Printer Friendly

East of the moon and west of the sun? Approaches to a land with many names, north of ancient India and south of Khotan.


Commemorating 50 years of Indian hospitality towards the Tibetan people, what could fit better than a contribution concerning the region at the very junction of Tibet and India: Ladakh? Once a colony of the Tibetan empire, independent for almost one millennium, Ladakh is now part of India, her crown, as some politicians would say. In the west, Ladakh is also known as Little Tibet, by which designation most people understand something secondary, a miniature replication of something more real, Tibet. This perception is wrong in two ways. First of all, the privilege of being called Little or Lesser Tibet goes to Baltistan, while Ladakh was known merely as Greater Tibet. This terminology reflects an ancient convention, attested in Indian, Chinese, and Tibetan nomenclature, by which Lesser means Closer to a particular reference point. This reference point could be a neighbouring state who applied the terminology from its own perspective, (2) it could be a more common, pan-national reference point, such as Mt Meru, the Central Asian axis mundis, and it could be the geographic or political Centre of the entity itself. The term Greater would thus apply to territories further away from the reference point or to politically peripheral regions, regions that were secondarily acquired and colonialised. In this latter way the Tibetans applied the term Bod for Central Tibet and Bodchen for Amdo and Khams.

Ladakh and Baltistan are commonly perceived as an intrinsic part already of Ancient Tibet, and as such their distinct linguistic, cultural, and political history, the Indian and Iranian influences, have often been underestimated. In general, apart from the 'nation'-building fictions of the royal genealogies, we do not know much about Tibetan prehistory and early history from independent sources. While there are ample studies (and good overview volumes) concerning Tibet's neighbours or more broadly South, Central, and East Asia, prehistoric Tibet apparently lies in the blind angle of any such approach. It is as if it never existed. The situation for Ladakh and Baltistan is even worse, if an augmentation of nothingness is thinkable, at all.

Examining the early history of Ladakh as well as that of the more fabulous than historically traceable Zarjzurj. one cannot avoid coming across the names Yangtong, Suvarnagotra (Gold Clan/Family), Nuguo (Women's Dominion), and Moluosuo (with its seemingly Tibetan equivalents Mard and Maryul) or Sanbohe. These names are used by Chinese historiographers and travellers in an all-too-often contradictory manner, and one may thus wonder whether these entities have any reality at all or whether they are just faeries or spiritual realms (like the Bonpo Holmo Lurjrirjs) beyond the reach of an ordinary, unenlightened human being.

The first name, Yangtong, bears a certain similarity with the Tibetan name Byajthaj (Changthang), and if there is some etymological relation, then the name must be quite old and certainly not signifying 'northern plain' (a designation that only makes sense from the later Tibetan perspective). There seems to be also a certain phonetical similarity between the designations Yangtong and Zanzun. and many scholars believe that the two names are, in fact, etymologically related.

I started this research with the presumption that it would hardly be possible to correlate the designation Yangtong from the 6th-7th c. (or even earlier) Chinese sources in a meaningful way with that of ZanZun in the somewhat later 7th - 9th c. (or even later) Old Tibetan sources, and that particularly the subdivisions could not be matched. It turned out, however, that, despite the contradictions in the sources, the congruence is quite substantial, and that, surprisingly, the main stage, where the above mentioned entities come together, does not lie in Guge but in various parts of Ladakh, Baltistan, and even Hunza. Hence it also turned out that phyidar and post-phyidar Bonpo references to a ZanZun as being part of, or at least as bordering on, 'Persia' (Tazig, Staggzig), or her Bolorian borderland, unbelievable as they had appeared to many scholars, must have been based on the knowledge of real geo-political facts.

0.1 Sources

The present study will be based mainly on the following sources:

* Ptolemy (2nd c. CE; Lindegger 1993): he gives precise coordinates that allow to draw maps relatively accurately.

* The Brhatsamhita (ca. 5th c. CE, based on the ca. 3rd c. Parasaratantra; Fleet 1973): a very general geographical overview, unfortunately completely confounded in the northern sections, furthermore without any indication of relations in terms of direction between the various peoples or which people live in the mountains and which in the plains.

* Various Tang annals (7th c. CE onwards; Rockhill 1891: 339-341, Chavannes 1900 (1969) Pelliot 1963: 688-720).

* The report by Xuanzang ([TEXT NOT REPRODUCIBLE IN ASCII],W.-G. Hsiian-tsang; ca. 640 CE; Beal 1881-84).

* The report by the Korean pilgrim Hyecho (chin. Huicha [TEXT NOT REPRODUCIBLE IN ASCII], W. G. Huei-ch'ao; ca. 730 CE; Fuchs 1938).

* The Old Tibetan Annals, Pt 1288, IOL Tib J 750 (OTA) and OR 8212.187 (OTA II; ca. 640-764 CE; Dotson 2009): short annalistic entries that may mention military campaigns, but usually not the exact whereabouts.

* A Khotanese 'prophesy' in Tibetan, the Vimalaprabhapariprccha or Drimamedpahihodkyis zuspa (ca. 700 CE, cf. Thomas 1935: 139-258): a rather confused and legendary narrative of limited historical value.

* The Hudud-al-'Alam (ca. 10th c. CE; Minorsky 1937): its geographical misconceptions in the spirit of the time hardly add to our knowledge, and the text can only be interpreted with the knowledge it implicitly presupposes.

* Kalhana's Rajataranginl (12th c. CE; M.A. Stein 1900): given its late composition, the historical facts of earlier centuries are reported rather summarily following the stereotypes of the time. Furthermore, there is not a single positive reference to Western Tibet, Ladakh, or Baltistan. (3)

* Some post-imperial and post-phyidar Tibetan sources (Buddhist or Bonpo), the most important of which is Dpaho Gtsuglag Hphrerba's Choshbyuj mkhaspahi dgahston (16th c. CE, cf. Tucci 1956: 91), which still draws upon imperial sources.

* The Ladvags Rgyalrabs (LDRR, ca. 17th c. CE; Francke 1926).

Among all sources, the Chinese sources are the most concise and reliable, although they are far from being consistent. Some fixed points are, however, given, from which to measure orientations and distances. These are: Khotan in the north, Jalamdhara (and Kulu) in the south, and Bolor to the east of our unknown entity. The inconsistencies can thus be reconciled to a certain extent with a rather small amount of interpretation. The most precise descriptions come from two pilgrims, the Chinese Xuanzang and the Korean Hyecho, but the sources in the Chinese annals that Pelliot (1963) displays and discusses in great extent, must also have been based on reports from foreigners (merchants as well as ambassadors) questioned by the Chinese authorities, as well as on reports from travelling Chinese officials.

Since these geographical entities overlap with present-day Ladakh and Baltistan, I will also discuss the name that has erroneously be taken for the old name of Ladakh: Maryul (Old Tibetan Mard, Chinese Moluosuo, Sanskrit Suvarnabhu) as well as the origin of the name Ladvags.

0.2 Visualisation of previous views

Previous attempts to pinpoint the ancient geography of Yangtong or ZanZun have been thwarted by basically three misconceptions: 1. the identification of Zanzun and Maryul with the post-phyidar entities of the same name, 2. the idea that the various geographical entities would be separate, albeit adjacent entities, 3. that the populations of the entities in question could be described as homogeneous groups or 'nations', basically of Qiang, that is, Tibeto-Burman origin. It may be useful to envisage the previous conceptions schematically as follows. The sketches are, naturally, only very rough approximations, which means that a particular place could be covered by a geometrical figure representing an entity whereto the place definitely does not belong. In the case of the Hudud-al-'Alam (Map 2), the localisations are based on mere guessing.







The scaled background map that will be used here and in the following was designed by Chr. Gigaudaut for Tournadre and Sangda Dorje (1998: 6, L'aire linguistique tibetaine). It should be noted that the left-most meridian is inclined by approximately 20[degrees], cf. the maps given in Stein 1981: 14-15 and 58-59. This will be adjusted in the detail maps. Where necessary, I will add a non-scaled map from Albinia (2008: xii-xiii, Map of the Indus Valley) for the adjacent western regions.

0. 3 Conventions and abbreviations

My perspective on Tibetan is that of a linguist (language-scientist) or a philologist (lover of the word) in the true sense. In transliterating Tibetan

1, therefore, will not represent the text in an unstructured flow of syllables (either separated by space, hyphen, or dot), but will represent 'words' as what they are, namely possibly polysyllabic intonational units. The Tibetan tsheg (the dot between the graphic syllables) does not serve to separate isolated monosyllables, nor does it necessarily mark a morpheme boundary. It is merely a graphical device to help identifying the syllable core, when no vowel sign does the job. Since the inherent vowel a, not represented in Tibetan, is 'transliterated' by convention, there is absolutely no need to take recourse to a syllabic representation. The Wylie system, however, while certainly justified in anthropological and related studies is unsuitable for the rendering of composite words as words, hence I will have to take recourse to diacritic letters.

Tibetan place names will be given either in their old (written Tibetan) or their modern form. In the first case, I will use the transliteration system with diacritics, if necessary. I will use the old form whenever discussing Old Tibetan place names, further when the written form corresponds to the modern pronunciation, otherwise I will use the modern forms, as used on maps etc. Tibetan personal names and titles are only given in transliteration.

Chinese names will be transcribed in simple Pinyin. Wade-Giles (= W.-G.) forms and Chinese characters will be given only in cases of high relevance.

BRGY Bod-Rgya tshigmdzod chenmo, Zhang et al. (1993)

EFEO Transliteration style of the Ecole Francaise d'Extreme Orient

LDRR Ladvags Rgyalrabs, Francke (1926)

KHAL Narrations and information by meme Tondup Tsering from Khalatse

OTA The Old Tibetan Annals, Dotson (2009)

Pt Pelliot tibetain (manuscripts of the Pelliot collection)

W.-G. Transliteration style according to Wade & Giles

1 Western Tibet and the Changthang-an ethno-geographical puzzle

The nomadic areas stretching along the northern steppes of Tibet must have seen an early Indo-European population, particularly in the West, which gradually mixed with people of Hunnic and/or Mongolic stock. At least, after the invasions of the Yuezhi, Tuyuhun/Haza, and Hephthalites/Huna into Northern India and Afghanistan, the people dwelling along the real upper course of the Indus (4) must have acquired many Iranian cultural traits, if only outwardly. Even after the Tibetan conquests in the West, people continued to pour in from various regions of the Pamirs and beyond.

The Korean pilgrim Hyecho, travelling in India around 726/727, most probably did not reach present-day Ladakh or Baltistan, but he apparently collected a great deal of information in Kashmir and Bolor. According to him or his informants, the people of Greater Bolor (Da-Bolu, [TEXT NOT REPRODUCIBLE IN ASCII] W.-G. Ta Po-lu, Po-lu-lo, or P'u-lu) and Yangtong ([TEXT NOT REPRODUCIBLE IN ASCII], W.-G. Yang-t'ung), as well as those from the unidentifiable Suoboci (HfffM, W.-G. So-po-tz'u) (5) were all Hu (quasi Iranians) and, therefore, believers, i.e. Buddhists. All three peoples are perceived as being clearly distinct not only from the people of Kashmir, but even more so from the inhabitants of Tibet (Tufan guo, [TEXT NOT REPRODUCIBLE IN ASCII] W.-G. T'u-fan kuo), who are described as non-believers and hence as comparatively less civilised (Fuchs 1938: 443f.). The Tibetans are described as nomads living in felt tents like the Tujue (that is, Turks; p. 443), alternatively also in caves dug out of the ground (p. 444). The Tibetans are further described as having a very dark complexion (p. 444). While Tucci (1977: 72) objects to the presence of Buddhism in Zanzun. (6) Petech (1977: 10) objects to the presence of Hu in this area. According to the latter,
      the term Hu applied to the Iranians of Central Asia; but
   its use was rather loose, and it appears that for Hui-ch'ao it
   applied generally to the Iranian populations, which would fit
   perfectly well with the Dards of Ladakh (but not with the
   people of Zan-zun).

While Dards and Iranians are not exactly the same, perhaps not even in the eyes of a Chinese traveller, both could have been comprised under the quite unspecific cover term Hu. While Mole (1970: 152f, n. 392) states that
      Hu originally designated the Huns, but in this epoch [618-626]
   had come to mean barbarian in general, and designated
   not only the T'u-yu-hun but also the population of the western
   lands (Hsi-yu),

J.E. Hill (2004, note 1.23), holds that
      Hu is a rather vague term used for northern and western
   peoples of non-Sinitic origin, usually, but not exclusively,
   for those of Caucasian appearance. It was commonly used
   for people of Persian, Sogdian, and Turkish origin, Xianbi,
   Indians, Kushans and even, occasionally, for the Xiongnu
   (who, however, are usually clearly differentiated from the

The term Hu could thus well have referred to Xianbi tribes, such as the Sumpa, which are usually counted among the aboriginal peoples of northern Tibet. Nevertheless, given the context of Hyecho's travels as well as the context of his statement, particularly the contrast with the Turks, one may be justified to take his 'Hu' as being similar to the Indo-European population of Khotan, the Pamirs, and the Hindukush. Therefore, I do not see any necessity to question a cultural or linguistic Iranian influence in Zanzun. especially not in its western parts or in the Changthang.

None of the regions in question is geographically well defined, the identification of each depending in almost perfect circularity upon the exact definition of the boundaries of the neighbouring regions. The correlation (if only partially) of Yangtong with 'Grand'-Zanzun or with the Changthang seems to be uncontroversial, but there is no consensus as to the exact location and extension. The only fixed points are that Lesser Bolor can be identified with Gilgit and/or Hunza (Bruza) (7) and that some directions and distances are given with respect to the Tarim basin and the Pamirs. The main problems are however:

* Where exactly lies Greater Bolor or how can it be defined?

* Where exactly are the north-western confines of Yangtong/Zanzun?

* Where exactly do we find the 'Gold Race Country' Suvarnagotra, the 'Women's Dominion' (Strirajya/Nuguo), and Greater Yangtong; and how can we explain the rather consistently conflicting data concerning these three names?

* Are the names Mard and Maryul, on the one hand, and Moluosuo, on the other, related to each other? And where would this entity (or these entities) be located?

1.1 Baltistan and Bolor

It had been commonly assumed that Baltistan is identical with the Greater Bolor of the Chinese sources. This would seemingly fit Hyecho's indication that Greater Bolor lies 15 days to the north-east (Fuchs 1938: 443) and Lesser Bolor 7 days to the north-west of Kashmir (ibid. p. 444). Hyecho's indication is not without oddities: Gilgit and Hunza lie due north of Srinagar, and the distance given between Kashmir and Gilgit appears to be much too short.8 This could imply that Lesser Bolor extended southwards along the Indus. On the other hand, the distance from Shrinagar to Skardo should have been either less or more than 15 days. The short summer route took 12 days "up the Gilgit road to the Burzil Chowki and thence over the Deosai Plains", the longer route, open the greater part of the year, led "in eighteen marches up the Sind valley, [...] over the Zozi [!] La, [...] and thence down the valleys of the Dras and Indus" (Workman 1905: 246). Hyecho describes both Bolors as being culturally and linguistically identical. The king of Greater Bolor had fled from the Tibetans to Lesser Bolor where he had managed to install himself as ruler. The rest of the nobility, however, had stayed in Greater Bolor (Fuchs 1938: 444).

Other Chinese sources are even more ambiguous. In the Tangshu it is stated that Greater Bolor was to the south-east of Lesser Bolor, an indication that could point to Baltistan, except if we follow Denwood's (2008: 15) suggestion that Lesser Bolor centred on the Yasin valley. Greater Bolor is further said to be straight west of Tibet, which fits better with Chilas than with Baltistan, and that it bordered in the west to Wuchang (Uddiyana), an indication that can by no means be reconciled with the location of Baltistan and thus necessarily refers to Chilas. Cf. Chavannes (1900 (1969): 149f.):
      Le grand Pou-lu (Baltistan) est aussi appele Poulou; il est droit
   a l'ouest des T'ou-po (Tibetains); il touche au petit Pou-lu; a
   l'ouest, il est voisin du territoire d'Ou-tch'ang (Oudyana) de
   l'Inde du Nord. [...] Il est assujetti aux T'oupo (Tibetains).
   Depuis la periode wan-soei-t'ong-tien (696) jusqu'a la periode
   k'ai-yuen (713-741), il envoya trois fois des ambassadeurs rendre
   hommage a la cour. C'est pourquoi on confera par brevet le titre de
   roi au prince de ce pays, Soufou-cho-li-tche-li-ni; a sa mort, on
   confera encore par brevet la succession royale a Sou-lin-t'o-i-tche
   (Sourendraditya?); en tout, celui-ci envoya deux fois de hauts
   dignitaires apporter en tribut les produits de son pays.

      Le petit Pou-lu est a plus de neuf mille li de la capitale; a
   trois mille li a l'est tendant un peu vers le sud, on arrive au
   campement du T'ou-po tsan-p'ou (le btsanpo de Tibet); a huit cent li
   du cote de l'est, (9) ce pays touche a Ou-tch'ang (Oudyana); a trios
   cents li au sud-est est le grand Pou-lu (Baltistan); a cinq cents li
   au sud se trouve le Kou-che-mi (Cachemire); a cinq cents li au nord,
   il y a la ville de So-le du (pays de) Hou-mi (Wakhan) ... (10)

According to Jettmar (1980: 120-122), Greater Bolor can only be identified with the valleys of Chilas and Astor, not with Baltistan, particularly because--according to the bizarre narration of the Vimalaprabhapariprccha, see below,--some kings of Skarrdo bore names that are not part of the tradition of Greater Bolor and because the capital of Greater Bolor is called either Hesalao or Pousalao in Chinese sources, which, according to him, could not be identical with Skardo. (11) Jettmar also disputes the strategic importance of Baltistan, as the main routes, including those from Western Tibet to Gilgit, are not to be found in this region (ibid. p. 121f.; but see Tucci, 1977: 79, 81-84). The main argument, which Jettmar receives from Tucci (1977: 80) is a passage in a comparatively late source, authored by Hu Sanxing (1230-1302):
   The great P'u-lu [...] is straight to the west of Tibet. North
   of it there is little P'u-lu (emphasis added, BZ).

Tucci concludes:
      This corresponds to the location of Chilas, having to the
   north little P'u-lu, Gilgit, but being in the west of Baltistan,
   subject to the Tibetans.

Evidently, Hu Sanxing did not know of a politically distinct Baltistan. Correspondingly, Denwood (2008: 15) takes Baltistan and Greater Bolor as two different political entities, stating that
      the core of Little Palkur at this time was the Yasin Valley,
   where its capital was located and its king based. Very likely
   the kingdom extended westwards to include the little-known
   but apparently fertile and populous Baushtaro Valley on the
   north side of the Gilgit/Ghizar River, and perhaps also the
   upper part of the Ishkoman valley to the northeast, reachable
   from Yasin by the easy Panji pass. [...]

      Great Palkur must have been the valley of the Gilgit River
   below about Hatuj. at least as far as [...] the Gilgit/Indus
   confluence [...] and extending down the Indus valley and/or
   across the Gilgit mountains to the Indus in Diamar.

As the present-day sub-district of Diamar comprises Chilas, Greater Bolor should have extended to the south at least as far as Chilas. According to Mirza Haidar (Elias 1895: 385), Bolor reached even further down south and west, namely as far as Lamghan (modern Jalalabad) and Kabul (or perhaps rather: as far as the Kabul river). This would mean that the whole valley of Chitral, west of Swat, was part of (Lesser) Bolor.

Xuanzang describes Bolor as lying on the right side of the Indus, somewhat above the Hanging Passages. He reached there from Daliluo (Kandia?), (12) the ancient capital of Uddiyana, after a march of 500 li (ca. 175 km). Unfortunately it is all but clear from where this distance was counted: from Daliluo, from the spot where he reached the Indus, from the beginning of the Hanging Passages at Jalkot, or from their end at Sazin or Shatial. The 500 li could roughly correspond to the distance between Shatial and Gilgit (cf. the map given as frontispiece in Jettmar 2002). If, however, the distance is reckoned from Daliluo or at least from the first approach to the Indus, the confines of Bolor would lie somewhere between Shatial and Chilas. Xuanzang describes the land as being extended from east to west and as narrow from north to south, which fits the course of the Gilgit river as much as the east-to-western course of the Indus between Gor and Shatial. The inhabitants of Bolor are said to use letters "nearly like those of India, their language [being] somewhat different" (Beal 1881-84: 177-179, 178).

Xuanzang further mentions a region Boluoluo (W.-G. Po-lo-lo), which is reached from a valley in the Pamirs by crossing a pass in the south (Beal 1881-84: 481). This should have referred to Lesser Bolor, rather than to Baltistan, as Beal suggests.

Pelliot (1959: 92) should have reached at a similar conclusion as Tucci, Jettmar (1980), and Denwood. According to Pelliot, the Baluristan of Mirza Haidar is to be understood as "the mountainous tract south of Badahsan, south-west of Yarkand, west of Balti, north and north-west of Kashmir, that is to say the valleys of Chitral, Yassin, and Gilgit" and "Mirza Haidar seems to leave Baltistan out of Baluristan." (13) But he goes on to say "that the localization may have been viewed differently by others at different dates", referring to Cunningham (1854: 34), who writes: "Balti or Balti-yul, is called Palolo, or Balor, by the Dards, and Nang-kod [i.e., Nangon (14)] by the Tibetans". The following names have been attested: Hunza Balotza or BaIoc (Baloi,a however, referring to 'Tibet'), Nagar Balots,a Chilas Paloye? Gilgit Shina Palole,a Poldle,h Pale,d or P[conjunction]le'c (a: Leitner 1890s: 72, 78, b: Leitner 1889: 60, c: Lorimer 1938 as cited by Bielmeier 1985: 14, d: T.G. Bailey 1924: 155). This designation might perhaps refer not so much to a location as to a certain element of the population, which could have migrated to Baltistan at some later time. In the case of the Palula speakers of Chitral, Ruth Leila Schmidt (email communication 04/2008) thinks that Chilas might have been only the centre of a larger unit Palola, because, on the one hand, "the Palula (Paaluulaa) speakers of Chitral trace their roots to Chilas, and on the other hand, 'Puluulii' ... is still a nickname for a Burusho girl."

In striking contrast to his 1980 conclusions, but resuming earlier considerations (1977), Jettmar (2002a [1993]) is more than convinced that Baltistan not only was part of Greater Bolor, but constituted its core area, and Skardo served as the seat of the Palola dynasty. His arguments, however, are not without contradictions, (15) and the main problem with this assumption is that no inscriptions relating to the Palola dynasty are found in Baltistan, while there are several inscriptions in Chilas and neighbouring areas, such as Alam Bridge, shortly above the confluence of the Gilgit river and the Indus. For unknown reasons, Jettmar seems to preclude that the Palola dynasty may have had its (secondary) seat in Chilas, and he brushes away the indications given by Xuanzang.

The problem may perhaps never be resolved, but I would tend to interpret the data in a similar manner as Denwood. Bolor must thus have comprised a rather large area. When Bolor split into two halves, Yasin became the centre of Lesser Bolor, which seems to have comprised at least the upper parts of the Chitral-Kunar valley (having the Swat valley as its eastern neighbour). The conventional use of the terms Lesser and Greater might indicate that Lesser Bolor was the core area of Bolor, and Chilas (and perhaps also Baltistan) constituted a secondary, peripheral region.

Since Hyecho mentions no other country between the two Bolors and Yangtong/Zanzun. Baltistan must have belonged to one of these, thus if not part of one of the Bolors, it must have been, by necessity, part of Yangtong/ Zanzun. A Baltistan being part of Zanzun is certainly hard to swallow, and it stands to reason to solve the problem in Pelliot's manner or by suggesting that the Chinese commentator was using the name Tibet somewhat loosely, referring to the places that were under the power of the Tibetans at a certain moment. One could even argue that the directions are not to be taken too seriously, after all, errors in direction are unfortunately not so uncommon in Chinese sources: Pelliot (1963: 695f.) cites examples where Nepal is located west of Tibet, and India even north-west of the Onion-Range (Congling, the Pamirs). (16)

Such inclusion is certainly also at variance with the later Tibetan perspective, as reflected in the LDRR, which never mentions Chilas and differentiates between Bruza/Gilgit, on the one hand, and identifiable parts of Baltistan, on the other, cf. LDRR 32.6, mentioning 'conquests' or raids in the west under Khri Dussron (appearing as Gunsron or Hdusron Hdurje in the LDRR):

nub... | Sbaltihi srajgi Najgoj | smadkyi Sidkar tshunchad mjahhogtu bsduste

'[In] the west... Najgoj (i.e. Skardo) (17) on the road to Sbalti, from Sidkar of the lower areas hither [all] were subdued and'

A similar wording is found in the Mjahris rgyalrabs (Vitali 1996: 29.12-13):

nub... Sbalyulgyi Najgoj | smadkyi Sijhgar [!] tshunchad mjar bsduste.

'[In] the west... Najgoj of the Sbal[ti] country, from Sijhgar (Sidkar) of the lower areas hither [all] were subdued and'

The separate mentioning of Sbalti could well be a reinterpretation in terms of the actual geography at the time of composition (17th c.). The preserved Old Tibetan documents refer only rather vaguely to Bruza and do not mention Baltistan at all, as it might be included in their notion of Zanzun. Nevertheless, the name appears in one of those later historiographic works that seem to draw upon relatively early sources, now lost. The Vth Dalai Lama's biography of Bsodnams Mchogldan Bstanpahi Rgyalmtshan contains an interesting division of Mjahris (the successor in title to Zanzun), where the second skor (district or county) is defined as comprising Li, (18) Gruza (!), and Sbalte (Tucci 1956: 73). No mention of Ladakh is made, and Maryul is still in the immediate neighbourhood of Purang and Zanskar, in the first skor, so that we may assume that the sources from which the Vth Dalai Lama drew his information reflect the situation of the late Empire. A similar division is found in BRGY (sub Mnahris skorgsum), where Khotan (Hchimmam Li), Bruza, and Sbalti form one of the skor.s. (19)

Given the fact that Hyecho's Yangtong grows grapes, and that grapes did grow in Lower, but not in Upper Ladakh (see below), one can at least definitely rule out the possibility that Bolor extended over Lower Ladakh. At the time of Hyecho's visit in Bolor, it is quite likely that Baltistan, if not an integral part of Yangtong anyhow, was already integrated into the administrative entity of Zanzun stod by the Tibetans.

Therefore, we cannot but accept that at least in terms of an international geography of the day, there was a part of Zanzun. an appendix of the Changthang: Lesser ZanZun (West)--actually, the ZanZun stod of OTA (see next section for the definition of the term), that comprised Lower Ladakh and perhaps also Baltistan. Perhaps some of the confusion in Chinese geographic accounts could be solved, if Lesser ZanZun (West) is taken to be identical with Lesser Yangtong.

Such a Lesser ZanZun (West) populated by Hu could no longer be rejected by Petech and a Lesser ZanZun (West) in which Buddhism is practised could likewise not be rejected by Tucci. That an Indian, Kashmirian, or Gandharl type of Buddhism was practised in Baltistan, Purik, and Lower Ladakh from the Kusana period until the 8th c. CE or later, as evidenced in many rockcarvings, is certainly uncontroversial (except perhaps for the exact dating of the artefacts) and does not need further elaboration here.

Taking a diplomatic stance, one could still argue that the borders between Bolor and Yangtong/ZanZun may have been subject to changes and that the principalities of Shigar and Skardo might have temporarily been part of Greater Bolor under the rule of the Palola Sahi (as seems to be suggested by von Hiniiber 2004: 8) or even provinces of Khotan (as seems to be indicated by the Vimalaprabhapariprccha, yet see the discussion below), but that their status remained undefined with alliances to all sides.


1.2 ZanZun

Like so many other place names, the designation ZanZun turns out to be highly ambiguous and therefore I should like to distinguish between:

(a) 'Grand-ZanZun' as a widely extended geographical term, assumed to be synonymous with the Chinese designation Yangtong and covering most of the Changthang-whether the historical designation ZanZun was ever applied in this sense is somewhat questionable.

(b) 'ZanZun Proper', more or less identical with the later West Tibetan kingdoms of Guge and Purang (Spuhrarjs)-this follows the identification by the later Tibetan sources, but again it is rather questionable whether the designation of the Old Tibetan sources applied to this region and only to this region.

