Ancient Vedic and Classical Composers and Writers Were Extremely
Illiterate in Textbook Etymology
I have written a lot about textbookworm etymologists who raise objections about identifications on textbookworm grounds. Although I am not really interested any more in endless and pointless discussions and debates, I thought it perhaps appropriate to give one example to specifically illustrate the case.
A few days ago, someone wrote me the following mail. Who wrote it is not important since, as I said, I have no intentions of indulging in any more endless and pointless discussions and debates.
The mail was as follows:
“Respected sir,
Please
tell me where you got the name Kaoša in Avestan (which you suggest is recorded
in the Rigveda under the name Kavaṣa RV 7.18.12). Is it a proper name? I
haven't been able to find it today. Also the cognate of Kaoša in Indic is Kōṣa,
not Kavaṣa.
Thank
you”
About the first part, I pointed out that a simple question on google gave the following:
“Inthe Avesta (the primary
collection of sacred texts of Zoroastrianism), Kaosha is a minor
historical figure mentioned in the Frawardin Yasht (Hymn to the
Guardian Angels). [1, 2, 3]
Specific
details regarding the mentions of Kaosha include:
·
Lineage: Kaosha is
explicitly named as the father of a devout, righteous follower (an ashavan)
named Fraora-ostra.
·
Context of the Text: His
name appears in a long, recited list of ancestral and early heroes, priests,
and devoted adherents. In the Frawardin Yasht (Yasna 13, verse 123),
followers invoke and worship the guardian spirits (Fravashis) of these early
holy men, including the "Fravashi of the holy Fraora-ostra, the son of
Kaosha”.
About the second point, that “the cognate of Kaoša in Indic is Kōṣa, not Kavaṣa”, I tiredly wrote back: “As for what is the cognate of
what, I have long ago given up replying to textbookwormish challenges to
identifications”.
However, it would perhaps be right to be a bit more specific, since there is a whole host of names regarding which textbookworm objections have been raised claiming that such-and-such non-Sanskrit names cannot have been Sanskritized in certain ways because as per the textbook rules of comparative cognate etymology, the Sanskrit forms should have been something else. So let me deal with this specific point in this short one page article.
When a person takes a name from another language and pronounces it in his own language, he does not go by textbook rules of cognate etymology and decide that he must be politically or etymologically correct in his pronunciation so that the resultant form (in his own language) is exactly as per the textbook rules of what it should be as per rules of “cognate” forms. He simply takes the name as he hears it and pronounces it in his own language in his own way.
When Indians writing in the second century BCE took the Greek name Menander and made it Milinda, it is not because they were following some textbook rule that a Greek “men” was cognate to a Sanskrit/Pali “mil”, because I am sure it is not. They simply pronounced the Greek name in their Sanskrit/Pali writings in the way that the name sounded to them on entirely arbitrary grounds. Likewise, when Greek Alexander became Iskandar and later Sikandar, it was not because there were textbook rules of cognate etymology which made a Greek “Alex” into “Isk’ or “Sik” in any language. These changes took place on purely arbitrary human grounds.
Likewise when Sanskrit writers accepted Iranian names like Mihira (by itself as well as e.g. in names like Varahamihira) they did not reason that Miθra/Mihira was actually a cognate of Sanskrit Mitra and therefore convert the name Varahamihira into Varahamitra. They simply accepted the Iranian name and pronounced it, then and later, in their own new Sanskritized way.
Similarly, when the Vedic composers referred to the Anu-Iranian Kaoša in the Rigveda, they did not open out their textbooks of cognate etymology and decide that they should pronounce the name as Kōṣa: they simply relied on their ears. It is sad that these illiterate Vedic and Sanskrit composers and writers did not have modern textbookworms to guide them.
As I said: just putting in my word on this textbookworm disease. No intentions of participating in endless and pointless discussions and debates on this point.
APPENDIX ADDED 25 June 2026:
I wrote above “It is sad that these illiterate
Vedic and Sanskrit composers and writers did not have modern textbookworms to
guide them”. Unfortunately
I am more happily (?) placed. The person who wrote the above mail to me seems
to believe in the never-say-die maxim, and continues to repeat (in emails to
me) his textbook arguments in spite of my repeated reiterations of the logic
behind how people pronounce foreign names (i.e. names from another language).
For the record, I will just give below the different ways in which I tried to
make him understand this logic.
He wrote:
“But yes, you have to realize that the language that the Rigveda
is preserved in (Śakala) was likely not the language it was composed in. There
is a lot of internal phonological & metrical evidence that shows this &
actually justifies the 2500-3000 year gap between Vedic & Classical. The
sound changes I am talking about would've happened from Rigvedic -> Śakala
& would follow the natural laws of sound changes that have been established
for Indo-Aryan languages (with Rigvedic looking more similar to
Proto-Indo-Iranian). Therefore, the Menander point wouldn't hold in this case.
