Having a Word with Angus Graham: At Twenty-Five Years Into His Immortality 1438468555, 9781438468556

Citation preview

HAVING A WORD WITH ANGUS GRAHAM

SUNY series in Chinese Philosophy and Culture ———­­—— Roger T. Ames, editor

HAVING A WORD WITH ANGUS GRAHAM At Twenty-Five Years into His Immortality

Edited by

CARINE DEFOORT AND ROGER T. AMES

Cover photo of Angus Graham courtesy of Dawn Baker Published by State University of New York Press, Albany © 2018 State University of New York All rights reserved Printed in the United States of America No part of this book may be used or reproduced in any manner whatsoever without written permission. No part of this book may be stored in a retrieval system or transmitted in any form or by any means including electronic, electrostatic, magnetic tape, mechanical, photocopying, recording, or otherwise without the prior permission in writing of the publisher. For information, contact State University of New York Press, Albany, NY www.sunypress.edu Production, Diane Ganeles Marketing, Anne M. Valentine Library of Congress Cataloging-in-Publication Data Names: Defoort, Carine, 1961– editor. Title: Having a word with Angus Graham : at twenty-five years into his immortality / edited by Carine Defoort and Roger T. Ames. Description: Albany, NY : State University of New York, 2018. | Series: SUNY series in Chinese philosophy and culture | Includes bibliographical references and index. Identifiers: LCCN 2017015616 (print) | LCCN 2017059539 (ebook) | ISBN 9781438468563 (ebook) | ISBN 9781438468556 (hardcover : alk. paper) Subjects: LCSH: Graham, A. C. (Angus Charles) | Zhuangzi. | Philosophy, Chinese. Classification: LCC B1626.G74 (ebook) | LCC B1626.G74 H38 2018 (print) | DDC 181/.11—dc23 LC record available at https://lccn.loc.gov/2017015616 10 9 8 7 6 5 4 3 2 1

Contents Introduction 1 Having a Word with Angus Graham: At Twenty-Five Years into His Immortality Carine Defoort and Roger T. Ames   1 Reading the Zhuangzi Anthology Esther S. Klein

11

  2  Reflections on Textual Analysis in the Post-Graham Era Liu Xiaogan

27

  3  Cognitive Attunement in the Zhuangzi Harold D. Roth

49

  4  Vital Matters, A. C. Graham, and the Zhuangzi Michael Nylan

79

  5  Remarks on Intertranslatability and Relativism Henry Rosemont Jr.

99

 6  Getting to the Bottom of “Things” (wù 物 ): Expanding on A. C. Graham’s Understanding 111 Robert H. Gassmann   7  Míng (名) as “Names” Rather than “Words”: Disabled Bodies Speaking without Acting in Early Chinese Texts Jane Geaney

137

  8  Unfounded and Unfollowed: Mencius’s Portrayal of Yang Zhu and Mo Di Carine Defoort

165

vi | Contents   9  Reconstructing A. C. Graham’s Reading of Mencius on xing 性: A Coda to “The Background of the Mencian Theory of Human Nature” (1967) Roger T. Ames

185

10  Reason and Spontaneity Reconsidered Lisa Raphals

215

11  Spontaneity and Marriage Paul Kjellberg

231

12  Rationalism and Anti-Rationalism in Later Mohism and the Zhuāngzǐ Chris Fraser

251

About the Contributors

275

Index

279

Introduction

Having a Word with Angus Graham At Twenty-Five Years into His Immortality

Carine Defoort and Roger T. Ames

When people die, they also live on in a variety of ways. This phenomenon is even more apparent in the case of great scholars and grand personalities. When Angus Charles Graham left us in March 1991, the most obvious way in which he was bound to survive was through his daughter, Dawn, and his two grandchildren, Calum and Holly, all of whom will carry on some of his genes. But since humans are storytelling animals, we also live on in the narratives woven around the events of our lives and personalities. A widely-known account of Angus Graham can be found on Wikipedia: it starts with his birth in Wales (1919), his studies of theology at Oxford, then his reading of Chinese at the School of Oriental and African Studies in London, his subsequent appointment as lecturer in Classical Chinese at SOAS (1950–1984), followed by visiting positions at various universities, including at Michigan, Hong Kong, Cornell, Brown, and Hawai’i.1 With an Englishman as wonderfully idiosyncratic as Graham, there is inevitably a wealth of unofficial stories: how he eluded a V-2 rocket during the Second World War by simply walking away from a boring companion just the moment before it struck, how he “got the year wrong” for his visiting appointment at Brown University, how he claimed to “hate Zhuangzi” in order to avoid a teaching assignment, and how he nevertheless agreed to hold a weekly class on the same subject under a banyan tree at Waikiki beach.2 A third way that one remains alive is through rituals: the yearly Angus Graham Memorial Lectures organized at SOAS since 2010 can be seen as a ceremonial occasion

1

2  |  Carine Defoort and Roger T. Ames to keep his memory alive in a continuing series of lectures and workshops that have been inspired by his work.3 But aside from blood ties, narratives, and rituals, the most obvious type of immortality achieved by scholars lies with their publications. Over a professional career of more than thirty years, Graham produced an impressive amount of scholarship on a wide array of topics, ranging from grammar and philology to poetry and philosophy. Among his most wellknown monographs are The Book of Lieh-tzu (1960), Later Mohist Logic (1978), the Chuang-tzu (1981), and Disputers of the Tao (1989).4 His combination of rigorous scholarship and philosophical originality has continued to inspire scholars to tackle related research topics and, in so doing, has required of them to respond to his views. Thus, Graham’s last and, perhaps for him, most gratifying version of longevity would lie not only in the enduring value of his own publications, but in the work of others who have been inspired by him. The various fields within sinology, including the history of Chinese thought, Chinese philosophy, Chinese philology, and the art of translation, are still replete with scholarship elaborating upon, disagreeing with, or reacting to Graham’s work. We can fairly claim that Graham’s ideas have become fundamental to the way in which Chinese philosophy is now read within the corridors of the Western academy. March 2016 was the twenty-fifth anniversary of Graham’s death. As a small contribution to his immortality, we invited some of his colleagues, friends, students, and admirers to continue the conversation with this grand old gentleman by sharing some of their current research as it has been inspired by his work. One of Graham’s self-declared “hobby horses” was the topic of spontaneity in Chinese philosophy in which he saw a novel solution to the Western fact/value dichotomy. Graham began to elaborate on spontaneity in an early monograph, The Problem of Value (1961), gave it a full reconsideration in his Reason and Spontaneity (1985), and ended up bumping into this topic wherever he looked, claiming: A point of interest in the Chinese tradition is that, various as it is, it seems everywhere to start from the assumption, quite foreign to at least the Kantian tradition in the West although familiar to common sense, that the ultimate springs of action are in spontaneous preference the value of which depends on the wisdom of the agent.5 This quote along with his tenacious interest in the topic of spontaneity strongly suggest that what Graham had found in the Chinese texts was himself. Whatever he read, even a text as neglected and corrupt as the Heguanzi, taught him something that he recognized as both very familiar and yet

Introduction | 3 philosophically intriguing: spontaneity. On an academic level, he proposed increasingly subtle reflections on this topic, thus engendering debates that have moved other scholars to join him, including some of the contributors to the present volume. At the same time, the topic was consistent with Graham’s own idiosyncrasy and his resolute conviction that, as a person as well as a scholar, he ought to follow his own preferences. Emotional support for this deeply felt commitment and his continuing theoretical reflections on the legitimate limitations of preferences (for example, the duty to enhance one’s awareness of any situation to the fullest) brought together the combination of rigorous thinking and personal intuition that characterizes much of Graham’s work. His hypothesis that Confucius’s changing views can be appealed to as explanation for inconsistencies in the Lunyu, his monumental reconstruction of later Mohist logic, and his reconfiguration of the book Zhuangzi are some examples of this peculiar combination of careful analysis and bold speculation. For the authors of this volume, Graham was a model of rigorous scholarship and creative reflection: each of the essays included herewith contains an original contribution on some specific topic that Graham had once worked on, from linguistic and textual matters to philosophical issues. We also encouraged the contributors to follow their hearts: to share their recollections, to ride their own hobby-horses, to rely on their own assumptions, to use their own orthography, and to disagree with each other, with us, and with Graham. The result is a kaleidoscope of twelve essays on spontaneity, the Zhuangzi, human nature, textual criticism, translation, uncommon assumptions, the use of metaphor, and much more. The essays do not divide neatly into textual matters and philosophy, but if there is a discernable pattern, it is perhaps in the gradual evolution of Graham’s interests from Chinese grammatical and textual matters to formal philosophy itself. Graham took painstaking scholarship on textual formation that called upon philological skills and philosophical insight as the ground of his own more speculative work. Indeed, one of his most enduring contributions was his reconstruction of texts such as the later Mohist canons, the Zhuangzi, and the Heguanzi. It is because of this commitment to evidential scholarship that we begin this collection of essays with the sustained reflections by Esther Klein and Liu Xiaogan on his reconstruction of the Zhuangzi, a project that Graham himself found exasperatingly inconclusive on most fronts, and who on many occasions insisted that he was “done with it.” Inspired by archaeological finds and recent scholarship, both scholars identify hidden assumptions in the field and nevertheless reach diametrically opposed conclusions. Continuing her earlier findings concerning the possibly late date of the Inner Chapters, Klein questions current expectations of authorship and textual unity. She argues that the received Zhuangzi is best read as one version of a

4  |  Carine Defoort and Roger T. Ames concatenated anthology belonging to a continuing lineage grouped under a single name rather than as the product of any particular author or identifiable period. While Graham saw himself as engaging in the reconstructive project of untying philological and philosophical knots in order to restore the more coherent “original” text, Klein is more interested in acknowledging uncertainties about authorship and attribution and in uncovering more or less explicit “castles in the air” that have, to some extent inevitably, constituted a major part of the currently dominant Zhuangzi account. Her insights into different possible kinds of textual production have many ramifications for how we read and understand these early canons. Liu Xiaogan agrees with Graham on the integrity of the Inner Chapters but disputes the chronological priority of the Zhuangzi and his reconstruction of the Outer and Miscellaneous Chapters. Appealing to recent archaeological discoveries, Liu challenges what he identifies as four hidden assumptions in the work of Graham, D. C. Lau, and Qian Mu, namely, that the paucity of extant texts can fully represent the whole ancient body of texts, that speculations are more reliable than imperfect historical records, that historical records must either be judged reliable or be discarded, and that small samples can be reliably used as evidence for general conclusions. Liu argues that, while absolute certainty is impossible, in order to draw more probable conclusions in textual studies, we have to reexamine these assumptions. His point is not that all early records are to be endorsed or that skeptical reasoning should be discouraged, but rather that we need to avoid mere speculations and critically construct a methodology combining the best findings of existing scholarship. Since Liu explicitly defends a position that implicitly still dominates the field, it is most appropriate that it be spelled out in this volume. Harold Roth and Michael Nylan take us from this discussion on what we can learn from the structure of the Zhuangzi to its philosophical import. Roth introduces the notion of “cognitive attunement” and advances his claim that this is in fact the main thematic of this composite text. He interprets the Zhuangzi via the intersection between cognitive science and a phenomenology of the varieties of contemplative experience advocated as a methodology by the newly established academic field called contemplative studies. Invoking William James’s “pure experience” and the “tacit understanding” derived from cognitive science research, Roth argues for a cultivated cognitive and perceptual experience. He appeals to an “effortless attention” that is spontaneous and nonintentional, and that can lead to “cognitive attunement” in a first-, second-, third-, and “no-person” or nonattached experience. Roth mounts a sustained textual argument that this transformed free and flowlike non-self-consciousness is the goal of the contemplative practices around which the Zhuangzi as a text has been woven. Michael Nylan focuses her analysis on how to characterize the kind of optimal experience advocated by the Zhuangzi. Nylan reiterates what

Introduction | 5 we observed earlier in this introduction: Graham imports and ascribes to Zhuangzi his own philosophy of life in which a hyperawareness enables one to act spontaneously and with optimal vitality in realizing the potential that the human experience brings. This being the case, Graham’s Zhuangzi seems to advocate for “heightened awareness” as a kind of superhuman omniscience that we might associate with sageliness and that, in her eyes, is more closely tied to cognitive reach than to emotions and intuition. Nylan takes exception to such a reading of Zhuangzi and proposes as a rather intriguing alternative to “be aware,” the injunction “to be fully present.” Zhuangzi with his xiaoyaoyou (逍遙遊), the oft-referenced expression used to capture the optimal Zhuangzian experience, recommends the cultivation of an attitude that not only recognizes but also affirms our entanglements in the ordinary human condition, where clarity and vitality, far from leading to “free and easy wandering,” in fact characterize a keen awareness of the complexity of the human experience and the resolutely muddled state of mind that attends it. The life of optimal vitality comes with an appreciation of the gift of life and a curiosity about it that welcomes new, befuddling experiences and pleasure. It is the openness to the complexities of the ordinary rather than insight into the extraordinary that enables us get the most out of life and to live it fully and well. Moving away from the Zhuangzi, Henry Rosemont Jr., Robert Gassmann, and Jane Geaney engage Graham critically as the translator and interpreter of several texts and concepts. Rosemont attempts to illuminate Graham’s views on both language and thought by examining the latter’s criticisms of Rosemont’s own work on the relations between the two, along with his collaborator Roger Ames. According to Graham, they both appear to confuse translation and interpretation at times, which is odd, given that he has elsewhere praised Ames and Rosemont for the way they employ the distinction between language and thought, with results that Graham has elsewhere praised them for. Moreover, as a student of Noam Chomsky and a proponent of his universal grammar, Rosemont finds Graham to be inconsistent in at once accepting claims about the uniqueness of conceptual schemes while at the same time assuming that there must be objective truth standing above cultural differences. But at the end of the day, Rosemont opines that the identifiable disagreements among largely consistent exegetes might amount more to speaking past each other than having substantive differences in our understanding of the language and philosophy of classical China. In any case, during the preparation of this volume Henry left us to join Angus in the great transformation of things. Now there are two old friends with their boundless affection for each other and for us too, arguing the fine points, making merry, and waiting for us to join the happy company. Robert H. Gassmann is also concerned about how we understand and translate the basic terminology that defines the classical Chinese world. To

6  |  Carine Defoort and Roger T. Ames make his point, he takes one key term—wu (物), conventionally rendered “things”—that is given a concise explanation by Graham in his work on the later Mohist canons. Gassmann uses Legge’s rendering of the Chunqiu zuozhuan as representative to show how elsewhere a proliferation of different ways of translating this single term produces a “semantic jumble” that he considers to be a demonstration of “lexicography by assumed contextual fit”—that is, a willingness on the part of less than careful translators to tailor their understanding of terms in order to make sense of particular passages. Through a careful application of analytical grammatical and lexicographical methods, Gassmann proposes a significantly alternative understanding of wu. He argues that the term references an “aggregate” or “assemblage”— that is, a variously structured union or group of two or more objects that can further be delineated into subtypes. Gassmann meticulously translates relevant passages, analyzes their contexts, and sets up the parameters of the meaning of the term to produce the evidence for his claim. Jane Geaney elaborates upon Graham’s insights into how “sounds” or utterances of “naming” (ming 名) function within an early Chinese nominalism, and upon her own work that distinguishes audible names from visible action. Making a distinction between detached and immersed views of language, she argues that discussions about “names” in early Chinese texts do not constitute linguistic theories about abstract language but reflect on speech as utterances in contextualized actions and grounded in situations. Body and, more specifically, disabled bodies enter the discussion because of the parallel in the literature between not knowing a name and a blind person’s not knowing color, or a lame person’s not being able to walk a long distance. In examining the relevant passages from the early texts, she finds that a crucial feature of reliable knowledge is the complementary relationship between sound and sight, and that a necessary condition of “naming knowledge” is missing when no visual action accompanies the sounds that constitute naming. These debates on interpretation bring us to the essays of Carine Defoort and Roger Ames that take the Mencius as a focus of reflection. Starting from a well-known passage in which Mencius criticizes Yang Zhu’s supposed egoism and Mo Di’s extreme altruism, Defoort argues that this assessment may to a large extent have been the result of Mencius’s rhetorical imagination rather than an account of the two actual figures. On the basis of the little textual evidence that we have, she tries to reconstruct Mencius’s portrayal of both masters as an inventive response to existing arguments in terms of “weighing” priorities in life. Even though Mencius’s double portrayal was probably neither founded on historical evidence nor followed by contemporary masters, it became increasingly influencial after the Han dynasty, to the extent that it still dominates interpretations of the three masters: Mencius, Yang Zhu, and Mozi. As a result, we have failed to notice how exceptional

Introduction | 7 and unreliable it is as historical testimony. Even though Graham was open to alternative readings of Yang Zhu’s portrayal, he also took part in the confirmation of the traditional Mencian picture. Graham’s Mencius scholarship is also the topic of Roger Ames’s essay. Ames celebrates Graham’s openness to continually question the received wisdom and to revise his own, always tenative conclusions. Beginning from a seminal article on the Mencian notion of “human nature” (xing 性) that Graham first published in 1967, Ames traces Graham’s evolving interpretation of this key philosophical idea as it continued to develop over his long career. Ames largely allows Graham to speak for himself, citing Graham’s own published work that over time evidences a growing sensitivity to and a deepening understanding of the cosmological assumptions that constitute the interpretive context within which these early texts must be read. For Graham, xing as a dynamic, gerundive concept references a process that is spontaneous and realizes its own potentialities when it is nourished and unimpeded. But taking his interpretation a step further, rather than assuming a doctrine of external relations where relations merely conjoin putatively discrete and independent “things,” Graham endorses an understanding of early Confucian cosmology that assumes a doctrine of internal relations where relations are themselves constitutive of “events.” Thus, his interpretation of xing’s “own potentialities” would make them radically contextual, historicist, particularist, and emergent—that is, Graham offers us an embedded, narrative understanding of human nature rather than the familiar ontological or developmental models associated with the idealism and teleology most often attributed to Mencius. Moving further toward Graham’s growing philosophical preoccupation with spontaneity as a key to how we might best think about moral action, we have reached the three remaining essays in this volume by Lisa Raphals, Paul Kjellberg, and Chris Fraser, all of whom explore Graham’s foray into formal philosophy with his commitment to spontaneity as an integral factor in achieving moral competence. Raphals sets out to track down the argument that motivates Graham’s Reason and Spontaneity written relatively late in his career, and that also provides structure for Disputers of the Tao. The argument is captured succinctly in what Graham calls his quasi-syllogism: In awareness of everything relevant to the issue (= everything which spontaneously moves me one way or the other), I find myself moved towards X, but overlooking something relevant I find myself moved towards Y. Be aware. Therefore, let yourself be moved towards X (= choose X as end).6

8  |  Carine Defoort and Roger T. Ames Graham rejects both Kantian rationalism and romantic irrationalism in favor of a notion of self in which awareness achieves an integration of reason with spontaneity. Because Graham’s argument is both philosophical and empirical, Raphals responds to it through the lens of recent arguments in philosophy and research that has been undertaken in psychology and biology. In her first section she argues that Graham’s account of inclination informed by awareness (rather than reason) as the basis for agency and the choices made do not require an account of Zhuangzi as “anti-rationalism.” She then turns to his empirical argument and defends it by appeal to recent research on the biology of agency. Having brought not only focus but also additional support to Graham’s argument, Raphals ends by critiquing his overdrawn distinction between humans and animals that puts him at odds both with the Zhuangzi and contemporary science. Paul Kjellberg thinks through the seeming tension between the vows of constancy that ground the institution of marriage, and the spontaneity and openness to new experience advocated by the Zhuangzi and firmly embraced by Graham. Kjellberg follows Graham through the Humean observation that no account of the facts by itself determines what people should do on the basis of those facts. Kant agrees with Hume’s “is/ought” distinction and brings reason into the discussion as an a priori foundation that can serve as a categorical ground for what is right and what people ought to do. Graham, and others, conclude that Kant’s rationalism has failed in only giving us what morality would have to be if it exists, rather than proving that it in fact does exist. This leaves us with what Graham calls “irrationalism” as the pure subjectivism of the romantic for whom values are purely volitional, and Graham’s own “anti-rationalism” as spontaneity in our inclinations (rather than pure rational deliberation) informed by a full awareness of the circumstances. He thus makes spontaneity the underlying foundation for both a more limited and contextualized function of reason and for morality itself. Kjellberg concludes that Graham’s “grounded rationality” is not telling us to do anything differently, but rather by illuminating the way that we choose, he means to help us do what we have done all along better, that is, with more intelligence and greater awareness. It is this more subtle reading of Graham’s project that for Kjellberg ultimately holds the key to reconciling spontaneity and marriage. Chris Fraser brings our volume to a close with a delightful dialogue between a reluctant and critical Zhuang Zhou and the “anti-rationalist” Graham, who would include Zhuangzi as his fellow sojourner. As Kjellberg has done, Fraser challenges the appositeness of Graham’s not always helpful distinctions—rationalism and anti-rationalism to begin with. Fraser in the persona of Zhuang Zhou makes a parody of the kind of simple rationalism that is sometimes assumed in Graham’s own reasoning and in his charac-

Introduction | 9 terizations of the early doctrines. Fraser summons Hui Shi, the Mohists, and Yang Zhu, too, in requiring that Graham provide a clearer and more nuanced account of spontaneity as the putative antithesis of reason. The disputes are sometimes quite technical, but at every step, Fraser allows Graham to speak for himself and is gracious even where he thinks Graham has led us astray. At the end of the day, Fraser allows Graham to save himself by redescribing his characterization of Zhuangzi as neither a rejection of rationality nor an endorsement of impulsive, spontaneous responses, but rather as someone who recommends an uncodifiable, adaptive virtuosity. Fraser then really speaks on behalf of all of the authors of this volume by thanking Graham for requiring that we abandon fixed interpretations of these early texts and philosophize for ourselves.

Notes 1. For more information, see also the bibliographical note by Henry Rosemont in Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus C. Graham, ed. Henry Rosemont (La Salle: Open Court, 1991), xii–xiii. For a more personal account, see A. C. Graham, Unreason within Reason: Essays on the Outskirts of Rationality (La Salle: Open Court, 1992), 1–6. 2. For some of those stories, see the interview by Connor Walsh made at SOAS, http://www.coffeeflavouredtea.net/wordpress/?p=1898. 3. Since the inauguration of the memorial lectures, the speakers have been Robert H. Gassmann (2010), Liu Xiaogan (2011), Carine Defoort (2012), William Nienhauser (2013), Roger Ames (2014), Michael Nylan (2015), and Christoph Harbsmeier (2016). Their lectures can also be accessed through the URL in note 2. 4. For a close to complete bibliography of his work edited by Henry Rosemont, see Chinese Texts and Philosophical Contexts, 323–328. 5. Graham, “A Chinese Approach to Philosophy of Value: Ho-kuan-tzu,” in Unreason within Reason, 121–135, esp. p. 121. The essays gathered in this posthumous volume attest to this lifelong interest in this topic. 6. A. C. Graham, Reason and Spontaneity: A New Solution to the Problem of Fact and Value (London: Curzon Press, 1985), 7.

1

Reading the Zhuangzi Anthology

Esther S. Klein

Introduction Scholars of the Zhuangzi have occasionally raised the question of authorship. The fragmentary nature of some parts of the text, as well as various idiosyncrasies and internal contradictions, could lead to the hypothesis that it was produced by multiple hands over more than one generation. This possibility was explicitly raised as early as the Song dynasty (960–1279),1 and was discussed by Ming (1368–1644) and Qing (1644–1911) scholars as well.2 Among modern scholars, A. C. Graham and Liu Xiaogan have researched this problem most extensively.3 Liu Xiaogan has generally defended the position that the entire Zhuangzi is the work of the historical Zhuang Zhou and his students or later followers, and that it has not suffered drastic editing or textual corruption since its original conception—merely some slight rearrangements.4 A. C. Graham, on the other hand, made distinctions among different philosophical lines within the text, arguing not only that they came from multiple authors but also that these authors belonged to different schools of thought.5 Both Graham and Liu have accepted the widely held belief that the first seven Inner Chapters (neipian) are, and have since the beginning been, distinct from the rest of the book, and that they are attributable to the historical Zhuang Zhou. This study begins by examining in detail Graham’s project of textually reconstructing the Zhuangzi according to his understanding of its authorship and analysis of its sometimes contradictory philosophical ideas. I argue that Graham’s philosophical and textual insights led him to conclusions that pose problems for his original assumptions. I then outline a general approach to

11

12  |  Esther S. Klein investigating the authorship of the text, as derived from recent work of several different scholars. This approach is one that questions traditional suppositions about the authorship of the Inner Chapters as they have come down to us today. Finally, I end with an outline of the motivations, assumptions, and methods that lie behind my study of the authorship of the Zhuangzi, together with some of my major findings. I will also argue that terminological evidence is not as persuasive or useful as it might seem, although the deployment of this and other types of evidence depends largely on one’s goals in studying the Zhuangzi. I conclude that it is most useful to prioritize the ideas within the text rather than the exact historical circumstances of their origination.

A. C. Graham’s Zhuangzi and the Authorship Problem In Chuang-tzŭ: The Inner Chapters, A. C. Graham describes the first seven chapters of the Zhuangzi as “homogeneous in thought and style and generally recognized as substantially the work of Chuang-tzu himself.”6 He also wrote an essay investigating the authorship of the Zhuangzi, wherein he acknowledged some challenges to the conventional attribution of the Inner Chapters but concluded that the attribution should stand.7 It is clear that Graham expended much time and effort on his study of the Zhuangzi, and therefore it is appropriate to take all the more seriously the sense of tentativeness and implicit frustration with which he presented his work. He prefaced his long and detailed scholarly study with the caveat that “the tests we propose are heterogeneous and of very unequal value” and expressed the hope that “future scholarship will develop a more sophisticated technique for dealing with [Chuang-tzu].”8 In the introduction to his translation, Graham complained that other translators “come to grief on a single basic error of policy,” that is, “they treat Chuang-tzu as though it were what is nowadays understood by a ‘book,’ ” while, in Graham’s view, “probably Chuang-tzu left behind only disjointed pieces, mixed up perhaps with his disciples’ records of his oral teachings.”9 In writing that the “ideal of integral translation is in this case meaningless,” Graham had in mind parts of the text he did not think of as “homogenous blocks” (i.e., sections other than the “Inner chapters and the Primitivist and Yangist sequences”). Nonetheless, according to his stated understanding of the authorial and editorial processes by which the Zhuangzi came to be, even the sequences he argued “must be presented complete” could be freely rearranged or supplemented. He admitted that “some of the worst problems are raised by the Inner chapters, which I am committed to translate entire.”10 One gets the sense that this commitment perhaps had more to do with publication conditions than with scholarly considerations.

Reading the Zhuangzi Anthology | 13 Graham characterized the Inner Chapters as a “collection of isolated episodes probably grouped together by a later editor,” including: Sequences of rhymed quatrains, stories in which speakers may burst into song, didactic verses with scattered prose comments, strings of aphorisms, provisional formulations of ideas followed by criticisms, propositions which Chuang-tzu (or a disciple perhaps) proceeds to annotate phrase by phrase.11 Graham made it clear that he believed an editor to be responsible for the current form of the chapters, but his presentation of the chapters shows that he also believed the chapters’ current state to be unsatisfactory—whether due to editorial shortcomings or later corruption is not made explicitly clear. Thus, in his translation he added material from chapters 14, 23, 24, and 32 to chapters 2, 3, 5, and 6, as well as rearranging material within many of the other chapters. As Harold Roth has pointed out, the scholarship justifying these rearrangements, though not transparent from the published version, is extremely solid. The only problem is that, as Roth puts it, “it is not clear how most of these textual dislocations actually took place”12—that is, whether there is a plausible theory of textual compilation underlying the project that Graham attempted. Authorship, as the word is generally understood, refers to more than mere generation of material. It usually also encompasses the ordering and plan of a work on both a large and a small scale, the narrative or logical flow, and the overall message. Graham’s approach to the Inner Chapters does not suggest a belief that Zhuangzi is responsible for any of these aspects, even within that core section. In short, even if one were to concede that Zhuang Zhou “authored” the Inner Chapters (to say nothing of the entire Zhuangzi), the authorial process Graham envisioned would be entirely different from what is meant by authorship today—so different that the use of the word “authorship” would arguably not be warranted.

Pre-Qin Authorship and the Zhuangzi Describing the formation of texts in the category of “Masters” literature (子 zi, or philosophical texts), Mark Edward Lewis wrote: The texts . . . survived only through active transmission and study, and they remained intellectually alive through the constant addition and adaptation of material. . . . The notion of authorship was weak. The writers of early texts would seldom, and the editors and transmitters never, have been the master

14  |  Esther S. Klein himself. Consequently, the master was always to some degree an invention of the text.13 Many scholars today subscribe to a similar theory of Warring States textual production, that is, generally presuming a weak or uncertain connection between the person for whom the text is named and the actual work of textual production. One reason for the prevalence of this belief today is that it seems most in keeping with evidence from recent archaeological finds datable to the Warring States and early Han eras, in which texts are not neatly grouped by author and do not seem concerned with attribution. The basic idea, however, is not new: Feng Youlan (1895–1990) traced it to Zhang Xuecheng (1738–1801)14 and earlier. As Feng described it, in part summarizing Zhang’s views, The texts of the various pre-Qin masters are the collected writings of an entire school of thought. Although a text may be named after a certain master, this does not guarantee that any given number of chapters were personally written by the master, much less the whole text.  .  .  .  If we do not fully understand the pre-Qin situation, we continually impose modern ideas of authorship onto pre-Qin compositions. In reality, the very idea of authorship was murky in pre-Qin times, and even more certainly there was no such thing as copyrights!15 As Feng pointed out, it is very easy to impose a modern understanding of authorship upon premodern authors. This is all the more true because these “modern” understandings have their roots in the Han or even earlier. Thus, even for a text like the Zhuangzi (i.e., one that so far has few excavated parallels to complicate its attribution), it is worth reconsidering the paradigm of authorship in light of the preceding arguments. A number of scholars have shown themselves open to a theory of the Zhuangzi’s compilation that does not necessarily assume Zhuang Zhou’s authorship of all or any particular part of the text. Though this has been dubbed by Liu Xiaogan the “No Specific Author” theory or the “Anytime Collection Theory,”16 I prefer to think of it as a model of text formation that considers the Zhuangzi as an anthology (borrowing the term from Chris Fraser17) that was compiled in a time when individual authorial attribution was still not a priority. Such compilation practices were far from unusual in ancient China. Sima Qian’s 司馬遷 (b. 145 BCE) account of how The Annals of Master Lü (Lüshi chunqiu 呂氏春秋) came to be18 and Ban Gu’s (班 固 32–92 CE) account of the Huainanzi’s 淮南子 (ca.141 BCE) compilation19 both follow such a model: they are the works of many hands grouped under a single name.20

Reading the Zhuangzi Anthology | 15 It is no surprise that Feng Youlan was also one of the first modern scholars to raise serious doubts about Zhuang Zhou’s authorship of the Inner Chapters.21 He introduced his view by explaining, “There is a traditional theory, that the Inner Chapters are Zhuangzi’s own work, while the rest were written by his disciples and later followers. This is nothing but a sort of speculation, and is not backed by evidence.”22 Feng proceeded to give various interesting pieces of evidence that, though minor and fragmentary in themselves, do seem to run contrary to the prevailing theory.23 Finally, Feng concluded that, based on his own sense of Zhuangzian “flavor” or “tone” (kouqi 口氣), “Xiaoyaoyou” (逍遙遊, chap.1) and “Qiwulun” (齊物論, chap. 2) alone should be taken as the core of the Zhuangzi, and the intellectual authenticity of the other parts should then be judged according to the standards set by just those two chapters.24 Wang Shumin (1914–2008) also doubted that the distinction between Inner and Outer and Miscellaneous Chapters could be used to draw conclusions about authorship and authenticity. He wrote: Many worthies of the past suspected that the Outer and Miscellaneous chapters were of late origin, and were forged by later people. The thirty-three-chapter version of the Zhuangzi does indeed have an authenticity problem, but we cannot rely on the Inner, Outer, and Miscellaneous Chapter [divisions] as [criteria] for judging. It is likely that the dividing lines among the Inner, Outer, and Miscellaneous Chapters of the current edition were fixed by Guo Xiang. Thus the Inner Chapters are not necessarily authentic and Outer and Miscellaneous Chapters are not necessarily dubious.25 He went on to give evidence for different editorial manipulations that Guo Xiang (or others) appear to have performed on the Zhuangzi text, all pointing to the conclusion that the Guo Xiang edition of the Zhuangzi (direct ancestor of the editions used today) looked very different from earlier recensions. • Among Western scholars, Christopher Rand gave one of the more thorough accounts of authorial issues with the Zhuangzi. He wrote, in an often-overlooked study, “Since the time of Han Yu . . . a growing number of Chinese scholars began to note anomalies . . . leading them to believe that the textus receptus may not have been the product of one man or even of one era.”26 Rand suggested that the Zhuangzi text we have is “a congeries of poetic parleys, diverting anecdotes, and philosophical disquisitions which have certain family resemblances to each other but also exhibit distinctive and

16  |  Esther S. Klein occasionally inconsistent opinions or ‘schools’ of thought.”27 He paints the following picture of how the text came to be. First, a certain amount of material was drafted by Zhuang Zhou and/ or his disciples. This material ultimately formed some (but not all) of the present Inner Chapters; some of it may also have ended up elsewhere in the text. Second, in subsequent generations (late Warring States to Han), various writers wrote pieces “elaborating on Master [Z]huang’s themes.”28 Third, both these types of material probably came together through the efforts of a “compiler or group of editors” in the Western Han—Rand cites evidence that this happened at Huainan, but does not insist on it. In either case, a version with distinct, named chapters was known to Sima Qian (since he lists some chapters), but it is not clear whether the version he saw was the same as the fifty-two-chapter one listed by Ban Gu in the “Treatise on Literature and the Arts” (Yiwenzhi 藝文志) (Hanshu 30.1730). The latter may well have undergone further editing at the hands of Liu Xiang 劉向 (77–6 BCE) or Liu Xin 劉歆 (46 BCE–23 CE). Fourth, various other versions were subsequently produced, often with different chapter and section divisions. Guo Xiang eventually produced a thirty-three-chapter version through what appears (from his own description) to be a fairly extensive editing process: Rand suggests that the Guo Xiang text was “a complete rearrangement and abridgement of earlier efforts.” It became “the most favored version” during the Tang and Song periods, “though it was by no means unaltered after the third century.”29 From the picture presented by Rand, it would seem that while some of the Inner Chapters (and other) material probably originated with Zhuang Zhou and/or his followers, it is not a good idea to assume that any particular part of the Zhuangzi is “authentic” in that way. Nor is the current organization of the text necessarily a reliable guide to its chronological age or original structure. Rand’s version of how the Zhuangzi came into existence seems highly plausible: it generally accords with archaeologically informed beliefs about early text formation and is also a fair interpretation of available evidence from transmitted sources. The only inference that seems potentially problematic—and it is one that a great number of scholars tend to make, implicitly or explicitly—is that because “the concepts of [the Inner Chapters] have been systematically merged into a well-integrated body of thought,” there was probably “a single initial author.”30 First, as A. C. Graham’s sensitive discussion and presentation of the Inner Chapters shows, the body of thought found there is not necessarily as systematic or well integrated as is generally assumed.31 And second, even if it were as well integrated as Rand suggests, it is just as plausible that this could be the work of a careful editor.32 For example, the philosopher Walter Kaufmann (1921–1980) selected pieces by multiple thinkers and brought

Reading the Zhuangzi Anthology | 17 them together into a well-known anthology (Existentialism from Dostoevsky to Sartre) that was “more than the sum of the parts” and was “intended to tell a story”33—that is, an account of the development and elaboration of the philosophical movement known as existentialism. Admirers will find the book does indeed give a systematic and well-integrated picture of this movement, and it is widely used as a teaching tool on the subject. Yet not all the authors included in the volume would have considered themselves to be members of a single “school” or lineage. It was up to Kaufmann to bring out “the growing variations of some major themes, the echoes, and the contrasts,”34 in part through a forty-page introduction that is in some ways comparable to the Tianxia 天下 (chap. 33 of the Zhuangzi).35 A similar process could well have happened with the Zhuangzi, except that the materials that went into it were written at a time when authorial attribution was a nearly nonexistent concern. One of the well-known text-critical rules of thumb is often expressed as lectio difficilior potior (the more difficult reading carries), that is, the less systematic and less well integrated a text is, the more likely it is to be original, old, and authentic. While this cannot be taken as a hard-and-fast rule—textual corruption, after all, does occur—it should certainly be admitted as a possibility: that we are not justified in assuming single authorship, merely because a number of textual units hang together reasonably well. • Chris Fraser has also argued, if only implicitly, for something like the anthology theory of the Zhuangzi’s authorship. His review of Liu Xiaogan’s Classifying the Zhuangzi Chapters (1994) is particularly useful in identifying a hidden assumption that makes Liu’s argument appear stronger than it actually is: namely, that “the division between Inner, Outer, and Miscellaneous chapters is related to authorship”—that “if Zhuangzi wrote any of the Zhuangzi material, he wrote one or more of these three distinct blocks of chapters.”36 As Fraser points out, this assumption is “not obviously justified” and is also question-begging, since it is exactly this assumption that various critics of the single-authored Inner Chapters theory are seeking to disprove. Fraser further critiques the inferences made in several of Liu’s main arguments. For example, Liu has identified three binomes only otherwise known in later materials. These binomes appear in some of the non-Inner Chapters but do not appear in any of the Inner Chapters. However, as Fraser points out, “the argument only shows that some of the Outer and Miscellaneous chapters could not have been written by Zhuang Zhou.”37 Indeed, I would go further and argue that in fact this evidence only shows that some parts of some non-Inner Chapters could not have been written by Zhuang Zhou (see further discussion in this chapter).

