The Peopling of the Eurasian Steppes
Sir Aurel Stein, the indefatigable British linguist, archaeologist, and explorer, led four expeditions into Central Asia between 1900 and 1930.1 There, in the buried caravan cities and caves of the Tarim Basin, Stein found troves of Buddhist and Manichaean documents between the fourth and eighth centuries AD, as well as mummies remarkably well-preserved due to the dry climate. The documents have illuminated the spiritual and commercial world of the Silk Road. They were written in a number of languages, and were recorded on different media and in different scripts. Among the languages were Sanskrit, Prakrit, Saka (an eastern Iranian dialect), Sogdian (another eastern Iranian dialect that was the lingua franca of the Silk Road), Tibetan, Tangut, Chinese, and a previously unknown language family that scholars have since dubbed Tocharian. The so-called Tarim Mummies, dating between 2000 and 300 BC, have shattered previous notions about the history of Central Asia. Subjected to DNA analysis, the mummies have now revealed, to the dismay of both Uyghur and Chinese nationalists, that the first inhabitants of the cities of the Tarim Basin were neither Turks nor Chinese.2 Instead, testing since 2008 confirms that the first inhabitants shared DNA with today’s European populations, while later immigrants from the Middle East and northern India arrived and intermarried with this indigenous population. Mummies of individuals with Mongolian features only date from the end of the first millennium BC. These scientific findings offer little comfort to Chinese or Uyghur nationalists today seeking historical justifications to claim Xinjiang or Uyghurstan as their homeland.3 Instead, they confirm that Stein’s discoveries of the manuscripts and mummies forever altered our understanding of the origins of the Indo-European languages, and the origins of the steppe nomads of Eurasia.
Sir Aurel Stein was the epitome of a Victorian scholar, gentleman, and confirmed bachelor who was always accompanied by a pet dog, invariably named Dash.4 He was, however, born in Budapest to an accomplished Jewish family in 1862, although he and his brother were baptized as Lutherans and received a Classical training in the gymnasium, mastering Latin, Greek, French, and English. Aurel then obtained his doctorate in Sanskrit and Persian at Tübingen in 1883. He departed for further study in England the next year. Three years later, Stein was convinced by the Indologist Rudolf Hoernlé, whom he had met at a conference in Vienna, to travel to India. Between 1888 and 1899, Stein served as the Principal of Oriental College in Lahore. Throughout his life, Stein passionately read and translated Sanskrit and Persian texts, meticulously recorded his travels, penned a voluminous correspondence with colleagues and family, and published extensively on the antiquities of Central Asia. In 1900, he was inspired to undertake his own expedition upon reading the account of the Swedish explorer Sven Hedin’s travels in Central Asia, which was published in 1898.5 For his outstanding contributions, Sir Aurel was showered with honors, obtained British citizenship, and was knighted. But Sir Aurel Stein was not the first to excavate the lost cities of the Tarim Basin, for he owed a debt not only to Sven Hedin, who mapped Central Asia, but also to Hedin’s mentor, the German archaeologist Ferdinand von Richthofen, who coined the name the Silk Road (Seidenstraße) and happened to be the uncle of the flying ace Manfred von Richthofen, the “Red Baron.”6 Sir Aurel was, however, the most celebrated, because he captured the imagination of the reading public of Great Britain, Europe, and North America by revealing the lost world of the Silk Road.
In 1907, during his second expedition, Aurel Stein excavated the Buddhist sanctuary the Mogao Caves, a complex of five hundred temples southwest of the Chinese garrison town of Dunhuang. Dunhuang stands at the eastern end of the narrow verdant Gansu corridor, flanked by the Gobi Desert to the north and the Tibetan highlands to the south. All caravans going to and from China passed through the city. Thankful merchants donated these temples between the fourth and fourteenth centuries. Within the complex, a great storeroom housed a trove of documents on palm paper, wooden tables, and Chinese paper. In return for offering a subvention to restore the temples, Stein convinced the Buddhist caretaker Wang Yuanlu to permit him to remove forty thousand documents, which are today the jewel of the Central Asian collection in the British Museum.7 Stein intimated to Wang Yuanlu that he was devoted to Buddhist Saint Xuanzang, and perhaps that he might even be a reincarnation of the Chinese pilgrim who had recorded his travels along the Silk Road to India during the seventh century AD. In fact, Stein pinpointed lost caravan cities by consulting Xuanzang’s itinerary. French, Japanese, Russian, and Chinese scholars followed with their own expeditions. Among the documents obtained by Stein were fragments in a hitherto unknown language written in a script adapted from Brahmi (the main syllabary used in Northern India). This new language, first deciphered by the German philologists Emil Sieg and Wilhelm Siegling in 1908, has been identified as Tocharian.8
Since the surviving Tocharian texts are primarily translations of Buddhist scriptures composed in Sanskrit, Prakrit, or Saka, philologists quickly identified two languages, Tocharian A and B, which were Indo-European in origin and descended from a mother language centuries earlier.9 In their core vocabulary and grammar, these two Tocharian languages exhibit many conservative features that had not survived in other Indo-European languages. The texts, dating between the fifth and eighth centuries AD, have been found in the caravan cities among the northern rim of the Taklamakan Desert, stretching east from the city of Aksu to Turpan. At the time of the composition of the texts, Tocharian A was already a liturgical language that speakers of Tocharian B could scarcely understand. A third dialect, Tocharian C, was spoken along the southern rim of the Tarim Basin between the cities of Niya and Krorän (Chinese Loulan). Excavations have so far failed to yield fragments of any texts in Tocharian C, but the language can be surmised from the numerous Tocharian loan words detected in Saka and Prakrit commercial documents.
