Wednesday, July 1, 2015

Do we know if Proto-Indo-European really existed?

In response to my piece tracing the first evidence found of Old Indic to Syria, Rohini Bakshi had this cogent piece. Please do read it since it adds context to this debate.

Let me, however, rebut this rebuttal.

Some of her points are not rebuttals and some answer themselves. I have clearly mentioned that the Mitanni spoke in Hurrian (#2) and worshipped non-Vedic gods (#3), so simply repeating them doesn’t serve much purpose. As for #4, I am not so sure why residents of northern Syria (for, literally, centuries) cannot be called Syrians. What would I call them? Either way, it is a trivial semantic point. And #6 uses BB Lal as a reference, who is, not to put too fine a point on it, a bit of a joke. Happy to see some real references to Rig Veda compilation dates.

The two main points are #5, #7.

Rohini's point #5 reads:

Indologists like Theodore Proferes quote the Mitanni treaty merely as a chronological marker - saying that if the Indo-Aryan gods (as opposed to Proto-Aryan;See Thieme ibid.) feature in the Mitanni treaty, and if they were worshiped for the same functions in the Rig Veda - the hymns are at least as old as the treaty. However Proferes is the first to admit that Max Mueller's dating is arbitrary and there is nothing which says the Rig Veda was not composed centuries before this date.

Yes the Rig Veda's compilation is "at least as old as the treaty” (probably). But that just proves the point. Here's how:

1) By the time the treaty was written, the Mitanni had stopped speaking Sanskrit. In fact, we know they had stopped speaking Sanskrit before 1500 BC (founding date of the empire) since they had switched to Hurrian.
2) This means that the Mitanni were speaking Sanskrit before 1500 BC.
3) This means that the Mitanni were speaking in Sanskrit before the Rig Veda was compiled (since no on dates the Rig Veda compilation before 1500 BC).
 4) QED

David Anthony has a great explanation of this in his book, The Horse, the Wheel, and Language.

Rohini's #7 reads:

Shoaib unqualified-ly quotes PIE (Proto Indo-European) as the founding language of Sanskrit. He doesn't consider it necessary to explain PIE itself is a scholarly construct and there is by no means any agreement on whether it actually existed. Linguistic scholars like N.S. Trubetskoy categorically deny its existence. Even a scholar like Ernst Pulgram who does believe that PIE existed admits, "...a procedure whereby we choose to arrive at a reconstructed proto-language, be it Proto-Indo-European or Proto-Athapaskan, by comparing languages in pairs and advancing by way of a pyramid until we have reached the summit, the reconstructed proto-language, does not in the least entitle us to believe, or postulate, that historically the reverse process-that is, a series of binary splits-actually occurred." So when the jury is still out on PIE I would be very uncomfortable with adding gravitas to an article by quoting PIE (founding language) as a 'Fact.'

Rohini is right that Trubetzkoy denied the existence of PIE. However, Trubetzkoy is an ancient chap – he died in 1938 (always useful to provide a date with an old citation). Almost everyone disagrees with him today including, Rohini’s other citation, Ernst Pulgram. This is what Pulgram has to say:

“Let me first mention, and dispose of, the view (held by few, notably by Trubetzkoy, though perhaps only for a time and for the sake of argument) that there never was a Proto-Indo-European language, but that the similarity of the attested Indo-European idioms is due to progressive assimilation through constant and enduring contact with one another.”

The paper that Pulgram quotes from was written in the 50s. In that he agrees that PIE "actually existed" (to use Rohini's own words). To quote from the same Pulgram paper Rohini quotes from: “there existed a Proto-Indo-European language whence the later attested Indo-European idioms derive their peculiar structure and much of their substance”.

In effect, Pulgram is in perfect agreement with me when I, to use Rohini’s words, “unqualified-ly quote PIE (Proto Indo-European) as the founding language of Sanskrit”.

Therefore, when Rohini says, “So when the jury is still out on PIE I would be very uncomfortable with adding gravitas to an article by quoting PIE (founding language) as a 'Fact'” she is not only disagreeing with me she is also disagreeing with her own citation, Pulgram.

In modern linguistics, PIE is very much a “fact”  with near "agreement on whether it actually existed" (agreement: it did). Pulgram agrees it was a "fact" in 1959 in the very same paper Rohini quotes from. And this fact has got stronger with age. Not only Pulgram, you would be hard put to find a modern linguist who would disagree with my statement that “The founding language of the family from which Sanskrit is from is called Proto-Indo-European”.


Now that I've cleared myself of all charges and have a free conscience, here's something extra.

The point that Pulgram was trying to make was somewhat different. He was questioning the method of reconstruction for PIE. 

Say your grandfather had a car which he sold before you were born. It was his favourite car and he spoke of it all the time. Today, though, all that's left of the car is a photo. Also, sadly, dadaji is dead in this tale.

Today if your wife asks you, "jaanam, what was your dada's car like that he bored us to death with?" what will you do? Look at the photo and tell her, right?

That's simple. 

But how good was the photo? Now if the camera was a B&W one you'd never be able to tell her the colour of the car. What if the photo ended above the wheels? You'd never be able to tell what the hubcaps were made off. If your little shit of a son cut out the back of the photo with a pair of scissors, you'd never be able to say whether it had a spoiler. And so on.

PIE is a bit like the car. It's lost since the people who spoke it were ganwaars and didn't write. But using a process of linguistic reconstruction (the camera), linguists have replicated PIE (the photo).

Like no one doubts the car existed, no one today doubts PIE existed. Pulgram doesn't and nor do I in my piece. And nor does any almost linguist today. Even the kooks like Elst agree that there was a PIE (he just says it came from India).

What Pulgram is saying in that paper is that we have a really bad camera which means we have a shitty photo. Hence Reconstructed Proto-Indo-European is a terrible replica of Real PIE. Just like the photo might be a terrible replica of the car.

Of course, PIE reconstructing is something that I don't even touch upon in my piece, so it's not relevant. I simply said the car (PIE) existed.

That said, Pulgram is also old. This paper we're quoting is 56 years old. By now almost everyone is convinced that the photo is a good one (say, some similar models still exist in the market and we've gone and checked them out at the showroom etc etc – this analogy is obviously being stretched now). 

Linguistic reconstruction has been so good that it has many, many times predicted old language features that have only later on been confirmed via archaeology. In effect, real world experiments have proved our camera works and have proven Pulgram wrong.


  1. Great discussion! Immensely enjoyed the whole thing. I am so much more informed after this whole episode. Thank you Shoaib Daniyal and Rohini Bakshi.
    BTW, was listening to the reconstructed PIE samples, sounds pretty surreal. Stuff of sci-fi, honestly. :)

  2. <<1) By the time the treaty was written, the Mitanni had stopped speaking Sanskrit. In fact, we know they had stopped speaking Sanskrit before 1500 BC (founding date of the empire) since they had switched to Hurrian.>>

    Are you sure they were speaking Sanskrit, rather Vedic Sanskrit? Also, you can never date the rig veda whether it was 1200 BC or 1800 BC or even prior to that