¹ Pāṇini, , V.iv.50: kṛ-bhv-asti-yoge sampadyakartari cviḥ. In his La grammaire de Pāṇini (Paris, 1966)Google Scholar, Renou, L. translated the Kāṣikāversion of this Sūtra, which prefixes to it the word abhūtatadbhāve ‘becomingthat whichdid not exist’ (see Kātyāyana*x0027;s vārttika 1), by: ‘Le suffixe secondaire invariant “cvi” ( = v avec ton sur la finale; amui par VI. 1,67)est valable (après un mot indèterminè) pour signifier: le faitde devenir quelque chose qui n'ètait pas, (le nouvel ètat) obtenu ètant celui de 1'agent, quand il y a jonction (subsèquente) avec les racines kṛ- “faire” ḍukrñ VIII.10, bhū - “devenir” bhū I.I et as- “ être” ásá 11.56.’Google Scholar

See also Cāndravyākaraṇa IV.iv.35: abhūtatadbhāve kr-bhv-astiyoge vikārāc cviḥ//, with Vṛtti: avasthāvato ‘vasthāntareṇâbhūtasya tadātmanā bhāve kr-bhv-astibhiḥ sambandhe sail vikāravācinaś cvir bhavati/… aśuklaḥ śuklo bhavati śuklibhavati/… abhū tatadbhāva iti kirn/ ghaṭaṃ karoti, ghaṭam bhavati/… yat punaḥ prayatnenâpi śuklaṃ na sampadyate tatrâbhūtatadbhāvābhāvād na bhaviṣyati/

² See Dignāga, , Pramāṇasamuccaya i.2abGoogle Scholar: pratyakṣam anumānaṃ ca pramāṇe; Dharmakirti, , Pramānavārttika, Pratyakṣa chapter, k. 63 f.Google Scholar (pramāṇadvitva). Cf. Dharmakirti, , Ṅyāyabindu i.2–3: dvividhaṃsamyagjñānam/ pratyakṣam anumānaṃ ca/These two forms of pramāṇa correspond respectively to the svalakṣaṇa, the efficient (arthakriyāsamartha) and paramārthasat, and to the sāmānyalakṣaṇa and saṃvṛtisat (Pramāṇavārttika, Pratyakṣa chapter, k. 3).Google Scholar

³ In his Pramāṇavāritikālaṃkāra i.7 concerned with the Bhagavant's being (a means of) correct knowledge (prāmāṇya)—i.e. his cognitive ‘normality’—, Prajñākaragupta has called attention to the difficulty caused by the fact that, in terms of usage as such, pratyakṣa and anumāna have prāmāṇya, but not the Bhagavant (ed. Rāhula Sāṅkṛtyāyana, 32: bhagavatas tarhi kathaṃ prāmāṇyamj pratyakṣānumānayor hi vyavahāramātreṇa prāmāṇyam, na bhagavataḥ/ tad dhi param pramāṇamj; P, f. 32b = D, f. 27b reads: ‘o na ci Itar bcom Idan ’das tshad ma yin/ mṅon sum daṅ rjes su dpag pa dag ni tha sñad tsam du tshad ma yin gyi/ bcom Idan ‘das ni ma yin te/ de ni tshad ma dam pa yin pa'i phyir ro že na). And in his Bhāṣya i.4ab, Prajñākaragupta observed: dhiya eva prāmāṇyam nânyasya.

⁴ For the translation of bhūta by ‘like’ see below.

⁵ In his Ṭikā (D, f. 112a), Jayānanda (late 11th to early 12th century) explains the compound *pramāṇabhūta-puruṣa as follows: … tshad mar gyur pa yaṅ yin la/ skyes bu yaṅ yin pas na tshad mar gyur pa'i skyes bu stej ‘phags pa klu sgrub žabs la sogs pa'oj jde rnams kyis by as pa'i bstan bcos te/ dbu ma rtsaba*x0027;i šes rab la sogs pa'o//

Here dgoṅs pa (the Tib. honorific form for bsam pa) probably translates Skt. abhiprāya (on which cf. D. Seyfort Ruegg, JIP, 13, 1985, 309–25, and 16, 1988, 1–4). And for speaker's abhiprāya (=bsam pa, mṅon par ‘dod pa) in connexion with śabda and śābdajñāna according to Dharmakirti, see Pramāṇavārttika, Pramāṇasiddhi chapter, k. 1 and Svārthānumāna chapter, k. 213. Tib. bsam pa can also translate Skt. āśaya.

⁶ On āgama as a pramāṇa according to Candrakīrti, see his Prasannapadāi. 1 (ed. La, ValléePoussin, 75.7).Google Scholar

⁷ See sGra tshad pa Rin chen rnam rgyal, Bu ston rnam thar, ff. 6b and 8a (and cf. f. 32b). On theexpression tshad ma'i skyes bu, see below, § 12.3.

⁸ The Sanskrit of this verse of the Pramāṇasamuccaya is found in Vibhūticandra's notes on Manorathanandin's Vṛtti on the Pramanavarttika, ed. Sāṇkṛtyāyana, R., JBORS, 26, 1940, 518 (cf. p. 108)Google Scholar; and the first hemistich is quoted by Prajñākaragupta, Pramāṇavārttikālaṃkāra, 3, and by Yaśomitra, Abhidharmakośavyākhyā, 1. For Jinendrabuddhi's explanation of this verse in his Visālāmalavati, see below, § 10.In both Tibetan versions this verse reads:

tshad mar gyur pa 'gro la phan par bžed//

ston pa bde gśegs skyob la phyag 'tshal nas//

tshad ma (b) sgrub phyir ran gi gžun kun las//

btus te sna tshogs 'thor rnams 'dir gcig bya//

The epithet pramāṇabhūta has been translated as ‘who is thepersonification of the means of cognition’ by M. Hattori, Dignāga, On perception (Cambridge, Mass., 1968), 23.Google ScholarNagatomi, M. has rendered it by ‘embodiment of pramāṇa’ and ‘pramāṇa incarnate’ in M., Nagatomi et al. (ed.), Sanskrit and Indian studies [Essays in honour of Daniel H. H. Ingalls], (Dordrecht, 1980), 245.Google ScholarSteinkellner, E. then rendered it by ‘who has become a means of valid cognition’ in his article ‘The spiritual place ofthe epistemological tradition in Buddhism’, in Nanto bukkyô, 49, 1982, 7. And in his Der Buddha und seine Lehre in Dharmakirtis Pramdnavdrttika (Vienna, 1984)Google Scholar, Vetter, T. has translated it by ‘der Erkenntnismittel ist’. Cf. also Inami, M. and Tillemans, T., WZKS, 30, 1986, 124 f.Google Scholar(‘who has become a means of valid cognition’); Jackson, R., JIP, 16, 1988, 335Google Scholar(‘who has become authoritative’); Franco, E., JIP, 17, 1989, 84Google Scholar; van Bijlert, V., Epistemology andspiritual authority (Vienna, 1989), 115 (‘whois a means of valid cognition’)Google Scholar; Krasser, H., Dharmottaras kurze Untersuchung der Gültigkeit einer Erkenntnis, Teil 2 (Vienna, 1991). 29 (‘der zum Erkenntnismittel geworden ist’) [and now T. Tillemans, Persons of authority, Stuttgart, 1993_.Google Scholar

On the epithet tāyin = skyob (pa) and its etymology see below, n. 65. And for the Buddha's function of ‘protection’ (tāya = skyob pa), see P V, Pramāṇasiddhi chapter, k. 145 f. and k. 280.

