Tibet Archaeology

and all things Tibetan

Flight of the Khyung

May 2013

John Vincent Bellezza

This month’s Flight of the Khyung soars to some of the highest and oldest points in Tibet. We will visit what may be the loftiest pilgrimage site in the world, looking for evidence of ancient human settlement. There is also a never before published Upper Tibetan megalithic site to discover in this issue. The third and longest feature reviews recent scientific evidence for the origins of the Tibetan people and their occupation of the most elevated plateau on Earth.

Thirteen Golden Reliquaries

Fig.1. The crystal southern face of holy Mount Kailas

Fig.1. The crystal southern face of holy Mount Kailas

The highest altitude pilgrimage in the entire world is very possibly Thirteen Golden Reliquaries (Gser-gdung bcu-gsum), located on the south side of Asia’s holiest peak, Mount Kailas (Ti-se). This pristine site is situated at 5820 m above sea level and a full 1000 m above Darchen, the burgeoning administrative and tourist center at the base of Mount Kailas. Traditionally, one is supposed to make thirteen rounds of Mount Kailas before entering the inner circle (nang-skor) and attempting the walk to Thirteen Golden Reliquaries. However, tourists who now visit the site often disregard this custom.

Fig. 2. The site of Thirteen Golden Reliquaries at the base of the Mount Kailas peak

Fig. 2. The site of Thirteen Golden Reliquaries at the base of the Mount Kailas peak

With the peerless pyramidal form of Mount Kailas as a constant companion, the walk to Thirteen Golden Reliquaries is immersed in beauty and sanctity. These chortens contain the body relics of religious masters  of the Drikung Kagyud, one of the dominant subsects of Buddhism at Mount Kailas. The line of thirteen chortens is set in a cleft at the base of the Mount Kailas peak, like pearls strung across black velvet. The final ascent to the hallowed site is along a very steep trail that actually traverses the face of the mountain. After being desecrated in the Chinese Cultural Revolution, the thirteen reliquary chortens were rebuilt in the 1980s. They occupy a fissure only two or three meters wide, forcing visitors to squeeze through a very narrow space. At a short distance away there is a fourteenth chorten that appears to be of an older design.

Fig. 3. The great cleft in the mountain face and the reliquary chortens

Fig. 3. The great cleft in the mountain face and the reliquary chortens

Fig. 4. The line of reliquary chortens rebuilt in the 1980s

Fig. 4. The line of reliquary chortens rebuilt in the 1980s

As it now stands, Thirteen Golden Reliquaries is a thoroughly Buddhist site. Before the 11th century CE, it appears that this belonged to the Bonpo. Presumably, these monuments in their original form were the thirteen reliquary chortens (sku-gdung mchod-rten bcu-gsum) noted in the Bon sacred geographic tradition. However, I have not found archaeological evidence to support this idea. Thirteen Golden Reliquaries is located well above the elevation of archaic temples and hermitages in the region, so no direct association with ancient residential patterns can be established. The climate at 5800 m is extremely harsh with snowfall and subfreezing temperatures occurring in all months of the year. Also, air pressure at this elevation is too low to sustain human life on a permanent basis. In ancient times, like today, the use of this location would have been limited to ritual and ceremonial functions of short duration.

Despite its contemporary architectural character, places near Thirteen Golden Reliquaries are rich in Bon mythology. Bon texts report that the front (south) side of the great crystal chorten that is Mount Kailas was graced by the founder of Bon (Tonpa Shenrab) and the first king of Zhang Zhung (Cakki Charucen, Lcags kyi bya-ru can) as well as other famous religious personages. Ancient scriptures written in the Zhang Zhung language are supposed to have been secreted in the vicinity of Thirteen Crystal Chortens as well. Given the high density of archaic habitational sites in the nearby area and its mythic renown, the cleft at the base of Mount Kailas must have been known to ancient inhabitants whether or not they had built shrines there.

Fig. 5. The southeastern view from Thirteen Golden Reliquaries. To the right of Mount Kailas, a tiny bit of Lake Mapang Yumtsho (Ma-pang g.yu-mtsho) can been seen

Fig. 5. The southeastern view from Thirteen Golden Reliquaries. To the right of Mount Kailas, a tiny bit of Lake Mapang Yumtsho (Ma-pang g.yu-mtsho) can been seen

Fig. 6. The archaic edifice in closest proximity to Thirteen Golden Reliquaries. We decided to name this structure Nandi Dokhang

Fig. 6. The archaic edifice in closest proximity to Thirteen Golden Reliquaries. We decided to name this structure Nandi Dokhang

Fig. 7. An intact portion of the all-stone corbelled roof on the uppermost tier of Nandi Dokhang. Note the long bridging stone precariously supported by two corbels

Fig. 7. An intact portion of the all-stone corbelled roof on the uppermost tier of Nandi Dokhang. Note the long bridging stone precariously supported by two corbels

The ruined all-stone corbelled edifice of Nandi Dokhang is located on a narrow rocky shelf approximately 15 m above a rushing stream. It is situated at 5330 m and has a southern aspect. The large rock formation of Nandi protects the site from northern winds coming off glaciated peaks, producing a fairly hospitable  microclimate. Measuring 12.3 m by 11 m, this structure was probably split in four levels and contained a number of rooms. The poorly preserved building seems to have been damaged by avalanches especially on its northeast and northwest sides. The southeast side has also entirely collapsed. The upper two tiers are better preserved than the lower part of the structure. External walls reach a height of 1.5 m and there are still some corbels clinging to them.

