Kamalasilas Interpretation Of Non Erroneous In Definition Of Direct Perception And Related Problems

Author(s): Toru Funayama
Publisher: Toru Funayama
Catalog link: https://jainqq.org/explore/269445/1

Page #1
KAMALASILA'S INTERPRETATION OF 'NON-ERRONEOUS' IN THE DEFINITION OF DIRECT PERCEPTION AND RELATED PROBLEMS . by Toru Funayama, Kyoto I. Introduction Dignaga (ca. 480-540) defined direct perception (pratyaksa) as the cognition which is free from conception. Dharmakirti (ca. 600-660) added to it the further element 'non-erroneous (abhranta); i.e., "(The cognition which is free from conception and non-erroneous is direct perception."! 'Non-erroneous' was added for the first time in the Pramanaviniscaya. Of course Dharmakarti was not the first person who mentioned it in the history of Indian epistemology. It corresponds to avyabhicarin (non-erroneous/non-deviating) in the Nyayasutra 1.1.42 and, in Buddhist literature, to avibhranta (non-erroneous) in the Yogacarabhumi? Also, the word abhranta is found in Asanga's Abhidharmasamuccaya. Therefore Dharmakirti's addition was the readoption of what had once been dropped in Dignaga's definition. I will examine in this article some special features remarkable in the chapter Pratyaksalaksanaparikna of the Tattvasamgrahapanjika by Kamalasila (ca. 740-795). It is a commentary on the Tattvasamgraha 1212-1360 by his teacher santaraksita (ca. 725-788). They followed Dharmakirti's philosophy as the main basis of their own philosophical investigations. In the chapter concerned, Tattvasamgraha 1311-286 is the section which explains the significance of 'non-erroneous'. I want to inspect Kamalasila's explanations in connection with other texts in order to understand what kind of theoretical changes were actually developed in the pratyaksam kalpanapodham abhrantam. Alternatively, the subject and the predicate of the sentence can be reversed (i.e., "Direct perception is free from conception and non-erroneous." Funayama (1992:46:59f., n. 39]). 2 NS 1.1.4: indriyarthasannikarsorpannam jnanan avyapadesyam avyabhicari vyavasayatmakam pratyaksam II. "Perception is the cognition which is generated through the contact of the sense organ with the object, which is not expressible, which is non-erroneous and which is essentially of a determining character." YBh (Yaita (1992: 517]): praryaksam katamat/yad aviparoksam anabhyuhitam anabhyuhyam avibhrantam ca/. "What is the object of perception? (It is the entity which is not beyond the sense organ and which is neither conceived nor to be conceived." I surmise that this is one of the main sources for Dignaga's and Dharmakirti's definitions of direct perception; anabhyuhitam anabhyuhyam in YBh corresponds to kalpanapodham, though the expression is different, and avibhrantam in YBh to abhrantam in Dharmakirti's definition. - AS 152,27; Schmithausen (1965: 153f.; 1972: 154). ASBh 152,30: abhrantagrahanam alatacakramayamaricikadivyudasartham iti /. "Non-erroneous' is employed in order to negate (erroneous cognitions) such as (a vision of a circle in a (rapidly whirling) fire-brand, a phantom and a mirage." 5 Stcherbatsky (1932: 155). Jap. tr. in Ishibashi (1993). Also see Jha's translation of the whole text.

Page #2
T. Funayama eighth century. It is an attempt to shed new light on what is called by Stcherbatsky? 'The Philological School of Commentators'. I will make my intention clear by summarizing here the basic structure of the concerned section. Santaraksita opens the section by pointing out that 'nonerroneous' is necessary in the definition in order to rule out from direct perception an erroneous cognition of 'hair-mesh' (/hair-net') and the like (kesondukadijnana) (1311). Such a cognition is considered to be a perceptual error (bhranti) caused by a sense organ indriya). This idea Santaraksita owes to Dharmakirti. Much more interesting is his disciple Kamalasila's elucidation of 'non-erroneous': Kamalasila takes it to be an equivalent of 'non-deceptive' (/non-belying consistent; avisamvadin/avisamvadaka) which is found in the general definition of a means of valid cognition (pramana). As pointed out in previous studies, this equivalency is also indicated by Vinitadeva, and his idea was severely criticized by Dharmottara. This is one of the most important features of Kamalasila's theory of perception. I believe that the same idea is presupposed by Santaraksita, too. I will take up this topic below and discuss the origin and the significance of this peculiar interpretation (see Section IV). After the above-mentioned introductory stanza, santaraksita introduces and criticizes the contrary idea that a cognition of 'hair-mesh', etc., is generated by the mind, and not by a sense organ (1312-13). This idea is found in the Pramanavarttika and Pramanaviniscaya of Dharmakirti, so santaraksita's opponents must have existed before Dharmakuti and not been his real contemporaries. Discussions are then expanded in the succeeding part through a further objection and reply [1314--1322], but I can't see any essential difference from Dharmakirti's intention. Hence I suppose that stanzas 1312-22 are more or less a reproduction of Dharmakirti's arguments. Next, santaraksita proceeds to a new point (1323-28]. The following objection is raised: A cognition of a yellow conch (pitasarkadijnana) for a white conch, etc., is direct perception. Such / a cognition is certainly erroneous in terms of its color. However, inasmuch as it is a cognition of 'conch' when its color is disregarded, it realizes the efficacy of a real conch, so that it is a means of valid cognition as direct perception (1323). This objection is more minutely explained in the Panjika, according to which the propounders were certain Buddhist insiders who followed only Dignaga's definition and insisted that 'non-erroneous' should not be added, because it contradicts the cognition of a yellow conch, etc. Kamalasila's opponents refer to a passage of Pramanasamuccaya, too, and interpret the disputable word sataimira of the Pramanasamuccaya I 8b. I consider that the interpretation is based on the statement of Jinendrabuddhi's Pramanasamuccayatika. I shall discuss this below in Section V. Against this, santaraksita's own idea is stated in 1324-28. He never admits the validity of the cognition of a "yellow' conch. To santaraksita what there really is is only the color (varna) of the object; the shape (samsthana) of the object does not exist as distinct from its color. That is why there is no validity in the cognition of a yellow conch' insofar as the shape is no more than the secondary construction through erroneous color-cognition (1326). In this way, the discussions of Tattvasamgraha 1311-28 and the corresponding Panjika are Sicherbatsky (1932: 39f.). 8 PV III 294-298 and PVin I 76,26-78,5.

Page #3
Kamalasila's Interpretation of 'Non-erroneous' divided into three sections: first, the general explanation for the necessity of non-erroneous' [ 1311); cecond, the refutation of the opponents' idea who attributed all perceptual errors to conceptual construction (1312-22]; and third, the criticism to those who did not admit the necessity of 'nonerroneous' with special attention to a cognition of 'a yellow conch' for a white conch. Among them, as pointed out above, the second group does not closely reflect the actual philosophical situations of Santaraksita's age. Therefore, the other two groups are more important, as representing the actual philosophical circumstances in the eighth century. Namely, why did Kamalasila consider that 'non-erroneous' is a synonym of 'non-deceptive' and in which way was the problem of yellow conch' propounded and refuted? II. Kinds of Illusion Generally speaking, erroneous cognition is expressed by the word bhranti (error; = bhrantijnana). It is often explained by the formula 'taking non-X to be X' (atasmims tadgrahah). The Yogacarabhumi enumerates five kinds of bhranti with illustrations: 10 1. Error regarding the representation (samjnabhranti); e.g., a cognition of water for a mirage of water. 2. Error regarding the number (samkhyabhranti); e.g., a cognition of many moons for the moon by a person who suffers from timira-disease." 3. Error regarding the shape (samsthanabhranti); e.g., a cognition of a circular form in a rapidly whirling fire-brand. 4. Error regarding the color (varnabhranti); e.g., a cognition of yellow for a non-yellow object by a person who suffers from jaundice. 12 5. Error regarding the movement (karmabhranti); e.g., a cognition of moving trees by a person who is running fast with his fists firmly clenched. Other texts refer to yet other examples: e.g., a cognition of silver for a white conch, and a cognition of a snake for a rope (these will correspond to samjnabhranti in the above list); and 'travelling on a boat' (nauyana), which means that the trees on the coast seem to move in the case of a person on a fast-moving boat [karmabhranti). Yet another stock example in many texts is a cognition of hair-mesh (kesondukajnana) by a person who suffers from timira-disease. 'A cognition of many moons for the moon' in the Yogacarabhumi is essentially the same as a more popular illustration: 'a cognition of a double moon' (dvicandrajnana), very often connected with timira-disease. However, some texts state it not as a result of sickness but as a cognition which occurs to a sound person. For instance, the Fang bian xin lun states, "One sees a double moon when he presses his eye with a finger."3 An eminent Tibetan scholar Bu ston "NBh 113.3f.: yad atasminis tad iti tad 1yablicari/yat tu tasmims tad iti tad avyabhicari pranyaksam iti / (see also Schmithausen (1965: 154]). PVSV 49,3-7:... l'astutpatter abhrantir iti cet/na/atatpratibhasinas tadadhyarasayat / naniprabhayam manibhrantidarsanena vyabhicarac ca/bhranter avastusamvada iti cet/na/yathoktenaiva vyabhicarat / vitathapratibhaso hi bhrantilaksanam/. PVin II lcd: atasinims tadgraho bhrantir api... VV 138.2: atadatmani tadatmyapratitir bhrantih/ 10 Yaita (1992: 520f.: 556f). Il For timira see Tillemans (1990: vol. 1, 236 n. 156; 275 n. 370) and Kanazawa (1987: 31-41). 1? This is essentially the same as 'a cognition of a yellow conch for a white conch' in TS(P). 17 1. (T vol. 32: 25b): I. . According to the Chu san zang jiji ( F : T vol. 55: 13b), the text was translated in 472 A.D.