(c) 'Local-ZanZun', a small province of the kingdom, perhaps its core part, or at least that region that continues to bear the name under the West Tibetan dynasty. 'Local-ZanZun' is only one of the elements of one of the districts (skor) of Mjahris.

(d) From 'ZanZun Proper' I should like to distinguish the peripheral or outer provinces in the west, which later became Baltistan and Ladakh. It is unclear whether, which parts, and for how long these regions may have been part of the historical kingdom/confederacy of Zanzun. or by which logic they were included under a merely geographical or even under a nostalgic notion of ZanZun. I will refer to these areas, for the sake of convenience, as 'Lesser ZanZun (West)'. (20)

Bonpo tradition knows of an Inner, Middle and Outer Zanzun. but the original references seem to be partly lost or of a merely speculative nature. According to those traditions that localise Inner ZanZun in 'Persia', (21) these designations could perhaps be translated into 'west' ('Persia' or the Iranian borderland in the Pamirs), 'central' (ZanZun Proper), and 'north-east' (the rest of Grand-ZanZun). According to Karmay (1998: 114), however, Outer ZanZun would have comprised a huge area from Gilgit to the Namtsho and from Khotan to Chumig brgyadcu rtsagnyis (Mukhtinath, according to Iida 2003), Inner ZanZun would have been 'Persia' (Staggzig), while the middle one remains unidentified.

It seems to be questionable whether a political entity ZanZun existed on Iranian territory and be it only at the borders, and I cannot avoid the feeling that the association of ZanZun with the Indo-Iranian Holmo Lujrijs (i.e. the Oxus valley and/or Chitral) is of a secondary, rather nostalgic origin, created by immigrant Bactrians or people from the Pamirs, known under the Tibetan appellation Rmu. While these Rmu apparently participated in the Bonpo re-invention of the past during the phyidar, they might have identified themselves, for lack of other alternatives, with a former political entity Zanzun. re-designed as the homeland of the holy teacher Gsenrab Mibo-associated with the past glory of Achaemenid or Sasanian Persia. On the other hand, if it were true that the Hephthalites or Huna or one of their associated tribes had been based also in the Kailash region (cf. Zeisler forthcoming a, Chapter 4 [section] 1.4.3), the subsumption of present-day Western Tibet with regions in present-day Afghanistan under the same 'national' or rather geographical heading would not be too surprising. It might perhaps be even due to this 'Huna connection', that some western areas, if not originally belonging to Zanzun. had been subsumed by the Tibetan administration under the designation ZanZun (stod) after their conquest in the 7th or 8th c. (see also below).

An interesting alternative division is transmitted in Ladakh. According to the historian Sonam Phuntsog (p.c.), one would have to distinguish between ZanZun Proper or ZanZun chunnn, located in Tibet, and a 'Greater' ZanZun or ZanZun chenmo, which would have comprised Ladakh, Baltistan, and Gilgit. This tradition might well reflect the addition of conquered areas under the administrative unit ZanZun.

Unlike many other scholars, R.A. Stein (1981: 13) thinks that ZanZun was in no way identical with any of the two, Lesser and Greater Yangtong, but was located between them. His indications are, however, contradictory. Lesser and Greater Yangtong, are, on the one hand, described as bordering on Khotan, Hunza, and Gilgit, on the other as being confined by Suvarnagotra or the Women's Dominion (p. 13), which should lie between Yangtong and Gilgit and Hunza. On his map (p. 58-59), ZanZun lies south of Greater Yangtong, and more or less in the same area as Lesser Yangtong. With R.A. Stein many other scholars look for ZanZun mainly in the area around Guge and Purang. This perspective, however, is not at all compatible with the Old Tibetan geography as presented by Dpaho Gtsuglag, and the attempts to accommodate the latter's descriptions to the current preconceptions add to the general confusion.

In a few Old Tibetan sources (cf. Takeuchi 2004: 54), and particularly in Dpaho Gtsuglag, we find the designations ZanZun stod and smad, which, according to a general convention, should have correlated to west (stod) and east (smad). According to Dpaho Gtsuglag, the two moieties are located at the boundary of Tibet and the Western Turks (Grugu for Drugu) and between Tibet and the Sumpa respectively (Ja 19 a, as cited by Tucci 1956: 91). The two moieties are divided into the following districts (sde): ZanZun stod: 1. Hoco, 2. Majma, 3. Gnema, 4. Tsamo, stonbuchun: Baga; ZanZun smad: 1. Gugge, 2. Cogla, 3. Sbyigtsan, 4. Yargtsan, stonbuchun: Cidi.

Both regions are identified by Tucci (ibid.) quite rashly with the Lob-Niya region and the region of Rgyalrong, while Gugge should be identical with Guran in Zims of Zanzun. mentioned in OTA (l. 64), for which, like anybody else, he is unable to give any localisation. These identifications are vehemently rejected by Yamaguchi (1970: 98, n. 1), who identifies Gugge in ZanZun smad with present day Guge. According to him (p. 100, n. 17), ZanZun stod would be the area of Ladakh, and ZanZun smad would be nothing more than ZanZun Proper. The same author associates the areas Spyirtsaj and Yarrtsaj (interpreted as Outer and Upper Rtsaj) of ZanZun smad with Rtsaj (p. 98), and his attempts to locate Rtsaj localities eventually leads him into Kham territory.

Although there might be other candidates around for the identification of Gugge (cf. the various place names with gog as an element), there is actually nothing that contradicts the identification with the more common form Guge, even more so as the other districts of ZanZun smad can be associated with place names in the closer or farther neighbourhood. The second element in the name of the third and the fourth district: Sbyigtsaj and Yargtsaj (in Dpahbo Gtsuglag) may well indicate that they were somehow related to, or part of, Rtsaj, as suggested by Yamaguchi. On the other hand, the first elements also show some similarity with Spiti and Yartse. The second district Cogla, may be identical with an area between Guge and Spiti, possibly the -lcog in the combination Spyilcog. Based on several sources, although not exactly specifying which is used for which identification, Hazod (2009: 168) gives the districts of ZanZun smad as "1. Gug-ge 2. Gu[g]-cog (~Cog-la) {between Guge and Spiti} 3. Spyi[r]-rtsang {Sato 1978: south of Khyunglung} 4. Yar-rtsang 5. Spyi-ti (~ Ci-de (= Spi-ti, also Sp[y]i-lcog; Petech 1997: 252)' (sic). (22)

Helga Uebach (p.c.) cautions that Dpaho Gtsuglag might have referred to some contemporaneous Drugu. Since the latter might have settled somewhere in Eastern Turkestan, but not somewhere on the Tibetan plateau, it would still follow from this reference that the boundary area between Tibet (proper) and the Drugu was at least Upper Ladakh and the western Changthang. But since Dpaho Gtsuglag lists a handful of obsolete place names in ZanZun stod, which must be drawn from some old documents, it is quite unlikely that he should have referred to contemporaneous Drugu in the same breath, and should not have based himself on an Old Tibetan description of ZanZun or on his knowledge as to where these places were to be located.

In OTA, the term Drugu refers to the Western Turks (cf. Beckwith 1987: 63f., n. 56). In the 7th c., the Western Turks had moved into the areas west of the Altai and north of the Tienshan and further west into Western Turkestan and into Afghanistan, where they replaced the Hephthalites, while the Eastern Turks mainly settled in present-day Mongolia and areas further to the south and east. By the mid 7th c., the Western Turks are found i.a. in Ferghana and Suryab, areas that could be accessed from Tibet directly via the Pamirs und thus via western ZanZun. This is what apparently happened: in the year 675, the minister Mgar Btsansna, after having carried out a registration (mkhos) (23) of ZanZun in Guran of Zims, went to Ltajyo in the land of the Drugu or went to the land of the Drugu for a (forced) ?trade agreement (Itanyo) (24) (OTA, 1. 64); in the following year, he led a military campaign into this area (OTA, 1. 67f.); in the year 687, after having proposed a campaign in the land of the Drugu in the previous year (which was postponed), the min ister Mgar Khrihbrij led a campaign to the lands of the Drugu and Guchen: Drugu-Guzanyuldu drans. (25)

The place names Guran and Zims are unidentified (cf. Thomas 1951: 268, for whom both places should be located in the region of Guge; Beckwith 1987: 42, n. 23, who abstains from guessing; Hazod 2009: 217, who states that they should be found in ZanZun). I suggest looking for them in the west. Gor in Chitral, below the confluence of the Astor river with the Indus, or Gurais (or Gurez), in the upper Kishenganga valley, may perhaps be the closest candidates for Guran, linguistically as well as geographically.

In 660, the Tibetans appear for the first time across the Pamirs, in Kashgar (Denwood forthcoming). Denwood, however, does not believe in a Pamir route, particularly because much of this area was still independent in the beginning of the 8th c.. According to him, the Tibetans would have reached the Western Turks by a route from (eastern) ZanZun via Shanshan or Miran through the Tarim.
      It would have been a simple matter to detach a small
   force and send it along the route south of the Tsaidam to
   the Charklik/Miran area (the old Shanshan kingdom on the
   southern silk route), then north and west along the course of
   the Tarim, Yarkand and Kashgar rivers, where water, grazing
   and supplies would be available in larger quantities than in the
   drier climate of today, and so to rendezvous with the Turkish
   armies near Kashgar. This route would avoid Azha armies as
   well as the Chinese-garrisoned cities of Khotan, Kucha and
   Aksu (ibid).

I do not think that such movements would have gone unchallenged by the Chinese, particularly as the routes were limited in number and certainly also surveyed. The mountainous regions of the Hindukush and the Pamirs were probably much more difficult to handle logistically, but they had the advantage of being hidden and, moreover, beyond Chinese dominance. The Tibetans do not need to have subjugated all the areas in question, the promise of sufficient profit for the respective rulers might have paved the way as well.26 We know that, at some later time, the Tibetans sought such an arrangement with the ruler of Lesser Bolor (cf. Denwood forthcoming).

Of course, one might think that a *Guran, due to its similarity in sound, could be found in the vicinity of Miran, Nevertheless, an eastern ZanZun could by no means constitute the border area towards the Western Turks, unless one assumes that Shanshan or Miran was Turk territory in the late 7th c. Geographically, the border region between the Western Turks and Tibet was Baltistan and Gilgit, even more so if "Suvarnagotra, Baltistan, Gilgit, and also presumably Ladakh" were not only independent of Tibet, but if "some or all of them may have been part of the Western Turk empire before 630" as Denwood (forthcoming) ponders, without, however, giving any evidence for this line of reasoning.

There seem to have been ethnic or political affinities between ZanZun and Bactrian Turkestan (most probably dating from a time before the rise of the Turk empire), which allowed confounding or merging the two. Thus we find, again in Dpahbo Gtsuglag's compendium, in one of several versions of the Four Border Kingdoms, a concept related with that of the Universal Ruler, the combination of ZanZun and Grugu for the western realm, usually occupied by Persia (Staggzig; cf. R.A. Stein 1959: 258). In another manuscript, dealing with the same concept, the king of Persia is called Muwer (ibid., p. 259), clearly a ZanZun appellation. (27) In the Btsunmo bkahi thanyig, it is even Mjahris Skorgsum that takes the place assigned for Persia (ibid., p. 257). While this confusion may be the result of an overall fusion of themes related to the Universal Ruler, due to scholarly playfulness or neglect, there is a similar confusion and anachronism in LDRR 30.28f.: during the time of Gnamri Slojmtshan, the Tibetans would have 'subdued king Gnazur etc. [of] the western regions, India (i.e. Kashmir) and Turkestan':

Rgyagar dan Grugu nubphyogs | Gnazur rgyalpo-lasogspa btul.

The name of the king is quite obviously related to the element Snasur, appearing in the name of the ZanZun king Lig Snasur, to whom a bride was given in 671 (OTA, 1. 53), and which is also found, a century or so later, as the name of a commander of the ZanZun stod stonsde Hotshopag ~ Hozopag and of the Rtsalmopag regiment (Thomas 1951: 454, 467). At the time of Gnamri Slojmtshan, the Drugu (and Kashmirians) in question could at best have been tribes of Indo-Iranian descent settling in ZanZun.

Given the geographical setting, the border regions of ZanZun stod should have comprised at least present-day Lower Ladakh (with Nubra and Purik). (28) At the very height of their power in the west, when the Tibetans were also in the possession of the Yasin valley and some outposts in the Pamirs, they might have spoken of these areas not only as stodphyogs 'western direction' (cf. OTA 1. 220) but perhaps likewise as parts of ZanZun stod. That something like this happened is indicated by a comparatively late source, the Vth Dalai Lama's biography of Bsodnams Mchogldan Bstanpahi Rgyalmtshan (Tucci 1956: 73), already mentioned.

The Vth Dalai Lama divides Old Mjahris into three skor.s, a term that seems to be an anachronism. (29) Nevertheless, he seems to have had access to much earlier sources, since his description of the three districts (skor) of Mjahris differs from all later sources, particularly with respect to the second skor, containing Khotan (Li), Hunza (Gruza), and Baltistan (Sbalte). Quite interestingly the first and the third skor refer to areas generally accepted to belong to Western Tibet: Purajs, Maryul, Zansdkar (first skor), ZanZun [Proper] and Khrite stod and smad (third skor). A similar division is given in BRGY (sub Mnahris skorgsum): Purang (Spuhren), Majyul, and Zajskar form one skor, ZanZun and Khrite stod and smad another, Khotan (Hchimmam Li), Bruza, and Sbalti the last. No reference is made to regions at the far eastern end of the Changthang. Strangely enough, Guge is not mentioned in either source. It must thus be implied in the designation Maryul~Majyul (see also below [section] 2.2). Ladakh simply does not exist.

One could thus argue that the ZanZun smad of the Old Tibetan sources corresponded to what is taken as the 'true' ZanZun kingdom, while ZanZun stod may have referred to the newly subdued regions further north and west. Whether Upper Ladakh was part of smad or stod may then remain an open question, which everyone will answer according to his or her own preferences. The linguistic evidence, however, discussed in [section] 3 below, seems to indicate that Upper Ladakh did, in fact, belong to ZanZun smad.

With ZanZun stod consisting at least of (Lower) Ladakh, it might be possible to identify Gnema (and the Nimobag district of Ms M. Tagh. c, iii 0019, Thomas 1951: 293) with Snemo in Lower Ladakh. One of the thousand districts of ZanZun stod or smad must have been based in Zanskar, but unfortunately the principality kept the administrative term and not its original name, so that we hear of a kingdom (today only a village) of Oonde (Stonsde).

It should be noted that the districts nr. 1 and 3 of ZanZun stod are attested in some Old Tibetan documents with the additional Iranian or Turkish element -pag (cf. Thomas 1951: 293: Nimobag; pp. 454, 460f.: Hotshopag ~ Hozopag; a further name is Rtsalmopag, p. 467), which is most probably related to the designation Baga for the stonbu, and with the location Beg, whereto the emperor went in 739 for a marriage arrangement with the previously subdued king of Bruza (OTA, ll. 276f., 281, 283f.). Beg must then have been either a place in Bruza or a place in the immediate neighbourhood (it could well have been identical with the stonbu Baga). This is another indication that ZanZun stod may have reached up to Hunza. As far as I know, the name element -bag or -beg is nowhere attested in West or Western Tibet. The only exception might be Mulbekh, but the spelling indicates an original -begs, and we do not know when the village was founded or when it acquired its present name.

On the evidence of a further -phag name element in the district of Gnamruphag in the Central Horn (Dburu) Denwood (2008: 10) assumes that Nimobag and Hotshopag ~ Hozopag were somehow linked with Gnamruphag, which would make it possible to identify Gnema or Nimo with "the present county headquarters of Nyima directly westwards of Namru" and Hotsho with "'O ma mtsho (Oma; Das's map: Huma Cho) further west again". While more or less communis opinio, this identification does not take into account the notion of stod 'west'. A region near the Gnammtsho would be clearly east of the eastern-most parts of ZanZun smad (defined as centring on Guge). This localisation would further imply that Turks were already settling in the present-day Hor district of Nakchu. But then Gnamru could have equally been defined as bordering to the Turks. And surely, the 'Turks' of Nakchu would have been part of Tibet. Should one thus assume that Turks settled further west throughout the Changthang? This is actually what Denwood (forthcoming in the present version) suggests, without, however giving any historical evidence for this claim. But if this would have been so, where should we locate the Sumpa in order to have ZanZun smad/Guge lying between them and Tibet? Is the latter indication simply an error as Denwood (2008: 12) suggests? By what evidence do we judge which of the statements is correct and which one not? By the contradictions that we have built up ourselves?

Dotson (2009: 87, n. 149) draws the attention to the fact that chief minister Mgar Stojrtsan performed a registration in Dugul of ZanZun just after and just before staying in the Haza country. One may easily conclude that this unknown place was not located in the west. This might indicate that like ZanZun stod, ZanZun smad could have been administratively enlarged far beyond the original boundaries so that it covered the whole northern Changthang up to the eastern-most end of the Kunlun range. This administrative enlargement could perhaps explain the localisation of Greater Yangtong by the Chinese envoy Liu Yuanding (see below). Nevertheless, this assumption, as much as the suggestion that ZanZun extended up to the Namtsho, faces the problem that the Namtsho and Gnamruphag belonged to the Central Horn. The territories to its west should thus have belonged to the Right Horn. Otherwise one would have to conclude that either the Right Horn was absolutely insignificant, occupying only a small territory along the Rtsajspo, or that the Right Horn overlapped with parts of ZanZun smad. Similarly, there should have been an overlap with ZanZun smad and the Sumpa Ru, if the former extended up to the Kunlun. If all this territory was, according to Bonpa claims, originally part of an old state of Zanzun, or perhaps merely of ZanZun-ian ethnicity, the Tibetan administration does not seem to have treated it as such, and while even Buddhist sources acknowledge the westward extension of Zanzun. they remain silent about a similar extension to the east.

Many of the above mentioned contradictions could be avoided if we allow ZanZun (stod) to extend up to the Pamirs, or to Gilgit at least. Denwood's idea of an ancient trade route through what he calls the 'Changthang Corridor' (2008: 10 and 18, map 1) does not depend on the localisation of ZanZun stod in the east, it would equally have its value if we allow ZanZun smad to extend into the said area.

I do not want to rule out the possibility that the Bonpo reminiscence of a ZanZun on Iranian territory or at least comprising Gilgit may be based on the 'official' notion of ZanZun at the height of Tibet's western-most extension. But it is likewise possible, that the designation of Tibet's western extension was actually based on an old, but perhaps more cultural than political, notion of (Western) Zanzun. inherited from the Hephthalite-Huna period (cf. Zeisler forthcoming a, Chapter 4 [section] 1.4.3).

The notions of high and low are likewise attested in Bonpo texts, referring again to western and eastern regions respectively, albeit on a much smaller scale. Mtho ZanZun is used for the castle of Khyujluj and the region Khayug to its south. The expression Dmah, by contrast, is apparently used for a Sumpa region or an area between the Sumpa and ZanZun Proper (Bellezza 2008: 284, 593). This might well refer to (parts of) the central Changthang, which borders on the Sumpa as well as on Tibet, but does not really lie between them, except if we admit that Sumpa tribes were also roaming on the northern fringes of the Changthang. (30)

Post-phyidar Bonpo sources also know of a ZanZun smar, and at least one text locates this in the area of Mt Tise and lake Maphaj (Bellezza 2008: 319). Given this location and given a ZanZun etymology, it must be a 'Golden ZanZun' not a 'Lower ZanZun' as Hoffmann (2003 [1990]: 48) suggests (cf. also Hummel 1974: 494). This epithet would well correspond to late Buddhist depictions of the Kailash area as the golden fundaments of earth and its identification with the axis mundis Sumeru (see below).


1.3 Yangtong

Chinese authors discriminate between Greater and Lesser Yangtong, but the localisation of each part seems to be, again, rather variable, except that Lesser Yangtong is always located west of Greater Yangtong, when both entities are mentioned. According to the conventions mentioned in the introduction, these designations should have reflected the spatial relation to China, obviously calculated from her western-most extension in Turkistan (or adopting the viewpoint of its Iranian neighbours, centring on Mt Meru as the axis mundis) or it might have reflected the ZanZunian perspective of a Centre in the west (Innermost ZanZun).

The east-west correlation is somewhat at variance with the identifications proposed by Japanese scholars as reported by Petech, who furthermore, maintains that an "excessive extension eastwards is not acceptable" (1998: 230). Basing himself on a 1981 article by H. Sato (and possibly also on the work of Yamaguchi; cf. the discussion just above), Petech (ibid.) states
      Greater Yang-t'ung corresponded to Upper Zhang-zhung,
   i.e. Guge and Purang, while the name Lesser Yang-t'ung
   indicated originally the upper valley of the Tsangpo from the
   Mar-yum pass to Lha-rtse. Later it was applied to the northwestern
   region, which after the "horns" (ru) reorganisation
   of the second half of the 8th century came to be known as
   Lower, i.e. Eastern Zhang-zhung.

It remains unclear which north-western region Petech refers to, if it should be part of an eastern ZanZun. Lesser Yangtong, albeit described in most Chinese sources as the western part, is here associated with ZanZun smad/dmah, by definition as much as by the localisation proposed, the eastern part. I am unable to imagine the reason for this inversion, except perhaps, that it is based on a itinerary, allegedly from north-eastern Tibet to Nepal, preserved in the Shijia fangzhi, which was completed in 650 (Pelliot 1963: 709f.). Denwood (2008: 12) would think that the localisation of Little Yangtong southwest of some part of Central Tibet and northeast of Nepal [...] is geographically just about possible if the Chinese Little Yangtong is the same as Tibetan Lower Zhangzhung.

But this assumption would contravene the conventions associated with the terminology of Lesser and Greater. Lesser Yangtong would no longer be closer to China (or Mt Meru) than Greater Yangtong, while Guge, commonly accepted as the core area of ZanZun would no longer be in the centre, but in the periphery of Greater Yangtong. Similarly the convention concerning the terminology of stod and smad would be broken, which would no longer relate to an east-west axis, but to the north and the south, respectively. Both ZanZun provinces would extend far into the east, and would overlap with Central Tibet. As a result, the western and north-western regions, present-day Ladakh and Baltistan would have been discrete entities, although they never show up as such in the Old Tibetan documents.

For me, at least, the itinerary does not constitute a reliable source, as it seems to have been mixed up with at least a second one, containing a description of the Hanging Passages near Chilas. I am not aware of any such gorges in Central Tibet, and it seems impossible to make any sense of this itinerary as the original road map, regardless of how one wants to define the ZanZun/Yangtong entities.

The itinerary, according to Pelliot (1963: 709-710) starts in Hezhou, and until reaching the Kokonor directions as well as distances are given. From that point onwards only the direction, mainly south-west is given, which is, of course, the general direction from Amdo to Nepal. Given this general route, one wonders how and where Yangtong could ever get into the way. The itinerary passes the frontier of the Tuyuhuj. reaches the kingdom of the Tomi, the kingdom of the Supi, the unidentifiable kingdom of Kan, then, with a slight turn towards the south-east, the kingdom of the Tufan.

Then to the south-west on reaches the kingdom of 'Lesser Yang-t'ung'. Then to the southwest, on crosses the Ta [?Chu]-ts'ang-ch'u Barrier, which is the southern frontier of the T'u-fan. Then to the east, slightly south, one crosses the Mo-shang-chia-san pi Barrier, to the south-east enters gorges (ku), crosses thirteen 'flying ladders' fei-ti) and nineteen 'plank-roads' (chan-tao, i. e. roads made of boards fixed more or less high on the wall of a vertical cliff), either south-east or south-west, snatching the creepers and grasping the lianas; after marching in the wilderness for more than forty days, one reaches the kingdom of Ni-po-lo (Nepal) of northern India. (Pelliot 1963: 710; all Chinese graphs as well as phonetic reconstructions have been omitted).

Southwest of Yarlujs or Lhasa lies Sikkim, from where one could reach Nepal or Darjeeling; the route would pass Yarhbrog G.yumtsho and Gyantse (Rgyalrtse). If a more western route along the Gnammtsho were followed through Tibet, one would have reached the Rtsajspo at Shigatse in Myaj. From there a south-western route would have ended directly in Kathmandu. Most likely, however, one would have followed the Rtsajspo further up westwards to Gujthaj or Majyul from where the route over Kyirong leads to Kathmandu in a roughly southern direction. This route was used by the Chinese embassies to Nepal in the mid 7th c. (cf. Sen Tansen 2002), that is, at the same time as this itinerary was documented. On the Kyirong route, and possibly all other routes, one would have to cross rope bridges, difficult enough to manage, but, as far as I know, neither flying ladders nor planks in the cliffs. The infamous Hanging Passages as known from the early 5th c. CE traveller Faxian have been variously located somewhere near Gilgit, sometimes in Hunza, but according to M.A. Stein (1942) and Jettmar (2002b [1987]) the passage that fits best Faxian's description is located at the foot of a spur of the Nanga Parbat between Sazin (near Shatial) and Jalkot.

[T]he party journeyed on in a south-westerly direction for fifteen days over a difficult, precipitous and dangerous road, the side of the mountain being like a stone wall ten thousand feet in height. On nearing the edge, the eye becomes confused; and wishing to advance, the foot finds no resting place. Below there is a river, named Indus. The men of former times had cut away the rock to make a way down, and had placed ladders on the side of the rock. There are seven hundred rock steps in all; and when these and the ladders have been negotiated, the river is crossed by a suspension bridge of ropes. [.] Having crossed the river, the pilgrims arrived in the country of Udyana (Swat) which lies due north of India (Faxian, as rendered by Giles and cited by M.A. Stein 1942: 54).

[W]e go up the course of the Sindu river; and then by the help of flying bridges and footways made of wood across the chasms and precipices, after going 500 li or so we arrive at the country Boluluo [Bolor] (Xuanzang, Beal 1881-84: 178).

On all the eleven trying marches [...] there was daily a constant succession of tiring ascents to be made. The track climbs up steeply ridge after ridge, each rising sometimes as much as 1000 feet or more above the river, in order to avoid impassable cliffs. From the heights thus gained there were invariably descents, often quite as tiring, to be made again towards the river. Nowhere was it possible to keep for any distance near to the river bank since masses of huge boulders line it wherever the river does not actually wash the foot of impassable rock walls. I have not counted all the climbs, but they must have been still more numerous before the recent track was constructed M.A. Stein 1942: 55).

[T]he most dangerous part [...] was practicable only because tree branches had been fixed in fissures on the rock supporting galleries, steps had been carved out, in many places there were logs with notches to be used as ladders (Jettmar 2002b [1987]: 179).

It may well be possible that, at the eve of its destruction, the dominion of ZanZun (smad), as the main competitor of the 'Tibetans' in Central Tibet, extended over the whole area of Rtsaj and Myaj. But it remains quite astonishing that the straight westward route section along the Rtsajspo is omitted in the itinerary. It is furthermore difficult to understand how one could reach the southern border of Tibet, after having crossed a territory to its southwest in a south-western direction. Apart from this, it is hardly believable that one could err for more than forty days in the wilderness between Lesser Yangtong and Nepal, that is, some 800 to 1200 km, counting 20 to 30 km a day, when the distance between Gujthaj and Kyirong as well as that between Kyirong and Kathmandu is about 100 km each, as the crow flies, thus at most 300 km or 10 to 15 marches altogether (cf. the map in R.A. Stein 1981: 14-15). (31)

Pelliot (ibid.) is likewise unable to make sense of this description, although for different reasons. Given the localisation of the sources of the Yellow River in Greater Yangtong in the Xin Tangshu (completed 1060), he is inclined to put back Lesser Yangtong "to an earlier stage, before T'u-fan, at least, and perhaps before the kingdom of Kan, if not even before those of To-mi and Su-p'i", but this is just mere guessing.

In the Tongdian (completed 801), Greater Yangtong is defined as bordering in the east to Tibet, in the west on Lesser Yangtong, extending in the north towards Khotan, it is more than 1000 li, thus roughly 350 km extended from east to west (Pelliot 1963: 708)-this corresponds roughly to the localisation of Suvarnagotra, Nuguo, and, ultimately, relatively well to present-day Upper Ladakh (plus some parts of the Changthang). However, the name of its king, Jiangge, which Pelliot reconstructs as *Kyajkar > *Kiang-kat, does not resemble any name of the ZanZun kings as preserved in Tibetan sources.