18  |  Esther S. Klein Another foundation block upon which Liu’s argument rests is that thematic connections among the Inner Chapters tie them together, while the rest of the chapters are less closely tied. As argued previously (and also by Fraser), this cannot necessarily be taken as evidence for single authorship; it could also be explained by literary influence, shared philosophical interests among different authors, or later editorial intent. Finally, Fraser addresses Liu’s arguments regarding textual borrowing. Because Liu thinks he has established an early date for the Inner Chapters, he uses textual parallels (which he describes as “borrowings”) to show that the rest of the book also came together not long after that supposedly early stratum to form a unified corpus at a fairly early date. Specifically, he relies on the fact that the Lüshi chunqiu, a relatively early text known to be compiled from even earlier sources,38 shares textual material with some of the Outer and Miscellaneous chapters. According to Liu this means that the entire Zhuangzi text must have been widely circulated.39 Fraser proposes a plausible alternate theory, that “groups of chapters may have existed separately for a long time before being collected together.”40 • David McCraw’s ambitious study, Stratifying Zhuangzi: Rhyme and Other Quantitative Evidence, explores a tantalizing source of potential information about the text: namely, the Zhuangzi’s “abundant rhyming evidence.” McCraw has identified 605 rhyming segments41 and attempts to use the properties of those rhymes (since rhyme groups vary over time and space) as evidence regarding the dating and/or authorship of the text. McCraw’s book is difficult: even someone familiar with the Zhuangzi, conversant with principles of ancient Chinese historical phonology, and comfortable with statistical methods still might find it challenging to follow.42 However, the general conclusion is that the Zhuangzi rhyming patterns resemble those of composite texts like the Lüshi chunqiu. Regarding the Inner Chapters, McCraw uses a variety of criteria for “inner links,” more complex and opaque than Liu Xiaogan’s.43 He also arrives at an opposite conclusion, namely: “Most likely many hands took part in forming Zhuangzi; indeed, allowing for later passages tacked onto various chapters, probably more than a dozen hands took part in forming the Inner Chapters alone.”44 McCraw explicitly distinguishes between the linguistic or stylistic coherence one might expect to be characteristic of a single-authored block of text and the philosophical coherence that, as I have also argued earlier, could just as well be attributed to the work of a thoughtful editor or anthologizer. I might add that terminological similarities (such as those listed by Graham and Liu) could also potentially be the work of an editor, given that in the

Reading the Zhuangzi Anthology | 19 early Chinese context editors did feel free to make small changes of wording. In any case, such distinctions are useful in approaching the complex phenomena of authorship and attribution. Another interesting aspect of Stratifying Zhuangzi is the section on the possible different layers of the text.45 As the author freely admits, the ideas he presents there are extremely speculative. Of course, the intuitions of traditional Chinese scholars, though more elegantly expressed, were not necessarily more rigorously supported.46 By current scholarly standards, this section—and perhaps the book as a whole—reads like the blueprint for an argument rather than the argument itself. Nonetheless, it is an interesting look at the kinds of conclusions one might reach by discarding externally based presumptions about authorship. As David Branner has pointed out, McCraw has overplayed his evidentiary hand (a possibility the author himself also admits) and could have done with a much more rigorous editorial process—preferably one that would have made his book at least somewhat usable by nonspecialists. That said, one wonders what the results would be if McCraw—with the considerable expertise he has now amassed regarding this issue—were to follow the more constructive of Branner’s suggestions and engage in a new study modeled on something like William H. Baxter’s methods.47 That is, I find it plausible that the flaws in McCraw’s study will not, in the long run, invalidate his insights.

Reading the Zhuangzi Anthology: Goals, Assumptions, Methods Elsewhere I have put forth my own discussion of the issue of the Zhuangzi’s authorship and compilation.48 My goal was to undermine the prevalent assumption that the Inner Chapters constitute the original core of the Zhuangzi, both in terms of their composition and compilation as a textual unit. Since certain aspects of that study appear to be susceptible to misunderstanding,49 it seems appropriate to continue the discussion here. First, a few words about why this argument matters. In approaching the study of a text like the Zhuangzi (or any text at all), almost all scholars will be working toward some greater end goal, whether this goal is explicitly acknowledged or not. Such goals vary considerably from person to person, but tend to share a certain “castles in the air” quality: rarely is the bridge of available evidence sufficiently solid to reach all the way there. It is this very circumstance that gives these larger goals their inspirational quality. A. C. Graham, for example, laid out a “wish-list” for future research on the Zhuangzi that included not only more sophisticated tests of authorship and

20  |  Esther S. Klein textual stratification, but also the desire to see scholars “restore the corrupt or dislocated text  .  .  .  make sense of the philosophical terminology  .  .  .  [and] relate the book to the other philosophical literature of the age.”50 Graham’s goals, not only for the Zhuangzi but for other texts as well, could be characterized as reconstructive: he wanted to “restore” the text, the philosophy expressed therein, and the entire intellectual context in which that philosophy once unfolded.51 My underlying goal, in this and other work, is somewhat different. The general point I want to make is that there was a different model of textual production in early China, particularly with regard to the issue of authorship and attribution. This in itself is nothing new, as various scholars have recognized and described the differences. What has not been attempted with sufficient thoroughness, however, is an exploration of the full ramifications that such a model should have on the way we study, analyze, and talk about early texts. It is common to first acknowledge uncertainties about authorship and attribution, but then proceed to write as if such uncertainties did not exist. One result of this is excessive optimism about the project of reconstruction (textual, philosophical, or socio-intellectual). Liu Xiaogan admits that there “are not sufficient grounds to discuss the dating and authorship of an isolated single essay or passage, because we have no reliable point of comparison or evidence.”52 This is a conclusion with which I would generally concur, whether the text in question is the Zhuangzi, the Lunyu, or any other early text.53 Where Liu and I differ is in what we do with this observation. Liu goes on to write that while isolated passages in the Inner Chapters (for example) could be of later date, “a few exceptions do not negate the general judgment.”54 He criticizes the use of a technique he refers to as “sample argumentation,” which he describes as follows: Scholars select certain characters, terms, sentences, or ideas as samples from which to argue their various positions or theories. In most cases, the isolated samples are not adequate to prove anything about the big picture, and so these arguments are not convincing.55 While this is certainly a technique scholars use in making their arguments, the term “sample argumentation” does not seem to exist outside of Liu’s work: the more usual term is, simply, “giving evidence.” One might criticize a tendency to overgeneralize or overplay the available evidence, but this is a problem of degree, not of kind. In the humanities generally—not to mention the study of ancient texts—almost no evidence is ever adequate to “prove” anything. This holds as true for Liu’s arguments in favor of early

Reading the Zhuangzi Anthology | 21 single-authored Inner Chapters as it does for any other theory about the Zhuangzi’s authorship. As Chris Fraser has pointed out, with some justice, “Liu seems to treat his single-author theory as an established conclusion that rival theories are obliged to disprove.”56 My approach toward conflicting evidence about the provenance of the Inner Chapters has been to consider it in terms of probability, with the added caveat that traditionally held beliefs should not receive a “free pass”—that is, no hypothesis about this issue should be considered “established until proven otherwise.” At this point, there is not enough evidence for any position to justify the use of the word “proof.” But if one considers how the evidence stacks up absent traditional assumptions about the Zhuangzi’s authorship, the case for single-authored early Inner Chapters is not necessarily strong enough to warrant its general acceptance as the default position. In my study of the Zhuangzi, I make three assumptions. First, I assume it is possible, even likely, that the basic textual units of Warring States texts were passages or sections rather than chapters. I assume, therefore, that a given unit of text below the chapter level could have come into existence before or after the other units of text now found together with it in any given chapter. I assume that the same is true, mutatis mutandis, of chapters within the larger divisions (Inner, Outer, Miscellaneous). Second, when there are textual parallels between the Zhuangzi and another early text, I assume that is evidence that the passage in question was either created by or known to the compilers of each text. I use the word “parallel” rather than “quotation” because I do not make assumptions about the direction of borrowing. Furthermore, there are always potentially more than two texts in the system: that is, it is always possible that both known texts borrowed the passage in question from a third (possibly unknown) text. Finally, I assume that editors in ancient times felt freer than those of today in altering the texts they were working with at the level of words and phrases. I assume this because it is observable in many different cases, of which Sima Qian’s reworking of his source material is just one example.57 • In my study of Zhuangzi parallels, I found a pattern that up through the Western Han significantly favors non-Inner Chapters.58 While parallels with Inner Chapters material are not entirely absent, they are relatively rare and equivocal. For example, a story about Yao and Xu You found in Lüshi chunqiu is most closely parallel with an anecdote found in Zhuangzi chapter 1 (“Xiaoyao you”).59 However, Zhuangzi chapter 28 (“Rang Wang”) contains a long sequence of similar anecdotes and even a sort of “stub” or abbreviated

22  |  Esther S. Klein version of the anecdote found in chapter 1.60 It is as if one of the best stories from chapter 28 had been copied into the portion of the text that would become chapter 1, and then a later editor, bent on eliminating redundancy, cut the redundant material from chapter 28. If Harold Roth (“Who Compiled the Chuang Tzu”) and others are correct, that the first stage of that editorial process described previously took place at Huainan (and if the second stage perhaps occurred at the time of Guo Xiang), the parallel with Lüshi chunqiu may just as well have been with chapter 28—or with the materials that would eventually form chapter 28—and not with chapter 1. This is just one of several examples. One of the few situations that can be confidently called a citation rather than a parallel is the chapter on Zhuangzi in the Shiji (63.2143–2145). There we find explicit citations of the Zhuangzi and a partial list of chapters. Oddly, no Inner Chapters are found on the list, nor is any Inner Chapters material cited or used in the Shiji biography. In writing Zhuang Zhou’s biography, Sima Qian does not include any of the several well-known Zhuang Zhou stories that are found in the Inner Chapters. Of course, one cannot draw firm conclusions from an absence of evidence. Nonetheless, we might consider the situation in terms of probabilities. Given the picture of the Zhuangzi text we find in the Shiji, would it seem likely that the version Sima Qian saw was structured like Guo Xiang’s version, with the Inner Chapters confidently presented as its core and most important division? I would tend to answer that question in the negative. A similar set of conditions obtains when we consider Zhuangzi parallels in excavated texts. There are not many of these. Mid-Western Han tombs at Fuyang (阜陽) and Zhangjiashan (張家山) have yielded some texts with Zhuangzi parallel passages. The Fuyang texts as known so far parallel passages from chapters 25, 26, and 28.61 The Zhangjiashan slips contain some parts of the story of Robber Zhi (Dao Zhi 盜跖) now found in chapter 29.62 There may well be any number of further Zhuangzi parallels yet to be discovered underground. Nonetheless, the pattern of evidence we have so far—no matter how fragmentary—tells a story similar to the one we find in the Shiji. There is no sign of a core Inner Chapters section, and that lack is coupled with an apparent preference toward the non-Inner Chapters. In general, regarding the authorship of the Inner Chapters or the general picture of the Zhuangzi’s compilation, there is not enough evidence to reach any firm conclusions. Nonetheless, various scholars including myself would argue that there is enough evidence to cast doubt on the traditional picture of a single-authored Zhuangzi Inner Chapters section. There may even be enough evidence to suggest that the Zhuangzi of the Miscellaneous Chapters has as much a claim to the title as the Zhuangzi of the Inner Chapters.

Reading the Zhuangzi Anthology | 23

Notes  1. As noted by Liu Xiaogan (William Savage trans, Classifying the Zhuangzi Chapters [Ann Arbor: University of Michigan, Center for Chinese Studies, 1994], 48), Su Shi’s 蘇軾 (1037–1101) “Zhuangzi citang ji” 莊子祠堂記 (Record on Zhuangzi’s memorial hall) casts doubt on the authenticity of four chapters: “Rang wang” 讓王 (Yielding the reign, chap. 28), “Shuo jian” 說劍 (Persuasion regarding swords, chap. 30), “Yu fu” 漁父 (Old fisherman, chap. 31), and “Dao Zhi” 盜跖 (Robber Zhi, chap. 29). Su Shi’s main reason for questioning these chapters, however, is his contention that the “real” Zhuang Zhou did not sincerely intend to criticize Confucius, while those four chapters seem to portray Confucius in a negative light (Su Shi wenji [Beijing: Zhonghua shuju, 1986], 11.347–8). While one might note a variance in the way Confucius is portrayed between sections of the book, this does not by itself allow one to conclude that one or the other portrayal is more or less authentic. An equally compelling story is that the syncretist editors of the Zhuangzi found the anti-Confucius material ideologically inconvenient and so relegated it to the outer reaches of the miscellaneous section.   2. See Liu, Classifying the Zhuangzi, 48–49.  3. A. C. Graham, “How Much of Chuang Tzu Did Chuang Tzu Write?,” in Studies in Chinese Philosophy and Philosophical Literature (Institute of East Asian Philosophies, 1986, rpt.; Albany: State University of New York Press, 1990), 283–321; Liu, Classifying the Zhuangzi.   4. E.g., Liu, Classifying the Zhuangzi, 155.   5. Graham, “How Much of Chuang Tzu.”  6. A. C. Graham, trans. and intro., Chuang-tzŭ: The Inner Chapters (London and Boston: George Allen & Unwin, 1981; rpt., Indianapolis and Cambridge: Hackett, 2001), 27.   7. Graham, “How Much of Chuang Tzu,” 283.   8. Ibid., 284.   9. Graham, Chuang-tzŭ, 30, 29. 10. Ibid., 33. 11. Ibid., 31. 12. Harold Roth, A Companion to Angus C. Graham’s Chuang Tzu: The Inner Chapters (Honolulu: University of Hawai’i Press, 2003), 188. 13. Mark Edward Lewis, Writing and Authority in Early China (Albany: State University of New York Press, 1999), 94–95. 14. Feng cites Zhang Xuecheng’s essay “Yan gong” 言公 (Wenshi tongyi 文史 通義 [Beijing: Zhonghua shuju, 1985, rpt. 2005] 2.169–217), which indeed contains fascinating insights on early Chinese authorship. 15. Feng Youlan, Zhongguo zhexue shi xinbian 中國哲學史新編, vol. 1 (Beijing: Renmin chubanshe, 1964), 364. 16. Liu Xiaogan, ed., Dao Companion to Daoist Philosophy (New York: Springer, 2014), 131–132. 17. Chris Fraser, Review of Liu Xiaogan, Classifying the Zhuangzi Chapters, Asian Philosophy 7.2 (1997): 157.

24  |  Esther S. Klein 18. Sima Qian 司馬遷, Shiji 史記 (Beijing: Zhonghua shuju, 1982), 85.2510. 19. Ban Gu 班固, Hanshu 漢書 (Beijing: Zhonghua shuju, 1962), 44.2145. 20. This is not to suggest that Zhuang Zhou was a patron like Lü Buwei 呂 不韋 (290–235 BCE) or Liu An 劉安 (ca. 179–122 BCE). There is even some doubt as to whether he had disciples per se (Graham, Chuang-tzŭ, 115). For a contrasting view regarding a potential Zhuangzi lineage, see Brian Hoffert, “Distinguishing the ‘Rational’ from the ‘Irrational’ in the Early Zhuangzi Lineage,” Journal of Chinese Philosophy 33.1 (2006): 159–173. 21. By contrast, Zhuang Zhou’s authorship of other chapters has been questioned far earlier, in Song times as mentioned previously, or perhaps even earlier (see Christopher C. Rand, “Chuang Tzu: Text and Substance,” Journal of Chinese Religions 11 [1983]: 5). In describing Guo Xiang’s 郭象 (d. 312) strategy for abridging the Zhuangzi, Lu Deming 陸德明 (556–627) wrote that “later people had added superfluous appendages” [後人增足] and Guo Xiang “included or excluded [material] based on the underlying meaning” [以意去取] (“Xulu,” Jingdian Shiwen 經典釋文 [Shanghai: Shanghai guji chubanshe, 1985], 66). Though Guo Xiang could perhaps have succeeded in lopping off all these appendages, there remains the suggestion that even in Lu Deming’s time it was understood that the Zhuangzi was not purely a single-author text. 22. Feng, Zhongguo zhexue shi xinbian. 23. Liu, Dao Companion, 138, dismisses this as “sample argumentation”—more on this later. 24. Feng, Zhongguo zhexue shi xinbian, 367. 25. Wang Shumin 王叔民, Zhuangzi jiaoquan 莊子校詮 (Taipei: Zhongyang yanjiu yuan Lishi yuyan yanjiu suo, 1988), 1436. 26. Rand, “Chuang Tzu: Text and Substance,” 5. 27. Ibid. 28. Ibid., 7. 29. Ibid., 14. 30. Ibid., 6. 31. Liu, Classifying the Zhuangzi, 25, also concedes this point. 32. Note that Chris Fraser makes a similar suggestion in saying that “thematic connections between chapters can also be explained by the hypothesis that the texts were written by different authors with shared philosophical interests” (Fraser, Review of Liu, 157). 33. Walter Kaufmann, ed., Existentialism from Dostoevsky to Sartre (New York: Meridian Books, 1956), 8. 34. Ibid. 35. A determined editor can also produce tremendous alterations in the overall message of the material being edited. Again, Kaufmann furnishes an excellent example in his diatribe against the irresponsible use of ellipses, deliberately (mis)representing Jesus’s views on war and arson as follows: “Do not think that I have come to bring peace on earth; I have not come to bring peace, but a sword.  .  .  .  I came to cast fire upon the earth.  .  .  .  Do you think that I have come to give peace on earth? No, I tell you.  .  .  .  Let him who has no sword sell his mantle and buy one” (From Shakespeare to Existentialism: An Original Study [Princeton: Princeton University Press, 1980], 99).

Reading the Zhuangzi Anthology | 25 This is only to say that even authentic Zhuangzian material could be slanted to say something different merely through the techniques of abridgement and rearrangement that are commonly employed in the traditional Chinese editorial tradition. 36. Fraser, Review of Liu, 156. 37. Ibid. 38. The Lüshi chunqiu is generally dated to around 239 BCE, based on the description of its compilation found in the Shiji biography of Lü Buwei (especially Shiji 85.2510). 39. Liu, Classifying the Zhuangzi, 52, 162. 40. Fraser, Review of Liu, 158. 41. David McCraw, Stratifying Zhuangzi: Rhyme and Other Quantitative Evidence (Language and Linguistics Monograph Series 41, Taipei: Institute of Linguistics, Academia Sinica, 2010), 1, 3. 42. E.g., David Prager Branner, Review of David McCraw, Stratifying Zhuangzi: Rhyme and Other Quantitative Evidence, Journal of the American Oriental Society 130.4 (2010): 653. 43. Liu, Classifying the Zhuangzi, 25–38. 44. McCraw, Stratifying Zhuangzi, 47. 45. Ibid., 87–100. 46. See for example the previously cited comments by Su Shi, or Wang Fuzhi’s 王夫之 (1619–1692) suggestion that “Tian dao” 天道 (Heaven’s way, chap. 13) was “probably written by a Huang-Lao scholar in the Qin or Han in order to assist the ruler” (蓋秦漢間學黃老之術，以干人主者之所作也) and that it was “certainly not written by Zhuangzi and furthermore not even written in imitation of Zhuangzi by someone who was a good scholar of the Zhuangzi, as should be clearly recognized by the reader” (定非莊子之書，且非善學莊子者之所擬作，讀者所宜辨也). Wang concluded that “there are many cases like this among the remaining chapters, as becomes clear if one extrapolates [from this]” (餘篇多有類此者，推之可見) (Zhuangzi jie 莊子解 [Beijing: Zhonghua shuju, 1964], 114). Like Su Shi, Wang Fuzhi based his argument on the philosophical content of the chapter, which was at odds with his Inner Chapters–derived notion of the core Zhuangzi. This is not to find fault with Wang’s argument, as most readers do find this block of chapters (at least chapters 12–14, often others as well) to be at odds with the rest of the Zhuangzi. Nonetheless, his suggestion about authorship (by a Huang-Lao follower) would be considered speculative by modern standards. 47. Banner, Review of McCaw, 653. 48. Esther Klein, “Were There ‘Inner Chapters’ in the Warring States? A New Examination of Evidence about the Zhuangzi,” T’oung Pao 96 (2010): 299–369. 49. Liu, Dao Companion, 131n1, does not, for example, accurately represent my actual presuppositions. 50. Graham, Chuang-tzŭ, 30. 51. This challenge is taken up to some extent in Roth, Companion, and Hoffert, “Distinguishing the ‘Rational’ ”—specifically in regard to establishing filiations. 52. Liu, Dao Companion, 144. 53. Contra E. Bruce Brooks and A. Taeko Brooks, The Original Analects: Sayings of Confucius and His Successors (New York: Columbia University Press, 1998).

26  |  Esther S. Klein 54. Liu, Dao Companion, 145. 55. Ibid., 37. 56. Fraser, Review of Liu, 159n4. 57. For interesting readings of Sima Qian’s work on his source material, see Wai-yee Li, “The Idea of Authority in the Shih chi,” Harvard Journal of Asiatic Studies 54.2 (1994): 345–405; Stephen Durrant, The Cloudy Mirror: Tension and Conflict in the Writings of Sima Qian (Albany: State University of New York Press, 1995); Griet Vankeerberghen, “Texts and Authors in the Shiji,” in China’s Early Empires: A ReAppraisal, ed. Michael Loewe and Michael Nylan (Cambridge: Cambridge University Press, 2010), 461–479. 58. Klein, “Were There Inner Chapters,” 355. 59. Chen Qiyou 陳奇猷, ed., Lüshi chunqiu xin jiaoshi 呂氏春秋新校釋 (Shanghai: Shanghai guji, 2001), 22/5.1524–1525; Guo Qingfan 郭慶藩, ed., Zhuangzi jishi 莊子集 釋 (Beijing: Zhonghua shuju, 1985), 1.22–24. 60. Zhuangzi jishi 28.965. 61. Han Zhiqiang 韓志強, “Fuyang chutu de Zhuangzi ‘Zapian’ Han jian” 阜陽 出土的莊子雜篇漢簡, Daojia wenhua yanjiu 道家文化研究 18 (2000): 10–14. 62. Liao Mingchun 廖名春, “Zhuangzi ‘Dao Zhi’ pian tanyuan” 莊子盜跖篇探源, Wenshi 文史 45 (1998): 49–59.

2

Reflections on Textual Analysis in the Post-Graham Era

Liu Xiaogan

A. C. Graham was the first Western scholar who made a dramatic impression on me. It was near the summer of 1985 in Peking University when I just completed my PhD dissertation, which entailed a textual and philosophical investigation of the Zhuangzi. I gave a copy of my dissertation to Professor Wei-ming Tu, who was then a visiting professor from Harvard. When Wei-ming spoke with me later, I was surprised at his opening words: “Your dissertation is well written. Your approach is similar to that of A. C. Graham. He also conducted linguistic analyses of the Zhuangzi.” This made me extremely curious. How could my approach possibly be similar to that of a Western scholar I had never heard of? Thanks to Professor Donald Munro’s kind invitation, in the spring of 1988 I went to the University of Michigan as a visiting scholar for three months, just before my two-year visit to Harvard. The first thing I did in Ann Arbor was to borrow and read Graham’s books. I discovered that, yes, there are certain similarities between his Zhuangzi studies and mine, though there are also significant differences. The next summer, when I met Graham in person during the East-West Philosophers’ Conference in Hawai’i, I hoped to exchange ideas and discuss issues in Zhuangzi studies with him. To my surprise, before I was able to raise my questions, he told me he was no longer interested in the Zhuangzi. So we talked about many different things, but not about the Zhuangzi. Though I was disappointed that he did not respond as I hoped, desiring to learn from him, I came to understand his style and disposition through personal observation and academic exploration. In my view, A. C. Graham

27

28 | Liu Xiaogan is not only a scholar, but a genius. Unlike most researchers, geniuses realize distinctive achievements more from their natural interest, talent, and dynamism, though certainly they also work hard. They don’t worry about or seriously adhere to the common practices of professional research and do not want to be a specialist in the usual sense. They freely take leeway in their studies to satisfy their great curiosity and interests. And indeed Graham is a genius, at least in my observation and understanding. He conducted path-breaking research on many topics and texts, including books and monographs on the Laozi, Zhuangzi, Liezi, Mozi, and on the Cheng brothers. His books also include Disputers of the Tao, translations of Poems of the Late Tang and Poems from West Lake. In addition, he published challenging and thoughtful books on much broader philosophical topics, including Unreason within Reason, The Problem of Value, Ethics, and Science, and Reason and Spontaneity. He always impressed academic colleagues with his original textual analyses, elegant translations, original insights in textual studies, brilliant speculations, and his philosophical vision. As Rosemont said, he “was known both for the depth of his Chinese textual scholarship as a philologist, grammarian, and translator, and equally for the breadth of his philosophical interpretations of those texts.”1 No wonder his is the only name frequently cited across the various chapters of my edited volume, Dao Companion to Daoist Philosophy.2 I am honored to have been invited to contribute a chapter to a volume dedicated to Graham, this great scholar and personal inspiration. But I believe this volume is not only a salute to a distinguished scholar, but also a place to showcase the work, areas, and approaches that he established as a model for all of us who have come after. The present essay will point to some unconscious assumptions common in our practice of textual analysis in the era since the great master left us, in hopes that it will inspire further reflection. This essay is essentially a reflection on our basic assumptions in doing textual analysis, some of which have not yet been clearly formulated, never mind discussed. My reflection has been stimulated by the evidence of the many texts excavated from ancient tombs that have turned on their head many of our prevailing impressions and conclusions about received Chinese texts. The critical issue here is not simply to modify and correct our positions according to that evidence, but also to reexamine the approaches and methods that have led us to wrong positions for many decades. Without these important excavations, we would still be accepting the prevailing mistakes and overlooking critical blind spots. These reflections touch on A. C. Graham’s work but are not limited to his personal approaches, since many of us have shared or adopted the same methodologies. Some may regard this essay as defending a traditional view, but this is not my goal. To defend or

Reflections on Textual Analysis in the Post-Graham Era  |  29 argue for any position is a different job and requires more work than allowed by the constraints of this essay. The focus of this essay is a methodological review, especially reexamining the prevailing assumptions in A. C. Graham’s argumentation against traditional positions, which are proved not valid and left us some lessons in the light of recent archaeological findings.3 • Here are the four unconscious assumptions we must reconsider in our practice of textual studies:

1. Extant texts and literature may be assumed to represent the body of all texts and documents in history.



2. Sound reasoning and speculation are more reliable than imperfect historical literature and records.



3. A historical text or record must either be judged reliable or discarded.



4. Samples can be used as evidence for general judgments.

Assumption 1: The texts we have available to study are the whole body of ancient literature. This is a common assumption in modern textual analysis, especially with regard to the dating of received works. Let us take the case of the Laozi as an example. Qian Mu (錢穆 1895–1990) was a highly respected scholar in China who was the first to strongly assert that the Laozi must have been written after the Zhuangzi.4 His position was taken as part of the influential scholarly movement known as “doubting antiquity” (古史辨), which was preeminent in the 1930s, but his arguments were quickly refuted by Liang Qichao (梁啓超 1873–1929), Hu Shi (胡適 1891–1962), and other scholars. In the end, leaving all argument, debate, and speculation aside, the discovery of the Guodian (郭店) bamboo-slip versions of the Laozi in 1993 proved that this text could not have been later than the Zhuangzi because the tomb from which the Guodian texts were excavated was dated to before 278 BCE by archaeologists. Zhuangzi is believed to have lived about 369–286 BCE. The fact that three versions of the bamboo-slip Laozi were possessed by the occupant of this tomb suggests that the work was already in circulation in a number of different versions. Thus, it may have been inscribed ten to twenty years earlier than the burial of these editions, and the master copies would certainly have been in circulation even earlier. Therefore, the text could hardly have been later than the work by Zhuangzi.

30 | Liu Xiaogan Qian’s main argument was based on his theory known as the “thread of thought,” in which the development of ideas of thinkers proceeds linearly from early to later. Specifically, he formulated the notion that a student’s simple outline must be derived from the master’s comprehensive and detailed statements. Thus, the Laozi, a brief work of five thousand characters, must be derived from the long and elaborate essays of the Zhuangzi. The premise here is questionable. The opposite seems rather more likely, namely, the master presents an outline of big ideas and his followers develop it into detailed statements and arguments, like Mencius’s elaborations on Confucius’s short dialogues. Qian’s “thread of thought” theory has gained little traction in English academic circles, and I will not discuss it further here. When I found that A. C. Graham had adopted Qian’s position and put the Laozi chapter after that of the Zhuangzi in his great book, Disputers of the Tao (1989), I was surprised and even felt a bit sorry for him, because Qian’s argument is clearly dubious, and I thought Graham would have had better judgment. Indeed, Graham devised a better strategy, though he shared Qian’s general position. He did not mention Qian’s assumptions at all but instead produced more sophisticated grounds for his position on the dating of the Laozi. Graham devised a brilliant maneuver by which to attack a problem that had been debated for nearly a century. In sum, his hypothesis is based not on the presence of positive evidence, evidence that support his hypothesis, but based on absence of negative evidence, evidence that contradicts his hypothesis. This argument works from the position that since no texts prior to the Zhuangzi quote the Laozi, this constitutes “evidence” that the Laozi probably appeared after the Zhuangzi. Graham claims, “Since the ‘Inner Chapters’ [of the Zhuangzi] show no clear evidence of acquaintance with Laozi, the book is conveniently treated after Zhuangzi, although there is no positive proof that it is later.”5 Graham’s method appears to be very scientific: everything should be based on positive evidence. Lacking evidence, namely, direct quotation from the Laozi in other pre-Zhuangzian works, he could challenge the traditional scholarship by provisionally putting the Laozi after the Zhuangzi. But this is paradoxical: by positioning the Laozi after the Zhuangzi, he bends the very rule of evidence he seems to follow at the outset, since he admits “there is no positive proof” to support his new chronological arrangement, and indeed, there are no quotations from the Zhuangzi in the Laozi. Graham’s method reminds us of Herbert A. Giles (1845–1935), who once contended that the Laozi must have been forged in the early Han dynasty (206 BCE–8 CE), after the composition of the Huainanzi (淮南子; ca. 141 BCE). Giles’s argument was based on his claim that Confucius, Zuo Qiuming (左 丘明), Mencius, Zhuangzi, Xunzi (荀子), Huainanzi, and Sima Qian (司馬遷)

Reflections on Textual Analysis in the Post-Graham Era  |  31 never saw or claimed to have seen the text of the Laozi.6 Hence, Giles came up with the hypothesis that the Laozi was forged in early Han times. This theory was nearly forgotten and definitively put to rest after the discovery of the Mawangdui silk versions in 1973, which have been confidently dated to before 195 BCE (Silk A) and 169 BCE (Silk B), respectively. Giles’s reasoning and argumentation proved invalid, but few of us have drawn lessons from his failure. Graham actually drew on Giles’s method when he based his hypothesis not on positive evidence but on a lack of negative evidence and a dismissal of the conventional literature. Giles was a seasoned and pioneering scholar. His arguments seemed logical given his reasoning, but his conclusion was clearly mistaken. Where does the pitfall lie in his argumentation? It is his unconscious assumption that he had access to all the texts available in ancient times. Actually, before the archaeological discoveries of the 1990s and onward, few scholars realized that 90 percent of ancient texts had been lost. Until recently, Giles’s assumption was actually one that had long been generally accepted by most scholars. Therefore, although one may meticulously examine all texts, she or he will likely see only a tiny portion of the texts, records, and books known to the Chinese of antiquity. Furthermore, Giles’s textual reasoning was rooted in more specific assumptions also shared by modern scholars. Without any one of them, Giles’s reasoning is invalid. They are:

1. All ancient texts and books must have been recorded, cited, or mentioned in other texts and books.



2. All those other texts and books must have survived over two millennia to be available to modern scholars.



3. As long as we cannot see X, we have grounds to suppose that it never existed.

Let me illustrate the invalidity of these assumptions. First, suppose we have a record by Mary that Susan wrote a book, but no one quoted that book and no other records make mention of it. This does not give us grounds to determine that Susan didn’t write a book. Second, if Bella quoted Susan, but both Susan’s and Bella’s books were lost and remain unavailable now, this does not provide sufficient grounds to say that Mary’s record must be wrong and Susan didn’t write the book. Third, if Mary, Cherry, and Berry in different times and various contexts wrote that they had seen or heard of Susan’s book, but we cannot find it now, this is not sufficient grounds to dismiss all three reports and determine that Susan didn’t write the book. After all, we cannot conclude that X never existed or was not possibly extant before a certain point in antiquity just because we cannot find X

32 | Liu Xiaogan now. More generally, we cannot suppose that the books and texts available to us today are the whole body of ancient works. Thus, the unconscious assumption that we already have at hand all evidence necessary to decide the dating and authorship of an ancient text is simply not valid.7 This is also why logical reasoning based on the books we currently have may lead to conclusions that are proved wrong by archaeological findings. No one actually lays out any of these assumptions, but we must be alert about them lurking behind similar reasoning in analyses of Chinese texts. They have remained firmly established in some researchers’ minds and methods—even in recent years.8 Scholars of early Chinese thought must recognize that the great numbers of unearthed texts have repeatedly proven these assumptions groundless, and any arguments based on them are highly risky. Giles’s method may be valid for a defined body of objects about which there is overwhelming consensus and enough related evidence and references, but the realm of ancient books and documents has proved to be an infinite, unknown kingdom for us: archaeologists cannot have excavated all antique books, records, and literature, but what they have found must give us serious pause. Conclusions based on these old assumptions are not only groundless; many of them have often proved to be false in light of new finds. To sum up, in recent decades, abundant and formerly unknown silk and bamboo-slip versions of both received and heretofore unknown works have come to light. Their discovery has revealed a simple truth: the texts we have today represent only a tiny part of the historical legacy of ancient Chinese literature. Actually, this should not be a surprise. After all, the bibliography section (“Yiwenzhi” 藝文志) of the History of the Western Han (Hanshu 漢書) mentions fifty-seven Confucian books, of which only seven survive; thirty-seven Daoist texts, with only five still extant; and ten Legalist books, only two of which survive today. Furthermore, 90 percent of the texts recently excavated are not recorded in the various bibliographies that have come down from antiquity. Historians in Han China must have seen many more texts and records that predated them than we can ever hope to. Certainly, they may have made mistakes, but they did not dream up historical figures and bibliographical matters. Newly excavated texts have shown that the histories and bibliographies from Han times are much more reliable than previously acknowledged. We simply do not have the resources and references to argue with them about the authenticity and historical sequence of ancient works, though our reasoning may sound more theoretical, consistent, and logical.9 Archaeological discoveries have also showed that some conventional ideas about ancient texts should be amended. Many of the books recorded in the Han bibliographies were widely considered to be fakes, according to

Reflections on Textual Analysis in the Post-Graham Era  |  33 the reasoning or guesses of modern scholars. But the unearthed bamboo-slip texts have proved some of them to have been in circulation even before the Western Han. Examples include Heguanzi (鶡冠子), Wenzi (文子), Liutao (六韜), Weiliaozi (尉繚子), and Yanzi (晏子), to name just a few. Contemporary scholars are inclined to trust their “rigorous” conjectures and reasoning more than the imperfect historical record. For example, Qian Mu denied historical records about the existence of the Sun Bin Arts of War (孫臏兵法) by Sun Bin (fl. ca. 350–320 BCE)10 and Sunzi Arts of War (孫子兵法) by Sun Wu (孫武 545–470 BCE), respectively. Because Sun Bin Arts of War had been lost for about two millennia, Qian claimed that there was only one Sunzi Arts of War by Sun Bin instead of Sun Wu, and Sun Wu is a name derived from the story of Sun Bin.11 However, the archaeological recovery of both Sun Bin Bingfa and Sunzi Bingfa in Shangdong Linyi (山東 臨沂) in 1972 illustrates that the speculation of great scholars is not as reliable as historical records. Another case is Zisizi’s work. According to early historical records, Zisizi (子 思子), a grandson of Confucius, composed certain works, but they were soon lost. So scholars believed he had not been actively writing at the time. But the bamboo slips excavated in Guodian in 1993 include one entitled “Duke Mu of Lu Asks Zisi” (魯穆公問子思), which suggests that Zisizi probably was a known figure and had his own recorded legacy. Examples like these show that ancient records and statements made by Han historians cannot be easily dismissed, and the “absence of a text” cannot be taken as sure evidence that it never existed, let alone as the basis of new theories or hypotheses about a text’s history. We must recognize that skeptical reasoning itself supplies no new evidence and no solid basis for knowledge.12 This is a lesson we should take from archaeological discoveries. By introducing archaeological findings and criticizing modern skepticism about the ancient Chinese textual tradition, however, I do not claim that the spirit of the “doubting antiquity” movement was entirely wrong and we should dismiss all its conclusions in favor of believing all antique texts and records. Instead, I mean to suggest that we really must improve our methodology in ways that will enhance the quality of our textual studies. This is critical to the healthy development of our fields.