Nandi Dokhang had upwards of ten rooms perhaps indicating that it was inhabited by a large family or other collective group. The time, skills and resources needed to build and maintain such edifices indicates that they were built by an economically prosperous faction of ancient society. The location of Nandi Dokhang amid holy landforms suggests that it was occupied by those of high prestige. This is borne out by the location of the residence far away from prime pastoral and agricultural lands. The occupants must have commanded whatever resources they needed thanks to a superior social position.

Nandi Dokhang is typical of detached archaic residential structures in Upper Tibet. This type of all-stone corbelled architecture (called dokhang) circumscribes the geographic extent of the paleoculture traditionally known as Zhang Zhung. As noted in last month’s newsletter and other publications, all-stone corbelled structures in Upper Tibet can be broadly attributed to 500 BCE to 500 CE (Iron Age and protohistoric period). They may have been continually occupied as late as 1000 CE and the definitive conversion of the region to Buddhism. While Buddhist ascetics made adventitious use of some dokhang for their meditations, the vulnerable physical location of Nandi Dokhang militates against such later occupancy.

Pillars of the ancients

Fig. 8. The megaliths or stelae known as Three Long Stones (Rdo-ring gsum). Note the orange climax lichen growing on the two larger pillars

Fig. 8. The megaliths or stelae known as Three Long Stones (Rdo-ring gsum). Note the orange climax lichen growing on the two larger pillars

Fig. 9. The highly eroded pillars of Three Long Stones

Fig. 9. The highly eroded pillars of Three Long Stones

The three ‘long stones’ of Doring Sum were surveyed in 2009 and are therefore not featured in Antiquities of Zhang Zhung volumes (www.thlib.org/bellezza). These three pillars are situated at 4640 m above sea level, on the southeastern fringe of the vast Upper Tibetan landscape modified by standing stones. There are  two types of pillar monuments most characteristic of ancient Upper Tibet: 1) array of pillars appended to a temple-tomb, and 2) pillars erected inside a quadrate enclosure. Both of these monument types possess funerary functions. However, Doring Sum does not exhibit morphological traits of either of these genres. Evidently, it can be classed among the category of pillars erected isolated from all other structures. The original functions of this less distinctive class of standing stones are unknown. According to the native oral tradition, they were used as border markers, ritual offerings to local deities, registers for oath taking, and as funerary ritual constructions.

The raising of stone pillars in Upper Tibet may have begun in the first third of the first millennium BCE and continued as an anachronistic cultural expression until the historic era, circa 650 CE. For more information on the chronology of the Upper Tibetan pillars, see my publications such as Flight of the Khyung, February 2012: www.tibetarchaeology.com/february-2012/

As is customary in highland Tibet, the three pillars of Doring Sum are uninscribed and otherwise unadorned. They were planted in now sandy, level terrain with wide vistas to the east and west. To the north is a rocky ridge that constitutes one side of the broad valley. The three pillars are evenly spaced, and form a row that is 1.7 m in length and oriented north to south. The north pillar (70 cm [height] x 95 cm [basal girth]) is tabular and tan in color. The middle pillar (1.3 m x 1.3 m) is tabular and white in color. A long strip along the axis of the stone has been removed. The white south pillar (1.3 m x 1.3 m) has four sides of more or less even width.

The origins of the Tibetan people

 A review of “Modeling the Neolithic on the Tibetan Plateau” by Mark Aldenderfer

This article introduces some of the latest research being carried out to determine when the Tibetan plateau was first peopled, a hotly debated topic among archaeologists and geneticists. The material and molecular evidence compiled to date indicates that humans reached the Tibetan plateau sometime in the Upper Paleolithic (12,000–30,000 year ago). A growing body of scientific evidence suggests that these people settled on a permanent basis, establishing a genetic legacy still discernable today.

 An excellent starting point on the subject of human origins on the Tibetan Plateau is a paper by Professor Mark Aldenderfer published several years ago: 2007. “Modeling the Neolithic on the Tibetan Plateau” in Late Quaternary Climate Change and Human Adaptation in Arid China, Developments in Quaternary Sciences, vol. 9, (eds. D. B. Madsen, F. H. Chen and X. Gao), pp. 149–161: Elsevier. Aldenderfer’s paper warrants close attention, as it is the first attempt to suggest the lines along which a more comprehensive picture of the Tibetan Neolithic (New Stone Age) will become feasible. The Neolithic ‘package’ is of course characterized by sedentary settlement, cultivation of food crops, animal husbandry and, in its fully developed form, ceramics production.

Aldenderfer begins by briefly reviewing archaeological evidence for Neolithic settlement in three major regions of the Tibetan Plateau: Amdo (Qinghai Plateau), Khams and Central Tibet. Neolithic sites in these three regions have been dated using chronometric means to a maximum of 6700, 5800, and 3800 years ago respectively. I estimate that around 40 sites identified as ‘Neolithic’ have been discovered on the Tibetan Plateau and studied to varying degrees.

For a review of some important Neolithic sites in Tibet, see:

1)     Chayet, A. 1994. Art et Archéologie du Tibet. Paris: Picard. Consult pp. 34–55.