In the Taiping huanyu ji (a geographical work, completed 983), Greater Yangtong is identified with a kingdom that was conquered by the Tibetans in 649 (Pelliot, ibid.), roughly corresponding with the elimination of the ZanZun king Lig Myirhya as attested in OTA, but this event is described as having lead to severe destruction and a redistribution of the apparently nomadic people, a fact which-perhaps for ideological reasons-does not seem to be corroborated by Tibetan sources.

The fall of Lig Myirhya, however, is described in OTA, year 644 (l. 13) quite drastically with the verb brlag 'was destroyed'. One year after the conquest of Globo and Rtsajrhya (possibly former vassals of ZanZun) in 652, there is mention of a great sale of fields (zingi tshonchen) in cooperation with minister Ridstagrhya of [the] Rasajrje [clan] and of the installation of a new fiscal governor for ZanZun (ll. 21f., 24f.). Since the Rasaj(s)rje are known as a ministerial clan of ZanZun (cf. the Catalogue of the Ancient Principalities, Pt 1286, l. 7), the great 'sale' of fields should have concerned territories of former Zanzun. a fact corroborated by the subsequent statement concerning the fiscal governor. Perhaps we have to understand the word 'sale' as a euphemism for 'redistribution'. It is also possible that what looks as a 'sale' from the perspective of the winners and the agents involved, namely the Tibetan and the collaborating ZanZun aristocracy, turns out to be a forceful disappropriation and thus 'redistribution' from the perspective of the losing ZanZun aristocracy and particularly their bondsmen. Their subsequent embassies to the Chinese court would certainly have described the events in the most accusatory tone.

Certain authors shifted Yangtong further to the east: a Chinese envoy, Liu Yuanding, visited the sources of the Yellow River in 822 and located them in Greater Yangtong (Pelliot, ibid. p. 710). More precisely, his description reveals that 300 li south of the sources lie three mountains, resembling copper coins with flat bottoms, said to be (part of) the Kunlun. It is these mountains that border on Greater Yangtong (Herrmann 1910: 60). We do not know the reason for locating Greater Yangtong so far east, but it seems that in this case, it referred more generally to the eastern Changthang and its various tribes. It may also be noted that the expedition took place comparatively late, and its report is found in the Xin Tangshu of even later date, whereas all other descriptions, although likewise from later sources, may have referred to a time shortly before the pre-imperial period.

Whatever the reason for the apparent mismatches, the 'historical' kingdom of Greater Yangtong/Zanzun. the one subdued by the Tibetans around 645 (or 649 according to the Chinese sources), cannot have extended throughout all of the Changthang, as (most of) this region was at the same time occupied by the Sumpa, testified as independent kingdom in Tibetan and Chinese sources. This does not preclude the possibility that an ethnic, linguistic, or cultural notion of Yangtong-ness might have overlapped with the Sumpa territories or that at some point in prehistory an unknown political entity encompassed both ZanZun and Sumpa territories. Bellezza's (2008) documentation of the cultural remains of what could perhaps have been the precursor of historical ZanZun points to Western Tibet as the core area and a possible extension of the same cultural features into the central Changthang.

The partial contradiction in the sources could indicate that the geographical or tribal entity in question was quite extended or scattered, while the political entities associated with the name were much smaller and more numerous than the division into two parts suggests. Such a scenario could perhaps explain the 'wrong' name of the king, the 'unattested' redistribution, and more generally, the fact that various dates are given for a 'complete conquest', while still envoys from Yangtong are registered at the Chinese court (cf. Denwood 2008: 9). (32)

(Greater) Yangtong is said to have been without script, so that, as in early Tibet, only tallies and cords with knots were used. The latter statement corresponds well with the fact that, despite respective Bonpo claims,33 no evidence of any kind of pre-Tibetan script has come into light in the area of ZanZun proper (Bellezza 2008: 187).34 The burial customs of Yangtong follow a Central Asian type:

When chiefs die, [their skulls are cleft] and the brain scooped out; [then] the space is filled with pearls and jade; the five viscera are [taken out] by cutting open [the abdomen] and replaced with gold; a false gold nose and [false] silver teeth [are put on]; men follow [the chiefs] in death (hsun). Having fixed by divination a propitious day, they bury the body in the cave of some cliff, in such a way that no other man should know about it, and slaughter many cows, ewes and mares (ibid., p. 708).

A further detail, given by Hyecho, but so far apparently overlooked, is that grapes were grown Kashmir, the two Bolors, and Yangtong, but not in Gandhara and the countries of India (Fuchs 1938: 445). Grapes grow or have grown in a semi-wild manner in Lower Ladakh up to Khalatse,35 and the production of wine or rgunchan 'grape beer' is known particularly from the Dardic areas further down the Indus. I have not heard about grapes being grown in Guge (36) or the upper valley of the Brahmaputra, but even if grapes grow or have been growing in these regions, it seems more likely that Hyecho's grape-growing Yangtong, as neighbouring the two Bolors comprised Lower Ladakh and perhaps (parts of) Baltistan, particularly since the Guge area was known to him as Suvarnagotra.


1.4 The 'Women's Dominion'

The earliest rumours about Amazons in the Himalayas (or Pamirs) seem to have been brought to Europe by Alexander's troops, who may have heard the stories and legends from Indian, perhaps also Iranian middlemen. Their accounts, mostly trivialised or lost in the subsequent literary processing, the various versions of the Alexander Romance, transmitted the Indian notions of an Amazon utopia (or paradise), forbidden to the all-too-worldly Alexander (Pseudocallisthenes, chap. 25-26, R.A. Stein 1959: 277). The Indian utopia was Uttarakuru, located unspecifically somewhere in the north of India, but according to the sources Ptolemy used, some traditions must have identified it with the Kunlun. the Altyn Tagh, and the Nanshan (Richthofengebirge).

Like the Hyperboreans of the Greek, the inhabitants of Uttarakuru were described (in Greek sources) as inhabiting an ideal country and together with other Northlanders, such as the Seres (the silk traders of the Tarim Basin), they seem to have been further idealised with respect to their society (cf. Lindegger 1993: 59f.). Such idealisation typically refers to an extraordinary quality of innocence and justness, but also, like other alternative conceptualisations of a better society, to libertinism and, more particularly, inverse social order: women free from the patriarchal yoke, women living without men, women even as rulers (cf. also Albinia 2008: 262).

Men's horror scenario of uncontrolled women went the by the name Strlrdjya in India, which corresponds exactly to the Chinese term Niiguo [TEXT NOT REPRODUCIBLE IN ASCII], W.-G. Nii-kuo). The common English translation as 'Kingdom of Women' is quite odd, since it is not intended to signal that a king ruled over an exclusively female population, and a female ruler would not be king, but queen. While the general parlance certainly permits that individual queens are rulers in a kingdom (which is generally ruled by kings), it is not just patriarchic world-view, but completely illogical to call an institution kingdom when there are no kings at all or when male rulers appear only exceptionally.

In accordance with the usual German translation: Frauenreich, I will use here the term Women's Dominion.

A Strlrdjya is mentioned unspecifically together with the wrongly assigned Kulkuta and with the Tala and Tukhara in the north-western division of the Brhatsamhitd (Fleet 1973: 10f). Tucci (1956: 101) cites some additional Sanskrit sources, which likewise point to a north-western direction: located between the Huna and Tangana, the country would be also mentioned in the Mahdbhdrata (III,51,1991 and XII,4,14) in connection with Kulkuta and Tukhara and in the Kdsyapa together with the Kosala, Tangana, Hala, and Bharukacchas. Since most of the associated peoples can be located in present-day Afghanistan, this should also be the main area, where one should look for the Strlrdjya. But one cannot preclude the possibility that a similar population was also found in Baltistan, Ladakh, the western Changthang, and perhaps even in the Guge region. There must have been more queendoms in the Indian or Iranian borderlands. Xuanzang mentions a Sutulisifalo (perhaps Strl-lsvara, Women's Paradise) as the capital of a western province in Sindh (Watters 1904/1905, II: 257), while other Chinese sources point to a, possibly only mythical, Women's Dominion west of the Pamirs (Pelliot 1963: 679).

Like the Greek notion of the Amazons, the Indian descriptions of the Strlrdjya can only be classified as phantasmagoria, serving the typical male erotic projections, as in the case of the submission of its queen to Lalitaditya-Muktapida (Kalhana's Rdjataranginl iv, 173f., M.A. Stein 1900 I: 138). This seems to be a rather crude echo of Cleophis, the Massagete queen's surrender to Alexander (for which cf. Tucci 1977: 49, 51). The fact that Lalitaditya-uktapida is said to have erected a Siva statue in this country (Rdjataranginl iv, 185, M.A. Stein 1900 I: 139), might perhaps point to the ddkinl land in Uddiyana/Swat, which continued to attract a religiously veiled erotic tourism throughout the centuries (cf. Tucci 1977: 69f.).

The Nuguo of the Chinese sources seems to have inherited the utopian notion of libertinism and a possible association with the Uttarakuru-Kunlun. but is otherwise, in striking contrast to the Indian myths, described in quite realistic terms. More particularly, since envoys and certain products are said to have arrived at the Chinese court, we are dealing with a country in real terms, wherever its location and whatever its social structures might have been.

By mischance, the Chinese historians confounded two such queendoms, one in Eastern Tibet and one in Western Tibet or at the confines of the latter. Even more regrettably, many Sinologists (e.g. Laufer 1918) and other scholars (e.g. Denwood 2008) have rejected outright the idea of a Western Tibetan or Bolorian queendom as due to a confusion with a purely fictional and utopian Western Dominion of Women in the Far West, that is, in the Roman Orient (for a legendary association with 'Rome' or what the Chinese thought Rome to be, cf. also R.A. Stein 1959: 277). This position is indefensible, since the Chinese sources repeatedly give quite exact coordinates, which lead us to Western Tibet or Bolor. This cannot be merely due to a confusion, or if so, there must be some specific reason for such confusion, lying in the reality on ground.

According to the Suishu (completed 636) and Beishi (completed 659), the Nuguo is located south of the Congling mountains (the Pamirs), (37) according to the Suishu it lies 3000 li south of Khotan (this is the same distance as for Suvarnagotra). (38) It exports salt to India and frequently fights with India as well as with the Dangxiang ([TEXT NOT REPRODUCIBLE IN ASCII], W.-G. Tang-hsiang; Eberhard 1942: 90, Pelliot 1963: 694f.), one of the later Tangut tribes, mainly located in Khams. While the location in relation to Khotan and the fights with India points to the 'Upper Indus' region in Pakistan and the adjacent eastern areas of Baltistan and Ladakh, the interaction with the Dangxiang points to Eastern Tibet. It is thus not surprising that we hear of a second (?) Eastern Dominion of Women, in Eastern Tibet (ibid. p. 82). This is located southwest of Chengdu and west of the Dangxiang. The country extends 9 days from east to west, 20 days from north to south (cf. Pelliot ibid., p. 699f.). It should thus be located at one of the great southwards bound rivers. According to Denwood (2008: 9)
      [i]t seems to have extended from the Gyarong area across
   Nyarong (Y along valley) towards the Yangtse, southwest of
   the kingdom of Fuguo and southeast of Sumpa.

The notion of an 'Eastern' Women's Dominion, Dongnuguo [TEXT NOT REPRODUCIBLE IN ASCII], W.-G. Tung-nu-kuo), is somewhat misleading, as this term was coined largely to distinguish this land from the legendary Western Women's Dominion in the Far West. The passage in the Jiu Tangshu (completed 945) would, in fact, give the impression that there is only one Eastern Women's Dominion, but on closer inspection, one can see that two distinct geographical entities have been entirely mixed up. (39) A region north- and southwest of Chengdu (cf. Pelliot 1963: 699f.), and extended over clearly less than 300 km from east to west can hardly be identical with an area south of the Pamirs or south of Khotan. But it is not the case that the Jiu Tangshu locates this queendom only in the east as Denwood (2008: 9), omitting the conflicting statements, insinuates:

The Eastern Kingdom of Women (Tung Nu kuo), also called Su-fa-la-na chu-chu-lo [=Suvarnagotra], is a division of the Ch'iang. There is also in the far west (Hsi hai) a country ruled by women, so this is called the eastern one. To the east it borders on the T'u-fan, the T'ang-hsiang [=Tanguts] and Mao chou [near the Min River, NNW of Chengdu]. To the west it touches the San-po-ho (Yaru tsangpo). To the north it is conterminous with Yu-tien [=Khotan], and to the southeast with Ya chou (in Ssu-ch'uan) [SW of Chengdu], the Lo-nu Man-tzu and the Pai-lang ['White Wolves'] savages. From east to west it is a nine days' journey, from north to south twenty days'. It has eighty towns, and is ruled over by a woman who resides in the K'ang-yen valley, a narrow precipitous gorge around which flows the Jo River in a southerly direction (Rockhill 1891: 340; identifications in square brackets added, cf. also Pelliot 1963: 699f. and 703).

Unfortunately, Rockhill does not give the Chinese characters for the place names, but as indicated by Pelliot (1963: 707), he must have been mistaken. The name Sanbohe does not refer to a river (see also note 75 below). Even if it had done so, this could hardly have been the upper Rtsarjspo, which runs dead straight from west to east, so that any particular country could only border on it either to the north or to the south. The only point where a region could border to the eastern shore of the Brahmaputra, is on its middle course, where it bends around the Himalayas. This is the area of Spobo or Poyul, and the eastern Dongnuguo would not have been all too far.

Rockhill apparently also overlooks the fact that the country cannot at the same time border to the east on the Tibetans and to the west on the Rtsarspo, assumed to lie within the borders of Tibet. It is clear that the borders of the more western Dongnuguo (including her border with Tibet in the east) are mixed up with those of the more eastern Dongnuguo (bordering on, e.g., Tangut and Bailang, another (south)-eastern tribe, classified either as Tangut or, linguistically, as Yi).

The 'Jo River' (Ruoshui) of the above statement is the notoriously peripatetic 'Weak River'. Although a river bearing this name is located in Gansu, "north of the desert above Shensi", that is, at the middle course of the Yellow River (E.J. Hill 2004, n. 10.2, with further references), the mythological connotations of the 'Weak River' take us further west. According to some traditions, the river has its source in the Kunlun (which seems to be identical with Mt Meru and the Pamirs on a mythological level) and is associated with the paradise of the mythical Queen Mother of the West, possessor of the elixir of immortality.

South of the western lake, by the shores of the flowing sands, behind the Red River and before the Black River there is a great mountain called 'The heights of K'un-lun'. [...] Below there are the depths of the Jo River which encircles the spot. [...] There is a person who [...] dwells in a cave and is named 'Queen Mother of the West' (Loewe 2005 [1982]: 32f. with further references).

This mythical mountain goddess might not only be connected with the peris of the Pamirs, Mt Meru and Kailash (the competing axis mundis), (40) but also with the Women's Dominion, since the Ruoshui is also thought to flow in or around the latter:

The [TEXT NOT REPRODUCIBLE IN ASCII] Jo-shui ("Weak River") has two springs, which flow north of the Kingdom of Women and south of the A-nou-ta Mountain (Anavatapta), (41) and have their confluence in the Kingdom of Women (Pelliot 1963: 696 with further references).

Pelliot (1963: 696-698) has discussed this legendary river (42) at length without coming to any conclusion, understandably, since, like in the description of the Tongnuguo, several regions have been conflated. The association with lake Anavatapta should lead us into the Pamirs, with Mt Meru (or Sumeru) or at least to the present-day Kailash as the source of the four great rivers, flowing in the cardinal directions (see also Zeisler forthcoming a, Chapter 2 [section] 3.3.4), arguably the 'western-most end' of the Kunlun. (43) One could perhaps think of the Gilgit-Indus and the 'real' Indus and their continuation after their confluence. The 'narrow precipitous gorge around which flows the Jo River in a southerly direction' may well correspond to the frightening gorges in Chilas. In fact, in the biography of Gao Xianshi, included in the Jiu Tangsu, the Weak River is explicitly identified with the Suoyi, i.e. Gilgit River (cf. Chavannes 1903: 153, n. 1). (44)

Nevertheless, the flowing sands might lead us into the Tarim basin west of Qarqan/Cherchen (cf. Hill 2009: 302f, n. 12.21; in which case the women's dominion should be located at Qarqan), while the five rivers of the mythical Kunlun (or at least some of them) lead us back to the real Kunlun and to Eastern Tibet and Gansu: the Red River (either the Chudmar, one of the Tibetan sources of the Yangzi Jiang (Yangtze), or the Yunnan and Vietnamese Red River), the Black River (Etsingol, Chunag), the Yellow River (Rmachu), (45) the said Ruo River, additionally also the Xiang River, a southern tributary of the Yangzi Jiang. Even from the mythological perspective, it does not seem possible to play off one of the two queendoms against the other and declare one of them as unreal or mere fiction.

The main corroboration for the existence of a Women's Dominion in the west comes from recorded history. In 445 CE, the Tuyuhun (H aza) raided Khotan and some regions down the Indus, where they entered into a political alliance with the Hephthalites or Huna of Gandhara. In the same year, they offered to the Chinese court, among other precious things, golden wine vases of the Nuguo (Mole 1970: 10). This is certainly not just a coincidence. Most probably, the wine vases were booty, not items of fair trade. From this note we learn two things: first of all, the Nuguo were producing and consuming wine, and secondly, rather indirectly, the Nuguo must have been located somewhere on the route of this expedition, and it is thus rather unlikely that the eastern Dongnuguo was meant. Both features link these Nuguo with the Dards or culturally like populations of the Pamirs and the Hindukush. As mentioned above, wine cultivation was observed in Bolor and Lesser Yangtong.

The second historical fact, the appearance of an envoy of the Women's Dominion at the Chinese court, cannot be correlated with either the eastern or the western country. The adversaries of a Western Tibetan or Bolorian Nuguo seem to conclude that such an envoy could have only come from the eastern region. But there is no reason to assume that an envoy could not have reached the Chinese court via Khotan or even via Ladakh and the Haza.

A decision for Eastern Tibet is thus rather arbitrary. Given all the contradictions and the lack of material witnesses, one could similarly come to the conclusion that there was only one Nuguo, but located at the borders of Western Tibet. At an unknown time and for unknown reasons, the geographical template of the lands to the north-west of India was transferred to the northeast, Bengal, Assam, and even Yunnan. If, e.g., Parthia (Nangxi), Gandhara, and Camp could be relocated in Yunnan, the (Upper) Sutlej region under the names of (Mahd)-Clna and Suvarnabhumi in Assam, and the Yavanas (Bactrians or Indo-Greeks) in Laos and Vietnam (cf. R.A. Stein 1959: 308, n. 77), then one should not be all too surprised to find, further to the north, a second Strlrdjya, as well. That concepts relating to Mt Meru became connected with the eastern Kunlun might be a result of this general geographic transfer. To be not misunderstood: unlike my predecessors I do not want to categorically rule out the existence of a second Nuguo. But if there had been only a single Nuguo, and if I were forced to decide which one was the real one, I would opt for the western one.

The description of both queendoms is more or less the same, in particular, both are associated with the elite's burial custom of removing the skin and keeping the bones, mixed with gold, in painted vases (Eberhard, ibid., pp. 82, 90, Pelliot, ibid. p. 695), a custom which seems to have parallels in Swat and, indeed, in Ladakh (Francke 1914: 64f., 71-74, Tucci 1973: 52f.). But the fact that up to ten followers were sacrificed points to Scythian or Central Asian traditions. (46) The queendoms are further associated with nine-storey stone houses (Eberhard, ibid., pp. 82, 90), a feature that seems to be typical for the Minag of Eastern Tibet. (47) Surprisingly, the more eastern Dongnuguo is said to use an Indic script (ibid., p. 82), which cannot be but the Karosthi or Brahmi, (48) and it is likewise said to celebrate New Year around the winter solstice (Pelliot ibid. p. 700, see also below).

The Suishu supplies the additional information that the people (of the more eastern Dongnuguo) venerate the asura and tree-gods (Pelliot, ibid., p. 695). The latter statement points to a predominantly Iranian or Iranianised population, which appears to be rather incompatible with the eastern region, but could correspond to the mentioning of Jatdsuras 'Asura with matted or twisted hair' in the north-eastern division of the Brhatsamhitd (Fleet 1973: 12). The hair style in question, matted, twisted, or braided tresses, could well refer to a particular Iranian hair style as found on many coins, from the Arsakides to the Sassanians. (49) A braided coiffure is mentioned also for the inhabitants of Greater Yangtong (Pelliot 1963: 708), while it is said about the Nuguo that either both men and women or only the men let their hair hang down, the women plaiting it and coiling it up (ibid. p. 695). But as already mentioned, the location of the Jatdsuras in the north-eastern section cannot be taken for granted, and it generally seems that all relevant references in the Brhatsamhitd end up on the 'wrong' side of the Indus in either Pakistan or Afghanistan.

It is thus an open question whether two different ethnic groups (IndoIranian or Burushaski in the west, Tibeto-Burman in the east) have been simply confounded because of their outstanding political features (or because of some mythological transfer) or whether the two originally formed a political or even ethnic unit, extending through the Changthang, and were disrupted only at a later time.

There should be no question that there was a Dongnuguo more to the west. But the association of the Nuguo with Suvarnagotra and even more so the identification of the latter with Greater Yangtong is more than questionable (see next section). If there was any relation at all between Nuguo, Suvarnagotra, and Yangtong, it should have been with Suvarnagotra III, and thus with Lesser Bolor, perhaps also with Lesser Yangtong. Francke's (1912: 269) idea of a Women's Dominion in Guge, followed also by Tucci (1956: 105, 1971a: 550) has possibly no other base than the non-discrimination between the three areas that have been associated, rightly or wrongly, with the designation Suvarnagotra.


1.5 The land of the 'Gold Clan'

Like the designation Women S Dominion, which referred to two distinct regions, the designation Gold Clan, in its various Indian, Khotanese, Greek, Tibetan, and Chinese versions, refers to at least three distinct regions, but the references tend to get mixed up. It will be useful to treat these regions separately.

1.5.1 Suvarngotra I (the Indian perspective)

The name Suvarnagotra is already attested by Ptolemy in the 2nd c. CE in the form of Eoravvayofjpa (Suannagura), located at 145[degrees]30720[degrees]30' (Lindegger 1993: 73, 164). According to Lindegger's map I, based on Ptolemy's coordinates (p. 223), the region might be roughly identified with the upper Sutlej area or the 7th to 8th c. Suvarnagotra of the Chinese pilgrims (see below).

While R.A. Stein (1981: 14-15, on his map "L'habitat et les habitants") identifies this land with present-day Ladakh or its central area around Leh, Denwood (2008: 9) locates it in the Rudok area, alternatively also in Nubra, whereas according to Thomas (1935: 153) it would be identical with HunzaNagar. Beckwith (1977: 174, n. 208, basing himself on Herrmann 1938) shifts the country to "the Kargil area, between Skardo and Leh". (50)

The Brhatsamhitd mentions a 'Goldland' (Suvarnabhu) in the north-eastern [!] division, together with Brahmapura, Palola, Kulkuta, and, among others, the Kasmiras and Darada, cf. Fleet (1973: 12). (51) The name of the country may either refer to the idealised (and fictive) area around the Kailash-Sumeru (see [section] 2.2 below) or may (additionally) refer to the same entity that was known under the somewhat different name Suvarnagotra. (52)

This country, according to Hyecho, is reached by a route from Jalamdhara across the Himalaya, turning towards the east (Fuchs 1938: 439), a direction that would point to Spiti or Guge. Similarly, according to Xuanzang (or an interpolated source), the country lies somewhat north of Brahmapura53 in the middle of the mountains (ed. Beal 1881-84: 227), which would likewise lead us to the realms of Spiti or Guge. But the country is further specified by Xuanzang as bordering in the north to Khotan, in the east to Tibet, and in the west to Sanbohe alias Moluosuo (see below, 2.1).

1.5.2 Suvarnagotra II (the Chinese perspective)

A quite different location is 3000 li (ca. 970 km) or merely 600 li (ca. 195 km) south of Karghalik (Thomas 1935: 152). Karghalik lies on the western rim of the Tarim basin. A crow flying almost straight southward from Karghalik could eventually reach Leh, after about 450 km (almost 4[degrees] of latitude) or very roughly 1500 li. After about the same distance, a little more to the east, our crow would reach the places in Northern India where both pilgrims obtained their information.

The routes between Khotan and Ladakh are far from being easy to negotiate. The straightest route would possibly lead over the Saser Pass into the valley of the Nubra River and then over the Khardong Pass above Leh, other routes lead further east via the Changla and Sakti or even further east via the Pangong range. Forsyth (1875, I: 249) gives the distance from Leh to Khotan as between 415 and 637 miles (i.e. 667.8 to 1025 km), and from Leh to Yarkand between 584 and 839 miles (939.8 to 1350 km), depending on the route chosen. (54) Ramsay (1890: 185-188) gives the following distances between Leh and Yarkand: summer route via Saser pass 482 miles (771.2 km), winter route via the Shayok river 520 miles (832 km), and the easiest but longest route 577 miles (923.2 km). The indication of 600 li (193.8 km) is more than questionable (leading into the middle of desert mountains), but 3000 li corresponds fairly well to the longest route between Leh and Khotan or the shortest route between Leh and Yarkand according to Forsyth or rather to the longest route between Leh and Yarkand according to Ramsay. To be exact, a journey straight south of Khotan would lead us to Rudok rather than to Leh. The direct distance (as the crow flies) between Khotan and Rudok appears to be the same as that between Khotan and Leh, although the actual route might then be somewhat shorter.

Xuanzang further mentions that the country "is extended from east to west, and contracted from north to south" (which, if referring to Ladakh or the Changthang, would be as true as the opposite, due to their north-west to south-east direction). He believes that the country is identical with that of the 'Eastern Women's Dominion'. His account had been summarised in the Shijia fangzhi with the additional note that Suvarnagotra was "not within the boundaries of India and [was] also called the Kingdom of [TEXT NOT REPRODUCIBLE IN ASCII] 'Great Yang-t'ung'" (Pelliot 1963: 699). Thus, if we were permitted to transfer the indication of 1000 li for the west-east extension of Greater Yangtong to Suvarnagotra, the distance would well cover both Leh and Rudok, whether we locate the country along the Indus River (in which case, the distance of 325 km or more would almost cover the entire distance from the source to Leh) or whether we locate it more to the north in the Changthang.

1.5.3 Suvarnagotra III (the Khotanese perspective)

Except for Tucci (1956: 98), who takes Suvarnagotra as the name of the ruling family of Suvarnabhu, nobody has as yet commented upon the fact that the Indian designation Suvarnagotra and its Tibetan counterpart Gserrigs or Gserrabs (55) do not refer so much to a landscape or a state, but to a people, clan, family, or 'race'. The Chinese word used by Xuanzang is Jinshi 'Gold Clan', but one may also find Jinxing 'Gold Surname' or Jinben 'Gold Origin' (Pelliot 1963: 696, 699). The Khotanese Saka original Suvarnaguttar- or ysarnai rrvl guttairi is used for a particular 'golden royal family', but the Khotanese Saka sources are rather reticent (cf. H.W. Bailey 1985: 82-83). Luders (as cited by Francke 1929: 148) has come across two kings of Kucha with the element suvarna as part of their name: Suvarnapuspa and Suvarnadeva. The people or the royal family were originally located in the Shazhou region (Dunhuang), which was accordingly also called the 'Golden District' (ysarrnai bdda) or with similar appellations (H.W. Bailey 1940: 602f. and 1949: 37). With the same word, ysarrnai, possibly in the sense 'imperial' the Khotanese also referred to their own kingdom (H.W. Bailey 1964: 10). Apart from the Khotanese texts, there seem to be also texts in Chinese and Tibetan mentioning a 'Gold Land' or 'Gold Family' in Shazhou (R.A. Stein 1951: 243f. n. 2; unfortunately without further references).

