Ignorance, epistemology and soteriology

Part II

by Vincent Eltschinger

For Jacques May

The ﬁ rst part of this essay (Eltschinger 2009) concentrated on the basic features and likely sources of Dharmakīrti’s understanding of ignorance (avidyā). Against the Vaibhāṣikas, but with Vasubandhu the Kośakāra, Dharmakīrti deﬁ nes ignorance as a “counter-” or “anti-knowledge,” i.e., as a cognition that counteracts true (perceptual) knowledge (vidyā) by displaying contrary/erroneous objectsupports and aspects (viparītālambanākāra). According to him, ignorance amounts to pseudo-perception (pratyakṣābhāsa), hence conceptual construction (vikalpa), superimposition (samāropa) and concealment (saṃvṛti). The core of Dharmakīrti’s philosophy, the so-called apoha theory, provides an exhaustive picture of both ignorance as conceptuality and inference as a corrective (though conceptual) principle. This conception of ignorance, however, fails to account for the most dramatic form of the Buddhist ignorance, viz. its being responsible for deﬁ lements, rebirth and suﬀ ering. In

1 This study has been made possible by the generous ﬁ nancial support of the Austrian Science Fund (FWF-Projekt P19862 “Philoso phi sche und religiöse Literatur des Buddhismus”). Most sincere thanks are due to Isabelle Ratié, Birgit Kellner, Helmut Krasser and Ernst Steinkellner. Lambert Schmithausen also deserves my wholehearted gratitude for having gone through this essay with incomparably great care and erudition. My most sincere thanks are due to Cynthia Peck, who kindly corrected my English.

order to account for this eschatologically valued form of ignorance, Dharmakīrti equates avidyā with the personalistic false view (satkāyadṛṣṭi). Consistently enough, ignorance as satkāyadṛṣṭi is but a specialization or instantiation of ignorance as conceptuality insofar as the satkāyadṛṣṭi exhausts itself in one’s superimposing such conceptual constructs as “self/I” (ātman, aham) and “one’s own/mine” (ātmīya, mama) on reality. Both Dharmak īrti and his commentators evolved exegetical strategies in order to argue for the orthodoxy of this equation of ignorance with a false view (dṛṣṭi), which

Vasubandhu clearly refuses in the Abhidharmakośa (but not in his commentary on the Pratītyasamutpādasūtra). As for the sources of Dharmakīrti’s conception, they are very likely to consist of the Pratītyasamutpādasūtra and its numerous “idealistic” interpre tations (Yogācārabhūmi, Vasubandhu’s Vyākhyā). In the second part of this essay, I shall ﬁ rst inquire into Dharmakīrti’s account of dependent origination (pratītyasamutpāda), viz. his interpretation of ignorance as the origin of deﬁ lements (craving, etc.), clinging and rebirth. I shall then turn to the philosophical core of this study by attempting to show how Dharmakīrti’s views on ignorance and the two truths/realities provide the basic frame work of his epistemological theory. This is tantamount to claiming that Dharmakīrti’s epistemology, in locating ignorance and deﬁ ning the cognitive means of opposing it and entering the path toward salvation, is Buddhistic in both its inspiration and its ﬁ nality. As a consequence, his philosophy should cease to be regarded as a dry academic endeavour deviating from the spirit of Buddhism as a salvation system.

2.1. Dependent origination

2.1.1. In his account of the future Buddhaʼs philosophical reﬂ ections on the eve of his career, Dharmakīrti presents the cause of suﬀ ering (duḥkhahetu) in the following way: “The cause [of suﬀ ering, i.e., of rebirth,] is attachment bearing upon the condi tioning fac tors, [an attachment that is] due to the belief in self and oneʼs own.”2

2 PV 2.135ac1: ātmātmīyagrahakṛtaḥ snehaḥ saṃskāragocaraḥ / hetuḥ … sneha = tṛṣṇā according to PVP D56a7/P64a4 and PVṬ D117b3–4/P143b7; Ignorance, epistemology and soteriology (2) 29

Accor ding to Devendrabuddhi, craving proceeds from one’s adhering to the painful conditioned fac tors that are intrinsically free from self and one’s own, under the aspects of self and oneʼs own.3 This is tantamount to saying that deﬁ lements such as craving only occur once unreal aspects have been super imposed on dharmas, speciﬁ cally on the ﬁ ve constituents one clings to, which lack these as pects entirely. While commenting on another passage, Devendrabuddhi claims that deﬁ lements such as desire (another equivalent for attach ment and craving) proceed from oneʼs superimposing aspects such as permanent, plea su rable, self and oneʼs own on the impermanent, painful, selﬂ ess and empty consti tuents.4 One may adduce here a huge number of passages presenting one and the same idea: The personalistic belief is responsible for oneʼs super imposing contrary aspects such as self and oneʼs own on the selﬂ ess and empty constituents.5 As Dharma kīrti himself has it, “desire [arises] from the superimposition of another [i.e., unreal] nature on something (dharma) that does not have this nature.”6 PV 2.270 pro vides us with Dharmakīrtiʼs most signiﬁ cant statement as to how cra ving takes place once unreal as pects have been as

Śākyabuddhi (PVṬ D117b4/P143b7–8) unambiguously ex plains gocara as viṣaya. 3 PVP D56b1/P64a5–6: sdug bsṅal du gyur pa’i ’dus byas bdag daṅ bdag gi daṅ bral ba la bdag daṅ bdag gi’i rnam par mṅon par źen pas ’jug pa źes bya ba’i don to //. 4 PVP D60b2–3/P69a4–5: mi rtag pa daṅ sdug bsṅal ba daṅ stoṅ pa daṅ bdag med pa’i phuṅ po rnams la rtag pa daṅ bde ba daṅ bdag daṅ bdag gir sgro btags nas ’jug pa ’dod chags la sogs pa de dag … 5 E.g., PVP D88a4–5/P101b4: ñe bar len pa’i phuṅ po lṅa la gaṅ rtag pa daṅ bde ba daṅ bdag daṅ bdag gi rnam pa yod pa ma yin no //. PVP D88a6/ P101b5–6: ñe bar len pa’i phuṅ po lṅa la rtag pa la sogs pa’i rnam par ’dzin pa’i śes pa yaṅ rnam pa med pa ’dzin pa can yin no //. 6 PV 2.196ab: ātmāntarasamāropād rāgo dharme ’tadātmake /. Devendra buddhi explains (PVP D84a7–b1/P97a1–2): ’dod chags la sogs pa’i raṅ bźin du yaṅ ’gyur ba ma yin te / ’di ltar de bdag med can te / rtag pa daṅ bde ba daṅ bdag daṅ bdag gi daṅ bral ba’i yul du gyur pa’o // chos la ste phuṅ po la sogs pa’i raṅ gi ṅo bo la’o // bdag gźan sgro btags phyir te rtag pa daṅ bde ba daṅ bdag daṅ bdag gi’i raṅ bźin gźan du sgro btags pa’i rgyu’i phyir mṅon par źen pa’i mtshan ñid kyi chags pa skye bar ’gyur ro //.

cribed to rea l ity: “Having[, due to ignorance,]7 super imposed sixteen unreal aspects, viz. ‘lasting,’ ‘pleasant,’ ‘mine,’ ‘I,’ etc., on the four [Noblesʼ] Truths,8 one experien ces cra ving [for superimposed objects such as delight, etc.].”9 According to Devendra bu ddhi and

7 PVP D116a1/P134b2: sgro btags nas ni mi śes pa’i phyir … 8 At least according to the Vai bhāṣikas, each of the four Nobles’ Truths is to be successively con tem pla ted under four diﬀ erent aspects: the Truth of suﬀ ering under the aspects “impermanent,” “painful,” “emp ty” and “selfless;” the Truth of origin under the aspects of “(distant/material) cause” (as a seed), “arising,” “(serial) causation” and “(joint) condi tion;” the Truth of extinction, under the aspects of “extinction,” “calm,” “excellent” and “salvation;” the Truth of the path under the aspects of “path,” “ﬁ tness,” “access” and “con ducive to release” (AKBh 343,16–19 on AK 6.17c1: duḥkhaṃ caturbhir ākāraiḥ paśyaty anityato duḥkhataḥ śūnyato ’nā tma taś ca / samudayaṃ caturbhir hetutaḥ samudayataḥ prabhavataḥ pratyayataś ca / nirodhaṃ caturbhir ni ro

dhataḥ śāntataḥ praṇītato niḥsaraṇataś ca / mārgaṃ caturbhir mārgato nyāyataḥ pratipattito nairyāṇi kataś ca /. The sixteen aspects are listed at PVP D62a3–7/P71a1–6). The AKBh records a lengthy discussion pertaining to four diﬀ erent ways of interpreting these sixteen aspects (see AKBh 400,1–401,17 on AK 7.13a, Kośa 7.30–39, Pruden 1988–1990: IV.1110–1116). Accor ding to the fourth exegetical pattern, each of these aspects aims at counteracting (prati pakṣa) a particular false view (dṛṣṭi): The aspects anitya, duḥkha, śūnya and anātman counteract the false views of permanence, pleasura ble ness, one’s own, and self; the aspects of hetu, samudaya, prabhava and pratyaya contradict the false views of the absence of a cause, of a unique cause such as God or primordial matter (according to AKVy 628,30–31), of an evolution of being, and of an intelligent creation; the aspects nirodha, śānta, pra ṇīta and

niḥsaraṇa oppose the false views that release does not exist, that release is painful, that the bliss of dhyānas is the most excellent, and that liberation, because it is subject to falling again and again, is not deﬁ nitive; as for the aspects mārga, nyāya, pratipad and nairyāṇika, they respectively counteract the false views that there is no path, that this is a wrong path, that there is another path, and that the path is subject to retrogression; see AKBh 401,11–17, Kośa 7.38–39, Pruden 1988–1990: IV.1115–1116. The explanations provided by Dharma kīrti’s commentators are too few to

allow us to determine which interpretation, if any, they favoured. Devendrabuddhi and Śākya buddhi content themselves with listing the four aspects superimposed on each of the last three Truths (see PVP D115b6–7/P134a8–b2 and PVṬ D147b3–5/P182a8–b2). On the sixteen aspects, see Wayman 1980. 9 PV 2.270: sthiraṃ sukhaṃ mamāhaṃ cetyādi satyacatuṣṭaye / abhūtān ṣoḍaśākārān āropya pa ri tṛṣyati //. Note PVṬ D147b5–7/P182b2–4: sgro Ignorance, epistemology and soteriology (2) 31

Śākyabuddhi, ignorance,10 i.e., the false view of self, has one grasp aspects that are contrary to the real ones, i.e., superimpose an “I” on what is selﬂ ess and a “mine” on what is empty. But ignorance is also responsible for delu ded persons taking mo men tary things to be lasting (sthira) or even unchangeably permanent (kūṭa stha nitya),11 or holding intrinsically painful things to be pleasurable, i.e., not to be under the sway of cankers (sāsrava) or dependent on causes (hetuparatantra) in each of their successive phases (pratikṣa ṇam).12 2.1.2. According to Dharmakīrti and his commentators, the

personalistic false view is the (principal) cause (nidāna), the origin (yoni, btags nas ni yoṅs su sred ces bya ba’i tshig gis log par sgro ’dogs pa sṅon du soṅ ba can gyi sred pa ñid gsal bar bstan pa yin no // sgro ’dogs pa’i yul la ’jug pa’i sred pa de yaṅ sgro ’dogs pa’i rnam pa ñid yin la / sgro ’dogs pa’i rnam pa can gyi yul can gyi ñon moṅs pa daṅ ñe ba’i ñon moṅs pa thams cad ñid ma rig pa ñid yin pa … “And with the pāda (= PV 2.270d) āropya pari tṛ ṣyati, [[[Dharmakīrti]]] clearly indicates craving, which presupposes erroneous

superimposition. As for this craving, directed [as it is] to an object of superimposition, it also has the aspect of superimposition, and all the kleśas and upakleśas, which bear on an aspect of superimposition, are [nothing] but ignorance …” 10 PVP D115b3–4/P134a4: ma rig pa des kyaṅ sdug bsṅal la rtag pa źes bya ba’i rnam par ’dzin par byed do //. PVP D115b6/P134a7–8: re źig de ltar sdug bsṅal gyi bden pa la mi śes pa mi rtag pa la sogs pa’i rnam pa las phyin ci log tu sgro ’dogs pa yin no //. See also PVṬ D147a1–2/P181b3–5. 11 According to Devendrabuddhi, all that is produced and lasts more than one

moment is permanent (PVP D115b4/P134a5–6: skad cig ma las dus phyis gnas pa’i ṅaṅ tshul can du skyes pa thams cad rtag pa ñid do //. To be compared with Vibh. 102 n. 1: nityam iti vācye kṣaṇāt paraṃ sthāyī sarvo nitya ity arthaḥ /). According to Śākyabuddhi, all that is either unchangeably permanent or lasts for at least a second moment is permanent (PVṬ D147a6–7/P182a2–3: ther zug tu gnas pa’i rtag pa gaṅ yin pa daṅ skad cig ma gñis pa la sogs par gnas pa’i ṅaṅ tshul can dus gźan du gnas pa can gaṅ yin pa de thams cad ni ’dir rtag par ’dod pa yin gyi ther zug tu gnas pa ñid ni ma yin no źes de bstan par ’gyur ro //). 12 According to PVP D115b5/P134a6: bde ba źes bya ba’i zag pa daṅ bcas pa ma yin pa’am skad cig ma re re la rgyu’i gźan gyi dbaṅ la[s] phyin ci log tu btags pa’o //. duḥkha(bhūta) is regularly explained as sāsrava in PVP; see, e.g., PVP D57b7/P66a1 and PVP D58a3/P66a5.

prabhava), or the root (mūla)13 of all (kinds of) moral faults (doṣa), deﬁ lements (kleśa, upakleśa) or moral impurities (mala).14 Among the expressions denoting the fact that deﬁ lements such as desire originate from the false view of self, one also meets with “cause” (kāraṇa, alone or with preceding ut patti°, pradhāna°; hetu),15 “arising” (jāti, utpatti)16 and suﬃ xal elements such as °pū r va ka, °maya,17 °hetuka, °ja, °mūla, or °kṛta. Deﬁ lements originate from the personalistic false view (satkāyadarśanaja, ’jig tshogs su lta ba’i raṅ bźin), are (ca usally) prece ded/ac companied by the false view of self or by the adherence to self and oneʼs own (bdag tu lta ba sṅon du soṅ ba can, ātmātmīyābhiniveśapūrvaka), arise from the false view of self (bdag

tu lta ba las byuṅ ba), or have ignorance for their cause (avidyā hetuka).18 They are all based on the beliefs in “I” and “mine” (ṅar ’dzin pa daṅ ṅa yir ’dzin pa dag la gnas pa) and arise in dependence on a mind that complies with the false view of self and oneʼs own (bdag daṅ bdag gir lta ba’i rjes su ’brel ba’i sems la ltos nas … ’gyur ba).19 2.1.3. As we have seen, the belief in self and one’s own is the cause of suﬀ ering, i.e., attach ment bearing on the conditioning factors. In other words, ignorance is the cause of craving (tṛṣṇā), which

13 Respectively PV 1.223ab (nidāna gl. pradhānakāraṇa PVSVṬ 402,23– 24), PV 2.211a, PVSV 111,11, PV 2.197ab1 (mūla gl. daṅ po’i rten PVP D84b2/ P97a4), PV 2.212c. 14 E.g., PV 2.197a (doṣa), PV 1.222a (sarvāsāṃ doṣajātīnām), PV 2.214d1 (sarvadoṣa), PVSVṬ 401,24–25 and PVP D91a2/P105a5 ([sarva]kleśa), PVP D60a2–3/P68b4 (ñon moṅs pa daṅ ñe ba’i ñon moṅs), PV 2.212c (malāḥ sarve). On upakleśa, see also PVṬ D133a4–5/P164a4. 15 E.g., PVSVṬ 50,28 (kāraṇa), PVSVṬ 401,29 and PVP D91a2/P105a5 (ut pat tikāraṇa), PVSVṬ 402,23–24 (pradhānakāraṇa), PVSVṬ 401,21 (he tu). 16 E.g., PV 1.222b (jātiḥ), PVSVṬ 401,22 and 26 (utpatti). 17 Rendered in Tib. as raṅ bźin (can). But note PVṬ D137b3/P916b6: raṅ bźin ni ṅo bo ñid dam rgyu yin no //. 18 Respectively PVSV 111,19, PVP D93b1/P108a1 (on raṅ bźin, see above, n. 17), PVP D60a2–3/P68b2–3, PVSV 8,20, PVP D93a5/P107b5, PVSVṬ 401,24 and 25. 19 Respectively PVP D93b1–2/P108a1–2 and PVP D67b4/P77a6–7.

is nothing but the traditional sequence of dependent origination, where both function as the cause of suﬀ ering: As deﬁ lements, they give rise both to other deﬁ lements (e.g., tṛṣṇā → upādāna) and to act(ion)s (kriyā, e.g., avidyā → saṃs kāra, or upādāna → bhava), the latter being in turn res pon sible for new foundations (vastu) of existence (e.g., saṃskāra → vijñāna, or bha va → jāti).20 In so far as they give rise to actions leading to new existential foundations, ignorance and cra ving21 are the two causes of (re)birth ([punar] janman) and transmi gration (saṃsā ra),22 which are the hallmarks of suﬀ ering.23 Whereas Devendrabuddhi simply deﬁ nes su ﬀ er ing

