From the Steppes to the Mountains: The Long Journey of Indo-Europeans into East Asia

For most of the 20th century, the prehistory of central Asia was reconstructed primarily from archaeological cultures, linguistic comparison, and the famously preserved natural mummies of the Tarim Basin in modern Xinjiang, whose European-looking features fueled speculation about an early Indo-European migration into East Asia. The genetic evidence, accumulated through the 2010s and culminating in the 2021 Wang et al. study in Nature (more than 500 ancient genomes from the eastern Eurasian steppe), has produced a far more textured picture than any of the earlier reconstructions allowed. The story turns out to be twofold. There is a real, demographically traceable Indo-European migration into East Asia, beginning around 3000 BCE with the Afanasievo culture, a direct eastern offshoot of the Yamnaya horizon, and continuing through multiple later waves during the Bronze and Iron Ages. But the famous Tarim mummies themselves are not part of this migration: they are descendants of the ancient North Eurasian (ANE) substrate that predates the Yamnaya by thousands of years, and their cultural connections to the Bronze Age steppe world were sustained without substantial genetic input from the migrating Indo-Europeans. The picture that emerges is one of repeated long-distance contact across the Eurasian steppe, with the Indo-European linguistic and cultural package reaching East Asia repeatedly but ultimately failing to establish a permanent linguistic presence, while the descendant populations carry small but recoverable traces of the migrations that did occur.

1. The Yamnaya and the Afanasievo: the first wave

The Yamnaya culture of the Pontic-Caspian steppe (modern southern Russia and Ukraine) developed around 3300 BCE from the admixture of Eastern European Hunter-Gatherer (EHG) populations with Caucasus Hunter-Gatherer (CHG) populations. The resulting Western Steppe Herder (WSH) genetic profile, combined with pastoralist economy, wheeled vehicles, and probably the earliest Indo-European language family, became the demographic and cultural source of multiple major migrations: westward into Europe (the Corded Ware horizon and the broader European Bronze Age), southward into the Caucasus and beyond, and eastward across the Eurasian steppe.

The eastward Yamnaya migration produced the Afanasievo culture, which appeared in the Altai Mountains and adjacent regions around 3300-3000 BCE. Afanasievo material culture is essentially identical to Yamnaya: same kurgan burial practice, same pastoralist economy, same metallurgy. The genetic evidence is even more striking. Afanasievo individuals are autosomally indistinguishable from Yamnaya. Sample SSGM16 from Songshugou in Xinjiang shows essentially pure WSH ancestry, with no substantial admixture from local East Asian or Central Asian populations. This is a direct demographic transplant: Yamnaya populations moved 4,000 km eastward across the Eurasian steppe, established themselves in the Altai region, and maintained their genetic identity for several centuries.

2. The Tarim mummies: not Indo-European

The natural mummies of the Tarim Basin (Xiaohe and adjacent sites in modern Xinjiang) have fascinated researchers since their discovery in the early 20th century. Their preservation in the arid desert, their European-looking phenotypes (light hair in some cases, prominent noses, tall stature), and their textile and material culture connections to the broader Bronze Age steppe world all suggested an Indo-European origin. The genetic evidence has now corrected this interpretation.

The Tarim mummies (Wang et al. 2021, and the more detailed Zhang et al. 2021 Nature paper specifically on the Tarim individuals) are genetically descended from the ancient North Eurasian (ANE) substrate, with closest affinities to the Botai culture of the Kazakhstan steppe (5500-3500 BCE) and to the Afontova Gora individuals from Siberia. They carry essentially no Yamnaya, Afanasievo, or BMAC ancestry. They are an isolated relic population descended from the pre-Yamnaya substrate, who maintained their cultural connections to the broader Bronze Age world through trade and contact without substantial genetic admixture.

