Tibetan and Himalayan Library - THL

THL Title Text
Antiquities of Zhang Zhung
by John Vincent Bellezza
Edited by Geoffrey Barstow, Mickey Stockwell and Michael White
Tibetan & Himalayan Library
Published under the THL Digital Text License.

Front Sections


1. The Archaic Archaeological Sites of Upper Tibet

The upper portion of the Tibetan Plateau, a land of large lakes, lofty peaks, interminable plains, and deep gorges, stretches north and west of LhasaLha sa for 1500 km. Bound by high mountain ranges on all sides and averaging 4600 m above sea level, Upper Tibet gave rise to an extraordinary civilization in antiquity. Beginning about 3000 years ago, a chain of mountaintop citadels, temples, and intricate burial complexes appeared in this vast region of some 600,000 square kilometers. These monuments were part and parcel of a unique human legacy, which flourished until the Tibetan imperium and the annexation of Upper Tibet by the PugyelSpu rgyal emperors (tsenpobtsan po) of Central Tibet. Gradually the unique beliefs, customs and traditions of archaic Upper Tibet yielded to a pan-Tibetan cultural entity that arose in conjunction with Vajrayāna Buddhist teachings.

A millennium ago, Buddhist domination of Tibet spawned a new civilization, one in which the celebrated Lamaist religions of BönBon and Buddhism came to hold sway. The inexorable march of time and the ascent of the new religious order slowly but surely clouded the memory of the earlier cultural heritage. As a result, many of the ancient achievements of the Upper Tibetan people were forgotten. All that remains are preserved in the impressive monumental traces of the region. Antiquities of Zhang Zhung attempts to reclaim these past glories by systematically describing the visible physical remains left by the ancient inhabitants of Upper Tibet.

The residential and ceremonial monuments of Upper Tibet, established by what can be termed the “archaic” cultures of the region (Zhang ZhungZhang zhung and SumpaSum pa of the literary records), strongly contrast with those built in the central and eastern portions of the plateau in the same span of time. There are very substantial differences between the archaeological makeup of the archaic cultural horizon (circa 1000 BCE to 1000 CE) and that of the Lamaist era (circa 1000 CE to 1950 CE) in Upper Tibet. The unique monumental assemblage of Upper Tibet delineates the bounds of a paleocultural complex squarely based in the uplands of the plateau. The special physical hallmarks and highland homeland of this ancient culture set it apart from other Bodic cultures, which arose in the central and eastern parts of the Tibetan Plateau. The paleocultural world of Upper Tibet is readily distinguished from those civilizations that appeared in adjoining lands to the south, west and north. In the archaic cultural horizon the Upper Tibetans constructed highly durable all-stone elite residences, temples and castles, developing stone working techniques particularly suited to their extremely harsh natural environment. They also designed and built elaborate burial complexes containing many types of ritual structures made entirely of stone. The use of stone corbelling for the construction of roofs and the erection of pillars in peculiar configurations for ceremonial purposes reached a very high level of proficiency in Upper Tibet. The eminently practical qualities of this architecture have helped to insure that the remains of a surprising number of monuments have endured to the present day.

Although the design and construction of the monumental assemblage of archaic Upper Tibet is highly distinctive, affinities with other archaeological cultures of the plateau and steppes certainly exist. During the first millennium BCE and first millennium CE, a tremendous amount of cross-fertilization occurred throughout Inner Asia. These manifold cultural links are explored in depth in my last book, Zhang Zhung: Foundations of Civilization in Tibet. This monograph furnishes the analytical framework and data necessary to begin to comprehend the chronological, economic and cultural dimensions of the sites surveyed in the present work.

Antiquities of Zhang Zhung systematically describes the physical remains of 404 Upper Tibetan monumental sites documented since 2001.1 It is an inventory of archaic or prospective archaic archaeological sites. These sites differ from Lamaist monuments in terms of morphology, function, mythology, and geographic orientation. This catalogue of archaeological sites should prove useful to scholars working in a variety of disciplines. As a reference work, it is well suited to provide a perspective for subsequent studies devoted to better understanding the archaic physical and cultural environment of Upper Tibet and other regions of Inner Asia. It presents uniform sets of physical and cultural data for each of the sites surveyed to produce a coherent view of the monumental vestiges scattered across the Upper Tibetan landscape. As a compendium of archaeological sites, this work is primarily quantitative (descriptions of the remaining physical evidence) in nature. To a lesser degree, it also provides qualitative information (analyses of the ideological groundwork underlying the physical manifestations) in order to elucidate various abstract aspects of the monuments. This methodological approach, borrowing from archaeological, literary and ethnographic sources of information, permits an integral picture of ancient Upper Tibetan archaeological assets to emerge. By bringing Upper Tibet’s fascinating past into clearer focus, we begin to acquaint ourselves with the formative elements in the development of Bodic civilization. In turn, this permits us to move one step closer to understanding the Tibetan Plateau’s place in the Eurasian cultural mosaic of yore.

An inspection of the sites surveyed opens a window onto a remarkable Tibetan heritage. Rather than a cultural backwater, upland Tibet emerges as a nexus of technological and cultural brilliance. A chain of citadels circumscribing the region reflects the existence of a vibrant social order in which agriculture played a vital role. From the first millennium BCE onwards, a warrior and priestly elite appears to have founded and occupied these citadels. The sheer number of fortified sites built on summits shows that martial struggle was a prominent preoccupation (which is mirrored in the Tibetan literary record). The top strata of ancient Upper Tibetan society constructed all-stone temples and residences in which the cultural life of the region reached a crescendo. Troglodytic communities sprang up wherever there were natural caves or where it was possible to excavate earthen formations. In the cultural hothouse environment of first millennium BCE and early first millennium CE Inner Asia, Upper Tibet appears to have been one of several regions with superior intellectual and military capabilities. The legendary status accorded Zhang ZhungZhang zhung in Tibetan literature buttresses the archaeological record, indicating that Upper Tibet had indeed reached a considerable level of human attainment before the spread of Buddhism.

