It has been known for 150 years now that the Indo-Aryan languages came from outside of India. The evidence is overwhelming, primarily linguistic, but there is also some archeological evidence. In scholarly circles, there is no debate on the Aryan Migration Theory (AMT) and there has been little debate for 150 years. It is only among Indian nationalists and a few hacks and kooks that it is not accepted.
1. There is a substrate of a language that looks like a Munda language in the Rig Vedas. A Munda language was probably spoken in the Punjab when the Aryans migrated there. About 4% of the words in Rig Vedas are these early Munda loans. None of these Munda loans are found outside of Indic.
They would be found all through IE if the Out of India Theory (OIT) was true. The OIT holds that Aryans inhabited North India for 8,000 years, all the while the Dravidians were in South India and Munda tongues were in East India. Obviously, the Aryans came into Punjab and there mass language shift from a Munda language to Indo-Aryan (IA). The language shift is evident in the sparse Munda loans into Vedic Sanskrit.
There are also a few place names left in North India from the original Munda language of the Greater Punjab area. There are some river names left in Eastern Punjab and Haryana where the local Indus Valley Civilization (IVC) continued for some time after the arrival of the Aryans. These names would not be there if the OIT was true.
There are a large number of IA words for local plants and animals and for agriculture that have been borrowed from the Munda language of Punjab. There would be no reason for the IA people to borrow these terms if the IA people were native to Punjab. Instead, this borrowing is precisely what we would expect to see when pastoralists from Central Asia move into the tropics, encounter new plants and animals and start farming – they borrow the terms for these new living things and technologies from the locals.
This is particularly so in the case of farming, which was left to the local people – the Sudra caste. The IA people only brought a few farming related words with them from Central Asia – the remainder were borrowed from the new locals.
40% of Hindi agricultural words still derive from an unknown pre-Munda language of the Indo-Ganges Plains. Nahali, a small language in Madhya Pradesh, at successively lower levels of its vocabulary, displays high levels of borrowing from earlier tongues. 36% of vocabulary is of Kurku (Munda) origin and 9% is Dravidian. At the oldest level, 24% have no cognates in any known language and appear to have derived from the oldest language known from India.
2. There is an old set of shared loans between proto-Indo-Aryan and proto-Iranian for a number of agricultural and other cultural items that appeared in the Bactria-Margiana (BMAC) 3700-4200 YBP. The BMAC is located more or less in present day Turkmenistan. Obviously, these shared loans were picked up by the proto-Indo-Iranian people as they moved down from the steppes of Kazakhstan and Russia into the BMAC, conquering the people who lived there.
There are references in the Rig Vedas to the conquest of the BMAC peoples by the Arya. For this sequence of events to have occurred, the Indo-Aryans would have had to have moved through the BMAC during this time period and later moved into Iran and India, not the other way around. The language of the loans is not known, but it is apparently the language of the BMAC people.
So there is a BMAC or Central Asian substrate in Indo-Iranian. A possible guess for the language of the BMAC people might be a relative of the Burushaski language of northern Pakistan.
The substrate of the Rig Vedas is a Munda language. The substrate of the earlier Proto-Indo-Iranian language is the language of the BMAC. This sequence is not possible under the OIT and is only possible under the AMT.
3. There are early Indo-Aryan loans into the Caucasian language of the Mittani, who lived in northern Iraq and Syria. These loans are dated to ~3400 YBP. These loans are from an earlier form of Indo-Aryan than is used in the Vedas. Therefore, the Vedas must have been composed 3000-3500 YBP and could not have been composed any earlier.
Also, the Mittani could not possibly have come out of India as the OIT demands, since the IA loans do not show any Indic influences. Nor could the loans have come from Iran, as there are numerous IA Gods in the Mittani texts who are marginalized or do not exist in Iran. The loans must have come from somewhere else, apparently the north.
4. We have numerous references in the Vedas to battles between the Arya with their stone forts, metals, horses and chariots against the more sedentary peoples living in South Asia at the time.
5. There are pottery shards in the BMAC that resemble that shards found in the steppe culture to the north. This indicates that there is cultural resemblance between the two cultures. The suggestion is that the shards are Indo-Aryan and appear first on the steppes and then again in the BMAC with its conquest by the Aryans.
6. The chariot appears in the Urals 4000 YBP and then spreads rapidly in many directions with the spread of IE languages, to Europe, to China via the Tocharians and of course to India and Iran via the Aryans. The horse also appears in South Asia (Pakistan) 3700 YBP in conjunction with the chariots. The modern horse is not native to South Asia, so obviously it came from outside, obviously from the Aryans. The indigenous horse of South Asia, the Siwalik horse, was long since extinct.
