For Sarvaastivaada

·期刊原文
FOR SARVAASTIVAADA

By David Bastow
Philosophy East and West
Volume 44, Number 3(July 1994)
P.489-499
(C) by University of Hawaii Press

--------------------------------------------------------------------------------

P.489
I

The Sarvaastivaadins believed that there is a

sense in which basic entities, dharmas, of the past

and future are real, that is, are real now. Dharmas

are short-lived or momentary happenings. The

Sarvaastivaadins claimed that the present happening

of a dharma is merely one phase in its existence.

Hitherto it has existed in its future phase; when

causes and conditions are ripe, it moves into its

present phase, and then when the moment of its

actualization is past it moves into the third phase

of its history and becomes a past dharma. So at the

present moment there exist not only present dharmas,

that is, dharmas in their present phase (bhaaba),

but also dharmas in their past and future phases.

The experienced difference between these three modes

of present existence is explained by a complex

theory of different types of causal efficacy; the

present bhaava is characterized by a particularly

strong kind of causality called kaaritra; dharmas in

past and future bhaabas are also causally

efficacious, but in a different way. So, in this

sense, past, present, and future dharmas are

contemporaneous; but it is a 'reduced

contemporaneity', for dharmas in past and future

phases are less forceful than dharmas in the present

phase. This doctrine was in direct opposition to the

opinion of the Sthaviras and the Sautraantikas, that

before its present moment of existence a dharma was

nothing, and after this moment it will become

nothing (abhuutvaa bhavati, bhutvaa ca

prativigacchati).(1)

Why did the Sarvaastivaadins hold their peculiar

view about the real existence of past and future?

This question expresses two linked types of

puzzlement: to the philosopher the Sarvaasti

position may well seem inherently implausible, even

bizarre; and the historian of Buddhist thought may

find it difficult to understand why the matter

should have been of any special concern to

Buddhists. Why as Buddhists should the Sarvaasti

philosophers have been concerned to assert the

present existence of past and future dharmas? While

I am addressing the question of the Buddhist

importance of the theory, I am also, especially in

the final section, concerned with the more general

philosophical question. A first step in

understanding why the Sarvaastivaadins held their

'sarvaasri' position must be to look at the

arguments they themselves produced to defend this

position.

Historically we know of three phases in the

development of their views: the original abhidharma

texts (second century B.C.); the great commentary on

these known as the Mahaa-vibhaa.saa (first century

A.D.); and the vigorous debate which arose from

Vasuhandhu's account of their views, with mainly

hostile commentary, in the Abhidharma-ko`sa (fifth

century A.D.).

P.490

Four types of argument are to be found in these

various sources. One of these types is the subject

of this essay, but the other three should first be

identified.

1. The one most commented on by modern writers

occurs in various forms in the abhidharma text

called the vij~naanakaaya,(2) and is the basis of

three of the four arguments presented in the

Abhidharma-ko`sa. This argues to the present reality

of past and future dharmas from present abilities to

know the past and future, or to have them as objects

of consciousness (vij~naana).(3)

2.The Vij~naanakaaya also gives a series of

arguments about the present true descriptions of a

person's mental state. On the one hand one may refer

to what is straightforwardly present, the thoughts

currently going through the person's mind. But one

can also truly say of a person that he has a certain

virtue or vice, or is at a certain stage on a path

of development, when what is currently going through

his mind may have nothing to do with these. Such

'dispositional' descriptions refer, the argument

implies, to past and future manifestations; for the

descriptions to be true now, these past and future

dharmas must in a sense be real now.

3. The final argument presented in the Ko`sa claims

that if past dharmas were not real, they could not

be causally efficacious in bringing about a present

effect.

Each of these arguments raises deep

philosophical issues; none can be summarily

dismissed. In the present essay I wish to

concentrate on a further type of argument, perhaps

simpler and more profound than any of the ones

above. As far as I know it has not been commented on

at all in the modern literature, perhaps because it

has been confused with one or other of the types of

argument presented in the Ko`sa.

