First steps in Pāṇinian grammar (1)

Series dedicated to Prof. Ramesh Chandra Panda, BHU, Vārāṇasī

For concise introductory summaries, a historical survey is George Cardona’s chapter on Indian Linguistics in G. Lepschy’s History of Linguistics I: The Eastern traditions of linguistics [London, New York: Longman 1994, 25-60]. The same author also wrote a short article Old Indic grammar for volume 17,1 of the Handbücher der Sprach- und Kommunikationswissenschaft (HSK) [Booij/Lehmann/Mugdan (Eds.): Morphologie / Morphology. Berlin, New York: de Gruyter 2000, 41-51]. In volume 9,1 of HSK there is also an article by Peter Raster on Die indische Grammatiktradition [Jacobs/von Stechow/Sternfeld/Venneman (Eds.): Syntax. Berlin, New York: de Gruyter 1993, 199-208]. Paul Kiparsky wrote a comprehensive lexicon article especially on Pāṇinian linguistics for the 6th volume of Asher’s Encyclopedia of Language and Linguistics [Oxford (etc.): Pergamon Press 1994, 2918-23, offprint]. In the same book there is also an article on the author Pāṇini by Frits Staal to be found [2916-18].

For what’s more extensive, a comprehensive description of Pāṇini’s grammatical system is Cardona’s Pānini – his work and its traditions I: Background and Introduction [2nd edition (revised and enlarged), Delhi: Motilal Barnarsidass 1997]. The 1st volume of of Sharma’s edition [see below] is an Introduction to the Aṣṭādhyāyī as a grammatical device. Kiparsky also presented a linguistic paper generally On the architecture of Pāṇini’s grammar on the CIEFL conference on the Architecture of Grammar, Hyderabad 2002. Comprehensive and useful as introductions – if no paṇḍit is available – are the English translations of the Kaumudīs like Śrīśa Chandra Vasu’s voluminous translation of the Siddhāntakaumudī [reprint, Delhi: Motilal Banarsidass 1995]. A rich resource also is Kshitish Chandra Chatterji’s Technical terms and technique of Sanskrit grammar [Calcutta: Visvanath Chatterji 1948].

There are some reference works, Kashinath Vasudev Abhyankar compiled the Dictionary of Sanskrit grammar [Baroda: Oriental Institute 1961], and there is Louis Renou’s Terminologie grammaticale du Sanskrit [Paris: Champion 1942, and again 1957]. A book couldn’t be missed on the shelf is Sumitra Mangesh Katre’s Dictionary of Pāṇini [3 volumes, Poona: Deccan College 1968], but absolutely indispensable is his A glossary of grammatical elements and operations in the Aṣṭādhyāyī [Mysore: Central Institute of Indian Languages 1981]. A useful vademecum also just is the Appendix F: Grammatical concordance in Apte’s dictionary.

As a text book of the Aṣṭādhyāyī the Böhtlingk [originally: Pâṇini's acht Bücher grammatischer Regeln. Bonn: 1839-40. The 2nd edition (Pâṇini's Grammatik) appeared 1887, and is the base for several reprints. Recently with corrigenda: Delhi: Motilal Banarsidass 2001] is just fine for just working with the Sanskrit, but perfect if you could also consider the short German translations resp. paraphrases of the sūtras. It also contains useful indices and also a dhātupāṭha and a gaṇapāṭha. Joshi/Roodbergen’s massive edition, Aṣṭādhyāyī of Pāṇini [Delhi: Sahitya Akademi 1991-, so far up to volume 13/2007?] contains English paraphrases next to explanations including valuable cross-references and several application histories. The same is true for Rama Nath Sharma’s edition in 7 volumes [2nd edition (revised and enlarged), Delhi: Munshiram Manoharlal 2002]. Both editions I would say could be right even for the beginner of vyākaraṇa if concentrating on what’s essential at the beginning, at best with some guidance. Plain etexts of the Achtbüchrige are also available, an analytical version could be found here.

Indian grammar as history of literature is treated in English at length in volume 5,2 of the History of Indian Literature (HIL), it’s Hartmut Scharfe’s Grammatical literature [Harrassowitz: Wiesbaden 1977, reviewed by Cardona in the Indo-Iranian Journal 21 (1979), 117-35]. Other disciplines of Sanskrit literature should have such rich research surveys as they were written by Cardona on traditional grammar, they are ideal starting points for getting deeper into different issues and generally to see what actually is an issue in scholarship [Pāṇini - a survey of research. The Hague (etc.): Mouton 1976, and the update: Recent research in Pāṇinian studies. 2nd edition (revised), Delhi: Motilal Banarsidass 2004]. I will give reference to these on the occasion using the sigla PSR and RRP.

