The Metaphysics of Dao

1. The Metaphysics of Dao
2. The Nature of Dao
3. The Conceptual Context
4. The Meaning of Dao
5. Dao-things
6. Dao-language
7. Dao-examples
8. Non-individuation of Dao
9. Dao analysis
10. Two philosophical "dead ends"
11. Conclusion
12. Bibliography

The Metaphysics of Dao

Tradition treats as Daoists those thinkers who addressed what we might call the metaphysics of dao.  I endorse that tradition with some qualifications about the appropriate implications to draw from it.  My argument will support the orthodox view that Daoists are appropriately distinguished from earlier moral philosophers (Confucius and Mozi) by their more metaphysical uses of the term dao.   However, this traditional way of fixing the reference of philosophical Daoism need not entail that that be a separate Daoist meaning for the word dao. Daoists could be said to address the metaphysics of dao—but the same dao that is in dispute in moral philosophy. It needn't have separate metaphysical and moral meanings. Indeed, a careful account of the metaphysics of dao removes the motivation to postulate a separate, Daoist meaning. Precisely because their concept of dao is the same, these reflections will be relevant to how Daoist metaethics informs their criticism of the Confucian-Mohist moral debate. We can explain the full range of Daoist usage, including its metaphysical use, without postulating a separate meaning.

The kinds of metaphysical comments I am targeting include observations that dao lacks beginning-end, is everywhere or inescapable, that we are in dao as fish are in water, that things come to be using dao, dao sustains/nourishes things, from a dao point of view, there is not shi-fei^{this-not this} (or some other distinction), as well as the frequent comments and questions about of dao and wu^thing-kinds (some contrasting the two and others seeming to imply that dao is a 物wu^thing-kind)[1]. If we reflect metaethically on the role and nature of dao as a central normative concept, we should be able to appreciate a motivation for these kinds of statements that does not require transporting them into a religious, God and creation story context.

Let us think of metaphysics as a component of the philosophical project of "thinking about thinking."  In effect, we are asking what is (must be) such that we can engage in thinking.  The different answers we may offer to this question reflect to some degree the different conceptions and contexts of  'thinking.'  Typically the most closely related Western context will be cognitive knowing, e.g., what must the world be like for us to know it (or think about it)? Somewhat less dominating the stereotype, Western thinkers ask about the metaphysical conditions of norms. What must be for us to be subject to them, know them, reason about them and care about them as we do?[2] Where the subject matter is a normative one, we call this second-order thinking "metaethics" of which moral metaphysics is a part (along with moral epistemology, moral semantics, moral logic and so forth).  I'm suggesting we understand Daoists best as doing metaethics in this sense. They do this as part of their project of assessing (and criticizing) the ancient first-order moral debates in China. Moral metaphysics will consist in reflecting on the nature of the subject matter and of our way of dealing with it. It should address why there are moral disagreements and explain the role or importance we give to the concepts used in ethical thinking. They should have theories not only about the metaphysics (nature) of dao, but also its knowability, its objectivity, decidability and so forth.

The Nature of Dao

Technically, I will not be seeking any specifically metaphysical account of dao. I prefer to think of the project as explicating the nature of dao and will not argue that my account fits any standard definition of 'metaphysical'. An account of the nature of dao will still differ from a theory of the meaning of the term dao^guide.  It would not be rebutted, for example, by evidence that some Daoist gave an account that contradicted it. Daoists, in my view, offered a variety of accounts so we need not conclude that knowing the meaning of the term was sufficient to settle the matter. 

However, an account of the nature of dao should be consistent with the meaning in the sense that it is plausibly specifies what different theories are talking about--what the ancient thinkers were disagreeing about. An account of the nature of dao thus amounts to taking a position on the issues ancient Chinese thinkers themselves were discussing.

Calling a position 'Daoist,' in the reading of the traditional reference-fixing formula proposed here, does not attribute a commitment to any particular answer to second-order questions.  Metaethical questions include reflections about the nature of dao but are not limited to these. In addressing these metaethical issues about dao, I am seeking to illuminate doctrines that are more characteristic of Daoist reflections than they are of first-order thinkers who also use the word dao (those dealing with simple casuistry or first order ethics). Daoists may be skeptics, relativists, monists or mystics. 

Daoists, on this proposal, may engage in first order dao theorizing and may even draw inferences about such commitments from their metaethical reflections. However, again I will not treat any particular first-order dao as distinctively Daoist.  Daoists may well disagree with each other in these first-order inferences. My version of traditional naming convention is that we treat given thinkers as Daoist philosophers in virtue of their centrally addressing metaethical questions. This way of explaining their focus on dao effectively distinguishes theirs from early Confucian, Yangist, Mohist and legalist discussions about dao in classical China.

A second reason I won't offer a straightforward account of the metaphysics of dao is that I leave open the possibility that a satisfactory purely metaphysical account of the nature of dao may be overwhelmingly difficult. That difficulty itself should explain why Daoist metaphysics turns out to be so obscure and does so without having to import Western mysticism as its explanation. The failure of traditional metaphysical categories will serve to explain the Daoist penchant for evasive and skeptical accounts of the nature of dao without the assumption that Daoists use the word differently, or that they used it to refer to some obscure object newly introduced into the discussion by their changing the meaning of the term. The obscurity, we will have discovered, is inherent in specifying the nature of the dominant normative concept in Ancient Chinese disputes about dao.

A final reason my account may not seem like straightforward metaphysics is my view that metaphysics play a radically different role in Chinese thought.  I can characterize the contrast best by borrowing recent talk about the "place" of meaning. Consider this statement of that place: "the meaning of words is determined by the role they play in the evidence-inference-action game.”[3] Meaning mediates between two poles of human interaction with the world and contributes to the mental process of inference.  Brandom calls the two interactions "entry and exit transitions with the world."  Traditional Western metaphysics has been preoccupied with the evidence or entry side, i.e., with the "passive knower" and a dominant appearance-reality form and setting of the problems of metaphysics. Metaphysics and epistemology motivate each other in Western philosophy.  Idealism and dualism are familiar and fairly obvious examples of the link of metaphysical and epistemic views. The epistemic preoccupation of Western metaphysics invites the positivist critique of metaphysics as cognitively meaningless--because it seeks descriptions of reality that are not subject to empirical testing.[4]

