Nothing in the concept of 'interpretation' makes Chinese philosophers subject to such radically different norms of assessment. Clearly, a representative reader from the ancient Chinese philosophical community could correctly take Zhuangzi or Mencius to have contradicted himself. Given the linguistic norms enshrined in Warring States practice, a sentence may validly commit Mencius to something of the form "P and not-P" whether or not he ever entertained the thought. I argue that Zhuangzi drew such an inference about Mencius. [17] Judging a writer to be committed to something incoherent is not a sign of failing to understand his writing. Such judgments are not claims about an author�s beliefs and desires.

This means that we can shift our interpretive focus from author to audience.� What does the expression of this passage commit the writer to according to the norms of linguistic use of some philosophical community? We are asking, "What role do these passages play in the philosophical disagreements, given these concepts and background arguments?" Our task is not to construct a theory of the mental states of a sage named Zhuangzi [18] but to interpret the sentences, arguments and theories written in the text using the community's shared language. We will want to know what a competent ancient Chinese philosopher reading the passage would correctly take the passage to mean, i.e., to what would it commit one who accepted it.� The writer himself may or may not accept or appreciate those commitments.

Communities, in the broad sense, can be at cross-purposes. Religious and philosophical discourse may diverge and the communities effectively cease to communicate for an extended period. Even philosophical communities may splinter and develop in relative discourse isolation, e.g. analytic and "Continental" philosophy. That philosophers should be particularly careful about norms of inference does not give philosophical communities pride of place where norms of discourse evaluation differ.

We should still reject the familiar claim that norms of discourse allow Chinese �logical errors� as long as we have evidence of careful contemporaneous Chinese thinkers who reliably conform to �valid� norms in reasoning. Even if a reasoning pattern is nearly universal, evidence that a single philosophical group recognized and diagnosed its error is enough to bar interpretive claims that the error is �correct in Chinese� inference practices. It can be true that some school trained nearly everyone in the community to be indifferent to these norms, but the sheer weight of such indifference would not change the norms.

That a community of careful, rigorous philosophy was thriving at the time of the text is enough to warrant calling a philosophical reading 'authentic'. Ancient Chinese literary, religious, historical and philosophically inclined communities may all have read Daoists texts differently. The historical possibility of other readings does not require a special justification from a philosophical reading.� I read these passages in the Zhuangzi as philosophical, in particular as informed by extensive contact with the school of names and ancient Chinese semantic theory. [19]

Regardless of authorship, we should formulate for ourselves an account of what a representative competent, thoughtful Ancient Chinese philosopher (a contemporary with sound training) should take to be the implied commitments of a passage or argument. We would be judging as a contemporaneous (not contemporary) philosopher would.� In effect, philosophical interpretation is joining those philosophers in philosophizing. [20]

The community I focus on here is the one engaged in philosophical discourse during the Warring States period. Hence, Graham's reconstruction of the Mohist Dialectical works is important (along with his claim that the Zhuangzi shows signs of familiarity with these kinds of issues).� Graham thus locates a relevant community of philosophically sophisticated inquiry. Given his friendship with Hui Shi, the link between a Zhuangzi persona and the community of linguistic theory is historically firmer than the fond and familiar religious Laozi‑as‑teacher tradition.

At the same time, we cannot rule out someone taking their representative reader of these Zhuangzi passages from some other contemporaneous audience. That is, we might construct the contemporaneous Huang-Lao reading or a contemporaneous authoritarian Confucian reading of the text. There may be reasons for interest in both--as there may be interest in the Han, Tang, Song, Japanese and other readings. We need not speculate about how good a case one could put for treating these other audiences as privileged for purposes of interpreting this text.� In principle, we may equally judge two or more to be historically authentic.� We need no further justification, however, for choosing the philosophical one--the reading guided by the careful attention to norms of inference recognized by the linguistic wing of the China's ancient philosophical community.�

Note that the combination of the philosophical community perspective and the access requirement allows us to interpret these passages as a philosopher would. It does not invite us to read them as either an ancient, enlightenment or modern Western philosopher would. [21] We can derive conclusions as a good philosopher would--i.e. using sound inference, imagination, insight, etc as long as we abide by the hypothesized norms behind ancient Chinese concepts, arguments and theories.� If interpreters a) use crudely Western formulations of skepticism or relativism (e.g., in terms of 'belief' or 'truth') or b) apply it to enlightenment or Greek rationalist theories of mind, reason, meaning, ideas and so forth then c) conclude that Zhuangzi contradicts himself in giving relativist skeptical arguments, their reasoning would give us no reason to dismiss the skeptical-relativistic passages. Any inference that the passages are incoherent must follow in ancient Chinese terms, and it should be a sound inference, not an impressionistic characterization. [22] Finally, even when correct, the conclusion may only be that the passage is incoherent with some other passages, not that the interpretation is wrong.

The Philosophical Context

I will summarize here some philosophically important differences in formulation and focus that distinguishes ancient Chinese schools of linguistic thought from those most familiar in the Western tradition. These differences are relevant to the proper formulation of Chinese skepticism, relativism and monism. My approach below on will be mainly analytical, not textual [23] . Readers interested in a more detailed account that stays closer to the text can consult my (1992). Here, I will concentrate on making theoretical distinctions to pinpoint the philosophical position I take these texts to express. With this warning, however, I will refrain from prefixing "according to the present interpretive hypothesis" repeatedly through the account. Readers please note that this analysis is controversial. I do not offer the following as observation reports, but as elements of my interpretive theory.

