... we find ourselves in the midst of a rude fetishism when we call to mind the basic presuppositions of the metaphysics of language-which is to say, of reason. It is this which sees everywhere deed and doer; this which believes in will as cause in general; this which believes in the 'ego', in the ego as being, in the ego and substance, and which projects its belief in the ego-substance on to all things-only thus does it create the concept 'thing'.... 'Reason' in language: oh what a deceitful old woman! I fear we are not getting rid of God because we still believe in grammar.... Nietzsche'

Why Nagarjuna and Dogen ...? Such a comparison is inviting because both are obvious and difficult. On the one hand, they are arguably the two greatest Mahayana thinkers, linked by their commitment to its understanding of the world and (if we accept the traditional account) by a transmission lineage that extends from Sakyamuni through Nagarjuna to Dogen and his successors. On the other hand, however, are vast cultural differences, due not only to the geography and the millennium that separate them but just as much to the disparity between their very different languages, Sanskrit and Japanese.

These linguistic differences are further reflected in their extraordinarily different---I am tempted to say opposite---textual styles. Sanskrit has sometimes been considered the archetypal philosophical language, for its easily formed substantives have encouraged a preponderance of abstract universals. Certainly Nagarjuna is a philosopher's philosopher, notorious for a laconic, knife-edged logic that wields distinctions that no one had noticed before and that many since have been unable to see the point of. In contrast, Chinese and Japanese both have a much more concrete flavor, with a preponderance of simile and metaphor. Dogen's major work, the Shobogenzo ..., written in his own very idiosyncratic Japanese, is as poetical and allusive as Nagarjuna's Mulamadhyamakakarika is dialectical and dry. Ddgen's text is mostly metaphor and Nagarjuna's has almost none. While Nagarjuna seems preoccupied with splitting what some see as conceptual hairs, Dogen is concerned with exploring the semantic possibilities of Buddhist texts to discover new meanings, willing and even eager to "misinterpret" certain passages to make his point.

What, then, can be gained from comparing them? My argument is, first, that Nagarjuna and Dogen nonetheless point to many of the same Buddhist insights because they deconstruct the same type of dualities, most of which may be understood as versions of our commonsense but delusive distinction between substance and attribute, subject and predicate. This will be demonstrated by analyzing the enigmatic chapter 2 (on motion and rest) of the Mulamadhyamakakarika and by examining Dogen's transgression of traditional Buddhist teachings in his Shobogenzo. The second part of this essay, however, is concerned with determining the limits of this similarity: for, although both texts work to undermine our dualistic ways of understanding ourselves "in" the world, they reach quite different conclusions about the possibility of language expressing a "true" understanding of the world---a disagreement that may reflect the different possibilities of their different languages.

What Does Nagarjuna Deconstruct?

[W]e do not only designate things with them [words and concepts], we think originally that through them we grasp the true in things. Through words and concepts we are still continually misled into imagining things as being simpler than they are, separate from one another, indivisible, each existing in and for itself. A philosophical mythology lies concealed in language which breaks out again every moment, however careful one may be otherwise. (Nietzsche)2

Few if any Buddhist scholars would dispute that Nagarjuna (second century C.E.?) is the most important Buddhist philosopher, and none of them would deny that the Mulamadhyamakakarika is his most important work. It is something of a scandal, then, that the basic meaning of this difficult text remains so obscure. This is not for want of interpreters-no Buddhist thinker has received more attention-yet there is little agreement among his Western expositors. It is curious, and more than a little suspicious, that Nagarjuna usually ends up expounding something quite similar to one's own favorite philosopher or philosophy: Shayer's Hegel, Stcherbatsky's Kant, Murti's Vedanta, Gudmundsen's Wittgenstein, Magliola's Derrida, Kalupahana's empiricism and pragmatism, and so forth. Does this mean that the Mulamadhyamakakarika is too foreign to our usual ways of understanding the world to be understood on its own terms?

The basic problem is not the nature of Nagarjuna's arguments themselves but their target; for, despite (or because of) the various opinions of traditional and contemporary commentators on this matter, it remains unclear from Nagarjuna's texts precisely what or whom he is criticizing. Since we have no other reliable access to Nagarjuna's intentions, this is an issue that may never be settled. From a postmodern perspective, the opportunity this ambiguity provides is not entirely negative, but then the onus falls upon each interpreter not only to offer a plausible account of Nagarjuna's motives but also to justify the continued importance of those motives for us.

Recently, for example, David Kalupahana made a strong case that the opponent in chapter 2 is the atomic theory shared by the substantialist Sarvastivadins and the momentarist Sautrantikas.3 This may well be true, yet that by itself does not go far enough to explain the significance of Nagarjuna's arguments today: for why should we be concerned about metaphysical debates between obscure Buddhist schools that thrived two thousand years ago?

The significance of those philosophical views increases for us, though, if they are attempts to resolve an inconsistency that plagues our ordinary "commonsense" way of understanding the world. If this is true, however, it may not be necessary or even worth our while to devote time and energy expounding those particular metaphysical systems; it may be more useful for us to turn immediately to that commonsense understanding and address its supposed aporia more directly. Accordingly, the target of this essay is not any developed philosophical position (such as the atomic theory of Abhidharma Buddhism ) but the more basic difficulties that plague our usual commonsense distinction between (what philosophers call) substance and attribute-which, Nietzsche would argue, may be traced back to our linguistic distinction between subject and predicate. In chapter 2, Nagarjuna attacks this distinction in terms of the duality we ordinarily make between a goer and his or her going.