⁹ For this double translation ‘correct/efficacious’, see Pramāṇavārttika, Pramāṇasiddhi chapter, k. 1. There pramāṇa is defined in terms of being non-disappointing, unfailing (non-defaulting and indefeasible) and congruent (avisaṃvādin = mi [bs]lu ba)—that is, as veridical in relation to a cognition and as non-delusive in relation to an object of cognition. And avisaṃvādana is described as being settled in the production of aneffect (arthakriyāsthiti = don byed nus par gnas pa), that is, efficaciousness. See below, n. 56. Dharmakirti's concept of cognitive correctness is thus linked with the pragmatic criterion of cognitive efficaciousness (rather than with the logical or legal, andhence formal, ones of validity). See also PV, Pramāṇasiddhi chapter, k. 4d-5a on prāmāṇya existing in virtue of vyavahāra or pragmatic transactional usage.

On avisaṃvādakam jñānam = samyagjñāna as bringing about the attainment of the apprehended object (gṛhitavastuprāpaṇa), see Dharmottara's, comment on Dharmakirti's Pramāṇaviniścaya, ed. Steinkellner, E. and Krasser, H., Dharmottaras Exkurs zur Definition gültiger Erkennthis im Pramāṇaviniścay (Vienna, 1989), 24–7 with 74–6Google Scholar, where the link with arthakriyā is made explicit; and Nyāyabindutṭiā (ed. , Malvania, Patna, 1955), 17Google Scholar: loke capūrvam upadarśitam artham prāpayan saṃvādaka ucyatej tadvaj jñānam api svayam pradarśitam artham prāpayat saṃvādakam ucyate. Cf. Durvekamiśra, , Dharmottarapradipa (ed. , Malvania), 17: avisaṃvādakaṃpravṛttiviṣayavastuprā-pakam samyagjñānam.Google Scholar

The word (a)visaṃvāda is found already with Dignāga (Pramāṇasamuccaya ii.5: āptavākyāvisaṃvādasāmānyād anumānatā), and also with Kumārila (ślokavārttika, Codanā°80: tasmād dṛḍhaṃ yadutpannam na visaṃvādam [?] ṛcchati/ jñānāntarena vijñānaṃ tat pramāṇam pratiyatām//). And the word saṃvāda is cited from Kumārila's Bṛhaṭṭ ikā (see Frauwallner, E., WZKSO, 6, 1962, 85–6Google Scholar; cf. Sāntarakṣita, Tattvasaṃgraha 2853).

In recent works avisaṃvādin has frequently been rendered as ‘reliable’ or ‘trustworthy’ (cf. V. van Bijlert, op. cit, 174, n. 12). But in Sāntarakṣita's Tattvasaṃgraha 2958 we read vaslusamvādah prāmānyam abhidhīyate ‘congruence with a real is termed prāmāṇya’; here the translation ‘trustworthy would hardly seem to be appropriate.

¹⁰ cf. recently V. van Bijlert, Epistemology and spiritual authority, 157 ff. (who on pp. 119–20 translates pramāṇabhūta by ‘who has become a means of valid knowledge’). In WZKS, 33, 1989, 180–1, E. Steinkellner has suggested distinguishing between ‘who is a pramāṇa’ (in Dignāga) and ‘who has become a pramāṇa’ (in Dharmakirti), rendering pramāṇa by ‘authority’.

The instruments of realization (sādhana, upāyd) in question in Pramāṇavārttika ii.7c that establish the Buddha as pramāṇa are cultivating compassion (karuṇābhyāsa) and engaging in teaching (śāstṛtva). See PV, Pramāṇasiddhi chapter, k. 34–131ab and 131cd–139ab with 145–6; also Prajñākaragupta, Pramāṇavārttikālaṃkāra, i.34 (p. 53); and Manorathanandin ad PV, Pramāṇasiddhi chapter, k. 36. and k. 284 (jagaddhitaiṣitvasya sugatatvaśāstṛtvatāyitvasahitasya prāmānyam āha—yato dayayā jagaddhitaiṣitvena śreya ācaṣṭe…).—Compare Dignāga's comment on Pramāṇasamuccaya i.l concerning the Buddha's hetusampad, viz. āśaya = jagaddhitaiṣitd (cf. PV ii.34–131ab) andprayoga = (jagacchāsanāc) chāstṛtvam (cf. PV ii. 131cd–139ab). (A comparable idea is to be found in different form already in the treatment of āptaprāmāṇya in Pakṣilaṣvāmin's Nyāyabāṣya, II.i.68.) See below, § 6.1 and § 17.

¹¹ See Manorathanandin: bhūtaśabdanirdeśo 'bhūtasya nityasya nivṛttyartham/ ‘nityaṃ pramāṇaṃ nâsti’ [ Pramāṇavarttika ii.8a]/… kasmāt punar nityam pramanam nâstij āha: vastuno ‘rthakriyākārinah saw ‘gateh’ [PV ii.8b] jñnanasya pramānyan nâsti nityam pramānam/ atrâiva kāranam āha: ‘jñeyasya’ vastuno rthakriyākāritvenânityatvat ‘tasyā vastusadgater’ api tajjanyāyā ‘adhrauvyād’ [PV ii.8cd] anityatvāt/; See also Jinendrabuddhi, , Viśālāmalavati, P, f. 2b(the corresponding Sanskrit fragment was discussed by Steinkellner, E., ‘Some Sanskrit fragments of Jinendrabuddhi's Viśālāmalavatt’, in A corpus of Indian studies [Essays in honour of Professor Gaurinath Sastri, Calcutta, 1980], 100) quoted below, n. 39.Google Scholar

In his sTon pa tshad ma'i skyes bur sgrub pa'i gtam, f. 8b, A lḁg ša Ṅag dbaṅ bstan dar (b. 1759) has also referred to this interpretation according to which bhūta excludes the idea of a permanent and spontaneously existing pramana, and he mentions an interesting difficulty that it entails.

¹² cf. Prajñākaragupta, , Pramānavārttikālaṃkāra i.7, quoted above in n. 3.Google Scholar

¹³ Edgerton, F., BHSD, listed the meanings ‘authority, evidenced and M., Hattori gave ‘authoritative’ or ‘standard’ (op. cit., 74).Google Scholar—Burrow, T., ‘Sanskrit MA- “ to ascertain”’, TPS, 1980, 135–40, has however distinguished between the roots mā-’ measure’ and mā -’ ascertain’, etc.Google Scholar

As for sākśibhuta, in his Abhisamayālaṃkārālokā ii (ed. , Wogihara, 207), Haribhadra following Candragomin has explained it as equivalent to sākṣin.Google Scholar

¹⁴ The Mahāyānasūtralamkārabhāsya is generally attributed in the Tibetan tradition to Vasubandhu, but Arya-Vimuktisena lias ascribed what appears to be a passagefrom this work (xii.9) to Asahga in his Abhisamayalamkaravrtti (ed. C., Pensa, Rome, 1967), 113–15.Google Scholar

Attention may be drawn to the distinction clearly made here in the Malāyānasūtrālamkarabhāsya between the form pramāna-bhūta as anepithet describing the very nature of the Buddha Bhagavant and the cvi-formation pramanl-krta, an epithet applicable to masters made standards and means of right knowledge (pramūna) by the Bhagavant (e.g. in a vyūkarana). This morphological difference is discussed below.

As the source for the doctrine of the ābhiprayika and the nitārtha/neyārtha in the context of the four pratisaraṇas, Sthiramati (f. 96a-b) refers to the Akṣayamatinirdeśasutra (cited in part by Candrakirti, Prasannapadā, ed. La Vallèe, Poussin, 43Google Scholar). Cf. Ruegg, D. Seyfort, ‘Purport, implicature and presupposition: Sanskrit abhiprāya and Tibetan dgonspa/dgoris gži as hermeneutical concepts’, JIP, 13,1985, 309–25Google Scholar; and ‘Allusiveness and obliqueness in Buddhisttexts’, in: C., Caillat (ed.), Dialectes dansles litteratures indo-aryennes (Paris, 1989), 295–328.Google Scholar

¹⁵ Nyāyābhaṣya, II.i.68. In NBh I.i.7 it is explained that the āpta, who is sākṣātkrtadhaṛman, then becomes an upadeṣṭṛ in virtue of his wish to expound. And expertness is there defined in terms of direct perception (sāksātkaraṇam arthasyāptih, tayā pravartata ity āptah).

