A Difference of Method: Buddhapalita, Bhavaviveka, and Chandrakirti
Around the turn of the sixth century, an important disagreement occurred about the method whereby the Madhyamika view was to be established in debate. This question had important ramifications concerning the manner in which the relative truth was to be explained and how, within that context, the ultimate truth
was to be presented. One side of the debate was represented by Buddhapalita, who had confined himself to the exclusive use of consequential arguments (prasanga). He was opposed by Bhavaviveka, who maintained that, in debate with non-Madhyamikas about the ultimate nature of phenomena and in order to
establish one's view beyond doubt, the adduction of mere consequences was insufficient. It was both possible and necessary to prove one's point positively by means of independent inferences (svatantra-anumana) adduced in syllogistic form. The fact that Buddhapalita returned no answer to this critique gave rise to the story that he had been intimidated by Bhavaviveka5s princely rank. However, there are reasons for believing that he was already dead by the time the latter launched his attack.
Bhavaviveka was a famous scholar with an encyclopedic knowledge of the different philosophical and religious schools of his time, both Buddhist and nonBuddhist. It is evident, too, that he was deeply interested in questions of formal logic, the study of which had been developing in India from the third century onward. And it has been suggested that the movement Bhavaviveka inaugurated was an attempt to create a bridge between the Madhyamika and the philosophical movement that reached its climax in the logical reforms of Dignaga and Dharmakirti.-
The division between the Prasangikas and the Svatantrikas is a large question and in certain respects highly technical. Fortunately, there exist a number of learned studies in English on this subject, and the interested reader is invited to refer to them.- For the present purposes, we will attempt a summary of the main issues.
To begin with, it is useful to bear in mind that the final aim of Madhyamika, as of all other Buddhist teachings, is soteriological. Its sole purpose is to lead beings to ultimate freedom. As Nagarjuna said, it is only through the understanding and realization of the ultimate truth that freedom from suffering
can be gained. Impelled by their vows of bodhichitta, Nagarjuna and the great Madhyamika masters who followed him were concerned, therefore, not only with the realization of the truth for themselves, but also with the communication of this truth to others. The disagreement between the Prasangikas and the Svatantrikas turns on precisely this question: how is the view to be established and what is the best and most effective way of indicating it to others?
The disagreement between Bhavaviveka and Buddhapalita arose out of the interpretation of the very first stanza of the karikas: "No things are produced
anywhere at any time, either from themselves, from something else, from both, or from neither.55 Cast in a form that harks back, it will be remembered, to the fourteen unanswered questions, the primary dilemma between the first two alternatives (production from self and production from other) is expanded into
a tetralemma (catuskoti) by the addition of two more alternatives intended to exhaust the entire range of possibilities. These four alternatives, which provide the framework for ChandrakirtPs later discussion of phenomenal emptiness in the Madhyamakavatara, were usually associated, emblematically, with
four schools of Indian philosophy: the Samkhya, the Buddhist Abhidharmika, the Jaina, and the Charvaka respectively. We are to imagine a discussion between a Madhyamika and the representatives of four types of philosophical realism, who believe that there is at least something, the intrinsic existence of which
must be accepted. The purpose of the Madhyamika critique is to demonstrate their mistake and to produce in their minds an understanding of the emptiness of all phenomena. How is one to go about this? The Prasangikas and Svatantrikas disagree as to the best approach.
Buddhapalita refuted the Samkhya theory simply by pointing out that it entailed an absurd consequence. His argument, which is reiterated exactly in the Madhyamakavatara, runs roughly as follows. The Samkhyas believed that everything arises through modulations occurring in the primal substance, prakriti.
All effects are therefore, in the most fundamental sense, identical with their causes. Buddhapalita argued that this assertion is untenable. To say that things arise "from themselves55 is absurd, because if they already exist (being identical with their causes), no further coming into existence is needed.