Chinese metaphysics inclines just as strongly to the exit-action end of the "transitions with the world."  It addresses how conceptions of reality fit with the project of guiding human action.[5]  My metaphysics will, accordingly, seem like distant cousins to the interpretive accounts usually found under the title "the metaphysical Dao."  Considering the possibility of this difference in fundamental outlook, we needn't read metaphysical passages about dao^guide as evidence that Daoists have changed the subject. Treating dao as a subject matter does not require that the word dao need be other than the one at home in the discussion of practical and normative issues. In addressing metaphysical issues, we need not infer that Daoists must be referring to a reality in a classical Western, (e.g., Parmenidean) sense—as something independent of or transcending sense experience, conceptions and beliefs. Keeping this difference in mind should help refrain from the temptation to situate metaphysical statements into our familiar appearance-reality structure. We needn't assume that meta-passages about the nature of dao must be about ultimate reality (or an ultimate source or creator of reality).  We needn't take claims about the ineffability of dao as claims about the ability to experience ultimate reality or of language to represent, define, denote, prove, confirm or "capture" some ultimate reality.

The Conceptual Context

One source of the difficulty in characterizing the metaphysics of dao is this deep difference in the role of metaphysics.  The other most important source is the wide gap between the conceptual schemes of ancient China and those of Indo-European (Buddhist, Islamic and Judeao-Christian) culture.  I have discussed these differences at length in my 1992 and will only restate the conclusions here. The important points for our purposes here includes the absence of a Western representational and propositional conception of 'belief', 'knowledge', 'inference' as well as (correspondence or coherence) 'truth', 'reason', or 'inference'.  The ways ancient Chinese grammar use to attribute commitments resembled de re belief ascriptions more than the de dicto ascriptions more familiar in English.[6] When ancient Chinese thinkers ascribed commitments to each other, they treated commitments as taking the form of a disposition to use a term or description of some part of the world. They seldom attributed commitments in a way that suggested "inner representations" of outer "facts" or pictures of things in the world. The writing may have been pictures, but the commitments were to use some term (written or spoken) of the given object.

In theory of language, I argued that they also did not otherwise highlight the sentence as a significant unit of language. They did not have a clear notion of the syntactic sentence, which would be located conceptually between the clearly recognized  ming^names (paradigmatically the ideographic character), the broadly construed ci^phrase (ranging from compound characters to long strings "with a yi^intent") and the discourse-like dao^guide.  A key to my view was my hypothesis that dao^guide was a linguistic unit--at the opposite end from ming^names.

Three other features of classical Chinese grammar bear on our account of the nature of dao. First, Ancient Chinese does not have singular-plural grammatical marking, so references to dao are comparatively mass-like. To skip a lot of controversy and detail, the important points are daos can be summed (your dao and my dao make our dao) and we may individuate them in several ways. Second, Chinese lacks articles—definite or indefinite. Again, traditional translations and accounts typically supply the definite article, “the” before occurrences of dao particularly in contexts where the nature of dao is concerned. I suggest instead using it as an implicit plural (e.g. like a mass noun) or an indefinite 'some' instead.  Third, dao is sometimes used as a verb in Chinese – and most famously in Daoist texts such as the first line of the Daode Jing – dao which can be dao-ed is not constant dao.

Against the prevailing practice of translating the verb as "to speak”, I argued that it (a) should incorporate the normative force of the noun, i.e., something like "to guide" and (b) that the range of denotation should include both speech and writing -- as well as gesturing and so forth. I suggest treating the verbal use as "to express as guidance." Again, I will not repeat these arguments here, but want to broaden them slightly in keeping with the entry-exit transition model. The notion of ‘expression’ may restrict the range of the verb in ways that underplay how many ways one may endorse a dao. Brandom's[7] formulation of a notion close to what I am after would be "to endorse a pattern of material practical inference" whether in intention, plan, expression or recommendation to another. I have reservations about explicating ancient Chinese commitments in terms (e.g., 'inference') that suggest a commitment to sentential analysis.  Still, we can rephrase Brandom's notion as expressing commitment to conform to a way (analogous to commitment to follow a practice), which leaves the reference to 'inference' implicit in the structure of the practice.  Zhuangzi's helpful metaphor here gets the vague effect of 'inference' via our "shooting out" shi-fei^{this-not this} 是非" distinctions and commitments to which we cling, as we would to "an oath or a treaty."  Talk of a 'practice', rather than to a specific intent, a principle or a norm is advised because of these worries about segmenting dao sententially. Individuation of dao, as I noted above is best left vague. To dao is to undertake a commitment to correctly effecting some (bit of?) dao in behavior.

The variety of ways to individuate dao stem not only from the mass-like segmentation, but also from the familiar difficulty of drawing the boundaries with a word like "community."  Implicitly in ancient China, the stereotype of dao in ethical debate was a large, nearly universal human community.[8] But Daoist literature regularly draw our attention to daos of thieves, of musicians, of carpenters etc. as well as to even more global daos of all natural kinds, of the world (natural and social) of nature and so forth. The notion of dao clearly is not limited to ordinary moral discourse, but to any practical discourse--including perhaps the discourse of "natural signs."

The Meaning of Dao

With those caveats, the question of meaning is relatively simple--despite the impression that dao is an impenetrable mystery of the East. The almost universal translation is one of the easiest and most familiar words of the English language--'way'. My view is that it is no accident that the translation sticks and works so well. The two concepts are remarkably close in meaning--except that 'way' has rather more explicit grammatical individuation and lacks a verb form.

To do the metaphysics of dao, we can start by thinking philosophically about ways while trying to reason about them (a) using a conceptual structure like that available to ancient Chinese thinkers and (b) considering the issues salient in their philosophical agenda.  We do not set out to do the impossible--reconstruct all features of their lives, all background religious beliefs or any of the other impossible feats of Verstehen that lead to hermeneutic regress and paradox. What we envision is reconstructing enough of their "manifest history" of the philosophical issues and the inference structure of their concepts to think along with them about the nature of normative ways.

Part of the appeal of 'way' as an explication of meaning is that, like dao, 'way' is indefinable--I do not mean ineffable but, in Hacking's helpful phrase it's one of those familiar, tiny, almost unnoticed words that "tend to be circularly defined."[9] Any synonym or attempted definition leads us back to 'way', which is a more primitive English term than any of its partial synonyms.

