Cang Jie pian was an elementary-education textbook of the Qin and Han periods. Lost for nearly a 1000 years, scholars knew it only through quotations in received texts. In 1977, a fragmented copy of Cang Jie pian dated to the early Western Han period was discovered in a noble tomb at Fuyang Shuanggudui. It provided a basic idea of how this ancient textbook was compiled. Recently, another much longer version was added to Peking University’s (Beida) collection of bamboo strips. The Guoxue approach to studying this ancient textbook focuses on orthography, word meanings and rhymes, as well as its values in relation to Western Han history and literary culture, whereas the Hanxue Sinological approach aims at discovering how a word and its concept in the excavated texts took root in Chinese culture.

Keywords: Cang Jie pian, bamboo strips, guoxue, Sinology (Hanxue)

In a conference dedicated to the theme of 21st-century Sinology (Hanxue漢學) and guoxue 國學, the organisers raised a difficult question that no simple answer can satisfy: How may the archaeological field illuminate the similarities and differences between Sinology and guoxue?[1] One may begin by first settling another question: Has archaeology been considered a part of Sinology? Technically speaking, it is not. Sinology is a discipline of language and philosophy, whereas archaeology is a discipline of stratigraphy and typology.[2] In China, however, excavated texts always attract tremendous attention. The contents of archaeological writing may either supplement what received historical texts left out – as found in the oracle-bone and bronze inscriptions – or present an early copy of a received text, so that a comparison with contemporary versions may reveal an interesting textual history. In this regard, archaeology supplies excavated texts that interest Sinologists. Indeed, texts are important for archaeologists, too. There are not many Chinese archaeologists who could have worked purely on excavated materials without making a reference, however minimal, to texts. China is a country of texts.

Cang Jie pian 蒼頡篇 (or 倉頡篇) was a primer adopted during the Qin and Han periods. It was a word list that gathered all or most of the words known by the time of its compilation, the late third century BC. Because it did not render defined meanings or pronunciation, Roger Greatrex described it as a proto-dictionary. This Western Han primer is believed to have heralded the early development of xiaoxue 小學, primary education or philology, which includes orthography, semantics and phonetics.[3] Most importantly, the later, revised and expanded editions of Cang Jie pian eventually formed the foundation for the monumental work Shuowen jiezi 說文解字.[4] Nevertheless, the original Cang Jie pian was not preserved. The book had been lost for nearly a millennium when fragments of it were discovered at Dunhuang 敦煌 in the early 20th century. Thereafter, several fragmented copies of Cang Jie pian were recovered archaeologically. The earliest copy known so far was dated to the mid-second century BC, excavated from the tomb of a Han noble member at Fuyang Shuanggudui 阜陽雙古堆 in Anhui 安徽 Province in 1976; another copy, which was dated to the same period but doubled the length of the Fuyang copy, was salvaged from the hands of antique dealers by Peking University (Beida 北大) in 2009.[5]

Cang Jie pian represents a less discussed text genre, which often serves as important reference work for cross-checking linguistic information that is useful for reading philosophical treatises. Yet scholars of intellectual history have frequently overlooked this genre’s significance. The last century, and particularly the last three decades, witnessed generations of scholars working on the archaeological copies of Cang Jie pian for its indications about the early developments of the Chinese language and writing. Interestingly, the works of Sinologists and their Chinese-based colleagues demonstrate quite different perspectives.[6] In the last section of this article, I will return to the very first question stated at the beginning. Definitions of Guoxue and Hanxue have never ceased to receive challenges and revisions; debates are ongoing as to what the terms may or may not cover. The studies of Cang Jie pian suggest that differences exist in perspectives, if not in the methodologies, adopted by Sinologists and Chinese scholars. Whether such difference may be attributed to how Guoxue or Hanxue are defined is difficult to tell. The two scholarly camps, if they may be defined loosely, have not the same priority preference in exploring China and her past.

Cang Jie Pian: Discoveries and Studies

Cang Jie pian was thought to have been named after a legendary figure, Cang Jie 蒼頡, who had, according to received texts, served the Yellow Emperor in remote antiquity and invented writing.[7] Archaeological copies have now revealed that the title usually adopted the first two characters in the text, written exclusively at the top of the first two strips.[8] In his monumental library catalogue included in Hanshu 漢書, Ban Gu 班固 (32–92) made a note about Cang Jie pian. There were originally three books written in Qin seal script. They were respectively known as Cang Jie, which contained seven sections; Yuan li 爰歷, six sections; and Bo xue 博學, six sections. Ban Gu also gave the names of the authors, who were all the highest-ranking officials in the Qin court. Cang Jie was attributed to Li Si 李斯 (284–208 BC), Prime Minister of Qin; Yuan li to Zhao Gao趙高 (258–207 BC), Director of the Livery Office; and Bo xue to Humu Jing 胡毋敬(active c.3rd c. BC), Grand Astrologer.[9] It was observed that the words in these three books were mostly collected from yet another character primer (zishu 字書), entitled Shi Zhou pian 史籀篇, which is now lost. At the beginning of the Han dynasty, folk teachers put the three books together and re-arranged them into 55 sections, each containing 60 words. This 3,300-word volume was then entitled Cang Jie pian.[10]

By the end of the Western Han dynasty, Cang Jie pian was found to contain too many difficult words that were no longer in use. It subsequently underwent several revisions during the Eastern Han dynasty (25–220), and was gradually combined with or incorporated into other synonymicons during and after the Western Jin dynasty (266–316). Wang Guowei 王國維 (1877–1927) suggested that Cang Jie pian was lost during the Sui dynasty (581–619), whereas Sun Xinyan 孫星衍 (1753–1818) argued that the text did not vanish until the end of the Tang dynasty (618–907), almost three hundred years later. Thereafter, Cang Jie pian was only mentioned in the form of quotations.[11] This textual history was established by the findings of Sun Xinyan and other Qing scholars, who compared several pre-Sui synonymicons in an effort to trace how the lines and words were collected and quoted in different books. Wang Chongmin 王重民 (1903–1975) offered a review of all Qing works in his article “Cang Jie pian jiben shuping” 蒼頡篇輯本述評, published in 1933. The article represents quite well how traditional Chinese scholarship handled textual studies.

In 1906, fragments of Cang Jie pian were found in Dunhuang. However, no more than 80 characters of the whole book were preserved. Luo Zhenyu 羅振玉 (1866–1940) and Wang Guowei obtained the photographs by the courtesy of Emmanuel-Edouard Chavannes (1865–1918) (see Fig. 1).[12] Wang observed that Cang Jie pian was compiled in a fashion of having “each line carrying four characters; rhythms were there to be found”[13] (see Fig. 2). Meanwhile, Wang Guowei was attempting to restore the lost Cang Jie pian from the words collected in Ji jiu pian 急就篇 by Shi You 史游, who was active during the reign of Emperor Yuandi 元帝 (r. 48–33 BC).[14] In 1930, the fragments of Cang Jie pian from the watchtower sites at Juyan 居延in Inner Mongolia preserved 110 words, giving a slightly better glimpse of this lost book (see Table 1)[15].

Table 1. Details of the remains of Cang Jie pian excavated from archaeological sites.

Year of Exc.


Site nature

No. of words preserved

Approx. dates


Gansu/Inner Mongolia[16]



Western Han


Gansu/Inner Mongolia[17]



Western Han


Inner Mongolia, Juyan[18]



Western Han


Inner Mongolia, Juyan

Pochengzi 破城子(Jiaqu   houguan 甲渠侯宫)[19]



Western Han


Gansu Yumen Huahai



Western Han


Anhui Fuyang Shuanggudui

(tomb of the Marquis of Ruyin 汝陰)



165   BC (date of the tomb)


Gansu Dunhuang Maquanwan





Xinjiang, Niya 新疆尼雅[22]



Western Han


Gansu, Yongchang Shuiquanzi 永昌水泉子[23]



Late   Western Han


Anhui or Jiangsu,

Beida Collection



Early   Western Han


FIGURE 1: Remains of Cang Jie pian excavated from Juran in the early 1900s. After Luo Zhenyu and Wang Guowei 1913 (reprint in 1993), p. 13.


 a                         b

FIGURE 2: Remains of Cang Jie pian from Juyan collected in the catalogue prpared by LaoGan in 1949. 2a. 蒼頡作書以教後詣(185.20); 2b. X嗣幼承詔謹慎敬戒X (167.4). After Lao Gan 1949.