By any standards, "the analysis of going and coming" is a peculiar and difficult text. Following the first chapter, which demonstrates our inability to understand the relationship between things and their causal relations, chapter 2 is evidently meant to exemplify the general argument presented earlier, by offering a more concrete instance of Nagarjuna's deconstructive approach to the relationship between things (in this case, movers) and their predicates/attributes (moving). In the process, however, Nagarjuna seems to engage in a kind of logic-chopping that is difficult to follow and whose import is unclear: exactly what is it that is being deconstructed? This chapter seems to exemplify Frederick Streng's objection to Nagarjuna's method, that it is "an analysis which appears to be rather arid and often simply a play on words."4 L. Stafford Betty points particularly to the reification of gamana ("act of going"): since the term is "empirically meaningless" and we do not need to grant that there is any such "thing" in the empirical world as a bare "act of going" without a goer, the argument fails.5 Yet isn't this looking in the wrong place? The Kirikas do not offer an analysis of the world itself but analyze our ways of understanding the world. It is these ways of thinking (which, according to Nagarjuna, are inconsistent) that make the world "empirical" for us. If so, we should look for a gamana in our categories of thought, and there we find it in our ingrained tendency (perhaps due to, and certainly enshrined in, the subject-predicate nature of language) to distinguish our experience into self-existing entities and their activities. We do think of ourselves, for example, as persons distinguishable from our actions, and this implies some sort of reification not only of ourselves but also of the act, as our substantives 'act', 'action', and 'activity' reveal.

The test of this approach is the light it can shed on the chapter, the whole of which may be summarized as follows.

Verses 1-7. Where does motion occur? Obviously not over the already-goneover, and not over the not-yet-gone-over, but it cannot be on the being-gone-over, because that would imply two movers: that there is a being-gone-over distinct from the goer that goes over it.

Verses 8-11. Who is going? We can't say "the goer is going" because that would imply two goers: that the goer is a goer even without going.

Verses 12-14. Where does going begin? Not on the already-gone-over, and not on the being-gone-over (in which case the going must already have begun). But it could not begin on the yet-to-be-gone-over (for beginning there would make it being-gone-over).

Verses 15-17. Similar arguments are made about coming-to-rest (becoming stationary): who comes to rest? Neither a goer (that would be a contradiction) nor a non-goer (who cannot become stationary). And where does coming-to-rest occur? Not on the already-gone-over, and not on the not-yet-gone-over, and it cannot happen on the being-gone-over (which would be a contradiction). So our usual understandings of going, beginning-to-go, and coming-to-rest have similar problems.

Verses 18-21. It doesn't make sense to say that the goer is going, for then we could not distinguish (as we normally do) between the agent and the action. But neither can the goer be different from the going, for then each could exist without the other. In short, describing what happens in terms of some relationship between a goer and its going is unintelligible.

Verses 22-25. (In summary:) A goer doesn't exist before going, for that would imply two goings. A goer cannot go on the three places of going (mentioned above), a non-goer cannot go on them, nor can someone who both is and is not a goer (a contradiction) go on them. "Therefore going, goer, and place of going do not exist."

Perhaps we can understand why some consider the arguments above to be a "logical sleight-of-hand" that "resembles the shell game"6-but such a conclusion nonetheless misses the point. The import of the arguments above is that our usual way of understanding motion-which distinguishes the goer from the going and from the place of going-does not really make sense when examined carefully, for the interdependence of the three shows that each is unreal when considered apart from the others. Nagarjuna's logic here (and in many other chapters) proceeds by demonstrating that once we have thus distinguished them-as ordinary language and "common sense" do-then it becomes impossible to understand their relation-a difficulty familiar enough to students of the mind-body problem. As Candrakirti points out in his commentary to verse 23, the same argument also refutes our usual notions that a speaker speaks something and that an agent performs an action (the latter dualism being the topic of chapter 9). Very similar arguments are employed in chapters 4, 5, and 8 to deconstruct our usual understanding of a perceiver perceiving a perceptual object; in chapter 6 to deconstruct the duality between persons and their affections; and in chapter 5 to deconstruct the duality in its most general terms, between things and their attributes.

In chapter 2, perhaps we see the problem most clearly by inquiring into the status of that-which-moves: in itself, is it a mover or not? Neither answer makes sense. For a mover to then be moving would be redundant ("a second motion"), and a non-mover moving is a contradiction. In contemporary analytic terms, we might say that Nagarjuna is pointing out a flaw in the ordinary language we use in describing (and hence in our ways of thinking about) motion and rest: our ascription of motion predicates to substantive objects is actually unintelligible. In everyday life we constantly fudge this, sometimes assuming that things exist apart from their predicates and at other times identifying things with their predicates (a good example is the relationship between me and "my" body). Nagarjuna's dialectics demonstrates this inconsistency simply by distinguishing clearly between the possibilities. It may be that this tendency to distinguish substance from attribute reflects the inherent dualism of language: a statement predicates something about something, for learning a language is learning what things there are (nouns correspond to things) and what these things do (verbs correspond to actions and processes) or have (adjectives correspond to attributes). But that such a dualism is widespread and even in a certain sense necessary (the "lower truth") does not make it a correct description of the way things really are ("the higher truth"), according to Nagarjuna.