Assumption 2: Sound reasoning and speculation are more reliable than imperfect historical literature and records. Another unconscious assumption is that historical truth must accord with our logical reasoning and conjecture. The danger here is that reasoned speculation might stand in for historical truth. Again, in the case of the Laozi’s dating, when scholars’ strong skepticism led them to put aside all traditional documentation, they could only offer logical inferences based on

34 | Liu Xiaogan one or two pieces of information selected from the very records they did not generally trust. D. C. Lau once wrote, “I am inclined to the hypothesis that some form of the Laozi existed by the beginning of the third century BCE at the latest.”13 Lau maintained that Sima Qian “had difficulty even with Laozi’s identity. He explicitly suggests that [Laozi] was probably the same person as Dan the historian (Taishi Dan 太史儋), though the latter lived more than a century after the death of Confucius.”14 This position can also be seen in the work of Arthur Waley (1889–1966), who described Sima Qian’s biography of Laozi as amounting to “a confession that for the writing of such a biography no materials existed at all.”15 But here Lau has misread Sima’s rhetoric: he neglects the distinction between the major part of Sima’s account, consisting of affirmative statements, and his additional recording of two rumors or pieces of hearsay. Lau focused on one of the two hearsay accounts and ignored the fact that this position faces even greater difficulties than the traditional one. The words “explicitly suggests” are used to support his reading, but they are not true. Lau omits key points in Sima’s comments: “Some say this Dan [Historian Dan of Zhou, Taishi Dan 周太史儋] was actually Laozi; others say no. Nobody knows which side is right.” Obviously, this is not quite the same as “explicitly suggests.” Lau also fails to see that, according to Sima Qian, Historian Dan’s prognostication about the Qin state’s destiny has nothing to do with the content of the Laozi or with the tenor of the exchange in the meeting between Laozi and Confucius, a story repeatedly cited in pre-Qin and Han literature. Lau’s hypothesis has more conflicts with the historical literature, broadly conceived, than do the major parts of Sima’s biography. Still, Lau shows admirable honesty when he says, “Indeed my whole account of Laozi is speculative, but when there is so little that is certain, there is not only room but a need for speculation.”16 Here we have to clarify the prevailing, if misconceived, impression that Sima Qian took no position regarding who Laozi was and when he lived. If we examine Sima’s Historian’s Record analytically and comprehensively, we find that Sima’s position on Laozi’s life is quite certain, and he was not at all confused by the two pieces of hearsay he mentions. In his biography of Confucius (Kongzi shijia 孔子世家) and that of the Confucian disciples (Zhongni dizi liezhuan 仲尼弟子列傳), Sima has no hesitation about Laozi’s identity and time; he sticks to the record in the main account of his biography of Laozi and makes no reference to the two rumors. Furthermore, scholars have examined the sequence of Sima’s seventy biographies of various masters (Liezhuan 列傳), which was clearly arranged in chronological order according to the first figure in each biography. The biography of Laozi is third, appearing after Guanzi (管子) but before Sima

Reflections on Textual Analysis in the Post-Graham Era  |  35 Rangqie (司馬穰苴). Guanzi served Duke Huan of Qi (齊桓公; ?–643 BCE), and hence came much earlier than Confucius. Sima Rangqie served Duke Jing of Qi (齊景公; fl. 547–490 BCE), who once consulted Confucius, which makes Sima Rangqie a contemporary of Confucius. The sequencing of the biography of Laozi, coming after Guanzi and before Sima Rangqie points to Sima Qian’s belief that Laozi was an elder figure to Confucius. He does not buy into the confusion between Laozi and Historian Dan of the late Warring States period.17 A. C. Graham contributed a brilliant conjecture about the meeting between Laozi and Confucius. In his essay “The Origins of the Legend of Lao Dan,” he proposes a scheme of five stages in the evolution of the story,18 which demonstrates a talent and vivid imagination fully engaged in the historical evolution of a key cultural narrative. For this job, Graham meticulously reexamined all materials related to any story about the relation between Laozi and Confucius, without discriminating between Confucian technical ritual records, Daoist fables, and historical notes. His purpose, however, was not to find the possible truth in these materials but rather to sweep away counter-evidence to the skeptical tradition regarding Sima Qian’s Laozi biography. He found that “from its origin in the fourth century B.C. the nucleus of the legend, and its only constant element, is the meeting of Lao Dan with Confucius.”19 Too many ancient texts referred to a meeting between Laozi and Confucius; thus, it became the “nucleus” of the tradition and a final obstacle to sweeping away the influence of Sima Qian’s account. Graham’s key point is his refutation of the possibility that the meeting story was created by the Daoist camp; instead, he argues that it was devised and used by Confucian groups. The following is his five-stages theory, the conclusion of this long essay: Stage 1: “A Confucian tale of the Master inquiring about the rites from a certain Lao Dan, very probably already known as an archivist of Zhou.”20 This tale, already current in the fourth century BCE, may have been either historical reminiscence or an exemplary story praising Confucius’s humility in seeking learning. This point is based on Graham’s painstaking textual analyses and personal resourcefulness in arguing that the tale of the meeting must have originated in the Confucian tradition instead of Daoist history. This reckoning depends entirely on a backdrop of antagonism between the two camps. But there is little in the historical record about a systematic distinction of the two sides before Han dynasty; even the term “Daoism” had not yet appeared (“Daoism” was probably coined by Sima Tan [司馬談], Sima Qian’s father, according to a survey of literature surviving from the period). Stage 2: “The adoption of Lao Dan in the Inner Chapters of Zhuangzi, towards 300 BC, as one of the characters in the life of Confucius exploitable as a spokesman of ‘Zhuangism.’ ”21

36 | Liu Xiaogan But there is no evidence in the Zhuangzi that Lao Dan was “one of the characters in the life of Confucius.” There is only one place in the Inner Chapters where Lao Dan makes an indirect comment about Confucius’s words, and in the Outer and Miscellaneous Chapters, there are seven dialogues between Laozi and Confucius. However, all these dialogues are obvious fables without any consistency about the two figures. Confucius is one of many “fable characters” in the Zhuangzi, and his identity is sometimes given as master, sometimes as disciple, and he even figures as both a positive and negative character. There is simply no clear relation between Confucius and Lao Dan. One should not try to find historical truth in fables. Stage 3: “The appearance of the Laozi under the name of Lao Dan, took advantage of his authority as a teacher of Confucius. From this point he represents a philosophical trend (‘Laoism’).” Graham further guesses that this stage, when the Laozi appeared “under the name of Lao Dan,” must have occurred by 240 BCE. This is certainly too late in light of the 278 BCE Guodian Laozi. Graham has also suggested that the Laozi became an important text because it was attributed to Lao Dan, who was credited as having given instruction to Confucius. This is a hypothesis that cannot be proved. Stage 4: “The identification of Lao Dan with the Grand Historiographer Dan of Zhou who in 374 BCE had predicted the rise of Qin, and [is also credited with] the invention of the story of the journey to the West and of the writing of the book for Guanyin. The purpose was to win favor for the Laozi from the [state of] Qin.” Here Graham seems to be proposing that the Laozi was at some point attributed to Historian Dan of Zhou (Taishi Dan 太史儋), and the intent of the person(s) who did this attribution was to persuade the ruler of Qin that this work was important. This development happened to coincide with the formulation of the story of Laozi’s retirement to parts west of the heartland China and the writing down of the Laozi at the request of Pass Keeper Guanyin (關尹). Be that as it may, the successful promotion of the Laozi to the Qin seems unlikely because the text clearly proposes natural harmony and a leadership style based in nonaction; it strongly opposes competition and aggressive wars. All these theories ran directly counter to the strategies and ambitions of Qin. Again, if one begins to rely on speculation to resolve complicated historical issues, one is also free to create stories and need present no evidence. Actually, there is no connection between the words and ideas attributed to Historian Dan and those in the Laozi. Graham’s focus on the story of the meeting between Laozi and Confucius is not the same as creating a story about who wrote the Laozi text. One must still ask, who first wrote the text? If Graham believes it was Historian Dan, why would words and deeds

Reflections on Textual Analysis in the Post-Graham Era  |  37 attributed to him in history have nothing to do with the ideas in the Laozi? If Graham thought the writing might have been done by another, how is it that Historian Dan so easily gets the credit and the true author disappears without a trace? Stage 5: “The graphic adaptation of the name of the Grand Historiographer [Dan] to Lao Dan after 206 BCE, by descendants of the former. The purpose was to make their ancestor welcome to the Han [dynasty] instead of the Qin. The personal details about the Grand Historiographer [Dan] thus became available as stiffening for the biography of Lao Dan by Sima Qian.  .  .  .  Since Lao Dan was earlier than Zhuangzi, he was established retrospectively as founder of the Daoist school.” I am afraid that Graham’s assertion that “the personal details  .  .  .  thus became available as stiffening for the biography of Lao Dan by Sima Qian” simply does not wash. It does not accord with Sima Qian’s original language or meaning in the biography. If it were true, why should Sima put the point so equivocally: “Some say this Dan was actually Laozi; others say no. Nobody knows which side is right. Laozi was really a gentleman who lived in retirement from the world.”22 It is not surprising that Laozi’s identity would become something of a mystery after he took up a hermit’s life. We know that in academic research, imagination and speculation can be important and productive; however, these are not the tools with which to discover historical truth. Any speculation about a person or situation must then put us in search of solid evidence, then subsequently tested, and in the end proved out by facts and documents. We also know that academic process and rational thinking must follow rules of logic; however, we have to remember that complicated events do not develop along a single path. Nor do events in the real world follow logical reasoning, especially developments that take shape over long historical periods and involve various people and elements. Mark Twain once remarked, “Truth is stranger than fiction, but it is because fiction is obliged to stick to possibilities. Truth isn’t.”23 Twain was right. Historical truth often goes in surprising directions without leaving clear hints about causes. Just one example of such a complicating element is the psychological phenomenon known as “Stockholm syndrome,” in which hostages express empathy and sympathy, and hold positive feelings toward their captors, sometimes to the point of defending and identifying with them. The FBI’s databases show that roughly 8 percent of victims exhibit signs of this syndrome, so its “illogic” may affect the course of events.24 Generally speaking then, truth does not unfold according to logical reasoning, though logical reasoning is frequently useful and in any case necessary to analysis. Both Lau and Graham have tried to interpret away the meeting of Laozi and Confucius against a backdrop of confrontation between the camps of

38 | Liu Xiaogan Confucians and Daoists. Intriguingly, Lau presumes the Daoist camp created the story to make a mockery of Confucius,25 while Graham proposes that it was the Confucians who invented the legend to praise Confucius’s humility in being willing to learn from anyone.26 They can come to opposite conclusions about the same story because both of them are speculative, with no positive evidence that may be examined. But neither Lau’s nor Graham’s account pass the test offered by the historical literature. We find no signs of struggle or conflict between Confucian and Daoist groups in pre-Qin records, where the term “Daoism” has yet to appear. The excavated texts provide even stronger proof against the modern assumption of antagonism between the two camps. Direct criticism in the Laozi, of the Confucian terms ren (humanity) and yi (righteousness) are missing in the earliest Guodian version, and there are clearly Daoist ideas in early versions of texts generally regarded as Confucian. Actually, up until the Song dynasty when the Neo-Confucian movement began to actively attack Buddhism and Daoism, there was little evidence to support the contentions of overt hostility between Daoism and Confucianism.

Assumption 3: A historical text or record must be either reliable or discarded. Many scholars hold a dualistic attitude toward ancient texts: a text is either counted reliable or it must be dismissed. If a text is judged unreliable according to a certain criterion, it can contain no truth or value and does not even deserve to be used in comparison. This attitude is like that of a judge toward a witness: if the judge finds that the witness has told a lie, all his testimony should be discounted and dismissed as proof. Sima Qian was certainly no liar, though he may have made mistakes in his biographies. Although his biography of Laozi includes elements from legend and hearsay, as a historian Sima was not in the business of creating fiction. His comments on and description of Laozi largely accord with the content and style of the Laozi and the record about him in other books. The meeting of Confucius consulting Laozi about ritual technicalities circulated widely, not only as story, but also as serious records, and it appears in many works of the time. More and more excavated documents point to the fact that Sima did base his great history on early records. He recorded unbelievable legends and rumors because that was what he read and heard, and he wanted to be faithful to the accounts of his time, even when he did not agree with or believe them—for example, the points about Laozi mentioned earlier, which he flags with expressions like “some say” or “nobody knows.” Occasionally, Sima also used stunning literary descriptions in his history because at that time, there was not a clear definition or distinction between literature and

Reflections on Textual Analysis in the Post-Graham Era  |  39 history.27 To state this is not to defend Sima’s mistakes. Indeed, we cannot take all his accounts as historical truth. We need to use discrimination in our analyses of the content, style, and credibility of the various sections in his work with reference to other period annals, literature, and archaeological findings. In addition to the three accounts of Confucius consulting Laozi about rites in the Shiji biographies, we also find records of the meeting in the eclectic anthology Lüshi chunqiu (Spring and Autumn of the Lu Family 呂氏春秋) and in Confucian works such as the Liji (Book of Rites 禮記), Hanshi waizhuan (Practical Annotation of Han’s Book of Songs 韓詩外傳), Kongzi jiayu (Confucian Family Teachings 孔子家語), and Shuoyuan (Collection of Anecdotes 說苑).28 Given the multiple records of this event, it is difficult to ignore it completely and date Laozi to an era sometime after Confucius. Still, this sort of evidence is not of itself sufficient to decide the date of the Laozi.29 A close reading of statements from the “Zengzi wen” (“Questions from Zengzi” 曾子問) section of the Book of Rites, which contains four passages regarding Confucius’s recall of Laozi, will be instructive. Under the influence of the “doubting antiquity” movement of the twentieth century, this work came to be considered an unreliable Han text. However, after the discovery of the Guodian bamboo slips, academic societies have come to recognize that it includes reliable historical materials, and at least some sections may preserve pre-Qin works. Confucius recalls in four passages the main topic of his exchange with Laozi—parts of an explanation of how to deal with specific issues in funerary practice. Here we can read their words as dialogue and experience the context and atmosphere of their meeting. One of the records runs like this: Zengzi asked: “Anciently when an army went on an expedition  .  .  .” Confucius replied: “When the son of Heaven (king, 天子) went on his tours of inspection  .  .  .  I heard the following statement from Lao Dan (吾聞諸老聃曰): ‘On the death of the son of Heaven or the prince of a state, it is the rule that the officer of prayer should take the tablets from all other shrines and deposit them in that of the high ancestor. When the wailing is over  .  .  .’ So said Lao Dan (老聃云).”30 Another reads: Zengzi asked, “At a burial, when the bier has been drawn to the path leading to the place, if there happens to be an eclipse of the sun, should there be any change or not?” Confucius said: “Formerly, along with Lao Dan, I was assisting at a burial in the

40 | Liu Xiaogan village of Xiangdang (昔者吾從老聃助葬於巷黨), and when we got to the path, the sun was eclipsed. Lao Dan said to me (老聃曰): ‘Qiu, let the bier be stopped on the right of the road; and then let us stop to wail and wait till the eclipse has passed.’ When it was light again, we proceeded.” (He) said: “This was the rite.” When we had returned and completed the burial, I said to him: “In the progress of a bier there should be no returning  .  .  .” Lao Dan said (老聃曰): “When the prince of a state is going to the court of the son of Heaven, he travels while he can see the sun  .  .  .” This is what I heard from Lao Dan (吾聞諸老聃云).31 These examples will suffice since the other two passages are very similar in style and content, though they deal with different technical details. All the citations of Laozi recall what he said to Confucius regarding the practical details of funeral ceremonies. We find no signs of praise or deprecation between them. Furthermore, we detect no expression of the author’s like or dislike of either. If we read this seriously and without bias, we cannot agree that the meeting of Confucius and Laozi is merely a story created to praise or belittle either figure. Interestingly—and meaningfully—the Laozi makes only one mention of the rites in a positive sense, and that treatment is reserved for funeral rites (“In victory, let us observe the occasion with funeral ceremonies”).32 Is this merely a coincidence? Obviously, these records preserved in Confucian texts concern the facts and techniques of funeral rites and are totally different from the dialogues between Confucius and Laozi in the Zhuangzi. The seven dialogues between the two in the Outer Chapters are among many other stories that show Confucius as a hero or in a supporting role. In the Zhuangzi where he serves as a convenient mouthpiece for the authors of the different chapters and passages, Confucius might promote either Confucianism or Daoism. Although in many of the stories, Laozi is superior to Confucius, their dialogues are exchanges of personal questions, advice, criticism, and talk of enlightenment, not antagonism or confrontation. Due to limited space, we cannot examine the examples from Lüshi chunqiu, Shuoyuan, Hanshi waizhuan, or Kongzi jiayu. However, in brief, their accounts of the meeting between Laozi and Confucius fall into at least three categories. The first is represented by Sima Qian’s biography, which retains historical elements mixed with legend. The second is the records in “Zengzi wen” from the Liji, which contain essentially direct memories about exchanges on technical details of certain rituals. The third is fables in the Zhuangzi, which have nothing to do with history, though sometimes they are based on the credible relationship of historical figures, such as Confucius and his pupils Yan Hui (顏回) and Zilu (子路). The historical literature is rich

Reflections on Textual Analysis in the Post-Graham Era  |  41 and complicated, and we should not take its accounts as entirely reliable or completely useless. Analysis and discrimination are necessary and useful for making the best historical sense of any source.

Assumption 4: Samples can be used as evidence for general judgments. One of Graham’s great contributions comes from his textual studies of the Zhuangzi, in which he convincingly demonstrates that the Inner Chapters were likely written by Zhuangzi himself. This work has influenced many scholars in Zhuangzi and other textual studies.33 In addition to his graceful language and original arguments, most impressive has been his undertaking of the linguistic comparison of exhaustive and detailed language data from the Inner Chapters and those he credited to “the Zhuangzi school” (chaps. 23–27, 32), as well as all the other chapters. This linguistic comparison was based on the distinctive idioms, philosophical terms, persons’ names and general terms, as well as phrases with special grammatical features, found in the Inner Chapters. I believe the persuasiveness of his argument rests on the completeness and sheer mass of the linguistic data. The coincidental similarity between his work and mine in Zhuangzi studies represents proof of the objectivity that can be achieved in linguistic analyses, though my comparison and statistics focus on three philosophical compounds (daode 道德, xingming 姓名, jingsheng 精神) and all passages with distinct and obviously similar language features across the whole book.34 It is sometimes claimed that my work challenges Graham’s, but it was not and could not be my intention, since I had not heard of him when I studied the Zhuangzi. As far as our differences are concerned, I agree with Harold D. Roth, who has said, Perhaps the most original and controversial aspect of Graham’s scholarship on the Zhuangzi is his reconstruction and rearrangement of the text. The reconstruction involves putting passages back together that he felt were originally whole but had been fragmented due to textual corruption and to the drastic editorial work of the fourth-century commentator Guo Xiang.35 Why this is controversial and not as convincing as Graham’s argument about the Inner Chapters rests on three points. First, the division between the Inner and the Outer and Miscellaneous Chapters is a historical heritage; thus, the findings of modern linguistic analyses are merely a test and proof of the general reliability of the conventional arrangement of the book. Second, the linguistic data Graham used to compare different sections of the

42 | Liu Xiaogan text consists of a great number of idioms, terms, and names, and these are used to compare all its sections. Third, the results of linguistic comparison among the different groups give either “yes” or “no” results, which are firm and clear. Thus, the results of this comparison are objective and leave little room for a different interpretation. However, Graham’s attempt to regroup the book, especially to move passages of the Outer and Miscellaneous Chapters “back” to the Inner Chapters, was a completely different project. First, there is little objective or historical reference or basis for his regrouping. Second, he judged the “similarity” between a few passages on the basis of personal observation without establishing a distinct “yes” or “no” line; hence, other scholars may not recognize the same degree of similarity. Third, the causes and origins behind the similarity of themes and certain words might be of many kinds: they might result from imitation by disciples, or arise coincidentally from an interest shared by someone writing much later; they do not necessarily indicate Zhuangzi’s authorship. Fourth, the Zhuangzi is not a book containing essays of one theme or even a consistent style. Its major sections are collections of fables, parables, discourse, narrative, prose, and stories; hence, they don’t have contextual structure by which we can find gaps and put back missing pieces. After all, examples of linguistic similarities and common themes by themselves cannot be the basis of serious textual analysis: they are simply not sufficient to prove an overarching textual structure and connection to a certain author. Samples, no matter how similar, are not adequate for making a decision about the author and structure of a whole chapter. Let us consider Graham’s attempt to move four passages from three Miscellaneous Chapters to reconstruct the introductory paragraph of chapter 3.36 Those four passages are A: Chap. 25 (51–54), B: Chap. 24 (105–111), C: Chap. 32 (50–52), and D: Chap. 24 (103–105).37 Graham eventually gave up entirely on the first passage from chapter 25 when he admitted, “I have since lost confidence in it.”38 Graham’s reconstruction takes both passage B and passage D from chapter 24, but separates and transposes them to come before and after passage C from chapter 25.39 I feel that this rearrangement involves too much speculation and deviates too drastically from the original text in the attempt to try to restore it. The “similarities as evidence” Graham listed40 for his regrouping include: (I) youya (with confines 有涯) and wuya (without confines 無涯) shared by chapters 3 and 24; (II) dai (dangerous [of knowledge] 殆) in chapter 3 and dai (dangerous [exercise of the senses and heart] 殆) in chapter 24; (III) jie (unraveling [of problems] 解) in chapter 24 and jie (dismemberment [of an ox] 解) in chapter 3; (IV) shi qi suo bu zhi (depending on what one does not know [good] 恃其所不知) in chapter 24 and shi qi suo jian (depending on what

Reflections on Textual Analysis in the Post-Graham Era  |  43 one sees [bad] 恃其所見) in chapter 32; (V) the parallel of shen (神), mu (目) in chapter 24, and shen (神), ming (明)41 in chapter 32.42 These five comparisons of linguistic similarity are weak evidence for arguing that the text is a single author’s work. Four of the examples are common words often used in various contexts; among them, in the third example, the usage and meaning of jie is quite different in chapter 3 and chapter 24; in the fourth example, the two phrases are respectively negative and positive, and none of them belongs to chapter 3. So these language materials are not adequate to support the regrouping. Furthermore, even in principle, apparent similarity does not provide sufficient basis to assert that the passages must have been written by the same author or originally belong to the same earlier essay. This is what I call the “weakness of sample argumentation.”43 And yet sample argumentation is common, which is useful within a defined and homogeneous data tank. However, in many cases, ancient Chinese texts are unchartered and heterogeneous; thus, the assumption that piecemeal examples can prove a general conclusion and negate a specific judgment is very risky. It has both quantitative and qualitative weaknesses. Samples usually occur in only a few instances; they cannot be seriously tested and so the reasoning behind invoking them is not logically valid. For example, since Qing times, on the basis of certain terms and phrases, scholars have argued that some chapters of the Zhuangzi were written in the early Han. This became broadly accepted, and its advocates included my supervisor, the late Zhang Dainian (張岱年 1909–2004), and Graham. However, few of the supporting samples and arguments have stood up under serious examination. Representative instances include the “six classics” ( 六經), the “twelve classics” (十二經), the “three August Ones and the five emperors” (三皇五帝), the “uncrowned king” (素王), “prime minister” (宰相), “super immortals” (上仙), “white clouds as hometown of the gods” (白雲帝 鄉), and “Confucius went west to deposit books with the court of Zhou” ( 孔子西藏書於周室). Even after serious linguistic investigation, there has been no positive proof that these terms and phrases can be definitively dated to the Han dynasty.44 Furthermore, even if one shows that these terms indeed appeared in certain Han texts, this is no basis for arguing that they cannot have been occasionally used earlier. Some scholars seem to have made an assumption that a word or phrase used in a historically first instance should appear in many contemporary texts; otherwise, it must be false and belong to a later time when it was broadly accepted. More to the point, a comprehensive linguistic investigation has shown that thirty passages from fourteen chapters of the Zhuangzi are cited in the Lüshi chunqiu and Hanfeizi. This means 42 percent of its thirty-three chapters

44 | Liu Xiaogan were cited in works of the pre-Qin period. This strongly suggests that the book must have been in circulation before the Qin dynasty. Few books have been cited so extensively. Intriguingly, the passages cited from these fourteen chapters are: three from seven of the Inner Chapters, six from fifteen of the Outer Chapters, and five from eleven of the Miscellaneous Chapters. Coincidently, 3:6:5 is approximate to 7:15:11, so the citations are proportional to the three sections of the work.45 This is based on a comprehensive survey of all related language data, not merely on certain samples. In light of the bamboo-slip Zhuangzi discovered at Jiangling (江陵) and Fuyang (阜陽), Li Xueqin (李學勤) also believes that the Inner Chapters were written by Zhuangzi himself, and the Outer and Miscellaneous chapters were also probably pre-Qin works (Li 1998).46 The pieces of bamboo texts are not whole chapters, thus they themselves alone may not be strong enough to prove the dating of the book Zhuangzi; however, if we don’t abandon all literary tradition and modern research of the Zhuangzi, including ­Graham’s work, we may have to admit that Li’s conclusion is right. Here the archaeological findings confirm again that sample argumentation is a risky and highly fallible method. We can use another example to demonstrate the weakness of sample argumentation. In the Laozi, chapter 63 has a sentence that calls for “repaying hatred with virtue” (報怨以德). This is a saying unique to pre-Qin texts. Interestingly, there is a dialogue directly opposing it in the Analects: Someone said: “What do you think of repaying hatred with virtue?” (以德報抱怨) Confucius said: “In that case, what are you going to repay virtue with? Rather, repay hatred with uprightness and repay virtue with virtue” (以直報怨, 以德報德).47 That someone asked Confucius about “repaying hatred with virtue” suggests that this was an idea in circulation that Confucius had not yet discussed. Among extant texts, only the Laozi advances the idea of “repaying hatred with virtue,” and according to the common method of textual analysis we criticized earlier, one might conclude that it was the Laozi that sparked Confucius’s comments, and this might further be taken as evidence that the author of the Laozi was a contemporary of Confucius. But this is a weak proof. First, the supporting examples are few. Second, we cannot exclude the possibility that someone else at that time or earlier had the same idea, though the text in which it appeared might no longer be available to us. Generally speaking, isolated samples are not adequate to support a general conclusion about a whole work. •

Reflections on Textual Analysis in the Post-Graham Era  |  45 The reflections in this essay are mainly based on the evidence of archaeological data. It is a pity that Graham did not know about the discovery of the bamboo-slip versions of the Laozi and, more generally, the many discoveries of ancient texts across China.48 Even so, Graham expected the research to move forward, which he expressed in the comment, “It is to be hoped that future scholarship will develop a more sophisticated technique for dealing with him [Zhuangzi].”49 The present author hopes the preceding reflections are useful toward developing “a more sophisticated technique” for dealing with not only the Zhuangzi, but more generally for strengthening the many fields that rely on textual analysis.50 Certainly, it does not follow from the previous reflection that one cannot or should not have any assumptions in scholarly works. The point is that our assumptions need examination, especially when new knowledge and references, such as archaeologists’ findings, are made available. The present author certainly also has some assumptions, which, however, have been examined and reexamined by himself, are subject to further examinations by colleagues, and would be corrected or abandoned if they are proved groundless.

Notes   1. Henry Rosemont Jr., Preface for A Companion to Angus C. Graham’s Chuang Tzu, ed. Harold D. Roth (Honolulu: University of Hawai’i Press, 2003), ix.   2. Liu Xiaogan, ed., Dao Companion to Daoist Philosophy (New York and London: Springer, 2015).   3. My purpose in textual studies is to evaluate various types of evidence and find new pieces of evidence for more plausible positions. If we don’t have enough evidence, we may remain silent. Certainly, scholars can present various speculations as a kind of academic exercise, though I am not interested in that game. Those who are interested in my arguments about textual issues of the Laozi and Zhuangzi, please refer to “Did Daoism Have a Founder? Textual Issues of the Laozi” and “Textual Issues in the Zhuangzi,” in Dao Companion to Daoist Philosophy, 25–45 and 129–157.   4. Qian Mu 錢穆, Zhuanglao tongbian 莊老通辨 (A reexamination on the relation of the Zhuangzi and the Laozi) (Taipei: Lianjing Press, 1994).   5. A. C. Graham, Disputers of the Tao: Philosophical Argument in Ancient China (La Salle: Open Court, 1987), 217–218. In this essay, I replace Wade-Giles romanization with pinyin in all quotes, and all the emphases in quotes are mine.   6. When Giles mentions Hanfeizi’s work, he writes, “[Hanfeizi] devotes the best part of two whole sections to ‘Explanations of Laozi’ and ‘Illustrations of Laozi’; and, in two places, writes as though he were consulting a written document.” Herbert A. Giles, “The Remains of Lao Tzǔ (Laozi),” The China Review 14.5 (1886): 231–281.   7. My idea that absence of evidence is not evidence of absence is specifically for the context of studies of Chinese texts. If we use it as a general rule, it needs

46 | Liu Xiaogan further discussion. For example, whether we should take the fact that there is no evidence of ghosts as evidence that there are no ghosts.   8. For example, in 2011 Esther Klein published an article suggesting that the conventional view that the Zhuangzi’s Inner Chapters contain the core and earliest sections of the Zhuangzi book may be totally wrong; those chapters could have been selected by a Han dynasty editor. She pays little attention to the huge number of texts excavated in recent decades. Her conclusions are similar to those of Giles and Graham, which rest on the presuppositions mentioned earlier. Esther Klein, “Were There ‘Inner Chapters’ in the Warring States? A New Examination of Evidence about the Zhuangzi,” Tong Pao 96 (2011): 299–369.   9. Liu Xiaogan, Foreword to the Reprint Edition, for Donald J. Munro, The Concept of Man in Early China (Ann Arbor: Center for Chinese Studies, The University of Michigan, 2001), xi–xxvi. 10. Sun Bin is believed to have flourished in roughly the same era as King Hui of Liang according to historical records related to him. 11. Qian Mu, “An Investigation of Tian Ji, Zou Yan, and Sun Bin” 田忌鄒衍孫臏 考, Chronology of Scholars in Pre-Qian 先秦諸子系年 (Taipei: Lianjing Press, 1994), 304– 305. 12. D. C. Lau, Tao Te Ching: A Bilingual Edition (Hong Kong: The Chinese University Press, 2001), 140. 13. Ibid. 14. Ibid., xi. 15. Arthur Waley, The Way and Its Power (London: George Allen & Unwin, 1934), 108. 16. Lau, Tao Te Ching,132. 17. Chen Guying 陳鼓應 and Bai Xi 白奚, Laozi Pingzhuan (A critical biography of Laozi) (Nanjing: Nanjing University Press, 2001), 9. 18. A. C. Graham, Studies in Chinese Philosophy and Philosophical Literature (Albany: State University of New York Press, 1990), 111–124. 19. Ibid., 111. 20. Ibid., 124. 21. Ibid. The following quotations are from the same page. The footnotes are omitted. 22. Sima Qian 司馬遷, Shiji (Records of the grand historian) (Beijing: Zhonghua Shuju, 1975), 2142. 23. Mark Twain, page from 1917 calendar published by Sully and Kleinteich of New York by arrangement with Harper & Brothers; from the Dave Thomson Collection, http://www.twainquotes.com/Fiction.html. 24. Wikipedia, http://en.wikipedia.org/wiki/Stockholm_syndrome. 25. Lau, Tao Te Ching, 130. 26. Graham, “The Origins of the Legend of Lao Dan,” in Studies in Chinese Philosophy and Philosophical Literature, 111–124. 27. Literary elements are not necessarily negative. They may be helpful in aiding understanding and revealing historical truth. Even modern historians rely on literary techniques. For example, Jonathan D. Spence and Ray Huang both write about Chinese history using their storytelling skills and literary talent, which strengthens

Reflections on Textual Analysis in the Post-Graham Era  |  47 their historical interpretation rather than weakens the trustworthiness of their work (and helped their books become bestsellers). 28. Kongzi jiayu and Shuoyuan were traditionally considered apocryphal late Han works. However, numerous bamboo slips from an early Han tomb unearthed in 1973 at Ding Xian 定縣, Hebei 河北, contain passages identical to those in both those works. This suggests that the contents of these books were collected from pre-Qin or early Han sources. 29. For this part of my argument I will leave out the Zhuangzi. It is common knowledge that the Zhuangzi is not a historical work, though it mentions by name many historical figures. In fact, historical figures are freely invoked throughout it for various authorial purposes. 30. Sun Xidan 孫希旦, Liji jijie 禮記集解 (Beijing: Zhonghua shuju, 1998), 523–524. James Legge, The Sacred Books of China: Texts of Confucianism, part 3, the Li Ki, I–X (Delhi: Motilal Banarsidass, 1968), 324–325. The translation of the Liji is adapted from Legge with minor amendments for readability and accuracy. 31. Sun Xidan 孫希旦, Liji jijie, 545–546. Legge, The Sacred Books of China: Texts of Confucianism, 338–339. 32. Wing-tsit Chan, A Source Book in Chinese Philosophy (Princeton: Princeton University Press, 1973), 155. 33. Harold David Roth, ed., A Companion to Angus C. Graham’s Chuang Tzu (Honolulu: University of Hawai’i Press, 2003), 3–4. 34. Liu Xiaogan, Classifying the Zhuangzi Chapters (Ann Arbor: Center for Chinese Studies, The University of Michigan, 1994). Liu, Dao Companion to Daoist Philosophy, 129–157. 35. Roth, A Companion to Angus C. Graham’s Chuang Tzu, 186. 36. Graham, Studies in Chinese Philosophy and Philosophical Literature, 296–297. 37. Ibid., 296–301. The number in parentheses is the line location in the Chinese text in his source edition, Zhuangzi yinde (莊子引得 Concordance to Chuang-tzu; HarvardYenching Institute Sinological Index Series, Supplement No. 20). 38. Roth, A Companion to Angus C. Graham’s Chuang Tzu,18. 39. Graham, Studies in Chinese Philosophy and Philosophical Literature, 296. A. C. Graham, Chuang-tzŭ: The Inner Chapters (Indianapolis: Hackett, 2001), 62–63. 40. In the following, the translations and explanation of the Chinese characters, including the use of brackets, are from Graham. 41. Graham glosses the parallel of shen and mu and shen and ming as “daemon preferred to eye” and “daemon preferred to eyesight,” respectively. These are difficult to understand out of context. 42. Roth, A Companion to Angus C. Graham’s Chuang Tzu,18. 43. Liu, Dao Companion to Daoist Philosophy, 36–40. 44. Liu, Classifying the Zhuangzi Chapters, 75–82. 45. Ibid., 50–61. 46. Li Xueqin 李學勤, “Zhuangzi zapian zhujian ji youguan wenti” 莊子雜篇竹 簡及有關問題 (The bamboo slips of miscellaneous chapters of the Zhuangzi and the related problems), Journal of the Shaanxi History Museum 陝西歷史博物館舘刊 5 (June 1998): 126–131. 47. Chan, A Source Book of Chinese Philosophy, 42.

48 | Liu Xiaogan 48. The Guodian bamboo texts were published in 1998, by which point Graham had passed away. 49. Graham, Studies in Chinese Philosophy and Philosophical Literature, 284. 50. The work described in this essay was partially supported by a grant from Peking University’s Project on the History of Chinese Hermeneutics (12&ZD109). I am also grateful that the Claremont School of Theology, especially its president, Professor Jeffrey Kuan, invited me to teach there for a semester in Spring 2015, where I had a pleasant time and completed the draft of this essay.

3

Cognitive Attunement in the Zhuangzi

Harold D. Roth

I suspect that some states that we call mystical are as natural as the relief of sorrow by tears and are suppressed by certain cultures, much as grown men may be forbidden to weep. —A. C. Graham, “Mysticism and the Question of Private Access”

I first met Angus Graham on a bright day in the late autumn of 1981. I was an unemployed PhD working on a series of radio programs for Ideas, a popular show produced by the Canadian Broadcasting Corporation.1 I had persuaded the CBC to send me to England to interview some of the great sinologists of the day and Graham, of course, was one of them. I had been a huge fan since my last year at university, when his initial translation of the Qiwulun, published in the History of Religions 1968 Bellagio Conference issue, had helped me crack the Zhuangzi’s coded secrets (or so I thought). After entering his office at SOAS, I picked my way carefully through the narrow pathway from door to desk, dodging wobbly stacks of book and piles of offprints and photocopied articles. Having been advised by Sarah Allan, whom I had interviewed the previous day, that the good professor was fond of cognac and cigars, I placed them before him on his desk and then proceeded to ready my battery-operated cassette recorder and handheld mic. Almost three hours later we were sitting in the near dark. The cognac bottle was entirely emptied, the cigars were gone, and a deep smoky haze filled the room. Professor Graham was holding forth on the topic of how Zhuangzi would have driven an automobile. I can still hear the twinkle

49

50  |  Harold D. Roth in his voice as he went on about Zhuangzi: “He would go here . . . and there . . . but never arrive anywhere.” Signaling the end of the interview, Graham stood up and said, “Let’s go to dinner. You do have an expense account, don’t you?” I saw the look of great anticipation disappear from this face as I shook my head back and forth. “Oh, very well,” he said, resignedly, “let’s get some pasta!” One of the topics on which he expounded that late fall afternoon was his idea that the insights from the Qiwulun about the relativity of things from the perspective of the Way, or the person who had embodied the Way, bore certain distinct resemblances to the twentieth-century discovery of the Heisenberg principle of indeterminacy in physics. In Graham’s mind, the insights of a Chinese philosopher from two and a half millennia ago could be relevant to modern philosophers as well as scientists because of certain basic elements of human experience and of the natural world that transcend culture and time. Graham’s wide-ranging curiosity and fascination with both scientific and humanistic knowledge enriched his own scholarship and broadened Chinese textual analysis. I have no doubt that he would have been a supporter of the new academic field of contemplative studies,2 which is devoted to the study of human contemplative experience across cultures and through history, drawing on both scientific and humanistic perspectives. Contemplative experience was most definitely a subject that Angus Graham and I discussed numerous times. He even wrote an essay on the subject, published in the posthumous collection edited by David Hall, Reason within Unreason.3 In this essay, Graham wrote of his inherent fascination with mystical experience yet expressed skepticism about the epistemological validity of something so subjective: “The great obstacle to the man of reason in coming to terms with mysticism is its appeal to the authority of an experience outside the public domain, and least accessible perhaps to the analytic cast of mind.”4 He then proceeded to describe some of his own experiments with what he ended up calling “The Common Felicity,” a deliberately cultivated, momentary, and direct cognitive apprehension of the fleeting moments of beauty in one’s environment. While not claiming them as precisely “mystical,” Graham saw in their spontaneity and profundity that “human capacity for awareness [is] raised to its highest pitch.”5 Using the language of reason, Graham was trying to identify the existence of a commonly accessible and immensely significant mode of experience that is, in itself, not inherently rational. I call this experience “cognitive attunement”6 and discuss a more recent and scientifically informed way of presenting it, via the lens of contemplative studies. I begin by elucidating a model for “contemplative phenomenology,” a set of philosophical and scientific ideas that provide the intellectual foundation for this emerging

Cognitive Attunement in the Zhuangzi | 51 field. Then I apply this model to demonstrate that cognitive attunement is the central theme around which the entire Zhuangzi collection is organized.