In the Tarim basin, 'Golden' could well have been an anthropological feature (the blond or reddish hair, not necessarily only among the IndoEuropean people, but also attested among the Xiongnu and Turks). If this is not the reason for the name, the designation might have referred simply to certain people dealing with gold. (56) The Darada had been commonly associated with trade or tribute of large amounts of gold, we should perhaps look for the Gserrigs in the 'Upper Indus Valley', i.e. in Gilgit or Hunza, which is definitely connected with legends of a rich gold production (see also Zeisler forthcoming a, Chapter 2 [section] 1.1). The equation with the Strlrdjya by Xuanzang would likewise point to this region. It is thus little surprising that Thomas (1935: 153), basing himself on the fact that already the Greek authors associated the gold production solely with the Darada, shifts the Gold Country to Hunza and Nagar.

On the other hand, the indication given by Ptolemy, most probably based on Indian sources, as well as the indications by the two pilgrims, quite certainly based on Indian testimonies, lead us to the upper Sutlej valley. This is the area where we find the Darpa or Darma, which may or may not have been originally belonging to the Darada or related tribes (see also Zeisler forthcoming a, Chapter 2 [section] 1.1). If they had, it might follow that the whole area from the Sutlej valley over Ladakh, Baltistan, Hunza, up to Khotan had been inhabited by Darada or related tribes.

Apart, perhaps, from the documents, R.A. Stein had alluded to, the Tibetan designation Gserrigs(gyi yul) '(Country of the) Gold Race' seems to be found only in the Vimalaprabadpariprccha (Drimamedpahihodkyis zuspa, Thomas 1935: 179-258), written and compiled not by Tibetan but by Khotanese authors. The designation is used for an unidentifiable region, somewhat peripheral to the main events. The text is, quite apparently, an amalgamation of various fragments of similar prophecies, not all of them referring to exactly the same situation. The historical information given, if historical at all, is hopelessly in disorder and partly contradictory.

Nevertheless, the main historical situation hinted to in the text seems to be located around the first Tibetan occupation of Khotan in 665/670. (57) The king of Skarrdo, who is to be the king of Li (Khotan), procures gold from the Gold Race in order to buy off Li from the Red Faces (Tibetans) (58) and hence becomes king of Li. This latter statement cannot easily be reconciled with the tradition that the royal lineage was unbroken (see also Xuanzang, Beal 1881-84: 490) and with the list of kings given in the Annals of Khotan (Liyulgyi lorgyus, Thomas 1935: 77-136). Possibly the prophesied usurpation did not last long enough to become official history. Or the official history, like the text, somehow accepted the usurper as legitimate member of the royal lineage. A further possibility could be that the principality Skarrdo had been a part of the Khotanese territory or became so afterwards. This Skarrdo does not necessarily have to be identical with the presentday Skardo/Iskardo in Baltistan. Given the possible derivation of the name from an original Iskandria/Sikandria (that is, Alexandria, see also Zeisler forthcoming a, Chapter 2 [section] 2.4.5), the name might well have been transferred to the present location at a much later time. (59) Albinia (2008: 181) mentions a village Sikundro in Chitral near Bajaur which is understood to mean 'Alexander Stopped' in Pashto. This could at least have been an alternative candidate for the designation.

In any case, the Gold Race Country seems to have been a vassal of Skarrdo or otherwise obliged to share its riches with Skarrdo. (60) Both countries are understood as being essential parts of Li, the king of Skarrdo proposing an alliance with the king of the Gold Race in order to unify all provinces of Li (ibid. p. 204, fol. 364a). From the Khotanese perspective (or from that of the Skarrdo usurper?), the royal lineages of all three regions are presented as being closely related, all persons being styled as reincarnations of the royal family of Ajatasatru, all male members bearing the name element Vijaya.

Li had first been raided by the Sumpa and Haza (61) and is now occupied by the Red-Faces (=Tibetans). (62) Sumpa and Red-Faces have destroyed many Buddhist buildings. The Chinese have made plans to invade the Gold (Race) Country, the Red Faces are alarmed, and a great battle seems to be imminent. While it is said once that the Gold Race Country is in the firm grip of the Red-Faces who do not let it go, in the same breath, it is also said that the Red-Faces are at a distance and do not approach to do harm (fol. 364a--to the Gold Race Country or to Skarrdo?). On another occasion it is merely stated that the Sumpa are threatening the Gold (Race) Country, but can be averted (ibid. p. 237, fol. 396a/b). For an unknown reason, the request for the gold and the alliance proposal must needs be kept secret (408a). (63) At some time, the king of Skarrdo and his son, the latter bearing the same name as the main hero, die in a battle (ibid. pp. 226-228, fol. 384a/b, 385b; it is not clear whether these passages refer to a secondary lineage or to events one generation earlier or later).

For Denwood (2008: 8) it follows from all this that an identification of this Gold Race Country with ZanZun can definitely be ruled out:
      In 649 Zhangzhung had been "destroyed by the Tibetans,
   who divided the people and scattered them in adjacent lands".
   True, Zhangzhung or elements of it staged a final revolt in
   678-9, but by then the Tibetans were firmly in control of
   Khotan. At the time of the composition of the Inquiry of
   Vimalaprabha, i.e. some time between 660 and 670, it is clear
   that both Khotan and Suvarnagotra, as well as Skardo, were
   independent of the Tibetans, while Zhangzhung was not.

Denwood does not seem to take into account the distinction into Lesser and Greater Yangtong (it was the latter that is reported to be annihilated in 649). Of course, he also believes that ZanZun ends southeast of Ladakh (see also [section] 2.2 below), but as I have shown ([section] 1.2 above), Ladakh was always a part of Zanzun. from the perspective of Old Tibetan documents. While I would likewise expect the Gold Race Country of this narrative to be located more to the west, I would nevertheless think that the ambivalent description of the Gold Race Country as being in the firm grip of the Tibetans and the Tibetans only threatening it from far would allow for an identification with (parts of) Yangtong, reflecting perhaps its ambivalent status until its final integration into the Tibetan Empire.

In a mythological side-narration (fol. 380 ff.), the Gold Race Country is connected with the Gold Mountain, which should have been located in Hunza (see Zeisler forthcoming a, Chapter 2 [section] 1.1). This mountain is characterized by a female mountain spirit, the rdksasl or devl Husa, who with her following keeps a batch of 500 merchants as lovers, devouring them, whenever a new batch arrives. The Gold Race originates from the offspring of the rdksasl and a last, rescued batch of merchants. There is some likelihood that, despite its Singhalese clothing, the story is to be connected with the Strlrdjya of Swat/ Uddiyana (Thomas, ibid. 224, n. 4). (64) The name Husa may be related to the river name Wakhsh (Oxus), rendered as Husha by Xuanzang with reference to the region on its upper course (cf. Beal 1881-84: 106). Jettmar (2002c [1995]: 199) takes this Husa as "the ancestress of a dynasty in Gilgit". He also points to a late reflection of this story in Shigar (2002a [1993]: 151f.). The legend speaks of a libertine the queen who used to eliminate her lovers whenever a child was born, which was likewise killed, except if it was a girl. When a Persian prince insisted on a legal Islamic marriage, she objected that she was a deity, and by a legal marriage she would become a laughing stock.

It seems thus that the Gold Race basically belonged to Hunza and its neighbourhood (as indicated via the reference to a Gold Mountain and the characterization as a Buddhist country (65)), and it stood in close connection with Li, on the one hand, and Skarrdo, on the other. The confusion (or doubling) of the names of the participants, the missing reference to a king from Skarrdo in the Khotanese Annals, the sudden appearance of a second king of Skarrdo assisting the main hero in becoming king of Li (ibid pp. 254f, fol. 415a) might perhaps point to Bolor as the main stage and a split or rivalry among its principalities even before the events related by Hyecho (Fuchs 1938: 444). As the names of the Skarrdo kings do not correspond to those of the Palola dynasty, the Skarrdo principality, whether identical with Baltistan or with (a region in) Chitral, seems to have been either independent or, according to the narrative, a dependency of Khotan. In any case, it seems to have kept close dynastic ties to both Bolor and Khotan (hence the close resemblance of the royal names).


It is noteworthy that the region H brusolona, which is certainly to be connected with Bruza (var. Brusa, Hbruzal, Hbrusal, Gruza, or Grusa), (66) is counted not only as part of Khotan in the Khotanese Annals, but is associated with the legendary establishment of Khotan rulership. The corresponding narrative yields a folk-etymology for Hbrusolona (cf. Thomas 1935: 100, fol. 428b f.), apparently based on the Burushaski word boso (Yasin) or busooso (Hunza-Nagar) 'calf' (cf. Berger 1974: 135, 1998 III: 65).

1.6 Preliminary conclusion(s)

Taking the evidences of the sources together, it appears that around the 7th to 8th c. CE, (Upper) Ladakh, more precisely a stretch of about 350 km, bordering in the north on Khotan, was known in Chinese sources as 1. Greater Yangtong, 2. Suvarnagotra, and 3. Nuguo. West of Greater Yangtong was Lesser Yangtong (including at least Lower Ladakh). (67) West of (parts of) Suvarnagotra was Sanbohe/Moluosuo (on which below). The name Suvarnagotra and its Greek, Khotanese, Chinese, and Tibetan equivalents, seem to have referred to three different regions in different contexts. While there is not much doubt about the location of Suvarnagotra I, attested from Indian, or rather Indian-based sources (Ptolemy and at least Hyecho), the further association in the Chinese sources with (one of the) Nuguo or Greater Yangtong (that is Suvarnagotra II) seems to be problematic. The error, if it was one, could have been caused by the non-discrimination of Lesser and Greater Yangtong, since the more western Nuguo should probably not be associated with Greater Yangtong, but at best with Lesser Yangtong and more likely with neighbouring Bolor. It is this latter area, where the Khotanese seem to have located their own Suvarnagotra III.

Nevertheless, lest the identical name should be coincidental, one can assume that there was a common ethnic substrate, which may have been Scythian (Sakean) in the beginning, (68) but may have been overlaid with a Pamirian (Dardic and/or Burusho) population. The ethnic continuum between Suvarnagotra III and Suvarnagotra I could well have been disrupted in the course of various Central Asian largescale migrations.

The following table summarises the main sources for the localisation of Suvarnagotra: Xuanzang, Vimalaprabhdpariprcchd (VP), and Hyecho in chronological order. The first row under each source specifies, whether a place is mentioned at all ("M"), the row headed by "V" specifies whether a place was visited in person, the row headed by "O" indicates the Tibetan occupation. Only Xuanzang gives the relations between the places in terms of orientation. His silence on the political engagement of the Tibetans in the area is certainly not surprising.

The distinction between Lesser and Greater Yangtong appears to be pretty much in concordance with the distinction between 'Little Tibet', i.e. Baltistan, and 'Great Tibet', i.e. Ladakh, as we find it in early western sources, based on the corresponding Indian and Persian designations, cf. Desideri's account in his first book:

We then found ourselves at the foot of a very high, steep, and formidable mountain called Kantel, on whose summit is the boundary between Kascimir and Lesser Thibet, called in Persian Khoval [Kalan] Thibet, in Hindustan Ciota [Chota] Thibet (both words bearing the same meaning) and called by the inhabitants "Baltistan" ...

A few days later we entered Second Thibet, known in Persian and Hindustan as Kalan and Bara Thibet which means Great Tibet, while by the natives it is called Lhata-yul.

Second Thibet, or Lhata-yul [Ladak], is two month's journey in length; it borders in the North with the Kingdoms of Kaskar [Kashgar] and Yarkand, to the south with the kingdom of Collahor [Kulu?], on the west with Lesser Thibet or Baltistan, and on the east with the great Desert of Ngnari Giongar [Ngari Jungar] (Filippi 1937: 82f., 84, 85, editor's additions in square brackets in the original).

A slightly different description is given in the second book:

Thibet consists of three distinct Kingdoms: The first is Lesser Thibet, or Baltistan, bordering on the South on Cascimir and on the East [North] with the Kingdom of Kaskar [Kashghar]; it was once ruled by several Kinglets, or rather Barons, and now forms part of the Mogol Empire.

Second or Great Thibet, also called Lhata-yul, takes two months to traverse from West to East. On the Western Frontier [North-North-Western] lies Lesser Thibet, on the Eastern the Great Desert of Ngnari Giongar [Hundes or Ngari Khorsum].

Third Thibet is called simply Thibet by Europeans and by Persians, but Hindustan-Mongolians call it Butant . The vast extend of this Third Thibet will be understood when I tell you that it takes more than six months of incessant travelling to go from the Western to the Eastern boundary, that is from Cartoa to Sciling [Sining] (ibid. p. 129f., editor's additions in square brackets in the original).

I should thus argue that the distinction between Little and Great Tibet, attested also in the 15th c. Rdjataranginl of Srivara with reference to contemporaneous events (III, iii 440-443, Dhar 1994: 546f.), is based on an ancient tradition, reflected in the Chinese designations Lesser and Greater Yangtong and possibly also in the Bonpo classification of an Innermost, Middle and Outer ZarZuj. In all these cases, the notion of 'Lesser' or 'Innermost', that is, 'Closest' refers to the ancient axis mundis, Mt Meru, to be looked for in the Pamirs (see Zeisler, forthcoming a, chapter 2 [section] 2.4.4). There are certain indications that the border between the two entities Greater and Lesser Yangtong as well as between the two Old Tibetan entities ZanZun stod and smad could have been situated just below Leh (see also [section] 3 below), although it is certainly possible that the border between the two areas had been subject to fluctuations or to a major change with the advent of the Namgyal dynasty. It is possible that the Chinese name Yangtong reflects a designation underlying also the Tibetan designation Byaqthaq. But the latter name can by no means have been the indigenous Tibetan name, since its etymology presupposes a southern centre in relation to which it is situated. Most probably the Tibetan name results from an attempt of etymologisation. (69)

As in the case of Little and Great Tibet, it seems that there must have been a 'third' or 'real' Yangtong of much larger extension and possibly greater historical impact. This 'real' Yangtong could thus have been identical with Grand-Zanzun. that is, the eastern parts of the Changthang. The fact that one of the Women's Dominions is located in south-eastern Tibet, and that a Greater Yangtong appears also in north-eastern Tibet indicates that neither the 'Ladakh'entity nor ZanZun Proper had a well-defined border, and that there were no clearly distinguishable ethnic groups or political entities along the Changthang and its western and eastern extension, but rather an amalgamation of various tribes that--due to similar economic and ecologic conditions--shared much of their material (and perhaps also linguistic) culture, despite their possible different origins. Such situation could perhaps explain the strange location of the land (=ZanZun) in south-eastern Tibet in the Hudud-al-'Alam (cf. Map 2 above).

The same would hold true for the apparently more western aspect of Suvarnagotra and the Strlrdjya, which implies an ethnic and cultural continuum up to Chilas, Hunza, (70) and Chitral, and further to the south, to Swat/Uddiyana, (71) but most probably not to Guge. The latter identification, confidentially proposed by Francke (1912: 269) and Tucci (1956: 105), is based on the location of Suvarnagotra I, the one that is historically well testified. Tucci (1971a: 550) adds that Clna 'China', an Indian designation, possibly referring to Guge, and the Strlrdjya were "often assimilated or quoted together". Unfortunately, he does not specify in which texts other than in Kautilya and in which context other than an enumeration of cloth exporting regions such 'assimilation' occurs. The mentioning together of two countries in an Indian text, the Brhatsamhitd exemplifies this clearly, is no proof that these countries are really found in close vicinity, not to speak of being identical.

Neither Francke nor Tucci realise that there might be a quite different Suvarnagotra III further north-west, and that the Chinese sources, other than Hyecho (who remains absolutely silent about women), rightly or wrongly, also refer to a Suvarnagotra II within, or congruent with Greater Yangtong (corresponding to Upper Ladakh). Given the identical coordinates with respect to Khotan, it must be this second Suvarnagotra with which the Nuguo is, perhaps wrongly, identified.

While these coordinates for the western Nuguo, making her practically congruent with Greater Yangtong, seem to be problematic, the queendom may nevertheless have stretched over (parts of) Lesser Yangtong or Lower Ladakh. Furthermore, the population associated with the queendom, most probably Dardic (or perhaps also Burusho) tribes, could once have spread much further to the east, and this could have caused the presumably wrong association with Greater Yangtong. The presence of 'emancipated' women among the prehistoric ruling elites of ZanZun or perhaps more generally on the Tibetan plateau seems to be corroborated by the representation of a few Bonpo female deities as wearing armours and helmets like warriors, and by the Bonpo lore of particular helmets and armours for queens, worn when they were observing battles (see Bellezza 2008: 241f., 325, 329; on p. 325, n. 361; he rightly points to similarities with warrior-like women among the Scythians and Xiongnu).

The mixed ethnic composition of the northern belt of Tibet seems to have been dominated by the Supi(ya)/Sumpa and Tuyuhun. the latter certainly, the former only possibly of Xianbi origin, (72) and while some of the characteristics of the Nuguo are typical of the Tuyuhun (and the later Tibetans), such as painting their faces (Pelliot ibid., p. 694, Tong Tao 2008: 173f.), other cultural traits show that they must have had absorbed, or must have been influenced by, various other tribes, some of which were part of the Indo-Iranian sphere. The use of an Indic script is perhaps the most prominent and most tangible feature, followed by the veneration of the Asuras, the two types of burial customs for Yangtong and Nuguo, which, although showing Central Asian features, are both quite distinctive from those of the later Tuyuhun (for which see Tong Tao 2008, chapter 6) and the Tibetans. A further possible Iranian trait is the celebration of the New Year at the winter solstice. (73) The bird divination connected with this celebration seems to be likewise quite peculiar and has nothing to do with the bird divination of Old Tibetan sources, based solely on the birds' utterances (cf. Laufer 1914). The Jiu Tangshu notes:

Our eleventh month is their first. Their custom is, as soon as the tenth month begins, to order a diviner to go into the mountains to make offerings of ch'u and to scatter in the air barley which has been steeped in wine. With great incantations he calls the birds. Suddenly a bird like a pheasant comes flying to the bosom of the diviner, who then splits its belly open and examines it. Whenever there is one grain of corn in it, the coming year will be fruitful; but if there is 'hoarfrost and snow' there will be calamities (Pelliot 1963: 700f.)

The description of the Suishu differs only slightly:

At New Year they sacrifice men or monkeys [...]. When the sacrifice is over, they go into the mountains, where [a diviner] makes incantations. A bird like a female pheasant comes and perches on the [diviner's] hand; he splits open the stomach and examines it. If there is grain, the year will be fruitful; if sand and gravel, there will be calamities (ibid., p. 695; additions in the original).

A reverberation of this divination technique might perhaps be found in the Chayraps, the 'genealogy of the beer', from a (Lower) Ladakhi cycle of marriage songs: a cultural hero kills various birds and examines their stomach in search of the first grain; finally barley is found in the stomach of a pigeon and disseminated for the first time (KHAL).

Strangely enough, like the celebration of New Year at the winter solstice, the use of an Indic script is located only in the eastern part of the Women's Dominion. As far as I know, no prehistoric inscriptions or writing systems have been reported from Eastern Tibet, not to speak of an Indic script. But we know of early inscriptions in Lower Ladakh, and of course further west, in the so-called Upper Indus Valley. Consequently, the literate 6th to 7th c. Women's Dominion should be sought in the west, not in the east.

In connection with the seemingly far spread extension of all these tribes, it may perhaps also be of interest that the LDRR (35. 15-16), when defining Maryul as inherited by Dpalgyimgon, mentions two goldmines, each time in combination with the word hgog: gserkha hgog, in the east, near Demchok, and the so far unidentifiable gserkha hgogpo (MS L: mgonpo), in the north. The reading of MS L as 'protector' is apparently based on the verb hgogs ~ hgegs, bkag, dgag, khog 'hinder, stop, prevent'. But the verb hgog, bkog, dgog, khog 'take away forcefully', perhaps in the sense 'exploit' might be more appropriate. The apparent Tibetan etymology might be misleading, however. There are several place names bearing the element gog ~ gug, first of all, of course, Gug(g)e (Goggadesa in Jonaraja's Rdjataranginl; Sahni and Francke 1908: 186) of ZanZun smad.

Another Gog (or Kog) is mentioned together with Bruza in OTA II for the year 747 (l. 10, the variant Kog appears in ll. 4), their inhabitants being reported to have fled, when the Chinese general Byimpo appeared in Gogyul, i.e. in the Hindukush. But the name is also attested in the northeast: Gogchu, near Dantig, rendering Chinese Gouzhou. If not related to metals or mining, the name could either have been a tribal designation or a designation for special types of settlements. In Western Tibet, there are several place names ending in -gog or -mkhargog (Bellezza 2008: passim). While the Tibetan dictionaries yield the meaning 'ruined' for gogpo, apparently productively used in several compounds, there remains the possibility that gog originally referred to the same entity as mkhar, and that the compound mkhargog was thus a translational compound, this meaning being lost, the actual meaning could perhaps have been derived from the dilapidated state of all places in question. The problem, of course, is: in which language could we find a corresponding designation?

2 An old name of Ladakh?

2.1 Moluosuo alias Sanbohe

There has been a certain consensus that Moluosuo ([TEXT NOT REPRODUCIBLE IN ASCII], W.-G. Molo-so) is the Chinese equivalent for Mard or Maryul, the alleged old name of Ladakh. Cf., among others, Cunningham (1854: 4), Francke (1908), Pelliot (1963: 706f.), and Uray (1990). Tucci (1956: 94, n. 1), however, opts for the merely emended form Moluopo, interpreted as the Malava, a tribe of northern India! (74) The country is mentioned only in Xuanzang's report, where, unfortunately quite problematic and contradictory coordinates are given. Moluosuo is said to lie about 2000 li north of Lahul (cf. Beal 1884: 210f.). Neglecting the distance, this direction is taken as evidence for the identity of Moluosuo with Ladakh, and thus also with the Old Tibetan place name Mard.

Moluosuo is said to lie west of Suvarnagotra, for which, as we have seen ([section] 1.5 above), there are three candidates. Moluosuo is further identified with Sanbohe ([TEXT NOT REPRODUCIBLE IN ASCII] W.-G. San-po-ho), a place so far not located, particularly not in Ladakh. (75)

Moluosuo/Sanbohe cannot be Ladakh, if lying west of either Suvarnagotra I (Upper Sutlej) or Suvarnagotra III (Hunza-Gilgit). If the country should thus lie west of Suvarnagotra II (Greater Yangtong, Upper Ladakh), we would quite naturally have to search for it west of Upper Ladakh. The name Sanbohe is reconstructed variously as *Sampaha (Beal 1881-84: 227 and Tucci 1956: 93), *Sampaha (Pelliot 1963: 706), and *Sampaka (Tucci 1977: 73). It could thus perhaps be identified with Sambi, known in Chinese sources also as She-mi or Syami, a place in Chitral (Thomas 1935: 142, Tucci 1956: 94: Sambi). Pelliot (1963: 706) arrives at a similar reconstruction, although he ultimately rejects it, perhaps because he thinks that Sanbohe should be Ladakh or because Xuanzang refers to this place as Shangmi (Beal 1881-84: 479). Petech (1947: 89) differentiates between She-mi, for which he suggests the Sanskrit equivalent Symaka, and Shang-mi, which he identifies with Mastuj in Chitral.

There might be also a relation with the place name Sam(bhu/cu)t(dnnd/ md) appearing in an inscription in Shigar (Hiniiber 2004: 68f. no. 33A, p. 68f.) in connection with a ruler of what might be the Hunza area. (76) While the Chinese name Moluosuo could well render a Tibetan *Mars, which is not too far away from the attested Mard, it could also render Buruso, cf. the parallel use of a transcriptional Mo for Hbal/Hbah and Molu ([TEXT NOT REPRODUCIBLE IN ASCII]) for Hbro (cf. Pelliot 1961: 134f., Richardson 1971: 434, and Petech 1994: 651, 657, n.12). In fact, since the distinction between (prenasalised) voiced oral and nasal stop consonants had been neutralised in Middle Chinese, Moluosuo is the only possible rendering of an original B(u)ruso (Buruso, Bruza).

If we take the indication 2000 li from Lahul to Molosuo to be correct, corresponding to ca. 650 km, (77) Moluosuo should be sought about 300 km further to the north (-west) of Leh. The distance from Leh to Lahul on the modern Manali highway is ca. 360 km up to Kyelang (the district headquarters of Lahul-Spiti). (78) Given the fact that this route is not particularly difficult and was constructed more or less along the ancient tracks, the latter should not have been much longer, perhaps even shorter (cutting the serpentines of the vehicular road at the Taglangla and the Baralachala).

There are thus several points that could speak for an identification of Moluosuo with Hunza, rather than with Ladakh, and for an identification of Sanbohe with a place in Chitral. One may perhaps even argue that Chitral and Hunza are not too much to the west, to make the indication north of Lahul completely wrong (compared to other much less precise coordinates). On the other hand, one may wonder why the countries lying between Lahul and Chitral or Hunza were not mentioned. It is also somewhat problematic to conceive of a route from Jalamdhara/Brahmapura to Chitral or Hunza via Ladakh or Zanskar instead of a route along the southern slopes of the Himalaya and via Kashmir. Finally, one could have expected both countries to be mentioned again by both pilgrims in the context of the description of Bolor. But these apparent contradictions could certainly be the result of a later interpolation (see below).

One conclusion could be that all indications of distance are completely unreliable, even more than the indications of direction. In this case, however, one has to give up all kinds of identifications, including that of Moluosuo with Ladakh, which would be simply arbitrary. This conclusion seems to be corroborated by the indication of a distance of 1800 to 1900 li between Kulu and Lahul, which would correspond to 580 to 610 km if counted in Tang miles. The distance on the modern road is 160 km from Kulu to Kyelang (the district headquarter of Lahul), and again, nothing indicates that the ancient route should have been four times as long.

Disregarding the problem of the distance, Pelliot (1963: 698f.) has questioned the whole passage concerning the route from Matipura up to Piluoshanna, in the middle of which the reference to Kulu, Lahul, and Moluosuo should be found. His reason is that the place names are difficult to identify, while in all other cases such problems would not exist. This is not really a strong argument, but if he were right, the designations Moluosuo and Sanbohe might be nothing than hoaxes.

The biography of Xuanzang, written by one of his disciples, mentions Kulu, but remains silent about Lahul and Moluosuo (John E. Hill, p.c.). This could be taken as an indication that Xuanzang did visit all the places questioned by Pelliot, except the latter two. The passage concerning these places, must have been added later, either by Xuanzang himself, or, more likely, by a different hand, whereas the biographer would have used the earlier report. The question remains why the distances are so utterly wrong.

Whatever the reason, it might be expedient to project the actual distance between Kulu and Lahul onto the almost equal distance given for Lahul to Moluosuo. If the 1800 to 1900 li correspond to the 160 km of the present distance between Kulu and Kyelang, the 2000 li between Lahul and Moluosuo should then correspond to another 168.5 (or 178) km, which could lead us, following the road after the Rothang pass first westwards and then towards the northeast and north, only up to the Tsokar and thus to the western and southern confines of the 'lake district'. If one would turn eastwards after the Rothang pass and then again towards the north and northeast, one would have a choice of several routes towards the northeast, stopping short before Tashigang or Demchok. One could also follow the Spiti valley down towards the Sutlej. The radius would be something like 45 km from Tabo. (79)

In these last cases, the location of Moluosuo would definitely be west of Suvarnagotra I. It could then be identified with the Old Tibetan province of Mard, which R.A. Stein has located just east of Spiti (see below). Perhaps someone better acquainted with the area than me could find there an appropriate candidate for the mysterious Sanbohe, but given the triplication of its neighbour Suvarnagotra, I can't help feeling that at least the name Sanbohe, belonging to Bolor, had been wrongly associated with the Guge area, due to a backseat driver's intervention in Xuanzang's original report. If the name Moluosuo rendered an original ZanZun-ian Mard, Mars, or perhaps rather Smar-(sa), this could have led to a confusion with a similar rendering of Burus(h)o, and Sanbohe, as referring to a neighbour of the latter, could then have been transferred to a non-existing neighbour of the former. The confusion could have been enhanced by a transfer of the mythic Maru(tse)/Maru(cina) from Chitral to the neighbourhood of Kulu (see below).