The existence of intricate burial rites is echoed in the many tombs and necropoli that dot the entire region. These architecturally diverse funerary sites allude to sophisticated eschatological concepts and practices prevalent in early Upper Tibet. The mortuary archaeological evidence also records yawning divisions in wealth and social status, a sign that the region possessed a hierarchical society with deep social, economic and political divisions. This puts the highland variant of Bodic civilization in line with surrounding civilizations of the Iron Age and the classical period, where social stratification, economic diversification and warfare were rampant. While many linkages between the empirical and textual perspectives remain hypothetical, the intellectual profundity of matters related to death in both the literary and archaeological records is unmistakable and very significant. In Zhang Zhung: Foundations of Civilization in Tibet, I examine the interconnections between the mortuary sites of Upper Tibet and the archaic funerary beliefs and rituals of the Tibetan texts.2

So much still needs to be discovered before we can find answers to even basic questions concerning the polity and people of ancient Upper Tibet. Nevertheless, the good news is that step-by-step an understanding of the region’s archaeological character is being secured. This increase in our knowledge should pave the way to new insights into the origins and development of Tibetan civilization, as well as to a more refined appreciation of the ancient cultural complexion of Inner Asia. It is in the service of such aims that the present work has been composed.

2. An Introduction to the Author’s Archaeological Exploration of Upper Tibet and Survey Methodology3

I began my travels in Upper Tibet (JangtangByang thang and Stod) in the mid-1980s, a golden period in the exploration of the plateau. This was an exciting time for discovery in Tibet, a time when a small group of explorers (curiously, they were mostly from English-speaking countries) reached places never before visited by foreigners. During my initial years of peregrination in Upper Tibet, I began to notice unusual manmade formations and ruins but did not pay much attention to them. In the early 1990s, having acquired the requisite cultural and linguistic skills, I turned much of my scholarly energy to the documentation of archaeological remains and the elucidation of the ancient cultural history of Upper Tibet. In the course of fieldwork, I have had the good fortune to visit every county and virtually every township in the great Tibetan upland north and west of LhasaLha sa. These archaeological surveys in the region have therefore proven geographically all-inclusive.

On earlier visits to Upper Tibet, an immense region of approximately 600,000 km², I spent a great deal of time on foot and solo. On more recent expeditions, I have depended on motor vehicles and crews to expedite reaching highly remote places and the process of documentation. Despite having vehicles, fairly long distances still had to be hiked or ridden on horseback due to the rugged nature of the terrain. Many sites located on mountaintops and escarpments, or in gorges and caves are only accessible on foot. The physical rigors of these expeditions should not be underestimated. Upper Tibet is a tough environment in which to work and the pace of study has been intensive.

List of Archaeological Survey Expeditions:

  • 1992: Four Fountains of Tibet Expedition (FFTE)
  • 1994: Divine Dyads Expedition, year one (DDE1)
  • 1995: Divine Dyads Expedition, year two (DDE2)
  • 1997: Changthang Phase II Expedition, year one (CPE1)
  • 1998: Changthang Phase II Expedition, year two (CPE2)
  • 1999: Changthang Circuit Expedition (CCE)
  • 2000: Upper Tibet Circumnavigation Expedition (UTCE)
  • 2001: Upper Tibet Antiquities Expedition (UTAE)
  • 2001: Shang Shung Institute Expedition (SSI)
  • 2002: High Tibet Circle Expedition (HTCE)
  • 2003: High Tibet Antiquities Expedition (HTAE)
  • 2004: High Tibet Welfare Expedition (HTWE)
  • 2005: Tibet Upland Expedition (TUE)
  • 2006: Tibet Ice Lakes Expedition (TILE)
  • 2006: Tibet Highland Expedition (THE)
  • 2007: Wild Yak Lands Expedition (WYLE)

In 2001, I launched the four-month long Upper Tibet Antiquities Expedition (UTAE), which clocked around 8500 km in vehicles and significant distances on foot and on horseback. On the UTAE, 90 archaeological sites were documented in BaryangBar yangs, PurangPu rang, KhyunglungKhyung lung, GugéGu ge lho smad, ChusumChu gsum, GötsangRgod tshang, northern RutokRu thog, Naktsang RongmarNag tshang rong dmar, and Dangra YutsoDang ra g.yu mtsho. In 2002, I set off on the High Tibet Circle Expedition (HTCE), which was of four months duration as well.4 This expedition yielded information on more than 100 archaeological sites, the overwhelming majority of which had never been documented. On the HTCE, I covered 13,200 km by motor vehicle, and trekked considerable distances on foot and on horseback. The main thrust of exploration included BaryangBar yangs, Langa TsoLa snga mtsho, Gang RinpochéGangs rin po che, ZarangZa rang, RutokRu thog, northern GertséSger rtse, Ngangla RingtsoNgang la ring mtsho, TsochenMtsho chen, Dangra YutsoDang ra g.yu mtsho, the TagoRta rgo range, and BartaBar tha. In 2003, I conducted exploration on the High Tibet Antiquities Expedition (HTAE), which lasted 48 days.5 On the HTAE, around 40 archaeological sites were documented by traveling more than 8000 km by motor vehicle. The geographic focus of exploration was the border areas situated in RutokRu thog, TsamdaRtsa mda’ and PurangSpu rang, marking the first access to many of these sectors by an outsider in 60 years.

