A note on the early Buddhist theory of truth

·期刊原文

A note on the early Buddhist theory of truth

Mark Siderits

Philosophy East and West 29, no.4 (october, 1979)

(c) by the University Press of Hawaii

p.491-499

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

p.491

Many scholars have held that the Buddha maintained a

pragmatic theory of truth. Jayatilleke argues

persuasively that this is not the case.(1) Rather, he

claims, the Buddha held something like the naive

version of the correspondence theory of truth. While

we agree with Jayatilleke's conclusion that, as a

matter of historical fact, the Buddha held a

correspondence theory of truth, this does not

completely settle the matter. for there remains a

significant philosophical question: Is the

correspondence theory consistent with the Buddha's

use of the notion of a category mistake? We shall

argue that early Buddhism is inconsistent on this

point, that a strict empiricist such as the Buddha

cannot employ the notion of a category mistake

without allowing for nonepistemic elements (such as

utility, coherence, and economy) in the conception of

truth. It is easily shown that the Buddha does bring

considerations of utility to bear on the question,

whether we should seek to ascertain whether a given

proposition is true or false; in this sense we may

say that early Buddhism includes a strain of

pragmatism. We may then make the following surmise

about the subsequent history of Buddhist philosophy:

that the problem of this inconsistency in early

Buddhism, along with such a pragmatic strain,

constitute important stimuli for the development of

the Maadhyamika dialectic.

Pragmatists have espoused a number of views

concerning the nature of truth. Peirce held a version

of the correspondence theory,(2) and thus need not be

considered here. Dewey saw truth as a kind of

warranted assertibility which grows out of

investigative operations. C. I. Lewis claimed that a

priori truths, such as the propositions of logic and

mathematics, are pragmatically grounded, that is,

true by virtue of their role in furthering human

conduct. (Lewis and Peirce also hold variants on the

ideal verification view of truth.) One element shared

by these views is the rejection of such extra

entities as facts or states of affairs which are

required to satisfy the correspondence theory of

truth. Pragmatists have traditionally preferred to

see truth not as dependent on some immutable

inhabitant of the world, but rather as deriving from

the practice of successful inquiry, especially that

of empirical science.

This attempt at ontological economy, which is one

of the chief attractions of the various pragmatic

theories of truth, is present as well in William

James' view. James held that that statement is true

which "works" or "pays," which has "cash value."

Truth is created not discovered; it is created

through the successful organization and manipulation

of what is given in experience. It might seem that

this agrees with the conceptions of other

pragmatists. The latter were careful, however, to

restrict the predicate 'true' to statements which

were first determined to be empirically verifiable.

Not so for James, who proposed that such statements

as 'God exists' could be shown to be true

p.492

if it could be demo^nstrated that the effects of

holding such a belief were predominantly beneficial.

This example brings out a second difference between

James and the other pragmatists as well: James'

notion of utility is far broader than the equivalent

notions of other pragmatists. Where Dewey, for

instance, thought of the utility of an idea as its

tendency to further the aims of empirical science,

James gives much weight to such subjective states as

personal satisfaction.

The difficulties involved in James' view are

notorious. Yet it is clear that it is just this

theory of truth, rather than the theories of other

pragmatists, which students of Buddhism have in mind

when they assert that the Buddha held a pragmatic

theory of truth. It might be thought that there is

evidence for this in the Nikaaya assertion that a

given theory is to be rejected as false if it is

found to lead to suffering (dukkha).(3) This might be

taken to mean that for the Buddha the truth of a

theory consists exclusively in its tendency, when

believed and acted upon, to produce beneficial

results; or it might be taken as support for the

weaker claim that utility is a criterion for the

truth of an assertion--that we test for truth by

looking to see which beliefs produce beneficial

results.

Jayatilleke presents evidence, in the form of

passages from the Nikaayas, which effectively refutes

both contentions. He cites the following passage as

conclusively establishing that the Buddha held a

version of the correspondence theory of truth:

When in fact there is a next world, the belief occurs

to me that there is no next world, that would be a

false belief. When in fact there is a next world, if

one thinks that there is no next world, that would be

a false conception. When in fact there is a next

world, one asserts the statement that there is no

next world, that would be a false statement.

It is difficult to imagine how this passage might be

construed in any other way than as an endorsement of

the correspondence theory of truth. Another passage

cited by Jayatilleke shows that utility was not

considered by the Buddha to be a criterion of truth.

