|p. 83-99||Liberating oneself from the absolutized boundary of language: a liminological approach to the interplay of speech and silence in Chan Buddhism||Philosophy East & West, Vol. 51, No. 1 (2001)|
This essay takes a "liminology of language" approach to the Chan Buddhist view of language and its linguistic strategy. The advent of liminology in contemporary thought has been inspired by the works of philosophers and thinkers such as Derrida, Foucault, Blanchot, Heidegger, and the later Wittgenstein. The central idea of the liminology of language is the relativization of any limits of language. The justification for this relativizing is the revelation of the dynamic interrelationship between the two sides of the limits of language. The consequence is an accepting and exploratory linguistic strategy as play at the limits of language. These ideas form the framework for a liminological analysis that is applicable to different views of language and different linguistic strategies. For instance, a liminological analysis will allow us to see what is beneath the claim of linguistic inadequacy. It will allow us to see the interchangeability of two sides of the limits of language such as silence and speech. Different linguistic strategies are then possible due to the claim of linguistic inadequacy.
What is meant by liberation from the absolutized boundary of language, as proposed by the title of this essay? From a contemporary point of view, it encompasses the following meanings. First, it breaks down the myth of any immobile, absolutely uncrossable boundary of language. The early Wittgenstein states: "There is indeed the inexpressible.... Whereof one cannot speak, thereof one must be silent."  Silence thus marks the boundary of speaking, the limits of language. Within this boundary, Wittgenstein confines the expressible to the use of propositional language. This view, considered an absolutization of the limits of language, has been challenged by postmodern thinkers. For example, Blanchot points out: "What is inexpressible is inexpressible in relation to a certain system of expression."  "The inadequacy of language runs the risk of never being sufficiently inadequate," otherwise "we would all have been satisfied with silence long ago."  This peculiar trait of never being sufficiently inadequate is determined by the double structure of language. To use Foucault's expression, language is always both excessive and deficient. This deficiency, this lack of language, means not only the absence of what is to be signified but also the absence of a center. To speak is to bring this kind of absence into play. At the same time, it is excessive. Deficiency becomes the condition of the possibility of more speaking, more signifying, more language. "Language can no longer avoid multiplying itself." "It is always beyond the limit in relation to itself" and "is fated to extend itself to infinity without ever acquiring the weight that might immobilize it."  In this view, the notion of an absolutized boundary of language simply cancels the possibilities of language.
Second, liberation from the absolutized boundary of language not only stimulates us to take a fresh look at the issue of the limits of language, but further to rethink the distinction between the two sides of the limits of language. In Heidegger's existential analysis of discourse as the articulation of the intelligibility of Being-in-the-world, the boundary of language or the borderline between silence and speaking is in no way absolute or static. He asserts: "To keep silent does not mean to be dumb.... K eeping silent authentically is possible only in genuine discoursing."  Thus silence involves speaking, while speaking involves silence. The two sides are interrelated and interchangeable. Echoing Heidegger, Merleau-Ponty also states: "The absence of a sign can be a sign."  "True speech ... is only silence."  Despite their contextual differences, these statements all radically blur the distinction between speaking and non-speaking, speech and silence, and show insights into the mutual connection and transition between the two sides. These statements therefore provide legitimation for relativizing the limits of language. As a result, what is inexpressible or silent, namely the Other of language, is no longer conceived in an isolated, non-relational manner.
Third, the consequence of overturning fixed binary divisions and relativizing the limits of language is that once philosophical discourse is freed from a static, non-relational understanding of the limits and function of language, options for novel linguistic strategies are given their due. Therefore we see, in Heidegger, experiments in philosophical tautology, a turn toward poetic and evocative language, a crossing out of words to mark their self-erasure, and innovative terminology. In Derrida, we see the effort to develop indeterminable concepts that can no longer be enclosed in traditional oppositions, a preference for neither/nor, a playful and elusive style. In Kierkegaard, there arise the strategies of indirect communication, using metaphor, parable, irony, and so forth. All these strategies can be regarded as a kind of play at the limits of language. The primary aim is to eschew the trap of a propositional, logical, or descriptive language. Language is transformed through a kind of linguistic twisting. Language turns against itself, and is inscribed with self-erasure, in order to negotiate the limits of language.
These three aspects -- the problematization of any absolute limits of language, insight into the mutual connection between the two sides of the limits of language, and linguistic twisting as play at the limits of language -- are the major elements of a liminology of language as addressed by postmodern thinkers. Stemming from its Latin root, the term liminology puts much weight on the meaning of the threshold that connects, or makes a transition between, the two sides, rather than on the meaning of an absolute borderline. 
The liminology of language is useful as an interpretive tool. For example, our understanding and analysis of the Chan Buddhist view of language can benefit from the notion that the inexpressible is always inexpressible in relation to a certain system of expression. This allows us to see more clearly what Chan Buddhists mean by the inadequacy of language. Second, we find that Chan Buddhism attains the deabsolutization of the limits of language by completely penetrating the nonduality between speech and silence.  In blurring or invalidating an absolute distinction between speech and silence, and in perceiving a dynamic connection and transition between them, the Chan masters, who were no less radical than our contemporaries, displayed their great wisdom and insight. Third, Chan Buddhism demonstrates a marvelous interplay between speech and silence, a skillful performance of linguistic twisting, along with various effective strategies for coping with the limits of language, far beyond the scope of the Western imagination. All these aspects indicate that Chan Buddhism is a great resource for developing alternative strategies, and for relativizing the limits of language.