In 2004, I launched a three-month mission to Upper Tibet called the High Tibet Welfare Expedition (HTWE). The HTWE was carried out with the purpose of reconnoitering areas of Upper Tibet not previously visited or where more inquiry was required. The main areas for research and exploration included Damzhung’Dam gzhung, YakpaG.yag pa, southern TsonyiMtsho gnyis, Dangra YutsoDang ra g.yu mtsho, TsochenMtsho chen, SenkhorBse ’khor, ZhungpaGzhung pa, RutokRu thog, GarSgar, and TsamdaRtsa mda’. In 2005, I embarked on the 45-day long Tibet Upland Expedition (TUE), in order to survey sites across the breadth of much of Upper Tibet not reached on earlier campaigns. By continually making forays, I have been able to close the gaps in the geographic coverage of the region. Slowly but surely, I have now visited most of the major basins and valleys of Upper Tibet south of the 33rd parallel.

In the winter of 2006, I conducted the four-week long Tibet Ice Lakes Expedition (TILE) in order to reach six islands in four different lakes. By traversing the frozen surfaces of the lakes, I was able to survey SemodoSe mo do (NamtsoGnam mtsho), DotagaDo rta sga and DodrilbuDo dril bu (DaroktsoDa rog mtsho), TsodoMtsho do (Tsomo Ngangla RingtsoNgang la ring mtsho), and DoserDo ser and DomukDo smug (Langak TsoLa ngag mtsho). In the spring of 2006, I completed the basic survey work, a 12-year enterprise. Known as the Tibet Highland Expedition (THE), the object of this 46-day 2006 excursion was to carry out reconnaissance in the northern JangtangByang thang, and to visit a few outstanding archaeological sites. In 2007, on the Wild Yak Lands Expedition (WYLE) (45 days in length), I reconnoitered parts of the northern JangtangByang thang and documented a handful of archaic sites in GugéGu ge and other locations.

In surveys conducted since 2001, I have endeavored to expand and strengthen the methodological tools at my disposal. It has been necessary to further systematize the collection of data and to articulate these in forms that make it accessible to a wider range of Tibetologists, archaeologists and cultural historians. The survey data thus compiled have permitted the various types of archaeological assets present in the region, their patterns of distribution, environmental context, and structural qualities to be elucidated in greater clarity. Another vital component of this appraisal of Upper Tibetan archaeological sites has been the compilation of chronometric data derived from the radiometric and AMS assaying of organic samples. To date, 21 samples have been submitted for chronometric testing and analysis, permitting the initial direct dating of a few documented sites. This augmented methodological approach to the survey work has enabled the positioning of the sites chronicled within a more refined chronological context, opening the way to new perspectives in the study of Tibetan textual sources. Generally speaking, these breakthroughs in the study of Upper Tibetan cultural development pertain to temporal controls, which encompass both the prehistoric and historic epochs.

The methodological regimen applied to the survey of monuments (residential and ceremonial) can be summarized as follows:

  1. The pinpointing of the geographic coordinates, elevation and administrative location of each site. The determination of latitude, longitude and elevation was accomplished with the use of a GPS. In locating sites, reference is made to toponymic nomenclature employed in both historical (traditional) and Communist (modern) political geography.
  2. A description of the geographic and ecological settings of archaeological sites. In order to understand the physical environment shaping the function and placement of monuments, attention has been paid to slope gradients, general soil conditions, prominent landforms in the proximity, geomorphologic changes, and the endowment of natural resources.
  3. The identification of the monumental types found at each archaeological site. This is carried out using a comprehensive typology of above-ground archaeological resources devised for Upper Tibet (see Section 5).
  4. An analysis of the morphological, design and constructional traits of each structural component of an archaeological site. A study of how monuments were built and the types of materials that went into their construction is vital in differentiating the various typologies. The investigative focus has been directed towards determining ground plans, wall fabrics, the rendering and presentation of structures, patterns of usage, and the spatial arrangements of the various structural components making up a site.
  5. The measurement of site dispersals and the dimensions of constituent structures. The overall extent of sites (measured in square meters), and the length, width, height, and girth of monuments and their respective components.
  6. The mapping of monuments (plans and topographic settings). Save for sketches of a few ground plans, the cartographic dimension has thus far been limited to overview and typological maps of archaeological sites.
  7. The photography of the general settings of sites, all visible archaeological remains, and the current cultural scene.
  8. The compilation of folklore, myths, legends, and historical accounts surrounding archaeological sites. I have endeavored to collect the local oral traditions attached to the monuments surveyed in order to gain a firmer understanding of the chronology, function and significance of sites as conceived by indigenous sources.
  9. The collection and translation of Tibetan textual sources pertinent to the function, cultural make-up, political affiliation, and chronology of monuments and the physical sites in which they are located. This facet of study defines the interface between empirical and traditional historiographic approaches to understanding Upper Tibet’s archaeological heritage.
  10. An assessment of contemporary anthropogenic and environmental risks to the continued survival of archaeological monuments. This proactive component of research concerns issues related to the conservation and sustainability of archaeological assets.

