Marsha a levine

---

See discussions, stats, and author profiles for this publication at: 
https://www.researchgate.net/publication/269403546
Domestication of the Horse
Chapter
· January 2012
CITATIONS
0
READS
964
1 author:
Marsha Levine
University of Cambridge
25
PUBLICATIONS
1,050
CITATIONS
SEE PROFILE
All content following this page was uploaded by 
Marsha Levine
 on 11 December 2014.
The user has requested enhancement of the downloaded file.

---

Marsha A Levine 
Reference
: 
2012 Levine, M.A. Domestication of the Horse, in Neil Asher Silberman (ed.), The Oxford Companion to Archaeology, 
Second Edition. Oxford University Press, USA, 978-0-19-973578-5, pp 15-19. 
Horse, Domestication of the
The impact of the earliest domestication of the horse on human society must have been as profound as that 
of the invention of the steam engine, and yet we know very little about when, where, or how it happened. 
The increased mobility provided by the horse would have enabled people to move further, as well as faster, 
and to take more with them than ever before. They could exploit larger and more diverse landscapes, 
maintain larger families, and increase the range of their trade contacts. And since a man on foot is no 
match for a man on horseback, the military implications of horse domestication would have been 
revolutionary. Ewers has shown how profoundly the introduction of the horse into North America changed 
Blackfoot culture. We should expect no less of its early domestication in central Eurasia. In recent years 
considerable attention has been paid to this problem, but the results have not taken us very far in solving 
it. This short discussion can do no more than outline some recent and current research approaches and 
results. 
It is important to note that wild horses, particularly as foals, can be captured and tamed and, as such, 
ridden, slaughtered, and eaten without being domesticated. Aboriginal peoples throughout the world are 
known to tame all kinds of wild animals to keep as pets. There is no reason to think that this would not 
have been the case at least from the time of the earliest anatomically modern 
Homo sapiens. And when the 
need arose, taming would probably have been the first step towards domestication. When and where the 
next step was taken is a matter of heated debate. 
Though customarily defined as the controlled breeding of plants or animals by humans, the real 
distinctiveness of domestication lies in the fact that it involves ownership and thus results in a completely 
different level of human commitment than does taming or hunting. The social and economic implications of 
horse taming would have been, at most, superficial and localized, and would have disappeared with the 
death of the animals involved. The repercussions of domestication would have been culture-wide. What this 
means is that we are not simply trying to identify horse riding, traction, milking, and meat eating in the 
archaeological record, but rather, we are looking for evidence of horse breeding, which is archaeologically 
invisible. It may be approached indirectly, however, through an investigation of population structure, 
archaeological context, and other characteristics of the data. 
Before Domestication 
We have no evidence that horses had been either tamed or domesticated during the Palaeolithic and there 
is no longer any significant support for the idea that horse domestication took place at such an early date. 
On the other hand, there is strong evidence, based upon kill-off profiles from various sites—including 
Solutré, Feldkirchen-Gönnersdorf, Combe Grenal and Pech de l'Azé—that throughout the Middle and Upper 
Palaeolithic in western Europe horses were, in fact, hunted.
Horses are relatively uncommon in western European Mesolithic and Neolithic archaeological deposits. The 
1