7. There are specific Punjabi and Uttar Pradesh loans in Vedic Indic that are not found in Iranian. Therefore, Iranian could not possibly have come from Indic as the OIT demands. The languages must have split in the BMAC, one line going to Iran and another line going to India.
8. The Soma ritual originates in the high mountains of Central Asia – the mountains of Iran, the Himalayas and the Pamirs – with the proto-Indo-Iranian peoples. The original name for the plant is a Central Asian term amsu . This term is borrowed into Indo-Iranian and eventually becomes soma, etc. Later, it moves down into Iran and India and appears in the Vedas. Therefore, the Aryans brought the Soma ritual with them from Central Asia to Iran and India.
9. There is tremendous evidence for a common Indo-Iranian language, mythology and ritual. This shared heritage is not possible with the OIT. It is only possible if there was an Indo-Iranian people, who then split into the Iranian and Indic branches.
10. The Vedic branch of IA becomes innovated and Indianized (in particular, the retroflex consonants) after its arrival in Punjab, while the Iranian branch escapes this development because it did not enter the subcontinent then. In addition, Iranian lacks any specific Indic terms. According to the OIT, the Iranian branch must be Indianized too, or else all of the Indic terms were somehow lost in Iranian.
Since it is not, both branches came from outside India, to the northwest. Iranian languages cannot possibly have come from the Punjab. An early date for Iranian to leave India is preposterous, and Old Iranian (Avestan) is too archaic to have left India after the Vedas. All this means that Iranian and Indic must have split before the Vedas and thus, not inside India. The OIT for Iranian lies in ruins.
11. Zero specifically Indic words are found in IE languages outside of India. For the OIT to be correct, many Indic terms should be found in all the other branches of IE. After all, the Gypsies left India 1000 years ago and took a large specifically Indic set of terms with them to Europe and beyond.
12. Retroflexion. According to the OIT, all branches of IE would have had to have lost their retroflexion after they left India. How likely is that? What we do find, though, is that those branches of Iranian which move east to abut the Indic languages do acquire retroflexion. Since retroflexion is in general not present outside Indic or languages abutting Indic, it must be a late development in IE specific to Indic and cannot have been part of the original IE language as required by the OIT.
Retroflexion only effects those moving into the Indic plain and the eastern Iranian lands, but everyone moving out of South Asia somehow loses it. This does not make sense.
13. Chariots. For the OIT to make sense, chariots must be exported from India 7,000 YBP. However, chariots only appear 4000 YBP in the Urals and NW Kazakhstan and from there spread from Ukraine to Mongolia. The western IE languages retain an IE root rotho for wheel because they had already moved away before the chariot had actually been developed in the Urals. Everyone to the east uses the IA form ratha. This could only be the case if the IA languages moved south from Urals.
Further, according to the OIT, chariots that appear in the Rig Vedas must show up in the text before they have even been invented. Linguistics shows that the word must have been innovated in proto IA at the Urals, for it is present in both branches of IA. This word, along with its invention, can be proved to have been innovated in the steppes and and then carried into India and vice versa could not possibly be the case.
14. Lack of tropical core vocabulary in IE. The core vocabulary of IE shows that the IE homeland was a temperate or even cold place. The plants and animals in the IE language include such cold weather animals, plants and weather words as the otter, beaver, wolf, bear, lynx, salmon, elk, red deer, hare, hedgehog, mouse, birch, willow, elm, fir, ash, oak, beech, juniper, poplar, apple, maple, alder, hazel, nut, linden, hornbeam, and cherry in addition to snow.
A few of these are found in South Asia, but most are not. There are no specific Indic plant and animal names found outside of India, even where these plants do occur outside of India. The OIT would assume retention of at least some of these terms, would it not? Instead, what we find is that a few core IE terms are modified inside India to apply to new plants and animals.
For instance, IE beaver bheber is adopted for the mongoose in South Asia, since beavers do not exist there. IE willow becomes reed, cane in India. So we see that IE temperate plant and animal terms are adopted for the newly encountered tropical living things in India. The flow is into India, not out of India.
For the OIT to be right, the IE languages would have had to have coined these terms after they left India. However, this is not what happened. Instead, the words were IE words from the core IE language itself, which, according to the OIT, was only spoken in India. But these plant and animal names could not possibly have been created in India because most don’t even exist there.
