Nāgārjuna's fundamental doctrine of Pratītyasamutpāda
By Ewing Chinn
Department of Philosophy, Trinity University
Philosophy East and West
v.51 n.1 (January 2001)
Copyright 2001 by University of Hawai'i Press
Nāgārjuna contends that the doctrine of Pratītyasamutpāda (dependent origination), properly understood, constitutes the philosophical basis for the rejection and avoidance of all metaphysical theories and concepts (including causation). The companion doctrine of śūnyatā constitutes the denial of metaphysical realism (or "essentialism") but does not imply an anti-realist, conventionalist view of reality (as Jay Garfield maintains).
It seems fitting that the very last verse of Nāgārjuna's challenging work, Mūlamadhyamakakārikā
(Fundamental Verses of the Middle Way), would present the reader with what seems to be a riddle: "I prostrate to Gautama, who through compassion, taught the true doctrine, which leads to the relinquishing of all views" (27:30). This should be read with an earlier verse (13:8): "The victorious ones have said that emptiness is the relinquishing of all views. For whomever emptiness is a view, that one will accomplish nothing."1 Since the last chapter deals with questions about the self and the world, it is understandable that so many commentators would assume that the "true doctrine" of which Nāgārjuna speaks is the doctrine of śūnyatā, but that gives a paradoxical character to these verses. Gautama recommends to us a doctrine that all this (we and all that we experience of the world) is śūnya, that is, empty of essence and thus of inherent, independent existence. And does this anti-realist, anti-essentialist view of things (as some would interpret śūnyatā) make possible and even constitute the relinquishing of all views, including his own? It cannot simply be the trivial and disingenuous claim that to accept the truth of śūnyatā is to relinquish all other wrong views. So is Nāgārjuna mysteriously refuting himself? Or is he uttering a paradox?
I will argue that there is no paradox or problem of self-refutation, because the true doctrine that Nāgārjuna refers to is not śūnyatā but the doctrine of Pratītyasamutpāda (dependent arising or origination). And what the verses assert is that if we understand and accept this doctrine, we will no longer have the need or inclination to hold any view about the nature of things, including the inclination to construct a view out of the declaration that "all this is śūnya." How this is possible, how accepting the doctrine makes possible the "relinquishing of all views," certainly depends on what kind of doctrine this is and how it is to be distinguished from a "view." Part of my argument, of course, is that neither pratītyasamutpāda nor the assertion that "all this is śūnya" should be taken as even implying any metaphysical views-either nihilism, absolutism, or anti-realism (the most common metaphysical interpretations). There is no doubt that a major target of Nāgārjuna's philosophical criticism is metaphysical realism. But this does not mean that the critic must hold some alternative metaphysical view.
The logical place to begin is chapter 1, "Examination of Conditions," where the subject of discussion is the concept of Pratītyasamutpāda, but it is not clear at all what Nāgārjuna is trying to establish about this concept. Jay Garfield, in his recent translation of and commentary on the Mūlamadhyamakakārikā titled The Fundamental Wisdom of the Middle Way, offers a unique interpretation of the point of this chapter,
and I will follow him to some extent in his analysis of the text.2 However, we will part ways at a crucial point in the chapter with regard to what Nāgārjuna is about.
On the Nature and Significance of Pratītyasamutpāda
In this first chapter, Garfield notes that "Nāgārjuna distinguishes two possible views of dependent origination or the causal process-one according to which causes bring about their effects in virtue of causal powers and one according to which causal relations simply amount to explanatorily useful regularities-and defends the latter."3 That is to say, according to Garfield, Nāgārjuna defends pratītyasamutpāda as a regularity or Humean theory of causation against an essentialist or realist interpretation. The latter contends that while we may cite one or more factors or events as the causal condition of some effect, it is the force or power in these conditions that is actually the cause of the effect. Nāgārjuna's basic criticism of this realist view is summarized in the fourth verse:
Power to act does not have conditions. There is no power to act without conditions. There are no conditions without power to act. Nor do any have the power to act.
Nāgārjuna had noted in the second verse that causal explanations without exception appeal to one or more of four kinds of conditions:
There are only four conditions, namely, primary condition, objectively supporting condition, immediately contiguous condition, and dominant condition. A fifth condition does not exist.4
So, for example, to explain the lighting of a match, we cite the striking of the match as the primary condition or efficient cause of that event. Thus, "there is no power to act without conditions." That is to say, talk about causal powers only arises in the context of the causal efficacy of one or more of these conditions. We say, "the match will light if it is struck," or "whenever a match is struck it will light." In other words, Nāgārjuna has no problem talking about conditions having the power to bring about an effect, if this is understood as just another way of expressing the Buddhist formula, "whenever this occurs, that will occur." What verse 4 warns against is confusing a functional property of the causal conditions with an existing, essential property, called "power to act." Not only is such an indescribable occult force or power neither evident nor detectable, but, more important, postulating its existence violates the fundamental tenet that nothing can exist without a cause. That is to say, an inherent, essential "power to act has no conditions." Therefore, while "there are no conditions without power to act," we are committing the error of reification if we maintain that conditions have the power to act. Where would we find the conditions necessary to explain why certain things have such occult powers and others do not? There can be no causal explanation for such powers. Causation is thus simply dependent arising: a certain kind of effect invariably comes into existence following upon or as the result of the existence of a certain set of relevant conditions.
But there are also commonly recognized problems with any regularity theory of causation. We know that there are regularities in nature involving causally unrelated events or objects, reported as contingently true generalizations as opposed to necessarily true causal laws. Causation, in short, cannot just mean regularity, the constant conjunction or succession of events. In fact, we quite commonly distinguish the one from the other by explaining that, in the case of causal regularities, the effect is produced by the causal conditions or that it necessarily comes into existence whenever the cause exists. In the latter case, causation is regarded as a necessary connection between two events or things.
Verses 11 and 12 seem to be attacking both of these commonsense explications of causation:
The effect does not exist in the conditions that are separated or combined[.] Therefore, how can that which is not found in the conditions come to be From the conditions?