Hence, I will focus on the rediscovery of the Chan tradition in relation to the postmodern concern with the limits of language and how these limits should be dealt with. I will attempt to provide a better understanding of the Chan Buddhist view of language and its linguistic strategy, toward the end of my attempt to clarify the relation between the Chan attitude toward language and its creative use of language. Finally, I will demonstrate how the Chan masters liberated themselves from the absolutized boundary of language, from the conventional fixation on either speech or silence, and how they displayed the great function of language in their soteriological and therapeutic practice. To avoid reading postmodern discourse into Chan Buddhism, our study will be firmly based on an examination of the relevant Buddhist texts, especially the Chan Buddhist "recorded sayings." This examination will reveal how some important sayings of the Chan masters regarding their view of language have been neglected by modern interpreters of Chan Buddhism in both the West and the East.
A liminological approach to the contradictory phenomena of the so-called Chan denial of language and the great Chan art of speaking is incompatible with some dominant modern interpretations of the Chan/Zen view of language. For instance, echoing logical positivism, Fung Yu-lan interprets Chan thought as a philosophy of silence, which is not to say anything about the noumenon--the unknowable.  This makes the Chan view somewhat close to the early Wittgenstein's absolutization of the limits of language. It fails to appreciate the trans-metaphysical meaning of the Chan insight into a complete interplay between speaking and silence. Similarly, D. T. Suzuki sometimes assumes a dichotomy between silence and verbalism, placing Chan at one extremity of this dichotomy.  The dominant modern interpretations of the Chan/Zen view of language have inevitably given rise to various criticisms.
Dale Wright and Bernard Faure have provided pioneer works in the recent rethinking of the relationship between Chan/Zen and language.  Both criticize the assertion that Chan/Zen simply transcends or negates language. Both attempt a reconfiguration of the role of language in Chan/Zen and provide textual evidence to justify it. Generally speaking, my investigation of the Chan view of language can be regarded as part of this tendency to shape a better understanding of that view. However, Wright's approach relies heavily on reflections drawn from contemporary Western philosophy. Faure's study contains extensive historical-textual materials from Chan/Zen, but does not engage fully in a systematic, theoretical interpretation. Both fail to provide a more systematic reconstruction, or a more comprehensive and coherent understanding, of the Chan view of language. A critical examination or systematic analysis of Chan notions of language has yet to be achieved. It is to this kind of analysis, reconstruction, or understanding that, I believe, my liminological approach will eventually contribute.
Undeniably, the Chan attitude toward language has its doctrinal or ideological foundation in Mahayana Buddhism. For this reason, our inquiry into the Chan notion of the inadequacy of language must start with an examination of this foundation. A careful contextual analysis of the Mahayana discourse on the inadequacy of language will reveal what is meant by the inadequacy of language.
A very influential notion of the inadequacy of language in Mahayana Buddhism derives from Madhyamika philosophy. In the Mulamadhyamakakarika (hereafter Karika) 18:7, Nagarjuna states: "When the object of thought is no more there is nothing for language to refer to. The true nature of things neither arises nor perishes, as nirvana does not."  This statement has been considered foundational to the so-called Madhyamika negation of language.  As David Kalupahana correctly observes, it is possible for interpreters of Nagarjuna to assume that the "ineffable" is identical to sunyata, paramartha, or nirvana. Nevertheless, Kalupahana states that "the first line of Nagarjuna's statement should prevent anyone from reaching such a conclusion."  A scrutiny of Nagarjuna's entire work, and of Candrakirti's interpretation of it, shows that Nagarjuna specifically negates a cognitive, entitative, or descriptive language, as well as a correspondence relation between language and object.
According to Nagarjuna and Candrakirti, soteriological terms such as "devoid of self-existence," "not devoid of self-existence," and so on should no longer be asserted as predicates (even though they have the form of predicates).  These terms should no longer serve the purpose of describing subjects, simply because noun words such as nirvana, paramartha, or sunyata, in the Buddhist discourse, are not entities, and, therefore, are not the objects of any cognitive activities. Words, sentences, and speeches are, in fact, prescriptions for curing people's illnesses by merely recommending the attitude and behavior of non-clinging to things. Madhyamika philosophy provides great insight into the intrinsic link between descriptive, imputative language and cognitive reification. For Madhyamika, the world of ordinary language is the realm of naming and things named, knowing and things known, and so forth.  Language constitutes the world, insofar as the external world as object always means the object grasped in thought, and insofar as language always imputes specific attributes to the object of thought.  Referred to as prapa~nca, this entified world constructed through cognitive language is considered by Madhyamika to be the root of all kinds of metaphysical thinking and all forms of clinging and suffering. A primary goal of the Madhyamika teaching of emptiness is to bring prapa~nca to an end.  It is within this context that Nagarjuna's verse 18:7 can be properly understood. In this verse, Nagarjuna asserts that when a referent no longer exists, referential language is no longer adequate. Elsewhere he also suggests, as we have seen, that when language no longer functions by reference to objects, it ceases to name things such that it no longer serves discursive thought or makes cognitive assertions. Both point to the inadequacy of the referential, cognitive function of language -- the predominant use of language.
