[page 39] If we are to believe later traditions, and there is in my opinion no reason not to do so, the first Tibetan historiographic writings date from Tibet's imperial period (seventh-ninth centuries), which coincided with her relations with the Nepalese, Indians, Arabs, Turks, Uighurs, 'A zha and, above all, Tang China. Only a fragment of this literary corpus, falling into two broad classes, has survived. The first of these constitutes those historical documents that were discovered as late as the beginning of this century in one of the caves of the famous cave-temple complex near the town of Dunhuang in Gansu Province in the People's Republic of China. Recent scholarship generally agrees that the cave housing these manuscripts was sealed sometime after the year 1002, the latest date found in the manuscripts, possibly around the year 1035 (Fujieda: 65), so that the terminus ad quem of these undated documents would fall in that year. Of signal importance are especially three untitled manuscripts that are known to English-language scholarship as:

1. (1-2) Royal Annals of Tibet (Bibliothèque Nationale, Paris, Pelliot tibétain no.1288, together with India Office Library, London, Stein no.8212, 187).
2. (3) Old Tibetan Chronicle (Bibliothèque Nationale, Paris, Pelliot tibétain no.1287)

[page 40] They have been studied in varying degrees of detail by a number of Western, Tibetan, Chinese and Japanese scholars.1 The first Tibetan to examine these was the great scholar and iconoclast dGe 'dun chos 'phel (1903-1951),2 who had gained access to these and a few other fragments while in Kalimpong sometime in 1939. As is related by H. Stoddard, his most recent biographer, the French Tibetanist Jacques Bacot visited Tharchin, a Christian missionary of Khunu descent, in Kalimpong and read with him several of these difficult manuscripts in Old Tibetan. Tharchin apparently solicited the help of dGe 'dun chos 'phel, who was able to aid him in deciphering a number of problematic readings. The results of Bacot's studies were published in 1946, but no mention is made there of either Tharchin or dGe 'dun chos 'phel, although he gratefully recorded his philological debt to another Tibetan, namely bKa' chen Don grub.3 The last tome of a recently published three-volume edition dGe 'dun chos 'phel's works contains inter alia three studies of a number of these Tibetan Dunhuang manuscripts. They include a reproduction of the Royal Annals with philological notes, an adaptation into Classical Tibetan of the Old Tibetan of the manuscripts of a large portion of a version of the celestial origin of the imperial families and other miscellaneous fragments, and a reproduction of the Old Tibetan Chronicle.4 Some of the results of these initial studies were subsequently incorporated into his incomplete work on Tibetan history, the Deb ther dkar po ("White Annals"). He was followed by such recent scholars as Khetsun Sangpo, Khang dkar sKal bzang tshul khrims, rDo rje rgyal po, and Chab spel Tshe brtan phun tshogs.

While most of Tibet's cultural institutions and literary canon derive from India or are based on one or other of her models, a notable exception is the intense preoccupation of Tibet's men of letters with history and historiography. In terms of literary genre, some of Tibet's historiographical writings bear a resemblance to, or are analogous with, the Indian vaṃśāvalī ("annals"), but her enormous historiographic literature, including that of biography and autobiography, bears testimony to an approach to history that is different from the Indian one(s) (see Warder, Subhrahmanian). As far as the secondary sources on this large corpus of literature are concerned, the premier study is still the one by A. I. Vostrikov.5 Now dated in a number of respects, it remains a classic and indispensable treatment of the various literary genres.

[page 41] Despite the fact that the dissolution of the Tibetan empire seems to have resulted in a virtual cessation of further literary developments for about a century, if we take the Tibetan Buddhist tradition at face value, there is ample evidence for affirming the existence in at least central and eastern Tibet of an unbroken transmission of historiographic texts, or quasi-historiographic documents like family chronicles, throughout this time and into the period of the so-called subsequent propagation, which the Tibetan Buddhist historians generally date to the middle of the tenth century. Indeed, we possess documents that trace the genealogies for such extended families or clans of the 'Khon and rLangs of, respectively, the Sa skya and gDan sa mthil/rTse[s/d] thang monastic principalities.6 Moreover, some sort of archives may also have been maintained, if only by the scattered descendants of the imperial family. A sample of the kinds of documents that may now lie buried somewhere in the vast collections of the Potala would be a series of "edicts" issued by Khri srong lde btsan (r. 742?-797?), which were preserved in the chronicle by the great sixteenth-century historian dPa' bo gTsug lag phreng ba (1504-1566).7 By the same token, the two recensions that are now available of the sBa bzhed, a virtual biography of the first Tibetan monk, sBa Ye shes dbang po (eighth century), suggest that the original text should by and large be considered a primary source on Khri srong lde btsan and his religious works, in spite of the fact that their transmission is beset with enormous complexity. In his chronicle of Buddhism in Tibet (and much else besides), Nyang ral Nyi ma 'od zer (1124-1192) refers to a number of very early works, in addition to numerous edicts, that have to do with the reign of the latter as well. Their descriptive titles are:8