Meanwhile, Derk Bodde offered a different view. Instead of examining how Cang Jie pian was transformed through historical times, Bodde studied the political bearing in the compilation of this text. Shi Zhou pian, which was compiled centuries before Li Si’s lifetime, was written in Large Seal. The writing was pictorial and complex. By compiling a new character primer, written in a new, simplified script – the Small Seal—Li Si was thus realising the writing unification project by means of, firstly, standardising the script, and secondly, making sure that a greater number of people would have been able to read it.[24]

In 1977, the tomb of the Marquis of Ruyin (Ruyin hou) 汝陰侯, who died ca. 165 BC, was found at Fuyang Shuanggudui in Anhui Province. His small library included Cang Jie pian, though in a fragmented form. This version preserved 541 characters.[25] Shiji 史記did not mention much about the owner. His name was Xiahou Zao 夏侯竈, son of Xiahou Ying 夏侯嬰 (d. 172 BC), who was one of Liu Bang’s 劉邦 (256–195 BC) early followers and assisted in the founding of the Han Dynasty. The Xiaohou family was consequently given a piece of territory at Ruyin (now Fuyang) in the Runan 汝南 Prefecture, where the tomb was discovered. Xiahou Zao inherited his father’s title. However, nothing else was known about him.[26] Apart from Cang Jie pian, the Marquis also had Shijing 詩經, Zhouyi 周易, Chuci 楚辭, almanacs, medical texts and recipes, annalistic history texts, a dog manual and numerous writings on craftsmen and agricultural management.[27]

The Fuyang copy of Cang Jie pian was rather fragmented. A total of 124 bamboo strips were identified, all broken. The longest strip preserved measures 18 cm long, carrying twenty consecutive characters (see Figs. 3 and 4).[28] The original length of the strips was probably 25 cm. Hu Pingsheng, a historian and specialist in early Chinese writing, and Han Ziqiang, the excavator of the tomb at Fuyang, relied on the philological studies by Luo Zhenyu and Wang Guowei to work out the broken strips’ original arrangement.

FIGURE 3: Remains of Cang Jie pian from the tomb of the second Marquis of Ruyin at Fuyang Shuanggudui, Anhui province. Nos 1–3 correspond to strips numbered C007–009 respectively; no. 4 (C034); no. 5 (C053); no. 6 (C054); and no. 7 (C037). After Fuyang 1983b, p. 31.

FIGURE 4: Transcriptions of the remains of Cang Jie pian from Fuyang. No. 4 above corresponds to bamboo strip C004, on which Hu Pingsheng and Han Ziqiang first mistaken a line as Ba Shu tu ting巴蜀荼朾that caused Roger Greatrex to stage a discussion on the meaning of the last word, ting and its relation with a more commonly used term, da. After Fuyang 1983b, p. 32.

Rhymes, therefore, were the first thing Hu Pingsheng and Han Ziqiang attempted to sort out with the Fuyang copy. They relied on the phonological table prepared collectively by 18th-century scholars. The words’ ancient pronunciations are divided into thirty groups, which may be further divided into three sub-groups based on the finals (the rhymes).[29] For easy reference, scholars had assigned a word to each category. In the Fuyang copy of Cang jie pian, words with zhi , yu , and yang rhymes were the most represented, i.e., words belonging to these three rhyme groups were preferred to form the last character at every second line to keep the flow, as found on the bamboo strip numbered C010:

Yuan, li, ci, yi爰歷次貤 [= change, undergo, secondary, shift]

Ji, xu, qian, tu繼續前圖 [= succeed, continue, preceding, plan]

Fu, jin, ke, X 輔廑顆X [= auxiliary, barely, kernel, X]

ba, dan, X, tu軷儋屠 [= sacrifice, weight, X, slaughter]

The last words in these four lines all belonged to the yu group.[30] These were the beginning lines of the Yuanli sections, as indicated by the first two words.

Hu and Han further argue that each of the three rhyme groups may have respectively corresponded to each of the three books, as Hanshu had recorded: zhi formed the rhyme of Cangjie; yu of Yuanli; and yang of Boxue.[31] Had this been the case, the Fuyang copy would have been quite close to the original Qin version of Cang Jie pian.[32] The two authors rendered two further pieces of evidence in support of this argument. First, the line chi, duan, xiu, biao飭端脩瀌 (C002) was still using duan as the alternative orthography for zheng , a practice of the Qin period to avoid using the First Emperor’s name.[33] Second, some Qin or Qin-related words were preserved here, as in the lines written on C029:

shi, nei 室內 [= room, interior]

Chiang, X , hu, fang 户房 [= window, ?, door, room ]

Fu, mei, cui, bi 桴楣榱㮰 [= pole, cross beam, rafter, small rafter]

X, X, qiao, liang X X橋梁 [= X, X, bridge, pillar] [34]

The third line was made up of words that referred to the inner compartments of a house. According to Shuowen jiezi, cui is the Zhou term for “house.” Qin called it chuan , whereas Qi and Lu people called it jue . This copy was using the old Zhou term, suggesting that the word would likely have been adopted from the Qin state, which was the former homeland of the Zhou. The same was found in the word (C028), which was a Zhou term meaning the gate to a regional district. Hu Pingsheng and Han Ziqiang also noted a few other examples of Qin terms, which find correspondence with the Qin bamboo strips excavated from Yunmeng Shuihudi 雲夢睡虎地in Hubei Province.[35]

Cang Jie pian was a collection of words used (or known) during the time of its compilation. The words were largely grouped on the basis of meanings. The most common rule adopted was to group words in pairs of similar or opposite meanings. Roger Greatrex calls them near-synonyms and near-antonyms respectively, as found in the lines on C001:

Jiao, han, jiao, ju傲悍驕倨 [= proud, ferocious, arrogant, haughty]

qu, ju, guan, wang 趣遽觀望 [= hurry, abrupt; observe, look] [36]

And in the words on C028 and C006 respectively:

kai, bi開閉 [= open, close]

xiong, ci [= male, female][37]

Another rule adopted was to form words of the same semantic areas into a cluster, as found on C015:

jiao, long, gui, she蛟龍龜蛇 [= dragon, dragon, turtle, snakes (which were all reptiles)]

And on C029 quoted above:

Fu, mei, cui, bi桴楣榱㮰 [= pole, cross beam, rafter, small rafter (which were different wooden components of a house)]

The words in the same clusters may sometimes share the same radical.[38] In the compilation of this text, the author or authors would have first grouped together words of related meanings, arranged them in blocks of four, according to specific meaning between pairs, and finally made use of the phonological elements to determine the word ordering.[39] Though found fragmented, the Fuyang copy of Cang Jie pian demonstrates the structure of this lost text clearly enough. After the Fuyang copy of Cang Jie pian was published, Chinese authors conducted a number of studies in orthography, phonetics and comparative analysis to rectify or elaborate on Hu Pingsheng’s first reading.[40] Indeed, Hu himself further elaborates his own finding on different occasions.[41] Chinese scholars value textual analysis in a microscopic manner and often revisit the same text for corrections or elaboration.

The Cang Jie pian collected by Beida came in a much lesser known context than the Fuyang copy. A private collector purchased it, along with other books, in 2009. The strips were sent to Beida stacked in nine plastic containers. After three months of consolidation, a total of 3,346 strips were registered. 1,600 of them were preserved intact. The books were found to have come in varying sizes. The biggest measured between 45 and 46 cm, belonging to the almanacs; the shortest ones were 23 cm, belonging to the medical texts. Apart from Cang Jie pian, the other texts were Laozi, Zhao Zheng shu 趙正書 (Biography of Zheng from Zhao), Zhou xun 周馴 (Teaching of the Zhou), almanacs and astrological texts; various writings by Warring States masters; a text on the spiritual world; and a fictional story about a household in which a man had two wives.42] The different handwritings suggest that they may have been written at different times by different people. Because the almanacs and other religious texts contained a remarkable number of references to the religious rites and practices of Chu, the collection is believed to originate in a region of the Chu kingdom of the Western Han period, i.e., from present-day Anhui and northern Jiangsu. The site would have been quite close to where the Fuyang copy of Cang Jie pian was found.[43]

The ruling years of Han Emperor Xiaojing 孝景 (r. 157–141 BC) were the only absolute dates found in this collection. Zhu Fenghan and his team note that the handwriting in this collection had already demonstrated some maturity in the use of the clerical script, compared with that script’s early development as found on the bamboo strips excavated from tomb M247, which was dated to ca. 186 BC, at Jiangling Zhangjiashan 江陵張家山 in Hubei Province. Yet the Beida script forms were not as angular as those found on the bamboo strips excavated from Dingzhou Bajiaolang 定州八角廊, which was dated to the reign of Emperor Xuandi 宣帝 (r. 74–48 BC). Zhu Fenghan and his team, therefore, suggested that these books were written during the later years of Emperor Wudi 武帝 (r. 141–87 BC). This Cang Jie pian copy was slightly later than that from Fuyang.[44]