2.2 Mard or Maryul

Although claimed to be the original name of Ladakh, (80) the designation Maryul and its possible variant Mayyul is not found as referring to present day Ladakh before the 11th c. or even later. The name appears first in an inscription in Alchi, dating from some time between the early 11th c. (according to the traditional dating) and the early 13th c. The dating of the Alchi temple group into the late 12th to the early 13th c. has been argued for, among others, by Luczanits (2005: 73-78). By the mid 13th c. the reference to Ladakh and Baltistan must have been firmly established among the elites. The biography of Orgyanpa Rinchen Sejge Dpal (1230-1293) contains the travelogue of his pilgrimage to Orgyan (Uddiyana/Swat) which he undertook some time after 1260. There he describes the river Sindhu (Indus) as arising from the Kailasa, flowing through Maryul (clearly Ladakh and Baltistan), then through Brusa on the north of Kashmir (which he describes as bordering on Zajsdkar and Purig) and Staggzig (here apparently referring to Chilas), before reaching Orgyan (which should here be the Buner valley; cf. Tucci 1971b: 396 with n. 9). It is interesting to note that in the earlier travelogue of Rgodtshajpa Mgonpo Rdorje (1189-1258), Maryul is reached via Zajsdkar, the latter being described as 'Tibetan' and thus not part of the kingdom of Ladakh (ibid., p. 417).

The Mayul or Muyul of the 10th c. Hudud-al-'Alam ([section] 11.3, fol. 16b, ed. Minorsky 1937: 93) has, as Minorsky (ibid., p. 256) rightly states, "nothing to do with Mar-yul 'the low country' which in the Tdrlkh-i Rashldl [of Mirza Haidar], pp. 410, 456, is applied to Ladak", despite Denwood's (2008: 14) suggestion to the contrary. Mayul refers to a tribe, and Muyul to a small town of the unknown N.zvan (or T.rvan ~ T.zdan) province ([section]11.3), which Denwood (2008: 8) identifies with Nubra, but which is located close to China ([section]5.3, fol. 6a, p. 61), possibly in the Kokonor area (Minorsky 1937: 258). The Muyul tribe would provide the Tibetan kings ([section] 11.3), and only if the lineage extinguishes the Tibetans would elect a chief from among the *Ajayul (Haza; [section] 11.7, fol. 16b, p. 93). Quite apparently, the Muyul is also not the land of the mothers (Minorsky, p. 257: ma for ama), but the land of the Rma (or Rmu) clan or tribe (on this tribal name see Zeisler forthcoming a, Chapter 4 [section][section] 2.4.2 and 2.4.4). Beckwith (1989: 68, cf. also p. 69, n. 29) suggest that M(u) yul could be a mistake for Mabud ~ Mabud ~ Mabd, Mahdbhota or Greater Tibet, a name that appears in the 9th c. Ahbdr as-Sin wa l-Hind. But this is quite unlikely, particulalry since Muyul is presented as a small town within a province within (Greater) Tibet.

The Old Tibetan Annals mention neither Ladakh nor Maryul or Majyul. Only one single entry in OTA 719 (l. 213) alludes to a province Mard, namely together with Zanzun. and in connection with a phalos, a conscription or registration of the male population: ZayZuyda) Mardkyi phalos bkug. The actual local pronunciation might well have been *Mars, as the alleged Chinese equivalent Moluosuo would suggest (Uray 1990: 220).

No further indication is given, but since locations are quite often enumerated from west to east, one should be able to preclude a location to the west of Zanzun. e.g. in present-day Ladakh. The latter identification is also precluded for the reason that Ladakh was included in either ZanZun stod or smad. Since Mard is mentioned besides Zanzun. it was probably an entity not yet fully incorporated into Zanzun. and this would make sense if it lay at the border to India. It is quite conspicuous that the name does not reappear.

The designation Mard as much as Maryul could reflect the ZanZun epithet smar ~ smra (81) 'golden'. In a more involved way, it could refer to a tribal group, the Rma (or Rmu), whose name is either related to the word smra 'speak' or to the 'monkey ancestor' of the Qiang (for these relations, see Zeisler forthcoming a, Chapter 4 [section] 2.4). Finally, the epithet could perhaps be related to s/dmad 'low'. The following possibilities are open to speculation:

(a) Lower Ladakh (Mard related to s/dmad 'low'). In this case, Upper Ladakh should have been part of ZanZun.

(b) Upper and Lower Ladakh, i.e. the whole area between ZanZun Proper and Baltistan, in which case ZanZun no longer comprised any part of Ladakh. The area could have passed as 'low' in reference to ZanZun.

(c) The gold fields of Manasarovar and/or Upper Ladakh plus some additional areas (Mard related to (s)mar 'golden'). This might imply that the northwestern part of Ladakh had been independent or, more likely, part of one of the two Bolors.

(d) The whole gold-producing area, starting with Mnasarovar, along the Indus and Shayok over Nubra and Baltistan to Hunza and then further down to Swat.

(e) The area where the Rma (or Rmu) tribe(s) (or clan(s)) settled. The form rma is a further development of smra. As the epithet Smar ~ Smra for ZanZun signals, this group must have been constitutive for Zanzun. and thus we should expect to find it in its core area. As in case (c), at least Lower Ladakh, but most probably also Upper Ladakh would have been a separate entity or separate entities.

(f) A fictive or idealised 'Golden Land' at the foot of the (transferred) axis mundis, the Kailash. (82) Again, the name would have had nothing to do with present day Ladakh.

(g) Finally, one could think of the mythic land of Maru(tse) or Maru(cina).

Possibilities (a) and (b) are quite probably ruled out by the conventions of the Tibetan administration, by which the western areas were 'Uplands' stodphyogs by definition. As we have seen, the convention seems to have been followed also in ZanZun. Nevertheless, possibility (a) has been suggested without any further comment by Hoffmann (2003 [1990]: 52).

Initially, I would have taken possibility (c) for the most likely one, with Maryul simply referring to the gold fields near Rudok and the adjacent areas. It might have, at times alternatively or additionally referred to the gold fields ascribed to lake Manasarovar. One could have seen here the Suvarnabhu(mi) of the Indian sources or the Suvarnagotra of the Chinese, which Fuchs (1938: 439, n. 4) locates north of lake Mnasarovar. (This would approximately correspond to Denwood's 2008: 9 localisation of Suvarnagotra). But cf. the discussion in the last section. The main goldfields of Western Tibet probably lay north of the Kailash "in the region of modern Thok Jalung", which are attested as "the main centre of gold digging in the 19th century" (Petech 1998: 246). "As early as the 15th century Guge appointed a 'master of the gold fields' (gser-dpon) in g.Yas-ru Byang-pa" (ibid., p. 255, note 99).

MTrzaHaidar (Elias 1895: 411) includes an enthusiastic remark concerning a goldfield in the Changthang, known to him under the Mongolian name Altunji 'goldsmith'. But the gold production of Guge would have been even richer and unlimited:

Again, Guga has two hundred forts and villages. It is three days' journey in length, and in it gold is everywhere to be found. Where ever they dig up the earth and spread it on a cloth, they find gold. The smallest pieces are about the size of a lentil [adas] or a pea [mash], and they say that sometimes [lumps] are found as large as a sheep's liver. At the time when I was settling the tribute upon Guga, the headman related to me that a man was lately digging a piece of ground, when his spade stuck fast [...] Having removed the earth, he saw that it was a stone, in the middle of which was gold. [...] A body of men went to the spot and extracted it, and having broken the stone, found in it 1,500 Tibetan mithkdls of pure [mohri] gold (a Tibetan mithkdl is worth one-and-a-half ordinary mithkdls) and God has so created this soil that when the gold is taken from the ground it does not diminish ..." (ibid., p. 411f.).

As far as Ladakh is concerned, the accounts vary. According to Francke (1907: 15), there seems to have been some substantial gold digging in Lower Ladakh:
      I have travelled along the Indus from Saspola to Dartsig,
   a distance of over fifty miles, and have seen but few parts of
   the ground which have remained untouched. It looks as if the
   ground had been worked with huge ploughs. In many places
   remnants of masonry can still be seen in the earth. There
   has been a break in the digging for several years, owing to
   political causes.

Individual gold digging is also remembered by people of Ladakh, but the existence of mines appears to be unknown. My informant from Gya in Upper Ladakh showed me pits about two metres deep, where not too long ago villagers had dug for gold, while according to an elder of Khalatse, in his youth, it was only poor Balti men, having no other income, who came to Lower Ladakh to dig up the soil. (83) The Buddhist clergy had eventually prohibited gold digging under various spiritual pretexts (KHAL; possibly these are Francke's political causes), but one may assume that this prohibition would hardly have been effective, if large quantities of gold could still have been obtained. The same may perhaps hold for Western Tibet; at least Elias (ibid., p. 412, n. 1) describes a more modest reality in the 19th c.
      The existence of gold in the western provinces of Tibet
   is well known, but the quantities found are very small and usually
   confined to dust--nuggets being seldom heard of. The quality is said
   to be good, and most of it finds its way to Kashmir and India.

This may be contrasted with a contemporary description of Gilgit and Nagar:

The Bagrot valley [...] was further of interest in that about the best gold of the country is found in its bed, and in the Gilgit river below, where the Bagrot stream runs in. Every year, after the summer floods are over, the deposits in back waters and sand banks are carefully washed, and a fair amount of gold, considering the rough methods and unskilled labour of the prospectors, is collected. When we first went to Gilgit pure gold was selling there at two-thirds the price in Srinagar [...]. The amount collected by the officials each year must have been considerable. Every petitioner or visitor, as a rule, presented his little nazzar [...] in gold dust (Durand 1899: 220).

The untold profusion of the apricots and the quantity of gold which can be washed out of every stream, has gained the country among the Nagaris themselves the name of "the land of gold and apricots" (ibid. p. 143).

This description, must have escaped Tucci's attention, otherwise it is not intelligible that he (1956: 98) claims "not [to] know that Hunza-Nagar was ever known as a country as rich in gold as Western Tibet".

Possibility (d) came into my mind when confronted with Tucci's remarks concerning the gold-production in Swat and the necessity, so to speak, to deBolorise Baltistan. However, as far as the name Maryul is concerned, such an extension is not supported by any textual evidence. In fact, all non-Ladakhi sources place Maryul in either ZanZun Proper or in what we could roughly call Upper Ladakh. Several post-phyidar sources describe Maryul as a land of lakes, surrounded by snow mountains or as a land surrounded by lakes, which according to Vitali (1996: 253, n. 365) refers to the Rudok area, but perhaps also to more eastward areas (see also below).

Possibility (e), namely that Maryul/Mard referred to an area where the Rma tribe or clan settled, is proposed by the Ladakhi scholar Bsodrnams Tshebrtan Yoseb Dgergan (1976: 184). According to him, both variants of the name, Maryul and Marjyul are related to an ancient East Tibetan tribe Rma in or from China. (84) As its members outnumbered the Menag and Ltoj, their tribal name happened to become the country's name Rmayul (the author seems to imply that the Rma tribe or perhaps all three (!) tribes had settled in present-day Ladakh). By and by the name changed into Maryul or Majyul (the latter form apparently as a reminiscence of the outnumbering). As this tribe later on had become extinguished, the country got the name Bladvags (Ladakh):
      dejsaj Bladvags byabahi sa Rgyahi-najna sjondussu
   Menagdaj Ltojgi rusla gtogspahi mi nujtsamdaj Rmahi
   ruslas chadpahi mi majtsam gnasstabskyis | ruskyi mij
   de yulmijdu somas Rmayul thogspalas | rimgyis gyurnas
   Maryuldaj Majyuldu soj | phyissu de yaj medpar sojnas
   Bladvags thogspar gyurto

That the Rma clan had some importance also in the western areas might be indicated by the place name Miru in Upper Ladakh, which is pronounced with high tone/Miru/by its inhabitants. The name thus represents a former *Rmiru, resulting from *Rmeru or directly from *Rmaru due to vertical vowel assimilation. (85) But given the complicated phonetic development, sketched in Zeisler (forthcoming a Chapter 4 [section] 2.4.5), it is not very obvious that the name of the landscape Mard could have been derived from the clan name Rma. Furthermore, a place name *Rmiru or *Rmaru could possibly also be connected with the mythic land Maru(tse) or Maru(cina) (see below).

R.A. Stein must have had an identification in mind that is close to possibility (f), since on his map "L'habitat et les habitants" (1981: 14-15) we find Maryul in "Ngarikorsum", east of Spiti, while the area around Leh is identified with Suvarnagotra and the Women's Dominion (West). R.A. Stein apparently did not accept that Maryul has anything to do with Ladakh. It is quite unfortunate that he did not find the opportunity to expound his position.

He does give, however, a further hint elsewhere (1959: 231). In the early 11th c., Jiaosiluo Qinan Lingwen (EFEO Kio-sseu-lo K'i-nan-ling-wen, Tibetan Rgyalsras ?Skyignam Glijdbon, cf. ibid., p 239f., n. 43) was installed ruler in Tsojka. He was claimed to be a descendant of the Tibetan imperial dynasty, a grand-grandson of Khri Bkrasis Brtsegsdpal, the latter being the elder brother of Skyidlde Nimamgon. Khri Bkrasis Brtsegsdpal could establish himself in Lhartse in Rtsajstod and his descendants became rulers of small principalities in western Rtsaj, the most important lineage being that of Majyul Gurthaq (cf. Petech 1994: 655). According to the Chinese sources, the Jiaosiluo hailed from Wusanmi (EFEO Wou-san-mi) or Gaochang Moyu (EFEO Kao-tch'ang Mo-yu). For the latter designation R.A. Stein (1959: 231) offers an interpretation in terms of Stodbyaj Maryul, that is, 'Maryul of the Northwest'. If this should be correct, this Maryul can by no means be identical with Ladakh, and most probably even not with the Maryul near Spiti, since the latter would belong to the lineage of Skidlde Nimamgon. It is, however, not impossible that the lineage of Khri Bkrasis Brtsegsdpal played the same political game as the former in intentionally confounding the names Majyul and Maryul (in order to make claims of inheritance across the lineages). Not being the crown prince of Majyul, it is also conceivable that the Jiaosiluo was brought up at the court of his cousins in Guge or Purang.

The location of Maryul in Guge or near Spiti is corroborated by the above-mentioned biography and the entry in BRGY, according to which the first skor of Mjahris contains Purajs, Maryul or Majyul, and Zansdkar (see above). Since Guge is not mentioned, it should be identical with Maryul/Majyul. (86) This corresponds well with the fact that Maryul/Majyul is mentioned between Purang and Zanskar. The enumeration was certainly not arbitrary, and also not merely triggered by alphabetic ordering as shown in the other two skor. Maryul/Majyul thus lay between Purang and Zanskar in a south-(east) to north-(west) direction. (87) It would probably also have included Spiti. In his Hdzamglijrgyasbsad Blama Btsanpo (ed. Wylie 1962: 3/55f) gives three alternative divisions of Mjah ris with 1. Stagmo (apparently in Baltistan or Purik) and Ladwags; 2. Majyul and ZanZun; 3. Guge and Buhraj. The second alternative is the one given by BRGY with slight variants in the names: 1. Spuraj, Majyul, Zajsdkar; 2. Hchiba (=Khotan), Blasa (Brusa), and Sbalti; 3. Zanzun. Khrigse stod and smad. The correct division, however, would be 1. Ladwags, 2. Ruthog, 3. Guge. Majyul/Maryul is thus closely connected with the core part of ZanZun and with Zanskar, but not with Central and Upper Ladakh. A location between Zanskar and Guge or Purang would match well with the location suggested for Moluosuo (see above). And it may also correspond to the suggested peripheral location of the Old Tibetan Mard. (88)

A location near the Kailash and more particularly in the vicinity of Guge and Spiti would also make sense in the context of Skyidlde Nimamgon's alleged division of his kingdom amongst his sons, (89) where we could expect that the eldest son, Dpalgyimgon, inherits a central or otherwise important part and not some marginal border area in the periphery. (90) It is uncontested in the Tibetan sources that the eldest son, Dpalgyimgon (var. Dpalmgon, Dpalgyi Lderig, Dpallde Rigmgon, Dpallde Rigpamgon, Rigmamgon) inherits Maryul (or Majyul) of Mjahris. There is some variation in the share of his two younger brothers, but in general the second son Bkrasismgon (var. Bkrasis Ldemgon) inherits Purang (Spuhrajs, var. Puhrajs, Puraj, Puroj, Phurajspa) and, depending on the source, also (parts of) Guge/Local-Zanzun. while the third son Ldegtsugmgon (var. Sdegtsugmgon, Sdebtsunmgon, Ldebtsunmgon) inherits either Guge/Local-ZanZun or Zanskar, Spiti, and Spilcogs, or some regions classified as Mon countries, probably Lahul or perhaps parts further east towards Nepal. (91) That the eldest son would split off from the main territory and set up a separate kingdom is rather unbelievable.

The location would also make more sense in view of the 'confusion' between Maryul and Majyul, which could have been rather intentional in order to legitimate possible claims over Majyul in Western Tibet, ruled by the descendants of Skyidlde Nimamgon's younger brother.

Given R.A. Stein's location, it appears to me that possibility (f) might come closest to the truth. Whether or not the gold production was more substantial in prehistory than in the more recent past, the notion of gold or a golden land seems to be part of the mythologic or symbolic geography of the Kailash as a representative of Sumeru (these concepts may have been transferred to this area with all other spiritual concepts concerning the cosmic function of the axis mundis). The golden fundament of earth (gsergyi sagzi) arises in the cosmic waters; upon this fundament the Sumeru comes into being. (92) A connected concept is that of a wish-fulfilling paradise: Vitali (2003: 44, n. 7) mentions a passage in the Dpagbsam ljonbzaj, according to which a Tree of Discrimination (Dpagbsamsij) is situated on the right side of lake Manasarovar. Its fruits are said to turn into pure gold, when falling into the water, hereby also producing the sound hdzambu, giving thus the base for the name of the southern continent (implicitly, this statement evokes the concept of Meru). The same spiritual concept seems to be reflected in a 17th c. dedication sheet from Western Tibet:

[parallel] kye legs [parallel] Rinchen Bzanpohi zabskyi lcagspahi [! for homophonous bcags-] gnas [parallel] Blobzan Gragspahi stanpa darbahi gnas [parallel] chuchen dalhbab Blanpo-Khahbab [!] hgram [parallel] Chahi-G.yanra ZanZunlhahi ljons [parallel] Khyunrdzons Spumtho Rinchen Lhunpohi nos [parallel] dgoshdod hbyun [! a syllable is missing] rinchen gsergyi glin [parallel] ...

'Khye legs! Place where the feet of Rinchen Bzajpo have walked! Place where the teachings of Blobzaj Gragspa spread! [At] the shore [of] the slow flowing great river, Descending from an Oxen's Mouth (Sutlej), [In] the divine region [of] Chahi G.yajra [in] Zajzmj. [On] the side of the Precious Mountain of the Universe, the Height of the Spu [and] the castle of JKrymj[fiu)], [There is] the continent of gold [and] jewels [for whatever] needs and wishes arise, .' (de Rossi Filibeck 2007: p. 168, text 1300, fol. 1a.6-7, with a somewhat differing translation on p. 170; emphasis mine).

Ultimately, such descriptions may go back to Abhidharmic concepts (Vitali ibid.), and this might in part also explain the Indian references to a Suvarnabhu(mi) somewhere in the north. Consequently, this may well have given rise to a place name Mard 'Gold(en)', derived from the ZanZun epithet smar 'Golden' for the region around the Kailash-Sumeru (see above). (93) The specifying element yul in the Ladakhi/Tibetan name Maryul itself is indicative of a non-Tibetan origin of the name.

An association with the mythic land of Marutse/Marucina (g) could help to establish the original location of Maryul or Mard in the vicinity of Kulu. It might further throw some light on the processes underlying the apparent confusion or transfer. The 14th c. Padmabkahi thangyig mentions a land Marutse between 'Persia' and Sambhala, together with the land of Khotan or with the Hasa and Brusa and other north-western regions. The concept of this land or rather a cemetery where the deity Begtse, the guardian spirit of Khotan abides seems to go back to Indian sources, which speak of the Marucina, that is 'Chinese of the Deserts', a name that should have referred to the Mongolians. The land Maru, however, had been located in Chitral, from where it was shifted into the neighbourhood of Kulu (R.A. Stein 1959: 522, 531, n. 47 with further references). If the Indians were responsible for this transfer, this might have happened because the goal area was already known under the names Cina and/or Mahacina. If the Tibetans were responsible, the transfer might have been caused by the existence of a Smaryul or Mard. It is tempting to think of a connection between this Maru (or Rmaru in Tibetan etymologised spelling) and the place Miru (*Rmiru), (94) although the latter is not in the immediate neighbourhood of Kulu, but at least on the road from Kulu to the later Maryul.

A location of Maryul near or in Guge implies that the original designation had been shifted by the later self-styled 'descendants' of Dpalgyimgon from Spiti and its neighbourhood first to Upper Ladakh and, with the extension of their power over Lower Ladakh, also to the lower parts. The first shift might have taken place with the construction of Nyerma (Narma, var. Myarma) monastery, south-east of Thiktse. It's foundation is ascribed to Rinchen Bzajpo, who is said to have laid the fundaments of three monasteries, Tholing in Guge, Khachar in Purang, and Myarma in Maryul simultaneously on the same day (Tucci 1988: 63, Snellgrove & Skorupski 1980: 91, n. 21). If we understand 'on the same day' as a close temporal connection, the apparent correlation with the original districts of Mjahris should rather preclude the location of the original monastery in Ladakh. But it is perhaps even more likely that the pious legend was invented only to veil the fact that the three monasteries were not founded at about the same time--and not by the same person. The art historians, however, assert that the plan of Nyerma corresponds well with the latest phase of Tabo, to be dated to the end of the 10th c. (Christian Luczanits and Holger Neuwirth, p.c., cf. also Luczanits 2005: 70, n. 25).

Perhaps the tripartition of Mrahris is a fiction, as well, since only one generation later, all parts of Mrahris are again in the hand of a single branch, the two sons of the youngest (or middle) son: Srorre (Yeses Hod) and Hkhorre.

In this context, it is interesting to observe that the above-mentioned large-scale division of Mjahris into three skor.s seems to have been a projection, based on the division of its central element, the 'first skof or Mjahris Skorgsum. The internal division corresponds almost exactly to the 'modern' division into three skor.s, as observed by Strachey (1853: 4). The latter describes the three provinces of Western Tibet as "Mangyul, Khorsum, and Maryul". Maryul would comprise Ladakh and Baltistan. 'Khorsum', that is, Skorgsum, consists of "Ruduk on the N., Guge on the S.W., and Purang on the S.E." (p. 12). Strachey adds that "Ruduk is said to be encircled by lakes" (ibid.). Apparently, the original Maryul and Zajsdkar of the three segments of the first or central skor had been replaced by Guge and Rudok, respectively. The description given for Rudok, as being encircled by lakes, is the same that one can found in post-phyidar descriptions of Maryul as a district of lakes (see above). Quite interestingly, BRGY, in an attempt to etymologise the apparently alternative appellation bskorgsum, associates the surrounding lakes with Zanskar--in striking ignorance of the geographic facts:

sras gsum byujbah i cheba Dpallde Rigpamgongyis bzujbahi mjahris Majyulte I Zajsdkar mtshoyis bskorbadaj \ Bkrasis Ldemgongyis bzujbahi mjahris Stagmoste \ Spuhrej gajskyis bskorba I Ldegtsug Mgongyis bzujbahi mnahris ZanZunste I Guge g.yahyis bskorba bcas gsum

'The mjahris (95) seized by Dpallde Rigpamgon, the eldest of the three sons who had arisen, Majyul, that is,96 Zanskar was surrounded by lakes, the mjahris seized by Bkrasis Ldemgon, Stagmo, that is, Spuhrej was surrounded by glaciers, and the mjahris seized by Ldegtsug Mgon, Zanzun. that is, Guge was surrounded by slate [mountains].'

This would indicate that the designation Maryul was first shifted to Rudok where it acquired the attribute of being surrounded by lakes. At an intermediate stage, when Maryul was still known to have referred to Guge (Spiti) and Zanskar or when the geographic concept of Rudok and Maryul had completely merged, the attribute of the surrounding lakes was transferred from Rudok to Maryul and Zanskar. At a later time, the designation Maryul was transferred to Upper Ladakh before it was finally applied to the whole dominion of Ladakh, comprising also Baltistan, as soon and as long as this was part of the kingdom.

Another (alternative or additional) vector for the transfer could have led through Miru, which may or may not have been associated or even identical with the mythic Marucfiia. The chiefs of Rgya (the neighbouring village) seem to have had played a certain role in the power struggle within Upper Ladakh before and after the installation of the first dynasty (allegedly of the btsadpo or imperial lineage). Apparently they were never strong enough to establish themselves as rulers of (Upper) Ladakh, and their aspirations finally shattered when the Rnamrgyal dynasty from Lower Ladakh took over power. In local oral tradition the kings of the latter lineage, particularly Sejge Rnamrgyal, are referred to rather grudgingly as/mat gyafo/smad rgyalpo 'the lowland king(s)', signalling that they were usurpers with no right to ruler over Upper Ladakh. In the 17th c. the chief of Rgya told Azevedo that the Ladakhi king had dispossessed him (or his ancestors?) of the kingdom of Mariul (cf. Hugues 1996: 189), which means that he considered himself the righteous king of Maryul (wherever this may have been located).

The background of this claim is all but evident. One may speculate that the chiefs of Rgya were descendants (perhaps of a side-lineage) of the original rulers of Maryul (Spiti-Guge) who managed to politically survive at some border area, never forgetting their former rights. Alternatively, and this is what their Gandhara-Turkic dynastic name Gyesar suggests, they might have been newcomers in the area (like so many others), but somehow managed to transfer a prestigeous land title to their own property. In any case, it might have been just their position as rulers of a fictive or actually existing Maryul that enabled them to compete with the Ladakhi kings. If so, the Ladakhi kings, on their part, would have had a heightened interest in acquiring better legitimation by transferring the name Maryul to their own dominion.


2.3 Lata (Ladvags)

I have to leave the dynastic or political implications to the imagination of the reader, but before coming to the final section, it should be briefly noted that the name of Ladakh, spelled as Ladvags may go back merely to a learned etymology, which, unlike in the case of other place names, apparently left its impact on the modern pronunciation(s). (97) The Zanskari people as well as some elderly people from Lower Ladakh (and the same might be true for the elder generation in Nubra) refer by the designation Ladak(s) only to the central region around Leh. However, Leh and its surroundings are not especially famous for 'high' or 'wild passes'.

We have already seen in the citation above that Desideri knew the region under the name of Lhata-yul. Its capital appears under the very name Lhata (Filippi 1937: 84). Desideri arrived in Lhata in 1715, but the same name had been used almost two hundred years earlier: Mirza Haidar represents the name of one of the two Ladakhi chiefs as Lata Jughdan. This king is most probably identical with Lodros Chogdan (Blogros Mchogldan), the last Upper Ladakhi king before the takeover by the Lower Ladakhi lineage. The name might have appeared as/Lotso'/in Tibetan and as/Lotre/in a Kenhat dialect (although it is all but granted that the cluster gr had already turned into/ dr/), but even, or rather: particularly, a Mongolian speaker could not have misperceived the vowels in such a dimension. Since Lata is quite obviously identical with Desideri's Lhata, the form might represent an old tribal or local name, while the learned 'etymology' might have been generated on the basis of some West(ern) Tibetan dialects where the final cluster -gs is completely deleted (e.g. Zanskar and Spiti). I have so far been unable to trace the origins of this name.

Other variant spellings show that the scholarly form cannot be the original one. The form Bladvags that is alternatively in use among scholars is not corroborated by the Kenhat pronunciation, which is low tone in Gya. The 16th c. Debther marpo of Bsodnams Gragspa (ed. Tucci 1971, fol. 39a) has the form Lastag, while Ladag may be found in some recent Bonpo texts (Vitali 2008: 386, 387, 400). Interestingly, the Lastag of the Debther marpo, in accordance with the above-mentioned local tradition, is mentioned as only one among several kingdoms in the area: Sel in Majyul (the lineage of Rigpamgon), Nubraj (!), Globa, Lastag, and Zajsdkar. According to the enumeration, this entity might have been more towards the south-east than the apparently Central Ladakhi Majyul. Lower Ladakh and Purik are not mentioned.