The decipherment of Tocharian as an Indo-European language astonished scholars at the beginning of the twentieth century, who had based their theories on the birth and dispersal of Indo-European languages on a proposal by Sir William Jones.10 In an address on February 2, 1786, Jones proposed to the Asiatic Society (which he and Governor General Warren Hastings had founded two years earlier) that Latin, Greek, and Sanskrit shared so many similarities that they must have descended from a mother tongue in the distant past.11 His address was published in 1788, and Jones went on to suggest that the Celtic, Germanic, and Iranian languages were likely members of this same family of languages, which has since been designated Indo-European. Sir William, who was appointed as one of three jurists to the Supreme Court of the Calcutta Presidency, had set sail for India in 1783. A precious linguist, he had, at a young age, mastered Greek, Latin, Hebrew, Arabic, and Persian, as well as understood the Celtic language because he had spoken Welsh as a boy. Once in India, Jones sought instruction in Sanskrit at Nadiya Hindu University, where he became enthralled with the beauty of the Sanskrit of the Vedas. In the nineteenth century, philologists, expanding upon Jones’s proposal, had classified the Indo-European languages into two discrete families, one named the Centum branch in Europe (comprising Greek and the Celtic, Germanic, Romance, Illyro-Thracian, and perhaps Armenian languages) and the other, the Satem branch, comprising the Baltic, Slavic, Iranian, and Indian languages. The discovery of Tocharian shattered this neat division. Less than a decade later, the philologists were again surprised when the Czech scholar Bedřich Hrozný identified Hittite as yet another ancient Indo-European language, which had been spoken in central Anatolia (today Turkey) in the Bronze Age.12 Hittite, along with its cousins Palaic and Luvian, was recorded on cuneiform clay tablets from the royal capital Hattusas (today Boğazköy), and these tablets dated between the sixteenth and thirteenth centuries BC.13
During his expedition to Krorän in 1907, Sir Aurel also discovered the burials of mummies whom he reported as people who appeared to be of Iranian or northern Indian origin. But he lacked the means of transporting the mummies safely to London, and so he reinterred them, erected over the site stone markers, and looked forward to a day when he or other explorers might return and recover the mummies for posterity.14 That day came twenty years later when the Swedish archaeologist Folke Bergman excavated the earliest mummies from the Tarim Basin in the prehistoric cemetery at Qäwrighul, which had once stood on the shores of Lop Nur, a vast lake fed by the Tarim River.15 Qäwrighul and its sister city Krorän now are ruins in the desert basin that was once the lake because of the river’s change of course and centuries of human destruction of the poplar forests for fuel and building material. The mummies, dating from the Middle and Late Bronze Ages (ca. 1800–1000 BC), exhibited Caucasian features, and the burial goods, use of ocher paints to adorn the deceased’s body, and clothing pointed to an origin far to the west. The so-called Cherchen Man, often dubbed Ur David, had reddish hair, blue eyes, an aquiline nose, and a fair complexion. He had died in his early fifties in ca. 1000 BC, and was interred along with his wife and infant son. He was clad in a red twill tunic and his body was painted with ocher pigments with sun patterns.16 The contemporary Princess of Xiaohe was unearthed by Chinese archaeologists from a cemetery just west of Krorän (Loulan) in 2003. She had flaxen hair, long eyelashes, and a fair complexion.17 She was wrapped in a white wool cloak with tassels and wore a felt hat, string skirt, and fur-lined leather boots—the garb suitable for life on the Eurasian steppes. The most controversial has been the Beauty of Krorän, discovered in 1980, who had died in her early forties, probably from lung failure after a lifetime of inhaling desert air mixed with sand and charcoal, in ca. 1800 BC. This lady, with lice-infested auburn hair and clad in modest clothing, lived a hard life as a member of a merchant community. Uyghur nationalists have hailed her the “Mother of the Nation,” but in speech, culture, and ancestry, she was neither Turkish nor Chinese.