1. (1) bKa'i yig rtsis che
2. (2) bKa'i yig rtsis chung
3. (3) bKa'i thang yig che
4. (4) bKa'i thang yig chung
5. (5) rGyal rabs rkyang pa
6. (6) Khug pa
7. (7) Zings po can
8. (8) sPun po

NYANGb wrongly collapses the titles of nos. 6 and 7, and reads Khug po zings pa [sic!] can. NYANGl has Yun po for no. 8, which is due to a misreading of the cursive ligature sp, which resembles[page 42] the graph for y. Moreover, the last four would appear to be historiographic texts per se, but none of these have been located so far if, indeed, they are still extant. One recension of the sBa bzhed, as do Nyang ral and, more elaborately, the chronicles of Buddhism by *lDe'u Jo sras and mKhas pa lDe'u,9 brings to attention the existence of five early historiographic texts from the imperial period, two of which appear to correspond to nos. 7 and 8 of the above titles. These have been briefly noted in a recent paper by S. G. Karmay.10

There are roughly three expressions which, when they occur in book titles, usually indicate that the books in question are historiographic in nature, and all of these are found in writings attested in Tibet for the period covering the eleventh to twelfth centuries, and one which in part may even go back as far as the seventh century. With their probable dates of inception, these are:

1. (1) Lo rgyus ("Records") (eleventh century)
2. (2) rGyal rabs ("Royal Chronology") (eleventh century)
3. (3) Chos 'byung ("Religious Chronicle") (twelfth century)

Due to limitations of space, we shall have to restrict ourselves, with one notable and fairly lengthy exception, to a bibliographic survey of historiographical texts belonging to these two centuries. However, it must be understood at the outset that those philological procedures that are fundamental to other branches of the humanities having to do with texts and their transmission have thus far mostly bypassed inquiries into Tibetan historiography, as they have virtually every other branch of Tibetan studies. Moreover, there are also considerable gaps in the literary corpus of available texts on the present subject. For these reasons, and also in the absence of "critical" texts, some of the remarks that follow are of necessity rather tentative.

The first instance of this expression in a historiographic context appears to be the famous but until now inaccessible Lo rgyus chen mo ("Grand Annals") by Khu ston brTson 'grus g.yung drung (1011-1075).11 The expression lo rgyus, literally "tidings of year[s]," is only very occasionally best rendered by "annals." It is far more often the case that works with this term in their title do not fulfill what[page 43] is promised by such a rendition, that is to say, they do not at all give a year-by-year account of their subject-matter, but rather present a narrative of events, historical, quasi-historical, or even ahistorical, in rough chronological sequence. It is well known that later historiographic sources abound in quotations from what appears to be Khu ston's work, although it does not seem to be extant.12 The fragments indicate that it was largely, if not entirely, written in verse. dPa' bo also often availed himself of this work in his study of Tibet's imperial period, and it functioned, for example, as one of his fundamental sources for information about the decades after Emperor Glang dar ma's assassination in 842 (or 846, the year which he assigns to this event), specifically about the insurrection of 869 against his two sons, 'Od srung and Yum brtan, which spread from central to eastern and northeastern Tibet (see DPA'1: 429-430; DPA': 432-433).