The Beida copy of Cang Jie pian was found on 86 bamboo strips measuring 30.2–30.4 cm long (see Fig. 5)[45]. A total of 1,325 characters were preserved.[46] Because it was a much longer version, the Beida copy provided some more details for what Hu Pingsheng and Han Ziqiang discovered about Cang Jie pian with the Fuyang copy. The zhi, yu and yang rhyme groups were once again in use, and all lines were rendered in four words. It is now clear that the entire book would have been divided into many sections. The longest contained as many as 152 characters, a fact known because the scribe noted the number of words at the end of each section. Moreover, each section employed the first two words as the title; these were distinguishably written horizontally at the top of the first two strips.[47] There is, however, no evidence that these sections had been grouped into the three books as described in Hanshu; and this Beida copy is certainly not the edited version prepared by the early Western Han village teachers.[48] From these observations, Zhu Fenghan believes that there would have been more than one or two versions of Cang jie pian copied from the Qin original.[49] His other findings in this copy may be summarised in a few notes as follows:

1. While the zhi, yu and yang rhyme groups formed the basic structure of the text, these groups alone could not have covered all words included in the text. Each of the three had been extended to work with other rhyme groups that shared the same initial or the same final. The zhi rhyme group, for example, was used interchangeably with the zhi rhyme group to form a composite. Elsewhere, the words of the you rhyme group were grouped together with words of the xiao rhyme group to form another composite rhyme.

2. While many lines contain two pairs of near-synonyms or near-antonyms as discussed above, some other lines may consist of pairs of unrelated meanings. The two lines on bamboo strip 013 read as:
xuan, yun; niang, zhi 泫沄孃姪 [= in which the first two words describe water in turbulence whereas the last two mean aunt and niece respectively] ;
meng, fo; die, xi 髳岪絰枲 [= where the first two words describe bush and hillside road, whereas the last two mean hemp].

3. Yet some other lines may be composed of words with no connection in meaning whatsoever. Bamboo strip 002 carried twenty such words, which read as:
bin, x, xiang, shang 賓X向尚;
Feng yi, qing, bei馮奕青北;
xi, sun, bao, su 係孫襃俗;
Ken, zheng, Ji, ji 貇鬵吉忌.

The only thing these characters had in common was the shared zhi-zhi composite rhyme.

FIGURE 5 a and b: Cang Jie pian in the Peking University collection. Strips 8 and 9 were inscribed with the title of the section at the top. Strip 11 also carries a line that reads Ba Shu tu zhu巴蜀荼竹, which confirms that the first reading of the same line in the Fuyang copy was mistaken, and Greatrex is correct in suggesting that the last word would have been a noun and carrying a meaning in relation to bamboo. After Beijing daxue cang Xi Han zhushu: Juan yi 2015, pp. 3–4.

In grouping the words, the compiler began by identifying words of related meanings, selecting them and grouping them in lines. He then moved on to work on the rest and grouped them into as many pairs as possible. Some “leftovers” remained, so he simply did the final grouping on the basis of phonetics. This was perhaps not the most systematic way of arranging a character primer – certainly not a way of compiling a dictionary. However, this compilation method ensured that the text could be easily chanted and recited.

Cang Jie pian: The Two Approaches in Studying

In the study of Chinese philology, Cang jie pian was never as popularly known and used as Shuowen jiezi. Although Cang Jie pian was probably adopted as an elementary school textbook for some time during the Western Han period, it was soon revised, incorporated, and eventually replaced by other character primers. When Derk Bodde discussed Cang jie pian and introduced it to the English readership, he considered it one of the tools developed in Li Si’s grand project of writing unification. For such, Bodde was concerned about how the book was connected with the high-ranking officials in the Qin court; how it came to simplify and standardise regional writings; and how it developed impacts by means of nation-wide distribution. To most Chinese scholars, by contrast, Cang Jie pian was a piece of valuable inheritance, especially because the book had been seen (or known) and recorded by Ban Gu, the 1st-century-AD historian. When Cang Jie pian was first discovered in the northwest, the last generation of Qing scholars paid particular attention to the arrangement of words and phonetics. Hu Pingsheng and Han Ziqiang inherited that approach and laid the focus on individual words and sounds. Zhu Fenghan and other contemporary Chinese scholars have a much longer version to work on; comparisons with other bamboo strip writings are also made possible, in light of many archaeological discoveries over the last decades. As the research materials’ scope has been expanding over time, generations of Chinese scholars seem to have come to the agreement that writing underwent a remarkable transformation in terms of both writing script and usage during the Qin and Han periods, i.e., the very beginning of the imperial times, and that Cang Jie pian may serve as physical evidence to demonstrate the details of this transformation. They thus aim their studies at producing a faithful word-by-word analysis, so that their fellow Chinese may gain an idea about what historical accounts left out.

Two recurring questions about Cang Jie pian often appear in Chinese scholarship. Firstly, how close is the archaeological copy to the edition compiled under the direction of Li Si, or to the revised edition used by the Han village teachers? Hu Pingsheng and Han Ziqiang were rather confident that the Fuyang copy preserved a certain number of Qin preference for words. Although Zhang Chunliang does not agree with them, his study of the Cang Jie pian versions from Yumen Huahai and Yongchang Shuiquanzi in Gansu Province followed the same approach to identifying the dates of different copies. His methodology revolves around rhymes, which he took as the key to distinguishing different editions, as well as the revisions made over time.[50] Banben 版本, or editions, and their developments have been an integral part in the studies of Cang Jie pian since the times of Wang Chongmin or even earlier. Contemporary scholars are privileged to have multiple archaeological copies to work on. From these, they develop methodologies such as comparing script forms, handwriting and contents with other bamboo strip collections or stone inscriptions. These are considered central to Chinese training and scholarship.[51]

The second question Zhu Fenghan raised was how Cang Jie pian may have been put to use in education. In his study, he discusses five examples of loan characters and explores their original and extended meanings. He believes that teaching with Cang Jie pian would have involved explanations not only of orthography, but also of the rules that governed the relationships between interchangeable words. In addition, he raises seven examples to show that Cang Jie pian had adopted many terms from the Classics, especially the Shijing and Chuci. For example, the Beida strip 032 carries a line “xi xiao pin ju 細小貧窶 [= thin, small; poor, impoverished], where pin and ju are found in the “Beimen” 北門 ode in Shijing Beifeng 北風, which reads as “zhong ju jie pin終窶且貧. Hence, Zhu suggests that “when the teachers of any ranks taught with Cang Jie pian during the Qin and Han periods, they were not only delivering knowledge about the words, but also directing pupils to look for the original sources and to read the Classics.”[52]

Zhu continued to look for details about education of the Western Han period. That led him to make a note about the training of historians and scribes mentioned in another set of bamboo strips, which were excavated from tomb M247 at Jiangling Zhangjiashan in Hubei Province in 1983. The tomb was dated to ca. 184 BC. A book entitled Shi lü 史律 stated that if the sons of literary officials, who had graduated at the age of seventeen, were to inherit their fathers’ titles, or to enter the official careers, of shi (scribe), bu (diviner), and zhu (invocator), they ought to spend another three years studying more difficult or bigger character primers under the instruction of the greater scribes, diviners or invocators.[53] Interestingly enough, the final evaluation of pupils would be conducted on the basis of the number of words they could master. The more words they knew, the higher the official ranks they could attain. The scribes, who were the most high-ranking class, were expected to know as many as 9,000 words. Although this number may be exaggerated, Li Xueqin looked up these details in Hanshu and Zhouli, and pointed out that they too carried an astonishingly similar set of notes about pupils’ training.[54] Zhu Fenghan, hence, concludes that Cang Jie pian, which was said to have contained 3,300 words, would have only been able to serve as an elementary text on writing for all pupils under the age of seventeen.