2)     Aldenderfer, M. and Zhang Yinong. 2004 “The Prehistory of the Tibetan Plateau to the Seventh Century A.D.: Perspectives and Research from China and the West Since 1950” in Journal of World Prehistory, vol. 18, no. 1, pp. 1-55. Springer: Netherlands. Consult: pp. 26–34.

Aldenderfer (2007) aptly points out that our knowledge of the Neolithic in Tibet is still very much in its infancy. There are only two sites, Kharro (Mkhar-ro in eastern Tibetan) and Chugong (Chu-gong in central Tibet), that have been excavated to any degree of thoroughness. Even scientific data from these two places are patchy, furnishing, for example, incomplete floral and faunal records of domestic crops and animals presumably reared there. Stratigraphic records from Kharro and Chugong are also less than meticulous. Although earlier Neolithic sites have been discovered in Amdo (their cultural affinities with lower elevation contemporaneous Neolithic sites Qinghai and Gansu are still being assessed), it is at Kharro and Chugong that attributes of a well developed sedentary agrarian society circa 4000–1100 BCE become fully manifested. Citing various paleoclimatological studies, Aldenderfer notes that in the early period of this era the climate was warmer and wetter than today, while assuming in the later period the dry and cold conditions still prevalent on the Tibetan plateau.

If Kharro represents an indigenous Tibetan culture, as theorized by its excavators, they reason it must have sprung from an antecedent hunting and foraging culture. Aldenderfer offers two caveats to this hypothetical ethnogenesis: 1) so far very few archaeological data has been compiled for the period from 10,000 to 5000 BCE, thus almost nothing concrete is known about any antecedent culture; and 2) millet and domestic pigs, ostensibly two staple food sources at Kharro, have lowland origins. This assessment serves as a backdrop to Aldenderfer’s main thesis: any ‘indigenous’ people of the Tibetan Plateau had to arrive originally from somewhere else (an allusion to the out of Africa theory and the spread of modern humans) and they are liable to have interacted with surrounding lowland peoples over a long stretch of time through demic diffusion and various forms of cultural transmission. This inclusive view of the peopling of the Tibetan plateau is a fairly safe one, for it is in line with much of the molecular and archaeological evidence thus far assembled. Operating from Aldenderfer’s perspective encourages the formulation of a dynamic model of human occupation on the Tibetan plateau, one that allows demic and cultural layering and interpenetration over many thousands of years.

Aldenderfer goes on to  examine several hypotheses proposed by Su et al., Ren, George van Driem for the spread of humans from lowland China, exposing certain inherent weaknesses in their positions. For their studies, see:

1)     Bing Su, Chunjie Xiao, Ranjan Deka, et al. 2000. “Y chromosome haplotypes reveal prehistorical migrations to the Himalayas” in Human Genetics, vol. 107, pp. 582–590.

The authors subscribe to the view that approximately 8800 years ago, the Yangshao (proto-Yangshao) culture cultivating millet served as the ‘mother’ culture of the Sino-Tibetans. Sino-Tibetan, as used in this paper, implies a genetic, linguistic and cultural interrelationship between Tibetans and the Han. The authors say that the proto-Tibeto-Burmans who dwelt in the upper Yellow River basin possessed characteristics of this Yangshao cultures. The paper holds that about 6000 years ago (after the split of the Sinitic and first Tibeto-Burman populations) the proto-Tibeto-Burmans left their homeland in the Yellow River basin due to population pressures and moved westward and then southward to occupy the entire Tibetan Plateau. Unfortunately, this narrative is flawed in a number of ways. First of all, it conflates languages, cultures and genes for which there is still little hard evidence. Secondly, it maintains that just one culture, the Yangshao, gave rise to all the many cultures of the Tibeto-Burmans. By the study’s own acknowledgement, no fewer than three major cultures, the Yangshao one among them, were the sources of Chinese civilization. On other possible routes used by Tibet-Burmans to gain access to the Tibetan plateau, see Wang et al., bibliography at the bottom of this review, paper no. 1. On the fallacy of millet farmers being the sole bearer of culture to a vast area of Inner Asia and the Himalaya, see below in the text.

2)     Ren, G. 2000. “Decline of the mid- to late Holocene forests in China: climatic change or human impact?” in Journal of Quaternary Science, vol. 15 (no. 3), pp. 273–281.

In this paper, Ren argues that the decline of forests in Tibet starting about 5000 years ago was mostly caused by human activities, as climate change factors in this period appear to be marginal. The actual role of humans in the depletion of forests in Tibet has yet to be determined, but it is unlikely to be as extreme as proposed by Ren. See Schlütz, F. and Lehmkuhl, F. 2009, bibliography at the bottom of this review, paper no. 10.

3)     van Driem, G. 2001. Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region containing an Introduction to the Symbiotic Theory of Language, vol. 1, Leiden: Brill.

4)     van Driem, G. 1998. “Neolithic correlates of Tibet-Burman migrations” in Archaeology and Language II. Archaeological Data and Linguistic Hypotheses (eds. R. Blench, M. Spriggs), pp. 68–102. London: Routledge.