First, I adapt A. C. Graham's metaphor for the global character of philosophical discourse in Classical China, i.e., I understand them as "disputes about dao." I treat Dao as an essentially normative concept�a way. Ways are the subject matter of practical thought as facts are the subject matter of descriptive thought. (This is my way of reformulating the commonplace contrast between the metaphysical focus of traditional Western thought and the practical focus of Chinese thought.)

The familiar translation of dao as 'way' is satisfactory, though I sometimes use 'guide' to emphasize the "normative" essence of dao-analysis.

One advantage of 'way' in the present context, however, is the rhetorical affinity of "way of life" with important Western notions, e.g., Wittgenstein's "form of life" and Rawls' "comprehensive view" (1993) as well as his earlier "conception of a good life" (1971).�� These all share important features with the Chinese conception of dao. They are wholesale guiding structures that include ways of describing and categorizing things for the purpose of action. The norms of meaning are included along with those of ethics. Making certain distinctions, having certain epistemic norms and rules of justification and so forth are components of a dao in the sense that they are part of a comprehensive view or a "form of life."

I divide the concept of dao into 'discourse' dao and 'performance' dao. The former is paradigmatically verbal and written instructions.� It embraces, however, whatever can guide us�whatever could give us knowledge of �how to� or �what to� do.� A "performance dao" is an interpretation into action�a practical interpretation.� It should be what a discourse way intends or points to.� My usage here borrows from a music model. [24] The score is the discourse dao and a performance is a "practical" interpretation.

The test of being a discourse dao is its being subject to normative practical interpretation. Norm-guided, practical interpretation takes the form of an execution or performance. The test of a performance dao, then, is that it is subject to evaluation, i.e., as a right or wrong interpretation.� One can attempt to follow a dao and get it wrong.

Many daos have both aspects. Discourse daos range from explicit instructions and algorithms to gestures, examples, paths, maps and descriptions of historical exemplars. Expressing, exhibiting, and gesturing are themselves performances--subject to evaluation.� However, they are also modes of discourse--subject to practical interpretation. Similarly, an example or model may be both performance and discourse. To follow the example, we have to extrapolate it to our situation and correctly identify the relevant similarities. We can go wrong following an example as surely as we can following an instruction book.

A path can count as discourse dao (something we may "wander off" or "stay on") or as performance dao (a "trace" of actual past "walkings"). A road is a dao, but we should not think of it as just a physical object like the grass, lakes, and trees surrounding it. A road guides is, i.e. it counts as a road (way) when regarded as something we should stay on�though we can wander off. [25]

My distinction between discourse and performance is interpretive--not a distinction made explicitly in Chinese texts. Examples of dual-aspect dao are common in the texts and even the best analyses blur correct interpretation and correct choice of discourse dao [26] "Disputing the dao" neatly captures how easy it is to slide between the two. What is interesting is that ancient Chinese thinkers formulated and focused on the interpretation question first and the model of interpretation provides the dominant form for discussing what discourse dao to choose. Choice of a dao is a performance interpretation of a "higher order" guiding dao�one for deciding which dao to follow here.

Confucius Analects 13:3 states the interpretive problem in the form of a doctrine of rectifying names. If names are not rectified, . . . language . . . social affairs . . . ritual and music . . . regulation and punishment . . . [all go wrong and] people will not know how to move hand or foot. Mohists first raised the selection problem. The traditional 禮 li^ritual might be the wrong guide. Mohists proposed utility (the benefit-harm distinction) as the guiding standard and used it to solve both the interpretation and selection problems.

Paradigm discourse daos consist of words. To follow a dao requires that we correctly assign the words used in it to things in the context of action. This rests on the ability to 辯 bian^distinguish and all distinguishing can be viewed as using shi-fei^{this (right)-not this (wrong)} 是 非 of objects following the norms implicit in the word�the dao of the word or the way the word is used in the relevant community. The Mohists speak of the dao of particular distinction pairs such as you-wu (being-nonbeing).� Thus, selecting a discourse dao is applying shi-fei^{this-not this} 是 非 to daos. The Mohist ur-norm is the distinction between �利 害 li-hai^benefit-harm.�

Mohists treat their disagreement with Confucians as showing the latter to be confused about how to distinguish between義 'moral' and不義 'immoral'. [27] This way of formulating disputes about dao made philosophy of language immediately relevant to ancient accounts of moral disputes. The disputes are about distinctions and Mohists called such disputes辯 bian^{distinction disputes}. The normative notion of a dao comprehends aspects of the Western concept of 'meaning.' A normative dao determines the right application and use of a guiding term�the right way to 辯 bian^distinguish using that term and its opposite.

仁 Ren^humanity� might be viewed as a Confucian answer to practical interpretation questions. The Mohists, however, used it as common currency to challenge the Confucian conventional dao.� They asked " . . . can we call this 仁 ren^humanity and 義 yi^morality?"� Thus, 仁 ren^humanity� works as a kind of dao of daos, i.e. a dao or way of formulating and settling disputes about dao. However, Mozi interpreted 仁 ren^humanity as universal concern and Confucius as (a notoriously undefined) something else. A standard for interpreting can itself be subject of interpretive dispute.