3 Lesser and Greater Yangtong, ZanZun stod and smad, Lower and Upper Ladakh--a dialect boundary

We have seen that the two Chinese regions of Lesser and Greater Yangtong and the Old Tibetan regions ZanZun stod and smad match fairly well. One could thus expect that the internal boundary reflects some important cultural, economic, or geographic boundary. Since all the entities discussed overlap with at least parts of present-day Ladakh, I should try to look for this boundary on Ladakhi ground.

As far as the Indus valley is concerned, the boundary between Lower and Upper Ladakh in the self-understanding of the Ladakhis today is to be found in the sandy hills between Snyemo and Leh. It falls together with the dialect boundary between the two main dialect groups, Shamskat (the dialects of Lower Ladakh, including Nubra, Purik, and Balti) and Kenhat (the dialects of Upper Ladakh and Zanskar), and it constitutes an important boundary between different climatic and thus economic zones: predominant nomadic pasturage and wool production vs. predominant farming (with double crops) and horticulture (cf. Zeisler forthcoming b [section] 1). Corresponding dialectal boundaries are found between the Zanskar (Kenhat) dialects and Purik in the northwest, and the Sham dialects spoken along the lower course of the Zanskar and in the Markha valley. The dialect boundary in the Nubra valley is yet to be established, but it must lie somewhere between the confluence of Nubra and Shayok and the bend of the Upper Shayok.

The Kenhat dialects are very closely related to Tibetan varieties spoken in Himachal Pradesh (Piti, Tot, and Nyamkat), which probably points to a common history. They are somewhat less closely related to the varieties spoken in Western Tibet. Likewise the Changthang dialects (as far as I could survey them) share many features with the Kenhat dialects, but also some features with the Central Tibetan dialects. The region of Western Tibet, Himachal Pradesh, Upper Ladakh, and Zanskar seems to constitute a relatively well-definable linguistic area with gradual transitions from one dialect to the other.

Similarly, the Shamskat dialects show gradual transitions, the Purik dialect being intermediary between Balti and the Sham dialects. But they differ radically from the Kenhat dialects, particularly on the grammatical level, notwithstanding some convergences due to the long lasting contact situation. The most striking difference is the distinction of agent s and possessors in the Shamskat dialect (/amas/'by mother' vs./ame/'mother's') and their nondistinction in the Kenhat dialects (/ame/'of, by mother'). This is not merely a question of phonetics, because the Leh speakers may well articulate final -s in other contexts, and some Kenhat dialects make use of the syllabic form of the agent marker/-se/equally for possessor s (/amase/'of, by mother'). For other differences see Zeisler (forthcoming b).

The differences between the two dialect groups are thus not restricted to mere sound changes, but also pertain to elementary grammatical features. What is perhaps more important, is that there is no gradual shift observable: grammatically, the dialects either belong to the Kenhat group or not. To my opinion this grammatical dichotomy strongly speaks against a linguistic continuum between the two areas in prehistoric times and against a common linguistic inheritance.

To me it appears that this dialect boundary, being at the same time a boundary between different ecological zones and having served at least temporarily as a political boundary, reflects an older ethnic distinction, which seems to be still reflected in subtle cultural differences and mutual prejudices among the two groups. Even foreigners may realise different attitudes towards linguistic straightness ('honesty') or 'crookedness' (sarcasm and linguistic wit) between Kenhat and Shamskat speakers, another important difference, which would deserve further study, concerns different attitudes towards hunting, the people from Lower Ladakh apparently following (and trying to bypass) values of sharing known from Dard communities. Other significant differences may concern patterns of landholding and taxation: it is quite noteworthy that the people of Gya remember their lord Gyapa Co as a bloodsucker, claiming 50% taxes on all products (wool, meat, grain), whereas the people of Khalatse do not remember any kind of substantial taxation by the king or by the monasteries (KHAL).

The close affinity between the Kenhat dialects and the Western Tibetan varieties of Himachal Pradesh (e.g., Spiti) and the affinity with the Western Tibetan varieties in general points to a common (pre)-history. Upper Ladakh and Leh might thus well have belonged to the 'core area' of Zanzun. that is, what most Tibetans (except the Bonpos) and most Westerners accept as the only possible Zanzun. and what seems to be reflected by the Old Tibetan designation ZanZun smad as well as by the Chinese designation Greater Yangtong. Lower Ladakh and Baltistan, on the other hand, would have belonged to ZanZun stod, which from the common perspective appears as rather Un-ZanZun-ian, but which might have represented the 'real', that is, Iranian ZanZun from a Bonpo perspective.


As we all know, nothing is permanent. This holds all the more for human artefacts, including cultural artefacts, such as languages and geographical or ethnic identities, not to speak about political boundaries. Peoples, tribes, or clans, particularly the nomadic societies, may migrate over long distances, taking their original name along to their new abode. Their original neighbours, however, may continue to refer to their ancient abode by the old ethnic name and thus transfer the name to a new incoming group. Names could also be adopted for political reasons. Furthermore, the tribes of the past were never homogenous ethnic groups but rather conglomerate entities, the single elements of which could disband and regroup to form a new tribe or 'nation'. In doing so they would sooner or later identify themselves with the new group and slowly forget (or even actively suppress) all memories of a different identity (and similar processes may happen in the way they are perceived by their neighbours).

Folk-lore as well as official history and its artefacts, such as legends, chronicles, or inscriptions, will always stress the notion of identity. This makes the task of the historian to reconstruct the 'real' events almost impossible. While one always needs independent witnesses, most help comes from conflicting data, because this makes it possible to reconstruct, if not the events themselves, so at least the way the events were perceived or interpreted. I would think, the greatest fallacy in the case of conflicting data concerning peoples and their locations is to make a decision as to which data is correct and which one is to be discarded, rather than to recognise that the contradictions may point to fluent and inhomogeneous identities.

The peoples and the provinces of Western Tibet give the best proof. What is now called Ladakh (via an Urdu interpretation of the scholarly name Ladvags in the pronunciation of Purik as/Ladax/, hence the spelling Ladakh) is a misnomer, as probably all previous names were. Particularly the name Maryul turns out to be an old, but certainly not the ancient name of the area. Parts of Ladakh were, rightly or wrongly, identified with the Gold Clan, most probably of Saka origin, and with the Women's Dominion. On the other hand, most, if not all of Ladakh was part of a greater geographical or tribal unit, called Yangtong in Chinese sources, but ZanZun in Tibetan (and perhaps original ZanZuj-ian) sources.

Although ZanZun fell victim to the phyidar and post-phyidar reinvention of a glorious Bonpo past (much as Tibet fell victim to a reinvention of a glorious imperial and preimperial Buddhist past), it turns out that the Bonpo traditions, vague or exaggerated as they may be, contain a kernel of truth. The Old Tibetan administrators, at least, used the term ZanZun stod, rightly or wrongly, to refer to regions even beyond Ladakh and Baltistan. At least with respect to Lower Ladakh, perhaps also with respect to Baltistan this identification was also accepted by the neighbours, where these regions were known as parts of (Lesser) Yangtong. We can thus safely conclude that the overlapping entities Yangtong and Zanzun. while perhaps not fully congruent, did border on Bolor (Gilgit), and thus, in the perspective of the greater empires and the transmitted memory of their past glory, did at least border on Achaemenid or Sassanian Persia.


Albinia, Alice. 2008. Empires of the Indus. The story of a river. London: Murray.

Alram, Michael. 1996a. Die Geschichte Irans von den Achaimeniden bis zu den Arsaikiden (550 v. Chr.--224 n. Chr.). In: Wilfried Seipel (ed.) Weihrauch und Seide. Alte Kulturen an der Seidenstrafie. Wien: Kunsthistorisches Museum Wien: 73-98.

--. 1996b. Die Geschichte Ostirans von den Griechenkonigen in Baktrien und Indien bis zu den iranischen Hunnen (250 v. Chr.--700 n. Chr.). In: Wilfried Seipel (ed.) Weihrauch und Seide. Alte Kulturen an der Seidenstrasse. Wien: Kunsthistorisches Museum Wien: 119-140.

Bailey, Harold W. 1940. 'Ttagutta.' BSOAS 10.3: 599-605.

--. 1949. A Khotanese text concerning the Turks in Kantsou. Asia Major, New Series 1.1: 28-52.

--. 1964. Sri Visa Sura and the Ta-Uang. Asia Major, New Series 11.1: 1-26.

--. 1985. Indo-Scythian studies: being Khotanese texts. Vol. VII. Cambridge: Cambridge University Press.

Bailey, Thomas Graham. 1924. Grammar of the Shina (Sina) language, consisting of a full grammar, with texts and vocabularies of the main or Gilgit dialect and briefer grammars (with vocabularies and texts) of the Khohistani, Guresi and Drasi dialects. (Prize Publication Fund, 3.) London: Royal Asiatic Society.

Beal, Samuel (transl.) 1881-1884. Si-yu-ki - Buddhist Records of the Western World. London: Trubner. Reprint/New edition 1958-1963 under the title Travels of Hiouen-Thsang, Calcutta: Susil Gupta.

Beckwith, Christopher I. 1977. A study of the early medieval Chinese, Latin, and Tibetan historical sources on pre-imperial Tibet. PhD dissertation, Indiana University.

--. 1987. The Tibetan empire in Central Asia: a history of the struggle for great power among Tibetans, Turks, Arabs, and Chinese during the early Middle Ages. Princeton, NJ: Princeton University Press.

--. 1989. The location and population of Tibet according to early Islamic sources. Acta Orientalia Scientarum Hungaricae 43.2-3: 163-170.

Bellezza, John Vincent. 2008. Zhang zhung. Foundations of civilization in Tibet. A historical and ethnoarchaeological study of the monuments, rock art, texts and oral tradition of the ancient Tibetan upland. (Osterreichische Akademie der Wissenschaften, philosophischhistorische Klasse, Denkschriften, 368. Band, Beitrage zur Kultur- und Geistesgeschichte Asiens, 61.) Wien: Osterreichische Akademie der Wissenschaften.

Berger, Hermann. 1974. Das Yasin-Burushaski (Werchikwar). Grammatik, Texte, Worterbuch. (Neuindische Studien, 3.) Wiesbaden: Harrassowitz.

--. 1998. Die Burushaski-Sprache von Hunza und Nager. Band III:

Worterbuch Burushaski - Deutsch, Deutsch Burushaski. (Neuindische Studien, 13.) Wiesbaden: Harrassowitz.

Bielmeier, Roland. 1985. DasMdrchen vom Prinzen Cobzan. Eine tibetische Erzdhlung aus Baltistan. Text, Ubersetzung, Grammatik und westtibetisch vergleichendes Glossar. (Beitrage zur tibetischen Erzahlforschung, 6.) St. Augustin: VGH Wissenschaftsverlag.

Blama Btsanpo (a.k.a. Smingrol Nomonhan, f1839). Ed. Turrell V. Wylie. 1962. The Geography of Tibet according to the 'Dzam-gling-rgyasbshad. (Serie Orientale Roma, 25.) Roma: Istituto Italiano per il Medio ed Estremo Oriente.

Bsodnams Gragspa (1478-1554). Ed. Giuseppe Tucci. 1971. Debther marpo gsarma. [The new red register.] (Serie Orientale Roma, 24.) Roma: Istituto Italiano per il Medio ed Estremo Oriente.

Bsodnams Tshebrtan Yoseb Dgergan. 1976. Bladvags rgyalrabs hchimed gter [The eternal treasure, the genealogy of Ladakh]. Shrinagar: S.S. Gergan.

Chavannes, Edouard. 1900. Documents sur les Tou-kiue (Turcs) occidentaux: recueillis et commentes suivi de notes additionnelles. Paris: Librairie d'Amerique et d'Orient. Reprint 1969, Taipei: Ch'eng Wen.

Cunningham, Alexander. 1854. Ladak, physical, statistical and historical, with notices of the surrounding countries. London: W.H. Allen and Co. Reprint 1970, New Delhi: Sagar.

Denwood, Philip. 2008. The Tibetans in the west. Part I. Journal of Inner Asian Art & Archaeology 3: 7-21.

--. Forthcoming. The Tibetans in the west. Part II.

Dhar, Kashi Nath. 1994. Srivara's Zaina Rajatarangini. English translation and annotations. New Delhi: Indian Council of Historical Research and People's Publishing House.

Dongrub Lhargyal. 2000. Gsenrab Mibocheh i mdzad rnambsdusmah i skorglejba [Discussion of the collection of the great Gsenrab Mibo's deeds]. In: Henk Blezer (ed.) Religion and secular culture in Tibet: PIATS 2000: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, Leiden 2000. (Brill's Tibetan studies library, 2/2.) Leiden, Koln: Brill: 379-401.

Dotson, Brandon. 2009. The Old Tibetan Annals. An annotated translation of Tibet's first history. With an annotated cartographical documentation by Guntram Hazod. (Osterreichische Akademie der Wissenschaften, philosophisch-historische Klasse, Denkschriften, 381.) Wien: Osterreichische Akademie der Wissenschaften.

Durand, Algernon. 1899. The making of a frontier. London: Murray. Reprint 1974 with an introduction by Garry Alder. (Quellen zur Entdeckungsgeschichte und Geographie Asiens, 5.) Graz: Akademische Druck- und Verlagsanstalt.

Eberhard, Wolfram. 1942. Kultur und Siedlung der Randvolker Chinas. (T'oung Pao, suppl. au vol. 35.) Leiden: Brill.

Elias, Ney (ed.) and Edward Denison Ross (transl.). 1895. The Tarikh-i-Rashidi of Mirza Muhammad Haidar, Dughldt: a history of the Moghuls of Central Asia. London. Reprint 1973, Patna: Academia Asiatica.

Filippi, Filippo de. 1937. An account of Tibet. The travels of Ippolito Desideri 1712-1727. Rev. ed. London: Routledge. Reprint 2005, New Delhi: Rupa.

Fleet, John F. 1973. The topographical list of the Brihat-Samhita. [Posthumous] edited by Kalyan Kumar Dasgupta. Calcutta: Semushi.

Forsyth, Thomas Douglas, Sir. 1875. Report of a mission to Yarkund in 1873, under command of Sir T. D. Forsyth, K. C.S.I., C.B., Bengal Civil Service, with historical and geographical information regarding the possessions of the ameer of Yarkund. Vol. 1. Calcutta: Foreign Dept. Press.

Francke, August Hermann. 1907. A history of Western Tibet. One of the unknown Empires. London: S.W. Patridge & Co. Reprint 1995, New Delhi, Madras: Asian Educational Services.

--. 1908. Note on Mo-lo-so. The Journal of the Royal Asiatic Society of Great Britain and Ireland: 188-189.

--. 1912. The Tibetan alphabet. Epigraphia Indica 11.2: 266-272, plates I-VII.

--. 1914. Antiquities of Indian Tibet. Vol. I. Personal narrative. (Archaeological Survey of India, New Imperial Series, 38.) Calcutta: Superintendent Government Printing. Reprint 1992. New Delhi, Madras: Asian Educational Services.

--. 1926. Antiquities of Indian Tibet. Vol. II. The chronicles of Ladakh and minor chronicles. Text and translations, with notes and maps. (Archaeological Survey of India, New Imperial Series, 50.) Calcutta: Superintendent Government Printing. Reprint 1992. New Delhi, Madras: Asian Educational Services.

--. 1929. Notes on Khotan and Ladakh. (From a Tibetan point of view.) I-III. Indian Antiquary 58: 108-112, 147-152.

--. 1930. Notes on Khotan and Ladakh. (From a Tibetan point of view.) IV-VI. Indian Antiquary 59: 41-45, 65-72.

Fuchs, Walter. 1938. Huei-ch'ao's Pilgerreise durch Nordwest-Indien und Zentral-Asien um 726. Sonderausgabe aus den Sitzungsberichten der Preussischen Akademie der Wissenschaften, phil.-hist. Klasse XXX: 426-469.

Gnahgon Dkonmchog Tshesbrtan [=Tshebrtan]/ Gao Rui. 1995. Bodkyi brdarnijyigchabdamsbsgrigs/ Tufan wenxian xuandu [A selection of Tibetan documents in old orthography]. Pecin/ Beijing: Krungdbyan Mirigs Slobgrwa Chenmo/ Zhonyang minzu daxue chubanshe [Central Nationalities University].

Grenet, Frantz. 2002. Regional interaction in Central Asia and Northwest India in the Kidarite and Hephthalite periods. In: Nicholas Sims-Williams (ed.), Indo-Iranian languages and peoples. (Proceedings of the British Academy, 116.) Oxford: Oxford University Press: 203-224.

Haarh, Eric. 1969. The Yar-lun dynasty. A study with particular regard to the contribution by myths and legends to the history of Ancient Tibet and the origin and nature of its kings. Kobenhavn: Gad.

Hazod, Guntram. 2009. Imperial Central Tibet. An annotated cartographical survey of its territorial divisions and key political sites. In: Brandon Dotson, The Old Tibetan Annals. An annotated translation of Tibet's first history. With an annotated cartographical documentation by Guntram Hazod. (Osterreichische Akademie der Wissenschaften, philosophisch-historische Klasse, Denkschriften, 381.) Wien: Osterreichische Akademie der Wissenschaften: 161-231.

Heller, Amy. 2007. Preliminary Remarks on the manuscripts of Gnas gsar dgon pa in northern Dolpo (Nepal). In: Amy Heller and Giacomella Orofino (eds.), Discoveries in Western Tibet and the Western Himalaya: essays on history, literature, archaeology and art. PIATS 2003: Tibetan studies: Proceedings of the Tenth Seminar of the International Association for Tibetan Studies, Oxford. (Brill's Tibetan studies library 10,8.) Leiden, etc: Brill: 129-149.

Herrmann, Albert. 1910. Die alten Seidenstraien zwischen China und Syrien. Beitrage zur alten Geographie Asiens. I. Abteilung. (Quellen und Forschungen zur alten Geschichte und Geographie, 21.) Berlin: Weidmannsche Buchhandlung.

--. 1938. Das Land der Seide und Tibet im Lichte der Antike. (Quellen und Forschungen zur Geschichte der Geographie und Volkerkunde, 1.) Leipzig: Koehlers Antiquarium.

Hill, John E. 2004. The peoples of the West from the Weilue by Yu Huan. A third century Chinese account composed between 239 and 265 CE, quoted in zhuan 30 of the Sanguozhi, published in 429 CE. Draft version. http://depts. texts/weilue/weilue.html (last accessed: X/2009).

--. 2009. Through the jade gate to Rome: a study of the silk routes during the Later Han Dynasty 1st to 2nd centuries CE; an annotated translation of the chronicle on the 'Western Regions 'from the Hou Hanshu. Charleston, SC: BookSurge.

Hinuber, Oskar von. 2004. Die Palola Sdhis. Ihre Steininschriften, Inschriften auf Bronzen, Handschriftenkolophone und Schutzzauber. Materialien zur Geschichte von Gilgit und Chilas. (Antiquities of Northern Pakistan, Reports and Studies, 5.) Mainz: von Zabern.

Hoffmann, Helmut. 2003. Early and medieval Tibet. In: McKay (ed.), The history of Tibet. Vol. I.: The early period--to c. AD 850, the Yarlung dynasty. London: RoutledgeCurzon, 45-68. [Reprint from D. Sinor (ed.), The Cambridge history of early Inner Asia. Cambridge: Cambridge University Press, 1990: 371-399, 490-492.]

Hugues, Didier (ed). 1996. Les Portugais ou Tibet. Les premiers relations jesuites (1624-1635). 2nd ed., Paris: Editions Chandeigne.

Hummel, Siegbert. 1974. Materialien zu einem Worterbuch der Zan-zunSprache. Monumenta Serica 31: 488-520.

Iida, Hiroya. 2003. The study of Zhang Zhung: attempts at establishing a new discipline. (Paper of presentation given at the 10th International Conference of the International Association of Tibetan Studies, 6-12 September 2003, Oxford, UK.) htm, as accessed in IV/2010.

Jettmar, Karl. 1961. Ethnological research in Dardistan 1958. Preliminary report. Proceedings of the American Philosophical Society 105.1: 79-97.

--. 1977. Bolor -- A contribution to the political and ethnic geography of North Pakistan. Zentralasiatische Studien 11: 411-448.

--. 1980. Bolor -- Zum Stand des Problems. Zentralasiatische Studien 14.2: 115-132.

--. 2002a [1993]. The Patolas, their govenors and their successors. In: Karl Jettmar, Beyond the gorges of the Indus. Archaeology before excavation. Ed. by Ellen Kattner. Oxford: Oxford University Press: 116-156. [Slightly revised version of the original publication in Antiquities of Northern Pakistan. Vol. 2. Ed. by Karl Jettmar. Mainz: von Zabern.]

--. 2002b [1987]. The Suspended Crossing, where and why? In: Karl Jettmar, Beyond the gorges of the Indus. Archaeology before excavation. Ed. by Ellen Kattner. Oxford: Oxford University Press: 174-181. [Slightly revised version of the original publication in India and the ancient world: history, trade and culture before A.D. 650. P.H.L. Eggermont jubilee volume presented on the occasion of his 70. birthday. Ed. by Gilbert Pollet. Leuven: Katholieke Univ. Leuven.]

--. 2002c [1995]. The Dards and connected problems. Giuseppe Tucci's last contribution. In: Karl Jettmar, Beyond the gorges of the Indus. Archaeology before excavation. Ed. by Ellen Kattner. Oxford: Oxford University Press: 189-201. [Slightly revised version of the original publication in Giuseppe Tucci nel centenario della nascita: Roma, 7-8 giugno 1994. Ed. nel Beniamino Melasecchi. Roma: Istituto Italiano per il Medio ed Estremo Oriente.]

Kaltenmark, Maxime. 1993. Sky and earth, sun and moon, stars, mountains and rivers, and the cardinal points in Ancient China. In: Yves Bonnefoy (ed.), Asian mythologies. Transl. under the direction of Wendy Doniger by Gerald Honigsblum et al. Chicago, London: University of Chicago Press: 236-241.

Karmay, Samten G. 1998. A general introduction to the history and doctrines of Bon. In Samten G. Karmay, The arrow and the spindle. Studies in history, myths, rituals and beliefs in Tibet. Kathmandu: Mandala Book Point: 104-156.

Laufer, Berthold. 1914. Bird Divination among the Tibetans. (Notes on document Pelliot No. 3530, with a Study of Tibetan Phonology of the Ninth Century). T'oung Pao 15: 1-110.

--. 1918. Origin of Tibetan writing. Journal of the American Oriental Society 38: 34-46.

Leitner, Gottlieb William. 1889. The Hunza and Nagyr handbook. Being an introduction to a knowledge of the language, race, and countries of Hunza, Nagyr, and a part of Yasin. Calcutta.

--. 1890s. Dardistan in 1866, 1886, and 1893. Being an account of the history, religions, customs, legends, fables, and songs of Gilgit, Ghilas, Kandid (Gabrial) Yasin, Chitral, Hunza, Nagyr, and other parts of the Hindukush, as also a supplement to the second edition of the Hunza and Nagyr handbook, and an epitome ofpart III of the author's The languages and races of Dardistan. Woking: Oriental University Institute. Reprint 1978, New Delhi: Manjushri.

Lindegger, Peter. 1993. Griechische und romische Quellen zum peripheren Tibet. Teil III: Zeugnisse von den Alexanderhistorikern bis zur Spdtantike. (Die 'Seidenstrafien'). (Opuscula Tibetana, 22.) Rikon/ Zurich: Tibetan Monastic Institute.

Loewe Michael. 1982. Chinese ideas of life and death: faith, myth and reason in the Han period (202 BC-AD 220). London, Boston: Allen & Unwin. Reprint 2005 under the title Faith, Myth and Reason in the Han Period (202BC-AD220), Indianapolis: Hackett.

Luczanits, Christian. 2005. The early Buddhist heritage of Ladakh reconsidered. In: John Bray (ed.), Ladakhi histories: local and regional perspectives. (Brill's Tibetan Studies Library, 9.) Leiden etc.: Brill: 65-96.

Minorsky, Vladimir (ed.). 1937. Hudud-al-'Alam. 'The regions of the world'. A Persian geography 372 A.H.--982 A.D. Translated and explained by V. Minorsky. With the preface by V.V. Barthold. London: University Press. Reprint 1993. (Publications of the Institute for the History of 'Arabic-Islamic Science, Islamic Geography, 101.) Frankfurt am Main: Institute for the History of 'Arabic-Islamic Science at the Johann Wolfgang Goethe University.

Mkhaspa Ldehu (13th c.). Ed. Chabspel Tshebrtan Phuntshogs. 1987. RgyaBodkyi chosbyuj rgyaspa [The religious history of India and Tibet in extensio]. S.l. [Lhasa]: Bodljorjs midmarjs dpeskrunkhang [People of Tibet publishing house].

Mole, Gabriella.1970. The T'u-yu-hun from the northern Wei to the Five dynasties. (Serie Orientale Roma, 41.) Roma: Istituto Italiano per il Medio ed Estremo Oriente.

Nikitin, Alexander B. 1996: Iran zur Zeit der Sasaniden (224-652 n. Chr.). In: Wilfried Seipel (ed.) Weihrauch und Seide. Alte Kulturen an der Seidenstrasse. Wien: Kunsthistorisches Museum Wien: 99-107.

Pandit, Ranjit Sitaram. 1935. Kalhana's Rdjataranginl. The saga of the kings of Kasmir. Translated from the original Samskrta and entitled the River of Kings with an introduction, annotations, appendices, index. Allahabad: Indian Press. Reprint 1968, New Delhi: Sahitya Akademi.

Parker, E.H. 1905. China and the ancient Cabul valley. The English Historical Review 20.80: 625-636.

Peissel, Michel. 1984. The ants' gold. The discovery of the Greek El Dorado in the Himalayas. London: Harvill.

Pelliot, Paul. 1959. Notes on Marco Polo, I, ouvrage posthume. Paris: Adrien-Maisonneuve.

--. 1961. Histoire ancienne du Tibet. (Oeuvres Posthumes de Paul Pelliot, 5.) Paris: Adrien-Maisonneuve.

--. 1963. Notes on Marco Polo, II, ouvrage posthume. Paris: Adrien-Maisonneuve.

Petech, Luciano. 1939. A study on the chronicles of Ladakh (Indian Tibet). Calcutta: Calcutta Oriental Press.

--. 1947. Alcuni nomi geografi nel La dvags rgyal rabs. Rivista degli Studi Orientali 22: 82-91.

--. 1977. The kingdom of Ladakh c. 950-1842 A. D. (Serie Orientale Roma, 51.) Roma: Istituto Italiano per il Medio ed Estremo Oriente.

--. 1994. The Disintegration of the Tibetan Kingdom. In: Per Kvserne (ed.) Tibetan Studies: proceedings of the 6th seminar of the International Association for Tibetan Studies, Fagernes 1992. (Occasional papers; Institute for Comparative Research in Human culture 1,1.) Oslo: The Institute for Comparative Research in Human Culture.

--. 1998. Western Tibet: historical introduction. In: Deborah E. Klimburg Salter (ed.), Tabo, a lamp for the kingdom. Early Indo-Tibetan Buddhist art in the western Himalaya. New York: Thames and Hundson: 229-255.

Ramsay, Henry Lushington. 1890. Western Tibet: A practical dictionary of the language and customs of the districts included in the Laddk wazarat. Lahore: W. Ball & Co. Government Printers.

Richardson, Hugh E. 1971. Who was Yum-brtan? Dans: Ariane Macdonald (ed.), Etudes tibetaines dediees a la memoire de Marcelle Lalou. Paris: Librairie d'Amerique et d'Orient: 433-439.

Rockhill, William Woodville. 1891. The land of the lamas. Notes of a journey through China, Mongolia and Tibet. New York and London: The Century Company. Reprint 2005, Delhi: Winsome Books.

Rossi Filibeck, Elena de. 2007. Poetical prefaces of manuscripts from Western Tibet. In: Amy Heller and Giacomella Orofino (eds.), Discoveries in Western Tibet and the Western Himalaya: essays on history, literature, archaeology and art. PIATS 2003: Tibetan studies: Proceedings of the Tenth Seminar of the International Association for Tibetan Studies, Oxford. (Brill's Tibetan studies library 10,8.) Leiden, etc: 151-172.