20 See AK 3.27 and AKBh 134,26–135,3, Kośa 3.69, Pruden 1988–1990: II.407. 21 PVP D56a6/P64a3: skye ba’i mtshan ñid can gyi sdug bsṅal gyi rgyu; PVP D57b3/P65b4: bdag daṅ bdag gi la chags pa’i mtshan ñid can gyi sdug bsṅal gyi rgyu; PVP D115b6/P134a8: sred pa’i mtshan ñid can sdug bsṅal gyi rgyu; PVP D116a1/P134b3: sred pa sdug bsṅal gyi rgyur gyur pa; PVP D115b2/P134a2–3: sdug bsṅal gyi rgyu ni sred pa yin no źes bstan zin to // de yaṅ ma rig pa las byuṅ ba … According to Śākya buddhi, craving is kun nas ’chiṅ ba’i rgyu, “the cause of bondage,” and according to PVP D58b1/P66b4, attach ment leads to

kleśas, punarbhava and janmaparigraha. 22 Dharmakīrti’s commentators provide us with various deﬁ nitions of saṃsā ra. (1) PVP D62b3–4/P71b2–3: ’khor bar ’khor bas na ’khor ba ste / skye ba daṅ ’chi ba’i rgyun no //, to be compared with PVV 62,11–12: jan ma maraṇaprabandhaḥ saṃsāraḥ /. (2) PVP D95b6/P110b3: (bdag gir yoṅs su ’dzin pa) rtsom pa la sogs pa’i mtshan ñid can gyi ’khor ba …, which Śākya buddhi (PVṬ D138b6–7/P171a7–8) comments as follows: bdag gir yoṅs su ’dzin pa la sogs pa rtsom pa la sogs pa’i mtshan ñid can gyi ’khor ba źes bya ba la bdag gi ñid du gzuṅ ba’i srid pa’i loṅs spyod kyi mtshan ñid can gyi dṅos po la mṅon par chags pa sṅon du soṅ ba can gyi ’dzin pa ni yoṅs su (P om. su) ’dzin pa’o // rtsom pa ni mṅon par bsgrub pa’o //. Tib. mṅon par

bsgrub pa may translate either abhinirhāra (BHSD s.v., 52b–53a) or (more surely) abhi saṃskāra (BHSD s.v., 57b): Deﬁ ning “[re]exis ten ce” (bhava) in the context of dependent origination, Vasubandhu (Vai bhā ṣi ka deﬁ nition, AKBh 132,20–21) says: sa … paunarbhavikaṃ karmopacinoti …, “he accumulates action(s) that is/are conducive to re birth.” Note also TSP Ś230,8–9/K184,21– 22 (unidentiﬁ ed quotation): cittam eva hi saṃsāro rāgādikleśavāsitam /. 23 PVṬ D148a1/P182b6: ma rig pa daṅ sred pa ni sdug bsṅal gyi rgyu ñid yin te / phyin ci log pa’i raṅ bźin can źes bya ba’i don to //. Suﬀ ering is also deﬁ ned in terms of duḥkhatātraya. PVP D62b4/P71b3–4: sdug bsṅal rnam pa gsum gyis dṅos sam brgyud pas sdug bsṅal ba yin no //, which Śākyabuddhi, having named the three “pain fulnesses” (PVṬ D120b5/

as (re)birth (skye ba’i mtshan ñid can gyi sdug bsṅal), Dharmakīrti characterizes it as the constituents under going transmigration (duḥkhaṃ saṃsāriṇaḥ skandhāḥ).24 It co mes as no surprise, then, that Dharmakīrti declares that “as long as (s)he adheres to a self, the [[[person]] who ex pe ri ences craving remains] in saṃsāra.”25 According to Devendra bu ddhi, for whom “the personalistic false view is the cause of the connection (pra tisandhi) to a new existen ce (punarbhava),”26 “the [[[person]]] who is under the sway of the false view of self has the notion of pleasure (sukhasaṃjñā) with regard to suﬀ ering [and] will be connected to a new existence.”27 The link between the false view of self, attachment and rebirth can be summarized as follows: “Thus when there is adherence to a self, a mul titude of [[[moral]]] faults such as attachment to oneʼs own arise, and the attachment to a self causes [one] to take a [new existential] place (sthā na).”28 2.1.4. Let us consider now the genealogy29 of deﬁ lements from the personalistic false view. As we shall see, Dharmakīrti provides a

P147b5), comments as follows (PVṬ D120b6–7/P147b5–7): (1) duḥ khā veda nā is suﬀ ering in a direct way as duḥkhaduḥkhatā (its causes and conditions being suﬀ ering in an indirect way); (2) sukhā vedanā is suﬀ ering in a direct way as pariṇāmaduḥkhatā (its causes and con di tions being suﬀ er ing in an indirect way); (3) asukhāduḥkhā vedanā is suﬀ ering in a direct way as saṃskā

ra duḥkhatā (its causes and conditions being suﬀ ering in an indirect way). On duḥkhatātraya, see Schmithausen 1977. 24 Respectively PVP D56a6/P64a3 and PV 2.146c. 25 PV 2.218cd (leaving tena untranslated): tenātmābhiniveśo yāvat tāvat sa saṃsāre //. 26 PVP D85a6–7/P98a3–4: ’jig tshogs lta ba yaṅ srid par ñiṅ mtshams sbyor ba’i rgyur gyur pa … Note also, referring to the sahajaṃ satkāyadarśanam (PV 2.200d), PVP D85b5/P98b2–3: de yaṅ srid pa’i rgyu yin no //. 27 PVP D85a6/P98a3: gaṅ la bdag tu lta ba yod pa de ni sdug bsṅal la bde ba’i ’du śes can yin te / yaṅ srid par mtshams sbyor bar ’gyur ro //. 28 PVP D58a7–b1/P66b3–4: de ltar na bdag tu mṅon par źen pa yod na bdag gir chags pa la sogs pa’i skyon gyi tshogs ’jug par ’gyur źiṅ / bdag tu chags pas kyaṅ gnas yoṅs su len par byed do //. 29 “Genealogy” as a free rendering of Karṇakagomin’s krama (lit. “sequence,” “succession;” PVSVṬ 401,25–26: kena punaḥ krameṇa doṣāṇāṃ satkāyadarśanād utpattiḥ /).

coherent picture of the sequence avidyā–(ṣaḍāyata na–sparśa– vedanā–)tṛṣṇā–upādāna–bhava–jāti, although some items in his ac count have no explicit equivalent in the traditional twelvemembered chain of dependent origination. In Dharmakīrtiʼs opinion, the false view of self may be held directly responsible for the rise of at least three factors: the notion of otherness, the belief in oneʼs own, and attachment/craving. In an interesting statement, Dharmakīrti points out that “once [the notion of] a self exists, the notion of the other (parasaṃjñā) [arises, and] from this distinction between self and other [is born] grasping and aversion; bound to the se two, all the moral faults arise.”30 For reasons that I shall explain below, I am inclined not to follow the traditional explanation that links grasping/attach ment to (the notion of) the self and aversion to the notion of the other.31 For the time being, let us leave this problem out of consideration and focus on the genealogy of other ness: “As long as the mind adheres to a self (ātmeti), [it has] the notion of a self (ātma saṃjñā), and once this [notion] exists, all that [the mind] does not grasp in this way is [held to be] other.”32 In 30 PV 2.219 (āryā metre): ātmani sati parasaṃjñā svaparavibhāgāt pa rigrahadveṣau / anayoḥ sampratiba ddhāḥ sarve doṣāḥ prajāyante //. Delusion (moha), covet ous

ness (lobha) and hatred/aversion (dveṣa) are tra ditionally held to be the three root-deﬁ lements (mūlakleśa) or roots of evil (aku śa lamūla); see AK 5.20c and AKBh 291,8. Note, e.g., AK 5.48a2b: rāgotthā āhrīkyauddhatyamatsarāḥ. “From out of lust there proceeds disrespect, dissipation, and avarice” (Pruden 1988–1990: III.843, Kośa 5.91). For de ﬁ ni tions of āhrī kya, auddhatya and matsara, see AKBh 59,19–20 (Pruden 1988–1990: I.200, Kośa 2.170), AKBh 312,17 (Pruden 1988–1990: I.194, Kośa 2.161) and AKBh 312,16–17 (Pruden 1988–1990: III.842, Kośa 5.90). AK 5.48a2b: krodher ṣye pratighānvaye. “From

out of hatred there proceeds envy and anger” (Pruden 1988–1990: III.843, Kośa 5.91). For deﬁ nitions of krodha and īrṣyā, see AKBh 312,16 (Pruden 1988–1990: III.842, Kośa 5.90) and AKBh 312,19 (Pruden 1988–1990: III.842, Kośa V.90). 31 PVP D95b1/P110b5–6: bdag ñid du bzuṅ ba la yoṅs su ’dzin pa ni mṅon par chags pa’o // gźan ñid du rnam par phye ba la sdaṅ ba yin te / yoṅs su dor ba’o //. PVV 87,15–16: svaparavibhāgāc ca kāraṇāt sva pa ra yor yathā kramaṃ parigraho ’bhiṣvaṅgo dveṣaḥ parityāgas tau bhavataḥ /. 32 PVP D95a7/P110b4–5: ji srid du blo bdag ces mṅon par źen pa de srid du bdag tu ’du śes pa daṅ de yod na de ltar mi ’dzin pa gaṅ yin pa de thams cad gźan yin no //.

another statement, Dharmakīrti declares that “the false view of self generates th e belief in oneʼs own (ātmīyagraha).”33 Persons deluded by the false view of self regard the con sti tuents of being both as a self and as belonging to the self, but this feeling of property may well be extended beyond the constituents and range over parts of the world that have been posited as other than the self. The personalistic belief is respon sible for yet another factor, which is variously termed “desire” (rāga), “craving” (tṛṣṇā), “grasping” (parigraha) or “attachment”/“love” (sneha), and clearly corresponds to the eighth link of dependent origination, i.e., craving. In spite of this functional equivalence, I am inclined not to consider these terms as (always) synonymous, and to believe that Dharmakīrti introduced a causal sequence between them, thus splitting the traditional eighth link into two. If I am correct, from the false view of self arises ﬁ rst attachment or love for the self and oneʼs own, and then craving for the things that are regarded as beneﬁ

cial or pleasurable to the self. This can be seen in the following stanza: “The one who sees a self has a constant love for this [[[self]], thinking of it as] ‘I.’ Because of [this] love [for the self] he craves for the delights [of this self, and his] thirst conceals [from him] the drawbacks [of the things he deems conducive to these delights].”34 Here, both Devendrabuddhi and Manoratha nandin interpret “love” as “love for the self.”35 Whereas attachment is directed to the self (but bears upon the conditioned factors), craving is directed to the delights (sukha) of the self,36 i.e., to the things that are deemed con du cive to these delights,37 or to impure (sāsrava) things that are (deemed) favour able (anugrāhaka) in that they are conducive to the delights (of the self).38 Besides the frequent occurrence of ex pres sions such

33 PVSV 111,18: ātmadarśanam ātmīyagrahaṃ prasūte /. 34 PV 2.217: yaḥ paśyaty ātmānaṃ tatrāsyāham iti śāśvataḥ snehaḥ / snehāt sukheṣu tṛṣyati tṛṣṇā do ṣāṃs tiraskurute //. Note that Śākyabuddhi interprets doṣa as jātijarāmaraṇa (PVṬ D138b1/P170b8). 35 PVP D95a6/P111a2, PVV 87,3. 36 PVP D95a6/P111a2: bdag gi bde la sred ’gyur … 37 PVV 87,3–4: sukhasādhanatvenādhyavasitānāṃ vastūnām … 38 PVP D95b1/P111a4–5: bde ba sgrub par byed pa ñid du ñe bar ’gro ba zag pa daṅ bcas pa’i dṅos po … On anugrāhaka, see also PVSVṬ 402,8:

as ātma sne ha,39 ātmātmīyasneha40 or even sat kāyasneha,41 we also ﬁ nd Devendra bud dhiʼs deﬁ ni tion of sneha: “[We call] ʽloveʼ an inclination for self and oneʼs own which presupposes the [aforemen tioned delusion].” 42 Ac cord ing to Dharma kīrti, self-love and attachment for what belongs (or ought to belong) to the self is in turn the cause of aversion (pratigha) and hatred (dve ṣa): “Indeed, the one who, without gras ping (pari graha), sees that there is neither I nor mine, does not love anything and, [being so] unattached, does not hate anything [either], for there is no [[[aversion]]] for that which does not hinder the self or one’s own, nor for that which op poses the [said] hindrance.” 43 One can show aversion or hatred only for that which hinders (< uparodha) or harms (< pīḍā) what has been taken as self and oneʼs own:44 “Hatred [arises] with regard to

ātmāt mīyatvena tadanugrāhakatvena parikalpya … 39 E.g., PVP D58a1–2/P66a3. 40 PVP D57b3/P65b4. Love for self and one’s own is said to be directed to the object that is clung to as self and one’s own (ātmātmīyatvābhiniviṣṭe viṣaye ātmātmīyasnehaḥ, PVSVṬ 401,26–27). 41 E.g., PVP D90b5/P104b7: ’jig tshogs la chags pa. 42 PVP D60a2/P68b2 –3: de sṅon du soṅ ba can gyi bdag daṅ bdag gir źen pa ni chags pa’o //. Note also PVP D94b7/P109b4–5: chags pa ni bdag tu mṅon par chags pa’o // (maybe: sneha ātmany abhiṣvaṅgaḥ). 43 PVSV 111,15–17: na hi nāhaṃ na mameti paśyataḥ parigraham antareṇa kvacit snehaḥ / na cāna nu rā giṇaḥ kvacid dveṣaḥ / ātmātmīyānuparodhiny uparodhapratighātini ca tadabhāvāt /. 44 According to PVSVṬ 402,12: ātmātmīyatvena gṛhītasya ya uparodhaḥ pīḍā /. Note also Devendrabu ddhi’s deﬁ nition of dveṣa at PVP D60a2/P68b3: de (= chags pa) sṅon du soṅ ba can rjes su chags pa’i yul la gnod par byed pa

la mnar sems pa ni źe sdaṅ ṅo //. “Hatred is maliciousness with regard to that which injures the object of attachment[, a maliciousness] that presupposes the [afore-mentioned love].” The Sanskrit original for Tib. mnar sems pa is unclear. I would conjecture vyāpannacitta, although, to the best of my knowledge, mnar (ba) is not attested as a translation of vyāpanna(/vyāpāda): vyāpannacitta = gnod sems at AKBh 251,10 and 12 on AK 4.81ac1 (“de pensée méchante” in Kośa 4.178) as well as in the Saṃcetanī yasūtra quoted in AKVy 400,9–15 on AKBh 237,18. Jaini 2001:221: “The kleśas are like roots which produce as well as sustain an evil volition. Abhidhyā, vyāpāda, and mithyādṛṣṭi are not called roots, but are recognized as intensive states of the three roots of evil (akuśalamūla), viz. lobha, dveṣa, and moha respectively.

that alone which oﬀ ers opposition (pratikūla vartin) by its hostility to that which love for the self and oneʼs own bears upon (viṣayabhūta). Therefore, there is no hatred without love for the self and oneʼs own.” 45 Dharma kīrtiʼs unambiguous deri va tion of aversion from love is the reason why I cannot agree with Devendrabuddhiʼs and Manorathanandinʼs interpretation of PV 2.219b (sva paravibhāgāt parigrahadveṣau), which presupposes that what is other than the self can only arouse hatred. In Dharma kīr tiʼs eyes, that which is other than the self gives rise to aver sion only insofar as it opposes love, but arouses craving as soon as it is regarded as plea surable to the self. Craving for the delights of the self and that which is conducive to them generally

implies oneʼs running around in search of pleasure. This is indeed the Vaibhāṣika deﬁ ni tion of the ninth link of dependent origination, ap propriation or clinging (upādāna),46 and what Dharmakīrti obviously has in mind in PV 2.218ab: “Seeing [but] qualities [to the things that he deems pleasurable to the self], he craves [for them, thinking of them as having to become] ‘mine,’ and appropriates (upā √dā) the means [that are conducive] to them.” 47 But Dharmakīrti also holds love for the self to be the cause of the three diﬀ erent kinds of craving that the oldest layers of Buddhist canonical literature have made respon si ble for rebirth (paunarbhavika): craving for (future) existence (bhavatṛṣṇā), craving for sen sual pleasures (kāmatṛṣṇā), and cra v ing for non-existence/annihilation (vibhava tṛṣṇā).48 According

All evil volitions are essentially rooted in and spring from one or another of these three basic passions (mūlakleśa).” 45 PVSVṬ 402,13–15: ātmātmīyasnehaviṣayabhūtavirodhena yaḥ sthitaḥ pratikūlavartī tatraiva dveṣaḥ / tasmān nātmātmīyasneham antareṇa dveṣa iti /. 46 AK 3.23cd: upādānaṃ tu bhogānāṃ prāptaye paridhāvataḥ /. 47 PV 2.218ab (āryā metre): guṇadarśī paritṛṣyan mameti tatsādhanāny upādatte /. 48 PVP D79b3–4/P91a7–8: de la sdug bsṅal kun ’byuṅ ’phags pa’i bden pa gaṅ źe na / gaṅ sred pa ’di ni yaṅ srid par ’byuṅ ba can dga’ ba’i ’dod chags daṅ bcas pa de daṅ de la mṅon par dga’ ba’i ṅaṅ tshul can / ’di lta ste ’dod pa’i sred pa daṅ srid pa’i sred pa daṅ ’jig pa’i sred pa yin no źes gsuṅs so //. PVA 134,33–135,2: uktaṃ hi bhagavatā tatra katamat samudaya āryasatyam / yeyaṃ tṛṣṇā paunarbhavikī nandīrā ga sahagatā tatratatrābhinandinī / yad

to him, craving for sen sual pleasures is to be interpreted as the actions (pravṛtti) of living beings to secure what they hold to be pleasurable (sukhāpti), whereas craving for annihilation refers to those of their actions that aim at avoiding suﬀ ering (duḥkhānāpti). This matches again perfectly with the Vaibhāṣika account of the tenth link of dependent origination, viz. bhava (literally “existence”), which is to be understood as the “act(ion) that results in future existence” (bhaviṣyadbhavaphalaṃ karma): bhava refers to the act(ion)s resul ting in rebirth (paunarbhavika) that are accu mulated by those who run around (under the sway of craving) in order to quench their thirst.49 In these stanzas, Dharmakīrti brings together both meanings of bhava, i.e., action to secure the pleasures of the self, and the (future) existence that they inevitably lead to: “The cause [of suﬀ ering] is the longing for

[re]ex istence, because human beings reach a speciﬁ c [[[existential]]] place [and con di tion] due to [their] hope of obtaining it. The [afore-mentioned longing for existence] is [called] the desire for [re]existence. And since a living being [only] acts with the desire of obtaining pleasure and avoiding suﬀ er ing, these two [i.e., craving for pleasure and cra ving for the avoidance of suﬀ ering,] are regarded as the desire for sensual pleasures and the desire for annihilation. And since love for the self is the cause [of it, this dual action] per tains to everything for [the living being] who has the notion of [something] pleasurable with regard to [something] unpleasurable. There fore, craving is the basis of existence [i.e., the cause of bondage].”50

uta kā[m]atṛṣṇā bhavatṛṣṇā vibhavatṛṣṇā ceti … PVV 74,10–11: nanū ktaṃ bhagavatā tatra katamaḥ samudaya āryasatyaṃ paunarbhavikī nandīrā gasahagatā tatratatrā bhi nandinī yad uta kāmatṛṣṇā bhavatṛṣṇā vibhavatṛṣṇā ceti … For the Pāli text, see Vetter 1990: 87, n. 1. 49 AKBh 132,19–21 (together with AK 3.24ab): sa bhaviṣyadbhavaphalaṃ ku rute karma tad bhavaḥ / sa viṣayāṇāṃ prāptihetoḥ paridhāvan paunarbha vikaṃ karmopacinoti so ’sya bhavaḥ /. 50 PV 2.183a2–185: hetur bhavavāñchā parigrahaḥ / yasmād deśaviśeṣasya tat prāptyāśākṛto nṛṇām // sā bhavecchā ’’ptyanāptīcchoḥ pravṛttiḥ sukhaduḥ khayoḥ / yato ’pi prāṇinaḥ k āmavibhavecche ca te mate // sa r va tra cātma snehasya hetutvāt sampravartate / asukhe sukhasaṃjñasya tasmāt tṛṣṇā bhavāśrayaḥ //.

