Reverberations of Dharmakirti’s Philosophy

by Birgit Kellner | 2020 | 264,305 words

This page relates ‘Jnanashrimitra on Viparyaya-badhaka-pramana’ of the study on the philosophy of Dharmakirti (6th century) and his predecessor Dignaga (5th century). This collection of articles reflects philosophical currents in India, China and Tibet during their time and investigates the Buddhist theories of Pramana (“instruments of trustworthy awareness”).

Jñānaśrīmitra on Viparyaya-bādhaka-pramāṇa

(By Kyo Kano)


It was really a new attempt when Dharmakīrti introduced sādhyaviparyaye [hetoḥ] bādha-kapramāṇa (hereafter VBP) as a methodological device with reference to svabhāvahetu, especially with respect to the proof of momentariness of all existence, in his logico-epistemological system.[1] Dharmakīrti’s followers discussed the role and logical content of VBP in the pramāṇa system mainly in the context of the proof of kṣaṇabhaṅga.

With reference to the theory, several studies have brought into focus its impact on the whole logico-epistemological theory of later Buddhist logicians.[2] Above all Tani, in his great work, focuses on the relation between VBP and prasaṅga/prasaṅgaviparyaya and their roles, and discusses how Dharmakīrti’s followers evaluate them in their proof of momentariness.[3] This theory, in fact, raised new complications with respect to the structure of the means of valid cognition and its logical and epistemic base.[4]

One of these complications is the relation between VBP and establishing anvaya(‑vyāpti) as well as vyatireka(‑vyāpti). It seems unclear in both traditional and contemporary inter-pretations whether VBP is valid for establishing either affirmative concomitance (anvaya) or negative concomitance (vyatireka), or whether it is valid for both at the same time. It is of course not easy to find a direct answer to this question by an Indian logician. But in an effort to do so, this paper will begin by focusing on Jñānaśrīmitra’s descriptions of VBP, especially in the vyatireka section of the Kṣaṇabhaṅga chapter of the Jñānaśrīmitranibandhāvali (JNA). At the beginning of this chapter, an opponent called viparyayabādhakapramāṇavā-din appears, whose claims are criticized by Jñānaśrīmitra. By analyzing the relevant texts, we may be able to discern where Jñānaśrīmitra stands on the issue.

As a preliminary observation, it may help us to keep in mind that anvaya-vyāpti and vyatireka-vyāpti are accepted as logically equivalent, at least by the thinkers after Dharma-kīrti. In the epistemic process, however, we cannot simply presume their equivalence, since it depends on what kind of independent pramāṇa or other subsidiary means of cognition the determination of anvaya or vyatireka is related to or based upon, as will be discussed in the following. We can suppose this is one of the reasons Jñānaśrīmitra and Ratnakīrti treat them separately in their discussion on the proof of momentariness.

Jñānaśrīmitra’s Kṣaṇabhaṅgādhyāya consists of three pādas. In the first pāda, pakṣa-dharmādhikāra, the theme is that in the formula “yat sat tat kṣaṇikaṃ yathā jaladharaḥ, santas tu bhāvā ime” (Whatever is existent is momentary, for instance a rain cloud. And these things are existent. [Therefore, these things are momentary.]), the reason sat/sattva is not “unestablished” (asiddha). In the second pāda, anvayādhikāra, it is not incompatible (viruddha). And in the third pāda, vyatirekādhikāra, which I take up in the following, it is not inconclusive (anaikāntika).

1. The relation between two kinds of pervasion, anvaya and vyatireka

The critique of the viparyayabādhakapramāṇavādin in Jñānaśrīmitra’s vyatirekādhikāra begins with the following argument:

(A) There is also no suspicion that [the reason “being existent” might be] inconclusive, because the proof is given based on the pervasion that comprises all [individual instances] (sarvopasaṃhāravatyā vyāpteḥ).

〈Opponent, i.e., viparyayabādhakapramāṇavādin〉 The pervasion is es-tablished by virtue of “the means of valid cognition that annuls [the reason] in the reverse [of that which is to be proven]” (viparyayabādhakapramāṇa). Therefore, we do not find any additional information (vārtā) by introducing it (i.e. the pervasion that comprises all individual instances). Then, why do you say that the pervasion is [already] established?

〈Opponent〉 In this way, there must only be an affirmative concomitance (anvayamātram). Then, how can we obtain the conclusive exclusion [of the reason] from the dissimilar domain?

〈Jñānaśrīmitra〉 It is because the pervasion is exactly established.

〈Opponent〉 How is the pervasion of the negative exclusion itself estab-lished?[5]

From the description above, we first notice that Jñānaśrīmitra evaluates sarvopasaṃhā-ravyāpti in the anvya-vyatireka context of their symmetrical roles, while the opponent, viparyayabādhakapramāṇavādin, does not.[6] This is one of the crucial points dividing Jñānaśrīmitra’s position from that of the viparyayabādhakapramāṇavādin.

I will come back to this point again in a later section. Jñānaśrīmitra then formally differentiates and defines the two pervasions, anvayarūpā and vyatirekarūpā.

(b)〈Jñānaśrīmitra〉 No, the pervasion is in fact of two kinds. One is that which has affirmative concomitance as its essential characteristic (anvayarūpā) and is a property of the subjective element (kartṛdharma). It is the inevitable presence of that which is to be proven in the property-possessor that has the proving [property]. And the other is that which has negative concomitance as its essential characteristic (vyatirekarūpā) and is a property of the objective element (karmadharma). It is the inevitable absence of the proving [property] when that which is to be proven is absent. This is because the following is a sound logic: The pervasion is such that the pervader presents there (i.e., in the pervaded) without fail, or else the pervaded presents only there (i.e., in the pervader).[7]

We do not find any original element in Jñānaśrīmitra’s definition of anvaya-vyāpti and vyatireka-vyāpti except the introduction of the expressions kartṛdharma and karmadharma. As to these terms, taking the sādhana-sādhya relation into consideration, I tentatively interpret kartṛ and karman as corresponding to sādhana and sādhya, respectively.[8] If my understanding is correct, then it follows that Jñānaśrīmitra regards a proof as a kriyā, of which sādhana and sādhya are constituent elements.

