Why was India as the homeland of PIE abandoned?

Question

I have recently become very interested in the linguistics in the problem of the Indo-Aryan migration controversy. I understand in the early 19th century India was favored as the Proto-Indo-European homeland, but then after the development of comparative-historical linguistics that fell out of favour and the Southern Russian one was preferred. In fact it seems there are three major homelands. In addition to the Southern Russian one, Anatolia and Armenia both seem to be taken seriously and have academic currency.

However, I don't understand why the Indian homeland theory is excluded from all discourse and considered fringe. It seems the Indian homeland was initially the most favoured one. The flora and fauna of reconstructed PIE lexicon all seem to point to India (elephants, lion, etc.); syntactically, Sanskrit is the closest to PIE having retained all eight cases, three genders and three numbers, and the original PIE culture only is preserved by India. However, phonetically, PIE is very distant from a Vedic language, mainly because it sounds like a Centum language. So it seems syntactically it is near identical to Vedic Sanskrit, and phonetically it is very different from Vedic Sanskrit, yet still retains archaic forms, like aspirated plosive sounds, such as Bha, lost everywhere else.

However, is it still not possible PIE originated in India, the Centum branches left early and then PIE changed into Vedic Sanskrit at home? Considering that Balto-Slavic is a Satam language and it originated where PIE use to be spoken, cannot the same apply to India?

I do not mean to be antagonistic with this question. I just want to genuinely understand what are the linguistic evidences that preclude India from being a homeland of PIE. All I know so far according to linguistics it is impossible, but I don't know why it is impossible.

Answer 1

From what you state, it seems you have read the papers of people like Shrikant Talageri.

Be aware that these people who side with the OIT (out-of-India) theory are generally completely incompetent as far as linguistics is concerned. The OIT is a politics-driven agenda that tries to find linguistic arguments. Its incompetence and hysterical politicization are what makes the OIT scientifically unappealing.

To be honest, it is quite difficult to reject the OIT on purely lexical reasons, because the Hindu Kush has a somewhat moderate climate that permits a fauna and flora consistent with the words we can hypothesize existed in PIE.

That being said, the OIT makes the whole tree grow out of a leaf.
Normally, families have a radiating pattern out of their maximal center of diversity. In the case of PIE, maximal diversity is located in Eastern Europe and Anatolia.

Like it or not, Old Indian is a subbranch of Indo-Iranian, itself a subbranch of Central Post-Anatolian, itself belonging to the Post Anatolian branch. In other words, Indo-Aryan is a leaf, it's neither the trunk nor the roots.

The OIT cannot account for the internal and geographic structure of Indoeuropean languages.

Another point is that Old Indian is the only language that has substratic words coming from languages like Burushaski, unique to South Asia. If the OIT were true, one would expect Burushaski words to exist in (almost) all Indo-European languages, and it's not the case.

Similarly, Dravidian only has loanwords from Indo-Aryan, not from PIE, which means Dravidian never met PIE.

Another point is that, if the OIT were true, PIE would expand in all directions: to the East of India, the south, etc. Why would PIE only expand thru the Afghan bottleneck? and then turn westward? All this just makes no sense.

In other words, for many reasons, the OIT just does not make sense. It fails to account for what the Indo-European family is, how it's structured, etc.

Answer 2

When a word can be reconstructed to a proto-language, it is generally assumed that there was such a word in the proto-language. Then if the meaning of the word can also be reconstructed, it is generally assumed that there is such a word with that meaning in the proto-language. Since a word meaning "axle" is reconstructable to at least late PIE, it is assumes that there was such a word, which also implies that the speakers of the language had axles (which in turn implies that they had wheels). If we know something about the development of the wheel, we might know something about the (late) Indo-European homeland. If it were a fact that the various branches of IE led to a reconstruction of words for elephants and lions, we could infer something about the homeland, given what we also know about elephants and lions. There is good linguistic evidence that "horse" is part of PIE vocablary, and that horses were important in the Eurasian steppes but not in Iran and India, at the relevant time (see Anthony The Horse, the Wheel, and Language). This depends on archaeology, not just linguistics. Likewise, there are reconstructed words for other temperate-weather flora and fauna, which are consistent with the Kurgan hypothesis but challenge the India hypothesis.

Another form of linguistic evidence for a homeland is the existence of loanwords into the proto-language from another language, where contact existed in one location but not another. The existence of loans from proto-Uralic into PIE would be very problematic for the "out of India" hypothesis; the existence of loans from proto-Dravidian on the other hand would be problematic for the Kurgan hypothesis.

The important thing to get from this is that there is never purely linguistic evidence for an Urheimat. All such arguments rely on non-linguistic archaeological knowledge (such as whether there were elephants at the time). There are a number of argument for the Kurgan hypothesis. Then an advocate of an out-of-india hypothesis would have to counter those arguments, and advance arguments in support of an Indian origin.