Is the yak a domesticated animal? The answer is cultural and political.

This is a lightly-modified excerpt from my newly published article, “Hacking the Yak: The Chinese Effort to Improve a Tibetan Animal in the Early Twentieth Century” (in East Asian Science, Technology, and Medicine no. 48). For the published article, go here; it will be open access starting in December 2019. For the author’s preprint version of the full article, click here.


If you know one word of Tibetan, it is probably yak—though as speakers of Tibetan are quick to point out, that word refers only to the male in their language, while the female is called a dri. (I will adhere to English usage here, using it to refer to both male and female.) The species has traditionally been associated with Tibet, and with Tibetan nomads in particular. But observers have produced contradictory accounts of the historical relationship between Tibetans and the yak. To some, Tibetan mobile yak pastoralism is a primitive livelihood marked by extreme conformity with nature, while others have described it as an elaborate technical exercise.

Published sources from the early twentieth century suggest a correlation between how non-Tibetans viewed Tibetans and how they viewed yaks: those who viewed Tibetan culture as primitive tended to also view the yak as an under-domesticated animal.

So is the yak a domesticated animal?

The question of whether (and to what extent) Tibetans domesticated the yak hinges on the definition of “domesticated”.

Recent debates highlight the tension between two competing paradigms of domestication. Natural scientists have conventionally invoked a “biological definition,” which posits that an animal is domesticated when its morphology and behavior meet a list of criteria such that they exhibit what some have called “domestication syndrome”. Docility (or “tameness”) is the hallmark of domestication syndrome, typically accompanied by the retention of juvenile features, including reduced brain size.[1] This biological paradigm is predicated on the expectation that domestication involves extreme human domination of domesticates, including near-total control over domesticates’ reproduction resulting in their genetic isolation and distinctness from wild ancestors.[2]

Some anthropologists point out that such a rigid understanding of domestication rules out cases in which animals have been thoroughly integrated into human society without such an extreme degree of domination. For example, Fiona Marshall and Lior Weissbrod observed that a single Maasai community paid great attention to horse pedigree while allowing its donkeys to procreate freely, which the researchers attribute to the fact that the community under study valued donkeys for their load-bearing strength and had little regard for their pedigree or sociability, in contrast with its attitudes towards horses. Yet commonsensically, both animals are “domesticated”.[3] Similarly, geneticist Ludovic Orlando notes that Mongolians often allow their horses to roam freely and “capture” them only as needed, because this is more practical than continuous domination.[4] Many social and natural scientists now advocate weighing the biological definition of domestication against a “cultural definition” that takes pragmatics into account.[5]

Early European encounters with yaks produced some confusion over their status with regard to domestication. A British travel writer observed in 1874 that “even the yaks of burden, which have been domesticated, or rather half domesticated, for generations, are exceedingly wild, and the only way they can be managed is by a rope attached by a ring through the nose.”[6] Only after 1875 did taxonomists register a distinction between the “domesticated yak” (Bos grunniens), and the “wild yak” (Bos mutus), corresponding to the Tibetan distinction between yak and drong.[7] Even contemporary observers note that the domesticated yak is not particularly docile, nor is it as genetically divergent from its wild ancestor as the typical farm animal.[8] By biological criteria that emphasize domination, yak domestication appears incomplete, especially when the transformation from mutus to grunniens is juxtaposed with the coeval transformation from the auroch to the ox—a common frame of reference for lowland peoples. Indeed, one historical paradigm situated the yak on a trajectory parallel to that of the ox or the pig, but with some catching up to do.

Figure 1

An alternate interpretation emphasizes human adaptation over domination.

Zooarchaeologist Juliet Clutton-Brock notes that the proximity of grunniens to its wild ancestor preserves its highland niche, so that “as a pack animal and provider of milk, the yak enabled humans to colonize the mountainous regions of Tibet and Nepal.”[9] We might speculate that, as with the Maasai donkey, the load-bearing role of the yak has disincentivized selection for docility among herders, given that such selection tends to produce smaller animals.[10] In this appraisal, the yak’s influence on Tibetan society is not incompatible with human ingenuity, in that adaptation to the yak is understood to be a highly technical accomplishment. Indeed, those who have closely investigated Tibetan yak production, including ethnologists, ethnographers and scientists, tend to emphasize its technical complexity. I synthesize some of their findings here, not to provide a comprehensive overview of the yak’s role in Tibetan society, but merely to establish three points that are of later relevance to this essay: (1) that yak domestication is ancient in origin, (2) that yaks were and are the subject of an elaborate system of indigenous knowledge, and (3) that yak production was not the domain of a monolithic “nomadic” lifestyle.

