Indo-European migrations

• 1. Introduction

  ›

  Indo-European migrations are among the most consequential population movements in human history, reshaping language, culture, and genetics across a vast sweep of Eurasia from roughly 4000 to 1000 BCE. Today, nearly 3 billion people speak languages descended from a single prehistoric tongue called Proto-Indo-European. That is more native speakers than any other language family on earth. How did one language, spoken by herders on the grasslands north of the Black Sea, become the ancestor of English, Hindi, Persian, Greek, Russian, and hundreds of other tongues? The answer involves wagons, horses, climate collapse, war, trade, and an extraordinary series of migrations that no one wrote down because writing had not yet been invented. What drove the first speakers to leave their homeland? How do we know any of this at all, given that they left no texts? And what happened to the peoples they encountered along the way? The search for those answers has drawn linguists, archaeologists, geneticists, and anthropologists into one of the most contested debates in all of scholarship.

• 2. Reading Silence: How We Know

  ›

  Sir William Jones delivered the observation that changed everything in his Third Anniversary Discourse to the Asiatic Society in 1786. After learning Sanskrit in India, he detected systematic correspondences between Sanskrit, Ancient Greek, and Latin, concluding they all originated from a common source. Jones was not the first to notice something unusual. The Dutch scholar Marcus Zuerius van Boxhorn, who lived from 1612 to 1653, had already noted extensive similarities between various European languages, Sanskrit, and Persian. But Jones gave the idea a rigorous foundation, and from it grew the hypothesis of a family consisting of several hundred related languages and dialects. The 2009 Ethnologue estimates about 439 Indo-European languages and dialects in total, roughly half of them belonging to the Indo-Aryan sub-branch of South Asia. With written attestations appearing from the Bronze Age in the form of the Anatolian languages and Mycenaean Greek, the Indo-European family holds the distinction of possessing the second-longest recorded history of any language family, after the Afroasiatic family. Because Proto-Indo-European left no written evidence, all knowledge of it comes from reconstruction. Historical linguists use techniques such as the comparative method and the method of internal reconstruction. The largest methodological advances of the 20th century were the discovery of the Anatolian and Tocharian languages and the acceptance of the laryngeal theory, each of which forced scholars to revise their picture of what PIE looked like. Most techniques of historical linguistics were themselves developed during the 19th century, when the vast majority of linguistic work was devoted to reconstructing Proto-Indo-European or its daughter proto-languages such as Proto-Germanic. Scholars estimate that PIE may have been spoken as a single undivided language around 3500 BCE, though estimates vary by more than a millennium depending on the authority.

• 3. The Steppe Homeland

  ›

  Marija Gimbutas formulated her Kurgan hypothesis in the 1950s, grouping together a number of related cultures at the Pontic steppes and identifying them as the most likely speakers of Proto-Indo-European. The term "Kurgan" comes from a Turkic loanword in Russian for a tumulus or burial mound. Since the early 1980s, mainstream scholars have favored this hypothesis or the closely related revised Steppe hypothesis developed by David Anthony, placing the homeland in the Pontic steppe between the Dniepr in Ukraine and the Ural river in Russia during the 4th and 5th millennia BCE. This is the time and place of the earliest known domestication of the horse, which according to this model was the work of early Indo-Europeans. Anthony regards the Khvalynsk culture, located at the middle Volga and dated to roughly 4700-3800 BCE, as the oldest phase of Proto-Indo-European. This culture kept domesticated sheep, goats, cattle, and possibly horses, and was connected to the Danube Valley by trade networks. According to Anthony, the archaic Proto-Indo-European language formed in the Volga Basin from a base of languages spoken by Eastern European hunter-gatherers, with influences from Caucasus hunter-gatherers who migrated from the Caucasus to the lower Volga. The critical catalyst for expansion was ecological and technological. Until around 5200-5000 BCE, the Pontic-Caspian steppes were populated by hunter-gatherers. The introduction of cattle herding, wheeled wagons, and horseback riding transformed these populations into mobile pastoralists capable of covering vast distances. The Yamnaya horizon, which spread rapidly across the Pontic-Caspian steppes between roughly 3400 and 3200 BCE, was an adaptation to a climate change that made the steppes drier and cooler, forcing herders to move their animals more frequently. Anthony describes this as "a new, more mobile form of pastoralism" that came with new social rules and a distinct cultural identity.