The expression rgyal rabs means something like "account/story of king(s)," and is perhaps best translated by "royal chronicle."13 As far as the rgyal rabs as a specific historiographic genre is concerned, the earliest ones that are presently available were composed by the third and fifth Sa skya pa patriarchs rJe btsun Grags pa rgyal mtshan (1147-1216) and 'Phags pa Blo gros rgyal mtshan (1235-1280).14 The latter's is dated to the year 1275. In addition to these two, there were also others that were written in the thirteenth century. Possibly dPa' bo but certainly the great Sa skya scholar Mang thos Klu sgrub rgya mtsho (1523-1596), in his study of the chronology of Buddhism in India and Tibet finally completed in 1587, have preserved several fragments of the rGyal po rabs phreng by U rgyan pa Rin chen dpal (1230-1309).15 According to bSod nams 'od zer's hagiography of U rgyan pa, the author wrote this work for Qubilai Khan (r. 1260-1294) as part of his attempt—his hagiographer and disciple states that he was successful—to dissuade the Mongol emperor from invading Nepal. While bSod nams 'od zer does not date this episode, evidence internal to the hagiography suggests that this may have taken place in the 1270s. This is now confirmed by the recent discovery of a thirteen-folio manuscript of U rgyan pa's rGyal po rabs kyi phreng ba, which is dated 1278.16 The still unavailable rgyal rabs is the rGyal rabs dpag bsam ljon shing[page 44] of 1286 by the elusive Byang ji ston pa Shes rab 'bum, which so far is first alluded to in Tshal pa Kun dga' rdo rje's extremely influential Deb gter/ther dmar po ("Red Book") (see TSHAL: 21a; TSHAL1: 45, Inaba-Satō: 103; Chen-Zhou: 41). The relevant passage states that his genealogy of Tibet's ruling families was for the most part taken from a summary of Byang ji ston pa's work, which had been written upon his request by a dPag thog pa Rin chen rdo rje or gSer thog pa Rin chen rdo rje.17

The Deb gter/ther dmar po, the earliest extant Tibetan example of an attempt at writing a global history, has so far been translated into Japanese and Chinese (see Inaba-Satō; Chen-Zhou; and also the papers of Bira, 1964, 1984). To be sure, its scope and the underlying conception of its composition can only be understood against the background of the Mongol conquest of Tibet in 1240 by Ögödei Khan (r. 1229-1241), the subsequent establishment of a central governing body under the 'Bri gung pa and Phag mo gru, and its inclusion into the Mongol empire. Under Qubilai Khan, Tibet became formally part of the Mongol empire in China, and the change of local government in the 1260s, headed this time by Sa skya, together with the preeminent position held by prelates from Sa skya, made it possible for Tibet, as during the imperial period, once again to make an entry onto the stage of world history, albeit this time of course not as a sovereign state, but under Mongol overlordship. The Mongol domination of Tibet from 1240 to 1368 had far-reaching effects on Tibet's religious and political institutions, as well as on the development of the Tibetan language and historiography. One of these was the adoption of numerous Uighur/ Mongol and Chinese loan words. Indeed, the very term deb gter/ther (gter and ther are homophonous) in the title of Tshal pa's work is an example of such a loan word; in fact, it is its first attestation in written Tibetan. It undoubtedly entered into the Tibetan lexicon from the Mongol debter which, in turn, ultimately derives from the Greek via the media of Persian and Old Turkish. While the introductory remarks in both recensions entitle it Deb gter dmar po, the chronicle is also known as the Hu lan deb gter/ther, where hu lan corresponds to Mongol ula'an/ulaghan, "red," a title which occurs at the very end of what may have been the original text (see TSHAL: 38b; TSHAL1: 149; Inaba-Satō: 194; Chen-Zhou: 128).

Tshal pa's notion of historiography is a traditional one, one which in another context Collingwood (257 ff.) has called the "scissors[page 45] -and-paste" approach to history, characterizing it as "...a kind of history which depends altogether upon the testimony of authorities." Tshal pa not only made use of a number of Indic and Tibetan sources, but also of treatises (originally) in Mongol and Chinese. A case in point of the former is the so-called Yeke tobčiyan ("Great/Large Records"), which, though they cannot be identified with any precision, could very well refer to the lost genealogical tables of the Mongol imperial family on which the relevant chapters of the Yuanshi are based, or perhaps even to the Dayuan tongzhi collection of legal documents.18 In this connection, we should note that for information on early Sino-Tibetan relations and for the royal/imperial genealogies of the Chinese, Xixia and Mongol empires, bLa ma dam pa, Yar lung pa, the chronicle of 1434 (with a few later interpolations) of sTag tshang pa dPal 'byor bzang po, alias Śrībhutibhadra, 'Gos lo tsā ba gZhon nu dpal's (1392-1481) Deb gter/ther sngon po of 1476-1478, Paṇ chen bSod nams grags pa's (1478-1554) Deb ther dmar po gsar ma ("New Red Annals") of 1538, and dPa' bo depend almost exclusively on the relevant texts in Tshal pa's compilation. Of interest is that dPa' bo contains a translation from a Chinese work on the spread of Buddhism in China which, he takes special pains to specify, is not met with in the so-called Deb dmar (see DPA'2: 567-572; DPA': 1391-1396). In his remarks that preface the reproduction of this work, he writes that it was first translated from Chinese into Uighur by a Uighur monk called Og zo at the order of Emperor Buyantu (r. 1311 to 1320); subsequently it was rendered into Tibetan in Sa skya Monastery by a Puṇyaśrībhadra (= bSod nams dpal bzang po), who was probably a Uighur as well.19