If that effect, being non-existent [in the conditions,] were to proceed from The conditions, why does it not proceed from non-conditions?
With respect to explaining causation as the product of an effect, Nāgārjuna maintained that "the essence of entities is not present in the conditions," so there is no conceivable sense in which the effect could exist in the conditions. If so, asserts verse 11, then causal conditions cannot produce their effect, unless we are willing to accept the miraculous.
As for the second option, what is this "necessary condition," this third thing that is supposed to exist between cause and effect and that is absent in cases of accidental regularity? Whatever is proposed must surely be as ad hoc as the notion of a power to act advocated by Nāgārjuna's opponents. He is thus faced with the dilemma of either going the route of his realist, essentialist opponents (claiming the existence of something unverifiable) or admitting that he has no answer to the question of what constitutes a causal relationship and is perhaps even forced to deny causation.
It is Garfield's contention that Nāgārjuna avoids these two extremes by an argument that turns the table on his causal-realist critics. I will show, however, that there are some serious problems with this contention and with the argument that he attributes to Nāgārjuna. The argument, he maintains, is based on verse 10:
If things did not exist without essence, the phrase, "When this exists so this will be," would not be acceptable.
As a first step, I shall lay out an argument based on a straightforward reading of this verse. This will turn out not to be the argument Garfield attributes to Nāgārjuna, but it is instructive that we begin there.
Verse 10 seems to be saying straightforwardly that causation exists or is accepted only because existing things lack an essence. It also seems reasonable to assume that to "lack an essence" is not to be an independently existing substance, not to have an identifiable self-nature or essence. And so, to "lack an essence or self-nature" is to
be śūnya. Moreover, it is apparent that the verse is referring to phenomenal things, the things of our ordinary experience, for Garfield takes Nāgārjuna to hold that all there can be are phenomenal things. If we accept all this, the view that everything that exists is śūnya is already implicit in this first chapter. Therefore, verse 10 provides us with the premise that
(a) Causation exists only because the things in the phenomenal world lack an essence (or are not independent substances).
As for the other key premise, recall that the causal realist's main objection to the regularity theory is that it is completely inexplicable how an effect follows from or is causally dependent on certain conditions, if the effect is neither present in some form in the conditions nor necessarily connected to the conditions. But, Garfield contends that this objection (as well as the causal-power theory) presupposes that both the conditions and the effect are not śūnya, that they must have an essence. That is to say, according to Garfield, Nāgārjuna responds "by drawing attention to the connection between a causal power view of causation and an essentialist view of phenomena." In other words, "if one views phenomena as having and as emerging from causal powers, one views them as having essences and as being connected to the essences of other phenomena."5 But now, the causal realist has a problem, because verse 10, the premise (a), denies precisely what the causal realist is insisting on. It denies the presupposition that causation is a relationship between independent substances with essential natures. Premise (a) claims that there can be no causal relationship, substantive or not, between things that are not śūnya. Therefore, Garfield contends, Nāgārjuna is arguing that
[causal realism] is ultimately incoherent since it forces one at the same time to assert the inherent existence of these things, in virtue of their essential identity, and to assert their dependence and productive character, in virtue of their causal history and power. But such dependence and relational character, [Nāgārjuna] suggests, is incompatible with their inherent existence.6
Consequently, the objection to holding a regularity theory of Pratītyasamutpāda fails, since
it is precisely because there is no such reality to things-and hence no entities to serve as the bearers of the causal powers the realist wants to posit-that the Buddhist formula expressing the truth of dependent arising can be asserted. It could not be asserted if in fact there were [such] real entities. For if they were real in the sense important for the realist, they would be independent. So if the formula were interpreted in this context as pointing to any causal power, it would be false. It can only be interpreted, it would follow, as a formula expressing the regularity of nature.7
The problem with this rejoinder to the causal realist is that premise (a) is both false and not what Nāgārjuna would hold. In fact, he would hold, with the causal realist, the exact opposite: to see the things in the phenomenal world as causally dependent or dependently arising is to see them as things with an essence. More-
over, Garfield admits as much when he acknowledges that in the phenomenal world of dependently arising things, in the realm of saṃvṛti satya (empirical truth and knowledge), "we typically perceive and conceive of external phenomena, ourselves, causal powers, moral truths, and so forth as independently existing, intrinsically identifiable and substantial."8 It seems, then, that Nāgārjuna would hold, quite correctly and contrary to premise (a), that what we experience as dependently arising are identifiable things, independent substances, or things with an essence. After all, these ordinary substances of our experience are not Spinozaistic self-caused substances.
But I have knowingly misrepresented Garfield's analysis of Nāgārjuna's rejoinder, because further examination reveals that premise (a) is not how Garfield wants to interpret verse 10. Premise (a) presents the verse as talking about phenomenal objects as phenomenal objects (as the things that we experience and describe), and as making the false claim (false from the standpoint of saṃvṛti satya) that such objects are without essence. Garfield, however, must be reading the verse as talking about phenomenal objects from the ultimate point of view. It is as an ultimate truth (paramārtha satya) that these phenomenal objects are "without essence." But by "without essence" Garfield does not mean what we normally mean by that phrase. What it means for Garfield is that the objects that we perceive as identifiable and substantial things are not things in themselves. Garfield is peculiarly using the term "without essence" to refer to the mode of existence of an object, and not to the nature of the object. This is indicated by the fact that he alternates between "with essence" and "inherently exists" or "without essence" and "to lack inherent existence."
Most important of all, if nothing in the phenomenal world inherently exists, if there are no things in themselves, this must mean that everything exists as a matter of convention. And it is this that Garfield takes to be the important meaning of "all is śūnya." For Garfield, the essential meaning of śūnyatā, what it means to say of phenomenal objects that they are śūnya, is that independent of our experience, outside our conceptual and linguistic framework, these things are nothing. They simply do not exist. Basically, to be a "conventional object" is to be lacking in "inherent existence." It is to be śūnya. The essence of Nāgārjuna's philosophy, as Garfield stresses, is the "dual thesis of the conventional reality of phenomena together with their lack of inherent existence."9 Therefore, it is important to keep in mind that, according to Garfield, what Nāgārjuna means by saying of a thing that it is śūnya is not that that thing lacks an essence or self-nature, but that it possess that essence or self-nature by virtue of conventions.