The inadequacy of language is also addressed by many Mahayana sutras. Despite various negative statements about language in divergent sutras, a close examination often discloses that the alleged inadequacy of language is always related to a certain mode of speaking, or a particular function of language. Among these Mahayana scriptures, the Lankavatara Sutra, well-known for its connection with the Tathagatagarbha theory and early Yogacara thought as well as its great impact on Chan, is deserving of special attention. Notwithstanding a general claim of the inadequacy of words and letters, the Lankavatara Sutra offers a specific account of what it emphatically opposes:
Said the Blessed One: Even when there are no corresponding objects there are words, Mahamati; for instance, the hare's horns, the tortoise's hair, a barren woman's child, etc.... are neither entities nor non-entities but expressed in words. If, Mahamati, you say that because of the reality of words the objects are, this talk lacks in sense.... W ords, Mahamati, are an artificial creation.... T he validity of all things has nothing to do with the reality of words. 
This is simply an explicit negation of the imagined correspondence relation between language and objects, and, therefore, of a descriptive, entitative, or cognitive use of language. The Lankavatara Sutra stresses the inner realization of supreme wisdom mainly by meditational practice and experience. This kind of "inner realization" cannot be achieved by any cognitive activity through discriminative language and dualistic thinking, since no such objective, isolated state of mind exists to which a cognitive language can refer. "What has been realized by the Tathagatas, that is my own realization...."  Everybody has to realize his or her own enlightenment by engaging his or her own subjectivity. Since no awakened state of mind stands behind each general term or expression, nothing can be grasped or gained through restricting oneself to "the words of the canonical texts." 
The Chan masters align themselves with the Mahayana critique of the descriptive, entitative, or cognitive use of language. While the Chan masters claim that language is inadequate, this claim is inseparable from their denial of the appropriateness of a cognitive maneuver (zhijian or zhijie ). For instance, Baizhang Huaihai states:
You must stop all cognition of being or of nonbeing, stop all desires and pursuits.... Nowadays there are cognitions or opinions about the Buddhas. But what people know about, what they seek after, or what they attach themselves to, all can be called the waste of the illusory knowledge produced by cognitive language. They can also be called "coarse language" (cuyan ) or "dead language" (siyu ). 
Why are cognitive discourses coarse (not refined) or dead (not living) language? Because they are the pitfalls of a deceptive correspondence relation between language and reality. Chan masters often advise their disciples that the Buddhist reality "is not something that can be obtained through descriptive words and language."  "Those who search for written words, and thereby look for the corresponding reality, become even more impeded by them."  A similar, but more forceful, critique appears in the Linji lu. Linji repeatedly points out that Buddha, dharma, dao -- "all are empty names and designate no true reality."  All Buddhist doctrines and teachings are "simply medicines to cure diseases of the moment"; "None of them designate any true reality."  He further declares:
The various phenomena in this world and other worlds are in all cases devoid of self-nature.... They are empty names, and the words used to describe them are likewise empty. But you think these idle names represent realities. This is a great error. 
His simile, "All sounds, names, words, phrases are like changes of robes," expresses his belief that language, like other useful things or tools, serves only practical purposes and is always in the process of change due to different contexts and situations. Similarly, one person can wear and change different robes, but you cannot claim that a robe defines the reality of the person.  Thus the radical Chan emphasis on non-reliance upon words, or, in Baizhang Huaihai's terms, on non-restriction of words,  aims particularly at freeing Chan Buddhists from the restriction of the descriptive, entitative, or cognitive use of language. The outcome of such an assertion of "the inadequacy of language" is not a turn completely away from language, as we will see, but a turn "within language.".
When analyzing the twofold truth -- samvrti (worldly convention) and paramartha (higher meaning or truth) -- Nagarjuna explains: "Unless worldly convention is accepted as a base, the higher meaning cannot be taught; if the higher meaning is not understood, nirvana cannot be attained."  Candrakirti clarifies that one of the meanings of samvrti is "the world of ordinary language."  Thus, for Madhyamika, to accept worldly convention as a base is to accept conventional language as a base. Nagarjuna's verse unmistakably shows his insight into the necessity of using language. Sengzhao, the Chinese Madhyamika thinker, who had a great impact on Chan, grasps Nagarjuna's point very well. He writes: "Though language cannot fully express the nameless dharma, without using language, the dharma cannot be conveyed." 
The situation a Mahayana Buddhist faces here is very much analogous to the one the Daoist thinker Zhuangzi faced a long time ago, namely how to find a way out between the conventional use of language and complete silence. However, the context of this problem now presupposes the understanding of both the insufficiency of language and the necessity of using it as primordial to the Buddhist philosophy of the Middle Way. The Middle Way maintains a nirvanic dimension in the everyday world without presupposing a transcendental realm. By the same token, it pinpoints the insufficiency of conventional language without postulating any sacred language (whether a metalanguage or complete silence). This position is like a thread running through the Buddha's teaching, Madhyamika discourse, and Chan practice. It advises Buddhists to avoid sliding into any extremist attitude toward language. The Middle Way thus provides a solid ground for a Buddhist liminology of language. If language use was not necessary and inevitable, the Buddha would have remained silent forever. Then the Buddha would never have been the Buddha of sentient beings. Only to the extent that neither retreat into complete silence nor ignorance of linguistic insufficiency are satisfactory does a liminological exercise become credible. The fact that Zhuangzi, as a precursor, prominently engaged in marginal speech qualifies him as practicing a kind of Daoist Middle Way. However, Chan Buddhists more plainly thematize the issue of the necessity or inevitability of using language, while simultaneously addressing the insufficiency or inadequacy of language.