Page #4
T. Funayama explains it in like manner in his Pramanaviniscayatika. 14 Further, Dharmakirti and his followers noticed the distinction between two kinds of erroneous cognition: a perceptual error caused by a sense organ (indriyabhranti) and that caused by the mind (manobhranti).15 They are paraphrasable in turn by indriyaja bhrantih and manasi bhrantih. 16 Here, indriya of course signifies five kinds of sense organ (pancendriya). Manasi bhrantih belongs to conceptual cognition (vikalpajnana). III. Dignaga and Dharmakirti Dignaga's definition of direct perception, pratyaksam kalpanapodham, suggests the denotative identity between 'being direct perception' and 'being free from conception'. According to this, any cognition that is free from conceptual construction is direct perception; any erroneous cognition is of the conceptual character. As far as the definition of direct perception (pratyaksa) and that of pseudo-perception (pratyaksabhasa) are concerned, Dignaga never admitted the existence of a cognition which is free from conception but still erroneous (I call such a cognition 'sensory illusion' hereafter). In the Nyayamukha, Dignaga enumerated, as pseudo-perception, cognitions such as recollection (smarta), inference (anumana), cognition related with wish (abhilasika), doubt (samsaya), illusory cognition (bhrantijnana) and cognition of empirical reality (samvrtisajjnana), and he attributed their causes to conceptual construction. The same is true of the corresponding passages of the Pramanasamuccayavrtti. However, the situation is different in the case of the Prananasamuccaya I 7c-8b, upon which that very commentary is made: bhrantisamvrtisajjnanam anumananumanikam //7cd smartabhilasikam ceti pratyaksabham sataimiram / 8ab18 14 Bu ston's Pramanaviniscayatika 128,21. (Tosaki (1990: 62 n. 20]): mig btsir ba sogs dbang po 'gyur na 'gyur ba'i phyir te/(a commentary on PVin I 76,30 = indriyavikare ca vikarat). 15 The exact meaning of manas (the mind) is a problem in the case of the Buddhist pramana-tradition. First of all, it would not be correct to distinguish manas from manovijnana in the very strict sense, as Dharmakirti sometimes uses manas in the sense of manovijnana; see, e.g., PV III 243 and 252. A similar tendency is traced in Ratnakirti's explanation of manas in the context of manasapratyaksa in SS (17,23): atrocyate / manahsabdena tavad asmakam anaksajam vijnanam evabhipretam/... "Here we reply: First of all, the word manas signifies in our theory the very cognition that is not generated through any sense organ...." Further, it is considered to be true that the five sense organs (pancendriya) are physically existent, but manas is not; see Jinendrabuddhi's explanation of manasapratyaksa in PSTI (P 30a6f.,cf.D 26b4f.): dbang po la mi bltos pa'i phvir zhes ste ( gzugs can gyi dbang po la bltos pa med pa nyid kyi phyir zhes pa dgongs pa'o Il gang gi rten yid kho na yin gyi dbang po gzugs can ma yin pa de yid kyi mngon par brjod do II. "Because (self-cognition of desire, etc..) is not dependent on any sense organ (= PSV ad PS I 6ab: indriyanapeksanvat). The intention of this is as follows: Because it does not depend on any sense organ that is made of substances. When the basis (asraya) of (a cognition) is manas, and not any sense organ that is made of substances, (that cognition) is stated to be a mental perception (manasapratyaksa)." And last, to Dharmakirti the mind (manas) could mean even the mental cognition of the present moment (PV II 243 and 520-521). This is not the same with the scheme of the Sarvastivada school which maintains that manas as an organ is actually the mind at the preceding moment as the samanantarapratyaya, and that manovijnana is the mind at the present moment (see the Abhidharmakosa I 17ab and 44cd). 16 E.g., TSP 480,19 and 481,17. 17 T vol. 32: 3bc; Katsura (1982: 90). 18 The verse is translated in Vetter (1964: 39]: Schmithausen (1965: 213): Hattori (1965: 122: 1968: 28] and

Page #5
Kamalasila's Interpretation of 'Non-erroneous In particular, sataimira is not commented upon in the Pramanasamuccayavrtti, and no direct explanation is available in Dignaga's works. Dignaga's silence was the starting point of all the subsequent history. It was a principle of Dignaga's epistemology to explain all wrong cognitions as being caused by the mind.20 Consequently even a cognition of a double moon is considered to be produced by the mind. However, such denial of sensory illusion was not consistently held throughout his writings. First, in the context of the criticism of Vasubandhu's Vadavidhi, Dignaga admits that a sense organ is the cause of such cognitions as the representations of something blue, etc., or of a double moon, etc.21 Second, his approval of sensory illusion is found in the Alambanapariksavrtti: "A cognition of a double moon (arises) owing to the defect of the sense organ indriyavaikalyat). Though it appears as that, (i.e., as a double moon), it is not the object of the cognition)."22 Here, Dignaga seems to consider a cognition of a double moon to be an error caused by some defect of the eyes. This is not compatible with Dignaga's epistemological principle. In this way, two different views about the origin of perceptual error are found in Dignaga's works: one, his unique epistemology that every erroneous cognition belongs to conception, including a cognition of a double moon; and two, the rather commonplace idea that a cognition of a double moon is caused by some kind of sensory defect. These two attitudes were not fully integrated by Dignaga himself. In the Pramanavarttika III 294a, Dharmakirti introduces an objection: "Certain people hold an idea that the (cognition of a double moon, etc.,) is also of the conceptual character" (manasam tad apity eke). It is not clear who 'certain people' were, but the passage suggests that certain Buddhists at the period between Dignaga and Dharmakirti insisted on the utter negation of sensory illusion.23 Tosaki (1979: 382). 19 For a detailed discussion see Franco (1986). 20 Such an attitude is evident when Dignaga states, in connection with 'avyabhicarin' (non-erroneous) of the Nyayasutra 1.1.4,"Nor is there a possibility of (sense-cognition's) having an erroneous object because an erroneous cognition (necessarily] has as object an illusion produced by the mind (manobhrantivisayatvad vyabhicarinah)." Hattori (1968: 36,96 n. 1.53; 122 n. 3.6f.) and Franco (1986: 90-94). 21 L.e., te 'pi hi paramarthato 'nyatha vidyamana niladyabhasasya dvicandradyabhasasya ca jnanasya karanibhavanti/. See Hattori (1968: 35; 120 n. 2.26). The passage is cited by Dharmakiriti, Devendrabuddhi, Manorathanandin (Tosaki (1979: 389 n. 22]) and Jinendrabuddhi (PST I P 33a5f.; D 29a6f.) as expressing the evidence of Dignaga's approval of sensory illusion. 22 APV ad AP 2b (Frauwallner (1930: 176]): dbang po ma tshang ba'i phyir zla ba gnyis mthong ba ni der snang ba nyid yin du zin kyang de'i yul ma yin no ll. 23 This assumption can be corroborated by a description in NP 7,18-20: kalpanajnanam arthantare pratyaksabhasam / yaj jnanam gharah pata iti va vikalpayatah samutpadyate tad arthasvalaksanavisayanvat pratyaksabhasam/. "Conceptual cognition for an object other than the cognition itself) is pseudo-perception. A cognition arises in the person who is conceiving (the object) in the form (This is) a pot' or '(This is) a cloth'. It is pseudo-perception, because it does not have as object the particular of the thing." This is the whole explanation of pseudo-perception in this handbook of Dignaga's logic. Needless to say, it was translated into Chinese by Xuanzang. a Chinese contemporary of Dharmakirti. Therefore, it is likely that not a few people before Dharmakirti's time got the idea, from this text, that all errors are due to conception. Further, see van der Kuijp (1985: 85) for Glo bo mKhan chen's idea that manasam tad apity eke (PV IU 294a) is a reference to Isvarasena's view.

Page #6
T. Funayama At the same period, however, there must have been yet another Buddhist who did approve of the existence of sensory illusion. I surmise that Dharmapala was such a person. He commented on the above-cited passage of the Alambanapariksavstti in the following way: "Like a double moon' TAP 2b] means: As the second moon (as a real entity) cannot generate the cognition's form of the second moon. (Objection:) If that is the case, through which cause such a manifestation arise? (Reply:) 'Owing to the defect of the sense organ(Vitti ad 2b). When the eyes as organs lack their property 'clearness' (and become obscure) through damage such as timiradisease, as a result, the cognition of a double moon arises through such defective organs." This is an explanation for the process of the generation of a double-moon-cognition, where conceptual construction is not mentioned at all. In the Pramanavarttika III 288-300, Dharmakurti enumerated four kinds of pseudo-perception: three kinds of conceptual cognitions and one perceptual error which is derived from the defects of a sense organ. This fourth type is, Dharmakirti states, the exception (apavada) in Dignaga's system of epistemology. According to Dharmakarti, the word timira implies all kinds of defects of the sense organ, without specifying eye-disease;26 its derivative form taimira signifies any cognition caused by the defects of the sense organ; and therefore sataimira means 'together with taimira'. Thus, Dharmakirti interpreted the Pramanasamuccaya I 7c-8b to mean that "erroneously conceptualized cognition (bhranti), cognition of empirical reality (samvrtisajjnana), inference (anunana), its result (anumanika), recollection (smarta), and cognition related with wish (abhilasika) are pseudo-perceptions, together with (sa-) cognition caused by the defects of the sense organ (taimira)." At this period, Dharmakirti had already propounded a new definition of valid cognition, pramanam avisamvadi jnanam, but followed Dignaga concerning the definition of direct perception, pratyaksam kalpanapodham. He added 'non-erroneous' for the first time in the Pramanaviniscaya. The first emergence of this addition runs as follows: "Among them, direct perception is free from conception and non-erroneous (= PVin I 4ab). Direct perception is the cognition which is not of the conceptual character and which is not affected by the illusions owing to the troubles (viz., the malfunctions of a sense organ)" caused by timira-disease, rapid motion, travelling on a boat, and so forth.":28 What then is the difference between the opinion expressed in the Pramanavarttika and that in the Pramanaviniscaya? In the Pramanavarttika, Dharmakiti criticized the idea of attributing 24 V PA 1 (T vol. 31: 890b): A. # 97 9. T 1 854. Nii R $i & Wi & #. 25 bhranti, samvrtisajjnana and anumananumanika, etc. 26 PV III 293cd: kevalam tatra timiram upaghatopalaksanam //. 27 I tentatively follow Dharmottara's commentary on samksobha ('khrugs pa) in PVinTIP 48a7f.: rab rib dang myur du bskor ba dang grur hugs pa mams kyis bya ba'i dbang po'i 'gyur ba ni 'khrugs pa stel. Cf. NBT(Dh) 55,18 (in this commentary Dharmottara's interpreteation is quite different from that in PVinT; see also Yaita [1984: n. 34]). 28 tatra / pratyaksam kalpanapodham abhrantam (= 4ab] / timirasubhramananauyanasamksobhadyanahitavibhramam avikalpakam jnanam pratyaksam/(= PVin I 40,1-5). A similar passage is found in NB I 4-6: tatra pratyaksam kalpanapodham abhrantam // abhilapasamsargayogyapratibhasa pratitih kalpana // taya rahitam

Page #7
Kamalasila's Interpretation of 'Non-erroneous all illusions to the mind and maintained the existence of sensory illusion. As a result, he had to state that sensory illusion is the exceptional case in Dignaga's system which principally admits only conceptual illusion. Next, in the Pramanaviniscaya, Dharmakirti modified the definition by adding 'non-erroneous'. At the same time, the notion of 'exception' was deleted from Dharmakirti's own system, because it was no longer necessary to resort to such an escape after the change of definition. To summarize, I think that Dharmakirti's original intention of adding 'non-erroneous' was very simple and clear. It was postulated in order to rule out sensory illusions from direct perception. 'Non-erroneous' functioned just as a subordinate condition, attached to the main condition 'free from conception'. There was not any essential change in Dharmakirti's own epistemology itself before and after the addition of 'non-erroneous'. IV. 'Non-erroneous': Kamalasila's Idea and Its Origin Dharmakarti's followers, as commentators, paraphrased 'non-erroneous' by 'non-deceptive'. First, I will trace an example of such a paraphrase in Kamalasila's statement in the Tattvasamgraha panjika:29 "Having thus demonstrated that direct perception is 'free from conception', (the master santaraksita now) states the motif of the employment of 'non-erroneous' in the following stanza: 'Non-erroneous' is employed in order to deny (i.e., exclude from direct perception) a cognition of 'hair-mesh' and the like. (Such a cognition) is not maintained (by us) to be a (means of valid cognition because it is erroneous (1311) Here, 'non-erroneous' should be understood in the sense of being non-deceptive', and not in the sense of having the form of the object as it is placed (in cognition)'. For, otherwise, the definition of direct perception which is to be maintained on the basis of both standpoints (viz., Sautrantika and Yogacara,) would be too narrow (avyapin), because the object of cognition (as being externally existent) is not established in the Yogacara viewpoint. "Being non-deceptive' means the efficacy to realize the attainment of the object which is appropriate for the fulfillment of a desired purpose, and not the attainment itself, because there may possibly be a hindrance (to the actual attainment, and so on..0 (Objection:) If so, only 'non-erroneous' should be employed in the definition of direct perception). What is 'free from conception to be employed for? (Reply:) That is not the case, because (without that) the undesirable conclusion would follow that the inference which is essentially) the conceptual cognition would also belong to direct perception." I find four peculiarities in this statement: 'non-erroneous' is here a synonym of 'non-deceptive' (point A); 'non-deceptive' means the efficacy to realize the attainment of the object which is appropriate for the fulfillment of a desired purpose' (point B); the definition of direct perception should be applied also to the Yogacara standpoint (point C); and 'free from conception timirasubhramananauyanasamksobhadyanahitavibhramam jnanam pratyaksam II. 24 For the text see Appendix I. The notion of 'hindrance (impediment; pratibandha) comes from PV I 8 and its PVSV. 'And so on' would mean the transformation (parinama) of the causal complex (karanasamagri) which is also assumed in the same stanza.