Śivasūtras and pratyāhāras

There is a set of sūtras called the Pratyāhārasūtras or the Śivasūtras [1], and they are going just like this: (1) aiuṇ (2) ṛḷk (3) eoṅ (4) aiauc (5) hayavaraṭ (6) laṇ (7) ñamaṅaṇanam (8) jhabhaj (9) ghaḍhadhaṣ (10) jabagaḍadaś (11) khaphachaṭhathacaṭatav (12) kapay (13) śaṣasar (14) hal.

It could be seen pretty quick that this a sound catalogue which is different than the one known from the western grammars of Sanskrit, that one is ordered strictly according to the sound classes and includes also the prolonged short vowels etc. Every one of the Śivasūtras ends with a consonant which is defined by the sūtra halantyam of the Aṣṭādhyāyī (1.3.3) as being called it. The sūtra ādir antyena sahetā (1.1.71) defines how to build a pratyāhāra, an abbreviation [2], from the Śivasūtras: any of the sounds here together with any it denotes a set including itself and all sounds which are in between (the other its are not included and the “a”s after consonants and semivowels are generally excluded from this) [3]. Thus ac denotes all vowels (includes the Śivasūtras 1-4), hal (the pratyāhāra, not Śivasūtra no. 14) denotes all consonants (includes the Śivasūtras 5-14) etc. Although 281 pratyāhāras are theoretically possible Pāṇini employs 41: al denotes all sounds (varṇāḥ), ak the simple vowels (samānākṣarāṇi: a, i, u, ṛ, ḷ) etc. [4]

The structure of the Śivasūtras is a fascinating subject. There are questions why two of the sūtras end on “ṇ”, which is open for ambiguities, why “h” appears twice etc. The Śivasūtras have become a field of Sanskrit computational linguistics and it has been found out that this arrangement of sounds is optimal, which is – like many things in vyākaraṇa – the work of a real genius [5].

Notes

[1] They are seen as originating from the drum of Śiva and it’s an interesting question, what this implies. It is striking that sound catalogue does not belong to the corpus of the Aṣṭādhyāyī and some authors see this as an evidence that they were not set up by Pāṇini himself [PSR III.1.3.4]. The designation as “śivāsutrāṇi” or “māheśvarasūtrāṇi” is found in later tradition, the classic grammarians refer to them as “akṣarasamāmnāya“ (for example in the Mahābhāṣya).

[2] On the term pratyāhāra which is not used by Pāṇini himself see Abhyankar’s dictionary, 248.

[3] There is a circle here: the sūtra halantyam defines it while ādirantyenasahetā uses this it to defines what is a pratyāhāra like hal which again is a component of halantyam. For mutually depended operations (anyonyāni kāryāṇi) are forbidden in the śāstra (this defect is called anyonyāśraya) this problem is tried to getting solved by the grammarians in understanding halantyam having different meanings (have fun! the key is Śs no. 14!).

[4] The “a” in Śs no. 6 is traditionally indicated as to be spoken nasal and therefore also being it by the sūtra upadeśe ‘janunāsika it (1.3.2). That marker is then used to build ra denoting “r” and “l”, although Pāṇini doesn’t employs that. In the tradition there are 44 pratyāhāras to be found, cf.  Sharma’s edition, II, 1 sq.

[5] Towards the architecture of the Śivasūtras [PSR 1.5.4c, RRP 2.1.4.4c] see for example Cardona: *Studies in the Indian grammarians. 1: The method of description reflected in the Śivasūtras [Philadelphia: American Philosophical Society 1969]; P. Kiparsky: Economy and the construction of the Śivasūtras [Deshpande/Bhate (Eds.): Pāṇinian studies (Festschrift S.D. Joshi). Ann Arbor: University of Michigan, Center for South and Southeast Asian Studies 1991, 239-61, offprint]; W. Petersen: A set-theoretical investigation of Pāṇini’s Śivasūtras [Proceedings of Mathematics of Language 8 (2003), 1-11], Mathematical analysis of Pāṇini’s Śīvasūtras [Journal of Logic, Language and Information 13 (2004), 471-89], On the construction of Śivasūtra-Alphabets [Kulkarni/Huet: Sanskrit Computational Linguistics (3rd ISCLS). Berlin, Heidelberg: Springer 2009, 78-97, offprint].