last month
next month

Centre for the Study of Manuscript Cultures (CSMC)

12/2013 manuscript  of the month

In Praise of a Chinese Emperor:

A Tangut Manuscript Fragment

The intricate script on the manuscript shown here is Tangut, one of the non-Chinese writing systems that had been invented in northern China in the medieval period. It contains a fragment of a translation of a lost Chinese text consisting of dialogues between the Tang emperor Taizong (r. 626–649) and his able ministers, attesting – along with other similar texts – to the popularity of the cult of Taizong in the Tangut kingdom. So if the Tanguts were so interested in the Chinese literary tradition, why would they try to distance themselves by creating their own script?

Fig. 1: Manuscript Or.12380/2579, recto.
© The British Library > Enlarge

The manuscript consists of a folded leaf about 22 cm tall and 13 cm wide, with writing on both sides. The pages used to be part of a notebook bound in the so-called butterfly form. There are also holes in the edge of the paper where the binding thread kept the leaves together. The manuscript has no punctuation marks, but there are two lacunae which indicate divisions within the text, similar to our modern way of starting a new paragraph. Although the script is highly uniform and was undoubtedly penned by a skilled copyist, even within this small fragment there are two correction marks, which were most likely inserted by the copyist himself who noticed these two mistakes immediately after making them.

The manuscript was found among the ruins of Khara-khoto (Inner Mongolia) by Sir M. Aurel Stein (1862–1943), better known for his acquisition of the Dunhuang manuscripts. The site was first excavated by the expedition of Pyotr K. Kozlov (1863–1935) whose discovery of an enormous library of books in Tangut and Chinese at this site in 1908–1909 caused a sensation in academic circles. Stein came to Khara-khoto in 1914, but was still able to retrieve thousands of fragments, taking these back to London and depositing them at the British Museum. The Tangut manuscript shown here is undated, it probably comes from the 12th or early 13th century.

The Tangut script was invented in 1036 upon the orders of Tangut emperor Li Yuanhao (1003-1048) as part of his state-building efforts. It is thought that the main reason for creating a native script was to be able to compile a Tangut edition of the Buddhist Canon. Another obvious task was to establish a ‘national’ identity that was different from Song China (960-1279). In this, the Tanguts were following the example of their Khitan neighbors who had created two different scripts a century earlier, one of which was largely based on the Chinese script. The Tangut script did not derive from the Chinese but was certainly inspired by it both visually and in terms of the principles according to which the characters were composed. Thus even though Tangut writing was trying to be different from that of China, at the same time it sought to continue the same tradition.

The same duality can be seen in the texts that survive from Khara-khoto. In addition to Buddhist scriptures, there were also many other Chinese works translated and written down with the newly created script, including Confucian classics, encyclopedias, military treatises or even works of literature. It is clear that the Chinese tradition acted as the most important source for Tangut books, even though there were also texts translated from Tibetan or written in Tangut from the start. The manuscript leaf shown here is one of the secular texts translated from Chinese. At first sight this is a rather unassuming leaf, which was perhaps the reason why it lay unnoticed in the Stein collection for a century without inciting scholarly interest. A closer reading reveals that it is a collection of anecdotes (one or two sentences) from the Chinese past, discussing issues related to morality and proper conduct. The stories are all from the Chinese tradition, revealing that the Tangut text is a translation, rather than an original composition.

Fig. 2: Manuscript Or.12380/2579, verso.
© The British Library > Enlarge

What makes this manuscript especially interesting is that the text is arranged in the form of a dialogue between the Chinese Emperor Taizong and his ministers. There are known examples of such a format in Chinese literature, most notably the Essentials of Statecraft of the Zhenguan Reign (Zhenguan zhengyao) which consists of questions asked by Tang Emperor Taizong on statecraft and answers provided by several of his ministers. This text was extremely popular not only in China, but also in other parts of East Asia, most notably Japan where numerous twelfth-century manuscripts survive to this day. More interestingly, the same work was also identified among the Tangut texts from Khara-khoto. This attests to the popularity of the figure of Taizong as an enlightened ruler in the Tangut kingdom. Another related text is a fragmentary manuscript that includes the title Essentials of Selections by Taizong (Taizong zeyao), and this also contains short Chinese stories of a few sentences about proper behaviour. It has been suggested that this may be a primer used by school children to learn to write. Now the name Taizong does not appear in the manuscript, but it does in the title at the end of the text, and this has puzzled researchers. It is possible, however, that this manuscript contains yet another section of the text that appears in the manuscript leaf shown here. The two do not overlap and have a slightly different layout, which shows that even if they were part of the same text, they came from two different copies. Yet their content is analogous and the name of Emperor Taizong in the title of one complements the mention of his name in the main text of the other.

In sum, this fragment is an intriguing manuscript which contains a Tangut translation of a text unknown in its Chinese original. Together with other Tangut texts related to Emperor Taizong, it attests to the popularity of the cult of Taizong among the Tangut population, even though they seem to have invented their native script partly in an effort to distance themselves culturally and politically from Song China. Yet a significant portion of Tangut texts was translated from Chinese, which points to the significance of the Chinese heritage in Tangut culture. The texts connected with the cult of Taizong are further examples of this strong affiliation with the Chinese past.


GALAMBOS, Imre (forthcoming): “A Chinese tract in Tangut translation (Or.12380/2579)”. In: Central Asiatic Journal.

KYCHANOV, Evgeny I. (2004): “Fragmenty perevoda na tangutskiy (Si Sya) yazyk sochineniya U Tszina Chzhen’guan’ chzhen yao”. In: Istoriografiya i Istochnikovedenie Istorii Stran Azii i Afriki, 22, 75–83.

NIE Hongyin 聶鴻音 (2003): “Xixiaben Zhenguan zhengyao yizheng 西夏本《貞觀政要》譯證”. In: Wenjin xuezhi 文津學誌, 1, 116–124.

NIE Hongyin 聶鴻音 (2012): “Xixiaben Taizong zeyao chutan 西夏本《太宗擇要》初探”. In: Ningxia shifan xueyuan xuebao (Shehui kexue) 寧夏師範學院學報 (社會科學), 4, 55–59.

YINGGUO GUOJIA TUSHUGUAN 英國國家圖書館, SHANGHAI GUJI CHUBANSHE 上海古籍出版社 and XIBEI DI’ER MINZU XUEYUAN 西北第二民族學院 (Hrsg.) (2005): Ying cang Heishuicheng wenxian 英藏黑水城文獻, Bd. 3. Shanghai: Shanghai guji chubanshe 上海古籍出版社.

The British Library, London
Shelfmark: Or.12380/2579
Material: Paper, 1 folio with writing on both sides, 8 lines per page with 17–20 characters per line
Dimensions: 22 x 13 cm
Provenance: Khara-khoto (Inner Mongolia), c. 12th century.

Text by Imre Galambos
Images reproduced with kind permission of The British Library.
Large size images are available at the website of IDP (International Dunhuang Project).