Page #8
T. Funayama excludes all the conceptual cognitions, and 'non-erroneous' excludes a cognition of 'hair-mesh and the like (point D). When compared with other texts, Kamalasila's statement has the closest relationship with Vinitadeva's Nyayabindutika.3! He states as follows:32 "Non-erroneous means the cognition which) does not err. It is in the sense of 'non-reversed (cognition) with regard to the object of attainment (prapanavisayain praty aviparyastam ity arthah). (Objection:) Why are the two qualifiers (visesanadvaya) stated? (Reply:) First of all, 'non-erroneous' is employed in order to exclude (vyavaccheda) the erroneous cognition of the person who suffers from timira-disease (and so on), and 'free from conception is employed in order to deny inference (as belonging to direct perception) (anumananirasartham). In this way, being non-erroneous' is admitted with regard to the object of attainment (prapanavisaye) and not with regard to the object of cognition (na tv alambanavisaye). If 'being non-erroneous were maintained with regard to the object of cognition, the view of the Yogacara school (yogacaramata) would thereby be negated, for the people of the Yogacara school maintain that all cognitions are erroneous with regard to the object of cognition (as externally existent). Consequently this would lead to the undesirable conclusion that the treatise (prakarana; i.e., the Nyayabindu) was composed by negating their view. However it is maintained (by us) that the treatise was composed in conformity with the views of the Sautrantika and the Yogacara schools (isgate ca sautrantikayogacaramatanusarena prakaranarambhah) and both parties admit 'being non-erroneous', which is defined to be 'being non-deceptive', with regard to the object of attainment (prapanavisaye visamvadakatvalaksanam abhrantatvam dvabhyam abhyupagamyate). (Objection:) If that is the case, even inference would come to be direct perception, for it is also nonerroneous because it is non-deceptive. (Reply:) For this reason, 'free from conception' is employed in order to deny inference (as belonging to direct perception)...." Vinitadeva's idea corresponds to Kamalasila's except for Point B among the above-stated four specific features. In other words, there is very little originality in Kamalasila's commentary on Tattvasamgraha 1311. Point A: abhranta = avisamvadaka It is known very well that Vinitadeva's idea on the equivalency between abhranta and avisamvadaka was severely criticized by Dharmottara.35 One of the main arguments made by Dharmottara was that 'non-erroneous' as a synonym of 'non-deceptive' would become redun Vinitadeva's life should be dated to about 690-750 A.D., because his Hetubindutika is composed on the basis of the Pramanavarttikarika by Sakyabuddhi (ca. 660_720), and some of Vinitadeva's statements therein are considered to be utilized in the Herubindulika by Arcata (ca. 710-770). Funayama (1995b: 55f.: 59f. n. 4). 2 For the text see Appendix II. And so on' (adi) should be added in this context. 34 The Skt. fragment pramanyavisaye (Tatparya <9> in Appendix II) should be emended to prapanavisaye with the help of the Tibetan translation thob par byed pa'i yul la. 35 E.g., Yaita (1982: 5f.) and Ishibashi (1993:4-6). According to Dharmottara, 'non-erroneous' should be interpreted regarding the essence of the thing. PVinTIP 44b1f.: don byed nus pa'i rang bzhin la ma 'khrul pa gang yin pa de ni ma 'khrul pa yin no II. NBT(Dh) I 41,5 42,2: abhrantam arthakriyaksame vasturupe 'viparyastam ucyate/ arthakriyaksamam ca vasturupam samnivesopadhivarnatmakam / tatra yan na bhramyati tad abhrantam/.

Page #9
Kamalasila's Interpretation of 'Non-erroneous' dant, because it is the same as the general definition of a means of valid cognition.36 This argument is so persuasive and so logical that one may get the impression that Dharmottara's idea was the one and only correct interpretation and the preceding teachers had just talked about nonsense. However, seen from the point of view of the history of the interpretations, Vinitadeva's and hence Kamalasila's interpretation of the term is in a sense much closer to Dharmakirti's original descriptions. Moreover, there is a very interesting point in Kamalasila's explanation. Taking it into account that Kamalasila knew Dharmottara's critical analysis, 37 Kamalasila's implicit assent to Vinitadeva's interpretation will suggest that Kamalasila intentionally adhered to the traditionally established interpretation of 'non-erroneous' after the activity of Dharmottara. The equation of 'non-erroneous' with 'non-deceptive' was not Vinitadeva's invention.38 Though it is not so clear in the statements of Dharmakirti's disciple Devendrabuddhi and granddisciple Sakyabuddhi, the equation of the two notions had possibly been made by the end of the seventh century. It is known from one of the purvapaksas on the interpretation of 'erroneousness' in Mandanamisra's Vidhiviveka: "But what is this erroneousness'?... (Buddhist opponent:) Non-erroneousness means non-deceptiveness in terms of the fulfillment of a (desired) purpose. (Mandana:) If it meant the attainment of pleasure and the avoidance of suffering as based on the object, then the cognition of the object), the taking and rejecting of which are not possible, would come to be erroneous.40 Mandana was active around 700 A.D. during the period between Kumarila and Umbeka, and he knew Dharmakirti's philosophy. Therefore, it is highly probable that some followers of Dharmakirti at this period actually construed abhranta in the sense of avisamvadaka. 36 NBT(Dh) I 47,3-7: na tv avisamvadakam abhrantam iha grahitavyam/yatah samyagjnanam eva pratyaksam nanyat / tatra samyagjnanatvad evavisamvadakatve labdhe punaravisamvadakagrahanam nisprayojanam eva/ evam hi vakyarthah syat / pratyaksakhyan yad avisamvadakam jnanam tat kalpanapodham avisamvadakam ceti/ na canena dviravisamvadakagrahanena kimcit / tasmad grahye 'rthakriyaksame vasturupe yad aviparyastam tad abhrantam iha veditavyam/. PVinTI P 45a5f.: gzhan yang mngon sum ni yang dag pa'i shes pa'i dbye ba nyid yin no Il de la yang dag pa'i shes partog pa dang bral zhing ma 'khrul pa zhes bya ba'i ngag la/ma 'khrul pa'i sgras yang dag pa'i shes pa'i rnam grangs mi slu ba'i don'nye bar len pa ni don med pa nyid do Il. 37 I follow the conclusion of Krasser (1992). 38 In not a few cases the sub-commentators on NBT(Dh) mention Vinitadeva by name as Dharmottara's target of criticism. But some references seem to be inappropriate. For example, Dharmottara criticizes the idea of taking the relation between pratyaksam and kalpanapodham abhrantam to be samjnasamjnisambandha. PVins IP 4365: ming dang ming can gyi 'brel pa yang skabs su bab pa ma yin pa nyid de /. The author of NBTT attributes this purvapaksa to Vinitadeva. NBTT 17.4f.: anena laksyalaksanabhavam darsayata vinutadevavyakhyanam samjna. samjnisambandharupam pratyukram/. However, such a statement is not found in NBT(V). In fact, it is a reference to Sakyabuddhi's PVT III P 222b3-5:... rlog pa dang bral pa can zhes bya ba'i mtshan nyid kyis mngon sum mtshon par bya ba yin pa'i phyir ro II de nyid kyi phyir rtog (corr.: rtogs) pa dang bral pa nyid bstan nas *anudya) mngon sum nyid ston par byed pa yin te / rog pa dang bral pa'i shes pa ni mngon sum gyi ming can yin no zhes ming dang ming can gyi 'brel pa byed pa'i phyir ro II (see also Funayama (1992: 59f. n. 39]). It can be safely said in this case that Vinitadeva is wrongly considered to be a target of Dharmottara in later periods. 39 E.g., it is impossible that one sees a moon and verifies the validity of that cognition by reaching the real moon. In that case, there would be an absurd conclusion that even a cognition of a moon would be erroneous. 40 VV 137,1-138,1: ka punar iyam bhrantata /... atharthakriyasamvaditvam abhranta vam/yady arthanibandhanau sukhaduhkhapraptipariharau, upadanaparityagayogyavedanasya bhrantarvaprasangahl. 41 Schmithausen (1965: 216 n. 150) and Thrasher (1979).

Page #10
T. Funayama I conjecture that the equation of the two notions was derived from some of Dharmakirti's statements. I am not speaking about his 'real intention' which nobody can determine exactly, but about his way of describing the issue. See, for example, the following statements: "A cognition which manifests itself vividly, such as of the person who is, say, asleep or awake, does not contain any conception. Any (cognition that appears) in another way in either case is conceptual. A cognition of that (i.e., a cognition of a double moon, etc.,) is negated as invalid because it is deceptive (visamvadat), even though it is (proved to be free from conception on account of that (= vivid manifestation). And the two kinds of pseudo-perception (i.e., savikalpaka- and nirvikalpakapratyaksabhasa) were stated for that purpose (i.e., for the purpose of negating a cog. nition of a double moon, etc., as direct perception)." [PV III 299-300) "(Objection:) Then, an undesirable conclusion would follow that a cognition in a dream, which manifests itself vividly because of its non-conceptuality, would be direct perception. (Reply:) That is not the case, because I stated 'non-erroneous' in the definition of direct perception). For this very reason, it is stated that pseudo-perception is of two kinds; conceptual cognition is (pseudoperception because it does not have the manifestation of reality, and illusion (upaplava) is (pseudoperception, because it is deceptive (visamvadat)." (PVin I 76, 8-14)42 Here, Dharmakirti introduces the viewpoint 'deceptive/non-deceptive' to the context of direct perception and pseudo-perception, and seems to use 'deceptive' in place of "erroneous' and 'non-deceptive' in place of 'non-erroneous'.43 Of course he did not state that the two notions were perfectly replacable. He may have meant that wrong perception's erroneousness is testified only with recourse to the criterion whether it is deceptive or not regarding 'fulfillment of a desired purpose' (arthakriya) at a later moment, because direct perception can never be proven either to be correct or to be false at the very moment of that cognition. In short, it is not at all a surprise that Dharmakirti's expressions as mentioned above came to influence the commentators to interpret 'non-erroneous' in the sense of 'non-deceptive'.44 42 PV III 299-300: suptasya jagrato vapi yaiva dhih sphutabhasini / sa nirvikalpobhayathapy anyarhaiva vikalpika //299// tasmat tasya vikalpe 'pi pramanyam pratisidhyate / visamvadat tadartham ca pratyaksabham dvidhoditam //30011. PVin 176,8-14: 'o na ni mam par rtog pa med pa'i phyir gsal bar snang ba'i rmi lam gyi shes pa mngon sum du 'gyur ro zhe na/ma yin te / ma 'khrul pa zhes bstan pa'i phyir ro II de nyid kyi phyir / rnam rtog dngos mi snang phyir dang // nye bar bslad pa slu ba'i phyir // mngon sum ltar snang (= v. 33 abc) ram pa gnyis su brjod do II. The importance of PV III 300 and PVin I 33abc in this respect has already been pointed out in Ishibashi (1993: 4). I follow his idea in that these two passages could have been the sources for the interpretation of nonerroneous' as paraphrasable by 'non-deceptive'. +3 Note that a yogin's perception (yogipratyaksa) is also discussed just in the same way, i.e., in terms of noncenceptuality ( 4 vividness of a cognition) and non-deceptiveness. PV III 285-286: rasmad bhutam abhutam va yad yad evabhibhavyate / bhavanaparinisparrau tat sphutakalpadhiphalam // tarra pramanam samvadi yat prarnirnitavastuvat / tad bhavanajam pratyaksam istam sesa upaplavah II. PVin I 28: bhavanabalatah spastam bhayadav iva bhasate /yaj jnanam avisamvadi tat pramanam akalpakam // (Skt. fragment by Steinkellner (1972: 203]). In particular, the last stanza suggests the closeness between 'non-deceptive' and 'non-erroneous'. ++ Jinendrabuddhi also seems to interpret 'non-erroneous' in the sense of 'non-deceptive'. See the text and translation of PST in Section V: gang rtog pa dang bral zhing 'dod pa'i don la mi bslu ba de ni mngon sum ste / "That cognition which is free from conception and is non-deceptive concerning a desired object is direct perception."

