DOWNLOAD THIS PAPER (PDF 88KB)
The manuscripts described in this Catalogue once formed part of a huge collection which was discovered about fifty years ago in a walled-up chamber adjoining one of the 'Caves of the Thousand Buddhas' (Qianfodong/Ch'ien Fo Tung) a few miles south-east of the Dunhuang oasis on the border of Gansu. Hollowed out in irregular tiers along the face of a steep cliff, these cave-temples were known in the Tang dynasty as Mogaoku or 'Grottos of Surpassing Height'. They are said to have owed their origin to a saintly monk, one Lezun, who began the work of excavation in AD 366. An account of him is given in an inscription of 698, and his name occurs again in a topographical fragment some 200 years later (see no. 7180V of this collection [Or.8210/S.788]) The manuscripts, mostly in the shape of long paper rolls, together with a large number of religious paintings, would appear to have been removed from different monasteries in the neighbourhood and hurriedly stowed here for safety on approach of an invading tribe, probably the Xixia early in the eleventh century. And here they lay, apparently undisturbed and forgotten, for a space of nearly 900 years. The news of the discovery soon spread abroad, and the late Sir Aurel Stein was the first foreigner to visit the spot with the definite object of acquiring by purchase some of the available treasures. And although he lacked the necessary knowledge of Chinese to make a wholly satisfactory selection, as is shown by the inclusion of far too many duplicate texts, the mass of Buddhist and other literature which he eventually sent home proved to be of outstanding value and importance. These were the fruits of his second great expedition (1906-08), jointly financed by the Government of India and the Trustees of the British Museum. By January, 1909, the whole collection had been safely deposited in the British Museum, to which all the Chinese MSS. were ultimately allotted. For many years no serious effort was made to deal systematically with the collection as a whole. It was not until 1919, just after the first world war, that it was placed under my charge, and I was able to embark on the formidable task of sorting it out and compiling a full descriptive catalogue. Meanwhile, Stein's third expedition had produced yet another large batch of manuscripts which came to swell the total in London, but again little or no attempt seems to have been made at judicious selection. Not that this now mattered very much, for only a year after Stein's first visit to the Caves the brilliant French sinologist, Professor Paul Pelliot, had spent three weeks on the site and carried away almost everything of first rate importance to the Bibliothèque Nationale in Paris. Shortly afterwards the Chinese Government, realizing at last the full value of the hoard, removed most of what was left to the security of the National Library of Peking.
A preliminary examination of what had reached the British Museum shows that the Buddhist works amounted to more than three-quarters of the whole, and included all the principal sutras, Vinaya and Abhidharma texts. It seemed best, therefore, to adopt the arrangement followed by Bunyiu Nanjio in his standard Catalogue of the Chinese translation of the Buddhist Tripitaka, supplementing it with the many uncanonical works and miscellaneous pieces which had been acquired with the rest. There were also a number of Taoist works and two very interesting Manichean texts. The rest of the Collection, which was non-religious, could then be reserved for separate treatment under the comprehensive heading of 'Secular Texts'.
The information given for each item in the Catalogue begins with a serial number, and contains the Chinese title, if any, with its transcription according to the Wade system [ed. note: these have been converted to pinyin system]; the juan or pin, or chapter and section, and any short description of the contents that may be thought necessary; a rough indication of the quality of the handwriting, grading from 'fine' to 'very bad'; the colour and texture of the paper; a measurement of length (in feet only, for ordinary rolls [ed.: the length is in the physical description and so is omitted from this electronic catalogue]); and finally, the 'Stein number' given to each MS. When first catalogued. Knowing the Stein number of a MS., one can also obtain its serial number by means of the Conversion Table at the end of this volume [ed.: not necessary in electronic version]. For all canonical Buddhist texts references are given to Nanjio's Catlogue (N.), to 大日校訂大藏經, known as the 'Kyoto reprint' of the Tripitaka (K.) and to 大正新修大藏經 Taisho Shinshu Daizokyo (T.). All the texts selected by Dr. Yabuki for this 鳴沙餘韻 Meisha Yoin are indicated by a Y. The width of the rolls does not vary enough to be worth recording except when it is less than 25 cm. or more than 30 cm. In the case of small fragments and booklets, both width and length are given in centimetres, the vertical measurement coming first. A dash at the beginning of a fresh entry indicated that the title is the same as the one preceding [ed.: the title in full is given in the electronic version]. This only applies of course to the first title recto, if there are more than one. An asterisk placed before any transliterated title, or description, or dash, indicates that the beginning of the text in question is wanting; coming after them, either that the end is lost of that the manuscript has been left incomplete. Incidentally, it may be noted that a large majority of the manuscript scrolls are imperfect at the beginning, where there has been constant wear and tear. Badly mutilated rolls have been repaired and mounted, either wholly or in part, in the binders' shop on the Museum premises, and special care has been taken to preserve any writing on the back. Many of the rolls and booklets, especially in the non-Buddhistic sections, will be found to contain a series of different texts, of which only the first may belong to the section under which its stands in the Catalogue. In all such cases, the texts are separately numbered, and cross-references are provided from all the sections to which they severally belong. These cross references have also been given serial numbers.