If, on the other hand, "coming into existence55 is part and parcel of the supposedly produced thing, its production must be as unending as the produced thing itself. The Samkhya account is therefore incompatible with causality, either in theory or in fact. It is not necessary, at this stage, to enter into
the details of Buddhapalita5s argument. The point is that he disposes of the Samkhya claim simply by showing that it involves contradictions and is therefore unviable. In refuting the Samkhya view, he does not substitute a theory of his own.
Bhavaviveka objected that Buddhapalita5s procedure was logically deficient. The latter, he said, should have supported his contention by supplying a reason
and an example. He complained, moreover, that the Samkhyas were being dealt with too summarily. They had their own arguments against the Buddhist critique, which Buddhapalita was failing to address. Finally, the simple negation of the Samkhya view by showing its untenable consequences was too open-ended. It could be taken to imply that Buddhapalita entertained an alternative position, which, since
he was a Madhyamika, was not the case. Simply to adduce a consequence, therefore, leaves room for doubt in the opponent's mind. In order to remedy these defects, Bhavaviveka argued that when refuting the Samkhya, it was both possible and necessary to prove Nagarjuna5s proposition (as given in the first stanza of the karikas) in terms of a syllogism一an independent syllogism, indeed, that expressed the contention in a self-contained manner, without
reference to the opinion of the interlocutor. Bhavaviveka evidently thought that this kind of approach would be more effective in convincing the opponent and helping him accept the Madhyamika view. The kind of formulation he proposed took the standard form of a syllogism as laid down in Indian logic:
subject, predicate, reason, and so on.-When used in debate, the syllogism is, or should be, founded on commonly accepted elements established by valid cognition, thereby deriving its cogency from basic premises shared by both parties. Such an argument is intended to convey real knowledge and induce conviction. To this standard format, Bhavaviveka added a touch of his own, namely, the rider "on the ultimate level55 or "ultimately."£ The reasons for this addition will become clear in due course.
Buddhapalita no doubt represented a conservative element in the Madhyamika tradition.- In confining himself to consequences, and in being evidently reluctant to involve himself in the sophistications of logic and epistemology as these were developing at his time, he emulated Nagarjuna, who had employed
consequential reasoning very often (though not exclusively) in the karikas, and who, in the Vigrahavyavartani, had been careful to confine the use of logic to the level of conventional truth, implying the illegitimacy of using it to establish anything transcending that sphere. Unlike Bhavaviveka, who was a popular and influential teacher, Buddhapalita appears to have had few disciples. It was left to Chandrakirti in the following century to defend him and to stem the Svatantrika tide.
As we have seen, the purpose of prasanga is to refute a position, not by stating a more plausible counterposition but by exposing a consequence unwanted by the proponents一on the basis of arguments that the proponents themselves accept. In adopting this strategy, the Prasangika debater is not committed either to the immediate conclusion of the argument or to the principles invoked in the course of the investigation. It is only necessary for the proponents to
accept them, the only object being to enable them to see for themselves the falsity of their position and to abandon it. The position of the adversary is not destroyed, as it were, from outside, by arguments adduced independently by the Madhyamika. It is shown instead to be intrinsically absurd, so that it collapses, so to speak, under its own weight. By using this technique in discussions concerning the ultimate status of phenomena, the Prasangikas are able to undermine the false notions of their opponents and to indicate the truth indirectly, without having to verbalize a position of their own.