Then, he refers to the relation between anvaya-vyāpti and vyatireka-vyāpti as follows:

(C) And the establishment of one of these two inevitably (niyamena) hints at the understanding of the second (i.e., the other). This is because otherwise even one of them cannot be established. The practical activity of the valid means of cog-nition is carried out only directly (sākṣāt) with reference to one [of them], and one can gain an understanding of the other as [logically] immediate (nāntarī-yakatayā). The expression “one [of the two]” is employed just as far as they are [related to each other in this way]. And in such a case, even if suspicion about one of these two arises before a valid means of cognition of the other functions, the suspicion will be expelled by it afterwards. Just as for those who advocate “the means of valid cognition that annuls [the reason] in the reverse [of that which is to be proven] (viparyayabādhakapramāṇavādin),” affirmative concomitance, even though it is suspected, since the suspicion is expelled by virtue of the conclusive exclusion [of the property] from the dissimilar domain (vipakṣād ekāntena vyāvṛttibalāt), concludes that the “being existent,” which never happens at all in [the things] that are not momentary, is really restricted by [the property of] momentary perishing.

Accordingly, there is no difference in content (na kaścid arthato bhedaḥ) between them; viz., the conclusive exclusion of the proving [property] in the absence of [the property] to be proven, and the inevitable concomitance of [the property] to be proven when the proving [property] is present [in the subject of the thesis].[9]

Here Jñānaśrīmitra stresses the incorporation of the understanding of anvaya and vyatireka into the epistemic process. In the last part of the passage cited above, he says that there is no difference in content between them. Here, we must draw attention to the underlined portion. According to Jñānaśrīmitra’s description, the VBP-vādin claims that the suspicion of anvaya can be expelled by the exclusion [of the property] from dissimilar domain. That is, VBP-vādin prioritizes vyatireka over anvaya.

Jñānaśrīmitra then adds:

(D) As to the function (vyāpāra) of the means of valid cognition, however, its practical activity (vyavahāra) is of two kinds, principally (mukhyato) and implicatively (arthataḥ).[10]

On the level of daily performance, however, we find two kinds of practical activities of the valid means of cognition, principally (mukhyato), or directly (sākṣāt), and implicatively (arthataḥ), which follows direct cognition as [logically] immediate (nāntarīyakatayā). For Jñānaśrīmitra, the difference between anvaya and vyatireka lies only in the level of practical activity (vyavahāra). In Dignāga’s system as well as Dharmakīrti’s, the relation between anvaya and vyatireka is sometimes expressed as arthāpattyā.[11] If one is proven, the other is automatically understood by arthāpatti. That is, either of these two is logically implied in the other. Jñānaśrīmitra’s expression arthataḥ accords with this idea.

2. sādhakapramāṇa and viparyayabādhakapramāṇa

Then Jñānaśrīmitra presents an argument of sādhaka- and viparyayabādhaka-pramāṇa.

(E) Precisely therefore, in every reasoning, if one seeks “the means of valid cognition that annuls the reverse of that which is to be proven” [and if there is such a means of valid cognition], there is an affirmative concomitance [of the reason] with that which is to be proven, in the case where the proving [property] (i.e., the reason) is present [in the subject of the thesis]. One should know that seeking “the means of valid cognition that establishes [that which is to be proven] (sādhakapramāṇa),” is implied (upalakṣita) by that (i.e., seeking VBP). Furthermore, [when one introduces VBP,] both [of these two pramāṇas, namely, sādhakapramāṇa and bādhakapramāṇa,] are really integrated (ubha-yasaṃgraha eva) by the term “annulling the reverse” (viparyayabādhaka).[12]

If we try to prove the pervasion by VBP, we should know that sādhakapramāṇa is implied in VBP, and accordingly, we should know that both pramāṇas are integrated in the concept bādhakapramāṇa. The ubhaya of ubhayasaṃgraha can also be interpreted as anvaya and vyatireka, but it seems more likely that Jñānaśrīmitra regards sādhakapramāṇa as corresponding to anvaya and bādhakapramāṇa to vyatireka. The above-quoted passage, however, throws into question any simple opposition between adopting sādhakapramāṇa and adopting bādhakapramāṇa. Though the two positions are ostensibly opposed, his argument makes little distinction between them. Rather, he seems to take an integrative or synthetic stance toward sādhakapramāṇa(vādin) and viparyayabādhakapramāṇa(vādin), even if he thinks one of them is prior to the other in a particular case. Then, the question arises what kind of pramāṇa is intended by the term sādhakapramāṇa? We can surmise that Jñānaśrīmitra’s answer is given in the subsequent discussion, so let us examine what follows.