Christopher Foster was inspired to follow the same approach, and continued to explore what Cang Jie pian reveals about the education framework and concepts of the Western Han period.[55] Chinese scholars and Sinologists frequently exchange ideas, especially in the present time when traveling and communications are easy to achieve. But comparing the work by Foster and those by Liang Jing and other Chinese scholars, it is rather obvious that Foster has been approaching the text primarily for its role and significance in social history. Chinese studies scholars of bamboo-strip writing usually undertake a close reading of texts, and exhaust all possibilities inherent in rhymes, root meanings, word inter-relations, etc. Whenever applicable, they cross reference with received historical texts and another set of excavated texts. Michael Loewe notes in a recent interview that guoxuejia國學家, which refer to Chinese experts, are highly knowledgeable about texts and historical details; their level of sophistication is unique. He also describes this approach to scholarship as very traditional.[56]

Roger Greatrex’s article was entitled “An Early Western Han Synonymicon: The Fuyang Copy of the Cang Jie pian,” published in the festschrift in honour of Göran Malmqvist on his 70th birthday. Throughout his study of the Fuyang copy of Cang Jie pian, Greatrex produces a systematic account of the editions and revisions made between the Qin and the Northern and Southern dynasties. Whereas he develops the first two sections on the findings of Wang Chongmin, Hu Pingsheng and Han Ziqiang, he uses the last section to provide a critical reading of the text. Here Greatrex offers two pieces of insights for us to think about: First, he remarks on the sizes of the bamboo strips of the Fuyang collection, and points out that the length of the Shijing bamboo strips was nearly half of what Wang Chong 王充 (27 – ca. 97), historian and critic, had come to believe. Wang Chong and his contemporaries believed that the classics were so important during the Western Han period that they would have been deliberately written on the extraordinarily long bamboo slips measuring two chi and four cun , i.e., approximately 56 cm.[57] Greatrex researched this for Wang Chong and found him erroneous, saying:

Wang Chong and his contemporaries incorrectly believed this phenomenon to be much older than it was, something which suggests the knowledge of the cultural history of the early days of the dynasty was not as strong after the hiatus of Wang Mang’s interregnum (AD 9–23) as we might believe today.[58]

This remark demonstrates clearly how Sinologists think and work. Accuracy and precision are of supreme importance when a historical fact is suggested. If possible, the exact time period and place ought to be sorted out if they are of relevance to a historical incident or concept. Once misinformed, the cause ought to be identified and the case rectified.

Greatrex adds his second note to the study of Cang Jie pian in the last section of his article, in which he engages in an interesting discussion about the origin of a word identified as ting , which scholars have long considered an old, alternative orthography of da , a commonly used term today. Although da was already in use during the time of the Eastern Han, it has never been found in any excavated texts from the Qin and Western Han periods.[59] Duan Yucai 段玉裁 (1715–1835) once quoted a line, which reads zhuang chu yue ting 撞出曰朾, in a rare book entitled Tong su wen 通俗文 by Fu Qian 服虔 (125–195), to argue that ting could be glossed as zhuang. Hence, ting could have been used as a verb and represented the action of hitting or pushing forward. Duan’s contemporaries soon took up his suggestion. When ting was found in the line “Ba Shu tu ting巴蜀荼朾 (C004) in the Fuyang copy of Cang Jie pian, Greatrex questioned this interpretation.

Greatrex argued that no simple equivalency could be established between the two words, ting and da. He began to research the meaning of the word preceding ting, i.e., tu and found that tu was a noun meaning “hollow bamboo.” Greatrex cross checked with Huainanzi 淮南子, which was composed in 140 BC in Anhui, the same region where the Fuyang copy was discovered, and confirmed that tu meant bamboo. Because Cang Jie pian was composed with the rule that nouns were always paired up with nouns and verbs with verbs, ting could only have served as a noun, in order to be paired with tu, bamboo. Ting, therefore, could not have been connected with the verb da.

Greatrex continued his search, which brought him to two Late Tang authors named Xu Xuan 徐鉉 (916–991) and Xu Kai 徐鍇 (920–974). These men had glossed ting as the term for chuang , which meant a pole or the length of a piece of wood. With that, Greatrex argues that tu and ting could have worked together to form two nouns following Ba and Shu in the same line. The later couple refers to bamboo and pole respectively.

In this reviewing process, Greatrex demonstrated the classical methods Sinologists adopt in reviewing texts and words. He also extended the search to include a Tang copy of Huainanzi discovered in Japan not long before he wrote his article. Interestingly, that Tang version omitted the word tu. Although the question as to whether da was in use during the Western Han is still open, Greatrex undertook a thorough search to identify how historical authors may have misinterpreted this character. Sinologists write to answer questions such as: What could have happened? What copies or editions did the historical authors use? Did they go wrong and how did the inaccurate interpretation come about? In his concluding line, Greatrex boldly claimed that “Duan Yucai’s emendation of the gloss on ting in Shuowen jiezi appears to be hasty,” and he continued: “[T]ing appears neither to have been a verb, nor an alternative orthography for da [] without further evidence [...] the suggestion that da may have been included in the original Shuowen jiezi seems to be quite untenable.”[60]

An additional word may be added to Greatrex’s finding. The same line was preserved in the Beida copy of Cang jie pian; here the word  zhu, bamboo, took the place of tin’. Together, the line reads as “Ba Shu tu (hollow bamboo) zhu” (bamboo) 巴蜀荼竹. This line makes much more sense, and it corroborates with a preceding line that reads as “Wu Yue gong (paper) zhi (weaving)” 吳越貢織. The first couples in both lines are names of places, whereas the following couples refer to the renowned productions from those places. Examining the photographed Fuyang version reveals that the word ting was not clearly preserved, so that the first reading of it in 1983 had mistaken it as ting. In his later writings, Hu Pingsheng rectified the reading and used zhu in place of ting.[61] As for whether da was in use during the Qin and Western Han periods, archaeological evidence is still pending.

Hanxue, Guoxue and Archaeology

Guoxue has been a very contested term since Liang Qichao 梁啟超 (1873–1929) and his contemporaries first coined it during the last years of Qing. Guoxue, the National Learning, did not specify any particular field of study, although it was conventionally accepted that the term refers to the studies of Chinese history, philosophy and language. Identifying Luo Zhenyu and Wang Guowei as guoxuejia would be uncontroversial. But guoxue, according to the historical studies about the Republican period, represented more of an academic campaign than a consolidated school of academic thought or training. It was originally advocated to safeguard and treasure something traditional and central in Chinese culture, in light of the overwhelming Xixue 西學, the Western Learning, during the early decades of the 20th century.[62] In the late 1990s, guoxue was once again promoted, but this time it was heavily intertwined with a political mission to promote Chinese culture both in China and overseas. A number of guoxue or guoxue-related research institutes were established in academia;[63] renowned Sinologists were invited to participate in grand academic events; the studies of classics and of excavated texts became what the cultural media called the guoxue re 國學熱, the National Learning Fever.[64]

Ironically, contemporary Chinese scholars central to National Learning largely oppose identification with the term guoxue. Qiu Xigui, an expert in ancient Chinese writing, emphasises that he and his generation of scholars are working on new (archaeological) materials, although their research ought to be established on a solid knowledge of the linguistic and semantic findings of the last generations of scholars of the early 20th century or before.[65] Throughout the interview, Qiu did not clarify why a difference exists between scholars who use excavated texts and those who do not. He repeatedly used the terms xin (new) and jiu (old) in referring to research materials, rather than to research methodologies. This reveals that Qiu Xigui considers guoxue an outdated practice, limited to pre-Republican times. Li Xueqin 李學勤 (1933–2019), by contrast, was less opposed to the idea of guoxue, or xin guoxue 新國學. He considered it a neutral term that describes the studies of China, much in the same way as sinology does. He did not comment on the difference, if any, between the two.[66]

In the field of Chinese archaeology, the historiography-oriented approach raised by Lothar von Falkenhausen in 1993 has been a rather widely accepted view about Chinese scholarship.[67] This applies in particular to the ongoing debates over the archaeological existence of a Xia dynasty, known from received texts. Most Chinese archaeologists believe that the Xia existed, and some even argue that the present site at Yanshi Erlitou 偃師二里頭 in Henan Province was its capital city. By contrast, most non-Chinese scholars show a tendency to believe that the Xia is yet to be identified by archaeological excavations. The arguments are built on the different interpretations of historical writings. The criticisms against Chinese scholars are inevitably intertwined with accusations of nationalism or of showing a strong preference for the centre-periphery model of developments in Chinese civilisation. Such debates have often led to extended prejudices and biases over other works. For as long as new archaeological finds about the Xia, or of those dated to and before the period of the Erlitou culture (ca. 1900 – ca. 1500 BC), are pending, such debates will produce nothing constructive.

The present article, therefore, hopes to offer a different view. Cang Jie pian is an example of Chinese scholars and Sinologists working together, even closely. The writings by Hu Pingsheng, Zhu Fenghan and other young scholars reveal a strong belief in their subject matter. The drive to exhaust every possible detail about the book is so strong that most scholars have focused on revealing what was lost or overlooked in antiquity. Except for the writing of Zhu Fenghan, authors’ works rarely extend to discussing the book’s meaning in its cultural contexts. It must also be noted that because they are writing in the Chinese language, their primary readership is their fellow Chinese. To most Chinese scholars, filling up the blanks in history is of primary importance.