Sahni, Pandit Daya Ram and August Herrmann Francke. 1908. References to the Bhottas or Bhauttas in the Rajatarangini of Kashmir. The Indian Antiquary: 181-192.

Sen Tansen. 2002. 'In search of longevity and good karma: Chinese diplomatic missions to Middle India in the seventh century.' Journal of World History 12.1: 1-28.

Snellgrove, David Llewellyn and Tadeusz Skorupski. 1980. The cultural heritage of Ladakh. Vol. 2. Zanskar and the cave temples of Ladakh. Warminster: Aris & Phillips.

Stang, Haakon. 1990. Arabic sources on the Amdo and a note on Gesar of Glin. Acta Orientalia Scientarum Hungaricae 44: 159-174.

Stein, Marc Aurel. 1900. Kalhana's Rdjataranginl: a chronicle of the kings of Kasmir, translated, with an introduction, commentary, & appendices. Vol. I, Introduction, Books I-VII. Vol. II, Book VIII, notes, geographical memoir, index, maps. Westminster: A. Constable & Co. Reprint 1961. Delhi etc.: Motilal Banarsidass.

--. 1932. On ancient tracks past the Pamirs. Himalayan Journal 4: 1-26.

--. 1942. From Swat to the Gorges of the Indus. The Geographical Journal 100. 2: 49-56.

Stein, Rolf Alfred. 1951. Mi-Nag et Si-Hia. Bulletin de l'Ecole Frangaise d'Extreme-Orient 44.1: 223-265.

--. 1959. Recherches sur l'epopee et le barde au Tibet. (Bibliotheque de l'Institut des Hautes Etudes Chinoises, 13.) Paris: Presses Universitaires de France.

--. 1961. Les tribus anciennes des marches sino-tibetaines. Legendes, classifications et histoire. (Bibliotheque de l'Institut des Hautes Etudes Chinoises, 15.) Paris: Presses Universitaires de France.

--. 1981. La civilisation tibetaine. Reed. rev. et augm. de l'ed. de 1962. Paris: Le Sycomore.

Strachey, Henry. 1853. Physical geography of Western Tibet. Journal of the Royal Geographical Society of London 23: 1-69.

Takeuchi, Tsuguhito. 2004. The Tibetan military system and its activities from Khotan to Lop-Nor. In: Susan Whitfield and Ursula Sims-Williams (eds.), The silk road. Trade, travel, war and faith. Chicago, Ill.: Serindica, 50-56.

Thomas, Frederick William. 1935. Tibetan literary texts and documents concerning Chinese Turkestan. Part I: Literary texts. (Oriental Translation Fund, New Series, 32.) London: The Royal Asiatic Society.

--. 1951. Tibetan literary texts and documents concerning Chinese Turkestan. Part II: Documents. (Oriental Translation Fund, New Series, 37.) London: The Royal Asiatic Society.

Thoussaint, Gustave-Charles. 1933. Le dict de Padma: Padma thang yig. Paris: Libraire Ernest Leroux. Reprint 1994: Le dict de Padma. Paris: Les Deux Oceans.

Tong Tao. 2008. The silk roads of the northern Tibetan plateau during the early middle ages (from the Han to the Tang dynasty) as reconstructed from archaeological and written sources. Tubingen: Univ. Diss. http://tobias-lib. 2008/3628, accessed XII/2008.

Tournadre, Nicholas and Sangda Dorje. 1998. Manuel de tibetain standard, langue et civilisation. Bodkyi spyiskad slobdeb [Textbook of colloquial Tibetan]. Paris: Langues & Mondes/L'Asiatheque.

Tucci, Giuseppe. 1956. Preliminary Report on two scientific expeditions in Nepal. (Serie Orientale Roma, 10, Materials for the study of Nepalese history and culture, 1.) Roma: Istituto Italiano per il Medio ed Estremo Oriente.

--. 1971a. Himalayan Cina. Dans: Ariane Macdonald (ed.), Etudes tibetaines dediees a la memoire de Marcelle Lalou. Paris: Librairie d'Amerique et d'Orient: 548-552.

--. 1971b. Travels of Tibetan pilgrims in the Swat valley. In: Giuseppe Tucci, Opera Minora II. (Studi orientali, 6.) Roma: Bardi: 367-418.

--. 1973. Tibet. Munchen: Nagel.

--. 1977. On Swat. The Dards and connected problems. East and West 27.1-2: 9-103.

--. 1988. Rin-chen-bzan-po and the renaissance of Buddhism in Tibet

around the millennium. First draft translation by Nancy Kipp Smith. (Satapitaka Series, Indo-Asian Literatures, 348.) New Delhi: Aditya Prakashan.

Uebach, Helga. 2003. On the Tibetan expansion from seventh to mideight centuries and the administration (kho) of the countries subdued. In: Alex McKay (ed.), Tibet and her neighbours. A History. London: Mayer, 21-27.

Uray, Geza. 1962. Old Tibetan dra-ma drans. Acta Orientalia Scientarum Hungaricae 14.2: 219-230.

--. 1990. The old name of Ladakh. Acta Orientalia Scientarum Hungaricae 44.1-2: 217-224.

Vigne, Godfrey Thomas. 1842. Travels in Kashmir, Ladak, Iskardo, the countries adjoining the mountain-course of the Indus, the Himalaya, north of the Panjab. With Map engraved by direction of the Hon. East India Company, and other illustrations. Vol. 2. London: Henry Colburn.

Vitali, Roberto. 1996. The Kingdoms of Pu.hrang. According to mNga'.ris rgyal.rabs by mkhan.chen Ngag.dbang Dharamsala: Tho ling gtsug lag khang lo cig stong 'khor ba'i rjes dran mdzad sgo'i go sgrig tshogs chung.

--. 2003. Tribes which populated the Tibetan plateau, as treated in the texts collectively called the Khungs chen po bzhi. Lungta 16: 37-63.

--. 2008. A tentative classification of the bya ru can kings of Zhang zhung. Revue d'Etudes Tibetaines 15: 379-420.

Watters, Thomas. 1904/1905. On Yuan Chwang's travels in India (A.D. 629 - 645). Ed. after his death by T.W. Rhys Davids and S.W. Bushell. Vol. I, II. London: Royal Asiatic Society. Reprint in one volume, 1973, Delhi: Munshiram Manoharlal.

Workman, William Hunter. 1905. From Srinagar to the Sources of the Chogo Lungma Glacier. The Geographical Journal 25.3: 245-265.

Yamaguchi, Zuiho. 1970. Su-p'i [TEXT NOT REPRODUCIBLE IN ASCII] or Sun-po [TEXT NOT REPRODUCIBLE IN ASCII]. A historicogeographical study on the relation between rTsan yul and Yan lag gsum pahi ru. Acta Asiatica 19: 97-133.

Zeisler, Bettina. Forthcoming a. Ethnic diversity, language contact, and the Old Tibetan lingua franca: early Tibetan history and the development of the modern Tibetan languages.

--. Forthcoming b. Kenhat, the dialects of Upper Ladakh and Zanskar. In: Mark Turin and Bettina Zeisler, eds., Himalayan Languages: Studies in phonology, semantics, morphology and syntax. Leiden etc.: Brill: 233-302.

--. Forthcoming c. For love of the word: a new translation of Pt 1287, the Old Tibetan Chronicle, chapter I. In: Imaeda, Yoshiro, Matthew T. Kapstein, and Tsuguhitu Takeuchi, eds., Studies on Old Tibetan Documents. (Old Tibetan Documents Online Monograph Series, 3.) Tokyo: Research Institute for Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.

Zhang, Yisun [Krarj Dbyisun] et. al. (eds.) 1993. Bod-Rgya tshigmdzod chenmo [The large Tibetan Chinese Dictionary]. Vol. 1-2. Pecin: Mirigs Dpeskrunkhar [Beijing: Nationalities Press].

Bettina Zeisler


(1) This article had originally been prepared as a chapter in a book on the ethnic composition of early Tibet and the history of Tibetan languages (Zeisler, forthcoming a, Chapter 2 [section] 1.2), where the present discussion will now be abridged. I am grateful to Roberto Vitali for the opportunity to present my findings on early Tibetan history to a more general audience, at the same time shifting the balance somewhat back towards linguistics, in the said publication. References to this publication will be kept at a minimum, but since it provides the necessary cultural and historical background, they cannot fully be avoided.

Since I am not a Sinologist I had to rely on the help of Mingya Liu and Thomas Preiswerk for getting the transcriptions right. To Thomas Preiswerk and John E. Hill I am specifically grateful for a lot of background information in matters of Chinese history. Thomas Preiswerk was particularly helpful in explaining Chinese characters, and all notes concerning Chinese renderings are due to him. Philip Denwood was kind enough to sent me his version of the story before publication (Denwood 2008, forthcoming), and I am much obliged, since his text served as a means of control against my own misconceptions as well as an incentive to improve my arguments wherever we disagree. Many thanks go also to John Bray for improving my English. I should further like to thank all those who, directly or indirectly, contributed to this article with comments or critics.

(2) Cf., e.g., the use of the term Da- and Siao-Yuezhi for the Tocharian tribes that moved to Bactria and those that remained in Gansu and in the Tarim Basin, or Qin 'China' vs. Da-Qin 'Ulterior China' for the Roman Orient (cf. R.A. Stein 1959: 304, n. 45).

(3) At a closer look, the name Bhautta (var. Bhatta or Bhatta), commonly taken to refer to the 'Tibetans' in Ladakh and Baltistan (M.A. Stein 1900 I, text ed., p. 47, note to i, 312-316; Pandit 1935 [1968]: 43, note to i, 312) cannot be identified with any precise location, more particularly, not with Western Tibet, Ladakh, or Baltistan (see Zeisler, forthcoming a, Chapter 4 [section]

(4) In antiquity, the Gilgit river was counted as the source of the Indus, and the section below the confluence of the Gilgit river with the Indus accordingly still bears the name Upper Indus Valley, a fact that has often been overlooked and has thus created quite some confusion. I will thus use the designations Gilgit-Indus and real Indus to discriminate between the ancient and modern understanding of Indus.

(5) Perhaps the western Sumpa: the Sobyi of the so-called prophesy of the Li country or the Supiya from the Central Asian Kharosthi documents (cf. Petech 1947: 87). Or perhaps even a variant form of Spiti (due to a palatalisation of the second syllable)? In his note 3, Fuchs conjectures the first syllable as H (thus Poboci, poboci), because he finds in the Yiqie jing yinyi the same name as WSM, (Poboci, but with a different tone: poboci), which he takes to be equivalent with Nepal. As Pelliot (1963: 709) states: "Although Nepal is well known in China as Nipo-lo (Nepala), we may perhaps suppose that Hui-ch'ao heard its Ti-betan name in Kashmir; but P'o-po-tz'u is uncertain, and So-p'o-tz'u (*Sa-pua-dz'i) is the reading of the only ancient Ms.; the phonetic equivalence of *B'ua-pua-dz'i is far from satisfactory."

It is not very likely that the Tibetan name of Nepal should have been current in Kashmir, and even if so, the Tibetan name for Nepal is Balyul or, as in Ladakh, Balpo.

(6) This is somewhat surprising in view of his earlier (Tucci 1956: 51-109) detailed discussion of the localisation Zajzuj.

(7) Occasionally, however, one may come across a commutation of Lesser and Greater Bolor.

(8) Meyers Konversationslexikon 1885-1892,, gives the distance between Srinagar and Gilgit as 22 daily marches. This may refer to a route along the Indus or, alternatively, to the route that went over Skardo. Gilgit could possibly also have been reached by a shortcut via Gurez and the valley of Astor, but even in this case, the indication of 7 days seems to be somewhat too optimistic.

(9) Chavannes takes this as an error for west. But Lesser Bolor, if consisting only of Gilgit, Hunza-Nagar, and Yasin, does not border on Swat, neither in the east nor in the west. The indication east could make sense, however, if Lesser Bolor also comprised (parts of) Chitral. This seems to be corroborated by the much later Mirza Haidar, see below.

(10) 'Greater Bulu (Baltistan) is also called Bulu; it is straight west of the Tubo (Tibetans); it touches on Lesser Bulu; in the west it is neighbouring the territory of Wuchang (Uddiyana) of Northern India. [...] It is subject to the Tubo. Since the period Wansuitongtian (696) until the period Kaiyuan (713-741), it sent three times ambassadors to render homage to the court. It is because of this that one conferred per diploma the royal title to the prince of this country, Sufushuolizhilini; after his death one conferred again per diploma the royal succession to Sulintuoyizhi (Surendraditya?); altogether the latter sent two times high dignitaries who brought the products of his country as tribute.

Lesser Bulu is more than 9000 li (away) from the capital; after 3000 li to the east, somewhat towards the south, one arrives at the camp of the Tubo zanpu (the Tibetan emperor); after 800 li on the eastern side [see note 9 above], this land touches upon Wuchang; 300 li to the south-east is Greater Bulu; 500 li to the south one finds Gushimi (Kashmir); 300 li to the north there is the town Suolei of (the country) Humi (Wakhan) ...'

(11) The name may perhaps refer to a place in present-day Gilgit district, mentioned in the Hatun inscription and reconstructed as Pusa[ta]ram by Gerard Fussman (Denwood 2008: 15).

(12) While Beal (1881-84: 177, n. 37) interprets the name Daliluo (W.-G. Ta-li-lo) as Darail (Darel), this is hardly possible, since the Darel and the Tangir river join the Indus above the Hanging Passages. Both valleys run parallel to the Swat valley, Tangir lying west of Darel. The Darel valley could thus hardly have served as seat for the rulers of Swat. Most probably Xuanzang came through the valley of Kandia.

(13) In fact, Mirza Haidar is quite explicit in separating Bolor and Balti, the latter being part of Tibet (cf. Elias 1895: 385, 405, particularly p. 417): "Tibet is bounded in the north, where it is called Balti, by Balur and Badakhshan."

(14) See note 17 below.

(15) Jettmar has to admit that Baltistan was not well enough connected with Gilgit to keep direct control over it or, the other way round, to be directly controlled by a ruler of Gilgit. Either Little Bolor had been a vassal of Greater Bolor or the other way round (2002a [1993]: 122, 125). Jettmar's suggestion that Hunza (Bruza) might have originally been an independent kingdom and had acquired the designation Lesser Bolor on being conquered by Greater Bolor (p. 126f.) is against the ancient conventions in the use of the terms Lesser and Greater: it is typically the added, the secondary territory that receives the designation Greater (see also p. 2 above).

Jettmar's scenario is also not corroborated by the earlier Chinese sources including Xuanzang's report, which present the kingdom as a single unit. Jettmar has to admit that several accounts of Bolor refer to a region that can only be identified with Hunza-Gilgit and the adjacent southern regions, but not with Baltistan, particularly the descriptions of the route by various Buddhist pilgrims, e.g. Zhimeng in 404 (p. 118) or Song Yun 518-522 (p. 119; here Bolor is described as a transit region to Swat). I cannot quite understand Jettmar's allegation that Xuanzang did not collect his information on the spot and remained rather vague (p. 119f.).

(16) This latter position, which is also in accordance with Faxian's description of reaching 'North India' from Khotan in a (south)-western direction (Beal 1881-84: 15), may refer, albeit anachronistically, to the Kusana Empire.

(17) The variation between Nangkot as in Cunningham (1854: 34; cf. Thomas 1935: 271, n. 5 Nanod) and Nanon might be due to copy errors. Cunningham also notes the name "Sagar-khoad" as a Ladakhi form given by Vigne (interpreted by Cunningham again as Skarkot 'starry building'). This might indicate that kot is the correct form, the origin of which we might perhaps have to seek in Central Asia; cf. also Skr. kot 'fort', which is commonly found in Indo-Aryan place names-Vigne (1842: 249), however, interprets the second part as khud 'valley'. Najgoj is also a place name in Amdo; it appears in an Old Tibetan report concerning the latter region (Thomas 1951: 146-148).

(18) It might be possible that the term Li referred here only to some peripheral dependent territories of Khotan rather than to the oasis itself.

(19) A further variant is found in Blama Btsanpo's Hdzamglijrgyasbsad (ed. Wylie 1962: 3/56) with Hchiba for Khotan and Blasa for Brusa.

(20) Actually, it seems to correspond to 'Innermost ZanZun'. This designation reflects a geography from a Iranian, Kashmirian, and even Chinese perspective, or a perspective that is related to Mt Meru as the cen-tral axis of the world, originally located in the Pamirs. It also corresponds to the distinction of Lesser and Greater Yangtong as well as to the much later distinctions of Little Tibet (Baltistan) and Greater Tibet (Ladakh). 'ZanZun-West' would be the exact counterpart of the Old Tibetan Zar/Zur/ stod, see below.

(21) Tazig or its etymologised form Staggzig (Tiger-Leopard) may have been originally be coined for the Arabs. But when used in the Bonpo sources, it seems to always refer to Iran and the whole of the Achae-menid empire, or, more specifically, to the Iranian borderlands along the Pamirs (Gilgit and Badakhshan) and the Dardic areas along the Hindukush and Indus (Chitral and Chilas). For the latter identification we can cite Buddhist and Bonpo sources alike: Chilas is identified with 'Persia' in Orgyanpa's itinerary (see also below), according to Mkhaspa Ldehu (ed. 1987: 222), the 'Persian' king is apparently a Kashmirian (at least he is called 'Morba of Kashmir'), speaking Shintrat, i.e. the Shina language, while the Bonpo text Drimedrtsabahirgyud (Dongrub Lha-rgyal 2000: 399) points to an identity of Holmofiuirings (in 'Persia') with Chitral (see Zeisler forthcoming a, Chapter 2 [section] 3.3.5).

(22) The identification of Cide with Spyiti is highly problematic, since the latter place name, modern pronunciation [piti], never underwent palatalisation, at least not in Western Tibet. It is thus extremely unlikely that the palatalised form ci should have occurred in an Old Tibetan document. Even if we admit that Dpahbo Gtsuglag wrote the name down from (Central Tibetan) hearsay, this would mean that the form Spyiti was not found in the documents at hand.

(23) According to Uebach (2003), mkhos were administrative measurements, carried out in a subjugated country mainly to integrate it into the Tibetan dominion. Unlike the phalos, they were not typically calls to arms, but in this case the military background is rather obvious.

(24) ltan means 'bale of goods, half load'. yo might perhaps be related to yobyad 'goods, necessities, furniture, household implements, etc', cf. Beckwith (1987: 42, n. 24). The latter, followed by Dotson (2009: 91), proposes to read the compound as 'plunder', but this seems to be contradicted by the statement concerning the following year, where a military campaign was conducted. It seems that first troops were levied, possibly close to the border region, with that threat prepared, a delegation went to the Drugu for negotiations, and as these turned out unsatisfactory, the threat was made reality in the following year. Against Beckwith (ibid), it is grammatically quite possible that Ltajyo refers to a place, but it is, of course, not possible that the place name contained the final -r as suggested by Thomas (1951: 268). According to Gnahgoj Dkonmchog Tshesbrtan (1995: 69, n. 25), the word ltar/yor would refer either to a cattle disease or to an internal turmoil, but these two meanings can hardly be reconciled with the fact that the word should be in the absolutive and followed by a movement verb.

(25) Dotson (2009: 96) translates this as 'the land of Kucha (Gu-zan-yul) in [Western] Turkestan'. Apart from the fact that Guzan might refer to Guchen in the eastern Tienshan north of Turfan (cf. Thomas 1951: 282(ff.)), rather than to Kucha, this is geographically absurd. Western Turkestan comprises the countries Kazakhstan, Kirghizstan, Tajikistan, Turkmenistan, and Uzbekistan. Like Guchen, Kucha lies at the northern rim of the Tarim Basin, and thus in the region commonly known as East(ern) or Chinese Turke-stan. In the period in question, the Western Turks, actually settled also in Eastern Turkestan, north of the Tienshan. They seem to have had dominated Kucha via Skaryab once in a while, but since Kucha was also claimed by the Chinese, and, in effect, recaptured several times, it seems to be unlikely that it was conceived of as a part of the Druguyul. Similarly, Guchen was at that time still under Chinese government (Thomas, ibid. p. 286). I should thus suggest analysing the compound as a tatpurusa compound, the first element of which consists of a dvandva compound: (Drugu & Guzan)-yul.

In reaction to the all too often rash conclusions of Thomas, R.A. Stein (1959: 293) rejects not only the former's localisation of Guzan, but suggests that the verb hdren could equally mean 'invite'. Unfortunately, R.A. Stein is mislead by his anger. First of all, he is unable to specify, where the 'land of Guzan' lay, to which the Drugu would have been invited. If Guzan lay in Tibet, we should have heard again of this province at some other occasion. If Guzan were not part of Tibet, why should the Drugu be invited to this province by a Tibetan minister? And why to a province (yul) and not to a particular place in this province? Secondly, while the meaning 'campaign, lead a campaign' is usually expressed by a collocation: dmag hdren (cf. OTA, l. 255), apparently referring to an ordinary troop, or drama hdren (OTA, l. 197), apparently referring to an elite (drama) expedition corps, the collocation is typically shortened in OTA. In 9 out of 11 cases (ll. 68, 94, 97, 127, 133, 135, 137, 274f., 276f.) we find only the place name plus drays, and except perhaps in our case (l. 97), it is not possible to read an invitee into the place name. There are similar examples in other Old Tibetan texts, but the most illustrative might be found in the Old Tibetan Chronicle, ll. 380-383:

Sbran Rgyalsgra Legzigskyis | stodphyogssu drar/ste | Muyur/su g.yul bzlognas | lungi rgyalpo Nui]kog manchad hbatjssu bsdus | Dbahs Btsanbzer Mdolodlastsogspas | mkhartshan yanchaddu dratjste | mkhar cupa brgyadphabnas | Dorpo bton te | hbatjssu bzesso |

'Rgyalsgra Legzigs from the Sbraj clan lead a campaign into the upper (i.e., western) regions, and after the enemies had been put to flight (lit. after the battle field had been overthrown) the kings of the valleys from Nujkog downwards were brought under control. Btsanbzer Mdolod from the Dbahs clan and others lead a campaign to [the region] from Mkhartshan upwards, and having overthrown eight fortified prefectures they made the Dorpo come forth and made them (lit. accepted them as) subjects.'

For a detailed discussion of the collocation and its elliptic form see Uray (1962).

(26) Unlike Denwood (ibid.), I do not take the impressing title of the Palola ruler to be a conclusive sign of power or military invulnerability. We can observe with the later petty chiefdoms of Ladakh a certain inflation of royal titles, which mirrored only the insignificance of their bearers. The Palola ruler, certainly not a major force, may well have been compelled to concede border areas or rights of passage to a power he could not contain.

(27) Muwer is the name of a king of ZanZun Khasgyor, a principality comprising Mt Tise, listed in Dkarru Grubdban Bstanhdzin Rinchen's Tisehi dkarchag and in Bstanhdzin Rnamdag's G.yundrun Bongyi bstan-pa-hi byunkhun nyunbsdu (Vitali 2008: 386, 387). A more prominent figure is Muwer btsanpo (ibid. p. 406 and passim).

(28) We do not know from which time exactly the Dpahbo Gtsuglag's description of ZanZun stod actually dates. Given the fact that no other region is mentioned between Tibet and the Western Turks, it appears to be most likely that the description belongs to a rather late date, when Baltistan and greater Bolor were already conquered.

(29) As an administrative term, skor is not attested in Old Tibetan source. Its usage may thus be related with the introduction of khriskor.s 'myriarchies' by the 13th c. Yuan dynasty.

(30) Yamaguchi (1970: 98) observes that the Yanglag (G)sumpahi ru "was established as though to hold Shanshunsmad between itself and Tibet proper." For him, Yanglag (g)sum-pahiru is to be translated as 'third additional horn' and has nothing to do with the Sumpa tribe(s). While one can never preclude that the majority of Tibetan scholars all fell prey to the same error, most probably because the Tibetans themselves identify the (G)sum-pahiru or Sumru with the Sumpa, it would be even more difficult to envisage ZanZun smad, being little more than the Guge province plus some areas along the upper Brahmaputra, as lying between some Tibetan secondary entity (presumably in the north) and Central Tibet, both east of it. Personally, while ready to sympathising with the nonconformist, I find difficulties in putting much weight on Yamaguchi's argument, as his paper is full of minor and major mistakes and contradictions: his further discussion reveals that many, if not most of the place names associated with the Sumru should actually lie far in the east, in the Khams region, while the Sumpa or rather Supi tribes would have dwelled in the west (p. 127, with n. 114). Particularly, the Supi would have been, according to Chinese sources, located west, the Sumru east of the Tomi, the Tomi being a tribe in eastern Tibet. But then the same Chinese sources are quoted as stating that the (allegedly western) Supi (or who else?) were named Sunba after the Tibetan conquest, where Sunba should be the Chinese transcription of Tibetan Sumpa. In any case, if the Sumru was established mainly in eastern Tibet, then how could a place supposed to be identical with presentday Guge lie in between the Sumru and Central Tibet?

(31) The distance between Khotan and Rudok or Leh is ca. 300 km as the crow flies, while the actual tracks have an approximate length of between 700 and 1100 km, see also below. The statement reminds me of a description of a route through the Pamirs: M.A. Stein (1932: 14, 20) mentions a passage in uninhabited wilderness, which may need up to forty days to cross, not so much due to its length, as due to the difficulty of the rugged terrain, particularly when covered by snow.

(32) The Tang annals speak more precisely first of the rendering of homage to the Tibetan Emperor (Pelliot 1961: 3 "Les royaumes voisins, comme celui de Yang-t'ong et les tribus des K'iang") and only later of the annexation of several territories (Pelliot 1961: 9: "du Yang-t'ong, des Tang-hiang et des divers K'iang") or of the submission of various Qiang tribes associated with Yangtong (Pelliot 1961: 89: "les K'iang Yang-t'ong et Tang-hiang"), leaving some room for interpretation, despite the use of entierement or completement.

(33) The ZanZun script, called Smar is most probably a phyidar invention, based on the 10th or 11th Landza and Wartu(la) scripts originating from Nepal. The alleged 'Persian' script, called Spuns could possibly refer to a script used in the Pamirs, although it appears somewhat strange, that no specimen of this script should have survived. However, I do not want to preclude the possibility that the Bonpos actually refer to the Brahmi manuscripts of Gilgit or to Karosthi inscriptions in the same area and in Khotan.

(34) The specimen from the Gnammtsho, given on p. 188, fig. 353 as an "archaic inscription" is definitively neither Karosthi nor Brahmi, but a strange type of Tibetan with some parallels to the dbumed letters of a 10th to 12th c. manuscript of Kyelang (a semi-circle-like ba and a similar ca with two 'horns' on top) as listed by Francke (1912: plate I-VII).

(35) Here, the grapes were consumed or sold only as fruits, sometimes also as dried fruits. The last plant was only recently removed to make place for a new house. According to the people of Khalatse, it is too cold for grapes further up the river, that is, in Nurla or Saspol, but it is quite possible, that more than thousand years ago, in a somewhat warmer climate, the cultivation area extended up to Bazgo and Snemo.

(36) Commercial grapes are currently grown in Kinnaur, but like apple cultivation, this might be a recent development. There is, however, also a wild variety (Anju Saxena, p.c.). As Anju Saxena further explains (e-mail communication 18.02.2010), "it seems that grapes have been grown in Kinnaur for a rather long time. This is mentioned abundantly in Alexander Gerard's works. He also provides the name of this fruit in different 'dialects' of Kinnaur."

(37) Or perhaps even south of the Hindukush, which would lead us to Uddiyana/Swat. The name Congling usually appears with reference to the Pamirs, particularly the mountain tract west and south-west of Kashgar, but it is possible that parts of the adjacent chains could be subsumed under this name. Parker (1905: 631) takes the Congling range to be idendentical with the Hindukush.

(38) The historian Zhang Shoujie gives the distance somewhat more precisely as 2700 li (ca. 870 km; Pelliot 1963: 698).

(39) I will thus use the terms (more) eastern Dongnuguo and (more) western Dongnuguo in order to distinguish the two elements of the Eastern Women's Dominion.

(40) The four rivers flowing from the Meru-Kailash are similarly said to encircle lake Anavatapta-Mapham seven times.