Following Nagarjuna and the Mahayana notion of upaya (skillful means), the Chan masters express their concern with the necessity of language use from a heuristic or pedagogic perspective. The Platform Sutra contains the following explanation: "All the sutras and written words ... are established for men.... Should deluded people ask the wise, the wise will expound the Dharma for the stupid and enable them to understand and gain a deep awakening."  The Jingde chuandeng lu, fascicle 7, records:
One day Mazu Daoyi () asked Zhizang (): "Why don't you read sutras?" Zhizang said: "What is the difference between a sutra and me?" Mazu said: "However that may be, you should get it in the future for the sake of other people." 
This use of language, including preaching and the reading of scriptures, "for the sake of other people," in Huangbo Xiyun's terms, is using "words for accommodating and guiding people" ().  The Chan masters are fully aware that they cannot avoid using language to accommodate and guide people: "When host and guest meet each other, there cannot but be exchanges of words and remarks." Therefore, they ask their disciples to "pay strict attention" to the use of language.  For this reason, Huineng, Baizhang, and Linji all formulate and hand down to their disciples a special use of language. Huineng teaches his disciples how to preach Dharma by "utilizing the thirty-six pairs of opposites and going around without attaching to either side."  Baizhang prefers using "the sentences that cut off the connection with two opposites."  Linji talks about "one phrase with three dark gates and three vital seals," and so on.  Nevertheless, these highly skillful uses of language in diverse contexts are not easy for most Chan students to practice. Oftentimes these skillful uses are the feats of enlightened ones. In the process of study and practice, in the encounter with Buddhist traditions, the task of keeping a focus on existential-spiritual awakening, on eschewing the trap of certain prevalent uses of language, remains. Hence, Chan masters fight tirelessly against any search through words and letters. But this position does not necessarily require the abandonment of language. The Platform Sutra shows an unusual discernment concerning the hidden relation between a negation of language and the inevitable use of language:
People who attach to emptiness ... simply say that we should not use written words. Since they have said that, language also becomes inappropriate for them. However, such linguistic expressions already have a form of written words. Again, they say that direct dao does not establish any written words. But merely these two words -- "not establish" -- are already written words. 
Huineng is explicating here that even if people negate language one way or another, when they convey this negation, they cannot help but use language. Chan masters have no illusion that they can get rid of language in their this-worldly enterprise of "curing people's illness." On this account, the Chan critique of the conventional use of language is by no means tantamount to the cancellation of language. It would be better understood as an effort to find an alternative way of communication, an alternative way of using language.
Detachment from dualistic thinking is one of the chief characteristics of Mahayana Buddhism. Nagarjuna's famous eight negations -- the negations of four pairs of opposites -- in the dedicatory verses of his karika set a pattern for the subsequent development of nondualistic discourse in various schools of Mahayana Buddhism.  The Lankavatara Sutra presents its account of nonduality as follows:
What is meant by non-duality? It means that light and shade, long and short, black and white, are relative terms, ... and not independent of each other; ... for the condition of existence is not of mutually exclusive character. Therefore, it is said that all things are non-dual as are Nirvana and Samsara. 
It is a logical step to include the pair of speech and silence in the Mahayana reflections on nonduality. The Vimalakirti Nirdesa Sutra seems to be on the verge of addressing this topic when it touches upon the relation between speech and silence in the discussion of "the dharma gate of nonduality." However, it leaves the impression that the best entrance into nonduality is silence, and therefore may lend itself to the privileging of silence over speaking.  Despite this, the Praj~naparamita tradition and other Mahayana sutras provide provocative views in blurring an absolute demarcation between speaking and non-speaking. For example, in the Vajracchedika Praj~naparamita Sutra (or Diamond Sutra) we read: "What do you think, Subhuti, is there any dharma which the Tathagata has taught?--Subhuti replied: No indeed, O Lord, there is not."  This view contradicts the conventional opinion that the Buddha taught or spoke something. The distinction between what is spoken and what is not spoken, between speaking and non-speaking, is virtually obscured.  This idea is further articulated by the Lankavatara Sutra: "It is said by the Blessed One that from the night of the Enlightenment till the night of the Parinirvana, the Tathagata in the meantime has not uttered even a word, ... for not-speaking is the Buddha's speaking."  All these expressions imply the nonduality of speech and silence, or the dynamic, transitionary relation between them. However, in most cases, they are implications and have no fully developed form. Only when we delve into Chan discourse do we find clear-cut statements about the nonduality of speech and silence.
In his Wanling lu, Huangbo Xiyun unequivocally states: "Speaking is silence (); silence is speaking (); speaking and silence are nondualistic ()."  Another Chan master, Dazhu Huihai, construes Vimalakirti's silence as being beyond speaking and non-speaking, a silence of the duality between silence and speech -- a strategy very similar to Madhyamika's emptiness of emptiness and Zhuangzi's nothingness of nothingness.  In saying this, Chan masters demonstrate that they apply the principle of pratityasamutpada (interdependent arising) to the issue of speech and silence, presenting a non-isolated, truly relational understanding of speech and silence. Speech and silence thus no longer have their self-identity. One always functions in relation to the other, and each always has its absent presence in the other. Each always retains traces of the other. Sengzhao's saying, "Speech always has something unspoken," might be a good footnote to the Chan notion of the speech-silence relation.  Chan masters might add one more point to Sengzhao's saying: silence always speaks.