Why is this last point so important? In order to answer this question, we must digress slightly. We have already seen that the Buddha himself had declared the ultimate truth to be beyond the scope of the ordinary mind. But though the ultimate is not to be expressed in thought and word, it can be indirectly indicated by demarcating the limits of conceptual construction and suggesting that there is, nevertheless, "something" beyond. In this procedure, logical
arguments are used to demonstrate that when reason attempts to give an accurate account, in absolute terms, of "the way things are,55 it leads to antinomies and contradiction. This is the method of Nagarjuna and of Chandrakirti. Even if reason is unable to encompass reality, it can at least convince itself that it is unequal to the task and that the ultimate is to be approached and realized by means other than philosophical cogitation. Reason
understands, inferentially, that the ultimate truth exceeds its powers of comprehension and expression. The Madhyamika approach is, in other words, a via remotionis, to borrow a term from Christian theology: the dialectic approaches its goal by showing all that the ultimate is not; its purpose is to demolish
the theories produced by the ordinary mind and to reveal the hollowness of their pretensions. The use of reasoning to demonstrate its own inadequacy is not, to be sure, an attractive prospect for the rationalist. In one sense, it is a bewildering discovery, and it did indeed prove the sticking point for
Kant. Having understood the limitations of pure reason, he found of course that this purely intellectual achievement was unable to remove what he called the transcendental illusion: the impression, and therefore the constant temptation to think, that thought is able to lay hold of perfectly perspectiveless
objectivity. He doubted that it could ever be removed, that the mind could ever pass beyond it.- He could never countenance the possibility of jnana, the nondual wisdom in which the ultimate is known directly without the mediation of thought.- He failed, in other words, to appreciate the immense spiritual
significance of his discovery and, as Murti aptly observes, ended by putting it to a trivial purpose.- This was a mistake that Nagarjuna and the Madhyamikas did not make. And they did not make it because they had at their disposal not just the intellectual tools of their own brilliant minds but also their spiritual training on the Buddhist path and the realization of the masters who had transmitted it to them.
A perception of the limitations of thought may seem, as we have said, a sorry conclusion to the philosophical enterprise一until one notices that the implications for the mind that reaches this conclusion are immense. The very fact that the discovery is possible points to something beyond the ordinary
intellect. To realize, by thought, that there is an ultimate truth that is not the object of thought is no ordinary finding. It is not just the negative conclusion of dialectical analysis, but also the discovery of a wholly new dimension in the mind itself. When the mind realizes emptiness, it overcomes the
subject-object dichotomy. It does not just break through the appearances that conceal the ultimate status of phenomena; it also penetrates the veils of mental construction that had concealed its own true nature and had made the misperception of phenomena possible. When the true nature of phenomena is
discovered, the mind's nature also stands revealed, for the realization of emptiness is the experience of nondual wisdom. Looked at from this point of view, the final outcome of Madhyamika analysis is not a negative but a profoundly positive experience.
Chandrakirti5s defense of Buddhapalita and his refutation of Bhavaviveka are to be found in his detailed commentary on the karikas entitled the Prasannapada.-Here he considers each of the objections brought against Buddhapalita by Bhavaviveka and refutes them all on technical grounds.- To begin with, he rejects as unfounded the charge that Buddhapalita5s use of consequences is inadequate because it fails to supply a reason and example. In fact a
consequential argument can be restated in the form of what is technically known as an inference accepted by the opponent,- whereby the import of the consequence can be expressed in a syllogism in which both reason and example are present by implication. In being an extension of the consequence, the inference accepted by the opponent is based on elements that need only be agreeable to the opponent, not to the proponent, in the debate. It is thus not to
be confused with an independent inference in the terms of which the acceptance of the proponent is implied. Like the consequence, the inference accepted by the opponent does not compromise the proponent in the way that an independent inference does. This fact, Chandrakirti argues, also acquits Buddhapalita of the second charge brought by Bhavaviveka, namely, that he fails to address the objections
advanced by the Samkhyas. In fact, these objections are of necessity only advanced against a position positively expressed, that is, in an independent argument. This does not occur in the case of the inference accepted by the opponent.-
Finally, that Buddhapalita5s consequential argument implies a contrary position that inadvertently undermines his Madhyamika stance is categorically denied. The meaning and purpose of the consequence are clear from the context, and Buddhapalita5s words are to be understood according to his evident
intention.- A consequential argument, Chandrakirti insists, is perfectly adequate to the task of refuting the false position. If the adversary refuses to accept defeat even after it has been shown, on principles already acceptable to him, that his view is untenable, it is clear that the further adduction of
an independent argument would serve no purpose. If the opponent still maintains his position even after its incoherence is laid bare, it is clear that he does so for motives that cannot be rational. Either he is too dull to understand the refutation or he clings to his position out of prejudice. This being so, it is futile to discuss further.