II

My interest is in the principal argument

propounded in the Mahaavibhaa.saa. It is put in

three different ways. I shall first present the

three variants with little comment, and then present

my own explanations of and philosophical

observations on them.

The first version of the argument is about the

relation between a karmic action and its consequence

in experience. I quote from Louis de la Vallee

Poussin's translation (Poussin [1937], pp. 9-10):

Quand existe une cause de retribution

(bipaakahetu) presente, le fruit obtenu par

cette cause est-il present, passe ou futur? Si

vous dites qu'il est passe, nous concluons que

le passe doit exister; si vous dites qu'il est

futur, nous concluons que le futur dolt exister;

si vous dites qu'il est present, nous con cluons

que la cause de retribution et son fruit doivent

etre simultanies. Contradiction, car la stance

dit:

P.491

Celui qui fair /e mal ne le sent pas

aussitot....

Si vous dites que re fruit de cette cause ne se

trouve dans aucune des trois epoques, nous

concluons que cette cause n'a pas de fruit, car

le fruit de retribution n'est pas inconditionne;

si elle n'a pas de fruit, la cause aussi

n'existe pas, comme une deuxieme tete ou une

troisieme main.

De meme nous demandons a quelle epoque

appartient la cause qui correspond au fruit de

retribution (vipaakaphala) priesent....

If it can be truly said of a present dharma that

it is cause or fruit, then its corresponding fruit

or cause cannot also be present; and to say that

this corresponding fruit or cause does not exist in

past, present, or future is tantamount to saying

that it does not exist at all. Being a conditioned

dharma, it certainly does not exist out of time. And

if the existence of past cause or future fruit is

denied, then the original hypothesis must be denied,

of the existence at the present time of dharmas

which are cause or effect, vipaaka-hetu or

vipaaka-phala. So, given this hypothesis--which

surely is noncontroversial--past cause or future

fruit must exist.

Several of the arguments for the Sarvaasti

position relate in one way or another to the karmic

relation between cause and effect, between action

and its fruition. One of the arguments in the

Vij~naanakaaya rests on a cognitive possibility,

that of seeing the future latent, as it were, in the

present. The fourth of the arguments in the

Abhidharma-ko`sa rests on the present efficacy of

past causes. But the Vibhaa.saa argument is even

simpler than these. Its point is that the very

nature of the present as cause (vipaakahetu)

incorporates the future as vipaakaphala. The very

nature of the present as vipaakaphala incorporates

the past as vipaakaheru. The present is what it is

only in virtue of its relations to past and future.

It exists as conditioned present only if there is

something nonpresent to which it is related.

The same idea is pursued in the next Vibhaa.sa

argument. It is explained not in abhidharmic terms

but in ordinary religious language. The argument is

again very straightforward (Poussin [1937], p. 10):

Si le passe et le futur ne sent pas reels, il ne

peut y avoir sortie du monde et profession

religieuse (pravrajyaa et upasa.mpad). II y a

une stance:

Si on soutient que le passe n'existe pas, le

Bouddha passe n'existe pas: done manqueront

sortie du monde et profession.

Encore, se le passe et le futur ne sent pas

reels, il faut que les religieux (pravrajita),

qui possedent le savoir exact, tiennent des

discours de mensonge. II y a une stance:

S'il soutient que le passe n'existe pas et si,

cependant, il parle d'annees peu nombreuses ou

nombreuses [depuis la profession], il dolt, de

jour en jour, accroitre son savoir

exact(samyagj~naana) et ses discours faux et

trompeurs.

P.492

The point here is not about the monk's ability to

grasp the past in his thoughts, but about what is

true--in particular what is true about the present.

In many ways the monk's present status, for example

that he is indeed a monk, a pravrajita, rests on the

reality of a past event, his own decisive action and

commitment. The fact that he is now ordained depends

on the possibility of tracing back a line of

ordination to the Buddha himself. In these ways, we

may say, the past lives on in the present.