In 1996, Professor Victor Mair of the University of Pennsylvania headed an international team of geneticists, archaeologists, and linguists who studied the mummies in the museums of Xinjiang. Their findings have been confirmed by subsequent genetic studies.18 Based on DNA analyses, the earliest inhabitants of the Tarim Basin were of European origin, although matrilineal descent for some individuals such as the Beauty of Krorän included Siberian ancestors. The burial goods, ceramics, and textiles pointed to a connection with the people of the Afanasievo culture, who had settled on the grasslands stretching from Lake Baikal to the Altai Mountains in ca. 3000 BC. The ancestors of the people of the Afanasievo culture, in turn, had migrated over one thousand two hundred miles across the central Eurasian steppes from the western or Pontic-Caspian (southern Russian) steppes, home to the Yamnaya culture (ca. 3500–2600 BC), where the mother tongue Proto-Indo-European (PIE) had evolved perhaps five centuries earlier.19 No writing of the Bronze Age has survived from sites of the Lopur Basin, but the only plausible explanation can be that these individuals whose remains were mummified spoke an archaic Indo-European language that evolved into the Tocharian speech represented in the Buddhist texts of the fifth through eighth centuries AD. Chinese chroniclers of the Han and Tang dynasties report tall, fiery-haired, and light-eyed barbarians with full beards and in felt caps and leather leggings in the Western Regions (today Xinjiang), who traded in jade and horses.20 Wall frescoes of the rock-cut Kizil Cave monasteries dating to the fourth through sixth centuries depict some native rulers, merchants, or Buddhist monks with red hair and fair-skinned features. The complex, often known as the Caves of the Thousand Buddhas, is seventy-five miles west of Kucha, a major oasis city on the northern side of the Tarim Basin. Kucha has yielded many documents in Tocharian B. The murals, dating between the third and sixth centuries AD, depict scenes from the life of the Buddha, and share close stylistic similarities to the contemporary art of Gandhara.21
The ancestors of the Tocharians, who dwelled in the cities of the Tarim Basin for nearly three millennia, had arrived from the Eurasian steppes in ca. 2000 BC. Yet these Proto-Tocharian speakers were not the first of the Indo-European speakers to depart from the original homeland on the Pontic-Caspian steppes in ca. 3300 to 3000 BC. The ancestors of the so-called Anatolian languages had migrated from the southwestern terminus of these same steppes into the Balkans perhaps five centuries earlier, in ca. 4000 BC.22 At this time, PIE speakers practiced a mixed economy of hunting, herding, and cultivating barley. Both Tocharians and Anatolians were genetically and culturally related to the people of the Yamnaya culture (ca. 3600–2500 BC), who devised the nomadic way of life on the Pontic-Caspian steppes between the lower Dniester and middle Ural Rivers.23 They already spoke dialects of a common language, dubbed PIE, the mother tongue of many of the languages of Europe, Iran, and India today. The immediate ancestors of the Yamnaya people had domesticated the horse as a source of winter meat, and later for riding.24 The Yamnaya nomad applied the large disc wheel to great carts (gers in Mongolian) conveying mobile felt tent homes. Soon they learned to travel rapidly over great distances across the six thousand miles of Eurasian steppes, and so they could exploit the endless grasslands for their ever-growing herds and flocks.25
PIE, spoken over a wide area of over one hundred thousand square miles, is best seen as a continuum of dialects of a koine, or common tongue, which itself must have undergone significant changes over the centuries. As families, clans, and tribes frequently moved in search of pastures and water, they encountered and exchanged words, ideas, and goods with each other. Over time, as speakers of this koine PIE migrated away from the steppes, they evolved their own distinct language as they encountered new peoples. Philologists have reconstructed a lexicon of one thousand five hundred root words of the Yamnaya people’s language, which is dubbed PIE based on comparison of cognates in ancient or conservative daughter languages (notably Greek, Sanskrit, Hittite, Tocharian, Early Germanic languages, Old Irish, and Old Lithuanian).26 Philologists have even reconstructed the principles of grammar, morphology (in the conjugation of verbs and declension of nouns, pronouns, and adjectives), syntax, and the rules of euphony for vowels and consonants that determined sound shifts of PIE words in the daughter languages. The relationship among the daughter languages is easily demonstrated from common words of kinship. Hence, the PIE root word for father, *pH2tér, gave rise to pater in Greek, pater in Latin, pitŕ in Sanskrit, and fædar in Old English.27 All Indo-European languages, both ancient and modern, share many such cognates for numerals, pronouns, common adjectives and verbs, and nouns denoting parts of the body, flora, fauna, geographic features, and direction. Even certain names of gods known in historical times descend from a common religious vocabulary. Most significant are the common words relating to the horse, wheel, and chariot. Six core words—two distinct words for wheel, axle, wagon, horse, and riding—shared by the daughter languages can be derived from the same PIE root word.28 Such words associated with the pastoral way of life can only date from ca. 3600 BC, when the Yamnaya people adopted a mobile way of life. At the same time, archaeological excavations since the late nineteenth century reveal a significant decline of settlements on the steppes stretching from the Dniester to the Ural Rivers. Furthermore, this region was then home to the vast majority of the horses in the world (perhaps 85 percent).29
Archaeologists have illuminated the material world of the Yamnaya people of the Pontic-Caspian steppes. They were likely descended from earlier foragers, who had sustained themselves primarily by hunting, fishing, and gathering wild grains, nuts, fruits, and vegetables in ca. 6500–5000 BC. These foragers traded with the Neolithic agriculturalists of the Çris culture (between the Dniester and Lower Danube) and Celshanka culture (between the Lower Volga and Ural Rivers), who had learned farming from immigrants arriving in considerable numbers from the Neolithic settlements of Anatolia.30 The inhabitants on the southern Russian steppes evolved a number of distinct local cultures with their own ceramics, material goods, and burial customs during the Eneolithic (or Chalcolithic Age), in ca. 4200–3800 BC. They had gained metallurgical skills in working gold, silver, and copper from the Neolithic villagers dwelling between the Bug and Danube Rivers, whose agrarian culture has been designated “Old Europe.” They also traded with peoples of the Russian forests (likely ancestors of the Finno-Ugric speakers). When in 3600–3400 BC the Yamnaya culture emerged, it marked a dramatic widespread shift to a pastoral way of life across the steppes between the Dniester and the Volga, although farming was still practiced in the river valleys. The Yamnaya proved skilled craftsmen, learning the bronze technology from southeastern Europe or the Near East. They built thousands of shaft graves, covered by tumuli known as kurgans, across the steppes. These graves reflected the emergence of a powerful warrior aristocracy who could afford to bury precious objects.31 The majority of the graves (80 percent) were of adult males, and the rest were likely those of ladies of high rank, so these kurgans were comparable to the later royal Scythian tombs of the sixth to fourth centuries BC.