3. Two kinds of viparyayabādhakapramāṇa[13]

Jñānaśrīmitra continues by explaining the latter position:

(F) In fact, with reference to the pervasion that has negative concomitance as its essential characteristic, [there is a means of valid cognition which] annuls the proving [property] in the reverse of that which is to be proven. (In the proof of momentariness, it annuls “being existent” in those that are not momentary.) With reference to [the pervasion] that has affirmative concomitance as its essential characteristic, on the other hand, it annulsthe reverse of that which is to be proven” [in the proving property] in the case where the proving property is present [in the subject of the thesis]. (anvayarūpāyāṃ tu sādhyaviparyayasya bādhakaṃ sādhane satīti) (It annuls “not being momentary” in those that are existent.) If these two [kinds of VBP] are absent (ubhayābhāve), the inevitable connection of the proving [property] with that which is to be proven would be impossible. The above is intended.[14]

According to Jñānaśrīmitra’s understanding, VBP proves not only vyatireka but anvaya as well, that is, it can be introduced in two ways. Furthermore, for anvaya, VBP annuls “the reverse of that which is to be proven” in the proving [property]. This is a new interpretation of the viparyayabādhakapramāṇa.

As is well known, in his Hetubindu, Dharmakīrti first refers to ‘sādhyaviparyaye hetoḥ bādhakapramāṇa’ as follows:

anvayaniścayo’pi svabhāvahetau sādhyadharmasya vastutas tadbhāvatayā sādhanadharmabhāvamātrānubandhasiddhiḥ. sā sādhyaviparyaye hetor bā-dhakapramāṇavṛttiḥ. yathā yat sat tat kṣaṇikam eva, akṣaṇikatve’rthakriyāvi-rodhāt tallakṣaṇaṃ vastutvaṃ hīyate.[15]

From the description, it is ambiguous whether or not Dharmakīrti was conscious of the relation between ‘sādhyaviparyaye hetor bādhakapramāṇa’ and the proof of anvaya-vyāpti or vyatireka-vyāpti. It seems that Dharmakīrti introduced ‘sādhyaviparyaye hetor bādhakapramāṇa’ as a subsidiary method for supporting the establishment of the inevitable connection of anvaya, stated as anvayaniścayo’pi, but remained strongly conscious of the logical equality of anvaya and vyatireka, an equality that is presupposed by the relation of the contradiction (virodha) between sādhya and sādhya-viparyaya. Logically, it is clear that ‘sādhyaviparyaye hetor (i.e., sādhanasya) bādhakapramāṇa’ can be understood as expressing vyatireka, since it proves that where sādhyaviparyaya is present, that is, where sādhya is absent, hetu is absent.

According to Jñānaśrīmitra’s understanding, however, VBP also annuls “the reverse of that which is to be proven” in the proving [property] (sādhane sādhyaviparyayasya bādhakapramāṇa).[16] This is a new interpretation of VBP, based on the interpretation of the compound (viparyayabādhakapramāṇa) as a genitive tatpuruṣa (hereafter VBP2), which is in contrast to the original interpretation of the compound as a locative tatpuruṣa (hereafter VBP1). According to Jñānaśrīmitra, at least, VBP1 annuls the reason “being existent” (i.e., hetos) in “those that are not momentary” (i.e., sādhyaviparyaye). That is, VBP1 principally or directly proves that if something is not momentary, it is non-existent, namely, vyatireka. It also proves that if a subject of a thesis is existent, it is momentary; that is, it proves anvaya indirectly, because being momentary and not being momentary are contradictory (virodha). In the second new interpretation, however, VBP2 annuls “being not momentary” (i.e., “the reverse of that which is to be proven”) in that which “is existent” (i.e., the proving property). It proves principally or directly that if a subject of a thesis is existent, it is momentary, that is, anvaya. Therefore, according to his understanding, not only vyatireka but also anvaya can and should be proven by VBP (VBP1/VBP2). In the descriptions above, however, we do not concretely find both of these bādhakapramāṇas, such as vyāpakānupalabdhi introduced by Dharmakīrti and other logicians.[17] Moreover, in the above passage, Jñānaśrīmitra emphasizes not only the logical equivalence of anvaya and vyatireka, but the methodological equivalence of sādhakapramāṇa and bādhakapramāṇa as well. By ubhayābhāve does he indicate that both of these VBPs are necessary? In the next paragraph, he states, “Those who present a proof should seek for both of these (i.e., anvaya and vyatireka) alternatively (vikalpena).” In other words, if one wants to prove a pervasion by VBP, one should introduce not only VBP1 but also VBP2, which principally establishes the anvaya relation. It is likely that what he really wants to emphasize is the latter, VBP2.

He continues:

(G) Otherwise, even though one cannot directly (sākṣāt) obtain the means of valid cognition that determines negative concomitance, if one shows a means of valid cognition that establishes the inevitability of the affirmative concomitance (anvayaniyama) of the proving [property] with that which is to be proven, then what is its (=the proving property’s) inevitable connection with reference to the establishment of that which is to be proven? It is be-cause [the means of valid cognition that establishes the inevitability of the affirmative concomitance] hints at the ability of negative concomitance as well (that is, there is no inevitable connection of affirmative concomitance that is independent from the ability of negative concomitance.)[18]

Otherwise [that is, if the means of valid cognition that establishes the in-evitability of the affirmative concomitance is only able to establish affirmative concomitance, irrespective of the inevitability of negative concomitance], it would ensue that the proof formulation of similarity should not really be pre-sented, since the proof does not directly express [the inevitability of] negative concomitance.

Accordingly, even though [an inevitable connection], the ability of which one can understand, does not have negative concomitance as its essential characteristic, exactly insofar as “wherever the proving [property is present], that which is to be proven [is also present]” [is the essence of pervasion], if there is the proving [property in the subject of the thesis], a wise person who is provided with the expectation of that which is to be proven [being established] achieves his purpose.