Clearly to affirm the "conventional reality of phenomena" is to reject metaphysical realism and to endorse anti-realism. We can assume that by characterizing our designations and descriptions of objects as "conventional," Garfield means to reject the realist tenet that there is only one true theory or description of reality. And to claim that no existent exists "from its own side," independent of the knower, is to oppose the tenet that reality is mind-independent.10 It is to deny the theory that there are things-in-themselves, to deny that metaphysical realism makes any sense.
With this clarification in hand, we can finally see how Garfield interprets verse 10. He understands it to be asserting that
(b) Causation exists only because phenomenal things (the things we experience) exist by convention and are not things in themselves.
We now realize-with (b) replacing (a)-that what Garfield means "by drawing attention to the connection between a causal power view of causation and an essentialist view of phenomena" is exposing the fact that the causal realist is looking for an objective, "inherently existing" causal connection or force between things that "inherently exist," that is, things-in-themselves. Therefore, Garfield conceives of the argument as follows:
1 .The causal realist maintains that causation exists only if there is some independent causal power or causal connection (a third thing) between the causal conditions and the effect.
2. And they assume that such forces and connections only exist in and between things that exist in themselves (i.e., things that exist inherently).
3. But, according to (b), this assumption is false and the contrary is true.
4. Therefore, causation can only be the regularity of "when this exists, that will exist," that is, pratītyasamutpāda.
This is a better argument, but still not persuasive. Even if we accept the heart of the argument, step 3, this does not answer the causal realist's challenge in step 1. Whether the things we experience are things in themselves or are merely conventional, it still seems legitimate to ask how causation is possible. The causal realist is asking for an explanation of why we would take one thing to be the cause of another, why anything should be regarded as a causal condition. Whether causal powers or necessary connections (or anything, for that matter) exist in themselves is a separate issue from whether it is legitimate to conceive of causality without conceiving of causal powers or necessary connections. In other words, the conclusion simply does not follow.
Moreover, (b) is problematic. While it is evident that (a) is wrong in denying that the phenomenal objects we experience are causally dependent, it is not clear what to make of U's claim that causation cannot apply to things-in-themselves. It would seem obvious that if it is legitimate to conceive of phenomenal objects as causally dependent, it would be equally legitimate to ascribe causation to objects that exist outside our experience, if there are such objects.
Furthermore, from the argument above, Garfield obviously thinks that (b) implies that pratītyasamutpāda is empty, that it is just a conventional and useful (and thus optional?) way of treating and conceiving of things in the phenomenal world. At the end of his analysis of chapter 1, as a way of summing up, Garfield states:
[T]o assert the emptiness of causation is to accept the utility of our causal discourse and explanatory practices, but [is] to resist the temptation to see these as grounded in reference to causal powers or as demanding such grounding.11
I will argue in the last section that this is close to what Nāgārjuna eventually recommends, but to "accept the utility of our causal discourse and explanatory practices" does not necessarily mean that causation is empty or a matter of convention. Moreover, once again, why would it not be just as useful to apply causation to things-in-themselves? Finally, if the important purpose of chapter 1 is to defend the emptiness of causation, it is odd that the term is never mentioned. How could Nāgārjuna expect the reader to know that "without essence" is a synonym for śūnyatā and, more important, that it is to mean "to exist as a convention?" In fact, I believe that it is not his concern in chapter 1 to defend a regularity theory of pratītyasamutpāda, empty or not, against the theories of a causal realist.
To find a different interpretation of the point of chapter 1 that does not have the problems of Garfield's, we must again reexamine verse 10. In a footnote, Garfield points out the interesting fact that his translation of this verse is in the minority, and he mentions that a number of prominent translations (by Inada, Streng, Sprung, and Kalupahana) are "diametrically opposed" to his.12 He cites, as an example (without giving the source), the following translation:
Since things exist without essence the assertion "When this exists, this will be" is not acceptable.
We see that this is indeed diametrically opposed to Garfield's:
If things did not exist without essence, the phrase, "When this exists so this will be," would not be acceptable.
We have seen that what Garfield's version asserts is:
There is no dependent arising (any causation) unless things (that we experience) exist by convention.
There is no dependent arising (any causation) between things-in-themselves.
But, if we follow Garfield in taking "exist without essence" to mean "conventional existence," the other version asserts, on the contrary, that
There is no dependent arising because things only exist by convention.
And this implies that
There is no dependent arising unless there are things that exist in themselves.
Clearly, while Garfield's version is a statement of Nāgārjuna's position, the other versions must be a statement of the causal realist's position (in light of Garfield's "conventionalist" interpretation of śūnyatā). The latter would mean that verse 10 initiates the causal realist's attack on Nāgārjuna that goes on to the end of the chapter, whereas Garfield has argued that 10 is the heart of Nāgārjuna's response to the attack, a response that is summarized in verse 14, the last verse of the chapter.13 Why are there such discrepancies between translations and interpretations?
Commentators have long realized that it is impossible to give a coherent reading of chapter 1, or any difficult text, without making some assumptions. Garfield's reading is governed by the assumption that Nāgārjuna had more in mind than just defending a regularity view of causation against the causal realist. What he sought to show in chapter 1, as his central purpose, is the emptiness of causation. Furthermore, Garfield maintains, "by showing the emptiness of causation itself, we understand the nature of emptiness itself ... [and] by showing causation to be empty we show all things to be empty, even emptiness itself."14 By the "emptiness of causation" he means, of course, that it has only a conventional reality, that it is something we introduce as a way of conceiving of the world. Therefore, the significance of showing that causation is empty at the very beginning is to establish what emptiness means, and this is a large step toward eventually arriving at his major philosophical thesis, the conventional reality of all things. It is based on this assumption that Garfield interprets verse 10 as he does, and this is the reason he takes "inherent existence" and "to exist with an essence" to mean "exist as a thing-in-itself independent of all conventions."