The bahuvrīhi compound sāksātkrtadharman could also be rendered by ‘having direct perception for [their] nature/quality’, an interpretation that would fit very well the contexts discussed in this article in which this expression and the expression pratyaksadharman have beenused. However, Indian commentators seem to have actually understood ḋharma (n) as the object of thedirect perception in question. See e.g. Uddyotakara, Nyayavārttika, II.i.68: ‘sākṣātkṛtadharmata’ yaṃ te padārtham upadiśanti sa taiḥ sākṣātkṛto bhavatîti; and Vācaspatimiśra, Nyāyāvārttikatātparyaṭīkā, I.i.7: sudṛḍhena pramānenâvadhāritaḥ sākṣātkṛta dharmāh padārthah (and the same commentary on II.i.68: pratyakṣisikṛtāheyopādeyatā). See also below, n. 17 (on sākṣātkṛtadharman in the Nirukta) and n. 19 (on pratyaksadharman). Compare e.g. the expression sākṣātkṛtatattva ‘having directly peceived reality’ in Manorathanandin's Vṛtti on PV, Pramāṇasiddhi chapter, k. 147 (p. 54).

The parallelism of the set of four factors constituted by hātavya, etc. in NBh, II.i.68—in other words the four arthapadas of NBh, I.i. 1 (heya, nirvartaka, hānam ātyantikam and upāya)—with the four factors enumerated in the Yogaśutra (ii.15 f.)—heya = duhkḥa/saṃsāra, heyahetu = pradhānapuruṣayor saṃyogah, hāna = kaivalya/saṃyogasyâtyantikī nivṛttiḥ, and hānopāya = vivekakhyātir aviplavā/ samyagdarśana—and with the four ‘principles of the Nobles’ (āryasatya) in Buddhism beginning with duḥkha is to be noted. (Cf. Wezler, A., ‘On the quadruple division of the Yogasastra, the Caturvyuhatva of the Cikitsasastra and the “Four Noble Truths” of the Buddha’. IT, 12, 1984, 289–337, who seeks to demonstrate that the four-fold division in question originated with the Buddha and then passed on the one side to the medical Sastra and on the other to the Yoga system whence it was adopted by the Nyāya.)Google Scholar

On āptopadeśa, see Oberhammer, G., ‘Die überlieferungsautorität im Hinduismus’. in his Offenbarung, geistige Realität des Menschen (Vienna, 1974), 49 f.Google Scholar (Oberhammer, 50, has rendered sākṣātkṛtadharmatā by ‘das den Dharma direkt Erkannthaben’, explaining his interpretation— ‘Dharma als Gegenstand der autoritativen überlieferung’—in his n. 29.) Cf. Chemparathy, G., L'autorite des Veda selon le Nyaya-Vaiśeṣika (Louvain, 1983), 19ff.Google Scholar

¹⁶ It is to be observed that according to Uddyotakara, Nyāyavārttika, I.i.7 (p. 174.6), the compound āptopadeśa is interpretable as a karmadhāraya (āptaś câsāv upadeśaś ca ‘instruction that is reliable’), not as a tatpuruṣa (āptasyôpadeśaḥ ‘instruction of a reliable person’) according to some (i.e. the Mimāṃsaka, who maintains the impersonality, apauruṣeyatva, of the Veda).

¹⁷ See also the commentary by Skanda, /Maheśvara, (ed. Sarup, L., Fragments of the commentaries of Skandasvāmin and Maheśvara [ch. i], Lahore, 1928), 114.Google Scholar

Concerning the interpretation of the bahuvrīhi compound sākṣātkṛtadharman, Durga remarks sākṣātkṛto yair dharmaḥ sākṣād dṛstaḥ prativśiṣṭena tapasā ta ime sākṣātkrtadharmānah, thus taking dharma(n) as the direct object of direct perception (cf. n. 15 above).—In his Jâska's Nirukta (Göttingen, 1852), p. xiii, R. Roth paraphrased this passage by the words ‘Die Weisen der Vorzeit, welche um Recht [dharma] zu tun selbst keiner Anweisung bedurft hatten’. PW lists sākṣātkṛtadharman on p. 892b but does not translate it; Monier Williams in his Dictionary gives ‘one who has an intuitive perception of duty’. And in his translation of the Nirukta (The Nighantu and the Nirukta, 20), L. Sarup rendered the compound sāksātkrtadharman by ‘had direct intuitive insight into duty’. But it is not established that dharma(n) here has the meaning of duty or virtue (cf. n. 15 above).

¹⁸ On bhūtavādin as an epithet of the tathāgata, see below, § 13.

¹⁹ Or: ‘whose nature is direct perception’? Bhartrṛari's Mahābhāṣyadipikā, ed. Bronkhorst, (Pune, 1987), 31Google Scholar, reads: ye ‘pratyakadṣaharmāṇaḥ’ <a?>lokadharmā ye parokṣā lokasya te pratyakṣās teṣām. But Abhyankar and Limaye in their edition (Pune, 1970), 38 read: dharmā ye parokṣā lokasya te pratyakṣds teṣām. According to Kaiyaṭa‘s Mahābhāṣyapradipa, the mode of knowledge in which ‘all’ is known by these Ṛṣis is direct perception born of Yoga (‘pratyaksadharmāṇa’ iti, yogajapratyakṣeṇa sarvam viditavantaḥ). (But this commentator's gloss does not make it altogether clear whether he took dharma(n) as the object of direct perception, or whether pratyakṣa is to be understood as the nature or quality (dharman-) of those who ‘know all’ in virtue of possessing such direct and immediate knowledge.)—Monier Williams gave ‘keeping in view the merits (of men)’, but this hardly fits the context.

²⁰ Kaiyata explains this as viāyavidyāvibhāgajñāh. Chatterji, K. C., Patanjali's Mahabhashya (Calcutta, 1964), 92Google Scholar, suggests emending parāparajñāh to parāvarajnāñh.

²¹ When explaining this epithet in his Pradīpoddyota, Nāgeśa speaks of the three stages of śravaṇa, manana and nididhyādsna.

²² Nāgeśa explains adhigata in terms of sākṣātkāra.

²³ The words yarvānas and tarvānas are stated by Patañjali to derive by mispronunciation from yad and tad. On the derivation from yadvānas and tadvānas, see Wackernagel, J., Ahindische Grammatik, i, § 189c) β) (p. 212)Google Scholar; and Bloomfield, M. and Edgerton, F., Vedic variants, ii (Philadelphia, 1932), § 272a (p. 142).Google Scholar And on this passage of Patañjali, see Chatterji, K. C., Patanjali's Mahabhashya, 92 f.Google Scholar A number of attempts at explaining the two words have been reviewed by Raja, K. Kunjunni, ABORI, 68, 1987, 537–9.Google Scholar

²⁴ Ed. F. Kielhorn, I, 39.10: pramấnabhūta ācāryo darbhapavitrapāṇiḥ śucāv avakāśe prāṅmukha upaviśya mahatā prayatnena sūtrāṇi praṇayati sma.