This helps us to understand the point of the general Madhyamika critique, by revealing what is being criticized: our usual, commonsense understanding of the world, which sees it as a collection of discrete entities (one of them myself) interacting causally "in" space and time. "Nagarjuna's rampage through the notions of the philosophers is directed at uncovering their ultimate nonsense with a view to releasing men from humiliating bondage to them."7 Yet Nagarjuna attacks more than the philosophical fancies of Indian metaphysicians, for there is a metaphysics, although an inconsistent one, inherent in our everyday view-most personally and painfully in the contradiction between my sense of myself as something nontemporal and unchanging (i.e., as distinct from my attributes, such as body) and the awareness that I am growing older and subject to death (indistinguishable from attributes such as "my" body). It is one or another aspect of this dualistic view that is made absolute in systematic metaphysics. This commonsense understanding is what makes the world samsara for us, and it is this samsara that Nagarjuna is concerned with deconstructing.

It is a consequence of our taken-for-granted distinction between things and their attributes that I now perceive the room I am writing in not nondually, but as a collection of books and chairs and pens and paper-and me-each of which is unreflectively taken to be distinct from all the others and to persist unchanged until affected by something else. The causal relation (Nagarjuna's primary example of an attribute) is what we use to explain the interaction among things that are distinct from each other. If causality explains the interaction between things, then these things in themselves must be noncausal, and, by no coincidence, this is precisely our commonsense notion of what an object is: a thing whose continued existence does not need to be explained, for once created it "self-exists." The objectivity of the world (including the "subjectification" of myself as a thing in it but apart from it) depends upon this dualism between things and their attributes/causal relations. This constitutes samsara because it is by hypostatizing such a "thingness" out of the flux of experience that we become attached to things-again, the primal attachment being (to) the sense of self. Yet what we experience as such self-existing objects (svabhava) are thought-constructed reifications, a shorthand way of remembering that our perceptions tend to have a certain stability, which allows us to relate them together and form expectations. This may be a necessary habit for us (which is why it is a lower truth), but such reifications create a delusive bifurcation between objects and their attributes (which is why it is a lower truth).

This point about the way we perceive the world is important because without it one might conclude that Nagarjuna's critique of self-existence svabhava is a refutation of something no one believes in anyway. One does not escape his critique by defining entities in a more commonsense fashion as coming into and passing out of existence. The logic of the Karikas demonstrates that there is no tenable middle ground between self-existence independent of all conditions-an empty set-and the complete conditionality of sunya phenomena. Nagarjuna's arguments against self-existence show the inconsistency in our everyday, taken-for-granted way of "taking" the world: while we accept that things change, we also assume that they remain the same-both of which conditions are necessary if they are to be things that have causal relations. Recognizing this inconsistency, previous Indian philosophers tried to resolve it by making one of these two aspects absolute at the price of the other. But the satkaryavada substance-view of Advaita and Sankhya emphasizes permanence at the price of not being able to account for change, while the asatkaryavada modal-view of Sautrantika Buddhism has the opposite problem of not being able to provide the connecting thread necessary for continuity. Chapter 1 of the Karikas argues, in effect, that any understanding of cause-and-effect that tries to connect these two separate things can be reduced to the contradiction of both asserting and denying identity. Nagarjuna concludes that their "relationship" is incomprehensible and therefore, from the highest point of view, unreal.

In sum, there is something confused and deluded about our ordinary understanding of the world, because it dualizes substance from attribute, subject from predicate, permanence from change. Instead of attempting to supply the "correct view," however, the Madhyamika simply deconstructs this commonsense understanding, a removal which allows something else-obvious but hitherto overlooked-to manifest.

With the benefit of hindsight, however, we can notice that Nagarjuna's critique of such dualisms itself generates another dualism, one that during the following millennium would become increasingly problematical: that between language and silence. This dualism became so important because it reflects an essential and perhaps inescapable dualism at the heart of Buddhism : between delusion (of which language is a vehicle) and enlightenment (to which silence is believed to point).

Nagarjuna, of course, is very sensitive to the dualism of samsara and nirvana, and its deconstruction in chapter 25 forms the climax of the Karikas: there is not the slightest difference between them, for the limits (kotih) of the one are the limits of the other (verses 19-20); that which arises and passes away (i.e., samsara), when taken noncausally and without dependence, is nirvana (verse 9). Its beatitude (sivah) is the coming-to-rest of all ways of taking things (sarvopalambhopasamah), the repose of named things (prapancopasamah), which is why no truth has ever been taught by any Buddha to anyone anywhere (verse 24).