The final version of the argument puts it in a

more abstract form (Poussin [1937], pp. 10-11).

Si le passe et le futur ne sent pas reels, le

present aussi n'existe pas. Car le present 'est

designe' (praj~napyate) en consideration

(apek.sya) du passe et du futur. Si les trois

epoques manquent, manquent les dharmas

conditionnies. Done manquent aussi les

inconditionnes qui sent etablis en consideration

des conditionnes. Les uns et les autres

manquant, manquent tous les dharmas. Done

manquent aussi la delivrance, la sortie, le

Nirvaa.na.

The use of 'praj~napyate' might suggest that the

argument was about the logical interdependence of

the concepts of present, past, and future; but the

surrounding text makes it clear that what is at

issue is the real, not the nominal, existence of

past and future. Past, present, and future are

essentially related, such that the reality of the

present is impossible without the reality of past

and future.

In fact, if we are to take seriously the

position adopted at the beginning of this section of

the Vibhaa.saa (and perhaps confirmed in the third

sentence of the quotation above) that time is not a

separate thing from temporally existing conditioned

(sa.msk.rta) dharmas, then this third variant is not

really about past, present, and future as such, but

about the interrelation of past, present, and future

dharmas. In that case the third variant relies on

the force of the first two variants.

Of course, the Sautraantikas and the Sthaviras

complained that if dharmas exist in a future phase,

then in a present phase, and finally in a past

phase, they have a history, in fact an endless

history; and this makes dharmas eternal substances,

quite contrary to the fundamental Buddhist doctrine

of anitya, transience.(4) The Sarvaastivaadins

always said in reply that they fully accepted the

conditionality, the sa.msk.rta nature, of ordinary

dharmas. In fact it seems to me that their clear

appreciation of the implications of conditionality

was a major motivation for their theory about the

reality of past and future. For present conditioned

dharmas are, by their very conditionedness,

internally related to the past conditioned dharmas

which together were their cause, and to the future

conditioned dharmas for which they are a

contributory causal factor.

III

It follows that if we are to take further our

understanding of the Sarvaasti position, we need to

appreciate the significance for the Buddhist

P.493

aabhidharmikas of the concept of conditionality.

'Sa.msk.rta', the standard term for this concept,

recurs throughout the Vibhaa.saa argument and its

context. The importance of the doctrine of

conditionality is emphasized at the beginning of the

relevant section of the Vibhaa.saa text (Poussin

[1937] p. 9):

L'auteur [here the Vibhaa.saa commentator refers

to the author of the original abhidharma text,

the J~naana-prasthaana] veut montrer clue la

'nature propre' du passe et du futur est reelle

(dravyasat) et que le present est conditionne

(sa.msk.rta).

The doctrine is also addressed specifically after

the abstract argument I have just described (Poussin

[1937], p. 11):

Le present n'est pas un dharma inconditionne;

car il nait des causes et conditions, car ii

possede activite (kaaritra) : tel n'est pas

I'inconditionne.

This characterization of (worldly) dharmas as

sa.msk.rta is fundamental to all abhidharma

theorizing. It expresses both sides of the Buddha's

basic metaphysical insight: that reality is

transient (anitya) and insubstantial (anaatman), but

also that this insubstantial reality is not entirely

fragmented and chaotic, but is bound together not by

abiding substance but by causal order

(pratiitya-samutpaada). This order makes possible

the explanation of suffering and the path to

liberation, the cause of suffering and the cause of

the extinction of suffering--the second and fourth

of the Noble Truths.

How can one concept express at the same time

these negative (anitya and anaatman) and positive

(pratiitya-samutpaada) doctrines? It is because the

causal relationships which provide the order of

pratiityasamutpaada are themselves an aspect of

anaatman, of insubstantiality. This is why, in the

early suttas, paticca-samuppaada is said to be a

middle way between existence and nonexistence.(5)

One way of expressing this is to contrast two

theories of causality. One is that entities are

basically self-subsistent; if left to themselves

they remain as they are, unchanged and unmoved.