As soon as the Chan masters bring speech and silence within the reach of relational, nondualistic understanding, the functions of speech and silence are liberated from the conventional fixation. As a consequence, Chan Buddhists acquire a better grasp of the Buddha's strategy and better guidance for their own soteriological practice. On the one hand, silence is no longer considered mere silence. "The Tathagata's silence speaks just as his speech does ()." "The Tathagata always speaks -- there has never been such a time the Tathagata does not preach."  One of the examples used to illustrate this point is the Buddha's silence in the face of fourteen metaphysical questions. This silence signifies the Buddha's refusal to take a stand in metaphysical debates. This case, as well as Vimalakirti's silence mentioned above, indicates that silence, in certain Buddhist contexts, is close to a special kind of negative expression that brings into effect the negation of dualistic thinking.
On the other hand, speech does not always or necessarily mean speaking. "Though the Buddha has preached for forty-nine years, he virtually does not say a word."  The Chan master here is clarifying that the Buddha's words are only intended to accommodate and guide people. Words simply cannot replace the realization of enlightenment, which involves going through one's own existential-spiritual transformation. There is no reality to which the words just correspond. In the entitative, reifying, or metaphysical sense, the Buddha says nothing. Therefore, Chan Buddhists regard their saying as non-saying and practice a sort of self-erasing saying in order to avoid being entangled by saying or misleading people.
Insight into the nonduality of speech and silence is significant to the Chan liminology of language. Once the absolute, impassable demarcation between silence and speech is obscured, the path for playing on the borders of language is opened. In other words, the liminological play of language is based on, and made possible by, a trans-conventional attitude toward the limits of language. This in turn is cultivated by the philosophy of the Middle Way, by the non-static, relational understanding of speech and silence, by the detachment from any duality, and so forth. However, freedom from fixation on either silence or speech enables Chan Buddhists, first of all, to relocate (or redefine) the positive role of language within the framework of the liminology of language.
If addressing the necessity or inevitability of language use still leaves the role of language somewhat negative, the Hongzhou sect sheds light on the positive relation between the Buddha mind and language. Hongzhou Chan tends to deconstruct the dichotomy of substance (ti ) and function (yong ) by canceling the quasi-metaphysical issue of substance and emphasizing that the everyday activities of the human mind are nothing but the function of the Buddha-nature itself.  The everyday activities of ordinary mind and the realization of the Buddha-nature or Buddha-mind are nondualistic. Accordingly, using language, as one of the everyday activities, is certainly relevant. When someone asked "How can we recognize our own mind (as the Buddha-mind)?" Huangbo Xiyun replied: "That which speaks (namely, asks the question) is your mind."  In other words, you should not attempt to attain enlightenment outside everyday activities. Speaking and writing, just like other everyday activities, can definitely be useful for triggering enlightenment. "Speaking, silence, move, rest -- all sounds and forms -- are the Buddha's business."  Dazhu Huihai also points out: "If separated from language, there would be no Buddha-mind."  Therefore, "The Buddha-mind, having no fixed form and characteristic (), can neither be separated from nor tied to language ()."  This is the best characterization of the Chan position concerning language. In terms of this position, all one-sided interpretations of the Chan view of language cannot stand up to examination.
Hongzhou Chan further claims:
(a) The Tathagata 's preaching is the Dharma (); the Dharma is the Tathagata's preaching (); the Dharma and the preaching are nondualistic (). 
(b) You just speak anytime and can speak of either events (shi ) or the principle (li ) without being hindered. The fruit of enlightenment is also like this. 
(c) The enlightened person's letters and words all come from the great wisdom and serve the great function right now and right here, having never been trapped by emptiness.  The enlightened person "always speaks in terms of function (), having no fixation whatsoever on either affirmation or negation()." 
These statements reveal, first of all, that the Chan masters' central concern is not whether silence or speech is preferable, but how to become enlightened. Once enlightened, hence free from any fixation, one is then a master of using language, a master of playing on and around the limits of language. There is no necessity to remain silent forever.  Second, when a logocentric hierarchy of silence and speech is completely abandoned, the function of language, or how to use language, in the soteriological practice, becomes fundamental. We should not misunderstand the Hongzhou Chan masters' view as a return to the logocentrism of speech. After saying the words "The Tathagata's silence speaks just as his speech does. While the Tathagata speaks all day long, no word is actually spoken," Huangbo Xiyun adds another sentence: "Though it is the case, we consider silence essential."  Here "essential" does not mean something metaphysical, but functional. Given the context, what Huangbo Xiyun refers to as silence is surely not complete silence opposed to speech, but a saying as non-saying or a self-erasing saying, a strategy of silencing or negating the duality between speech and silence. This is a unique Chan usage against conventional usage, a liminological play.
The Chan saying as non-saying, or its self-erasing saying, also involves two major aspects. On the one hand, fully aware of the necessity of using language for guiding people and of the risk of misleading some, the Chan masters invoke an interplay between speech and silence. By sustaining the position that their words are not different from silence, and that no word has been spoken about any hypostatizable reality, the Chan masters move away from entifying and help people to detach from their words. On the other hand, by underlining the non-saying or silence, by treating their saying as something like the finger pointing to the moon (as they always say), pointing to what is absent within language, pointing to what has not been spoken or what cannot be adequately spoken, Chan masters virtually say a great deal. In this way, Chan masters play on and around the boundary of language without being obstructed. As exhibited by Huangbo Xiyun's well-known maxim "walking all day long without touching the ground,"  Chan masters walk on the boundary of language without falling to either side. They therefore achieve their great freedom in the use of language. In terms of this understanding, the Chan masters' radical objection to reliance on words and their creative use of language can be placed within one framework of the liminology of language without contradiction. They are simply two sides of one single coin.
1. Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. C. K. Ogden (London: Routledge and Kegan Paul, 1922), pp. 187-189.
2. Maurice Blanchot, The Infinite Conversation, trans. Susan Hanson (Minneapolis: University of Minnesota Press, 1993), p. 337.
3. Maurice Blanchot, The Gaze of Orpheus and Other Literary Essays, trans. Lydia Davis, ed. P. Adams Sitney (Barrytown, New York: Station Hill Press, 1981), p. 129.
4. Michel Foucault, Language, Counter-Memory, Practice, ed. Donald F. Bouchard, trans. Donald F. Bouchard and Sherry Simon (Ithaca: Cornell University Press, 1977), p. 65.
5. Martin Heidegger, Being and Time, trans. John Macquarrie and Edward Robinson (New York: Harper and Row, 1962), p. 208.
6. Maurice Merleau-Ponty, Signs, trans. Richard C. McCleary (Evanston: Northwestern University Press, 1964), p. 44.
8. For the coinage of the term liminology and the emphasis on the meaning of threshold, see David Wood, Philosophy at the Limit (London: Unwin Hyman, 1990), pp. 53, 150; Mark Taylor, Tears (Albany: State University of New York Press, 1990), p. 224; and my article "An Inquiry into the Liminology of Language in the Zhuangzi and in Chan Buddhism," International Philosophical Quarterly, June 1997.
9. "Chan Buddhism" here refers to the main line of Huineng through Mazu/Shitou to Linji/Caodong, etc. See Heinrich Dumoulin, Zen Buddhism: A History, vol. 1, trans. James W. Heisig and Paul Knitter (New York: Macmillan, 1988), chap. 9.
10. Fung Yu-lan, A Short History of Chinese Philosophy, ed. Derk Bodde (New York: Macmillan, 1948), p. 341.
11. D. T. Suzuki, Studies in Zen (London: Rider and Company, 1955), pp. 141-142.
12. See Dale S. Wright, "Rethinking Transcendence: The Role of Language in Zen Experience," Philosophy East and West 42 (1) (January 1992): 113-138, and Bernard Faure, Chan Insights and Oversights (Princeton: Princeton University Press, 1993), chap. 7, "Chan and Language: Fair and Unfair Games.".
13. Mervyn Sprung's translation in his Lucid Exposition of the Middle Way: The Essential Chapters from the Prasannapada of Candrakirti (London: Routledge and Kegan Paul, 1979), p. 177. Cf. David J. Kalupahana's translation of this verse: "When the sphere of thought has ceased, that which is to be designated also has ceased. Like freedom, the nature of things is non-arisen and non-ceased" (Nagarjuna: 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. 268). It is interesting here to read Gadjin Nagao's translation of this verse (from the Sanskrit): "When there is a quiescence of mental activity, / Then the need for discourse ceases and / Reality, like unto cessation, / Neither arises nor passes away." Evidently, his translation is still close to the original Chinese translation. Both, if compared with Sprung's and Kalupahana's, more easily give the impression that Nagarjuna here is expressing a total negation of language. See Gadjin M. Nagao, The Foundational Standpoint of Madhyamika Philosophy, trans. John P. Keenan (Albany: State University of New York Press, 1989), p. 67.
14. Nagao, The Foundational Standpoint of Madhyamika Philosophy, pp. 67-68.
15. Kalupahana, Nagarjuna: The Philosophy of the Middle Way, pp. 268-269.
16. Nagarjuna, Karika 22:11; Candrakirti, Prasannapada. See Sprung, Lucid Exposition of the Middle Way, pp. 155, 201.
17. Sprung, Lucid Exposition of the Middle Way, pp. 230-232.
18. Ibid., pp. 155, 178.
19. Ibid., p. 229.
20. The Lankavatara Sutra, trans. D. T. Suzuki (London: Routledge and Kegan Paul, 1932), pp. 91-92.
21. Ibid., p. 124 (my brackets).
22. Ibid., p. 166. A further analysis may distinguish between two kinds of inadequacy of language, which are both reflected in the Lankavatara Sutra. One is the cognitive, reifying use of language, and another is the discriminative nature of language itself, which is particularly linked to the issue of how to designate the nondualistic dimension of meditational, especially enlightenmental, experience. In my view, these two are closely related to each other. The problematic of the latter becomes prominent only in relation to the problematic of the former. Giving heed merely to the latter makes both Suzuki's and Kalupahana's accounts of the notions of linguistic inadequacy in the Sutra inaccurate, though they differ in that Suzuki's offers a positive estimate and Kalupahana's a negative. Cf. D. T. Suzuki, Studies in the Lankavatara Sutra (London: Routledge and Kegan Paul, 1930), pp. 105-110, and David J. Kalupahana, A History of Buddhist Philosophy (Honolulu: University of Hawai'i Press, 1992), chap. 18. For a more recent exposition of the view of linguistic inadequacy in the Sutra, cf. Florin G. Sutton, Existence and Enlightenment in the Lankavatara-Sutra (Albany: State University of New York Press, 1991), pp. 156-163.
23. Guzunsu yulu , fascicle 2, reprinted in Chanzong jicheng (Taibei: Yiwen Yinshu Guan, 1968), vol. 11, p. 7324. Throughout this essay, Chinese terms and proper names are rendered according to the pinyin system, except for a few titles (or names) from English sources where I leave the original rendering unchanged.