2.1.5. Although the standard formulation of dependent origination is traditionally held to range over three (Vaibhāṣika) or two (Yogācāra, Sautrāntika) lifetimes,51 at least some of its members can also be seen at work on the much shorter sequence of a few inter dependent psychological events. According to Vasubandhu, desire follows (anuśete, or: is connected to, samprayukta) a pleasant sensation (sukhā vedanā), whereas ave r sion follows (or: is connected to) an unpleasant sensation (duḥkhā veda nā).52 Dharmakīrti agrees with this common sense state ment.53 De pe n ding on whether a given tangible object

(spraṣṭavya) is considered fa vou rable (anugrā haka) or unfavourable to the self, the pleasant or unpleasant sensations born from the contact between this object and the sense faculties are conducive to the rise of deﬁ le ments such as desire or hatred.54 This obviously conforms to the pratītyasamut pā da sequen ce linking a sensory basis (āya tana), contact (sparśa) between the former and an ob ject, sensation, and craving. But as we have seen, to deem a given object favourable or unfavourable to the self belongs to the personalistic false view. Note should be made here that the erro neous aspects which the perso nalistic false view consists of overlap in part with those traditionally called “wrong notions” or “misconceptions” (viparyāsa), which amount to four55 and

51 For a useful overview, see Kritzer 1999: 67–72. 52 AKBh 312,1 –2: tri vedanāvaśāt trīṇi bandhanāni / sukhāyāṃ hi vedanā yāṃ rāgo ’nuśete ālambanasam pra yo gābhyām / duḥ khā yāṃ dveṣaḥ /. AK 5.55ab + AKBh 316,6 and 8: sukhābhyāṃ samprayukto hi rāgaḥ / sukha sauma nasyābhyāṃ rāgaḥ samprayuktaḥ / dveṣo viparyayāt / duḥkhābhyām ity arthaḥ / duḥkhena daur manasyena ca /. 53 See PV 2.151c2d: rāgāder vikāro ’pi sukhādijaḥ /, and the discussion below. 54 According to PVP D66a5–6/P75b5–6: reg bya’i khyad par gyi don phan ’dogs par byed pa daṅ de las gźan pa’i rjes su byed pas bde ba’am sdug bsṅal lam (sic) ’dod chags la sogs pa skye ba daṅ rjes su mthun pa yin pa … 55 To take the impermanent as permanent, the painful as pleasant, the impure as pure, and the selﬂ ess as a self (AKBh 283,5–7: catvāro viparyāsāḥ / anitye nityam iti / duḥ khe sukham iti / aśucau śucīti / anātmany ātme ti /). With the exception of the (im)pu re, they corres pond to the erroneous aspects one superimposes on the Truth of suﬀ

ering (see above, n. 8).

are regular ly held to be caused by imagina tion (saṅ kalpa).56 Śāntarakṣita and Kamala śīla provide interest ing mate rials regarding the rise of deﬁ lements from wrong notions. According to Śānta rakṣita, “deﬁ lements such as desire ari se once [erroneous aspects] such as beautiful, oneʼs own, lasting [or pleasant] have been superimposed on a woman, etc.”57 A little later, he says: “[A sensation] such as a plea

sant or unpleasant [one] arises in the pre sen ce of a [sensory] object[, say a woman]. For those who despise [sus pending] wisdom (prati saṅ khyā na) [and] are subject to improper reﬂ ection, this [[[sensation]]] gives rise to deﬁ le ments such as desire or hatred, which are [them selves] born from the ripe ning of a homologous latent tendency.”58 What does this amount to? The contact be tween an ob

56 On saṅkalpa, see May 1959: 181n. 586, PrP 451,9 ﬀ ., and the following excerpts: PVP D68a4–5/P77b8–78a1: ci ste ’di la yaṅ kun tu rtog pa yaṅ yan lag ñid du rtog par ’gyur ba de’i tshe kun tu rtog pa yaṅ bdag daṅ bdag gi daṅ gtsaṅ ba daṅ bde ba la sogs pa’i miṅ can gyi mtshan ñid kyi sa bon yin no //. PVP D67a3–4/P76b5–6: gaṅ gis bud med ’ga’ źig gi gzugs la sogs pa la kun tu rtog par byed ciṅ ’dod chags kyis gduṅs pa de ni … TSP Ś666,25–667,9/ K547,8–9: atītānāgate ’pi viṣaye saṅkalpavaśād abhivṛddhasu khā divi paryā sa sya puṃsaḥ pratisaṅkhyānanivṛttau teṣāṃ r āgādīnāṃ prabalatvaṃ dṛśyate /. MMK 23.1: saṅkalpa prabhavo rāgo dveṣo mohaś ca kath yate / śubhāśubhaviparyāsān sambhavanti pratītya hi //. PrP 452,4–5: tatra hi śubham ākāraṃ pratītya rāga

utpadyate / a śubhaṃ pratītya dveṣaḥ / viparyāsān pratītya moha utpadyate / saṅkalpas tv eṣāṃ trayāṇām api sādhāraṇakāraṇam utpattau /. PVSVṬ 166,29–167,2 gl. saṅkalpita (PV 1.70d) as āropita. To sum up, saṅkalpa is the bīja of the wrong notions or, equivalently, of the erroneous aspects, which in turn form the bases (āśraya < āśritya) or conditions (pratyaya < pratītya) of the deﬁ lements; to put it as shortly as Candrakīrti, saṅkalpa is the common cause (sādhāraṇakāraṇa) for the rise of the deﬁ lements. On saṅkalpa, see also below, nn. 68 and 69. 57 TS Ś1951ac/K1952ac: śubhātmīyasthirādīṃś ca samāropyāṅganādiṣu / rā gā dayaḥ pravartante … “Plea sant” accor ding to TSP Ś667,13–14/ K547,12–14 thereon: ātmā*tmīyanityasukhādyākārān abhūtān evā ro

payanto ’ṅga nā di ṣu pra vartante, na ca śubhādirūpā viṣayāḥ /. *TSPK with no equivalent of ātmā°. 58 TS Ś1953–1954d1/K1954–1955d1 (leaving tu untranslated): viṣayopani pāte tu sukhaduḥkhā di sam bhavāḥ / tasmāt samā na jātīya vāsa nā pa ripā ka jāḥ // r āga dveṣā dayaḥ kleśāḥ prati saṅkhyāna vidviṣām / ayo ni

śoma nas kā ravidhe yā nām … Note also PV 2.157ac: sajātivāsanābhedapra tibad dha pra vṛtta yaḥ / … rā gā dayaḥ … PVV 66,8–10: sajātivā sanā

ject and a sense faculty generates an aﬀ ec tive sen sation (pleasant, unpleasant, or neutral). People who do not devote themselves to medi tative practices such as the contemplation of the loathsome (aśu bhabhāvanā),59 and are therefore under the sway of improper reﬂ ection, su per impose erroneous aspects on the ob ject: that it is a women, of course, but also that she is attractive, desirable, (at least virtually) oneʼs own, etc. Aﬀ ective sensation as well as the superimposed aspects is in turn responsible for the actualization of the latent tendency of desire.60 Com men ting on his mas terʼs two

stanzas, Kamalaśīla provides us with a more systematic account of the sequence at stake: “For such is the sequence [of events]: When an object is present, a pleasure born of the sense faculty arises. And for those who, in the absence of any [sus pending] wisdom, abide in the improper reﬂ ection consisting of wrong notions such as self, this pleasure brings to maturity (vipāka) the latent tendency imprin ted by previous desire, etc. From this [coming to] maturity, deﬁ le ments such as desire arise. Therefore, the objects [themselves] are not directly the cause [of deﬁ le ments].”61 How should we un ’’

t mā tmī ya gra ha mū lasya sajāteḥ (Vibh. 66 n. 1: satkāyadarśanasya) pūrvapūr vā

bhy astasya rāgāder vāsanā ’parāpararāgādijanikāḥ śaktayas tāsāṃ bhe

daḥ pa rasparatas tatra prati baddhā pravṛttir janma yeṣāṃ te tathā … Here, sa jāti vāsanā is analysed as a genitive tatpuruṣa: “latent ten dencies of the homo lo gous [deﬁ lements which are rooted in the belief in self and one’s own].” But accord ing to Devendrabuddhi and Śākyabuddhi, the compound is to be analysed as a dvandva (PVṬ D123a2–3/P150b7): sajāti refers to the sat kā ya dṛṣṭi (āt mā tmī yadṛṣṭi in PVP D68a6–8/P78a3– 5) whereas the vāsa nā (bhe

da) consists in the pūrvarāgādyāhitabīja. 59 TSP Ś666,22–23/K547,6: aśubhādipratisaṅkhyāna. According to PVP D67a6–7/P77a1–2, rāgādi do not occur in those who have the aśubhādisaṃjñā. Note also Kamalaśīla’s deﬁ nition at TSP Ś666,23/K547,6–7: aśubhā dy ālam

ba nā r āgādipratipakṣabhūtā prajñā pratisaṅkhyānam /, which may be compared with AKVy 389,13 on AKBh 226,13–14: pratisaṅkhyānasya tatpra tipakṣabhāvanālakṣaṇasya, where tat = kleśa (con text: nirvāṇa). Note also AKBh 4,1 on AK 1.6ab1: duḥkhā dīnām ārya sa tyā nāṃ pratisaṅkhyānaṃ pra ti saṅ khyā prajñāviśeṣaḥ … (see also Kośa 1.8, and AKVy 16,4–7). 60 On latent tendencies and their actualization, see Eltschinger 2009: 57– 58, nn. 53–55. 61 TSP Ś667,19–22/K547,26–548,2: eṣa hi kramaḥ – viṣayopanipāte satīn driyajaṃ sukham utpadyate, tas māc ca sukhāt pra tisaṅkhyānavaikalye

derstand this strong in sis tence on the responsibility of improper reﬂ ection in the rise of deﬁ lements? 2.1.6. That improper reﬂ ection62 is closely connected with ignorance/personalistic belief and is part of the process leading to the rise of deﬁ lements can be easily substantiated.63 The problem raised by the source materials is rather that they testify to contradictory views regarding the relationship between improper reﬂ ection and ignorance/personalistic belief. Some sources (mainly Yogā cāra) introduce improper reﬂ ection in the deﬁ nition of the personalistic belief, which is held to be the manner deluded people improperly consider the ﬁ ve constituents of being as self and

saty ātmādiviparyāsalakṣaṇāyoniśomanaskāre sthitānāṃ pūrvarāgā dy āhita vāsanā paripāko bhavati, tato rāgādayaḥ kleśāḥ pravartanta iti na sākṣād viṣa yāḥ kāraṇam /. Note also Prajñākara gupta’s remarks while commenting on Dharmakīrti’s polemics against a Materialist uphold ing medical ideas (PVA 122,22–23): sukhādijo hi rāgādir na ka ph[ā ]di bhāvī / sukhaṃ ca kasyacit kathaṃ cid upalabdham ānta ra vāsanāprabodhāt / tato na rāgādayo do ṣebhya iti yuktam /. Though Śānta ra kṣi ta and Kamalaśīla cannot be suspected of allegiance toward Vaibhāṣi ka thought, their views are reminiscent of an interesting passage in the AK(Bh), accor ding to which a deﬁ lement arises out of three factors: ﬁ rst, its propensity (anuśaya) has not been eliminated; second, an object

(viṣaya, dharma) that is conducive to the actu ali zation of desire for sensual pleasures (kāmarāgaparyavasthānīya) is pre sent and perceived (ābhā sa gata); third ly, an improper reﬂ ection occurs with regard to the said object. AK 5.34, together with AKBh 305,19–20: aprahīṇād anuśayād viṣayāt pratyupasthitāt / ayoniśomanas kārāt kleśaḥ – tad yathā rāgānuśayo ’prahīṇo bhavaty apa ri jñātaḥ kāmarāgaparyavasthānīyāś ca dharmā ābhāsagatā bhavanti tatra cāyoniśo ma naskāra evaṃ kāmarāga utpadyate /. 62 AKBh 54,23: manaskāraś cetasa ābhogaḥ /. AKVy 127,33–128,2 thereon: manaskāraś cetasa ābhoga iti / ālambane cetasa āvarjanam / ava dhā raṇam ity arthaḥ / manasaḥ kāro manaskāraḥ / mano vā karoty āvar ja yatīti ma naskāraḥ /. PVSVṬ 50,29–51,12: ayoniśa ityādy asyaiva samarthanam / yoniḥ padārthānām anitya

duḥ khānātmādi / samyagdarśana pras[ū]ti he tutvāt / taṃ śaṃsaty ālambata iti yoniśaḥ / yoniṃ yoniṃ ma nas

ka ro tīti saṃkhyaikavacanād vīpsāyām (Pā 5.4.43) iti śaspratyayo vā / tathābhūtaś cāsau manaskāraś ceti yoniśomanaskāro nairātmyajñānam /. 63 On ayoniśomanaskāra and avidyā, see La Vallée Poussin 1913: 8–9, and especially Mejor 2001. On the improper reﬂ ection’s conditioning and reinforcing dṛṣṭis, see the passage of AN I.31 alluded to by Mejor (2001: 50 + n. 5); see also AKBh 5.32–33 in Mejor 2001: 51.

one’s own.64 Some materials regard improper reﬂ ection as caused by ignorance: this is the case of the Sūtra quoted in the AKBh, according to which, “depending on the eye and visible [[[objects]],] an incorrect (āvila) reﬂ ection born of delusion (moha ja) arises.”65 Much more common seem to be sources view ing improper reﬂ ection as the cause of ignorance/perso na listic belief: this is the case in a Sutta of the MN and two Suttas from the AN,66 in the MS,67 in the Sa t ya dvayanirdeśa(sūtra) as it is quoted by Kamalaśīla in BhK 1,68 and in the Sa he

tusapratyaya ni dānasūtra as it is quoted in AKVy 288,26–29 and used by Bhadanta Śrīlāta to demonstrate that ignorance (as the ﬁ rst link of dependent origination) has indeed a cause.69 Having quoted and commen ted on various excerpts

64 See Eltschinger 2009: 68–69, nn. 92 and 110. 65 AKBh 135,13–14 and AKVy 288,30–31: cakṣuḥ pratītya rūpāṇi cotpadyate āvilo manaskāro mohaja iti /. Note also AKBh 135,7 (in a quotation): avidyāhetukaś cāyoniśomanaskāraḥ /. 66 MN I.6 ﬀ . (no. 2, Sabbāsavasutta). Here, the ayoniso manasikāra is held to be responsible for the rise (uppajjhanti) and the increase (pavaḍḍhanti) of the three cankers (kāmāsava, bhavāsava and avijjāsava), which are in turn responsible for the rise of false views (diṭṭhi) concerning personal identity in the past (atītaṃ addhānam), in the future (anāgataṃ addhānam) and in the present (paccuppannaṃ addhānam), such as atthi me attā ti and na-tthi me attā ti. On this passage, see Collins 1982: 118–119; for similar expressions of the satkāyadṛṣṭi/sakkāyadiṭṭhi, see Eltschinger 2009: 73–75. AN V.113 ﬀ . (no. 61, Avijjā sutta) and V.116 ﬀ . (no. 62 Taṇhāsutta). According to the

