When every Chinese book has been catalogued digitally, and according to consistent standards, it will be much easier to take a global view of Chinese literature in the imperial period. At present, we must rely on what is printed, and in my view the most useful tools by far (and they appeared at much the same time) are Shanghai Library’s congshu index (中國叢書綜錄), which was first published in 1959, and the two printed catalogues of the Chinese collection in Kyoto University’s Institute for Research in Humanities.
The first edition of the Kyoto catalogue (京都大學人文科學研究所漢籍分類目録) was published in 1963, and within it, the contents of congshu are classified; it thus corresponds with the second volume of the Shanghai work. But in the second edition (京都大學人文科學研究所漢籍目録), which appeared in 1979, the congshu contents are not classified; it thus corresponds with the first volume of the Shanghai work.
These two monumental bibliographies give us what the computer screen never will: conspectus, and a three-dimensional view of their contents. Simply by counting the pages devoted to each subject in the classified volumes we can discover how much was written, and about what, and say with some certainty what interested traditional Chinese scholars.
I have just looked at the “classics” division of each of them, and confirmed what has long been my impression, that the greatest number of individual works is to be found in the yijing 易經 section. This is not surprising in view of the primacy of that text, which has always come first in the traditional Chinese classification systems from the time that they were first devised, over two thousand years ago.
A good example of the voluminous traditional scholarship in this area is an edition in the Backhouse collection which I first encountered in the late 1970s, and which has long fascinated me. It is a collectaneous work by the Ming dynasty scholar Yang Shiqiao (明)楊時喬, a jinshi 進士 of 1565 (嘉靖44), who as part of the conservative reaction to the ideas of Wang Yangming 王陽明 (1472–1529) sought to reaffirm the orthodox interpretation of the Yijing by the brothers Cheng Hao 程颢 (1032–1085) and Cheng Yi (1033–1107), and Zhu Xi 朱熹 (1130-1200).
I have catalogued the edition as follows:
周易全書 / (明)楊時喬撰
線裝32冊 ; 28公分
周易古今全書論例 : 二卷
周易全書古文 : 二卷
周易全書易學啟蒙 : 五卷
At much the same time as I was examining this edition, another example came to my notice, for sale in the catalogue of Tsi Ku Chai in Hong Kong. The contents of the Tsi Ku Chai copy were tantalisingly different, and as I had already decided to develop our pre-modern collections so as to demonstrate the nature and variety of Chinese book production during the last two dynasties (building on very considerable strengths), I bought it.
The book reached us in spring 1979, and cost HK$2,000, then worth £195. It was in rather bad condition, and had to be repaired. This was done by Judy Segal, extraordinarily well in view of the fact that the only English guide to the structure of Chinese books available to her was the famous photographer Hedda Morrison’s article Making books in China (Canadian geographic journal 39, 1949, 234-45) – Xiao Zhentang’s 肖振棠 manual on Chinese book restoration 中國古籍裝訂修補技術, which I later translated for Robert Minte, hadn’t even appeared in Chinese at this time (it was published in Peking in 1980). I now catalogue this copy as follows:
周易全書 / (明)楊時喬撰
線裝20冊 ; 27公分
周易古今文全書論例 : 二卷
周易全書古文 : 殘一卷. – 全二卷, 殘卷一
周易古今全書傳易考 : 二卷
周易全書龜卜考 : 一卷
The edition was published over a period of nine years in the 1590s, and not many complete copies survive. The reproduction in Siku quanshu cunmu congshu (四庫全書存目叢書, 經部8-9) has been pieced together from partial copies in Peking University Library, the Chinese Academy of Social Sciences Library, and the National Library of China. There is a complete copy in the Harvard-Yenching Library which has been described by Shen Jin 沈津 (美國哈佛大學哈佛燕京圖書館中文善本書志, 1999, 14-15).
The cunmu congshu copy, the Harvard copy, and the Sinica copy all have a general preface 周易古今文全書總序 dated 1590 (萬曆18). The total number of works that were ever included is as follows:
周易全書古文2卷 — 萬曆20(1592)序
周易全書今文9卷 — 萬曆24(1596)序
周易全書易學啟蒙5卷 — 萬曆20(1592)序
周易古今全書傅易考2卷 — 萬曆23(1595)序
周易全書龜卜考1卷 — 萬曆27(1599)序
The cunmu congshu copy also has a general table of contents, which must be contemporaneous with the publication of the final work in 1599:
I will try to make some sense of this jumble of evidence.