Page #11
Kamalasila's Interpretation of 'Non-erroneous Point B: prapanasakti The notion of prapanasakti (the efficacy to realize the attainment of an object) was mentioned in the Pramanasiddhi chapter of Devendrabuddhi's Pramanavarttikapanjika.45 In explaining pramana as non-deceptive cognition, he says in effect that the validity of cognition consists in the efficacy to realize the attainment of an object (prapanasakti) and not the attainment itself (na tu prapanam eva). Subsequently, sakyabuddhi stated that 'efficacy' (sakti) is significant because its attainment may not be realized owing to the possible emergence of a hindrance (pratibandhasambhavat).46 Thus, Devendrabuddhi and Sakyabuddhi used the notion of prapanasakti in the context of the general definition of pramana. Kamalasila's expression abhimatarthakriyasamarthaprapanasakti was made with this background, via the equation of 'non-erroneous' with 'non-deceptive' in the context of direct perception. Further, his appreciation of this notion might have been influenced by Dharmottara who emphasized it shortly before his activities. 47 Point C: Sautrantika-Yogacara Problem Vinitadeva and Kamalasila held the common view that the definition of direct perception was made from the standpoint of both the Sautrantika and Yogacara schools. The reason was - their belief that Dharmakirti composed his works not only as a Sautrantika but also as a Yogacarin. Vinitadeva expressed this idea in his commentary on the Nyayabindu, but we should not limit his idea only to this handbook, because it was criticized by Dharmottara as a problem of interpretation of the Pramanaviniscaya. Therefore, this is a problem relating to the whole work of Dharmakarti, namely, in what sense was he a scholar of what is called the Sautrantika-Yogacara school. The definition of direct perception is the basis for all the epistemological discussions and in some cases Dharmakirti shifts his position to the Yogacara standpoint, 48 nevertheless, he does not speak about any change in the definition. This is presumably the reason for the arguments of Vinitadeva and Kamalasila. On the other hand, Dharmottara criticized this, and propounded that the definition issues only from the Sautrantika standpoint. Dharmottara maintains that Dharmakarti was indeed a scholar of the Yogacara school, but his ideas were not always expressed from that standpoint; for example, Dharmakirti's statements, such as "Direct perception is also a means of valid cognition, because it does not exist when its object does not exist" (PVin I 3ab), are expressed only from the view of assuming the existence of external objects (*bahyarthanaya).49 Dharmottara's claim is understandable enough, but it leads to the following question: Did 45 PVP II P 3a8-b2, D 3a2f. The passage is translated into Japanese in Inami (1993: 92f.). 46 PVT II P 88b7-89a1, D 73a7-62. The passage is translated into Japanese in Inami [1993: 98). +7 For Dharmottara's theory of prapanasakti see Krasser (1991: vol. 1, 54, 1-5; 106, 4-6; vol. 2, 7f.; 1995: 264). 48 E.g., PV III 320ff. and PVin 1 841ff. "DhP 44,17-20: nanv evam abhrantane yogacaramatam asamgrhitam spat/grahyagrahakakarataya pravstteh sarvasyaivasarvajnavijrianasyalambane bhrantavat / tat katham purvavyakhyana vajna na kriyata iti cet/ucyate / na yogacaranaye laksanam idam/kim tu sautrantikanaya eva /na ca sarvan vijnanavade yojayitum sakyam/ tasya visavah svalaksanam (NB I 12) irydder asakyayojanatat /. NBTT 19,10f.: nanuktam yogacaramatam asamgrhitam syrad iti/ucyate / bahyanayena (sic!) sautrantikamatanusarenacaryena laksanam krtam ity adosah / PVinTIP 44b8-45a3: gshan dag mdo sde pa bzhin du mal 'byor spyod pa'i gzhung bsdu ba'i don du mi slu bar byed pa la ma 'khrul par shes pa de ni mi rigs soll gang gi phyir slob dpon gyis ma 'khrul pa'i sgrai don de Ita bur gyur pa yin pa

Page #12
84 T. Funayama Dharmakirti indicate his Yogacara position without the special definition of direct perception in that position, if pratyaksam kalpanapodham abhrantam is limited to the Sautrantika standpoint? Unfortunately I have not found any concrete answer to this problem. Point D: The Roles of Two Qualifiers According to Vinitadeva and Kamalasila, the qualifier 'free from conception' excludes all conceptual cognitions, and the other qualifier 'non-erroneous' excludes a cognition of 'hairmesh', etc. This idea is also criticized by Dharmottara who maintains that the wrong ideas (vipratipatti) on the definition are negated by a set of two qualifiers.50 Kamalasila presents further viewpoints regarding the function of the two qualifiers. He introduces the objection that 'non-erroneous' is employed in order to exclude a cognition in a dream from direct perception, and he gives his assent to this idea.51 It is remarkable here that a cognition in a dream is considered to be free from conception but still erroneous, just like a cognition of a double moon, etc. Similar ideas can be traced in Dharmakirti's works and Sakyabuddhi's commentary.52 After the commentary of the last stanza (1360) of the chapter Pratyaksalaksanapariksa, Kamalasila adds a discussion, independently of Santaraksita's statement. It is an excursus on the definition of direct perception in general. Kamalasila therein examines the opponent's idea that 'non-erroneous' is not necessary in the definition because whether an indeterminate cognition (e.g., a cognition of water) is erroneous or not cannot be determined at the very moment of cognition; for it is ascertained by the criterion 'fulfullment of a desired purpose' (e.g., quenching the perceiver's thirst) only at a later moment. This viewpoint reveals Kamalasila's idea about the relationships between direct perception (pratyaksa), impression (vasana), cognition of ascertainment (niscayajnana) and action (pravstti). It seems to be connected with the ideas of Devendra ni mi bzhed (D:gzhed PJ de / 'khrul pa'i mtshan nyid ni phyin ci log tu snang ba yin no zhes de dang der gsungs pa'i phyir ro Il dngos po mi snang ba'i phyir rnam par rtog pa ni mngon sum ltar snang ba'o // (PVin 1 33ac) zhes gsungs na yang /mngon sum ni dngos po snang ba don byed pa'i dngos po'i rang bzhin la ma 'khrul pa yin par brjod pa yin no Il. Ibid. 45a6-b1: gal te rnal 'byor spyod pa'i gzhung yang thams cad du rjes su "jug par byed pa yin no ll slob dpon gyis ries su dpag pa nvid szung ba la phvin ci log pa nvid vin no (unidentified) zhes bya ba dang/don gyi mthu stobs kyis skye bas rios pa med pa nvid du (PVin 1 4cd) gsungs par ci'i phyir rjes su 'jug par mi byed / de'i phyir mam rig tsam du smra ba ni slob dpon bag chags brtan pa dang mi brtan pa dag gis rnam par 'jog pa mthong la/ 'di ni don med na mi 'byung ba las (Ibid. 3ab) zhes de la bu la sogs pa gsungs pa na/phyi rol gyi don gyi tshul nyid bzhed pa yin no II de'i phyir sngar bshad pa nyid ma 'khrul pa'i don yin no ll. For an exposition of the passage see Yaita (1982: 6). 50 E.g., PVinTI P 44a7f.: 'dir yang mtshan nyid gzhan rnam par bcad pa ni ma yin gyi l'on kyang log par rtogs pa (*vipratipatti) bsal ba'i don yin no II, as well as NBT(Dh) 1 44,1 and 46,6f. For this topic in detail see the previous studies in n. 35. SI TSP 480,12-15: nanu ca bhavatu nama manasam tathapy abhrantagrahanam kartavyam eva / na hy anenendriyajnanasyaiva (anena MKS: om. Tib.) pratyaksalaksanam kartum arabdham/kim tarhi/manasasyapi yogijnanadeh* / tatra ca svapnanrikasyapi nirvikalpakarvam asti / spastapratibhasitvat / na tv abhrantanam iti tannivrityartham abhrantagrahanam yuktam eva/satyan etat /... *The opponent here assumes yogin's perception to be mental perception. This is an idea of santaraksita and Kamalasila, too; e.g., TSP 484,19: yad va yogijnanam iha manasam. For a previous study on this problem see Hattori (1968: 93f. n. 1.46). 52 For Dharmakirti see PV and PVin in n. 42; for sakyabuddhi see Inami (1993: 102). 53 TSP 492,16-494,20.