The problem, however, is that this solution to the dualism of delusion and enlightenment resolves the tension between them only by displacing it onto another dualism between the manifold world of named things (prapanca) and its coming-torest in silence (prapancopasamah). If nirvana involves realizing the sunyata of sarisara, for Nagarjuna that " emptiness " involves the cessation of thought-construction. Some translations of 25 :24 de-emphasize this cessation,8 but many other passages in the Karikas leave no doubt as to Nagarjuna's perspective on this matter: from the ultimate point of view no predication is possible. The dedicatory verses that begin the Karikas also emphasize that prapancopasamah is the way things truly are (pratttyasamutpida), a claim echoed in 18: 9 (where tattva "suchness" is characterized by lack of mental fabrication), in 13 :8 (sunyata is sarvadrstinam prokta nihsaranama "the relinquishing of all views"), and again even in the final verse of the Karikas, 26:30, where the author bows to Gautama, whose compassion "taught the true doctrine which leads to the relinquishing of all views."9

Nagarjuna is well aware of the tension intrinsic to the claim that the true characterization of the nature of things is that things cannot be conceptually characterized. His solution, of course, is the two-truths doctrine. All predication is part of the lower truth. Candrak-rti's commentary on 13:8 quotes the Ratnakuta Sutra to make the point that sunyata is a medicine that must itself be expelled in order for the patient to recover fully. Since sunyata is itself sunya, one uses that lower truth to climb up a ladder that, finally, is kicked away. The Wittgenstein analogy is appropriate because Nagarjuna would also agree with the conclusion to the Tractatus: "7. What we cannot speak about we must pass over in silence."10 For the Buddhist tradition as it developed thereafter, however, this solution to the dualism of samsara and nirvana left a legacy that bifurcated too sharply between the lower and the higher truths, between means and ends, between thought/language and the peace that surpasses understanding. In the centuries that followed, these bifurcations reappeared in various doctrinal forms, especially in East Asian controversies about our "Buddha-nature." Significantly, the crux of these debates may also be expressed in terms of substance and attribute, subject and predicate: is enlightened mind intrinsic or adventitious, something we already have or something we need to gain?

By no coincidence, this is precisely the issue in the dialectic between original enlightenment (hongaku) and acquired enlightenment (shikaku) that is said to have preoccupied the young Do gen: if we are endowed with the Dharma-nature by birth, why did all the Buddhas strive for enlightenment by engaging in spiritual practice? Hongaku seems to encourage a self-satisfied quietism complacent in its delusions, shikaku a self-stultifying split between means and ends, as we strive to become what we are. We shall see that Dogen's solution to this dilemma not only transformed the understanding of the relationship between practice and enlightenment; it also led to a radically new appreciation of how language can combat its own mystifications.

What Does Dogen Deconstruct?

Language and symbols circumscribe; but, as living forces, they are dynamic enough to open up, constantly re-expressing, renewing, and casting-off, so as to unfold new horizons of their own life. In this way language and symbols know no limits with respect to how far they can penetrate both conceptually and symbolically. No Buddhist thinker was more intensely and meticulously involved with the exploration of each and every linguistic possibility of Buddhist concepts and symbols-even those forgotten, displaced onesthan Dogen who endeavored to appropriate them in the dynamic workings of the Way's realization. (Hee-Jin Kim)11

Nagarjuna's dialectical arguments are foreign to Dbgen. In fact, the Shobogenzo is interested not in Buddhist philosophy as such, but in semantic analysis of passages from Buddhist Sutras and Ch'an texts. Such analyses are not inspired by any conventional piety toward such scriptures, for Dogen offers many deliberate, and often brilliant, "misinterpretations" of these passages. By his readiness to transgress the traditional readings and contradict orthodox teachings, Dogen is able to challenge our usual understanding and generate a new way of "taking" the world freed from our usual linguistic dualisms, including conventional Buddhist ones such as that between language and silence.

Hee-Jin Kim's exegesis of Dogen's analytical methods distinguishes seven different techniques in the Shobogenzo.12 Although these overlap and are not exhaustive, we begin by summarizing what Kim says about how each of these functions, followed by an attempt to understand what these techniques imply about language and how language can be utilized from the enlightened point of view. Below are only a few of the many examples that could be cited for each technique.

Transposition of Lexical Components

A simple example is Dogen's discussion of to-higan ... ("reaching the other shore") in the Bukkyo ... fascicle, which transposes the two characters into higanto ..., "the other shore's arrival" or "the other shore has arrived." The original meaning of higan ("the other shore," i.e., nirvana) dualizes between a future event and one's present practice aimed at attaining that event. The transcribed term no longer refers to a future event but emphasizes the event of realization right here and now.

In the Mujo-seppo ... fascicle, seppo ... "preaching the dharma" is reversed in the same way to become ho-setsu tS "the dharma's preaching." This allows Dogen to say: "This 'discourse on the Dharma' is 'the Dharma's discourse."' There is no duality (trinity?) between the speaker, the speaking, and the Dharma that is spoken about.

Semantic Reconstruction through Syntactic Change

Perhaps the best-known example is in the Bussho ... fascicle, which quotes from the Nirvana Sutra: "All sentient beings without exception have Buddha-nature" (issai no shujo wa kotogotoku bussh6 ari). Dogen rearranges the syntactical components to make them mean: all sentient beings, that is, all existence, is Buddha-nature (issai shujo shitsuu-bussho ...). Buddha-nature is no longer an attribute of sentient beings, something that needs to be actualized. Sentient beings and "their" Buddha-nature are nondual.

Another example of this reconstruction, to the same end, occurs in the Juki ... fascicle. Juki refers to the Buddha's prediction of a disciple's future enlightenment, but Dogen refigures the phrase masani anokutara-sammyaku-sambodai o ubeshi ("They shall attain supreme, perfect enlightenment") into totokuanokutara-sammyakusambodai ... ("They have certainly attained supreme, perfect enlightenment"). The assurance of a future event is transformed into testimony to a present condition.

