(I am grateful for the following learned and thought-provoking guest post I received from Christophe Vielle, dealing with an emic way to define “Philosophy” in the Indian context. EF)
Rājaśekhara’s classification of śāstras in the Kāvyamīmāṃsā, adhyāya 2 (śāstra-nirdeśa)*, is the following:
1. apauruṣeya = śruti (mantra-brāhmaṇa) + upavedas + vedāṅgas
2. pauruṣeya = purāṇa, ānvīkṣikī, mīmāṃsā, smṛti-tantra
2.1. purāṇa (+ itihāsa)
2.2.1. pūrvapakṣa : arhat, bhadanta, lokāyata
2.2.2. uttararapakṣa : sāṅkhya, nyāya, vaiśeṣika :
(ta ime ṣaṭ tarkāḥ | tatra ca tisraḥ kathā bhavanti vādo jalpo vitaṇḍā ca |
madhyasthayos tattvāvabodhāya vastutattvaparāmarśo vādaḥ |
vijigīṣoḥ svapakṣasiddhaye chalajātinigrahādiparigraho jalpaḥ |
svapakṣasyāparigrahitrī parapakṣasya dūṣayitrī vitaṇḍā |)
2.3. mīmāṃsā : (nigamavākyānāṃ nyāyaiḥ sahasreṇa vivektrī mīmāṃsā | sā ca dvividhā vidhivivecanī brahmanidarśanī ca |)
The matter is in fact a bit more complicated, with a discussion on the number of vidyāsthānas occuring between 2.4. and 2.2.1-2 (these vidyāsthānas are numbered 14 if 4 vedas + 6 vedāṅgas + 4 [types of pauruṣeya] śāstras are counted; different views are quoted), and the special place given here and there to the art of poetry. At the end of the chapter, there are interesting definitions of formal types of works: sūtra, in which the śāstras were originally composed, vṛtti, paddhati, bhāṣya, samīkṣā, ṭīkā, pañjikā, kārikā, vārtika and prakaraṇa.
Rājaśekhara’s distinction between mīmāṃsā and ānvīkṣikī as two different types of śāstras is noteworthy. For him (dated to the early 10th century) the former is of two kinds, corresponding to karma- (‘pūrva”) and brahma- (“uttara”) mīmāṃsā, the aim of both being to discriminate among/to interpret the Vedic vākyas by means of/on the basis of thousand of nyāyas. The latter corresponds to six tarkas “logics” (tarka can also be here equated to darśana), three belonging to the pūrvapakṣa: the Jaina (arhat), Bauddha (bhadanta) and materialistic ones; three to the uttarapakṣa: the Sāṅkhya, Nyāya and Vaiśeṣika ones. These 6 tarkas use 3 types of kathās “[argumentative] talks”: vāda (the sound argument for convincing mediators in disputes), jalpa (the attempt to defeat one’s rival by means of unsound arguments) and vitaṇḍā (destructive criticism, without substantiating one’s own point of view).(**)
I would be interested by textual parallels to, or scholarly comments on, such a classification.
(*) For the GOS edition, see: https://archive.org/details/kvyammsofrajasek00rjauoft
there is an input of the text on GRETIL, with a few typos:
(**) On these three kinds of kathā “debate“ (vāda “discussion“, jalpa “disputation”, vitaṇḍā “wrangling”) according to the NS (1.2.1-3) and their commentaries, see Esther A. Solomon, Indian Dialectics I, pp. 101-134: https://archive.org/details/SOLOMONIndianDialecticsI . Nadine Stchoupak and Louis Renou, La Kāvyamīmāṃsā de Rājaśekhara, traduite du sanskrit, Paris: Imprimerie Nationale, 1946 (Cahiers de la Société Asiatique, 8), translate here by “controverse”, “argutie” and “calomnie”.