At the outset of the Deb gter/ther dmar po, in his statement of intent, Tshal pa writes that what follows is "the first of three Deb gter dmar po [texts]"; unfortunately, the other two, if they were ever written, are wanting. However, Dung dkar Blo bzang 'phrin las, the editor of the Beijing recension, does observe that he wrote in addition to other works (which include two biographies) a supplement to a/the Deb gter/ther dmar po,20 a rgyal rabs entitled Deb ther khra po ("Multicolored Book"), and a catalogue of the so-called Tshal pa bKa' 'gyur, which bore the subtitle of Deb ther dkar po ("White Book").21 Of some interest is of course the use of color terms in the titles (or subtitles) of books. This was unprecedented in Tibetan historiography and is something that is very Mongolian indeed.

[page 46] The third historiographic genre is that of the Chos 'byung ("Origin of Buddhism"). The very first of such texts may have been the one written by the eleventh-century scholar Rong zom Chos kyi bzang po of which only a few fragments have surfaced so far.22 Although the reasons are still far from transparent, it is possible that with the proliferation of various doctrinal cycles a need was felt to place these in historical perspective and thereby legitimate them. In any case we find, starting with the twelfth century, an enormous upsurge of interest in Indo-Tibetan religious history in particular. Unfortunately, only a fraction of the potentially available literary corpus of such texts has been located and published to date. For, while those authored by the bKa' gdams pa masters Phya pa Chos kyi seng ge (1109-1169) and his student gTsang nag pa brTson 'grus seng ge have yet to be discovered, the earliest extant text belonging to this genre is the Chos la 'jug pa'i sgo ("Introduction to Buddhism") by the second Sa skya pa patriarch Master (slob dpon) bSod nams rtse mo (1142-1182), a work which he completed towards the end of 1167 or the beginning of 1168. The overall approach to its subject-matter and its architecture typifies many subsequent chos 'byung texts such as those by *lDe'u Jo sras, mKhas pa lDe'u and Bu ston Rin chen grub, but we do not know whether he was indebted for these to his teacher Phya pa. His work was succeeded by the magnificent text of Nyang ral's chos 'byung which, however, bears little resemblance to it in terms of its scope and structure. bSod nams rtse mo's text deals in the main with the life of the Buddha, while Nyang ral principally deals with the religious environment of Tibet's imperial period. The thirteenth century, too, knew of a considerable number of such treatises, the sole information concerning which is owed to a very brief remark by Bu ston as well as potentially to a number of quotations in his own chos 'byung. He notes the existence of such treatises by Khro phu lo tsā ba Byams pa'i dpal (1172/73-1236), Chag lo tsā ba Chos rje dpal (1197-1264) and mChims Nam mkha' grags (1210-1285) to which he apparently had access when writing his own well-known work sometime between 1322 and 1326. The present whereabouts of these treatises, if they are still extant, is unknown. As few as two bona fide chos 'byungs that probably belong to this century have come down to us, namely those by *lDe'u Jo sras and mKhas pa lDe'u. Ne'u Paṇḍita Grags pa smon lam blo gros' sNgon gyi gtam[page 47] me tog phreng ba ("An Account of the Past, A Garland of Flowers") of 1283 (Chab spel, NE'U), while often referred to as a chos 'byung, styles itself in the introductory lines as a rgyal rabs. There is much in the manner in which the subject-matter is treated that is strongly reminiscent of a chos 'byung, so that we may characterize it as a text that falls midway between these two other genres.23