Explication of Semantic Attributes In the Uji WW fascicle, Dogen takes the common term arutoki ("at a certain time," "sometimes," "once") and reinterprets its components aru or u ("to be") and toki or ji ("time," "occasion") to make uji, "being-time," which he uses to signify the nonduality of existence and time, that is, things and their temporal attributes. In other fascicles Dogen makes the same point by reducing each of these two concepts to the other, saying that objects are time (objects have no self-existence because they are necessarily temporal, in which case they are not objects in the usual sense) and, conversely, that time is objects (time manifests itself not in but as the ephemera we call objects, in which case time is different from what it is usually understood to mean). "The time we call spring blossoms directly as an existence called flowers. The flowers, in turn, express the time called spring. This is not existence within time; existence itself is time."'3 If there are no nouns, there are no referents for temporal predicates. When there are no things that have an existence apart from time, then it makes no sense to speak of things as being young or old. Nagarjuna had drawn the same conclusion: "Becoming other is not comprehensible either of the same thing [for then it is not the same thing] or of another thing [for then it is not the same thing]. So the young man does not grow old nor does the old man grow old."14

In a famous passage in the first fascicle of the Shobogenzo, "Genjo-koan" JA , the image of firewood and ashes is used to make the same point about things and "their" time:

Firewood becomes ash, and it does not become firewood again. Yet, do not suppose that the ash is future and the firewood past. You should understand that firewood abides in the phenomenal expression of firewood, which fully includes past and future and is independent of past and future. Ash abides in the phenomenal expression of ash, which fully includes future and past. Just as firewood does not become firewood again after it is ash, you do not return to birth after death.

This being so, it is an established way in buddha-dharma to deny that birth turns into death. Accordingly, birth is understood as no-birth. It is an unshakeable teaching in Buddha's discourse that death does not turn into birth. Accordingly, death is understood as no-death.

Birth is an expression complete this moment. Death is an expression complete this moment. They are like winter and spring. You do not call winter the beginning of spring, nor summer the end of spring.5

The beginning seems to echo Nagarjuna's deconstruction of the duality between fire and fuel in chapter 10 of the Mulamadhyamakakariki, but Dogen's explication brings the issue home more directly to our own lives. Because life and death, like spring and summer, are not in time, they are timeless. And if there is no one nontemporal who is born and dies, then there are only the events of birth and death. But if there are only those events, with no one in them, then there is no real birth and death. such is the consequence of the nonduality between me and that most uncomfortable attribute of all, "my" birth/death.

Reflexive, Self-causative Utterances

Dogen uses repetition (ji-ji ... "time," sho-sho ... "birth," butsu-butsu ... "buddha," etc.) and identity statements ("mountains are mountains" and " emptiness is emptiness ") for emphasis, and, taking advantage of the facility with which the Japanese language allows nouns to become verbs by adding the suffix -su, he delights in such Heideggerian-type expressions as "the sky skies the sky." These techniques are used to exemplify his notion of ippo-gujin --..., "the total exertion of a single dharma." This key term embodies his dynamic understanding of interpenetration, according to which each dharma in the universe is both the cause and effect of all other dharmas. This interfusion means that the life of one dharma becomes the life of all dharmas, so that (as Zen masters like to say), this is the only thing in the whole universe. The application of ippo-gujin to language allows words, too, to transcend dualism, as we shall see.

The Upgrading of Commonplace Notions and the Use of Neglected Metaphors

By Dogen's time, a number of metaphors had become traditional as ways to contrast this world of suffering with the realm of enlightenment: for example, gabyo ... (pictured cakes, which cannot satisfy our hunger), kuge ... (literally, sky-flowers, seen when the eye is defective, and hence a metaphor for illusory perceptions), katt ... (entangling vines, meaning worldly attachments), and mu ... (a dream, as opposed to being awake). In this way, too, Buddhist teachings that work to deconstruct dualisms created new ones, and in the thousand years between Nagarjuna and Dogen these images had ossified to become more problematical. Here, too, Dogen's "misinterpretations" revitalize these depreciated terms by denying the dualism implicit in each. Instead of dismissing pictures (i.e., concepts), the Gabyo fascicle emphasizes their importance by transforming gabyo wa ue ni mitazu ("pictured cakes do not satisfy hunger") into gabyo ... ("pictured cakes are no-satisfaction-hunger"), escaping the dualism of hunger and satisfaction into the nondualism of a hunger that, because it is itself ultimate reality, lacks nothing: "Because the entire world and all dharmas are unequivocally pictures, men and dharmas are actualized through pictures, and the buddhas and patriarchs are perfected through pictures."

The Kuge fascicle revalorizes kuge, usually castigated as illusions, into "flowers of emptiness "; in place of the typical Buddhist duality between reality and delusion, "all dharmas of the universe are the flowers of emptiness ." Instead of the usual admonition to cut off all entangling vines, the Katto fascicle emphasizes the importance of worldly relationships such as the dharmic connection between teacher and student, which leads to ever-increasing understanding of the Dharma. And "all dharmas in the dream state as well as in the waking state are equally ultimate reality.... Dream and waking are equally ultimate reality: no largeness or smallness, no superiority or inferiority has anything to do with them."16

The Use of Homophonous Expressions

In addition to employing associative techniques such as interweaving shozan ... "all the mountains" with shosui ... "all the waters" to vividly present the nonduality of mountain and water in the Sansuikyo ... fascicle, Dogen uses homophonous word pairs-puns-to reinforce his meaning. In the Gabyo fascicle, for example, the phrase shobutsu kore sho naru yueni shobutsu kore sho nari ...) ("Because all the Buddhas are verification, all things are verification") identifies shobutsu "all the Buddhas" with shobutsu "all things."

