, who is documented

The contrast is instructive, because the figure Xu Ziping supposedly superseded is a real historical person with a real paper trail.

Li Xuzhong 李虛中, roughly 762–813, courtesy name Changrong, a Tang official who held the post of Palace Censor. We know this because Han Yu — one of the most important prose writers in Chinese literature — wrote his tomb epitaph.

The epitaph praises Li as an expert in Five Agents theory who used a person's birth year, month and day, with their stems and branches, to predict fortune, rank and lifespan with unsettling accuracy.

Read that list again. Year, month, day. No hour. No "four pillars." And Han Yu does not say Li wrote a book.

This matters because practitioner accounts routinely claim Han Yu's epitaph documents Li using the four-pillar system. It doesn't. The Tang bibliography lists no such work by him; a Li Xuzhong mingshu appears only in Song catalogues, the original is lost, and the version we have was reconstructed by the Siku Quanshu compilers from quotations preserved in the Ming Yongle Dadian.

And that reconstructed text does use birth hours — contradicting Han Yu directly, which is how we know those layers were added later. The term 四柱 itself is a Song coinage.

So the clean "Li Xuzhong used three pillars, Xu Ziping added the fourth" story is a retrospective tidying of something messier.

The old law and the new law

Strip away the biography and a real transition remains. Chinese scholarship organises it as 古法 and 今法 — the old method and the new.

The old method read from the year pillar, and its primary machinery was 納音 nayin — a system assigning each of the sixty stem-branch pairs a five-element quality distinct from the elements of its own characters. Alongside it ran an elaborate apparatus of 神煞, spirit-stars: named auspicious and inauspicious configurations, hundreds of them.

The new method reads from the day stem — the day master — and analyses the chart through five-element generation and control, with the four pillars read as root, sprout, flower and fruit. The spirit-stars recede; the Ten Gods and strength assessment come forward.

That shift is what the label "Ziping" actually denotes. It is visible in the texts whether or not a man named Xu Ziping made it.

And the bridge text survives. The Wu Xing Jing Ji 五行精紀, by Liao Zhong, carries a preface dated 1196 and was first printed in 1228. In thirty-four juan it cites around fifty-one earlier fate-calculation works, and it documents the old method — year pillar, nayin, spirit-stars — in full. It is the best evidence we have of what the system looked like before the shift, and it is securely dated in a way nothing attributed to Xu Ziping is.

The book that bears his name

The Yuanhai Ziping 淵海子平 is not by Xu Ziping, and the people who made it are known.

Its cores are two Southern Song works — the Yuanyuan 淵源 and the Yuanhai 淵海 — by Xu Dasheng 徐大升, a Qiantang man of the Baoyou era, 1253–1258. That is the same Xu Dasheng whose Ziping Sanming Tongbian Yuanyuan became a Korean state examination text in 1466.

The compilation then passed through three named Ming hands:

The most-circulated text carries a preface dated 1634 — roughly seven hundred years after the traditional Xu Ziping lifespan. The sinologist Chao Wei-pang, writing in 1946, described it plainly as an earlier work enlarged and re-edited by a Ming fate-calculator.

The book itself is honest about this. It presents itself as synthesising the accumulated results of earlier masters, Li Xuzhong and Xu Ziping among them. It is a compilation that says so.

What arrived when

One more correction to the standard account. The analytical machinery people think of as "Ziping" did not arrive all at once with him.

The Ten Gods originate with the Song reform and appear in the Yuanhai Ziping, but they reach their mature form in the Sanming Tonghui 三命通會 — by Wan Minying, printed 1578.

Strong and weak day master as a framework is Song. As the rigorous balancing doctrine practitioners use today, it is a Ming–Qing systematisation, crystallised in the Qing commentary on the Di Tian Sui and in the Ziping Zhenquan.

The favourable element, 用神, appears as a word early — but the whole-chart doctrine, with the month command primary, was systematised in the Ziping Zhenquan 子平真詮 by Shen Xiaozhan, preface dated 1776.

Climatic adjustment, 調候, is later still — formalised in the manuscript that became the Qiong Tong Bao Jian, a late-Ming text in its Qing printed form.

So a modern reading is Song in its frame and Ming-to-Qing in most of its analytical apparatus. Four to eight centuries after the man it is named for supposedly lived.

Why this is the better story

The hagiography asks you to believe that one Song master invented a complete system that has needed no development since. That is both unlikely and duller than what the record shows.

What the record shows is a genuine methodological revolution — the move from year to day, from nayin to elemental relation — that happened somewhere in the Song, was consolidated over four centuries by named and datable compilers, and acquired a founder's name the way traditions do.

If you want to know when your chart's analytical framework was actually built, 1196 and 1228 date the bridge text that records what it replaced, and 1548, 1578, 1599, 1634, 1776 date the building itself — each with a name attached. All seven are checkable, and that is worth more than a founder.