The interrelated methodological regimen used in the surveys of rock art can be summarized as follows:

  1. The pinpointing of the geographic coordinates, elevation and administrative location of each rock art site.
  2. A description of the geographic and ecological settings of rock art sites.
  3. An analysis of the physical characteristics, relative locations and techniques of manufacture of rock art.
  4. The measurement of rock panels and individual compositions.
  5. The mapping of rock art sites (geographic locations).
  6. The photography of the general settings of rock art sites, individual compositions and the current culture-scape.
  7. The compilation of folklore, myths, legends, and historical accounts surrounding rock art sites.
  8. The collection and translation of Tibetan textual sources pertinent to the function, cultural orientation, political affiliation, and chronology of rock art sites and individual compositions.
  9. An assessment of contemporary anthropogenic and environmental risks to the continued survival of rock art.

I have undertaken to document every visible archaeological site of the archaic cultural horizon on the vast Tibetan upland and, while falling short of this ambitious goal, more than 600 monumental sites and 50 rock art sites have been surveyed throughout the region. How many other archaic sites with visible above-ground footprints exist in Upper Tibet remains to be determined. In particular, there must be many dozens of ancient burial grounds that have yet to be charted. This is indicated by the sheer number of tombs already documented, the oral tradition that holds that tombs are distributed all over Upper Tibet, and the practical difficulties in locating structures with little or no protrusion above the ground surface. The geographical thoroughness of the survey work, however, indicates that a statistically significant cross-section of monument types and rock art has been documented.

Over 90% of the sites chronicled in this inventory have not been identified or studied by other research teams. Rather than the application of remote sensing tools and aerial surveys to identify archaic cultural horizon archaeological assets in Upper Tibet, I took upon myself the laborious and time-consuming task of comprehensive field inspections. Visible detection of sites was facilitated in most places in Upper Tibet by poorly developed alpine and steppe soils, sparse vegetation cover, and high rates of surface erosion. As in any region, a percentage of the total number of Upper Tibetan archaeological assets is not amenable to surface detection. The percentage of sites that were overlooked because of the lack of visual apprehension, however, appears to be relatively low in the JangtangByang thang (given its prevailing topographic and vegetative features). Conversely, in the badlands of GugéGu ge, a region of thick alluvial deposits and the regular occurrence of landslides, a much higher percentage of archaeological remains are probably obscured from view. A significant number of archaeological sites may have been overlooked in the still active agricultural communities of far western Tibet and Lake DangraDang ra. In these regions it is plausible that successive layers of human occupation have been hidden from view by the structural overlay of contemporary settlement.

The field inspection of archaeological remains has the advantage of furnishing positive identification and the procurement of definitive empirical information. The field surveys entailed visiting virtually every one of the approximately 250 townships (reckoned according to the number of townships existing prior to the 1999-2001 period of administrative consolidation in the TAR) in the 17 counties that comprise Upper Tibet west of NakchuNag chu city. During this twelve year campaign, I have spent nearly four years in the field, and covered more than eighty thousand kilometers by vehicle, and at least another eight thousand kilometers on foot and on horseback. In order to locate archaeological sites, individual and collective interviews were conducted in all county seats, as well as in many township headquarters, monasteries, local villages, and pastoral settlements. In the course of interviews with over 5000 people, I have met with a wide range of civil officials, monks, lay practitioners, farmers, and herders. Special emphasis was placed on allocating time to speak to those locally recognized as the most knowledgeable in their respective communities. The meticulous geographic coverage of the surveys, accomplished by canvassing large swaths of territory upwards of three to seven times each, has proven invaluable in understanding the geographic distribution of the various types of archaic cultural horizon archaeological assets in Upper Tibet.

3. Criteria Used in the Determination of Archaic Archaeological Sites6

Before presenting an analysis of the various types of monuments, it is crucial to revisit what constitutes an archaic cultural horizon archaeological site in Upper Tibet. In brief, these are structures exhibiting physical and cultural qualities that predate the introduction and spread of Lamaism (systematized BönBon and Buddhism) in Tibet. The establishment and particularly the usage of these archaeological sites, however, may have persisted for centuries after Buddhism gained a foothold in imperial Tibet (early seventh century to mid-ninth century CE). The term ‘archaic’, therefore, is employed to describe archaeological sites that exhibit non-Lamaist cultural and architectural characteristics, and not to refer to a specific time period as such.7

The provisional identification of archaic monuments in Upper Tibet is made on the basis of the following physical and cultural criteria using inferential means:

  1. Sites in BönBon literature attributed to personages, events, facilities, and locations associated with the Zhang ZhungZhang zhung and SumpaSum pa kingdoms
  2. Monuments attributed in local oral traditions to the ancient BönBon, the MönMon, personalities in the Ling GesarGling ge sar epic, and the pantheon of genii loci
  3. Monuments exhibiting early design, constructional and morphological features
  4. The siting of monuments in now abandoned environmental niches
  5. Monuments and rock art comparable to those documented in other regions of Tibet
  6. Monuments and rock art comparable to those documented in other Inner Asian territories
  7. Art and artifacts that exhibit primitive stylistic and fabrication traits

1) Sites in BönBon literature attributed to personages, events, facilities, and locations associated with the Zhang ZhungZhang zhung and SumpaSum pa kingdoms