Avijjāsutta, ayonisomanasikāra belongs to the eight aliments (āhāra) of avijjā; see Mejor 2001: 52–55. 67 MS 2.20.9 (Lamotte 1973: I.34): mṅon par źen pa’i rnam par rtog pa ni ’di lta ste / tshul bźin ma yin pa’i yid la byed pa las byuṅ ba’i ’jig tshogs la lta ba’i rtsa ba las byuṅ ba lta bar soṅ ba drug cu rtsa gñis daṅ mtshuṅs par ldan pa’i rnam par rtog pa gaṅ yin pa’o //. See also Lamotte 1973: II.115. 68 BhK 1.215[/525],7–14: kathaṃ mañjuśrīḥ kleśā vinayaṃ gacchanti / kathaṃ kleśāḥ parijñātā bhavanti / mañjuśrīr āha / paramārthato ’ty antājātānutpannābhāveṣu (sic, <Tib, but °nabhā° ms) sarva dhar meṣu

saṃvṛtyāsadviparyāsaḥ / tasmād asadviparyāsāt saṃkalpavikalpaḥ / tasmāt saṃ kalpa vikalpād ayoni śo ma na sikāraḥ / tasmād ayoniśomanasikārād ātma samāropaḥ / tasmād ātmasamāropād dṛṣṭi pary utthānam / tasmād dṛṣṭi paryutthānāt kleśāḥ pravartante /. 69 anyaḥ in AKBh 135,12, Bhadanta Śrīlāta according to AKVy 289,23; AKBh 135,12–17: anyaḥ punar āha / ayoniśo manaskāro hetur avidyāyā uktaḥ

of the Sūtra, Yaśomitra comes to the conclusion of a circularity (cakra ka), i.e., that improper reﬂ ection and ignorance condition each other.70 This is indeed the position most clearly exhibited by the Paramārthag āthās.71 To the best of my knowledge, Dharmakīrti alludes only twice to improper reﬂ ection in the context of the rise of deﬁ le ments. Unfortunately, both statements are far from unambigu ous. In PVSV 8,20–21, Dharmakīrti says that “[[[moral]] faults] such as desire presuppose [one’s] adhe rence to self and one’s own, for the rise of sūtrāntare / sa cāpi sparśakāle nirdiṣṭaḥ / cakṣuḥ pratītya rūpāṇi cot padyate āvilo manaskāro mohaja iti / vedanākāle cāvaśyam avidyayā bhavitavayam /

avidyā saṃ sparśajaṃ veditaṃ pratītyotpannā tṛṣṇeti sūtrāntarāt / ataḥ sparśakāle bhavann ayoniśomanaskāro vedanāsahavartinyā avidyāyāḥ pratyayabhāvena siddha iti nāsty ahetukatvam avidyāyāḥ … See Kośa 3.71n. 4. The whole discussion starts with the Sthavira Vasubandhu’s (AKVy 289,6: sthaviro vasubandhur ācā rya manorathopādhyāya evam āha…) claim that ignorance is not causeless on the basis of a Sūtra (the Sahe tusa pratyayasanidānasūtra according to AKVy 288,25–26; AKBh 135,7: ayo ni śomanaskāra hetukā ’vidyo ktā sūtrāntare /). As quoted by Yaśomitra (AKVy 288,26–29), this Sūtra runs as follows: avidyā bhikṣavaḥ sahetukā sapratyayā sanidānā / kaś ca bhikṣavo ’vidyāyā hetuḥ kaḥ pratyayaḥ kiṃ nidānam / avidyāyā bhi kṣavo ’yoniśomanaskāro hetur

ayoniśomanaskāraḥ pratyayo ’yoniśomanaskāro nidānam iti sūtre vacanāt /. This passage is also quoted in PrP 452,7–9 (avidyāpi bhikṣavaḥ sahetukā sapratyayā sanidānā / kaś ca bhikṣavo ’vidyāyā hetuḥ / ayoniśo bhikṣavo manaskāro ’vidyāyā hetuḥ / āvilo mohajo manaskāro bhikṣavo ’vidyāyā hetur iti), but as coming from the Pratītyasamutpādasūtra (PrP 452,6 [but see n. 3 thereon], Kośa 3.70n. 3). Immediately after the quotation, Candrakīrti remarks (PrP 452,9): ato ’vidyā saṅkalpaprabhavā bhavati /. Note also Yaśomitra’s (AKVy 289,1) reference to the Pratītya sam ut pādasūtra. Mejor (2001: 61–65) has translated

Vasubandhu’s polemics against Śrīlāta (AKBh 134,20–25 and 135,7–27). 70 AKVy 290,5–7: tad etac cakrakam uktaṃ bhavati / ayoniśomanaskārād avidyā / avidyāyāś cāyoni śo manaskāra iti /. This is, indeed, the position of the Sahetusapratyayasanidānasūtra (1. [[[moha]]] → ā vi lo ma na sikāra →

ayoniśo ma naskāra → avidyā → tṛṣṇā → karman → cakṣus [but also ear, nose, tongue, body and mind]; 2. cakṣus → karman → tṛṣṇā → avidyā → ayoniśomanaskāra); see above, n. 69, and Mejor 2001: 58 and 65–69 (Mejor’s translation of the Sūtra from Tibetan and Chinese sources). 71 Paramārthagāthā 20 (Wayman 1961: 170): ayoniśomanaskārāt saṃ moho jāyate sa ca / ayoniśomanaskāro nāsaṃmūḍhasya jāyate //.

all moral faults presupposes improper reﬂ ection.”72 A little later, he refers to a “speciﬁ c con dition for the rise of desi re, viz. improper reﬂ ection that consists in the false view of self/viz. the false view of self and improper reﬂ ection.”73 Commenting on the ﬁ rst passage, Śākyabuddhi and Karṇaka go min clearly equate the per sonalistic belief with improper reﬂ ection.74 But com menting on the second passage, they allow both a dvandva and a karmadhāraya ana ly sis of the com pound ātma darśanā yoni śomanaskāra.75 Though I am inclined to interpret these two passages as involving an equivalence between the false view of self and improper reﬂ ection, I would like to refrain from any conjecture regarding Dharma kīrti’s position on this issue.76 In the same way, I would like to postpone any attempt at orga ni z ing the above-mentioned (§2.1.5–6) psychological events into a sequen ce of phases exhibiting their mutual relationships. At

72 PVSV 8,20–21: ātmātmīyābhiniveśapūrvakā hi rāgādayo ’yoniśo manas kā rapūrvaka tvāt sarvadoṣot pat teḥ /. 73 PVSV 10,11: rāgotpattipratyayaviśeṣeṇātmadarśanāyoniśomanaskāreṇa yogāt /. 74 PVṬ Je D23b1–2/P28a1–2 = PVSVṬ 51,12–13: ātmādijñānam ayo ni śoma naskāras tatpūrvakatvāt sa rva rāgādidoṣotpatteḥ /. 75 PVṬ Je D27a2–3/P32a5–7 = PVSVṬ 55,29–56,12: ātmadarśanaṃ satkā yadṛṣṭiḥ / nityasukhādi viparyāso ’yoniśomanaskāraḥ / dvan dva samāsaś cā yam / ātmadarśanam evāyoni śomanaskāra iti viśeṣaṇasamāso vā /. Interest ing ly enough, Śākyabuddhi and Karṇa ka gomin explain “improper reﬂ ection” as a “wrong notion such as permanent or pleasant,” which matches perfectly Kamalaśīla’s deﬁ nition of “improper reﬂ ection” as “wrong notion such

as self.” According to these authors, then, improper reﬂ ection and wrong notions are conceptually equivalent. See above, n. 61. 76 Lambert Schmithausen (personal communication) has drawn my attention to the possibility that in the ﬁ rst passage (PVSV 8,20–21), Dharmakīrti may not be providing a logical justiﬁ cation, but rather a legitimation of his position by resor ting to a more traditional phraseology involving a co-extensivity of the two concepts: “d.h. weil sie [be kann ter maßen] ayoniśomanaskāra voraussetzen(, und dieser in nichts anderem besteht als eben dem ātmātmīyābhi niveśa).” By interpreting the compound in the second passage (PVSV 10,11) as a karmadhā raya, one may, then, read the two passages as exhibiting a homogeneous perspective. Ignorance, epistemology and soteriology (2) 47

the present state of research, such an attempt would only be idle speculation.77 2.1.7. Both wrong notions and the personalistic false view consist in the superimposition of erroneous aspects. Both are born of the actualization of a homogeneous latent ten den cy, which is the hallmark of conceptual construction. In other words, they are but concep tual constructs distorting both internal (the upādānaskandhas) and external reality. Dharma kīrti’s understanding of the personalistic belief harmonizes perfectly well with his overall conception of ignorance as the concealing conceptuality. As for his commentators, they seem to be justiﬁ ed in holding the satkāyadṛṣṭi to be a part, a branch or a speciﬁ c case of ignorance as a whole. That all conceptual constructs misre present reality, and some times are even deceiving from a practical point of view, does in no way mean that they are morally and (hence) eschatologically harmful. The su

7 To the best of my knowledge, no study has ever been dedicated to the issue of the Buddhist episte mologists’ way(s) of dealing with the Abhidharmic citta samprayuktasaṃskāras. Their assent to Vasuban dhu’s treatment of them cannot be taken for granted. To adduce but one example: saṃjñā is classiﬁ ed as a mahābhūmika, and as such, should occur together with vijñāna/citta/ manas; but niścaya(jñāna), the Buddhist epistemologists’ equivalent of saṃjñā, takes place after the sensory awareness (the latter giving rise to the vāsanā prabodha of the conceptual construct). In the present context, I think we should refrain from modelling the epistemologists’ conception of ignorance and improper reﬂ ection on Bhadanta Śrīlāta’s above-mentioned (see n. 69) elaborations on this topic. According to him, the improper reﬂ ection that is present at the moment of contact (sparśakāle) is the condition (pratyaya) for the

ignorance that coexists with sensation (vedanāsahavartiny avidyā) and in turn gives rise to craving. On the contrary, an Arhat’s unbiased (aviparīta) contact does not give rise to a deﬁ led sensation (kliṣṭā vedanā), which in turn does not provide a condition for craving. As both Śrīlāta (at least Vasubandhu’s Śrīlāta) and Yaśomitra describe it, Arhats do have sensations, but these do not generate craving, for only sensations that are accompanied by ignorance (sāvidya) give rise to craving (AKVy 290,13–15: arhatām asti ve da nā / na ca sā tṛṣṇāyāḥ pratyayībhavatīti / sāvidyaiva vedanā tṛṣṇāpratyaya iti gamyate /.) Śrīlāta adduces a reasoning (yukti) in order to make his point (AKBh 135,20–22): kayā yuktyā / na hi niravadyā vedanā tṛṣṇāyāḥ pratyayībhavaty arhatāṃ na cāviparītaḥ sparśaḥ kliṣṭāyā vedanāyāḥ / na ca punar niravadya syārhataḥ sparśo viparīta ity anayā yuktyā /).

perimposition of ego-related aspects alone re sults in the rise of deﬁ lements and reinforces one’s entanglement in saṃsāra. Dharmakīrti singles out this kind of harmful conceptual distortion as the personalistic belief.78 2.2. Ignorance, inference, and the path toward salvation 2.2.1. Like most Indian systems of salvation, Buddhism traces human beingsʼ unsatis factory condition back to ignorance, and presents itself as a cleansing and illuminative the rapy aimed at uprooting ignorance and the evils it is responsible for. Though the Buddhist epistemologists do not (even pretend to) bring any doctrinal or practical innovation into traditional Buddhist soteriologies, they lay

strong emphasis on the means of valid cognition (pramāṇa) as being instrumental in salvation. As is well known, Dignāga reduced the number of genuine pramāṇas from three (perception, inference, and scrip tures [[[āgama]]]) to two (perception and inference). At the pre sent state of our knowledge about Dignāga, however, it is diﬃ cult to estimate the extent to which non-epistemo logical, i.e., religious (lato sensu) considerations played a role in this epistemo logical reduction. If one cannot question Dharmakīrtiʼs endorse ment and consolida tion of Dignāgaʼs two-headed system as far as the epistemology is concerned, one might still argue that Dharmakīrtiʼs religious ideas, as they are known to us, provided, if not the basic framework, at least a strong additional motivation for sticking to this episte mo logy. This two-headed system could, after all, lay no claim to traditionally sanctio ned au tho rity be fore Dignāga.79 In my opinion, Dharmakīrti was deeply convinced that

78 PVSV 110,20–21: te [= doṣāḥ] vi kalpaprabhavāḥ /. PVSVṬ 398,23–25 thereon: vikalpād ayoniśo ma na si

kāra vikalpāt prabhava utpāda eṣām iti vigra haḥ / tathā hy ayoni śo manaskāram antareṇa saty api bāhye ’rthe not padyante rāgādayaḥ … 79 According to Frauwallner (1959), Vasubandhu had already restricted the number of pramāṇas from three to two in his Vādavidhi. But this might well be another case of Frauwallner’s use of the argumentum ex/a silen tio: the fact that no fragment dealing with (āpt)āgama is available to us does not mean that the original Vādavidhi did not address scripture as a third genuine means of valid cognition. At any rate, Vasubandhu seems to acknowledge

per cep tion and inference are enough both to shape and bring about the path to salvation and to provide the basic gnoseological features of the liberated yogin. To put it in a nutshell: Although it is conceptual in nature and thus belongs to ignorance, inference is the means through which per cep tion, which is nothing but “knowledge,” can be brought to function in its most genuine manner. Dharmakīrtiʼs system is in a way analogous to Tathāgatagarbha patterns of thought: though polluted by (ultimately adventitious) false views and deﬁ lements, the condition of the liberated mind is already here at hand. To be more precise, perception is basically the same with regard to its operation and objects before and after the revolution of the basis (āśraya parivṛtti). The only (but admittedly crucial) difference is that, at the completion of the path, it is no longer adulterated and contradicted by the counteracting cognitive factor called “ignorance.” Correcting erroneous superimpositions of all kinds and substituting them with true/validated intellectual contents is

the basic task of inference. Far from being a means of investigating the world and improving know ledge, inference aims ﬁ rst and fore three means of valid cognition in his AKBh (76,24–25: pramā ṇābhāvāt / na hi … pramāṇam asti pratyakṣam anumānam āptāgamo vā …) as well as in VY 173,16–17: mdor na rigs pa ni ’dir tshad ma rnam pa gsum po mṅon sum daṅ rjes su dpag pa daṅ yid ches pa’i gsuṅ ṅo //). Buddhist eristic-dialectical treatises are at great variance concerning the number (and deﬁ nitions) of the pramāṇas: four (or ﬁ ve) in the Hetuvidyā Section of the YBh (see, e.g., HV [§3.2] 4*,15–16,

where the last ﬁ ve items of the list deﬁ ning sādhana must be considered as pramāṇas because of their functional similarity (providing evi dence [yuktivāda] for the hetu, see HV [§3.22] 5*,3–5): sārūpyaṃ vairūpyaṃ pratyakṣam anumānam āptā gamaś ca; to the best of my knowledge, the HV only uses the term pramāṇa with regard to pratyakṣa; therefore, the number of the pramāṇas here is either ﬁ ve [or four if we consider that sārūpya and vairūpya occur once in a singular dvandva compound] or only one), four in the *Upāyahṛdaya/*Prayogasāra (*pratyakṣam anumānam upamānam āgamaś ca; see *UH 6,10–11 and 13,5 ﬀ .), three in Asaṅga’s Abhi dhar ma samuccaya (which, maybe on the basis of the BoBh and the Madhyāntavibhāga, sets the standard number for all subsequent Yogācāra treatises), i.e., pratyakṣa, anu mā na and āptāgama (see ASBh 152,27, 153,1 and 153,5).

most at discarding the erroneous superimpo si tions that ignorance is ultimately respon sible for.80 2.2.2. As we have seen, ignorance basically amounts to superimposition, con ceal ment/co ver ing, conceptual construct and pseudoperception. As such, ignorance is of a cogni tio nal character and consists in an “anti-knowledge,” in a mental event counteracting, con tra dicting or conﬂ icting with “knowledge.” What does, then, “knowledge” consist in? As we have seen, Dharmakīrtiʼs commentators deﬁ ne it as the “vision/perception of a real object” (bhūtārtha°/sadarthadarśana), or the “grasp ing of a real object” (bhūtārtha gra ha

ṇa).81 In these expressions, darśana and grahaṇa hint at perception and direct co gni tion (vijñāna), two terms denoting immediate sensory awa re ness of an object.82 Accord ing to Dharmakīrti, the nature of an object is undivided and amenable to sense percep tion.83 This is tanta mount to claiming that a single act of perception is enough to grasp this nature, and that it grasps it in its entirety (sarvātmanā), in all its aspects (sarvākā reṇa), so that no other means of valid cognition is needed for cognizing this nature in a positive way (vidhinā): Perception leaves no part of this undivided nature unknown, so that, say, inference or verbal knowledge might be needed in order to gain access to it.84 In other words, a single perception grasps an object as selﬂ ess and momentary, or, to be more precise, grasps a selﬂ ess and momentary thing.85 This can,