Hacking's phrase occurs in his discussion of a "loose" distinction between object words, idea words, and elevator words. This distinction will also be handy for our purposes in explaining a philosophical difference between dao and 'way'.

"In addition to 'objects" and "ideas" we need to take note of a group of words that arise by what Quine calls semantic ascent: truth, facts, reality. Since there is no common way of grouping these words, I call them elevator words, for in philosophical discussions they raise the level of discourse." (Hacking 1999:21.)

Hacking notes that elevator words typically are familiar, unproblematic words that have quite innocent uses -- until we employ them for semantic ascent or with philosophical emphasis. Hacking, however, clearly would not consider listing 'way' among his elevator words. It's not that Western philosophers do not use the word; they use it a lot, but usually innocently. Hacking's own account of looping shows how "handy" the term can be in philosophical discourse.

People classified in a certain way tend to conform to or grow into the ways that they are described; but they also evolve in their own ways, so that the classifications and descriptions have to be constantly revised. (Hacking 1995)[10]

Ironically, Hacking's (or anyone's) short list of Western "elevator words" ('truth', 'facts', 'reality') contains words that ancient Chinese thinkers seldom or never used as terms of philosophical ascent.

We could expand the list of philosophically pivotal "familiar" terms which we all learn to analyze deeply as we learn philosophy: 'know', 'reason', 'true', 'believe', 'represent', 'refer', 'mean', along with some longer but equally central notions in our left-side world-view such as 'conscious', 'experience', 'sensation', 'perception'.  Finally, we also learn to analyze a cluster of other simple structural words like 'the', 'some' and 'a'.  To my recollection, however, I have never seen a philosophical article on the concept of a 'way.' The irony, then, is that the concepts in Hacking's and my "short" list of Western elevator terms include those that are usually either absent or "innocent" in Chinese philosophical literature.  The counterpart of dao, the most widely used term of philosophical ascent in ancient China, is an almost unnoticed "innocent" in the West.

The contrasting lists of elevator words manifests the contrast in emphasis noted earlier.  Western elevator terms cluster on the entry side of the transitions between language and the world.  'Way' belongs to the exit side. While 'way' is not in Hacking's list, we can find some close relatives--partial synonyms and situational counterparts--in Western ethical reflection and in the anti-Platonic strains of Western thought, e.g., when Western philosophers analyze 'practices', 'conventions', 'games', 'forms of life', 'conceptions of the good life', 'traditions', 'processes', and 'modes'.  We may view 'way' as a more comprehensive term that embraces all of these practical concepts--as a genus of which these others are species. The network of words we might use in defining 'way' includes another core term of the language--the question word ‘how'.  We point to a way in answer to either a "how to" or a "what to (do)" question. We may think of knowing ways as unlike the routinely analyzed "knowing-that" of Western epistemology, and more like Ryle's "knowing how"[11] or "knowing to."

A possible motivation of meaning-change hypotheses regarding dao may stem from this ironic contrast.  Interpreters, cognizant that dao is a major elevator word in Daoist meta-theorizing, naturally seek a counterpart elevator word from Western philosophical discourse. This explains the interesting coincidence that interpretation of dao makes it a Chinese philosophy counterpart of 'reason', 'truth', or 'ultimate reality'--borrowing terms from Hacking's short list of Western elevator terms. If 'way' had become a target of Western philosophical analysis and a mainstream elevator word, one motivation for this interpretation would shrink.

Dao-things

Motivation aside, my goal (reasoning about the nature of ways using the conceptual structure available to ancient Chinese thinkers) counts against the temptation to explain dao in terms such as 'experience', 'perception', 'belief', '(propositional) knowledge', 'reality' (as philosophically contrasted with appearance), 'truth', 'facts', and so forth. Our emphasis on the "exit transitions" and the kinds of practical nouns that do show up regularly in Western philosophy reminds us that 'way' is essentially a normative term.

Equally important, perhaps, is that in being the central normative elevator term of ancient Chinese discourse, dao signals an important difference in the conception of normativity. Ancient Chinese thinkers did not make talk of sententials, such as 'laws', 'rules', 'principles', or 'norms' central to their meta-discussion of normativity. Dao was the crucial way they referred to the normative realm (although without the familiar Western supposition that normativity implies a separation from common-sense nature--as did, e.g., Plato's "realm of forms," Kant's categorical imperative or Moore's open question). A dao is some aspect of the natural context that invites us to perform or "implement as guidance" for our action.

Both 'way' and dao share concrete uses as 'road' or 'path'. Roads seem fully object-like-- highways are concrete (or asphalt) objects. However, if we picked up the asphalt "object" that runs between Las Vegas and the Grand Canyon and put it back down between Las Vegas and Provo, a brown, ragged scar would then be the way to the Grand Canyon, not the original ribbon of asphalt. We can characterize a road as an information storage and retrieval device.[12] Drivers with "broadband access" read the information (in the form of the contrast between the black of the road and the earthy colors on both sides) to retrieve the architects' knowledge of the way to get to the Grand Canyon.

The abstract nature of ways is made clearer when we consider less technologically developed roads. A path is blazed by the frontiersman who goes through the forest hacking a bit of bark of a tree every 100 yards or so.  Such a "path" would not strike us as being much more like an object than is a written instruction. Similarly, a boy scout who constructs piles of rocks leaves a fragment of a conventional scout "language" behind for other scouts to interpret in their own purposive activity. To understand the notion of a dao or a way is to see the continuity between road-making, path-marking, drawing a map or writing a list of directions. This underlying continuity motivated my earlier analogy of the metaphysics of dao to that of normative language, which I called "guiding discourse." The linguistic notion of discourse was intended to capture the common normative content of roads, maps and instruction manuals. The linguistic metaphor should neither limit the explanation to the paradigmatically linguistic end of the continuity nor rule out concrete highways.

The focus on the instruction book model of a dao has the advantage of launching us our "manifest history" of the setting of the Daoist meta-doctrines about dao.  The paradigm initial form of Confucius' dao was extremely text-like--a book of ritual (the rough counterpart of a series of books by Emily Post). Still, the Confucius of The Analects clearly is engaged in a study of ritual that is not exhausted in library work. The text has him placing importance on examples and the study of history (not merely of rule-books but of histories of the ways others acted in the past).