All this is assumed because there is no explicit textual evidence in the chapter, so the only justification or proof that this is Nāgārjuna's real purpose is that it gives a coherent interpretation of the text. This is a bit question-begging, but I doubt if any commentator can avoid this with respect to a difficult, if not obscure text. The problems I have with Garfield's interpretation, however, are that it portrays Nāgārjuna as making a very weak and problematic argument, and that it demands reading a great deal into the text. It is only much later in the text that the terms śūnya and śūnyatā appear and only then that Nāgārjuna talks about everything being śūnya. It is taking a great deal of liberty to read "without essence" as "śūnya" and to take the latter to mean "exists by convention."
On the other hand, the opposing translation of verse 10 gives us a simpler and cleaner reading of the text. That verse is just an introduction to the realist's objections to the regularity theory in verses 11 to 13, and the last verse, 14, concludes that both approaches, the realist and the regularity, have their problems and are thus unsatisfactory. But I think that there is more going on than just giving two sides of a debate. Why would Nāgārjuna begin his important treatise with such a weak and inconsequential chapter? Garfield's interpretation at least has the merit of giving significance to the chapter.
Since I have been arguing that verse 10 is the key to understanding chapter 1, there is one last version I want to examine and see where it leads, David Kalupahana's translation:
Since the existence of existents devoid of self-nature is not evident, the statement: "When that exists, this comes to be," will not be appropriate.15
Kalupahana routinely takes "not evident" (na vidyate) to have the epistemological meaning of "not found in our experience," and uses the words "existence" and "to exist" in the realist sense of "existing in itself." So, the first half of the verse says, "our experience does not reveal that there really are (or 'exists in itself') 'existents
devoid of self nature."' And if we again follow Garfield in taking "existents devoid of self-nature (i.e., essence)" as referring to "things that exist by convention" or "conventional objects," we can construe Kalupahana's version as asserting essentially the same thing as the one Garfield cites in that footnote. It is asserting that
Causation is not possible (nothing dependently arises), because there is nothing in our experience of these conventional objects to reveal or prove that they exist in themselves.
And this again implies that
There is no dependent arising unless there are things that exist in themselves.
But there is an alternative and simpler reading that is faithful to Kalupahana's use of all the relevant terms. He takes "without essence or self-nature (svabhāva)" to mean exactly what it says: that the things referred to lack an essence or self-nature. Because I prefer Kalupahana's translation and agree with his use of these important terms, "without essence," "to exist," and "evident," I contend that what verse 10 is asserting is:
(c) Causation is not possible, because (or if) there is nothing in our experience that reveals or proves the existence of things that lack an essence or self-nature.
What is Nāgārjuna talking about? These things that lack an essence or self- nature are certainly not the ordinary things that we experience-not phenomenal objects-for I agree with Garfield that Nāgārjuna regards the things we experience as things identifiable by their essences. In order to answer this question and to see what is really going on in chapter 1, I must assume something that is seemingly innocuous, but certainly safe, for it is backed by independent and historical evidence. I will assume that the background and motivation for this chapter was a concern with the philosophical debate within Buddhism over the metaphysical doctrine of dharmas. This was the theory that imperceptible elements underlie our experiential world of phenomena, and it is these elements that are ultimately real, that ultimately exist. And I will assume that chapter 1 is discussing the problem of causation with respect to dharma-like entities and not phenomenal objects. Consequently, the term "things that exist" or "existents" appearing in this verse and elsewhere in the chapter refers to theoretical, metaphysical entities like dharmas.
The main antagonists were two schools of Buddhist realism. One was the Sautrāntika School, which held that these elements, these "existents," are "durationless," existing not even for one moment of measurable time-the theory of "momentariness" (kṣaṇika-vāda). Thus, a dharma must be without a permanent essence, for the elements themselves are not permanent. The other is the school that Kalupahana calls "one of the most explicit and unqualified essentialist views ever to appear in the Buddhist philosophical tradition," the Sarvāstivāda School.16 The Sarvāstivādins criticized the Sautrāntikas for their inability to make sense out of the formula of pratītyasamutpāda, for reducing causation to the mere succession of durationless events. In order to solve the problem that the Sautrāntikas could not solve, these essentialist philosophers developed a sophisticated theory of the "self-
nature" (svabhāva) of dharmas. So, each dharma possesses an unchanging, inherent essence, the ground for the changes that one observes in the phenomenal world.
The criticisms of the causal-power theory would certainly apply to Sarvāstivāda realism, to their efforts to devise a theory of Pratītyasamutpāda that is based on the presumption that dharmas possess a self-nature or essence. But Nāgārjuna is attacking all realist views of causation that are based on the assumption that there are things in themselves or substances with a self-nature (svabhāva) that underlies phenomena. In the second half of the chapter, beginning with verse 10, Nāgārjuna is not just returning the favor for the Sarvāstivādins and giving their criticisms of these nonessentialist rivals. Neither verse 10 nor any other verse in the chapter contains what we would expect to be an essentialist's criticism. The problem the Sarvāstivādins have with applying pratītyasamutpāda to durationless and essenceless dharmas is that it does not work. The criticism of verse 10, in Kalupahana's translation, is a Nāgārjuna-like criticism that it does not make sense to talk about the dependent arising of entities whose existence is not evident In other words, there is no empirical evidence for the existence of essenceless elements, for all that experience reveals are things identified by their essence.17 Finally, the last verse, 14, is not just concluding that both realist views are unacceptable theories of causation. The gist of that verse is that neither can explain the fact of pratītyasamutpāda. In Kalupahana's translation, the verse says:
An effect made either of conditions or of non-conditions is, therefore, not evident. Because of the absence of the effect, where could conditions or non-conditions be evident?