80 On the corrective function of inference, see Kellner 2004: 4–9. 81 See Eltschinger 2009: 41–42, n. 6. 82 AK 1.16a: vijñānaṃ prativijñaptiḥ; AKBh 11,7: viṣayaṃ viṣayaṃ prati vijñaptir upalabdhir vi jñā na skan dha ity ucyate /. AKVy 38,24: upalabdhir vastumātragrahaṇam /. 83 PV 1.43: ekasyārthasvabhāvasya pratyakṣasya sataḥ svayam /; PVSV 26,4: eko hy arthātmā / sa pratya kṣaḥ … 84 PV 1.45: dṛṣṭasya bhāvasya dṛṣṭa evākhilo guṇaḥ /; PVSV 26,5–6: tasya pratyakṣeṇaiva siddheḥ sarvā kā rasiddheḥ / tadanyasyāsiddhasyābhāvāt /; PVSV 26,9–11: tasmāt pratyakṣe dharmiṇi tatsvabhā va sā ka lyaparicchedāt tatrānavakāśā pramāṇāntaravṛttiḥ syāt /; PVSVṬ 121,17–18: pratyakṣadṛṣṭāt svabhāvāt ko ’nyaḥ /. Through perception, bare particulars are grasped in their entire true nature (dṛṣṭasarvatattva PVSV 26,14). 85 PVSV 43,8–11: nāpi svalakṣaṇasyānityatvādyabhāvaḥ / yas mān nā

of course, be traced back to Dharmakīrtiʼs “Sautrānti ka” assumption that a perceptual awareness results direct ly from a real thing’s causal eﬃ cien cy. According to a well-known state ment, “experts in reason(ing) hold that [for a given thing] to be a graspable [[[object]]] con sists in being a cause capable of casting (arpaṇa) [its own] aspect into cognition.”86 That real things cast their own aspect into the consciousness, thus giving rise to perceptual awareness, is the basic meaning of the description of this aware ness as arising by the force of something real (vastubala pra vṛtta). Dharmakīrti makes it especially clear in the following statement: “The property of a [[[Wikipedia:perceptual|perceptual]]] cognition is to grasp an object; [as for] this [[[object]], it] is grasped as it is, and it generates this [[[cognition]] of itself] through [its truly] existing nature. Such is the nature [of the cognition and of the object].”87 Devendra buddhi (as well as Śākyabuddhi and Kamalaśīla) exhibits the rationale behind Dharmakī rtiʼs (pro vi sio nitya tvaṃ n āma kiṃcid anyac ca lād vas

tunaḥ / k ṣa ṇa praty upa sthā nadhar matayā tasya tathābhūtasya grahaṇād etad evaṃ bhavaty anityo ’yam anitya tvam asyeti vā /. “Neither does the bare particular lack impermanence, etc., for what we call ‘im per ma nence’ is nothing other than the transient entity [itself. But] this is so because [those who see the last phase* of a continuum] grasp such an [[[entity]]] as having the property of being present [during only one] phase, [and thus say, ascribing properties]: ‘This is impermanent,’ or: ‘This has impermanence’.” *The last phase (antya kṣaṇa) is deﬁ ned in the following way by Śākyabuddhi and Karṇakagomin (PVṬ Je D48a7/P56b8 = PVSVṬ 95,30): sadṛśa kṣaṇāntarāpratisandhāyī kṣaṇo ’ntyakṣaṇaḥ … “The last phase [of an entity] is the phase which is not connected with a new (antara) similar phase.” Ac cor ding to PVSVṬ 184,5–6, PVSV 43,8–11 answers the objection formulated in PVSV 42,11–12: svalakṣaṇe cāni tya

tvā dyapra tīter atād rū pyam / teṣāṃ cāvastudharmatā /. “And since one does not cognize impermanence, etc., in the bare parti cu lar, [the bare par ticular] does not have this nature[, viz. im per manence, etc.], and [hence imper manence, etc.] are not pro per ties of [real] entities.” Note also PVSV 21,4–6: sa eva hi bhāvaḥ kṣaṇa sthitidharmā ’nityatā vacana bhede ’pi dharmidharmatayā ni mit taṃ vakṣyāmaḥ /. 86 PV 3.247b2d: grāhyatāṃ viduḥ / hetutvam eva yuktijñā jñā nā kā rār paṇa kṣamam //. See Hattori 1968: 53. 87 PV 2.206–207a1: viṣayagrahaṇaṃ dharmo vijñānasya yathāsti saḥ / gṛh yate so ’sya janako vidyamānā tmaneti ca // eṣā prakṛtiḥ … See also below, §2.2.6 and n. 136.

nal) position as follows:88 “When he is asked about the property of a cognition, the one who accepts that a cognition really grasps an object must answer that the property of a [[[Wikipedia:perceptual|perceptual]]] cognition is to grasp an object (= PV 2.206ab1). [And] if the property of all the cognitions possessing an object is to grasp an object, then they grasp [their] objects as they [really] are, (…) under an aspect such as impermanence, not under an unreal aspect. For in this way, if it is rationally established that a cognition cognizes (viṣayīkaroti) an object as it [really] is, that which is not cogni zed in this way is due to an external89 or inter nal90 adventitious cause of error, just as the [erroneous] cognition of a snake in the case of a rope in a dark place abundant in/suitable for snakes. Therefore, to grasp the real aspect of an object is the nature of a cognition. If on the contrary (atha ca) [its] na ture were to

grasp [an object] erroneously, then it would not have the property of grasping any object [at all]. Because in this way the object would not be as the cognition cognizes [it], and because [the co gni tion] would not cognize the object as it [really] is, cognitions would be devoid of object, (…) [and] hence all entities would be unestablished (…) There fore, the one who accepts a re la tionship between object and object-possessor has to hold that the property of a cogni tion is to grasp an object, [and] thus the nature of this [[[cognition]]] is to grasp the real as pect of an object. That which is other than this [i.e., un real,] is pro duced by a [purely] adventi tious con di tion.”91 This argument draws a sharp delineation

88 In an introductory statement, Śākyabuddhi reminds his audience that the following argument does not match Dharmakīrti’s ﬁ nal, Yogācāra position in epistemological matters. PVṬ D133b2–3/P164b3–5: don dam par rnam par śes pa ni don ’dzin par ’dod pas źes bya ba la / don dam par rnam par śes pa don ’dzin pa ñid ni ma yin te / gzuṅ ba ma grub pa’i phyir ro // ’on kyaṅ re źig phyi rol gyi don yod par ’dod pa gaṅ yin pa des ’di ltar ’dod par bya’o źes bstan pa’i phyir de skad du brjod pa yin no //. 89 PVṬ D133b3–4/P164b5: phyi rol lam źes bya ba ni ’dra ba gźan daṅ gźan ’byuṅ ba la sogs pa’i ’khrul par byed pa’i rnam pa’o //. See below, nn. 116 and 139. 90 Tib. cig śos = Skt. itara, lit. “other [than external].” 91 PVP D87b5–88a4/P101a2–b3: rnam par śes pa’i chos kyaṅ gaṅ źe na / źes dris pa na don dam par rnam par śes pa ni don ’dzin par ’dod pas rnam śes yul ’dzin pa’i chos śes brjod par bya’o // gaṅ gi tshe rnam par śes pa yul

between non-erroneous cognitions, which result directly from their objects’ causal eﬃ ciency, and erroneous cognitions, which result from a cause of error (bhrāntinimitta, pra tyaya). Whe reas the former are termed vastubalapravṛtta, true (bhūtārtha), and (being the mind’s) nature, the latter, which arise, among other factors, from the la tent ten dencies of erro neous conceptual constructs,92 are de can du gyur pa thams cad kyi chos yul ’dzin pa yin pa de’i tshe / mi rtag (D rtag: P rtag rtag) pa la sogs pa’i rnam pa gaṅ gis … yul yod pa de bźin du ’dzin ’gyur gyi med pa’i rnam pas ni ma yin no // de de ltar na śes pa don ji lta ba bźin du yul du byed par rigs pas thob pa na / de ltar na (D na: P om. na) rtogs pa ma yin pa gaṅ yin pa de ni phyi rol lam cig śos glo bur ba’i ’khrul pa’i rgyu mtshan gyis yin te / dper na sbrul du ’dris pa’i phyogs mi gsal bar thag pa la sbrul gyi (D gyi: P mi) śes pa lta bu’o* // de bas na yul gyi rnam pa yod pa ’dzin pa gaṅ yin pa de ni sems kyi raṅ bźin no // ci ste yaṅ

log par ’dzin pa ñid raṅ bźin yin pa de’i tshe yul ’dzin pa’i chos ma yin no // de ltar na ji ltar śes pas yul du byed pa de ltar don de ma yin źiṅ ji ltar don de yin pa de ltar yul du byed pa ma yin pa’i phyir / śes pa dag yul med pa can du ’gyur bas … de ltar na dṅos po thams cad ma grub pa yin te … de bas na yul daṅ yul can gyi dṅos po ’dod pa ñid kyis rnam par śes pa’i chos yul ’dzin pa yin par brjod par bya’o // de ltar na ’di’i raṅ bźin ni yaṅ dag pa’i yul gyi rnam pa ’dzin pa yin no // de rnam pa gźan du ’gyur ba gaṅ yin pa de ni glo bur gyi rkyen gyis byas pa ñid yin no //. * Vibh. 82 n. 4: mandamandaprakāśe sarpopacite pradeśe /. Note also TSP Ś1056,21–1057,5/K872,27–873,7: tathā hi – vi ṣa ya vi ṣayibhā vam icchatā cittaṃ

viṣayagrahaṇasvabhāvam abhy upe yam, anyathā vi ṣa ya jñānayor na viṣaya viṣayibhāvaḥ / arthagrahaṇasvabhāvatvenāṅgīkri ya māṇe yas tasya svabhāvas tenai vā tma no ’ṃśo ’rthas tena gṛhyata iti vak tavyam / anyathā katham asau gṛhītaḥ syāt / yady asatākāreṇa gṛhyeta ta taś ca viṣayaviṣayibhāvo na syāt / tathā hi – yathā jñānaṃ viṣayīkaroty arthaṃ na tathā so ’rthaḥ, yathā so ’rtho na tathā taṃ vi ṣa yī ka ro tī ti nirvi ṣa yāṇy eva jñānāni syuḥ / tataś ca sarva padā rthā siddhi prasaṅgaḥ / tasmād bhūtaviṣayā kā ra grāhitā ’sya svabhāvo nija iti sthitam / bhūtaś ca svabhāvo viṣayasya kṣaṇikānātmādirūpa iti pratipāditam etat / tena nairātmyagrahaṇasvabhāvam eva cittaṃ* nātma gra ha ṇa svabhāvam /. * TSPK reads eveti tan against TSPŚ and TSPTib eva cittaṃ; both the Jaisalmer ms and the Pāṭan ms read eve! cittaṃ. On this passage of the TSP, see McClintock 2010: 213–214. 92 PVṬ D133b4/P164b5–6: cig śos źes bya ba ni naṅ gi bdag ñid can gyi phyin ci log gi rnam par rtog pa’i bag chags źes bya bas bslad pa’o //. In an etymologizing vein, Devendrabuddhi explains āgantuka as follows (PVP D89a5/P103a2): rkyen gźan gyi rgyu mtshan las ’oṅs pa ñid yin pa’i phyir

scribed as avastubala pra vṛtta, as not agreeing with (means of) valid cognition (pramāṇāsaṃvādin) an d as adventitious (āga ntu[ka]). Accord ing to Dhar ma kīrti’s followers, this delineation only holds good provi ded per ceptual cognitions cog ni ze their objects in their real aspects. Claiming that a perceptual cognition grasps the real aspect of an object93 is tantamount to saying that it grasps aspects such as impermanence or selﬂ ess ness.94 As Kamalaśīla nicely puts it, “it is ﬁ rmly established that the intrinsic nature of the [[[mind]]] is to grasp the real aspect of an object; but it has been explained [earlier] that the real nature of an object consists of [its being] momentary, selfless, etc.; therefore, the mind has the grasping of selﬂ essness for its nature.”95 In other words, the nature of the mind is to perceive reality/the true nature (tattvadarśana) of things.96 And granted that selﬂ ess ness is the true nature of things, the mind turns out to be nothing other than discern ment (vi paśyanā) itself,97 which Śākyabuddhi deﬁ nes as wisdom (prajñā) bearing upon selﬂ ess

(*pratyayāntara nimittād āgatatvāt). Erroneous cognitions and deﬁ lements are due to raṅ daṅ rigs mthun pa’i ñe bar len pa’i rgyu (PVP D89a5–6/P103a2–3; *svasamānajātīyopādānakāraṇa; note PVṬ Je D251b6/P299a4–5 = PVSVṬ 400,30–431,9: upādāna

balabhāvīti vitathavikalpavāsanābalabhāvi). 93 PVP D87b7/P101a6: yul gyi rnam pa yod pa …; PVP D88a3/P101b2 = PVP 89a1/P102b3: yaṅ dag pa’i yul gyi rnam pa … 94 PVP D88b3–4/P102a3–4: mi rtag pa la sogs pa’i rnam pa yod pa’i yul …; PVP D87b6/P101a4 = PVP D90a4/P104a4: mi rtag pa la sogs pa’i rnam pa …; PVP D89a6/P103a3: bdag med pa …; PVP D89b3/P103a8: bdag med pa ñid … 95 TSP Ś1057,2–5/K873,5–7: bhūtaviṣayākāragrāhitā ’sya svabhāvo nija iti sthitam / bhūtaś ca svabhāvo viṣayasya kṣa ṇi kānātmādirūpa iti pra tipāditam etat / tena nairātmyagraha ṇasva bhāvam eva cittam … For the context of this statement, see above, n. 91. 96 PVP D87a7/P100b3: sems kyi raṅ bźin ni de kho na ñid mthoṅ ba’i bdag ñid can yin … (PVṬ D133a3–4/P164a2–3: de kho na ñid mthoṅ ba’i bdag ñid can yin gyi źes bya ba ni dṅos po ji lta ba bźin du gnas pa’i ’dzin pa’i* bdag ñid can źes bya ba’i don to) *Cf. PVV 82,14: yathāvasthitavastugrahaṇam; PVP D89b1/P103a6: sems ni ṅo bo ñid kyis de kho na ñid mthoṅ ba’i bdag ñid can yin … 97 PVP D90a1/P103b8: raṅ bźin yaṅ lhag mthoṅ yin … Ignorance, epistemology and soteriology (2) 55

ness.98 This “Sautrāntika” epistemology forms the back ground of Dhar m a

kīrti’s well-known al lu sion to the canonical topos of the mind’s being radiant (pra bhāsvara) by its very nature (prakṛtyā). “Radiant” is to be understood as “having the nature of grasping [entities] as they really are” (yathābhūtagrahaṇasva bhā va), or “con sis ting in the percep tion of rea li ty/the true nature [of things]” (tattva darśa na sātmaka).99 “Knowledge” is nothing but direct perceptual awareness, i.e., the mirror-like mind grasping the true nature of real entities.100 What can be regained from Dharmak īrtiʼs under standing of “knowledge” seems to mirror a signiﬁ cant shift from the

ideas held by his Yogā cāra predecessors. Deﬁ ning a threefold ignorance, the YBh declares its antidotes (vipakṣa) to be the insights born of audition, reﬂ ection and (mental) cultivation.101 In his PrSVy, Vasubandhu deﬁ nes “knowledge” as the insight born of reﬂ ection and (mental) cultivation.102 Dharmakīrti assents, of course, to the fact that ignorance can only be eliminated by the practice of the path and its three (or at least two) successive types of insight. But according to him, soteric practice does not aim at developing entirely new cognitive modalities, but rather, at freeing from all counteracting factors a type of cognition that has already been here at hand.

98 PVṬ D134b3/P166a1: lhag mthoṅ ba yin la źes bya ba bdag med pa la (D la: P la bya ba) dmigs pa’i śes rab bo. Discernment is described in BhK 1.219,23–220,4 as sarvadharmaniḥsvabhāvatālambana, and deﬁ ned in BhK 3.5,17–20 as follows: bhūtapratyavekṣaṇā ca vipaśyanocyate / bhūtaṃ punaḥ pud gala dhar ma nairātmyam / tatra pudgalanairātmyaṃ y ā skandhānām ātmātmīyarahitatā / dharmanairātmyaṃ yā te ṣām eva māyopamatā /. For a French translation, see Lamotte 1987: 340. On vipaśyanā/prajñā, see Eltschinger 2009: 57–58 (§1.2.5) and nn. 26–27. 99 PVP D89a5/P103a1: ’od gsal te / yaṅ dag pa ji lta ba bźin du ’dzin pa’i raṅ bźin yin no //; TS Ś3434ac1/K3435ac1: prabhāsvaram idaṃ cittaṃ tattvadarśanasātmakam / prakṛtyaiva sthitam … 100 On this point, see Eltschinger 2005: 190–192. 101 YBh 206,6–7: śrutamayyāś cintāmayyā bhāvanāmayyāś ca prajñāyā vi pakṣeṇa trayaḥ paryāyā yathā kra maṃ yojyante /. 102 PrSVy 9a1: bsams pa daṅ bsgoms pa las byuṅ ba’i śes rab ni rig pa źes bya’o //.

2.2.3. Contrary to “knowledge,” which, qua perception, is a cognition that is free of con cep tual construction (kalpanāpoḍha) and non-erroneous (abhrānta),103 the realm of ignorance is coextensive with conceptuality and error. “Error,” however, is not necessarily synonymous with “unreliability” (visaṃvāda, visaṃvāditva): Whereas “erroneous” is to be said of any cognition that does not arise from and hence display a bare particular, “un reliable” denotes those cognitions that are not conducive to a successful practical inter action with the particulars (or, as Dharmottara will say, that do not allow one to reach/obtain [pra√āp] the concrete particular).104 All con cep tual con structs are erroneous by their very nature and origin, but some of them are re liable (and hence valid cognitions, pramāṇa),105 whereas others are not. Śākya buddhi and Karṇakagomin have an op po nent ask the following question: “[But] if every conceptual con struct is simply er ro neous, why [do you hold] conceptual constructs such as [being] impermanent or selﬂ ess [to be] valid cognitions, but not conceptual constructs such as [being]

103 PVin 1.4ab1 ≈ NB 1.4: pratyakṣaṃ kalpanāpoḍham abhrāntam … On kal pa nā po ḍha, see Funayama 1992; on abhrānta, see Funayama 1999. 104 See Krasser 1995. Note also TSP Ś479,23–24/K392,7: avisaṃvāditvaṃ cābhimatārtha kriyā samarthā rtha prāpaṇaśaktiḥ /. “‘Being non-deceptive’ means the eﬃ cacy to realize the attainment of the object which is appropriate for the fulﬁ lment of a desired purpose.” Translation Funayama 1999: 79. On the diﬀ erences between Dharmottara’s and Kamalaśīla’s interpretations of abhrānta, see Funayama 1999: 80–81. 105 PV 2.5a: vyavahāreṇa prāmāṇyam … “Epistemic validity [is known] through practical activity.” Most important in this connection is the case of inference. PVin 2 46,5–8 (including PVin 2.1cd): tad etad atasmiṃs tadgrahād bhrāntir api sambandhataḥ pramā // svapratibhāse ’narthe ’rthādhyavasāyena pravar ta nād bhrāntir apy arthasambandhena tad avya bhicārāt pramāṇam /. “Die (Schlußfolgerungserkenntnis) ist wegen der Verbindung [mit dem Gegenstand] eine gültige Erkenntnis (pramā), obgleich sie wegen des Erfassens von etwas als etwas, was es nicht ist, Irrtum ist. (Das heißt:) Obwohl sie Irrtum ist, weil sie in der Weise auftritt, daß sie ihr eigenes Erkenntnisbild, das nicht der (wirkliche) Gegenstand ist, als [diesen Gegenstand] bestimmt, ist sie als mit dem Gegenstand verbundene (dennoch) gültige Erkenntnis.” Translation Steinkellner 1979: 26–27. See also PV 3.55–63.

