Dignāga on Āvīta and Prasaṅga

Journal of Indian and Buddhist Studies Vol. 61, No. 3, March 2013 （171） DignAga on AvIta and PrasaGga Watanabe Toshikazu 0. In the SaSTitantra, the fundamental text of the SAGkhya school before the time of the SAGkhyakArikA, two modes of reasoning for proving invisible things such as the existence of primordial matter（pradhAna）etc. are explained, i.e., vIta and AvIta. Of these two, AvIta reasoning is formulated after vIta reasoning in order to support the same conclusion drawn by the vIta reasoning through a reductio ad absurdum-style argument（prasaGga）. 1） In the third chapter of his PramANasamuccaya, DignAga（ca. 480–540）mentions the SAGkhyas’ AvIta reasoning as being a prasaGga argument, and criticizes its definition, its way of being formulated, and its independence from vIta reasoning. While other prasaGga arguments are classified as refutation（dUSaNa/parihAra）because they are formulated by provisionally relying on the opponent’s position, AvIta reasoning is presented on the basis of the proponent’s own view and hence its logical reason is regarded as fulfilling the three characteristics of a proper logical reason（trairUpya）, at least for the proponent. In this paper I will first attempt to clarify DignAga’s explanation of the structure of AvIta reasoning and its reformulation into vIta reasoning, and then examine the influence of DignAga’s analysis of AvIta reasoning on DharmakIrti’s theory of prasaGgaviparyaya, as well as on BhAviveka’s criticism of BuddhapAlita. 1. In PS 3.16, DignAga mentions two characteristics of AvIta reasoning, and argues that AvIta reasoning is not different from vIta reasoning. hetvabhAvaprasaGgas tu yatrAvItena kathyate | sa dRSTAntadvayAt siddhes tasmAd vItAn na bhidya­ te || 2） On the other hand, in a certain proof［a-1］the negation of the logical reason（hetvabhAva） ［adduced in the vIta reasoning］is stated as an undesired consequence（prasaGga）by means of AvIta reasoning. Such a proof is［a-2］based on the establishment［of the pervasion］by two modes of exemplification（i.e., positive and negative concomitance）. Therefore,［AvIta reasoning］is not differentiated from vIta reasoning. Because of these two characteristics, i.e.,［a-1］and［a-2］, AvIta reasoning can be reformu─ 1229 ─ （172） DignAga on AvIta and PrasaGga（T. Watanabe） lated into vIta reasoning. DignAga presupposes the following vIta formulation in the process of exemplifying this. 〈VIta-D〉bhedAnAm ekakAraNatvam, anvayadarZanAt. Various individual things possess one and the same cause; Because homology（anvaya） ［concerning three constituents, i.e., pleasure（sukha）, pain（duHkha）, and confusion（moha）］is observed. 3） DignAga formulates this〈VIta-D〉by revising the following original formulation found in the ST in order to make the logical reason “homology”（anvaya）possess the three characteristics of a proper logical reason 4）: 〈VIta-1〉asti pradhAnam, bhedAnAm anvayadarZanAt. 5） The AvIta reasoning derived by DignAga from〈VIta-D〉is as follows: 6） 〈AvIta-D 1〉＊na vyaktasyaikapUrvakatvAbhAvaH, bhedaprasaGgAt. It is not the case that evolutes do not arise from one and the same cause, because［if they would not arise from the same cause,］it would follow that［all evolutes would］differ［from each other completely］. This〈AvIta-D 1〉is also an improvement on the following AvIta reasoning set forth by the SAGkhya 7）: 〈AvIta-1〉na vyaktasyAsata utpattiH, bhedaprasaGgAt. 8） With〈VIta-D〉and〈AvIta-D 1〉in mind, DignAga explains his reformulation of AvIta reasoning into vIta reasoning as follows: ［1］yasmAd anvaya ekapUrvakatve dRSTaH, tadabhAve ca na dRSTaH, ata ekapUrvakatvam.