Contemplative Phenomenology Contemplative studies examines the philosophy, psychology, and cognitive activity of contemplative experience. It focuses on the many ways human beings have found to concentrate, broaden, and deepen conscious awareness as the gateway to cultivating more meaningful, ethically responsible, and personally fulfilling lives. Scholarship in this emerging field attempts to identify the varieties of contemplative experiences, find meaningful scientific and philosophical explanations for them, cultivate first-person knowledge of them, and critically assess their nature and significance.7 As we define it, “contemplation” includes the focusing of the attention in a sustained fashion leading to deepened states of concentration, tranquility, insight, and “contextualizing” orientations. Such experiences occur on a spectrum from the rather common experiences of absorption in an activity such as reading a book or playing a sport to the profound and transformative experiences that are deliberately cultivated within religious traditions.8 When intentionally cultivated, contemplative experiences can become the basis of a clear and spontaneous cognition that is able to attend effortlessly to whatever presents itself. Such experiences can also serve as the basis for the development of various “other-regarding” ethical orientations, such as empathy and compassion, love and loving-kindness. Results of this type have become the basis for serious scientific research in areas such as effortless attention, flowing cognition, mindfulness, and compassion.9 Contemplative studies derives its orientation to experience from the work of the late cognitive neuroscientist Francisco Varela, one of the founders of the Mind and Life Institute. He was influenced by both phenomenology and the cognitive sciences. Varela makes a crucial distinction between firstperson and third-person events and experiences: By first-person events we mean the lived experience associated with cognitive and mental events. Sometimes terms such as “phenomenal consciousness” and even “qualia” are also used, but it is natural to speak of “conscious experience” or simply “experience.” These terms imply here that the process being studied (vision, pain, memory, imagination, etc.) appears as relevant and manifest for a “self” or “subject” that can provide an account; they have a subjective side.

52  |  Harold D. Roth In contrast, third-person descriptions concern the descriptive experience associated with the study of other natural phenomena. Although there are always human agents in science who provide and produce descriptions, the contents of such descriptions .  .  .  are not clearly or immediately linked to the human agents who come up with them.  .  .  .  Such “objective” descriptions do have a subjective-social dimension, but this dimension is hidden within the social practices of science. The ostensive, direct reference is to the “objective,” the “outside,” the content of current science that we have today concerning various natural phenomena such as physics and biology.10 Contemplative studies as a field embraces the recommendations of Varela and the philosopher Evan Thompson that serious scientific and philosophical study of contemplation must take into account first-person subjective experience.11 A second significant distinction that arises from this fertile interface between cognitive science and phenomenology is the interlinked concepts of second-person experience and intersubjectivity. This elaboration of the “I-You” hypothesis asserts that human beings are fundamentally and biologically social animals—something not at all foreign, of course, to Mencian Confucianism. And that, as Evan Thompson argues, “individual human consciousness is formed in the dynamic interrelation of self and other, and is therefore inherently intersubjective.”12 Danish philosopher Dan Zahavi adds: “The phenomenologists never conceive of intersubjectivity as an objectively existing structure in the world which can be described and analysed from a third-person perspective. On the contrary, intersubjectivity is a relationship between subjects which must be analyzed from a first-person and secondperson perspective.”13 In other words, intersubjectivity is not merely an abstract intellectual concept: it can only be known and understood by doing, by astute observations of lived interactions with others. Furthermore, Thompson concludes that intersubjectivity inherently involves empathy: “The concrete encounter of self and other fundamentally involves empathy, understood as a unique and irreducible kind of intentionality.”14 Cognitive scientists and neuroscientists have been very interested in developing measures of empathy and compassion and studying ways to entrain it. Their work in the past few decades has been copious.15 Therefore, from these perspectives, third-person experience entails an activity of consciousness that objectifies an apparently external world of things and facts as well as an internal world of ideas and feelings. This third-person objectification of an internal world of experience is significant in the construction of the idea of a fixed or permanent self-identity; it is also

Cognitive Attunement in the Zhuangzi | 53 the root of self-consciousness. Second-person experience involves intersubjective communication that at its best is spontaneous and unself-conscious and is grounded in empathy inherent in the I-You relationship. First-person experience is the basis of subjectivity; the experience of what we commonly perceive as the internal contents of our consciousness or minds. Consistent with this model, I will add a fourth dimension of consciousness: what I call “no-person experience.” “No-person experience” is immediate experience in the present moment that is nonintentional and non-self-referential. It occurs within the experiencing subject, within first-person experience on a moment-to-moment basis. I have derived this concept from a number of extant sources, both Western and Eastern. The first is William James’s notion of “pure experience” as expressed in the following passage: “The instant field of the present is always experienced in its ‘pure’ state, plain, unqualified actuality, a simple that, as yet undifferentiated as to thing and thought.”16 Yoko Arisaka goes on to say: “According to James, ‘pure experience’ is not a subjective experience, but a ‘a simple that,’ an immediate thisness of experience which can be taken as subjective or objective states.”17 When one reflects on the activity of knowing, experience splits into two terms, one of which becomes the knowing subject and the other the known object. Thus, for James, prior to the retrospective experience that divides into subject and object, pure experience is itself a unity that is neither subject nor object.18 Influenced by James and his understanding of Zen metaphysics, Kitaro Nishida also put forth his own concept of “pure experience.” For him, in contrast to James, while pure experience is both prereflective and immediate, it transcends both subject and object in such a way that it does not contain any specific individual content: “The difference between ‘my’ flow and ‘your’ flow is not a fundamental fact, but is the result of the different ways in which the same pure experience is abstracted.”19 Thus, in pure experience, one cannot make absolute distinctions between oneself and another. So, for Nishida individuality emerges from a continuum of undifferentiated experience: “It is not that experience exists because there is an individual; but that an individual exists because there is experience.”20 In my analysis, the difference here is that James, as an empiricist, does not question the given nature of the external world, while Nishida, influenced by the Yogācāra-based idealism that lingers in Zen metaphysics, does. For James, pure experience has content, it’s noetic. For Nishida, pure experience is devoid of any specific content, as is true for the Yogācāra concept of cittamatra (mind-ground).21 Recent research in the brain sciences supports James’s understanding of pure experience.22 This research has identified two types of cognition that recognize environmental data spontaneously and without intentionality. First, Aaron Seitz and Takeo Watanabe have pioneered research on what they call task-irrelevant perceptual learning (TIPL). Second, Arthur Reber has

54  |  Harold D. Roth researched a major mode of cognition, “tacit knowledge,” that was initially suggested by philosophers Gilbert Ryle and Michael Polanyi. What is taskirrelevant perceptual learning? In their abstract the authors state: “The basic phenomenon is that stimulus features that are irrelevant to a subject’s task (i.e. convey no useful information to that task) can be learned due to their consistent presentation during task-performance.”23 TIPL does not stand in opposition to theories of attention but operates instead in concert with attention. Where attentional learning is best to enhance (or suppress) processing of stimuli of known task relevance, TIPL serves to enhance perception of stimuli that are originally inadequately processed by the brain. In this phenomenon low-level sub (attentional) threshold cues are perceived without the subject deliberately attending to them. If these cues are raised to the attentional threshold, they are not noticed or learned because of the greater strength of the cues to which the subject is paying attention. Thus, TIPL is a form of passive cognition in which information is processed while deliberately attending to a different, much stronger, stimulus. As Seitz writes, Task-irrelevant perceptual learning is distinguished from other forms of perceptual learning in that it is not dependent on attention being directed toward, nor having awareness of, the stimulus array during the period of learning. Stimulus arrays that are unnoticed, and even subliminally presented, can be learned through task-irrelevant perceptual learning when presented in conjunction with appropriate reinforcing events.24 “Tacit knowledge” is literally “knowledge that cannot be put into words.”25 This broad definition covers four general types: knowledge or know-how that

1. Often cannot be made explicit or codified (e.g., language learning);



2. May be made explicit or codified, but we don’t understand how to achieve it or how to explain it (e.g., problem-solving);



3. We exercise without being aware that we are exercising it, such as in automatic habits or routines that we carry out, often very skillfully, but with a minimum of conscious attention (e.g., driving, catching a baseball);



4. We exercise without being able to explain how we do it (e.g., walking, running).

Cognitive Attunement in the Zhuangzi | 55 Tacit knowledge is often contrasted with propositional knowledge, which is derived from logical analysis that separates itself from what it is analyzing. It is a “knowing how” that involves embodied understanding often put into action; whereas propositional knowledge is a “knowing that.” The contrast between tacit and propositional knowledge parallels the contrast between first-person and third-person experience. The research on TIPL points to perception and cognition that occurs without being actively pursued within the context of a self or of self-identity. It is nonintentional and hence spontaneous. I postulate that something like this is going on in “no-person experience” as well as in James’s concept of “pure experience.” A perceptual world is cognized on a moment-to-moment basis before we bring that cognition to full awareness. The research on tacit knowledge points to cognition and action that occurs spontaneously and nonintentionally that leads to the automatization of a whole range of tasks we perform daily, from brushing our teeth to driving a car. It is impossible to pay attention to every single item of perceptual data that arises in performing these tasks. When we are first learning them we inevitably have to be much more intentional about taking in these data. But after they become automatic we sometimes perform them without much intentionality or selfconsciousness at all. Psychologists have been researching a related area of experience that they refer to as “effortless attention.” To quote Brian Bruya: Under normal circumstances, the expectation is that expenditure of effort increases with the level of demands until effort reaches a maximum point at which no increase is possible.  .  .  .  Sometimes, however, when the level of demand reaches a point at which one is fully engaged, one is given over to the activity so thoroughly that action and attention seem effortless.26 Mihalyi Csikszentmihalyi’s extensive research focuses on the broader phenomenon of effortless, autotelic (self-reinforcing) experience across a wide range of human activities spanning all cultures and times. Ranging from sports to games of chance to deep listening and intimate conversations, he identifies a common and rewarding experience achieved in all of them he calls “flow,” or “optimal experience.” “Flow states” are characterized by an often nonintentional focusing of the attention, an awareness centered in the present moment, a complete absorption of body and mind in the activity at hand that is infinitely self-rewarding. Psychological research on effortless attention and flow states ranges from self-reporting measures to designing environments and methods that induce such states deliberately, as can be seen in the range of scientific and philosophical articles that Bruya has

56  |  Harold D. Roth c­ollected. In most instances, meditation and yoga techniques are not used in this research. Finally, the neuroscientific study of meditation has made an important distinction between “states” of experience and “traits” induced by these states.27 “State” refers to the altered sensory and cognitive awareness that can arise during meditation practice. “Trait” refers to the lasting changes in these dimensions that persist in the meditator irrespective of being actively engaged in meditation. Regular meditation practice can produce relatively short-term states as well as long-term changes in traits. According to Rael Cahn and John Polich: State changes from the meditative and religious traditions are reported to include a deep sense of calm peacefulness, a cessation or slowing of the mind’s internal dialogue, and experiences of perceptual clarity and conscious awareness merging completely with the object of meditation, regardless of whether a mantra, image, or the whole of phenomenal experience is the focal point. . . . Trait changes from long-term meditation include a deepened sense of calmness, increased sense of comfort, heightened awareness of the sensory field, and a shift in the relationship to thoughts, feelings, and experience of self.28 Neuroscientific research has identified two basic categories of meditative practices: “concentrative” and “receptive.”29 Concentrative meditation techniques involve a top-down focusing of the attention on a specific mental or sensory activity: a repeated sound, an imagined image, or specific body sensations such as the breath. Repeated practice over time can lead to one-pointed modes of attention. Receptive meditation is a more effortless, involuntary, and inclusive form of practice in which one learns to allow any thoughts, feelings, and sensations to arise while remaining an unattached observer without judgment or analysis. It taps into spontaneous pre-attentive intuitive mechanisms and can lead to forms of “choiceless awareness” free of thoughts.30 The results of these practices include a shift from what James Austin calls a self-referential “egocentric” cognitive orientation to an otherreferential “allocentric” orientation. In addition to improved attentional focus, these results include increased bodily awareness, improved emotional regulation through decreased reactivity, and “detachment from identification from a static sense of self.”31 Neuroscientific research often focuses on studying meditative states by examining brain activity during meditation. Third-person measuring devices such as the EEG (electroencephalogram) fMRI (functional magnetic resonance

Cognitive Attunement in the Zhuangzi | 57 imaging) and MEG (magnetic encephalography) are used. Research on meditative traits focus on the long-term implications of meditative practice by comparing the brains of experienced and novice meditators to nonmeditating controls or asking these groups to react to various stimuli during and after meditation. In addition, Varela and Thompson have pioneered the field of “neurophenomenology,” which pursues research that combines these thirdperson measures with simultaneous first-person reports.32 To sum up: the contemplative phenomenology outlined identifies four modes of experience: third-person, second-person, first-person, and no-person experience. “No-person experience” is nonintentional, non-self-conscious, and immediate. It resembles James’s notion of “pure experience” in that it is prereflective, yet it is unlike Nishida’s notion in that it is noetic and provides specific content. Cognitive scientific research on “task-irrelevant perceptual learning” and “tacit understanding” support these models of consciousness in demonstrating that perception and cognition can occur that is momentary, spontaneous, nonintentional, and nonpropositional. Psychological research on “effortless attention” and on “flow” explore states of nonintentional cognition, and spontaneous and concentrated action that could be viewed as similar to “no-person experience” and to a range of experiences of “tacit knowledge.” Finally, neuroscientific research on meditation differentiates between states (temporary experiences of concentration) and traits (longer-lasting alterations in cognitive functioning). It identifies two modes of contemplative practice: concentrative and receptive; intentional attention-focusing that leads to a narrower, one-pointed awareness; and an opening of the field of attention to embrace thoughts, feelings, and perceptions without judgment that leads to a choiceless awareness. Results of these practices lead to a shift from egocentric modes of information processing to allocentric and to increased detachment from a static sense of self-identity. These philosophical and scientific perspectives lead to what I call a condition of “cognitive attunement” that enables one to live free of attachment to rigid, egocentric third-person objectifications of self and other, free from excessive self-consciousness, able to act effortlessly, without intentionality, in a mode of no-person orientations to first-person, second-person, and third-person experience.

Cognitive Attunement in the Zhuangzi I approach the Zhuangzi presupposing a number of hypotheses about the text and composition that Angus Graham presented and that my own work elaborated upon:33

58  |  Harold D. Roth

• That this is a collection written by a number of distinct authorial voices compiled over perhaps one and one-half centuries;



• That the text contains a pre-Han core, much of which represents the ideas of (but not necessarily the writings of) an initial author or authors and several later generations of disciples;34



• That the work was compiled into a fifty-two-chapter original recension at the court of Liu An, second king of Huainan and sponsor and contributor to the work that bears his name, the Huainanzi;35



• That the apparent disparate authorial voices in the text, for the most part, share not only a very specific intellectual lineage of master and disciple, but they also share a common set of contemplative practices and goals;



• That identifying the nature and results of said practices is essential for understanding the organizing principles of the text and constitutes an important missing piece of the jigsaw puzzle that is the pre-Han classical Daoist tradition. I propose and then argue for the following hypotheses:



1. “Cognitive attunement”36 is the central theme and main goal advocated in most of the Zhuangzi, including the sections Graham identifies—and to which I concur—as being representative of the historical Zhuang Zhou (chaps. 1–7); the sections created by the people whom I have argued are his lineal descendants: the “Primitivist” (chaps. 8–11.5), the “Syncretist” (chaps. 12–16, 33), and the authors of the “School of Zhuangzi” materials (chaps. 17–22 and parts of chaps. 24–27).



2. Cognitive attunement is a trait change that involves a transformation of cognitive processing away from a self-conscious reliance on third-person doctrines, dogmas, and self-analyses to an increasing reliance on the phenomenology of the “noperson” perspective. It also includes the development of effortless attention and flow-like optimal experience.



3. Cognitive attunement is a trait change that results from the apophatic contemplative practices discussed in the text.

Let’s examine the evidence in the text for these hypotheses. The goal of a free and fluid consciousness that derives from a fundamentally different

Cognitive Attunement in the Zhuangzi | 59 mode of cognition can be found throughout the text; motifs and exemplars abound in all sections. To begin with, we have the famous “free and easy wandering” of the first chapter that, I would contend, epitomizes a freedom from being confined to the narrow set of third-person conceptual analyses of oneself and the world that restrict our vision and creativity like the cicada, the turtle dove, and the quail, who cannot conceive of the possibility of— nor imagine the perspective of—the giant Peng bird.37 This freedom that comes from being unconfined to one third-person way of looking at things enables the imagination that permits the stranger who recommended the hand salve to the king of Wu to be enfeoffed, Zhuangzi to see the value in his friend Huizi’s gourds, and Xu You to refuse the empire offered to him that all others would value. Zhuangzi herein links this freedom of the spirit to being selfless: The utmost (zhiren 至人) are selfless (wuji 無己), The daemonic (shenren 神人) take no credit for their deeds, The sagely (shengren 聖人) are nameless (wuming 無名).38 I would aver here that being selfless means these exemplars are exceptional human beings (ren 人) who are completely free of attachment to a fixed selfidentity, which, from the standpoint of a contemplative phenomenology, is a third-person construction abstracted out of the continuous flow of experience. They are able to allow it to continuously arise and to pass away and thus have no interest in gaining credit or fame. To be selfless here implies that they are unself-conscious in their actions in the world. The characteristic of selflessness associated with this altered cognition is found scattered throughout the text. For example, in chapter 7 we find the exemplar of the Nameless Man, who gives this advice: “let your heart roam with the flavourless, blend your energies with the featureless, in the spontaneity of your accord with other things leave no room for selfishness . . .”39 In chapter 11 we find the teaching that Great Persons (daren 大人) [i]n describing and sorting out shapes and bodies remain joined with them in ultimate sameness. In ultimate sameness they have no self, and without a self, from where would they get to have anything? The man who perceives something is the “gentleman” you were yesterday; the man who perceives nothing is the friend of heaven and earth.40 This is dense but worth unpacking. If we take “having no self” to mean “being without self-consciousness,” then we can see that these cognitive exemplars do differentiate things and persons as objects in their environment.

60  |  Harold D. Roth Yet they do not divide the world up into rigid ideas of self and other. People who perceive others in opposition to a self (the “gentleman” of yesterday) are contrasted with people who do not do this and who are therefore friends of all things. This contrast is also found in chapter 12 wherein Lao Dan instructs Confucius that “to forget things and to forget heaven is called forgetting the self. The person who forgets himself may be said to have entered heaven.”41 From the viewpoint of contemplative phenomenology, to “forget things and forget the self” means to not objectify them from the third-person perspective. This lack of self-consciousness allows one to not become attached to one’s constantly changing self-identity. We see here and repeatedly throughout this text, despite the different authorial voices contained within it, that this liberated way of cognizing the world and oneself is referred to by being “heavenly.” It is frequently contrasted with the human. In chapter 4 after Confucius teaches Yanhui about the contemplative practice he calls “the fasting of the mind,” (xinzhai 心齋) he makes this contrast: What has man for agent is easily falsified; what has Heaven for agent is hard to falsify. You have heard of using wings to fly. You have not yet heard of flying by being wingless; you have heard of using the wits to know, you have not yet heard of using ignorance to know.42 If you act from the usual human perspective, you usually act falsely; acting from the heavenly perspective means you usually act in truth. This genuine action is metaphorically equated with flying wingless and using ignorance to know. These are not just meaningless paradoxes, they point to the uniquely transformed cognition that is advocated throughout the text: cognition that knows completely without self-consciousness. In terms of contemplative phenomenology, it’s acting in the world from the no-person perspective. This contrasting of human and heavenly perspectives is found in many other passages throughout the various authorial voices in the text. In chapter 6 we find the “True Men of Old,” who clearly grasped “what is Heaven’s doing and what is man’s.” In contrast to common people who breathe from their throats, these sages breathed “from their heels,” had few desires and cravings, and in them “the dynamism of the Heavens” (tianji 天機) was deep. They did not know how to be pleased that they were alive, did not know how to hate death . . . were pleased with the gift they received, but forgot it as they gave it back. It is this that is called “not allowing the thinking of the heart to damage the Way, not

Cognitive Attunement in the Zhuangzi | 61 using what is of man to do the work of Heaven” . . . when unified they were of Heaven’s party; when divided were of man’s. Someone in whom neither Heaven nor human is victor over the other, this is what is meant by the True Person.43 This contrast is brought out further in the final passage of the previous chapter in which Zhuangzi schools his Logician friend Huishi in the “essentials of being human.” For Zhuangzi, these consist in the very problematic dividing up of the external world into arbitrary categories of “that’s it” (shi 是) and “that’s not” (fei 非), which come about from the attachment to third-person judgments of value based on a fixed self-identity. These cause ordinary humans to “inwardly wound themselves by likes and dislikes.” By contrast true sages “constantly go by the spontaneous and do not add anything to the process of life.” [These sages] buy at the market of Heaven . . . (which means) to be fed by Heaven. Having received their food from Heaven, what use have they for the human? They have the shape of human beings but are without what is essentially human  .  .  .  and therefore “that’s it, that’s not” are not found in their persons. Indiscernibly small, that which attaches them to humans. Unutterably vast, the Heaven within them which they perfect in solitude . . .44 For Zhuangzi, the “human” is associated with bias, selfishness, selfinjury through personal preferences and desires, one-sided judgments, shallow breathing, overintellectualization, and lack of spontaneity. The “heavenly” is associated with the opposite characteristics, including deep breathing, lack of self-injury through preferences and desires, and a spontaneity and dynamism that come from not allowing thinking to block the Way and from a contemplative experience of complete solitude. We will return to these foundations in contemplative practice. For the moment, in terms of contemplative phenomenology, the state and trait that is being commended in these passages is one that we well recognize from the scientific literature on effortless attention and flow. It suggests precisely the kind of nonintentional subjectivity that allows true sages to abandon objectification of both self and world via third-person analysis and to manifest the unself-conscious mode of “no-person experience.” This recalls the description of the Great Person in chapter 11: “In describing and sorting out shapes and bodies remain joined with them in ultimate sameness. In ultimate sameness they have no self, and without a self, from where would they get to have anything?”45 Ultimate sameness without self refers to this being united with the flow of experience via this unself-conscious “no-person experience.” I see it as also

62  |  Harold D. Roth linked to “tacit knowledge,” in which cognition occurs and activities are performed effortlessly and without intentionality. Throughout the Zhuangzi this commended cognitive trait is also associated with manifesting de (Power; Potency 德). Graham defines this in early Chinese philosophy as “the capacity to respond without reflection according to the Way.”46 In the inner chapters we see this emphasis most clearly in the deformed and mutilated exemplars of fearless true sages in chapter 5, people like chopped-foot Wang Tai, who is “aware of the Flawless” . . . not moved by death and life and who “lets his heart go roaming in the peace which is from the Power.” Another example is “Uglyface T’o,” as Graham translated his name, instantaneously perceived as trustworthy enough to hand over a state to because he “keeps the Power whole” inside himself. An emphasis on the importance of cultivating a cognition based on “the Power” is one of the defining philosophical tenets of the “Primitivist” writings, one that also links it to the cognitive attunement we have seen elsewhere in the text. In chapter 8 the Power is basically defined in the phrase Graham sees as quintessentially associated with this author: xingming zhi qing (性名之情), which he translates as “the essentials of our nature and destiny.” I would explain this further as “the spontaneous responses of our instinctual nature in response to things that happen to us in life.” In chapter 8 our essential human nature is for all the senses to respond spontaneously and harmoniously on their own without the interference of social and ethical dicta, such as those provided by Confucians: When I call someone a fine man, it is not Goodwill and Duty that I am talking about, but simply the fineness in his powers; nor when I call someone is it the Five Tastes that I am talking about but simply a trust in the essentials of our nature and destiny.47 This is another way to talk about the kinds of “tacit knowledge” and “effortless activity” that apprehend the world clearly when not interfered with by fixed and rigid categories and dicta about human moral behavior. Chapter 9 talks about “The Age when Power was at its utmost,” a primitive Utopia when people lived in sameness with the birds and animals, side by side as fellow clansmen with the myriad creatures . . .  In sameness, knowing nothing! Not parted from their Power In sameness, desiring nothing! Call it “simple” and “unhewn.” In the simple and unhewn, the nature of the people is found.48

Cognitive Attunement in the Zhuangzi | 63 The “simple” (su 素) and the “unhewn” (pu 樸) are famous images from Laozi 19 that are metaphors for states of mind in which the Way is present. I see them as symbolizing freedom from desires and from dualistic knowledge, an unself-conscious state of pure experience in which, for the Primitivist, our instinctive tendencies to respond spontaneously and harmoniously to whatever situation arises are allowed to operate without the encumbrance of social conventions. As we see in chapter 9: “In the time of the House of Hexu: the people when at home were unaware of what they were doing and when travelling did not know where they were going, basked in the sun chewing a morsel or strolled drumming on their bellies.”49 So, for the Primitivist, how does one create a society in which the basic unself-conscious tendencies of human nature are allowed to develop? Be led by an exemplary person (junzi) who practices “Doing Nothing” (wuwei 無為): Only by Doing Nothing will he find security in the essentials of nature and destiny. . . . If then a gentleman does prove able not to dislocate his Five Organs and stretch his eyesight and hearing, then sitting as still as a corpse he will look majestic as a dragon, from the silence of the abyss he will speak with a voice of thunder, he will have the promptings which are daemonic and the veerings which are from Heaven, will have an unforced air and do nothing . . .50 Of course the locus classicus of wuwei in the classical Daoist tradition is the Laozi. It is the one classical Chinese philosophical concept that has been studied by scholars informed by cognitive science, where it is frequently linked to ideas of “effortless attention,” “effortless action,” and “flow.”51 In the context of the Primitivist chapters I would take it to mean not interfering with the spontaneous functioning of cognition that arises from contact between our innate natures and the environments in which we find ourselves. In our contemplative phenomenology this is associated with “no-person experience,” unself-consciousness, and nonintentionality. It is most definitely linked to the cognitive scientific ideas of “effortless attention” and “tacit knowledge.” These different ways of presenting the transformed trait of cognitive attunement are not just found in the Inner Chapters and the Primitivist writings, they abound throughout the rest of the text. For example, in chapter 12 we find the exemplar of the gardener who is a “man in whom Power is whole.  .  .  .  Someone who by illumination enters into simplicity, by Doing Nothing reverts to the unhewn, who identifies himself with his nature and protects his daemon . . .”52 In chapter 14 we find Lao Dan teaching Confucius, a common trope in the “Outer Chapters”:53

64  |  Harold D. Roth The Utmost Persons of old borrowed right of way through the benevolent and lodged for a night in the dutiful to roam in the emptiness where one rambles without a destination, eat in the field of the casual and simple, stand in the orchards where one can keep all the fruit. To ramble without a destination is Doing Nothing, to be casual and simple is to be easily nurtured, to keep all the fruit is to let nothing out from oneself. Of old they called this the roaming in which one plucks out only the genuine.54 This passage brings together several of our metaphors for cognitive attunement and adds to them the concept of “the genuine” (zhen 真), another symbol for this state of unself-conscious no-person experience that is influential throughout the classical and later Daoist traditions.55 Commending this state is a common theme in the dialogues between Confucius and Lao Dan that Graham has collected. In chapter 21 we see the Daoist sage, who, after his bath was “so still he seemed other than human . . . as motionless as withered wood, as though he had left everything behind and parted from man to take a stand in the Unique . . .” He manifests the Power, which, in the perfected or “Utmost” Persons (zhiren 至人) cannot be trained through even the most “far-reaching words.” It is simply there: “It is like heaven being high of itself, earth being solid of itself, the sun and moon shining of themselves . . .”56 Along with some of the other passages we have presented to this point, this suggests the link between a state of complete stillness and tranquility and the development of the trait of transformed cognition of which we have seen many exemplars and symbols. This becomes clear when we examine two important and related passages from the authorial voice Graham labels as the “Syncretist”: Chapter 13: “Emptiness and stillness, calm and indifference, quiescence, Doing Nothing, are the even level of heaven and earth, the utmost reach of the Way and the Power; therefore, emperor, king, or sage finds rest in them.  .  .  .  Emptiness and stillness, calm and indifference, quiescence, Doing Nothing, are at the root of the myriad things. . . . To have these as your resources in high estate is the Power which is in emperor, king, Son of Heaven; to have these as your resources in low estate is the Way of the obscure sage, the untitled king. . . . In stillness a sage, in motion a king, you do nothing yet are exalted, you are simple and unpolished yet no one in the empire is able to rival your glory.”58

Cognitive Attunement in the Zhuangzi | 65 Emptiness, stillness, calm and indifference, and quiescence are states that result from the kinds of concentrative practices we see in contemplative phenomenology. Such practices—often linked to focusing the attention on the breath or a bodily location, or some other object—gradually reduce the occurrences of random thoughts, feelings, and perceptions that roll through the consciousness of most people from moment to moment resulting in a state of one-pointed concentration. Many of these are tied up in third-person conceptualizations of first-person and second-person experiences, creating mental structures that reinforce fixed notions of self-identity, desires, preferences, and beliefs. Concentrative practices take the focus of consciousness off these third-person conceptualizations and gradually reduce their occurrence. This Zhuangzi passage argues that the states of emptiness that result from these practices enable one to “Do Nothing,” that is, to allow cognition to function without effort, for sages and rulers to develop the “simple and unhewn” (supu 素樸) Power that enables them to act in the world through “tacit knowledge” in complete spontaneity and harmony. This result echoes those that scientists include under the category of “receptive” meditation. These ideas from chapter 13 are echoed, almost verbatim, in the following passage from chapter 15: Hence it is said that calm and indifference, quiescence, emptiness, and nothingness, Doing Nothing, they are the even level of heaven and earth, the substance of the Way and the Power; therefore sages find rest in them. At rest they are even and unstrained; being even and unstrained they are calm and indifferent.  .  .  .  Hence it is said that sages . . .  .  .  .  Only when stirred will they respond. Only when pressed will they move, Only when it is inevitable will they rise up Rejecting knowledge and precedent They take their course from the patterns of the heavens (tianli 天理).58 Here the states of calmness and emptiness leading to Doing Nothing allow sages to respond spontaneously and without deliberation or intentionality. They are able to do this because they do not impede in the effortless automatic responses that emerge because they do not interfere with the underlying “patterns of the heavens.” These are the various natural guidelines that organize the activities of the myriad things enabling them to stay true on the courses laid down by their distinctive natures in response to the

66  |  Harold D. Roth circumstances in which they find themselves.59 While there is a great deal of evidence for these illumined cognitive states and traits, what about the practices that contemplative phenomenology posits create them? While there are fewer passages that discuss these practices, they are present in some of the most renowned narratives in the Inner Chapters. Perhaps the most famous of them is the one in which Confucius instructs his favorite student Yan Hui on the best preparation for addressing a local tyrant: Unify your attention. Rather than listening with your ears, listen with your mind Rather than listening with your mind, listen with your ­breathing (qi 氣). Listening stops at the ears; the mind stops at what it can objectify As for your breathing, it becomes empty and waits to respond to things. The Way gathers in emptiness. Emptiness is attained through the fasting of the mind (xin zhai 心齋)60 This closely resembles the concentrative meditation with its attention focused first on sounds, then on mental objects, and finally on the breath. This leads to the emptying out of the usual contents of consciousness, as one gradually “fasts” or eliminates them. The other passage that refers to contemplative practice is this equally famous one in chapter 6 in which Yan Hui gives the following advice on “sitting in forgetfulness” (zuowang 坐忘) to his teacher, Confucius: I let organs and members drop away (duo zhi ti 墮支 體), dismiss eyesight and hearing (chu zong ming 出聰明), part from the body and expel knowledge, (lixing chüzhi 離 形屈知), merge with the Great Pervader (tongyu datong 同於大通).61 To “let organs and members drop away” means to gradually lose visceral awareness of the emotions and desires, which, for the early Daoists, have “physiological” bases in the various organs or orbs of qi circulation in the body-mind complex (wuzang 五臟).62 To “dismiss eyesight and hearing” means to cut off awareness of sense perception. To “part from the body and expel knowledge” means to lose bodily awareness and banish all thoughts from consciousness. To “merge with the Great Pervader” is to become united with the Dao.

Cognitive Attunement in the Zhuangzi | 67 Shorter references to such practices are sprinkled throughout the text. In chapter 2 “heart like withered wood; mind like dead ashes” refers to the deep stillness of the mind that develops from a one-pointed attentional focus that comes from concentration meditation. These two metaphors occur seven more times in the text, particularly in chapters 19–24.63 In chapter 19 the cicada catcher becomes completely still and totally focused: “it is only the wings of the cicada of which I am aware. I don’t let my gaze wander or waver, I would not take all the myriad things in exchange for the wings of a cicada . . .”64 Confucius characterizes this as “using the will in an undivided fashion and concentrating the spirit.” The mind as a mirror is another important cognitive trait that is praised in the text. In chapter 5 Wang Tai, the sage who “regards losing his own foot as he would shaking off mud  .  .  .  and who uses his mind to discover the unchanging mind beyond it” is someone with such a mind. “None of us finds his mirror in flowing water, we find it in still water. Only the still can still whatever is stilled.” “If your mirror is bright dust will not settle, if the dust settles it’s that your mirror isn’t bright . . .”65 In chapter 7 we read about the Utmost Persons who “use their minds like a mirror: they do not escort things as they go or welcome them as they come, they respond and do not store” (ying er bu zang 應而不藏).66 In the very beginning of the Syncretist, chapter 13, we find a long passage commending the still mind of the sage that “is the reflector of heaven and earth and the mirror of the myriad things.”67 Finally, chapter 33 contains the advice of the sage Guanyin: Within yourself, no fixed positions; Things as they take shape disclose themselves. Moving, be like water. Still be like a mirror. Respond like an echo . . .68 So, the consciousness that is emptied and stilled through concentrative meditation can mirror things with complete clarity, a symbol for the effortless, spontaneous, and unself-conscious functioning of cognitive attunement. This is an expression of “no-person experience.” The dialogue between the Yellow Emperor and the Daoist sage, Guang Chengzi, in chapter 11 contains fairly specific instructions on the concentrative dimension of contemplative practice: . . . look at nothing, listen to nothing cling to the spirit and be still, and the body will align itself,

68  |  Harold D. Roth always be still, always be pure, Do not strain your body Do not allow your vital essence to waver, And then it will be possible to live on and on. When the eye has nothing it sees, The ear has nothing it hears, And the mind has nothing it knows, Your spirit will abide in the body And then the body will live on and on. Be watchful over what lies within you Shut it off from being externalized Excessive knowledge is ruinous . . .69 In commending the removal of all thought and perception through the practice of developing stillness through focused internal attention, this passage touches on contemplative practices that the early practitioners of the Way regarded as longevity engendering. The goal, while seemingly specific to the Daoist tradition, has echoes in the stress reduction research of the past three decades.70 Finally, most of chapter 23 is devoted to an extended search for the most effective contemplative practices in which two characters, Geng Sangchu and Laozi, instruct Nanrongzhu on methods of cultivating emptiness and stillness that are not at all dissimilar to those commended by Guang Chengzi. Rid yourself of the perversities of the will (e.g., honor and fame) Eradicate the absurdities of the mind (e.g., false gestures and phone attitudes) Relinquish the attachments of Power (e.g., preferences and feelings) Break through the blockages of the Way . . .  And you will be aligned, still, lucid, empty and you will take no action yet leave nothing undone.71

Cognitive Attunement in the Qiwulun Two of the most evocative statements about the cognitive attunement that derives from the contemplative practice of emptying and stilling the activity of consciousness in the Zhuangzi come from two diverse contexts, yet they nicely symbolize the states and traits so clearly spelled out in the Qiwulun (Essay on sorting things without bias). In the teaching of Lao Dan to Confucius in chapter 22 we find the following statement made about the sage who

Cognitive Attunement in the Zhuangzi | 69 closely guards “that which no increase increases and no reduction reduces”: The Sage neither misses the occasion when it is present, nor clings to it when it is past. He responds to it by attuning himself, that’s the Power; he responds to it by matching with it, that’s the Way. From this course the emperors arose and kings began . . .74 This is the epitome of the flowing consciousness that is advocated throughout the text; it is literally cognitive attunement. This attunement is well captured in the central metaphor of the results of the attentional training practice of “the fasting of the mind.” After his pupil demonstrates to him that, indeed, he has attained a complete clearing out of his consciousness and concomitant opening it up to the Way, Confucius says: “It is easy to stop walking; but much more difficult to walk without touching the ground” (wuxingdi 無行地).73 “To walk without touching the ground” is Zhuangzi’s metaphor for living in the world without the support of a fixed and attached self-identity. This is living in cognitive attunement to the impulses of the Way as one engages in daily living. This process is detailed in the Qiwulun. This most detailed chapter in the attunement of cognition questions subjective bias in the categories of knowledge human beings create: For there to be “That’s it” (shih) and “That’s not” (fei) before they are formed in the mind would be to “go to Yue today and have arrived yesterday.” This would be crediting with existence what has no existence; and if you do that, even the Divine Yu could not understand you; how can you expect to be understood by me?75 Categories of knowledge are true relative to the limited perspectives that create them: “What is it is also other; what is other is also it: There someone says “This is ‘it’ (i.e. ‘true’), ‘that’s not (i.e. false)’ ” from one point of view; here we say, “ ‘That’s it’; this is not,’ ” from another point of view. Are there really it and other? Or really no it and other?”75 Here the ultimate veridicality of third-person experience is seen to be arbitrary and relative to the viewpoint of the observer. However, there is a perspective from which all these relatively true and false viewpoints can be “seen as equal”: Where neither It nor Other finds its opposite is called the axis of the Way. Once the axis is found at the center of the circle, there is no limit to responding with either, on the one hand no limit to what is It, and, on the other, no limit to what is Not. Therefore I say: “The best means is Illumination (ming 明).”76