Further, it will require some ingenuity to blend the linguistic core conception with a connected common use of 'way' will become important when we leave early Confucianism for more "naturalistic" arguments about dao. The human paths a Confucius studies in history (e.g., the strips of bare ground surrounded by grass) may not have been caused by an explicit intent to communicate guidance. Prior walkers were seldom intending to cause or elicit any belief in some later "reader" of their tracks.[13] Similarly, there are "natural" ways (paths) that are created by animals. Consider a mountaineer finding his way by following the paths of mountain goats.[14] The goats' walking was certainly not intended as discourse.  Still, we can and do read and follow such paths. 

We can even talk about there being "a way through the forest" without supposing that either human or animal has gone that way before. That way consists in a configuration of objective components that contribute to their being a practical solution to a "how to" problem, i.e., the dangers and opportunities between the starting point and exit, e.g., narrow spots in the river, fallen logs across ravines, and places where bears are not feeding at the moment. There are also ways that we read for guidance in dressing for the day -- red skies at night. We will return to these later, but note that while they are non-language like in being natural rather than social, they are still language like in being "read" for guidance. We spot natural "signs" and "interpret" their significance for our actions.[15]

Dao-language

Though we want to be able to take in these less language-like cases, the philosophical story of Daoism requires a note on a passage in The Analects that does emphasize language--the "rectifying names" sorites.

This passage initiates[16] a philosophical reflection on puzzles about dao in China. In effect, the direction of reflection signals the dominance in ancient China of questions of interpretation of dao over questions about the content of dao.  Cognitive revision and originality are thought of as expressing themselves in adapting guidance to real-time situations rather than as revising the explicit words of the guidance. Chinese tradition focused on the problem of games and interpretation before they learned to doubt their traditional value systems--they did their Wittgenstein before their Socrates. This problem arises most acutely within Confucian traditional orthodoxies with their commitment to following conventional, ceremonial patterns.  Confucius' students regularly discovered that, in concrete contexts, they disagreed about what to do even when they agreed on using the same "rulebook" as their guide.

A natural metaphysical treatment for this discourse-like situation is the type-token model from theory of language. We have dao-types (discourse dao) and dao-tokens (performance dao). A discourse dao is analogous to a play or a musical score. I coined the term 'performance dao' for that implicit token sense of 'way' that is the "real" goal of a discourse dao.  A performance dao consists of a concrete, particular series of actions or behaviors. The score-performance model motivates a cluster of helpful insights. First, it illuminates Confucius' constant pairing of ritual and music in his formulations. Second, it explains the implicit “method” of rectifying names—which consists not of giving definitions or doing analysis, but simply of using them correctly yourself—as a model. Third, it points us to an important difference in interpretive processes as each end of our language spectrum. Rather than a theory of the meaning of a word as related to the experiences or evidence for its use, we have a “performance” of the discourse as the “way of interpreting.” This is, however, different enough from our evidence-side notion of interpretation to that may be helpful to distinguish it. Let us call it “performance interpretation.” Finally, it motivates my translation of 德de^virtuosity as “virtuosity” and reminds us that interpretation in Confucianism (and for most of Chinese philosophy) is not simply a matter of a bi-polar right-wrong but a more continuous aesthetic evaluation. This gives us one way to construe Hall and Ames' assertion that Chinese thought represents an aesthetic rather than a logical order[17].

Despite these advantages, however, a type-token analysis is a more stark (and Platonic) metaphysical structure than is suggested by the contextually shifting individuation we have highlighted in ancient Chinese talk of dao.  The type-token model presents a particular problem for my approach since I find little motivation in ancient Chinese concepts of background beliefs for the kind of universal-particular model characteristic of ancient Greek (and Indic) thought. While handy for us today in understanding what a dao is, we are unlikely to find any echoes of this metaphysical structure in Daoist writing from the classical period of Chinese thought.  Thus, while we have no overt reason to reject this metaphysical analysis of dao, it departs from the conditions on a solution we began with.  Arguably, it is not what would occur to a Chinese thinker with the conceptual structure and philosophical agenda of ancient China.[18]

Dao-examples

It's hard to avoid some version of a type-token analysis, however, if we concentrate on the problem of metaphysical individuation. Let us agree not to seek a definitive answer to how we should individuate daos.  That will simply be another sense in which we will not seek an "orthodox" metaphysics of dao. We can still draw some interesting observations about the complexity Chinese thinkers encountered when they considered issues about the nature of dao.

To focus our issue, consider the use of example, or what Munro calls “model emulation”[19] in Confucianism. Take Confucius in the act of rectifying names. He simply uses language correctly. The rectify names passage ends:

Thus when an exemplary person uses a name, it can surely be spoken, and when spoken it can surely be acted upon. There is nothing careless in the attitude of the exemplary person to what is said.[20]

The guidance that allows us to rectify names has the metaphysical status of an action (which we may analyze as a series of events, as a process and so forth).  It is a bit of the world—a concrete speech act. It's in the world as much as a road is. However, it is intended as a guide—e.g. for our use of the word. When I “follow” the junzi’s example, I extrapolate or interpret to my different situation of action. The junzi uses the proper term to speak to someone younger, I fail to follow his example if I use the same term of that same person. The relation of that concrete particular and my action is interpretively like the relation between a play and performance, instruction and action. (This corresponds to Wittgenstein's point about pointing.)

Non-individuation of Dao

So rather than specifying exactly how to individuate dao, we start by observing that we can take two perspectives toward the same speech act.  It may be regarded as a performance interpretation of something (as the junzi saying what is permitted in the circumstances), or as something to be interpretively performed (the junzi being careful to set a model for us). These are different points of view to take of the same series of events or actions. While they are explained here for speech, they clearly apply also to gesture, dance, chant, ritual and other behavior according to a practice. A community relative to that practice exists when its members regularly take both kinds of attitudes towards each other's actions.

The two points of view are somewhat like normative and descriptive. The way you did the action is regarded as descriptive when I have no commitment to follow you—i.e., to subject your way to practical interpretation in my own behavior. It is normative when I do accept or endorse such a commitment or when you interpret your own action as thus binding on me. We routinely intend our actions both to conform to duties and to exemplify the practices to which we take others to be committed.  We may well disagree on whether I should acknowledge your performance as a guide for me.  The undertaking reflects a judgment that another should use the performance particular as a "score" or guide. It is analogous to pointing to a road--another bit of the furniture of the world.  The speech act is a bit of our surroundings to which we can endorse, pick out or point to as a guide.