Other historiographic texts, that are sometimes styled, or that sometimes incorporate, smaller texts variously called lo rgyus, rgyal rabs, or chos 'byung, would be a limited number of so-called treasure-texts (gter ma) (see Gyatso, in this volume). A case in point is the bKa' chems ka khol ma, putatively Srong btsan sgam po's (?-649/50) testament (bka' chems), which was allegedly retrieved from a hole in a pillar (ka khol ma) by Atiśa (982?-1054?) in ca. 1049. It figures among the earliest such treasure-texts, and a number of particulars of its textual history were delineated by Vostrikov (28-32) and recently by Eimer (1983a). Although two versions were published some years ago, the best recension appears to be the one that was issued a few years ago by sMon lam rgya mtsho on the basis of two handwritten manuscripts, one at the Central Institute of Minorities, Beijing, and one written in silver on dark blue paper that belongs to the library of bLa brang bKra shis 'khyil Monastery in A mdo. In the colophon, the text elicits the following course of its transmission: Atiśa; Bang ston [Byang chub rgyal mtshan]; sTod lung[s] pa [Rin chen snying po] (1032-1116); sPyan snga ba [Tshul khrims 'bar (1033-1103)]; sNe'u zur pa [Ye shes 'bar (1042-1118/19)]; 'Bri gung pa [read here ?"lHa (chen) 'Bri sgang pa"]24; rGya ma ba; Rwa sgreng pa; dKon [mchog] bzang [po]; rDo rje tshul khrims25; "me." Who is this "me"? Obviously, he must be one with strong ties to the bKa' gdams pa school and he must have flourished sometime towards the end of the thirteenth or the beginning of the fourteenth century.

The bKa' chems ka khol ma contains a great deal of interest concerning the reign of this first Tibetan religious king, and also contains a number of prophecies in the sixteenth chapter. It served as a primary source for later Tibetan accounts of that period, including, it would appear, the narratives of Thon mi Sambhoṭa's alleged invention of the Tibetan script and the arduous journeys to,[page 48] and sojourns at, the Nepalese and Chinese courts by the minister mGar sTong btsan yul bzung for purposes of escorting a lady of their ruling houses to Tibet for betrothal to Srong btsan sgam po. Although it is supposed to be the work of the latter, it contains some information which perhaps impeaches the veracity of this imputation. For one, it mentions the date in which he passed away to the exact day and includes a number of prophecies in its sixteenth chapter that most of us would consider to be evidence for much later compilation, since Atiśa is mentioned in them!26

Of the twelfth and thirteenth century chronicles known to date, the text is only mentioned in the works of Nyang ral and mKhas pa lDe'u. The latter, if he is indeed to be placed in the second half of the thirteenth century, refers to it in passing just prior to his narrative of the building of the Ra sa phrul snang, the gTsug lag khang temple in the center of Lhasa.27 While he does not explicitly cite it in his account of the life and times of Srong btsan sgam po (although there can be no doubt that he must have used one or another recension of this work) Nyang ral records a few details of its retrieval in the bibliographic remarks at the end of his chronicle.28 There he writes that the document (yi ge) of the rGyal po bka' chems was of difficult access, and that Atiśa retrieved three works from a central beam (gdung bar) of the Ra sa phrul snang temple, namely the "bKa' rtsis chen mo written by the kings, the Dar dkar gsal ba'i me long written by the queens, and the Zla ba'i 'dod 'jo written by the ministers." He furthermore appears to hold that these three are known together as rGyal po bka' chems, which in turn seems to refer to the bKa' chems ka khol ma. This might be confirmed by what may be the best recension of the text itself, the first chapter of which states that Atiśa and two assistants excavated three scrolls (shog dril) from atop a jug-shaped pillar, or a (hollow) pillar containing a jug within it (ka ba bum pa can gyi steng nas), where the first and the third, here noted as the Zla ba 'dod 'jo and the bKa' chems kyi yi ge, are described as being lo rgyus.29 In addition to these texts, the bKa' chems ka khol ma also signals the existence of several other early treatises on which it may be based.30 A detailed study of this highly significant work, which perforce needs to include a comparative analysis of the various recensions (at least three are known to me) that have come down to us and the various recensions of the Maṇi bka' 'bum, is one of the many urgent desiderata in the area of Tibetan historiographic research.

[page 49] Of course, because gter ma texts are considered to date from Tibet's imperial period, many came to be considered crucial sources for this period in later historiographic works. A case in point is Nyang ral's chronicle, for not only is its discussion of the life and times of Srong btsan sgam po largely based on the bKa' chems ka khol ma, but Nyang ral also incorporated into it significant portions of the Zangs gling ma biography of Padmasambhava, a gter ma in its own right, which he himself had retrieved earlier.

As has hopefully become evident, the earliest Tibetan historiographical materials are extremely diverse and, regrettably, to a large extent still unpublished. Investigations into the literary sources used by authors of those texts that are available to us are also in their infancy, as is, consequently, research into the particular ways in which they have made use of them. This renders it particularly difficult to determine the original contributions made by these early authors in terms of how they interpreted them when they were not simply incorporating large portions of their sources into their own work.

Notes