²⁵ For the meaning ‘like a pramāna’, see Nāgeśa's comment in his Pradipoddyta on the form sāmānyabhūta in the Paspaśāhnika of the Mahābhāṣya (ed. Kielhorn, i, 1): vṛddhisaṃjñasūtrasthabhāsyaprayogasya ‘pramānabhūta’ ity asya svayam karisyamāṇavyakhyānarityâsyâpi vyākhyānasambhavāt/ pitrbhūta ity atrâpi anyatrânyaśabdaprayogaḥ sādṛśyapara itisādṛśyapratitir, na tv asya sādṛśyavācakatve mānam astîti dik/ As noted at the the beginning of this paper, following Pāṇini, v.iv.50 the form to be expected could be pramāṇtbhūta rather than pramāṇabhūta. But in his Pradipoddyota on vārttika 2 to this sūtra (prakṛtivivakṣāgrahanaṃ ca) Nāgeśa has noted a restriction, explaining that when the meaning of likeness (sādṛśya) is conveyed by the compound ending in °bhūta the grammatical operation in question does not take effect. Paribhāṣā 15 in Nāgeśa's Paribhāṣenduśekhara specifies moreover that a grammatical operation is realized when a word has its primary meaning, but not when the word has a secondary meaning: gaunamukhyayor mukhye kāryasaṃpralyayaḥ. A form quoted in this context is mahadbhūtaś [instead of *mahābhūtaś] candramāḥ ‘the moon become as it were great’. (For this last reference I am indebted to J. Bronkhorst.) On further commentarial discussions, see below § 8.3.

Following many commentators who understand the expression pramāṇabhūta as meaning prāmāṇyam prāptaḥ (see below, § 7.3), Filliozat, P.-S. has translated by ‘le Maître qui possède l'autoritè’ in his Le Mahābhāṣya de Patañjali, Part I (Pondichéry, 1975), 376Google Scholar; but in Part v (Pondichéry, 1986), p. vii, he has adopted the meaning ‘qui est devenu, est transformé en’ for bhūta. Earlier, in L'Inde classique, i (Paris, 1949) 96Google Scholar, L. Renou had rendered Patañjali's expression by ’ le maitre-fait-norme’.

The term pramāṇabhūta is found also in an old commentary on the Sāṃkhyakārikās, the Yuktidipikā {c. 600?) 4, where we read (ed. Pandeya, 31): āptavacanaṃ tu pramānabhūtadvārako ‘tyantaparokṣe ‘rthe niścayaḥ. See also Pārthasārathimiśra's Nyāyaratnākara on Ślokavārttika, Codanāsūtra 64, on the pramāṇabhūtapauruseyavacas.

²⁶ See lastly Mahābhāṣyadïpikā of Bhartrhari, Ahnika, 3, ed. Palsule, G. B., Pune, 1985.Google Scholar

²⁷ On the dependence of Dignāga's Traikālyaparikṣā on part of the Sambandhasamuddeśa of Bhartṛhari's Vākyapadiya (Trikndl), see Frauwallner, E., WZKSO, 3, 1959, 145–52.Google Scholar For Bhartṛhari's influence on Dignāga in general, including in grammatical matters, see Hattori, M., ‘The Sautrāntika background of the apoha theory’, in: Buddhist thought and Asian civilization: Essays in honor of Herbert V. Guenther (Emeryville, 1977), 47–58Google Scholar, and ‘Apoha and pratibhā’, in M. Nagatomi et al. (ed.), Sanskrit and Indian studies, 61–73; and Hayes, R., ‘Jinendrabuddhi’, JAOS, 103, 1983, 714Google Scholar; Dignāga on the interpretation of signs (Dordrecht, 1988) (on pp. 30 f. Hayes discusses R. Herzberger's Bhartrhari and the Buddhists, Dordrecht, 1986)

²⁸ Mahābhāṣyadipikā on the Paspaśāhnika, ed. Bronkhorst, J. (Pune, 1987), 3Google Scholar: ‘sāmānyabhūtam ‘iti/ bhūtaśabda upamāvāci/tatra yeṣām arthāntarabhūtā jātiḥ teṣāṃ viśiṣṭeṣu gopiṇḍeṣv anuvartamānam gotvam teṣām sāmānyaṃ, nâśvadinām sarvesām iti tadapeksa upāmdsambandhaṃ/ For a discussion of this difficult passage, see Bronkhorst's note on p. 111 and his translation on p. 43.

²⁹ Concerning the use upamārthe of °bhūta, L. Renou tentatively dated it from the time of Kālidāsa in his Grammaire sanskrite (Paris, 1961), 113. Renou further noted its use as a copula to express the predicate and, especially ‘en bouddhique’, its pleonastic use. Cf. Speijer, J. S., Sanskrit syntax (Leiden, 1886), 154.Google Scholar

³⁰ Pradipa: prāmānyam prāpta ity arthahj bhū prāptāv ityasyâdhṛṣād vêti ṇijabhāve rūpamj vṛttiviṣaye ca pramāṇaśabdaḥ prāmānye vartate/

Concerning pramāna and āpta-prāmānya ‘trustworthiness/authority of the expert/fit’, cf. Nyāyabhāsya, II.i.68 with I.i.7 discussed above, § 6.1.

³¹ Pradipoddyota: nanu bhavater janmārthatvenâbhūtatadbhāvapratityā cvau sati pramāṇibhūta iti syāt, tadavivakṣāyāṃ tu pramāṇam ācāryaḥ prakārāntareṇa bhūta ity arthaḥ syād ata āha ‘prāmāṇyam’ iti/ ‘vṛttiviṣaye’ iti prakṛtābhiprāyam/—Compare also the discussions in Rāmacandra Sarasvati's Vivaraṇa, Annambhaṭṭa's Uddyotana and Pravartakopādhyaya's Mahābhāṣyapradipaprakāsa (ed. Narasimhacharya, M. S., Mahabhāṣya-Pradipa-Vyākhyānāni, I, Pondichery, 1973)Google Scholar and in the Mahābhāṣyapradipaprakāśa (ed. Narasimhacharya, M. S., Pondichéry, 1986), 56.Google Scholar

³² Contrast what was said in the Nyāyabhāṣya, II.i.68 on the āptas as well as the āptopadeśa being pramāṇa (see above § 6.1).

³³ See the Mahābhāṣya-Pradlpa-Vyākhyānāni, ed. Narasimhacharya, M. S., I, 230: ⃛ yathā pramāṇasya pramāṇāntaranairapekṣyeṇârthagrāhakatvaṃ tathā pratyakṣādipramāṇanairapekṣyeṇa tapomahimnaiva sarvān sādhuśabdān sākṣātkṛtavān iti tasyâpi prāmāṇyasambhayāt/ ata evôktaṃ bhagavatā paspaśāhnike ‘pratyakṣadharmāṇa’ iti/Google Scholar—In this publication, p. xiii, śivarāmendra has been dated to the second half of the seventeenth century by P.-S. Filliozat.

³⁴ See above, n. 25, for Nāgeśa's comparison of the forms pramāṇabhūta and pitṛbhūta when commenting in his Pradipoddyota on the word sāmānyabhūta.

³⁵ This is according to Pāṇini's rule V.iv.50.

³⁶ See the Nārāyaṇiya in the Mahābhāṣya-Pradipa-Vyākhyānāni (ed. Narasimhacharya, M. S., I), 232. On these commentators, see the editor's Introduction in vol. X of the series (Pondichéry, 1983).Google Scholar

For further discussions on the use or non-use of the cvi’—suffix taught in Pāṇini, V.iv.50, see for example the commentaries on Pāṇini I.iv.74; II.i.59 (concerning the form śreṇikṛta); III.i. 12 vārttika 4; III. 1.92 vārttika 2; III.ii.56. I am indebted to K. Bhattacharya for the opportunity of discussing some of these passages with him.