70  |  Harold D. Roth By direct experience of embodiment of the Way, it is possible to break free of relativistic truths and experience a mode of cognition that can, without hesitation or limit, respond to whatever realities are present in any situation. Zhuangzi conceives of this mode of experience as an illumined cognition based in a consciousness centered in the Way. Such a consciousness is utterly without bias and able to fluidly and spontaneously respond to all the viewpoints and limited visions of reality that it encounters. This “illumined” mode of cognition is so freed from bias that it has no subjective preferences: “No thing is not Other; no thing is not It. If you treat yourself as Other, they do not appear.  .  .  .  This is why sages  .  .  .  open things to the lucid light of Nature; theirs too is a ‘That’s it which goes by circumstance’ (yinshi 因是).”77 To rephrase: judgments of true and false, it and other, are third-person perspectives based in individual biases. They are not in any way “objective.” Throughout this chapter Zhuangzi sets up the contrast that only Graham among the myriads of translators really gets, that between a “That’s it which deems” (weishi 為是) and a “That’s it which goes by circumstance.”78 The former is a “fixed cognition”; the latter is a “flowing cognition.” To “shi things” is essentially the activity of the third-person objectification of the internal and external worlds involved in rational thought. When we use “fixed cognition” to “deem,” we are locked into rigid self/other, true/ false distinctions—and attendant emotions—that cause all the arguments among the different philosophers of the day and cause a great deal of confusion for most everyone. When we use “flowing cognition” to “adapt,” we can function without the interference of attachment to a self-identity, thus yielding a freedom from adhering to fixed and rigid positions that enables us to alter in the moment as things alter. As the previous quote from chapter 22 says: “The Sage neither misses the occasion when it is present, nor clings to it when it is past. He responds to it by attuning himself.” In terms of our contemplative phenomenology, it is as if the activity of first-person subjective consciousness, because it has previously apprehended the nondual Way, is able to manifest its unself-conscious, non-self-referential aspect as “no-person experience” so it can pursue a completely unbiased third-person analysis that is free of attachment to any one particular position. This creates an illumined cognition from which all other perspectives are seen as equally valid or invalid: you have no more attachment to your own personal preferences and separate point of view than to anyone else’s. This is symbolized by the distinctive phrase “to treat yourself as ‘other,’ as in: “No thing is not Other; no thing is not It. If you treat yourself as Other, they do not appear.  .  .  .  This is why sages  .  .  .  open things to the lucid light of Heaven; theirs too is a ‘That’s it’ which goes by circumstance (i.e., a flowing cognition)  .  .  .”79

Cognitive Attunement in the Zhuangzi | 71 Keeping an unbiased Way-centered focus amid a constantly changing world attunes subjectivity to the deeper underlying power of the Way: Therefore when the “That’s it which deems” (fixed cognition) differentiates a stalk from a pillar, a leper from the beauty Xi Shi  .  .  .  the Way pervades and unifies them (道通為)—(and sees them as equal).  .  .  .  Only those who see right through things realize how to pervade and unify . . . the “That’s it which deems” (fixed cognition) they do not use, but lodge their attention in the ordinary (flow of experience)  .  .  .  To lodge in the ordinary flow of experience is to manifest   it; when you manifest it you can pervade things; when you pervade things you attain it and once you attain it you are just about there; Then the “That’s it which goes by circumstance” (flowing cognition) comes to an end, and when it does, That of which you do not know what is “so” of it you call the Way.80 Epistemologically, the Way is realized through a “pure experience.” We cannot know it as an object via third-person experience in which we separate from it (and know it as “so”); we can only know it through “no-person experience” in which the objectifying self collapses. For Zhuangzi, to return to this pure experience of the Way from time to time attunes subjectivity so that it can pervade and unify different perspectives, enabling a spontaneous cognition that is completely unbiased and purely spontaneous. We become directly connected to the situation in which we are embedded via the Way, which pervades all underlying Patterns (li 理) present in this situation. To do this is later in this text called “Embodying the Way” (ti Dao 體道).81 In terms of contemplative phenomenology, “embodying the Way” implies a non-self-referential, unself-conscious “no-person” experience in the present moment free from self-objectification; it is experience that is immediate, unbiased, spontaneous, and intuitive. So why does this mode of experience lead to cognitive attunement and not to chaos? In the Zhuangzi it is because embodying the Way connects you with a deep unitive power or source that underlies you and your environment. Throughout the text there are many metaphors for this source: “the Ancestor” (zong 宗);82 “the Unique” (du 獨);83 “the maker of things” (zaowuzhe 造物者);84 “the Great Pervader” (datong 大通);85

72  |  Harold D. Roth “the Flawless” (wujia 無假).86 Currently, the phenomenologists and cognitive scientists we have studied do not make this seemingly metaphysical claim. One thing about which most of our sources agree is that this cognitive attunement is quite often the source of intuition and creativity, as many passages in chapters 3 and 19 of the Zhuangzi detail. Cook Ding carving an ox, the cicada catcher reeling in insects, the skilled swimmer over the Spinebridge Falls, and the following passage about a bellstand carver are all examples: Woodcarver Qing chipped wood to make a bellstand. When it was finished viewers were amazed.  .  .  .  The Marquis of Lu asked him his secret. “Your servant is a mere artisan, what secret could he have? However there is one point. When I am going to make a bellstand I take care never to waste my energies, and I make sure to fast in order to still my mind. After three days of this I do not care to keep in mind congratulations and reward  .  .  .  after five days I do not care to keep in mind your blame or praise, my skill or clumsiness. After fasting seven days, I am so intent that I forget that I have a body and four limbs. During this time my lord’s court does not exist for me . . . the only thing I do is go into the forest and observe the nature of the wood as Nature makes it grow . . . only then do I have a complete vision of the bellstand and put my hand to it. Otherwise I give the whole thing up  .  .  .”87 This evocative passage reflects the techniques and results of the cognitive attunement we find throughout the text. Here there is an apophatic practice of “fasting” similar to the “fasting of the mind” in chapter 4. This is followed by a complete loss of self-consciousness and total focusing on the task of going into the forest to find the tree linked to his “vision of the bellstand.” Then he can bring it home and carve it out. As contemplative phenomenologists and cognitive scientists gradually come to include sources that are not European or Eurocentric, they will be able to develop increasingly sophisticated theories and hypotheses about the nature of human experience. And as scholars of traditional Asian thought come to see the value of phenomenological and scientific models of human experience, they will begin to recognize certain distinctive modes of experience reflected in traditional texts such as the Zhuangzi. Being able to identify the distinctive goal of cognitive attunement that is advocated in the Zhuangzi has helped us see that despite the various authorial viewpoints found in this work, there is a compelling continuity of interest and insight that motivated the people who authored and compiled it.

Cognitive Attunement in the Zhuangzi | 73 Angus Graham was convinced from his own experience that contemplative states are part of our potential as human beings. He was drawn not only to direct experience of contemplative states but also to the challenge of working out their philosophical significance. This is without a doubt what he found so fascinating about the Zhuangzi. He would have been pleased to know that the spontaneous flowing cognition that he recognized as a central theme in the text is now being studied and researched by philosophers and scientists so that many more people can access the significant benefits he saw in such experiences. It is, indeed, “human capacity for awareness raised to its highest pitch.”

Notes   1. The series of four radio programs for Ideas was broadcast on the CBC in February 1982 when I was in Sendai, Japan. I still use copies of them today in my teaching.   2. Harold D. Roth, “Contemplative Studies: Prospects for a New Field,” Teachers College Record 108.9 (2006): 0161-4681, 1467-9620 2006, 1787; “Against Cognitive Imperialism,” Religion East and West 8 (2008): 1–23; “A Pedagogy for the New Field of Contemplative Studies,” in Contemplative Approaches to Learning and Inquiry across Disciplines, ed. Olen Gunnlaugson and Heesoon Bai (Albany: State University of New York Press, 2014), 97–118.   3. A. C. Graham, “Mysticism and the Question of Private Access,” in Reason within Unreason: Essays on the Outskirts of Rationality, ed. David Lynn Hall (La Salle: Open Court Press, 1992), 265–282.   4. Ibid., 265.   5. Ibid., 277.   6. I have borrowed this idea of “attunement” from Jung H. Lee, The Ethical Foundations of Early Daoism (London: Palgrave-Macmillan, 2014). Lee speaks of “ethical attunement” as a result of practices advocated in the Zhuangzi.   7. Roth, “A Pedagogy for the New Field of Contemplative Studies,” 98–99.   8. The former have been studied extensively as examples of the “optimal experience” he calls “flow” by psychologist Mihalyi Csikszentmihalyi. See Flow: Towards a Psychology of Optimal Experience (New York: Harper and Row, 1990). The latter have been studied during the past century under the general category of “mysticism or “mystical experience.” For a superb overview article, see Jerome I. Gellman, “Mysticism and Religious Experience,” in The Oxford Handbook of the Philosophy of Religion, ed. William Wainwright (Oxford: Oxford University Press, 2005), 138–152.   9. See, for example, the collection of scientific research in Brian Bruya, ed., Effortless Attention: A New Perspective in the Cognitive Science of Attention and Action (Cambridge: MIT Press, 2010). See also two books by Edward Slingerland: Effortless Action: Wu-wei as Conceptual Metaphor and Spiritual Ideal in Early China (Oxford: Oxford

74  |  Harold D. Roth University Press, 2007) and Trying Not to Try: The Art and Science of Spontaneity (New York: Crown, 2014). 10. Francisco J. Varela and Jonathan Shear, “First-Person Methodologies: What, Why, How?” in The View from Within: First-Person Approaches to the Study of Consciousness, published as a special issue of the Journal of Consciousness Studies 6.2–3 (1999): 1–14 (Bowling Green: Imprint Academic). 11. Francisco Varela, Evan Thompson, and Eleanor Rosch, The Embodied Mind (Cambridge: MIT Press, 1991). 12. Evan Thompson, “Empathy and Consciousness,” in Between Ourselves: Second-Person Issues in the Study of Consciousness, ed. Evan Thompson. Published as a special issue of the Journal of Consciousness Studies 8, 5–7, 2001 (Bowling Green: Imprint Academic), 1. 13. Dan Zahavi, “Beyond Empathy: Phenomenological Approaches to Intersubjectivity,” in Between Ourselves, 166. 14. Thompson, “Empathy and Consciousness,” 1. 15. For an excellent recent summary of a plethora of this research, see Richard Davidson, The Emotional Life of the Brain (New York and London: Plume, 2012). 16. William James, Essays in Radical Empiricism and a Pluralistic Universe (New York: Peter Smith, 1967; first edition 1929), 74. 17. Yoko Arisaka, “The Ontological Co-Emergence of ‘Self and Other’ in Japanese Philosophy,” in Between Ourselves, 202. 18. Ibid., 202. 19. Ibid., 203. 20. Kitaro Nishida, preface in An Inquiry into the Good (Zen no kenkyu 善の研究), trans. Christopher Ives and Masao Abe (New Haven: Yale University Press, 1990), xxx. 21. The ultimate ground of consciousness and the world in the foundational Yogācāra work, The Laňkāvatāra Sutra, is “mind only” or the ultimate Storehouse Consciousness (Pāramālaya vijñāna). For an excellent summary of this complex work, see Mark A. Ehman, “The Laňkāvatāra Sutra,” in Buddhism: A Modern Perspective, ed. Charles S. Prebish (University Park: Pennsylvania State University Press, 1975), 112– 117. 22. The “brain sciences” includes psychology, cognitive science, and neuroscience, and various overlapping fields of research such as “cognitive neuroscience.” 23. Aaron Seitz and Takeo Watanabe, “The Phenomenon of Task-Irrelevant Perceptual Learning,” Vision Research 49.21 (2009): 2604–2610. 24. Aaron Seitz, “Task-Irrelevant Perceptual Learning,” in Encyclopedia of the Sciences of Learning, ed. Norman M. Seel (New York: Springer, 2012), 3270–3272. 25. Paul Hager, “Tacit Knowledge,” in Encyclopedia of the Sciences of Learning, 3259–3261. 26. Brian Bruya, “Introduction: Towards a Theory of Attention That Includes Effortless Attention and Action,” in Effortless Attention: A New Perspective in the Cognitive Science of Attention and Action, ed. Brian Bruya (Cambridge: MIT Press, 2010), 1. 27. Rael B. Cahn and John Polich, “Meditation States and Traits: EEG, ERP and Neuroimaging Studies,” Psychological Bulletin 132.2 (2006): 180–211. The information about neuroscientific studies of meditation presented here is taken from this excellent overview article.

Cognitive Attunement in the Zhuangzi | 75 28. Ibid., 181. 29. Ibid., 181. I have replaced their category of “mindfulness” with that of “receptive meditation,” as in the following source: James Austin, “The Thalamic Gateway: How Meditative Training of Attention Evolves toward Selfless Transformations of Consciousness,” in Effortless Attention, 375–377. “Mindfulness” is a technique and result that can be used for either concentrative or receptive meditation. 30. Austin, “Thalamic Gateway,” 377. 31. Britta K. Hölzel, Sara W. Lazar, Tim Gard, Zev Schuman-Olivier, David R. Vago, and Ulrich Ott, “How Does Mindfulness Meditation Work? Proposing Mechanisms of Action from a Conceptual and Neural Perspective,” Perspectives on Psychological Science 6.6 (2011): 537–559. 32. Francisco Varela, “Neurophenomenology: A Methodological Remedy for the Hard Problem,” Journal of Consciousness Studies 3.4 (1996): 330–350. 33. Angus C. Graham, “How Much of Chuang Tzu did Chuang Tzu Write?” in Studies in Chinese Philosophy and Philosophical Literature (Albany: State University of New York Press, 1990). Harold D. Roth, “An Appraisal of Angus Graham’s Textual Scholarship on the Chuang Tzu,” in A Companion to Angus C. Graham’s Chuang Tzu: The Inner Chapters, ed. Harold D. Roth (Honolulu: University of Hawai’i Press, 2003), The Society for Asian and Comparative Philosophy Monograph 20, 181–220. 34. Three recent works address many issues surrounding the text and composition of the Zhuangzi. Most relevant for our concerns is the issue of whether the first seven Inner Chapters of the Zhuangzi contain the original core around which the text formed and whether they can be attributed to the historical figure of Zhuang Zhou, as maintained by Graham and myself. Liu Xiaogan’s newest work on this topic examines a wide range of opinions and concludes that the Inner Chapters should be taken as the core of the text and represent the work of the historical figure, and that the Outer and Miscellaneous Chapters can be roughly attributed to the followers of Zhuang Zhou. See Liu Xiaogan, “Textual Issues in the Zhuangzi,” in Dao Companion to Daoist Philosophy, ed. Xiaogan Liu (Dordrecht: Springer, 2015), 129–157. David McCraw, in Stratifying Zhuangzi: Rhyme and Other Quantitative Evidence, Language and Linguistics Monograph Series, vol. 41 (Taipei: Institute of Linguistics, Academia Sinica, 2010) uses his own system of rhyming patterns in the Zhuangzi to argue that the presence of many “non-canonical cross-rhymes” in the Inner Chapters proves that they could not have been written by a single author. Reviewers have commented on the confusing presentation of his methodology and results, calling this “opaque to the point of incomprehensibility” (Richard Lynn, Journal of Chinese Studies 54 [January 2012]: 335–359). Linguist David Branner (Journal of the American Oriental Society 130.4 [2010]: 653–654), further questions both McCraw’s methodology and conclusions: “In the case of the Zhuāngzǐ, the fact that there are a great many noncanonical ‘cross-rhymes’ could indeed point to multiple authorship, with each hand characterized by a different cross-rhyming pattern. Could it also point to a single author writing in a style that allows for non-canonical rhyming, or even simply trying to sound non-canonical?” He concludes: “In short, this book should not have been published in its present form. It deals with an important question but is confusing and awkwardly composed. . . . its rambling and hesitant presentation makes me wonder about the soundness of its findings.”

76  |  Harold D. Roth In “Were There Inner Chapters in the Warring States? A New Examination of Evidence About the Zhuangzi,” T’oung Pao 96 (2011): 299–369, Esther Klein has put forth a well-researched and well-argued analysis of the composition of the Zhuangzi. In it, she theorizes that while there must have been a pre-Han version of the text that contained most of the materials that are in the work today, the compilation of the original fifty-two-chapter recension, later excised to thirty-three by Guoxiang, is likely a Western Han event, possibly—as I hypothesized in the original Festschrift for Angus Graham—at the Huainan court of Liu An (see Harold D. Roth, “Who Compiled the Chuang Tzu?” in Chinese Texts and Contexts: Essays in Honor of Angus C. Graham, ed. Henry Rosemont Jr. [La Salle: Open Court, 1991], 79–128). She contends further, contrary to accepted opinion, that the Inner Chapters are likely to have been the last section of the text compiled, put together because of—or to form—a coherent philosophy. She concludes that the time has come to stop thinking of there being one author of the Inner Chapters and most certainly to stop thinking of this author as the reputed historical personage of Zhuang Zhou. While there is something to commend in each of these sources, all fail to seriously consider the experiential dimension to which many passages in the text attest. (To be more precise, McCraw summarily dispenses with this in a comment on pp. 48–49.) If, as I am arguing, a contemplative practice and its results are attested throughout the text, the reliance on a historical tradition in which this practice was handed down from teacher to students is a necessity. Then as now, it is clear that one cannot derive an effective contemplative practice from reading about it in a book. As I have argued in the first Graham Festschrift, there is reason to theorize that this tradition continued into the Han and was present at the Huainan court. While it is not possible here to fully assess the impact this argument has on the evidence for the compilation of the text, it is possible to take something from each of their works: 1. With Liu, I think that the material now in the Inner Chapters was part of the original stratum of the book. 2. With McCraw, I think that it is possible that not all the Inner Chapters were written by one person; I do think that there is a consistency in literary styles and in philosophy that makes these chapters the core of the text. 3. With Klein I think that the work was likely compiled at the court of Liu An from materials that existed in the late Warring States. 35. Roth, “Who Compiled the Chuang Tzu?” 36. I am using “cognitive” to encompass the entire range of responses of consciousness to its environment, including intellectual, emotional, intuitive, aesthetic, and even “spiritual.” All these aspects of consciousness contribute to the total apprehension or cognition of the whole from moment to moment throughout the full range of third-, second-, first-, and no-person experience. This full range is included in classical Daoist ideas about the activity of consciousness or of the “heart” or “mind.” 37. A. C. Graham, Chuang-tzŭ: The Inner Chapters (London: George Allen & Unwin, 1981), 43–44. I generally follow Graham’s translation and indicate those passages in which I have modified the translation based on my own reading. All references to the text are from the edition in: D. C. Lau et al., eds., Zhuangzi suizi soyin, Ancient Chinese Text Concordance Series, Chinese University of Hong Kong, Philosophical works no. 43 (Hong Kong: Commercial Press, 2000), 1/1/15–21 (Chapter 1/page 1/ lines 15–21.).

Cognitive Attunement in the Zhuangzi | 77 38. Graham, Chuang-tzŭ, 45 (modified); Zhuangzi suizi soyin 1/2/2–3. 29. Graham, Chuang-tzŭ, 95; Zhuangzi suizi soyin 7/20/16. 40. Graham, Chuang-tzŭ, 150; Zhuangzi suizi soyin 11/28/31–32. 41. Graham, Chuang-tzŭ, 131–132; Zhuangzi suizi soyin 12/31/19. 42. Graham, Chuang-tzŭ, 69; Zhuangzi suizi soyin 4/10/5–6. 43. Graham, Chuang-tzŭ, 84–85 (modified); Zhuangzi suizi soyin 6/15/29–16/9. 44. Graham, Chuang-tzŭ, 82; Zhuangzi suizi soyin 5/15/20–25. I have emended Graham’s translation to be gender neutral whenever possible. 45. Graham, Chuang-tzŭ, 150; Zhuangzi suizi soyin 11/28/31–32. 46. Graham, Chuang-tzŭ, 81. 47. Graham, Chuang-tzŭ, 202–203; Zhuangzi suizi soyin 8/23/9–10. 48. Graham, Chuang-tzŭ, 205; Zhuangzi suizi soyin 9/24/1. 49. Graham, Chuang-tzŭ, 205; Zhuangzi suizi soyin 9/24/10. 50. Graham, Chuang-tzŭ, 212; Zhuangzi suizi soyin 11/26/25–27. 51. See, for example, Brian Bruya, “The Rehabilitation of Spontaneity: A New Approach in the Philosophy of Action,” Philosophy East and West 60.2 (2010): 207–210, and Edward Slingerland, “Towards an Empirically Responsible Ethics: Cognitive Science, Virtue Ethics, and Effortless Attention in Early Chinese Thought,” in Effortless Attention, 248–286. 52. Graham, Chuang-tzŭ, 186–187; Zhuangzi suizi soyin 12/32/22–24. 53. So common, in fact, that Graham has put together a section of these passages, seven in all. They come from chapters 12–14 and 21–22. Graham, Chuang-tzŭ, 126–134. 54. Graham, Chuang-tzŭ, 129–130; Zhuangzi suizi soyin 14/39/27–28. 55. See, for example, the second chapter of the Huainanzi entitled “Activating the Genuine.” The Huainanzi: A Guide to the Theory and Practice of Government in Early Han China, translated, annotated, and introduced by John S. Major, Sarah Queen, Andrew S. Meyer, and Harold D. Roth (New York: Columbia University Press, 2010). 56. Graham, Chuang-tzŭ, 130–131; Zhuangzi suizi soyin 21/57/20–25. 57. Graham, Chuang-tzŭ, 259–260; ZZSY 13/34/16–22. 58. Graham, Chuang-tzŭ, 265; Zhuangzi suizi soyin 15/41/24–28. 59. For an analysis of this important concept of li (patterns), see Harold D. Roth, “The Classical Daoist Concept of Li and Early Chinese Cosmology,” in Studies in Honor of Li Xueqin, ed. Wen Xing (Berkeley: The Society for the Study of Early China, 2012–2013), 157–184, Early China 35–36. 60. Graham, Chuang-tzŭ, 68–69 (modified); Zhuangzi suizi soyin 4/10/1–4. 61. Graham, Chuang-tzŭ, 92; Zhuangzi suizi soyin 6/19/17–22. 62. For details, see Manfred Porkert, The Theoretical Foundations of Chinese Medicine (Cambridge: MIT Press, 1974), 108–152. 63. Zhuangzi suizi soyin 2/3/15; 19/50/14; 20/55/12; 21/57/21; 22/61/1; 23/65/23; 24/70/19. 64. Graham, Chuang-tzŭ, 138; Zhuangzi suizi soyin 19/50/12–16. 65. Graham, Chuang-tzŭ, 77; Zhuangzi suizi soyin 5/13/18–21. 66. Graham, Chuang-tzŭ, 98; Zhuangzi suizi soyin 7/21/21 (modified). 67. Graham, Chuang-tzŭ, 359–360; Zhuangzi suizi soyin 13/34/1–24. 68. Graham, Chuang-tzŭ, 281; Zhuangzi suizi soyin 33/99/29–30. 69. Graham, Chuang-tzŭ, 178; Zhuangzi suizi soyin 11/27/23–28.

78  |  Harold D. Roth 70. Concentrative meditation is one of the techniques included in the toolbox of the practice of mindfulness-based stress reduction. See Jon Kabat-Zinn, Full Catastrophe Living (New York: Delta, 1990). 71. Zhuangzi suizi soyin 23/67/8–11. For an alternate translation, see Victor Mair, Wandering on the Way (New York: Bantam, 1994), 234. 72. Graham, Chuang-tzŭ, 132–133; Zhuangzi suizi soyin 22/61/16, 21–23. 73. Graham, Chuang-tzŭ, 69; Zhuangzi suizi soyin 4/10/5. 74. Graham, Chuang-tzŭ, 51 (modified); Zhuangzi suizi soyin 2/4/9–10. 75. Graham, Chuang-tzŭ, 53; Zhuangzi suizi soyin 2/4/18–19. 76. Graham, Chuang-tzŭ, 53; Zhuangzi suizi soyin 2/4/19–20. 77. Graham, Chuang-tzŭ, 52; Zhuangzi suizi soyin 2/4/16. 78. He first cracked this in the following article: A. C. Graham, “Chuang Tzu’s Essay on Seeing All Things as Equal,” History of Religions 9 (October 1969–February 1970): 137–159. 79. Graham, Chuang-tzŭ, 52; Zhuangzi suizi soyin 2/4/16. 80. Graham, Chuang-tzŭ, 53–54. Zhuangzi suizi soyin 2/4/26–5/3 81. Zhuangzi suizi soyin 22/62/14–15 82. Zhuangzi suizi soyin 5/13/13; 7/21/15; 13/34/22; 22/61/20; 33/97/15. 83. Zhuangzi suizi soyin 6/17/14; 7/21/19; 11/27, 28; 20/54/3; 33/99/27. 84. Zhuangzi suizi soyin 6/17/25, 27; 6/18/19; 7/20/14; 19/50/1, 3; 32/95/19; 33/100/10. 85. Zhuangzi suizi soyin 6/19/21; 17/46/27. 86. Graham, Chuang-tzŭ, 76–77. Zhuangzi suizi soyin 5/13/11–15. 87. Graham, Chuang-tzŭ, 135. Zhuangzi suizi soyin 19/52/4–8.

4

Vital Matters, A. C. Graham, and the Zhuangzi

Michael Nylan

The subject of this essay is a subject dear to the heart of Angus Graham, the eminent sinologist. Graham not only translated the first seven chapters in the long Zhuangzi text of thirty-three chapters. He wrote his own manifesto for living well entitled Reason and Spontaneity, largely inspired by his readings in the Zhuangzi. As readers will recall, Graham hoped to go down in history as a philosopher, not a sinologist. Reason and Spontaneity therefore offers a lively, whirlwind tour of Western philosophy from the ancient Greeks up to the twentieth-century analytic philosophers. And since Graham’s understanding of the Zhuangzi has colored nearly all subsequent readings of the Zhuangzi in modern Euro-America (and hence those of China as well), the plausibility of Graham’s analysis is worth assessing here, especially since Graham is now revered as something of an eccentric demi-god. And for those who don’t know the Zhuangzi well, that book (which I will treat today as one thinker’s work, for convenience sake) represents perhaps the sharpest thinking in antiquity about what it means to be and act fully alive, while holding a potentially dangerous court position with access to power. I will argue here that the Zhuangzi uses as its central concept clusters a series of terms that I translate as “vital spirit” (jingshen 精), “vital energy” (qi 氣), and “vitality” (sheng 生, also sometimes “life,” “living,” and “livelihood”). As a student of early manuscript culture, I am, of course, aware that the Zhuangzi is a composite text, like every other early text in our possession, and in a separate publication I discuss the implications of that obvious fact in sufficient detail. I am not inclined to repeat myself in print, and obviously enough, this essay’s subject, A. C. Graham, made important contributions

79

80 | Michael Nylan to this line of research, all of which are well known to students of early China.1 At the same time, Graham’s decision to cordon off the so-called Inner Chapters was disastrous, insofar as it allowed a majority of scholars claiming expertise in the Zhuangzi to confine their inquiries mainly or solely to those chapters.2 More importantly here, to worry overmuch in this essay about the compilation and structure of the text, at any given time in the past, would distract attention from this essay’s central implication: that early readers of the text understood better than some modern academics what seemed to hold the Zhuangzi text together as a unit. I argue here for a more historically situated reading of the Zhuangzi than that held by many others (including Graham), who have presumed that the Zhuangzi urges people to be “spontaneous.”3 As I think or write about the Zhuangzi, I have in mind comparisons with the medical treatises of the time, and also with the Xunzi, for Xunzi was an avowed Confucian who shared the Zhuangzi’s analysis of the human condition, especially regarding human desire, even if Xunzi intended to challenge some of the historical Zhuangzi’s prescriptions.4 Supplementing these comparisons are allied treatments, including the late synopsis of vitalist theories offered in the “Jing shen” 精神 chapter of the Huainanzi (ca. 138 BCE).5 • There is, as Aristotle famously remarked, a great deal more to living than survival. Pleasure in being alive, a capacity associated with feeling secure and at ease,6 as well as freedom from both humiliation and self-preoccupation7—all this depends upon developing a certain kind of clarity, most would argue. In the opening passage of the chapter entitled “Supreme Pleasure,” the Zhuangzi confronts us with the question of what it takes to feel more alive, on the assumption that we intuitively move toward whatever we take (or mistake) for pleasure-enhancing activities, objects, and people, while avoiding whatever we think will work against pleasure. As the passage raises so many of the issues treated in this essay, it seems best to quote more than one paragraph of it: In our world, is there such a thing as supreme pleasure or is there not? Is there something that may be used to enliven our persons or not? In today’s world, how are we to act, and how are we to make a secure foundation? What are we to shun and what are we to make our place? What are we to go for and what are we to eschew? What are we to regard as pleasure and what are we to disdain?

Vital Matters, A. C. Graham, and the Zhuangzi | 81 Now, as we all know, the world esteems riches and high rank, long life and others’ approval. What it regards as pleasurable is bodily ease, rich flavors, gorgeous robes, attractive sexual companions, and fine music. What it denigrates is poverty, low station, early death, and social censure.  .  .  .  Deprived of such pleasures, people feel greatly aggrieved and cowed. This is a stupid way act on behalf of the body, however.  .  .  .  When it comes to maximizing pleasures and imparting vitality to the physical person, one is nearly there if one acts without fixed goals or polarizing effects.  .  .  .  For such activities are nearly enough to sustain or preserve vitality.8 Here Zhuangzi argues for a certain kind of activity that is often quite misleadingly called “nonaction” or wuwei 無為. For most of his life Graham failed to correct that bad translation, though his own prose showed that Zhuangzi does not advocate doing nothing. Zhuangzi talks instead of doing what is needed to resolve pressing problems, but perhaps no more than that, so that the person may husband his vitality and élan, as shown in the following passage that I have retranslated, from the chapter “Mastering Vitality”: “Mastering Life” means never devoting time and energy to live in ways that do not promote life and vitality. “Mastering One’s Lot” means never devoting time and energy to trying to understand what is not under one’s control or has no real alternatives. When nourishing the body, one must always prioritize certain things [such as food, shelter, and clothing]. But there are cases, certainly, when things may be in great surplus but the body remains undernourished. . . . Then, too, there are cases, certainly, where a person still resides in his body but he lacks any vitality.  .  .  .  How pathetic that the people in this world think that merely nourishing the body suffices to preserve vitality, when it does not suffice, in fact. So how or why do they think this suffices? [They think this] because even though it does not suffice, it is imperative that it be done, being that it is unavoidable [if life is to be sustained at all].9 In sum, feeding and clothing and sheltering the body are necessary but insufficient conditions for vitality. Zhuangzi10 says this quite plainly. Here is one reason—there are others—why Zhuangzi’s project is not identical with that outlined by Guo Xiang 郭象 (d. 312), the earliest extant commentator on

82 | Michael Nylan the Zhuangzi, who plainly is more preoccupied with questions of political survival and transcendence than with vitality.11 Turning to what Graham saw as Zhuangzi’s project, let me state the obvious: Graham’s understanding of the Zhuangzi was a sophisticated adaptation of the early Chinese text for Graham’s own philosophy of life and attempts at world-making. Graham assumed that parts of the human condition were hardwired, and so incapable of change. He moreover accepted a set of beliefs about humanity that he ascribed to Zhuangzi also: (a) that humans hardly qualify as human beings, unless they act purposefully toward ends, bestowing values upon things and relations in the process; (b) that their purposes and values are rooted in their basic inclinations, innate or acquired; (c) that those inclinations arise spontaneously, a term Graham used in this precise sense: that “reason alone cannot give rise to them or make them go away”;12 and, hence, (d) that pure reason has a highly circumscribed role to play within human lives. Hence, the title of Graham’s chief foray into straight philosophy: Reason and Spontaneity. In Graham’s view, reason may help people decide upon a certain means to a specific end, but reason alone can in no way finally determine a person’s inclinations toward life. For, as Graham remarks, “In the choice of means [or instrument], the person is an agent; but in the choice of value, our whole lives are part of the weighing mechanism (the ‘arm of the balance itself’).”13 Graham’s proposed solution was deceptively simple: that the wisest people (the “sages”) shape their choices by following the injunction “Be Aware.” In Graham’s telling, the wise person is aware of every relevant factor and agent before acting, so as to obviate problems and misunderstandings.14 Vividness, pleasure, and spontaneity (a kind of audible hum of life)—all these imagined goods Graham ties to this principle of hyperawareness. Hence, for Graham the closer a person moves to “supraawareness,”15 the more he exemplifies the best and most vital in human life, because he maximizes his own potentials. My own critique of Graham stems mainly from the unidirectional quality of the message “Be Aware” and his descriptions of the state he called “heightened awareness.” By Graham’s reading, one’s awareness was to act upon the things and people in the surrounding world, making sense of it through this heightened awareness. Four key assumptions underlie Graham’s reading:

1. Graham thinks that we can be aware of many things that Zhuangzi argues are hidden (including, often, our own reasons for acting—here, Graham follows Nietzsche more than Zhuangzi, I think).

Vital Matters, A. C. Graham, and the Zhuangzi | 83

2. Graham therefore makes the mistake of positing a nearly omniscient being he calls a “sage,” thinking in this that he is consciously imitating Zhuangzi’s language. The Zhuangzi by turns praises and mocks the sages, so it is hard to believe that “to be a sage” is Zhuangzi’s central message.16



3. Graham moreover thinks the sage by definition surpasses the human condition to become heavenly, a person of true spontaneity. But Zhuangzi, I contend, never confuses the heavenly or godlike with the human. Humans can cultivate within themselves that “vital spark,” the animating force that seems to be divine and is certainly “inexplicable,” but they remain profoundly human throughout their journeys through life, and that is as it must be.



4. Then, too, Graham described “Be Aware” mainly in brainy terms, not alluding much to emotions or intuition. In the end, Graham took the individual autonomous rational being as the foundation for his philosophy. In so doing, Graham ignored all the talk in the Zhuangzi that insists upon people’s profound embeddedness in social, political, and cosmic relations.