Especially when used in conjunction with other archaeological criteria, Tibetan literature is a precious indicator of the location and identity of archaic monuments. BönBon (and to a lesser degree Buddhist) texts are an excellent and extensive source for mythic and quasi-historical accounts relating to places in Upper Tibet supposed to have been important centers of the ancient Zhang ZhungZhang zhung and SumpaSum pa kingdoms. These texts provide biographical data about the lives of Zhang ZhungZhang zhung and SumpaSum pa saints, including information regarding their residences and political dealings with local potentates and foreign enemies. These literary accounts are framed in both the prehistoric epoch and early historic period, but their historicity remains obstinately difficult to corroborate. For the most part, BönBon literary sources postdate the eleventh century CE (centuries after the historical events they purport to chronicle) and are heavily colored with mythic and hagiographic content, significantly limiting their value as prosaic historical documents. This literature names geographic locations, some of which can be confidently correlated to the contemporary toponymic picture (places such as MamikMa mig, PurangPu rang, GugéGu ge, DangraDang ra, TagoRta rgo, TiséTi se, NamtsoGnam mtsho, TanglhaThang lha, etc.), while the identity of others has either not been established or only tentatively. As the chronology of BönBon mythic and quasi-historical materials pertaining to Zhang ZhungZhang zhung and SumpaSum pa is uncontrolled (by associative events such astronomical phenomena, natural disasters, cross-cultural references, calendrical lore, etc.), it limits their use as indexes of time, except in the broadest sense.

Moreover, BönBon sources have been subjected to an ongoing process of textual revision, altering the portrayal of early historical events. This modification of contents expresses itself in two major ways: the idealization of past patterns of settlement and cultural achievement, and the reconfiguration of the archaic cultural heritage using the language and concepts of Buddhism. Nonetheless, BönBon literature furnishes us with valuable contextual information on major centers of early settlement and their cultural and religious complexion. For one thing, a comparison of textual-based geographic lore related to Zhang ZhungZhang zhung with the patterns of archaic monumental distribution in Upper Tibet reveals a strong positive correlation.

2) Monuments attributed in local oral traditions to the ancient Bönpobon po, the MönMon, personalities in the Ling GesarGling ge sar epic, and the pantheon of genii loci

The oral traditions surrounding the archaic monuments of Upper Tibet tend to contrast with these accounts connected to Buddhist monuments, in which piety and otherworldliness prevail. Since the domination of Lamaism in Upper Tibet, circa 1000 to 1250 CE, religious attitudes developed that altered perceptions of the earlier cultural heritage of the region. Generally speaking, this recasting of history led to the archaic past being viewed with a considerable sense of fear and denial. As Buddhism and systematized BönBon gradually took hold in Upper Tibet, transforming its culture and ethos, the push to reinterpret history gained momentum in society. The major effect of this historical reformulation has been to make the ancient past increasingly resemble Lamaist thought and practice. In the contemporary socio-cultural setting, the archaic monumental wealth of Upper Tibet has been compressed into just four major themes. This thematic compression involves the reduction of the ancient cultural legacy into stereotypic narratives, which now stand as supposed factual representations of the past. This has led to the loss of much historical information once associated with the archaic archaeological assets in the oral tradition of Upper Tibet. The cognitive and affective forces enmeshed in this cultural transformation were not directed at highland archaeological sites alone, but came to express themselves in manifold social and political ways across the Tibetan world.

It is within these four legendary themes that clues pointing to the identification of archaic monuments must be sought: (1) the ancient BönBon, (2) the MönMon, (3) the GesarGe sar epic, and (4) the pantheon of local spirits. These legendary and mythic attributions are generally applied to sites that do not fall under the architectural ambit of Lamaist culture. They function as convenient intellectual categories to relegate awkward bits of early heritage (which by their very physical presence cannot be simply brushed aside) to a safe and distant ideological realm. While the oral tradition provides associative evidence of early settlement, it is not well suited to the collection of archaeological facts concerning archaic monuments and rock art. The oral tradition, therefore, is best applied as a non-specific and broadly inclusive interpretive anthropological tool.

3) Monuments exhibiting early design, constructional and morphological features

An excellent indicator of the archaic status of archaeological monuments in Upper Tibet is the presence of distinguishing features in substance and form. These physical properties reflect different architectural conceptions and modes of execution than those exhibited by familiar Lamaist monuments. Of special note are the various funerary pillars (menhirs) and necropoli of Upper Tibet. These types of monuments embody distinctive forms of abstraction and construction which do not appear to have been adopted by Lamaist adherents. A different religious ethos required an alternative assemblage of monuments: rather than large burial complexes, Buddhism and systematized BönBon saw fit to cover the landscape with chötenmchod rten (a type of shrine) and walls with inscribed plaques, which are of a different order of architectural magnitude. In the domain of residential monuments, great structural contrasts are seen between the all-stone corbelled edifices of early times and BönBon and Buddhist buildings built with high walls and wooden rafters. Aside from the very different methods and materials used in construction, the former structures are small-chambered, windowless and semi-subterranean, while Lamaist halls and temples have larger rooms and frequently windows or skylights, and are set prominently above the ground.

4) The siting of monuments in desolate environmental niches

The specific geographic setting of archaeological sites provides some clues to their cultural identity. Many archaic residential monuments were built at high elevation and in special environmental niches that have long since been abandoned. These sites were not the objects of sustained sedentary settlement in any way associated with the Lamaist cultural milieu of later times. A significant number of archaic sites are concentrated in defunct agricultural enclaves in far western Tibet, and on headlands and islands across the breadth of the Upper Tibetan lake belt. Archaic residential sites are also found on lofty, inherently defensible summits and ridges, or at the heads of valleys at elevations sometimes exceeding 5000m. Environmental degradation and changed cultural realities appear to be the motive forces behind the geographic shift from these specialized locations to the patterns of population distribution witnessed in more recent centuries. For the most part, the Lamaist religions chose lower-elevation basins and valleys for their major residential sites. Even when escarpments and mountain slopes were selected for the establishment of religious and political edifices, these are consistently located at a lower elevation than their archaic counterparts. Gang TiséGangs ti se is an excellent case in point: all around this sacred mountain one must climb well above the existing Buddhist sites in order to reach those established in earlier times. The same patterns of settlement hold true for Dangra YutsoDang ra g.yu mtsho where the archaic cultural horizon looms over the contemporary BönBon villages.