³⁷ ed. L., Sarup, Commentary of Skandasvāmin and Maheśvara on the Nirukta, Chapters II-VI (Lahore, 1931), 173.Google Scholar

³⁸ Maheśvara, (ed. Sarup, L., 11): ‘sattvabhūtam’: sattvaṃ dravyam, bhūtaśabdaḥ pitṛbhūta ityādivad upamāyāṃ draṣṭavyaḥ.Google Scholar

³⁹ In the Tibetan translation of Jinendrabuddhi's Viśālāmalavati Pramāṇasamuccayaṭikā, the whole relevant passage reads (P, f. 2b): tshad ma dan 'dra bas bcom Idan 'das tshad ma'o/ /gaṅ gi phyir ji Itar mṇon sum la sogs pa'i tshad ma ni skyes bu'i don la ñe bar mkho ba sṅar ma rtogs pa'i gsal bar byed ciṅ mi bslu ba yin pa de Itar/ bcom Idan 'das kyaṅ gaṅ la skyes bu'i don gyi mchog rag las pa 'phags pa'i bden pa bži'i mtshan ñid kyi de kho na ñid la de'i yul can gyi šes pa bskyed nas/ ṅes par legs pa don du gñer ba rnams la/ khoṅ du ma chud pa de gsal bar byed pa daṅ mi bslu ba de'i phyir tshad ma daṅ chos mtshuṅs pa ñid kyi phyir tshad ma'o/ /gyur pa ni skyes pa ste byuṅ ba ces pa'i don to/ /gyur pa'i tshig ni ma byuṅ ba rtag pa dbaṅ phyug la sogs pa'i tshad ma gžan gyis yoṅs su brtags pa dgag pa'i don du'o/ /tshad ma yaṅ 'di yin gyur pa'aṅ yin pas tshad mar gyur pa ste/ tshad mar gyur pa de la'o//

For the corresponding Sanskrit fragment from the Viśālāmalavati found in Vibhūticandra's notes, see their edition by Rāhula, Sāñkṛtyāyana, JBORS, 26, 1940, 518–19Google Scholar, and by Steinkellner, E., in: A corpus of Indian studies, 100. The Sanskrit of Jinendrabuddhi's further comment giving the alternative explanation of pramāṇabhūta (in which bhūta = anitya) reads: bhūta utpannaḥ/ bhūtava-canam aprajātasyêśvarāaeḥ paraparikalpitanityasya [?] pratiṣedhārtham ivârthas tu sāmarthyagata iti na tadartham iti vakṣyate/ pramāṇaṃ câsau bhūtaś cêti pramāṇabhūtaḥ/ tasmai pramāṇabhūtāya praṇamyêti yojanam/Google ScholarSee also Steinkellner, E., WZKS, 33, 1989, 180–1.Google Scholar

On the question whether the author of the Viśālāmalavati is identical with Jinendrabuddhi, the grammarian author of the Nyāsa commentary on the Kāśikā, see Hayes, R., JAOS, 103, 1983, 709–17.Google Scholar

⁴⁰ Tib. (D, f. 27b): bcom loan 'das ni mṅon sum gyi raṇ bžin gyis tshad ma yin te.

Yamāri in his voluminous commentary on Prajñākaragupta's work, the Pramāna-vārttikālaṃkāraṭikā Supariśuddhā, has also mentioned similarity with pramāṇa (D, phe, f. 203b6: tshad mar chos mtshuṅs pa). But in his detailed exegesis in which he has some very interesting things to say, he has also introduced a further line of explanation. Although the entire passage (D, f. 202a ff; cf. also f. 188b) needs to be considered in detail, this is not possible in the frame of the present article. Suffice it to say here that in his analysis of the proposition tshad ma yaṅ bcom Idan 'das ñid Yamāri has discussed the suitability of employing pramāabhuta as an epithet of the Bhagavant and the applicability to the Bhagavant, as attribute-possessor (viśeṣya), of the attribute (viśeṣaṇa) of being pramāṇa. This Yamāri has done (f. 203a) with reference to two fundamental syntactic-semantic principles: (i) predication by the exclusion of connexion of the attribute (here pramāṇa) with any possessor of this attribute other than the one specified (here the Bhagavant: gžan dan Idan pa mam par gcod pa=anyayogavyavaccheda, in which case the particle eva [if used] follows the possessor of the attribute, as in e.g. pārtha eva dhanurdharaḥ where it is asserted not only that Arjuna is a bowman but also that he alone is the perfect [phul du byuṇ ba] bowman); and (ii) predication by the exclusion of non-connexion between the attribute and the possessor of the attribute (ma 'brel pa rnam par gcod pa = ayogavyavaccheda, where the particle eva [if used] follows the attribute possessed, as in e.g. sa paṇḍita eva where it is asserted that a person is learned indeed, but without asserting that this person alone is a learned person). It is explained that it is predication by ayogavyavaccheda that applies in the present case, and that what is in question is having the character of pramāṇa (tshad ma'i mtshan ñid). On the contrary, in the case of pratyakṣa—i.e. of a pramāṇa in the strongest and strictest sense—what is in question is perfect pramāṇa; and it is then predication by anyayogavyavaccheda that applies. Reference is further made to the content (brjod bya = abhidheyd) of the Pramāṇa-śāstra being indeed directly (dṇos su) pramāṇa as such, but dominantly and preeminently (gtsor) the Bhagavant.

⁴¹ 'khruṅs pa. Cf. Steinkellner, E., in Steinkellner, and H., Tauscher (ed.), Contributions on Tibetan and Buddhist religion and philosophy (Vienna, 1983), 276.Google Scholar And compare bhūta utpannah: gyur pa ni skyes pa ste byuṅ ba in Jinendrabuddhi's Viśālāmalavati.

⁴² On this, in addition to v. Bijlert, op. cit., 150 ff. see most recently Katsura, S., ‘Dharmakirti's theory of truth, JIP, 12, 1984, 215–35;Google Scholar G. Dreyfus, ‘Dharmakirti's definition of prāmana and its interpreters’, in: Steinkellner, E. (ed.), Studies in the Buddhist epistemological tradition, 19–38;Google ScholarFranco, E., ‘The disjunction in Pramāṇavārttika, Pramāṇasiddhi chapter verse 5c’Google Scholar, ibid., 39–51; Lindtner, C., ‘The initial verses of the Pramāṇasiddhi chapter in the Pramāṇavārttika’Google Scholar, ibid., 155–9.

⁴³ cf. Bhattacharya, K., Dr. B. M. Barua birth centenary commemoration volume (Calcutta, 1989), 67, 69.Google Scholar

⁴⁴ Edgerton, F., BHSD, has rendered caityabhūta in the Vajracchedikā by ‘of the nature of an object of veneration’. Conze's rendering ‘a true shrine’ (Glossary, 105) does not follow the Tibetan translators’ interpretation of °bhūta, and it is somewhat ambiguous (though it has been retained by G. Schopen in his recent translation of the Gilgit Manuscript inGoogle Scholar: Gómez, L. O. and Silk, J. A., The Great Vehicle: three Mahāyāna Buddhist texts, Ann Arbor, 1989, 124).Google Scholar On this passage, cf. Schopen, G., ‘The phrase “sa prthivipradeśaś caityabhūto bhavet” in the Vajracchedikā notes on the cult of the book in Mahāyāna’, in IIJ, 17, 1975, 147–81.Google Scholar

The same expression is attested in Aṣṭasāhasrikā prajñāpāramitā iii (ed. Rājendralāla, Mitra), 56–8 (where caityabhūta is found in the proximity of āśrayabhūta [p. 58]).—But in Haribhadra's Abhisamayālaṃkārālokā iii (ed. Wogihara, , 207) caityabhūta has been explained following Candragomin as meaning ‘in nature a caitya’, in the same way that sākṣibhūta is stated to have the meaning of sākṣin (sākṣy eva sākṣibhūta iti tatsvabhāyatve caityabhūta iti candragomi). The Tibetan translation of Haribhadra's work (ascribed to rṄog Bio ldan šes rab) accordingly has mchod rten ñid la mchod rten du gyur pa (which is comparable with Ye šes sde's translation mchod rten du gyur pa in Vajracchedikā, Section 12, but not with mchod rten lta bur in Section 15c).Google Scholar

⁴⁵ On the other hand, in e.g. Nyāyabindu iii.30 where the compound expression avidyāmānasarvajñạtāptatāliṅgabhūtapramānātiśayaśāsanatva is found, and where liṅgabhūtaḥ, pramāṇātiśayaḥ is explained as liṅgātmakah pramāṇaviśesaḥ by Dharmottara, the meaning ‘like’ would clearly not be appropriate for °bhūta. This meaning ātmaka is of course also frequent for °bhūta.

⁴⁶ Aggavaṃsa (Saddaniti, 554) adds the explanation: bhāveti bhāvayati pabhāveti pabhāvayati, itthambhūto ‘cakkhubhūto ñāṇbhūto… brahmabhūto’. (Here pabhāvayati curiously recalls prabhāvanā in the above-mentioned passage of the Aṣṭasāhasrikā, 58, where caityabhūta is attested.) Itthambhūta means according to Aggavaṃsa imaṃ pakāraṃ bhūto patto.—Cf. K. Bhattacharya, loc. cit.Google Scholar

⁴⁷ Compare e.g. ’Jam dbyaṅs [phyogs/]mchog lha ’od zer (1429–1500), Tshad ma mam ‘grel gyi bsdus gžuṅ šes by a ba'i sgo ‘byed rGol ṅan glaṅ po joms pa gdoṅ lṅa'i gad rgyaṅs rgyu rig lde mig (Rva stod bsdus grva), f. 187a; A kya yoṅs ‘dzin dByaṅs can dGa’ ba'i bio gros (c. 1800), Blo rigs kyi sdom tshig blaṅ dor gsal ba'i me loṅ, f. 1b; and Yoṅs ‘dzin Phur bu lcog pa Byams pa Tshul khrims rgya mtsho (1825–1901), Tshad ma'i gžuṅ don ‘byedpa'i bsdus grva'i rnam bžag rigs lam ‘phrul gyi Ide mig ces by a ba las Rigs lam che ba yul yul can daṅ bio rig gi rnam par bšad pa, f. 12b.

⁴⁸ cf. Pramāṇavārttika, Svārthānumānā chapter, k. 215–16, and Parārthānumāna chapter, k. 108. For the testability–or, to keep the established metaphor, the ‘assayability’—of the Buddha's teaching, see śāntarakṣita, Tattvasaṃgraha 3587:

tāpāc chedāc ca nikaṣāt suvarṇam iva paṇḍitaiḥ/

parikṣya bhikṣavo grāhyaṃ madvaco na tu gauravāt//

See also 3344 and Kamalaśila, Pañjikā 3.

⁴⁹ Manorathanandin on Pramāṇavārttika iv.50 mentions svarga as an example of the atyantaparokṣa, and nairātmya as an example of the parokṣa.

This threefold scrutiny—corresponding to the domains of pratyakṣa, anumāna and āgama—constitutes the fundamentally important dpyad pa gsum of the Tibetan tradition. See e.g. Tsoṅ kha pa, sDe bdun la ‘jug pa'i sgo don gñer Yid kyi mun sel (gSuṅ ‘bum, vol. tsha), f. 3b, 6b, 23a-b; rGyal tshab Dar ma rin chen, Thar lam gsal byed (gSuṅ ‘bum, vol. cha), f. 99a ff.; Guṅ thaṅ dKon mchog bsTan pa'i sgron me, Legs bšad Sñin po'i dka ‘grel, f. 24a ff.; Ruegg, D. Seyfort, La théorie du tathāgatagarbha et du gotra (Paris, 1969), 229.Google ScholarCf. recently Tillemans, T., ‘Dharmakirti, Aryadeva and Dharmapāla on scriptural authority’, Tetsugaku, 38 (Felicitation Volume for Professors A. Uno and K. Ogura, Hiroshima, 1986), 31–47Google Scholar; and Materials for the study of āryadeva, Dharmapāla and Candrakirti, I (Vienna, 1990), 23 ff.

In relation to the triad of pratyakṣa, anumāna and āgama, it is to be noted that Dharmakirti has himself alluded to a pramāṇatraya several times in his Svavṛtti on the Pramāṇavarttika (pp. 102 [cf. Pramānaviniścaya ii.34, p. 17*], 107, 176). Cf. Steinkellner, E., Dharmakirtis Pramānaviniścayaḥ, II (Vienna, 1979), 62.Google Scholar

⁵⁰ See rGyal tshab Dar m a rin chen on PV, Pramanasiddhi Chapter, in Thar lam gsal byed, f. (126b ff.,) 134a (under the rubric of yoṅs gcod or positive determination). Cf. ‘ Jam dbyaṅs mchog lha ‘od zer, Rva stod bsdus grva, f. 212 f.

⁵¹ A reference to a purusa as saṃvādaka and as pramāna is found in a passage of Divākara's, SiddhasenaTattvabodhavidhāyini (ed. Sanghavi, and Dośi, , Ahmedabad, 1924–1931), 468Google Scholar (cited by Steinkellner and Krasser, Dharmottaras Exkurs zur Definition gültiger Erkenntnis im Pramāṇaviniścaya, 27: loke pratijñātam arthaṃ prāpayan puruṣaḥ saṃvādakah pramāṇam ucyate).

On the concept of *pramāṇa-puruṣa and its sources, reference can be made to E. Steinkellner's article in: E. Steinkellner and H. Tauscher (ed.), Contributions on Tibetan and Buddhist religion and philosophy, 275–84, where tshad ma'i skyes bu is rendered as ‘person of authority’. [See now T. Tillemans, Persons of authority.]Google Scholar

⁵² For the etymology of tathā-gata by gad-, see e.g. Haribhadra, , Abhisamayalaṃkārālokā ii (ed. Wogihara, 183)Google Scholar: yathaiva te dharmā vyavasthitās tathaiva gadanāt tathāgata iti anenâviparitadharmadaiśikatvād vaktṛtvasampad uktā. (Cf. sGra sbyor bam po gñis pa [ed. M. Ishikawa, Tokyo, 1990], 7. And compare Bodhisattvatvabhūmi vi, 384: yat kiṃcid anena bhāṣitaṃ lapitam udāhṛtaṃ sarvaṃ tat tathā avitathêti tasmāt tathāgata ity ucyate with Dighanikāya, III 135 [cited below in n. 53 and containing implicitly also the etymology by tathā+gam-‘understand’], Aṅguttaranikāya II 24 and Itivuttaka, 121.)

According to Haribhadra, through this vaktṛtvasampad of the Buddha the appellation tathāgata relates to his vaktṛtvalakṣaṇā śāstṛtvasampad. Concerning the śāstṛtvasampad—characterizeded by this vaktṛtva of the Buddha and his pratipattṛtva (viz. the jñānao and prahāṇa-sampad, connoted respectively by the appellations arhant and samyaksambuddhd)—compare the idea of the Buddha's sampad in Dignāga's comment on his Pramāṇasamuccaya i.l (referred to in n. 10 above).—For secondary literature on the etymology of tathāgata, see Lamotte, E., Le traité de la Grande Vertu de Sagesse, i (Louvain, 1944), 126Google Scholar, n. 1; and for sugata see ibid., 131.

Concerning the more frequent etymology of tathā-gata by gam-, the semantic link between tathāgata and adhigatayāthātathya, the epithet of certain Ṛṣis in the Mahābhāṣya passage cited above (§ 7), was noted by Ruegg, D. Seyfort, ‘Védique addh et quelques paralléles à Tathāgata’, JA, 1955, 168 and 169 note 5.Google Scholar

On adhigata and sākṣatkṛta/sākṣātkāra, see above § 5 and n. 22.

