Pointless Discussions on the Nature of the Mitanni Data
Shrikant G. Talageri
Someone just sent me a tweet by Koenraad Elst:
Twenty years ago this discussion was unthinkable, for most of the anti-AIT camp pulled up their noses for a "pseudo-science" like linguistics & a "ghost language" like PIE. Today they can't be ignored anymore & their knowledge of Indo-Iranian is proving superior.
2:38 AM· Mar 20, 2025
I wondered what it was all about, and it was followed up by a twitter thread of a current discussion on Twitter on the identity of the Mitanni and the nature of the Mitanni data. The participants in the debate seemed to be:
And it seems to be a resumption of an earlier discussion a year ago in March 2024, in which the participants, apart from these three, were:
As I went through the thread, I found myself nonplussed by the thrust of the debate, which seemed to be all about whether the Mitanni
data represented Indo-Iranian, Iranian or Indo-Aryan,
and about “Prakritisms” in general or particularly in relation to the Mitanni
data.
Agreed these are interesting points of discussion and debate, and, although it is almost a settled fact that the Mitanni data represents specifically Indo-Aryan, trying to show that there are Iranian elements as opposed to Indo-Aryan elements in the Mitanni data is a perfectly legitimate exercise. Giacomo Benedetti even gave the URL of an article written by him on 24 May 2017, titled “Were the Mitanni Aryans really Indo-Aryans”, where he attempted to show that the Mitanni data represented Iranian rather than Indo-Aryan:
https://new-indology.blogspot.com/2017/05/were-mitanni-really-indo-aryans.html
All very interesting, but already much
discussed in detail (in my articles on the subject as well by countless western
academicians including Michael Witzel, who agrees that it is Indo-Aryan),
and, in fact, Benedetti begins his above article with the admission that
“The most common theory is that they were more precisely
Indo-Aryans, and it is repeated everywhere”
before beginning to dispute it.
Apart from the exact identity (Indo-Iranian, Iranian or Indo-Aryan) of the Mitanni, another peripheral issue deeply discussed on this thread is the question of “Prakristisms” in the Mitanni data. Of course there have always been Prakritisms at every stage of the Sanskrit language, whether the Vedic stage or the Classical stage, and this has also been much discussed by eminent Indologists and Sanskritists. I do not have (and have no right to have) objections to these discussions, but they are misleading if they are carried on under the impression that they are the key to the main issue. And the way in which one tangential discussion leads to another, all leading to confusion can be seen from the fact that among the “Prakritisms” mentioned in (what I saw of) the thread are the following: “Royal names ending in -sena are a late classical Sanskrit invention. virya > biriya has two different Prakritisms etc. Too many Prakritisms in Mittani”.
I have written and been writing on this subject since my book in 2008, and in internet debates from 2002, and have repeated it in many articles, notably the following:
https://talageri.blogspot.com/2024/03/the-finality-of-mitanni-evidence.html
Royal names ending in –sena are certainly a late invention, but not a “late classical Sanskrit invention”: they are a late invention in the latest part of the period of the New Rigveda, where we already have one important name ending in –sena: in X.98.5,6,8, we have Ṛṣṭiṣeṇa, the father of Devāpī (and therefore of his brother Śantanū).
And “virya > biriya” is a wrong and now rejected speculation. It is now accepted that the correct equivalence is “priya > biriya”, and we have the direct counterpart of the Mitanni name Biriamasda in the Rigvedic name Priyamedha.
As Yajnadevam points out in the discussion: “Skt > Prkt laws as a set are not seen in other languages. Sound laws are due to random changes that occur when children acquire language. If laws were universal instead of random, then languages would change in lockstep and the tree model of language families would be impossible.” Kṛṣṇa, in Konkani is krūṣṇu, but has also become (and among my acquaintances alone) ki:sanu and ki:ṭṭɑ (and in baby-talk, even kū:ṭṇu). Precise historical conclusions cannot be derived from such “phonetic changes”.
But all these are peripheral discussions about the Mitanni: peripheral to the main issue. The main issue is that − regardless of whether the Mitanni data is purely representative of Indo-Aryan or Iranian (or, as someone suggested, Nuristani!) elements or contained mixed elements from these groups, and regardless of whether their language contained specific “Prakritisms” or not (a very speculative exercise given the somewhat clumsy or ambiguous nature of the exact phonetic symbols used in the Cuneiform script in which the Mitanni words and names are found written) – the geographical location and chronological point of time at which the ancestors of the Mitanni parted from their eastern brethren is very clear and unambiguous from the data.
As Witzel makes very clear, the final redactions resulted in changes in the sounds in the original hymns, but not changes in the words. (as in “an ancient inscription”, the words of the RV “have not changed since the composition of these hymns c.1500 BCE, as the RV has been transmitted almost without any change”, but in certain “limited cases certain sounds — but not words, tonal accents, sentences — have changed”. WITZEL 2000a:§1). So any comparison of the Vedic and Mitanni IA data should be on the basis of words and not sounds.
1. The ancestors of the Mitanni very definitely parted from their eastern brethren from the area of composition of the hymns of the New Rigveda (i.e. the area stretching from easternmost U.P and Haryana in the east to Afghanistan in the west), since all the names and name-types, and the few other common words (maṇi, for example), are very definitely innovations of this new period.
2. They very definitely parted from their eastern brethren after the period of composition of the hymns of the Old Rigveda, these hymns of the Old Rigveda being completely lacking in all these new words (the Mitanni elements being, in fact, just a drop in the total huge new vocabulary which separates the New Rigveda from the Old Rigveda):
https://talageri.blogspot.com/2022/08/final-version-of-chronological-gulf.html
3. The geographical area of the Old
Rigveda, in a period far earlier to the
period during which the ancestors of the Mitanni parted from their eastern
brethren, is originally to the east of the Sarasvati river
flowing through Haryana; and the narration of the expansion of the
ancestors of all these people from there, westwards into the area
of composition of the New Rigveda (right up to Afghanistan),
is also recorded in the hymns of the Old Rigveda.
4. Since the Mitanni kingdom flourished in Syria-Iraq from at least 1500 BCE, and their presence there is known from at least three centuries before that, their ancestors must have left the area of composition of the hymns of the New Rigveda (i.e. the area stretching from easternmost U.P and Haryana in the east to Afghanistan in the west) well before 2000 BCE, and this places the Old Rigveda (located to the east of this area) many centuries before that. I leave open the problem of deciding how many centuries earlier.
5. And, in this period, many centuries before 2000 BCE, the Old Rigveda was being composed in and around Haryana, and this area in this period had rivers and local animals with full-fledged Indo-Aryan names, was regarded by the composers of the hymns composed in this period as their beloved ancient homeland, and the hymns give no indication of any non-Indo-Aryans in the vicinity then or earlier.
No discussions about the exact linguistic
identity (within the “Indo-Iranian” spectrum) of the Mitanni
data, or about the presence or absence of “Prakritisms” in that data,
can make any alteration in the above facts.