These are undoubtedly complex questions. The point to retain, however, is that Chandrakirti5s objective is to defend a method of communication whereby proponents of Madhyamika can debate on matters concerning the ultimate truth without having to verbalize positions of their own, thereby betraying the Madhyamika's most important principle, namely, that the ultimate status of things is ineffable.
Having vindicated Buddhapalita, Chandrakirti turns to his assailant. "Bhavaviveka wishes only to parade his knowledge of the logical treatises. He adduces independent syllogisms, despite the fact that he claims to hold the Madhyamika view. The Madhyamika system, to be sure, creates lots of difficulties for
such a would-be logician. He makes one mistake after another.55- According to the rules of logic, when an independent syllogism is framed, its validity depends on the fact that its terms denote exactly the same thing for both parties in the debate. This, Chandrakirti argued, is impossible in any discussion
between a Madhyamika and a realist philosopher (such as the Samkhya) when the subject of discussion is the ultimate status of phenomena. A viable independent syllogism presupposes the existence of objects that both sides accept. But the whole purpose
of the Madhyamika is to show that no such objects exist. In situations of this kind, therefore, the Madhyamika debater cannot use independent syllogisms without being fatally compromised. The Madhyamika teaching on the two truths, which is not accepted by the opponent, necessarily excludes any community of
understanding with the realist concerning the existential status of phenomena. This being so, the Madhyamika is unable to advance an independent syllogism, without the syllogism itself, according to the rules of logic, being defective. If the two parties use the same terms but in effect mean different things by them, it is obvious that they are talking at cross-purposes; common understanding is ruled out. In brief, therefore, Chandrakirti castigates Bhavaviveka not only for compromising his Madhyamika principles but also for being an incompetent logician.
In defense of Bhavaviveka, it may be said that he was not unaware of the difficulties involved in his position, and it seems clear that the characteristic orientation of the Svatantrikas with regard to logic and the conventional truth is adopted with a view to consistency. The fundamental teaching of
Madhyamika is the rejection of the ultimate existence of all entities. Like the Prasangikas, the Svatantrikas are concerned to communicate this view to non-Madhyamikas. Unlike the Prasangikas, they seek to do this not by consequences alone but by the use of logic and the making of positive statements一
adducing, as we have seen, independent syllogisms that are based on elements commonly acceptable, on the conventional level, to both parties. In so doing, the Svatantrikas take a conciliatory step toward the opponent. In other words, they introduce the Madhyamika view in terms easier for ordinary people to understand. The motivation, as we have seen, is a good one, but the step cannot be made without compromise, and this consists in the creation of a provisional separation of the two truths.
Bhavaviveka and those who followed him say that whereas, on the ultimate level, phenomena have no reality whatever, on the conventional level, they do possess a certain existence (though not a true existence), and this is proved by the operation of conventional reasoning. Common sense can, for instance,
distinguish a "real" object from an optical illusion. It does so on the basis of functionality (real water is drinkable, mirage water is not, and so on). Therefore, when phenomena are said to be without inherent existence, the Svatantrikas add the proviso "on the ultimate level.55 Conventionally, for the
Svatantrikas (at least those who follow Bhavaviveka), phenomena do have a kind of "natural existence according to their characteristics.^^- When investigated by conventional reasoning, they are "found"; one can discuss them and entertain theories about them. For Bhavaviveka, therefore, meaningful
discourse is still possible on the level of the conventional truth. Although ultimately empty, conventional phenomena can nevertheless be talked about without absurdity. It is still possible to philosophize, and this can be utilized to good purpose, in giving disciples a correct orientation and leading
them gradually on the path. Thus the theory of partless particles, as presented in the Abhidharmika schools, is provisionally accepted. What the Abhidharmikas had taken to be ultimate truth remains valid, but only conventionally valid, for the Svatantrikas.