Causal influences, if there are any, irrupt on them

contingently, from the outside. Causal relationships

are external, do not concern an entity's innermost

nature. This is perhaps the metaphysics of everyday

common sense, which inhabits a world of (relatively)

stable, middle-sized objects such as stones and

tables, and looks for causal explanations only when

these objects suffer some change.

The other theory, more realistic and more

scientific as well as more abhidharmic, is that an

entity's causal relations with other entities

express its innermost nature. What it is is

determined by a multiplicity of causes and

conditions; and its nature is to be understood as

functional, interacting with other beings and

leading always to future effects.

P.494

A similar point is made by Nyanaponika Thera

about the pali Abhidhamma, in his Abhidhamma

Studies.(6) There he emphasizes how what he calls

the "two methods" of abhidhamma, the analytic and

the synthetic, complement each other. The two

principal books of the Abhidhamma, the

Dhammasanga.nii and the Pa.t.thaana, respectively

represent these two methods, the analysis of the

experienced world into its basic elements, but also

the specification of the conditionality linking

these elements. Nyanaponika examines the basic forms

of the propositions in the two works, and concludes

(ibid., p. 21):

The mere juxtaposition of these two basic

schemata of the Abhidhamma already allows us to

formulate an important axiom of Buddhist

philosophy:

A complete description of a thing requires,

besides its analysis, also a statement of its

relations to certain other things.

This causal reinterpretation of the notion of

substance is particularly important when the basic

elements of existence are momentary. If things are

long-lasting, it may be thought that although causal

operations are necessary to bring them into

existence, thereafter they survive by a kind of

existential inertia. But if the world has constantly

to be remade, it is the more obvious that every

aspect of its existence rests on the complexities of

conditioned coproduction. There are then no

self-subsistent entities; everything exists in and

as a network of interrelationships. These are as

much in operation when things are apparently stable

as when change is occurring. On this, the

abhidharmic view, causal relations are internal; the

very being of a dharma is interpenetrated by those

other dharmas which are causally responsible for its

arising; it partakes in the being of those dharmas

to whose future arising it contributes.

The significance of the concept of

conditionality as interdependence is brought out in

several places in the Abhidharma-ko`sa. Vasubandhu

introduces the notion of sa.msk.rta dharma thus:

sa.msk.rta, conditionne, s'explique

etymologiquement: "qui a ete fait (k.rta) par

les causes en union et combinaison (sametya,

sa.mbhuuya)." II n'y a aucun dharma qui soit

engendre par une cause unique.(7)

The discussion of pratiitya-samutpaada in the

Abhidharma-kosa includes, among various accounts of

the meaning of the doctrine, the "saa.mbandhika"

interpretation that it concerns the (general)

relationship between causes and effects

(Abhidharma-ko`sa III.24d-25b). Vasubandhu quotes

from the abhidharma texts the view that

pratitya-samutpaada is to be identified with all

sa.msk.rta dharmas, not just those relating to

living beings;s that is, it is not just a

twelve-limbed account of the origins of human birth

and rebirth, but constitutes a general theory of

causality, And pratiitya-samutpaada and

conditionality are explicitly linked to anaat-

P.495

man, to insubstantiality at least as regards the

self, by a quotation from the Suutra:

Quiconque connait par la Praj~naa le

Pratiityasamutpaada et les dharmas produit en

dependance, il ne se tourne pas vers le passe en

se demandant s'il a existe....(9)