---

widely held view that they became extinct in Europe at the end of the Pleistocene, however, does not stand 
up to close scrutiny. Horses were present throughout Europe, though perhaps in reduced numbers, or 
alternatively, were hunted less frequently and/or with less success. In any case, there is, at present, no 
credible evidence for horse domestication during these periods - 
circa 10,000 BC to 4,500 BC - give or 
take a couple of thousand years.
The Evidence for Early Horse Husbandry 
Relatively large quantities of horse bones and teeth have been recovered from Eneolithic (
c 4
th
millennium 
BC) sites on the central Eurasian steppe. This is where discussions become heated. Although other data, 
such as tooth wear and morphology, representation of anatomical elements, and taxonomic distinctions 
based upon measurements, have been credited as evidence for horse domestication; researchers seeking 
the origins of horse domestication invariably target sites with large quantities – both in absolute numbers 
and by comparison with other taxa – of horse bones. In other words, a strong bias was in built into the 
research from its inception. 
Until around 2000, Dereivka, an Eneolithic habitation site, was central to the problem of horse 
domestication. It was situated on a tributary of the Dnepr River in Ukraine and dated 3380-4570 BC (Sredni 
Stog, Phase IIa). Excavated by D Y Telegin (Institute of Archaeology, Kiev), 1960-1983, Dereivka had been 
regarded as the site with the earliest evidence for horse domestication. It was argued that horses had been 
raised there first of all for meat, but also for riding. The methodological framework used to reach these 
conclusions was conventional but unsound. Criteria used as evidence that the horses from Dereivka were 
domesticated included the following: (1) the absence of old horses, (2) the presence of a high proportion of 
male skulls, (3) the presence of objects identified as bridle cheek pieces, (4) the results of a morphological 
analyses comparing the Dereivka horses with other equid material, (5) the association of horse with other 
domesticates—cattle, sheep, goat, pig and dog—and (6) the relatively high percentage of horse bones and 
teeth in the deposit. 
In reality, on the basis of archaeological, ethnographic and ethological comparisons, the absence of old 
individuals is much more likely to indicate hunting than herding. Adult males would outnumber females 
either if bachelor groups or stallions protecting their harems were targeted in the hunt. The cheek pieces 
might not have been cheek pieces at all. The morphological study involved very small and disparate 
samples and produced contradictory results. The association of horses with other assumed domesticates is 
not evidence of horse domestication. In any case, they were also associated with the remains of wild 
animals. In fact, the most important criterion was the relatively high proportion of horse bones and teeth 
present at the site. However, this could have resulted from an increase in horse hunting by comparison with 
earlier sites. Levine's reassessment of the population structure data indicates that the vast majority of the 
Dereivka horses had probably been killed in the hunt. This leaves the possibility open that at least some of 
the horses could have been ridden and domesticated. However, no credible evidence has been presented to 
prove this.
Just to be clear, we have good evidence – even for a born sceptic – that by around 2000 BC horses from the 
Sintashta-Petrovka culture (South Urals) were well and truly domesticated and had been so for a 
considerable period of time. However, the problem of when and where horse domestication first took place 
has proven to be very intractable. Domestication is a process that involves an important change in human-
2

---

horse relationships. Mortality profiles or butchery evidence, for example, may strongly suggest that the 
majority of the horses at a site were wild, but it does not prove that none were domestic. At a site where 
most remains were from wild horses, evidence for a few domestic individuals might be difficult to extract. 
The challenge of creating new methodologies to do this is of critical importance. 
Some of the new developments in archaeozoological research – such as, ancient DNA, stable isotopes, 
palaeopathology, environmental analyses – are very promising, as long as they are applied critically. Ideally, 
what we seek, is evidence that would be explicable only in terms of domestication. Such was Anthony and 
Brown’s objective in the 1980s when they set out to prove that the bevel they observed on the 2
nd
lower 
premolar of a horse skull from Dereivka was caused by the horse chewing its bit. This they took as evidence 
for horse riding and domestication. Although the results of their experimental research on bit wear were 
inconclusive, the underlying concept is not unreasonable with several caveats: 1) not all bitted horses chew 
their bits; 2) horses can be ridden without bits; 3) bevelling on the lower 2
nd
premolar can be caused by an 
abnormal bite. In any case, when the Dereivka skull - up to this point, a key piece of evidence in the 
domestication debate- turned out, on the basis of new radiocarbon dates, to be intrusive, attention shifted 
to Botai.
Botai 
Botai caught the attention of archaeologists and archaeozoologists studying the origins of horse 
domestication because of the stupendous number of bones recovered from the excavations. It has been 
estimated that, during the 1980s, 10 tons of bones had been excavated, over 99 percent of which were 
horse. The rest were, mostly or entirely, from wild species. Archaeologists working on this material 
estimated that bones from 70,000 individuals were recorded from 2.5 percent of the Botai site. 
Botai is an Eneolithic settlement site, dated to 
c 3500 cal BC, and located in the forest-steppe region of 
northern Kazakhstan (Kokchetav Oblast). The site covers approximately 15 hectares on the bank of the 
Iman-Burluk, a tributary of the river Ishim. It comprises around 300 semi-subterranean, polygonal 
‘dwellings’, packed together in a kind of honeycomb pattern. As of 1995, about 10,000 m
2
had been 
excavated, including 217 ‘dwellings’. Since then further excavations have taken place. 
Since work began at Botai, researchers have variously determined that the horse bones were from either 
wild or domestic animals or a mixture of both. It is not possible here to discuss all the arguments for these 
determinations. However, Outram et al ( 2009) is illustrative of the kind of pitfalls awaiting researchers who 
use data to prove a theory rather than to test a hypothesis. In the very first paragraph the co-authors 
identify the Eurasian steppe, and specifically Botai, as the “prime candidate” for the locality “where horse 
domestication first took place”, noting that “the faunal assemblages consist almost entirely of horse 
remains”. In other words, this site was targeted because of the enormous quantities of horse bones. They 
state that “sedentary settlement structure” is incompatible with “hunting mobile wild herds”. In truth, the 
size of the herds at Botai would have been so large, and the grazing so limited, that they could only have 
been kept near the settlement for a very short time. The ethnographic record provides plenty of evidence 
that seasonally occupied hunting sites in the northern latitudes often have ‘permanent’ structures, 
especially during winter; and that both hunters and herders are often semi-nomadic or semi-sedentary, 
moving, for example, between customary winter and summer occupation sites.
3