With this background, we can preview Zhuangzi's skeptical position. Choosing and interpreting a dao are themselves performances. They presuppose some norm, standard, distinction or dao. The Zhuangzi raises the obvious question: What is the correct standard (dao) to use in choosing and interpreting a dao and what is the correct dao of practical interpretation of that standard? This initiates the relativistic regress that informs Daoist skepticism.

Political Theory

Confucians and Mohists viewed the role of government as that of ensuring broad compliance with a shared conception of the moral life--unifying the dao of the social world. For Confucius, the ruler was like the father in inculcating and reinforcing a conception of morality by example. Notoriously, Confucius' wing rejected the rule of law and institutions. Good government essentially consisted of putting a virtuous person in the father-ruler role. The Mohists rejected the Confucian conception of morality and appealing to its traditional status as a standard. A Mohist discourse dao would have more emphasis on language. However, the Mohist conception of the moral education role of government was largely congruent with that of the Confucians.

The Mohist did offer a more complex account of that role. They started with the observation that people have different conceptions of morality (義 yi^morality).� A warlike "state of nature" results from people�s sincere, moralistic attempts to promote their different conceptions of "the good." [28] The resulting disorder and the inefficiency convinces everyone of the wisdom of "selecting" as "governing [29] elder" one whom we all regard as wise. His task is to harmonize society's 義 yi^morality .�

It is tempting to treat Mohism, qua utilitarianism, as initiating a Daoist trend toward neutrality among conceptions of the good. It does explicitly advocate non-partiality or universality concerning Confucian family ties, locality, etc. [30] Further, the Mozi bases its account of utility on objective benefit and harm, not on private or subjective pleasure or desire satisfaction. This suggests that valuing benefit over harm is something universal and natural and does not depend on, e.g., location or time. Still, Mohist political structure has the purpose of selecting and imposing a single conception of shi-fei^{this-not this} 是 非 on the polity. [31]

The political mechanism for harmonizing 義 yi^morality does start with everyone "reporting" their evaluative judgments to their superior but it then gives the superior the performative authority to declare which shi-fei^{this-not this} 是 非 reports all should accept.� After the ruler settles on the shi-fei^{this-not this} 是 非, the Mohist conception of government resembles that of the Confucians more than it does that of a rule of law.� The goal is a leader-guided acceptance of a single social dao�the Mohist difference is in the content of that dao. Theirs would be a universal utilitarian one where Confucians advocated a traditionalist, �partial-love� dao. It has little relation to what Rawls would characterize as institutional structure of ongoing neutrality among comprehensive views (daos).

The quasi-historical account of philosophy in the "Tianxia" chapter of the Zhuangzi [32] suggests a similar analysis. [33] � Disorder follows on the loss of a common traditional ethical conception. This introduces the treatment of Mohism, the first named school discussed. This internal Zhuangzi account identifies other stages on the way to the Zhuangzi. It explicitly identifies the others as endorsing impartiality or lack of doctrinal bias in some way. Let us call them 1) the few-real-desires-group 2) the can't-miss-dao group and 3) the reversibility group.

The few-real-desires-group treats most of our desires as the product of our accepting a certain conception of the good life. Particularly, they think of insults and shame as harms only for those who accept an optional, evaluative point of view.� We could choose not to take that point of view that values status or social honor.

The can't-miss-dao group [34] , in effect, undermines the appeal to naturalness (or 天 tian^nature:sky) in selecting one way of life over another. If following the natural dao is the goal, you do not need to work at it�you simply will. There is no such thing as getting nature wrong.� All "live options" for us now are equally natural�any we choose will amount to following the Great dao. A negative way of putting the point is: 天 tian^nature:sky� gives us no basis for selecting among rival ways of life.

The reversibility group suggests that we can take any existing guidance scheme of concepts and get an equally compelling dao by reversing all its values. Thus, Laozi values passivity, submission, the lower position, declining, the moist, and other female values over their Confucian-Mohist counterparts. The distinctions and names that constitute our ways of life are not attached in any constant way to correct guidance. This exercise, presumably, should undermine our attachment to any particular dao.

Zhuangzi's relativistic skepticism emerges in the context of this broad social-political movement toward impartiality and it implicitly supports that trend. The Zhuangzi's skepticism lends support to ancient Chinese political advocacy of neutrality among conceptions of the good life.

The Lack of Sentential Context.�

Before elaborating on this political point, let us look at a major contrast in the structure we have been elaborating for disputing dao. When Chinese thinkers elaborate on normative dao, they focus on words, which mark distinctions. Western discussions of morality center on the role of sentences--particularly the universal normative sentence called rules, laws, principles or maxims.

Sentential concepts pervade other aspects of Western thought. Philosophers give reasons to construe thoughts, beliefs, propositions, judgments and even desires as sentential (propositional) in form. Only such "sentential" items, we learn, can be true or false. Inference, reason, and logic proceed by moving from sentence (premise) to sentence (conclusion). Valid thinking is thinking that preserves sentential-truth.