To the southwest, the nomads of the Yamnaya culture likely raided the settlements of agriculturalists in the Balkans, because archaeology has revealed that many villages of the Tripolye culture in the lower basin of the Dniester and Danube Rivers were fortified.32 To the south of the Yamnaya culture was the contemporary Maykop culture (named after its principal site) on the Kuban steppes, which stretched to the Caucasus and beyond into northeastern Anatolia.33 The people of the Maykop culture were of mixed origin, and while their linguistic identity is a matter of scholarly controversy, they most likely spoke a non-Indo-European Caucasian language (which is today represented by Georgian). They too raised kurgans, produced fabulous “animal-style jewelry,” and shared in the material culture of their nomadic neighbors to the north. They also had trade links across Transcaucasia to the earliest cities of Sumer (ca. 3600–3100 BC), so they too might have transmitted the smelting of bronze to their neighbors, the Yamnaya people.
By adopting their very way of life, the nomads of the Yamnaya peoples possessed in their ox-drawn mobile homes and horses the means to trek across great distances. Just as important, they traveled and traded widely so that they gained a far greater understanding of geography than any of their contemporaries. These two advantages have proved vital for the success of all later Eurasian nomads down to the Mongols. In response to drought, overpopulation, or invasion, nomads could quickly set out in search of new homes.
Speakers of the Proto-Anatolian languages (Hittite, Luvian, and Palaic) were the first to migrate from the Indo-European heartland. The Dniester River had long been the frontier between the steppe peoples and the agriculturalists of Old Europe (today southeastern Europe). Fortification of settlements of the Late Tripolye culture and a shift in material culture likely mark the arrival of the ancestors of the Anatolian languages in the Balkans in ca. 4000–3800 BC. Gradually the Anatolian speakers moved south, and in ca. 2300 BC, they crossed the Bosporus and entered Asia Minor. They might have sacked the royal citadels of Troy II and Alaca Hüyük.34 The Hittites, who called themselves Neshites, settled on the Anatolian plateau, while their kinsmen the Luvians settled in western and southern Asia Minor, and the Palaic speakers in northwestern Asia Minor.
In the fashion of the later Seljuk conquerors, the Hittites imposed themselves as a ruling class over a far larger indigenous population called Hattians. They introduced wheeled vehicles and the horse, and at some point learned chariot warfare from their distant kinsmen on the central Eurasian steppes. Also, just like the later Seljuk Turks, Hittites found the Anatolian grasslands on the central plateau similar to those of the Eurasian steppes and so ideal for herding their livestock and for breeding horses. By 1900 BC, a dozen Hittite kingdoms had emerged in central Anatolia. At each royal center, Assyrian merchants, in search of silver, horses, mules, woolens, and leather, established a commercial colony (karum).35 The Assyrians who had settled at Nesha (Kanesh in Assyrian records, and today Kültepe) maintained a regular correspondence with the home office in Ashur. From the Assyrian merchants, Hittite dynasts learned cuneiform writing and the arts of Mesopotamian civilization. Abruptly, in ca. 1750 BC, the Assyrians abandoned their colonies, because Hittite monarchs incessantly warred against each other and disrupted trade. The most successful, Labarnas (1680–1650 BC), was remembered as the progenitor of the Hittite imperial family, and united central Anatolia. His successor and namesake lifted his father’s curse on the citadel of Hattusas and built a new capital there. He assumed a new royal name, Hattusalis, “the man of Hattusas.”36 The Hittite rulers, styling themselves as the Great Kings of Hatti, forged the first imperial state in Asia Minor based on the power of their chariots.
The Hittites adapted the cuneiform script to their Indo-European language. Thousands of clay tablets recovered from the royal library at Hattusas are the oldest written records in an Indo-European language. The language of the texts reveals that the ancestors of the Hittites had departed at a time when PIE was still spoken as the koine across the Pontic-Caspian steppes. Hittite was not affected by shifts in pronunciation that gave rise to the division of PIE into Centum and Satem language families in ca. 2500 BC. Hittite, along with Luvian and Palaic, exhibits many archaic features that were common in early PIE.37 Hittite grammar lacks the dual. It recognizes only two genders for nouns and pronouns (animate and inanimate rather than the three of masculine, feminine, and neuter in other Indo-European languages), and two tenses (as opposed to four to six in later Indo-European languages). Most remarkably, laryngeals, long posited by philologists as a class of consonantal sounds in early PIE, are found in Hittite, whereas they were abandoned in other Indo-European languages. Furthermore, written Hittite vocabulary, syntax, and perhaps accent were significantly modified by the indigenous languages of Asia Minor.
In ca. 3300–3000 BC, the ancestors of the Tocharians migrated next. They departed from the upper Ural basin, crossed 1,200 arid miles of Kazakhstan’s steppes, and reached the grasslands between Lake Baikal and the western foothills of the Altai Mountains. There, in their new homeland, they created their own distinct Afanasievo culture.38 The immigrants would have found the steppes between the Volga and Lake Baikal thinly settled, so they were soon isolated from their kinsmen of the Yamnaya culture. Hence, Tocharian, just like Hittite, never experienced many of the later modifications common to other Indo-European languages. The conservative, even archaic, character of the Tocharian languages was still evident in the Buddhist texts over thirty-five centuries later.39 Furthermore, contrary to popular perceptions since the Middle Ages, the earliest nomads migrated not from east to west, but the reverse, from west to east. The Tocharians, who headed east from the original homeland, were later followed by the ancestors of the Indo-Iranian languages. Hence, Indo-European-speaking nomads brought the horse, the wheel, and the nomadic way of life to the central and eastern Eurasian steppes.