This is because, in fact, when the proving [property] is present [in the subject of the thesis], even if [a property that is to be proven] is absent when it (i.e., the proving property) is absent, if the [property] that is to be proven is inevitably present when it (i.e., the proving property) is present, then it is fruitful to accept the proof. Hence, a proponent who presents a proof should seek even for both of these (i.e., anvaya and vyatireka) alternatively (vikalpena), and even in either of these styles [a proof] which is established by a means of valid cognition must be shown by a proponent. In the proofs of the opponent, however, even an affirmative concomitance as having inevitability, is really difficult to be obtained, as a negative concomitance is. We should know in this way.[19]

From the above description, the position of the opponent, the viparyayabādhakapramāṇa-vādin, seems to be that VBP should be introduced for the proof of vyatireka and it is the only way to prove the pervasion, while Jñānaśrīmitra equates VBP of vyatireka with that of anvaya, which is the substance of his new interpretation. Here, Jñānaśrīmitra claims that it is needless to state both of the means of valid cognition that prove anvaya and vyatireka; that the statement of either one is enough because they are logically connected to each other. Therefore, if the pervasion of anvaya is proven, it is unnecessary to state the proof of vyatireka.

The final statement above is particularly noteworthy. Jñānaśrīmitra points out a disad-vantage in the opponent’s view: according to the opponent’s proof by VBP, the relation between anvaya and vyatireka is not clear. The point of his criticism lies not in VBP1 itself as a methodical device, but in the opponent’s position whereby the proof of vyatireka by VBP1 is the one and only way to prove a pervasion.

According to Jñānaśrīmitra’s understanding, the difference between his position and that of the VBP-vādin is as follows:

  viparyaya-bādhaka-pramāṇavādin Jñānaśrīmitra’s interpretation of viparyaya-bādhaka-pramāṇa
anvaya   sādhane sādhyaviparyayasya bādhakapra-māṇa (VBP2) or
vyatireka sādhyaviparyaye hetoḥ bādhaka-pramāṇa (VBP1) sādhyaviparyaye hetoḥ bādhakapramāṇa (VBP1)

We can say that the position of the opponent, viparyayabādhakapramāṇavādin, is a propounder who claims VBP1 has the exclusive ability to establish pervasion only through proving the vyatireka-vyāpti by VBP1. Meanwhile, Jñānaśrīmitra regards VBP (VBP1/ VBP2) as having the synthetic ability to prove pervasion.

3. Priority of prasaṅga and prasaṅgaviparyaya

Jñānaśrīmitra continues:

(H) However, for us, with reference to the discussed proof, affirmative concomi-tance with inevitability (niyamavān anvayaḥ) is really shown by prasaṅga and prasaṅgaviparyaya prior [to other means of valid cognition] (paura-styābhyāṃ). And even if both are applied to the cloud presented as a similar example, if one thing (A1) has the essential property of creating another (B1), the former (A1) inevitably produces the latter (B1), and one thing (A2) does not produce another (B2), then the former (A2) does not have the essential property of creating the latter (B2). Thus, [anvaya and vyatireka], which are functioning in this way, show that all [entities], indeed, that are endowed with the ability of activity (kriyāśaktiyuktaṃ) are restricted to being momentary. Therefore, comprising all cases (sarvopasaṃhāro) is the ground for hinting at negative pervasion (vyatirekākṣepabījaṃ),[20] such as smoke and the like.[21]

Against the VBP-vādin, Jñānaśrīmitra argues for the inevitability of the affirmative con-comitance (anvayaniyama) first and foremost through prasaṅga and prasaṅgaviparyaya. In the above statement, the first point to note is the expression paurastyābhyāṃ; the second is that the statement is not concerned with pervasion in general, including vyatireka(-niyama), but only with anvayaniyama. As to the first point, the term paurastya means “prior to,” “first,” or “preceding,” which suggests that he does not necessarily exclude the logic of VBP, but insists only that prasaṅga and prasaṅgaviparyaya should be introduced before introducing VBP. That is, for Jñānaśrīmitra, prasaṅga and prasaṅgaviparyaya are enough for establishing affirmative concomitance with reference to the proof of momentariness, whereas the VBP-vādin considers VBP to be the sole means for establishing the pervasion. Moreover, the “comprising [of] all [individual instances] (sarvopasaṃhāro)” as a concept having a connecting function between anvaya and vyatireka is highly esteemed by Jñāna-śrīmitra. According to him, “if anvaya is established by prasaṅga and prasaṅgaviparyaya,” sarvopasaṃhāra is regarded as “the grounds for hinting at negative pervasion (vyatirekā-kṣepabījaṃ),” whereas the VBP-vādin does not make much of this. The VBP-vādin’s low evaluation of sarvopasaṃhāra suggests that, according to his/their view, vyatireka can be proven independently (concerning each subject of the thesis) as having inevitability and is sufficient for establishing a pervasion.

As mentioned at the beginning of this paper, the basic difference between Jñānaśrīmitra and the VBP-vādin is the evaluation of efficacy of the proof based on the pervasion that comprises all [individual instances] (sarvopasaṃhāravyāpti). The opponent, the VBP- vādin, apparently takes a negative stance toward sarvopasaṃhāravyāpti. On this point it is likely that the opponent, the VBP-vādin, is not Ratnākaraśānti, because Ratnākaraśānti takes a positive stance toward the significant role of the sarvopasaṃhāravyāpti for the proof of momentariness and his antarvyāpti-theory. He says, for instance, “as this pervasion comprises all [individual instances], it depends upon the universal,”[22] and, “for pervasion comprising all [individual instances] is in fact an indispensable constituent (aṅga) of the establishment of that which is to be proven.”[23]

4. Priority of anvaya over vyatireka

Furthermore Jñānaśrīmitra argues:

(I) Then, in this way, [the essence of] the pervasion [is that] “the pervading [property] is inevitably present in those which have the [property of being] pervaded.” And the establishment of momentary perishing cannot be repudi-ated on account of the efficacy itself of the reason “being existent” brought about by such a pervasion as that which has affirmative concomitance as its essential characteristic (anvayarūpā) and which is a property of the subjective element. Thus, it is considered “what is the use of introducing the means of valid cognition that annuls [the proving property] in the reverse [of that which is to be proven] in this case? (kim atra viparyaye bādhakapramāṇopanyāse-neti)” Precisely for this reason, even the impossibility of introducing this (=the means of valid cognition that annuls [the proving property] in the reverse [of that which is to be proven]) does not hurt [the inference].