For the OIT to be correct, IE core vocabulary should indicate a tropical climate.
15. Early loans in very early IE. The earliest loans in IE are from Semitic languages of the Middle East. This is possible with an IE homeland in SW Russia or Anatolia, but not possible if the IE homeland is in India, as the OIT requires.
16. Typological features of IE. The typological features of IE are between Kartvelian in Georgia and Uralic in the Urals, as we would expect with an IE homeland in SW Russia, and unlikely with an IE homeland in India.
17. Skeletons. Where are the Indian bones? The OIT requires not a trickling out of India, but a massive migration out of the Punjab. Yet Indian bones look remarkably different from Middle Eastern and European bones. With the massive migration out of India required by the OIT, we should find Indian bones in all of the branches of IE. One would have to argue that the IE speakers who left India did not look like the rest of the Indian people.
18. Facial characteristics. DNA analyses of burials in the Kurgan area near the IE homeland 6000 YBP shows that 60% of the early IE people there had light hair and green or blue eyes. How many Indians, even North Indians, have light hair and light eyes? Almost none. Clearly, the Kurgan peoples were a European type of people. They moved down into Iran and India and mixed with the darker folks already there, creating the present day swarthy peoples of South and Central Asia.
19. Very early Proto IA loans in Finno-Ugric. The homeland of the Finno-Ugric people is somewhere in the Urals. The homeland of the Proto Indo-Aryan people is also somewhere in Urals, especially at the very southern end. The only way for these early PIA loans to get into Finno-Ugric is if the PIA homeland is in the Urals. It’s not possible with the OIT, which generally makes a separate Indo-Aryan branch impossible anyway.
20. Vedic is later than Hittite. For the OIT to be correct, Vedic must be the most ancient branch of IE of them all, very close to Proto IE itself. Yet Hittite, attested from 4000 YBP, is earlier than Vedic. In fact, it is later than Eastern IE, Proto IA, and even pre-Vedic, so Vedic must be a fairly late development in IE. In fact, Vedic is even later than the early forms present in Mittani 3400 YBP.
21. Sanskrit is the most ancient language in all of IE and looks a lot like the original IE language. This is the OIT claim. In fact, IE does not look much like Sanskrit at all. And Sanskrit is not even the oldest attested IA language. Vedic comes first, then Epic Sanskrit and then Classical Sanskrit, and Vedic itself cannot possibly be older than 3500 YBP. The IE language is dated to 6500-8000 YBP (I favor the earlier date). Epic Sanskrit appears only 2500 YBP and Classical Sanskrit comes even later.
22. Lack of IA archeological sites. This is a classic OIT argument. Actually, we do have quite a few site. From the original Proto-Indo-Iranian sites in Sintashta southeast of Urals to the BMAC in Turkmenistan to the Yaz Culture in northeast Iran to the Swat Culture in the Swat Valley of Pakistan to the Cemetery H Culture in Punjab to the Copper Hoard Culture to the south, to the Painted Grey Ware Culture to the south and east, we have a long stretch of cultures that have long been associated with the AMT by archeologists.
Cemetery H in particular shows a possible move away from IVC culture. While the pottery is of course the same, there is a new design on the pottery. On funeral urns we see a small picture of a man with a bird inside of him. This seems to indicate the Vedic belief that the souls of men could fly like birds. Cemetery H also shows a new burial style – cremation and deposit of remains in burial urns. These changes in culture are probably due to Aryan influence.
The ideal Aryan archeological site, however, has typically not yet been found. The ideal site would have the remains of horses and their furnishings, chariots, A Vedic ritual site with three fireplaces west of a river, a flimsy and primitive building pattern of bamboo huts, tools made of stone, copper and bronze, gold and silver ornaments, food consisting of barley, milk products and the meat of cows, sheep and goats. However, the pottery style would remain local, as the Aryans did not innovate pottery.
Such a site has continued to elude searchers, but one has been found in Swat. Swat is mentioned in the Vedas as Aryan territory – suvastu.
23. No Aryan bones. Another OIT argument. It’s quite common for migrations to not be represented by skeletal remains. The remains of the Huns, a large force of proven invaders who conquered Hungary have only just been found in the past 20 years. The most recent research indicates that the Aryans left language, but few genes, in India.
This is reasonable and is often the case with many migrations and invasions. The Huns left as little genetic imprint on the Hungarians as the IA people did on Indians. The Magyars also left their language in the Danube, as the IA people left their IA language in India.