I propose a startlingly different reading of the Zhuangzi, wherein I introduce three major changes. The precept “Be Fully Present” replaces Graham’s “Be Aware,” and “clarity” refers not to Graham’s hyperawareness of all the relevant facts confronting the person but to “knowing that you never know enough” when facing an unfolding situation that requires action. So “Be Aware” could only mean “Be Aware of how unaware you usually are, actually,” when spouting pronouncements.17 Finally, in my reading, the highly cultivated person (the “sage,” if you like) remains profoundly human throughout life, though cultivating the divine spark of vitality within his physical person. Quite reasonably, Zhuangzi insists that even the sage (i.e., the ideal person) cannot get rid of family ties and emotions, nor the necessity to make a living. Taken together, these three changes to Graham’s reading produce a hugely different Zhuangzi than Graham’s version: a fallible person who aims neither to transcend the world nor hide himself away from it, a person who continues to be deeply tied to the world, yet not hopelessly entangled by the lures of conventional success—a Zhuangzi that proves the rule that a measure of “detachment does not require disengagement.”18 Hence, the Zhuangzi promise that learning to operate within constraints can eventually school the person in “making [his or her] nature one,” and thereby prevent the dissipation of vital energies.19 That said, the issue of

84 | Michael Nylan heaven versus human in the sage is hardly only Graham’s problem. Every single commentator whom I have read, whether Chinese or European (most recently, Brook Ziporyn) thinks the ideal person should imitate the cosmos in its spontaneity (or ziran). The Zhuangzi text does not offer serious support for such readings, however, since the term ziran is never used to describe human beings, ideal or not. Ziran is an epithet applied solely and consistently to the cosmos (tian di). The degree to which the smartest person remains profoundly human and liable to humanity’s foibles and misjudgments, given that even the “fully present person” can only sustain that level of insight for a short period of time—the refusal to accept this insight lies, I warrant, at the very core of many misreadings of the Zhuangzi. As I read Zhuangzi, the text is very clear on this one point: that we human beings never escape the human condition, nor should we want to.20 Let us first look at some definitions of vitality and of clarity, in order to provide some groundwork for my claims; then we’ll turn to the single story that started me thinking about all of this, in terms not of post-Kantian philosophical theories but early resonance theories in China. Zhuangzi’s definition of “vitality” clearly identifies it as the one and only goal of wise men; as the Zhuangzi says: The unalloyed Way is simply and only to preserve and guard the vital spirit. If it is preserved and guarded, it will not be lost. If the whole person becomes one with the divine, so that his vital spirit communicates, he can then harmonize with heaven’s orders. As the proverb says, “Ordinary people prize profit; the man of honor prizes reputation; the worthy men in service uphold their commitments, [but] the sage values the vital spirit.”21 Often we read in modern secondary studies that Zhuangzi has no standard, that he is a relativist. That I dispute. The text may use relativist insights (say, a seemingly useless tree is not useless after all) as a preliminary strategy, but not as its final vision or its initial premise.22 Zhuangzi is no purist, but to deride Zhuangzi as a go-with-the-flow sloppy relativist because he dislikes rules (as did Confucius!) is the greatest indignity to fling at his therapeutic challenges, and all because the Zhuangzi is invariably entertaining and often quietly disorienting. For the Zhuangzi sooner or later explodes all possible rules in one anecdote after another, including the much-vaunted utility of being useless; in one of my favorite stories, Zhuangzi is asked to explain why a carpenter ignores the most useless tree when he goes to chop down trees, but the local householder refuses to slaughter the useful goose: “The tree that was not good lumber got to live out its allotted days  .  .  .  while the goose that lacked a certain capacity was the very one slaughtered.”23

Vital Matters, A. C. Graham, and the Zhuangzi | 85 The lesson should be clear: no hard-and-fast rules can insure even that we survive, let alone live vital lives.24 (This resonates perfectly with Timothy Chappell’s lucid book entitled Knowing What to Do.)25 Each unfolding situation represents a unique concatenation of too many factors and considerations for rules to finally be of much help.26 Context (i.e., the conjunction of things, aka “fate” and “timing” in the early Chinese sense) finally determines what goes, what works, no matter how aware or erudite we aim to be. Go figure! Zhuangzi is no less clear about the benefits clarity can bring: clarity cannot induce “free and easy” wandering through life, for the simple reason that the Zhuangzi text associates “clarity” rather with “being in a daze.” Please note: I do not deny that the famous binomial term xiaoyao, thought to describe the “free and easy wandering” the sage enjoys, gains that delicious meaning, during the Six Dynasties period, roughly mid-third through the sixth century. But I would draw your attention to this: that in the few Han readings we’ve got, xiaoyao means instead “befuddled, in a daze.” To take one famous passage, the Shiji, compiled approximately 90 BCE, in its chapter on the sage Confucius uses this very phrase to describe Kongzi as a very old and decrepit man shuffling to the door in considerable confusion. The sage is shown there paradoxically in a fog, as he cannot think who has come to knock at his door, though it is his disciple Zigong.27 To confirm that there is no living thing that can operate in perfect clarity, we need only turn to the opening passage of the Zhuangzi, which describes the marvelous Peng bird (transformed from something like a gigantic whale). The Peng bird has such an enormous wing span that whole quadrants of the earth grow dark for the duration of its flight as it passes over. But when an ordinary spectator looks up at the sky to watch the Peng bird as it flies away, all appears to be limpid blue. In other words, the human spectators, as we are explicitly told, see none of the bits of “wavering heat, bits of dust, and living things blowing each other about.”28 Odder still, the Peng bird sees no more than the most ordinary of spectators: it sees none of the things the flapping of its wings has stirred; “all it sees is the same blue” as that of the spectators, or so we are told.29 This is the part of the story that few have noticed: that no matter how great a creature is, it has real-time limitations. Living creatures, in particular, are limited in what their sensory equipment and hearts and minds allow them to perceive. Hence, my contention that none of us, no matter how we try, can be very successful, pace Graham, at the job of “being aware.” And that, I would argue, for Zhuangzi, is perfectly fine. The Zhuangzi-butterfly tale moves in the same direction: one can only be in one state at one time, it tells us, either Zhuangzi dreaming he is a butterfly or a butterfly dreaming it is Zhuangzi, as Hans-Georg Moeller has kindly taught us.30 It seems curious, but one of the closest people to echo Zhuangzi’s views may be Virginia Woolf, who

86 | Michael Nylan wrote during World War I, long before her suicide, “The future is dark, and that is the best thing for it to be, I think.”31 Here is what I mean by the foregoing: Zhuangzi supplies us with the paradox that seems more and more self-evident as I grow older: far from celebrating freedom from care or the enviable transcendence or omniscience of the sage, Zhuangzi reminds us again and again that the only clarity that is open to us is to realize, with brutal clarity, that we humans travel through life nearly always in a virtual daze. We humans can hope for, besides boundless curiosity about unfolding events, an exquisite sensitivity to our own muddled state and that of others who claim to know things and facts. As Zhuangzi remarks in the “Supreme Pleasure” chapter, “Such confusion and muddle, with no single origin. Such confusion and muddle, with no particular form. The myriad things nonetheless attend to carrying out their own businesses. What person among us [the privileged] can possibly manage that much?”32 Reliant as we are on others, both the living and the dead, for giving us slices of our perspectives, which we then glimpse through a jumbled labyrinth of memories and experiences, how could we ever hope for more than contingent and provisional flashes of insights into the sublime totality of phenomenal existence? Nevertheless, we might ask ourselves, How much do we actually need to know about complex realities for us to figure out how we had better act? Not much, says Zhuangzi in “Levelling All Things” or Qiwulun: If a man follows the complete mind and makes it his teacher, then who could truly be said to lack a teacher? Why must a person fully comprehend the process of change and form her mind on that basis before she can have a teacher?33 Like the Peng bird story, this passage hints that we have all that it takes to flourish and even soar. Our limitations are fine, since we also have the capacities we need to live a vital life.34 It behooves us to recall that these lines I just mentioned are to be found in the “Supreme Pleasure” chapter. It behooves us also to remember that most of the best modern philosophers have arrived at similar places to Zhuangzi’s. In other words, they do not believe that we humans fully understand how and why we act.35 But unlike most thinkers in the Western literary and philosophical traditions, Zhuangzi evinces remarkably little interest in the reality/appearance dichotomy (itself an outer/inner conflict). As with the Peng bird story mentioned before, there is a vague nod toward sense skepticism, the idea that a person’s sensory equipment does not allow him or her to see the world clearly. But the far greater interest for Zhuangzi lies elsewhere, judging

Vital Matters, A. C. Graham, and the Zhuangzi | 87 from how often the text returns to one recurring theme. It has to do with making distinctions and assigning relative value, and then choosing words to reify the values we assign to those distinctions; also with recognizing that we ourselves, like the world around us, are transforming constantly and hence never fully complete. A small selection of passages teach us that knowledge of any kind is also never complete. Perhaps the most succinct is, “Your life has a limit but knowledge has none. To use what is limited to pursue what has no limit will undo you.”36 That says it all. Clear-sightedness about our own propensity to misidentify value is the precondition for taking greater pleasure in things, not to mention more acute observation. Doubting ourselves, we examine other things beside ourselves with greater focus and curiosity. But no matter how great the focus or curiosity, we cannot be absolutely clear about anything. All significant things remain undecided, including and most especially who we are ourselves. As Zhuangzi observes, “The sage would confine [box in] the dazzlingly brilliant light that comes from what is actually in slippery doubt.”37 To lodge in the ordinary, to make insignificant moves, this the sage finds suffices for vitality; thus, to move in that direction edges the person closer to the only truly divine (shen) aspects that humans share with heaven, earth, and the myriad things. For robust reality wreaks havoc with any system of logic yet devised. I have already cited the memorable line, “Man’s life has always been a muddle like this. How could I be the only muddled one, and the other men not muddled?”38 Accordingly, words, even very precise words, do not serve us particularly well as we go through life, and not only because the very preoccupation with careful modes of categorizing often distracts us from a full appreciation of vital things and people as they are. Meanwhile, the central fact of human existence, our mortality, can and often does prevent people from wanting to experience things over which they suspect they have no control. Let us read Zhuangzi’s analysis of the frantic and fragmentary lives human beings are wont to lead: Once a man receives his complete bodily form, he keeps it, waiting for the end. Sometimes clashing with things and sometimes bending before them, he completes his course like a galloping steed, and nothing can stop him. Is he not pathetic? To the end of his days he sweats and toils,  .  .  .  utterly exhausting himself and never knowing where to find a true refuge. Can you help pitying him? “I’m not dead yet!” he says, but what good is that? His body decays and his heart and mind follow. Can you deny that this is a great sorrow? Man’s life has always been a muddle like this.39

88 | Michael Nylan That said, no ontological problem prevents us from living fully. The goal for Zhuangzi is not to produce new knowledge, then, but to prompt new people to be open to new experiences, and therefore be prepared to live life fully. “Brightness is born” at the place where the vital pulse resides, analogized to an indwelling god, and at the place where enlivening contacts and vital exchanges occur between things and people in phenomenal existence. Most of all, the Zhuangzi means to persuade us that we’ve wonders enough as we go through life, that only greedy-guts ask to be blessed with more and more gifts in the form of more and more life, like a rude child or immature adult who cannot be placated or satisfied with a single treat. The journey through terra incognita will present us with pleasures galore. This belief Zhuangzi would have us embrace as an article of faith. We cannot test the proposition, if we fail to attempt the journey. But there are reasons to trust Zhuangzi’s promise that it can be “spring with everything.”40 Reverie makes it a possibility, for example. One who spends her life in rapt contemplation of the incredible process of “things changing”—such a person can be said to have “never tasted death,” though she will most certainly die in due course, like all the rest.41 In Zhuangzi’s way of putting it, that is because she doesn’t let fear of death damage her or alter her activities, because she knows that “life and death are the same story.”42 Three and a half more stories to look at and then we shall have the essentials of Zhuangzi’s message, which I claim is “Be Fully Present.” I mention a half story, because probably most of you know the Butcher Ding story. Let me recap it briefly: Butcher Ding has a marvelous skill for butchering oxen and for living life well. Everyone knows that because of his skill Butcher Ding’s knife never grows dull, no matter how often he uses it. As Butcher Ding puts it, “There are spaces between those joints, and the knife blade really has no thickness to speak of. If you insert something with no thickness into a space, then there turns out to be more than enough room for a knife to play around in.” But Butcher Ding never transcends the human condition, and this is the part we tend to forget. Butcher Ding’s explanation of his skill continues, Whenever I come to a complicated place, where I see I am bound to have trouble, I am on my guard and most cautious. I keep my eyes fixed on what I’m doing, I work very slowly, and I nudge the knife the smallest little bit, until flop! The whole thing falls apart like a clod of dirt crumbling to earth.43 An even finer version of an artisan story describes Woodworker Qing, and this is the story that prompted my own thorough rethinking of the Zhuangzi. In it Woodworker Qing reveals how he acquires what appears to others to

Vital Matters, A. C. Graham, and the Zhuangzi | 89 constitute godlike powers.44 Essentially, Woodworker Qing undergoes many days of preparation, which ready him to perceive the barely perceptible: the inlaid patterns in the trees in the nearby forest. Once he perceives a pattern that he is drawn to, he is ready to receive what he has glimpsed into the very core of his being, allowing an exchange between what is divine in Woodworker Qing (his vitality, plus his endowed talents and learned capacities) with what is divine in the tree (its fundamental patterns or “grain”), thereby changing them simultaneously. By antique resonance theories, I would remind you, a real exchange means an indelible change in both parties. I quote: Only then do I enter the mountain forests, to contemplate the natural patterns the trees were given. With the bodily forms fully present, then and only then do I see a bell stand with perfect clarity, and afterwards I apply my hand to it. Otherwise, it’s no go! In this way, I join something of heaven to something of heaven.45 Neither the mountain nor the carpenter leaves unchanged by their encounter; both have gained and lost a measure of qi.46 At the same time, Woodworker Qing could no more live well as a tree in a forest than the chosen tree could live well by wielding the woodworker’s tools on behalf of an admiring court. Admittedly, to persuade readers to retain the very vitality they are born with, reckless words must be broached. Zhuangzi does this, evincing no regrets for the usual formalities. Hence, the imperative for Zhuangzi and his would-be followers to glorify everything that people typically abhor, including old age, deformity, ugliness, excrement, and death. Simply put, our usual sense of values must be turned upside down if we are to come to see that it is always and ever “enough that morning and evening we have things and they are the means by which we live.”47 Note here my strong implicit objection to those who persist in making Zhuangzi advocate a “natural” course of doing things, if by that they mean “spontaneous” (not needing thought, going with the flow).48 (Graham was himself one to insist upon this, and the sheer silliness of that proposition the Director Jia Zhangke mines, to the delight of his audience, when one of his protagonists quotes the Zhuangzi in a futile attempt to get his girl in bed.)49 For Zhuangzi, only the most unnatural of activities—the leap to formal logic and frequently the reductio ad absurdum—may possibly appeal enough to snap us out of our customary attitudes, so that we discover ingenious ways, strategies if you will (always geared to our unique orders and experiences, in continual flux), to deprogram our most basic learned proclivities. The goal is meanwhile to encourage a degree of mingled curiosity and steady confrontation with our current predicaments via the laughter of absurdity.50 Yet “the sage  .  .  .  merges his person with things.”51

90 | Michael Nylan Zhuangzi would never be so foolish as to hazard a neat division between the heavenly and the human. After all, the human derives from the heavenly, insofar as life comes from God knows where. The opening lines of chapter 6 show the sort of typical complicating move of which the Zhuangzi is so fond: He who knows what it is that Heaven does, and what it is that man does has reached the peak. Knowing what it is that Heaven does, he lives with Heaven. Knowing what it is that man does, he uses the knowledge of what he knows to help out the knowledge of what he doesn’t know, and so lives out the years that Heaven gave him. However, there is a difficulty: knowledge must wait for something, if it is to be applied and what it waits for is never certain. How, then, can I know that what I call “Heaven” is not really “man” and vice-versa?  .  .  .  When man and Heaven do not defeat one another, then we have a genuine person.52 Evidently, the main thing for humans is not to neglect either the heavenly or the human, both of which they have already within themselves. “Do not tire of the heavenly, nor neglect the part that is human. Then people will come very, very close to being able to deploy their true forms.”53 Besides, as the Zhuangzi shows in another justly famous passage, we humans do not have to get rid of either the heavenly or the human, in order to learn to live well. As the story goes, Huizi, the logician and one of Zhuangzi’s best friends, believes he has finally gotten Zhuangzi in a logical contradiction. Far from it, as the following summary suggests: Zhuangzi: “Clearly you are applying conventional notions to my talk about feelings. When I talk about ‘having no feelings,’ I refer to a person who does not allow likes or dislikes to worm their way in and so harm his person. By habit, he just lets things be the way they are; he doesn’t try to help life along or increase his lifespan.” Huizi: “Well, if he doesn’t try to facilitate his living or livelihood, then how is he going to possess a live and vital body?” Zhuangzi: “The Way gave him a look and a bearing; heaven gave him a form. He doesn’t allow likes or dislikes to get at him and into him, so they do no harm to him. You, now, you treat your spirit like a total stranger. You deplete your vital energy; you

Vital Matters, A. C. Graham, and the Zhuangzi | 91 lean on a tree and moan or slump at your writing table and doze. Heaven picked out a body for you and you use it to gibber about logical puzzles.” In the story, Huizi has just failed to understand where Zhuangzi was going with things, since Huizi’s customary playground is dichotomous either-or language. He doesn’t use language as a conveyer of the wondrous; Huizi prefers to use language as an iron clamp. For that reason, Huizi simply mistook Zhuangzi’s playful slogan “no feelings” to mean a person could actually come to lack any and all emotions, after having aspired to that state.54 Zhuangzi replies that we human beings have all we need to sustain our vital energies: we have a physical body, which is a boon. That is what we are to pay due heed to and nurture, rather than “winning.” Of that Zhuangzi is sure. Perhaps the most sublime pleasure to be had from reading antique authors is that the Ancients saw different things (they did not just see them differently); thus, they can awaken us to the peculiarities of today’s “regime of truth” that we take for granted, beckoning us to explore thrilling new ideas with utmost care.55 Zhuangzi is no exception. A great deal of what the Zhuangzi enjoins is “getting rid of things,” so that we are less burdened by our expectations for the future and drives for success, as well as those of others, less weighted down with information overload.56 The hope is that by traveling somewhat lighter through the world we can proceed to operate with respect to the world with the somewhat lighter touch that it and we require. But some things humans cannot rid themselves of: the desire or need to love and be loved, the desire or need to make sense of the unintelligible, the desire or need to communicate these mostly inchoate ideas to others. Significantly, in the Zhuangzi, at no time do the sages and wise men (Zhuangzi included) not have wives and children, homes and jobs, friends and disciples, and more friends and more friends. The conclusion: unnecessary entanglements (lei in Chinese) can be loosened, but to try hard to rid ourselves of certain attachments and ties is not human, as human beings are social beings and sociality is learned rather than “natural”; thus, were we to rid ourselves of the attachments and ties that are natural to human beings, we would be acting at our peril. Zhuangzi is no Buddhist, in other words. Desires and emotions we have. We cannot gain anything by getting rid of them, any more than we can hope to retain our vitality and animation by holding our breath. (On this point, the Zhuangzi text shares more with the Xunzi text than with the Laozi.) The eyes of human beings continue to see, their minds to think, and their mouths to jabber. (In this we resemble nothing so much as the baby birds that go cheep-cheep-cheep all the live long day; hence, the metaphor

92 | Michael Nylan that threads through the Zhuangzi.) The only thing that is different is this: once we have attained that state of “unknowing” that affords us a bit of welcome repose (i.e., that state where we are truly comfortable having little confidence in our judgments), we are—paradoxically—apt to see and engage a bit more, if more selectively.57 Why? Because we retain that sense of curiosity, of trying to make sense of things, particularly those multiple contradictions and possibilities that make “everything come alive” in our mind’s eye. “Spirit, being empty, waits upon all things.”58 “Nothing is as good as following orders.”59 The alternate is only “things condemning things”—and what sort of a life is that?60 To the degree that we live less in our own narrow constructs where X always follows from Y, we live far more fully in the world, noticing the marvels of phenomenal existence for what they are: an endless stock of diverse and shifting patterns impressing themselves on our senses, our emotions, and our intellects as we wander through life. Touring is Zhuangzi’s trope, and it’s a capacious one, insofar as simple aspects of life tend to strike our fancy more when we have left our usual haunts.61 Then we may learn the vital and sustaining pleasures that come with the gift of life. To disdain those pleasures, in whatever form they present themselves to us, is to lose for all time the habit and promise of vitality. Put another way, the goal for Zhuangzi is not to produce new knowledge but to spark the production of a vital person open to new experiences, one both willing and prepared to live life as fully as possible. His Dao adept goes beyond “what he knew and how he knew,”62 developing a knack for accommodating new things. He is more inclined to drink in scenes whole, since he has erected fewer barriers than usual.

Conclusion The promise that Zhuangzi holds out is this: so long as humans endeavor to “try on” Zhuangzi’s vision, they may come to see that, as Zhuangzi says, “it is not possible to live—let alone, live well—while continuing to behave as we have been behaving.”63 (This insight should ring especially true in the era of global warming and multiple ethnic wars, cold and hot.) There are but two feasible responses to our own severe limitations and those of others: on good days one may “burst out laughing”64 and continue roaming, or one may “go home and forget everything.” With either response, the wryly self-deprecating or the quietly dismissive, we end up in a state where we are less likely to use the paltry powers at our command to pummel other people or trumpet our own superiority. We end up with something not so far from Socrates’s aporia: we will admit how much we cannot know, and see how much we do not see. This dollop of self-knowledge and a curios-

Vital Matters, A. C. Graham, and the Zhuangzi | 93 ity about others is a good start, at any rate, one that may allow not only for mere survival, but for true vitality. I rest my case: go out and read the Zhuangzi, preferably reading backwards from chapter 33 toward the Inner Chapters and not in the most recent Penguin translation.65 Why do I advise readers to approach the text by reading it backwards? Partly because, as Esther Klein (and Wang Shumin centuries before) noted,66 we have focused far too much on the Inner Chapters (ergo Graham’s decision to translate none but those), and partly because, as my own review of the secondary literature in Western languages proves, we have devoted nearly all of our attention to a mere five stories and themes (four from the Inner Chapters).67 Reading the Zhuangzi backwards is a bracing experience. Once we get outside the Inner Chapters, we realize how much more there is to the Zhuangzi, and how often it complicates its messages. That is an excellent beginning.

Notes   1. For recent opposing views on this issue, see Shuen-fu Lin, “Confucius in the ‘Inner Chapters’ of the Chuang Tzu,” Tamkang Review 18.1–4 (1994): 379–401, versus David McCraw, Stratifying Zhuangzi: Rhyme and Other Quantitative Evidence (Taipei: Institute of Linguistics, Academia Sinica, 2010).   2. My able undergraduate research assistant Scott Davis compiled a database for me of citations in modern scholarship on the Zhuangzi. I am happy to share the results of his work in an Excel spreadsheet. The sheet shows how narrow the base of Zhuangzi secondary scholarship has been for the last few decades in English, with many scholars seldom, if ever, going in their citations outside five main passages (the Peng bird, the butterfly dream, Butcher Ding, “catch the fish and discard the fish trap,” and discussions of mirrors). I would argue that little secondary scholarship in Chinese ventures far beyond these workhorses either, but given the volume of recent Chinese scholarship on this issue, I cannot survey the field.   3. I reject as well the “school” affiliations given to Zhuangzi/Zhuangzi, given that Xunzi thinks Zhuangzi more like Mencius than Laozi. The Buddhist-inflected versions of the Zhuangzi are also anachronistic.   4. As William G. Boltz, “The Composite Nature of Early Chinese Texts,” in Text and Ritual in Early China, ed. Martin Kern (Seattle/London: University of Washington Press, 2005), 50–78, has observed, nearly all early Chinese texts are composite texts, whose authors are often ascribed by tradition, rather than actual. At the same time, it is tedious in every essay to try only to disentangle the historical Zhuangzi from the Zhuangzi compilation we hold in our hands today, and since we have no other basis on which to discuss the text, I use the name Zhuangzi as a placeholder for the historical figure, author of part of the Zhuangzi, whom tradition says lived at the time of two kings (one in Liang and one in Qi), reigning between 370–301 BCE. I do not mean to suggest that the Zhuangzi existed in its present form during Xunzi’s lifetime,

94 | Michael Nylan merely that some form of the text certainly existed, enough for Xunzi to believe he had discerned the gist. Important work has been done by Liu Xiaogan on this topic.   5. On this, see Nylan, “Logical Connectives in the Huainanzi,” Text and Context: New Perspectives on the “Huainanzi,” ed. Michael Puett and Sarah Queen (Cambridge: Harvard East Asian monograph, 2013), 225–265.   6. See, for instance, chapter 5 of Shuoyuan: 人情安則樂生. D. C. Lau, ed., Shuoyuan zhuzi suoyin (Hong Kong: Commercial Press, 1992), 5.14/35/1. A similar phrase is found in Qian Hanji, 17.172, and Hanshu, 51.2370. References to Yuan Hong, Qian Hanji (Taibei: Taiwan shangwu yinshuguan, 1971), and Ban Gu, Hanshu (Beijing: Zhonghua shuju, 1962).   7. See chapter 11 of Lüshi chunqiu: 誠辱則無為樂生. 若此人也，有勢則必不自私矣. References to D. C. Lau, ed., Lüshi chunqiu zhuzi suoyin (Hong Kong: Commercial Press, 1992) 11.3/54/22.   8. Wang Shumin, ed., Zhuangzi jiaoquan (Taibei: Shangshu yinshuguan, 1988), 639–642, hereafter, Zhuangzi. The author has consulted, here and later, Burton Watson’s translation, The Complete Works of Zhuangzi (New York: Columbia University Press, 1968), 139–140.  9. Zhuangzi, 667; Watson, Complete Works, 145. 10. See note 3. 11. On Guo Xiang, see Brook Ziporyn, The Penumbra Unbound: The Neo-Taoist Philosophy of Guo Xiang (Albany: State University of New York Press, 2003). Note, however, that the very last line of Guo Xiang’s commentary seems to come to much the same conclusions about Zhuangzi’s project that I do, for Guo argues that ordinary people are bored with morality, so a transformation can only occur through the skillful use of mind games and logic. 12. How central the idea of “spontaneity” is to Graham’s version of Zhuangzi becomes clear when we see Graham’s treatment of it in the introduction to his translation, Chuang-tzŭ: The Inner Chapters (London/Boston/Sidney: George Allen & Unwin, 1981), 6–8. At the same time, Graham was far more careful in his use of the word “spontaneous” than many who have followed; many would loosely equate it with a type of “transcendence” that correlates with Wei-Jin discourse, but not with that of the pre-Qin or early Western Han periods. 13. A. C. Graham, Reason and Spontaneity (Totowa: Barnes & Noble Books, 1985), 8–9. 14. Ibid., 6. 15. By this Graham means that we as human beings can be cognizant not only of our own inclinations but also those of others. But how does this supra-awareness develop, and can too much of it become a problem? Graham answers neither question. Graham argues that one can equate “the most rational” with “the best,” even if full awareness and being drawn to those ends is not fully articulable, except by poetry and myth. Graham, Reason and Spontaneity, 221. 16. I argue in a forthcoming essay “Zhuangzi: Closet Confucian?” that Zhuangzi consistently praises the virtues of being aware of one’s own befuddlement. 17. Many passages could be adduced in evidence here, among them Zhuangzi, chap. 2, where the protagonist asks, “Why must a person fully comprehend the pro-

Vital Matters, A. C. Graham, and the Zhuangzi | 95 cess of change and form her mind on that basis before she can have a teacher?” (奚 必知代而心自取者有之); chap. 2: “Human life has always been a muddle like this. How could I be the only muddled one, and the other people not muddled?” (人之生也固若 是芒乎。其我獨芒。而人亦有不芒者乎); chap. 23: “Out of the murk, things come to life. With all your cleverness, you declare, ‘We have to analyze this!’ You try putting your analysis into words, though it is not something to be put into words. You cannot, however, attain understanding”; and chap. 19: “Our life has a limit but knowledge has none. To use what is limit to pursue what has no limit will surely undo you!” (吾生也有涯，而知也无涯。以有涯隨无涯。殆已。已而為知者。殆而已矣). The last statement is repeated many times. 18. This saying is ascribed to David S. Nivison, “Hsun Tzu and Chuang Tzu,” in Chinese Texts and Philosophical Contexts: Essays in Honor of A.C. Graham (La Salle: Open Court, 1999), 132. 19. This charisma somehow develops when “returning one’s vital spirit to the time before birth.” See the “Da sheng” (Mastering life) chapter for one full exposition of this theory. That “oneness” (and the allied concepts of “intactness,” “wholeness,” and “integrity”) are central preoccupations of the late Zhanguo and Western Han thinkers I have already demonstrated in a variety of publications, including my translation of the Taixuan jing and my essay on Xunzi, entitled “The Politics of Pleasure,” Asia Major n.s. 14.1 (2001): 73–124. 20. This insight has definite political repercussions, and Zhuangzi is astutely aware of politics and political dangers, in my view. The portraits of bad rulers from Zhanguo and Western Han consistently show them deluding themselves into thinking they are superior beings. 21. Zhuangzi, 560–561; Watson, Complete Works, 121. 22. McCraw, Stratifying Zhuangzi, speaks of “an irreducible pluralism” pervading the very organization and structure of the text. 23. Zhuangzi, 719–720; Watson, Complete Works,156. 24. See, for example, Zhuangzi, 687n7; Watson, Complete Works, 201, on two figures: “In Lu, there lived a man called Shan Bao. Shan lived in a cave and drank only water. He refused to contend for profit with others. Even at seventy, he had the complexion of a baby, but, unfortunately, he met with a hungry tiger, which promptly killed and ate him. Then there was Zhang Yi. Zhang had a good name, since he paid his respects to each and every household, whether wealthy or commoner. At the age of forty, he died of internal heat [possibly related to his exertions]. Shan Bao nurtured his inner spirit but a tiger consumed his external form. Zhang Yi nurtured the externals, yet illness invaded his internal organs.” 25. Timothy Chappell, Knowing What to Do: Imagination, Virtue and Platonism in Ethics (Oxford: Oxford University Press, 2014). 26. For example, ibid., 150. 27. Shiji, 47.1944. References to Sima Qian, Shiji (Beijing: Zhonghua shuju, 1982). 28. Zhuangzi, 6; Watson, Complete Works, 1. 29. Zhuangzi, 6; Watson, Complete Works, 1. 30. Hans-Georg Moeller, “Zhuangzi’s ‘dream of the butterfly’: A Daoist Interpretation,” Philosophy East and West 49.4 (October 1999): 439–450.

96 | Michael Nylan 31. Quoted in Rebecca Solnit, Men Explain Things to Me (Chicago: Haymarket Books, 2014), 85. 32. Zhuangzi, 642; Watson, Complete Works, 140. 33. Zhuangzi, 56; Watson, Complete Works, 9. 34. See Zhuangzi, 1293; Watson, Complete Works, 362: “There are many who have mastered an art or two, and each believes that he has made something that is sheer perfection itself. But where, in fact, does there exist what in olden days would have been called ‘arts of the Way’? I say, ‘There is no place it does not exist.’ ” 35. Among the modern philosophers who do not believe that we fully understand how and why we act are Elizabeth Anscombe (“Mental states [desires, intentions, etc.] are not associated with actions via empirical laws, and so they cannot be said to cause actions”); Gilbert Ryle (“When we observe a person’s acts, we observe his intentions and desires, so the latter are not causes, but characteristics”); and Martin Heidegger (“A mood assails us. It comes neither from ‘outside’ nor from ‘inside’ but arises out of Being-in-the world, as a way of such Being.  .  .  .  [Understanding] has in every case already gone astray [yet there exists] the possibility of first finding  .  .  .  again in its possibilities”). For Anscombe and Ryle, see Essays on Anscombe’s Intention, ed. Anton Ford, Jennifer Hornsby, Frederick Stoutland (Cambridge: Harvard University Press, 2011), 14; for Heidegger, Being and Time, trans. John Macquarrie and Edward Robinson (New York: HarperCollins, 2008), 176. 36. Zhuangzi, 99; Watson, Complete Works, 18. 37. Zhuangzi, 66; Watson, Complete Works, 12. 38. Zhuangzi, 53; Watson, Complete Works, 9. 39. Zhuangzi, 52–53; Watson, Complete Works, 9. 40. Zhuangzi, 190; Watson, Complete Works, 39. 41. Zhuangzi, 174; Watson, Complete Works, 35. 42. Zhuangzi, 184; Watson, Complete Works, 37. 43. Zhuangzi, 105; Watson, Complete Works, 20. 44. Zhuangzi, 706–707; Watson, Complete Works, 152–153. 45. Zhuangzi, 707; Watson, Complete Works, 152–153. Forever lost to us are the wondrous opportunities for ordinary experiences to be sacralized, if we fail to fully attend to the present moment. Similarly, chapter 4 of the Zhuangzi (“Ren jian shi”) contrasts the attitude activating those godlike spirits that dwell within to the distraction or fear prompting the retreat or dormancy of the vital spirits in daily life. 46. Similarly, in the famous Butcher Ding story, it takes a moment for the Butcher to recover sufficiently from the powerful exchange that has taken place between the ox and him. 47. Zhuangzi, 49; Watson, Complete Works, 8. 48. Thus, my general agreement with the characterization of Zhuangzi made in Lee Yearley’s “The Perfected Person in the Radical Chuang-tzu,” in Experimental Essays in Chuang-tzu, ed. Victor Mair (Honolulu: University of Hawai’i Press, 1983), 125–139. 49. Unknown Pleasures (Director: Jia Zhangke). 50. For example, Zhuangzi, 2; Watson, Complete Works, 28. “Running around accusing others is not as good as laughing, and enjoying a good laugh is not as fine as going along with things. Be content to go along and forget [to dread] change, and then you can find a way to enter the mysterious oneness of Heaven.”

Vital Matters, A. C. Graham, and the Zhuangzi | 97 51. Zhuangzi, 213; Watson, Complete Works, 47. Indeed, chap. 4, like some later chapters, tries to reconcile the demands of both free roving (for one’s vitality) and public service (for one’s survival and livelihood). 52. Zhuangzi, 205; Watson, Complete Works, 42. 53. Zhuangzi, 674; Watson, Complete Works, 147. 54. Zhuangzi, 200–201; Watson, Complete Works, 40–41. 55. The term “regime of truth” is Foucault’s, cited in many texts. 56. Zhuangzi, 1205; Watson, Complete Works, 262. 57. Hence, the all-important metaphor of travel and touring in the Zhuangzi, a theme too large to take up in this essay. When exploring for pleasure, there is remarkably little striving; there is no particular schedule. As a result, the unexpected is met as often with delight as with outrage. Naturally, the traveler in a new land observes the local things and people more closely (guan 觀, used seventy-nine times alone). Thus the traveler notices things that may ordinarily strike a person at home as of no interest, seeing things not ordinarily perceived. Du 睹/覩, jian 監, jian 見, congming 聰明 are just a few of the words the Zhuangzi employs throughout the text. More than any other Chinese text, in my view, Zhuangzi emphasizes the pleasures attending curiosity, though the text would distinguish this from idle curiosity. The best analogy for life becomes that of the gift, unexpected and unrequested, but packed with potential pleasures. 58. Zhuangzi, 130; Watson, Complete Works, 25. 59. Zhuangzi, 138; Watson, Complete Works, 27. 60. Zhuangzi, 361; Watson, Complete Works, 61. 61. He replied, “I was born on dry land and feel safe on land; that’s what I am used to. I grew up in the water and came to feel safe in it; that’s what I meant when talking of a second nature. I don’t understand how I do what I do; that’s the destiny I was dealt” (see Zhuangzi 19; Watson, Complete Works, 152–153). 62. Solnit, Men Explain Things to Me, 96. 63. Pierre Hadot, What Is Ancient Philosophy? (Cambridge: Belknap Press of Harvard University Press, 2002), 31. 64. This is said of Song Rongzi. Zhuangzi, 17; Watson, Complete Works, 3. 65. I just spent the last year and a half undertaking this exercise and I believe my views of the Zhuangzi were substantially changed (improved?) by the effort. This is the place to thank my devil’s advocate, Trenton Wilson, for his efforts on my behalf. 66. Esther Klein, “Were There Inner Chapters in the Warring States? A New Examination of Evidence about the Zhuangzi,” T’oung pao 96 (2001): 299–369. (At the same time, I believe that some of Klein’s assertions about Han dynasty reading habits have been skewed by her assumption that citations from the Zhuangzi were always marked as explicit citations.) Wang Shumin, Zhuangzi jiaoquan, 306, is very clear that not everything in the Inner Chapters makes for superb reading, and not everything in the so-called Outer and Miscellaneous Chapters is worth rejecting. 67. Secondary studies of the Zhuangzi in Western languages seldom stray from a small selection of scenes (mostly from the Inner Chapters). See note 2.

5

Remarks on Intertranslatability and Relativism

Henry Rosemont Jr.