The Tarim individuals in the G25 panel (Tarim_EMBA samples like 11KBM1, the Xiaohe BA samples L5209 through L6106, the XHM samples) sit in a distinct region of the PCA, far from the Afanasievo samples (AYIM, G218M5, SSGM16) despite sharing some material culture connections.

3. The Iron Age: Sintashta, Saka, Scythians

The Sintashta-Petrovka horizon of the Trans-Ural region (2000-1700 BCE) produced a second major eastward demographic pulse. Sintashta populations were themselves a mixture of Corded Ware-derived European Bronze Age populations and Yamnaya-derived steppe populations, with their own distinctive material culture (fortified settlements, chariot burials, the earliest documented horse-drawn vehicles). The Sintashta and the related Andronovo horizon (1800-1500 BCE) expanded eastward into central Asia, where they contributed to multiple subsequent populations including the early speakers of Indo-Iranian languages.

The Saka of the central Asian steppes (700 BCE onward) and the broader Scythian-related populations of the Iron Age inherited substantial Sintashta-Andronovo and earlier Yamnaya-derived ancestry, combined with East Asian and Central Asian local substrates. Their burials extend from the Black Sea to Mongolia and contain material culture that shows extensive long-distance connections. The Xiongnu confederation (3rd century BCE to 1st century CE) included elite individuals whose autosomal ancestry preserved a recognizable Indo-European-related component, demonstrating that the demographic legacy of the earlier migrations persisted into the historical period and seeded the later Hun-Avar-Magyar westward migrations.

4. The five major phases

5. Why no Indo-European language survived

The disappearance of Indo-European languages in East Asia, despite the documented demographic input, illustrates the asymmetry between genetic and linguistic continuity. Several factors contributed to the linguistic outcome. First, the Afanasievo and later Indo-European migrants were always demographically small relative to the East Asian and Central Asian populations they encountered. Second, the political and economic dominance of Sinitic civilizations to the east and of Iranian-speaking civilizations (Sogdian, later Persian) to the west created institutional pressure favoring those linguistic worlds. Third, the relatively short duration of Indo-European political organization in the region (the Tocharian Buddhist kingdoms of the Tarim oases were absorbed by Uyghur Turkic populations in the 9th century CE) gave less time for linguistic consolidation than the parallel Indo-European expansion in Europe.

The Tocharian languages preserved in 4th-9th century Buddhist manuscripts from the Tarim oases are the last Indo-European linguistic legacy in East Asia. After their disappearance, the region's languages became (and remain) Sino-Tibetan, Mongolic, Turkic, and Tungusic, with no Indo-European survival. The genetic legacy persists as small recoverable components in some Central Asian and Iron Age populations, but the linguistic story has effectively ended.

6. References

1. Wang, C. C., Yeh, H. Y., Popov, A. N., Zhang, H. Q., Matsumura, H., Sirak, K., et al. (2021). Genomic insights into the formation of human populations in East Asia. Nature, 591(7850), 413-419. DOI: 10.1038/s41586-021-03336-2 East Asia aDNA
2. Zhang, F., Ning, C., Scott, A., Fu, Q., Bjorn, R., Li, W., et al. (2021). The genomic origins of the Bronze Age Tarim Basin mummies. Nature, 599(7884), 256-261. DOI: 10.1038/s41586-021-04052-7 Tarim
3. Jeong, C., Wang, K., Wilkin, S., Taylor, W. T. T., Miller, B. K., Bemmann, J. H., et al. (2020). A dynamic 6,000-year genetic history of Eurasia's Eastern Steppe. Cell, 183(4), 890-904. DOI: 10.1016/j.cell.2020.10.015 Mongolia
4. Damgaard, P. de B., Marchi, N., Rasmussen, S., Peyrot, M., Renaud, G., Korneliussen, T., et al. (2018). 137 ancient human genomes from across the Eurasian steppes. Nature, 557(7705), 369-374. DOI: 10.1038/s41586-018-0094-2 Steppe genomes
5. Haak, W., Lazaridis, I., Patterson, N., Rohland, N., Mallick, S., Llamas, B., et al. (2015). Massive migration from the steppe was a source for Indo-European languages in Europe. Nature, 522(7555), 207-211. DOI: 10.1038/nature14317 Yamnaya
6. Mallory, J. P. (1989). In Search of the Indo-Europeans: Language, Archaeology and Myth. Thames and Hudson. Standard introduction to the Indo-European problem. Linguistic context
7. Davidski, A. (ongoing). Global25 PCA modern and ancient population averages. eurogenes.blogspot.com G25 panel