(41) Commonly identified with lake Manasarowar/Mapham near Mt Kailash!

(42) The notion 'weak' means that the water was perceived as being not able to carry even a swan's down (Pelliot 1963: 697). Nothing could float on its surface, everything would immediately sink. If this rather fanciful idea had a natural basis (as most such ascriptions have), the only realistic explanation I can imagine is that the river in question had a very strong undercurrent or strong turbulences leading to a downward pull.

(43) The Kunlun is elsewhere conceptualised as an axis mundis, although rather an incomplete one, since all five rivers flow into a southern direction, cf. Paul Kekai Manansala,, accessed III/2010. Kaltenmark (1993: 238) thinks that "the mythical mountain of the west" called Kunlun "has nothing to do with the mountain known by geographers" and "is a purely legendary mountain". This is not exactly true. All five rivers that are said to flow from the purely legendary mountain exist in reality, and at least three of them can be linked up with the Kunlun of the geographers. Even the legendary aspects can be localised. Apart from the encircling river (see note 40 above), the mountain is described as having four gates opening in the four directions and a walled city of nine tiers (ibid.). The former feature corresponds to the four rivers of Mt Meru (or Sumeru), flowing in the four cardinal directions. While the 'original' Mt Meru cannot be traced, this description is approximately true for the Pamirs, the original Roof of the World, with the Tarim, Gilgit-Indus, Oxus, and Jaxartes flowing roughly towards the cardinal directions, but it does not match any other river system, particularly not that of the Kailash. The latter feature corresponds to the nine tiers of the mountain in Holmofiujrhjs on which Gshenrab Mibo descends. One of the later Chinese sources explicitly connects the Kunlun with Mt Meru, stating that "in the western countries, Kunlun is given the name of Sumeru" (ibid. p. 239).

(44) For the localisation of the Nuguo of the Tang historiographies it plays no role that the earliest notes con-cerning the Ruoshui could not have referred to Bolor (cf. Chavannes, ibid. p. 154 n. 1 k). It is quite possible that the mythology of a 'weak' river reached China from the west together with the mythology of libertine queens, but for lack of better geographical knowledge the respective regions were located close to China.

(45) Here, the word rma does not mean 'peacock' as commonly assumed, but is either a dialectal form for dmar, ZanZun mar or smar 'golden' or refers to the Rma tribe (see Zeisler forthcoming a, Chapter 4 [section] 2.4).

(46) See Haarh (1969: 347-348) for the various parallel quotations.

(47) But cf. also Jettmar (1961: 85f.) for two types of tower-like structures in the Dardic area, one of which is from a quite recent date, but might still follow an ancient pattern.

(48) Laufer (1918: 41, n. 26) strictly opposes the possibility that an Indic script could have been found on (Western) Tibetan ground. The information of Xuanzang would have been "misplaced and smuggled into the New History of the T'ang". But Xuanzang is silent about the script, and what might have been smuggled, namely into Xuanzangs report, is the identification of Suvarnagotra (I, see next section) with the Nuguo. Laufer's somewhat harsh reaction is motivated by Francke's (1912: 269) quite simplistic identification of the Nuguo with Guge.

(49) Cf. the coins presented by Alram (1996a): p. 93, fig. 73: Mithradates II, p. 95, fig. 75: Dareios (?), fig. 76: Artabanos II, p. 96, fig. 79: KamnaskiresOrodes III, p. 97, fig. 83: Ardashir V (I); as well as those presented by Nikitin (1996): p. 100, fig. 85: Ardashir I, fig. 86: Shapur I, fig 87: Ohrmazd I, p. 101, fig. 89: Narseh. Quite often, particularly in the later epoch, the tresses are also formed into a one or two hair balls. The hair style is copied by some of the Kusana kings, the Kidarites, and Hephthalites, cf. Alram (1996b): p. 128, fig. 109: Gondophares, p. 137, fig. 125: Vahram Kushanshah, and with the hair ball: p. 137, fig. 127: Kidarites, p. 139, fig. 132: Hephthalites. The Huna, which seem to be related to, if not identical with, the Hephthalites, were located in the northern division of the Brhatsamhitd (Fleet 1973: 11), and more precisely at the Kailash in the Harsacarita (chapter v).

(50) Herrmann (1938: 12), however, does not talk about Suvarnagotra, at all. What he does, is to identify Kargil with the place where the legendary ant's gold were produced. There is no reason for doing so, except that August Herrmann Francke, in a tone of utter conviction, had projected all references to neighbouring and not-so-neighbouring regions onto Ladakh, so also in the case of the gold digging ants (1907: 12f.). Herrmann (1938: 22f.) has well noted that Herodotus refers to a place in Afghanistan, but takes this as a gross misconception. There are many places in the Hindukush, the Pamirs, the Himalaya, and the Tibetan plateau where gold could have been dug out of the ground or washed out of the rivers, and much better candidates than just Kargil. Unfortunately, Francke is not the kind of authority that one could rely upon. Beckwith's rash identification is not much better: apart from the word gold, there is not much reason to associate the Gold Clan with the legendary gold-digging ants. By the same kind of reasoning, one should better locate the Gold Clan in the Altai, veritable 'Gold Mountains'.

(51) It is quite apparent that the geographical coordinates must have been confounded, and that all non-Indian countries and peoples of the three northern sections are to be sought in the north-western division, e.g. the Hematala (referring to the Hephthalites in western Badakhshan, cf. Grenet 2002: 214 with further references) of the northern division, or at least in the northern section as the Kulkuta, which are found in the north-western and north-eastern division, instead.

(52) The name is transcribed as subanajudaluo ([TEXT NOT REPRODUCIBLE IN ASCII]) by Hye-cho (Fuchs 1938: 439) and as sufalanaqudaluo by Xuanzang (Fuchs, ibid, n. 4 gives the transcription as sudailanaqudaluo [TEXT NOT REPRODUCIBLE IN ASCII], but W dai is apparently a mere mistake for the quite similar [TEXT NOT REPRODUCIBLE IN ASCII] fa. Pelliot 1963: 703 has sufalanaqujuluo [TEXT NOT REPRODUCIBLE IN ASCII], emended as [TEXT NOT REPRODUCIBLE IN ASCII], where [TEXT NOT REPRODUCIBLE IN ASCII] ju is likewise a mere mistake for BB da).

(53) I was able to locate a Brahmapura in present-day Uttarakand. Denwood (2008: 8) prefers Pelliot's (1963: 699) reconstruction *Balahimapura or *Varahimapura, which he suggests to identify with modern Barkot north of Har(i)dwar, which is likewise in Uttarakand. Against this, Laxman Thakur (p.c.), opposes that Brahmapura must be identical with Brahmaur in Himachalpradesh. According to him, there is a foundation inscription clearly indicating the name Brahmapura, dating, however, from as late as ca. 680 CE. It would thus be later than Xuanzang's visit. The distance between the two locations in question would be around 360 km, and Jalamdhara would be situated between the two. From this perspective, it would seem unlikely that the more eastern region (Suvarnagotra in relation to Moluosuo) would be reached via the more western location (Brahmaur in relation to Jalamdhara), but in the end, for the purpose of this paper, there would not be much difference, all tracks leading through Lahul to Spiti or Guge.

(54) A reproduction can be found in the Digital Archive of Toyo Bunko Rare Books under http://bameduniya.tripod. com/ moreroutes.html, as accessed II/2009. It gives the distance from Leh to Ilchi and Kiria over the Pangong range, along the Karakash river as 700 km and 891 km. The distances are said to be taken "from an 1897 book called 'Routes in Jammu and Kashmir' written by Maj. Genl Le Marquis de Bourbel, also former Chief Engineer of the Jammu & Kashmir State."

(55) The designation Gserrigs is found so far only in one of the Khotanese 'prophecies', the Vimalaprabhdpariprcchd (see below). While Hazod (2009: 169) states "Gserrigs [it is registered as part of Tibet c. 726, but apparently was not included in the Zhang-zhung stong-sde structure as it was probably never part of Zhang-zhung]", he does not give any reference for this 'registration', nor does he specify his dating. I am not able to find any only approximately similar term in the Old Tibetan indices, so far published. Gser-rabs is mentioned by Hoffmann (2003 [1990]: 48), although without indication of sources. Quite apparently this form goes back to a mere conjecture, cf. Thomas (1935: 152, n. 1).

(56) There are, of course, two closely related (if not identical) peoples that combined the anthropological feature of 'golden' hair with a tremendous mastery of working with gold: the Scythians and the Sakas. Their fame might have been transferred to other groups when the Sakas moved to India. Or certain regions on a common trade route to the Scythians or Sakas might have been wrongly associated with them. Something like this had happened with the designation Khrom for Rome when it was applied to mere 'market places' on the Rome route (cf. Stang 1990: 172).

(57) Some of the accounts would rather point to the conflicts between Lesser and Greater Bolor and the imminent Chinese Hindukush expedition in 747. However, the text is completely silent about the dramatic events around 740: the expulsion of the Buddhist community from Khotan, their stay in Tibet, their expulsion from Tibet, their flight to Gandhara, their mutual massacre, which figure so prominently in all other prophecies, which means that it must have been compiled before this date. It is thus impossible that the narrative refers to the period after the decline of the Tibetan empire and before Khotan became a Muslim state, as Jettmar (2002a [1993]: 149) suggests. Moreover, at that late period there would no longer have been any Haza and Sumpa to threaten Khotan.

(58) Gdorjdmar; most probably this was nothing than tribal designation: Red Ldo) (the name appears also in the form Gdo), Hdo), Sdo), R.A. Stein 1961: 18, 31ff.), and had nothing to do with the application of red colour on the face. This was performed by several Central Asian groups, e.g., by the Haza, and hence not a distinctive feature. The colour term identifies a moiety or smaller segment. Later, the White Ldo) was the dominant clan of the Minag and of Glirj in Khams (cf. R.A. Stein 1959 passim).

(59) Nevertheless, the Khapulu area is still referred to in Skardo folklore as Liyul (CDTDn), pointing to the fact that parts of Baltistan had been associated in one way or another with Khotan.

(60) The prophecy mentions several times that the Skarrdo king should appropriate some of the gold for his own expenditure.

(61) According to the Weishu and the Beishi, the Tuyuhun in 445, being attacked by the Northern Wei, turned westward, overran Khotan, killed its king and thousands of Khotanese people, and finally even subdued Kashmir. In this connection they sent an ambassador to Emperor Wendi of the Liu Song with golden wine vases form the Women's Dominion (Tong Tao 2008: 25f.). They seem to have had a greater impact in the region. In the Padma thayyig Hasa (!) and Brusa are mentioned together in the neighbourhood of Khotan (cf. Toussaint 1933 [1994]: 105, 215, 410), and similarly, in Lower Ladakh, the Haza (/Asa/) are always mentioned in one breath with the Burusho and are thus conceived of as their immediate neighbours.

(62) Most probably, this refers to the practise to decorate the face with red circles, a custom common among several Central Asian tribes, not only the Tibetans.

(63) Is the Gold Race Country perhaps occupied by the Tibetans? And could it thus be identified with Ladakh? But the undercover nature could also have something to do with the rivalries among the rulers of Bolor and, perhaps, Khotan or other neighbours. There is also the possibility that the affair had to be kept secret, simply in order to keep the "Wild Men", the robbers and thieves (an allusion to some unruly, anarchic tribes) out of the game. The 'Wild Men', their changing loyalties, and even fights amongst them are mentioned several times, although mostly in connection with the side-narration concerning the Do-Good and the Do-Evil (fol. 384b ff.). These two designations are clearly borrowed from a common fairy tale, and it is questionable, whether this part has any historical value at all, except demonstrating indirectly that Skarrdo and Li must have had a long history of intense enmity, despite the official claim of Buddhist and family solidarity dominating the text.

(64) See also Jettmar (1961: 89) for ancient sexual rituals or 'black masses' in the context of the worship of female mountain deities, peris, among the Dards; one may perhaps also think of Sivaitic cults and human sacrifices to the Devi A prominent case is mentioned in Kalhana's Rdjataranginl, 331-335, were a sorceress, that is, tribal priestess managed to sacrifice king Baka, Mihirakula's son and successor, together with almost all of his male descendants to the circle of goddesses (M.A. Stein 1900: 49).

(65) At least as a country where financial exploitation can be veiled with Buddhist arguments.

(66) Note that the BRGY defines Bruza ~ Brusa as 'an old name for a part of Uigur Xinjang': Brusa | Sinca) Yugu rigs ra)skyo) ljo)skho)skyi lu)pa-ziggi mi) rni).

(67) This is quite at variance with Denwood's (2008: 9) conclusion "that Suvarnagotra, the Eastern Women's Kingdom and Zhangzhung/Yangtong were separate entities". According to him, we would have the following separate entities from the west to the east: besides Bolor, and Baltistan, 1. Ladakh = Moluosuo/Sanbohe, 2. Rudok = Suvarnagotra, 3. ZaqZuq = Lesser/Greater Yangtong, and 4. a Kingdom of Women in Eastern Tibet. According to his own arguments, the Gserrigs of the Vimalaprabhdpariprcchd could not have referred to a place all to close to Zaqzuq, which was already occupied by the Tibetans. I would also think that Rudok could not have been tributary to Skardo with an independent Ladakh in between. Furthermore, one cannot simply dismiss Hyecho's location of Yangtong as neighbouring Greater Bolor or his 'omission' of a separate Ladakh and Baltistan as mere error. After all, Hyecho is the latest source, and these areas are correctly described as being occupied by the Tibetans. Nor can the distances be ignored. Denwood argues that the designation Khotan may have referred to a larger area extending eastwards, not just to the city of Khotan, so that Rudok could have bordered on it. But he remains silent about the fact that the distance to Suvarnagotra had also been calculated from western-most rim of Khotan, Karghalik.

Denwood is, of course, right when he separates the more eastern Nuguo from Yangtong and Suvarnagotra. The difference in our interpretations of the same facts or sources seems to be mainly due to his taking Sanbohe and Maryul as doubtlessly identifiable with present-day Ladakh and thus as a fixed corner stone. A further difference is his attempt to delimit discrete ethnic or political entities, dismissing all conflicting statements as errors or confusions, while I am ready to accept them as evidence for the non-discreteness of the entities in question. I would certainly not want to claim that Yangtong, Suvarnagotra, and Nuguo were congruent, but they might have been overlapping considerably.

(68) Given the fact that the name is known in India at least since the early 2nd c. CE. it might reflect the southward migration of one of the Saka tribes that was triggered off in the late 2nd c. BCE by the Yuezhi migration to Baktria.

(69) It might be worth mentioning the place Yangthang in Lower Ladakh, not very far from Tijmozgaj where the Lower Ladakhi kingdom once had its base. There is likewise a village Yangthang on the Spiti river in Kinnaur. If this is not just a linguistic accident (as is probably the case with the Yangthang in western Sikkim), this might well indicate that the boundaries of the Yangtong culture were absolutely fuzzy and that there was a more intricate connection between Ladakh and ZanZun Proper.

(70) Hunza and Nagar as well as the Indus around Chilas are the only narrow valleys that, in fact, lie in a truly east-west direction. And this applies in general for the two Bolors. Greater Bolor is described by Xuanzang as being "long from east to west, and narrow from north to south" (Beal 1881-84: 178), which is quite in accordance with Durand's (1899: 199) measurement for 'Dardistan', i.e. the two Bolors together as extending "roughly two hundred and fifty miles from east to west, and a hundred and fifty from north to south". The upper reaches of the Sutlej river likewise run in an almost exactly east-west direction, but I find difficulties with a description of Guge as being "contracted from north to south".

(71) This even more so, if the Congling range comprised also the Hindukush, see note 37 above.

(72) The Xianbi (also spelled Xianbei) are typically counted as a proto-Mongolian tribal confederation, but 'proto-Mongolian' may mean nothing more that they lived for some time on Mongolian territory. They may or may not have been composed of quite different ethnic groups. Cf. Michael Weiers' script 4 Abrisse zur Geschichte innerasiatischer Volker: 'Turken, Protomongolen, und Prototibeter im Osten' and Xianbi.PDF, last accessed II/2010.

(73) As a rather minor trait, continued also among the Tibetans, one may also mention the extremely long sleeves (Pelliot, ibid., p. 700), borrowed from, or shared with, the Sogdians (Tong Tao 2008: 188).

(74) The alternative reading Moluopo ([TEXT NOT REPRODUCIBLE IN ASCII]), however, is based on a mere conjecture by Cunningham, who wanted an equivalent for Tib. dmarpo 'red' (cf. Uray 1990: 217). The letters suo and po are apparently quite similar, inviting misinterpretations by western scholars, as could be observed also in the case of Suoboci vs. Poboci, cf. n. 5 above.

(75) In his highly speculative paper, Francke (1930: 67, 71), following Cunningham (1854: 4), takes San-bo-he for the Ladakhi name of the Indus, Gtsayspo, which already at his time was pronounced as/L. tsajspo/(~/R. tsajspo/) in Leh and Lower Ladakh and as/Sagpo/or/Safo/in Upper Ladakh. First of all, this idea does not account for the third syllable [TEXT NOT REPRODUCIBLE IN ASCII] he, which has only a phonetic value (based, in fact, on the word [TEXT NOT REPRODUCIBLE IN ASCII] he for 'river'). Secondly, when arguing that this name did not refer to the Tibetan Rtsayspo, Francke's main argument is that the initial consonant is still pronounced in Ladakh but not in Central Tibet. But if the retention of the initial should be the deciding fact, then it should have been reflected also in the Chinese transcription. Francke further fully ignores the fact that the Brahmaputra as well as the western province through which it flows have been spelled Rtsays(po) in Old Tibetan, quite obviously reflecting a pronunciation similar to that in modern Ladakhi.

It is likewise not very feasible that Sanbohe should have been the equivalent of a Kulu name and Moluosuo that of a Kashmirian name of one and the same ethnic or geographical entity as Pelliot (1963: 707) suggests. It is far more likely that Xuanzang or the interpolator mixed up the hearssay concerning two countries, roughly located in the same direction (from the perspective of North India), and reachable only on difficult routes.

(76) The inscription mentions a king (nrpatti) from a so far unidentified tribe or location Kunjana or Kuljana, governor (visayapatti) of Sam(bhu/cu)t(anna/ma). The spelling Kunjdna is fairly close to the Burushaski name for Hunza: Kanjuut. This name appears also as Kdncudlya in the Hatun inscription and as Kakvd or Kajvd <Ka(m)jud (gen. pl.) <Ka(m)jua with omitted dental in a Saka itinerary (v. Hiniiber 2004: 75, n. 73). The vowel metathesis of the inscription is in agreement with the form 'Hunza' itself.

(77) A li measured 415.8 m in Han times, which would yield 830 km. In Tang times, when the historical works relevant for our topic were written, it measured ca. 323 m, cf. Changingvalues, as accessed in II/2009. The same entry, contains a remark that during the dynasties of the Wei and the Western Qin, the li was considerably shorter, namely merely 77 m. Thomas Preiswerk and John E. Hill both think that this is unbelievable (p.c.), and the latter has thus added a "citation needed" tag in I/2010. 2000 li of 77m would yield ca. 155 km, which is again much to short for the distance to Ladakh, but may nevertheless correspond to the almost equal distance between Kulu and Lahul (see below).

(78) gives the distance as 358 km, http:// gives the distance as 273 km, http://www.indiaepostoffice. com/kullumanali/link7.html gives the distance between Leh and Manali as 475 km, the distance from Manali to Kyelong as 115 km, the difference being 360 km, ManaliHighway gives the distance from Leh to Manali as 479 km, from which 115, 120, or, according to Manalito-Keylong.html even 129 km have to be subtracted for the distance between Manali and Kyelong. All sites have been last accessed in II/2010.

(79) Tabo, which lies on a route towards the Changthang, is 284 km from Kulu on the vehicular roads. One could thus have continued for another 45 (or 54) km from Tabo. The first value is calculated on the base of 160 km corresponding to 1900 li, the second, given in brackets, is calculated on the base of 160 km corresponding to 1800 li. In the first case, a li would be equal to 84,21 m, in the second to 88,89 m, a value that is astonishingly close to the above mentioned questionable measurement of 77 m per li. The real distances on the ancient tracks were most probably somewhat shorter than the distances on the highway, as the tracks were cutting the long serpentines on the passes. This would bring us even closer to the supposed ratio. The ratio would match almost perfectly if based on 1900 li for a distance of 150 km (0,079). If one would thus accept the ratio 1 li = 77 m, the whole distance of 3800 or 3900 li would yield 292 or 300 km from Kulu.

(80) Since Uray's (1990) paper, this assumption has become standard acceptance, and the identification is typically uncritically cited, cf. Dotson (2009: 111, n. 256) with reference to Mard. Hazod (2009: 166, map 2) represent this as Mar-yul with the addition "(719 A.D.)"! at Leh.

(81) According to Petech (1947: 85), also smu. Unfortunately Petech does not say, from where he has got this reading, but only remarks in note 8: "nell'edizione del Francke erroneamente sMra" ('erroneously sMra in the edition of Francke'). Petech (1939: 2) mentions a London manuscript of the LDRR, the first part of which was not used by Francke, and announces to add "either in a foot note or in appendices the pieces supplied by this manuscript, that are missing by Francke's edition", however, on p. 13, he only gives the name ZagZug (with the indication that the original spelling would be ZagZug in contrast to Francke's 1926 edition) without any epithet and without any remarks about its correct or incorrect spelling. R.A. Stein (1961: 50), who refers to Petech, does not seem to have seen the London manuscript. Nor did I. The form smu is found, however, in the Bkahchems kakholma, 1989 edition, Kansuhu mirigs dpeskrunkhang.

(82) See Zeisler (forthcoming a, Chapter 2 [section] 2.4.4), where it is argued that the geographical concept of the Kailash as the source of the four cardinal rivers must have been transferred from an original place in the Pamirs.

(83) In the early 1980s, Peissel (1984: 73) could still observe traces of earlier digging activities along the road from Alchi to Khalatse. He similarly notes that Zanskaris and Ladakhis alike consider panning for gold as a "lowly, unclean task, one best left to the blacksmiths".

(84) Rgyahi-najna. In some Ladakhi dialects locative postpositions are ambiguous and allow an ablative reading.

(85) Assimilation over two opening grades does not seem to be common in Ladakh (actually I know of no case), but has been attested in Spiti, cf., e.g./simgul/sahgul 'earth quake' (Veronika Hein p.c.). There is a direct parallel to the place name concerning a figure of the Gesar epic, bsanpa Rmaru(rtse) ~ Rmeru-(rtse), which is known in the Ladakhi version as Sankra Miru (cf. R.A. Stein 1959: 522 and passim for this figure).

(86) A Majnaj 'Inner Man' of Guge is to be located to the south-east of Toling (Tucci 1971b: 377 with n. 5).

(87) It should be remembered that the main Zajskar valley cannot be accessed from Ladakh (except in mid-winter, when the river is frozen), but only from the eastern and western extensions, namely from Garza or from Purik.

(88) Given the close neighbourhood to India, it could well have been possible, that Mard had special ties with India, which could have helped to escape Tibetan dominion a short while longer than ZanZun. Such a scenario would motivate the mentioning of Mard as a discrete entity. It would further also motivate the fact that Mard is mentioned only once, but never again.

(89) Although attested in various historical texts, I have some difficulties in accepting uncritically, first of all, that a legitimate heir of the Tibetan brtsanpo-s was adopted as son-in-law and main heir into a local ruling family at the periphery of the empire. The prince does not seem to have been much more than a puppet in the hands of some ambitious leaders, cf. LDRR 35.4f., who could well have 'embellished' his descent. Even if the underlying political ambitions should have been based on real descent, it is even more unlikely that Skyidlde Nimamgon himself divided the small empire he had just built up.

(90) Adherents of the Maryul-Ladakh identity would claim that Ladakh constituted an important trade thoroughfare. However, the scarcity (if not non-existence) of sources concerning early Ladakh seems to indicate that the area was much less important than even sceptics would believe.

(91) Cf. Tucci (1956: 51-63) for a synopsis of all the relevant sources.

(92) Cf. also the wording in two West Tibetan manuscripts:

gsergyi sabzi [! for -gzi] hdalzin khajmnam mdzespar bkod [parallel] rirgyal lhunpo brjidcin glin bzi glinbran [! for -phran] bcas ...

'[On] the golden fundament of the earth [the cosmic water or great flood dba3chen] got absorbed and together with houses [everything] was established nicely. The king of the mountains, the cosmic mountain shone and was accompanied by (lit. possessed) the four continents (and) the islands ...' (Heller 2007: 140 with a differing trans-lation, p. 140, n. 40 for the text).

destens dbanchen gsergyi sagzi chags [parallel] destens Rirab dpagtshad bzikhri chags [parallel] gsergyi ri bdun rol-pahi mtshoyi [!] bskor [parallel] glin bzi glinhphren [! for -phran or hphren?] byeba-phrag brgyad chags [parallel] ...

'On that [mandala of cosmic water], [in] the great flood, arouse the golden fundament of the earth. Thereon the Sumeru, measuring 40,000 yojana arouse. The seven golden mountains were surrounded by the ocean (lit. magical lake). The four continents and the eighty million islands arouse....' (Heller 2007: 140 with a differing translation, n. 42 on p. 142 for the passage).

(93) A relation between Maryul and Suvarnagotra or Suvarnabhumi as western and south-western areas of ZanZun has been suggested already by Hummel (1974: 491).

(94) Depending on when the transfer did take place, and whether the Tibetan imperial or post-imperial elites came across an already established designation or whether they were actually the agents of the transfer, Miru might have been named after the mythic land or the place name Miru could have served as catalyser.

(95) The first attestation of the word mjahris in Pt 0016, quite apparently refers to an official document that is executed by cutting off an identical half or a smaller part for the purpose of validation and identification. Unlike in other cases, where such documents are called thaj or mjahthaj, referring to their property as scrolls (cf. Zeisler forthcoming c, n. 73), the designation here refers to the act of writing or drawing, based on the root "ri, Old Tibetan hdri, bris, Classical Tibetan hbri, bris (cf. also the word rimo 'drawing, painting'). It appears that in an unknown period, the West Tibetan provinces were given as a special fiefdom to a dependant ruler. After some time, the administrative entitlement (in the above passage still visible) apparently turned into the very designation of the provinces. Something similar happened with the administrative unit stojsde, which became the name of a village and former kingdom in Zanskar.

(96) The lhagbcas morpheme {ste} after nouns has an introductory function, like our colon. It can often be found in commentaries, separating a word or passage to be explained from the explanation.

(97) This is not so uncommon, as the people of Zanskar typically call the Balti/Walti/, a form that evidently goes back to the written representation as Sbalti, but not to the original ethnonym. Similarly the written Tibetan word bskalpa for Skr. kalpah has eventually become/skalpa/in KHAL.
Table 1. Suvarnagotra and its neighbourhood

                   Xuanzang ~650              VP~670       Hyecho~730
Name               mentioning/
                   location            V   M       O       M   V   O

 1. Tibet/Tufan    E. of 3=4           -   +       /       +   -   /
 2. Khotan         N. of 3=4           +   +       +       +   +   -
 3. Suvarnagotra   3=4, S. of 2,       -   +   ?[+ or -]   +   -   +
    (Gserrigs)       E. of 5=6, W.                           NE. of
                     of Tibet, N.                          Jalamdhara
                     of ?Brahmapura
 4. Strirajya
 5. Sanbohe        5=6, W of 3=4,      -
 6. Moluosuo       N. of Lahul
 7. Bololo         S. of Pamir             -       -       -   -   -
                     valley,=8 (?)
 8. Lesser Bolor                                           +   +   -
 9. Polulo/        W. bank of Indus    +                   +   +   +
10. Skarrdo (?)                            +       -
11. Soupoci                                                +   -   +
12. Yangtong                                               +   -   +
COPYRIGHT 2009 Library of Tibetan Works and Archives
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Cultural Geography
Author:Zeisler, Bettina
Publication:The Tibet Journal
Article Type:Geographic overview
Geographic Code:9INDI
Date:Sep 22, 2009
Previous Article:Notes on the thirteenth Dalai Lama's confidential letter to the Tsar of Russia.
Next Article:Sculptural production during the bstan pa phyi dar and its stylistic nomenclature: some examples from Khu nu of Rong Chung.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters