As I noted last time, I think the disregard of ethics by Indian-philosophy scholars like Dan Arnold is a problem in itself: it’s a misconception of what philosophy is, and one that harmfully shrinks the field of the study of Indian philosophy. But I think this neglect would still be a problem even for people who do decide to restrict their study of Indian philosophy to the theoretical realms of metaphysics, epistemology, philosophy of language. For it seems to me that at least in Arnold’s case, the neglect of ethics leads to a misinterpretation of the metaphysics.
Arnold’s misinterpretation is focused above all around the relationship between the famous Buddhist “two truths”: conventional truth (saṃvṛti) and ultimate truth (paramārtha). Consider Arnold’s description (again in his review of Karen Lang) of the second chapter of Candrakīrti’s Catuḥśatakaṭīkā. “Candrakīrti develops (contra Vasubandhu) a characteristically Mādhyamika point to the effect that the conventional reality of pleasure is not denied, only its being the ‘inherent nature’ of life.” From this description, Candrakīrti’s chapter sounds like it is all about acknowledging pleasure and making room for it. You would not be able to tell that the point of this chapter, very explicitly stated at its beginning, is “rejecting the illusion of regarding the painful as being pleasant” – or that in this chapter, pretty much everything that we would normally consider pleasant turns out to be painful.
That point is an ethical point. But it seems to me that it feeds into Arnold’s interpretation of Candrakīrti’s theoretical philosophy. For when Arnold describes Candrakīrti’s theoretical philosophy in his book, he makes it sound as if Candrakīrti is rejecting the very idea of ultimate truth in favour of conventional truth – valuing the conventional in a way that his Pramāṇavādin opponents do not. But I would argue that Candrakīrti devalues the conventional as much as they do; he just happens to disagree with them about what the ultimate is. And I think Candrakīrti’s ethics help us see that point.
On Arnold’s account, Candrakīrti would seem to be advocating more or less the same view as Jayarāśi, the Lokāyata skeptic, and I cannot imagine Candrakīrti being anything but horrified by that conclusion. Yes, Jayarāśi and Candrakīrti do use similarly skeptical means – but to diametrically opposite ends, ends which seem to play very little role in Arnold’s discussion.
So Arnold says Candrakīrti “defers” to the conventional (saṃvṛti), even to “ordinary intuitions”; “there is a sense in which Candrakīrti’s deference to the conventional is itself the argument.” (65, emphasis in original) When Arnold spells out this claim further, it is more cautious and avoids the language of “deference”: he says that for Candrakīrti
there is nothing “more real” than the world as conventionally described – or, more precisely, that there can be no explanation that does not itself exemplify the same conditions that characterize our conventions.
While the “more precise” characterization of Candrakīrti is not wrong, it is misleading, and doubly so when it is described as “deference”. What Candrakīrti is doing is not saving the conventional world but disparaging explanation itself. Arnold himself (page 183) acknowledges Candrakīrti’s claim that the ultimate (paramārtha) “cannot be taught, nor can it be known” (nopadiśyeta na câpi jñayeta, Prasannapadā 493.10-11). It is beyond learning and language, but qua ultimate truth it is more real than conventional truth – to the extent that anything can be said about it at all.
So I think it is absolutely false when Arnold claims that “for Candrakīrti, the only ultimate truth is that there is no ultimate truth – that the ‘ultimate truth,’ in other words, is the abstract state of affairs of there being no set of ‘ultimately existent’ (paramārthasat) ontological primitives like the dharmas of Abhidharma.” (184) Here, immediately after acknowledging Candrakīrti’s claim that the ultimate “cannot be taught, nor can it be known”, Arnold turns around to express the ultimate truth as a clause of a sentence, “that there is no ultimate truth” – exactly the sort of thing that can be taught and known!
John Dunne had offered a more convincing account of Candrakīrti’s thought in his “Thoughtless buddha, passionate buddha” (JAAR 54.3), an account that Arnold mentions but never actually refutes, just batting it aside to offer his own interpretation. Arnold’s interpretation, bizarrely, relies on adopting what he calls a “type-theoretical approach” even though “nothing like this approach is taken by Mādhyamikas” and it is “not being in the sprit [sic] of Madhyamaka…” (185)
It seems to me far more plausible when Dunne tries – and I think succeeds – to take an interpretive approach that is in the spirit of Madhyamaka, and has the distinct advantage of being in line with what Mādhyamikas actually say. Dunne (541-2) rightly points out Candrakīrti’s etymology of the term saṃvṛti (Tibetan kun rdzob, what we are rendering “conventional”) in the Madhyamakāvatara and its commentary:
Since confusion (moha / gti mug) obstructs one’s realization of essence, it is called “saṃvṛti“. The Sage said that a saṃvṛtisat thing is a fabricated (thing) that appears to be real due to saṃvṛti; a fabricated thing that appears to be fabricated is saṃvṛti. “Confusion” is that which confuses beings about seeing things as they really are; that confusion is ignorance. It makes one superimpose a non-existent essence onto things, and by nature, it obscures one’s experience of (actual) essence; hence, it is “saṃvṛti”. (MA 6:28, MAB 54b)
Etymologically, Candrakīrti uses the literal meaning of saṃvṛti as “covering” to explain it – as something which obstructs understanding of how things really are. (For that reason Dunne even translates saṃvṛti, with considerable justification, as “the spurious”.) It is no closer to the way things really are than the ultimate truth is; indeed it seems likely to be further away.
Now Arnold is correct that Candrakīrti criticizes the view of Dignāga, Dharmakīrti and the Ābhidharmikas according to which atomized parts are ultimately real. Against Dignāga, Candrakīrti would deny that any given thing or set of things is ultimately real. But that is a very different thing from “deferring” to the conventional. The conventional is the source of our problems. For Candrakīrti just as much as for Dignāga, we must get away from the conventional and to the ultimate; it’s just that what constitutes the ultimate for Candrakīrti is something very different, something unsayable. Likewise Candrakīrti does say, and Arnold is right to point out, that “worldly conventions are not to be totally annihilated…” (Madhyamakāvatāra, quoted in Arnold 165-6) This is true. But this is not because they’re truer than anything ultimate; it’s merely because you need them as expedient means to deal with people who haven’t realized they’re ultimately false.
Over and over, Arnold attributes to Candrakīrti a metaphysical approach deeply at odds with most of what Candrakīrti actually says. How could such a distinguished scholar, who has clearly read the Prasannapadā and Madhyamakāvatāra at such great length, get Candrakīrti so wrong? I suspect that Arnold would have been able to offer a far more plausible interpretation of Candrakīrti if he had paid attention to Candrakīrti’s ethical claims – which are present in the works Arnold studies as well as in the Catuḥśatakaṭīkā. (For example, at Prasannapadā 155b-156a, Candrakīrti says first to reject karmically bad actions, then to reject theories that identify the self with the aggregates, and then reject all attachment – by understanding that all things are empty of inherent nature.) For Candrakīrti’s ethics is built on the idea that our everyday conventional understandings are gravely in error – our everyday metaphysical understandings, our understandings of the nature of things. The source of our suffering is in the conventional understanding of things as permanent, pleasurable, pure and self. If we are to end suffering – which is the whole point – then we must get away from that convention.