Chapter 2: Candrakirti — The Consequence Line Completed¶

Candrakīrti takes over the case of Chapter 1, finds for Buddhapālita, and supplies the half-sentence that Buddhapālita had left unsaid. Neither of his two principal works, the Prasannapadā and the Madhyamakāvatāra, has an ancient Chinese translation; this chapter restates their skeleton throughout from the critical editions and the modern translations, and gives no quotations.¹

His grounds for refuting independent inference may be reconstructed as follows. For an independent inference to stand, the terms of the thesis and the property serving as reason must be admitted by both parties; and in the domain of ultimate discourse, what the realist admits is exactly what is under examination, while what the Mādhyamika admits is exactly what the other party does not have, so that there is nowhere for a base of common admission to be laid. If Bhāviveka's inference carries its qualification out strictly, then the conditioned in the thesis is already the conditioned as qualified with respect to true nature, and that is a term the opponent does not recognize; and if it does not qualify, then what is laid down is not his own position. In this domain, therefore, the very ticket of admission for a positive inference cannot be issued. The shape of this criticism deserves notice: it is of one form with the discipline of warrant in Chapter 2 of Part 18. That discipline says that the target must be set up by the opponent himself if the proof is to have any force against him; Candrakīrti turns the same discipline round and applies it to the framing of one's own thesis: common admission is warrant, and where there is no common admission there is no right to a thesis.

As to why a reductio is exempt from the duty of laying down an inference, Candrakīrti speaks the half-sentence that Buddhapālita did not: one who draws a reductio lays down no counter-thesis, and to fasten a commitment onto his words is the opponent's misreading. Translated into notation, this is Lemma 1 word for word: ⊣A does not entail ⊢¬A. To reject your proposition is not to assert its negation; hence there is no thesis to be inverted, and the danger of inversion lifts of itself. Only at this point is the consequence line complete in its self-understanding: Buddhapālita supplied the practice, and Candrakīrti supplied its instructions for use.

The sevenfold analysis of the Madhyamakāvatāra is the standard exhibit of this line. The chariot and its parts are taken through seven relations one by one (identity, difference, this depending on that, that depending on this, possession, the collection, and the shape), and the chariot is not to be found; while at the level of names and conventions the chariot takes the road as usual, and one calls for a chariot and it comes, mounts it and it goes. The form is the same as that of the two samples of Chapter 4 of Part 18: the goer is individuated by the going, the chariot is individuated by its parts and its uses, the E3 reading is empty under all seven searches, and the E2 level is not damaged by a single step.¹

Candrakīrti's attitude to the conventional takes its settled shape from this, and may be called the lawfulness clause: ultimate analysis does not come before the court of the conventional, worldly names and conventions run by the world's own standards of acceptance, and within those names and conventions the Mādhyamika conforms to the world. This corresponds almost item for item with direction (a) of the Preservation Theorem: the rejections carry a type mark and are applied to D3-type claims alone, and not a hair of the truth conditions of everyday assertions is touched. The difference must be recorded as well: Preservation (a) is a theorem that has been proved, whereas the lawfulness clause is in Candrakīrti a methodological position, its place of proof being filled indirectly by the overall success of reductio.

The verdict in two sentences. First, the completed form: the warrant is reduced to a single source, ⊢¬π is obtained by reductio alone, and, fitted out with a self-explanation of the kind Lemma 1 supplies, the line has no crack left in it from within. Second, the price: reducing the warrant to a single source shuts the gate on dialogue with the community of Buddhist logic, and the public checkability that Bhāviveka wanted is left with nowhere to lodge. Whether that gap is a real gap turns on whether the two words common admission can be supplied from some other quarter; the inventory of resources of Part 18 bears directly upon this, and the reckoning is held over to Chapter 7.

References¶