There’s a lengthy paper by the late Ole Holten Pind entitled Studies in the Pali Grammarians. The author goes through all the discussions of grammatical points in those commentaries whose attribution to Buddhaghosa is undisputed and in which the commentator alludes to akkharacintakā or other words denoting grammarians. In nearly every case Pind traces the grammatical point under discussion back to either Pāṇini’s Aṣṭādhyāyī or Patañjali’s Mahābhāṣya.
The paper is published in two parts in the 1989 and 1990 editions of the JPTS.
He says, “It is quite evident that Buddhaghosa did not know Sanskrit”, with the following footnote:
There are more than a score of instances which prove this point, but the following is especially interesting. The word for " whole, entire" vissa is extremely frequent in Vedic and Sanskrit ( = visva), but unknown in Pali (where sabba takes its place), except for one passage in the Dh. (266). Had Bdhgh known Sanskrit, he would have explained it as " sabba," but instead of that he takes it as *visra (musty), which (as a lexic. word) was current in late Pali, but does not fit the passage mentioned. — Among other errors B explains “stiffness” (swoon) by “calati” (see under chambhita and mucchancikata) ; in parajita he takes para as instr. of para ( = parena DhA 111.259); he connects Pali pineti with pinvati (DA 1157, cp. Vism 32 pinana), and he explains attamana as “saka-mana” (DA 1.255) equalling atta = atman.
Stede is basing his view on vocabulary, whereas Pind is looking at grammar, so perhaps that explains the difference?
Just coming back to that, I wonder what your thoughts are?
We seem to have, with regards the Pali commentaries:
reliance on Sanskrit grammars before any attested Pali grammars
wrong explanation of words that an elementary knowledge of Sanskrit would have fixed
systematic lack of understanding of details of Vedic rituals and customs in the suttas beyond the basics
This raises the question of authorship. It’s possible that, eg. Buddhaghosa knew Sanskrit but some of his sources did not, so certain examples slipped through by oversight.
It’s not unreasonable to suppose that an intellectual like him would have studied the Sanskrit grammars but not the rituals.
Sheldon Pollock talks about how across South (and Southeast) Asia, there was a general recognition that Sanskrit grammar was exceptionally beautiful, and studying Sanskrit grammar became ubiquitous among the elites, even those with no interest in Brahmanism per se.
The notion that Buddhaghosa might not have known Sanskrit is rather new to me, but fwiw here’s my current thinking on it.
For discussion’s sake, let’s provisionally suppose that Buddhaghosa really was a brahmin convert, trained in the six Vedaṅgas and therefore a competent Sanskritist.
So, following his conversion he’s ensconced in the Mahāvihāra translating the atthakathās from Sinhala into Pali. From time to time he meets with some etymological gloss that’s at odds with what he learned as a brahmin student of the Nirukta Vedaṅga. What then?
Would this discovery necessarily, or even probably, lead him to the judgment: “The atthakathācariyas have got it wrong.” ?
And suppose that it did, would such a judgment necessarily, or even probably, result in his wanting to correct the atthakathācariyas by drawing upon his Brahminical learning?
And suppose that it did, is it likely that the elders of the Mahāvihāra would have granted him a mandate to make such corrections?
If we can confidently answer all three questions with a yes, then arguments like the one made by Stede in his PED footnote will probably hold water.
On the other hand, if even one of these questions warrants a no, then Stede’s argument will flounder. It will fail to indicate anything at all about Buddhaghosa’s knowledge (or lack thereof) of Sanskrit.
I’ll give my own answers to these questions in a later post (probably tomorrow) In the meantime perhaps you’d like to post yours.
Edit: I’ll post them in brief now and in detail later:
Viśva is less frequent in later Sanskrit than in Vedic. Sarva is used in the same sense usually in classical Sanskrit, though viśva is also generally well understood although not used as much.
In virtually all or most of the younger (prose) Pali, the underlying language is Classical Sanskrit, not Vedic, so it follows that the vocabulary then current in classical sanskrit is used in prose Pali.
However Buddhaghosha’s reading of vissa as (the Pali form equivalent to) Saskrit ‘visra’ does not mean that he didnt know Sanskrit, rather it means that it is Pali that didnt make sense to him. Similarly the other examples of misinterpreted Pali words given by Stede show the same thing (Buddhaghosha misinterpreting Pali). If it was not his own misinterpretation, but that of one of his predecessors in the scholastic tradition that he was reusing, the same argument would hold for them instead of Buddhaghosha.
This misinterpretation of Pali vocabulary (due to misidentification of the underlying Sanskrit form) is not unique to Buddhaghosha, it is found throughout the early Buddhist scholarship. That is how there exists so many incorrectly reSanskritized words in Buddhist Hybrid Sanskrit - while such a thing is unknown in texts composed directly in classical Sanskrit. It means Pali wasnt generally that well understood by early Buddhists and they were therefore making guesses, some educated-guesses, and others not so educated.
In SN11.12 , some etymologically nonsensical explanations of some names of Śakra (i.e. Maghavan, Purandara, Śakra, Vāsava, Sahasrākṣa etc) are put by the creators of the canon in the mouth of the Buddha.