This erudite study correlates the convoluted linguistic history of Tibeto-Burman languages with the tortuous and exceptionally long peregrinations of the Majiayao Neolithic people as envisioned by the author. Van Driem brackets his study by readily acknowledging that his proposed migration routes are speculative and open to further revision and even rejection, should additional archaeological and linguistic evidence demand it (pp. 69, 86). He also explicitly recognizes that the spread of a material culture does not necessarily indicate the dissemination of a language or people (p. 69). Nevertheless, van Driem seems to ignore this axiom when moving a singular culture across Xinjiang, Kashmir, Swat, Tibet, and the Himalaya. According to the author, the Majiayao people culture (3900–2700 BCE), possibly because of climate change, ecological distress or the allure of a better life in foreign lands, left Gansu on an epic journey (Völkerwanderung) that took them in two main directions: 1) west across Xinjiang to Kashmir and Swat to establish the ‘Northern Neolithic’, before turning southeast to Bhutan via the cis-Himalaya; and 2) south to eastern Tibet, Bhutan and Sikkim. The author maintains that these monumental migrations of the Majiayao population account for all Bodic languages and those who speak them.

As Aldenderfer makes clear, Ren’s ideas on anthropogenic deforestation of the Tibetan plateau and van Driem’s historical linguistic hypothesis are impaired by an outdated  assumption, which equates archaeological cultures with real, once living and breathing ones. An archaeological culture (as defined through coherent assemblages of material objects, esthetic motifs and burial practices) cannot in itself be used as a marker of ethnic and linguistic affiliation. Simply put, this is because archaeological materials over wide areas are often fungible entities exchanged in a variety of ways by different groups of human beings. Ren’s and van Driem’s transferring  of an entire set of cultural, genetic and linguistic traits present in lowland China to the Tibetan plateau or to Kashmir because the people living in these regions share interrelated archaeological cultures is therefore unwarranted.

To digress a bit: the archaeological axiom I have just described explains my resistance to recent claims that Iron Age cultures of Xinjiang displaced native variants in western Tibet. Likewise, I view with skepticism that four golden burial masks recently discovered on both sides of the Himalayan divide all belong to the ‘same culture’. In short, these kinds of claims are to me examples of cultural reductionism by virtue of not taking into account the rich ethnological mosaic of the region in the Iron Age, as supported by the extremely diverse archaeological and anthropological records. Rather, given their distinctive stylistic traits and geographic sources, the four golden masks are more likely to be examples of cultural variability. While the various makers probably did share common cultural traditions, trade connections and perhaps even political and linguistic ties, it not prudent to lump them into a singular ‘culture’ exhibiting precisely the same language, ethnicity, life-ways, cultural traits, and self-perceived identity.

Returning to the Neolithic, van Driem’s model (and the notion of a Proto-Bodic Majiayao culture) postulates the mass-migration of Majiayao Neolithic peoples (3400–2800 BCE) as the mechanism accounting for the diffusion of Tibetic languages across the Tibetan plateau. Nonetheless, as Aldenderfer mentions in passing, large-scale demic diffusion is only one of a number of ways in which languages may possibly spread. I would observe that small-scale demic diffusion associated, say, with a foreign ruling elite, can also potentially introduce a new language, as could possibly a priestly class propagating new religious traditions. It must also be pointed out that archaeological evidence for the Majiayao culture on the Tibetan Plateau above 3500 m in elevation is scant at best. Moreover, other sources for Tibetic languages may also be implicated, such as the penetration of Neolithic Tibeto-Burman languages and peoples from the east directly into southern Tibet.

For the reasons outlined above, van Driem’s belief (1998: 83, 84) that double-perforated rectangular harvesters and semi-lunar shaped knives discovered in northern Sikkim are solid evidence for the migration of Majiayao proto-Tibeto-Burman speakers to that region cannot be sustained on its own merits alone. A bevy of alternative theories on how and why these stone implements reached Sikkim can be developed encompassing trade, tribute, borrowing technologies, independent innovation, etc. The same general arguments also compel us to seek alternatives to the wholesale movement of a single population to explain similarities between the Majiayao Neolithic and those of Swat and Kashmir.

Aldenderfer (also see Aldenderfer and Zhang 2004, pp. 38, 39) expends considerable effort to demonstrate that the models proposed by Su Bing et al. (based on the distribution of Y haplotypes in East Asia), Ren and van Driem concerning the march of humanity across the great breadth of the Tibetan plateau in the Neolithic, as if filling a void, are little concerned with the adaptive strategies necessary for the successful  migration to the highlands and the lasting occupation of the entire Tibetan plateau. As Aldenderfer points out, much of the Tibetan plateau is only suited for pastoralism and this could hardly have been adopted by sedentary agriculturalists genetically and technologically ill-suited to the high altitude environment of the highlands. Essentially, Aldenderfer argues that subsistence strategies and interrelated physiological responses develop in tandem over time in specific physical environments, and cannot be picked up and moved en masse.

I concur with this general principle articulated by Aldenderfer. While economic, technological and ideological components of a culture can be highly mobile and have a huge impact on others, cultures in toto are not interchangeable. The inexorable advance of low-elevation populations onto the Tibetan plateau advocated by Su Bing et al. Ren and van Driem may indeed help explain the current genetic and linguistic traits of Tibetans, however, they are incomplete reckonings of their origins and in need of revision in light of the matters discussed in this review.