China_Tarim_EMBA2:11KBM1,0.104717,-0.118817,0.078441,0.199938,-0.088016,0.05271,-0.041127,-0.053998,-0.035587,-0.086015,0.022572,-0.012739,0.02334,-0.052985,0.002443,0.02612,-0.003651,0.007095,-0.003771,-0.01013,-0.057773,0.011376,0.019843,0.02663,0.002155
China_Xinjiang_Ayituohan_Afanasievo_BA.SG:AYIM22BN.SG,0.111547,0.028435,0.048271,0.12694,-0.052933,0.051595,-0.014806,-0.016384,-0.039064,-0.06925,0.006658,-0.009741,0.004608,-0.037158,0.030266,0.022408,-0.003651,-0.000507,-0.004777,0.008254,-0.009733,0.001978,0.003204,0.020846,-0.002395
China_Xinjiang_Ayituohan_Afanasievo_BA.SG:AYIM22BY.SG,0.105855,0.013202,0.05242,0.114666,-0.054164,0.038208,-0.012221,-0.007384,-0.047245,-0.073441,0.003897,-0.004346,0.012933,-0.038534,0.039087,0.014585,-0.017732,0.004307,0.002137,-0.004377,-0.006114,0.005317,0.016885,0.018918,-0.011735
China_Xinjiang_Nileke_Afanasievo_BA.SG:G218M5-2.SG,0.108132,-0.010155,0.059208,0.123387,-0.063396,0.039324,-0.014571,-0.025384,-0.046018,-0.074899,0.008931,0.004796,0.004757,-0.048443,0.029858,0.018165,-0.005607,-0.000633,-0.011313,0.005378,-0.020838,0.010634,0.021322,0.023377,-0.005149
China_Xinjiang_Nileke_Afanasievo_BA.SG:G218M5-3N.SG,0.104717,-0.036559,0.056568,0.115312,-0.060627,0.03514,-0.021386,-0.022384,-0.037428,-0.069432,0.00341,0.002847,0.02661,-0.040736,0.035016,0.008088,-0.017602,0.000253,0.000251,0.004627,-0.021712,0.011747,0.012571,0.025184,-0.005628
China_Xinjiang_Beifang_Xiaohe_BA.SG:GMGM1.SG,0.09675,-0.116786,0.072784,0.197677,-0.112021,0.038487,-0.043947,-0.059074,-0.044586,-0.101688,0.031016,-0.015586,0.028989,-0.069224,0.019137,0.028772,-0.029206,-0.00114,-0.00993,0.003877,-0.043922,0.003586,0.023417,0.005061,-0.008143
China_Xinjiang_Xiaohe_BA:L5209,0.104717,-0.100537,0.072784,0.203491,-0.117253,0.0502,-0.051232,-0.062767,-0.043564,-0.098954,0.030691,-0.014837,0.033003,-0.073766,0.028908,0.026651,-0.014864,-0.003547,0.003897,-0.003377,-0.050661,0.013231,0.032784,0.021449,-0.003473
China_Xinjiang_Xiaohe_BA:L5213,0.101303,-0.102568,0.072784,0.199938,-0.11356,0.052153,-0.049117,-0.065766,-0.049904,-0.099501,0.027444,-0.01169,0.021556,-0.075004,0.030809,0.024662,-0.01369,-0.001267,0.002765,-0.011255,-0.049662,0.007419,0.016515,0.013616,-0.003712