Jianlin Han of the Chinese Academy of Agrarian Sciences, summarizing the findings of an international constellation of scientists, notes that initial domestication of the yak must have occurred approximately ten thousand years ago, and that DNA evidence points to a single domestication event on the Qinghai-Tibetan Plateau (meaning that Mongols and other yak herders beyond the plateau were eventual beneficiaries of this feat).[11] Domestication of the wild yak (Bos mutus) is usually attributed to the Qiang 羌, an ancient civilization thought to be ancestral to Tibetans (as well as modern Qiang) and which the Food and Agriculture Organization of the United Nations describes as “quite possibly the earliest animal husbandry culture of excellence in the world.”[12] DNA and archaeological evidence points to a major expansion in yak production about five thousand years ago. The breeding of female cattle (Bos taurus) with male yaks (Bos grunniens) to produce hybrids has been traced to 1100 BCE, though according to Han, it intensified in the middle of the twentieth century—precisely the period with which this study is concerned.[13] Yak pastoralism is thought to have originated on the southeast of the plateau, corresponding to the western portion of modern Sichuan Province (and the overlapping Tibetan region of Kham); today western Sichuan is also host to the world’s largest yak population (4,084 as of the year 2000, not including hybrids).[14] In other words, yak husbandry has benefited from millennia of experience plus many hundreds of generations of passive and active selection, even if these techniques did not conform to the scientific method per se.

Tibetan mobile pastoralists often identify (and are identified by their Tibetan neighbors) as drokpa (Wiley: ‘brogs pa), a term that is typically translated as “nomad.”[15] But those who picture the Tibetan nomad as moving erratically or, as Chinese observers often put it, “chasing the water and the grass” (zhu shui cao er ju 逐水草而居) are mistaken; unlike peripatetic or “wandering” nomads (such as Gypsies and Thuggees), the drokpa are transhumant nomads, meaning that they move methodically between a certain number of pre-selected locations according to the season. The most basic principles of yak transhumance in high-altitude settings are that animals must move higher as the temperature rises and lower as it drops, and that animals must move shorter distances when they are most vulnerable to effects of the weather (the onset of spring); the drokpa clan will have designated pastures whose spatial distribution adheres to these principles. From a sedentist perspective, such mobility is unusual, and drokpa may appear to be unduly influenced by their animals. Recent scholarship, however, has inverted this narrative by arguing that settled agriculture reflects a high degree of plant influence on humans in exchange for dubious benefits to individual people.[16]

In fact, sedentary agriculture appears to have facilitated high-altitude nomadism such that it arguably represents an evolution of sedentary agriculture, or an escape from it. The farmer and nomad are linked by a continuum of production methods through which individuals may transition from the more temperate conditions of the lowlands to the harsher conditions of the highlands. Robert Ekvall, who conducted extensive field research in Amdo (northeast Tibet) between 1926 and 1941, noted the existence of an intermediary social category called the sa ma drok (Wiley: sa ma ‘brog meaning literally, “neither soil nor high pasture”) that practiced a semi-nomadic lifestyle.[17] Ekvall was able to observe the evolution of households from one mode of production to the other–not from nomadism to farming, as sedentary outsiders have often assumed, but from farming to semi-nomadism and then (sometimes) to full nomadic pastoralism.[18]

The multi-step process Ekvall describes implies the transmission of detailed technical knowledge across generations. Farm households typically kept a small number of cattle, but once the herd of a farming household reached a certain size, the farmers might establish a separate summer pasture within a day’s reach, both to give the herd ample grazing room and to protect the year’s crop from those same animals; the herd returned to the farm after the harvest. Some households retained their farms while others abandoned them entirely, rotating between pastures with the seasons. While Ekvall notes that sa ma drok households never achieved an identity as drokpa regardless of their livelihood (and perhaps predominated at a lower altitude than the latter), the sa ma drok lifestyle provided excellent technical training in nomadic pastoralism and many individuals were incorporated into drokpa lifestyles through marriage or other means. Further, these seminomadic households occupied an important niche in the pastoral economy as an intermediary between farmers and drokpa; for one thing, with access to both cattle and yaks they were ideally positioned to breed the hybrids known in Tibetan as dzo upon which many drokpa relied for dairy and other purposes.[19]