Vasubandhu's explanation of the various kinds of

causal relationships-hetu, pratyaaya, phala--gives

the different dimensions of the intertwining of

conditioned dharmas. The most important

relationships are of course between karmic action

and its fruition, and between the various dharmic

participants in the causal complex which leads from

perception to motivated action. But though the

causal relations of any particular dharma link it

'thickly' to a limited number of other dharmas, its

'thin' causal relationships extend much further. In

fact the abhidharmic position is that the causal

relations of every dharma extend throughout reality;

kar.ana-hetu (Abhidharma-ko`sa II.50a) or

adhipati-pratyaaya (Abhidharma-ko`sa II.62d) is a

type of causality which relates every conditioned

dharma with every other. The point here is that the

very existence of a dharma shows that there is a

causal compatibility between it and every other

coexisting dharma. The web of causal

interrelationships is seamless.

IV

It is against the background of this fundamental

Buddhist insight about conditionality that the

author of the Vibhaa.saa makes the simple point that

causal relations unite dharmas from the three times;

they link past, present, and future. How, then, can

a complete account of reality contain only the

present! The reality of past and future is necessary

to make the present what it is. The opposing view,

that the present dharma has emerged from and passes

immediately into nonexistence, dramatically

expresses the transience of dharmas; but at the

price, the Sarvaastivaadins would say, of obscuring

the intimacy of the relations between what is now

and what has been and will be.

We may now represent the form of the Vibhaa.saa

arguments as follows: (1) The present (that is,

present sa.msk.rta dharmas) exists only in and

through its relationships with past and future

(sa.msk.rta dharmas). (2) The existence of these

internal relationships requires the reality of their

relata--that is, not only the present, but also

past and future dharmas. (3) Since the existence of

the present as vipaakaphala etc. is

non-controversial, the relevant past and future

dharmas must exist, must be real.

Is this a sound argument? It seems to me, as I

have tried to indicate in the previous section, that

there is much to be said for the Buddhist theory of

conditionality, and so for the present argument's

first premise.

P.496

The second step in the argument gains what force

it has from the extension of a principle which in

general seems to be noncontroversial: that the

existence of a relation demands the existence of its

terms.(10) Of course we are concerned here with

internal and therefore 'real' rather than merely

'nominal' relationships; that is the relationship

concerns how the relata are in themselves, it exists

'in re' and not just 'in mente'. (I can in my

thoughts contrast what is with what is not, but this

'relationship' does not imply the existence of what

is not.) How, for example, can there be a 'being the

husband of', unless there are in existence a man and

a woman to be so related? In this case, of course,

the satisfaction of the relation demands the

coexistence of man and woman in the strong sense of

contemporaneity. But what should be said about

relations which allow, or even require, the

existence of their terms at different points in

time; such as being the great-great-grandfather of,

or the karmic cause of?

It would seem inappropriate to abandon

altogether the force of the principle, to say that

the operation of a relation between present and past

things or events requires only the existence of the

present thing. There must surely be some sense in

which we must demand the existence of past (and

future) relata, for such relations to hold.

It will not do to classify these relata as

nonexistent. There was never a time at which my

great-grandfather and I were both alive, but it does

not do justice to the relationship between us to say

that it links myself, who exists, with him, who does

not exist. The fact that he is (was) my

great-grandfather implies that his hold on reality

is more positive than that. We may say that the fact

that there 'is' such a relationship between him and

me implies our 'co-reality'; even though he died

before I was born. How should this co-reality be

understood? The Sarvastivadin position as expressed

in the Vibhaa.saa arguments provides an answer that

is in some ways straightforward: co-reality consists

of a 'reduced' form of contemporaneity.

But surely, one may say, an alternative and less

metaphysically startling account of co-reality is

ready at hand. Simply, A and B are co-real if there

is some time at which A exists or occurs and there

is some time at which B exists or occurs. But this

interpretation also contains its metaphysical

surprises. The verbs in the expressions 'are

co-real' and 'there is some time' (in the previous

sentence) are obviously not referring to the present

time; contemporaneity is not being claimed. Rather

the claims about the existence of these two times

must themselves be timeless. The theory of time

implied here gives primacy to what modern

philosophers have referred to as McTaggart's

B-series, in which time-orderings are unchanging and

are to be expressed using the tenseless felations of

'earlier than','later than','simultaneous with'. The

alternative view is that the primary time

relationships involve the tense-notions of past,

present,

P.497

and future (the A-series), and what falls into these

three classes is always changing.(11)

Interpreting co-reality in terms of the B-series

provides us with a clear (timeless) sense of

'everything exists', and also a clear alternative to

the 'abhuutvaa bhavati' position.