---

The taphonomic analyses of skeletal element representation and butchery patterns, suggest that: 1) the 
carcases were not moved far from the kill site, and 2) the butchery methods used were very wasteful and 
thus the opposite of those usually employed by herders, such as Kazakhs and Mongols. The most 
straightforward explanation for this would be that Botai had been located on or near a horse migration 
route and the horses had been killed in herd drives. Herd driving is a common hunting technique and it is 
also the best explanation for the observed age and sex structure, based upon analyses of the teeth and 
pelves from Botai. As Outram et al admit, the age structure manifested at Botai is not really compatible with 
herding.
Outram et al also claim, without providing any evidence, that the presence of what were possibly leather 
working tools could be used as proof for horse domestication. However, leather is used for all kinds of 
things, not just horse tack; and the tools found at Botai indicate that the inhabitants of the site carried out 
a wide range of activities, including hunting and fishing.
Outram et al focuses on 3 “three independent lines of evidence demonstrating domestication” (p. 1332) at 
Botai. However, each of these three lines of evidence could have alternative explanations.
1) The co-authors claim that the relative gracility of Botai metacarpals vis-à-vis several other assemblages 
is evidence that they were from domesticated horses. However, differences in metacarpal robusticity 
between assemblages could have many different explanations: for example, diet, climate, genetic drift or 
isolation. This paper does not consider alternative explanations. The number of sites studied was small and 
no independent evidence is provided to justify the claim that any of the bones from any of the assemblages 
were from domestic horses.
2) The co-authors claim to have identified bit-wear on 5 out of 15 lower P2s from Botai, based upon a new 
analytical method. However, this method is deeply flawed. The proxy group for the earliest domestic horses 
comprised 32 work animals (including Thoroughbreds, Shires, mules and hinnies), 24 of which were over 
15 years old; and the non-work group comprised 28 unworked Przewalski’s horses (no provenance given), 
11 of which were over 15 years old. Only 9 of the work animals (3 hinnies and 6 horses) had the ‘abrasion’, 
described as evidence of bit-wear. All these animals were more than 11 years old and 7 were over 20 years 
of age. Essentially what is the authors are saying is that a miscellany of equids – of different shapes, sizes, 
ages and even species - can be used as proxies for the earliest domesticated horses, while Przewalski’s 
horses – of various ages - can be used as proxies for ancient wild horses. What they fail to take into 
account is that in order for this kind of analysis to mean anything, the proxies must be as close as possible 
in every way to their targets. The Botai horses had similar conformations to ponies – not to Thoroughbreds, 
Shires or mules. Their diets, life expectancies and lifestyles would not bear the slightest resemblance to 
those of the Botai horses. Only about 1% of the Botai teeth were from horses over 15 years of age; while at 
least 75% of the proxy work animals were that old. Moreover, the riding horse proxies, such as 
Thoroughbreds, would have, most probably, worn stainless steel bits, had their teeth rasped regularly, and 
consumed high energy foods from raised feeders in stalls. These horses bare so little resemblance to the 
ancient steppe horses of Botai, wild or domestic, as to make comparisons meaningless. Moreover, no 
evidence has been presented that the proxy work equids ever chewed their bits. Or that chewing the bit 
could cause the kinds of dental and bony changes they attribute to bit-wear.
4

