Xu Shen: The Man Who Took Chinese Writing Apart
Around the year 100 CE, a scholar in the Eastern Han empire finished a book that did something nobody had thought to do: it broke every Chinese character in the written language into its component parts and explained why each one meant what it meant. Then he sat on it. The *Shuowen Jiezi* was not presented to the throne until 121 CE, and it was his son Xu Chong who carried it to Emperor An. By then Xu Shen was at the end of his life, and he had already given Chinese literacy the organising principle it would use for the next nineteen centuries.
A Provincial Scholar of the Five Classics
Xu Shen lived roughly from 58 to 148 CE, born in the Zhaoling district of Runan prefecture, in what is now Luohe in Henan province. He studied under Jia Kui (30–101 CE), a respected scholar of the Old Text tradition, and built a reputation for mastery of the Five Classics — the canonical Confucian texts that formed the whole intellectual foundation of Han officialdom.
That reputation carried him upward through the modest channels available to a provincial scholar. He held administrative posts at the prefecture level, and eventually secured a position in the royal library. It was the right room to be in. The library held the disputed manuscripts, the variant editions, the texts recovered after the Qin dynasty's suppression of learning — the raw material of the argument that would consume his life.
The Fight Over What the Classics Actually Said
To understand why Xu Shen wrote a dictionary, you have to understand that in Han China a dictionary was a weapon. Scholarship was split between two hostile camps. The New Text school worked from versions of the classics reconstructed after the Qin book burnings, transmitted in the contemporary clerical script. The Old Text school worked from manuscripts said to be older, written in archaic characters, and claimed those were the authentic transmissions. What was at stake was not antiquarian: it was who had the right to say what Confucius meant, and therefore who had authority in the empire.
Xu Shen came out of the Old Text tradition, and his method was to attack the question at its root. If you could establish what each character had originally meant — by showing how it was built — you could adjudicate between readings of the classics on evidence rather than assertion. He framed this in explicitly Confucian terms, in service of the Rectification of Names: the doctrine that correct government depends on words meaning what they should. He also wrote the *Wujing Yiyi* ("Different Meanings of the Five Classics"), an attempt to reconcile the two traditions, which has been lost.
Nine Thousand Characters and 540 Doors
The *Shuowen Jiezi* — the title means roughly "explaining simple graphs and analysing compound characters" — runs to fifteen chapters with a preface and a postface. It holds 9,353 character entries, plus 1,163 variant forms. Every one is analysed: its structure described, its meaning explained, often its pronunciation indicated.
The innovation that made it immortal was the filing system. Xu Shen sorted his 9,353 characters into 540 sections, each headed by a shared graphic component — what we now call a radical. This was, as scholars have put it, a major conceptual innovation in understanding the Chinese writing system. Before it, there was no principled way to find a character you did not already know, and no way to see that characters sharing a component shared something about their meaning. After it, Chinese had an index and a theory at the same time. The system was later condensed — the Kangxi dictionary would use 214 radicals — but the idea was Xu Shen's.
He also worked from the small seal script, the form standardised under the Qin dynasty rather than the clerical script of his own day. That choice made the book a paleographic handbook as well as a dictionary, preserving graphic forms that were already becoming archaic when he wrote.
The Six Categories
Xu Shen set out the *liushu*, the six categories of character formation, and gave them their durable definitions: pictograms, simple indicatives, compound indicatives, phono-semantic compounds, loangraphs, and derived characters. Only the first four are worked through extensively in the dictionary itself, and the last two have been argued about ever since. But the framework is still how Chinese writing is taught. Any student who has been told that a character has a meaning element and a sound element is receiving Xu Shen's analysis at second hand.
Why Xu Is Called a Genius
The claim for Xu Shen is not that he was a great thinker in the philosophical sense. It is that he performed an act of abstraction that nobody before him had managed and that everybody after him depended on. He looked at nine thousand individually memorised symbols and saw that they were not individual at all — that they were built from a finite set of recurring parts, according to a small number of construction rules. Extracting 540 shared components from that mass, and a six-part typology of how characters get made, is a feat of analysis performed without any prior model to copy. It is the same class of move as devising a periodic table: the material was public, the pattern was not.
The honest counter-case is that a great many of his specific analyses are wrong. Xu Shen worked from seal script, believing it to be the oldest form available. It was not. The discovery of oracle bone inscriptions, centuries older, superseded many of his etymologies — he was reconstructing origins from a stage of the script that had already been through significant distortion. His framework was also ideological: he was arguing an Old Text position in a partisan scholarly war, and his readings sometimes serve that argument. Even the text we have is not quite his, having required substantial restoration by Xu Xuan in the tenth century.
So the genius is architectural rather than evidentiary. He got the system right and a good share of the details wrong — and the system has outlasted every correction made to the details.
Legacy
The *Shuowen Jiezi* founded Chinese lexicography and never stopped being read. Xu Xuan restored the text in the tenth century; Duan Yucai's annotated edition, produced in the Qing dynasty, remains the standard scholarly version. Its radical-based organisation is the ancestor of every Chinese dictionary since, and its influence, as one summary puts it, is still felt today — in the structure of dictionaries, in how children are taught to read, and in the simple fact that a Chinese character is understood as something assembled rather than something merely remembered.
.jpg)