• 4. Competing Origins Theories

  ›

  Colin Renfrew advanced the primary competing hypothesis, arguing that Indo-European languages spread peacefully from Asia Minor, modern Turkey, beginning around 7000 BCE, carried by the earliest farmers who migrated into Europe via demic diffusion. The strength of this Anatolian hypothesis is that it ties language spread to an archaeologically well-documented event, the spread of farming. Genetic evidence, however, has increasingly undermined it. Damgaard et al. (2018) found no large-scale steppe ancestry in Bronze Age Anatolia, which complicates a steppe origin for the Anatolians, but this finding cuts both ways: it also suggests the Anatolian branch may have separated from Proto-Indo-European before the steppe populations built up their characteristic genetic profile. A third proposal, the Armenian hypothesis of Gamkrelidze and Ivanov, places the original homeland south of the Caucasus, specifically within eastern Anatolia, the southern Caucasus, and northern Mesopotamia, in the fifth to fourth millennia BCE. Their argument rested partly on the disputed glottalic theory of Indo-European phonology and on evidence of Semitic loan-words in Proto-Indo-European. Mallory and Adams consider the proposed Semitic loans for words meaning "bull" and "wine" or "vine" to be more likely than some other proposed borrowings. David Anthony, however, rejects the idea that the Bronze Age Maykop people of the Caucasus were a major southern source of either language or genetics for the Indo-Europeans, pointing out that Maykop paternal lineages differed substantially from those found in Yamnaya remains and that Maykop populations had absorbed significant Anatolian Farmer ancestry that the Yamnaya lacked. The debate has been revived by recent DNA research that led scholars such as David Reich to propose a Caucasian homeland for an archaic or proto-proto-Indo-European stage, from which speakers migrated both into Anatolia and northward onto the steppes. Anthony criticizes these Caucasian origin proposals, concluding that the roots of Proto-Indo-European were mainly in the steppe rather than the south.

• 5. The Hittites and the Anatolian Branch

  ›

  The Anatolians' earliest linguistic and historical attestation comes as personal names in Assyrian mercantile texts from 19th-century BCE Kanesh, making them among the first Indo-Europeans to appear in the written record. The archaeological discovery of the Hittite archives and the subsequent classification of the Hittite language to a separate Anatolian branch caused what the source describes as "a sensation among historians," forcing a re-evaluation of both Near Eastern history and Indo-European linguistics. Anatolian is generally considered the first Indo-European branch to split from the main group, and its archaic features have led some scholars to suggest it might be a "cousin" of Proto-Indo-European rather than a direct daughter. According to Anthony, if the Anatolian branch separated from Proto-Indo-European, it likely did so between 4500 and 3500 BCE. The phylogenetic analysis by Kassian et al. (2021) placed Hittite as the earliest language to split off, around 4139-3450 BCE. The Hittite empire reached its height during the mid-14th century BCE under Suppiluliuma I, when it encompassed most of Asia Minor as well as parts of the northern Levant and Upper Mesopotamia. Though firmly in the Bronze Age, the Hittites were forerunners of the Iron Age, developing the manufacture of iron artifacts from as early as the 14th century BCE. After 1180 BCE, amid the Bronze Age Collapse associated with the arrival of the Sea Peoples, the Hittite kingdom disintegrated into independent Neo-Hittite city-states, some of which survived until as late as the 8th century BCE. By the Middle Ages, all the Anatolian languages were extinct, absorbed over centuries of successive invasions by the Phrygians, Medes, Persians, Greeks, Galatian Celts, Romans, and Oghuz Turks.