Chronological and functional aspects of models proposed by Madsen et al. (paper no. 1 in the bibliography below) and by Brantingham et al. (paper no. 2 in the bibliography below) for the colonization of the Tibetan plateau are also scrutinized by Aldenderfer. These researchers advocate a three-step process, whereby settlers first gained a foothold on the northern fringe of the Tibetan Plateau up to an elevation of 2000 m prior to 25,000 years ago, before ascending to a higher altitude as part of specialized forays, circa 14,000 to 25,000 years ago. According to these authors, this sets the stage for the permanent tenure of the entire Tibetan Plateau in the early Neolithic by pastoralists. Unlike the models proposed by van Driem and Su Bing et al., the one that Madsen et al. and Brantingham et al. champion is underpinned by a sophisticated analysis of the physical and cultural factors incumbent in biogeographical adaptation, a process driven by the mechanisms of ‘directional selection’ and ‘adaptive radiation’. In light of the archaeological and molecular evidence, the potential problem with their model is not one of dynamics per se, but on their insistence that humans did not people the Tibetan plateau on a permanent basis until approximately 10,000 year ago. In his paper, Aldenderfer questions the timeline of permanent occupation of the Tibetan plateau they proffer by referring to Chusang, located 80 km north of Lhasa, where uranium-thorium isochron and OSL analyses of sediments suggest that permanent human occupation of the site occurred 21,000 to 28,000 years ago. As Aldenderfer observes, the site is situated too far away to have been reached by bands of Paleolithic foragers following a migratory circuit that began and ended on the northern fringe of the Tibetan plateau.

1)     Madsen, D. B.,  Ma Haizhou, Brantingham, P. J. et al. 2006. “The Late Upper Paleolithic occupation of the northern Tibetan Plateau margin” in Journal of Archaeological Science, vol. 33, pp. 1433–1444.


This archaeological study breaks new ground in understanding Upper Paleolithic settlement patterns and cultural characteristics around Qinghai Lake (Mtsho-sngon). It is founded on impeccable fieldwork methodology all the more remarkable for the logistical challenges posed by the studied area.

2)     Brantingham, P. J., Xing Gao, Olsen, J. W. 2007. “A short chronology for the peopling of the Tibetan Plateau” in Late Quaternary Climate Change and Human Adaptation in Arid China Developments in Quaternary Sciences, vol. 9, (eds. D. B. Madsen, F. H. Chen and X. Gao), pp. 129-150. Elsevier.

3)     Paper no. 2 cited by Aldenderfer was not available to me at present, however, I was able to access its predecessor online: Brantingham, P. J and Gao Xing. 2006. “Peopling of the northern Tibetan Plateau” in World Archaeology, vol. 38, (no. 3: Archaeology at Altitude, pp. 387–414.


This latter paper introduces scientific data gathered through extensive fieldwork in northwestern China to adduce the origins of the first Tibetan population, using the signature three-step model of the authors. It is argued that the first fulltime occupation of the Tibetan plateau occurred after 8200 BCE. Such a late date, say the authors, is the result of the time required for humans to develop the evolutionary responses needed for the biogeographic adaptations. The authors hypothesize that these early foraging migrants were probably pushed out of the lowlands by an expanding agricultural population.

Recent human genomic studies tend to support Aldenderfer’s contending view by also placing permanent human settlement of the Tibetan Plateau in the Late Paleolithic. Genetic inflows into Tibet are traced to roughly 22,000 years ago but only become very pronounced in the Epipaleolithic and Early Neolithic, approximately 7000 to 9000 years ago. The prospect of a mass introduction of human genes into Tibet in the last 6000 years is a matter of ongoing debate among geneticists and evolutionary biologists. See references to recent genetic studies at the end of this newsletter.

Archaeological studies conducted by the Chinese Academy of Sciences in the 1970s and 1980s based on the typological analysis of stone tools (core and blade technology) discovered on the surface, in the middle and western portion of the Tibetan plateau, also attribute human occupation to the Paleolithic. While these older studies suffer from serious methodological deficiencies, they must also be considered in light of the total picture. For a review of archaeological data indicating a Paleolithic occupation of the Tibetan plateau, see:

1)     Chayet, A. 1994, pp. 25–34.

2)     Aldenderfer and Zhang. 2004, pp. 15–21.

The Madsen et al. and Brantingham et al. papers cited above discount much of the lithic evidence taken by some researchers to corroborate a Paleolithic presence on the Tibetan plateau before the Last Glacial Maximum (LGM: 18,000 to 23,000 years ago). According to them, most of these stone tools probably date to after the LGM and reflect the existence of small foraging groups operating from lower elevation residential bases. Madsen et al. (2006: 1441, 1442) hold that Paleolithic foragers were traveling 300 to 400 km from their permanent lowland homes to the Tibetan plateau on a revolving basis. To support this hypothesis, the authors cite studies indicating that Upper Paleolithic dwellers in eastern and central Europe transported stones used to make tools similar distances to where they lived more generally. I am not convinced though by this argument. The plains of Europe are one thing, the Tibetan plateau another. In order to travel from any Paleolithic sites in the middle of the Tibetan Plateau northward to the lower climes of Qinghai, these itinerants would have to trudge at least 1200 km over the highest tableland and traverse some of the highest passes in the world. These distances are simply too great, the terrain too tough, the altitudes to high, and the energy expenditures involved too excessive to make any such foray worthwhile and productive.