By contrast, those who are working outside China have no direct access to the primary study materials. As such, they are not burdened with the responsibility of making a detailed account of the excavated texts. Above all, the studies delivered in Japanese, English (or other European languages) are intended to introduce and explain what these texts have revealed about early China. Whether or not there existed an older or a more standardised banben, for example, may not be their intended readers’ primary concern. These scholars set out to discover how a word, an event and a concept came to take root in Chinese culture. The process of creation and its impacts is far more appealing to Sinologists than to Chinese scholars themselves. In the Encyclopedia of Modern China, Harriet Zurndorfer defines Sinology as the “study of pre-modern Chinese civilization through philological and literary analysis.”[68] To that it must be added: Sinology is not a study of China’s past for the sake of historical interests; it is a study of the past to understand how Chinese culture and China arrived at their present. The difference between the two scholarly camps, if they could be loosely defined, derives not from thought or methodology, but from whom they intend to speak to.

In 2018, Edward Shaughnessy published a book on the intellectual history of sinologists who work on excavated texts. The book was first published in Chinese as Xi guan Han ji 西觀漢記, which the author himself translated as Chinese Annals in the Western Observatory.[69] Guan, which means observing or looking from a distance, is perhaps the best word to describe how sinologists work. Another book by the same author published years ago reads as: Wen gu zhi xin 溫故知新, “review the old and get to know the new.” The phrase suggests that the new ought to be built on the foundation of the old, much in parallel with what Qiu Xigui comments about his profession. Guan and wen are two separate attitudes – and they explain fairly well how Sinologists and Chinese scholars differ respectively.


Primary Sources

Beijing daxue cang Xi Han zhushu: Juan yi 北京大學藏西漢竹書——卷一. 2015. Beijing daxue chutu wenxian yanjiusuo 北京大學出土文獻研究所 (ed.). Shanghai: Shanghai guji chubanshe.

Dunhuang Hanjian 敦煌漢簡. 1991. Gansu sheng wenwu kaogu yanjiusuo甘肅省文物考古研究所. (ed.). Beijing: Zhonghua shuju.

Hanshu 漢書. 1962. Ban Gu 班固 et al. (comp.). Beijing: Zhonghua shuju.

Juyan xinjian: Jiaqu houguan yu disisui 居延新简——甲渠候官与第四燧. 1990. Gansu sheng wenwu kaogu yanjiusuo甘肃省文物考古研究所, Gansu sheng bowuguan 甘肃省博物馆, Wenhuabu guwenxian yanjiushi 文化部古文献研究室 and Zhongguo shehui kexueyuan lishi yanjiusuo 中国社会科学院历史研究所 (eds.). Beijing: Wenwu chubanshe.

Liusha zhuijian 流沙墜簡. 1914, reprint in 1993. Luo Zhenyu 羅振玉and Wang Guowei 王國維. Beijing: Zhonghua shuju.

Zhongguo jiandu jicheng中國簡牘集成, vol. 18. 2005. Zhongguo jiandu jicheng bianji weiyuanhui 中國簡牘集成編輯委員會 (ed.). Lanzhou: Dunhuang wenyi chubanshe.

Secondary Sources

Barbieri-Low, Anthony J. – Robin D.S. Yates. 2015. Law, State, and Society in Early Imperial China: A Study with Critical Edition and Translation of the Legal Texts from Zhangjiashan Tomb no. 247. Leiden: Brill.

Bodde, Derk. 1967. China’s First Unifier: A Study of the Ch’in Dynasty as Seen in the Life of Li Ssŭ. Hong Kong: Hong Kong University Press.

Bottero, Françoise. 2002. “Revisiting the wen and the zi: The Great Chinese Characters Hoax.” BMFEA 74 (2002), pp. 14–33.

———. 2006. “Cang Jie and the Invention of Writing: Reflections on the Elaboration of a Legend.” In: Christoph Anderl and Halvor Eifring (eds.), Studies in Chinese Language and Culture. Oslo: Hermes Academic Publishing, pp. 135–155.

Bottero, Françoise – Christoph Harbsmeier. 2008. “The Shuowen jiezi Dictionary and the Human Sciences in China.” AM 21 (2008) 1, pp. 249-271.

Chen Fuhua 陈复华 He Jiuying 何九盈. 1987. Guyun tongxiao 古韵通晓. Beijing: Zhongguo shehui kexue chubanshe

Dai Yan 戴燕 (ed.). 2017. Zhibi jingshan: Shiyi wei Zhong-wai xuezhe fangtanlu陟彼景山——十一位中外学者访谈录. Beijing: Zhonghua shuju.

Falkenhausen, Lothar von. 1993. “On the Historiographical Orientation of Chinese Archaeology.” Antiquity, vol. 67, issue 257 (1993), pp. 839–849.

Foster, Christopher J. 2017a. “Study of Cang Jie pian: Past and Present.” Unpublished Ph.D. thesis. Harvard University.

———. 2017b. “Introduction to the Peking University Han Bamboo Strips: On the Authentication and Study of Purchased Manuscripts.” EC 40 (2017), pp. 167–239.

Fukuda Tetsuyuki 福田哲之. 1989. “Seoketsu henno naiyo to kozo: Fuyo Kankan SoketsuHen wo chushin to shite” 『蒼頡篇』の内容と構造——阜陽漢簡『蒼頡篇』を中心 として.” In: Nihon Chūgoku gakkai hō 日本中国学会報 / Bulletin of the Sinological Society of Japan 41 (1989), pp. 223–239.

Fuyang 1983a, 1983b, see Wenwuju guwenxian yanjiushi – Anhui sheng Fuyang diqu bowuguan fuyang Han jian zhenglizu.

Greatrex, Roger. 1994. “An Early Western Han Synonymicon: The Fuyang Copy of the Cang Jie pian.” In: Joakim Enwall (ed.), Outstretched Leaves on His Bamboo Staff: Studies in Honour of Göran Malmqvist on His 70th Birthday. Stockholm: The Association of Oriental Studies, pp. 97–113.

Hu Pingsheng 胡平生. 1996. “Han jian Cang Jie pian xinziliao de yanjiu” 漢簡《倉頡篇》新資料的研究. In: Zhongguo shehui kexueyuan jianbo yanjiu zhongxin 中國社會科學院簡帛研究中心 (ed.), Jianbo yanjiu 簡帛研究, vol. 2. Beijing: Falü chubanshe, pp. 332–349.

———. 2007. “Yingguo guojia tushuguan cang Sitan’en (Stein) suo huo jiandu zhong de Cang Jie pian canban yanjiu” 英國國家圖書館藏斯坦因所獲簡牘中的《蒼頡篇》殘版研究. In: Wang Tao 汪濤, Hu Pingsheng 胡平生, and Frances Wood (eds.), Yingguo guojia tushuguan cang Sitan’en (Stein) suo huo weikan Hanwen jiandu 英國國家圖書館藏斯坦因所獲未刊漢文簡牘. Shanghai: Shanghai cishu chubanshe, pp. 62–75.

———. 2012. “Du Shuiquanzi Han jian qiyanben Cang Jie pian讀水泉子漢簡七言本《倉頡篇》. In: Hu Pingsheng胡平生. Hu Pingsheng jiandu wenwu lungao胡平生簡牘文物論稿.  Shanghai: Zhongxi shuju, pp. 42–51

———. 2016. “Du Beida Han jian Cang Jie pian zhaji” 讀北大漢簡《蒼頡篇》札記. In: Chutu wenxian yanjiu 出土文獻研究, vol. 15. Shanghai: Zhongxi shuju, pp. 280–297.

Hu Pingsheng 胡平生 – Han Ziqiang 韩自强. 1983. “Cang Jie pian de chubu yanjiu” 《苍颉篇》的初步研究. WW 1983/2, pp. 35–40.

Huang Xiaoxiao黄潇潇. 2017. “Chutu jiandu wenxian Cang Jie pian yu Shuowen duibi yanjiu” 出土简牍文献《苍颉篇》与《说文》对比研究. Lantai shijie兰台世界 2017/6, pp. 25–28

Lao Gan 勞榦. 1949. Juyan Han jian kaoshi: Yiwen zhi bu 居延漢簡考釋——釋文之部. Shanghai: Shangwu yinshuguan.

Li Quanxing. 李權興. 2011. “20 shiji shang ban shiji guanyu guoxue dingyi de zhenglun” 20世紀上半世紀關於國學定義的爭論. Published online, July 17, 2011, accessed July 20, 2020).

Li Xueqin 李学勤. 2002. “Shishuo Zhangjiashan Shilü试说张家山《史律》. WW 2002/4, pp. 69–71.