Few scholars of Chinese studies in the twentieth century can match the combination of breadth, depth, and originality of Angus Graham. His Disputers of the Tao1 is still selling well as it passed its twenty-fifth birthday and remains unparalleled as a textbook on Chinese thought. No one has translated and deciphered the Zhuangzi,2 Liezi,3 or the later Mohists4 better, nor approached the beauty of his very different translations in Poems of the Late T’ang5 or Poems of the West Lake.6 His philological articles remain on the “must cite” list of everyone who works with the syntax or semantics of classical Chinese.7 He also wrote straightforward philosophy both early in his career (The Problem of Value, 1961)8 and late (Reason and Spontaneity, 1985).9 Certain themes appear and reappear regularly in much of his writings no matter what the subject matter, particularly his abiding interest in the nature of language, the nature of thought, and the relation(s) between the two. The shadow of Benjamin Lee Whorf10 frequently looms in the background of much of Graham’s work, and is occasionally center stage: we see it in the title of one of Graham’s later essays: “Conceptual Schemes and Linguistic Relativism in Relation to Chinese,”11 and it opens an earlier essay on a closely related topic: “The concept of Being is a good test for the thesis of Whorf that the grammatical structure of language guides the formation of philosophical concepts.”12 In this essay I will follow Angus’s interests in these themes, especially as he took them up in some detail in the thirty-nine-page Appendix 2 in Disputers, supplemented by his remarks made in the “Reflections and Replies” section of the Festschrift dedicated to him and his work, and to

99

100  |  Henry Rosemont Jr. some of his essays gathered together by David Hall the year after Graham’s death.13 More specifically I want to address certain features of his views on these themes as they bear on the understanding of comparative philosophers and sinologists on the nature of the classical Chinese language, and the consequent relation between translations of classical Chinese texts and interpretive exegeses of them. I confess immediately to a personal interest in the matter. In the course of advancing his views on these matters, Graham proffers criticisms of some of mine on the same subjects, and those of my good friend and (later) collaborator Roger Ames. This has long puzzled both Roger and myself because as we read him, we both hold pretty much the same views on these matters that he did. Moreover, he frequently praised both of us for our efforts to translate and interpret classical Chinese texts as much as possible in their own lights, minimizing the imposition of a Western orientation on them, a point on which he has insisted throughout his career: “One explores Chinese philosophy by comparing and contrasting Western and Chinese concepts.”14 What I will be doing herein, then, is very much in the spirit of what Angus said to me when we were putting together the “Reflections and Replies” he wrote for his Festschrift: “I have very probably managed to bite all of the hands that have fed me.”15 The bite was quite gentle, however, and mine is intended in like manner, both for myself and for Roger. I have long held that all translation presupposes an interpretation of what the work is about. The more we think that a text is basically about morality, for instance, the more we will be inclined to employ our own moral vocabulary as the basis of our translation, which should be resisted (for reasons I will take up later).16 Among Roger’s many signal contributions to the field has been his consistent insistence that we try at all times to understand the Chinese in their terms, not ours, frequently requiring the use of English terms not usually found in Western philosophical writings.17 Hereafter, I will have little to say about the praise but hope to illuminate some of Graham’s ideas on language and thought by focusing on his criticisms of Roger and myself while responding to them. Most of the criticisms, I will suggest, will be seen to be based on misunderstanding more than disagreement. To begin, Angus writes with a touch of sarcasm: The game of demonstrating that some important concept of ours (“morality,” “ethics,” “rights,” “philosophy,” “civilization,” “science,” “art,” what you will) is missing in Chinese thought, although still popular, is quite pointless, since it can be played with any philosophical term one chooses, and not only with philosophical . . .18

Remarks on Intertranslatability and Relativism  |  101 He has tarred a number of comparativists past and present with this brush, but may well have had me in mind in particular, because I have long made much of the fact that Confucius should not be called a “moral philosopher,” nor read as doing “moral philosophy,”19 because there is no term in classical Chinese that straightforwardly (or even crookedly) can be rendered as “moral” as that term is used in the West today. And contra Angus, I did not (and do not) believe the exercise was “trivial.”20 Rather was I issuing a warning not to interpret the Lunyu in our terms, because it is not just “morals” that has no near lexical equivalent in classical Chinese; there are no equivalent Chinese terms for almost all the other concepts clustered21 around it in contemporary English-language moral discourse either: “rights,” “duty,” “choice,” “freedom,” “autonomy,” “dilemma,” “justice,” “liberty”—even a cognate for “ought” is not in the Master’s lexicon. Now if we can’t talk about moral issues without using these terms, and if none of them have clear analogues in Chinese, then it does seem to follow non-trivially that Confucius did not think about or do contemporary moral philosophy as we think about and do it, and we will therefore have to seek another interpretation of the Lunyu when translating it.22 To continue to interpret Kongzi as doing “moral philosophy” the way it is done in the contemporary West—using our lexicon in civil society no less than in philosophy—pretty much guarantees that we will see him as a well-meaning preacher, perhaps, but certainly not much of a philosopher, and as not really attending to our moral concerns as we think about them—insuring that we will not learn the important lessons I believe he has to teach us. Ames also believes we have much to learn from the classical Confucian writings, and in the course of his efforts to have those texts speak to us in their own terms he receives good marks from Graham, but a fair amount of criticism as well, one of which concerns Roger’s exposition of his interpretation of the Zhongyong. In many respects it is a mirror image of the criticism of me. Graham writes: Ames says that the opening of the Zhongyong (天命之謂性) “is often conceptualized and translated with strongly essentialistic assumptions to assert that ‘What is decreed by Heaven is called the xing,’ ” and decides that this phrase “might be more appropriately rendered ‘the relationships that obtain between man and his world (tian ming) are what is meant by xing,” To me, the second version is not a translation at all.23 Graham then continues, “The key terms are tian ‘Heaven’ and ming ‘decree,’ . . . inked to make a metaphor from the edicts of human rulers;

102  |  Henry Rosemont Jr. why is this crucial metaphor missing in the second [Ames] version?”24 In the first place, Graham has charged Roger with confusing translation and exposition in these paragraphs, then squishes some materials together to support his claim. But Ames was engaged in making an interpretive point, not a “translation”: If we take Tang Junyi’s counsel on ming to the opening passage of the Zhongyong, the phrase “what is decreed by Heaven is called the xing,” might be more appropriately rendered “the relationships that obtain between man and his world (tian ming) are what is meant by xing.”25 The point Ames is making here (grounded, paradoxically, in the analysis of xing proffered by Graham in earlier work on Mencius)26 is that the standard translation implies an essentialism that is not to be found in the Chinese, wherein, according to Ames, processes and relationships, not substances, make up the basic ontology. Moreover, on this interpretation, we must abandon the notion of transcendence as applicable to early Chinese thought, and abandon as well the idea of a God as He Who Dwells in Heaven. This is indeed a novel interpretation, not only of the Chinese text but our own basic conceptual scheme as well, and Graham finds Roger’s analysis an “attractive thesis.”27 Altogether unsurprisingly, Ames thus does not at all like “Heaven” as translation for tian (天)—all its Christian connotations and Western metaphysics—nor for ming (命) as “fate” or “destiny” (or worse, “God’s will”). But these misleading renderings of tian and ming from the Chinese do not seem to trouble Angus much, for he says, “One must insist that there is no latitude for more than trivial variations on ‘It is the decreed by Heaven which is called xing.’ ”28 Yet the only justification for this reading seems to be that it is the standard reading,29 and consequently has precedent on its side. Of course, a translator cannot ignore the specific terms in a textual passage being translated and replace them with terms not found in the language of the text. Eisegetics are notoriously poor translators. But liberties of this magnitude have not been taken by translators for quite some time now, even by fervent missionaries of one stripe or another. Graham continues the criticism of Ames: There are few if any terms in Chinese philosophical discourse sufficiently close to risk “relation” or “relationship” as English equivalent in a context where it will carry philosophical weight. Ames’s English will do as exposition, but presented as translation

Remarks on Intertranslatability and Relativism  |  103 it conveys the false impression that the Chinese says that xing consists of relationships. (emphasis in the original)30 Graham has an important point here, but it is one that Ames and I have been urging for years—as have other comparativists. Graham usually tends to denigrate it, however, as in the quote from him documented in note 18. But surely his comment there is out of place. If it is “pointless” for comparativists to note that there is no concept in classical Chinese equivalent to “morals” or “ethics,” why would he find the lack of a lexical equivalent for “relation” noteworthy for criticizing Ames? He is right on target when he insists on the highly misleading nature of using Western terms unreflectively when translating from the Chinese, as when he comments on the French sinologist Le Gall’s translation of qi (气) as des atomes in a quote from the Song Neo-Confucian Zhang Zai: “A reader asking the important question, ‘Is there atomism in Chinese philosophy?’ would find the wrong answer embedded in an actual quote from a Chinese philosopher.”31 But if asking about atomism is “an important question,” so, too (and more immediately relevant for Ames), is the question, “Is there a concept of transcendence, or Heaven in Chinese philosophy?” No small number of missionary and other translators have asked these latter questions, and many of them have given affirmative answers. Such a reading is not, however, confined to literalist Christians or other true believers. It is inherent in Western conceptual schemes as well as in the grammar of English, wherein “Heaven” is a place (a transcendent place conceptually, suggested by the capital letter), just as the surface grammar of English requires that “God” be a person, place, or thing dwelling there. Thus the following is not at all unusual: If a confessional note be allowed, I have somehow arrived at the point where I think “God” in association with the grounding mystery, but when I hear or read the word “God,” I immediately associate the word with a being, Our Father in Heaven.32 This is not the confession of a naïve undergraduate in an Introduction to Religion class, but of a distinguished professor of the philosophy of religion and of Buddhism, Sallie B. King. She is by no means alone in her reading, and Ames has been working long and hard to develop a different reading, just as I have been attempting to do with the concept of morals, especially as suggesting moral choice. Paradoxically perhaps, in the end Angus did seem to understand what we were about and think we’ve done it fairly well:

104  |  Henry Rosemont Jr. In treating [conceptual] schemes as equal, it becomes possible to use one to criticize something in another, as Fingarette uses Confucius to undermine our inner/outer dichotomy, Rosemont the problematic of moral choice, Hall and Ames the concept of transcendence.33 Turning to another theme taken up by Angus in some detail, we must appreciate at the outset that interpretation is not only about the content of the Chinese text(s) being translated, it also involves the translator’s answer to the question of how the Chinese language presents that content, and in order to answer this question I must have previously answered another, more basic one: how is meaning conveyed in classical Chinese, and to what extent do its obvious grammatical differences with English affect those meanings and, consequently, translation? This, too, is interpretation, and every translator from the Chinese must address it, because its grammatical structures do indeed differ dramatically from those of English, which raises once more the Whorfian hypothesis of linguistic relativity. In 1985 Donald Davidson published an influential paper, “On the Very Idea of a Conceptual Scheme,”34 claiming, nuances aside, that the idea of cognitive relativism—with truth relative to particular conceptual schemes— was incoherent. Davidson’s argument hinges on his rejection of what he calls “The third dogma of empiricism,” namely, that we can meaningfully distinguish conceptual schemes from an uninterpreted reality—taken as “a neutral ground” or “coordinate system”—by means of which we could compare and contrast differing conceptual schemes. He says: In giving up dependence on the concept of an uninterpreted reality  .  .  .  we do not relinquish the concept of objective truth—quite the contrary. Given the dogma of a dualism of scheme and reality, we get conceptual relativity, and truth relative to a scheme. Without the dogma, this kind of relativity goes by the board. Of course truth of sentences remains relative to language, but that is as objective as can be.35 Graham rejected Davidson’s claim, saying that the postulation of conceptual schemes different from our (modern Western) own was “inescapable” for scholars of the language and thought of other cultures, and indeed, was an “indispensable tool” for such inquirers, to which Davidsonian concerns and objections did not apply.36 At the same time, Graham seems to have had a genuine fear of the ideas of epistemological or linguistic relativism, for he regularly used the word “chaos” with respect to those ideas.37 The idea that truth might be relative to a

Remarks on Intertranslatability and Relativism  |  105 conceptual scheme was as abhorrent to him as it was incoherent to Davidson. To be sensitive to the differing thought patterns of another culture while yet not descending into philosophical unintelligibility (to him), Graham turned to syntax, thus embracing the weaker form of the Sapir-Whorf hypothesis: Since the interrogative words of different languages are not synonymous, and connect with their syntactic structures, it seems inevitable that Aristotle or anyone else asking questions with them will be categorizing along lines initially set by the language in which he thinks.38 But similarities and differences in our common world can be linked differently in different cultures on the basis of the prelogical correlative thinking that is characteristic of all cultures, according to Graham.39 They will thus be conceptualized in different ways, largely in keeping with the syntactic and semantic patterns of the grammars of their respective languages. This does not imply, however, that statements of fact are dependent on a specific conceptual scheme, but that when you want to confirm or refute them by reason and observation, it is enough to “understand their place in the appropriate conceptual scheme, you do not have to share the scheme.”40 In other words, although I link and distinguish things differently than you do, I can come to understand your linkages and distinctions. To take a simple (and simple-minded) semantic example, I can learn that and why you link whales with sharks and halibut while I categorize them with lions and chipmunks. And we can do this, according to Graham (and Davidson, on this score), because we can both point to what we call whales, sharks, halibut, lions, and chipmunks. So far, so good: Ames and I would, minor details aside, concur with Graham on these points. But in the course of making them and related others, he criticizes both of us for explaining some Chinese conceptualizations on the basis of the grammar of the classical language that he says are “totally obsolete assumptions.”41 In my case he picks out a few examples from my “On Representing Abstractions in Archaic Chinese” for specific criticisms, beginning with my claim that the classical language was not to be seen strictly as a transcription of speech, because it does not seem altogether capable of unambiguously expressing grammatical relations, which every spoken (and thus natural) language is capable of doing. Hence, in some respects at least, it should not be seen as a natural language. As Graham and everyone else admits, classical Chinese is dependent on word order (and later, “particles”), but not uniformly so, and consequently semantics must play a larger role in the interpretation of classical Chinese sentences than syntax.

106  |  Henry Rosemont Jr. Now perhaps Angus and I simply have different ideas of what a “grammar” is, because he otherwise seems to admit the accuracy of my claim himself a few pages later: The parallelism so noticeable in Chinese style is not mere decoration but an indispensable aid to syntax. Given a language in which sentences are structured by word-order, and not only can verbs stand in nominal positions but nouns have causative and putative uses in which they stand in verbal positions, a sentence or clause of any length will be structurally ambiguous unless clarified either by particles or by parallelism with another similar in structure.42 But as I studied linguistics, syntax was syntax and style was style, and they were not to be intermingled. Moreover, for myself the rules of grammar are just that, rules, not to be violated, as Angus knew was the case in classical Chinese word order interpretation. A simple rule of English, for example, is that whenever a noun uses a definite article it must always precede and never follow the noun. Or articles and modifiers are optional in noun phrases, but the nouns themselves never are. I simply do not know what prompted Angus to criticism on this point. But much more important on this score, having accepted the basic Chomskian claims of universal grammar, I must also accept that we are all “hardwired” for language in the same way, and thus must accept in turn that whatever can be said in any one natural language can be said in any other if the speakers are capable of fully utilizing the resources of their language.43 But Angus did not believe in the full intertranslatability of languages.44 Yet if he rejects Davidson’s claims and insists on a multiplicity of conceptual schemes while also accepting Davidson’s arguments for keeping the idea of objective truth, then, if Chinese is not fully intertranslateable with English, and there are true statements in English, it must follow that there are at least some objective true statements that monolingual Chinese (or English speakers?) cannot come to know. I doubt seriously that Graham would accept this conclusion; I certainly don’t. But I am not forced to, even though both he and I accept and reject the same elements of Davidson’s arguments. For me, what cannot be translated from classical Chinese can be put into spoken Chinese, and the truth may thus be told. How Angus would deal with that unpalatable (to me) conclusion I am not sure; to reject the Chomskian orientation on this score would take him from the frying pan into the fire, for without that orientation we must entertain the possibility that the more impoverished a language might appear to be, the less intellectually capable its speakers.

Remarks on Intertranslatability and Relativism  |  107 Roger Ames appears to base his (and David Hall’s) accounts of the relation between the classical Chinese language and Chinese thought on even more obsolete assumptions about the nature of the former, according to Graham, despite his frequent commendation of their views on the latter.45 But he insists that the latter views have nothing to do with the grammar, asking, “[W]hat have they to do with such essentially grammatical matters?”46 But he himself has just quoted Hall and Ames on the same page with their answer to his challenge: “The claims we wish to make about classical Chinese are themselves consequences of the broader cultural contrasts for which we have repeatedly argued throughout this work.”47 That is to say, given the more event-oriented ontology Hall and Ames interpretively posit for ancient China and contrast with the much more substance-oriented nature of most Western ontology, it should be reflected to some extent in the grammar of the language (in both cases): Subject-Predicate (NP VP) for Indo-European substance languages, but not Chinese, wherein verbs (processes, events, actions) are regularly the default reading. One major task of the translator from the Chinese who interprets the text as reflecting a process ontology is thus how best to capture at least the flavor of the language that reflects that ontology, and not put it only into substance—qualities/attributes grammatical form. If translators reject the Hall and Ames process ontology for the Chinese, or it never occurs to them, they will continue to translate many passages in the Lunyu as, for example, “so-and-so asked about government.” Employing the process/event/activity ontology for the Chinese, however, Ames and I tried to reflect it when we later did our translation of the Lunyu: “so-and-so asked about governing effectively.”48 This specific passage requires elaboration in the context of the importance of interpretation for translation. The term zheng (政), invariably rendered simply as “government,” is strictly word for word. When Ames and I changed it to the dynamic “governing,” however, we added “effectively.” What Graham would have said about that we do not know. But on our interpretation of what the text is about, it was clear to us that the component graph zheng (正), while clearly phonetic in zheng (政) maintains its semantic content as well: upright, straight, correct, and so on. Like many items in the lexicon of the Lunyu it is used normatively; hence, our addition of an appropriate adverb for the English version. Another, related example of where we differ from other translators because of our interpretation(s) of the text is the well-known passage on the “Rectification of Names” (12.11), which we rendered: “The ruler must rule, the minister minister, the father father, and the son son.” The last, of course, is ungrammatical in English but nevertheless conveys information, we believe, in a way other translations do not. (And these terms, too, are

108  |  Henry Rosemont Jr. clearly being used normatively in this passage.) We do need to invoke our interpretation of the text in order to defend our reading of its language, but this does not mean that either of us—especially Roger in Graham’s critique—was reading it eisegetically. To conclude these brief reflections, if I had to make a guess, I’d suggest that Graham’s concerns about some of our views of classical Chinese were grounded in the fact that our views differed just enough from his to make our ontologies just enough different from his—in the Whorfian sense—as to make the “quasi-syllogism,” for which he argued long and hard as an interpretive measure in Disputers of the Tao, more difficult to generate except with a substance ontology, a point that cannot be taken up herein.49 But this point in no way diminishes the debt we both owe him for all that he taught us over many enjoyable years together, nor the abiding affection with which we continue to remember him.

Notes   1. A. C. Graham, Disputers of the Tao (La Salle: Open Court, 1989).   2. A. C. Graham, Chuang-tzŭ: The Seven Inner Chapters and Other Writings from the Book Chuang-tzŭ (London: George Allen & Unwin, 1981).   3. A. C. Graham, The Book of Lieh-tzu (London: John Murray, 1960; rpt. Columbia University Press, 1990).   4. A. C. Graham, Later Mohist Logic, Ethics, and Science (Hong Kong: Chinese University Press, 1978).   5. A. C. Graham, Poems of the Late T’ang (London: Penguin Classics, 1965).   6. A. C. Graham, Poems of the West Lake (London: Wellsweep Press, 1990).   7. A full list of Graham’s published articles is in his Festschrift, Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus C. Graham, ed. Henry Rosemont Jr. (La Salle: Open Court, 1991).  8. A. C. Graham, The Problem of Value (London: Hutchinson’s University Library, 1961).   9. A. C. Graham, Reason and Spontaneity (New York: Barnes & Noble, 1985). 10. The locus classicus for Whorf’s work is Language, Thought, and Reality (Cambridge: MIT Press, 1964). 11. Included in the anthology of Graham’s essays edited by David Hall, Unreason within Reason (La Salle: Open Court, 1992). 12. Ibid., 85. 13. See note 11. 14. Graham, Disputers of the Tao, 396.On supportive words for our work, see ibid., 390–393; Graham, Chinese Texts and Philosophical Contexts, 288; Graham, Unreason within Reason, 78, 83. 15. Graham, Chinese Texts and Philosophical Contexts, xiv.

Remarks on Intertranslatability and Relativism  |  109 16. I began harping on this theme in “Against Relativism,” in Interpreting Across Boundaries, ed, Gerald Larson and Eliot Deutsch (Princeton: Princeton University Press, 1987). 17. Beginning with his first collaboration with David Hall, Thinking Through Confucius (Albany: State University of New York Press, 1987). 18. Graham, Disputers of the Tao, 396. 19. I first advanced this position in “Against Relativism.” I have not seen a citation of this particular article in any of Graham’s later writings, however. I later came to refer to this view as a methodological requirement for translation, using “concept-cluster” to denote a set of terms (clusters) gathered around a central one. The fullest statement of my position here is in Against Individualism (Lanham: Rowman & Littlefield, Lexington Books, 2015). 20. Graham, Disputers of the Tao, 396. 21. On concept-clusters, see note 19. 22. After we began our collaboration with the Lunyu, Ames and I began to develop our ideas on classical Confucianism as an ethics of roles. 23. Graham, Chinese Texts and Philosophical Contexts, 288. 24. Ibid. 25. Ibid., 154. 26. Discussed by Ames in his contribution to Chinese Texts and Philosophical Contexts. 27. Graham, Disputers of the Tao, 394. 28. Ibid., 289. 29. Graham, Unreason within Reason, 66. 30. Graham, Chinese Texts and Philosophical Contexts, 289. 31. Graham, Unreason within Reason, 61. 32. Sallie B. King, “Concepts, Anti-Concepts and Religious Experience,” Religious Studies 14.4 (1988): 458. 33. Graham, Unreason within Reason, 78. We must note in passing while leaving this point that Ames (and Hall) published a translation of the Zhongyong a decade after Graham’s death, in a manner that the latter’s shade should not have objected to: “What tian 天 commands (ming) 命 is called natural tendencies. (xing 性)” 34. Donald Davidson, “On the Very Idea of a Conceptual Scheme,” Proceedings and Addresses of the American Philosophical Association 47 (1973): 5–20. 35. Ibid., 20. 36. Graham, Unreason within Reason, 59. My own views on truth in the classical Chinese context are sketched in “Does the Concept of ‘Truth’ Have Value in the Pursuit of Cross-Cultural Philosophy?,” Confluence: Online Journal of Cross-Cultural Philosophy 1.1 (2014): 150–217. 37. As, for example, on 321 of Chinese Texts and Philosophical Contexts; 428 of Disputers of the Tao; and 59 and 67 of Unreason within Reason. 38. Unreason within Reason, 78. 39. Ibid., 67. 40. Ibid., 78. 41. Graham, Chinese Texts and Philosophical Contexts, 392.

110  |  Henry Rosemont Jr. 42. Graham, Disputers of the Tao, 404; Graham, Unreason within Reason, 78. 43. The key word in this sentence is “natural.” Because classical Chinese is not, in my opinion, strictly a transcription of speech, despite the numerous uses of “The Master said” in the Lunyu. Writing is not “natural” in the way that speech is because it must be taught and practiced using the eye and hand, rather than simply acquired—beginning in the bassinet. Graham seems to have accepted my overall phonetic argument on this point as well. Graham, Disputers of the Tao, 290. 44. Graham, Unreason within Reason, 65. 45. Graham, Disputers of the Tao, 393. 46. Ibid. 47. Ibid. 48. Roger T. Ames and Henry Rosemont Jr., The Analects of Confucius: A Philosophical Translation (New York: Random House/Ballantine Books, 1998). 49. The quasi-syllogism was the major theme of Graham’s Reason and Spontaneity as well, and he came to believe it could/did solve the problem of the fact/ value split. The quasi-syllogism originated with Graham’s studies of Zhuangzi, not grammar or language proper, so it would take us far afield to consider it herein. A fuller account of it and its implications was the subject of my contribution to Chinese Texts and Philosophical Contexts (“Who Chooses?”), and he replied at some length to it in his remarks on all the contributions.

6

Getting to the Bottom of “Things” (wù 物) Expanding on A. C. Graham’s Understanding

Robert H. Gassmann

1. Introduction Angus C. Graham, under the heading “Technical Terminology,” presents the following summary of the way the word wù (物 ) was used by the later Mohists, the writings of whom he dates to about 300 BCE: Wu 物 “thing”’ [.  .  .]. The later Mohists use wu when they wish to refer generally to the thing meant by a name, not to a particular object which the name may single out. However, the word has a particular reference at least once (B9); wu is not exclusively universal as shih “object” is exclusively particular. Wu may refer to such abstract entities as love and benefit (EC 2). Similarly in older chapters of Mo-tzŭ we find objectors dismissing universal love as an “impracticable thing” (ch. 15, Sun 70/5 不可行之物 ) and recommending aggressive warfare as a “profitable thing” (ch. 19, Sun 95/–1 利物 ).1 This tidy characterization of the meaning of the word wù reflects its use in a systematic text of comparatively late date and is therefore basically synchronic. The untidy side of the lexicography of wù presents itself in the various dictionaries (usually “character books”) proposing equivalents for ancient Chinese words written with the same character. Here, the entries dealing with wù are a fairly distracting challenge even for seasoned users,

111

112  |  Robert H. Gassmann for large is the array of texts drawn on, many are the centuries covered, and not many are the characters presented as the written form of such a number of diverse words. Two random examples: 503 h–i: [*miwət / miuət / wu] thing, object, article (Shï); colour of cattle (Shï); quality (Shï); divide acc. to quality, to sort, classify (Tso); class, sort (Tso); variegated pennon (Chouli). [. . .].2 Ding, (Lebe-)Wesen (2), Wesen, Art; Kennzeichen; Gegenstand, Objekt; Sache, Aufgabe; die Dinge (wie sie sind) / Verhältnisse; Sachverhalt, Fall; Artikel (von Verordnungen); sortieren; als Ding behandeln, verdinglichen. (134)3 But these two lists are far from really mirroring the diversity to be found in translated texts. In the eighty-four instances of wù in the Chūn Qiū/Zuǒ Zhuàn (according to my count in an electronic version), James Legge offers the following variety of equivalents or quasi-equivalents in English (some of them are questionable due to uncertainties of identification and therefore marked with a question mark): 18–19: creatures, ramifications, materials; 40: articles, great objects and offices, things, rank, day; 120: offerings, appearances (of the clouds); 145–146: tribute; 217: array (of provisions); 244: character; 283: unfriendly (?); 317: signals, dresses; 328: productions (of the earth), case; 346: character (of the ground); 395: everything; 580: information; 580–581: hands (?), desire (?), signification; 625: order (of the kingdom); 634: circumstances; 655: use (according to ability); 667: robes; 674: physic; 708: sphere; 726: Being; 741: situation (of the ground); 754: appendages (of State); 778: mark; 794: all that (of old) belonged; 813: (great-class) offering.4 Anyone with a linguistic background or with a sensibility toward the use of words would, I believe, tend to assume that this semantic jumble is primarily indicative of a flawed understanding of the basic meaning(s) of this word.5 As a matter of fact, it appears that the third list is simply another bad case of “lexicography by assumed contextual fit”—the translations suggested reflecting the translator’s understanding of a specific instance in a specific context, rather than being based on an analysis of the meaning by application of grammatical and lexicographical methods. As no attempt is made in the first two lists to integrate the various meanings or to show the lines along which derivation possibly ran, we are left without a reasoned assumption as to the basic meaning—which might be “thing,” but may not necessarily be so. Where is the link connecting this word to the so many, prima facie differing contexts

Getting to the Bottom of “Things” (wù 物) | 113 it is used in? And how do we bridge the gap between A. C. Graham’s neat characterization of the word in later Mohist texts and the medley of meanings apparently necessary when translating a text like the Zuǒ Zhuàn? An attempt to better this situation is therefore the main incentive for the present note,6 and the Zuǒ Zhuàn has been chosen as a mirror-text because it comprises many occurrences and usually supplies sufficient nonargumentative context and, it is to be presumed, predominantly normal usages.

2. Tracking the Semantics of wù The only safe way to approach the meaning(s) of words in a language confined to received textual remnants is to find sufficiently rich contexts that allow one to methodically and safely fix the basic outlines and then to test these in less obvious contexts. Such an informative and, as I would suggest, benchmark-setting instance is the following (translations are mine, Legge is given as a reference for the reader wanting to look up the context, and wù remains transcribed for the moment): Zhāo 32.6: [. . .] When wù were created,7 (they) existed as twos, threes, and fives, and they existed as correspondents or complementaries. Thus heaven has the Three Celestial Bodies, earth has the Five Agents, and skeletons have a left and a right (side). Each one has consorts or partners. The king has the (three) ministerial dukes, all lords have qīng-ministerials: they all are such who have complementaries. Heaven brought the Jì-lineage into existence in order to give it as complementary to the marquis of Lǔ. [.  .  .]. (Legge 741) [. . .] 物生, 有兩、有三、有五、有陪貳. 故: 天有三辰, 地有 五行, 體有 左右. 各有妃耦. 王有公, 諸侯有卿, 皆有貳也. 天生季氏, 以貳魯侯. [. . .] Table 6.1 subject verb object subject verb object



[ 物 ]

有

兩

7

地

有

五行

2

[ 物 ]

有

三

體

有

左右

[ 物 ]

8

3

有

五

各

有

妃耦

[ 物 ]

9

有

陪

王

有

公

5

[ 物 ]

10

有

貳

11

諸侯

有

卿

6

天

有

三辰

12

皆

有

貳

1

4

114  |  Robert H. Gassmann Let us first examine the predominant grammatical structures of this excerpt. Excluding the first clause wù shēng 物生 (a passive), which furnishes the formal subject wù of clauses 1–5, table 6.1 shows the simple and uniform syntactical build of the statements. Although yǒu is the verbal nucleus in all clauses, the case relationship between the subjects and the nucleus is of two kinds: in clauses 6–8 (9) the subject is a locative, in clauses 10–11 (12) a possessive. This implies that in clauses 6–8 the object is conceptually or even physically part of the subject, that is, located “in” the subject (indisputably clear in clause 8); in clauses 10 and 11, however, the object is attached to the subject or in a positive relationship with it. The part-whole relationship in clauses 6–8 is a fact produced by “nature,” the (asymmetrical) part-part relationship in clauses 10 and 11 is a matter of social convention. As clauses 1–3 are related to clauses 6–8, and clauses 4 and 5 related to clauses 10 and 11 (12), the subject wù in clauses 1–3 turns out to be the general term for composite aggregates (6–8), in clauses 4 and 5 for combined aggregates (10–11).8 That wù in clauses 4 and 5 refers to the aggregate, that is, to the combination of the subjects and objects in clauses 10 and 11, is evidenced, on the one hand, by the use of the totalizing reference jiē (“all” 皆) in clause 12, and on the other, by the fact that èr (“complementary” 貳 ) cannot but work both ways, that is, inclusively referring to subjects and objects. The conclusion to be drawn is as follows: if clauses 1–5 are to work according to one and the same rule (which for formal reasons one clearly must assume), they must be presenting a generalization extending over all cases in clauses 6–12. The extracted implicit modifications “composite” and “combined” (italicized in the preceding paragraph) indicate subtypes of aggregates; this leaves the word wù as the bearer of the semantic information “aggregate.” The generality of this meaning (and the plausibility of the preceding result) is evidenced by the fact that the case relationship between the subjects wù and the nucleus yǒu (clauses 1–5) is not affected by the difference between locative and possessive: whether one translates “(the class of) aggregates has duals (etc.)” (possessive) or “in (the class of) aggregates there are duals (etc.)” (locative)—both statements not only work without jarring but also have the same meaning. It turns out that the syntactic simplicity of the clauses discussed is not only quite misleading but also invites imposing a preunderstanding on them (i.e., that wù cannot mean anything else but “thing”), as Legge apparently did: “Things are produced in twos, in threes, in fives—in pairs.” Talking of (material) things being produced, he is clearly not talking of the locative or possessive subject but of the objects coming in twos, threes, and so on. Now the objects are the abstract nouns liǎng, sān, wǔ, péi, and èr. These designate, or act as class designations for, multiples, that is, for pairs or duals (twos), for triads (threes; compare sān 三/叄/參), for pentads (fives; compare wǔ 五/伍),

Getting to the Bottom of “Things” (wù 物) | 115 for correspondents (péi), or for complementaries (èr). A triad, as an example, would therefore be a designation for three elements in some way belonging together (i.e., forming an aggregate); the noun does not refer to the material things but to the relationship between them. Legge’s translation ought to be at least changed to: “Multiples are produced in sets of two, three, five—in pairs.” A look at one of the subtypes of wù, that is, the one with a pair of elements or a union of two and specifically named èr (“complementary” 貳 ) (clauses 5 and 12), can round off the arguments. In the statement “The king has the (three) ministerial dukes, all lords have qīng-ministerial(s): all of them are such who have complementaries,” the expression “ministerial duke” (gōng 公 ), actually referring to the “Three Ministerial Dukes” (sān gōng 三公 ), clearly denotes one part of an èr (cf. clause 12: “they all have complementaries” jiē yǒu èr [ 皆有貳 ]). An èr is a wù (clause 5); wù therefore is the class name of èr in its quality as a subtype of the class of multiples or aggregates. The èr is a fundamental organizational structure of ancient Chinese society: it designates “dyads,” that is, complementary role pairs such as ruler-ministerial (as here), father-son, among others. These èr are dyadically paired individuals, but I believe that this subtype also comprises paired entities on a more generalized, collective, or abstract level, that is, binomial expressions of the type “princely family and important families” (guó jiā 國家 ), “consanguineous and affinal relatives” (qīn qī 親戚 ), “noble and base” (guì jiàn 貴賤 ), and so on.9 To summarize: The arguments presented so far show that in the Zuǒ Zhuàn passage just discussed, wù does not mean “thing,” but rather “aggregate” or “assemblage,” that is, a variously structured union or group of two or more objects.10 The elements or objects of the (sub)group, that is, the “things,” are themselves not designated as wù (exception, see p. 121), and the object expressions “dual,” “triad,” among others here are not counting these “things” but marking the (no doubt quantitative) differentia specifica of certain subtypes of aggregates when compared to others.11 The clause, for example, [wù] yǒu (noun) is therefore to be understood as: “(the class of) aggregates comprises n-multiples” (n = 2, 3, or 5). This conforms to, as well as confirms, the indirect assertion in the passage discussed that there are no aggregates comprising only one single element (which would clearly be strange). The lower limit for the number of elements in an aggregate being two (i.e., a pair), the upper limit is, according to this context, five; according to Āi 7.3 there apparently could also be up to twelve. This also confirms that the expression “the Three Celestial Bodies” (sān chén 三辰 ) designates the aggregate of these three bodies, that is, one (type of) ensemble, and not three (separate) “things.” We shall shortly return to the implications of this finding and try to explain the surprising “vanishing act” of “thing” or “object.” The preceding translation can now be plainly put as follows:

116  |  Robert H. Gassmann Zhāo 32.6: [.  .  .] When aggregates were created, they existed as twos, threes, and fives, and they existed as correspondents or complementaries. Thus heaven has the Three Celestial Bodies, earth has the Five Agents, and skeletons have a left and a right (side). Each one12 has consorts or partners. The king has the (three) ministerial dukes, all lords have qīng-ministerials: they all are such who have complementaries. Heaven brought the Jì-lineage into existence in order to give it as complementary to the marquis of Lǔ [.  .  .]. Note, now, that despite the obvious fact that this passage is referring to aggregated entities, there is no mention of the important type of assemblage designated by “class, category, kind, type” (lèi 類 ).13 A lèi-class is ordered according to a logical or linguistic feature, for instance, a bird is an animal, animals are living beings, living beings are “things.” Clearly, such classes are not confined to groupings of two, three, or five members or elements but comprise simply either the naturally given (and usually unknown or indefinite) number or then totalizing numbers such as, for example, one hundred elements (cf. the “Hundred Officers” bǎi guān 百官 ). How, then, are wù-aggregates ordered? Based on the examples given, we may note that aggregates are of two designs: one subtype is basically related to lèi-classes as it is ordered according to a common feature and usually has its own “class”-designation (e.g., chén [celestial body], xìng [agent], tǐ [(side of the) body], zhū-hóu [all lords], etc.), but it is restricted to an artificially fixed number of tokens, which usually shows up in a modification with a number expression (e.g., the Three Celestial Bodies, i.e., sun, moon, stars). Such a wù, such an aggregate—which we shall term a “set”—may have correspondent sets (péi 陪 ) comprising the same number of tokens. Sets comprising five tokens are probably the best known: the five elements or agents, seasons, quarters of the empire (sometimes mistaken for directions), colors, organs, and so on. The tokens that correspond to each other in the different sets (e.g., water, winter, north, black, liver) also constitute sets cutting across the sets of origin (cf. Zhāo 29 fù 4). These arrays or aggregates we shall call “sets of correspondent tokens.” The second subtype of aggregate comprises correlated and complementary pairs (èr 貳 ), for example, “prince” (jūn 君 ) and “ministerial” (chén 臣 )—and pairs only.14 These pairs constitute functional sets based on an inherent feature of the word that imperatively demands its complement: a father is only a father if he has a child, an older person can only be termed old when younger ones exist, and so on. This subtype is usually based on dyadic social roles. The most important of these “dyadic sets” or “dyads” are termed lún (倫 ), “(dyadic) relationships.” It is interesting to note that only

Getting to the Bottom of “Things” (wù 物) | 117 members of such pairs are termed “disassembled; orphaned” (pǐ 匹 ) when the dyad has, willfully or fatefully, been dissolved. Having established that wù designates aggregates or sets, we can proceed to the following observation: wù are counted—grammatically speaking: wù can be modified by an adjective denoting quantity, for example, “one hundred wù” (bǎi wù 百物 ) (s. Xuān 3.5), or “ten thousand wù” (wàn wù 萬 物 ). According to received knowledge these expressions are to be understood as the “one hundred things” or the “ten thousand things.” Whatever that may mean in a given context, the fact that wù has the meaning “aggregate” naturally triggers the question whether we should be talking of two, five, one hundred, or ten thousand things, or whether such expressions denote two, five, one hundred, or ten thousand aggregates, sets, or dyads! An illustrative example for making aggregates of sets (and thus counting aggregates) is found in Āi 7.3. Here the ruler of Wú (吳 ) demands one hundred láo-sets of animals from the marquis of Lǔ (láo 牢 designates a set of sacrificial animals, usually an ox, a sheep, and a pig). Lǔ protests this exorbitant demand and argues for a lower number (at most eleven) in the following way: [.  .  .] When the Zhōu had become kings, they regulated ritual behavior. The topmost aggregate (due) did not exceed twelve (láo-sets), because they fixed this in accordance with Heaven’s largest number. Today, (the viscount of Wú) is abandoning the ritual behavior of the Zhōu and saying that there must be one hundred láo-sets, [. . .]. (Legge 813) Āi 7.3: [.  .  .] 周之王也, 制禮. 上物不過十二, 以為天之大數也. 今棄周禮 而曰必百牢, [. . .]. Twelve láo-sets (shí èr láo 十二牢 ) are the largest aggregate (shàng wù 上物 ). This clearly demonstrates that expressions like bǎi wù and wàn wù do not on principle resist an interpretation as “n sets.” Any sorts and numbers of elements, that is, “things” or sets and aggregates, can, for a specific purpose or reason, be gathered into a (superordinate) “aggregate.” This we could compare to a motor car: all the parts that in the course of assembly result in a car are not necessarily logically predestined to become a car. Most of them (e.g., the wheels) could also be used for different aggregates. But once they are brought together in the assembled entity, that is, in the car, they become constituent parts of it. Although aggregates may be, and mostly are, to a certain extent ruled by logical or language-based common features and are thus similar to “aggregates” of a special type termed “class” (lèi), they can also be randomly assembled and submitted to a given purpose (there is

118  |  Robert H. Gassmann nothing to force a ring, a crown, a scepter, and an orb into a set symbolizing the status and power of a king, but once this has happened, they become a set; for ceremonials sets, cf. Zhuāng 24.6). “Aggregate” (wù) is, therefore, the general term; “class” (lèi) is a defined aggregate. In uncertain cases (i.e., when we are not sure whether sets or dyads are being referred to) we shall use the general translational equivalent “aggregate.” That the two words “aggregate” and “class” are categorically related is demonstrated in the following example: Wén 18.9: [. . .] In ancient times, the August One, the Chief of the lineage of the Hóng, had a wayward son. He refused correct behavior, committed hidden misdeeds, and was fond of practicing unlucky obligations. He (thus) disgraced (his) class and corrupted (his social) ensembles or dyads.15 [.  .  .] (Legge 283) 昔帝鴻氏有不才子. 掩義隱賊, 好行凶德; 醜類惡物. [. . .]

What is meant by “disgracing his class and corrupting (his social) ensembles or dyads”? According to the understanding outlined here, the two ways of categorizing hinge on the dual embedding of the word “son” (zǐ 子 ): as a member or token of the class of children (comprising sons, daughters, elder and younger brothers) he by his behavior brings disgrace on his class (lèi), that is, on other sons, daughters, and others; as a member of a dyad, for example, the “father-son” dyad or possibly the “elder brother– younger brother” dyad (wù), he by his behavior corrupts such fundamental relationships. Let us now return to the “vanishing act.” In the following passage, wù is the class name “set” of six objects belonging to an aggregate. Based on our preceding insights and on the new meaning of wù, we conclude that the term “set” in the class name characterizes the six objects included in it, that is, these six objects are themselves aggregates comprising fixed numbers of tokens. As a matter of fact, common sense confirms this: the year has a fixed number of days (or seasons, months, hours), there are four seasons with a fixed number of months (or days), the day has a fixed number of hours, the month a fixed number of days, the stars have their constellations, and in the celestial bodies we have sun, moon, and stars: Zhāo 7.7: The Patriarch asked: “What does (the expression) ‘Six Sets’ mean?” (He) answered: “(solar) years, seasons, days, months, stars and celestial bodies—these are the designated (sets).” (Legge 619) [. . .] 公曰:「何謂六物?」對曰:「歲、時、日、月、星、辰, 是謂也.