A prominent feature in the Buddhist section is the tail-piece of 'colophon' (as it is here somewhat loosely termed) which is sometimes appended to the copy of a sutra or some other holy text. Its main purpose is to make known the person who has acquired 'merit' by having the copy made at his own expense, and the beneficiary (usually deceased) in whose direction he wishes the merit to flow. An exact date is generally included. Many of these colophons run to a few words only, while others are long and elaborate compositions with a pronounced literary flavour about them. Most of them are translated here in full, though in a few cases, owing to the difficulties of decipherment or interpretation, only the main facts have been extracted. A list of the colophons, excluding very short notes, is given on page xv [ed: not included here: Advanced Search, Subject/Keyword = 'Colophon'). Twenty four of them, marked t, are drawn up in a carefully tabulated form, where the names of the copyist, the revisers, the dyer of the paper, and various other persons concerned in the production are all duly noted. The number of sheets of paper used is also stated, which enables us to tell how much of the imperfect text is missing. The earliest of the tabulated colophons (no. 2580 [Or.8210/S.5319]) is dated 671, and exceptionally good specimens will be found under nos. 2298 [Or.8210/S.3361], 2454 [Or.8210/S.3094] and 3011 [Or.8210/S.3348]. It is due mainly to these tail-pieces that the date of copying has been preserved for us in so many cases — about 380 altogether. A detailed account of all the dates MSS. in chronological order, from A.D. 406 to 995, is given in a series of six articles in the Bulletin of the School of Oriental Studies [ed.: these are being added online as a catalogue record, linked to the relevant manuscripts]. Of the undated MSS. it is quite possible that some may be a little earlier than 406, and it is still more likely that several are a few years later than 995.
Another occasional appendage to Buddhist sutras, occurring for some reason with special frequency in copies of the Jinguangmingzuishengwangjing (N.126) is what I have ventured to call a phonetic glossary. This consists of just a few words selected from the preceding text, with the fanjie (initial plus final) pronunciation. See, for example, nos. 1932 [Or.8210/S.1177] and 2051 [Or.8210/S.2227] .
The fact that paper was still unknown in Europe during the whole period covered by the Dunhuang MSS. has made a regular notice of it desirable in most of the following entries. For a general survey of the subject of early paper-making, as well as an expert examination of certain papers of different dates in the Stein collection, those interested should consult R. H. Clapperton's Paper: An Historical Account pp. 1-26 [Clapperton_1934]. A list of the papers thus analysed, to which both serial and Stein numbers have been added, appears below on p. xvi [IDP Note: details of the results of the analysis along with images of the paper fibres are being added to the IDP database and being made available online]. Broadly it may be said that even the earliest fifth-century papers known to us are of a remarkably good quality; they are generally a dull or brownish buff in hue, for the application of colouring matter does not appear to have been practised much before AD 500 (See no. 5524 [Or.8210/S.2106]). An agreeable lemon-yellow is characteristic of the earlier part of the sixth century: see 1646 [Or.8210/S.6476], 1690 [Or.8210/S.2876], 1714 [Or.8210/S.6510], 1723 [Or.8210/S.3316] and 1735 [Or.8210/S.4366] among others. Then, a little later on, a beautiful golden-yellow paper, thin and crisp, makes its first appearance. More uncommon shades of colour are orange-yellow, slate-blue, greenish-buff, pink or pinkish, sulphur-yellow, etc.; see especially 2530 [Or.8210/S.3796], 4722 [Or.8210/S.6486], 5784 [Or.8210/S.6020], 5844 [Or.8210/S.1568], 6053 [Or.8210/S.6492], 7076 [Or.8210/S.575], 7456 [Or.8210/S.3753]. And one roll (no. 1262 [Or.8210/S.4324]) is actually composed of twelve variously coloured sheets. The staining fluid was evidently of an oily nature; sometimes it has been applied on both sides of the paper, but usually on one only. During the seventh and part of the eighth, the texture of the paper used in the monasteries continues to be fine and smooth, and its colour a bright or brownish-yellow; but after the An Lushan rebellion a marked deterioration sets in, and most of it now becomes coarse and drab-coloured.
The handwriting of the copyists also passes through a series of changes which are similarly useful in furnishing a rough clue to the date. Throughout the fifth century and perhaps even a little later, a stubby kind of brush seems to have been in common use which was incapable of producing the fine, delicate strokes that are characteristic of Chinese calligraphy at its best. It was during the Sui dynasty that the art of handwriting appears to have reached its zenith — at any rate, as far as the present Collection is concerned. For a series of particularly fine manuscripts see those of the Avatamsaka-sutra, nos. 1612-1647. Only the most carefully trained scribes were entrusted with the task of copying sutras (supposed to be the pronouncements of the Buddha himself), as opposed to Vinaya and Adhidharma texts, commentaries and the like; and that is probably the main reason why we find the production of sutras suddenly diminished almost to vanishing point after the disastrous upheaval mentioned above, which must have affected every monastery in the country. Handwriting, then, is another of the features that receive particular attention in the pages of this Catalogue (the term 'MS.' being used, for the sake of brevity, in its stead).
