The domestication of the horse: four versions of the story.

Why there is no single answer

The question has two parts that are easy to run together: when people first kept and managed horses, and when they first rode them. The evidence for both is indirect. Researchers have looked at wear on ancient teeth that might come from a bit, fat residues in pottery that might come from mare's milk, the shape of horse bones, the skeletons of the people who may have ridden, and, more recently, ancient horse DNA. Each kind of evidence has been read in more than one way, and the genetic studies of the last decade have changed what the question even means.

The four versions below are the main positions in our research. Each is given with its authors, its journal or publisher, and its date.

Version A: Botai, about 3500 BCE (2009)

On 5 March 2009 a team led by Alan Outram of the University of Exeter, working with Richard Evershed of the University of Bristol, published a paper in Science titled "Earliest Horse Harnessing and Milking". The University of Bristol's release of that day says the team argued that horses kept by the Botai Culture of northern Kazakhstan, about 5,500 years ago, were domesticated.

The paper rested on three lines of evidence, according to the release:

• damage on horse teeth consistent with a bridle or harness;
• traces of fat from mare's milk in Botai pottery;
• bones closer in shape and size to Bronze Age domestic horses than to the wild horses of the area.

The release said this would put domestication about 1,000 years earlier than had been thought, and about 2,000 years before domestic horses are known in Europe. It was careful on two points: it said riding was only "possibly" indicated, and it called the finding the earliest known evidence, not proof of the very first domestication.

Version B: Botai horses led to Przewalski's horses (2018)

Genetics changed the argument in 2018. Summaries of a Science paper by Gaunitz and colleagues say the team sequenced 42 ancient horse genomes, 20 of them from Botai, and concluded that Przewalski's horses descend from the horses herded at Botai and are feral rather than wild. The same summaries say horses from about 4,000 years ago onward carry only about 2.7 percent Botai-related ancestry, which the authors read as a large turnover in the horse population.

We saw this study only through summaries of its abstract, so these details are unconfirmed. A short 2018 commentary held in the Smithsonian's repository, also seen only as a summary, argued that the Botai horses may only have been tamed and that the evidence for Przewalski's horses being feral was not strong.

Version C: Botai was hunting, not herding (2021)

In Scientific Reports on 2 April 2021, William Taylor and Christina Barron-Ortiz challenged the tooth evidence. They compared Botai teeth with those of wild North American horses from the Pleistocene, which no one ever bridled, and argued that the supposed bit wear is better explained by natural variation in how teeth grow and wear. On their reading, Botai looks like regular mass hunting of wild Przewalski's horses.

The original team answered. In July 2021 Outram, Bendrey, Evershed, Orlando and Zaibert posted a rebuttal, held on Zenodo and in the University of Exeter's repository, saying the critique misrepresented their methods, misread the isotope evidence and left out newer protein evidence. In a paper in Frontiers in Environmental Archaeology on 24 April 2023, Outram restated that Botai, about 3500 to 3000 BCE, shows early horse husbandry even though its horses are not the ancestors of modern domestic horses. He described domestication as a process with several centres and several stages.

Version D: the western steppe, from about 2200 BCE (2021 and 2024)

A Nature paper published on 20 October 2021, led by Ludovic Orlando of the French research body CNRS with 162 co-authors, analysed 273 ancient horse genomes. The Czech Academy of Sciences, announcing it the same day, said it placed the origin of modern domestic horses in the Pontic-Caspian steppe of the northern Caucasus. Between about 2200 and 2000 BCE, the release says, one genetic profile spread out from that region and replaced wild horse populations from the Atlantic to Mongolia within a few centuries. It also says Anatolia, Siberia and Iberia, all proposed earlier as centres of domestication, were not the origin. Summaries of the study name the lower Volga-Don region as the source area, but we did not read the page that says so.

A follow-up paper in Nature, "Widespread horse-based mobility arose around 2,200 BCE in Eurasia", published online on 6 June 2024 by Librado and colleagues, used 475 ancient and 77 modern genomes. Sci.News reported on 7 June 2024 that it found a domestication bottleneck, with horses breeding at younger ages, beginning about 2700 BCE, then a sharp change in breeding practice about 2200 BCE. The paper proposes two episodes of domestication: the first about 5,500 years ago, the Botai lineage, and the second about 4,200 years ago, which produced the modern domestic horse.

The Yamnaya question: who were the first riders?

A separate argument concerns the Yamnaya, steppe herders of about 3300 to 2600 BCE. On 3 March 2023 Martin Trautmann, Volker Heyd and colleagues published a study in Science Advances that looked for six skeletal markers they call "horsemanship syndrome" in human remains from Romania, Bulgaria and Hungary. Science News reported that they judged five Yamnaya-culture men from about 3000 to 2500 BCE to have been frequent riders, with nine more probable riders. Ursula Brosseder, who was not involved, cautioned against reading this as proof that riding was fully developed in Yamnaya culture.

The sources part ways on what this means:

• Hartwick College (6 March 2023) says co-author David Anthony's work found more than 30 percent of Yamnaya people sampled rode habitually, and describes the 2021 Nature study as showing that Yamnaya-culture horses were the direct ancestors of modern domestic horses.
• The 2024 Librado paper, as Sci.News reports it, says its timeline does not fit Yamnaya migrations being accompanied by large herds of modern domestic horses.
• The study's own limits. The 2023 work rests on stress marks on human bones, not direct evidence of riding, and summaries of Archaeology magazine's coverage report that Trautmann said these may not be the earliest riders. We saw that coverage only as a summary.

Evidence that has been withdrawn

A stallion from Dereivka in Ukraine, from a site of about 4500 to 3500 BCE, was once cited as evidence of riding around 4000 BCE because its teeth seemed to show bit wear. Outram's 2023 paper notes that radiocarbon dating later showed it was an Iron Age animal buried into the older layers, so it no longer supports early domestication. Hartwick College records that David Anthony and Dorcas Brown's research on bit wear began with a visit to the Institute of Zoology in Kyiv in 1989 to study ancient horse teeth.

Chariots

Outram's 2023 paper gives the earliest physical evidence of chariots as burials of the Sintashta culture in the late third millennium BCE, and the earliest pictures of chariots as Assyrian, about 2050 BCE. Whether riding or chariot driving spread first is, it says, still debated.

How riders came to sit on saddles and stand in stirrups is the subject of the stirrup and early tack. The dates on this page also appear in the timeline.