Here the Buddha tells us that he "does not assert a

statement which he knows to be true, factual, and

useless" (bhuuta.m, taccha.m, anatthasa.mhita.m)

mine.gif border=0 align=middle>.(5) It would appear from this

statement that pragmatic considerations govern only

what the Buddha teaches, not what is actually true. A

true statement is one which agrees with reality, and

reality is here conceived as objective and

independent, unrelated to questions of human utility.

This stance accords well with the naive realism of

early Buddhism.

That the claim that early Buddhism embraces at

least a pragmatic criterion of truth is not entirely

devoid of merit may be seen, however, from an

examination of the so-called avyaakata, the

'indeterminate questions'. These are given as ten in

number: Whether the world is eternal or not eternal;

p.493

whether the world is finite or infinite; whether the

soul (jiiva) and body are identical or different; and

whether the enlightened one exists after death, or

does not exist after death, or both exists and does

not exist after death, or neither exists nor does not

exist after death.(6) These are regularly described

in the Nikaayas as 'theories' (di.t.thi) which the

Buddha has 'set aside' (.thapita) and 'rejected'

(patikkhita) .(7) The Buddha gives a number of

different explanations about why he does not

elucidate these questions. Several of these are

decidedly pragmatic in character, for example, "The

theory that the world is eternal, is a jungle, a

wilderness, a twisting, a writhing, and a fetter, and

is coupled with misery, ruin, despair, and agony, and

does not tend to aversion, absence of passion,

cessation, quiescence, knowledge, supreme wisdom, and

Nirvana."(8)

Similarly there is this: "Whether the dogma

obtain ... that the world is eternal or that the

world is not eternal, there still remains birth, old

age, sorrow, lamentation, misery, grief, and despair,

for the extinction of which in the present life I am

prescribing."(9) This is further illustrated by the

famous arrow parable. Suppose a man is wounded by a

poisoned arrow. His friends propose that he be

treated by a physician, but he refuses until it be

determined whether the man who shot him was a

K.satriya or a Brahman, and so on, what his name and

clan are, whether he is tall or short, and so on.

This is just the same as the case of the man who

refuses to enter the path to nirvaa.na until he has

determined whether the world is finite or infinite,

and more. The path to nirvas.na which the Buddha has

taught does not depend on any of these theories.

Such explanations of the avyaakata do show that

early Buddhism is, in some sense, pragmatic.

Metaphysical theories which have no bearing on the

quest for enlightenment and the cessation of

suffering are not to be pursued. This is not

tantamount to an endorsement of either the pragmatic

theory of truth or a pragmatic criterion of truth,

however; utility for the religious life is not here

being equated with truth, nor is such utility being

employed as a test for the truth of the avyaakata

questions.

With the fire parable,(10) however, we encounter

a different type of explanation as to why the Buddha

does not answer these questions. A sama.na or

wandering ascetic by the name of Vaccha is puzzled by

the fact that the Buddha has denied, in turn, each of

these four alternatives--that the enlightened one is

reborn after death, that he is not reborn, that he is

both reborn and not reborn, and that he is neither

reborn nor not reborn. Thus he asks these questions

again, but this time the Buddha replies, "To say that

he is reborn does not fit the case.... To say that he

is not reborn does not fit the case," and so forth.

At this point, Vaccha confesses that he is totally at

a loss about what to think. The Buddha then proposes

this simile: Suppose a fire which had been burning

before you were to go out. If someone were to ask in

p.494

which direction the tire had gone, north, south,

east, or west, what would you reply? "The question

would not fit the case." answers Vaccha. What we have

here is something akin to a category mistake.

Certain types of physical objects are properly

spoken of as 'going in direction---- ', for example,

birds. rocks, and clouds. Thus we may explain the

fact that a bird which was formerly present to the

senses, is so no longer, by saying that the bird 'has

gone north'. Under certain conditions such assertions

are also possible with respect to fires--for example,

in the case of a brush fire which was burning behind

my house thirty minutes ago. This predicate cannot be

applied in the case of extinguished fires, however,

to do so would be to commit a category mistake. If a

physical object is to be spoken of as 'going in

direction ' it is necessary that it be an enduring

physical object. It is of course arguable whether a

fire counts as a physical object at all; but it is

perfectly clear that a fire which has gone out due to

lack of fuel cannot be thought of as an enduring

physical object. That a fire was formerly present to

our senses, but is no longer, might lead us to think

that, as in the case of the bird, we may use the

'going in direction ---' predicate to account for the

facts. But an extinguished fire is simply not

subsumed under the appropriate category for the

application of this predicate.