24. "" (Huangbo Xiyun , Chuanxin fayao , Chanzong jicheng, vol. 13, p. 8976. Cf. Denshin hoyo, Ui Hakuju yakuchu Zenseki shusei maki 2 vol. 2 Tokyo: Iwanami Shoten, 1990 , pp. 22-23).
25. "" (Dazhu Huihai ; see Jingde chuandeng lu , fascicle 6, Taisho, vol. 51, p. 248, no. 2076. Cf. The Transmission of the Lamp: Early Masters, trans. Sohaku Ogate Wakefield, New Hampshire: Longwood Academic, 1990 , p. 199).
26. Guzunsu yulu, fascicle 4, Chanzong jicheng, vol. 11, p. 7359. Cf. Yanagida Seizan , Rinzai roku (Tokyo: Daizo Shuppan Kabushiki Kaisha, 1972), pp. 160-161.
27. Guzunsu yulu, fascicle 4, Chanzong jicheng, vol. 11, p. 7361; Rinzai roku, p. 171.
28. Guzunsu yulu, fascicle 4, Chanzong jicheng, vol. 11, p. 7355; Rinzai roku, pp. 129-130. Also cf. Burton Watson, The Zen Teachings of Master Lin-chi: A Translation of the Lin-chi Lu (Boston: Shambhala, 1993), p. 47.
29. Guzunsu yulu, fascicle 4, Chanzong jicheng, vol. 11, p. 7359; Rinzai roku, pp. 156-157. There is no strong evidence to support Faure's accusation that Linji is logocentric and that Linji regards written words as robes and considers speech superior to writing and mental activities superior to language (Bernard Faure, Chan Insights and Oversights, p. 223). As a matter of fact, Linji regards all writings, speeches, and mental activities as robes, because, in his view, all these things are devoid of self-existence and in the process of constant change. See Watson's translation of the relevant sentences in the Linji lu: "Because of mental processes thoughts are formed, but all of these are just robes...." (Watson, The Zen Teachings of Master Lin-chi, p. 61). For an important interpretation of Linji's view of language, see Ronald L. Burr, "Lin-chi on 'Language-Dependence, an Interpretive Analysis," in Whalen Lai and Lewis R. Lancaster, eds., Early Ch'an in China and Tibet, Berkeley Buddhist Studies series, 5 (Berkeley: Asian Humanities Press, 1983), pp. 207-227.
30. Baizhang Huaihai's saying "" (not being restricted by words) (Guzunsu yulu, fascicle 1, Chanzong jicheng, vol. 11, p. 7313).
31. Karika 24: 10. Cf. Sprung, Lucid Exposition of the Middle Way, p. 232.
32. Sprung, Lucid Exposition of the Middle Way, p. 230.
33. Sengzhao , Zhao lun , Taisho, vol. 45, p. 153, no. 1858. Cf. Richard H. Robinson, Early Madhyamika in India and China (New York: Samuel Weiser Inc., 1978), p. 216.
34. The Platform Sutra of the Sixth Patriarch, trans. Philip B. Yampolsky (New York: Columbia University Press, 1967), pp. 150-151. I have changed the translation of the first sentence. The original Chinese reads: ".".
35. Jingde chuandeng lu, fascicle 7, Taisho, vol. 51, p. 252, no. 2076. Also see Ogata, The Transmission of the Lamp: Early Masters, p. 224.
36. Huangbo Xiyun, Chuanxin fayao, Chanzong jicheng, vol. 13, p. 8979. Cf. Denshin hoyo, Ui Hakuju yakuchu Zenseki shusei maki 2 (vol. 2), pp. 36-37.
37. "" (Linji lu; see Guzunsu yulu, fascicle 4, Chanzong jicheng, vol. 11, p. 7357, and Rinzai roku, pp. 145-146. Cf. Watson, The Zen Teachings of Master Lin-chi, p. 55).
38. "." Cf. The Platform Scripture, trans. Wing-tsit Chan (New York: St. John's University Press, 1963), pp. 120-121.
39. "" (Baizhang guanglu, Guzunsu yulu, fascicle 1, Chanzong jicheng, vol. 11, p. 7316. Cf. Sayings and Doings of Pai-chang, trans. Thomas Cleary Los Angeles: Center Publications, 1978 , p. 34. Cleary's translation misses the meaning of "" sentence , and, therefore, Baizhang's instructions about the special use of language).
40. Guzunsu yulu, fascicle 4, Chanzong jicheng, vol. 11, p. 7350; Rinzai roku, p. 67. Cf. Watson, The Zen Teachings of Master Lin-chi, p. 19.
41. The Platform Sutra, ed. Zongbao ; see Guopeng , Tanjing duikan (Jinan: Qilu Shushe, 1981), p. 143.
42. Cf. Kalupahana, Nagarjuna: The Philosophy of the Middle Way, p. 101, and Sprung, Lucid Exposition of the Middle Way, p. 32.
43. Suzuki, The Lankavatara Sutra, pp. 67-68.
44. See The Holy Teaching of Vimalakirti: A Mahayana Scripture, trans. Robert A. F. Thurman (University Park: Pennsylvania State University Press, 1976), p. 77. Cf. The Vimalakirti Nirdesa Sutra, trans. Lu K'uan Yu (Berkeley: Shambala, 1972), p. 100.
45. In Buddhist Wisdom Books, trans. Edward Conze (London: George Allen and Unwin, 1958), p. 52.
46. This view is obviously close to Nagarjuna's statement of Karika 25:24, in which he says "no Truth has been taught by a Buddha for anyone, anywhere" (Sprung, Lucid Exposition of the Middle Way, p. 262).