Page #13
Kamalasila's Interpretation of 'Non-erroneous' buddhi and Sakyabuddhi on the general definition of pramana,54 as well as Kamalasila's own discussion developed in the opening section of the Tattvasamgrahapanjika. This interesting issue will need further examination. V. The Relationship between Jinendrabuddhi and Kamalasila Jinendrabuddhi's Pramanasamuccayatika is an indispensable source for the study of the eighth century developments of the pramana school. In a previous paper I have assumed him to be a contemporary of Santaraksita.55 The assumption that the Pramanasamuccayatika was written before Kamalasila's composition of the Tattvasamgrahapanjika holds good in the following examinations, too. I will show further relations between Devendrabuddhi, Jinendrabuddhi and Kamalasila. 85 The pratyaksabhasa-section of Jinendrabuddhi's Pramanasamuccayatika56 has connections with other texts. First of all, it is evident that some of its passages are identical with those of Devendrabuddhi's Pramanavarttikapanjika." This means that Jinendrabuddhi composed this section of the work on the basis of Devendrabuddhi's commentary. Second, some passages of Jinendrabuddhi on the interpretation of sataimira (PS I 8b) are identical with Kamalasila's passages. Kamalasila introduces the opponents' idea, which is wellknown among modern scholars, 58 in the following way: kecit tu svayuthya evabhrantagrahanam necchanti/bhrantasyapi pitasankhadijnanasya pratyaksatvat/tatha hi///// (TSP 482,24-483,13 ad TS 1323),60 "On the other hand, certain Buddhist insiders do not wish to employ 'non-erroneous' (in the definition of direct perception), because (they consider that) a cognition of a yellow conch (for a white conch in reality), etc., is direct perception, even though it is erroneous. Namely, it is not inference because it is not generated from an inferential mark>; it is a means of valid cognition because it is not deceptive. For this very reason, (they state,) 'non-erroneous' is not employed in the master Dignaga's definition. Because pseudo-perception is indicated by the statement 'Erroneous cognition, cognition of empirical reality, inference....61 such a kind of (definition) as '(cognition which is) non-deceptive and free from conception' is maintained to be the definition of the master. Page #14
86 T. Funayama the other hand, sataimira (of PS I 8b) means as follows: the word timira is a synonym of 'ignorance b> 62 163 Note that this is a purvapaksa for Kamalasila. In the subsequent discussions Santaraksita and Kamalasila criticize this idea. As I see it, Kamalasila is here referring to Jinendrabuddhi's own idea in his Pramanasamuccayatika, so that certain Buddhist insiders' signifies Jinendrabuddhi and his followers. Now I will show this in detail. Jinendrabuddhi presents two kinds of interpretation of the word sataimira of the Pramanasamuccaya I 8b. His first interpretation is that the word means the cognitions caused by the defect of a sense organ, such as timira-disease. 64 This is based on Dharmakirti's interpretation in Pramanavarttika III 29365 and the above-cited Pramanavarttikapanjika by Devendrabuddhi. After this, Jinendrabuddhi introduces Dharmakirti's notion of avisamvadaka66 and refers to seyeral kinds of cognition which can be controversial in terms of this notion: (a) cognition which has the manifestation of 'hair-mesh', etc., owing to timira-disease; (b) cognition of a yellow conch for a white conch by the person whose eyes are affected by jaundice; (c) cognition of the object as having the non-vivid blue form, etc., for a vivid blue entity caused either by the perceiver's eyes being affected by timira-disease for a long time or by his viewing the object from afar; and (d) cognition in which the trees are manifested as being possessed of movement, the movement being wrongly superimposed upon those stationary trees as a result of trouble (in a sense organ) caused by the perceiver's travelling on a boat.07 Thus, making much of the notion of avisamvadaka in the context of direct perception, Jinendrabuddhi develops a special discussion on the cognition 62 Panini 4.3.53: tatra bhavah // (=Siddhantakaumudi $1428. An affix comes after a word in the 7th case in construction, in the sense of 'who stays there'). 63 A similar expression is found in Dharmottara's statement, too. PVinTI P 45b1-3 (Yaita (1984:25]): kha cig zla ba gnyis pa la sogs pa'i shes pa yang cha 'ga' zhig la mi slu ba des na tshad ma yin pa de'i phyir de bzlog pa'i don du ma 'khrul pa smos pa ni rigs pa ma yin te/de nyid kyi phyir slob dpon phyogs kyi glang pos kyang ma mdzad pa yin no ll ... de yang rigs pa ma yin te l. Cr. TSop 280,10-13: abhrantagrahanenapi suklasankhadau pitasarkhadivijnanam nirasyate / saty api bhrame 'rthakriyavisamvadabhavat / napi tad anumanam yujyate 'lingajarvat/atah pratyaksam iti. (This is probably based on Kamalasila's statement.) 64 PSTI P 32a7f.:D 2862: rab rib bcas zhes pa 'dis dbang po la nye bar gnod pa las skyes pa rab rib la sogs pa'i shes pa (indriyopaghatajam rimiradijnanam) mngon sum ltar snang ba bzhi pa gsungs so zhes pa'o II, elucidated by Hattori (1968: 95 n. 1.53). See also Appendix III. 69 PV III 293: apavadas caturtho 'tra tenoktam upaghatajam/kevalam tatra timiram upaghatopalaksanam II. 66 PST IP 33a7f.:D 29b1f.: 'dir brtsad par bya ste [V] 'dir bslu [P:slu D] ba med pa la 'jug par byed pa'i shes pa ni tshad mar (D: ni added P]mngon par 'dod de/'di dag gis don jongs su bcad nas 'jug pa ni don bya ba la bslu [P:slu D] ba med pas so* zhes gsungs pa'i phyir ro Il. *The citation is PVin I 30,17f. = na hy abhyam artham paricchidya pravartamano 'rthakriyam visamvadyate. 67 PSTIP 33b1-4;D 29b3-5: de'i phyir gang 'di skra shad la sogs par snang ba rab rib la sogs pa'i shes pa ni gang si yang dag par 'dod pa skra shad la sogs pa'i dngos po la mi [P: om. D) bslu [P:slu D] ba med cing de tshad ma nyid du mi 'gyur mod/ gang 'di mig ser gyis nye bar bslad pa'i mig can gyi (P:gyis D] dung dkar po la sogs pa dung ser po la sogs par snang ba dang/gang yang yun ring du rab rib kyis mig bsgribs pa nyid kyi phyir rum thag ring ba nyid kyi phyir gsal bali sngon po la sogs pa mnams la mi corr.: mig PD] gsal ba' sngon po la sogs pa'i rnam pa dang/gang yang gru dang bzhon pa bskyod pas byas pa'i 'khrul pa shing sngong la sogs pa gnas pa rnans la yang 'gro ba'i bya ba dang Idan pa'i rkang 'thung la sogs par snang ba dang 'di Ita buli rnam pa gzhan yang /

Page #15
Kamalasila's Interpretation of 'Non-erroneous' of a yellow conch (b) in the following manner: ... de'i phyir de mams kyang yang dag par 'dod pa'i don bya ba nus pa dung la sogs pa'i dngos po tsam la gang du mi bslu ba de la tshad ma nyid du mi 'gal bar mthong zhing / smra bars nus te / shes pa gang la rab tu zhugs pa'i skyes bus gegs med na nges par mngon par 'dod pa'i don sgrub pa de de la de'i tshad ma ste/dper na mngon par 'dod pa'i mngon sum dang rjes su dpag pa bzhin (ji skad bshad pa'i shes pa las rab tu zhugs pa'i skyes bu yang gegs med na nges par mngon par 'dod pa dung la sogs pa'i dngos po tsam sgrub par byed do Ilzhes pa rang bzhin no // /...rab rib can76 de dang Ihan cig 'jug go zhes pa ni rab rib bcas pa ste (U] Page #16
T. Funayama "... Therefore, if these (cognitions of a yellow conch, etc.,) are also non-deceptive regarding a certain (object) such as the thing just as a conch, which has the efficacy to realize the desired purpose, they are understood without contradiction to be the means of valid cognition for that (aspect of the object). (The same) can be stated as follows: (Major premise:) If the person who acts on the basis of a cognition (of an object) necessarily attains a desired object when there is no hindrance (pratibandha), such a cognition is for him the means of valid cognition for that (object), just like any direct perception and inference. (Minor premise:) The person who acts on the basis of the cognition as stated above necessarily attains a desired (object) such as the thing just as a conch when there is no hindrance. (Conclusion: A cognition of a yellow conch, etc., is the means of valid cognition for the conch, etc.) This is a svabhava(hetu inference). This (cognition of a yellow conch) is not an inference, because it is not generated from an inferential mark (na caitad anumanam/alingajatvat / and because it is not conceptual cognition. A third means of valid cognition is not admitted, either. Therefore, it should be admitted that the cognition is certainly direct perception. (Major premise:) That cognition which is free from conception and is non-deceptive concerning a desired object is direct perception, just like any direct perception. (Minor premise:) So is the cognition of the present topic. (Conclusion: A cognition of a yellow conch, etc., is direct perception.) (This is) a svabhava(hetu inference). One may assume as follows: (A cognition of a yellow conch, etc.) is maintained to be the means of valid cognition for the very object because (the cognition) is non-deceptive regarding the object inasmuch as it is intended to be the thing just as a conch, etc.; it is not maintained to be nondeceptive concerning (the object's) specific attribute (visesa) such as having the form of yellow, etc. (pitadyakara). Therefore, the following should not be stated: "By the word sataimira (which shows) the exception (apavada) (in Dignaga's system), the cognition whose sense organ becomes injured through external and internal defects as the cause (pratyaya) is stated to be pseudo-perception; and timira in the word sataimira implies (upalaksana) all the cognitions that are caused just by the defects of sense organs.79 (This statement of Dharmakirti and his followers is not correct) because in any case any indeterminate) cognition whose sense organ is defective is direct perception. Moreover, the qualifier (visesana) 'non-erroneous' (abhranta) should not be added in the definition), because in any case any indeterminate) cognition, even though it is erroneous, is maintained to be direct perception. (Reply:) Therefore, the word sataimira (which shows) the exception (in Dignaga's system) is explained in a different way. (Namely,) Page #17
Kamalasila's Interpretation of 'Non-erroneous' 89 bhavam taimiram) 80 (...)81 Sataimira is that which occurs along with taimira.82 taimira > sataimira is taken into account. Cf. Kamalasila's parallel wording: timire bhavam taimiram/visamvadakam ity arthah /, where the grammatical explanation for sa- is omitted. For the grammatical explanation of the compound sataimira see Panini 2.2.28: tena saheti tulyayoge // (= Siddhantakaumudi SS848. The word saha 'together' is compounded with a word ending with the third case-affix and the compound is Bahuvrihi; provided that, the companion and the person accompanied are equally affected by any action or thing, in the same manner) and Panini 6.3.82: vopasarjanasya // (= Siddhantakaumudi $849. Sa is optionally the substitute of saha, when the compound is a Bahuvrihi). 8.3 PV III 293cd: kevalam tatra timiram upaghatopalaksanam //.

Page #18
90 T. Funayama 1. Sataimira indicates the 'exception' (apavada) in Dignaga's system. It is a reference to a special kind of pseudo-perception, in addition to the explicitly enumerated three kinds of pseudo-perception (viz., [a] bhranti, [b] samvrtisajjnana and [c] anumana, anumanika, smarta and abhilasika). However, there are two possible interpretations concerning its meaning. 2. According to one interpretation, taimira signifies that any cognition that is caused by the defect of a sense organ is implied in pseudo-perception. The whole stanza (PS I 7c-8b) means as follows: "Erroneous cognition (= erroneously conceptualized cognition), cognition of empirical reality, inference, its result, recollection and cognition related with wish are pseudo-perceptions, together with (sa-) cognition caused by the defects of a sense organ (taimira)." [This is the same with Dharmakirti's interpretation.] 3. According to the other interpretation, sataimira signifies that any deceptive cognition is pseudo-perception. It is the exception in Dignaga's system, since the notion 'deceptive/ non-deceptive' was not used by him. Hence the stanza means as follows: "Erroneous cognition, cognition of empirical reality,... and (ca) any deceptive cognition (sataimira) are pseudo-perceptions." [This is a modification on the basis of Dharmakirti's notion of non-deceptiveness.) Jinendrabuddhi's idea would not reflect the original intention of Dignaga. It is only a product of the integration of Dignaga and Dharmakirti's theories. The supposition that Jinendrabuddhi's own idea, as seen above, is the purvapaksa in Kamalasila's text would be important on several points. In the first place, there is no room to doubt Jinendrabuddhi's historical precedence to Kamalasila.84 Word-for-word correspondences in the five passages between their texts clearly show that Kamalasila is indebted to the Pramanasamuccayatika and nothing else. It also reveals Kamalasila's criticism of Jinendrabuddhi. As far as I know, Vinitadeva's Nyayabindutika does not mention a cognition of a yellow conch as a problem in the definition of direct perception. In this respect, the cognition of a yellow conch may have been taken up as an important topic about the middle of the eighth century. It was a special problem raised in the context of direct perception when people began to interpret Dignaga's wording sataimira by using Dharmakirti's notion of non-deceptiveness. Those people who maintained the validity of a cognition of a yellow conch, etc., are called by Durvekamisra Amsasamvadavadin or 'propounders of the partial agreement/non-deceptiveness.85 It should be noted here that Jinendrabuddhi's discussion on this point was constructed on the basis of the somewhat astonishing thesis that a cognition of a yellow conch is non-deceptive. Therefore, his second interpretation of sataimira no longer holds when this thesis is denied. This is the reason why Kamalasila introduced Jinendrabuddhi's commentary only as a purvapaksa. Santaraksita, Kamalasila and Dharmottara do not think such a cognition is direct perception as pramana, 86 so that to them only Dharmakirti's interpretation of sataimira is correct. 84 Here I mean by Kamalasila the author of the Tattvasamgrahapanjika. The text is considered to have been composed between ca. 760-780/790. Schmithausen [1965: 216 n. 150] and Krasser [1992: 157]. 85 Tosaki [1984: 489f. n. 12] and Krasser [1991: vol. 2, 73-75]. 86 TSP 483,14-19 (ad TS 1324f.): pramanyam hi bhavad dvabhyam [MK: dvabhyasam S] akarabhyam bhavati/yathapratibhasam avisamvadad yathadhyavasayam va/tatreha na yathapratibhasam avisamvadah/pitasya pratibhasanat tasya tathabhutasyaprapteh [tatha- MS: yatha- K]/napi yathadhyavasayam avisamvadah/pitasyaiva