The Buddha is suggesting that similar

considerations apply to the case of the arahat after

death. While the framework of ancient Indian thought

allows us to ask of any deceased individual whether

he or she is to be reborn, the question is

meaningless with respect to the arhat. Vaccha was

confused by the Buddha's rejection of each of the

four logically possible alternatives; it would seem

that by necessity one of these must be true. Once we

see, however, that such predicates as 'reborn' simply

do not apply to the arhat and that the deceased arhat

is subsumed under a different category, the seeming

oddity of the position vanishes.(11)

The same line of thought is to be found in the

following. The Buddha is teaching that formulation of

pa.ticcasamuppaada (dependent coorigination) known as

the bhavacakka or 'wheel of becoming': On ignorance

depends karman, on karman depends consciousness... on

rebirth depends old age and death. He is then asked,

What are old age and death, and what is it that has

old age and death? The Buddha replies:

The question is not rightly put.... O priest, to

say,'What are old age and death, and what is it that

has old age and death?' and to say, 'Old age and

death are one thing, but it is another thing which

has old age and death', is to say the same thing in

different ways. If, O priest, the theory obtain that

the soul and the body are identical, then there is no

religious life; or if, O priest, the theory obtain

that the soul is one thing and the body another, then

there is no religious life. Both these extremes, O

priest, have been avoided by the Tathaagata, and it

is a middle doctrine he teaches:'On birth depends old

age and death'.(12)

p.495

Two points require explanation here. First. by

the second sentence the Buddha merely means to make

explicit a presupposition involved in the original

question. To ask, `What is it that has old age and

death'!' is it assume that there is a soul or self

(jiiva--anima is a fairly close Western equivalent)

which has old age and death as characteristics.

Second, if this point is accepted, then we must

assume that the self is either identical with the

body or different from the body. These positions were

held, respectively, by the Materialist and the

Saa^mkhya schools of Indian philosophy. But on either

assumption, according to the Buddha, the quest for

nirvaa.na becomes impossible. If the self and the

body are identical, then the self dies with the body

and there is no karmic retribution in the next life;

thus it becomes impossible to follow the path to

nirvaa.na over a succession of lifetimes, timespan

that is often required according to Buddhist

doctrine. It is not clear why the Saa^mkhya position

should be thought to make nirvaa.na unattainable, but

the Buddha might very well want to argue thus: If

self and body are separate, then the self or soul

would seem to be intrinsically unknowable(13); in

this case the reason why suffering attaches to the

self cannot be known, and thus its cure cannot be

found.

Having explained these points, let us return to

the Buddha's original response, 'The question is not

rightly put' (No kallo pannho). What is meant by

this? Both the grammatical structures and the usage

patterns of paali and English make it natural to think

of such terms as 'old age' and 'death' as denoting

qualities which inhere in a substance or subject. We

may think of this tendency as deriving from the

commonsense categorical framework ordinarily assumed

by the speakers of these languages. Here the Buddha

is, in effect, challenging that categorical

framework. We should not think of a discrete enduring

entity, the individual person, which suffers old age

and death; instead we should think of old age and

death as states or conditions which arise dependent

on other states and conditions and which, together

with these antecedent states and conditions,

constitute without residue the individual on which

they are predicated. This is tantamount to the claim

that the notion of an individual (at least in the

sense of 'individual person'(14)) should be subsumed

not under the category of substance but under the

category of logical fiction. If this claim is

accepted, then the question, 'What is it that has old

age and death?' is indeed 'not rightly put'--for it

depends on a category mistake.

We have seen two cases in which the Buddha makes

use of the notion of a category mistake. On what

grounds may such claims be made? The second example

is misleading, in that the Buddha appears to be

basing his assertion on purely pragmatic

considerations: The religious life cannot be lived

unless we accept this alternative categorcial

framework. Much empirical evidence is elsewhere

adduced, however, in favor of the claim being made

here: nowhere

p.496

in our experience do we discover anything other- than

transitory states and conditions, and so on, thus we

have no grounds for asserting the existence of a self

or person which is the bearer of such personal

attributes as old age and death. It is true

nonetheless that such evidence cannot be used in

support of the proposed change in category

subsumption. This is true for the reason that there

is no empirical evidence in favor of the notion that

personal identity is maintained through causal

relations of the sort described in the doctrine of

pa.ticcasamuppaada. Thus there might well be

significance in the fact that the Buddha here adduces

only pragmatic grounds in justification of his claim.