Everyone who has catalogued this book to date has treated it as a single work, with one title, so that the complete work (and of these there are very few extant examples) is treated something like this:
But for a number of reasons, I regard the work not as a single entity, but as a congshu. The titles within it were printed and circulated at different times, and indeed are often found individually in library collections.
In his general preface, Yang says that he started to work on the contents in 1570 (隆慶庚午), and that they were to be known collectively as Zhouyi quanshu 周易全書. This is the collective title used in the general table of contents (illustrated above), and it also appears in the banxin 版心 of every single leaf. And most interestingly, the Backhouse copy, which I will presently describe in detail, shows that even if the work is to be taken as a single entity, the standard title (by convention taken from the first juan 卷) is in any case doctored.
I also have a small quibble with Shen Jin’s description of the edition as having been made by Wang Qiyu 王其玉, one of Yang Shiqiao’s followers (門人). He does this on the basis of evidence found only in two later and relatively small sections of the work (傅易考2卷, 1595; 龜卜考1卷, 1599) – there is no evidence that Wang was responsible for the entire thing, although it is quite possible that he might have been.
The Backhouse impression is very fine. As indicated above, the three sections present (論例二卷﹑古文二卷﹑易學啟蒙五卷) were all cut in 1592 – the rest followed three or more years later. In fact, the quality of the impression is so fine that we can reasonably infer that it was made soon after the blocks were cut, and that at the time of printing, the edition was complete (in a catalogue it would be incorrect to describe such a copy as incomplete simply because more blocks were cut subsequently).
But one thing in particular confirms that the Backhouse copy is probably the earliest surviving impression of this edition. This is the title of the first work to be cut, which is 周易古今全書論例. In all the other copies, the title is 周易古今文全書論例 (that is, with the addition of the character wen 文).
When we compare the Backhouse copy with the Sinica copy, it is clear that they are both from the same blocks, but that the characters 全書論例 in the former have been excised and replaced with 文全書論例 in the latter, as well as all other surviving copies. But as ever, when doctoring blocks, they never do it properly, so that in all the later impressions, the old title survives without the wen 文 at the end of juan 2.
We can only guess why the character wen 文 was inserted. At first I thought it might have been done after the cutting of the jinwen 今文 section in 1596, to make it quite clear that the work is concerned specifically with the “old text” and “new text” versions of the Yijing, and not with the general history of the Yijing throughout the ages. However, the general preface also bears the longer title, and it is dated earlier than any of the other works in the set, perhaps being contemporary with the lunli 論例 section, which is the only work to have no preface of its own.
The blocks of the Backhouse edition may well be an undoctored and unique survival, but the same cannot be said of the copy. Like so many books in Chinese collections, it has passed through the hands of a dealer who has tried to pass it off as something else. When I was first teaching myself about the tricks of this particular trade, I was puzzled by the preface, which is the Tang dynasty scholar Li Dingzuo’s 李鼎祚 preface to his commentary on the Yijing entitled Zhouyi jijie 周易集解. The preface has four leaves, but in the banxin of each are found the title of Yang’s work 周易全書 (I didn’t know that Yang had written his own prefaces to each of the sections until I bought the Tsi Ku Chai copy; they are in a more rounded calligraphic style than the text).
The large-scale reproduction of historic editions in recent decades (not to mention my acquiring and cataloguing them for the Bodleian) has now enabled me to establish that this preface in the Backhouse copy is taken from a copy of the following edition:
周易集解 : 十七卷 / (唐)李鼎祚撰
It is from the same blocks (but later, and possibly themselves doctored) as the copy in what is now the National Library of China, which is the very first work to be reproduced in the series 北京圖書館古籍珍本叢刊 (Peking, 1988).
Taking a closer look at Li’s preface in Backhouse 276, we find that the banxin of all four pages has been carefully doctored. The words 「聚樂堂」 in the upper banxin 版心 have been excised from the leaves, and replaced with 「周易全書」, undoubtedly taken from leaves removed by the dealer (along with Yang’s dated prefaces) in an attempt to pass the copy off as something earlier.
The detailed ascriptions at the beginning of the guwen 古文 section, which clearly identify Yang Shiqiao and other Ming dynasty personages involved in producing the book, have also been excised from the leaf in the Backhouse copy and replaced with something else (I don’t know where from) which might suggest an earlier date; they are present in the undoctored (but later) Sinica copy.
Gilding the lily, the dealer has also applied the seal 「乾隆御覽之寶」 (“imperially perused by Qianlong”) to the first leaf of each of the three sections, so that the first impression appears on the unrelated preface by Li Dingzuo that had been supplied to pass the book off as something earlier. It is inconceivable that the emperor would have examined such a copy, and the seal must clearly be considered a fake.