• 6. East to the Tarim Basin: The Tocharians

  ›

  The Tocharian languages are known from manuscripts dating to the 6th to 8th centuries CE, found in oasis city-states on the northern edge of the Tarim Basin in what is now Xinjiang, China. These languages are Indo-European, and their presence thousands of kilometers east of the steppe homeland is one of the more striking puzzles in the whole migration story. The Tocharians are thought to have descended from the Afanasevo culture of Siberia, which arose around 3300-2500 BCE from a migration of people out of the pre-Yamnaya Repin culture at the Don river. The phylogenetic dating by Kassian et al. (2021) placed the Tocharian split from the main Indo-European trunk at roughly 3727-2262 BCE. For a long time, the Tarim mummies, dated from around 1800 BCE, were thought to represent Tocharian speakers from the Afanasevo culture who had moved into the Tarim Basin. A 2021 genetic study overturned this interpretation, demonstrating that the mummies were remains of locals descending from Ancient North Eurasians and ancient Northeast Asians. The same study suggested instead that Tocharian may have been introduced to the Dzungarian Basin by Afanasievo migrants from the Altai-Sayan region of southern Siberia, who in turn had close genetic ties with the Yamnaya of the Pontic-Caspian steppe. The Indo-European eastward expansion left material traces on Chinese culture, introducing wheeled vehicles and the domesticated horse. By the end of the 2nd millennium BCE, populations as far east as the Altai Mountains and as far south as the northern outlets of the Tibetan Plateau showed what Christopher I. Beckwith described as "Europoid" physical traits. A gradual shift in the population balance began around the 5th to 4th centuries BCE, but Eastern Central Asia remained predominantly Indo-European-speaking well into the 1st millennium CE.

• 7. Genetics, Plague, and Old Europe

  ›

  Research by Haak et al. (2015) found that roughly 75% of the ancestry of Corded Ware-related people came from Yamnaya-related populations, while Allentoft et al. (2015) showed that the people of the Sintashta culture were genetically related to those of the Corded Ware culture. These findings give the migration hypothesis concrete biological support. Between roughly 4000 and 3000 BCE, Neolithic populations in western Europe declined, probably due to the plague and other viral hemorrhagic fevers. Into this demographic vacuum moved Indo-European-speaking populations from the steppe. The study of Britain's ancient DNA found that more than 90% of the Neolithic gene pool was replaced with the coming of the Beaker people, who carried around 50% Western Steppe Herder ancestry. Danish archaeologist Kristian Kristiansen stated he is "increasingly convinced there must have been a kind of genocide." Evolutionary geneticist Eske Willerslev put it more precisely: "There was a heavy reduction of Neolithic DNA in temperate Europe, and a dramatic increase of the new Yamnaya genomic component that was only marginally present in Europe prior to 3000 BC." Marija Gimbutas had characterized the process as essentially cultural rather than physical, describing the Yamnaya as military victors who imposed a new administrative system, language, and religion on indigenous groups she called Old Europeans. She noted that Old European societies had neither a warrior class nor horses, and that the Yamnaya social structure was patrilinear, patriarchal, hierarchically organized, and warlike. Edgar Polomé argued that as much as 30% of modern German derives from a non-Indo-European substratum spoken by the Funnelbeaker culture indigenous to southern Scandinavia, a remnant of the languages that Indo-European did not entirely erase. Climate likely played an independent role throughout. Around 4200-4100 BCE a cold snap in Europe pushed steppe herders into the lower Danube valley. The second millennium BCE brought widespread aridization that triggered water shortages and the collapse of sedentary urban cultures in south-central Asia, Afghanistan, Iran, and India, fueling new waves of migration.

• 8. Indo-Iranian and the Path to India and Iran

  ›

  The Sintashta culture, dated roughly 2100-1800 BCE at the eastern border of the Abashevo culture, is where the Indo-Iranian language and culture are thought to have emerged. It is also where the chariot was invented. The Sintashta culture grew into the Andronovo culture, dated roughly 1900-800 BCE, whose two earliest phases were the Fedorovo Andronovo culture, running approximately 1900-1400 BCE, and the Alakul Andronovo culture, running approximately 1800-1500 BCE. Indo-Aryan groups moved into the Bactria-Margiana Archaeological Complex, dated approximately 2400-1600 BCE, and then spread to the Levant as the Mitanni and to northern India as the Vedic people around 1500 BCE. Iranian languages spread throughout the steppes with the Scyths and into Iran with the Medes, Parthians, and Persians from around 800 BCE. David Anthony's revised Steppe hypothesis holds that this spread did not happen mainly through large-scale folk migrations but through a process he calls "elite recruitment," in which small but powerful groups of Indo-European-speaking migrants were attractive to local leaders because they offered a compelling social system, good weapons, and luxury goods that marked high status. According to Parpola, local elites joined these small groups and were incorporated through matrimonial alliances. The phylogenetic study by Kassian et al. (2021) placed the Indo-Iranian split at around 2044-1458 BCE, consistent with the Sintashta archaeological culture. Approximately 60% of all Indo-European language speakers globally today speak an Indo-Iranian language, making it the largest branch of the family by number of speakers, and the linguistic descendants of those steppe-born chariot cultures now include some of the most widely spoken languages on earth.