In the case of the fire parable we seem to have

an instance of argument by analogy: like fire,

samsaric existence continues in dependence on causes

and conditions. If these causes and conditions are

removed, both will cease. In both cases, this is

known as 'extinction' (nirvaa.na). Now when a fire

has gone out through exhaustion of its fuel, we do

not ask where it has gone; to do so would be to

commit a category mistake. So when the arhat reaches

'final extinction' (parinirvaa.na) through extinction

of the conditions for samsaric existence, it would

likewise be a category mistake to ask whether he is

reborn, and so on.

The relative weakness of this argument is

instructive. It is extremely difficult to imagine

strictly empirical evidence for the change in

category subsumption being proposed here. It is easy

to understand why the Buddha would insist that the

arhat not be thought of as being reborn, and so on:

any talk of parinirvaa.na as a 'state' into which the

arhat passes upon death would give rise to the

notions of ego and enjoyment, notions which must be

abandoned if nirvaa.na is to be attained, If the

Buddha claims that 'rebirth' and so on do not apply

to the arhat after death, however, then it must be

pointed out that there is no possible experience

which could verify this claim.

This brings us to an important point about the

notion of a category mistake. A consistent empiricist

who presupposes the naive form of the correspondence

theory of truth cannot always effectively argue that

such a mistake has been committed, This is so because

the claim involved in such an assertion is that

experience should be construed in one way rather than

another. There exist many cases, however, in which

each of two or more ways of construing the data is

equally adequate to what is given in experience. In

such cases we regularly appeal to such things as

utility, coherence, and parsimony in order to decide

between competing theories. This route is not open to

the empiricist who maintains a strict form of the

correspondence theory of truth however. To assert

that a category mistake has been made is to say this:

term x, ordinarily subsumed under category c, is

rather to be subsumed under category d. And if truth

is correspondent to states of affairs, then there

must exist facts to which this statement either

corresponds or fails to correspond. Now a consistent

empiricist cannot allow that knowledge of facts or

states of affairs is to be had by any other means

than experience, Thus

p.497

in those cases in which experience is silent as to

which style of category subsumption is to

preferred, (15) our empiricist must likewise be

silent.

If we are to accommodate the existence of cases

such as these, we shall have to revise our notion of

truth along the following lines: Once a categorical

framework is in place, we may speak of truth as

correspondence to facts-since the categorical

framework tells us in effect what is to count as a

fact. When an alteration in the categorical framework

is proposed, however, there are no longer any 'facts'

to which to appeal; thus our decision must be based

on grounds other than correspondence, such as

utility, coherence, and economy. And this is, of

course, tantamount to accepting nonepistemic elements

in our conception of truth.

It is difficult to determine if there is any

awareness within early Buddhism of the difficulties

which arise in connection with the notion of a

category mistake. It is true that pragmatic

considerations are sometimes advanced as grounds for

accepting one style of category subsumption over

another. But this need not be interpreted as

acceptance of a pragmatic theory of truth, that truth

is what 'works'. We have already seen evidence of a

marked pragmatism of another sort in early

Buddhism--the tendency to avoid speculation

concerning issues not directly related to the project

of enlightenment. This latter attitude is perfectly

consistent with the native form of the correspondence

theory of truth. And that this is the limit of the

Buddha's socalled pragmatism would seem to be borne

out by the passage already cited, in which he

proclaims that he does not assert a statement which

he knows to be true, factual, and useless."

We may then surmise that the Buddha's use of

pragmatic criteria in his rejection of the question,

What is it that has old age and death?, represents no

more than an instance of this tendency to avoid

fruitless speculation. There is evidence for this

supposition in the fact that the Buddha refers to

this and similar questions as 'twistings' (visuuka)

and 'writhings' (vipphandita)(16)--terms which he

also applies to the avyaakata.(17) This suggests that

the Buddha has failed to notice one important respect

in which these questions and the avyaakata are

dissimilar, namely, that the former are dismissed as

pseudoproblems while the latter are not. (It is said

that upon the dissolution of ignorance, questions of

the former sort 'become nonexistent'; we should not

expect the same thing to be said of the

avyaakata--only interest in the latter should be

dissolved in enlightenment.) If this is indeed what

the Buddha has done, then this and similar instances

cannot count as evidence for pragmatic elements in

the early Buddhist theory of truth.