Reinterpretation Based on the Principle of Absolute Emptiness

Dogen "misinterprets" some of the most famous Zen stories to give them a radically different meaning-often one diametrically opposed to the traditional understanding. In the Katto fascicle, for example, Dogen challenges the traditional view of Bodhidharma's dharma transmission to his four disciples Tao-fu, Tsung-chih, Tao-yu, and Hui-k'o. According to their different responses to his challenge, Bodhidharma says that they have attained his skin, flesh, bones, and marrow, respectively-the last because Hui-k'o demonstrates the highest attainment by saying nothing at all. So it is, at least according to the usual view that sees these four attainments as metaphors for progressively deeper stages of understanding, indicating a hierarchy of rank among the disciples. Dogen, however, repudiates this common view by adopting the absolute point of view:

We should know that the patriarch's saying "skin, flesh, bones, and marrow" has no bearing on shallowness or deepness.... The patriarch's body-mind is such that the skin, flesh, bones, and marrow are all equally the patriarch himself: the marrow is not the deepest, the skin is not shallowest.17

Kim cites many other instances to demonstrate these "transgressive" techniques, but what we need to do now is characterize their function. Two points stand out.

First, Dogen is doing more than twisting traditional texts to make them say whatever he wants them to mean. In the examples above, he is using the freedom of a poet to conflate a problematic dualism, that is, a deluded way of thinking that causes problems for us; and, despite the fact that this literary approach to language is so different from Nagarjuna's dialectical one, in each case there is a parallel with deconstructions in the Mulamadhyamakakarika. For example, ho-setsu denies any duality between the one who preaches the dharma and the dharma that is taught, even as many chapters of the Karikas challenge the duality between an agent and his or her action. Uji denies any duality between beings and their temporality, between springtime and its flowers, between us and our birth/death; this parallels Nagarjuna's deconstruction of the difference between time and things in chapters 19 and 13. The Bussho fascicle denies the duality between sentient beings and their Buddha-nature, which may be seen as another instance of Nagarjuna's repeated attack on the duality between things and their attributes. Higan-to (like many other reconstructions) denies the usual duality between practice and realization (means and ends), just as Nagarjuna's nirvana chapter deconstructs the usual Buddhist duality between sarm sara and nirvana.

In each case Dogen, like Nagarjuna, does not allow himself to be limited by the usual dualisms of our language, and of our thought. While Nagarjuna's dialectic exposes the unintelligibility of these dualisms by showing how we cannot relate the two terms back together, Dogen exploits the different resources of the Japanese language to concoct expressions that leap out of the bifurcations we get stuck in. For both thinkers, however, these deconstructions may be understood as conflations of various recurrences of the subject-predicate dualism: nirvana is not something I can attain; the dharma is not something I can preach; Buddha-nature is not something I have (or do not have); "my" time is not something distinguishable from me. This is all the more striking because, although Dogen sometimes refers to Nagarjuna (Jpn: Ryuju), these references are largely confined to quotations and passages from various Chinese collections, and so far as I know they do not reveal any familiarity with the arguments in primary texts such as the Mulamadhyamakakarika.

However, this basic similarity also serves to highlight the differences between them. Part of this difference is emphasis, a shift in focus necessary to respond to the historical development of Buddhist teachings in the thousand years between thema development due in no small part to Nagarjuna's enormous influence. As we have seen, the dualisms that most preoccupy Dogen are versions of the practice/enlightenment-means/ends bifurcation. Granted, nirvana is not something that can be attained, but it still needs to be realized, and by his time many traditional Ch'an/Zen stories and metaphors designed to encourage this process had themselves become more problematical than helpful, in his view.

Dogen's revaluation of commonplace Buddhist metaphors in particular leaves us no doubt about his understanding of language-which is where the difference of emphasis between Nagarjuna and Dogen becomes a more significant difference of perspective. Concepts, metaphors, parables, and so forth are not just instrumental, convenient means to communicate truth, for they themselves manifest the truth-or rather, since that is still too dualistic, they themselves are the truth that we need to realize. "Metaphor in Dogen's sense is not that which points to something other than itself, but that in which something realizes itself," summarizes Kim. "In short, the symbol is not a means to edification but an end in itself-the workings of ultimate truth."18 As Dbgen himself puts it in the Muchu-setsumu ... fascicle: "The Buddha-dharma, even if it is a metaphor, is ultimate reality." If I do not try to get some graspable truth from the metaphor, it can be a way my mind consummates itself: although symbols can be redeemed only by mind, the mind does not function in a vacuum but is activated by-or as-symbols.