permanent?”106 Dharmakīrtiʼs answer is as follows: “And since all this is an error due to the latent tendencies imprinted by [pre vious] per ceptions of the parti culars themselves, [those] conceptual constructs whose arising is [indirectly] bound to these [particulars] are reliable with regard to the thing [itself] although they do not display it, just like the error [consisting] of [cogni zing] a gem [is reliable] with regard to the radiance of that gem. [But] others [such as perma nence] are not [reliable with regard to the thing itself] because, (…) disregarding (pari tyajya) the conformity107 with the speciﬁ c [property] as it has been per cei ved, they superimpose another[, erroneous] speciﬁ c [pro perty] by [[[Wikipedia:arbitrarily|arbitrarily]]] grasping any sort of universal

(kiṃcitsāmānya). [These conceptual constructs are as unreliable with regard to the thing itself] as the notion of a gem [is unreliable] with regard to the radiance of a lamp.”108 Inasmuch as they do not display bare particulars and owe their existence to latent tendencies, all conceptual constructs are error. Some of them, however, are valid cognitions: Because the aspect they ascribe to the thing exists in it,109 and because they are indirectly related (pratibaddha) to the bare particular, they are reliable with regard to the thing itself, i.e., allow a successful practical interaction with it.110 Other con cep tual constructs are not valid cogni tions: because

106 PVṬ Je D95b3–4/P112a8–b1 ≈ PVSVṬ 183,9–10: yadi mithyārtha eva sarvo vikalpaḥ kasmāt … anityā nātmādivikalpāḥ pramāṇaṃ nityā[di] vikalpās tu neti … 107 PVṬ Je D96a3/P113a1 = PVSVṬ 183,23–24: anusaraṇaṃ niścayaṃ parityajya … 108 PVSV 43,2–7: sarvaś cāyaṃ svalakṣaṇānām eva darśanāhitavāsanākṛto viplava iti tatpratibaddha ja nma nāṃ vikalpānām atatpratibhāsitve ’pi vastu ny avisaṃvādo maṇipra bhāyām iva maṇibhrānteḥ / nānye ṣām / … ya thādṛṣṭaviśeṣānusa raṇaṃ parityajya kiṃcitsāmānyagrahaṇena viśe ṣānta rasamāropād dī pa pra bhāyām iva maṇibuddheḥ /. On maṇibhrānti, see Krasser 1991: 65–66n. 121. 109 PVṬ Je D95b6–7/P112b4–5 ≈ PVSVṬ 183,16–17: anityādirūpasya vastu ni vidyamānatvāt … 110 PVin 2 48,1–5 (together with PVin 2.7a): ata eva prāmāṇyaṃ vastuvi ṣa yaṃ dvayoḥ pratyakṣānumāna yoḥ, arthakriyāyogyaviṣayatvād vicārasya / sukhaduḥkhasādhane jñātvā yathārhaṃ pratipitsavo hi kiṃcit parī kṣante prekṣāpūrvakāriṇaḥ, na vyasanitayā /. “Eben daher bezieht sich

they superimpose an aspect that is not found in the thing itself,111 and because what they ascribe to it is not even indirectly related to it,112 they are un reliable with regard to the thing itself, i.e., are decei ving in practice. 2.2.4. In our philosophersʼ linguistic usage, however, “error” (bhrānti) quite often occurs as a shorter term for “unreliable cognition,” and is equated with “wrong notion” or “misconception” (viparyāsa). A si mi lar seman tic shift can be observed in connection with “super impo si tion” ([sam]āropa), no longer used in the general sense of conceptuality and con ceal ment, but in the sense of a mista ken identiﬁ cation barring determinate cognition (niś ca ya). In the present context, “error,” “superimposition” (both in this specialized meaning), “wrong notion,” and “lack of deter minate cognition” can be considered to be equivalent. Two kinds of situation are res

ponsible for the rise of error: the presence of a cause of error (bhrāntinimitta)113 and the lack of the causal conditions needed for determinate cognition (niścayapratyaya vaika lya).114 Together with Śākyabuddhi, we may consider the cause of error as twofold: The internal cause of error consists in the latent tendency of a contrary con ceptual construct (viparītavikalpavāsanā);115 as for the external cause of error, it is most often exempliﬁ ed as the arising of ever new similar phases (sadṛśā pa rā pa rotpatti) in a conti nuum,116 die Gültigkeit der beiden, Wahrnehmung und Schlußfolgerung, auf das Wirkliche, denn eine prüfende Erkenntnis hat ein Objekt, das fähig ist einen Zweck zu erfüllen. Vernünftig handelnde Leute, die (auch nur) ein wenig abwägen, (tun dies), wenn sie die Mittel für Lust und Leid (einmal) erkannt haben, aus der Absicht, [diese] nach Vermögen zu erreichen, aber nicht aus [bloßer] Neigung.” Translation Steinkellner 1979: 29 (slightly modiﬁ ed). 111 PVṬ Je D96a1/P112b7 ≈ PVSVṬ 183,20–21: teṣāṃ [= nityā di vi kalpānām] … vastuny avidyamāna syai vā

kā rasya samāropāt /. 112 PVṬ Je D96a4/P113a2 ≈ PVSVṬ 183,26–27: pāramparyeṇāpi … apra ti baddhatvāt /. 113 PV 1.44a, PVSV 26,15, and passim. 114 PVSV 26,19. 115 See above, n. 92. 116 PVSV 26,20–21; sadṛśāparotpatti at PVṬ Je D61a3/P72a2 = PVSVṬ 122,10–11, PVṬ Je D61a5/P72a5 = PVSVṬ 123,8–9; note also PVṬ Je

D61b4/P72b5 = PVSVṬ 123,27–28: sadṛśasya dvitīyasya kṣaṇa syot pattyā bhrāntinimittena … See also above, n. 89. Locus classicus for sadṛśāparotpatti is PVSV 21,6–9: tāṃ punar asya kṣaṇasthitidharmatāṃ svabhāvaṃ svahetor eva tathotpatteḥ paśyann api mandabuddhiḥ sat to palambhena sarvadā tathā bhā vaśaṅkāvipralabdho na vyavasyati sadṛśāparotpattivipralabdho vā /. Trans lated according to Śākyabuddhi’s explanation (PVṬ Je D46b2– 47a1/P54b6–55a6): “However, al though (s)he experiences this property of lasting [only] one phase[, a property which is] the nature of the [[[entity]]] since [this entity] is produced such [i.e., momentary,] by its own cause, a [[[person]]] of weak intellec t fails to determine [it in the same way as (s)he has just experienced it; this failure occurs] either [because this person,] due to having perceived the existence [of this entity at one phase, is] mistaken by the supposition that

it permanently (sarvadā) exists in this [very] way, or [because this person is] mistaken by the rise of a new (apara) phase similar [to the former one].” According to Śākyabuddhi’s interpretation (PVṬ Je D47b6–48b1/ P56a6–57a1), the ﬁ rst cause of error (*vipralambhanimitta) is proper to the outsiders (tīrthika) professing the doctrine of non-momentariness (akṣa ṇika vāda), and points to their in ter nal *ku dṛṣṭy abhiniveśavāsanābīja (or else: *an ādikudṛṣṭy abhiniveśabīja), which is reinforced by the false views propagated by wrong treatises (*kuśāstradṛṣṭi). As for the second cause of error, it is aimed at ex plain ing why the Buddhists, who follow sound reasoning and scripture (yuktyāgama) professing momentariness, still do not ascertain

momentariness upon perceiving the real entity. Karṇakagomin’s explanation (PVSVṬ 91,23) of manda buddhi is worth noticing: anādisaṃsārābhyastayā nityā di rūpā vi dyā vāsanayā mandā bud dhir yasya … “Whose intellect is [made] weak by the latent tendency, repeated [and reinforced] in the beginningless saṃsāra, of ignorance in the form of [mistaken aspects] such as ‘perma nent’.” This ignorance (or rather, its latent tendency) being the internal cause of error, the two causes men tio ned by Dharmakīrti point to external causes of error (bāhyam api bhrāntibījam, PVSVṬ 91,27). Note also PVSV 100,4–7 = PVin 2 82,7–9: tam asya mandāḥ svabhāvam ūrdhvaṃ vyavasyanti / na prāk / dar śa ne ’pi pāṭavābhāvād iti tadvaśena paścād vyavasthāpyate / vikāradarśane neva viṣam ajñaiḥ /. “Weak[-minded people] identify this [transient] nature of the [[[entity]] only] later [i.e., at the time of the interruption of the continuum, but] not before [i.e., at the time of the existence of the entity], because even though they [directly] experience [this nature], they lack [[[intellectual]]]

sharpness. Therefore, [this transient nature] is ascer tained [only] later on account of this [[[determination]]], just as ignorant [persons identify a poisonous substance that they have seen only] by experiencing a [morbid] aﬀ ection [such as over-salivation].” See also Steinkellner 1979: 98. Note Karṇakagomin’s explanation of mandāḥ in PVSVṬ 366,27: ā saṃsāram avidyānubandhān mandāḥ … This explanation is borrowed from Dharmottara’s PVinṬ Dze D249b5/P301b3–4: ’khor ba ji srid par ma rig

hich leads to the superimposition of aspects such as permanent (nitya), enduring (sthira), and non-momentary (akṣaṇika).117 Be it internal or external, this cause of error impedes determinate cognition (niścayapra ti ro dhin, °vi ban dhaka).118 The lack of (conce ptual) habitus (abhyāsa) is most often quoted as being among the conditions that, when lacking, prevent determinate cognition from arising.119 Just as de ter minate cognition bears upon one speciﬁ city (bheda) or aspect (ākāra) of a previously cognized particu lar, wrong no tion superimposes one partial erroneous/contrary aspect

(aṃśasamāropa)120 and associates (< samyojyeta, PV 1.44b) another, i.e., a false quality (guṇa; glossed as rūpa, dharma),121 to the thing. As Dhar makīrti himself has it, “though it has been per ceived as distinct from all [other entities], an entity is not [necessarily] recognized in this way [i.e., in all its as pects], because an obstruction (vyavadhāna) to [the recognition of] a certain speciﬁ city [such as mo menta riness] may occur.”122 Deter mi na te cognition (niś ca ya, °jñāna, °manas) and su pe r imposi tion (samāropa, °jñāna; āropamanas) are mutually exclusive and stand in a relationship of mu pa daṅ rjes su ’brel pa źan pa … Note also the various inter pre tations of the fact that the determinate cognition arises only at the time of

pravāhavicche da: (1) PVṬ Je D227a3–4/P263b7–8: mthoṅ ba’i dus su ṅes pa yod pa ma yin te / ma rig pa’i mun pa ñid kyi phyir daṅ gźan rgyun ’dra ba skye ba’i phyir ro // mthoṅ ba gsal ba med pa’i phyir ro //. (2) PVSVṬ 366,28–29: na darśanakāle ’dhyavasāyo ’sti / avidyā(sāma)rthyāt sadṛśāparot pat tyā ca dar śa napāṭavasyābhāvāt /. (3) PVinṬ Dze D249b6–7/P301b5: ma rig pa daṅ ldan pa’i źan pa rnams la mthoṅ ba gsal ba med pa’i phyir ro //. Here again, both Śākyabuddhi and Karṇaka gomin suggest that the absence of niścaya proceeds from an internal (ignorance) and an external (the rise of a new similar phase) cause. 117 See also above, §2.1.1. and n. 11. 118 PVSV 26,14, PVṬ Je D61a5/P72a5 = PVSVṬ 123,8. 119 PVṬ Je D61b2–3/P72b3 = PVSVṬ 123,21. On niścayapratyayas, see Kellner 2004: 19–32. 120 PV 1.50a; PVSV 27,22–28,1: ākārasamāropa; PVṬ Je D62b3/P73b6 = PVSVṬ 125,28–29: tadviparītā kā rasamāropī viparyāsaḥ. 121 PV 1.44b, PVṬ Je D64b7/P76a8–b1 = PVSVṬ 131,11, PVV 306,6. 122 PVSV 28,13–14 (leaving hi untranslated): na hi sarvato bhinno dṛṣṭo ’pi bhāvas tathaiva pratyabhijñā ya te / kvacid bhede vyavadhānasambhavāt /.

tual annulment (bā dhya bā dhakabhāva):123 When aspects such as lasting, endowed with a self (sātmaka), or un con ditioned (akṛtaka) are superimposed, (real) con tra ry aspects such as imperma nent/ mo mentary, selﬂ ess (nirā tma ka), or condi tioned (kṛ ta ka) are not made the objects of deter minate cognitions.124 2.2.5. According to Dharmakīrti, the function (vyāpāra) and aim (phala, artha) of inferen ce125 (anumāna, liṅga, sādhana) as a means of valid cognition is not to cognize something in a positive way (vidhinā) or to determine the nature of an entity (vastusvabhāva niśca ya),126 but to rule out, negate, or exclude (vyavaccheda, niṣedha, pratiṣedha, nivṛtti, apo ha) un reliable super impo sitions and wrong

123 PV 1.49ab: niścayāropamanasor bādhyabādhakabhāvataḥ /; PVSV 28,16–17: samāropaniścayayor bā dhyabādhakabhāvāt /. 124 niścitākāras: kṛtakatva (PVSVṬ 124,26 and 125,23–24), anitya tva (PVSV 26,5), kṣaṇikatvādi (PVSVṬ 130,28), kṣaṇikatvānātmādi (PVSVṬ 124,12), asthira (PVSVṬ 129,28), nirātmaka (PVSVṬ 129,28); sam āro pi tākāras: sthira (PVSV 28,11, PVSVṬ 122,12), sātmaka (PVSV 28,11), sthiti (PVSV 26,21), akṛtaka (PVSVṬ 125,23–24), nityādi (PVSVṬ 124,13 and 125,24). 125 Note should be made that inference is itself strictly of a conceptual nature, and as such is basi cally on the side of error and ignorance. An inference indeed mobilizes two properties (dharma, a pro bans [[[sādhana]] dharma, hetu, liṅga] and a probandum [[[sādhyadharma]]]) that are thought

to be long to a single property posse ssor (dharmin, or “subject”). Both of these two properties are universals (sāmānya) unduly ascribing a single unitary aspect to the many. At the same time, these two diﬀ erent properties are tied to one and the same subject, thus unduly dividing the indivisible. To unify the many (the seed of the use of universals*) and divide the undivi ded (the seed of co-reference [sāmānādhikaraṇya]**) are indeed the two main psycho logical operations giving rise to conceptual constructs. *According to PVṬ Je D101a6/P119a4 and D101a7/P119a5: spyi’i tha sñad kyi sa bon … (sāmānya vya va hā ra bī ja); the psychological genesis of uni ver sals is presented in a nutshell in PV 1.82. **According to PVṬ Je D101b4/P119b3: gźi mthun pa ñid … [kyi] sa bon (sāmānādhikaraṇyabīja); the psychological genesis of co-reference is presented in a nutshell in PV 1.83. See also Eltschinger 2009: 59–62 (§1.2.10). 126 Resp. PVSV 27,10 and PVSV 28,20.

notions:127 “Superimpositions endowed each with its own cause are as many as the alien natures (parabhāva) [wrongly ascribed] to the [en ti ty]. In that they exclude these [superimpo sitions], the means of valid cognition [na med ‘inferences’] can therefore be useful. But these [inferences,] aiming (°phala) [as they do] at the exclusion [of superimpositions,] are not employed in order to cognize a [supposedly still] uncognized part of the entity, because this [part has already been] perceived, and because an indivisible [[[entity]]] cannot be perceived in a partial way (ekade śena).”128 Dharmakīrti spells out the same argument in the following three stanzas: “[If] the undivided (eka) nature of an object is in itself perceptible, which other unperceived part [of it] would there be left for [further positive] investigation by the [other] means of valid cognition [i.e., by inferen ce]? [There would be none,] if another [unreal] quality were not associated [with this na ture] due to [some] cause of error, just like the aspect of silver [is associated] with a conch-shell due to oneʼs observing a similarity of colour [between them]. Therefore, all the qualities of the perceived entity are perceived, [but] due to some error, they are not deter mined. Thus one undertakes an [inferential] proof [in order to determine what the error has left undetermined].”129 To be more precise, in fe ren ces, like con ce p tual constructs and words, perform both a direct, positive (< vidhinā, vidhirūpeṇa)

127 vyavacchedaphala (PVSV 26,24); samāropavyavaccheda (PVSV 27,13; 27,14); vyavacchedakṛt (PVSV 27,10); anyavyava ccheda (PVSV 27,14); vyavacchedaviṣaya (PVSV 28,9; PV 1.56a); anyavyavaccheda viṣaya (PVSVṬ 127,10); anyasamāropavyava cchedaphala (PVSV 31,12–13); sam āro papratiṣedhaphala (PVSVṬ 124,16); bhrāntinivṛttyartham (PVSV 31,12); apohagocara (PV 1.48d; PVSV 28,19); apohaviṣa ya (PV 1.47a); anyā pohaviṣaya (PVSV 31,13). See Kellner 2004: 4–9. 128 PVSV 26,22–27,2: yāvanto ’sya parabhāvās tāvanta eva yathāsvaṃ ni mit tabhāvinaḥ samāropā iti tad vya va cchedakāni bhavanti pramāṇāni sa phalāni syuḥ / teṣāṃ tu vyavacchedaphalānāṃ nāpratītavastv aṃśa pratyāyane pravṛttis tasya dṛṣṭatvāt / anaṃśasya caikadeśena dar śa nāyogāt /. 129 PV 1.43–45: ekasyārthasvabhāvasya pratyakṣasya sataḥ sva yam / ko ’nyo na dṛṣṭo bhāgaḥ syād yaḥ pra māṇaiḥ parīkṣyate // no ced bhrān ti nimit tena saṃyojyeta guṇāntaram / śuktau vā rajatākāro rūpasādhar myadarśanāt // tasmād dṛṣṭasya bhāvasya dṛṣṭa evākhilo guṇaḥ / bhrānter niś cīyate neti sādhanaṃ sam pravartate //.