Nāgārjuna conceives of these rival Buddhist schools, the Sarvāstivāda and the Sautrāntika, as representing, respectively, the identity and nonidentity approaches to the explanation of causation. He claims to have shown that these theories would actually make it impossible to accept what is so evident to our experience, the fact of dependent arising. In different ways they would both bring into question or render problematic any attempt to cite certain conditions as the causes of an existing effect, when these very things are facts of common experience. Thus, Nāgārjuna warns, "if there are no such effects [due to the failure of the realist to explain how causation is supposed to work according to their theories], how could conditions or nonconditions be evident." It is his contention, in short, that Buddhist realism (or any realist approach to causation) will run aground in its efforts to explain causation on a metaphysical level.
Nāgārjuna, in this first chapter, rather than defending a particular theory of pratītyasamutpāda, is trying to convince the reader of the futility of speculating about the real nature of causation, with the implication that we should renounce any philosophical theory of causation. I will show in the next section that this includes the anti-realism or "conventionalist theory of pratītyasamutpāda" that Garfield attributes to Nāgārjuna.
In addition to preparing the ground for similar demonstrations of the futility of abstract, theoretical speculation and debate over various philosophical topics taken up in succeeding chapters, the further importance of chapter 1 is that it points the
way to his positive position on causation and other basic categories of human experience. We shall see that the notion of pratītyasamutpāda is far richer than the ordinary notion of causation. Most important of all, we shall see that to understand this notion and its implications fully is to understand Nāgārjuna's philosophical orientation. And finally, we shall have the answer to how the true doctrine of pratītyasamutpāda enables us to relinquish all views.
Pratītyasamutpāda and Śūnyatā
We are used to dealing with all kinds of causal questions, such as, for example, "Why do you think that a short in the electrical system caused the fire?" or "What brought on his nervous breakdown?" The answer we give to any causal question clearly depends on our knowledge of the particular facts of the situation and our general knowledge of such things as electrical systems and the symptoms and causes of a nervous breakdown. I would think that there are many of us who have no idea how to answer one or the other or both of these questions. Anyone who lacks the basic conceptual or theoretical understanding of a phenomenon would not know what to look for or even where to begin. Nevertheless, as a result of past experience, we are certain that there must be, in each situation, some actual set of conditions that best explains the existence of any phenomenon. Pratītyasamutpāda can be taken, on the simplest, most rudimentary level, to refer to this universally accepted practice and belief that everything has a cause, that every phenomenon is dependently arisen. We might call this the scientific principle of pratītyasamutpāda.
Philosophers, of course, have no interest in such mundane "factual" problems as the particular cause of someone's mental condition. Their concern is with serious questions about the foundations of things, believing that we cannot take even the most familiar of things for granted or at their face value. In this case, they would be concerned with the very idea of causation or cause and effect, with the general philosophical question of "what makes anything the cause of something." They would worry about the possibility that there is really no such thing as a cause-and effect relationship, no objective fact of the matter behind our causal discourse.
Gautama, the Buddha, chose to remain silent on this and all philosophical questions, to take no side in any dispute. There has been a great deal of speculation about whether his silence represents being skeptical or being disinterested, or something more esoteric. Nāgārjuna, on the other hand, goes much further in urging that no one should hold any views on these matters. He claims that the Buddha has taught the tattva, the true doctrine, or, more literally, "the exact or real nature of the case" in order that we may relinquish all views (dṛṣṭi). So, is Nāgārjuna, the philosopher, telling us that we do not have to philosophize? Is he, like Wittgenstein, telling us that there is a cure for the disease?
To answer these and other questions, we need to have a deeper understanding of this true doctrine of pratītyasamutpāda, this doctrine that teaches that everything is dependently arisen. What is being taught, I contend, is a two-sided principle, the first side of which is this scientific and regulative principle that literally every phenome-
non is dependently arisen. This pertains not just to the objects of our world, physical things or material form, but to any of our experiences in the broadest sense of that word. It holds importantly to all our ideas and beliefs of our world, even if an idea or belief does not seem to be experientially derived-for example, the beliefs that the world is eternal or not eternal, created or not created, that there is an indestructible self, and so on. This seems to be what Nāgārjuna intimates in 4:7: "The method of treatment of all existents such as feeling, thought, perception, and dispositions is in every way similar to that of material form."
I imagine that a crucial implication of the principle is that there are no transcendental categories of experience in the Kantian sense of innate ideas, with the possible exception of pratītyasamutpāda itself. But even in this case, as we shall see, Nāgārjuna makes no distinction between empirically contingent ideas and rational, necessary ideas. And there are no ideas or beliefs inherent in the nature of the mind or in the way we think, no matter how abstract or universal. Whatever idea or belief we possess must have a causal explanation in a set of conditions that may include the circumstances and content of our experience, the background of ideas and beliefs, whatever biases we may have, and other psychological factors.
We would not have the rudimentary notion of cause and effect, for example, unless we experienced not only the regularity with which sequences of events occur but also the particular kinds of effects associated with particular kinds of things. We experience the heat or burning of fire, the sounds of animals, the breaking of glass by hard objects, and so forth. Our conception of causation, in terms of its origin, is connected not only to temporal succession but also to the observable properties of various kinds of objects and of our ideas of the nature of these objects and properties. The crucial implication of the contingent origin of our ideas and beliefs about causation is that the very meaning of the term, indeed of any term, is constituted by its place in a web of other concepts and beliefs.
This is the semantic side of the doctrine of pratītyasamutpāda, the doctrine of the mutual dependency of concepts and beliefs in both the systematic and historically contingent sense.18 But this is not just a coherence or holistic view of meaning. It is, more appropriately, a kind of pragmatic understanding of meaning. To "possess" an idea or to understand a term is to have the ability to use that term appropriately. It is to know how to navigate within an intellectual sea of ideas and beliefs. To ask for the "meaning" of a term is to ask for instructions about what to do with that term, for example what kinds of objects or features the term should be used to refer to, or on what occasions or circumstances that term should be used and how it should be used on those occasions. Giving a synonym or dictionary definition for a term is also appropriate, but these are either compact sets of instructions or mediates between the definiendum and its "meaning," that is, the knowledge of how to use that term. Meaning, then, can change and vary from time to time, place to place, and person to person. It is a dynamic theory of meaning.