China_Xinjiang_Xiaohe_BA:L6101,0.106994,-0.092413,0.076555,0.196062,-0.106481,0.035977,-0.049352,-0.050536,-0.039064,-0.087838,0.022734,-0.009741,0.01888,-0.064545,0.021172,0.025192,-0.014994,0.003547,0.003645,-0.000625,-0.053531,-0.003462,0.020582,0.013978,-0.003592
China_Xinjiang_Xiaohe_BA:L6103,0.104717,-0.100537,0.075801,0.199615,-0.115406,0.049085,-0.043477,-0.058613,-0.0452,-0.095674,0.025495,-0.015886,0.022299,-0.07294,0.024973,0.017369,-0.012908,-0.00038,0.001257,-0.012256,-0.052532,0.006306,0.020336,0.013134,-0.017962
China_Xinjiang_Xiaohe_BA:L6105,0.09675,-0.102568,0.07467,0.207367,-0.1231,0.047133,-0.050292,-0.058151,-0.037632,-0.094398,0.039136,-0.014087,0.025124,-0.073766,0.025515,0.014187,-0.030771,-0.001014,-0.003645,-0.012381,-0.05091,0.01694,0.024773,0.019762,-0.006107
China_Xinjiang_Xiaohe_BA:L6106,0.097888,-0.099522,0.067505,0.207044,-0.113252,0.051874,-0.049117,-0.056536,-0.03211,-0.088385,0.027606,-0.013188,0.026016,-0.05849,0.03013,0.019358,-0.013038,0.0019,0.004274,-0.01113,-0.052408,0.01014,0.029826,0.018918,-0.012214
China_Xinjiang_Songshugou_Afanasievo_BA.SG:SSGM16.SG,0.120652,0.026404,0.051288,0.136307,-0.048009,0.039602,-0.005875,-0.016153,-0.05379,-0.078179,0.01023,-0.001349,0.007582,-0.039222,0.043159,0.017104,-0.013169,-0.003294,-0.006159,0.004502,-0.016471,0.003586,0.011216,0.024943,-0.003952
China_Xinjiang_Xiaohe_BA.SG:XHM100.SG,0.105855,-0.105615,0.073161,0.208659,-0.112636,0.041834,-0.045122,-0.064613,-0.035178,-0.10515,0.029392,-0.013038,0.017988,-0.0706,0.024158,0.024264,-0.017602,-0.00076,-0.00352,-0.013006,-0.048789,0.013231,0.020582,0.016508,-0.002994
China_Xinjiang_Xiaohe_BA.SG:XHM110.SG,0.108132,-0.093429,0.067505,0.199292,-0.111405,0.044623,-0.051937,-0.049152,-0.044177,-0.096585,0.021922,-0.014987,0.018434,-0.071701,0.032301,0.024131,-0.021644,-0.000887,0.00176,0.002876,-0.052657,0.007048,0.020706,0.01217,-0.00958
China_Xinjiang_Xiaohe_BA.SG:XHM135.SG,0.106994,-0.100537,0.071276,0.192509,-0.119407,0.051595,-0.047942,-0.059536,-0.046222,-0.096221,0.033452,-0.016186,0.023786,-0.062343,0.021172,0.018828,-0.020079,-0.008361,-0.001006,-0.001501,-0.050161,0.004946,0.023664,0.015062,-0.008861
China_Xinjiang_Xiaohe_BA.SG:XHM75.SG,0.111547,-0.094444,0.068636,0.208982,-0.118484,0.048248,-0.045357,-0.065074,-0.033542,-0.098772,0.02582,-0.009292,0.027948,-0.077482,0.026058,0.02148,-0.026207,0.002407,-0.002514,-0.003001,-0.044422,0.019908,0.032661,0.016749,-0.011855