Cangjiepian 倉頡篇

Cangjiepian 倉頡篇 "Book of Cangjie" was a character dictionary originally written in small seal script (xiaozhuan 小篆) during the Qin period 秦 (221-206 BCE) by Counsellor-in-chief Li Si 李斯.

With the unification of the empire, it had become necessary to standardize the script, in order to make bureaucracy somewhat easier. The compilation of the character handbook Cangjiepian was overseen by Li Si, the director of the livery office Zhao Gao 趙高 oversaw the compilation of another part called Yuanlipian 爰歷篇, and the grand scribe Humu Jing 胡毋敬 that of a book called Boxuepian 博學篇. It is not known what the particular differences between the three books were.

The title of the Cangjiepian is derived from the name of Cang Jie 倉頡, a minister of the mythological Yellow Emperor 黃帝 and putative inventor of the Chinese script.

The handbook was written in easy-to-remember four character verses. During the Former Han period 前漢 (206 BCE-8 CE), the three books were merged into one, and the text was divided into 55 stanzas of 60 characters each. The book was then called Sancang 三倉 "The three Cang". The characters were at the same time transcribed into the then-common chancery script (lishu 隸書). The book went lost at the end of the Tang period 唐 (618-907). Part of it was rediscovered from a Former Han period tomb library at Fuyang 阜陽. Quite interestingly, the names of Qin rulers are für taboo reasons replaced by other characters. Han-period versions of the text in turn might replace the names of Han rulers. Fragments of the Cangjiepian were also found in Juyan 居延 (Pochengzi 破城子), and Yongchang 永昌 (Shuiquanzi 水泉子), both Gansu. A 40-words long preface is also preserved.

At the end of the Former Han period, the Cangjiepian was revised by Yang Xiong 揚雄, who rewrote a supplement called Xunzuanpian 訓纂篇. During the Later Han period, Jia Fang 賈魴 added a supplement called Pangxipian 滂喜篇. The two supplements were then again united into one book with the title Cangjiepian, in three juan, each juan corresponding to one book.

The name Sancang, a designation for this combination of texts, appeared during the Jin period 晉 (265-420). During that time, Guo Pu 郭璞 wrote a commentary to the Sancang, which is also lost.

The text of the Cangjiepian is compiled in a very interesting manner: The characters are not forming sentences, but are grouped according to meaning, forming a semantic coherence. This shows neighbourhood to the glossary Erya 爾雅, which is part of the Confucian Canon. Characters with common radicals are also united in one verse group. Furthermore, the sequence of characters that are radicals themselves is largely identical to the sequence in the Later Han period 後漢 (25-220) character dictionary Shuowen jiezi 說文解字, which gives an impression of the long tradition upon which the latter is based.

There are also verses where characters with similar pronunciation or even similar phonetic parts are grouped (e.g. 杞芑, 㛍挾, 賞勦向尚), a method which was first discussed by the Song period 宋 (960-1279) scholar Wang Shengmei 王聖美 and then used by the Qing period 清 (1644-1911) scholar Zhu Junsheng 朱駿聲 in his new arrangement of the Shuowen characters in his Shuowen tongxun dingsheng 說文通訓定聲.

During the Qing period, various scholars tried to reconstruct at least that part of the Cangjiepian, of which fragments were scattered in other literature, like the dictionary Yupian 玉篇, the phonetic commentary Yiqiejing yinyi 一切經音義, the anthology Wenxuan 文選 or the encyclopaedia Taiping yulan 太平御覽.

The most important scholars engaging in a reconstruction were Sun Xingyan 孫星衍, Ren Dachuang 任大椿, Ma Guohan 馬國翰, Tao Fangqi 陶方琦 and Gu Zhenfu 顧震福. The Dainange congshu 岱南閣叢書 version of Sun Xingyan is included in the Congshu jicheng 叢書集成, another version in Long Zhang's 龍璋 Xiaoxue soushi 小學搜佚. The fragments of all these versions are arranged according to the radical system in the Shuowen jiezi.

Figure 1. Fragments of the Cangjiepian from Dunhuang

Three fragments of bamboo slips inscribed with a text from the Cangjiepian. It is reproduced in Luo and Wang (1993).