Page #19
Kamalasila's Interpretation of 'Non-erroneous' 91 Thus it is certain that the Pramanasamuccayatika was written at the period between Devendrabuddhi's Pramanavarttikapanjika and Kamalasila's Tattvasamgrahapanjika. This will raise another question: Did Jinendrabuddhi precede Santaraksita or vice versa? The problem of the cognition of a yellow conch is discussed in Tattvasamgraha 1323-28: "(Purvapaksa:) Other people hold an opinion that a cognition of a yellow conch, etc., is the means of valid cognition (as direct perception), even though it is erroneous, because it is not deceptive in terms of the fulfillment of a purpose.87 [1323] (Uttarapaksa:) This is not true, (because) also in the above case there is not such a realization of the fulfillment of a purpose as representing the conceptually determined appearance (i.e., yellow). Otherwise this fallacy of over-extension would follow: Since (even) a cognition which has the manifestation of 'hair-mesh', etc., is (certainly) non-deceptive concerning light and the like, such a cognition would irresistibly come to be the means of valid cognition! (However this is not the case, of course). [1324-25] Further, it cannot be assumed that a shape exists as distinct from its color.88 And non-deceptiveness is not found in the manifesting color (i.e., yellow). If (opponents) assume, independently of the shape (of the object), the validity of cognition on the basis of non-deceptiveness in terms of the fulfullment of a purpose, then it would be surely contradictory to all the statements (of the master Dignaga), such as 'because the ascertainment of the object conforms with that (i.e., the manifestation in cognition)'89 and so forth.90 Therefore, non-deceptiveness comes about, as generated through the cause which ripens the subconscious impression."'91 [132728] Santaraksita's idea is possibly a more developed one than that of Jinendrabuddhi, because Santaraksita shares the idea in common with his younger contemporaries in not accepting a cognition of a yellow conch as direct perception, while Jinendrabuddhi accepts it. This seems to suggest that Jinendrabuddhi was an elder contemporary of Santaraksita. visistarthakriyakaritvenadhyavasayat/na ca tadruparthakriyapraptir asti/na canadhyavasitarthavisamvadenapi pramanyam/atiprasangat/kesadijnane 'pi hy anadhyavastalokadiprapteh/. For Dharmottara see Yaita [1984: 2327] and Krasser [1991: vol. 2, 73-75]. 87 It would be a problem whether this puryapaksa is a reference to Jinendrabuddhi's own idea. This point is related with another problem as to whether Jinendrabuddhi's purvapaksa in the same context ('One may assume as follows:...) is a hypothetical one or based on the existence of those who maintained such a purvapaksa. Santaraksita mentions here just the validity of the cognition concerned, while Kamalasila positively relates it with Jinendrabuddhi's interpretation of PS I 8b. 88 This is a statement from the Sautrantika viewpoint in contrast to the view of the Sarvastivada school which admits the existence of both samsthanarupa and varnarupa (Mimaki [1972: 86]). 89 PS 19b: tadrupo hy arthaniscayah. See Hattori [1968: 103 n. 62]. 90 Kamalasila refers to PSV ad PS I 9d as an example of 'and so forth' (TSP 483,24f.). 91 pitasankhadibuddhinam vibhrame 'pi pramanatam [K: pramanata MS] / arthakriyavisamvadad apare sampracaksate //1323// tan nadhyavasitakarapratirupa na vidyate / tatrapy arthakriyavaptir anyathatiprasajyate // 1324// kesadipratibhase ca jnane samvadabhavatah [S: samvadi- K]/alokader atas tasya durnivara pramanata // 1325//na varnavyatiriktam ca samsthanam upapadyate/bhasamanasya varnasya na ca samvada iksyate [S: isyate K (M is not clear)] //1326// yady akaram anadrtya pramanyam ca prakalpyate/arthakriyavisamvadat tadrupo hy arthaniscayah //1327// ityadi gaditam sarvam katham na vyahatam bhavet / vasanapakahetutthas tasmat samvadasambhavah //1328//

Page #20
T. Funayama I don't think that Jinendrabuddhi flourished far before santaraksita (ca. 725-788 A.D.), because both authors were possibly influenced by Arcata's theory on the aim of a treatise (prayojana) in his Hetubindutika.92 As for the date of the composition of this text, I would like to assume it to be approximately 740-750 A.D.; Arcata is considered to be a master of Dharmottara (ca. 740-800) so that his life would be about 710-770 A.D. On the other hand, Santaraksita's Tattvasangraha is considered to have been written about 760 A.D.93 Jinendrabuddhi's Pramanasamuccayatika would have been composed slightly before the Tattvasamgraha. This is related with another problem as to whether our Jinendrabuddhi is identical with the grammarian Jinendrabuddhi who was also a Buddhist 94 and who was the author of the Nyasa, a commentary on the Kasikavrtti. Prof. Richard P. Hayes pointed out the possibility of such identity.% Prof. Nobuhiko Kobayashi pointed out that the Nyasa-kara who is mentioned in Bhamaha's Kavyalamkara is the grammarian Jinendrabuddhi, and that three stanzas of Bhamaha's same work are cited (TS 911-13 on apoha theory) and criticized in the Tattvasamgraha.97 These convincingly demonstrate that Jinendrabuddhi the grammarian flourished first, then Bhamaha quoted his passages, and the latter was in another context attacked by Santaraksita. If I am allowed to conjecture that the Nyasa-kara is identical with the Tika-kara and that the composition of the Nyasa was done bofore the Pramanasamuccayatika, Jinendrabuddhi's life would possibly be about 710-770 A.D. with some margin of error. However, this consequence is not compatible with the previous studies which place Bhamaha before Dharmakirti.98 At any rate, this important issue will have to be examined further from every possible viewpoint in order to reach a final settlement. VI. Concluding Remarks Dharmakirti added 'non-erroneous' (abhranta) to Dignaga's definition of direct perception. He intended to rule out sensory illusion (indriyabhranti, indriyaja bhrantih) from direct perception and enumerated it as a kind of pseudo-perception. Dignaga's inconsistent dealings with a cognition of a double moon guided this interpretation of Dharmakirti. In the context of direct perception, 'non-erroneous' was construed in the sense of 'nondeceptive' by Vinitadeva and Kamalasila, and possibly by Jinendrabuddhi, too. I suppose it was a general tendency of interpretation among the people who belonged to 'Philological School. It was Dharmottara who first criticized such an interpretation and opened a new phase of epistemology 92 This is one of the conclusions of Funayama (1995a). 9 Frauwallner (1961: 143) and Krasser (1992: 157). 94 He is called bodhisattvadesiyacaryasrijinendrabuddhipada in the colophone of the Nyasa. 95 Hayes (1983: 716). 46 Kobayashi (1978) 97 Kobayashi (1965). 98 E.... Tucci (1930). Cf. According to Kobayashi (1978: 470], Jayaditya, one of the two authors of the Kasikavrti which Jinendrabuddhi commented upon, is stated to have died in ca. 660 A.D. by Yijing in his Nan hai ji gui nei fa zhuan, so that Bhamaha flourished possibly after Dharmakirti, though the latter's idea is not mentioned in the Kavyalamkara. On the ohter hand, the reliability of Yijing's statements is doubted by Brough (1973: 255f.).

Page #21
Kamalasila's Interpretation of 'Non-erroneous' Jinendrabuddhi interpreted the word sataimira of the Pramanasamuccaya I 8b in the following two ways: His first interpretation is based on that of Dharmakirti and Devendrabuddhi, and his second interpretation was motivated by the idea of some of his predecessors' idea that a cognition of a yellow conch is direct perception. In this article, I also examined the possibility that Jinendrabuddhi flourished shortly before santaraksita. Kamalasila constructed his own interpretations of 'non-erroneous' by using preceding texts: Vinitadeva's Nyayabindutika as being agreeable with Kamalasila's main standpoint, Jinendrabuddhi's Pramanasamuccayatika as being his purvapaksa, and probably the works of Dharmottara, to say nothing of the master Santaraksita's teachings. In this respect, Kamalasila's statement is weak in originality. Its importance would lie in the point that it is a well-systematized theory which reflects several differenct phases and actual debates in the eighth century. Appendix I: Kamalasila's Interpretation of abhranta in the Tattvasamgrahapanjika evam kalpanapodhatvam pratyaksasya prasadhyabhrantagrahane prayojanam aha kesonduketyadi / . kesonduka'divijnananivrttyartham idam krtam/ abhrantagrahanam tad dhi bhrantatvan nesvate prama //1311// abhrantam' atravisamvaditvena drastavyam/na tu yathavasthitalambanakarataya102 / anyatha hi yogacaramatenalambanasiddher ubhayanayasamasrayenestasya pratyaksalaksanasyavyapita syat / avisamvaditvam cabhimatarthakriyasamartharthaprapanasaktih 103 / na tu prapanam eva/pratibandhadisambhavat / yady evam abhrantagrahanam evastu /kim kalpanapodhagrahaneneti cet104 / na / anumanavikalpasyapi pratyaksatvaprasangat/ (TSP 479,20-480,9). Appendix II: Vinitadeva's Interpretation of abhranta in the Nyayabindutika 'khrul ba med pa ni ma khrul ba ste / thob par byed pa'i yul la phyin ci ma log ces bya ba'i tha Ishig go // // // 'di ltar ma 'khrul ba nyid ni thob par byed pa'i yul la khas blang bar bya'i /dmigs pa'i yul la ni ma yin nos ll gal te dmigs pa'i yul la ma 'khrul pa nyid khas len na/des na 'dir /'di ltar mal 'byor spyod pa'i ni diigs pa'i o kesonduke- M: kesondrake. KS. I follow M;both mean the same. 100 kesonduka- M: kesondraka-KS. 101 Cf. Tib. tr. ma 'khrul ba nyid ni (*abhrantarvam). 102 C. Tib. tr. dngos po ji ltar gnas pa'i rnam pa dmigs pa nyid kyis ni ma yin te /(*vastuno yathavasthitakaralambanataya / *yathavastusthitakaralambanataya?). Compare also a similar expression in TSP 492,20:... api tu yathavasthitam eva vastusvarupam anudya... (Tib. fon kyang ji lia ba bzhin du gohag pa nyid kyi dngos po'i rang bzhin rjes su brjod nas). 103 -saktih SM(saktir)T(nus pa yin gyi/): sakrikanam K. 104 cer MK : om. S.

Page #22
94 T. Funayama shes pa thams cad ni 'khrul ba yin par 'dod de">/de bas na de'i lugs bsal bar rab tu byed pa rtsom par gyur te/<8rab tu byed pa ni mdo sde pa dang/rnal 'byor spyod pa'i rjes su snyegs par rtsom par 'dod la>/<9gnyi ga yang thob par byed pa'i yul la mi bslu ba'i mtshan nyid ni ma 'khrul ba yin par khas len to> // <10de lta yin na go/rjes su dpag pa yang mngon sum nyid du thal bar 'gyur tel>/ //<12de'i phyir rjes su dpag pa bsal bar bya ba'i phyir rtogs pa dang bral ba zhes bya ba smos te1/... (NBT(V) 40,1-19) In the following passages, words in bold letters are the original Skt. of the Tibetan words of NBT(V) with the same index number <1><12>. DhP 42,18f: ihavisamvadakatvam abhrantatvam abhipretam, tena dvicandradijnanam vyavacchinnam yogacaramatam api grhitam bhavatiti purvavyakhyanam...; ibid. 44,23f.: ihabhrantapadam taimirikadijnanavyavacchedartham3> iti yat purvakair vyakhyatam... NBTT 18,14-19, 1: tena (= Vinitadevena) tv evam vyakhyatam/abhrantam iti yad visamvadi na bhavatievam saty anumanasyapi pratyaksalaksanam prapnotiti ? / taimirikajnanavyavacchedartham abhrantapadam3>, / 'nyatha yogacaramatam amatam syat>// yatah "pramanya(sic!)visaye 'visamvadakatvalaksanam abhrantatvam dvabhyam abhyupagamyate/ <10 saty evam anumanasyapi pratyaksatvaprasangah10>//<12atah tannirakaranartham kalpanapodhagrahanam2> iti anyavyakhyam niracikirsur aha-abhrantam ity adi/. Appendix III: Devendrabuddhi and Jinendrabuddhi Some of the passage of the pratyaksabhasa-section of Jinendrabuddhi's Pramanasamuccayatika are identical with those of Devendrabuddhi's Pramanavarttikapanjika. The following two cases are such examples, where the Tibetan words in bold letters were possibly identical in Sanskrit. 1. PST IP 32a7-b5, D 28b2-7: rab rib bcas105 zhes pa 'dis dbang po la nye bar gnod pa las skyes pa rab rib la sogs pa'i shes pa mngon sum ltar snang ba bzhi pa gsungs so zhes pa'o //...de ltar ni smig rgyu la sogs pa la chu la sogs pa'i shes pa de kho na rnam par gcad% pa'i ched du ma 'khrul ba zhes mngon sum gyi mtshan nyid la khyad par nye bar byas pas so // de ni dbang po la nye bar gnod pa las skyes pa'i shes pa bzlog pa'i don du'o zhe na/ma yin te / don dang phrad pa 'dzin pa kho na las don med pa las byung ba zla ba gnyis la sogs pa'i shes pa bzlog pa'i phyir ro // gzhan du na dbang po las skyes pa shes pa nyid brjod par bya bar 'gyur ro //rjes su dpag pa la sogs pa'i shes pa ni/ji ltar sngar nyams su myong ba'i brda dran pa'i stobs las 'jug pa dang ingon sum ma yin pa'ang yin par grub pa kho na'o zhes pa'o // PVP IP 249b1-5, D 212b6-213a2: dbang skyes min pa bsgrub don dul07 //zhes bya ba smos she 105 PS I 8d: sataimiram. 106 gcad D: bcad P. 107 PV III 289a: anaksajatvasiddhyartham.

Page #23
na/khrul pa bstan phvir ro108 // phyi rol pa 'di la 'khrul pa snang ste/smig rgyu la sogs pa dang kun rdzob tu yod pa dag la shes pa dbang po las skyes pa nyid yin par 'dod pa lta bu'o // de ltar na de nyid rnam par gcad pa'i phyir mngon sum gyi mtshan nyid kyi khyad par 'khrul pa med pa can zhes brjod do // dbang po la109 gnod pa byas pa'i shes pa bzlog pa'i don du zhes bya ba yang brdzun yin te/don dang phrad nas 'dzin pa nyid kyis don ma yin pa las 'byung ba zla ba gnyis pa la sogs pa'i shes pa zlog!10 pa'i phyir ro // de lta ma yin na dbang po las skyes pa zhes bya ba nyid brjod par bya bar 'gyur ro // rjes su dpag pa la sogs pa yang de yin la/grub pa yin pas zhes bya bar tshig rnam par sbyar ro // 'di Ita ste/rjes dpag rjes su dpag las byung // dran byung brjod las byung ba vill // shes pa sngon nyams su myong ba brda dran pa'i stobs kyis 'grub par 'gyur ba nyid yin pa lta bu'o // Kamalasila's Interpretation of 'Non-erroneous 2. PST IP 3267-33a2, D 29a2-4: mngon sum ltar snang ba bzhi pa gang de ni 'dir dmigs kyis bsal bar blta bar bya'i [/] mtshan nyid kyi tshig gis don gyis go bas gsal113 ba'i dper brjod pa ni ma yin no // gzhan du na mtshan nyid 'khrul par 'gyur ro // de'i phyir dmigs kyis bsal ba'i tshig des phyi dang nang gi nye bar gnod pa'i rkyen gyis bcom pa'i dbang po'i shes pa rtog pa dang bral pa nyid na'ang mngon sum ltar snang bar brjod do // rab rib bcas zhes pa 'dir/rab rib ni dbang po nye bar bcom pa'i rkyen thams cad nye bar mtshon pa tsam du blta bar bya'o lo // gang zhig zla ba gnyis la sogs pa'i shes pa dbang po las skyes pa ma yin pa kho na ste [/] 'on kyang yid kyi kho na'o zhes smra'o // PVP IP 250b5f., D 213b6f.: 'dir bzhi pa ni / mngon sum ltar snang ba rnam pa bzhi po 'di la mngon sum ltar snang ba bzhi pa/mngon sum Itar snang rab rib bcas14 //zhes15 bya ba gsungs pa gang yin pa de dmigs kyis dgar ba yin parl16 rig par bya ba'am mtshan nyid lhag ma yin par rig par bya'i/mtshan nyid kyi tshig las don gyi gsal ba'i dpe ni ma yin no // PVP IP 250b7-251a3, D 214a1-4: 'di Itar rtog pa yongs su rdzogs par byas nas 'di'i rtog pa nyid kyi rgyu ma bshad par/rab rib bcas11 zhes!18 gsungs so // de nyid kyi phyir de ni rtog pa dang bral ba'i mngon sum ltar snang ba bstan pa yin no zhes par bya'o // de lta ma yin na/mtshan nyid 'khrul pa can du 'gyur ro // de bas na dmigs kyis dgar ba'i tshig/des bshad nver gnod pa las skyes 119/phyi dang nang gil20 gnod pa byed pa'i rkyen gyis 21 dbang po la122 gnod pa byed pa las skyes pa'i shes pa thams cad rtog pa dang bral pa nyid yin na yang / mngon sum Itar snang ba zhes brjod do // de la rab rib 'ba' zhig ni // nyer gnod mtshon par byed pa yin //23 de la rab rib dang 108 PV III 289b: (ukte dve) bhrantidarsanat. 109 la D: om. P. 110 zlog P: Idog D. Iba yi D: ba'i P. The citation is PS I 7d-8a: anumananumanikam // smartabhilasikam ceti/. 112 gis P: gi D. 113 gsal P: bsal D. IPS I 8b: pratyaksabham sataimiram. 115 //zhes D: shes P. 116 yin par D: om. P. 117 PS I 8b: sataimiram. 95 118 zhes D: ces P. 119 PV III 293b: tenoktam upaghatajam. 120 ogi P: gis D. 121 gyis P: gyi D. 122 la D: om. P. 123 The passage is PV III 293cd: kevalam tatra timiram upaghatopalaksanam //.

Page #24
T. Funayama bcas pa'i tshig gi rab rib ni dbang po la gnod pa byed pa'i rkyen thams cad kyis124 nye bar mtshon pa yin par blta bar bya'o // de yang vid kyi!25 shes kha cig!26 // zla ba gnyis la sogs pa'i shes pa yang dbang po las!27 skyes pa ma yin gyi/'on kyang yid kho na yin zhes kha cig smra'o // Abbreviations and Literature AS Abhidharmasamuccaya (Asanga): See ASBh. ASB Abhidharmasamuccayabhasya: N. Tatia, ed. Abhidharmasamuccaya-bhasyam. Patna, 1976. AP Alambanapariksa (Dignaga) (Tib.): See Frauwallner (1930). APV Alambanapariksavitti (Dignaga) (Tib.): See Frauwallner (1930). Brough, John. 1973. l-ching on the Sanskrit grammarians. Bulletin of the School of Oriental and African Studies (Univer sity of London) 36 (2): 248-260. Derge edition of Tibetan Tripitaka: SDE DGE TIBETAN TRIPITAKA BSTAN HGYUR, preserved at the Faculty of Letters, University of Tokyo. TSHAD MA. 20 vols. Tokyo, 1981-84 DhP Dharmottarapradipa (Durvekamisra): Dalsukhabhai Malvania, ed. Pandita Durveka Misra's Dharmoftarapradipa: Being a sub-commentary on Dharmottara's Nyayabindutika, a commentary on Dhannakiri's Nyayabindu. Tibetan Sanskrit Works Series 2. Patna, 1955. Fang bian xin lun ITC (anonymous), tr. Jijiaye ( 7): T 1632 (vol. 32). Franco, Eli. 1986. Once again on Dharmakirti's deviation from Dignaga on pratyaksabhasa. Journal of Indian Phi: losophy 14: 79-97. 1987. Perception, knowledge and disbelief: A study of Jayarasi's scepticism. Stuttgart. Frauwallner, Erich 1930. Dignagas Alambanapariksa, Text, Ubersetzung und Erlauterungen. Wiener Zeitschrift fur die Kunde Morgenlandes 37: 174-194. 1961. Landmarks in the history of Indian logic. Wiener Zeitschrift fur die Kunde Sud- und Ostasiens 5: 125-148. Funayama, Toru. 1992. A study of kalpanapodha: A translation of the Tattvasamgraha vv. 1212-1263 by Santaraksita and the Tattvasamgrahapanjika by Kamalasita on the definition of direct perception. Zinbun (Annals of the Institute for Research in Humanities, Kyoto University) 27: 33-128. 1995a. Arcata, santaraksita, Jinendrabuddhi, and Kamalasila on the aim of a treatise (prayojana). WZKS 39: 181-201. 1995b. 8 seiki Nalanda shusshin chushakuka oboegaki: Bukkyo chishikiron no keifu (Notes on the 8th century commentators from Nalanda: A lineage in the Buddhist theory of knowledge). Nippon Bukkyogakkai Nenpo 60:49-60. Guan suo yuan lun shi TELPTAST (r) (Dharmapala's commentary on the Alambanapariksa), tr. Yijing (MP): T 1625 (vol. 31). Hattori, Masaaki. 1965. Pratyaksabhasa: Dignaga's view and Dharmakirti's interpretation. Indogaku Shironslui 6/7: 122-128. 124 kyis D: kyi P. 125 kyi P: kyis D. 126 PV III 294a: manasam tad apity eke. 127 las D: la P.