Meanwhile, on the Pontic-Caspian steppes, the PIE speakers of the Yamnaya culture proved ever more successful given the advances of their pastoral way of life. Success bred periodic surges in population; seasonal conditions produced spring flooding or summer droughts. In a search for water and pastures, families, clans, and tribes continually migrated beyond the horizons of the Yamnaya world, to the southwest, west, and east. By 2500 BC, the koine of spoken PIE had diverged into two daughter branches known as Centum and Satem.40 Each family of languages, in turn, represented a broad spectrum of dialects that evolved further into the main Indo-European languages known today. This proliferation of mutually unintelligible languages resulted from the great distances traversed by descendants of the Yamnaya culture. The wide-flung communities of immigrants lost touch with each other. But from the very beginning, given the absence of writing, there never was a standard version of either a spoken Centum or Satem language. Instead, speakers in each region modified consonants and vowels according to ease of pronunciation, innovated on grammar and syntax for more precision, and adopted loan words from speakers of other languages to express new ideas or objects. In daily conversations, speakers of any language continually alter pronunciation. For example, in modern English, the conservative rules of spelling retain the hard guttural sound k before the nasal consonant n in such words as know or knight, even though speakers have ceased to pronounce the initial k sound since Chaucer’s day. During the last three generations, many English speakers have swallowed the aspirate sound h of the relative pronoun which so that the spoken word is indistinguishable from the noun witch. Context rather than sound now determines the meaning of two words that have become homonyms. In the preliterate world of the Yamnaya culture, such changes occurred frequently, and each community agreed to its own rules of pronunciation. A broad range of such changes in the second half of the third millennium BC is represented by one such major shift of the PIE initial guttural consonant kw to a simple k/[h] sound in Centum languages (either a simple k or, in Germanic languages, a hard h) and in Satem languages into the initial s/sh sound. Among speakers of Indo-European languages in Central and Southeastern Europe, the initial kw sound was modified to a simple guttural sound k, made at the back of the tongue, when it preceded an open vowel such as a, e, or i, which is formed at the tip of the tongue. Hence, the word centrum resulted in Latin for the numeral one hundred—a change shared by the Italic, Celtic, and Germanic languages, as well as Greek.41 Many centuries later, perhaps in ca. 750–500 BC, speakers of Proto-Germanic then altered this hard c/k sound into an aspirate; hence, centrum became hundred in modern English, hundert in modern German.42 Among the speakers of what became the Balto-Slavic and Indo-Iranian languages, the initial hard k gave way to a softer sibilant s or sh, which was spoken at the tip of tongue just like the following vowel; hence, Sanskrit satem for the numeral. Speakers of Proto-Sanskrit also swallowed the medial nasal n in an original santem before dental consonants (d, t, dh, and th) so that the earlier santem was pronounced satem when the language was first put down in writing.43
The linguistic changes resulted from the far-ranging migrations of Yamnaya descendants out of the Pontic-Caspian steppes. The migrations are reflected in the archaeological record and recent DNA analysis of large samples of human remains recovered from graves. The distinct material culture of the Yamnaya culture gave way to a far more widely dispersed material culture designated Corded Ware, extending from Central Europe and southern Scandinavia across eastern Europe to the forests and steppes as far as the Volga River. The culture is named after its signature pottery with corded rings of lateral decoration—today a favorite image on Danish stamps and European bank notes.44 These changes in material culture resulted from migrations by Indo-European speakers of nomadic ancestry. DNA evidence presented by Nicolas Patterson has now confirmed that the steppe peoples of the Yamnaya culture migrated in large numbers into central and northern Europe from 2600 BC on.45 DNA analyses of skeletal remains indicate that well over half of the DNA among individuals possessed the same genes, on both the patrilineal and matrilineal sides, as those of the Yamnaya people. Such a dramatic change in demography resulted when Indo-European speakers moved first into the basin of the upper and middle Danube, and then into the rest of Central Europe and Scandinavia in the second half of the third millennium BC. The newcomers displaced or assimilated the indigenous population, who were themselves a mix of early hunter-gatherers of the Neolithic Age and agriculturalists from the Middle East who had crossed the Bosporus, and gradually migrated along the Danube into Central and Western Europe. From them sprang the language families of Celtic, Italic, and Germanic.46 This major migration of steppe nomads, recovered by over a century of the exacting labors of philologists, archaeologists, and geneticists, added the third and most significant genetic and linguistic component in the making of the European populations of today.