---

3) The evidence for milking, based upon the biochemistry of fatty acid residues found in potsherds from 
Botai, is the most interesting part of Outram et al. The paper claims that the hydrogen isotope, deuterium, 
allows milk (produced in summer) to be distinguished from other horse fats (produced throughout the 
year), because rain falling during summer has a higher deuterium signal than that falling in winter.
However, the isotope values of fatty acids extracted from archaeological potsherds seem sufficiently 
different from the modern reference horse fats to warrant further scrutiny. The co-authors suggest that 
this variability could result from climate change. Alternatively, the elevated deuterium values might not be 
related to milking at all. For example, they might instead reflect variability in environmental conditions 
experienced by different horses. For example, variability in the deuterium signal might reflect variability – 
geographical or temporal - in the home ranges of the migrating horses which comprised the bone 
aggregations at Botai. 
Conclusion 
Botai might well fit within the time frame expected for the origins of horse domestication. However, none of 
the evidence presented so far supports the hypothesis that domestic horses were present at Botai, while the 
evidence of population structure, butchery practices, palaeopathology incidence, as well as contextual 
evidence, best supports the hypothesis that the Botai horses were mostly or wholly wild and killed in herd 
drives. 
Focusing the search for the origins of horse domestication solely on sites with high frequencies of horse 
remains is not the best way to approach this problem. A much wider net needs to be cast, including a 
broader geographic and temporal range, the careful development of new methods and rigorous testing of 
new hypotheses. Researchers whose explicit objective it is to identify the earliest domestic horses are 
almost certainly doomed to failure.
References 
Anthony, D. W. & D. R. Brown, (2003). Eneolithic horse rituals and riding in the steppes: new evidence, in 
Prehistoric Steppe Adaptation and the Horse, eds. M. Levine, C. Renfrew & K. Boyle Cambridge: McDonald 
Institute for Archaeological Research, 55-68. 
Ewers, J. C., 1955. 
The Horse in Blackfoot Culture, Washington, D.C.: Smithsonian Institution Press. 
A. M. Kislenko & N. S. Tatarintsev, (1999). The eastern Ural steppe at the end of the Stone Age, in 
Late 
Prehistoric Exploitation of the Eurasian Steppe, eds. M. A. Levine, Y. Y. Rassamakin, A. M. Kislenko & N. S. 
Tatarintsev Cambridge: McDonald Institute, 183-216. 
P. L. Kohl, 
The making of Bronze Age Eurasia, Cambridge: Cambridge University Press (2007). 
L. Koryakova, & A. Epimakhov, 2007. 
The Urals and Western Siberia in the Bronze and Iron Ages, 
Cambridge: Cambridge University Press. 
5

---

6
M. A. Levine, 
The origins of horse husbandry on the Eurasian Steppe. Late Prehistoric Exploitation of the 
Eurasian Steppe. M. A. Levine, Y. Y. Rassamakin, A. M. Kislenko and N. S. Tatarintseva (1999). Cambridge, 
McDonald Institute: 5-58. 
A. K. Outram, N. A. Stear, R. Bendrey, S. Olsen, A. Kasparov, V. Zaibert, N. Thorpe & R. P. Evershed, 2009. 
The Earliest Horse Harnessing and Milking. 
Science, 323(5919), 1332-5. 
D. Y. Telegin, 1986, 
Dereivka, a settlement and cemetery of Copper Age horse keepers on the Middle 
Dnieper, Oxford: BAR. 
View publication stats
View publication stats