Le 8 sept. 2024 à 01:34, Harry Spier <vasishtha.spier@gmail.com> a écrit :

Dear members,
Gerard Huet wrote:
. . .. Thus a serious Sanskrit grammar ought to use the Paninian notions, like Pr Filliozat’s « Grammaire Sanskrite Pâninéenne ».

What resources are there for an english speaker to learn Paninian grammar and how to use it.
 I'm aware of works on  Pāniṇian grammar  in english:
1) The Laghukaumudī of Varadarāja translated by James Ballantyne written in 1849
2) The Siddhānta-kaumudī translated by Chandra Vasu over 2300 pages.
3)  Pāniṇi, His Work and Its Traditions, vol. 1 background and introduction by George Cardona

Thanks,
Harry Spier

Dear Harry and Indology members.

First of all, let me clarify that what I call grammar is different from a Sanskrit manual, of which there are numerous excellent ones available in English.
Many of these manuals give you rudiments of the linguistic structure of the language, and discuss separately topics like phonology/phonetics, morphology,
syntax and semantics. They are intended as instruction material for students interested in reading and understanding Sanskrit text, or for general linguistics students
interested in comparative linguistics.

Pāṇini's grammar is on the other hand a completely differently organized description of the language, intended as a generative grammar, giving instructions on
how to construct Sanskrit utterances that are not only correct grammatically (by construction), but meaningful, and corresponding to the communication intention of the speaker.
One of its striking features is that the description is wholistic, and its operations mix the various linguistic levels. In a sense, this is necessary for Sanskrit, since the notion
of compound involves a mutual interplay between morphology and syntax. 
Panini’s grammar is utterly formal, it looks like a computer operation manual. What is hard to understand is the flow of control of instructions during sentence elaboration,
which is not fully specified, and must be guided by meta-rules (paribhāṣā) which specialists do not always agree on, as a recent controversy showed. 

This grammar is not used for teaching beginners, as far as I know, even in gurukulas the teacher uses generally the Siddhāntakaumudī, which is inspired from Pāṇini’s grammar
but with some duplications of instructions allowing a simpler control flow, or its lighter version Laghukaumudī. 
If you want to understand the basic principles of Panini’s grammar, it is better to ignore these simplified methods, and face the complexity of the full Aṣṭādhyāyī.
However, operating the grammar needs more than just the 3996 rules (sūtra) of the Aṣṭādhyāyī, you need auxiliary resources, first of all the roots recitation (Dhātupāṭha),
then the annex of primary nominal stems not easily obtainable by the sūtras (uṇādisūtrāṇi), and an annex of word classes admitting idiosyncratic treatment by certain sūtras (gaṇapāṭha). Other annexes give indicatiions on gender and accent. The historical development of the whole system is described in:
«  Pāṇini, a survey of research »  by George Cardona.
The main problem with these auxiliary materials is that, contrarily to Aṣṭādhyāyī for which there is a relative consensus on the text, available on various Web sites,
these texts have many versions, which may be more or less well adapted to the standard Aṣṭādhyāyī, and often appear in poor editions with many typos or inconsistent scripts.

The grammar can be understood only with copious commentaries. English ones have been listed here, like Saroja Bhate's Pāṇini, indeed an excellent introduction.
Another one is « Pa:ninian linguistics » by P. S. Subrahmanyam, Rashtriya Sanskrit Vidyapeetha, Tirupati, 2010 . 
A more complete commentary is « The Aṣṭādhyāyī of Pāṇini » by Pr Rama Nath Sharma, Munshiram Manoharia, 1987-2003, a splendidly edited work in 6 volumes.
To which I would add « Four vṛttis in Pāṇini », a very interesting short description (124 pages) of Pāṇini’s system by Pr Korada Subrahmanyam (self-published, 2002). 

Best
Gérard