Liang Jing 梁靜. 2011a. “You Beida Han jian chakao Cangjie pian liuchuan zhong de yichu yiwen由北大漢簡查考《蒼頡篇》流傳中的一處異文. Published online, July 19, 2011, (accessed Aug. 14, 2020).

———. 2011b. “Chutu ben Cang Jie pian duidu yize” 出土本《蒼頡篇》對讀一則. Published online, Aug 5, 2011, (accessed Aug. 14, 2020).

———. 2015. “Chutu Cang Jie pian yanjiu” 出土《苍颉篇》研究. Beijing: Kexue chubanshe.

Lin Suqing 林素清. 1987. “Cang Jie pian yanjiu” 蒼頡篇研究. Hanxue yanjiu 漢學研究 5 (1987) 1, pp. 53–74.

Luo Jianqiu 羅檢秋. 2000. “Ye shuo ‘guoxue’” 也說「國學」. Wenshi zhishi 文史知識2000/1, pp. 103–108.

Luo Zhitian 罗志田. 2002. “Guoxue bushi xue: Xifang xueshu fenlei yu minchu guoxue dingwei de kunhuo” 国学不是学——西方学术分类与民初国学定位的困惑. Shehui kexue yanjiu 社会科学研究2002/2, pp. 117–121.

Ning He 宁赫. 2005. “Cang Jie pian yanjiu” 《仓颉篇》研究. Unpublished M.A. thesis. Dongbei shifan daxue.

Nivison, David. 2002. “The Xia-Shang-Zhou Chronology Project: Two Approaches in Dating.” Journal of East Asian Archaeology 2002/4, pp. 359–366.

Onishi Katsuya. 2018. “An Investigation of Clerical Script in Chu Regions during the Qin and Han Periods, and Its Relationship to ‘Scribal Writing’.” Bamboo and Silk 1 (2018), pp. 359–402.

Rosthorn, Arthur von. 1975. “The Erh-ya and Other Synonymicons.” Journal of the Chinese Language Teachers Association, 10 (1975) 3, pp. 137–145.

Shaughnessy, Edward. 1997. Wen gu zhi xin lu: Shang Zhou wenhuashi guanjian 溫故知新錄——商周文化史管見. Taipei: Daoxiang chubanshe.

———. 2018. Xi guan Han ji: Xifang Hanxue chutu wenxian yanjiu gaiyao 西觀漢記——西方漢學出土文獻研究概要. Shanghai: Shanghai guji chubanshe.

———. 2019. Chinese Annals in the Western Observatory: An Outline of Western Studies of Chinese Unearthed Documents. Boston – Berlin: de Gruyter Mouton.

Wang Chongmin 王重民. 1933. “Cang Jie pian jiben shuping” 蒼頡篇輯本述評. Furen xuezhi 輔仁學志 4 (1933) 1, pp. 1–16.

Wang Guowei 王國維. 1919; reprint in 1983. “Chongji Cang Jie pian” 重輯《蒼頡篇》. In: Wang Guowei yishu 王國維遺書, vol. 7. Shanghai: Shanghai guji chubanshe.

Wang Xiangtian 王襄天 – Han Ziqiang 韩自强. 1978. “Fuyang Shuanggudui Xi Han Ruyinhou mu fajue jianbao 阜阳双古堆西汉汝阴侯墓发掘简报. WW 1978/8, pp. 12–31.

Wang Yue 王樾. 1998. “Lüeshuo Niya faxian de Cang Jie pian Han jian” 略说尼雅发现的《苍颉篇》汉简. Xiyu yanjiu 西域研究 1998/4, pp. 55–58.

Wenwuju guwenxian yanjiushi 文物局古文献研究室 – Anhui sheng Fuyang diqu bowuguan Fuyang Han jian zhenglizu 安徽省阜阳地区博物馆阜阳汉简整理组. 1983a. (Fuyang 1983a). “Fuyang Han jian jianjie” 阜阳汉简简介. WW 1983/2, pp. 21–23.

———. 1983b. (Fuyang 1983b). “Fuyang Han jian Cang Jie pian 阜阳汉简《苍颉篇》. WW 1983/2, pp. 24–34.

Zhang Biao 张标. 1990. “Fuyang chutu Cang Jie pian de ruogan wenti” 阜阳出土《仓颉篇》的若干问题. Hebei shifan daxue xuebao (Shehui kexue ban) 河北师范大学学报(社会科学版) 1990/4, pp. 7–11.

Zhang Chuanguan 張傳官. 2016a. “Du Beida Han jian Cang Jie, Jie, Yu Shu deng pian zhaji” 讀北大漢簡《倉頡》、《節》、《雨書》等篇札記. In: Chutu wenxian yanjiu 出土文獻研究, vol. 15. Shanghai: Zhongxi shuju, pp. 298–308.

———. 2016b. “Ju Beida Han jian pinzhui, bianpai, shidu Fuyang Han jian Cang Jie據北大漢簡拼綴、編排、釋讀阜陽漢簡《倉頡》. Chutu wenxian 出土文獻, 8 (2016), pp. 165–189.

———. 2019. “Beida Han jian Cang Jie pian pinhe yize北大漢簡《蒼頡篇》拼合一則. Published online, Nov 28, 2019, (accessed July 27, 2020).

Zhang Cunliang 張存良. 2010. “Shuiquanzi Han jian qiyanben Cang Jie pian lice” 水泉子漢簡七言本《倉頡篇》蠡測. Chutu wenxian出土文獻研究, 9 (2010), pp. 60–75

———. 2012. “Shuiquanzi Han jian Cang Jie pian de wenzi ji shufa tedian” 水泉子漢簡《蒼頡篇》的文字及書法特點. Jianbo yuyan wenzi yanjiu 簡帛語言文字研究 6 (2012)pp. 64-69.

———. 2015. “Cang Jie pian de banben, liuchuan, wangyi he zai faxian” 《苍颉篇》的版本、流传、亡佚和再发现. Gansu shehui kexue 甘肃社会科学2015/1, pp. 89–94.

Zhang Cunliang 張存良 – Ju Hung 巨虹. 2015. “Cang Jie pian yanjiu de xin jinzhan” 《蒼頡篇》研究的新進展. Chutu wenxian yanjiu 出土文獻研究 14 (2015), pp. 232–258.

Zhang Cunliang 张存良 – Wu Hong 吴荭. 2009. “Shuiquanzi Han jian chushi” 水泉子汉简初识. WW 2009/10, pp. 88–91.

Zhao Yingyi. 2020. “Interview of Michael Loewe.” Dated June 21, 2020, (accessed June 21, 2020).

Zhou Fei 周飛. 2016. “Chutu Cang Jie pian banben tantao” 出土《蒼頡篇》版本探討. Chutu wenxian 出土文獻 8 (2016), pp. 190-200.

Zhu Fenghan 朱凤瀚. 2011. “Beida Han jian Cang Jie pian gaishu” 北大汉简《苍颉篇》概述. WW 2011/6, pp. 57–63.

———. 2015. “Beida Han jian xin qishi” 北大漢簡新啓示. In: Beijing daxue chutu wenxian yanjiusuo (ed.). Beijing daxue cang Xi Han zhushu: Juan yi 北京大學藏西漢竹書——卷一. Shanghai: Shanghai guji chubanshe, pp. 170–180.

Zhu Fenghan 朱鳳瀚 – Han Wei 韓巍 – Chen Kanli 陳侃理. 2011. “Beijing daxue cang Xi Han zhushu gaishuo” 北京大學藏西漢竹書概說. WW 2011/6, pp. 49–56.

Zurndorfer, Harriet. 2009. “Sinology.” In: David Pong (ed.), Encyclopedia of Modern China. Detroit: Scribner, pp. 1673–1675.

Notes on Contributor

Celine Y.Y. Lai 黎婉欣, D.Phil. in Archaeology, is currently Associate Professor and Head of the Museum and Cultural Heritage Stream in the School of Archaeology and Museology, Peking University. Her research interests include Ancient Chinese Bronzes, Archaeological Theories, Chinese Art History, and History of Museums. Some of her recent publications include “On the Writing of Museum Texts: Inspirations from International Museum Label Writing Contests” (Chinese Museum, 2020/1, written in Chinese); Contacts Between the Shang and the South: Resemblance and Resistance (Archaeo Press, 2019); and “UNESCO and Chinese Heritage: An Ongoing Campaign to Achieve World-Class Standards” (in Poul Duedahal ed., Making a Difference: Seventy Years of UNESCO Action, Palgrave Macmillan, 2016).