24. European appearance of Indo-Iranian peoples. There is no getting around it. The speakers of Indo-Iranian (II)languages often look strikingly European. This is particularly the case of Iranians, who consider themselves White, or Europeans outside of Europe. The speakers of II languages in Afghanistan often look very European. People in northern Pakistan are some of the most European looking people in the region. Punjabis often look very European, and they look much different from the South Indians to the south.
For the OIT to be correct, this should not be the case. All across the region, all II speakers should look like South Indians, and so should the Punjabis of North India. That II speakers look so European is evidence that they are partly descended from the very European looking peoples of the Kurgan culture of southern Russia. They moved south and east into Central and South Asia, bred in with darker locals, but still retain a strong resemblance to their European roots.
25. LANDSTAT photos indicate the drying up of the Sarasvati River 3900 YBP. A stable of the OIT argument. Since the Sarasvati is mentioned as “the great river” in the Vedas, this proves that the Vedas are much older than 4000 YBP, despite the copious linguistic evidence. The problem is that LANDSTAT photos cannot indicate geographic times.
Further, the Sarasvati River situation is very tricky. The situation as represented in the Vedas is the same situation as exists today. The upper Sarasvati is a significant river, and this is where the settlements were. The lower Sarasvati had already begun to dry up, and by the time of the Vedas it emptied into an inland lake. In a few places, the lower river goes underground in the alluvial Punjabi plains and disappears.
Archeological investigation indicates that settlements along the lower river were abandoned as the river dried up around this time. As you can see, the “Sarasvati River dried up” meme is a huge red herring.
26. No memories of an Aryan migration. Another OIT line. First of all, it is quite typical of most people to have no memories or false memories of wherever they came from. The Romans said they came from Greece. The Gypsies say they came from Egypt.
However, the Vedas do contain vague references to former habitations, such as what appears to be the BMAC and there are references to journeys over mountains and mountain passes. Many place names in Afghanistan are from proto-II words from Central Asia and often lead back to ancient Central Asian enemies of the Arya referred to in the Vedas. One of these is the Parni, associated with the BMAC and later with a northern Iranian group. They had stone forts and well-built cow stables in northern Iran that look a lot like earlier BMAC structures.
The route of migration did not take place over the high passes of the Himalayas and the Pamirs. Few groups have migrated over these treacherous mountains in the last 2000 years. Instead, the migration went from the BMAC down through northern Iran to Herat in West Afghanistan to the Gomal River in near Ghazni in East Afghanistan to the Swat Valley.
There are frequent references in the Vedas to southward and eastward movements of various groups of Arya. There are no references to westward groups as would be required by the OIT. Some of these movements to the south and east are described in military terms as victorious conquests. There are also references in the late Vedas of movements of the Arya east from the Afghan/Pakistan border to Haryana, Uttar Pradesh and all the way to Bihar.
27. Archeoastronomy. OIT proponents like to push this theory. Supposedly, the positions of stars are mentioned in the Vedas. By analyzing the positions of stars in the Vedas, we can make claims about when the Vedas were written via tracking the movements of stars in ancient days.
However, archeoastronomy is a field in poor standing. All we can learn for sure from archeoastronomy is that the Vedas were written some time in the past 8,000 years. All else is up in the air.
The Indian Astrophysicist Rajesh Kochhar has clearly mentioned that the astronomical data in the Vedas is not reliable.
28. The association of Andronovo culture with Indo-Iranians is controversial. So say the OIT proponents. This is not true.
Andronovo is a culture associated with the proto Indo-Iranians that stretched, in its formative location, around northern Kazakhstan and and west into Russia to near Samara, then down to the Caspian Sea, covering most of the northwest quadrant of Kazakhstan.
Later its borders enlarged. At maximum, its northern boundary was from Samara in the Volga Basin east to Anzhero Sudzhensk northeast of Novosibersk in southern Siberia.
The eastern boundary bordered on the Afansevo Culture in eastern Kazakhstan, southern Siberia and Xinjiang. Andronovo did include part of Xinjiang in the far north where the Altai Mountains come down.
The eastern border then encompassed most of eastern Kazakhstran except the area east of Balquash Koli, moving down to the border with Kyrgyzstan in the south, encompassing most of Uzbekistan except the far south, the northern half of Turkmenistan all the way to the the southeastern shore of the Caspian Sea. The Aral Sea was the realm of the Andronovo People.
The relation of Andronovo to the Indo-Aryan people in particular, as opposed to Indo-Iranians in general, is more controversial, but has been suggested by some experts.