It seems that if there were Paleolithic peoples in the middle of the Tibetan Plateau that is where they lived and died. Furthermore, Brantingham and Gao (2006: 408) tender evidence from the lithic assemblage that I read as possibly indicative of permanent Paleolithic settlement rather than the roving humans of their ‘competitive exclusion’ model. This pertains to what appears to be indigenous technological innovations exhibited in stone blades. In the words of these authors: “The large, flat blade and bladelet technologies of the Chang Tang are derived from a late Upper Palaeolithic (sic) substrate, but present attributes that are unknown in the low-elevation environments that surround the plateau.”

In his paper, Aldenderfer goes on to comment on the “northward bias” of the models discussed above. He appeals to readers to consider alternative avenues of entry onto the Tibetan Plateau during the Neolithic, which do not entail migrations from northwest China to the Qinghai portion of the tableland. His call to ponder other routes of entry is reinforced by recent genetic studies indicating that early Neolithic Tibet-Burman groups migrating to the Plateau did so via the southeastern flank of the Tibetan plateau (see Wang et al. 2011, in bibliography at end of newsletter, paper no. 1). As Aldenderfer suggests, even migrations via the western fringe of the Tibetan Plateau during the Neolithic should not be dismissed out of hand. Indeed, if evidence for areal genetic differentiation among Tibetans (as some studies seem to indicate) is confirmed, this may be attributable to early settlers coming to the Tibetan plateau from different directions and/or to successive waves of migrants using the same geographic gateway. As regards the ethnogenesis of the Metal Age ‘Zhang Zhung’ polity, I suggest multifarious geographic origins in light of the clan and lineage history of Upper Tibet (Zhang Zhung 2008: 259, 260).

Aldenderfer also briefly discusses the origins of barley cultivation on the Tibetan plateau and the prospects that it constitutes a secondary center of domestication for this cultigen. To buttress his arguments that humans permanently inhabited the Tibetan Plateau before the developed Neolithic (circa 3500 BCE), he suggests that genetic changes regarding lactase persistence, adaptation to hypoxic conditions and barley cultivation may best be explained by a longer period of tenure. As pertains to genetic changes associated with an oxygen-poor environment, his position is in agreement with findings made by Shuhua Xu et al. 2011 (see bibliography below, paper no. 4).

In conclusion, Aldenderfer calls for a more nuanced model to account for the human occupation of the Tibetan plateau. Given the evidence he puts forth in his paper, a more comprehensive analysis and model is certainly called for if research on the origins of the Tibetans is to be optimally productive. Most certainly, further scientific exploration of early human sites on the Tibetan plateau is the order of the day. Sites previously identified as Paleolithic should be revisited and subjected to systematic  surveying and excavation. This must be done using rigorous analytical techniques drawn from a wide range of scientific disciplines.

Since the late 1990s, one of the most promising scientific approaches to Tibetan origins is human, animal and plant genomic research. However, this is a field still in the making. Certain findings from studies on the human genome are not infrequently contradictory and subject to ongoing analyses and revision. It is also fairly common for the experimental methodology of one research team to be called into question by another. In fact, very little already published in the field of genetics on the historical origins of Tibetans may stand the test of time.

Of high priority, are further assessments of changes in allele (genes responsible for hereditary variation) frequencies associated with both mitochondrial DNA and Y chromosomes, in order to determine the time of arrival of particular human lineages into Tibet and their geographic distribution. Nevertheless this remains a very challenging proposition because scientists must grapple with extremely small genetic differences (most alleles are widely distributed) to determine the history of human populations. Improving the technology and methodology currently being employed as well as new breakthroughs in understanding human ancestry on a molecular level are required.

The good news is that the pace of research on the origins of the Tibetan people and their livestock and crops has accelerated in the last five years. This should lead to better answers for many of the questions raised in this review and of course will throw up new ones as well.

I shall now cite a few recent scientific studies for those thirsting to know more about the Tibetan genome. Looking at these studies as a whole it becomes apparent that they often apply linguistic terminology to explain their findings, and they sometimes even correlate the genetic evolution of Tibetans with the historical development of languages. Terms like Sino-Tibetan or Tibeto-Burman to describe genetic source populations conflate genes with languages, for which there is little or no direct scientific evidence. I think it would be helpful if those working on the human genome were to devise a terminology shorn of linguistic connotations to denote the historical trajectories of human populations. This would reflect a maturing of the field.

1)     Wang B, Zhang Y-B, Zhang F, Lin H, Wang X, et al. 2011. “On the Origin of Tibetans and Their Genetic Basis in Adapting High-Altitude Environments” in PLoS One, vol. 6 (no. 2).


This paper summarizes recent research findings on the Tibetan genome. The paper concludes that Tibetans have developed distinctive biological traits to offset the stress of hypoxia, a process involving an increase in the rate of fixed adaptive alleles over approximately 1100 generations (22,000 years). This study also finds that the overwhelming bulk of the Tibetan genome is of more recent origins. The authors provide a résumé of the genetic ontology of the Tibetans. In their own words: “Population genetic structure analyses suggested that Tibetans share the common ancestors with East Asian populations, but not Central/South Asian populations who settled on the western and southern side of Himalayas. Our finding is consistent with the results of a previous study which suggested gene-flow inhibition caused by the Himalayas… After the ancestors of Sino-Tibetans reached the upper and middle Yellow River basin, they divided into two subgroups: Proto-Tibeto-Burman and Proto-Chinese… The ancestral component which was dominant in Tibetan and Yi arose from the Proto-Tibeto-Burman subgroup, which marched on to south-west China and later, through one of its branches, became the ancestor of modern Tibetans.” The authors hypothesize that the major migration route to the Tibetan plateau was via the valleys of the Hengduan mountains.