A new departure of another kind also begins to be noticeable during the tenth century at Dunhuang. Ever since the invention of paper by Cai Lun eight hundred years before, books had been written and circulated in the form of long paper rolls; hence the use, down to the present day of 卷 juan, the word for 'roll', to designate the section or chapter. By far the greater part of the Stein Collection consists of such rolls; they are made up of a number of sheets, each about 1.5 feet in length, very neatly and efficiently fastened together with glue. The earliest dated roll, a Vinaya text of AD 406 (no. 4523 [Or.8210/S.797]), is 23 feet long, which is slightly less than the average length of ordinary sutra scrolls. The longest of all the rolls are to be found among the commentaries, of which nos. 5587 [Or.8210/S.4052] (99 feet), 5597 [Or.8210/S.6789] (90 feet), 5287 [Or.8210/S.4495] (86 feet) , and 5523 [Or.8210/S.6713] (80.75 feet) head the list in this respect. The handling of such rolls, or even those of more moderate length, cannot but be awkward for the reader, who has to be constantly unrolling and rolling up again as he goes along, and any reference to a required passage may involve serious loss of time. It is surprising, therefore, that the invention of some more convenient device for the construction of books should have been so long delayed. The first step in advance was taken when, instead of being rolled up, the paper was folded into leaves of a reasonable size thus forming a volume that could be quickly opened at any point and shut up after consultation. A fairly good specimen, containing as many as 211 leaves, is no. 5591 [Or.8210/S.5603] in the Catalogue. The next innovation was the stitching of all the leaves together at one size, so that they should no longer fly apart in a long chain. This is the style in which Chinese books are still produced. It has the disadvantage, however, of leaving one side of the paper unused; and as a scarcity of paper was becoming a problem in the Dunhuang region, booklets were generally made up of small separate sheets, intended for writing both sides, as with us in the West. Of these there are several hundred specimens, varying greatly in size, in the Stein Collection; but it is fairly clear that even by the end of the tenth century they were far from having superseded the roll, for of the last fifty dated manuscripts only two are in booklet form.
About 70 of our MSS. may be confidently assigned to the fifth century, and for the next 200 years or so the output continues steadily to increase. The great majority of these early MSS. are copies of Buddhist sutras that happened to be in particular flavour at the time. During the following period, when the Dunhuang district passed temporarily under Tibetan rule, the flow of Buddhist literature rapidly dried up, and never returned to anything like its earlier level. On the other hand, after the re-establishment of Chinese sovereignity in 848 by the famous leader Zhang Yichao (for whose autograph as a youthful copyist, see no. 5656 [Or.8210/S.5835]), miscellaneous texts of a 'secular' kind begin to appear in ever-increasing numbers: parts of the Confucian classics, literary, topographical, and didactic fragments, verses, legal documents, club circulars, account notes, and so on. Divination, medicine, and calendar-making have always been regarded as specifically Daoist activities by the Chinese, and are therefore included in the Daoist section of the Catalogue. A number of manuscripts in which other languages appear as well as Chinese are listed in nos. 8049–8092, besides which there are some 50 or more elsewhere, which contain miscellaneous scraps in Sanskrit, Sogdian, Uighur, Khotanese, and especially Tibetan.
The Buddhist paintings brought back from the Qianfodong by Sir Aurel Stein have been catalogued in a separate volume by Dr Arthur Waley (see Waley_xx), and apart from these there is very little of pictorial nature that demands special attention. However, it has been considered worth while to draw up a list of cross references to any drawings or diagrams, however crude, that may occur here and there [Advanced Search, Subject/keyword=Illustrated].
A small but very precious part of the Collection consists of 20 early specimens of block-printing, one of which, a complete and remarkably well-preserved copy of the Diamond Sutra, bears a date corresponding to the 11th May, AD 868. Among the other printed documents are two calendars of AD 877 and 882, several prayer sheets from the middle of the tenth century, and a very well-printed set of Buddhist verses on the 24 examples of filial piety. In the year 689, during the reign of the usurping Empress Wu, new forms are know to have been officially substituted by her command for a certain number of Chinese characters, variously given as 12, 16 or 19; and that their use was rigidly enforced until her abdication in 705 is proved during their regular appearance in the manuscripts copied during this period. I have found examples occurring in at least 47 different rolls, and there may well be others that have escaped my notice. [Advanced Search, Subject/Keyword=Empress Wu characters] There are certain markings of an enigmatical kind, such as the 'shooting stars' which appear in 7380 (1) [Or.8210/S.1604] , 7381 [Or.8210/S.5747] , and 7408 [Or.8210/S.5810] , and a curious 'bird-character' (the figure of a bird perched on radical no. 162), which may possibly represent the signature of the Commissioner or some other important official: see 5231V(1) [Or.8210/S.2069] , 7399 [Or.8210/S.4453] , and compare also 7444 [Or.8210/S.376] , 7511 [Or.8210/S.3728].