Like the type-token contrast, the normative-descriptive contrast is another of the Western conceptual dichotomies that would be controversial to attribute to ancient Chinese thinkers.[21] Attributing the full conceptual apparatus of our distinction is surely suspect, though we may find closely related counterparts in a different conceptual context.  We can, therefore, mark the two perspectives on a manifestation of a dao in a different but related way. One bit of the concrete world we take as a correct subject to performance interpretation--we 是 shi^this:right as something to be 行 xing^"walked".  We may also evaluate the concrete act as a performance of a something we take in the first way--again 是 shi^this:right it as a good "walking" of that dao or 非fei^{not-this:wrong} it as a bad or wrong one.

To regard something as a discourse dao is to take it as subject to interpretation (a dao-type).  To regard it as performance dao is to take it as subject to evaluation (a dao-token).  Since they may take the different evaluative attitudes toward the same speech or ritual act, we should not be surprised that Chinese thinkers do not think of them as different metaphysical types.  They also do not clearly distinguish the two evaluative attitudes. They are both evaluative attitudes that can be expressed in terms of a shi-fei^{this-not this}是非 judgment.  Chinese thinkers need not sense in these two attitudes anything like a descriptive-prescriptive distinction.

Routinely, we take a concrete document (a book of ritual or a musical score) as a discourse dao--something to be interpreted or executed in performance. We can, however, also treat it as an extended performance—a writing act by some sage-king.  We can treat it as subject to selection--as Mozi does when he initiates skepticism about the Confucian confusion of the traditional and the 義yi^morality.  When we pick (是 shi^this:right) a physical road out from the surroundings (or from other roads) we undertake to "read and execute it."  The behavior we subsequently produce in each case either correctly follows the road or loses it. Note that the shifting-perspective insight does not solve the problem about what is dao, but it helps us appreciate how Chinese thinkers might have come to regard dao as being in nature without reducing it to descriptive history and while giving us an alternative, contextual and pragmatic distinction to replace a type-token metaphysics.

The ancient Chinese notion of human (moral) conduct is simply 行 xing^behavior. Now we can see how a normative-perspective shift analysis instead of a type-token analysis can account for how dao works in human life. A great many behavior particulars, book tokens, physical structures etc. may be regarded as guides to performance--as something "to be performed." Discourse dao is discourse or language-like essentially in being viewed as inviting 行 xing^behavior interpretation. It is unlike normal human discourse in not necessarily in employing human conventions of communication. The "authors" of the concrete entities that we treat as discourse might not intend to communicate anything. Nature might not have any intentions as to how we use its "signs."

Recent pragmatists (e.g., Brandom (1994) and McDowell (1994)) face the challenge to show how norms can be enshrined in facts but still transcend those facts and moralists (Gibbard 1990) address the way our moral practices rule out appeal to the fact of our moral practices in practices of justification. When China's Socrates, Mozi, develops his similar point, he doesn't see himself as leaving tradition in making the distinction. His argument is that we would not call something 義yi^morality or 仁ren^humanity simply because it was a society's tradition.[22]  Mozi uses the traditional language to express his philosophical approach to questions about dao and his doubts about Confucian traditional guiding dao.

Having made that point, Mozi initiates the use of the terms of analysis of dao that are crucial to understanding the mature Daoist meta-ethics--shi-fei^{this-not this}是非 and 辯 bian^{distinction dispute}. As we saw, the general notion of evaluation is shi-fei^{this-not this}是非 shi-fei^{this-not this}. The capacity to use shi-fei^{this-not this}是非may be described as knowing how to 辯 bian^{distinction dispute}. He also uses the concept of 明ming^discern to describe the achievement of mastery of a shi-fei^{this-not this}是非distinction, i.e., as being able to 辯 bian^{distinction dispute} correctly.[23]

The interpretive and normative way of explicating guidance and action that we see emerging in ancient China together with the shifting and vague patterns of individuating can help us understand the emergence of skeptical, relativist and mystical or monist talk about dao in Chinese meta-ethics. Confucians themselves noticed the puzzle about interpretation of a dao into action in changing circumstances. This puzzle affects even as explicit a dao as a ritual ceremony. Mohists then draw our attention to the use of terms, distinctions, and standards in selecting some dao as our basis of guidance. It's not clear that Mohists distinguish the two ways (choice and interpretation) that any explicit dao presupposes other standards. As I argued (Hansen 1992:118) Mozi seems to find a similar dynamic in the two normative activities—both count as 辯 bian^{distinguishing} shi-fei^{this-not this}是非.  We distinguish among different ways of making shi-fei^{this-not this}是非judgments.

Dao analysis

Mozi depends on 天  tian^nature:sky in this context and he makes appealing to 天  tian^nature:sky more central to normative theorizing. Tian^nature:sky天 is not "escapable" the way the edicts of a ruler or the conventions of a social group are. He also treats 天  tian^nature:sky as a solution to the twin problems because of its measurement-like operationality.  The standard exemplified in 天  tian^nature:sky  is the distinction between 利 害  li-hai^benefit-harm (and the natural preference for li^benefit). This step extends the apparatus of dao to natural inclinations. Any learning of a way presupposes some natural way—namely our capacity to acquire and follow it. So the idea of a 天  tian^nature:sky 道 dao^guide becomes important in ancient Chinese dao talk.

The Zhuangzi takes up the theme of daos depending on daos in the form of talk about dependency. Neither the Mohists nor the Zhuangists make a clear distinction between justification and motivation, reasons and causes. Tian^nature:sky's天 operational standard is not evidence for the daos it leads us to choose, as much as a cosmic metaphor for our having a way to choose and follow daos. So we have daos of choosing and interpreting dao both built into identifiable discourse daos (like books of ceremonial ritual or existing conventions) and we have daos built into us as natural ways of acquiring, selecting and mastering socially generated discourse daos. In any normal dao-guided activity, much dao is presupposed (depended upon).