In the Sansuikyo fascicle, Dogen criticizes those who have only an instrumentalist view of language and who think that koans are simply nonsensical ways to cut off thought: "How pitiable are they who are unaware that discriminating thought is words and phrases, and that words and phrases liberate discriminating thought." What a challenge to the traditional Buddhist dualism between language and reality: the goal is not to eliminate concepts but to liberate them! Despite their problematical aspects, "words are not essentially different from things, events, or beings-all 'alive' in Dogen's thought."19

In an important essay on language in the Ch'an/Zen experience, Dale Wright has argued that such awakening is not from language but to language. As in Gadamer's hermeneutics, language is less an obstructing barrier than a reservoir of possibilities becoming available to those not trapped within its dualistic categories, not a clothing that hides truth but a medium that manifests it-in short, not a veil but a window. "Far from being a transcendence of language," concludes Wright, "this process would consist in a fundamental reorientation within language [that] would require training to a level of fluency in distinctive, nonobjectifying, rhetorical practices."20

Within the Buddhist tradition, this move from transcendence of language to reorientation within it is perhaps best exemplified by the difference between Nagarjuna and Dogen. The latter shows us that words and metaphors can be understood not just as instrumentally trying to grasp and convey truth (and therefore dualistically interfering with our realization of some truth that transcends words), but as being the truth-that is, as being one of the many ways that Buddha-nature is. To the many dualisms that Nagarjuna deconstructs, then, Dogen explicitly adds one more: he denies the dualism between language and the world. If we are the ones who dualize, why blame the victims? A birdsong, a temple bell ringing, a flower blooming, and Dogen's transpositions, too, blossoming for us as we read them: if we do not dualize between world and word, then we can experience the Buddha-dharma-our own "empty" nature-presencing and playing in each.21

A Scheme We Cannot Throw Off?

Now we read disharmonies and problems into things because we think only in the form of language-and thus believe in the "eternal truth" of "reason" (e.g., subject, attribute, etc.).... Rational thought is interpretation according to a scheme that we cannot throw off. (Nietzsche)22

Both Buddhist thinkers exploit the very different strengths of their respective languages. The complex syntax of Nagarjuna's sophisticated Sanskrit permits precise and terse philosophical analysis. The looser syntax of Dbgen's Japanese, due to the greater flexibility and ambiguity of its Chinese ideographs, allows a poetic allusiveness that lends itself to his semantic transpositions. We have seen that this difference is further reflected in their respective attitudes toward language: to Nagarjuna it seems to be fundamentally problematical, for he limits himself to employing it negatively, solely to deconstruct the dualities that are delusive (from the higher point of view) although necessary in daily life (from the lower point of view). In contrast, Dogen views and uses language more positively, by emphasizing the innovative possibilities that Chinese and Japanese encourage but Nagarjuna's philosophical Sanskrit apparently did not.

I wonder how much the languages themselves contribute to this difference. Do Nagarjuna's and Dogen's different approaches perhaps reflect different "mental spaces" created in employing the different types of script? The meaning of an alphabetic script is derivative (or representational) because it converts letters into sounds, while Chinese and Japanese ideographs express their meaning more directly, without speech. How such a non-oral/aural meaning could arise is suggested by the peculiar origin of Chinese characters. According to Simon Leys, the earliest Chinese inscriptions "did not record language, but meanings-directly, and speechlessly: they transcended language."

This Chinese emblematic meta-language developed independently from contemporary speech. For convenience, however, the written characters were progressively given conventional sounds; thus, eventually the inscriptions did not merely convey silent meanings, they could also be read aloud. In the end, they themselves generated a languagemonosyllabic and non-inflected (features that remain as the special marks of its artificial origin)-and since this language carried all the prestige of magic and power, it gradually supplanted the vernacular originally spoken.23

Perhaps an alphabetic script is more likely to suggest a representational understanding of meaning and truth: as letters represent sounds, so words re-present things, implying that language is something superimposed on the world. In contrast, an ideographic script seems to de-emphasize such a duality between thought and words, between meaning and reality, encouraging instead the view that thought is (part of) reality.

Finally, however, what was more important for Buddhism is that the very different resources of these different languages-Nagarjuna's alphabetic Sanskrit and Dogen's ideographic Japanese-could be tapped for the same end: deconstructing the dualisms implicit in our usual ways of "taking" the world, most of them variations of the fundamental one between subject and predicate, substance and attribute. By dividing up the world into things and their relations, and most of all by distinguishing my sense-of-self from the world I live and act "in," I overlook something important about the actual nature of that world.

This parallel suggests that Nietzsche was wrong when he reflected that "rational thought is interpretation according to a scheme that we cannot throw off." Nagarjuna and Dogen both demonstrate, in their different ways, that language at its best can work against its own mystifications. However, neither of them believed that such conceptual deconstructions are, in themselves, sufficient to escape the disease that plagues us insofar as we feel separate from the world (from our bodies, our actions, our death). Both took for granted a religious context that provided the situation for their philosophical enterprises, a rich heritage of ethical and meditative practices provided by the Buddhist tradition to help us transform our mode of experiencing the world. They knew that the most important deconstruction extends beyond language to deconstruct the delusive duality between my sense-of-self and the world.24