［2］ tadanabhyupagame cAnvayAbhAvaprasaGgaH. sa ca nAsti［3］iti sa evAnvaya ukto bhavati. . . . ko hy atra viZeSo bhedaprasaGgAnvayayoH. 9） ［1］Since homology［among evolutes］is seen［only］when［they］come from one and the same cause and is not seen when［they］do not,［evolutes that possess homology as their property］come from one and the same cause.［2］And if it（i.e., their coming from one and the same cause）is not accepted, the negation of homology would follow. But there is no［negation of homology in reality］.［3］Therefore,［in the AvIta reasoning,］that very［undesired consequence of negation of homology（anvayAbhAvaprasaGga）］becomes spoken of as “homology［in vIta reasoning］.” . . . For, in this case what difference is there between “bhedaprasaGga” and “anvaya”? In passage［1］, it is shown that “homology,” which is the logical reason of〈VIta-D〉, is pervaded by “coming from one and the same cause.” Since “ekapUrvakatva” and “ekakAra­ Natva” are regarded as conveying the same meaning, this pervasion（vyApti）is the same one as presupposed by the〈VIta-D〉. If the pervasion of the logical reason（H）by the property to be proved（S）is represented as “H → S,” the positive concomitance（an­ ─ 1230 ─ DignAga on AvIta and PrasaGga（T. Watanabe） （173） vayavyApti）of this pervasion can be described as follows: 〈VyApti-A〉anvaya → ekapUrvakatva And if the fact that pervasion “H → S” held with regard to the subject “P” is represented as “P（H → S）,” then〈VIta-D〉can be described as follows: 〈VIta-D 1〉Individual things（anvaya → ekapUrvakatva） In contrast, the contraposition of〈VyApti-A〉, i.e., negative concomitance（vyatirekavy­ Apti）, can be described with the negative sign “¬” as follows: 〈VyApti-B〉¬ekapUrvakatva → ¬anvaya Passage［2］expresses〈VyApti-B〉in the form of a prasaGga-style argument and denies the possibility of the negation of homology（¬anvaya）in reality. This is the explanation of the characteristics of AvIta reasoning referred to in PS 3.16. Comparing these descriptions, we can get clearer view of DignAga’s understanding of AvIta reasoning: it has the same pervasion as〈VIta-D〉 （see above PS 3.16［a-2］）because〈VyApti-B〉 is logically equivalent to〈VyApti-A〉and results in the undesired consequence of the negation of “homology,” i.e., the logical reason of〈VIta-D〉 （see above PS 3.16［a-1］）. Moreover, it is known from passage［3］that the negation of “homology” means the complete difference （bheda）of every evolute. Therefore, the undesired consequence of complete difference （bhedaprasaGga）is derived from the negation of coming from one and the same cause （¬ekapUrvakatva）. Now let us return to〈AvIta-D 1〉. This formulation apparently consists of the thesis and its reason. However, the phrase “bhedaprasaGgAt” should not be regarded as the logical reason for proving the thesis “na vyaktasyaikapUrvakatvAbhAvaH” because it cannot prove the thesis by itself. In order to derive the thesis, “bhedaprasaGga” requires two things: First, the negation of coming from one and the same cause（¬ekapUrvakatva）must be understood as the condition of “bhedaprasaGga,” because, as seen above, the latter can be held only if the former is established. In addition, since “bhedaprasaGga” is an undesired consequence, it must be negated. Consequently, in〈AvIta-D 1〉the following process of proof is seen with regard to evolutes（vyakta）: 〈AvIta-D 2〉Evolutes（¬ekapUrvakatva → bheda）∧ ¬bheda ∴ ¬¬ekapUrvakatva Since for the SAGkhya there