and an indirect (< arthāt), negative function.130 In its positive function, inference aims at the con ceptual determination of those aspects of the perceived parti cular that have escaped determina tion (aniścitaniścaya).131 But inference ipso facto ne ga tes the conce ptual constructs wron g ly ascribed to the perceived entity, and such is its indirect function. In this respect, in fe rence does not diﬀ er from words and concepts, which refer simultaneously to positive intel lectual constructs and indirectly exclude other, unﬁ tting constructs. It is hardly surprising, then, that Dharmakīrti repeated ly describes inference, too, as the exclusion of an o ther (anyāpoha): Inference aims at determination, but to determine amounts to holding oﬀ superimposition (samāro pa vi veka), i.e., to excluding another, superim po sed aspect. That in ference always presupposes a wrong notion is the point at stake in the follow ing discus sion: “[Objection:] The [infe rential] deter mination of [something previou s ly] un co

gni zed does not necessa rily pre sup po se a wrong notion, as [in the case of one] sudde n ly (akasmāt) knowing from [the presence of] smoke [that there is] ﬁ re [in a certain place], for in this case, the [previous] super im po sition of the absence of ﬁ re (anagni) [in this place] is not possible. Therefore, [infe rence] does not always (sarvatra) exclude [a pre vious superimposition]. [Answer:] (…) In this case too, the [[[person]]] who sees this [spot] lacks a determinate cogni tion of its nature [i.e., of this spotʼs indeed possess

130 This is made especially clear by Karṇakagomin, who regularly (e.g., PVSVṬ 124,14, 124,22, 124,24, 125,14, 125,15, 125,21, 126,9) adds vidhinā/ vidhirūpeṇa after words denoting niścaya or adhyavasāya, and arthāt after words denoting vyavaccheda, etc. Interestingly enough, close comparison with the PVṬ reveals that this is never done by Śākyabuddhi. Commenting on PV 1.45d (sādhanaṃ sampravartate), Karṇakagomin (PVSVṬ 124,21– 22) says: tanniścayārthaṃ sādhanam anumānaṃ vidhirūpeṇaiva pravartate …, whereas Śākyabuddhi (PVṬ Je D62a2/P73a4) has: sgrub pa źes bya ba ’khrul pa sel bar byed pa’i rjes su dpag pa rab tu ’jug pa yin /. For a similar observation, see Kellner 2004: 5n. 3. 131 Note, e.g., PVSVṬ 184,8–11 (with no equivalent in PVṬ): tena pra tyakṣeṇa svalakṣaṇe gṛhyamāṇe ’ni tya

tvaṃ g ṛhītam eva ke va laṃ bhrān tinimi ttasadbhāvād aniścitam / atas tanniścayamātre ’nu mā na vyāpāras / tena tan niścaya eva svalakṣaṇe ’nityatvapratītir iti siddham /.

ing ﬁ re. And] why [does he lack it]? Because of a wrong notion!132 And [insofar as] this [[[person]]] deter mines this place as free of [ﬁ re] (tadvivi ktena rūpeṇa) through a cogni tion that does not presume [by any means] that ﬁ re exists [there], how can it be said [that this person is] not mistaken (avipar yasta)? And a [[[person]]] who would neither superimpose this aspect nor doubt [the existence of ﬁ re] would [cer tain ly] not resort to an inference (liṅga) in order to know that [there is ﬁ re in this place].”133 2.2.6. We are now in a position to grasp one of the fundamental trends of Dharmakīrtiʼs philosophy. Perception provides an unmediated and unbiased access to reality, especially to the so-called vastu dharmas (impermanence, selﬂ essness, painfulness, emptiness), tho se ultimately real as pects that entities themselves cast into the conscious ness. But ignorance (qua con ceptuality and concealment) ﬁ rst has us ascribe erroneous intellec tual con structs to reality, both by unifying the many and by dividing the indivisible. Second,

ignorance (es pecially as the personalistic false view) has us fail to identify, recognize, or de ter mi ne the entities’ real aspects by superimposing contrary qualities. Now, aspects such as self, pleasure, or one’s own are the root causes of craving, appro priating, act ing and ﬁ nal ly being reborn, i.e., suﬀ ering. From this perspective, the value of inference as a cor rec ting, error-eliminating principle cannot be overestimated. In a very interesting passage in PV 3, Dharmakīrti clearly connects error, its elimination by inference, and the (yogic, i.e., Buddhist) strengthening of an (inferentially based) conceptual habitus: “Be cau se of the error that is due to the [immediate] occur

132 I.e., be cau se this person grasps this place as identical with a spot without ﬁ re. 133 PVSV 27,15–28,1: nanu nāvaśyaṃ viparyāsapūrvaka evāpratītaniścayo bhavati / yathā ’kasmād dhūmād agnipratipattiḥ / na hi tatrānagnisamāropaḥ sambhāvyate / tan na sarvatra vyavacchedaḥ kriyate / … tatrāpi taddarśinas tatsvabhāvāniścayaḥ / kutaḥ / viparyāsāt / sa ca taṃ pradeśaṃ tadviviktena rūpeṇa niścinvann agnisattābhāvanā*vimuktayā buddhyā katham aviparyasto nāma / tadākārasamāropasaṃśayarahitaś ca tatpratipattau na liṅgam anu saret /. *On bhāvanā, see Gnoli 1960 (= PVSV): 27–28n. 22. This passage has also been translated and discussed by Kellner (2004: 10–19).

rence of a new (apara) similar [phase, some one] fails to see [i.e., determine] the diﬀ erence [between two phases as long as the conti nuum is not interrupted; this person thus] lacks the [[[Wikipedia:determinate|determinate]]] knowledge of a certain [aspect like impermanence, although (s)he has grasped it perceptually (…) But if the con ti nuum is interrupted by an interval of non-existence,] it is indeed without [res ort ing to any] inference that down to a child, [any] person determines, upon seeing the rise of a new (uttara) [phase of light] disconnected [from the preceding one], that the light [of a lamp], etc., is perishable. [Or,] failing to see the eﬀ ect [of an entity] because of the inter val [implied by the causal process], an ascertainer [can also], due to dullness (apāṭa va), be mistaken with regard to [this entity’s very] capacity [to bring about its eﬀ ect,] al though it is inherent to the entity [itself]. It is in order to remove just this [kind of error] that inference is [so] minutely described. [As for] those of great understanding, they deter mi ne all aspects [of an entity] by [just] seeing [it].”134 The intimate con nec tion between inference and the search for the structure of ultimate reality and hence soterio logy is emphasized in the following statement by Dharmakīrti: “The differentiation bet ween the probandum and the probans is used by/ allows wise people to penetrate ultimate reali ty.”135 In deter min ing

134 PV 3.104ac and 105–107: kvacit tad aparijñānaṃ sadṛśā para sambhavāt / bhrānter apaśyato bhedam … // tathā hy aliṅgam ābālam asaṃśliṣṭottarodayam / paśyan paricchinatty eva dīpādiṃ nāśinaṃ janaḥ // bhā va svabhāvabhūtāyām api śaktau phale ’dṛśaḥ / anāntaryato moho viniś cetur apāṭavāt // tasyaiva vinivṛtty artham anumānopavarṇanam / vyava syantīkṣaṇād eva sarvākārān mahādhiyaḥ //. See PVP D162b6–163b5/ P189a7–190b1 and PVV 148,19–149,17. Note that both Devendrabuddhi and Mano ra thanandin ana ly ze the compound asaṃśliṣṭottarodayam as a bahuvrīhi. Whereas Devendrabuddhi does not elaborate on mahādhiyaḥ, Śākyabuddhi (PVṬ D178a6/P219b7) explains: blo gros chen pos źes bya ba ni dbaṅ po las ’das pa’i don mthoṅ ba’o (*mahādhiya ity atīndriyārthadarśinaḥ), and Manorathanandin (PVV 149,16), more con vin cing ly: mahādhiyo viparītavyava sāyānākrāntapratyakṣā yoginaḥ. 135 PV 1.86bd: sādhyasādhanasaṃsthitiḥ / paramārthāvatārāya vidvadbhir avakalpyate //. Skt. saṃsthiti is not entirely clear, but must be semantically near vyavasthāna (PV 1.85). Manorathanandin explains saṃsthiti (PV 3.214, 3.315, 3.319, 4.15, 4.64) as vyavasthā (PVV 182,25, 213,14–15, 214,22, 419,11–12, 437,3), “settlement, establishment; statute; ﬁ xed rule.” On

what had remained unidentiﬁ ed and hereby excluding wrong notions, infe rence indeed restores, still on a purely conceptual level, the most funda men tal fea tures of reality. The sequence linking the obliteration of perception and an infe rence’s corrective function is outlined by Dharmakīrti in a highly suggestive statement of PV 2: “The property of [all] cognition is to grasp an object; this [[[object]]] is grasped as it [really] is [i.e., as impermanent, etc.], and it generates this [[[cognition]] of itself] by [its] real natu re. And such is [the object’s and the cognition’s original] nature [i.e., that the ob ject ge nerates a cognition that grasps it as it really is, and that the cogni tion grasps a real aspect of the object. But] on account of another cause [i.e., on ac count of a cause of error], the [[[mind]]] shifts (skhalat) from this [inherently veracious nature, super imposing such er ro neous aspects as per ma nen ce on the object,] and becomes uncer tain, requiring a [[[Wikipedia:cognition|cognitive]]] condition for the removal [of this state], like the cognition of a piece of rope

[as a snake].”136 There is little doubt that the condition alluded to here, explained by Deven dra buddhi as “a means of valid cognition annulling error,”137 is none other than infe rence. And given the soteriological context (description of the ﬁ nal revolution of the basis, āśraya pa rivṛtti) in which this sta tement occurs, it is no less obvious that Dharma kīrti holds that this condition provides the ﬁ rst impe tus toward establishing the mind (vijñā na, i.e., perception), at the completion of the path, in its genuine radiant condition. Taking Dharmakīrti’s epistemological interpretation of the mind’s natural radiance seriously, but also his insistence on perception’s non-erroneousness and its giving access to the ultimate structure of reality, we are left with no other possibility than to hold percep tion before and after the āśrayapari vṛtti to be one and the same with regard to its content and operation. As we have seen, ignorance as “anti-knowledge” avatāra, see BHSD s.v., 71a. 136 PV 2.206–207: viṣayagrahaṇaṃ dharmo vijñānasya yathāsti saḥ / gṛhya te so ’sya janako vidya mā nā tma neti ca // eṣā

prakṛtir asyās tan nimittān tarataḥ skhalat / vyāvṛttau pratyayāpekṣam adṛḍhaṃ sarpa bu ddhivat //. See above, §2.2.2 and n. 87. 137 PVP D89a2–3/P102b5–6: rkyen la ltos pa yin te / de ltar … ’khrul pa gnod pa can gyi tshad ma la ltos pa daṅ bcas pa yin no //.

neither impedes nor obliterates perception itself, but is responsible for subsequent errors and superimpositions. The main diﬀ erence between cognition before and after the āśrayaparivṛtti, i.e., between cognition-cum-ignorance and cognition-sine-ignorance, does not pertain to perception itself, or, as Dharmakīrti himself would have it, to the nature of the mind, but to the sub se quent treatment of perceptual data. Inference is responsible for bringing out the intellectual contents that correct erroneous superimpositions; it makes determinate cogni tion possible, and further, endows the yogin with true conceptual counterparts of the en ti ties’ real aspects. In other words, inference sets the path in motion138 that will ﬁ rst enable the yogin to determine the real aspects of entities upon perceiving them,139 and then free his mind from all those adventitious factors that counteracted perception. To the best of my under

138 Note PVṬ Je D252a1–2/P299a8–b1 = PVSVṬ 401,12–13: pramāṇāny anityādibhūtākāragrā hī ṇi prati pakṣamārgam āvahanti /. 139 Note, e.g., PVṬ Je D70b4–5/P83a4–5 = PVSVṬ 142,15: yathā yogināṃ buddhipāṭavād darśanamātreṇa kṣa ṇi katvādiniścayaḥ /. That perception as such does not diﬀ er between ordinary people (pṛthagjana) and yogins is also Karṇakagomin’s opinion in two interesting statements. (1) PVSVṬ 91,24–25: yogināṃ saty api sadṛśadarśane mandabuddhitvābhāvāt kṣaṇikatvaniścayo bhavati … “The yogins do determine momen tariness because, though [their perceptual] experience is the same [as that of ordinary persons], they lack [this] being of weak intellect.” (2) PVSVṬ 92,19–21: mandabud dhir (PVSV 21,7) iti / tena bāhyā dhyā tmikavipralambhanimittasadbhāvāt pṛthagjanānāṃ [na] niśca yaḥ / yogi nāṃ tu saty api sadṛśadarśane paṭubuddhitvān niścayo bha vaty eva /. “By ‘of weak

intellec t,’ [[[Dharmakīrti]] means the following:] Because of the presence of both external [i.e., the rise of a new similar phase, etc.,] and internal [i.e., ignorance,] causes of error, ordinary persons fail to determine [[[momentariness]] in the same way as they have experienced it], but the yogins, though [their perceptual] experience is the same [as that of ordinary persons], do indeed determine [[[momentariness]]] because they are of sharp intellect.” According to Karṇakagomin, then, perception itself does not diﬀ er between those who have reached the darśanamārga and those who have not; what indeed diﬀ ers is the degree of their intellectual sharpness, the increase of which can only be due to the habitus (abhyāsa) or cultivation (bhā vanā) that comes along the path. On the context of these statements and the issue of internal as well as external causes of error, see above, n. 116; on abhyāsa as a condition for determinate cognitions to arise, see Kellner 2004: 19–32.

standing, the perception of the liberated saint is to be equated with the paramārthika pramāṇa that Dharmakīrti tou ches upon at the end of PVin 1.140 I do not intend to claim, in contrast to most scholars and the textual evidence, that Dharmakīrti’s inference has only soteriological meaning and relevance. By pointing out Dharmakīrti’s insistence upon the vastudharmas in his treatment of both perception and inference, and by putting to the fore the corrective function of inference, I would like to emphasize the fact that Dharmakīrti never lost sight of soteriology in his elaborations on epistemology. According to him, there is at least one set of cases (the most important ones indeed) in which the use of inference coincides with, or impinges upon, the precincts of the wis dom born of rational reﬂ ection (yukticintāmayī prajñā).141 The wis dom born of rational reﬂ ec tion traditionally consists (at least in connection with the socalled upa patti sā dha nayukti) in an analysis carried out on the basis of the means of valid cognition. This holds true of the Buddhist epistemologists, according to whom rational reﬂ ection basi cally aims at bringing out intellectual contents that have been thoroughly examined

and made imma cu late by means of valid cognition (pra mā ṇa pa ri dṛṣṭārtha, pra mā ṇa pari śud dhā rtha), i.e., by infe rence.142 Though still strictly concep tual in nature, the se contents (the vastu dharmas again) “co-function” as the antidote (pratipakṣa = nairātmyadarśana, etc.) to the cause of suﬀ ering, i.e., ignorance in the form of perso na listic belief. Most ordinary people may

140 PVin 1 44,4–5: cintāmayīm eva tu prajñām anuśīlayanto vibhrama vive kanirmalam anapāyi pāramārthi ka pramāṇam abhimukhīkurvanti /. On this passage, see Krasser 2004: 142–144 and Eltschinger 2005: 155–158. That liberated perception comes about through the yogin’s initially resorting to inferences is clear. How it can be equated with omni scien ce remains, however, obscure. But does not Dharmakīrti himself term “unfathomable” (acintya) the cognition of (libera ted) yogins and the Buddha’s omniscience? PV 3.532d: acintyā yogināṃ gatiḥ //; SAS 94: bcom ldan ’das kyis don thams cad thugs su chud pa ni bsam gyis mi khyab ste / rnam pa thams cad du śes pa daṅ brjod pa’i yul las ’das pa’i phyir ro //. 141 PVin 1 27,9. 142 On the cintāmayī prajñā in the Buddhist

well show no interest at all for evolving determi nate cognitions of momen ta ri ness and selﬂ essness. But to the Buddhist yogin still in the stage of being an ordinary person, inves ti ga ting the most intimate structure of reality by means of inferences is the ﬁ rst signiﬁ cant step towards the path of vision and liberation

References

Abbreviations BHSD Franklin Edgerton: Buddhist Hybrid Sanskrit Grammar and Dictio nary. Volume II: Dictionary. Delhi 1970: Motilal Banarsidass. D Jikido Takasaki/Zuiho Yamaguchi/Noriaki Hakamaya: sDe dge Tibetan Tripiṭaka bsTan ’gyur preserved at the Faculty of Letters, University of Tokyo. Tokyo 1977–1981. ms Manuscript P Daisetz T. Suzuki: The Tibetan Tripitaka, Peking Edition, Kept in the Library of the Otani University, Kyoto. Tōkyō/Kyōto 1957: Tibetan Tripitaka Research Institute. s.v. sub voce Tib Tibetan

Primary sources AK(Bh) – Prahlad Pradhan: Abhidharmakośabhāṣyam of Vasubandhu. Patna 1975: K.P. Jayaswal Research Institute (Tibetan Sanskrit Works Series 8). AKVy – Unrai Wogihara: Sphuṭārthā Abhidharmakośavyākhyā, the Work of Yaśomitra. Tokyo 1989: Sankibo Buddhist Book Store (The Publishing Association of Abhidharmakośa vyākhyā). AN – R. Morris/E. Hardy/M. Hunt/C.A.F Rhys Davids: Aṅguttara Nikāya. 6 volumes. London 1885–1910: The Pali Text Society. ASBh – Nathmal Tatia: Abhidharmasamuccayabhāṣyam. Patna 1976: K.P. Jayaswal Research Institute (Tibetan Sanskrit Works Series 17). BhK 1 – First Bhāvanākrama (Kamalaśīla). Pp. 497/187–539/229 in Giuseppe Tucci: Minor Buddhist Texts. Delhi 1986: Mo ti lal Banarsidass.

