Archive for March, 2019


12 March 2019

In my previous blog entry (which I partly wrote as a mise-en-scène for this one), I confessed to having left our albums of model calligraphy (fatie 法帖) until last because they were difficult. This, of course, is what librarians do: shove anything difficult into a cupboard and forget about it – my own cupboard was pretty full when I was dismissed eighteen months ago.

It’s some consolation to find that even Thomas Hyde wasn’t above doing this sort of thing. In his manuscript notes of our Chinese holdings (British Library Sloane Or.853, increasingly quoted in my blog) he has a section headed Praetermissa in Arch. A, literally “Things put aside in Arch[ivum] A” (I won’t repeat the facts about the storage and handling of our earliest Chinese accessions – it’s all explained in an earlier blog entry). I think it’s pretty clear what he meant by “put aside” – one of the items is Sinica 91, something I have put aside these past forty years.

In his notes, Hyde describes the item as follows:

Praetermissa in Arch. A …
182. Liber Fa-tie, continens Calligraphiae exemplaria nitida pro addiscentibus scribere linguam Sinensem.

Things put aside in Arch[ivum] A …
182. A Fa-tie, containing fine examples of calligraphy for those who are learning to write the Chinese language.

In the Bernard catalogue (p.152), we learn a little more about it; it is a roll, and the text appears as white on black:

Rotulae in Archivo A …
2969.18 Liber Sinensis impressus Characteribus albis in charta nigra, continens exemplaria Calligraphiae nitida pro addiscentibus scribere linguam Sinensem.

Rolls in Archivum A …
2969.18 A Chinese printed book with white characters on a black background, containing fine examples of calligraphy for those who are learning to write the Chinese language.

Here is the item as it is currently preserved:


It is bound in a codex, probably by Nicholson. The composition of the codex and the order in which the leaves are presented suggest that in Hyde’s time the Chinese leaves were rolled up in a protective sheet of western paper, and that the whole thing was then rolled in a piece of limp vellum inscribed by Shen and Hyde in the usual way. The inscription is for the most part illegible, that is unless you are Will Poole, for it is he who kindly transcribed and translated it for me within minutes of receiving my e-mail:


A 182
fa Formularius
tie Libellus seu charta
Est libellus pro Institutione eorum qui / primò addiscunt scribere linguam Chinensem, / continens varia Exemplaria rariores / Scripturae tam quadratae quam cursivae. / Anglicè A China Copy-booke.

A 182
法 fa A model
帖 tie album
This is a book for teaching those who are first gaining knowledge of how to write the Chinese language, containing various uncommon examples of writing both squared and cursive. In English: A Chinese copy-book.

The codex contains the first 17 pages of the fatie, but they are not bound in order. Here is the first, which clearly bears its title, Mingshu jixuan fatie 名書集選法帖 (“An album of collected works by famous calligraphers”), and appropriately the very first example is by Wang Xizhi 王羲之 (see my previous blog entry):


The leaf, like all the others, is clearly divided into three panels, with pagination in the lower right corner. Originally they would have been pasted together and folded to make an “accordion” binding (zhezhuang 折裝).

According to Madan and Craster’s Summary catalogue [1], the item was “acquired about 1618”, but I don’t know where this information came from. This means that it must have been printed during the Wanli period at the latest, so I have described it thus in my catalogue:

名書集選法帖 不分卷殘十七葉
洋裝1冊(原活葉) ; 60 x 30公分
Sinica 91

I can find no record of the title in any catalogue, whether printed or online. A Google search for “名書集選法帖” will at the time of writing find only two things: this text in my online list Chinese books in Europe in the 17th century, and a work entitled Mingshu jixuan fatie qianzi wen 名書集選法帖千字文 (the “Thousand character classic”) by the Tang dynasty monk and calligrapher Huaisu 懷素 (737–799) in Osaka Prefectural Nakanoshima Library 大阪府立中之島圖書館 in Japan.

This can only mean either that Sinica 91 is of exceptional rarity, or it isn’t a discrete work, but part of another which I have failed to identify. Either way, it is extraordinary that a work of such quality should have arrived along with the rather cheaper productions of the Jianyang and Jinling commercial printers, and it must surely be the very first calligraphic manual to reach Europe.

1. Summary catalogue of Western manuscripts in the Bodleian Library at Oxford, vol.2 pt.1 (Oxford, 1922), p.558.

Model calligraphy

9 March 2019

I have almost finished cataloguing the pre-modern Chinese collections in the Bodleian Library, but have left two things until last because they are difficult. These are so-called “ink-squeezes” (tapian 拓片) and the model calligraphic albums known as fatie 法帖. They are related in how they are produced, but are different in nature and function.

Ink squeezes (sometimes inaccurately termed “rubbings”) are used to reproduce inscriptions on rock faces and stelae, in tombs, or even on bronze vessels and other artefacts. They are “one-off” products, usually presented as single sheets, often extemely large. They are classified as epigraphy.

Fatie on the other hand are collections of model calligraphy used for self-instruction or teaching. These are published as printed books, but are produced in the same way as ink squeezes. That is, the calligraphy is engraved on stone or wooden blocks in intaglio, not relief, so that the end product is white on black, not black on white. They are usually presented in so-called “accordion” bindings (zhezhuang 折裝), and are classified as text-books.

The two types of material are sometimes confused. Ink squeezes are occasionally cut up and bound accordion-style to make them easier to store and consult, so that they look like fatie. And sometimes, epigraphical material is re-engraved on blocks and published as fatie.