is no difference between “individual things”（bheda）and “evolutes”（vyakta）, and since the negation of homology（¬anvaya）means complete difference（bheda）, it is obvious that〈AvIta-D 2〉does not differ from〈VIta-D 1〉in its ─ 1231 ─ （174） DignAga on AvIta and PrasaGga（T. Watanabe） content. Therefore, insofar as the phrase “bhedaprasaGgAt” in〈AvIta-D 1〉indicates the logical reason “homology” in〈VIta-D〉which is, for the SAGkhya, accepted as possessing the three characteristics of a proper logical reason（trairUpya）, it can also be regarded as equivalent to a proper logical reason. Even though for the Buddhist the logical reason “homology” is not accepted as a proper logical reason, 10）it is worth noting that DignAga accepts the possibility that AvIta reasoning, i.e., a type of prasaGga argument, becomes a proper proof（sAdhana）. 2.〈AvIta-D 1〉, even though it can be reformulated into the〈VIta-D〉, is not a proper proof（sAdhana）if it is put forth to the Buddhist disputant, because the logical reason “homology,” which is adduced in〈VIta-D〉and which guarantees the adequacy of the assumed logical reason “bhedaprasaGgAt” in〈AvIta-D 1〉, is accepted only by the SAGkhya proponent as possessing the three characteristics of a proper logical reason（trairUpya）. However, DignAga’s explanation of the reformulation of AvIta reasoning into vIta reasoning opens up the possibility that AvIta reasoning, a type of prasaGga argument, can thus be regarded as a proper proof. This had a great influence on the later Indian logicians, including Buddhist philosophers BhAviveka（ca. 6th cent.）and DharmakIrti（ca. 600–660）, as well as within Tibetan Buddhism. Let us first consider the process of formulating prasaGgaviparyaya（the contraposition of prasaGga）from prasaGga. This process is often used in the treatises of DharmakIrti’s successors. As earlier studies have shown, the structures of prasaGga and prasaGgavipary­ aya can be described as follows: 11） PrasaGga: A（B → C）∧ ¬C ∴¬B PrasaGgaviparyaya: A（¬C → ¬B） Both prasaGga and prasaGgaviparyaya are presented with regard to the same subject（A）. The pervasion presupposed in prasaGgaviparyaya（i.e., ¬C → ¬B）is the contraposition of that which is presupposed in the prasaGga argument（i.e., B → C）. Even though DharmakIrti does not clearly explain these two structures, because his successors commonly refer to the above-mentioned structure, it is likely that these structures represent DharmakIrti’s own idea. And it is easily recognized that the structures of DharmakIrti’s prasaGga and prasaGgaviparyaya correspond completely to those of DignAga’s vIta and AvIta arguments, respectively. 12）The identification of these structures is supported by the fact that Jinendrabuddhi explains the reformulation of AvIta reasoning into vIta reasoning in the PS（V） ─ 1232 ─ DignAga on AvIta and PrasaGga（T. Watanabe） （175） by using the term “prasaGgaviparyaya.” 13）Accordingly, it can be said that the germ of DharmakIrti’s theory of prasaGgaviparyaya can be found in DignAga’s interpretation of AvIta reasoning. Another Buddhist philosopher who owes much of his theory of proof to DignAga is BhAviveka. As has been pointed out by Ueda［1995］, it seems that BhAviveka makes use of DignAga’s interpretation of AvIta reasoning in order to criticize BuddhapAlita’s prasaGga argument. As an example, we will examine the discussion of the negation of arising from other. On the other hand, the Master BuddhapAlita says: It is not the case that things arise from other, because［if they arise from other,］it would follow that everything would arise from everything.