Simultaneously, the ancestors of the Greek, Thraco-Illyrian, Macedonian, Phrygian, and Armenian languages, dwelling on the steppes between the Dniester and lower Don Rivers, either crossed the Carpathians and settled on the Transylvanian grasslands or pressed south across the lower Danube into the Balkans.47 These newcomers’ spoken languages, while classified within the Centum branch, also shared certain features with the Satem languages. By 1900 BC, Greek speakers entered the Hellenic peninsula, where they adapted the aesthetics, mores, and syllabary of the Minoans of Crete. The lords of Mycenaean Greece employed chariots, built citadels, and left the earliest records of the Greek language on clay tablets in the syllabary known as Linear B.48 Thracians, Illyrians (who were likely the ancestors of the modern Albanians), and Macedonians remained in the Balkans. The Phrygians, whom Herodotus called neighbors of the Macedonians in the distant past, crossed from the Balkans into Asia Minor in the wake of the collapse of the Hittite Empire shortly after 1200 BC.49 The Phrygians spread their language across Asia Minor. In the Roman Age, funerary monuments reveal that Phrygian was still spoken as late as the third century AD. The Armenians, likely kinsmen to the Phrygians, migrated farther east to their final homeland in the seventh and sixth centuries BC.
The speakers of the Satem languages migrated in two different directions. The ancestors of the Slavic speakers moved off the steppes into the forests of Russia, while the Proto-Baltic speakers (today represented by Lithuanian and Latvian) pressed northwest, settling along the shores of the Baltic Sea. In these dense forests, both groups encountered Finno-Ugric speakers (ancestors of modern Finns, Estonians, and Hungarians), who taught the newcomers the skills of hunting, trapping, and fishing in the northern climes.50 By ca. 2000 BC, Proto-Indo-Iranian speakers, who might have migrated east from regions of the Corded Ware culture of Central Europe, already dwelled on the steppes straddling the lower Volga valley. There they had acquired spoke-wheeled light chariots and superior domesticated horses bred for endurance and a docile nature from local inhabitants whose material culture is dubbed the Sintashta culture (ca. 2050–1900 BC).51 A royal kurgan at Sintashta dating to ca. 2050 BC has yielded the earliest intact chariot, destined to become the primary weapon of warfare in the Middle and Late Bronze Ages from the Near East to China. Over the next three centuries, Indo-Iranian-speaking nomads of the so-called Andronovo culture (ca. 2000–1450 BC) settled the central Eurasian steppes of Kazakhstan, the future heartland of the Scythians, nomads of the Iron Age who spoke eastern Iranian dialects.
The Indo-Iranians pressing east soon reached the western borders of the Tocharian-speaking peoples of the Afanasievo culture. The arrival of the Indo-Iranians might well have triggered the migration of the ancestors of the Tocharians to quit the steppes and seek new homes in the oases of the Tarim Basin, where they learned to cultivate the river valleys, and traded in jade and horses with the Han Chinese of the Shang Dynasty.52 Other Indo-Iranian tribes, reaching the northern shores of the Aral Sea, pressed southeast and soon arrived at the banks of the Jaxartes River (Amu Darya). Once across the Jaxartes River, they grazed their horses, cattle, sheep, and goats on the grasslands of Ferghana, Sogdiana, and Bactria between the Jaxartes (Syr Darya) and Oxus (Amu Darya). There they encountered sedentary agriculturalists who were members of a culture known as the Bactria-Margiana Archaeological Complex (BMAC).53
The fortified towns and villages in the oases and middle and lower Oxus and Zeravshan valleys had long prospered as the nexus of trade routes extending to the central Eurasian steppes to the north, and west along the future Silk Road to the cities of Sumer and Akkad, and southeast across the Hindu Kush to the cities of the Indus Valley civilization called Meluhha in contemporary cuneiform texts, and Mleccha in the Rig-Vedas.54 Transoxiana provided the literate cities of Iraq and India with a number of exotic products, notably lapis lazuli, tin, copper, spices, timber, slaves, and horses, prized by Sumerians and Akkadians for breeding mules. They had originally obtained horses in limited numbers from the people of the Sintashta culture on the lower Volga. The newcomers from the steppes entered into an economic symbiosis with the indigenous inhabitants that was to be repeated many times in the history of Transoxiana.55 The Indo-Iranian nomads provided superior horses in far greater numbers to merchants engaged in the international trade to Mesopotamia and India. They traded leather, woolens, salted meat, and dairy products for grain, bread, fruits, and vegetables, and the finished goods of craftsmen. The newcomers gained from their hosts the domesticated Bactrian camel. In the Avestan and Vedic texts, the speakers of Iranian and Sanskrit used the same loan word for camel, ushtra, which is unrelated to the word for camel in other Indo-European languages.56 In Transoxiana, the ancestors of the Indo-Iranians prospered and perfected their use of chariots, perhaps shifting from javelins to the composite bow as the favored missile weapon. They also learned the trade routes to the west and southeast.
The ancestors of the Iranians and Indo-Aryans long dwelled together as neighbors in Transoxiana. They shared a common speech and many cultural and religious practices. In historic times, they still worshipped the same gods known from the later Rig-Vedas and Avesta. They both practiced horse sacrifices, imbibed the same fermented beverage (soma), and were organized by occupations in a social hierarchy that crystallized in the caste of Hinduism. The Rig-Vedas, hymns to the early Indo-Aryan gods, and the Garthas, the oldest verses in Avesta (the scriptures of Zoroastrianism), represent the earliest forms of the Sanskrit and Iranian languages, which are very close in vocabulary, grammar, and syntax. These verses perhaps were first composed in the Late Bronze Age (1500–1200 BC). The Vedic hymns were long recited orally until they were written much later, perhaps in the sixth century BC. But the hymns must have been repeatedly edited and redacted.57 The earliest surviving Sanskrit documents are later still: the edicts of the Mauryan emperor Ashoka (269–232 BC), written in Brahmi script, which was based on the Aramaic alphabet of the sixth century BC.58 The earliest surviving document of any Iranian language is the Persian text, written in cuneiform script, on the trilingual inscription at Behistun.59 The Great King Darius I (521–486 BC), the victor in a civil war, commissioned the awesome monument in Akkadian, Elamite, and Persian, to proclaim his legitimacy.