It is certainly beyond the scope of this paper

to attempt to decide between the two theories, that

giving logical primacy to the A-series and that

giving primacy to the B-series. But should we say

that if the latter theory prevails, then an account

of co-reality is available which makes the

complications of the Sarvaasti theory unnecessary?

An alternative would be to claim that the

B-series account of coreality is in fact that

adopted by the Sarvaastivaadins. Something like this

position is taken by Paul Williams (Williams

[1981]). If I understand Williams correctly, his

theory is as follows: The Sarvaastivaadin notion of

svabhaava, as the defining characteristic of a

dharma, which it possesses whether or not it is

presently actualized, should be understood as a kind

of atemporal essence, rather like a Platonic form,

which legitimizes talk about past and future

individuals, and also timeless discourse about the

general nature and taxonomy of dharmas.

The fact that x is a primary existent which can

always be referred to and thought of was felt to

require some sort of existence, but not the sort

of existence that a present entity enjoys. Such

is a perfectly reasonable and defensible

position, albeit perhaps mistaken. In fact for

Sarvaastivaada existence sasvabhaava is of a

different type, on a higher level, a

metalinguistic or metasystematic category

necessary for the atemporal systematisation of

primary existents in the dharmic list.(12)

It is unclear how far Williams wishes to press

this interesting interpretation as a motivation for

the Sarvaastivaada traikaalyavaada. He says that in

their theory of svabhaavataa, "the Sarvaastivaadins

were half-consciously[my italics] indicating a

difference in the given between talking tenselessly

about x and talking in the present tense" (ibid., p.

241); and later: "The critics of the Sarvaastivaada

failed to realise that the real existence of past

and future dharmas was in part [sic] a derivative of

an atemporal use of the verb 'to be'" (ibid., p.

245). It would be difficult to prove that this

consideration played no part in the svabhaavataa

theory, but it is surely not the whole of the

matter. It seems to me that most if not all of the

arguments by which the Sarvaastivaadins explicitly

defended their position, and in particular those

which have been considered in this essay, relate not

to timeless talk, ranging indifferently over past,

present, and future, but rather to the intimate

relations which the 'sa.msk.rta' doctrine postulates

between present dharmas and specific past and future

dharmas, that is, between dharmas which are all very

much in time, a part of history, of sa.msaara.

P.498

In fact, there seems little doubt that the

classical Sarvaasti theory fits best into a tensed

theory, in which change is real, so that dharmas are

definitely thought of as becoming present, and there

is a real difference between dharmas in the three

bhaavas or phases. In that case, the form of the

Vibhaa.sa arguments may be reexpressed as follows:

1) the sa.msk.rta doctrine shows that present

dharmas are intimately related to past (and future)

dharmas; 2) these relations, without which present

dharmas could not be what they are generally

acknowledged to be, imply the co-reality of past,

present, and future dharmas; 3) the Sarvaastivaadin

theory of 'reduced contemporaneity' offers an

account of co-reality which does justice to the

intuition that what is future becomes present, and

what is present becomes past.

NOTES

1 - This powerful summary of the anti-Sarvaastivaada

position comes in Abhidharma-ko`sa V.27 (Louis

de la Vallee Poussin, trans., L'Abhidharmako`sa

de Vasubandhu, 6 vols., new ed. [1st ed.