Chandrakirti and the Prasangikas will have none of this. For them no compromise is possible. The ultimate, being ineffable, is falsified by any attempt to express it. To separate the two truths is to deviate from Nagarjuna5s meaning. Therefore, when establishing the view and in debate, the Prasangikas express
no position, no thesis. In debate, they confine themselves to consequential arguments, the reduction to absurdity of the opponents position; the ultimate truth is indicated only indirectly by the demolition of theories. For the Prasangikas, therefore, it is neither desirable nor possible to elaborate a
theory of the conventional truth. Unlike Bhavaviveka, who discusses along Sautrantika lines, and unlike Shantarakshita, who presents the conventional truth in terms of the Yogachara view, Chandrakirti refers to the conventional as being simply the unexamined phenomena of ordinary experience, accepted as true
by the common consensus. The Prasangikas do not care to theorize about the conventional. They do not philosophize. This does not, of course, mean that they acquiesce in the ignorant opinions of worldly people, who believe firmly in the reality of the phenomenal and personal selves. It does mean, however, that,
as a method of approach to the ultimate truth and as a medium with which to communicate with worldly people, the Prasangikas simply accept, without analysis, the things and events occurring in everyday experience.
As forms of philosophy, the four theories of production given in the tetralemma all claim to give an accurate account of conventional experience. All can be shown to be logically incoherent and are, the Prasangikas say, a source of confusion. Far from giving a sensible explanation of the world, their
solutions are obscure and far-fetched. In ChandrakirtPs opinion, they are quite irrelevant (as philosophy often is) to the perceptions and concerns of ordinary folk. No ordinary person consciously advocates either the theory of the Samkhyas or that of the Buddhist Abhidharmika一production explained in
terms either of identity or difference of material causes and effects. A man who deposits a drop of semen in the womb of his wife will point to the baby nine months later and say, "I produced this child.55 The difference between baby and semen is routinely overlooked. In the same way a gardener points to
the flowers that "he planted,55 whereas in actual fact he planted only seeds. In practice, therefore, people do not acknowledge a separation between material cause and material effect. On the other hand, if you ask someone whether the food they eat and the feces they excrete are the same, they will
certainly say that there is a difference. They are very far from accepting the Samkhya theory. On the level of what actually happens, it is impossible to say that cause and effect are either the same or different. The only thing one can and must allow is that, in experience, production does occur. Everyone is agreed about this and, as an account of the conventional, this is, for the Prasangikas, quite sufficient.
Indeed, in situations where one is trying to penetrate to the ultimate status of phenomena, the introduction of theories as a means of explaining the working of the phenomenal world fogs the issue and actually undermines the correct approach to the conventional truth. Far from elucidating the
conventional, Chandrakirti says, theories actually undermine it. It is the conventional itself— what actually happens一that is the means of entering the ultimate. To create a theory as a way of explaining the mechanics of the conventional does not help to introduce the ultimate; it merely complicates the
matter. Therefore theories are dangerous, for they obscure the conventional; they hinder the procedure whereby one can "see through55 the conventional appearance of phenomena and perceive their lack of intrinsic "thingness." Chandrakirti says that to create a theory about the conventional is in a sense to
"destroy" the conventional; it produces an account that, however coherent it may be, is always at variance with what we actually experience. As such, it is at best irrelevant to the task in hand, namely, to perceive the true nature of phenomenal appearance. At worst it is a hindrance and a trap. The image
often evoked is that of a man climbing a tree. Before he has caught hold of the branch above, it is inadvisable for him to move off the one below. In weaving their theories, this is precisely what philosophers do. To create a theory about the conventional is in a sense to move away from the conventional as experienced (which alone is the gateway to the ultimate). The progression from the conventional to the ultimate is rendered more difficult by the invention of ill-conceived hypotheses.