5) Monuments and rock art comparable to those in other regions of Tibet

Comparative study of Upper Tibetan archaeological assets, with their counterparts in Central Tibet and Eastern Tibet, is another tool for ascertaining relative age and cultural affiliations. Unfortunately, very little reliable chronometric data has yet been assembled for archaic residential and ceremonial sites located in other regions of Tibet. Moreover, comprehensive archaeological surveys have yet to be launched outside Upper Tibet. The poorly organized archeological data compiled in other regions of the plateau impedes studies based on cross-referenced archaeological comparisons. As a result, the extent and nature of paleo-cultural affinities between Upper Tibet and Central Tibet and other regions of the plateau have not been adequately determined.

In Upper Tibet and Central Tibet, quadrate burial tumuli with inwardly sloping walls were built in the early historic period and most probably in the prehistoric epoch as well. However, the all-stone corbelled residential edifices and pillar monuments that define the Upper Tibetan paleo-cultural territory are not represented in Central Tibet. KhamKhams and AmdoA mdo have varying assemblages of monuments (these are still not well catalogued). Nevertheless, the pastoral regions of AmdoA mdo were host to a rock art tradition that is thematically and stylistically related to that of Upper Tibet. The areal variability marking archaeological assets is acknowledged in the Tibetan historical tradition, which assigns prehistoric Central Tibet and DokhamMdo khams to different proto-tribal or quintessential groupings. Central Tibet is recorded as being dominated by Bod, KhamKhams by MinyakMi nyag, and AmdoA mdo by AzhaA zha.

6) Monuments and rock art comparable to those in other Inner Asian territories

Cross-cultural Inner Asian study is a fecund methodological approach for the determination of the identity and chronology of Upper Tibetan archaeological assets. This method has proven best suited to the interregional comparison of funerary sites that possess substantial above-ground structural elevations. Archaic funerary pillars and slab wall structures are a case in point, where comparisons between the Upper Tibetan, Mongolian, Altaian and south Siberian types have borne good results. These basic monumental forms are dispersed throughout Inner Asia. As in other spheres where the technologies and cultural traditions of Inner Asia were disseminated widely, chronological and cultural parallels between the funerary monument traditions of Upper Tibet and adjoining regions are indicated. The comparative study of Inner Asian rock art is useful in delineating the amalgamative processes that brought Upper Tibet into functional and aesthetic congruity with its northern neighbors. The biggest drawback to cross-cultural analyses remains the general shortage of good chronological controls for sites in Upper Tibet. This will be remedied only when chronometric studies gain sufficient ground.

7) Art and artifacts that exhibit archaic stylistic and fabrication traits:

The aesthetic and technical analysis of art and artifacts is best used in conjunction with collateral archaeological data, but even alone it is a helpful method for estimating chronological values. The rock art record provides one of the best indexes of cultural evolution from the archaic to the Lamaist. The prehistoric Upper Tibetan rock art tableaux are rich in compositions that depict economic, environmental and cultural matters related to the way of life in the region. These petroglyphs and pictographs are largely unrelated to Buddhist-inspired art and design as they developed in Tibet. Rock art exhibiting archaic themes (such as hunting scenes, the isolated portrayal of wild animals, and iconic motifs) continued to be produced well into historic times. This suggests that there was a good deal of cultural continuity between the prehistoric and historic epochs in Upper Tibet. Nonetheless, analogous subject matter reveals different modes of manual execution and stylistic presentation, valuable evidence in any attempt at chronological differentiation. As compared to rock art made in the prehistoric epoch, the later variants exhibit their own set of production qualities and aesthetic refinement. Rock art of the historic epoch is either cruder or more polished. This inferred chronological progression is also discernable in other spheres of material culture. Copper alloy artifacts such as amulets, implements and weaponry possess aesthetic and technical features indicative of relative age and cultural affiliation as well.

In addition to these indirect means of assessing archaic cultural status, the radiometric and AMS assaying of organic remains recovered from sites constitute the direct approach to dating. The criteria outlined above are all dependent on inferring chronological information from evidence that does not intrinsically lend itself to scientific verification. For these criteria to be validated, the conclusions drawn from the cultural identity, appearance and location of monuments and rock art must ultimately stand the test of chronometric verification. Over the last four years, I have begun the process of independent corroboration of the suppositions set forth above. I am intent on presenting the identification of the corpus of archaic structural and aesthetic forms in Upper Tibet in a more objective and reproducible fashion. In pursuance of this goal, 20 samples have been submitted for radiometric and AMS analysis (derived from both residential and ceremonial sites). The recovery and archaeometric assaying of far more samples from many more sites is demanded to definitively chart the chronology (and other objective values) of the Upper Tibetan archeological assemblage. Archaeometric inquiry is also essential in weeding out those sites that may not have an archaic cultural horizon status. It is on a good footing that chronometric data assembled thus far have begun to corroborate the presumptions made concerning the temporal orientation of the sites surveyed.