I take Nāgārjuna to be opposing the idea that there is something called the "meaning" of a term, something denoted by that term without variation through time. According to this kind of denotative or essentialist theory, the "meaning" of a
term serves as a criterion or model, for it purports to represent the nature or essence of the things to which the term applies. This is not to say that for the contextual and functional theory of Nāgārjuna there cannot be definitive criteria for some terms. But the criterion, in this case, must consist of an open-ended set of beliefs about the character and behavior of the objects that fall under the term, and it is a matter of historical consensus as to what belongs to that set. However, essentialist theories of meaning can be very seductive.
For example, when thoughtful people reflect on their experience with causality, that is, reflect on the extensive record of the successful use of that idea in their intellectual and practical lives, there arises the desire to understand more deeply why everything has a cause. We become obsessed with the question "What is cause and effect?" Or "What is the real meaning of one thing causing another?" It is the desire to understand what the term "causation" really stands for or represents by itself. It is thus that we become captives of an essentialist or denotative view of meaning. We want to know the essential nature (svabhāva) of causation, the "thing" denoted by that term.
The term dṛṣṭi (view) is used to refer to the answer we give to this kind of question, and it is understood that the answer, the view given, must be rationally compelling, either logically demonstrable or self-evident. This is because, according to the essentialist view of meaning, the relationship between a term and its designatum must be necessary and eternal. Nāgārjuna realizes that no dṛṣṭi can satisfy this condition, but it remains an obsession nevertheless. Frederick Streng has it right when he explicates dṛṣṭi as standing for "illusory mental effort-a view, or doctrine that claims absolutely validity on the grounds that it asserts a self-evident truth."19 Streng could be talking about treating causation as a necessary connection or causal powers when he remarks on the "inappropriateness of our acting as if we could discern a self-evident reality either in the conditioned 'thing' or in some identifiable 'element' of our experience (like 'origination,' 'duration,' or 'cessation')." The problem, he notes, is that "[b]y seizing on one aspect and making decisions ... on the assumption that it is an ultimate (self-existent) reality, human beings mistake their judgments for the nature of existence."20
The problem with this powerful psychological tendency to engage in speculative metaphysics, this need to discover the nature of the thing itself, is eloquently and poignantly voiced by Kant in the very first sentence of the first preface to his Critique of Pure Reason:
Human reason has this peculiar fate that in one species of its knowledge it is burdened by questions which, prescribed by the very nature of reason itself, it is not able to ignore, but which, as transcending all its powers, it is also not able to answer.21
Nāgārjuna understood this natural desire of the human mind to be contingent (dependently arisen) and not inherent in the nature of reason as Kant thought. This desire to possess an absolute understanding of the basic contours of experience creates the obsessive delusion of thinking that we are dealing with an existing substance (svabhāva) when what we really have is a construction of our minds. The
philosopher's idea of causation as a "power" or a "necessary connection between events" is related to, but distinct from, our ordinary notion of causation or cause and effect. All we can say, when pressed for a definition, is that "cause and effect" just means "cause and effect." It is just the knowledge that "when this exists, that exists." The former, the philosophical idea or belief in causation as a power or existent connection, Nāgārjuna contends, is a saṃkalpa ("mental fabrication") and a prapañca ("obsession" according to Kalupahana and "phenomenal extension" according to Streng). Calling it saṃkalpa indicates that there is no empirical support, nothing in our experience even to suggest anything like a "causal power." Calling it prapañca locates the source of the idea in our desire to "go beyond" the ordinary use of the term, to leave the "home" of the notion in the experiential world, to forget that its real meaning is its function. We are driven by the obsessive desire to grasp the objective reality or real designatum of an idea; but in effect, as Streng again puts it, we have mistaken our "[illusory] judgment for the nature of existence."
It is in this context that Nāgārjuna invokes the notions of śūnya and śūnyata. He seems at times to be using the idea of "emptiness" as a tool to free us from a picture that has held us captive (to borrow a phrase from Wittgenstein), to release us from the entrapment of a deviant and perverse theory of meaning. He declares that "the Victorious ones have announced that emptiness is the relinquishing of all views," but warns that "those who are possessed of the view [dṛṣṭi] of emptiness are said to be incorrigible." He fully expected his remarks about śūnya and śūnyatā to be misinterpreted as expressing a metaphysical view of reality, as playing their metaphysical language game. To those individuals who are so enamored by the game he admonishes, "you do not comprehend the purpose of emptiness. As such, you are tormented by emptiness and the meaning of emptiness."22 To ask if he is an absolutist, nihilist, or conventionalist is to assume wrongly that he is holding a view concerning what is real with his talk about śūnyatā .
On the contrary, he is not engaged in metaphysical theorizing or speculation, but is urging a disengagement from this way of thinking. He is trying to release us from being held captive by a "picture of reality" for which our language must assume a large burden of responsibility. For example, the claim that "the nature of women makes them different from men" makes perfectly good sense in ordinary discourse when we take it to mean "there is (are) something(s) about women that makes) them different from men, that is, some observable features) of women." And we can continue the conversation by pointing out what these might be. Unfortunately, it has also been taken to mean that "the essential self-existent nature of all women is what causes there to be the differences we observe in women and men." Nāgārjuna would confront us with the illusion expressed in this last statement by asserting that the subject of the statement is empty. Ascribing a nature to all women, moreover, tells us nothing about what differentiates men from women, so there is no natural way to continue the conversation. Another example is "the power of X to cause Y," leading a person into thinking that we are talking about something that X has rather than about what X does or what X has the ability to do.