29. Chariots could not go over the Hindu Kush. Another OIT argument. But as noted above, the Aryans did not move down through the Hindu Kush; instead, they came east from the BMAC through northern Iran to Herat in west Afghanistan east to around Ghazni over to the Bannu region in the NWFP of Pakistan. That’s a much easier route than the Hindu Kush.
30. There was no invasion. The invasion scenario has been replaced in the past 40 years to a migration scenario. It seems more likely that instead of defeating the Dravidian people and pushing them to the south, or destroying the IVC, instead the Aryans merely profited from the collapse of the IVC that was already underway.
31. There was no genocide of the Dravidian people, all Indians look alike genetically. No one ever said there was a genocide of the Dravidians by the Aryans. Instead, the Aryans moved in, and there was intermingling and intermarriage with the Dravidians, the combined result being the culture of the Vedas.
32. The linguistic evidence. The case for the AMT and the total non-case for the OIT is made by the linguistic evidence. Everything else is secondary. The case was clinched by Hock 1999 (see references).
33. Indians descend overwhelmingly from the Paleolithic population of India. It’s true that 80% of Indian genes go all the way back to the Paleolithic era. But 80% of European genes go all the way back to the Paleolithic too. Same in Britain. Therefore, Europe and Britain has never experienced any migrations of invasions in the past 10,000 years. The Aryan genetic footprint on Indian genes, if it exists, is doubtless less than 10% of the total. It’s well known by now that the Aryans left language, but few genes, in India. Identifying genetic history with linguistic history is naive.
Keep in mind that the Aryans were probably installed a superstrate over the existing Dravidian population. The Aryans were probably no more than 10-15% of the population genetically, and the remaining 85-90% were Dravidians.
34. How could a more primitive people like the Aryans replace the language of the more civilized people, the IVC Dravidians? So ask the OIT theorists. However, let us note that Greek speakers in the Levant, Aramaic speakers in Mesopotamia, Coptic speakers in Egypt and Romans in northern Africa all got their languages replaced by the culturally inferior Bedouins of Arabia. This sort of thing happens all the time.
35. There is no solid proof an Aryan migration to India in archeological terms. This is true as far as it goes, but all it means that is that archeologists typically refuse to characterize migrations in terms of who is migrating where. While there is no archeological proof for an Aryan migration, there is also no proof for Greek, Germanic, Italic, Celtic or Armenian migrations in those branches of IE either.
36. The Rig Veda says that the Sarasvati River flows to the sea. According to OIT folks, since the river dried up 3900 YBP, if the Vedas discuss it flowing the sea, they must have been written before 4000 YBP. However, this statement is only in one sentence of the Vedas, and the word “sea” in question is actually samuda, which Sanskrit experts say can mean lots of thing, but in this case means and inland sea or lake as formed by a river emptying into a desert. Which is what the Sarasvati did. The Sarasvati never emptied into the sea at any time.
37. Horses. OIT proponents keep claiming that they have found horse bones or evidence of horses on seals or objects at some early date. None of this has been confirmed, and some cases have involved overt fraud by Indian nationalist “scholars.” The earliest confirmed horse in the region is at Pirak 3800 YBP. Many horse remains have been found after that, but none earlier.
38. The AMT was invented by Max Muller in 1848. Muller as a British spy – agent – whatever who was sent by the British to falsify the history of India so the Indians would lose their national pride. Hence, the AMT is a British conspiracy. Yes, OIT supporters actually say this. The long version is that he was hired by the British East India Company as part of a nefarious plot to denigrate Hinduism.
First of all, the theory was not invented in 1848 nor was it invented by Muller, as it substantially predates 1848 and Muller was not the first to come up with it.
There is no evidence at all that the AMT was hatched as a British conspiracy (other popular theories say that the entire linguistic community was in on this conspiracy), nor has anyone offered any reason how or why the British could profit by making up the theory of a Bronze Age culture in India. Or why the British, who supposedly hated Indians and thought they were inferior, would invent a theory that said that Indians were in part related to the great British people.
- Hock, Hans H. 1999. Out of India? The Linguistic Evidence. In: J. Bronkhorst & M. Deshpande, Aryan and Non-Aryan in South Asia: Evidence, Interpretation and Ideology, 1-18. Harvard Oriental Series. Opera Minora, Vol. 3. Cambridge, MA.
Kochhar, Rajesh. 2000. The Vedic People: Their History and Geography. New Delhi: Sangam Books. 2000.