Western moral theory comes to focus on finding some universal "ought" sentence�the first premise in moral reasoning. Historical Western figures speak of moral laws, maxims, rules or principles�all of which they typically analyze as sententials. They understand normative correctness mainly in terms of sound normative inference. A central puzzle of Western ethical theory is what is the "axiom" or "first premise" of moral deduction.�

Ancient Chinese thinkers, while correctly noticing the irreducibly normative and integrated character of language, reference, knowledge, and ways of life, did not develop this insight using sentential concepts nor did they have any clear notion of sentential inference. They did not theorize about the norms of sentential inference. [35] The Chinese conception of zhi^know was, thus, non-propositional. Years ago, I analyzed it as know-how, but have come to prefer the conception of know-to as the key to understanding Chinese talk about zhi^know.� (An example is the rare idiomatic use in English when we describe someone as 'knowing to come in out of the rain') Know-how presupposes a goal and the conception of a 道 dao^guide includes both the goal and how to reach it. Knowing is knowing a dao; to know is to have a comprehensive answer as to what to do. [36]

Chinese discourse about 心xin^heart-mind uses a belief context (S 以 yi^with T 為 wei^deem:do P¹) that invites term rather than sentential analysis.� The seemed not to construe such states of mind as a relation of a 心xin^heart-mind and a sentential (a proposition, fact, state of affairs etc.).� Where Western accounts would have a person as believing that S is P, ancient Chinese wrote of a person as tending to treat S as P or to (say) "P" of S. [37] � It is closer to Western accounts of de re belief. They linked this pragmatic belief context to their theory of dao, shi-fei^{this-not this}是 非and 辯 bian^distinction. To 為 wei^deem:do an object is to assign some term in a discourse dao to it. This is the core of practical interpretation�a case of rectifying a name. What corresponds to Western propositional beliefs are dispositions to act toward (to treat) a thing in a certain way.

Chinese "knowing" relates to de re believing in a consistent way�zhi^know is a success verb. It is wei-ing something correctly. [38] It amounts to knowing to (say) P in the presence of S or knowing to treat S as P. It underlies the evaluative judgment that our practical interpretation is correct. Clearly, these sentential-term contrasts in talk about belief and knowledge will give a distinctive character to Chinese monism, skepticism and relativism.�

Monism, Skepticism and Relativism in the Zhuangzi.�

I shall treat monism as a metaphysical posture [39] , skepticism as an epistemological one [40] and relativism as a posture about standards of shi-fei^{this-not this} 是 非 and 辯 bian^{distinction dispute}.� We use the standards to guide how we characterize situations or things as shi-fei^{this-not this} 是 非.

Metaphysics: Monism and Realism

The context I have outlined signals that skeptical and relativist attitudes should apply mainly to normative contexts. However, the norm analysis implicitly presupposes a real context (realities to which we assign terms from our daos).� I think we can appropriately describe the Daoist view as metaphysically realist. Our conceptions of that real world may be relative to our comprehensive views or ways of life, but the argument proceeds against a background assumption of a reality independent of our rival conceptions. I also think we can call the Daoist view monism�in one sense.

Naturalistic v. Parmenidean Monism.

The dominant Western sense of 'monism' is a denial of dualisms. Robert Brandom (1994) makes a helpful distinction between distinctions and dualisms. A distinction becomes a dualism when accounts of it entail that one cannot explain the relation of the two elements in terms of something common. Explaining the interaction between the two realms typically becomes a philosophical puzzle. Classical mind-body metaphysics is such a dualism; transcendent supernaturalism, Plato's theory of forms, Kantian noumena-phenomena, and the like are other familiar examples. Common-sense distinctions, however, such as male-female, old-young, living/non-living etc. are distinctions, not dualisms.�

Notoriously, Chinese metaphysics lacks much evidence of the Indo-European mind-body dualism and (somewhat more controversially) has little or no transcendent supernaturalism. [41] If these anti-dualist characterizations are correct, then we can describe Chinese metaphysics as 'monist' while accepting that they are committed to many ordinary distinctions. I think Daoists not only accept this conception, as naturalists, they rely more strongly on it than most other ancient schools.

Positively, a naturalistic monism envisions a continuity of nature. Like the notorious butterfly that, by flapping its wings in Beijing, causes a blizzard in New York [42] , our actions, causally ripple throughout the universe. This Chinese view of the continuity of the human social world and nature is a characteristic of Daoist reasoning that we can take as background in reading these Zhuangzi passages and of Daoist references to "oneness" and "unity". Their view has much of the flavor of modern physicalist monism�e.g., when he describes his wife's death as merely a "transformation" of the same sort that gave her birth (Ch. 18). Of course, it is not materialist or physicalist in the sense of denying a mind-body dualism.� It is naturalistic monism in virtue of neither presenting nor presupposing such a dualism.

Parmenidean monism, by contrast, is the denial not merely of dualisms but of distinctions. It commits one to a peculiarly strong version of one-ness, i.e., not merely of some underlying natural continuity. Parmenidean monism denies all qualitative and quantitative difference. Besides Parmenides, Neo-Platonism, Spinoza, and some idealist views of "the absolute" may approach this kind of Monism about absolute reality. It seems to find expression in Indian Upanishadic and Buddhist metaphysics as well. It tends to be associated with mystical epistemic doctrines and semantics since, typically, it is hard to prove or to state without paradox. The traditional religious view of the Zhuangzi (and Daoism generally) reads references to unity, oneness, etc. as Parmenidean rather than naturalistic monism and links it to the mysticism of Indo-European traditions.