Nāgārjuna declares that these unobservable metaphysical entities, these mental
constructions, are nonexistent by asserting that the world we experience is devoid of such entities. All this is empty. Therefore, nothing we say about the nonexistent, śūnyatā , makes any sense, and worse, may lead to "defilements of action" or misguided courses of action. So, Nāgārjuna says, "[w]hen he is empty of self-nature, the thought that the Buddha exists or does not exist after death is not appropriate" (22:14). And, "[d]efilements of action belong to one who discriminates, and these in turn result from obsession. Obsession in its turn, ceases within the context of emptiness" (18:5).
It is not difficult to see that that "all this is śūnya" declares that we should relinquish the view of metaphysical realism. We should resist being tempted by the view that what we experience and believe about our world are mere representations that may or may not correspond to (or be "made true by") the real existents, the "inherent existents," that exist independent of and beyond our experience. The danger of operating from this view, Nāgārjuna advises, is that "If you perceive the existence of all things in terms of their [inherent] essence, then this perception of all things will be without the perception of causes and conditions" (24:16). The consequence of seeking for the real, independent essence of things, of thinking that we only experience the appearance of what really exists, is to overlook or confine to the background the causal and semantic relationships that are actually at work in giving us the world that we have, all that should count as real.
We should not forget that Nāgārjuna declared that "śūnyatā is the relinquishing of all views." Therefore, he would not endorse any side in the realist-anti-realist debate. He would not hold the opposing view that the objective world is a vacuous notion, that "what there is" is determined by the conventions of our language. Jay Garfield exemplifies this kind of interpretation when he has Nāgārjuna maintaining that, on the level of ordinary, conventional knowledge and truth (saṃvṛti satya),
Conventional phenomena are typically represented as inherently existent. We typically perceive of external phenomena, ourselves, causal powers, moral truths, and so forth as independently existing, intrinsically identifiable and substantial.... Mo see them this way is precisely not to see them as conventional.23
But on the level of ultimate truth, absolute knowledge (paramātha satya), what we experience as real entities and processes are, in truth, conventional existents. "Emptiness" has more than a negative meaning in the rejection of metaphysical realism. For Garfield, it has the positive meaning of pointing to "the fact that conventional dependent phenomena are conventional and dependent... [and] it is simply the only way in which anything can exist." He takes the statement in 24:18, "that [the dependent arising], being a dependent designation is itself the middle way," to mean, "existence depends on designation" or verbal convention. And most important of all, Garfield contends that "our conventions and our conceptual framework can never be justified by demonstrating their correspondence to an independent reality.... [Therefore, Nāgārjuna] suggests, what counts as real depends precisely on our conventions." Garfield's articulation of this anti-realist view of Nāgārjuna verges at times on nihilism. For example, he says, in reference to ordinary things, "their
ultimate nonexistence and their conventional existence are the same thing." But, for Nāgārjuna, "a wise person does not say 'exists' or 'does not exist."'
The true doctrine of Pratītyasamutpāda, as a two-sided semantic and causal principle, frees us from the extremes of realism and anti-realism, absolutism and nihilism. Richard Rorty is talking about the bankruptcy of the extremes of realism and anti-realism, when he says:
"Determinacy" is not what is in question-that neither does thought determine reality nor, in the sense intended by the realist, does reality determine thought. More precisely, it is no truer that "atoms are what they are because we use 'atom' as we do" than that "we use 'atom' as we do because atoms are as they are." Both of these claims ... are entirely empty. Both are pseudo-explanations.24
The truth, Rorty explains (and this is the causal side of the doctrine of pratītyasamutpāda), is that
our language, like our bodies, has been shaped by the environment we live in. Indeed, he or she insists on this point-the point that our minds or our language could not (as the representationalist skeptic fears) be "out of touch with reality" any more than our bodies could.25
Nāgārjuna would approve of Hilary Putnam's metaphorical remark that "the mind and the world jointly make up the mind and the world."26 That is to say, the existence of the world is just as dependent on language as the language that we use is dependent on the world. Or, as Nāgārjuna asserts, "if characteristics do not appear, then it is not tenable to posit the characterized object. If the characterized object is not posited, there would be no characteristics either" (5:4).
His contention in 24:18 that "whatever is dependently arisen, that is explained to be emptiness," is simply the rejection of metaphysical realism, by declaring that there is nothing but the dependently arisen. And his further remark "that [the world as we know it] is dependent on convention [and] that is itself the middle way" is a warning against transgressing the boundaries of ordinary discourse, the boundaries of our natural language, and thus "losing the world."
We would be transgressing the bounds of actual discourse if we worry about the correspondence of our designations to some reality that stands unspecified, or if we take the opposite view that reality is what we specify and that there are alternative, unconstrained ways of conceiving of what is real. The implication of pratītyasamutpāda is that our language, like anything else in the world, is shaped by the environment we live in, and that our language cannot be "out of touch with reality" any more than we can be.
Human beings cannot live without ideals, without something to strive for that gives their existence authenticity. Many have searched for the ultimate meaning of their existence, some point to the exigencies of their mundane life, in a reality that transcends that life. Nāgārjuna tries to discourage this kind of search by assuring us that whatever truth and meaning there is to one's existence can be and must be found within the confines of our human world.
Nāgārjuna asserts in 24:10: "Without relying upon convention, the ultimate fruit [paramārtha satya) is not taught. Without understanding the ultimate fruit, freedom is not attained." As I read this verse, the real distinction between saṃvṛti satya (ordinary, conventional truth) and paramārtha satya is not an ontological distinction, where the latter is the knowledge and truth about ultimate reality. It is a distinction between ordinary consciousness, which is fraught with the dangers of misconceptions influenced by all kinds of desires (the desire for certainty, for eternal existence, etc.), and an enlightened consciousness purged of all the needs and fears that bring about suffering. The teachings of pratītyasamutpāda and śūnyatā are Nāgārjuna's way to put us on the path to nirvana, to the nurturing of an enlightened consciousness and a way of living. It is, for Nāgārjuna, the philosophical foundation of the middle way of Buddhism.