In ca. 1500 BC, Indo-Aryan speakers departed from Transoxiana and traversed northern Iran, and settled on the grasslands of the al-Jazirah, a new homeland congenial to nomads arriving from the Eurasian steppes. When they arrived soon after 1500 BC, they found the Near East divided into warring states in the aftermath of the collapse of the Babylonian Empire. They imposed themselves as a warrior elite known as the mariyanna (“young warriors”), the equivalent of the Vedic Kshatriya caste.60 These Indo-Aryans ruled over a far larger Hurrian-and Semitic-speaking population. They adopted cuneiform writing and Mesopotamian bureaucratic institutions to forge an imperial state in northern Syria and Mesopotamia called the Mitanni. These Indo-Aryan kings of the Mitanni corresponded on terms of equality with the Pharaoh of the New Kingdom, the Great King of Hatti, and the Kassite King of Babylon. In a treaty between the Mitannian King Kurtiwaza and Hittite Emperor Suppiluliumas (1344–1322 BC), the Mitannian gods invoked were Mitrasil, Arunasil, Indar, and Nasattyana. They are, respectively, the Vedic divinities Mitra, Varuna, Indra, and the Nasatya twins.61 An Indo-Aryan master horse trainer, Kikkuli, in the service of the Hittite emperor, composed in his native language and Hittite a manual for training warhorses. The text, dating to the early fourteenth century BC, is the oldest record of the Indo-Aryan language.62 The Hittite king Suppiluliumas invaded and overthrew the kingdom of Mitanni, and sacked its capital Washukanni (today an unidentified tell, or an artificial hill, in the vicinity of Tell Halaf in northern Syria). The Indo-Aryans of the Middle East abruptly disappeared from history.
Meanwhile, other Indo-Aryans migrated southeast, crossed the Hindu Kush, and entered the Punjab and Gandhara as immigrants, mercenaries, and invaders. Over the course of perhaps five centuries (ca. 1500–1000 BC), they settled in the lands of Meluhha. Scholars are still divided over whether these newcomers arrived as invaders or as successors to an urban civilization long in decline.63 To be sure, the cities of Meluhha were already in decline; hence, many scholars posit ecological or environmental causes for the urban collapse. As yet, the limited number of ideograms stamped on Meluhhan glyphs are too few and too brief to be deciphered.64 Records and historical accounts were likely written on palm leaf paper that disintegrated long ago. But if we had such historical texts, they would have revealed a concert of powers, rather than a single state, that had long warred among themselves in the fashion of the contemporary Near Eastern empires. Furthermore, the Rig-Vedas celebrate how the Aryans warred against the Dasas, the Dravidian-speaking people of Mleccha, who were subjected as the Sudra caste. Therefore, it is best to compare the Indo-Aryan migrations to those of the Germanic tribes who entered into the Roman Empire by invitation and invasion between the third and sixth centuries AD. In their new homeland, the Indo-Aryans recorded their ancestral rites of sacrifices and deeds of their gods in the verses of the Vedas, composed in an archaic Sanskrit and recited orally, only to be put down in writing centuries later. The newcomers from the Eurasian steppes also brought the horse, the spoked wheel, and the chariot, and they defined the linguistic and cultural future of India.65
In the early Iron Age (ca. 1000–750 BC), the Iranian-speaking nomads, collectively known as the Scythians, dominated the central Eurasian steppes and Transoxiana.66 They replaced the light chariot with the mounted archer, equipped with the composite bow and astride a superior high saddle with leather toe loops as stirrups. This major shift in warfare enabled a new wave of migrations by Iranian-speaking nomads into the Near East, Iran, and the Tarim Basin from the ninth century BC on. The Scythians represented a confederation of diverse tribes, and later classical sources report among the many tribes the Dahae, Massagetae, and Saka. The Scythians had gained many benefits from their trade with the sedentary populations of Transoxiana and the forest peoples of Siberia. Foremost, they possessed a deep knowledge of geography and trade routes.
Eastern Iranians pitched their tents on the grasslands of Ferghana, which offered ideal grazing for their horses and livestock, and then crossed the Pamirs and settled in the western oases of the Tarim Basin, notably in the later caravan cities of Kashgar and Khotan. They brought their nomadic way of war and their distinct language, Saka.67 A group of mummies from the Tarim Basin dating from the Iron Age (700–1 BC), tested for DNA, has confirmed this migration. The individuals were members of a population who shared ancestry with peoples of northern India and the Middle East. Murals of the Kizil Cave Monasteries of the fifth and sixth centuries AD depict many men and women with dark hair and slender figures typical of Iranians.68 Men are shaven save for a handlebar mustache. Chinese and Persian literary sources report how Turkish kaghans, Persian shahs, and Chinese emperors prized Saka musicians, dancers, and courtesans. Some Saka-speaking tribes migrated even farther east into the Ordos triangle of the Upper Yellow River, where they came into contact with the warring states of Zhou China.69 These Iranian nomads likely taught the steppe way of life, and way of war, to Altaic-speaking peoples of Mongolia, whose common tongue evolved into the later Turkic and Mongolian languages. The nomads dwelling on the Mongolian steppes proved adept pupils, for they would forge the confederation of the Xiongnu that would challenge imperial China.