I argue that classical China offers some candidates for rough [43] counterparts of this stronger kind of Monism. Something like this may be the conclusion of Zhuangzi's philosophical companion, Hui Shi. [44] The metaphysics of what I called the "can't-miss-dao" group also suggests an intermediate kind of monism. Its core is continuity-of-nature monism, but from that premise Shen Dao seems to derive a prohibition on our making any distinction between shi-fei^{this-not this} 是 非.�

I have argued that a philosophical reader could reasonably take passages in the Zhuangzi to reject both views. Graham drew our attention to the fact that Zhuangzi seems to formulate and then reject Hui Shi's position (an attempt to derive 'all is one' from observations about concept relativism).

"Nothing in the world is bigger than the tip of an autumn hair, and Mount T'ai is small; no one lives longer than a doomed child, and P'eng-tsu died young; heaven and earth were born together with me, and the myriad things and I are one.

Now that we are one, can I still say something? Already having called us one, did in succeed in not saying something? One and the saying make two; two and one make three. Proceeding from here even an expert calculator cannot get to the end of it, much less a plain man." (Chapter 2 Graham 1981:56)

Zhuangzi's line here relates closely to arguments we find in the Mohist dialectical chapters. They pointed out that eschewing distinctions is unacceptable. "Denying denial is perverse." (Canon B79)� The Mohists also rebutted, on coherence grounds, a closely related anti-language view."� "To take language as exhaustively perverse is perverse." (Canon B71, see Graham 1978:445-6) Their acceptance of the then common analysis of language as consisting in distinctions unites these two paradoxes. Rejecting all distinctions is tantamount to rejecting language and vice-versa. Hui Shi, from the fragments available, tended to conclude from the relativity of distinctions to perspective that there were no distinctions in reality. Someone familiar with linguistic work would have good grounds for rejecting this inference. Zhuangzi's point here is that rejecting distinctions is implicit in the claim that "all is one." We cannot make it without making a distinction. "Having said, "all is one" have we managed to say anything?"� Zhuangzi saw that we cannot soundly draw an absolute conclusion about reality from relativist premises about words, distinctions and reference.

The Zhuangzi writer(s) show no signs of being relativist or pluralist about the world (i.e., committed to many metaphysically real worlds). The text, however, certainly expresses a descriptive norm-pluralism (many norms systems exist, including higher level norms for choosing/interpreting lower-level norms).� Moreover, it is skeptical (agnostic) about normative realism. He is unsure that one and only one of the existing norm systems is correct.

Skepticism about our ability to choose from the many available guiding systems and skepticism about our ways of distinguishing things are two manifestations of skepticism about daos. We get access to the reality that is relevant to acting through our daos or "comprehensive views." These "forms of life" include our norms of discrimination, of distinction, and of classification into "things" or "types" for purposes of action.

My conclusion is that the Zhuangzi shares with many writers of the period an implicitly naturalistic-monist conception of the world. However, it rejects the invitation to treat this monism as Parmenidean�as warranting a blanket denial of distinctions. However, speculation about what is ultimately real is not the central philosophical project of the period. [45] The Zhuangzi aims its skeptical attitude at conceptions of the world that inform our various, rival concepts of what counts as the good life--dao.

Skepticism: Theme, Scope and Strength. [46]

Skepticism is a complex epistemological stance. Following Goldman (1986), we can distinguish among many varieties of skepticism by focusing on its different themes, its scope and its strength. [47] We can use these distinctions with appropriate modifications in the different conceptual context of Chinese thought. To do so, we should avoid formulating them in terms of sententials or sententially linked concepts such as truth, belief, or inference. The appropriate counterparts would be distinctions or shi-fei^{this-not this} 是 非 assignments (approximating judgments), dispositions to treat as shi-fei^{this-not this} 是 非 (approximating thoughts) and appeal to standards (a 法 fa^standardor a 道 dao^guide) for treating as shi-fei^{this-not this} 是 非 (approximating inference).� These concepts underlie ancient accounts of knowing-to, knowing how-to, and contact-knowing (acquaintance).

Gaptooth asked Wang Ni.� 'Do you know what all things treat as是 shi^this:right?'� 'How could I know that?'� 'Do you know what it is you do not know?'� 'How could I know that'?� 'So in that case, does no one know anything?'� ' How could I know that? Still, let me try to state it - "How do I know that what I call knowing is not non-knowing? How do I know that what I call non-knowing is not knowing?"� HY 6/2/64-66.

Theme: non-epistemic goals v. doubt about epistemic achievement.

Several recent writers deny that Zhuangzi's is direct skepticism about our epistemic achievements�whether we have knowledge. They draw parallels to ancient Greek Pyrrhonian skepticism and construe Zhuangzi's theme as non-epistemic�as a posture aiming at some envisioned good. They aver that Zhuangzi's apparent skepticism is not really a negative thesis about our epistemic achievements but a recommendation about an attitude we ought to adopt (suspension of judgment) to get peace of mind. They justify this reinterpretation as a way to avoid an inconsistency in skepticism�the skeptic cannot know that he does not know. [48] If one claimed to know that no one knew, there would be a problem, but it can hardly be a problem with skeptical doubt as a philosophical attitude. The problem arises if the skeptic asserts the doubt in the form of a skeptical conclusion "no one knows anything" and acknowledges the norm of asserting that acknowledges the familiar challenge "how do you know that" as one that binds him to withdraw the assertion if he cannot vindicate that he does know it.� The content of the assertion itself is fully consistent.

