Chapter 10: Dependent Origination¶

This chapter is written in the first person, and stands outside the proof.

As before, a word at the outset. This chapter is my own history, not an argument. Whether the theorem of this part holds has nothing whatever to do with how I came to stand in front of it; lift this chapter out entire and not a word of the preceding nine needs changing. The last chapter of Part 13 recorded a stretch of my history with the early teaching, and closed owing a sentence: the other vertigo Madhyamaka gave me, and the other medicine, would be left to the parts on Madhyamaka. This is that debt paid.

The title carries two senses. Dependent origination is what this part has been about from beginning to end; it is also, in Chinese, the ordinary word for how a thing came about, that is, the origin story of a book or a piece of writing. This chapter takes the second sense.

Like most readers of the Chinese, I grew up in the territory of the northern transmission. Buddhism was not, at first, something I chose; it was the air: the look of the monasteries, the names of the bodhisattvas, a few verses everyone can recite — all of it there before I knew what I was taking in.

Toward the Mahāyāna I carried an unease I could not name. It was not directed at any one school. It had the shape of a question: a religion that takes as its charge the rescue of beings from suffering — why does it end up, every time, watching how a person talks? Watching how he words it, how smoothly he can turn a view, and ruling on whether he has awakened from that. How does a language examination come to stand between awakening and its absence?

The other face of the same question is the pair of verses. Shenxiu says: wipe it clean, hour by hour. Huineng says: from the first, not one thing. The second is ruled the higher, and the ruling is that it goes further. But wiping is something to be done, and not-one-thing is something to be said. Why should the sentence said rank above the act done? (For the textual layers of that famous contest, see Part I, Part 10, Chapter 8: the Dunhuang manuscript has Huineng's verse read "the buddha-nature is ever pure," not the received "from the first, not one thing," and the contrast between the two verses is nothing like as sharp as it later became. I learned this philology only afterwards; what pressed on me then was the received line.)

And a third layer, the sharpest. If from the first there is not one thing, then there is nothing. Yet whenever anyone takes it literally and asks whether there is or is not, the fluent turn around at once: it does not mean that there is nothing. Then what does it mean? One sentence, which reads the same way said forward and said backward, with both ends held by the one saying it. That does not look like insight. It looks like rhetoric.

Carrying these three, at some point I ran into Madhyamaka.

Reading the Mūlamadhyamakakārikā the first time, I did not understand what it was saying. All I had was a coarse sense that something was off. That sense turned out to be right; I had simply put it in entirely the wrong place.

So I did something that looks fairly comic now and that I do not regret: I set out to refute it. The aim was definite: find a self-contradiction in emptiness. If everything is empty, is that sentence empty? If the claim is absence of intrinsic nature, is that absence not itself a nature? I did not invent these objections; I had simply asked my own way to them.

Three months. Discussion, question and answer, attack and defence, argument, round after round.

I failed.

The quality of the failure needs stating precisely. I did not find a self-contradiction in the text. Every stroke I took for fatal, once it landed, met a distinction of syntax already prepared: rejection is not the assertion of a negation, a negation does not posit its counter-thesis, the emptiness of emptiness is not a hole but a clause. Each of these later found its place in this part: Lemma 1 in Chapter 3, the hierarchy clause in Chapter 5. The blade was blunt; the target was not dodging.

Here is the strange part: I lost, and the unease grew deeper. By rights, a system that takes three months of attack without a scratch should have won me over. It did not. The unease was not dispelled but sharpened, except that I still could not say what it was pointing at.

So I did something stupider still: I took the text and worked it through as a modern proof, clause by clause. Definitions written as definitions, derivations written as derivations, every premise used at every step listed out, to see whether it would stand. What came of that is the part the reader now holds.

It stands. It stands more solidly than I had expected: the definitions are taken word for word from criteria the opponents apply themselves, the derivation draws only on logic, and the Buddhist material serves two offices in the whole part and no more — provenance for the definitions, historical actuality for the reductios. After that I had no reservation left about the logical rigour of the theorem.

And it was precisely by finishing that work that I finally saw where the unease lay: not in the conclusions. The conclusions are correct. What troubled me was something else.

For a while after that I fell into a pit, and it is the deepest pit the text digs for itself.

With a tool like that in hand it is very hard not to use it. In any argument, whatever the other party puts forward, I could dissolve it: this thing you are asserting — is it established in dependence on other things, or does it stand on its own? If in dependence, it is not the thing you wanted; if on its own, it has cancelled same and different, and you cannot so much as pick it out. It never failed. And when the other party fell silent, I knew perfectly well what rose in me.

After more times than I can count, I stopped one day and looked at myself. What I saw was this: I did not care what was left standing after each dissolution. What I cared about was the dissolving.

This is functional attachment. It does not cling to a view. On the contrary, what is being held is the technique for cancelling views: not a thesis clutched, but the tool called "no thesis." It is harder to detect than clinging to a view, because on the outside it is empty: what is talked is emptiness, what is said is emptiness, what is done is the dissolving of the question. It does not show. It is also undeniable.

The matter has its formal place in this book: Volume II, Madhyamaka, Part Two, Chapter 7 treats functional attachment and the ultimate truth, and Chapters 1 and 2 of the same, on suffering and on its arising, set out the mechanism: Madhyamaka moves the working surface from pre-reflective craving to view at the reflective level. I had seen that migration happen in myself first. Not one measure of my craving had gone; it had only changed object, from "I want to be right" to "I want to be able to dissolve you."

I dare state this judgement precisely because I am someone who walked the derivation through again from outside Buddhism. The excursion of Chapter 8 tested more than the theorem's scope; it tested its quality. A tool that works just as well where there are no Buddhist premises at all is thereby shown to be a logical instrument and not a soteriological one. It is a philosophy, a constraint, marking out what may not be said and what combinations may not be posited. It is not what its advertising promises. It is not a road that ends in the ending of suffering. This is no disparagement. The law of non-contradiction is extremely useful too, and no one expects to be released from birth and death by it.

With that groundwork, the Diamond Sūtra and the Heart Sūtra opened where they had been closed. "What is called X is not X, and that is why it is called X" is no longer mystifying; it is a layering: reject the real-existence reading of X, preserve the force of X at the level of names. The string of negations in the Heart Sūtra stops being frightening: it is a list of rejections, not a proclamation of annihilation. That fluent speaker of my youth had in fact been half right (it truly does not mean that there is nothing); he simply could not say at which level the other half was being spoken, which is why it sounded like rhetoric. The Mūlamadhyamakakārikā writes that level out.

And because it writes it out, I could finally see what that level can do and what it cannot.

The result of this whole part is that my confidence in the logical core is greater than it was, not smaller. It is a genuine theorem, not a bound volume of reductios. And I have put my name to the sentence that closes Appendix II of Volume II: the more perfect its philosophical form, the more respectable — and greater — the distance from the original question "how does suffering cease?" Madhyamaka stands far from soteriology; far enough that an entire volume is needed to account for the distance.

More than that: for a certain kind of person it may not be help but hindrance. My kind: intellectually avid, good at argument, delighting in working something out. In the last chapter of Part 13 I wrote that toward logic and intelligence themselves I still carry a great craving. When a person like that is handed this text, the readiest destination is not cessation but promotion: attachment moves up from content to method, from view to the technique for dissolving views. After the promotion it is harder to notice, because it looks exactly like letting go.

So the verdict of this chapter is a single sentence. The text remains an extremely effective tool, and in this part I have made the strongest defence of it I am able to make; but it should stop at the position of a tool, at the position of a constraint.

Chapter 9 closed with the simile of the raft: one needs a raft to cross the river, and does not carry it on one's back after landing. When I wrote that passage I was speaking of the theorem's self-limitation. Writing this chapter, I admit that the sentence was addressed first of all to me. I carried the raft on my head for a long time, and carried it rather pleased with myself.

The preceding nine chapters are the proof I made for this theorem. This chapter is the proof this theorem made of me.

References¶

This chapter is a personal record in the first person and cites no new material. The theorem, lemmas, clauses, and verdict referred to are set out in Chapters 1 through 9 of this part and in Appendix A; for the textual layers of the two verses see Part I, Part 10, Chapter 8; for functional attachment and the migration of the working surface see Volume II, Madhyamaka, Part Two, Chapters 1, 2, and 7; the sentence "the more perfect its philosophical form, the more respectable — and greater — the distance" closes Appendix II of that volume.