Moving from lowland farm to high pasture was not simply a matter of preference in that it required a high degree of technical knowledge. Anthropologist Rinzen Thargyal provides unique insight into this body of indigenous knowledge through his interviews with nomadic households that were exiled from Kham’s Zilphukhog Valley by the Communists in 1957. While in Kham, his informants had practiced strategic gelding, retaining some virile bulls (or chu-yak) for breeding but castrating other bulls for use as docile pack yaks. Keeping track of livestock fertility was key to successful husbandry, and Thargyal’s informants had an elaborate scheme for identifying the age of yaks, including unfamiliar ones, first by counting their teeth and later, after the age of seven years, by counting the lines (or trü) on their horns.[20] The naming scheme that Thargyal records for yaks at each age is corroborated by Namkhai Norbu, who conducted his field research in Dzachuka in 1951 and who reports that the average yak there reached 10 trü, corresponding to 17 years.[21] Drokpa took care not to milk yak cows (dri) too early since this could starve their calves and put an early end to milk production (“to starve dri calves is a self-destructive act for a pastoral nomad” notes Thargyal) but were more likely to eventually starve the calves of yak-ox hybrids (dzo), which are known to be less valuable for most purposes than their parents.[22] We should recognize that in addition to these token details, Tibetan nomads have developed a vast amount of tacit knowledge that would be difficult or impossible to transcribe, and that would be lost if their way of life were to disappear.

Figure 3.JPG


Popular discourse outside of Tibet has tended to ignore the technical complexity of yak transhumance and has instead constructed it as primitive and simplistic.

Observers often painted a caricature of Tibetan society as neatly divided between nomads and sedentary Tibetans; for example, a journalist for the North-China Herald and Supreme Court & Consular Gazette pontificated in 1914 that “the population of Tibet falls naturally into two divisions, namely, those who live in tents and those who live in temples. The people may roughly be divided into priests and nomads.”[23] “Nomads” figured in Chinese-language discourse as youmu min 游牧民 or “ranging herders”. The notion of the ethnos or minzu 民族 remained ill-defined during this period (in comparison to its later Communist usage), and nomads throughout Inner Asia were sometimes even referred to as youmu minzu 游牧民族, a sort of para-ethnic category.

The term niuchangwa 牛場娃was further impemented as a rough loan-translation of drokpa, niuchang being the Chinese term for yak pastures. The Chinese journal New Asia (Xin Yaxiya 新亞細亞) remarked that “the ethnos (minzu 民族) of Tibet can be approximately divided into two groups. The first are the house-dwellers (fangshewa房舍娃), and the second are the niuchangwa.” The author claimed that Tibetan house-dwellers were “almost the same as us Han,” while the niuchangwa was “entirely unlike us Han”[24]

Such writers were wont to focus on “pure” nomads, ignoring the complex gradations between farmer and nomad as observed by ethnologists, or relegating them to impurities within an otherwise binary economy. Concern with the “pure” nomadic Tibetan took on eugenic dimensions, and paralleled the rise of the “pure Han” as a concept in Chinese and international discourse.[25] Ren Naiqiang, the preeminent Han scholar of Kham, contended that the Han lived in the lowlands of Kham while the “pure Tibetans” lived on the high plains herding livestock, and their mixed-blood offspring inhabited the spaces in between.[26] The very concept of the “nomad” seems to encourage this kind of binary, since it is a simplification that obscures the complex range of lifestyles is attempts to describe.[27] In the imagination of sedentary peoples, the “nomad” is essentially primitive, violent and a relic of the premodern past—prejudices that infiltrate even contemporary historical scholarship.[28]

International appraisals of nomads as barbarians living in something like a state of nature made it easy to interpret their livestock as undomesticated or under-domesticated. In 1914 the North China Herald imagined Tibetan nomads as passive beneficiaries of the easy predisposition of their animals, since “the yak gives the nomad little or no trouble and though undomesticated [my emphasis], is a quiet and timid animal.”[29] Positivist theories of social evolution generally posited that nomads were at a more primitive stage of development than sedentary peoples, and the ingenuity involved in domestication must have seemed beyond the nomad, prompting the (mis)conception that Bos grunniens was docile yet unaltered by its human hosts.

It is in this milieu that the monolithic “nomad” became the target of Chinese nationalist discourse on the yak.

The entire species was characterized its participation in a backwards mode of development and thus its potential for improvement by the Han was self-evident; the key was extraction from nomadism. For example, in a 1943 essay titled “Nomadic and Sedentary Pastoralism,” one Dong Mi凍沁traces the divide between sedentism and nomadism to the ancient sage king Fuxi 伏羲, whose guidelines for the sustainable raising of livestock were articulated in the Rites of Zhou (Zhou li周禮). The Han were those who had received the transformative influence (hua 化) of Fuxi, while the nomads on the western frontier were those who had not. Dong faults the decline of the Qing dynasty over the previous two hundred years with facilitating nomadism, because “the people have not been opened to knowledge.”[30]

Dong’s nomads, which seem to be based mainly on secondhand reports, are utterly devoid of science. He writes that “many of the residents of the northwest have been stuck in a rut, following the water and the grass, and I hear that the livestock must live and perish of its own accord for they don’t know how to improve it.” He then cites Chiang Kai-shek as saying that “Argentina in South America is the world’s greatest producer of meat, and yet the pastures of our own northwest could displace Argentina if they were opened up and properly developed.”[31]