The doubt illustrated in the Gaptooth discussion entails no inconsistency and, arguably, is carefully drawn to avoid it. Its root is worry that the standards for using terms like "know" and "don't know" could be other than they are. The anti-skeptical interpreters do not soundly derive the incoherent conclusions from things in the Zhuangzi and, even if they had, their derivation would not have been accessible to a Zhuangzi (i.e. he would not have formulated it in terms of 'truth', 'belief' or 'meaning'). The "irony" of their therapeutic solutions to the bogus problem is that they are plainly incoherent since the content of the judgment is "we should not make any judgment."� Further, its incoherence is precisely the one identified and criticized by the ancient Mohists and the Zhuangzi passage (D.1 above ). Both would easily spot the "perversity" in recommending that we not make judgments.

Last night Chuang Chou dreamed he was a butterfly; spirits soaring he was a butterfly (is it that in showing what he was he suited his own fancy?), and did not know about Chou. When all of a sudden he awoke, he was Chou with all his wits about him. He does not know whether he is Chou who dreams he is a butterfly or a butterfly who dreams he is Chou. Ch. 2 HY 7/2/94-6 [Graham]

These accounts of suddenly seeing things differently remind us that the challenge of correctly wei-ing has a "metaphysical" dimension filtering through our background categorizations. Zhuangzi the hunter in a grove feels confident (and smugly superior) as he stalks a bird that is also engaged in stalking a mantis that is about to pounce on a cicada. Little does each know he is the hunted as well as the hunter. When the "keeper" surprises Zhuangzi and chases him from the grove, he sees his confidence and sense of superiority as foolishness. He recognizes his likeness to the chain of prey he had been observing. An unappreciated feature of the context made him suddenly reclassify himself and other things differently for purposes of guiding action.

The Zhuangzi also targets norms of judgment that seem beyond challenge. We normally "know to do" things that keep us alive, but some ways of life (conceptions of the good life) sometimes treat accepting death as rational. Is this an objection to them?� The Zhuangzi asks, "How do I know that avoiding death is not a similar mistake?"� Without knowing that I should value life, I cannot claim to know to do any of the things in which pragmatic survival (individual or social) is a motivating standard. Zhuangzi's conversation with the skull illustrates by portraying even this judgment as one that we could not motivate from the rival point of view.

When Chuang Tzu went to Ch�u, he saw an old skull, all dry and parched. He poked it with his carriage whip and then asked, �Sir, were you greedy for life and forgetful of reason, and so came to this? Was your state overthrown and did you bow beneath the ax and so came to this? Did you do some evil deed and were you ashamed to bring disgrace upon your par�ents and family, and so came to this? Was it through the pangs of cold and hunger that you came to this? Or did your springs and autumns pile up until they brought you to this?�

When he had finished speaking, he dragged the skull over and, using it for a pillow, lay down to sleep.

In the middle of the night, the skull came to him in a dream and said, �You chatter like a rhetorician and all your words betray the entanglements of a living man. The dead know nothing of these! Would you like to hear a lecture on the dead?�

�Indeed,� said Chuang Tzu.

The skull said, �Among the dead there are no rulers above, no subjects below, and no chores of the four seasons. With nothing to do, our springs and autumns are as endless as heaven and earth. A king facing south on his throne could have no more happiness than this!�

Chuang Tzu couldn�t believe this and said, �If I got the Arbiter of Fate to give you a body again, make you some bones and flesh, return you to your parents and family and your old home and friends, you would want that, wouldn�t you?�

The skull frowned severely, wrinkling up its brow. �Why would I throw away more happiness than that of a king on a throne and take on the troubles of a human being again?� it said. Ch. 18� HY46/18/22-9 [Watson]

Other species -based doubt expands on the theme. Clearly, the extinction of humans would seem threatened species not to be such a bad thing.

Mao-ch'iang and Lady Li were beautiful in the eyes of men; but when the fish saw them they plunged deep, when the birds saw them they flew high, when the deer saw them they broke into a run. Which of these four knows what is truly beautiful in the world? In my judgment the principles of Goodwill and Duty, the paths of "That's it, that's not", are inextricably confused; how could I know how to discriminate between them?'� Ch. 2 HY 6/2/69-70 [Graham]

Dream passages illustrate the theme by showing how, in dreams, we classify others and ourselves in one way with little hesitation then on awakening, abandon that classification instantly. We cannot convince the "dream persona" that it is wrong.�� Am I Zhuangzi dreaming being a butterfly or a butterfly dreaming Zhuangzi? Do we know how to treat the dream context?� Some ways of life treat dreams as important identifications of one's religious status, some as prophesies, some as revelations of deep structures of desire, etc. Are we right merely to dismiss what we "knew" during the dream?

The main target of doubt in these passages is conventional wisdom, especially our shared, conventional patterns of discrimination or distinction making. A key claim is that any time we make a discrimination, we fail to see something (as Zhuangzi fails to see the gamekeeper while discriminating the bird as unwitting prey).

That the scope is broad, however, does not mean that it is total. It does not, as we noted above, signal any doubt about the real-world context in which our ways of life operate nor that the ways operate in the same world nor that we have different ones and disagree about them.

Besides reminders of how we can go wrong, the stories mainly underline the irresolvability of our disputes about different ways of life and the difficulty of appreciating alternatives.� They remind us that our confidence in our own comprehensive view is neither reliable nor unique to us. The text naturally implies that we are to learn (come to know) something from these stories. This knowing about knowing is valuable. Ignoring it would bad from many, if not all, points of view.

Broad but not total skepticism of this type can support a normative conclusion � as it does for modern liberalism. The conclusion would be that we ought not to participate in social structures designed to oppress, mutilate, or kill those who do not share our way of life. Engaging in coercive activities to impose a single way of life on people requires a level of confidence that we have identified the correct comprehensive vision to which we are not entitled.� The Zhuangzi passages remind us of that we have frequently been wrong in our classifications and action attitudes toward things.

Strength

The strength of a skeptical doubt is independent of its scope. A very broad skepticism might still be weak and a narrow skepticism might be quite strong. The stronger the demand the skeptic requires to justify a claim to knowledge, the weaker the skeptical considerations have to be to support the judgment that we do not or cannot have it.� There are various dimensions of strength.�

a)      Doubt of perfect knowledge or any knowledge?

Denying that we can be certain of something is a weak version of skepticism because 'certainty' is a strong demand. Similarly, denying that we now know X now weaker than modal versions, which deny any possibility of knowing X. The Gaptooth discussion suggests that the Zhuangzi position is weak in that it allows we might in fact know something, but do not know that we know it. [56] Thus, it does not commit us to the claim that knowledge is impossible.�

The passages under discussion do not use notions like 'certainty', but the skepticism still looks weak in a vaguely similar sense. The surprise and dream passages suggest that skepticism about how to act is comparative. We may know a way to do a thing, but not plausibly claim to know the way or the best way to do it. There are always possibilities we have not yet even contemplated. We learn from our experiences of coming to realize a better way to view and act that we typically tend to have too much confidence that our way is the way of doing something.

Consider the story of the protective hand-salve used by its inventor for bleaching silk, which was later adapted to a military use by an enterprising stranger. (Ch. 1) It worked as well or better to that purpose and brought the entrepreneur greater status and monetary rewards. The inventor did know a way to use of the salve, but he clearly did not know the correct use. A better use, one that he had never even thought of, was available.

The dream sequences make a similar point. We learn that one way of life is not correct when we "awaken" to one that seems better. Each such "awakening" is a kind of knowing--that our past confidence was misleading about our epistemic status. We know then that we did not know before what we thought we knew. The repeated experience of coming to know should make us cautious, modest about, or present state of knowledge. Even when our prior ways have seemed to work fine, they have turned out to come to seem inferior to some alternative. Typically, we had not even imagined the alternative.

The guidance-skepticism is thus weak in that it denies mainly that we can ever have perfect, ultimate, the highest or best knowledge. The demand is a strong one, so the skepticism is weak. We cannot intelligently treat or regard our way of doing things as unqualifiedly best. This is, of course, precisely the kind of dao skepticism that justifies neutrality among comprehensive views and encouraging the spread of different daos. However, in making this point, the skeptic acknowledges that we can come to realize (know) that one dao is better than another. We simply cannot conclude that with this realization we have found the way.

Again, notice that this skepticism does not arise from any analysis of our representations or of memory v. some reality.� It takes both memory and our past for granted in appealing to our common past experiences of coming to see as dubious what we had thought obvious.

One may object that the skepticism is so weak that it really has no opponents. However, the tradition defends us from accusations of triviality. We find many claims in Confucian [57] and other authoritarian (Proto-legalist) writing that "sages" are capable of transcending the ordinary limited perspectives and achieving something like total or fully rounded knowledge. In any case, the interpretative claim is not trivial since the dominant rival interpretation is that Zhuangzi himself claims access to Confucian-like perfect knowledge. Traditional interpretations take the skill stories as endorsing just such a claim. [58]

The Zhuangzi also observes that skillful performers are all limited to their skill. Inevitably, other things they cannot do as well.� Perhaps their focus in gaining the skill hinders them from getting some other valuable ability. Typically, the stories note that the skillful practitioner cannot convey his skill to his son. A musician cannot play all notes at once; a debater cannot convince everyone etc. He sums up by noting that 成cheng^completion always entails 虧kui^defect.

May men like this be said to be complete? Then so am I. Or may they not be said to be complete? Then neither am I, nor is anything else. Ch. 2 [Graham] HY 5/2/46-7

The no-completion-without-defect insight reminds us that no matter how much we perfect our skill, we will be unable to do other things. This is rooted in the linguistic point.� Anytime we learn to make and follow a distinction, there will be things we are unable to appreciate or respond to. To make a distinction is to choose not to make other possible distinctions in the context.