⁵³ e.g. in Saddharmapundarika, ch. ii (ed. Kern and Nanjō), 39: śraddadhata me śāriputra bhūtavādy aham asmi tathāvādy aham asmy ananyathāvādy aham asmi/ durbodhyam śāriputra tathāgatasya saṃdhābhāṣyam/ tatkasya hetoḥ/ nānāniruktinirdeśabhilapanirdesanair mayā śāriputra vividhair upāyakauśalyaāśatasahasrair dharmaḥ samprakāśitaḥ/ atarko tarkāvacaras tathāgatavijñeyah śāriputra saddharmaḥ/ See also ch. xv, 315 (abhiśraddadhadhvam tathāgatasya bhūtām vācaṃ vyāharataḥ), and xv.23. Cf. Bodhisatttabhūmi vi (p. 384) quoted above in n. 52. The canonical Pali sources can easily be identified from the Pāli Tipiṭakaṃ concordance. See in particular the Pāsādikasuttanta (DN, III. 135): yaṃ kho sadevakassa lokassa samārakassa sabrahmakassa sassamaṇabrāhmaniyā pajāya sadevamanussāya diṭṭhaṃ sutaṃ mutaṃ viññātaṃ pattaṃ pariyesitaṃ anuvicaritaṃ manasā sabbaṃ tathāgatena abhisambuddhaṃ tasmā tathāgato ti vuccati. yañ ca … rattiṃ tathāgato anuttaraṃ sammāsambodhiṃ abhisambujjhati, yañ ca rattim anupādisesāya nibbānadhātuyā parinibbāyati, yaṃ etasmiṃ antare bhasati lapati niddisati sabbaṃ taṃ tath’ eva hoti no aññathā. tasmā tathāgato ti vuccati. yathāvādi … tathāgato tathākāri, yaihākāri tathāvādi. iti yathāvādi tathākārī yathākārī tathāvādī tasmā tathāgata ti vuccati … For the tathāgata being so called because of his aviparitārthavāditva, see Candrakirti's Prasannapadā i, p. 12.1; and i.3, p. 78.14.

The word evaṃvādin is used beside the term tathāgata in the summary pratityasamutpāda formula ye dharmā hetuprabhavāh …, and the question arises whether this expression also belongs in the present context. In rendering evaṃvādi mahāśramaṇah in this summary stanza, Lamotte, E. once used the translation ‘le veridique grand ascete’ (see Le traité de la Grande Vertu de Sagesse, II, Louvain, 1949, 631 note).Google Scholar But op. cit., 1076, Lamotte, has translated by ‘et leur suppression également il l'a révélée’; and later, in his Histoire du bouddhisme indien (Louvain, 1958), 547Google Scholar, Lamotte translated ‘telle est la doctrine du grand śramane’. It is indeed to be noted that evaṃvādin often stands beside evaṃdiṭṭhi(ka) and modifies samaṇabrāhmaṇa, and that it there evidently means’ having such (and such) a [possibly wrong] doctrine’. But in the pratityasamutpāda formula evaṃvādin may well have the special sense of speaking exactly and just as things are, i.e. truly.

Concerning addhā + vid-, see the present writer's remarks in JA, 1955, 163–4. And for a connexion between evam vid- and tathāgata, see Renou, L., Etudes védiques et pāninéenes, i (Paris, 1955), 83Google Scholar, n. 1, who refers to tathā ‘ à nuance “transcendante” (cf. tathātā)’. And for tathāgata: tathā pratipad-, see J. Filliozat, JA, 1952, 266, along with Franke, R. O., Dighanikāya (Göttingen, 1913), 143Google Scholar, n. 2 (on tathattāya paṭipajjanti).

⁵⁴ It is to be observed, nevertheless, that in Dharmakirti's Nyāyabindu i.2–3 (cited above, n. 2) samyagjñāna refers to pratyakṣa and anumāna, and therefore in this passage has t he same meaning as pramāṇa.—In the Nyāyasūtra, where the pramāṇas are four (pratyakṣa, anumāna, upamāna and śabda, I.i.3) and the prameyas 12 (beginning with ātman and ending with apavarga, I.i.9), tattvajñāna relates to 16 things the first and second of which are pramāna and prameya (I.i.l)

⁵⁵ Among the kinds of cognition recognized in the Tibetan Pramāṇa-school bead šes and yia dpyod can be described as being in accordance with fact (don mthuri). Nevertheless, since yid dpyod does not ascertain the elimination of imputation (sgro ‘dogs gcodpa) on the basis of either direct empirical perception or an inferential reason, and because it is conceptual fresh knowledge (gsar du žen pa'i rtogpd) bearing on a real object rather than non-conceptual fresh cognition (gsar du nogs pa'i bio) or non-conceptual fresh and non-disappointing/congruent knowledge (gsar du mi bslu ba'i šes pa), it is not counted as tshad ma= pramana (even though it is not log šes either).— On pramāṇa defined as fresh knowledge, see recently G. Dreyfus, ‘ Dharmakirti's definition of pramāna and its interpreters’, in: E. Steinkellner (ed.), Studies in the Buddhist epistemological tradition, 19 ff.

⁵⁶ PV, Pramāṇasiddhi chapter, k. lac:

pramāṇam avisaṃvādi jñānam, arthakriyāsthitiḥ/

avisaṃvādanam

On this relation between (avi)saṃvāda and arthakriyā, see also Tattvasaṃgraha 2958. And on the concept of arthakriyā in relation to knowledge as well as to a thing, see M. Nagatomi, ALB, 31–32, 1967–68, 52–72; and E. Mikogami, JIP, 7, 1979, 79–94.

The epistemological term avisaṃvādin = mi (b)slu ba—literally non-disappointing, unfailing (indefeasible), congruent—could be rendered by ‘veridical’ when relating to a cognition, and by ‘non-delusive’ when relating to an object of cognition. For avisaṃvādin as a quality that relates to both the yul can = visayin and the yul=visaya, see Devendrabuddhi, Pañjikā on PV, Pramanasiddhi chapter, k. 1 (P, f. 2a-b, D, f. lb). In PV, Svārthānumāna chapter, k. 216 avisamvāda is defined in terms of abaāhana ‘non-cancellation, non-sublation, non-invalidation, non-defeasibility’; and in PVSV, 217 Dharmakirti has equated it with avaiparitya ‘nonmistakenness’. (On the translation of avisamvādin see above, n. 9.)

Admittedly, an inference (anumāna) as a process of reasoning may legitimately be described as valid, and anumāna is of course also a pramāna. But since svārthānumāna and parārthānumāna are distinguished, there in fact exist two forms of anumāna which are not to be simply equated because only the first is a pramāna. And even if validity in the proper sense could somehow, by extension, attach to the inferential knowledge that is anumāna, it does not automatically follow that the concept of validity can be extended to pratyaksa-pramāna. In any case, rather than the pramāna of inferential knowledge known as (svārtha-) anumāna, it is doubtless the inferential process in a parārthānumāna that can philosophically speaking be appropriately described as formally ‘valid’.

Concerning the translation of pramāna by ‘validating cognition’ which has also been used in some scholarly work, it is not clear just what cognition is considered to validate. But it could be said that a pramāna is a validated correct cognition in so far as it is certified in terms of avisaṃvādana and arthakriyā.

Validity can of course attach in addition to a legal or administrative act, such as a law or a visa. But a law can remain valid even if it proves ineffective, that is, even if it does not successfully fulfil its purpose intended by the legislator. And a visa can be valid even if the intention with which it was issued (e.g. travel to a given destination) is, for some reason, not possible or realized. Thus there clearly exists a difference between validity and effectiveness, i.e. the realization of the envisaged effect. Here too the concept of validity on the one side and that of being cognitively efficacious (arthakriyāsamartha) and non-defaulting and unfailing (avisamvādin) on the other side are certainly not identical.

In French, a rendering for pramāna that is gaining currency is ‘connaissance droite’. For pramā Matilal, B. K. opted for the translation ‘true cognition’ in his Logic, language and reality (Delhi, 1985), 203Google Scholar, while noting that this rendering is ‘purely arbitrary’; and about pramāna he remarked that its meaning ‘seems to be so comprehensive in some contexts and so limited in other contexts that it defies all our attempts at finding a happy English translation’.—Concerning the translation equivalents and the problems discussed in this and the next section of this article, I am indebted to Ernst Steinkellner for his valuable observations.