［To this,］the Master BhAviveka states the refutation［as follows］: Then, in this case, because［it is］a statement of［undesired］consequence, when the contraposition of the［property］to be proved and the logical reason is made, contradiction［of your prasaGga argument］with［your］original position［become evident because the contraposition states that］things arise from self, or from both（i.e., self and other）, or without any cause, because a particular thing arises from a particular thing.（PrasannapadA 36,11–37,2）. As the structure of〈AvIta-D 1〉is described as〈AvIta-D 2〉, the structure of BuddhapAlita’s prasaGga argument is described as follows 14）: 〈PrasaGga-BP〉Things（parata utpadyante → sarvataH sarvasambhava）∧ ¬sarvataH sar­ vasambhava ∴ ¬parata utpadyante BhAviveka reformulates this〈PrasaGga-BP〉as the following argument, and points out that it contradicts the MAdhyamika’s central thesis that things do not arise from self, from other, from both, or without any cause. 〈Viparyaya-BV〉Things（kutaZcit kasyacid utpatti → svata ubhayato ’hetuto votpadyante） What is the “viparyaya” by way of which BhAviveka derives〈Viparyaya-BV〉from 〈PrasaGga-BP〉? Judging from the fact that he regards〈Viparyaya-BV〉as conveying the same meaning as〈PrasaGga-BP〉, it must express the contraposition of the pervasion. That is, “kutaZcit kasyacid utpatti” in〈Viparyaya-BV〉is the negation of “sarvataH sarvasam­ bhava” in〈PrasaGga-BP〉, and “svata ubhayato ’hetuto votpadyante” in〈Viparyaya-BV〉 is the negation of “parata utpadyante” in〈PrasaGga-BP〉. Taking this point into consideration,〈Viparyaya-BV〉can be formulated as follows: Things（¬sarvataH sarvasambhava → ¬parata utpadyante） Although in the case of BhAviveka’s reformulation of〈PrasaGga-BP〉into〈Viparyaya─ 1233 ─ （176） DignAga on AvIta and PrasaGga（T. Watanabe） BV〉the negation is interpreted as a implicative negation（paryudAsa）, while in the case of DignAga it is taken as being a simple negation（prasajyapratiSedha）, it can be said that the process of BhAviveka’s reformulating〈PrasaGga-BP〉into〈Viparyaya-BV〉has a close similarity to that of DignAga’s reformulation of AvIta reasoning into vIta reasoning. 3. When discussing AvIta reasoning, DignAga accepts a possibility of reformulating pra­ saGga argument as a proper proof（sAdhana） . This enables him to incorporate the prasaGga argument into the framework of the trairUpya theory. It can be said that his view about prasaGga marks a turning point in the development of the theory of prasaGga, which until then had merely been regarded as a means for refuting the views of opponents. By making use of DignAga’s explanation of AvIta reasoning, BhAviveka and DharmakIrti further developed their theory of prasaGga. DignAga’s theory of trairUpya became widely accepted, causing the SAGkhya to relinquish their leading position in Indian logic to DignAga and his followers. Since DignAga developed his theory of prasaGga by criticizing the SAGkhya theory of AvIta, it can be said that DignAga is their most ingenious successor and, at the same time, their most severe critic. ―――――――――――――――― ＊ I would like to thank Ms. Peck-Kubaczek for correcting my English and Dr. Anne MacDonald for giving me many invaluable suggestions. Notes: １）For the characteristics of “vIta” and “AvIta,” see Watanabe［forthcoming］. ２） See Katsura［2009: 158］. Words of PS（V）in Roman typeface are attested in Jinendrabuddhi’s Sanskrit manuscript or in fragments, whereas those in italics have been reconstructed from Tibetan translations. ３）See NMu（Katsura［1978: 110］）. ４）For DignAga’s modification of vIta reasoning, see Watanabe［2008］ . ５）See Frauwallner［1958: 125］ . ６）Since DignAga does not give this formulation, this is my conjecture based on his description. ７）For DignAga’s criticism of the formulation of AvIta reasoning propounded by the SAGkhya, see Watanabe ［forthcoming］. ８）PSV 3.16（K［P128a8］, V［D44a5–6, P47a5–6］）. ９）PSV 3.16 （K［P129b1–3］, V［D44a6–44b1, P47a6–8］）. 