2)     Mian Zhao, Qing-Peng Konga, Hua-Wei Wang et al. 2009. “Mitochondrial genome evidence reveals successful Late Paleolithic settlement on the Tibetan Plateau” in PNAS, vol. 106 (no. 50).


This paper maintains that the vast majority of matrilineal genetic components of Tibetans can be attributed to Epipaleolithic and Neolithic immigrants from northern East Asia in the mid-Holocene. The Mian Zhao et al. study finds that Paleolithic inhabitants of the Tibetan plateau (some 21,00 years ago) made a limited genetic contribution to the modern Tibetan genome. In particular, haplogroup M16 appears to be an authentic Paleolithic relic of an autochthonous Tibet plateau population. Each haplogroup shares a common distant ancestor as indicated by the interrelated group of haplotypes upon which it is composed. A haplotype is comprised of adjacent loci on a chromosome marked by the same set of single-nucleotide polymorphisms (an SNP consists of a single variation in the DNA sequence of chemical bases), which are inherited together. This study claims that most Paleolithic humans somehow disappeared before the time of Neolithic genetic inflows.

3)     For a synopsis of a lecture given by Su Bing touching upon the Paleolithic antiquity of components of the Tibetan genome, entitled “Genetic studies of human origins and adaptation of the Tibetan Plateau”, see  www.tibetarchaeology.com/september-2011/

4)     Shuhua Xu, Shilin Li, Yajun Yang et al. 2011. “A Genome-Wide Search for Signals of High-Altitude Adaptation in Tibetans” in Molecular Biology and Evolution, vol. 28 (no. 2), pp. 1003–1011.


This is a technical analyses of recent studies that examine natural selection pressure of hypoxia on the Tibetan Plateau. Of special note is the criticism of highly publicized claims made by Yi et al. (2010:5) that the Chinese and Tibetans genetically diverged as distinct populations just 2750 years ago. Simply speaking, Shuhua et al. argue that the Yi et al. study is compromised by exclusively modeling exome (the very small portion of the genome made up of exons) data, as well as by several fallacious technical assumptions and by quality control issues. They hold that the genetic basis of physiological mechanisms for coping with a high altitude environment is best seen as part of a very long-term process, beginning with the first settlers of the Tibetan plateau in the Paleolithic.

5)     Yi X, Liang Y, Huerta-Sanchez E et al. (65 co-authors). 2010. “Sequencing of 50 human exomes reveals adaptation to high altitude” in Science, vol.  329, pp. 75–78.


6)     Yi Ping, Zhaohui Yang, Hui Zhang et al. 2011. “Genetic Variations in Tibetan Populations and High-Altitude Adaptation at the Himalayas” in Molecular  Biology and  Evolution, vol. 28 (no. 2), pp. 1075-1081. http://mbe.oxfordjournals.org/content/28/2/1075.full

This study furnishes further confirmation of the detection of very ancient genetic responses to hypoxia among Tibetans. A sequence analysis of the hypoxia-related gene EPAS1 suggests strong selection pressure of long-term duration. This is observed to be consistent with a Paleolithic peopling of Tibet still recognizable in the modern genome of Tibetans. The authors say they are planning a more detailed analysis of their findings.

7)     Aldenderfer, M. 2011. “Peopling the Tibetan Plateau: Insights from Archaeology” in High Altitude Medicine & Biology, vol. 12, (no. 2), pp. 141–147.

This paper relying on an analysis of archaeological data, also contradicts the Yi et al. 2010 findings as per the recent genetic divergence of Tibetans and Chinese.

8)     Hong Shi, Hua Zhong, Yi Peng, et al. 2008. “Y chromosome evidence of earliest modern human settlement in East Asia and multiple origins of Tibetan and Japanese populations” in BMC Biology, vol. 6 (no. 45).


This study reconstructs the phylogeography of the D-M174 patrilineage, which defines a sub-haplogroup shared primarily by today’s Tibetans, Japanese and Andaman islanders. The authors state that D-M174 constitutes an extremely ancient lineage of modern humans in East Asia, which may be as much as 60,000 years old. Moreover, the divergence of these peoples is also very ancient, as they belong to different sub-lineages of D-M174, each of which is marked by varying gene mutations. The Tibetan sub-lineage appears to have reached the Tibetan plateau about 30,000–40,000 years ago. To explain the current fragmented geographic distribution of the D-M174 lineage, the authors vaguely suggest that adaptation to extremely cold conditions during the last glacial epoch (LGM) and the Neolithic population expansion of other groups may have been the causes.

9)     Stanyon, R., Sazzini, M., Luiselli, D. 2009. “Timing of the first migration into eastern Asia” in Journal of Biology, vol. 8 (no. 18).


A review broadly endorsing the findings of Hong Shi et al. The review authors however also sound a note of caution: “… dates inferred from present-day genetic diversity can vary greatly, as a result of unknown differences, in variables such as population size, rates of genetic drift, gene flow, and the presence of selection.” It is important to note that the factors outlined above can more generally hamper any historical reconstruction of human populations based on genetic data.