Although there were antecedents, none of which is extant, the archetypal fatie is the famous Chunhuage tie 淳化閣帖 which was commissioned by the Song emperor Taizong 宋太宗 in 992, the third year of his reign period (淳化三年), from which it takes its name. He ordered the Hanlin academician Wang Zhu 王著 to make a compilation of the calligraphy which was preserved in the imperial collection and engrave it on blocks. Copies of the resulting publication were then given to the imperial princes and senior officials.

The work consists of ten juan 卷. The first five contain the calligraphy of emperors, officials, and others. The last five contain the work of Wang Xizhi 王羲之 (303–361) and his son Wang Xianzhi 王獻之 (344–386); these two figures from the Eastern Jin dynasty 東晉 (317-420) are generally regarded as the founding fathers of calligraphy as an art form. The title of each juan follows the formula「…法帖第…」, and it is from this that the term fatie 法帖 to describe this genre is believed to be derived. At the end of each juan, in seal script, is a reminder of when and under what circumstances the work was produced:「淳化三年壬辰歲十一月六日奉聖旨摹勒上石」. Although this colophon ends with the words 「上石」”engraved on stone”, there is reason to suppose that the calligraphy was actually engraved on wood. [1]

The Chunhuage tie was recut many times, especially during the Song and the Ming dynasties, and it is very difficult to identify the various editions, as a mere glance at any catalogue whether published or online will confirm.

There are three copies of the work in the Bodleian Library; two of them are incomplete copies of the same edition, which I catalogue as follows:

淳化閣帖 殘七卷
折裝7冊 ; 32公分
全書十卷, 殘卷四~十
Sinica 2758

法帖第六. 王羲之書一
法帖第七. 王羲之書二
法帖第八. 王羲之書三
法帖第九. 晉王獻之一
法帖第十. 晉王獻之二

淳化閣帖 殘五卷
折裝5冊 ; 32公分
全書十卷, 殘卷一~五
Backhouse 604


The only difference between the two copies (apart from the sections preserved) is that Sinica 2758 was printed earlier than Backhouse 604, as indicated in my statement of the imprint.

The history of the edition is almost common knowledge, and can be found in many Chinese sources both printed and online. I summarise it as follows.

In 1392, the Hongwu 洪武 emperor Zhu Yuanzhang 朱元璋, founder of the Ming dynasty, made his fourteenth son Zhu Yang 朱楧 king of Su 肅, a fiefdom centred on Lanzhou 蘭州 in Gansu Province 甘肅省. To mark the occasion, he presented him with a copy of a Song edition of the Chunhuage tie for him and his heirs to treasure. In 1615, his successor Zhu Shenyao 朱紳堯, king Xian 憲, ordered Wen Ruyu 溫如玉 and Zhang Ying 張應 to make a copy of the edition, as recorded in the colophon:


The work was completed in 1621 天啟元年. Over 140 stone blocks were used to make the edition, and most of them were engraved on both sides, so that there are over 250 pages. Already by the early Qing some of the blocks had got damaged, so some 40 were re-cut in 1654, as recorded in another colophon:


The blocks are extant. In 1910 the Hanlin academician Liu Erxin 劉爾忻 moved them into the Zunjingge 尊經閣, a pavilion of Lanzhou Confucian Temple 蘭州文廟 (now Lanzhou No.2 Middle School 兰州市第二中学), and in 1966 they were moved into Gansu Provincial Musuem 甘肃省博物馆, which is where they are to this day.

Some images of these blocks can be found on the web, although I’ve had difficulty in finding a good one. Here is an image of the block for the opening of juan 2, together with one of the impression in Backhouse 604 that has been taken from it:



Xu Guoping 许国平 has made a detailed study [2] of datable impressions of the Su edition of Chunhuage tie in the Palace Museum collections, in which he carefully notes how the stone blocks have deteriorated over the course of time. It is from his notes that I have, tentatively at least, been able to date the Bodleian impressions to the Kangxi 康熙 period or possibly a little earlier (Sinica 2758), and the Qianlong 乾隆 period or possibly later (Backhouse 604).

For example, Xu notes that in the following leaf from juan 4, the block was intact in all impressions up to and including the Kangxi period, but broken in the Qianlong period:


Sinica 2758


Backhouse 604

The break is vertical, for as we would expect, a piece of stone is more likely to fracture at its smaller dimension. With wood blocks the reverse is the case, as wood splits along its grain, not across it, and in printing blocks the grain is invariably horizontal.

The third copy in the Bodleian Library is of what I suppose is a later edition, but it could well be earlier. It is one of many fatie and ink squeezes among the books that were given to the Library by Dr William Lockhart, a medical missionary, in 1879. It is larger in size than the Backhouse copy, and of very much finer quality. Its heavy wooden boards are covered in brocade, which may help in dating the copy, but unfortunately I have no expertise in this area. I describe it as follows:

淳化閣帖 十卷
折裝10冊 ; 39公分
Sinica 465

歷代帝王法帖第一 (有缺, 殘末六葉)
法帖第六. 晉王羲之書
法帖第七. 晉王羲之書
法帖第八. 晉王羲之書
法帖第九. 晉王獻之書
法帖第十. 晉王獻之書




1. Rong Geng 容庚 (叢帖目, 14) quotes Zhang Boying 張伯英 who refers to a passage in Ouyang Xiu’s 歐陽修 Jigulu 集古錄 which says that copies became scarce as a result of a fire which destroyed the blocks.
2. 许国平: 肃府本《淳化阁帖》版本考略. In 中国书法学术 271(2015:11), 177-181.