However, in order to help your side, you can argue that the original Indo-Iranian name was *Kawša (rendered in Śakala as Kaʋaṣa to fit the meter, as glides are perfectly valid in Rigvedic, ex: goagrāḥ). However, it would be difficult to explain the name Kavaṣa Ailūṣa, the author of a hymn in the 10th maṇḍala.”
I replied back:
“Apart from the fact that Mandala 10 follows long after the dasarajna battle, by which time the name would have been long established as a known one in India, you will note that Kavasha Ailusha is regarded as an outsider to the Vedic tradition (expressed of course in later myths in the Brahmanas, as being because he was born of a low-caste mother).
But all the verbal gymnastics trying to reconcile "natural laws of sound changes" into your arguments cannot explain why the Sanskrit references to Varahamihira do not refer to him as Varahamitra, recognizing mihira to be a cognate form of Sanskrit mitra. Unlike people obsessed with "cognate" forms, normal people simply accept the names of people speaking another language in the forms in which the names appear before them; they do not open out textbooks of cognate equivalence charts to see how those foreign names can be converted into "cognate" forms. As, moreover, there was no name "Kosha" among the Rigvedic people, why should they be expected to have found it necessary to convert the name Kaosha as it appeared before them into a "cognate" form instead of simply Sanskritizing it into "Kavasha"? Everything does not fall into rigid rules in real life, so I see no need whatsoever to avail of your "help" and "argue that the original Indo-Iranian name was *Kawša (rendered in Śakala as Kaʋaṣa to fit the meter, as glides are perfectly valid in Rigvedic, ex: goagrāḥ)" as I see no need for finding rules to explain every nuance.
[And why do you say "the Menander point wouldn't hold in this case". Doesn't this mean that it is senseless to expect to find rules to explain everything? Regardless of which stage of Sanskrit was involved, why should "Mena" become "Mili": doesn't it show that acceptance of foreign names is always based on arbitrary practice and not on immutable rules?]”
After this, he still writes:
“You are not at all getting the point. You've mentioned this multiple times before, and even I have.
Rigvedics
would've "Sanskritized" (I honestly wouldn't use the term
"Sanskrit" for Rigvedic, the language of the composers of the Rigveda
was closer to Proto-Indo-Iranian. RV just happened to get preserved in a later
language) the name Kaoša as *Kawša itself. This /aw/ later turns into /ō/ in
the Śakala recession (of the Brāhmaṇa-period), several centuries after
book 10 of the Rigveda. The question was why didn't it turn into Kōṣa in Śakala
(because we only have "Late Brāh? Makes a lot more sense if the name was
borrowed as *Kawaša just to fit the meter, rendered in Śakala as Kaʋaṣa.”
As I repeatedly pointed out that I was not going to go into endless and senseless discussions on this point, I did not bother to do so in reply to his mail. I am merely ending the “discussion” by putting the above dialogue on record.
APPENDIX ADDED 1 OCTOBER 2026:
Today, just on a whim, I put the following question on
Google:
“Why did Sanskrit borrow the Iranian name Mihira without
turning it back into the original Mitra?”
I was pleasantly surprised to see that the Google AI Overview answer has taken note of my above logic:
“Sanskrit borrowed the Iranian name Mihira instead of converting it back to its original native cognate, Mitra, because ancient and classical writers treated it as a distinct foreign proper noun associated with a specific, imported religious tradition rather than viewing it through the lens of modern etymological linguistics. [1, 2, 3]
When the worship
of the Sun god Mihr (from the Middle Persian evolution of Mithra)
entered India—principally via the Sakaladvipi or Maga
Brahmins who arrived from the Iranian world—they brought with them unique
iconographies and solar rituals distinct from traditional Vedic solar worship.
[1, 2]
Why Sanskrit
Writers Kept the Iranian Form
·
Absence of "Reverse
Etymology": As noted by historians and linguists like Shrikant
Talageri, ancient and classical writers simply relied on their ears. They did
not consult comparative grammar books to deduce that Middle Persian Mihr/Mihira was the
historical twin of Vedic Mitra. They
heard a distinct proper noun and preserved it as they received it. [1, 2, 3]
·
Preserving Specific Identity: Changing Mihira back to Mitra would have
stripped away the specific cultural identity of the imported deity. By keeping
the name Mihira, Sanskrit texts clearly denoted
the specific solar figure introduced by the Maga priests, differentiating it
from native Vedic concepts. [1]
·
Phonetic Adaptability: Instead of
reverse-translating the root, Sanskrit scholars used native phonetics to
naturally integrate the loanword. Over time, traditional Sanskrit grammarians
even generated folk etymologies or reverse-engineered derivations (such as in
the Uṇādi-sūtras) to fit the borrowed word Mihira directly
into Sanskrit root systems (e.g., deriving it from the root √mih, meaning
"to sprinkle water" or "cause rain"). [1, 2]
This seamless integration explains how the loanword became native enough to appear standard in famous compound names like Varāhamihira (the celebrated 6th-century astronomer) and Mihirakula (the Alchon Hun ruler).”