The following parable, borrowed from Bertrand Russell, may further illustrate this important point.- If I go up to a nuclear physicist and ask him to describe for me the physical constitution of a table, I will receive a long and learned answer, all about magnetic fields and atomic and subatomic
particles moving around at great speed. These, he assures me, are the real constituents of the table; the object in the corner is little more than an optical illusion. On the other hand, if I approach the same scientist unannounced and simply ask whether there is a table in the room, he will, without a
momenfs hesitation, point and say: "It's over there, can't you see it?" However accurate the scientists earlier description may be, it has clearly not interfered with his perceptions. But now let us extend the parable further and imagine the same physicist trying to use his bank card to get money from a
cash machine outside a bank, and let us suppose that there is something wrong with the card, with the result that the machine swallows it and produces no money. Before long he will become annoyed and start beating on the machine with the same degree of frustration as any ordinary nonscientist. And I would be ill advised, at this point, to try to comfort him by reminding him that, after all, the bank card he has lost and the bank notes he has failed to receive
are no more than a mass of subatomic particles. Sophisticated as the physicists theory may be, it has done nothing to free him from the suffering and perturbation always liable to manifest in the course of conventional transactions. In the same way, the propounding of theories about the conventional does
nothing to remove the tyranny of phenomenal appearance. And the use of independent syllogisms, and the acceptance of conventionally existent entities, which this entails, necessarily implies a theoretical explanation of the conventional一of the kind that, in the above example, seemed only to intensify (when mentioned inappropriately) the impotent fury of the frustrated scientist.
Therefore, in discussions about the reality or otherwise of phenomena, the Prasangikas restrict the terms of discussion to the position propounded by the non-Madhyamika opponent. They do not allow themselves, by the use of logical
arguments, to become involved in an exchange that might give the impression that they believe in the real existence of the topic under discussion. It must be stressed that in the debates between the Madhyamika and other philosophies, the only point of issue is real existence. The opponents, Samkhya, Buddhist,
and so on, all contend in one way or other that something exists. The Madhyamikas deny this. Therefore, for Madhyamikas to discourse about phenomena as if they believed in their real existence would, the Prasangikas say, necessarily weaken the force of their argument.
It is important to be aware that a discussion about a thing5s existence is radically different from a discussion about a thing5s attributes. The standard example used to illustrate this point is the debate about the nature of sound. Buddhists find themselves in disagreement with certain Hindus who believe
that sound is permanent, part of the primordial structure of the universe, and so on. The two positions are in total opposition. But in the discussion, both parties are agreed on one thing, namely, sound itself. Sound as a phenomenon can be observed by Buddhist and Hindu alike, irrespective of the ideas
they have about it. However complicated the discussion may become, the situation is clear: both parties are referring to sound; they are disagreeing about its properties.
Discussions about existence, by contrast, are much less straightforward. And it may be observed in passing that the problem at hand evidently concerns the question of whether existence is a predicate. This topic has had a long and interesting career in the history of Western philosophy, and the matter is
still not settled. But since Western Buddhist scholars never seem to advert to it, and since the traditional texts formulate the matter differently, it would perhaps be hazardous to insist upon it too much in the present context. Briefly, the point is that when two people are debating the qualities of
sound, for instance, they can both accept sound as the basis of the discussion without preempting the issue and committing themselves to conclusions that are yet to be established. Whatever the facts of the case, no illogicality is involved in saying, uSound is either permanent or impermanent.^^ It must be
one or the other, of course, but this remains to be demonstrated. There is nothing, however, in the notion of sound itself that logically excludes either permanence or impermanence; and in an inquiry of this kind, one may analytically separate a subject from its properties, even though they are not separable
in experience. One might suppose that the situation is exactly parallel in the statement "Sound is either existent or nonexistent.55 But this is an illusion created by the verbal structure of the sentence. Whereas sound, as a fact of experience, can be considered separately from its permanence or
impermanence, it cannot be considered, with the same propriety and in the same way, in isolation from its existence. We may conceivably have a permanent sound, or an impermanent sound. But we cannot conceivably have a nonexistent sound一that is, a sound that has no existence一since a nonexistent sound is not a sound; it is just nothing. On the other hand, as soon as an object is consciously indicated, existence, or belief in existence, is logically implied.
Consequently, the Prasangikas conclude, in a debate about the existence of phenomena, if instead of confining oneself to an examination of the validity of the opponent's view, one makes an assertion about the phenomenon in question, this very fact is liable to imply that one acquiesces in the thing5s existence. In such debates, therefore, the Prasangikas say that one must abstain from expressing an independent position of one's own on pain of already falsifying one's own position and misrepresenting the case.