47. Suzuki, The Lankavatara Sutra, pp. 123-124.
48. Huangbo Xiyun, Wanling lu , Chanzong jicheng, vol. 13, p. 8994. Cf. The Zen Teaching of Huang Po, trans. John Blofeld (New York: Grove Weidenfeld, 1958), p. 121.
49. Yuezhou Dazhu Huihai heshang yu , Jingde chuandeng lu, fascicle 28, Taisho, vol. 51, p. 442, no. 2076. Cf. Tongo yomon , Ui Hakuju yakuchu Zenseki shusei maki 3 (vol. 3) (Tokyo: Iwanami Shoten, 1990), pp. 106-107. Also cf. The Zen Teaching of Instantaneous Awakening, trans. John Blofeld (London: Buddhist Publishing Group, 1962), p. 81.
50. "" (Sengzhao, The Reply to Liu Yimin , Taisho, vol. 45, p. 157, no. 1858). In the transition from Zhuangzi's liminology of language to a Chan Buddhist liminology of language, the role that Sengzhao's view played and the contributions that Sengzhao made to the liminology of language are significant topics for further research.
51. Huangbo Xiyun, Wanling lu, Chanzong jicheng, vol. 13, p. 8994.
52. Huangbo Xiyun, Wanling lu, Taisho, vol. 48, p. 385, no. 2012b. Cf. Denshin hoyo, Ui Hakuju yakuchu Zenseki shusei maki 2 (vol. 2), pp. 70-71. The edition printed in Taisho misses some sections that the Changzong jicheng edition includes. Therefore, sometimes I quote from different editions.
53. For an account of the Hongzhou sect's radical shifting of the Chan focus from substance to function, see Yanagida Seizan and Umehara Takeshi , Mu no tankyu: Chugoku Zen : (Tokyo: Kakukawa shoten, 1969), pp. 156-159. The term quasi-metaphysical is used here to assert the difference between Western metaphysics and the Chinese use of substance (ti). The Chinese character ti originally involves the meaning of body or organic whole. Even if in Chinese philosophy ti is distinguished from yong, at the very beginning it is made clear by Wang Bi, the Neo-Daoist, that ti and yong are united in dao or nonbeing. It has never had the meaning of essence as opposed to accidents in Western metaphysics. Nor has it had the meaning of self-existence. These meanings are absent from both Neo-Daoist and Chinese Buddhist uses of the term. At worst, ti is relatively static in contrast to the more dynamic feature of young. In the present context, Hongzhou Chan Buddhists are obviously more concerned with how one's mind can respond freely to the change and flux of daily lives, and overcome all forms of quietism and dualistic separation. This concern underlies their claim that outside young there is no ti. Therefore, what they oppose might be called a quasi-metaphysical fixation on ti, a tendency to see the ti, the Buddha-nature, as separable from everyday activities, from the dynamic state of living flux.
54. Huangbo Xiyun, Wanling lu, Taisho, vol. 48, p. 386, no. 2012b. Cf. Denshin hoyo, Ui Hakuju yakuchu Zenseki shusei maki 2 (vol. 2), pp. 74-75, and Blofeld, The Zen Teaching of Huang Po, p. 87.
55. "" (Huangbo Xiyun, Wanling lu, Taisho, vol. 48, p. 385; Denshin hoyo, pp. 68-69).
56. Dazhu Huihai, Yuezhou Dazhu Huihai heshang yu, Jingde chuandeng lu, fascicle 28, Taisho, vol. 51, p. 444, no. 2076. Cf. Tongo yomon, Ui Hakuju yakuchu Zenseki shusei maki 3 (vol. 3), pp. 128-129, and Blofeld, The Zen Teaching of Instantaneous Awakening, p. 96.
57. Dazhu Huihai, Yuezhou Dazhu Huihai heshang yu, Jingde chuandeng lu, fascicle 28, Taisho, vol. 51, p. 444, no. 2076.
58. Huangbo Xiyun, Wanling lu, Chanzong jicheng, vol. 13, p. 8994.
59. Mazu Daoyi's saying; see Jingde chuandeng lu, fascicle 6, Taisho, vol. 51, p. 246, no. 2076.
60. Dazhu Huihai's saying; see Jingde chuandeng lu, fascicle 6, Taisho, vol. 51, p. 247, no. 2076. Cf. Tongo yomon, Ui Hakuju yakuchu Zenseki shusei maki 3 (vol. 3), pp. 80-81.
61. Yuezhou Dazhu Huihai heshang yu, Jingde chuandeng lu, fascicle 28, Taisho, vol. 51, p. 441, no. 2076. Cf. Tongo yomon, Ui Hakuju yakuchu Zenseki shusei maki 3 (vol. 3), pp. 96-97. One of the reasons I afford these quotations so much space is that they have long been neglected by modern scholars and interpreters both in the West and in the East.
62. The Hongzhou Chan masters would definitely disagree with Gadjin Nagao's interpretation that the paramartha should remain silent forever. Cf. Gadjin Nagao, "The Silence of the Buddha and Its Madhyamic Interpretation," in his Madhyamika and Yogacara, trans. Leslie S. Kawamura (Albany: State University of New York Press, 1991), p. 42.
63. Huangbo Xiyun, Wanling lu, Chanzong jicheng, vol. 13, p. 8994.
64. Ibid., p. 8995.