Further support for my contention is found in the

fact that nowhere in the early Buddhist literature,

nor in the contemporary non-Buddhist Indian

philosophical literature, does there appear to be any

discussion of alternative conceptions of truth. Now

the correspondence theory would seem to come closest

to the commonsense conception of truth. And it is a

good rule to

p.498

follow in historical reconstruction that, in the

absence of an articulated problematic, endorsements

of what would seem to be the commonsense attitude

toward a philosophical issue should be taken at face

value. (This principle could also be used in support

of the claim that, metaphysically, early Buddhism

represents a form of native realism.) Thus it would

seem we should conclude that the Buddha was not aware

of the problems involved in, on the one hand,

maintaining a strict empiricism and a naive form of

the correspondence theory of truth, and, on the other

hand, utilizing the notion of a category mistake.

I have not discussed this matter solely with an

eye toward establishing a precise philosophical

reconstruction, however. The issues involved here are

of great importance for the subsequent history of

Buddhist philosophy. We have concluded that early

Buddhism is empiricist, that it holds a form of the

correspondence theory of truth, and that it employs

the device of the category mistake. The Abhidharma

schools such as Theravaada and Sarvaastivaada follow

the lead of early Buddhism on these points. Indeed it

is possible to view the dharma theories of these

schools as alternative categorical frameworks which

are designed to replace that framework which is

presupposed by common sense. Here again, however, the

chief justification advanced in favor of these

alternative modes of categorizing experience is that

they correspond to the nature of reality.

It is just this aspect of orthodox Buddhist

thought which Naagaarjuna singles out for criticism

in his dialectical refutations. What the Maadhyamika

critique comes to is the assertion that such notions

as alternative categorical frameworks and revisions

in category subsumption require the incorporation of

nonepistemic elements such as utility into our

conception of truth. In other words. Naagaarjuna

exploits the tension between empiricism and the

correspondence theory on the one hand, and the use of

a notion of categories on the other hand. He sees the

theories of early Buddhism and Abhidharma as

empirically unverifiable and thus confirmable only on

pragmatic grounds; but this gives those theories

precisely the same status as that commonsense

conception of the nature of the world which they were

meant to replace. The final point I wish to make is

that the possibility of viewing early Buddhism in

this manner was probably suggested to Naagaarjuna by

the presence of pragmatic elements in early

Buddhism. Thus while the Buddha may not have solved

the problem which he set for himself by employing the

notion of a category mistake, he did provide the

elements which led to its ultimate solution.

p.499

NOTES

1. K. N. Jayatilleke. Early Buddhist theory of

Knowledge (London: George Allen and Unwin Ltd.,

1963), pp. 351-359.

2. Collected papers of Charles Saunders Peirce, eds.

Charles Hartshorne and Paul Weiss (Cambridge:

Harvard University Press, 1960), 2.652.

3. A^ngutara-nikaaya (hereafter cited as A), ed. R.

Morris (London: Pali Text Society. 1955). 2.191.

4. Majjhima-nikaaya (hereafter cited as M), ed. V.

Trenckner (London: Pali Text Society, 1948, 1.

402-403.

5. M 1.395.

6. M 1.426, M 1. 483, and so on.

7. M 1.426. 8. M 1.485. 9. M 1.429. 10. M 1.486-488.

11. It should be pointed out that the Buddha is not

saying that the state of the arahat after death

is indescribable or ineffable. This possibility

is represented by the fourth of the four

alternatives, which the Buddha rejected.

12. Sa.myutta-nikaaya (hereafter cited as S), ed. M.

Leon Feer (London: pall Text Society, 1960),

2.61.

13. This is the Saa^mkhya notion of puru.sa or self

as pure subjectivity. The conclusion that

puru.sa is intrinsically unknowable appears to

have resulted from a line of thought not unlike

the progressive refinement which the notion of

self undergoes in Descartes, Locke, and

Berkeley.

14. It is the chief purpose of early Buddhism to

make this point with respect to the individual

person. Only in Abhidharma is the point made

with respect to all commonsense 'things'.

15. Confer Quine's essay "Ontological Relativity" in

Ontological Relativity and Other Essays (New

York: Columbia University Press, 1969) , pp.

26-68, for examples of such cases. 16. S 2.61.

17. M 1. 485.