4. The Chronology of Archaic Archaeological Sites8

The assembled chronometric and collateral data indicate that Upper Tibetan archaic monuments and rock art were produced over a wide spectrum of time, in both the prehistoric and historic settings. Two major epochs, each with two cultural phases, are provisionally indicated. The archaic cultural horizon spans both the prehistoric epoch and the early historic period:

  • I) Prehistoric epoch
    • i) Iron Age
    • ii) Protohistoric period
  • II) Historic epoch
    • i) Early historic period
    • ii) Vestigial period

I) Prehistoric epoch (early first millennium BCE to seventh century CE)

The first phase of the prehistoric epoch includes those sites that were founded in the early Iron Age (first half of first millennium BCE), and the developed Iron Age (middle and late first millennium BCE) of Inner Asia. Possibly, late Bronze Age (circa 1200 to 800 BCE) affiliations are also indicated in the first phase of prehistoric Tibetan civilization, but this remains difficult to corroborate.9 A treatment of more remote prehistoric epochs (Paleolithic, Mesolithic, Neolithic) falls outside the purview of the current study.10 The second or later phase of the prehistoric epoch corresponds to an anachronistic extension of the Iron Age, marked by the Central Tibetan line of kings (late first millennium BCE to the seventh century CE). This second phase of the prehistoric epoch can be termed the protohistoric or legendary monarchal period, due to the many Tibetan literary records that refer to the Central Tibetan kings of that time. There are also BönBon texts purported to have been written in this time frame, though solid evidence for this allegation is lacking. These literary records include some assumed to have been first written in the ZhangzhungZhang zhung and SumpaSum pa languages, which came to be translated into Tibetan during the imperial period. According to the Tibetan historical tradition, the plateau of the Iron Age was divided into a number of petty states and governed by a succession of demigod chieftains. The protohistoric period in turn, is marked by the rise of the yar lungYar lung or PugyelSpu rgyal dynasty beginning with King Nyatri TsenpoGnya’ khri btsan po (traditional chronologies place him in the circa 200 BCE period).

II) Historic epoch (early seventh century CE to present)

This first phase of the historic epoch, the early historic period, chronologically corresponds with the Tibetan empire or imperial period and its troubled aftermath (seventh century to the end of the tenth century CE). It was in the imperial period that the definitive introduction of Buddhism (tenpa ngadarbstan pa snga dar) into Tibet, the development of the Tibetan system of writing (bö yigébod yi ge), and the expansion of Tibetan political power across the entire plateau and beyond occurred. The Upper Tibetan proto-states of Zhang ZhungZhang zhung and SumpaSum pa were absorbed into the pan-Bodic polity of this period as well. The vestigial period includes all archaic style monuments and rock art that continued to be founded in Upper Tibet (late tenth century to mid thirteenth century CE). The production of some archaic cultural horizon archaeological assets appears to have continued for some centuries after the collapse of imperial Tibet. Certain surveyed tombs, strongholds and religious edifices are likely to fall into this category. These architectural anachronisms seem to have been a cultural counterpoint to the inexorable process of Lamaist transformation. This period in Tibetan history is characterized by political reconsolidation, such as the formation of the Buddhist GugéGu ge state in western Tibet in the late tenth century CE, and the ascendancy of the Sakyapasa skya pa in the early thirteenth century CE.

At this juncture, the chronological values proposed above remain largely hypothetical, and with the exception of those few sites where chronometric data have been forthcoming, inexact and open to amendment. Nevertheless, this provisional chronology indicates that archaic cultural horizon archaeological monuments in Upper Tibet are a highly diverse group in terms of age and composition. By virtue of straddling the prehistoric and historic divide, the sites surveyed represent a heritage of varying environmental dimensions, social forces, religious persuasions, and political orders, which are emblematic of cultural change in Upper Tibet over a period of no less than two millennia.

This work primarily treats the typological aspects of the study of archaic monuments and rock art as the basis for their periodization. Additional study, involving the vigorous application of chronometric methodologies, will be needed to create a precise chronology for each of the monument and rock art types surveyed. It is through such study that the cultural development of Upper Tibet and the nature of its intercourse with adjoining territories will come to be known in the kind of detail that such an important piece of the world’s ancient heritage deserves. In addition to providing a model of cultural transition and adaptation in Upper Tibet, chronometric inquiry is required to determine the impacts of Late Holocene (circa 2000 BCE to present) climatic deterioration on the various archaeological sites. Material culture studies are another area of archaeological research that has barely begun. The scientific recovery and study of utilitarian and ritual objects is of the utmost importance if we are to flesh out the cultural specifications, periods of usage and environmental determinants at work at each of the sites catalogued.

5. A Typological Outline of Archaic Monuments and Rock Art

Herein is an outline of the archaic cultural horizon monument and rock art typologies distributed above the ground in all areas of Upper Tibet. The monument typologies fall into two major divisions: residential (structures in which people resided or temporarily lived) and ceremonial (non-residential structures chiefly used for religious and burial purposes). Residential monuments are further divided according to their primary design traits and situational aspects. Ceremonial structures are subdivided according to their morphological and functional aspects. In Upper Tibet there are also minor physical remains associated with the ancient agricultural economy. Earthworks located in Damzhung’Dam gzhung and NyingdrungSnying drung may have had a residential and/or ceremonial function. Rock art of all types forms the aesthetic or graphic division of Upper Tibetan archaeological assets, while rock inscriptions are the epigraphic component.