For preliminary scientific studies suggesting that human activities began to modify the vegetative geography of Tibet 8800 to 7200 years ago, see:

10)  Schlütz, F. and Lehmkuhl, F. 2009. “Holocene climatic change and the nomadic Anthropocene in Eastern Tibet: palynological and geomorphological results from the Nianbaoyeze Mountains in Quaternary Science Reviews, vol. 28 (nos. 15, 16), pp. 1449–1471. Elsevier.

11)  Georg Miehe, Sabine Miehe, Knut Kaiser. 2009. “How old is pastoralism in Tibet? An ecological approach to the making of a Tibetan landscape” in Palaeogeography, Palaeoclimatology, Palaeoecology, vol. 276, pp. 130–147.


12) Xuebin Qi, Chaoying Cui, Yi Peng et al. 2013. “Genetic evidence of Paleolithic colonization and Neolithic expansion of modern humans on the Tibetan Plateau” in Molecular Biology and Evolution, first published online May 16, 2013.


This highly ambitious study appeared after the May newsletter had been published. The paper is not yet available as a print copy. I present the authors’ abstract below, as given on the Molecular Biology and Evolution website. This study adds to the growing body of molecular evidence indicating that the Tibetan population retains a genetic signature originating in the Upper Paleolithic, some 30,000 years ago. The genetic data provided furnishes solid evidence for Paleolithic settlers on the Tibetan Plateau having survived the LGM to contribute to the formation of the present-day Tibetan population. A rapid growth in population accompanied by gene flow from outside the Tibetan plateau occurred in the early Neolithic, giving rise to the bulk of the Tibetan genome as it now stands. Referring to recent molecular and archaeological studies on the subject, the paper concludes that both barley and yaks were domesticated on the Tibetan Plateau, indicating that this region produced a distinctive Neolithic cultural complex, and was not merely the recipient of agricultural technologies carried by lowland peoples from East Asia. It should be noted, however, that the study’s attribution of yak domestication to the early Neolithic is still uncertain. This paper also provides divergence times for the Han and the Tibetans from founding Y chromosome and mtDNA haplogroups, the most extensive analysis of its kind to date.

Tibetans live on the highest plateau in the world, their current population size is nearly 5 million, and most of them live at an altitude exceeding 3,500 meters. Therefore, the Tibetan Plateau is a remarkable area for cultural and biological studies of human population history. However, the chronological profile of the Tibetan Plateau’s colonization remains an unsolved question of human prehistory. To reconstruct the prehistoric colonization and demographic history of modern humans on the Tibetan Plateau, we systematically sampled 6,109 Tibetan individuals from 41 geographic populations across the entire region of the Tibetan Plateau and analyzed the phylogeographic patterns of both paternal (n = 2,354) and maternal (n = 6,109) lineages as well as genome-wide SNP markers (n = 50) in Tibetan populations. We found that there have been two distinct, major prehistoric migrations of modern humans into the Tibetan Plateau. The first migration was marked by ancient Tibetan genetic signatures dated to around 30,000 years ago, indicating that the initial peopling of the Tibetan Plateau by modern humans occurred during the Upper Paleolithic rather than Neolithic. We also found evidences for relatively young (only 7-10 thousand years old) shared Y chromosome and mitochondrial DNA haplotypes between Tibetans and Han Chinese, suggesting a second wave of migration during the early Neolithic. Collectively, the genetic data indicate that Tibetans have been adapted to a high altitude environment since initial colonization of the Tibetan Plateau in the early Upper Paleolithic, before the Last Glacial Maximum, followed by a rapid population expansion that coincided with the establishment of farming and yak pastoralism on the Plateau in the early Neolithic.

13) Gayden, T, Cadenas, A. M., Regueiro, M., et al. 2007. “The Himalayas as a Directional Barrier to Gene Flow” in The American Journal of Human Genetics vol. 80 (no. 5), pp. 884–894.


This study provides a broad view of the Tibetan genome, which is useful in understanding how Tibetans fit into the phylogeographic patterns of Asia. Based on the fairly wide sampling of DNA, this paper also examines the effect of the Himalayan Range on human migration. It concludes that this physical barrier was permeable to various Tibeto-Burman groups (mostly of northeastern Asian origins) periodically but effectively blocked migration northwards of peoples with Indian Y haplogroups (many of which are of Eurasian origins). The most pronounced genetic difference between Tibeto-Burman populations and the Newar and other South Asian populations of Kathmandu is the high frequency of haplogroup D in the former and haplogroups J and R in the latter. According to the study, the presence of haplogroups N and Q in Tibet (commonly found in Siberia and Mongolia) indicate gene flows into Tibet from the north and/or east. However, genetic contributions from Mongolia have been shown to be minimal in more recent genetic studies. The study also postulates that the penetration of haplogroup D subclades into Tibet occurred in two separate demographic events dating to the Epipaleolithic and Neolithic. This proposed chronology however has also been called into question by more recent genetic studies. See for example study No. 12, which places the inflow of certain D subhaplogroups as far back as 28,000 years ago.

Coming soon: thoughts on Neolithic antecedents in Upper Tibet, Ladakh’s prehistory and other bits and bobs

Previous Post

Next Post