BhK 3 – Giuseppe Tucci: Minor Buddhist Texts, Part III: Third Bhāvanākra ma. Roma 1971: Istituto Italiano per il Medio ed Estremo Oriente (Serie Orientale Roma 43). HV – Hetuvidyā Section of the Yogācārabhūmi. Hideomi Yaita: Three Sanskrit Texts from the Buddhist Pramāṇa-Tradition: The Hetuvidyā Section of the Yogācārabhūmi, the Dharmottara ṭippa naka, and the Tarkarahasya. Narita 2005: Naritsan Shinshoji (Monograph Series of Naritasan Institute for Buddhist Studies 4). The Sanskrit text of the HV can be found on pp. 98/1*–124/27*. MMK – See PrP. MN I – V. Trenckner: The Majjhima-Nikāya. Vol. 1. London 1935: Pali Text Society. MS – See Lamotte 1973: I. NB(Ṭ) – Th. I. Ščerbatskoj: Nyāyabindu. Buddijskij učebnik’’ logiki sočine nie Dharmakirti I tolkovanie na nego Nyāya binduṭīkā sočinenie Darmo ttary. Osnabrück 1970: Biblio Verlag. Pā – Pāṇini (Aṣṭādhyāyī). PrP – Louis de La Vallée Poussin: Madhyamakavṛttiḥ:

Mūlamadhyamaka kāri kās (Mādhyamikasūtras) de Nāgārjuna avec la Prasannapadā Commentaire de Candrakīrti. Delhi 1992: Motilal Banarsidass. PrSVy – Pratītyasamutpādavyākhyā (Vasubandhu). D no. 3995, Chi 1b– 61a, P no. 5496, Chi 1–71a. PV 1–4 – Yūsho Miyasaka: Pramāṇavārttika-kārikā (Sanskrit and Tibetan). Acta Indologica 2 (1971–1972), pp. 1–206. See also PVV; for PV 2–3, see also PVA; for PV 1, see also PVSV; for PV 2.131cd–285, see also Vetter 1990. My numbering of the verses in PV 2 follows that of Vetter. PVA – Rāhula Sāṅkṛtyāyana: Pramāṇavārttikabhāṣyam or Vārtikā laṅkāraḥ of Prajñākaragupta (Being a Commentary on Dharmakīrti’s Pramāṇa vā rti kam). Patna 1953: K.P. Jayaswal Research Institute. PVin 1–2 – Ernst Steinkellner: Dharma kīrti’s Pramāṇaviniścaya, Chapters 1 and 2. Beijing/Vienna 2007: China Tibetology Publishing House/Au s

trian Academy of Sciences Press (Sanskrit Texts from the Tibetan Au to nomous Region 2). PVinṬ – Pramāṇaviniścayaṭīkā (Dharmottara). D no. 4229, Dze 1b1–Tshe 178a3/P no. 5727, Dze 1b1–We 209b8. PVP – Pramāṇavārttikapañjikā (Devendrabuddhi). D no. 4217, Che 1–326b4/P no. 5717, Che 1–390a8. PVSV – Raniero Gnoli: The Pramāṇavārttikam of Dharmakīrti. The First Chap ter with the Auto-Commentary. Roma 1960: Istitu to Italiano per il Medio ed Estremo Oriente (Serie Orientale Roma 23).

PVSVṬ – Rāhula Sāṅkṛtyāyana: Karṇakagomin’s Commentary on the Pra mā ṇa vārttikavṛtti of Dharmakīrti. Kyōto 1982: Rinsen Books Co. PVṬ – Pramāṇavārttikaṭīkā (Śākyabuddhi). D no. 4220, Je 1b1–Ñe 282a7/P no. 5718, Je 1b1–Ñe 348a8. Unless otherwise stated, all references to the PVṬ belong to Ñe. PVV – Rāhula Sāṅkṛtyāyana: Dharmakīrti’s Pramāṇavārttika with Com mentary by Manorathanandin. Published as an appendix to the Journal of the Bihar and Orissa Research Society 24–26 (1938–1940). SAS – F.I. Ščerbatskoj: Tibetskij pere vo’’ sočinenij Saṃtānāntarasiddhi Dharmakīrti i Saṃtānāntara si ddhi

ṭīkā Vinītadeva. Delhi 1992: Moti lal Banarsidass. TS(P) – K = Embar Krishnamacharya: Tattvasaṅgraha of Śāntarakṣita With the Commentary of Kamalaśīla. 2 vols. Baroda 1984: Oriental In sti

tute. *UH – *Upāyahṛdaya/*Prayogasāra. Giuseppe Tucci: Pre-Dignāga Buddhist Texts on Logic from Chinese Sources. Baroda 1929: Oriental Institute (Gaekwad Oriental Series 49). Vibh. – Vibhūticandra’s notes to PVV. See PVV. VY – Jong Cheol Lee: The Tibetan Text of the Vyākhyāyukti of Vasubandhu. Tōkyō 2001: The Sankibo Press (Bibliotheca Indologica et Buddholo gica 8). YBh – Yogācārabhūmi, or, followed by page/line numbers: V. Bhattacharya: The Yogācārabhūmi of Ācārya Asaṅga. Calcutta 1957: Univer sity of Calcutta.

Secondary sources Collins 1982 – Steven Collins: Selﬂ ess Persons. Imagery and thought in Theravāda Buddhism. Cambridge/New York 1982: Cambridge University Press. Eltschinger 2005 – Vincent Eltschinger: Études sur la philosophie religieuse de Dharma kīrti: 2. L’āśrayaparivṛtti. Journal Asiatique 293/1 (2005), pp. 151–211. Eltschinger 2009 – Vincent Eltschinger: Ignorance, epistemology and soteriology – Part I. Journal of the International Association of Buddhist Studies 32/1–2 (2009 [2010]), pp. 39–83. Eltschinger 2010 – Vincent Eltschinger: Studies in Dharmakīrti’s Religious Philo sophy: 4. The Cintāmayī Prajñā. Pp. 553–591 in Piotr Balcero wicz (ed.): Logic and Belief in Indian Philosophy. Delhi 2010: Motilal Banarsidass.

Frauwallner 1959 – Erich Frauwallner: Dignāga, sein Werk und seine Entwicklung. Wiener Zeitschrift für die Kunde Süd- und Ostasiens 3 (1959), pp. 83–164. Funayama 1992 – Toru Funayama: A Study of kalpanāpoḍha. A Translation of the Tattvasaṃgraha vv. 1212–1263 by Śāntarakṣita and the Tat tva

saṃ gra hapañjikā by Kamalaśīla on the Deﬁ nition of Direct Perception. Kyoto 1992: Zinbun Kagaku Kenkyūsho, Kyōto University. Funayama 1999 – Toru Funayama: Kamalaśīla’s Interpretation of ‘Non-Erroneous’ in the Deﬁ nition of Direct Perception and Related Problems. Pp. 73–99 in Shoryu Katsura (ed.): Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy. Proceedings of the Third International Dharmakīrti Conference (Hiroshima, November 4–6, 1997). Vienna 1999: Verlag der Österreichischen Akademie der Wissenschaften (Philologisch-historische Klasse, Denkschriften, 281). Jaini 2001 – Padmanabh S. Jaini: On the Ignorance of the Arhat. Pp. 167–179 (= Chapter 9) in Padmanabh S. Jaini: Collected Papers on Buddhist Studies. Delhi 2001:

Motilal Banarsidass. Kellner 2004 – Birgit Kellner: Why Infer and not just Look? Dharmakīrti on the Psychology of Inferential Processes. Pp. 1–51 in: Shoryu Katsura/Ernst Steinkellner (eds.): The Role of the Example (dṛṣṭānta) in Classical Indian Logic. Vienna 2004: Arbeitskreis für tibetische und buddhistische Studien Universität Wien (Wiener Studien zur Tibe tolo gie und Buddhismuskunde 58). Kośa – Louis de La Vallée Poussin: L’Abhidharmakośa de Vasu bandhu. 6 vols. Bruxelles 1980: Institut Belge des Hautes Études Chinoises (Mé lan ges Chinois et Bouddhiques 16). Krasser 1991 – Helmut Krasser: Dharmottaras kurze Untersuchung der Gültigkeit einer Erkenntnis – Laghuprāmāṇyaparīkṣā. Teil 2: Übersetzung. Vienna 1991: Verlag der Österreichischen

Akademie der Wissen schaften (Beiträge zur Kultur- und Geistesgeschichte Asiens 7). Krasser 1995 – Helmut Krasser: Dharmottara’s Theory of Knowledge in his Laghu prāmāṇyaparīkṣā. Journal of Indian Philosophy 23 (1995), pp. 247–271. Krasser 2004 – Helmut Krasser: Are Buddhist Pramāṇavādins non-Buddhistic? Dignāga and Dharmakīrti on the impact of logic and epistemology on emancipation. Hōrin: Vergleichende Studien zur japanischen Kultur 11 (2004), pp. 129–146. Kritzer 1999 – Robert Kritzer: Rebirth and Causation in the Yogācāra Abhidharma. Vienna 1999: Arbeitskreis für tibetische und bu ddhi stische Studien Universität Wien (Wiener Studien zur Tibetologie und Buddhismus kunde 44).

La Vallée Poussin 1913 – Louis de La Vallée Poussin: Bouddhisme, études et matériaux. Théorie des douze causes. Gand 1913: Librairie scientiﬁ que E. van Goethem (Université de Gand, Recueil de travaux publiés par la Fa cul té de philosophie et des lettres 40). Lamotte 1973 – Étienne Lamotte: La Somme du Grand Véhicule d’Asaṅga (Ma hā yā na saṅgraha). 2 vols. Louvain-la-Neuve, 1973: Uni ver sité de Louvain, Institut Orientaliste (Publications de l’IOL 8). Lamotte 1987 – Étienne Lamotte: Le troisième Bhāvanā-krama de Kama laśīla. Traduction de la version tibétaine. Pp. 336–353 in Paul Demiéville: Le Concile de Lhasa, une controverse sur le quiétisme entre bouddhis tes de l’Inde et de la Chine au VIIIe siècle de l’ère chrétienne. Paris 1987: Collège de France,

Institut des Hautes Études Chinoises (Publi cation de l’Institut des Hautes Études Chinoises 7). May 1959 – Jacques May: Candrakīrti: Prasannapadā Madhyamakavṛtti. Douze chapitres traduits du sanscrit et du tibétain, accompagnés d’une intro duction, de notes et d’une édition critique de la version tibétaine. Paris 1959: Adrien Maisonneuve (Collection Jean Przyluski 2). McClintock 2010 – Sara L. McClintock: Omniscience and the Rhetoric of Reason. Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority. Boston 2010: Wisdom Publications (Studies in Indian and Tibetan Buddhism). Mejor 2001 – Marek Mejor: Controversy on the mutual conditioning of avidyā and ayoniśomanas(i)kāra in Vasubandhu’s Abhidharmakośa. Journal of the International College for Advanced Buddhist Studies 4 (2001), pp. 49[/292]–78[/263]. Pruden 1988–1990 – Leo M. Pruden: Abhidharmakośabhāṣyam by Louis de La Vallée Poussin. English Translation by Leo M. Pruden. 4 vols. Berkeley 1988–1990: Asian Humanities Press. Ś = Swami Dwarikadas Shastri: Tattvasaṅgraha of Ācārya Shāntarakṣita with the Commentary ‘Pañjikā’ of Shri Kamal shīla. 2 vols. Varanasi 1981: Bauddha Bharati (Bauddha Bha ra ti Series 1). Schmithausen 1977 – Lambert Schmithausen: Zur buddhistischen Lehre von der drei fachen Leidhaftigkeit. Pp. 918–931 in Wolfgang Voigt (ed.): Zeit schrift der Deutschen Morgenländischen Gesellschaft, Supplement III,2 (XIX. Deutscher Orientalistentag). Wiesbaden 1977: Franz Steiner Verlag. Steinkellner 1979 – Ernst Steinkellner: Dharmakīrti’s Pramāṇa viniśca yaḥ. Zweites Ka pi tel: Svārthānumānam. Teil II: Übersetzung und Anmerkungen. Vienna 1979: Verlag der Österreichischen Akademie der Wissen schaf ten (Veröﬀ entlichungen der Kommission für Sprachen und Kul tu ren Südasiens 15).

Vetter 1990 – Tilmann Vetter: Der Buddha und seine Lehre in Dharmakīrtis Pra mā ṇavārttika. Der Abschnitt über den Buddha und die vier edlen Wahr heiten im Pramāṇasiddhi-Kapitel. Vienna 1990: Arbeitskreis für tibetische und buddhistische Studien Universität Wien (Wiener Stu dien zur Tibetologie und Bu ddhi smuskunde 12). Wayman 1961 – Alex Wayman: Analysis of the Śrāvakabhūmi Manuscript. Berke ley/Los Angeles 1961: University of California Press. Wayman 1980 – Alex Wayman: The Sixteen Aspects of the Four Noble Truths and Their Opposites. Journal of the International Association of Buddhist Studies 3/2 (1980), pp. 67–76.

The Journal of the International Association of Buddhist Studies (ISSN 0193-600XX) is the organ of the International Association of Buddhist Studies, Inc. As a peer-reviewed journal, it welcomes scholarly contributions pertaining to all facets of Buddhist Studies. JIABS is published twice yearly. As announced at the XVIth IABS Congress in Taiwan, the JIABS is now available online in open access at http://archiv. ub.uni-heidelberg.de/ojs/index.php/jiabs/ index. Articles become available online for free 60 months after their appearance in print. Current articles are not accessible online. Subscribers can choose between receiving new issues in print or as PDF. We are kindly requesting all authors that could be opposed to this decision to inform the Editors by June 2012. Manuscripts should preferably be submitted as e-mail attachments to: editors@iabsinfo.net as one single ﬁ le, complete with footnotes and references,

in two diﬀ erent formats: in PDF-format, and in Rich-Text-Format (RTF) or OpenDocument-Format (created e.g. by Open Oﬃ ce). Address books for review to: JIABS Editors, Institut für Kultur- und Geistesgeschichte Asiens, Apostelgasse 23, A-1030 Wien, AUSTRIA Address subscription orders and dues, changes of address, and business correspondence (including advertising orders) to: Dr Jérôme Ducor, IABS Treasurer Dept of Oriental Languages and Cultures Anthropole University of Lausanne CH-1015 Lausanne, Switzerland email: iabs.treasurer@unil.ch Web: http://www.iabsinfo.net Fax: +41 21 692 29 35 Subscriptions to JIABS are USD 55 per year for individuals and USD 90 per year for libraries and other institutions. For informations on membership in IABS, see back cover.

EDITORIAL BOARD

KELLNER Birgit KRASSER Helmut Joint Editors

BUSWELL Robert CHEN Jinhua COLLINS Steven COX Collet GÓMEZ Luis O. HARRISON Paul VON HINÜBER Oskar JACKSON Roger JAINI Padmanabh S. KATSURA Shōryū KUO Li-ying LOPEZ, Jr. Donald S. MACDONALD Alexander SCHERRER-SCHAUB Cristina SEYFORT RUEGG David SHARF Robert STEINKELLNER Ernst TILLEMANS Tom Cover: Cristina Scherrer-Schaub Font: “Gandhari Unicode” designed by Andrew Glass (http:// andrewglass.org/fonts.php) © Copyright 2011 by the International Association of Buddhist Studies, Inc. Print: Ferdinand Berger & Söhne GesmbH, A-3580 Horn JIABS Journal of the International Association of Buddhist Studies Volume 33 Number 1–2 2010 (2011)

Articles

William CHU The timing of Yogācāra resurgence in the Ming dynasty (1368–1643) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Vincent ELTSCHINGER Ignorance, epistemology and soteriology – Part II . . . . . . . . . . 27

Richard F. NANCE Tall tales, tathāgatas, and truth – On the “privileged lie” in Indian Buddhist literature. . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Alexander WYNNE The ātman and its negation – A conceptual and chronological analysis of early Buddhist thought . . . . . . . . . . . . . . . . . . . 103

Indian Buddhist metaethics Contributions to a panel at the XVth Congress of the International Association of Buddhist Studies, Atlanta, 23–28 June 2008 Guest editor: Martin T. Adam

Peter HARVEY An analysis of factors related to the kusala/akusala quality of actions in the Pāli tradition. . . . . . . . . . . . . . . . . . . . . . . . . 175 Abraham VÉLEZ DE CEA Value pluralism in early Buddhist ethics. . . . . . . . . . . . . . . . . . 211 Martin T. ADAM No self, no free will, no problem – Implications of the Anattalakkhaṇa Sutta for a perennial philosophical issue. . . . . . . . . . 239 Bronwyn FINNIGAN Buddhist metaethics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 Stephen JENKINS On the auspiciousness of compassionate violence . . . . . . . . . . 299 Jay L. GARFIELD What is it like to be a bodhisattva? Moral phenomenology in Śāntideva’s Bodhicaryāvatāra. . . . . . . . . . . . . . . . . . . . . . . . . 333 Tom J. F. TILLEMANS Madhyamaka Buddhist ethics . . . . . . . . . . . . . . . . . . . . . . . . 359

Miracles and superhuman powers in South and Southeast Asian Buddhist traditions Contributions to a panel at the XVth Congress of the International Association of Buddhist Studies, Atlanta, 23–28 June 2008 Guest editor: David V. Fiordalis David V. FIORDALIS Miracles in Indian Buddhist narratives and doctrine . . . . . . . . . 381 Bradley S. CLOUGH The higher knowledges in the Pāli Nikāyas and Vinaya. . . . . . . 409 Kristin SCHEIBLE Priming the lamp of dhamma – The Buddha’s miracles in the Pāli Mahāvaṃsa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435 Patrick PRANKE On saints and wizards – Ideals of human perfection and power in contemporary Burmese Buddhism . . . . . . . . . . . . . . . 453 Rachelle M. SCOTT Buddhism, miraculous powers, and gender – Rethinking the stories of Theravāda nuns. . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 Luis O. GÓMEZ On Buddhist wonders and wonder-working. . . . . . . . . . . . . . . . 513 • Notes on the contributors . . . . . . . . . . . . . . . . . . . . . . . . . . .

Source

Ignorance, epistemology and soteriology

Source

Navigation menu

Personal tools

Namespaces

Variants

Views

More

Search

Navigation

Categories

Tools