Simultaneously, other Scythian tribes, perhaps under pressure from the Massagetae, migrated west, back to the ancestral homeland of the Pontic-Caspian steppes, in the eighth and seventh centuries BC.70 They expelled a nomadic people who also spoke an Iranian dialect and were called in Greek sources Cimmerians. The Cimmerians fled across the grasslands of the Kuban, crossed the Caucasus by the Dariel Pass, and invaded the Near East in ca. 720 BC. The Cimmerians, the first nomadic horse archers to burst into the Near East from north of the Caucasus, quickly overran the Armenian plateau. In 714 BC, they defeated King Rusa I of Urartu, who ruled over the Hurrian-speaking kingdom centered around Lake Van that was the political heir of the Hittite Empire. Rusa had suffered a decisive defeat at the hands of the Assyrian king Sargon II the year before; in despair, he committed suicide.71 The Cimmerians thereupon crossed the Upper Euphrates and entered Asia Minor, ravaging the northwestern provinces of the Assyrian Empire. In 705 BC, King Sargon II fell fighting in a great battle on the Cappadocian plateau against the rebel Hittite vassals of Tabal, supported by Cimmerian horsemen.72 The rebels and their Cimmerian allies, however, had suffered heavily and failed to exploit their victory. In 696 BC, the Cimmerians, in alliance with King Rusa II of Urartu, invaded the Phrygian kingdom. The Phrygians, speaking an Indo-European language remotely related to Greek and Macedonian, were newcomers from the Balkans. In the Early Iron Age (ca. 950–750 BC), a succession of Phrygian kings, ruling under the dynastic name Midas (called Mita in Assyrian sources), united western and central Anatolia. Their massively fortified capital at Gordion, today seventy-five miles west of Ankara, succeeded to the role of Hittite Hattusas. Phrygian kings consulted Greek oracles and promoted trade with the Greek colonies on the Euxine and Aegean shores of Asia Minor. The last Midas, who was buried in the great tumulus of Gordion, suffered a catastrophic defeat at the hands of the Cimmerians, and he too committed suicide, reportedly by drinking bull’s blood.73 For the next fifty years, the Cimmerians rampaged across Asia Minor, raiding as far west as the Greek colonies on the Aegean littoral. In response, the Lydian king Gyges (ca. 680–644 BC), ruling from Sardes, made common cause with the Assyrian kings Esarhaddon and Ashurbanipal II against the invaders.74 By 656 BC, the Cimmerians were driven back into Cappadocia and northeastern Asia Minor, where they settled and imposed their Iranian language and mores of the steppe way of life on the indigenous populations.
Finally, the most successful Iranian speakers migrated out of Transoxiana west across northern Iran and settled in the Media and Persia east of the Zagros mountains in the tenth and ninth centuries BC.75 These Iranian-speaking nomads too excelled in horsemanship and archery. The Medes clashed with the Neo-Assyrian kings, and so they were compelled to organize an effective state centered on their capital, Ecbatana (today the massive mound Tappe-ye Hagmatān outside of Hamadan and still to be excavated). In 612 BC, King Cyaxares of the Medes allied with Nabopolassar, the Chaldean King of Babylon, against the Assyrians. The two monarchs captured and sacked the Assyrian capital, Nineveh, and so ended the Assyrian Empire.76 In the fashion of the first Mesopotamian conqueror, Sargon of Akkad, Nabopolassar went on to unite the Fertile Crescent for the last time.77 Cyaxares, commanding the invincible Median cavalry, expelled the Scythian invaders from northwestern Iran and forged a new imperial order comprising the lands of Iran, Transoxiana, and Afghanistan. Yet in the next generation, Cyrus the Great (559–530 BC), the vassal king of Persia, overthrew his overlord the Median King Astyages.78 The Persians, kinsmen of the Medes, had settled in southwestern Iran, Persis (today Fars). They succeeded to the urban, bureaucratic state of Elam and its capital Susa, because the Assyrian monarch Ashurbanipal II had shattered Elamite power in 639 BC. In less than a generation, between 550 and 515 BC, Cyrus and his two immediate successors, Cambyses and Darius I, built the greatest empire of the ancient Near East, spanning three thousand miles from the shores of the Aegean to the Indus valley, and embracing perhaps forty million residents.79 The Achaemenid kings of Persia ruled over an imperial order that rested on the bureaucratic institutions of Babylon and the horse archers of the Eurasian steppes. The Persians had proved to be the most successful of all the Iranian nomads who had migrated across and then off the Eurasian steppes in search of new homes.
In 500 BC, over the course of thirty-five centuries, Indo-European nomads had spread their languages from Europe to India. Those who peopled the central and eastern Eurasian steppes propagated a pastoral way of life based on the horse and the wheel. They transformed the steppes into a great land corridor, which would be traversed by future generations of merchants and missionaries, immigrants and conquerors. Those nomads such as the Anatolians, Tocharians, Indo-Aryans, and Persians who eventually settled new lands outside of the steppes exchanged their nomadic way of life for a settled one, based on agriculture and cities. They altered the linguistic and cultural map of Europe, the Middle East, and India. In so doing, they enriched, and even defined, the classic civilizations of Greece, the Middle East, and India.