It is important to recognize that the Mohist doctrines, while giving us the language of analysis of daos, differ from Daoist analyses. The appeal to 天  tian^nature:sky  道 dao^guide and the ideal of a constant dao are Mohist innovations, not specifically Daoist doctrines (in fact, as I argue below,  Daoist theory effectively critiques both notions). Mozi speaks of "constanting" (常 chang^constant  used as a verb) language (Mozi 80/46/37-8, 82/47/18) that is adequate to promote, pick out or restore行 xing^behavior. The idea if a 常 chang^constant  道 dao^guide goes together with the idea of a 天  tian^nature:sky  道 dao^guide --constancy being the salient feature of 天  tian^nature:sky and its operational, inescapable standards. 

So we have ample basis for understanding talk of dao being everywhere, inescapable, incapable of being made fully explicit, even for understanding  Zhuangzi's metaphor of humans being in dao as fish are in water. None of this talk requires that dao has changed its meaning. All the Mohist apparatus of normative analysis of dao applies to both the activity of interpreting a discourse into performance and evaluating a performance as a particular "walking" of a discourse dao. The language is also highly context specific, e.g., "choosing this and not that," interpreting in ways that "benefit."  And finally his dao seems to place much more emphasis on the performance side since a test applied to the effect of performance determines what discourse dao is right. In some sense the real dao for Mozi is the way it works out in performance--which, via the standard for judgment (benefit-harm) brings nature into the discussion both as the operational standard and as the context in which success is achieved or not.

These are other themes developed in the Zhuangzi with the suggestion that the context is as much a part of any process of dao guidance as is the explicit discourse. In the course of any behavior, we presuppose other dao besides the one in view--dao of how to select a first order dao and dao of how to interpret it. The idea of 德 de^virtuosity in performance pulls in the dependence on context and criteria of success in real time.  Clearly any act involves lots of dao (many daos) and they are likely to be intertwined in a "bloomin' buzzin confusion" -- not of experience, but of guidance. [24] This iteration of dao together with the individuation, blurring and the naturalness of summing to include context can even explain the talk of a holistic or single dao. 

The Mohist analysis itself promoted this view of the unity of dao.  The natural standard of benefit-harm is shared in every natural kind and everywhere--it is constant nature. The standard and context are set in constant nature and the ideal of finding a natural dao to guide all human action seems the solution that will yield an integrated constancy in human affairs. The ideal, however, quickly required adjustment to fit the complexity noticed in the natural world. Mohist utilitarian analysis put the emphasis on the results of actual performance and stimulated a trend toward emphasizing the uniqueness of situations of performance at the same time the universalized reasoning encouraged summing. Other moral thinkers, like Mencius and Yang Zhu followed suit (or influenced Mozi) in appealing to 天  tian^nature:sky's authority but tilted away from discourse toward a special intuition tuned to guiding concrete process. They still disagreed in their accounts of what behavior was dictated by natural standards—their answers ranging from Mozi's universal utility to Yang Zhu's egoism.

The move toward intuition was thus both a product and a betrayal of the Mohist ideal of a 常 chang^constant 道 dao^guide. The original goal was a formula or norm that could be a standard for choosing, reforming, etc. a constant discourse dao whose content will not have to vary with time or location. It also envisioned a discourse together with internal operational standards of interpretation into practice that were both unequivocal. Neither the content nor the interpretation should vary with schools of emphasis as the traditionalist Confucian dao did. Implicitly, the goal was to evade the normal play of intuition, but the emphasis on performance made special intuitionism the dao of choice--prior to Zhuangzi.

Two philosophical "dead ends."

The history of Daoism in the Zhuangzi mentions neither Mencius nor Yang Zhu. It places two other developments between Mozi and Laozi. We will discuss only the second here. Shen Dao represents a stage in the dialectic that Zhuangzi comes to view as a reductio of the Mohist appeal to nature--as well as the consequent intuitionism. Shen Dao's doctrine and the mature Daoist rejection of it are crucial steps in the meta-ethical dialectic in ancient China and the ultimate complexity of the metaphysics of dao.

Shen Dao coins a term, 'Great dao' to refer to the actual, total process--i.e., the actual history of the world. Then he observes that we do not have to "know" any guidance to follow Great dao. We can relax and mimic him in "floating like a leaf on the water." I have been suspicious of the common analysis on which Shen Dao's doctrine represents a kind of fatalism. His importance, I believe, lies in revealing that the Great dao, as he conceives it, has no normative force. The history in the Zhuangzi puts the point by saying that Shen Dao's is a dao that is not dao--a dao for the dead.

The important point about the nature of dao is that token performance of itself has no normative force--offers no guidance. If we take the sum of all actual performance of all things (the great Dao) as a normative guide, we have no guide. To get guidance, we need some other source of normative direction that guides us in selecting a performance. That meta-ethical insight clearly would also undermine the idea of a natural constancy. The injunction to conform to natural constancies gives us no guide. There is a dao of nature (the course of nature), but it is not one of the options among which we are to make a dao choice. I interpret the opening of Qiwu Lun as expressing this insight. All the normative daos that compete for our real choice as guides are similarly "pipes of 天  tian^nature:sky." They are natural by virtue of being "puffed out of us" by our nature, i.e., it is our nature to generate these ways of speaking and making distinctions to guide our behavior. The appeal to nature does not settle anything once we have noticed that the problem is a choice among rival actually occurring daos.

Thus, there is insight in the traditional assertion that Daoists take 道 dao^guide  rather than 天  tian^nature:sky as the guide. Without some dao other than natural constancy, we have no guidance. However, as the reflections so far have shown, without "reading nature" accurately as we execute or perform some dao, we also don't get any guidance. Extracting guidance always requires a mixture of nature and "discourse."  Themes in the Zhuangzi develop this insight into an extended reflection on the complex and puzzling inter-relation between the natural and the social--the discussion of 天  tian^nature:sky  and 人ren^human. It expands on the point we noticed earlier in discussing Mozi, that anything we do with a dao presupposes some other dao --of interpretation, or of evaluation. In any dao activity, we depend on some meta-dao. 

A similar line of reasoning yields the Zhuangzi's rejection of intuition. We could undermine Chinese versions of intuitions as we do Western ones by noting that the intuitionist owes us an account of why we should care about the "simple moral property" grasped by the intuition. That an object possesses that property makes it analogous to any other natural property--potentially irrelevant to our practical concerns.  The Zhuangzi expresses this point by noting that we have to choose the intuition as a guide--from among alternative potential "ways" of guiding our behavior and we have to choose which intuitions (of the conflicting kinds available in any medium sized community) to grant this authority. His conclusion, as in the case of nature and the Great dao is that we cannot get a 是 shi^this:right out of our 心xin^heart-mind  without presupposing one. 