This is because that which is to be proven is established by [the means of valid cognition] that annuls nothing but [the existence of ] the reverse [of that which is to be proven in the proving property] (viparyayasyaiva bādhakena). Even in the texts of our tenets, …, the pervasion that has affirmative concomitance as its essential characteristic (anvayarūpā) is established in detail, appearing as excluding the inconclusiveness [of the proving property], as with seeds and the like…[24]

Here we see Jñānaśrīmitra’s efforts to defend the affirmative relation, namely, anvaya.

  Viparyaya-bādhaka-pramāṇavādin Jñānaśrīmitra’s interpretation of viparyaya-bādhaka-pramāṇa Jñānaśrīmitra’s position
anvaya   sādhane sādhyaviparyayasya bādhakapramāṇa (VBP2) or prasaṅga and prasaṅgavi-paryaya (=?)
      sādhane sādhyaviparya-yasya bādhakapramāṇa (VBP2)
vyatireka sādhyaviparyaye hetoḥ bādhakapramāṇa (VBP1) (independent) sādhyaviparyaye hetoḥ bādha-kapramāṇa (VBP1)  

Concluding remarks

A cursory glance at the passages cited above seems to suggest that Jñānaśrīmitra criticizes the VBP maintained by the VBP-vādin. A detailed examination of the passage as a whole, however, leads us to the conclusion that the difference between his position and that of the VBP-vādin lies, in fact, above all in their evaluation of sarvopasaṃhāravyāpti rather than of VBP itself. Both of these issues are, of course, closely related to each other, at least in Jñānaśrīmitra’s theory. The VBP-vādin’s low evaluation of sarvopasaṃhāravyāpti suggests that the VBP-vādin is not Ratnākaraśānti, because Ratnākaraśānti takes a positive stance toward the significance of the sarvopasaṃhāravyāpti. Secondly, with reference to VBP itself, Jñānaśrīmitra claims the priority of prasaṅga and prasaṅgaviparyaya over VBP, whereas for the VBP-vādin VBP1 is the only way to establish pervasion. Here Jñānaśrīmitra tries to interpret VBP as a more synthetic and interrelated theory by expanding his interpretation of VBP (such that VBP is not only used for the proof of vyatireka but for that of anvaya as well). So it would seem that in all of his argumentation on VBP examined above, Jñānaśrīmitra seeks to establish the superiority of anvaya over vyatireka.

By this interpretation, we can easily understand and accept his positive statement concerning VBP in other contexts. For instance, in his Vyāpticarcā, after criticizing the bhūyodarśana of the Naiyāyikas, he says:

However, nothing but the viparyayabādhakapramāṇa, an inference, should be inducted (unneyam). If it (i.e., viparyayabādhakapramāṇa) is absent, it would be impossible to negate the occurrence (vṛtti) [of reason] in the dissimilar domain (vipakṣa).[25]

In the above statement, where the determination of causal relations is discussed, Jñānaśrī-mitra really evaluates VBP. In the Īśvara discussion as well, he states:

The means of valid cognition that establishes a pervasion is of only two kinds: perception and non-cognition or viparyayabādhaka, because both of these have perception and inference as their essential characteristics.[26]

In these passages Jñānaśrīmitra accepts the role of VBP, even if he regards it as a having a limited secondary function.

As to the relation between VBP and prasaṅga and prasaṅgaviparyaya, Jñānaśrīmitra gives priority to the latter over the former in the proof of the affirmative concomitance of momentariness. Is VBP2, as a genitive tatpuruṣa, different from prasaṅga or prasaṅgavipa-ryaya? If so, in which aspects is it different? And if the VBP-vādin is not Ratnākaraśānti, who might he be? These issues require further investigation.

References and abbreviations

Primary literature

AVS Antarvyāptisamarthana (Ratnākaraśānti): See Kajiyama 1999.

HB Dharmakīrtis Hetubinduḥ, Teil I: Tibetischer Text und rekonstruierter Sanskrit-Text. Teil II: Übersetzung und Anmerkungen, ed. Ernst Steinkellner. Vienna 1967.

HBṬ Hetubinduṭīkā of Bhaṭṭa Arcaṭa, with the sub-commentary entitled Āloka of Durveka Miśra, ed. Sukhlalji Sanghavi and Muni Shri Jinavijayaji. Baroda 1949.

JNA Jñānaśrīmitranibandhāvaliḥ, ed. Anantalal Thakur. Patna 1959,[27] 1987.

NMu Nyāyamukha (因明正理門論). Taisho 1628.

PSṬ Pramāṇasamuccayaṭīkā.

PVin2 Dharmakīrtis Pramāṇaviniścaya: Chapters 1 and 2, critically edited by Ernst Steinkellner. Beijing/Vienna 2007. Kyo Kano 189

PVSV Pramāṇavārttikasvavṛtti (Dharmakīrti): The Pramāṇavārttikam of Dharmakīrti, the first chapter with the autocommentary, text and critical notes, ed. Raniero Gnoli. Roma 1960.

RNA Ratnakīrtinibandhāvaliḥ, ed. Anantalal Thakur. Patna[28] 1975.

TBh(M) Tarkabhāṣā and Vādasthāna of Mokṣākaragupta and Jitāripāda, ed. H.R. Ran-gaswami Iyengar. Mysore 1952.

VC Jñānaśrīmitras Vyāpticarcā, Sanskrittext, Übersetzung, Analyse, ed. Horst Lasic. Vienna 2000.

VNṬ Vādanyāyaprakaraṇa of Acharya Dharmakīrti with the commentary Vipanchitārthā of Acharya Śāntarakṣita and Sambandhaparīkṣā with the Commentary of Acharya Prabhachandra, ed. Swami Dwarikadas Shastri. Varanasi 1972.

Secondary literature

Kajiyama 1966 Yūichi Kajiyama, An Introduction to Buddhist Philosophy. An Annotated Translation of the Tarkabhāṣā of Mokṣākaragupta. Kyoto 1966.

Kajiyama 1999 Yūichi Kajiyama, The Antarvyāptisamarthana of Ratnākaraśānti. Tokyo 1999.

Katsura 1981 Shoryu Katsura, Inmyo-Shorimonron-Kenkyu (4) [A study of Nyāyamukha (4)]. Hiroshimadaigaku-Bungakubu-Kiyo [A Bulletin of the Faculty of Letters, Hi-roshima University] (1981) 62–82.

Sakai 2014 Masamichi Sakai, Dharmottara on the viparyaye bādhakapramāṇa and trai-rūpya in Dharmakīrti’s sattvānumāna. Draft only, paper presented at the Fifth Inter-national Dharmakīrti Conference, Heidelberg 2014.

Shiraishi 2005 Tatsuhiko Shiraishi, Jñānaśrīmitra’s two interpretations of bādhakapra-māṇa. Draft only, paper presented at the Fourth International Dharmakīrti Conference Vienna, 24 August 2005.

Steinkellner 1967 See HB.

Steinkellner 1982 Ernst Steinkellner, Remarks on viparyaye bādhakapramāṇamDhar-makīrti’s Development of a Theorem. Draft only, paper presented at the First Inter-national Dharmakīrti conference, Kyoto 1982.

Steinkellner 1991 Ernst Steinkellner, The logic of the svabhāvahetu in Dharmakīrti’s Vādanyāya. In: Studies in the Buddhist Epistemological Tradition. Proceedings of the Second International Dharmakīrti Conference Vienna, June 11–16, 1989, ed. Ernst Steinkellner. Vienna 1991, 311–324.

Tani 1999 Tani Tadashi, Setsunametsu no Kenkyu [A Study of the Momentary Perishing of Existence]. Tokyo 1999.

Woo 1999 Woo Jeson, The Kṣaṇabhaṅgasiddhi-anvayātmika: An eleventh-century Bud-dhist work on existence and causal theory. Diss., University of Pennsylvania. Penn-sylvania 1999.

Footnotes and references:


Cf. Steinkellner 1982, 1991.


Cf. Tani 1999, Woo 1999, Shiraishi 2005, Sakai 2014.


Tani’s understanding of the VBP’s interpretations by Dharmakīrti’s followers, such as Dharmottara and Prajñākaragupta, are scattered throughout his work (1999), but are found in chapter 3 in particular.


For instance: First, VBP is sometimes identified with or regarded as being based on vyāpaka[dharma]-anupalabdhi, at least by Dharmakīrti and some logicians after him. Is VBP an inference based on vyāpakānupalabdhi or it cannot necessarily be identified with such an inference? Second, how is VBP related to prasaṅga and prasaṅgaviparyaya? If VBP is different from these, what is the difference between prasaṅga/prasaṅgaviparyaya and VBP formally, logically, and content-wise? Or is the latter the same as the former, especially prasaṅgaviparyaya, since both of them have the same style of proof formulation. The late Prof. Kajiyama assessed the matter as follows: “I think that Kamalaśīla was quite right when he identified viparyaye bādhakapramāṇam with a prasaṅga. … With Kamalaśīla, I also regard bādhakapramāṇa as prasaṅga in essence” (Kajiyama 1999: 37). He did not, however, mention the passage upon which his estimation depends.


JNA p. 60, 3–10: nāpy anaikāntikaśaṅkā, sarvopasaṃhāravatyā vyāpteḥ prasādhanāt. nanu viparya-yabādhakapramāṇavaśād vyāptisiddhiḥ, tasya ca nopanyāsavārtā. tat kathaṃ vyāptiḥ prasādhitety ucyate? …nanv evam anvayamātram astu, vipakṣāt punar ekāntena vyatireka iti kathaṃ labhyam iti cet. vyāptisiddher eva. vyatireke vyāptisiddhir eva katham it cet. (Ci’ RNA p. 70, 9–11 and 15–16: na cāyam anaikāntika, atraiva sādharmyavati dṛṣṭānte sarvopasaṃhāravatyā vyāpteḥ prasādhanāt. nanu viparyayabādhakapramāṇabalād vyāptisiddhiḥ, tasya copanyāsavārtāpi nāsti. tat kathaṃ vyāptiḥ prasādhiteti cet. … nanv evam anvayamātram astu. vipakṣāt punar ekāntena vyāvṛttir iti kuto labhyata iti cet. vyāptisiddher eva. vyatirekasandehe vyāptisiddhir eva katham iti cet.) [Different readings are in bold font.]


The relation between VBP and sarvopasaṃhāra has been studied recently by Sakai (2014).


JNA p. 60, 10–13: a→na, dvividhā hi vyāptiḥ, anvayarūpā ca kartṛdharmaḥ sādhanavati dharmiṇi sādhyasyāvaśyambhāvo yaḥ, vyatirekarūpā ca karmadharmaḥ sādhyābhāve sādhanasyāvaśyamabhāvo yaḥ,←a vyāptir vyāpakasya tatra bhāva eva vyāpyasya vā tatraiva bhāva iti nyāyāt. (a: Ci’ RNA p. 70, 17–20: na. dvividhā hi vyāptisiddhiḥ. anvayarūpā ca kartṛdharmaḥ sādhanadharmavati dharmiṇi sādhyadharmasyāvaśyambhāvo yaḥ, vyatirekarūpā ca karmadharmaḥ sādhyābhāve sādhanasyāva-śyamabhāvo* yaḥ.) [Different readings are in bold font.] (*Emendation. RNA p. 70, 19: sādhanasyāva-śyambhāvo.)


Another interpretation is as follows: Taking the context of the vyāpaka-vyāpya relation into consideration, kartṛ and karman can correspond to vyāpaka (i.e., sādhya) and vyāpya (i.e., sādhana), respectively.


JNA p. 60, 13–18 and 61, 2–3: a→enayoś caikasiddhir niyamena dvitīyapratītim ākṣipati. anyathā ekasyāpy asiddheḥ.←a kevalaṃ sākṣād ekatra pramāṇavyāpāro’nyatra nāntarīyakatayā pratītir iti tāva-taivaikavyapadeśaḥ. tathā ca saty ekatra pramāṇapravṛtteḥ pūrvam anyatra saṃśayaḥ pravartamāno’pi paścāt [tanyā?] tayā nirasyate. yathā viparyayabādhakapramāṇavādinām anvayaḥ sandihyamāno’pi vipakṣād ekāntena vyāvṛttibalāt nirastasaṃśayaḥ paryavasyaty akṣaṇike sarvathā’nupapadyamāneyaṃ sattā kṣaṇabhaṅganiyataiveti, …tasmāt sādhyābhāve sādhanasyaikāntiko vyatirekaḥ, sādhane sati sādhyasyāvaśyam anvayo veti na kaścid arthato bhedaḥ. (a: Ci’ RNA p. 70, 19–20: enayoś caikatara-pratītir niyamena dvitīyapratītim ākṣipati. anyathaikasyā evāsiddheḥ.) [Different readings are in bold font.]


JNA p. 61, 3–4: pramāṇavyāpāras tu mukhyato’rthataś ceti dvidhā vyavahāraḥ.


NMu 3a1–3 (Katsura 1981: 71–72): 若有於此一分己成、随説一分亦成能立、若如其聲両義同許、俱不須説、或由義准一能顕二。 PSṬ (Ms) 178b6 ad PS 4.5: arthāpattyā vetyādi. …; PVSV p. 18, 17 (=PVin2, p. 53, 10): arthāpattyā vānyatareṇobhayapradarśanād iti.


JNA p. 61, 4–6: ata eva yatra yatra sādhyaviparyaye bādhakapramāṇaparyeṣaṇā, tatra sādhane sati sādhyānvayaḥ, sādhakapramāṇaparyeṣaṇāpi tayopalakṣitā veditavyā. viparyayabādhakaśabdena punar ubhayasaṃgraha eva.


At the last Dharmakīrti conference in 2005, Shiraishi drew attention to this passage. Regretfully, his paper, titled “Jñānaśrīmitra’s two interpretations of bādhakapramāṇa” was not included in the proceedings of the conference, Religion and Logic in Buddhist Philosophical Analysis, 2011.


JNA p. 61, 6–8: vyatirekarūpavyāptau hi sādhyaviparyaye bādhakaṃ sādhanasya, anvayarūpāyāṃ tu sādhyaviparyayasya bādhakaṃ sādhane satīti. tadubhayābhāve sādhanasya sādhyapratibandhānupa-pattir ity ayam arthaḥ. Cf. Shiraishi 2005: 6ff.


HB p. 4*, 5–7. Skt. is reconstructed by Steinkellner. “Die Feststellung des gemeinsamen Vorkommens (anvayaḥ) ferner besteht beim Eigenwesen als dem Grund in dem Nachweis, daß die zu beweisende Beschaffenheit sich an das bloße Vorhandensein der beweisenden Beschaffenheit anschließt, weil [die zu beweisende Beschaffenheit] der Sache nach das Eigenwesen (bhāvaḥ=svabhāvaḥ) von etwas ist [das zugleich die beweisende Beschaffenheit zu seinem Eigenwesen hat]. Dieser [Nachweis] besteht im Auftreten eines Erkenntnismittels (pramāṇam), das den Grund im Gegenteil des zu Beweisenden aufhebt. Z.B.: Was seiend ist, das is ausschließlich augenblicklich; wäre es nicht augenblicklich, würde es, da [die Nichtaugenblicklichkeit] mit der Wirksamkeit in Widerspruch steht, die Dingheit verlieren, da [die Dingheit eben] durch diese [Wirksamkeit] bestimmt ist.” (Steinkellner 1967: 37)


We should also not overlook the expression sādhane sati. The expression is presumably intended for those who claim that pakṣadharmatā is not necessarily required, such as Ratnākaraśānti. AVS p. 86, 2–4: bādhakāt sādhyasiddhiś ced vyartho hetvantaragrahaḥ. bādhakāt tadasiddhiś ced vyartho dharmyantaragrahaḥ.


Cf. Steinkellner 1982: 2; 1991: 318. After Dharmakīrti, for instance: HBṬ p. 44, 24: etac ca bādha-kaṃ pramāṇaṃ vyāpakānupalabdhirūpam uttaratrāvasaraprāptaṃ svayam eva vakṣyati. VNṬ p. 10, 27–28: idam uktaṃ bhavati. vyāpakānupalabdhir eva sahabhāvaṃ bādhate hetoḥ sādhyābhāvena. RNA p. 83, 8: na ca viruddhānaikāntikate, vyāpakānupalambhātmanā viparyaye bādhakapramā-ṇena vyāpteḥ prasādhanāt. TBh(M) (Kajiyama 1966: 115, n. 309) … zhes pa khyab byed mi dmigs pa’i mtshan nyid can bzlog pa la gnod pa can gyi tshad mas (… iti vyāpakānupalabdhilakṣaṇa(/ rūpa)viparyayabādhakapramāṇena…) [This part is dropped in the Sanskrit text to which we have access. The Sanskrit above has been reconstructed by Kajiyama.]


JNA p. 61, 8–10: anyathā vyatirekaniścāyakapramāṇasya sākṣād alābhe’pi yadi sādhanasya sādhyenānvayaniyamaprasādhakaṃ pramāṇam upadarśayet, kas tasya sādhyasiddhau pratibandho vyati-rekasyāpi* sāmarthyākṣepāt. [*Corrected with Ms.; JNA: vyatirekasyānvayasyāpi.] Concerning the last part, various readings are possible. The difference in the interpretation of the sentence lies in the understanding of “tasya sādhyasiddhau” [tasya: (a) sādhanasya, (b) pramāṇasya; sādhyasiddhau: (a) when … is proven, (b) with reference to sādhyasiddhi].


JNA p. 61, 11–17: anyathā sādharmyaprayogo’nupādeya eva syāt, sākṣād vyatirekānupadarśanāt. tad yadi sāmarthyagamyo’pi na vyatirekātmā, tadā’pi yatra yatra sādhanaṃ tatra tatra sādhyam itīyataiva sādhane sati sādhyapratyāśāprayuktaḥ kṛtī kṛtārthaḥ. tadabhāve’bhāviny api hi sādhane sati yadi tadbhāve’vaśyaṃbhāvi sādhyaṃ tadā phalitaḥ sādhanasvīkāraḥ. tasmāt sādhanavādinā dvayam api paryeṣaṇīyaṃ vikalpena, vādinā caiko’pi prakāraḥ pramāṇasiddha upadarśayitavyaḥ, parasādhaneṣu ca vyatirekavad anvayo’pi niyamavān durlabha eveti veditavyam.


As to the usage of ākṣepa in the context of the relation between anvaya and vyatireka, see RNA 67.4: ākṣiptavyatirekā yā vyāptir anvayarūpiṇī / sādharmyavati dṛṣṭānte sattvahetor ihocyate // [The pervasion that has anvaya as its essential characteristic and by which [its] vyatireka is hinted at ]; 83.4: vyatirekātmikā vyāptir ākṣiptānvayarūpiṇī / vaidharmyavati dṛṣṭānte sattvahetor ihocyate // [The pervasion that has vyatireka as its essence and has the essential characteristic that [its] anvaya is hinted at … ] It seems that these two verses, which appear at the beginning of each chapter of Ratnakīrti’s Kṣaṇabhaṅgasiddhi, in the anvaya and vyatireka chapters, respectively, and have symmetrical construc-tion, were composed by the author (Ratnakīrti or someone else) (see Woo 1999: 141), being strongly conscious of the discussion of the relation between anvaya and vyatireka, which Jñānaśrīmitra makes here.


JNA p. 61, 18–22: asmābhis tu prakṛtasādhane niyamavān anvayaḥ prasaṅgaviparyayābhyāṃ paurastyābhyāṃ darśita eva, tau ca yady api sapakṣiīkṛte jalabhṛti pravartitau, tathāpi yo yatkaraṇasvabhāvaḥ sa taj janayaty eva, yo yan na janayati na sa tatkaraṇasvabhāvaḥ, ity evaṃ pravartamānau sarvam eva kriyāśaktiyuktaṃ kṣaṇikatve niyataṃ darśayata iti sarvopasaṃhāro vyatirekākṣepabījaṃ dhūmādivat.


AVS p. 64, 4: sā ca sarvopasaṃhārāt sāmānyam avalambate /


AVS p. 66, 3: sarvopasaṃhāravatī hi vyāptiḥ sādhyasiddher aṅgam.


JNA p. 63, 10–14: tad evaṃ vyāptir vyāpakasya vyāpyavati bhāva eveti kartṛdharmānvayarūpavyā-ptisaṃpāditasāmarthyād eva sattvahetoḥ kṣaṇabhaṅgasiddhir apratihateti kim atra viparyaye bādha-kapramāṇopanyāseneti samarthitam? ata eva tadupanyāsāśakyatāpi na pīḍayati, viparyayasyaiva bādhakena sādhyasya siddhatvāt. śāstre’pi pratha[mani??]yame ca vyāptir anvayarūpā bījādivad anekāntaparihāravyājena vistarataḥ prasādhitā. …


VC p. 33*, 7–9: viparyayabādhakam eva tu pramāṇam anumānam unneyam, tadabhāve vipakṣavṛttini-ṣedhasyāśkyatvāt,


JNA p. 293, 1: tac ca pramāṇaṃ vyāptisādhakaṃ dvividham eva. pratyakṣānupalambhaṃ vā, viparyayabādhakaṃ vā, anayoḥ pratyakṣānumānasvabhāvatvāt,....


Cf. Tani 1999, Woo 1999, Shiraishi 2005, Sakai 2014.


Cf. Tani 1999, Woo 1999, Shiraishi 2005, Sakai 2014.

Help me keep this site Ad-Free

For over a decade, this site has never bothered you with ads. I want to keep it that way. But I humbly request your help to keep doing what I do best: provide the world with unbiased truth, wisdom and knowledge.

Let's make the world a better place together!

Like what you read? Consider supporting this website: