Chapter 0: Guide to This Part¶

The philosophy of the Sarvāstivāda has one convenience about it: it states its own assertions very plainly. All exists means that whatever is admitted to be a dharma exists in the past, the present, and the future alike. The sentence itself is not hard to understand. What is hard is the train of supports standing behind it: by what does the substance of a dharma keep on existing always, by what is the division of the three times drawn, by what do action and defilement hang upon a single continuum. It is that train of supports that this part sets out to look at.

This guide first explains why this part must exist, then gives an account of who is speaking to whom inside the library that lies before us, and last sets up a single table running through the whole part.

Why this part must exist.

Begin with a difference between this part and Part 14. Part 14 had to take some trouble to explain why it should exist at all: seen from a distance, Theravāda Abhidhamma is a vast taxonomy, dense with entries, turned toward insight practice, and its metaphysical outline lies buried in the procedures of cataloguing, invisible unless one takes the cells apart one by one. On the Sarvāstivāda side the case is exactly the reverse. The real existence of the three times and the everlasting existence of the substance of a dharma are set out in the open, and anyone who has read in the sectarian period knows that such a set of assertions exists; the school has never concealed what it claims to be or what it means to accomplish. The tasks of the two parts are therefore different. Part 14 had to trace an outline that was buried. This part has to see whether an outline already drawn in the open can hold itself up.

First, half of the opponents of the parts that follow are here. The intrinsic nature that Madhyamaka sets out to demolish, the seeds and the continuum that Yogācāra sets out to establish, and the terminus that tathāgatagarbha thought sets out to supply all became problems only because this set of Sarvāstivāda assertions came first. To read the Theravāda part alone is to have seen half a base map; only with the two together do the opponents of the later parts acquire faces.

Second, the verdict at the close of Part 14 can be sustained only against this part. That chapter found three vacancies standing on the Theravāda list. Whether those vacancies are the oversight of one school or the boundary of the Abhidharma style of analysis itself cannot be decided by looking at one school alone. The Sarvāstivāda filled something into each of the same places, and both the manner of the filling and the consequences of it lie there in the open. Only in the comparison does it become clear whose problem those three places really are.

Third, readers are too familiar with this vocabulary, and familiarity itself gets in the way of examination. The doctrinal groundwork of Chinese and East Asian Buddhism descends from the line of the Abhidharmakośa, and dharma, substance, function, cause, condition, unmanifest, and seed entered ordinary Buddhist Chinese long ago. They come readily to hand, so readily that few people stop to ask what each of them presupposes. This part means to put these familiar words back into the dispute out of which they were made.

How to read this part. Part 14 offered three ways through it, because four of its chapters were reference lists from which the conclusion alone could be taken. This part has no such stratification. Details of certain items among the seventy-five dharmas may be passed over, that roll having been given in full in Part I, Part 3, Chapter 2 and not set out again here. Apart from that, the chapters are a single argument pushed through from one end to the other, in which the conclusion of one section is often the premise of the next, and reading by jumps will break the thread.

The temper of that argument is a good deal unlike the temper of Part 13, and it does no harm to be prepared for the difference before beginning. Part 13 faced a body of texts unwilling to say everything, and the labour went into silence and simile: where the Buddha gave no answer the metaphysical outline had to be recovered by working backwards, and the writing accordingly kept close to the scriptural sentence and to the occasion. This part faces a body of texts that say everything—thesis, ground, refutation, and reply, with every side leaving a written argument behind it. The labour therefore goes into examining item by item: watching who concedes at which point, which sentence reports another's doctrine and which states the author's own, and what one concession disturbs elsewhere. To read Part 13 is to supply what was not said. To read this part is to audit what has been said.

Textual layers.

The Sarvāstivāda is not a book but a library, and the library is not of one mind throughout. Before reading any sentence of Sarvāstivāda doctrine one has first to know which layer it comes from.

What the Chinese tradition calls Kośa doctrine is often taken for the representative work of the Sarvāstivāda, and it came in fact from the hand of a critic. Wherever this part cites the Kośa it must be made clear which sentence is reporting Vaibhāṣika doctrine and which is Vasubandhu speaking for himself; the latter are very often exactly where the school is sore, and Saṅghabhadra's replies fall in the same places. The two layers of broad exposition and turn are held over until after this part: first the question of how an orthodoxy came to be adjudicated into existence, then the question of why a man who had finished arranging the whole system walked away from it. This part does no textual history, and draws on those two only where a statement of doctrine is wanted.

The handling of technical terms. This part takes the received Chinese technical vocabulary as primary and renders it in English by settled equivalents, glossing the Sanskrit at first occurrence and keeping to the one equivalent thereafter. Modern words are charged with nothing beyond showing the way; having brought the reader to the door they ought to withdraw. To say that the substance of a dharma is like a substratum, or that possession is like a title of ownership, will help a reader take the first step, and two steps further it will have carried in presuppositions that do not belong to this system. The introduction to Part II set down a warning on this point, and this part in particular has to keep to it.

One table for the whole part.

Part 14, Chapter 7 listed three groups of vacancies on the Theravāda list. Those three groups are copied into the left column, and the right column holds what the Sarvāstivāda filled into the same places. This table is the road map of the part, and the chapters walk through one or several of its rows in order.

The table still lacks a column: the problem that each of these fillings itself gives rise to. That column is held over to Chapter 7, because only when all seven chapters have been walked through does it become clear whether these fillings are compatible with one another.

Data and basis. The numbers of dharmas, the nomenclature, and the classification of causes, conditions, and fruits in this part follow the Abhidharmakośa and the Mahāvibhāṣā, with divergences between the two marked in place. Secondary scholarship follows chiefly Collett Cox and K. L. Dhammajoti, the former working on the dharmas dissociated from mind and the judgement of real existence, the latter having given the most complete survey of the Sarvāstivāda system as a whole. Any item this part has not yet settled is marked in the notes as to be checked; nothing is made up to fill a gap.

Interface with Part I. Part I, Part 3 dealt with the historical position of the school, the descent of its literature, and the nomenclature of the seventy-five dharmas, and none of that is repeated here. This part does one thing only: it takes what has already been listed as an assertion awaiting examination, and asks of each in turn on what ground it stands and at what cost. For the course of the sectarian division, the background to the compilation of the Mahāvibhāṣā, and the personal relations between the Sarvāstivāda and the Dārṣṭāntikas, please return to Part I, Part 3.

To Chapter 1. Chapter 1 begins by fixing limits on the phrase all exists. That slogan has been read in a good many senses over the centuries, and not a few of them are not what the Sarvāstivāda meant; unless one first marks off what it does assert and what it does not, all six chapters after it will fall on empty ground. Having fixed the limits, Chapter 1 has also to hand over the two measures used throughout this part: how substance is divided from function, and how the real is told from the nominal. Once those two measures are set up, there is something to be said about why the seventy-five dharmas are these seventy-five.