10）See Watanabe［2008］. 11）See Tani［1983: 7–12］and Iwata［1997: 427–428］. 12）Since Ueda［1995: 113–117］takes the phrase “bhedaprasaGgAt” in〈AvIta-D 1〉as the direct reason for proving its thesis, his interpretation of the structure of〈AvIta-D 1〉is different from mine. 13）See PST B122b4–6（D159a5–7, P182b3–6）: tadanabhyupagame ceti tasya vyApakasyaikapUrvakatvasyAnabhyupagame. anvayAbhA­ vaprasaGga iti vyApyasyAnvayasyAbhAvaprasaGgaH syAd ity arthaH. . . . sa cAnvayAbhAvaprasaGgo nAstIti prasaGgaviparyayeNa sa evAnvaya ukto bhavati-anvayAd ekakAraNapUrvakA bhedA iti. etc. ─ 1234 ─ DignAga on AvIta and PrasaGga（T. Watanabe） （177） 14）Avalokitavrata gives different interpretation. See Ueda［1995: 119］. Abbreviations: K: PS（V）. Trans. KaNakavarman and Dad pa’i shes rab. P5702. NMu: NyAyamukha. See Katsura［1978］. PrasannapadA: MUlamadhyamakakArikAs de NAgArjuna avec la PrasannapadA. Ed. L. de La Vallée Poussin. St. Pétersbourg. 1903–1913. PST: PramANasamuccayaTIkA. D4268, P5766. PS（V）: PramANasamuccaya（vRtti）. See K and V. ST: SaSTitantra. See Frauwallner［1958］. V: PS （V） . Trans. VasudararakSita and Seng rgyal. D4204, P5701. Frauwallner, Erich. “Die Erkenntnislehre des klassischen SAMkhya-Systems.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 2（1958）: 84–139. Iwata Takashi. “On PrasaGgaviparyaya in DharmakIrti’s Tradition-PrajJAkaragupta and gTsaG nag pa.” In Tibetan Studies: Proceedings of the 7th Seminar of the International Association for Tibetan Studies, Graz 1995. Vol. I, 427–437. Wien: Verlag der Österreichschen Akademie der Wissenschaften, 1997. Katsura ShōryU 桂紹隆. “Inmyō shōri monron kenkyU（2） ” 因明正理門論研究〔二〕 . : 110–130. Katsura Shoryu. “ReHiroshima daigaku bungakubu kiyō 広島大学文学部紀要 37（1978） discovering DignAga through Jinendrabuddhi.” In Sanskrit Manuscripts in China. Ed. E. Steinkellner, 153–166. Beijing: China Tibetology Publishing House, 2009. Tani Tadashi 谷貞志. “PrasaGga-sAdhana （kibyU ronshō）dōnyU ni yoru ronrikei no kōzō henkan: Dharmottara to PrajJAkaragupta no kaishaku no sai” プラサンガ・サーダナ（帰謬論証）導入による論理系の構造変換: ダルモッタラとプ : 1–27. Watanabe Toshikazu ラジュニャーカラグプタの解釈の差異. Bukkyōgaku 仏教学 15（1983） 渡辺俊和. “DignAga to SAGkhya gakuha to no ronsō: PradhAna no sonzai ronshō o megutte” ディグ ナーガとサーンキヤ学派との論争: プラダーナの存在論証をめぐって. Indogaku Bukkyōgaku kenkyU 印度学仏教学研究 57, no. 1（2008）: 295–291. Watanabe Toshikazu. “On the Sanskrit Fragments of the Early SAGkhya Theory of Proof.” In the proceedings of the Japan-Austria International Symposium on Transmission and Tradition, Matsumoto, 20–24 August 2012, forthcoming. Ueda Noboru 上田昇. “Viparyaya kō: Bukkyō ronrigaku ni okeru kibyU ronshō” Viparyaya考: 仏教論理学に おける帰謬論証. Tōhōgaku 東方学 90（1995）: 119–110. （Work on this paper has been generously supported by the Austrian Science Fund（FWF）in the framework of the FWF project P23196-G15 “Buddhist literature and its context.”） 〈Key words〉 DignAga, SAGkhya, AvIta, vIta, prasaGga, prasaGgaviparyaya, BhAviveka （Research Fellow, The Institute for the Cultural and Intellectual History of Asia of the Austrian Academy of Sciences） ─ 1235 ─