⁵⁷ For these definitions, see The shorter Oxford English dictionary, s.v. authority.

⁵⁸ That is, the luṇ la brten pa'i rjes su dpag pa or yid ches rjes dpag, as distinct from the dṇos stobs rjes dpag = vastubalapravṛttānumāna which relates to the (cuṇ zad) Ikog gyur=parokṣa—i.e. that which owing to distance or some other impediment is not accessible to direct perception (pratyakṣa), but is nevertheless not transempirical—and which is thus in the domain of regular inference for oneself (svārthānumāna). See above, § 12.2; and cf. PV, Svārthānumāna chapter, k. 213 f., 309 f. See n. 49 for Manorathanandin ‘s distinction.

⁵⁹ See Pramāṇavārttika, Pramāṇasiddhi chapter, k. 29 f.

⁶⁰ See PV, Pramāṇasiddhi chapter, k. 132cd ff. See also k. lcd: avisaṃvādanaṃ, śābde ‘py abhiprāyanāedanād//; and PVSV 216.—Compare the concepts of pratyayita and āpta=yid ches pa/yid brtan pa. Thus, it is through his reliable teaching that the Bhagavant is a means of correct knowledge for those not yet able to achieve such knowledge directly for themselves. Compare further the old canonical idea that he who sees the dharma sees the tathāgata, where the two are as it were identified.

⁶¹ If the semantic force of pramāṇa in Pramāṇasamuccaya i. 1 were thus to be understood as secondary rather than primary, the form pramāṇabhūta (instead of pramāṇibhūta according to Pāṇini V.iv.50) might perhaps be explained in terms of Nāgeśa ‘s remark on likeness (sādrśyd) in his Pradipoddyota on vārttika 2 on V.iv.50, and by invoking paribhāṣā 15 in Nāgeśa's Paribhāṣenduśekhara specifying that a grammatical operation concerning a word is realized only when the word conveys its primary meaning, but not if it has a secondary meaning. See above, n. 25.

⁶² See above, nn. 9 and 56.

⁶³ See the present writer's article ‘Védique addh et quelques parallèles à tathāgata’ in JA, 1955, 163 ff. In Buddhist usage the terms ṛṣi and its derivative ārṣa have occasionally been applied to the Tathāgata.

⁶⁴ According to Nyāyabhāṣya, II.i.68, the prāmāṇya— ‘being a means of correct knowledge’ or ‘trustworthiness/authoritativeness’—of a teaching derives from being taught by an expert teacher (see above, § 6.1). And following the grammarian Nārāyana's explanation of the relation between prāmāṇya and the state of an ācārya (cited above, § 8.3) also, this cognitive ‘normality’, or trustworthiness/authoritativeness, results from being taught by an expert teacher. Here, then, we evidently have a three-stage process: (1) a person being sākṣātkṛtadharman or pratyakṣadharman, (2) being a teacher, and (3) the cognitive normality, normative force or trustworthiness/ authoritativeness (prāmāṇya) of this teacher/his teaching. This series would then seem to support the differentiation argued for above between direct cognition of reality or truth in stage 1 and having force/being reliable/trustworthy/authoritative in stage 3 (see above, § 6.1). (But, according to the theory in question, stage 1 might be preceded by a person's study of the āptāgama in a previous life; cf. Skanda/Maheśvara on Nirukta i.20, p. 114. In respect to the Veda, a Mimāṃsaka's view of the matter will of course be quite different because for him the eternal and timeless Veda is apauruṣeya and authorless.)

⁶⁵ Etymological uncertainty affects in particular the word tathāgata in the undetermined question (avyākṛtavastu) as to whether a tathāgata exists after death or not. (In this case the word is translated into Tibetan sometimes by de bžin ‘oṅs pa and sometimes by de bžin gšegs pa.) In this avyākṛtavastu, the word tathāgata is apparently being used not as an epithet of the buddha/bhagayant/sugata only but as an expression for ‘being, entity’. (It may be recalled that E. W. Hopkins pointed out that in Epic usage tathāgata sometimes has the meaning of ‘in so (grievous) a condition’, i.e. ‘dead’; see American Journal of Philology, 32/2, 1911, 205–9.)

In the case of tathāgata = de bžin gšegs pa used as an epithet of the buddh, it is not altogether clear whether tathā was understood as a noun meaning ‘reality, truth’, or whether the meaning ‘thus’ predominated. For addhāti ‘sage'addh —perhaps derived from the pronominal base a(d)- (see Renou, L., Grammaire de la langue védique, Paris, 1952, 329)Google Scholar, although Iranian azdā with which addhā has often been connected etymologically has another derivation according to Szemerényi, O., Die Sprache, 12, 1966, 202–05Google Scholar, and M. Mayrhofer, Etymologisches Wörterbuch des altindoarischen, s.v., with bibliography)—meaning ‘manifestly, certainly, truly’ (and originally no doubt ‘thus’), see the present writer's article in JA, 1955.—Very interestingly, some Indian lexicographers have given pratyakṣa as a meaning of addhā.

The words taāi(n) and tāyin (used e.g. in the introductory verse of Dignāga's Pramāṇasamuccaya cited above, § 3, where it has been understood as ‘protector’, Tib. skyob (pa); cf. Pāṇimya-Dhātupāṭha 489: tayR saṃtānapālanayoh, and sGra sbyor bam po gñis pa, p. 12), and sometimes also tādisa/tādṛ (n)/tādṛśa(ka), are epithets of a sage, and of the tathāgata = buddha, a link between the tādin (and the nāga) and tathatta being attested in Suttanipāta 520 and 522. See Franke, R. O., Dighanikāya (Göttingen, 1913), 88Google Scholar, n. 2; Bapat, P. V., ‘Tāyin, tāyi, tādi’, in: Law, B. C. (ed.), Bhandarkar, D. R. Volume (Calcutta, 1940), 249–58Google Scholar; Lüders, H., Beobachtungen ü'ber die Sprache des buddhistischen Urkanons (Berlin, 1954), 92 f.Google Scholar; Edgerton, F., BHSD, s.v. tādin/tadr(n)/ tādrśa(ka)/tāyinGoogle Scholar; Roth, G., ‘“A Saint like that” and “A Saviour” in Prakrit, Sanskrit and Tibetan literature’ (1968), reprinted in Indian studies (Delhi, 1986), 91–107Google Scholar; and Caillat, C., ‘Pronoms Et Adjectifs De Similarité En Moyen Indo-Aryen’, in: Indianisme et bouddhisme, Mélanges offerts à Mgr étienne Lamotte (Louvain-la-Neuve, 1980), 33 ff.Google Scholar

⁶⁶ See e.g. Manorathanandin (ed. Swami Dwarikadas Shastri), 95–6: pramāṇasaṃvādinaḥ parokṣārthasyopadeśas tatsākṣātkarapūrvaka eveti yuktaṃ tāyitvāt sugatatvānumānam bhagavataḥ/ … upadeśasya tathābhāvaḥ saṃvādakakatvaṃ prāmāṇyam…

⁶⁷ See Dignāga Pramāṇasamuccaya- Vṛtti i. 1. This fourfold scheme is again found in the Pramāṇasiddhi chapter of Dharmakirti's PV, even though it is there worded slightly differently in terms of the cultivation of compassion (karuṇābhydsa, k. 35), proclamation (ākhyāna, k. 132), being a Sugata (k. 139) and being a protector (tāya, i.e. svadṛṣṭamārgokti or catuḥsatyaprakāśana, k. 145–6).

⁶⁸ See above, n. 52.

⁶⁹ Abhisamayalāṃkārālokā ii, 183–4.