Correspondence to: Celine Lai, School of Archaeology and Museology, Peking University Beijing, China, 100871. Email:


[1] This article was first presented at the international workshop “Sinology – China/Chinese Studies – Guoxue: Their Interrelations, Methodologies and Impacts,” held at Siegburg, Germany on October 21–22, 2019. In the opening speech, Professor Zbigniew Wesołowski and Dr. Huang Mei Ting addressed the purpose of the workshop and observed that “shifts in perspectives were and are accompanied by thorough reflections on the self-conception of the study of China as an academic discipline, its merits and limits, intrinsic and extrinsic motivations, aims and methodology. Which changes in methodology can be observed? Are there mutual influences and exchanges in methodology between the different perspectives on China?” The organisers intended to offer an “assessment of the state of the research on China.” I was responsible for sharing thoughts in the field of archaeology. Sinologists and Chinese scholars often collaborate on translation projects of archaeological works or on excavation projects. Despite the frequent academic exchange, their approaches to studying excavated materials still differ significantly, as this paper will show. Here I am using a recently discovered text for comparative discussion. For other renowned cases in the field of archaeology, which I will return to in the last section, see, for example, David Nivison’s article, which questions China’s state-funded project in calculating absolute dates of the early dynasties, Nivison 2002, pp. 359-366.

[2] Archaeology was introduced to China at the end of the 19th century. About the same time, Sinological ways of thinking and doing research were becoming established in the circle of Chinese scholars in the 1920s, culminating in the founding of Academia Sinica, which included an Institute of History and Philology (IHP). The institute’s name suggests that researchers adopted a new approach to exploring China’s past, opening up to areas other than history. The IHP was set up to focus on four areas of study, namely history, philology, archaeology and anthropology. Fu Sinian 傅斯年 (1896–1950) was the first director as well as a key excavator of the archaeological site at Anyang, which was confirmed as the last capital of the Shang dynasty (ca. 1500–ca. 1050 BC). Most importantly, the Anyang excavation team unearthed pits fully buried with inscribed oracle bones, which is still the earliest form of Chinese writing known so far. The inscriptions may corroborate some of the historical accounts in Shiji 史記 (Historical Records). Because of this background, archaeology was, and still is, understood as a branch or a twin brother of history in China.

[3] For the study of other synonymicons of the same period, see Rosthorn 1975, pp. 127–145.

[4] The early Western Han copy of Cang Jie pian was said to have contained 3,300 words. Yang Xiong 揚雄 (58 BC – 18 AD) developed a Xun zhuan pian 訓纂篇based on it, and it contained 5340 words. Wang Chongmin suggests that Xu Shen 許慎 (58–148) was consulting Yang’s work, which incorporated Cang Jie pian. Thereafter, Du Lin 杜林 (d. 47 BC) further expanded it and developed a combined work of three books, which was entitled the Sancang 三蒼, that contained 7,380 words altogether. This San Cang was known in later historical writings until the end of the Tang dynasty. For the textual history of Cang Jie pian, see Wang Chongmin 1933, pp. 1–3 and Greatrex 1994, pp. 100–104.

[5] Sinologists have often doubted the authenticity of purchased bamboo strips; it is also argued the purchase may bear ethical issues in relation to looting and antique trade. Christopher Foster has conducted a thorough study of Peking University’s purchase, as well as many other previous cases involving other museums and universities; Foster 2017b, pp. 167–239. The ethical issues fall outside of the present discussion’s scope. The Beida manuscripts appeared to have just been looted in light of their condition on arrival; authenticity can be guaranteed by scholars’ expertise, scientific tests and the contents and correlation of the texts in the same collection.

[6] Roger Greatrex’s article (1994) was almost the only work to offer a detailed explanation of Cang Jie pian in the English language. In his 2017 doctoral thesis, Christopher Foster recently delivered an excellent study of all Cang Jie pian excavated so far. Foster discussed the arguments over the authenticity of the Beida collection of Han bamboo strips, the details of the copy of Cang Jie pian excavated from Yumen Huahai 玉門花海 in Gansu Province, and the use of vocabulary in this elementary school textbook. The author would like to thank Zhu Fenghan for the courtesy of reference.

[7] Françoise Bottero examines this legend and traces the earliest sources to Xunzi and Han Feizi, hence suggesting that the legend was likely composed or stabilised in the third century BC; Bottero 2006, pp. 135–155.

[8] Zhu Fenghan 2015, p. 173.

[9] The English translation of the official titles follows Greatrex’s version, 1994, pp. 100–104; see also Bodde 1967, pp. 147–161.

[10] Hanshu 30.1721. Hanshu called the folk or village teachers lüli shushi 閭里書師, who were the first teachers of elementary schools. This remark suggests that Cang Jie pian would have been used and distributed widely across the Han territory. The Hanshu quotations here are adopted from the Zhonghua shuju edition published in 1962.

[11] According to the findings of Hu Pingsheng and Han Ziqiang, Jiu Tangshu 舊唐書 made a note of Sancang and the same book was again mentioned in Xin Tangshu新唐書. Although San Cang had been developed from the Eastern Han editions of Cang Jie pian and much removed from the copy of the Western Han period or earlier, Hu and Han argue that the date of the Xin Tangshu, which is ca. 1060, may serve as the last seen date of Cang Jie pian. Hu – Han 1983, pp. 35–40.

[12] The authors remarked that the bamboo strips and other documents were discovered at the sites located at 40 degrees north and 93.10–94.30 degrees east; Luo – Wang 1914, reprinted in 1993, pp. 75–78.

[13] Wang Guowei 1919, reprinted in 1983, pp. 3–14.

[14] Hanshu 30.1719. Of the 541 words found in the Fuyang copy of Cang jie pian, 270 were also found in Ji jiu pian. Lin Suqing argues that Ji jiu pian may not have followed Cang jie pian as faithfully as Wang Guowei had presumed, because Ji jiu pian appeared to have intentionally excluded some words collected in Cang jie pian. The two works, as Lin suggests, followed quite different approaches in selecting characters. Ji jiu pian showed a tendency to collect words that were still in use at the time of its compilation; Lin Suqing 1987, pp. 69–71. However, because Cang Jie pian and Ji jiu pian were set apart by nearly two hundred years, the omission or revision of outdated words would have been a natural decision. The similarity in the two texts is still quite obvious. For the studies about Cang Jie pian and the attempts at restoration from other synonymicons by Qing scholars, see also Lin Suqing 1987, pp. 53–58.

[15] Lao Gan, 1949, pp. 559–561.

[16] Luo – Wang 1913 (reprint in 1993).

[17] Wang Tao et al. 2007.

[18] Lao Gan 1949.

[19] Gansu 1990.

[20] Gansu 1991.

[21] Gansu 1991.

[22] Wang Yue 1998.

[23] Zhang – Wu 2009.

[24] Bodde 1967, pp. 147–161.

[25] The bronzes and lacquered articles from the tomb were inscribed “Ruyin”; other absolute dates inscribed suggested the tomb would have fallen in the early reign of the second Han Emperor, Wendi 文帝 (r. 180–157BC); Wang – Han 1978, pp. 17–18.

[26] Hanshu 41.2076–2077 and 28.1561–1562.

[27] His tomb was found underneath a tumulus; his wife’s tomb was buried next to his, also covered by a tumulus. The tumuli would have served as obvious landmarks, because the area is called Xuanggudui – “two old mounds.” Both tombs had been plundered before excavation. Only the Marquis was buried with bamboo strip scrolls; the lacquered remains found among them suggested that the books would have been placed in lacquered boxes before burial, Wang – Han 1978, pp. 12–17; and Fuyang 1983a, pp. 21–23.

[28] The longest line is found on the bamboo strip numbered as C001 (where C represents Cang Jie pian, same below), Fuyang 1983b, pp. 24–34. This copy of Cang Jie pian was again published in Zhongguo jiandu jicheng (Jiandu 2001), vol. 18, pp. 1656–1674, with updated footnotes prepared by Hu Pingsheng and Han Ziqiang.

[29] Chen – He 1987, pp. 129–131.

[30] The last word of the third line carried a juon the right hand side, but the radical was unclear; Fuyang 1983b, pp. 25 and 28–29.

[31] Hu and Han 1983, pp. 37–38.

[32 Zhang Cunliang and Ju Hung, however, do not agree with this argument, because the handwriting of the Fuyang copy shows signs of the clerical script used during the Han. They therefore suggest that the Fuyang copy may have been one of the early Western Han copies, rather than the copy handed down directly from Qin; Zhang – Ju 2015, pp. 239–240.