1923-1931; Bruxelles: Institut Belge des Hautes

Etudes Chinoises, 1971], vol. 4, p. 59 n. 3),

where it is said to be quoted from the

Paramaartha`suunyatasuutra: the text is

discussed by Sa.mghabhadra (Louis de la Vallee

Poussin, "Documents d'abhidharma-la controverse

du temps," Melanges Chinois et Bouddhiques 54

[1937]: 56-57).

2 - Translated in Louis de la vallee Poussin, "La

controverse du temps et du pudgala dans re

Vij~naanakaaya," Etudes Asiatiques publiees a

I'occasion du 25me anniversaire de I'Ecole

Francaise d'Extreme Orient, 1 (Paris:

Publications de I'Ecole Francaise d'Extreme

Orient, 1925), pp. 343-376.

3 - Several modern commentators, including Bareau

and de la Vallee Poussin, have claimed that the

Sarvaastivaada position rested on the naive

assumption that if x can be thought about, x

must have real existence. Bareau speaks of "that

coarse and even puerile realism on which their

doctrine down the centuries has been based"

(Andre Bareau, "The notion of time in early

Buddhism," Philosophy East and West 7 [1956]:

355-356). Similarly Poussin: "etrange confusion

de I'ideal et du real, curieuse survivance d'une

'psychologie primitive,' on affirme I'existance

'objective' de tous les objets (aalambana) de la

pensee" (Poussin, Melanges, p. 134) . This

assumption may play a part in some versions of

this type of argument; but it by no means

exhausts the argument's force.

P.499

4 - Kathaavatthu 1.6, $7 (Shwe Zan Aung and Mrs.

C.A.F. Rhys Davids, Points of Controversy, Being

a Translation of the Kathaavatthu [London:

Published for the Pall Text Society by Luzac,

1915; reprint, 1960], pp. 87-88) ; and

Abhidharma-ko`sa V.27 (Poussin,

L'Abhidharmako`sa, vol. 4, pp. 57-58).

5 - "Everything exists--this is one extreme. Nothing

exists--this is the other extreme. Not

approaching either extreme the Tathaagata

teaches you a doctrine by the middle [way]:

Conditioned by ignorance...." (Sa.myutta Nikaaya

II.17, translated in Mrs. C.A.F. Rhys Davids,

The Book of the Kindred Sayings, Part II: The

Nidaana Book [London: Pali Text Society, 1972],

p. 13).

6 - Nyanaponika Thera, Abhidhamma Studies (Kandy:

Buddhist Publication Society, 1965), chap. 2.

7 - Abhidharma-ko`sa 1.7, in Poussin,

L'Abhidharmako`sa, vol. 1, p. 11.

8 - Ibid., vol. 2, p. 65.

9 - Abhidharma-ko`sa III.25c-d, in Poussin,

L'Abhidharmako`sa, vol. 2, p. 68; compare, for

example, Nidaana Sa.myutta 12, Sa.myuttanikaaya

II.25-27, in Rhys Davids, Book of the Kindred

Sayings, p. 22.

10 - This principle is stated explicitly by

Sa.mghabhadra: "En effet, il n'y a pas de

relation possible de I'existant avec le

non-existant" (Poussin, Melanges, p. 115). This

is noted by Paul Williams in his article "On

the Abhidharma Ontology, " Journal of Indian

Philosophy 9 (1981) : 232 n. 26. Williams

himself does not discuss the truth of this

principle, for his concern is to show, with

respect to the Sarvaastivaadins' `vij~naana'

argument, that the grasping of something in

thought is not a relation between the thought

and the object grasped. In fact, Sa.mghabhadra

uses the principle only in passing, in the

course of a more specific argument about how

one can presently be bound by past passions.

But the principle does seem to lie behind the

Vibhaa.saa arguments.

11 - The distinction is very clearly stated in the

Introduction to Robin Le Poidevin, Change,

Cause and Contradiction (Basingstoke, England:

Macmillan, 1991).

12 - Williams, "On the Abhidharma Ontology," p. 245.