  • I. Residential Monuments
    • 1) Residential structures occupying summits (fortresses, breastworks, religious buildings, palaces, and related edifices)
      • a. All-stone corbelled buildings
      • b. Edifices with roofs built from timbers
      • c. Solitary rampart networks
    • 2) Residential structures in other locations (religious and elite residences)
      1. a. All-stone corbelled buildings
      2. b. Other freestanding building types
      3. c. Buildings integrating caves and rock overhangs in their construction
  • II. Ceremonial Monuments
    • 1) Stelae and accompanying structures (funerary and non-funerary)
      • a. Isolated pillars (doringrdo ring)
      • b. Pillars erected within a quadrate stone enclosure
      • c. Quadrangular arrays of pillars appended to edifices
      • d. Domestic pillars
    • 2) Superficial structures (primarily funerary superstructures, burial and non-burial in function)
      • a. Single-course quadrate, ellipsoid and irregularly-shaped constructions (slab wall and flush-block)
      • b. Double-course quadrate, ellipsoid and irregularly-shaped constructions (slab wall and flush-block)
      • c. Heaped-stone wall enclosures
      • d. Quadrate mounds (bangsobang so)11
      • e. Terraced constructions
    • 3) Cubic mountaintop tombs
    • 4) Shrines and miscellaneous constructions
      • a. Stone registers (totho)
      • b. Tabernacles (lhatsuklha gtsug, sekhargsas mkhar, lhatenlha rten, and tenkharrten mkhar)
  • III. Agricultural Structures
    • 1) Stone irrigation channels
    • 2) Terracing
      • a) Retaining walls
      • b) Partition walls
  • IV. Earthworks
    • 1) Rampart-like walls and platforms
  • V. Rock Art and Epigraphy
    • 1) Petroglyphs
    • 2) Pictographs
    • 3) Inscriptions and ciphers


[1] For the findings of my earlier expeditions see: Bellezza, Antiquities of Upper Tibet; Bellezza, “Gods, Hunting and Society: Animals in the Ancient Cave Paintings of Celestial Lake in Northern Tibet,” East and West 52 (2002): 347-396; Bellezza, Antiquities of Northern Tibet; Bellezza, “Bon Rock Paintings at gNam mtsho: Glimpses of the Ancient Religion of Northern Tibet,” Rock Art Research 17, no. 1 (2000): 35-55; Bellezza, “A Preliminary Archaeological Survey of Da rog mtsho,” The Tibet Journal 24, no. 1 (1999): 56-90; Bellezza, “Notes on Three Series of Unusual Symbols Discovered on the Byang thang,” East and West 47, nos. 1-4 (1997): 395-405; Bellezza, Divine Dyads.
[2] Another crucial archaeological asset of Upper Tibet is rock art, which provides a rich record of the archaic way of life in the region. Dozens of sites in which petroglyphs and pictographs document social, religious and economic facets of early life are distributed over much of Upper Tibet. This graphic evidence also reveals the existence of a distinctive paleoculture, one with strong affinities to surrounding peoples but with it's own idiosyncratic qualities, setting Upper Tibet apart from the steppes and more eastern regions of the plateau. Rock art, a prime indicator of aesthetic values, defines the uniqueness of early Upper Tibet as much as its monumental assemblages. The rock art tableaux spectacularly depict the vitality, resourcefulness and stamina of the past inhabitants of the region. This is certainly something that modern day Tibetans can take pride in and something in which the rest of the world can marvel. A comprehensive inventory of Upper Tibetan rock art was also conducted and will constitute the contents of another volume in the present series in due course.
[3] Much of this section of the work was taken from the text of Bellezza, Zhang Zhung.
[4] On the autumn phase of this expedition, I was accompanied by Döndrup LhagyelDon grub lha rgyal, a highly skilled researcher at the Tibet Academy of Social Sciences.
[5] On this expedition, I had the good fortune of being accompanied by Könchok GyatsoDkon mchog rgya mtsho, a research scholar at the Tibet Academy of Social Sciences.
[6] This section of the work was derived from Bellezza, Zhang Zhung.
[7] Bellezza, Zhang Zhung
[8] This section of the work is also derived from Bellezza, Zhang Zhung.
[9] At present the scant chronometric data do not demonstrate that any of the archaeological sites surveyed date to the late second millennium BCE or earlier. I suspect, however, that this current age limitation will be overcome as the pace of archaeological research intensifies and Bronze Age (especially late Bronze Age) structures can be positively identified. As in Central Tibet, some Upper Tibetan monuments may even prove to date to the Neolithic. An earlier periodization is particularly likely for tombs, because in all adjoining regions where chronometric and collateral archaeological data have been assembled, there are burials that predate the first millennium BCE. Another possible exception to an early Iron Age chronological basement are certain Upper Tibet rock art sites and compositions, which in terms of the techniques of manufacture and style conform to what some Central Asian rock art specialists would consider to be Bronze Age schema.
[10] For reviews of these earlier epochs see Aldenderfer, “The Prehistory of the Tibetan Plateau”; Chayet, Art et Archéologie du Tibet. Sites attributed to the Paleolithic, Mesolithic and Neolithic have been discovered in Upper Tibet, but far more research is needed to determine when the high plateau was first peopled and how these earlier occupations contributed to the later course of civilization in the region.
[11] For the purposes of this study, the Tibetan term bangsobang so is only used to denote burial mounds. In the Tibetan language this term can also be applied to a larger range of burial structures.

Note Citation for Page

John Vincent Bellezza, Antiquities of Zhang Zhung: A Comprehensive Inventory of Pre-Buddhist Archaeological Monuments on the Tibetan Upland (Charlottesville, VA: Tibetan & Himalayan Library, 2010), .

Bibliographic Citation

John Vincent Bellezza. Antiquities of Zhang Zhung: A Comprehensive Inventory of Pre-Buddhist Archaeological Monuments on the Tibetan Upland. Charlottesville, VA: Tibetan & Himalayan Library, 2010.