The Daoist theme clearly cautions against choosing daos that conflict with nature -- "do not let the human harm the natural" but it otherwise removes the authority of 天  tian^nature:sky and intuition.  Ultimately our authority for choices and interpretation are some presupposed dao. Tian^nature:sky  天  by itself makes no choices.  The Zhuangzi tries to balance its emphasis on the theme of the dependency of social daos on nature with the insight that when we use any dao, our dependencies are on dao all the way down. We can't escape presupposing some guidance that is not fixed by nature or history.

The insight, I argued, is expressed negatively in the Laozi in the famous opening passage. Any dao that can dao is not a constant dao. I believe it surfaces in many of the allegedly metaphysical passages precisely in the insistence that dao is somehow like a thing but not a thing--not a natural kind. Many of the allegedly "metaphysical" passages in both texts read better as denials that dao is a natural kind than as affirmations. That daos guide our language (stereotypically in guiding our classification of features that we count as constituting a thing-kind for the purposes of acting) together with the observation that natural kinds have to do things in certain ways to live, usually makes as much or more sense of any alleged "creation" passages than does an Indo-European story of  supernatural creation.  This view of the nature of dao has trouble neither with passages that deny the "visibility" of dao nor with those that note that dao can't be exhausted in language or ideas.

The skill passages fit this account better than they do those that portray dao in terms of supernatural perfection. A recurring theme in these passages is how someone having the skill might fail at teaching it to others. That yields a far more plausible sense in which dao cannot be "captured" in language. The famous "difficult" part of the Cook Ding story illustrates how dao outruns any content we acquire in past practice--not only in language.

Finally, though we did not find a straightforward metaphysical solution to the problem of individuating dao, we can make sense of its nature. Much of the account comes via negation. Dao is not simply an object, though aspects of dao may take the form of a road, a map, or a written score or book of instructions. Process objects, such as the Great Dao are daos in one sense, but are irrelevant to our living choices--not daos in another. Similarly Natural processes are daos but represent mainly the limits on our possible choices, not guidance in making them. Dao is not limited to language or discourse except in an extended sense -- where we consider as "language" everything that is interpretable into 行 xing^behavior performance.

Dao always involves a mix of nature but can never consist in mere "receptivity."[25]  Social or explicitly discourse dao always requires interpretation in ways that entail reading some natural dao along with it.  Natural and normative dao are inextricably blended in all dao guidance. The normative side does not follow from the mere existence of a social practice--nor does it require one. When social practice is an aspect in dao, it is subject to interpretation in some act--including the speech acts that constitute pronouncing dao. These speech acts can be evaluated as correct or incorrect as performances of a dao of advocating or pronouncing (writing) dao. Choice of which dao to perform, qua act, can also be evaluated as a performance of some dao of choosing a dao.  Interpreting a chosen dao is itself an action, for which there are several ways among which we may choose.  The we should choose among them is a dao and so forth. Dao questions require dao answers all the way down--they do not come to rest in either history or nature.

The picture could be a counterpart of McDowell's claim that humans experience the world conceptually. Our rephrasing of his Sellarsian picture using a normative conception of concepts translates it into a claim that humans encounter a world via dao.  Zhuangzi's picture of humans being in dao in the way fish are in water starts to seem fully apt. To be human is to be in a reality viewed as open invitations to ways to "carry on."

Conclusion

I am proposing to refine a traditional account of the character of Daoist thinking that I take that tradition to have mistaken for a meaning change. The refinement does not deny that they may give accounts of the nature of dao. To the degree that we consider such an account metaphysical, Daoists may be said to address the metaphysics of  dao^guide, but the "metaphysical" viewpoint is of a metaphysics of action, not a way of turning their backs on the philosophical issues of ancient China to talk about a religious view of supernatural creation and divine nature. Rather than asking, "What is, such that we may know it?" the Daoist asks, "What is, such that humans can act in it?" His answer does not use the familiar concepts of Western sentential philosophy--no propositions, truths (including moral truths), beliefs, practical reasoning, or most of their close and distant conceptual relatives. Nor does a Daoist conception of either language or mind shift focus to the experiential or entry side of meaning. It does not start with mystical experience, consciousness, evidence, or either a successful or failed attempt to picture some ultimate being. 

The elevator terms of this conception include this-not this (right-wrong) (of actions of contextual selections), distinctions, discrimination, desires, deemings, and dao itself.  Dao is the main elevator term but is otherwise close to 'way' in English--a term that never had been a dominant elevator term of Western philosophy. Chinese meta-ethics uses a radically different conceptual scheme in a different philosophical agenda, and with different ways of setting of its problems. I have not, as I predicted, provided a definitive individuation of this confused practical concept, but we can begin to appreciate why (a) Daoists express the view that the subject matter leads to obscurity and paradox and (b) why their conclusions cover the range from monism to relativism and skepticism.

Bibliography

Alt, Wayne (1991) "Logic and Language in the Chuang-tzu" Asian Philosophy v. 1 n. 1 p. 61-76

Ames Roger T and Rosemont Henry Jr. tr. (1998). The Analects of Confucius: a Philosophical Translation. New York: Ballantine Books.

Ames, Roger and Hall, David (1987). Thinking Through Confucius. Albany: State University of New York Press.

Brandom, Robert B. (1994). Making it Explicit. Cambridge: Harvard University Press.

__________. (2000). Articulating Reasons: An Introduction to Inferentialism. Cambridge: Harvard University Press.

Brooks, Bruce & Taeko (1998). The Original Analects. Princeton: Princeton University Press.

Carus, Paul (1913). The Canon of Reason and its Virtue. Chicago: Open Court.

Chen, Guying 陳鼓應(1977). Lao Tzu: Text, Notes, and Comments. San Francisco: Chinese Materials Center.

Cook, Scott (1997) "Zhuang Zi and his carving of the Confucian ox." Philosophy East and West v. 47 n. 4 p. 521-554

Creel, Hurlee G. (1970). What is Taoism?. Chicago: University of Chicago Press.

Cua, Anthony S. (1970) "Reflections on the Structure of Confucian Ethics" Philosophy East and West v. 21 n. 1 p. 1-16

__________ (1989) "The Status of Principles in Confucian Ethics" Journal of Chinese Philosophy v. 16 n. 3 p.

Defoort, Carine (1997). The Pheasant Cap Master (he guan zi): A Rhetorical Reading. Buffalo: SUNY Series in Chinese Philosophy and Culture.