[33] Zhang Cunliang points out that C003 of the Fuyang copy carried the word zheng , which was the name of the First Emperor. This may contradict Hu Pingsheng’s and Han Ziqiang’s suggestion that the Fuyang copy would have been one of the Qin editions; Zhang – Ju 2015, pp. 251–252 and Zhang Cunliang 2015, pp. 93–94. In the same line, both the Beida copy and that from Shuiquanzi in Gansu had adopted dan in place of zheng, which was likely to have been a practice handed down from the Qin edition. The several copies of Cang Jie pian obviously have inconsistent word choices. Zhu Fenghan, therefore, suggests that scholars attempted corrections during the early Western Han, but may not have executed them strictly; Zhu Fenghan 2015, pp. 176–177.

[34] The first two words of the fourth line had mu as the radicals, but the rest were unclear; Fuyang 1983b, pp. 29–30.

[35] See, for example, the word die , which referred to joint dislocation, on C025; and qiang , which was an alternative orthography of qiang , meaning walls, were also found on the bamboo strips from Yunmeng, Fuyang 1983b, pp. 29–30.

[36] According to Shuowen jiezi, qu meant ji, meaning quick or hurry.

[37] Greatrex 1994, p. 105.

[38] Hu – Han 1983, pp. 38–39; and Greatrex 1994, pp. 104–106.

[39] Among the 541 words found in the Fuyang copy, nine of them were found repeated. Nearly 95% of the preserved lines were composed on the basis of word meanings, while a few were “written” to make an introduction. One example comes in the opening lines on C002, read as []兼天下,海內并廁 – “Han united the empire. [It] encompassed the world within its borders (Greatrex’s translation).” Hu Pingsheng and Han Ziqiang identified these introductory lines as the “declarative” type (chenshu shi 陳述式), while the rest are the “listing” type (luolie shi 羅列式, both Foster’s translation). The authors of Cang Jie pian, however, may not have planned for this division. When turning a character primer into a textbook, adding a few meaningful lines to connect sections would add some sense to the text. See also Fukuda 1989, pp. 223–239.

[40] See, for example, Zhang Biao 1990, pp. 7–11; Ning He 2005; Zhang Cunliang 2015, pp. 89–94; and Liang Jing 2015, pp. 26–46.

[41] In 1996, pp. 332–349, for example, Hu Pingsheng recaps the compositional style of Cang Jie pian, in light of the bamboo strips excavated from Juyan in Inner Mongolia and Yumen Huahai in Gansu during the late 1970s. In 2007, pp. 62–75, Hu renders his transcription of the fragments of Cang Jie pian in the bamboo strips collection of the British Library, and offers a comparison of the corresponding lines in the Fuyang copy. See also Hu Pingsheng 2012, pp. 42–51 and 2016, pp. 280–297.

[42] Zhu Fenghan et al., 2011, pp. 49–56.

[43] Zhu Fenghan 2011, pp. 59–61.

[44] Zhu Fenghan et al. 2011, pp. 49–56.

[45] A few of the 87 found pieces were rejoined, so the official publication gives 79 labels. Zhang Chuangguan has recently proposed rejoining a few more of the pieces; Zhang Chuangguan 2019.

[46] Seven words were found to have been repeated in this copy. They were Han (Beida 008, 0057); Li (008, 011), ju (008, 028), li/hui (012, 023), cong (008, 014), zhan (043, 048) and kui3F3(016, 072); Zhu Fenghan 2015, p. 177. Han means both the Han River and the Han dynasty; ju to means both mince and the pickled. But the others may not carry more than one meaning. It is still unclear why these repetitions occurred,

[47] The same method of entitling a text was also found in the bamboo strips excavated from the Qin tomb at Yunmeng Shuihudi in Hubei Province; Zhu Fenghan 2015, p. 173.

[48] By contrast, the fragments of Cang Jie pian found in Dunhuang and Juyan likely belonged to the Han edited version, in which each section contained sixty words. Unlike the copies of Cang Jie pian found in the tombs in the east, those discovered at the watchtower sites contained many repeated words, and the handwriting looked immature. The fragments from the northwest borders, therefore, were likely to have been remains of practice books. Cang Jie pian was quite popularly employed in elementary education during the Western Han; Zhang – Ju, 2015, pp. 233–245.

[49] Zhu Fenghan 2015, pp. 176–177.

[50] Zhang Chunliang 2015, pp. 238–251.

[51] In Chinese academia, master-level and doctoral students are similarly trained to conduct word-by-word analysis in textual studies; see, for example, Ning He 2005, who was trained at Northeast Normal University; Liang Jing 2015, at Peking University; Zhou Fei 2016, at Tsinghua University; Zhang Chuanguan 2016a, pp. 298–308 and 2016b, pp. 165–189, at Fudan University; and Huang Xiaoxiao 2017, at Beijing Language and Culture University.

[52] Zhu Fenghan 2015, pp. 178–179.

[53] For details about the accounts on Zhangjiashan bamboo strips, see Foster 2017a, pp. 233–248.

[54] Li Xueqin 2002, pp. 69–72.

[55] Foster 2017b. Foster spent an exchange year at Beida, where he got in close contact with the team working on the Beida bamboo strip collection. His study perspective, however, differs from theirs. Compare with, for example, the articles published in the same journal as Beida’s Cang Jie pian, Wenwu 2011/6. Foster’s discussion of scribal training also owes tremendously to Anthony Barbieri-Low’s and Robin Yates’s work on the Zhangjiashan manuscripts (2015).

[56] The term guoxue jia in the interview was raised by the interviewer. Michael Loewe uses it to refer to what he calls the Chinese colleagues in general: “Interview of Michael Loewe,” prepared by Zhao Yingyi, dated Jun 21, 2020, (accessed June 21, 2020).

[57] One chi equals 10 cun and gives approximately 23.1 cm.

[58] Greatrex 1994, pp. 98–99.

[59] Da was not found in Shuowen jiezi, but it was included in Guangya 廣雅 by Zhang Yi 張揖 as well as in Fangyan 方言by Guo Pu 郭璞 (276–324). Da was also found in the Han texts, as in the Meng fu 夢賦 (The Rhapsody of Dreams) by Wang Yanshou 王延壽. These were what Greatrex discovered about da, Greatrex 1994, p. 107.

[60] Greatrex 1994, p. 112. Similar arguments against established understanding are also found in the works of other Sinologists; see, for example, how Françoise Bottero traced the concepts of wen and zi in the preface of Shuowen jiezi, Bottero 2002, pp. 14–33 and the development of the legend about Cang Jie in the writings of Xun Zi, Bottero 2006.

[61] Hu Pingsheng 2007, p. 69; see also Zhang Chunliang 2015, pp. 243–245. The anonymous reviewer of this paper points out that ting is found on C071 of the Beida copy. The line reads as ta, jian; bi, ting, (蹋) 踐 (奰)朾, which contain two synonym pairs, (the characters in brackets are the interchangeables) meaning step, tread; strength, pole, respectively. Here, ting works with bi to indicate support and strength; see Zhu Fenghan 2015, p. 138. Ting is unconnected with da, as Greatrex verifies.

[62] For the origin of the term guoxue and the discussion about it in the early 20th century, see Luo Zhitian 2002, pp. 117–126 and Li Quanxing 2011. Although both authors offer a good account of the Republican resources related to guoxue, neither describes what guoxue was in terms of research subjects or methodology. Whether archaeology could be included was ambiguous. Compare with, e.g., Luo Jianqiu (2000), who argues that guoxue was, and still is, referring to jing, shi, zi, ji (Classics, History, Philosophy and Literature). However, Luo also argues that any research related to China could be identified as a part of guoxue, even though the research may have been undertaken under the inspiration of a Western-developed theory or methodology. Guoxue, therefore, is ambiguous as it was in the early 20th century.

[63] See, for example, the establishment of the International Sinological Research Centre of Shandong University, which is dedicated to the projects of collecting and publishing different editions of the classics collected in the libraries around the world. A new School of Chinese Classics was established in Remin University in 2014.  

[64] Semi-private websites and organisations (e.g.; were also established to promote not only the study of classics, but also to take it to an extreme of dressing up in Hanfu 漢服 (Han costume), and bowing to teachers and classmates as ancient scholars would do. Sichuan University published a volume entitled xin guoxue 新國學 to promote the study of Chinese culture. Dai Yan (2017) of Fudan University, formerly the editor of ancient texts at Zhonghua shuju 中華書局, interviewed eleven guoxuejia for their views about their profession and this New National Learning trend.

[65] In addition, he suggests calling his profession gudianxue 古典學 (Studies of Classics); Dai Yan 2017.

[66] Dai Yan 2017.

[67] Falkenhausen 1993, pp. 839–849.

[68] Zurndorfer 2009, p. 1673.

[69] Needless to say, the book’s title was inspired by the renowned Dong guan Han ji 東觀漢記, written by Ban Gu. The same work was published in English in 2019.

*原文刊载于Monumenta Serica: Journal of Oriental Studies, 2021, 69.1