Notes

1 - Friedrich Nietzsche, Twilight of the Idols [1889], trans. R. J. Hollingdale (Harmondsworth: Penguin Classic, 1969), p. 38; his own emphasis here and elsewhere.
2 - Friedrich Nietzsche, The Wanderer and his Shadow [1880], in Human, All Too Human, trans. R. J. Hollingdale (Cambridge: Cambridge University Press, 1986), p. 306.
3 - David J. Kalupahana, Mulamadhyamakakarika of Nagarjuna: The Philosophy of the Middle Way (Delhi: Motilal Banarsidass, 1991), Introduction, pp. 35-36.
4 - Frederick J. Streng, Emptiness : A Study in Religious Meaning (Nashville, Tennessee: Abingdon, 1967), pp.181-182.
5 - L. Stafford Betty, "Nagarjuna's Masterpiece-Logical, Mystical, Both, or Neither?" Philosophy East and West 33 (2) (April 1983): 125-126.
6 - Ibid., p. 135.
7 - Mervyn Sprung, in the introduction to his translation of Lucid Exposition of the Middle Way (Boulder, Colorado: Prajna Press, 1979), Candrakirti's classical commentary on the Mulamadhyamakakarika, p. 6. My references to the original Sanskrit text are from J. E. de)ong's edition (Madras: Adyar Library and Research Centre, 1977).
8 - For example, Kalupahana translates the first half of 25 :24 as "The appeasement of all objects, the appeasement of obsession" (p. 369), and Jay Garfield (in The Fundamental Wisdom of the Middle Way: Nagarjuna's Mulamadhyamakakiriki [ New York: Oxford University Press, 1995]) translates the same as "The pacification of all objectification/And the pacification of all illusion" (p. 76). But elsewhere (e.g., in his discussion on pp. 353-359), Garfield, too, emphasizes the cessation of predication and conceptual construction. My version of 25:24 follows Sprung (Lucid Exposition, p. 262). The basic difficulty is that, although important both in Pali Buddhism and in Madhyamaka, the meaning of prapanca is unclear and controversial. Etymology yields pra + panc, "spreading out" in the sense of manifoldness and ramification; it seems to refer to some indeterminate "interface" between perception and thought. In his book on the concept, Nanananda defines its primary meaning as "the tendency towards proliferation in the realm of concepts," but this loses the connection with perception. See his Concept and Reality in Early Buddhist Thought (Kandy, Sri Lanka: Buddhist Publication Society, 1971), pp. 3-4.
9 - See also 18:7; 22:11, 12, 15; and 24:18.
10 - In this context the final verse is often quoted, yet the previous one is just as relevant as an elucidation of the Madhyamaka position: "6.54. My propositions serve as elucidations in the following way: anyone who understands me eventually recognizes them as nonsensical, when he has used them-as steps-to climb up beyond them. (He must, so to speak, throw away the ladder after he has climbed up it.) He must transcend these propositions, and then he will see the world aright" (Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. D. F. Pears and B. F. McGuinness [London: Routledge and Kegan Paul, 1961 ], p. 151).
11 - Hee-Jin Kim, "Method and Realization: Dogen's Use of the Koan Language" (unpublished paper presented at a conference on "The Significance of Dogen," Tassajara Zen Mountain Center, 8-11 October 1981), p. 9.
12 - For more examples, see Hee-)in Kim, "'The Reason of Words and Letters': Dogen and Koan Language," in William R. LaFleur, ed., DC*gen Studies (Honolulu: University of Hawai'i Press, 1985), pp. 54-82. The typology is on pp. 61-78. My analysis of Dogen's deconstructive techniques largely follows this essay, with the exception of the discussion of time (in technique number 3), which offers my own understanding of Dogen's view of time. 13 - As quoted in Masunaga Reiho, The Soto Approach to Zen ( Tokyo: Layman Buddhist Society Press, 1958), p. 68.
14 - Mulamadhyamakakarika XIII.5, in Sprung, Lucid Exposition, p. 147.
15 - "Genjo-koan," trans. Dan Welch and Kazuaki Tanahashi, in Moon in a Dewdrop: Writings of Zen Master Dogen, ed. Kazuaki Tanahashi (San Francisco: North Point Press, 1985), pp. 70-71.
16 - In the Much u-setsumu fascicle; see Kim, "The Reason of Words and Letters," pp. 71-72.
17 - As translated in Kim, "The Reason of Words and Letters," p. 75.
18 - Hee-Jin Kim, Dogen Kigen-Mystical Realist ( Tucson: University of Arizona Press, 1975), p.110.
19 - Kim, "The Reason of Words and Letters," pp. 57, 58. 20 - Dale Wright, "Rethinking Transcendence: The Role of Language in Zen Experience," Philosophy East and West 42 (1) (January 1992): 125, 131, 133. 21 - For a more detailed comparison between Dogen and Eckhart, with reference to Derridean deconstruction, see David Loy, "Dead Words, Living Words, and Healing Words: The Dissemination of Dogen and Eckhart," in David Loy, ed., Healing Deconstruction: Postmodern Thought in Buddhism and Christianity (Atlanta, Georgia: Scholars Press, 1996).
22 - Friedrich Nietzsche, The Will to Power, trans. Walter Kaufmann and R. J. Hollingdale (New York: Vintage, 1968), no. 522, p. 283. 23 - Simon Leys, "One More Art," New York Review of Books, 18 April 1996, pp. 29-30.
24 - For more on this type of deconstruction, see David Loy, Lack and Transcendence: The Problem of Death and Life in Psychotherapy, Existentialism, and Buddhism (Atlantic Highlands, New Jersey: Humanities Press, 1996).