1 - Since I will be addressing Garfield's interpretation of Nāgārjuna in the first part of this essay, I will be using Garfield's translation of the Fundamental Verses of the Middle Way (New York: Oxford University Press, 1995), unless otherwise noted. [Back]
2- Essentially the same analysis is in an earlier essay: Jay Garfield, "Dependent Arising and the Emptiness of Emptiness: Why Did Nāgārjuna Start with Causation?" Philosophy East and West 44, no. 2 (April 1994). The analysis and argument is presented in more detail in Garfield, The Fundamental Wisdom of the Middle Way. [Back]
3 - Garfield, "Dependent Arising," p. 222. [Back]
4 - Nāgārjuna is following the standard classification of kinds of causal conditions or causes in the Indian philosophical tradition. [Back]
5 - Garfield, Fundamental Wisdom, p. 118.[Back]
6 - Ibid. [Back]
7 - Ibid., p. 119. [Back]
8 - Garfield, "Dependent Arising," pp. 232-233. Garfield nevertheless seems to be committed to the assumption that things with essences do not causally arise. He contends that "essences are by definition eternal and fixed. They are independent. And for a phenomenon to have an essence is for it to have some permanent independent core. So, neither essences nor phenomena with essences can emerge from conditions" (Garfield, Fundamental Wisdom, p. 111). But surely, it depends on the essence. Some essences, for example the essence of a kind of plant or animal, are such that things with that essence are dependently arisen, but not so with the essence of being the greatest of all beings. [Back]
9 - Verses 18 and 19 of chapter 24 are crucial to Garfield's conventionalist interpretation of śūnyatā :
Whatever is dependently co-arisen, that is explained to be emptiness. That being a dependent designation, is itself the middle Way. Something that is not dependently arisen, such a thing does not exist. Therefore, a nonempty thing does not exist.
As Garfield reads these verses, to assert that an object in the phenomenal world, a causally dependent object like a table, is śūnya "is to say ... that it does not exist "from its own side"-that its existence as the object that it is-as a table-depends not on it, nor on any purely relational characteristics, but depends on us as well" (Garfield, Fundamental Wisdom, p. 89). And to say that is to say that it is a dependent designation, that is, that "whatever is dependently co-arisen is verbally established ... that its identity as a single entity is nothing more than its being the referent of a word" (Garfield, "Dependent Arising," p. 229). In short, all that exists is determined by our theoretical and conceptual apparatus; that is, "it depends on us." Without that apparatus, independent of us, nothing exists. [Back]
10 - Garfield is careful to avoid being misread as subscribing to either of two extremes, nihilism and absolutism. He thus insists that "to view the dependently originated world [as conventionally real] is to see it neither as nonempty nor as completely nonexistent." In particular he sees 24:11 as expressing the concern with the mistake of an absolutistic interpretation of śūnyatā : "By a misperception of emptiness a person of little intelligence is destroyed, like a snake incorrectly seized or like a spell incorrectly cast." Thus he follows Candrakīrti in defending the emptiness of emptiness, a way of expressing the conventionalist thesis, denying that "emptiness" refers to something beyond description and convention.
11 - Garfield, Fundamental Wisdom, p. 122. [Back]
12 - Ibid., p. 119. In this footnote 33, Garfield intimates that these opposing translations are taking liberty with the text by remarking that "[t]hey may be suggested by Candrakīrti's comments to the effect that this phrase ["When this exists, this will be"] would make no sense were it asserted by the realist." Garfield has no problems here with his presumably literal translation, because he sees the verse as expressing Nāgārjuna's position and not the realist's. [Back]
13 - Verse 14 reads: "Therefore, neither with conditions as their essence, nor with non-conditions as their essence are there any effects. If there are no such effects, how could conditions or non-conditions be evident?" It is difficult to see that as summarizing Nāgārjuna's response to the causal realist, and Garfield's explication tends to confuse rather than clarify. See pp. 121-122 of Garfield, Fundamental Wisdom. [Back]
14 - Garfield, "Dependent Arising," pp. 237-238. [Back]
15 - Nāgārjuna: The Philosophy of the Middle Way, introd., Sanskrit text, English trans., and annot. David J. Kalupahana (Albany: State University of New York Press, 1986), p. 113. [Back]
16 - David Kalupahana, A History of Buddhist Philosophy (Honolulu: University of Hawaii Press, 1992), p. 128. [Back]
17 - There is perhaps a secondary motive for the way Nāgārjuna argues. It is possible that the Sarvāstivādin realists interpreted Nāgārjuna's criticism of their position as implying that he held a metaphysical view similar to the Sautrāntikas, indeed that he might even be a spokesman for that Buddhist school. Therefore, Nāgārjuna would want to correct that misunderstanding, by attacking the latter with equal force. [Back]
18 - A precursor of this interpretation may be Candrakīrti. See Mervyn Sprung's discussion of Candrakīrti's commentary on chapter 1 in Lucid Exposition of the Middle Way: The Essential Chapters from the Prasannapadā, trans. from the Sanskrit by Mervyn Sprung, in collaboration with T.R.V. Murti and U. S. Vyas (London: Routledge and K. Paul; Boulder: Prajna Press, 1979), pp. 32-52. [Back]
19 - Frederick Streng, "The Significance of Pratītyasamutpāda," in Mervyn Sprung, ed., The Problem of Two Truths in Buddhism and Vedanta (Dordrecht and Boston: D. Reidel, 1973), p. 32. [Back]
20 - Ibid., p. 30. [Back]
21 - Immanuel Kant, Critique of Pure Reason, trans. Norman Kemp Smith (New York: St. Martin's Press, 1963), p. 7. [Back]
22 - This is Kalupahana's translation of 24: 7, as well as of 13:8 above. All the translations hereafter are Kalupahana's. [Back]
23 - Garfield, "Dependent Arising," pp. 232-233. [Back]
24 - Richard Rorty, Objectivity, Relativism, and Truth (Cambridge and New York: Cambridge University Press, 1991), p. 5. [Back]
25 - Ibid. [Back]
26 - Hilary Putnam, The Many Faces of Realism (La Salle: Open Court, 1987), p. 1. [Back]