Hence, divisions have that which they do not divide; distinctions have that they do not distinguish. You say, "What does this mean?" Sages embrace things, ordinary people discriminate things so they can see them in relation to each other; so I say, those who distinguish have that they do not see. Ch. 2 HY 5/2/57

Normally, we regard our way of life as better than the ways of others. We might even be right about it. However, the skeptical passages warn us that our evaluation to this effect is suspect. It may be question-begging because others may not share our standards of evaluation. We might accidentally be right, but we cannot conclude that we are so merely from the strength of our conviction, nor can we merely because we appeal to higher standards. A rival dao comes complete with its own different norms of warrant.

Is man's life really as stupid as this? Or is it that I am the only stupid one, and there are others not so stupid? But if you go by the completed heart and take it as your authority, who is without such an authority? Why should it be only the man who knows how things alternate and whose heart approves its own judgments who has such an authority? Ch.2 [Graham] HY4/2/20-2

We would be unwise to ignore that the person with the rival form of life probably finds hers superior on functionally similar grounds. That is, though she does not appeal to the same norms, she does appeal to norms of warrant and they may justify and convince her in as complex and complete a way.� She may equally cite her dao's success in the actual world as judged by those norms. She too will have a skill, albeit not the same as ours. The crucial skeptical realization is that we can no more easily convince her that our way of life is better than she can us.

I referred to how dream passages illustrate the effect of conversion from one comprehensive view to another.� When we are "within" one way of life, we find it obviously correct.� When we have awakened to a different way, we equally clearly realize how we were wrong before. We become as strongly convinced of our new view [59] That we can so easily come to see things "in a new light" should lead us to realize how real the danger is that we may be misled merely by the reflective elaboration of a dao we accept.

How should the philosophically alert reader then make sense of the interspersing of these skeptical passages with talk about "perfected" people and absolutes? What they might find striking about such passages is that they often suggest that such perfection is unintelligible and irrelevant to us. [60] Unintelligible, some examples suggest, even to the highest religious sage�the Yellow emperor (HY 6/2/71-6).�

Nieh Ch'ueh said, "If you don't know what is profitable or harmful, then does the Perfect Man likewise know nothing of such things?" Wang Ni replied, "The Perfect Man is godlike. Though the great swamps blaze, they cannot burn him; though the great rivers freeze, they cannot chill him; though swift lightning splits the hills and howling gales shake the sea, they cannot frighten him. A man like this rides the clouds and mist, straddles the sun and moon, and wanders beyond the four seas. Even life and death have no effect on him, much less the rules of profit and loss!

. . .

The sage leans on the sun and moon, tucks the universe under his arm, merges himself with things, leaves the confusion and muddle as it is, and looks on slaves as exalted. Ordinary men strain and struggle; the sage is stupid and blockish. He takes part in ten thousand ages and achieves simplicity in oneness. For him, all the ten thousand things are what they are, and thus they enfold each other. Ch. 2 [Watson] HY� 6/2/71-6

Typically, this kind of text makes its point by asserting a kind of reversal. Perfection may be the very opposite of what we now take it to be. Perfection from one perspective may appear as stupidity from another! The perfect musician may be one who does not play a note(?!). [61] � The judgments and behavior of a perfect man are just irrelevant, incomprehensible to us and of no use when we, ordinary people, think about what to do.

Our lives are bounded and knowledge is unbounded. To use the bounded to pursue the unbounded is dangerous�stop it and you can be deemed a knowledgeable person. Ch.3 HY 7/3/1

We can also read passages like "A man like this rides the clouds and mist," [62] as contemplating what might be possible with different daos. Some comprehensive way of life that is possible for us, might give us powers over nature and our lives that we can now barely conceive of.� That is another sense that should lead us to doubt that we now know how to live. There almost certainly are far better possible ways to organize our lives, whose superiority we could appreciate even using our present norms.

(1)   Arguments

Besides theme, scope and strength, we can classify dominant arguments for skepticism. Here I will strategically focus on the two most relevant to understanding skepticism in the Zhuangzi. One line of argument draws on the fallibility of human cognitive faculties the other on the question of the grounds or warrant. Clearly, my analysis is that the core argument for skepticism in the Zhuangzi is of the latter kind. The skepticism appears to flow mainly from the dependency analysis of knowing-to as presupposing standards and which opens up a regress of knowing-to use certain standards and how to use them.

The most common alternative in the literature on Zhuangzi's skepticism is human fallibility�particularly of the senses. Do these passages suggest that the ground of the skepticism is a view of the limitations of the human mind (Say as contrasted to the transcendent supernatural)? Aside from the implicit claim that the judgments of a perfect being are irrelevant to our deciding what to do, there is scant evidence in this texts or the tradition for any negative analysis of human cognitive psychology. [63] The main limitation we face is "our lives are bounded."� We find none of the familiar appeals to relativity or fallibility of sense perception.

The Zhuangzi does sometime makes its skeptical point in physiological terms. It refers to a 成cheng xin as a physiological tendency that respond to things with shi-fei^{this-not this} 是 非. It "grows" in the heart just as the heart grows in the body. The point there, however, simply recapitulates the regress-of-norms argument. All shi-fei^{this-not this} 是 非 emerging from the heart rely on having some other shi-fei^{this-not this} 是 非 already present there. As I have argued (Hansen 1992), the discussion of the heart approximates (i.e., translates with appropriate conceptual adjustments) the Humean skeptical point "you can't get an ought from an is." The heart-mind's judgments are dependent it its having absorbed the shi-fei^{this-not this} 是 非 presuppositions of a "way of life."