Dong’s polemic reflects a synthesis of Han chauvinism with the global paradigm of “improvement.” An early watchword of scientific agriculture, “improvement” typically referred to increasing yields per unit of livestock or unit of land, and was inspired by the marketization of agricultural products in Europe beginning in the 18th century. Its infiltration of Chinese discourse as gailiang 改良 or gaijin改近gave Chinese and international agrarians a common idiom for discussing Tibet. In an article on the founding of Xikang Province (Xikang sheng西康省), The Christian Science Monitor had this to say:

Tibet’s nomads have been ignorant of how to care for the Yak. In Minya, west of Tasienlu [Kangding], many cows give only a quart or so a day and the price of highly-watered milk is high. To the nomad “a yak is a yak, and who ever heard of improving a yak? And why improve the yak?” As the nomad sees it, an improved yak may give more milk, and more milk may give more cheese, which will produce more money. But more money will mean more expense… so why improve the yak?[32]

 “Value” (jiazhi價值) became the arena in which industrialized, scientific production sought to compete with nomadic production. “The value of the yak is indeed great,” wrote a contributor to Xikang Economy Quarterly in 1944, “but knowledge among the people is shallow and they do not recognize its value.”[33] Nomadic pastoralism was doubly troublesome to sedentary observers because of its lower economic output and its problematic mode of social organization. [End of excerpt.]

For the full article, visit the links at the top. The references below refer to entries in the bibliography of the full article.

[1] Clutton-Brock (2012), p. 8; Ritvo (2004), p. 214.

[2] Ingold (2000), p. 75; Saey and Engelhaupt (2017), p. 22.

[3] Marshall and Weissbrod (2011), p. S405.

[4] Quoted in Saey and Engelhaupt (2017), p. 23.

[5] For example, see Alan Outram (2014), p. 759; Saey Engelhaupt (2017), p. 22-23. The implication of “cultural” here is not so much that we should defer to a given culture’s definition of domestication, but that we should frame our understanding of domestication in reference to a species’ practical role in a given cultural context.

[6] “The Yak of Thibet” (1874).

[7] Wiener, Han, and Long (2003), p. 9.

[8] Clutton-Brock (2012), p. 85.

[9] Clutton-Brock, Juliet (2012), p. 85.

[10] Tibetans do produce more docile bovids by breeding yaks with oxen, a process I shall examine shortly.

[11] Han (2013).

[12] Wiener, Han, and Long (2003), p. 3.

[13] Wiener, Han, and Long (2003), p. 3.

[14] Wiener, Han, and Long (2003), p. 7.

[15] Robert Ekvall more accurately translates drokpa as “high pasturage ones” See Ekvall (1968), p. 3.

[16] For example, see Diamond (1987); Harari (2015), pp. 87-109; Scott (2017).

[17] Anthropologist Rinzen Thargyal has also described the sa ma drok, but Ekvall was the first scholar to describe their transition from sedentary to nomadic pastoral production in detail.

[18] Ekvall (1968), pp. 21-23.

[19] Ekvall (1968), pp. 20-21.

[20] Thargyal (2007), pp. 76-78.

[21] Norbu (1997), pp. 39-40.

[22] Thargyal (2007), pp. 78.

[23] “Among Tibetan Nomads: A Picturesque People, the Ways of the Yak”.

[24] Wang Dufeng (1930), p. 23.

[25] Zhihong Chen Traces the notion of the “pure Han” to the writings of Ellsworth Huntington and Zhang Qiyun in the early twentieth century. See Chen (2012), pp. 85-88.

[26] Ren Naiqiang (1990), pp. 219-220.

[27] In their critique of the term ‘nomadism,’ Carolyn Humphries and David Sneath note that “pastoralism in the vast grassland region of Inner Asia is not timeless ‘nomadism’ but is a series of local knowledges and techniques located in particular historical circumstances.” See Humphrey and Sneath (1999), p. 1.

[28] For instance, historian Michael Khodarkovsky describes the nomadic groups along the steppes of the Russian empire as “societies organized for war” (2002, p. 17), while Jane Burbank and Frederick Cooper write of Eurasian nomads generally that “nomadic life meant that the whole society could be mobilized for war” and that “the point of war was plundering, sharing out the booty, and moving on to get more” (2010, p. 99).

[29] “Among Tibetan Nomads: A Picturesque People, the Ways of the Yak”.

[30] Dong Xin (1943), p. 34.

[31] Dong Xin (1943), p. 34

[32] “China Builds ‘Inner Empire’ With ‘Go-West’ Movement”.

[33] Xu Deming (1944), p. 111.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s