Duyvendak, J. J. L. (1954). Tao Te Ching. London: John Murray.

Eno, Robert (1990). The Confucian Creation of Heaven. Buffalo: SUNY Series in Chinese Philosophy and Culture.

Feng , Gia-Fu and English, Jane (1972). Lao Tsu: Tao Te Ching. Taiwan: Caves Books.

Fung, Yu-lan (1952). History of Chinese Philosophy. Princeton: Princeton University Press.

Gibbard, Alan (1990). Wise Choices, Apt Feelings: A Theory of Normative Judgment. New York: Oxford University Press.

Girardot, Norman J. (1983). Myth and Meaning in Early Taoism: The Theme of chaos (hun-tun). Berkeley: University of California Press.

Graham, Angus (1969) "Chuang-tzu's Essay on Seeing Things as Equal," History of Religions v. 7 n. p. 137-159

__________ (1981). Chuang tzu: The Inner Chapters. London: Allen & Unwin.

__________ (1983). “Daoist Spontaneity and the Dichotomy of 'Is' and 'Ought'.” In Mair, Victor, ed. Experimental Essays on Chuang-tzu. Honolulu: University of Hawaii Press. 3-23.

__________ (1985). Reason and spontaneity. London and Maomö: Curzon Press.

__________ (1986). Yin-Yang and the Nature of Correlative Thinking. Singapore: Insitute of East Asian Philosophies.

__________ (1989). Disputers of the Tao: Philosophical Argument in Ancient China. La Salle, IL: Open Court.

__________ tr. (1960). The Book of Lieh-tzu. London: John Murray.

Hacking, Ian (1995). “The Looping Effects of Human Kinds.” In Sperber, Dan, and Premack, David and Ann, ed. Causal Cognition: A Multidisciplinary Approach. Oxford: Clarendon Press. .

__________ (1999). The Social Construction of What?. Cambridge: Harvard University Press.

Hansen, Chad (1983). “A Tao of Tao in Chuang Tzu.” In Mair, Victor, ed. Experimental Essays on Chuang-tzu. Honolulu: University of Hawaii Press. 24 55.

__________ (1992). A Daoist Theory of Chinese Thought. New York: Oxford University Press.

__________ (1997). “The Zhuangzi: A Historical Introduction.” In Tsai Chih Chung, ed. The Dao of Zhuangzi. Garden City, NY: Anchor Books (Doubleday and Co.). 9-22.

Henricks, Robert G. (1989). Lao-tzu: Te-Tao Ching: A New Translation Based on the Recently Discovered Ma-wan. New York: Ballantine Books.

Hsiao, Kung chuan (1979). A History of Chinese Political Thought, Volume I: From the beginnings to the Six Dynasties. Princeton: Princeton University Press.

Kaplan, David (1990) "Words" Aristotelian Society, Supplementary Volume v. 64 n. p.

Kjellberg Paul and Ivanhoe Philip J ed. (1996). Essays on Skepticism, Relativism and Ethics in the Zhuangzi. Buffalo: SUNY Series in Chinese Philosophy and Culture.

Lau, D. C. tr. (1963). Lao Tzu: Tao Te Ching. Baltimore: Penguin Books.

Liu Xiaogan (). Classifying the Zhuangzi Chapters. Ann Arbor: University of Michigan Center for Chine.

Mair, Victor ed. (1983). Experimental Essays on Chuang-tzu. Honolulu: University of Hawaii Press.

__________ tr. (1990). Tao Te Ching: The Classic Book of Integrity and the Way. New York: Bantam Books.

Maspero, Henri tr. (1981). Taoism and Chinese Religion. Amherst: University of Massachusetts Press.

McDowell, John (1994). Mind and World. Cambridge: Harvard University Press.

Munro, Donald J. (1969). The Concept of Man in Early China. Stanford: Stanford University Press.

Needham, Joseph (1956). Science and Civilization in China. New York: Cambridge University Press.

Peerenboom, R. P. (1999) "Ruling the Country in Accordance with Law" Cultural Dynamics v. 11 n. 3 p. 315-51

__________. (2000) "The Limits of Irony: Rorty and the China Challenge" Philosophy East and West v. 50 n. 1 p. 56-89

Raphels, Lisa (1992). Knowing Words: Wisdom and Cunning in the Classical Traditions of China and Greece. Ithaca, New York: Cornell University Press.

Rhett Young and Roger T. Ames tr. (1977). Lao Tzu: Text, Notes, and Comments. San Francisco: Chinese Materials Center.

Rosemont, Henry, Jr. (1988). “Against Relativism.” In Larson, G. and Deutsch, E., ed. Interpreting Across Boundaries: New Essays in Comparative Philosophy. Princeton: Princeton University Press. 36-70.

__________. (2003). Confucian Alternative. : Forthcoming.

Roth, Harold D. (1992). The Textual History of the Huai Nanzi. Ann Arbor: Association of Asian Studies.

Schwartz, Benjamin (1985). The World of Thought in Ancient China. Cambridge: Harvard University Press.

Shun, Kwong Loi (1997). Mencus and Early Chinese Thought. Stanford: Stanford University Press.

Smullyan, Raymond (1977). The Tao is Silent. New York: Harper and Row.

Van Norden, Bryan V. (1996) "Competing interpretations of the Inner Chapters of the 'Zhuangzi.'" Philosophy East and West v. 46 n. 2 p. 247-269

Waley, Arthur tr. (1938). The Analects of Confucius. London: Allen & Unwin.

Watson, Burton tr. (1963). Mo Tzui: Basic Writings. New York: Columbia University Press.

Welch, Holmes (1966). Taoism: The Parting of the Way. Boston: Beacon Press.

Wong, David (1984) "Taoism and the Problem of Equal Respect" Journal of Chinese Philosophy v. 11 n. p. 165-183

Yearley, Lee (1983). “The Perfected Person in the Radical Chuang-tzu.” In Mair, Victor, ed. Experimental Essays on Chuang-tzu. Honolulu: University of Hawaii Press. 125-139.

Zhang Longxi (1992). The Tao and the Logos: Literary Hermeneutics, East and West. Durham & London: Duke University Press.