Page #25
Kamalasila's Interpretation of 'Non-erroneous' 97 1968. Dignaga, on perception: Being the Pratyaksapariccheda of Dignaga's Pramanasamuccaya, from the Sanskrit fragments and the Tibetan versions, translated and annotated. Harvard Oriental Series * 47. Cambridge: Harvard University Press. Hayes, Richard P. 1983. Jinendrabuddhi. Journal of the American Oriental Society 103 (4): 709-717. Inami, Masahiro. 1993. Bukkyo ronrigakuha no shinri ron: Devendrabuddhi to sakyabuddhi (Theory of truth in the Buddhist epistemological school: Devendrabuddhi and Sakyabuddhi). In Watanabe Fumimaro hakase tsuito ronshu: Genshi bukkyo to Daijo bukkyo, ed. E. Mayeda, vol. 2, 85-118. Tokyo. Ishibashi, Saka'e. 1993. Tattvasamgraha-panjika, pratyaksa-pariksa kk. 1311-1328, daini kiteiji 'meiran ga nai' no kento, shiyaku (A tentative translation of TSP ad TS 1311-28 on the examination of the second regulation 'non-erroneous'). Ryukoku Daigaku Bukkyogaku Kenkyushitsu Nenpo 6:2-33. Krishnamacharya's edition of TS/TSP: See TS. Kanazawa, Atsushi. 1987. Kuge: timira ganbyo (gen'ei) tono kakawari de (Sky-flower: With relation to eye-disease timira). Bukkyogaku 23: 29-56. Katsura, Shoryu. 1982. Inmyo shorimonron kenkyu (A study of NMu). Part 5. Hiroshima Daigaku Bungakubu Kiyo 42: 82-99. Kobayashi, Nobuhiko. . 1965. Tattvasamgraha ni hihan sareru Bhamaha no imi ron (Bhamaha's view on sabdartha as criticized in TS). Indogaku Shironshu 6/7: 86-92. 1978. Bhamaha ni in'yo sareru "Nyasa" no chosha ni tsuite (On the author of the Nyasa as quoted by Bhamaha). In Ashikaga Atsuuji hakase kiju kinen: Orientogaku indogaku ronshu, ed. Oriento Gakkai, 461-474. Tokyo. Krasser, Helmut 1991. Dharmottaras kurze Untersuchung der Gultigkeit einer Erkenntnis. Laghupramanyapariksa. 2 vols. Wien. On the relationship between Dharmottara, santaraksita and Kamalasila. In Tibetan Studies: Proceedings of the 5th Seminar of the International Association of Tibetan Studies, NARITA 1989, 151 158. Narita. 1995. Dharmottara's theory of knowledge in his Laghupramanyapariksa. Journal of Indian Philosophy 23:247-271. van der Kuijp. Leonard W.J. 1985. Studies in the life and thought of mKhas-grub-rje I: mKhas-grub-rje's epistemological oeuvre and his philological remarks on Dignaga's Pramanasamuccaya I. Berliner Indologische Studien 1: 75 105. M Jaisalmer Mss. of TS TSP. Mimaki, Katsumi. 1972. Shoki yuishiki shoronjo ni okeru Sautrantika setsu (The Sautrantika theories in the early Yogacara treatises). Tohogaku 43: 92-77. NB Nyayabindu (Dharmakirti): See DhP. NBh Nyayabhasya (Vatsyayana): See NS. NBT(Dh) Nyayabindutika (Dharmottara): See DhP. NBT(V) Nyayabindutika (Vinitadeva) (Tib.): L. de La Vallee Poussin, ed. Tibetan translation of the Nyayabindu of Dharmakirti (A treatise on Buddhist logic) with the commentary of Vinitadeva. Bibliotheca Indica 171. 1913. Reprint, Calcutta, 1984 NBTT Nyayabindurikarippani. Th. Stcherbatsky, ed. Nyayabindutikatippani. St.-Petersbourg, 1909. NMU Nyayamukha (Dignaga): In ming zheng li men lun ben (EWE FI P9A4]), tr. Xuanzang (*): T 1628 (vol. 32). 1992.

Page #26
T. Funayama Nyayapravesakasutra (Sankarasvamin): The Nyaya-pravesa, part I: Sanskrit text with commentaries. Ed. A. B. Dhruva. Baroda, 1930. Reprint, Delhi, 1987, as Nyayapravesa of Dirinaga with commentaries of Haribhadra Suri and Parsava-deva. Nyayasutra (Gautama): Nyayadarsanam Bhasya-Varttika-Tatparyarika-sahitam, with Varsyayana's Bhasya, Uddyotakara's Varttika, Vacaspati Misra's Tatparyarika and Visvanatha's Vitri. Vol. 1. Ed. Taranatha Nyayatarkatirtha and Amarendramohan Tarkatirtha. Calcutta Sanskrit Series 28. Calcutta, 1936. Vol. 2. Ed. Amarendramohan Tarkatirtha and Hemanta Kumar Tarkatirtha. Calcutta Sanskrit Series 29. Calcutta, 1944. Reprint, Kyoto, 1982. Peking edition of Tibetan Tripitaka. Panini Astadhyayi (Panini): S. Ch. Vasu, ed. and tr. The Ashadhyayi of Panini. 2 vols. Allahabad, 1891. Reprint, Delhi, 1962. PSI Pramanasamuccaya (Dignaga), chapter 1 (Pratyaksa) (Tib.): See Hattori (1968). PST Visalamalavati-nama-pramanasamuccayatika (Jinendrabuddhi) (Tib.): P 5766 (vol. 139), D 4268 (vol. 20). PSV Pramanasamuccayavrtti (Dignaga), chapter 1 (Pratyaksa) (Tib.): See Hattori (1968). PV Pramanavarttika (Dharmakirti). PVI Chapter 1 (Svarthanumana): See PVSV. PV III Chapter 3 (Pratyaksa): See Tosaki (1979, 1985). PVin Pramanaviniscaya (Dharmakarti). PVin I Chapter 1 (Pratyaksa): Tilmann Vetter. Dharmakirti's Pramanaviniscayah. I. Kapitel: Pratyaksam, Einleitung. Text der tibetischen Ubersetzung. Sanskritfragmente, deutsche Ubersetzung. Osterreichische Akademie der Wissenschaften, philosophisch-historische Klasse, Sitzungsberichte, 250. Band, 3. Abhandlung. Veroffentlichungen der Kommission fur Sprachen und Kulturen Sud- und Ostasiens, Heft 3. Wien: Herman Bohlaus Nachf., 1966. PVin II Chapter 2 (Svarthanumana): Ernst Steinkellner. Dharmakirti's Pramanaviniscayah, zweites Kapitel: Svarthanumanam, Teil 1. tibetischer Text und Sanskrittexte. Osterreichische Akademie der Wissenschaften, philosophisch-historische Klasse, Sitzungsberichte, 287. Band, 4. Abhandlung. Veroffentlichungen der Kommission fur Sprachen und Kulturen Sudasiens, Heft 12. Wien: Verlag der Osterreichischen Akademie der Wissenschaften, 1973. PViT Pramanaviniscayarika (Dharmottara) (Tib.): P 5727 (vol. 136), D 4227 (vols. 14-15). PVP Pramanavarttikapanjika (Devendrabuddhi) (Tib.): P 5717(b) (vol. 130), D 4217 (vol. 2). PVSV Pramanavarttikasvavrtti (Dharmakirti): Raniero Gnoli, ed The Pramanavarttikam of Dharmakirti, the first chapter with the autocommentary, text and critical notes. Serie Orientale Roma 23. Rome, 1960. PV? Pramanavartrikarika (Sakyabuddhi) (Tib.): P 5718 (vol. 131), D 4220 (vols. 3-4). Shastri's edition of TS/TSP (based on Jaisalmer Mss.) Schmithausen, Lambert. 1965. Mandanamisra's Vibhramavivekah: mit einer Studie zur Entwicklung der indischen Irrtuinslehre. Wien. 1972. The Definition of pratyaksam in the Abhidharma-samuccayah. WZKS 16: 153-163. SS Sarvajnasiddhi (Ratnakirti): A. Thakur, ed. Ramakirtinibandhavalih. 2nd ed. Patna, 1975. Stcherbatsky. Theodore. 1932. Buddhist logic. Vol. 1. Leningrad. Steinkellner, Ernst. 1972 New Sanskrit fragments of Pramanaviniscayah, first chapter. WZKS 16: 199-206. Taisho Shinshu Daizokyo. Thrasher, Allen W. 1979. The dates of Mandana Misra and Samkara. WZKS 25: 117-139. S

Page #27
Kamalasila's Interpretation of 'Non-erroneous' 99 TS Tillemans, Tom J. F. 1990. Materials for the study of Aryadeva, Dharinapala and Candrakirti. 2 vols. Wien. Tosaki, Hiromasa. 1979. Bukkyo ninshikiron no kenkyu (A study of Buddhist epistemology). Vol. I. Tokyo. 1984. Kamalasila saku Nyayabindupurvapaksesamksipta: Genryo sho no text to wayaku (Kamalasila's Nyayabindupurvapaksesamksipta: the chapter direct perception'. Text and Japanese translation). Indo Koten Kenkyu 6: 477-494. Narita. 1985. Bukkyo ninshikiron no kenkyu (A study of Buddhist epistemology). Vol. 2. Tokyo. 1990. Hosho cho "Pramanaviniscaya" dai I sho, genryo (chikaku) ron no wayaku (6): yogin no genryo to jigenryo (Japanese translation of PVin by Dharmakirti, the first chapter on the theory of direct perception (6): Yogin's perception and pseudo-perception). Nishinihon Shukyogaku Zasshi 12:58-62. Tattvasamgraha (santaraksita): S. Dwarikadas Shastri, ed. Tattvasangraha of Acarya Shantaraksita with the commentary 'Panjika' of Shri Kamalashila. 2 vols. Varanasi, 1968. (=S). Cf. E. Krishnamacharya, ed. Tattvasangraha of santaraksita with the cominentary of Kamalasila. 2 vols. Baroda, 1926. Reprint, Baroda, 1984, 1988. (= K). TSop Tarkasopana (Vidyakarasanti): G. Tucci, ed. Minor Buddhist texts. Parts I & II. Delhi, 1986. (Original edition, Rome 1956 and 1958). TSP Tattvasamgrahapanjika (Kamalasila): See TS. Tucci, Giuseppe. 1930. Bhamaha and Dinnaga. Indian Antiquary 59: 142-147. Vetter, Tilmann. 1964. Erkenntnisproblemie bei Dharmakirti. Osterreichische Akademie der Wissenschaften, philosophisch historische Klasse, Sitzungsberichte, 245. Band, 2. Abhandlung. Veroffentlichungen der Kommission fur Sprachen und Kulturen Sud- und Ostasiens, Heft 1. Wien: Herman Bohlaus Nachf., 1964. Vy Vidhiviveka (Mandanamisra): Vidhiviveka of Sri Mandana Misra with the commentary Nyayakaaika of Vachaspati Misra. Ed. M. L. Gosvami. Varanasi, 1984. Wiener Zeitschrift fur die Kunde Sudasiens. Yaita, Hideomi. 1982. Dharmottara no ichi kosatsu: Dharmakiti no genryo teigi ni kanshite (A study of Dharmottara: On Dharmakirti's definition of direct perception). Sanko Bunka Kenkyujo Nenpo 15: 1-14. 1984. Dharmottara no Pramananiniscayatika no ichi shiryo: Sanskrit danpen to shiyaku (A source for Dharmottara's PVins: Sanskrit fragments and tentative translations). Mikkyogaku Kenkyu 16: 17-37. 1992. Yugaron no inmyo: Bonbun text to wayaku [The hetuvidya of YBh: Text and Japanese translation). * Naritasan Bukkyokenkyujo Kiyo 15: 505-576. YBh Yogacarabhumi, the hetuvidya-section: See Yaita (1992). WZKS