Getting to the Bottom of “Things” (wù 物) | 119 Table 6.2 set

year

object

day



day

set

hour object

The expression “Six Sets” (liù wù 六物 )16 opens up a new view, which we can exemplify with the word “day(s)” figuring in the list. This word, on the one hand, is the designation of the set “day(s),” comprising a fixed number of hours that are structured in a certain way, on the other hand, it is the designation of a member object or token of several of the sets listed, that is, of the object or token “day” as occurring in a fixed number and being an element of the listed sets “month,” “season,” or “year.” This double classification, on the one hand object name (left column), on the other set name (right column), can be transferred and applied to the word wù: On the one hand, it designates the superordinate aggregate or set, for example, the set “season”; on the other hand, it can designate an object of this set, for example, the object or token “month” or “day.” In other words: wù designates objects when and if they are aggregates. So, the motor of the motor car is an object because it is an aggregate within an aggregate. But, and this is the absolutely crucial point in our search for the vanished “thing,” the motor car as an aggregate, as a consciously conceived cohesive form, is without doubt also an object, and it is so no matter whether it is explicitly integrated into a superordinate set or aggregate. Probing the possible limits in a thought experiment leads to the following insight: on the one hand, it is conceivable that a specific aggregate is the biggest possible aggregate and thus no longer an element within an even higher-level aggregate. This would be the case for the universe; it would have the status of an object without a parent aggregate. At the other extreme, we would find minute objects no longer perceivable as aggregates with our senses but—due to the experience with linear regressions—no doubt conceivable as such. The biggest aggregate is a thing; the smallest things most probably are or must be aggregates—this entitles us to consider every aggregate an “individual” object and—yes—to call it a “thing”—and vice versa.17 This also strongly suggests that “things” in ancient China were conceived as “aggregates” (see further arguments). The following example dealing with medicines confirms this result and takes us one step further. The noun wù in the expression yào wù (藥物) in the following excerpt clearly shows that a medical potion was understood to be an aggregate of medical substances.18 This reminds us that an aggregate is the result of an assembly process, of a mixing together of elements or substances (some perhaps themselves aggregates). This statement sounds

120  |  Robert H. Gassmann so innocently self-evident that it would not seem to deserve mention, but it is important because linguistically it translates into the insight to be expanded on later, that an aggregate (noun) is the result of an assembling process (verb). In other words: we shall be on the lookout for evidence of a verb wù written 物. Zhāo 19.2: Summer: The Dào-Patriarch of Xǔ [.  .  .] drank the medicine (presented by) the crown prince Zhǐ and died. The crown prince fled to Jìn. (The Chūn Qiū) has the following entry: “(The crown prince of Xǔ) assassinated his ruler.” (To this) the princely squire says: “Whoever serves his ruler by exhausting his cardial senses and corporeal strength is one who is permitted to leave the aggregates of medical substances19 to someone else.” [.  .  .] (Legge 674) 夏，許悼公 . . . 飲大子止之藥 卒. 大子奔晉. 書曰「弒其君」. 君子曰: 「盡心力以事君, 舍藥物可也.」

From here it is only one obvious step to concluding that every aggregate that is (a) a consciously conceived cohesive form and (b) the bearer of a name qualifies as an object or thing, no matter whether it is or is not an element or token of a superordinate aggregate or set. Or to put it conversely (and debatably): whatever we find designated as a wù in an ancient Chinese text must necessarily be considered to be an aggregate, an assembled object or a thing. Table 6.1 (coming now full circle, and in awareness of the hermeneutic trap) shows that the subjects tiān 天 (clause 6), dì 地 (clause 7), tǐ 體 (clause 8), wáng 王 (clause 10), and zhū-hóu 諸侯 (clause 11), being in the same position as the general term wù in clauses 1–5, qualify as aggregates. The feature of name-bearing is addressed in the following passage: Huán 6.5: [.  .  .], to take (a name for someone) from an (ordinary) aggregate-object is regarded as borrowing (the name), [. . .]. Therefore, as for (the names of) important aggregate-objects, it is not permitted to give them as names (to persons). [. . .] The Patriarch said: This (boy), when he was born, (he) and I shared the same (calendrical) set.20 Name him: “Same.” (Legge 50) [. . .], 取於物為假, [. . .]. 是以大物不可以命. [. . .] 公曰: 是其生也, 與吾同物. 命之曰同. “Objects” that should not be drawn on for names are listed in the context: “The name must not be taken from the name of the State; or of an office;

Getting to the Bottom of “Things” (wù 物) | 121 or of a mountain or river; or of any malady; or of a [domesticated or sacrificial, my explication–RHG] animal; or of a utensil, or of a ceremonial offering” (Legge 50). There follow a few examples illustrating that certain hills and offices lost their names because they were improperly appropriated for persons. All these “objects” are then, in the text quoted in the original, designated as “important” (wù), as important “aggregate-objects,” or “thing”! And one of the proper sources for names (see quote) was also (ordinary) “aggregate-objects.” In other words: most entities that we, based on our world-conception, would be inclined to understand as, and to unreflectingly call, “objects,” “things,” or “individuals” were for the ancient Chinese, in fact, “objects” in the sense of “aggregate-objects” or “composite things.” The concept of “thing” in ancient China was primarily that of a structured and ordered aggregate of parts—the parts at times themselves being aggregates at a lower level of a more complex aggregate. The fact that things representing a whole, and normally conceived as such, are actually multipart aggregates is evidently one of everyday experience—human beings have arms that have hands that have fingers that have nails.21 Surprisingly, referring with the word wù to “things” in the sense of an assembled whole not explicitly related to a superordinate set or aggregate, that is, referring to a—in our understanding—“individual object” is exceedingly rare in the Zuǒ Zhuàn— the only passage with two, in my opinion, clear and indisputable instances is the one just quoted from Huán 6.5! Two out of eighty-four, a bare 2.4 percent of the word’s occurrences—this imperatively calls for a change in our perspective, in our way of reading and translating the “remaining” 75.6 percent of its occurrences. This I shall attempt in section 3. In Zhāo 19.2 the noun phrase yào wù (藥物 ) occurred, reminding us that aggregates are the result of an assembly process, that is, that we should be on the lookout for instances in which the character 物 represents a verb or different verbs. The following interesting pair of examples shows the derivational transition from noun to verb.22 In the first passage, the expression “earth aggregates” or “aggregates (made of) earth” (tǔ wù 土物 ) is of the same construction type as the expression yào wù just mentioned. Earth moving and earthwork is a basic activity when building; earth aggregates would comprise structures in the guise of platforms, walls, terraces, and certain types of buildings—made of earth, of course. These structured aggregates have to be planned, surveyed, marked out, and finally built: Xuān 11 fù 1: Chancellor Ài-liè of the Wěi-lineage erected a fortified wall round Yí. (He) allotted materials and instruments, [.  .  .] set out the earth aggregates, [. . .]. The (compulsory) service was completed within three ten-day weeks, [. . .]. (Legge 310) 令尹蒍艾獵城沂. [. . .] 分財用, [. . .] 程土物, [. . .]. 事三旬而成, [. . .].

122  |  Robert H. Gassmann In the second passage, the expression tǔ wù “earth aggregates” appears in the reverse form, that is, “assembling earthen squares” (wù tǔ fāng 物土方 ). In this construction, wù writes a causative verb with the meaning “to make N into an aggregate” or “to make a structure out of N.” As I understand it, “forming earthen squares into a structured aggregate” signifies that the responsible builder—in a way well known from ancient city plans—formed structural units out of squares, that is, living quarters (sic!), markets, orthogonal networks of streets, and so on: Zhāo 32.5: [.  .  .] Mí-móu of the Shì-lineage surveyed Chéng-zhōu, [. . .] measured (the depth of) the moats and ditches, formed earthen squares into structured aggregates, [.  .  .]. (Legge 741) [. . .] 士彌牟營成周 [. . .] 仞溝洫, 物土方, [. . .]. In the Chūn Qiū, persons who commit robbery of estates are dealt with in two ways: if it is a case of insurrection and appropriation of the principality by a high ministerial, this is marked by using the dishonorable epithet “robber” (dào 盜 ) and by suppressing the name of the perpetrator; if it is a case of defection to another principality with estates, this leads to entries naming the perpetrators and the estates, but suppressing the epithet “robber.” Despite this difference, the two different forms of entry had one and the same purpose, that is, to punish and to inhibit others from following these examples. Of the four cases mentioned in the passage from which the following excerpt is taken, one belongs to the first category and three belong to the second. Grouped under the same purpose, they are different tokens (entries) of the same type, the type being designated as wù “set.” This word occurs in the following statement: Zhāo 31.6: [.  .  .] These two (types of entries) that form a set (my italics—RHG) are those in which (the Chūn Qiū) reproves lack of restraint and keeps its distance from greed. [.  .  .]. (Legge 738) [. . .] 此二物者所懲肆而去貪也. [. . .] An (uncritical reading of the) context could tempt us to translate cǐ èr wù zhě (此二物者 ) as “the two cases” (which Legge opts for), but the text does not read cǐ èr zhě (此二者 )—which would be the normal way of expressing this notion—but adds a wù (物 ). Even if the normal reading had been realized, supposing that the pronoun zhě was referring to “case,” it would not really be acceptable because the text is talking of four cases of one type of crime, that is, robbery of estates, but of two ways of phrasing the entries, that is,

Getting to the Bottom of “Things” (wù 物) | 123 shū (書 which is mentioned earlier in the passage). As the construction needs to be parsed in cǐ èr // wù zhě (此二 // 物者 ), that is, into the two modifiers cǐ èr and the head expression wù zhě, this forces us to analyze the latter part either as a relative clause or as a genitive construction. In the first case, wù would have to be analyzed as a verb, in the second as a noun. The topic of the passage being the form of entries, I would argue shū is the reference of zhě; and these shū entries form two types, but—due to the similarity of the underlying crime—attribution is to one and the same set. We therefore are justified in giving preference to reading wù as an (intransitive) verb with the meaning “to form an aggregate/set.” The derivational map of wù that can be safely extracted from the passages discussed up to now contains a noun (“aggregate/set” > “thing”), an intransitive verb (“to be/form/behave like an aggregate/set”) and a causative verb (“to make N into an aggregate” or “to make a structure out of N”). This profile is quite familiar and corresponds to that of nomina agentis, as the comparison in table 6.3 shows. Table 6.3

nomen agentis

nomen agentis

“nomen agentis”

noun wáng 王 chén 臣 wù 物 “king” “ministerial” “aggregate/set” > “thing” intransitive wáng 王 “to behave wáng “to behave verb like a king” like a king”

“to behave like an aggregate”

causative verb

wàng zhī 王之 wàng “to make “to make N a king” N a king”

“to make N into a set”

putative verb

wàng zhī 王之 “to consider N a king”

wàng “to consider N a king”

see Gōng-sūn Lóng 6 following

verbal noun (s)

the behaving  .  .  . the making  .  .  . the considering . . .

the behaving  .  .  . the making  .  .  . the considering . . .

see Gōng-sūn Lóng 6 following . . . 

At first, this comes as a bit of a surprise, but reflecting on the matter leads to the conclusion that wù as a class name for living beings, also, for example, for a noun like wáng (cf. clause 10), would “naturally” have these derivational qualities. Still, it takes some adjustment to “think” of an “aggregate” or “set” as “behaving like an aggregate (or set),” and there are some quite interesting lines of (speculative) thought (both ancient Chinese as

124  |  Robert H. Gassmann well as sinological) that issue from this. The social world (as defined by the well-known roles of the lún (倫 , i.e., from father to friend) was, and is—as we know—governed by the corresponding rules of behavior, but if now the larger world(s) of “beings” and “objects”—that is, the material world, as we would be apt to call it—were also governed by “rules of behavior”—if, therefore, all (aggregated) objects “act” according to appropriate rules, then we could not but assume that any aggregate with a name (míng 名 ) realizes (shí 實 ) its behavioral program. Gōng-sūn Lóng 6: Heaven and earth are the ones that grant those they produce in them [the quality of] aggregates. Aggregates are those that actualize (their aggregateness) by treating what they form into aggregates as aggregates, but not going beyond this. [. . .]23 [. . .] 天地與其所產焉物也. 物以物其所物而不過焉實也. [. . .] The actualization of a specific behavior, that is, of aggregateness, is not only theoretically conceivable, but would open up a new perspective for the explanation of disorder in the form of natural calamities and catastrophe: the sun not behaving as sun would produce untimely eclipses, the river not behaving as it should would cause inundations, the rain not falling at the right time would result in droughts, and so on—these would all be evidence of abnormal or wrong role behavior. And because of the all-pervasive system of correspondences, the behavior of things would tie up with and mirror the—orderly or disorderly—behavior of human beings. Looking at the material world through the lenses of behavioral patterns also would seem to fundamentally influence our understanding of ontology more sinico: A horse would not be a horse only because it can be physically described as such, but much more because it behaves like a horse. We are reminded of the notorious duck test: If it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck. In ancient China this duck would not probably be a duck, it would certainly be a duck! This also extends to material objects: Water in its normal aggregate state is liquid and clear (remember that this state is called xìng 性, the word also used for human beings); it “behaves” with a change of this state when subjected to cold (either in or out of season) or when it is choked with dirt. The material world, that is, heaven and earth (tiān dì 天地 ), would not be an aggregate of nonsentient physical substances, but of entities acting specific roles adapted to, or at times going against, their characteristic way of being. Dealing with such entities in a proper way means taking into account their physical and behavioral “programs”: being or becoming cups and bowls is not part of the

Getting to the Bottom of “Things” (wù 物) | 125 natural design of willows; only by mutilation can their wood be brought to behave in such a way (cf. Mèng 6A.01).

3. The Evidence of the Zuoˇ Zhuàn In the following sections, selected evidence from the Zuǒ Zhuàn (including the Chūn Qiū) shall be presented and, where necessary, commented on. I am restricting myself to a few significant examples per subtype, that is, to structured sets (wù; further relevant examples are appropriately mentioned in endnotes). The passages are assigned to the subtypes discussed earlier, that is, to sets (3.1), to sets of correspondent tokens (3.2), to dyads (3.3), and to assemblages of objects and random sets (3.4).

3.1 Sets “Sets’ are aggregates ordered according to a common feature—like a lèiclass, but restricted in number (e.g., the Three N). Sets comprising five tokens are best known: the five elements or agents, seasons, quarters of the empire, colors, organs, and so on. Only three out of eighty-four, 3.6 percent of the word’s occurrences in the Zuǒ Zhuàn, can be justifiably allotted to this category (the third being Zhāo 20 fù 4, Legge 684, with lèi and wù co-occurring). Although not modified by a numeral, the following occurrence clearly refers to sets. Comparing to the set of (four) seasons, this could be the set of the (five) agents (cf. Zhāo 32.6, clause 7 in table 6.1): Xuān 15.3: To rely on abilities and superiority in numbers is a guiding principle for doom. Zhòu of the Shāng followed it. As a consequence, he was extinguished. When Heaven reverses seasons, (this) produces calamities; when Earth reverses sets (of correspondences), (this) produces inauspicious (behavior); when the mín go back on their obligations, (this) produces disruptions of services. (Legge 328) [. . .] 夫: 恃才與眾, 亡之道也. 商紂由之. 故滅. 天反時, 為災; 地反物, 為 妖; 民反德, 為亂. [. . .] In the second passage, wù refers to the many diverse sets structuring the fabric of kinship, society, and administration (the vicinity of wù to these domains is remarkable and can be observed in many instances). It would not be imprecise to call these the “settings” of the empire:

126  |  Robert H. Gassmann Zhāo 9 fù 1: [.  .  .] When the (later) Wén(-King) was Elder (of the princes in the west of the empire), how could he have managed to change the (various) sets (of the empire)? He not only carried the son of Heaven as the (western) wing, but also did this with all due respect. [.  .  .] (Legge 625) [. . .] 文之伯也, 豈能改物? 翼戴天子, 而加之以共. [. . .]

3.2 Sets of correspondent tokens “Sets of correspondent tokens” are an aggregate (wù) of tokens coming from sets and being in correspondence (péi 陪 ) with each other (e.g., water from the five elements or agents, winter from the five seasons, the northern quarter from the five quarters of the empire, black from the five colors, liver from the five organs, etc.).24 There are examples for this type of aggregate, for instance, in the following passage, which broaches the subject of dragons. The crucial statement for understanding the meaning of wù here is the following: “Dragon is (correspondent member of) the water-set” (lóng shuǐ wù yě 龍水物也 ). Wù is referring to the set of tokens corresponding with water, the principal token of this set. In this set we therefore find assembled all tokens regarded as systematically corresponding to the element water. The official (guān 官 ) responsible for this set—named after the principal token (shuǐ wù 水物 )—is called “First officer of water” (shuǐ zhēng 水正 ). He was versed in the art of rearing dragons and of catering to their needs. But over the years—the passage tells us—this art was neglected (evidently due to bad government) and dragons gradually disappeared, which is explained as follows: Zhāo 29 fù 4: As for sets (of correspondent tokens), (such) sets have their (responsible) official. [.  .  .] If the official tends continuously to his work, (the tokens of) his set will then approach. If they neglect and abandon their work, the (tokens of the) set will then retreat and go into hiding. [.  .  .] The First officer of water [was called] Xuán-míng, [.  .  .]. Dragon is (correspondent token of) the water-set. Officials (responsible) for the water(-set) have already been relinquished. Therefore dragons are (no longer) born and obtained. [. . .] If dragons do not appear morning and evening, who is able to identify them as (correspondent tokens of) the water-set? [.  .  .] (Legge 731) 夫物，物有其官. [. . .] 官宿其業, 其物乃至. 若泯棄之, 物乃坻伏. [. . .] 水正曰玄冥, [. . .]. 龍水物也. 水官棄矣. 故龍不生得. [. . .] 若不朝夕見, 誰能物之? [.  .  .]

Getting to the Bottom of “Things” (wù 物) | 127 Related to medical aggregates, that is, “medicines,” are such compiled of corresponding and assorted foods and dishes, as in the following passage: Xī 30.8: If, in the case of a ruler of a principality, (his) exemplariness is something worth drawing attention to, and if (his) martiality is something that may be feared, then there takes place a banquet with full token sets (of corresponding dishes and foodstuffs), in order to thereby represent his capacity for obligating. The Five Flavors were presented. (Legge 217) [. . .] 國君, 文足昭也, 武可畏也, 則有備物之饗, 以象其德. 薦五味. [. . .]

3.3 Dyads The excerpt from Zhāo 32.6 (clauses 5 and 12 in the table) cited at the beginning of this essay mentions a special type of “aggregate” (wù 物 ): “complementaries” or—more specifically—“dyads” (èr 貳 ). As a significant example, the passage refers to the dyadic relationship between the princes of Lǔ and the lineage of the Jì, the first being in the position of ruler (jūn 君 ), the second in the position of ministerial (chén 臣 ). Understanding wù as “dyad” is not only clearly possible and at times highly illuminating in many passages from the Zuǒ Zhuàn, it is surprisingly so in a solid majority of the cases: a total of fifty-one out of eighty-four instances (60.7 percent) can be allotted to this subtype. This, in fact, dovetails neatly with the “quantity” implied by the expression wàn wù (萬物 ), which does not appear in the material presented here, but which in the contexts I have analyzed denotes “the ten-thousand dyads/relationships.” At second glance, this turns out to be less of a surprise if one takes the paramount importance of roles in ancient Chinese society into due account. In keeping with this observation, the high co-occurrence of wù and dè (“obligation” 德 ), the prerequisite for successful dyadic relationships, is to be remarked in many a passage.25 Let us turn to significant examples in the material: Zhāo 25.2: [.  .  .] Therefore [ritual conduct] produced (the roles of) ruler and ministerial as well as (the spheres of) high and low, and thereby they imitated the correct behavior of Earth; [ceremony] produced (the roles) of husband and wife as well as (the spheres of) inside and outside, and thereby made the two dyads into warp threads. [.  .  .] Therefore [the model kings] examined behavior and made commands trustworthy, used misfortune and good fortune, rewards and punishments to thereby regulate life and death. (They) let live because they judged aggregate-specific behavior as good; they let die because they judged aggregate-specific behavior as

128  |  Robert H. Gassmann contrary to the role. To judge aggregate-proper behavior as good is (a matter of) joy; to judge aggregate-proper behavior as bad is (a matter of) grief. (Legge 708) [. . .] 是故 [. . .] 為君臣上下以則地義; 為夫婦外內以經二物; [. . .] 是故審 行信令, 禍福賞罰以制死生. 生好物也; 死惡物也. 好物樂也; 惡物哀也. [. . .] In the first instance (jīng èr wù 經二物 ), we have a noun wù with the meaning “(dyadic) aggregate.” This clearly fits with the two major dyadic relations mentioned: ruler and ministerial, and husband and wife. In the following four instances, we have verb-object phrases (hào wù 好物 and è wù 惡物 ) where the object wù is a derived verbal noun (“to behave [in a way proper to] a [specific] aggregate” > “aggregate-specific behavior”). The connection with “behavior” is made manifest by the use of xìng (行 ) in the preceding context.26 The founders of dynasties established principalities. These they assigned to worthy personages, and they also staffed them with officials bound to the ruler in dyadic relationships:27 Wén 6 fù 2: [. . .] The founders of dynasties of antiquity knew of the finiteness of the mandate. Because of this they set up at their side role-coining and wise (personages), implanted customs and tunes in them, allotted them apanages and dyads, acquainted them with speeches and counsels, made for them the standard tubes and measures, [.  .  .]. (Legge 244) [. . .] 古之王者知命之不長. 是以並建聖哲, 樹之風聲, 分之采物, 著之話 言, 為之律度, [. . .]. Another case related to office appears in the verb-object expression wù guān (物官 ) Zhāo 14 fù 2: Dān of the Rán-lineage [.  .  .] gave responsibility to the proven, and established officers in dyads. [.  .  .] (Legge 655) [. . .] 然丹 [. . .] 任良, 物官. [. . .] If officers fail to fulfill their obligations, they step down. In the following case this is called “surrendering the dyad”: Zhāo 17.2: [.  .  .] When the sun has passed the equinox, but has not yet culminated, when (then) a disaster happens to (one of

Getting to the Bottom of “Things” (wù 物) | 129 the) Three Celestial Bodies, the Hundred Officers, in consequence of this, surrender (their) dyads. [.  .  .] (Legge 667) [. . .] 日過分而未至, 三辰有災, 於是乎百官降物. [. . .] An important dyad is that of male-female relationships in various combinations: Zhuāng 24.6: [. . .] Males make ceremonial gifts (to a new wife of the ruler), the important ones (being) gems and silks, the lesser ones (being) birds and animals, because these illustrate (their respective) dyads. (Legge 107). [. . .] 男贄, 大者玉帛, 小者禽鳥, 以章物也. [. . .] The males mentioned here are members of the same clan as the ruler of Lǔ, so the new wife not only becomes a member of this kinship group but also enters one or more new male-female relationships. Such relationships are clearly of the type “dyad.”28 Knowledge of the functioning and the state of dyads is important: Zhāo 1 fù 7: [.  .  .] The marquis of Jìn having been informed of the remarks squire-Chǎn had made (about a personage), said “He is a princely squire with vast knowledge of dyads.” (Legge 580)29 [. . .] 晉侯聞子產之言, 曰: 「博物君子也.」 Dyadic relationships are also important between the living and the dead. In the following excerpt we encounter two important markers: “(feeling of) proximity; relatives” (qìn 親 ) and “to make true, to realize” (shí 實 ). The first marker qìn is indicative of the dyad “father-son” (which is typical in cases of ancestorship), the second one is the verb shí denoting fulfillment of a (dyadic) role: Xī 5.9: [.  .  .] When the guǐ-spirits of the dead, the shén-spirits, (i.e.) non-humans, implement (their dyad with) relatives, only such who assume (their) obligations, on them do they rely. Hence in the Documents of the Zhōu [.  .  .] it is also said: “The mín do not take dyadic relations lightly; only such who assume (their) obligations make true (their) dyads.” (Legge 146)30

130  |  Robert H. Gassmann [. . .] 鬼神非人實親, 惟德是依. 故周書 [. . .] 又曰:『民不易物, 惟德繄 物.』[. . .] The marker qìn, “(feeling of) proximity,” also appears in the next excerpt, indicating that wù is to be interpreted within the framework of dyadic ­relationships between the presiding power and the minor members of an alliance: Zhāo 11.8: [.  .  .] We are unable to come to Chén’s rescue, and we are also unable to come to Cài’s rescue (against Chǔ). In the dyads (with them), we because of this have made the feeling of proximity disappear. (Legge 634) [. . .] 不能救陳, 又不能救蔡. 物以無親. The following, final, example relates to a passage in the Mèngzǐ, where in 1B.01 the ruler is encouraged to enjoy music with the mín, and in 1B.04 where the ruler asks the music master to compose a piece of music celebrating the harmony between ruler and ministerial, that is, in both cases within a dyad: Zhāo 21 fù 1: [.  .  .] If the less important (music) is not too inconspicuous, and the more important music not too overwhelming, then they create harmony in the dyads. If the dyads are harmonious, then the excellent is accomplished. [.  .  .] (Legge 687) [. . .] 小者不窕, 大者不摦, 則和於物. 物和, 則嘉成. [. . .]

3.4 Assemblages of objects and random sets Assemblages of objects and random sets are arrays of objects that do not, or are not meant to, fit into the generally accepted sets previously described. The characterization of such sets as “random” may be disputable due to a lack of knowledge about what was at the time regarded as random or rather as systematically corresponding (see 3.2). Such assemblages can be of two types: (a) wù can refer to an unexpected and unplanned for large or vast aggregate of individuals of a certain kind (e.g., vermin) that seem to behave according to a specific program, that is, a swarm of locusts (cf. Zhuāng 29.3); (b) wù can refer to a deliberate whole consisting of fitting parts, for example, the elements of a (military, ceremonial) outfit, a military formation,

Getting to the Bottom of “Things” (wù 物) | 131 administration system, the layout of architectural quarters, the structuring of a text, and so forth. In the Chūn Qiū, there is one example for (a): Autumn quarter: There were fěi-vermin. It was (an occurrence) that was regarded as a plague. In all (cases) of (great) aggregates (of vermin recorded in the Chūn Qiū the following rule) applies: If they were not regarded as a plague, they were not recorded. (Legge 116) Zhuāng 29.3: 秋: 有蜚. 為災也. 凡物: 不為災, 不書. Examples for (b): Dìng 10.7: The buff-coats of the Shú-sūn-lineage have their (specific) aggregate. [.  .  .] (Legge 778) [. . .] 叔孫氏之甲有物. [. . .] In the next passage tracts of land are assembled: Chéng 2.4: [.  .  .] If you command the lords to neglect filial duties, and if they let (these) vanish, are they then (members) of the category of those who take obligation-bound (dyadic) behavior to be wrong? The ancestor-kings defined the boundaries in the empire and partitioned it, they formed suitable (tracts) of land into aggregates and proclaimed their usefulness. (Legge 346) [. . .] 若以不孝令於諸侯, 其無, 乃非德類也乎? 先王疆理天下, 物土之宜, 而布其利. [. . .] The following passage belongs to the field of prognostication. It informs us that the ruler had to take note of the aggregates formed by clouds at these times of the year, presumably also the way they started in one aggregate and developed into another, thus indicating the course of future events: Xī 5 fù 1: [. . .] At equinoxes, solstices, commencements, and closings of (calendrical) quarters, (the ruler) had to record the aggregates of clouds. This was for reasons of making preparations (for events thus announcing themselves). (Legge 144) [. . .] 凡分、至、啟、閉, 必書雲物. 為備故也.

132  |  Robert H. Gassmann Gifts presented by both sides on the occasion of missions could also constitute aggregates: Xuān 14 fù: [.  .  .] Your ministerial has learned how small principalities save (themselves) from large principalities. (You must) go on a friendly mission of enquiry and present as gift an (appropriate) aggregate. (Legge 324)31 [. . .] 臣聞小國之免於大國也. 聘而獻物. [. . .] In a similar way, the signs of excessive government, that is, speaking stones, discontent among the mín, or ministerials who had hitherto kept silent, form a random set of tokens: Zhāo 8 fù 1: [.  .  .] The marquis of Jìn enquired (the following) of music master Kuàng and said: “Why should a stone speak?” He answered: “Stones are not capable of speaking. Somebody is availing himself of it. If not, then it is because the mín are hearing (the sounds of) excessiveness. But, being compelled (to speak), your ministerial has also heard them saying: ‘When the launching of (compulsory) services is untimely, and resentment and discontent stir among the mín, then it happens that (things) voicing criticism form sets (of corresponding tokens) and speak.’ ” (Legge 622) [. . .] 晉侯問於師曠曰:「石何故言?」對曰:「石不能言. 或馮焉. 不然, 民聽 濫也. 抑臣又聞之曰:『作事不時, 怨讟動于民, 則有非言之物而言.』[. . .]

4. Pointing to Things as They Are Now— And Other Things Being Equal Chapter 3 of this paper may be regarded as a rather tedious and taxing thingamabob. But to my mind it is unavoidable and necessary. Only by sifting and presenting enough of the evidence available, by meticulously translating corresponding passages, by analyzing their context, and by attempting to set up the necessary specifics of the meaning of such an important word does one avoid jumping to premature conclusions and overlooking opportunities for relevant insights. We simply cannot blindly rely on our usual linguistic reflexes which instinctively react to deceptively similar thoughts and the bias of which is again and again reaffirmed by an established (and mostly unquestioned) tradition of translation. I am sure that Angus C. Graham would

Getting to the Bottom of “Things” (wù 物) | 133 appreciate the critical vein of this paper and bear with a (duly respectful) disputer of “his Tao.”

Notes   1. A. C. Graham, Later Mohist Logic, Ethics, and Science (Hong Kong: The Chinese University Press, 1978), 210. For the sake of completeness, the importance of the word in other major studies and translations by Graham should be mentioned: “Kung-sun Lung’s Essay on Meaning and Things,” Journal of Oriental Studies 2.2 (1955): 282–301 (Zhǐ Wù Lùn 指物論); and “Chuang-tzu’s Essay on Seeing Things as Equal,” History of Religions 9 (1969–1970): 137–159 (Qí Wù Lùn 齊物論). I wish to thank Carine Defoort for insightfully pointing out several instances in my essay needing clarification.   2. Bernard Karlgren, Grammata Serica recensa (Stockholm: The Museum of Far Eastern Antiquities, 1972).  3. Ulrich Unger, Glossar des Klassischen Chinesisch (Wiesbaden: Otto Harrassowitz, 1989).   4. James Legge, The Chinese Classics, vol. 5, The Ch’un Ts’ëw with the Tso Chuen (Hong Kong: Hong Kong University Press, 1960, rpt.).   5. Sinolinguists could also assume a bad case of homographs, that is, the writing of different words with the same character. But this avenue is usually not very promising.   6. This, again, was originally fueled by the need to reach a clearer understanding of instances of wù in the Mèngzǐ. See my study and new translation Menzius: Eine kritische Rekonstruktion mit kommentierter Neuübersetzung (Mencius: A critical reconstruction and new translation with commentary) (Berlin: De Gruyter Mouton, 2016).   7. The verb shēng 生 “to give birth to” is transitive. In the passive construction wù shēng 物生 the agent is not named. Due to the examples that follow, we can assume “Heaven” in this role (not in the sense of “sky,” but in the sense of “ruler over fate” or “creator”).   8. Note that the text itself makes a caesura between these two subtypes: clause 9 concludes the first subtype, clause 12 the second.   9. For a comment on what kind of subtype we find in péi 陪, see later discussion and section 3.3. 10. In German I would suggest translating wù as “Ensemble,” which in Gerhard Wahrig, Deutsches Wörterbuch, means: “Gesamtheit; [. . .]; wirkungsvoll zusammengestellte Gruppe (von Gegenständen, z.B. Häusern) totality; [.  .  .]; grouping (e.g of things, houses) meant to achieve a certain effect.” Gerhard Wahrig, Deutsches Wörterbuch (München/Zürich: Bertelsmann/Ex Libris, 1986/1989). 11. I would reject the assumption that the object expressions are the “orphaned” adjectives of phrases reconstructed as liǎng [wǔ], sān [wǔ], and so on. This would mean counting aggregates, as is shown later in the chapter. 12. The references of gè 各 are tiān 天, dì 地, tǐ 體. The object expressions are not included. 13. It is rather surprising to note that Anna Ghiglione, La pensée chinoise ancienne et l’abstraction (Paris: Éditions You-Feng, 1999) has no special treatment for this

134  |  Robert H. Gassmann important term. Neither is there an entry for lèi in Zhang Dainian, Key Concepts in Chinese Philosophy (New Haven/Beijing: Yale University Press/Foreign Language Press, 2002). What’s more: wù is also ignored as a possible troublemaker. 14. The five grades of lords, that is, duke (gōng 公), marquis (hóu 侯), count (bó 伯), baron (zǐ 子), and viscount (nán 男) (the translations are conventional), are, on the one hand, tokens of the set of hóu 侯 “lord” (cf. zhū-hóu 諸侯 “all lords”); each one of them, on the other hand, is member of a dyad with his respective complementaries (èr 貳), that is, his ministerials or “mates.” 15. Cf. section 3.3. 16. Legge renders this expression as “six things.” Even admitting that “thing” is a highly abstract and general term, I find it rather difficult to conceive of months and days as things, objects or even entities. 17. I wonder whether the fact that many abstract terms, such as, for example, “animal,” which are constructed not as single words but as binomial expressions, that is, “aggregated” (e.g., qín-shòu 禽獸 “birds + land-borne beasts” = “animals), could be identified as the underlying paradigm. This would shed light on the relationship between wù 物 as dà gǒng míng 大共名 and qín-shòu as dà bié míng 大別名 in Xúnzǐ 22. 18. Medicines, we should note here, were not random aggregates. They were mixed with regard to correspondences, so that the substances assembled corresponded to elements conforming, usually, to the system of, for instance, Five Agents. 19. Analyzing the expression yào wù 藥物 as “aggregate(s) of medical substances,” that is, as a normal genitive with the nucleus wù and the modification yào, is based on the observation that the form “the assembling of medical substances,” that is, with a verbal noun and its object, would normally be in the reverse sequence wù yào 物藥. 20. Wù is here referring to one (or a combination) of the liù wù “Six Sets” (cf. Zhāo 7.7). It could be season, month, or day. 21. This, I would strongly argue, is not the same concept that led the Greek to speculate about the point where something becomes atomos, that is, indivisible. This type of division happens without regard for the internal structuring of the whole: it, so to speak, halves and halves again, until it can no longer be halved. The ancient Chinese concept of aggregate is probably best illustrated by the story of Cook Dīng 丁, who dismembers an ox by following the natural borders and cavities without injuring the parts constituting the ensemble (Zhuāngzǐ 3; Graham 63–64). 22. The direction of the derivation is important to note, because it indicates whether the source word is a noun, from which verbs are derived, or a verb, from which nouns are derived. This aspect will be expanded on later. 23. Gōngsūn Lóng 公孫龍; Míng shí lùn 名實論; cf. Graham, “Kung-sun Lung’s Essay on Meanings and Things.” In Zhuāngzǐ 19, wù are defined as having manner and voice (fán yǒu mào xiàng shēng sè zhě jiē wù yě 凡有貌象聲色者皆物也 “whatever has manners, shape, voice, and color is in every case a wù”)—another clear indication of a biomorphic and “behavior-based” view of the workings of physical nature. 24. Further interesting examples and contexts are to be found in Zhuāng 32 fù (Legge 120); Dìng 4.2 (Legge 754, with the expression bèi wù 備物 “full sets of corresponding tokens”; and Zhāo 1 fù 8 (Legge 580–581, two corresponding situations are here designated as tokens of the same set).

Getting to the Bottom of “Things” (wù 物) | 135 25. Interestingly, Graham implicitly recognizes the dyadic background in his definition when he notes, “Wu may refer to such abstract entities as love and benefit (EC 2). Similarly, in older chapters of Mo-tzŭ we find objectors dismissing universal love as an ‘impracticable thing’ (ch. 15, Sun 70/5 不可行之物) [. . .].” The emotions named here are all characteristic of certain dyadic relationships. 26. Further interesting examples and contexts with the key word, that is, the verbal noun “(role-informed) aggregate-specific behavior,” deriving from “to behave [in a way proper to] a [specific] aggregate,” are to be found in Yīn 5.1 (Legge 18–19), Xiāng 9.3 (Legge 440), Zhāo 25.8 (Legge 711), and Zhāo 26.7 (Legge 718). 27. There are many passages in which wù refers to the important dyadic aggregate of the ruler and his officers: Huán 2.6 (Legge 40, dyadic aggregates accompanying offices to which officials could ascend to or be deposed from), Xuān 15.6 (Legge 329), Xiāng 3 fù 2 (Legge 420, by establishing a single officer accomplishing three dyads), Zhāo 7 fù 6 (Legge 618, the use of dyads for maintaining power), Zhāo 9 fù 5 (Legge 626), Āi 1 fù 1 (Legge 794, restoration of dyads of old jiù wù 舊物). In Xuān 3.5 (Legge 293) we find the designation “Hundred (dyadic) Assemblages” bǎi wù 百物 referring to the many types of dyadic aggregates of officials, that is, to ruler-official relationships; the bǎi guān 百官, “Hundred Officials” in Xuān 12.3 (Legge 317) is an alternative designation for these types of dyadic relationships. In Dìng 4.2 (Legge 754) the allotment of dyads fèn wù 分物 by the king on the occasion of enfeoffment is mentioned, cf. also Xī 7.4 (Legge 149); Dìng 1.2 (Legge 744) mentions differing social aggregates/dyads in the Three Dynasties. 28. Further examples referring to the male-female dyad: Zhāo 1 fù 8 (Legge 580–581, excessive sexual intercourse disrupting the male-female dyad), Zhāo 28 fù 1 (Legge 726, ruin of the Three Dynasties because of corrupted male-female dyads). 29. A further avenue of prognosticating the state of a dyad is by listening to the words spoken by their members: “Words let one perceive (the quality of a) dyad” (Zhāo 1.2, Legge 576). 30. That wù and dè “obligation” co-occur is not accidental. Diligently serving one’s obligations within the framework of a dyadic relationship is a prerequisite, if not an outright guarantee, for a successful life; to violate such obligations is to invite disaster, cf. Xī 15.14 (Legge 169). 31. In Xiāng 31 fù 3 (Legge 564) officials present aggregates to arriving guests.

7

Míng (名) as “Names” Rather than “Words” Disabled Bodies Speaking without Acting in Early Chinese Texts

Jane Geaney

It is impossible by any convention, to make a blind man understand what is meant by the words red, green, or blue. —Antoine Arnauld and Pierre Nicole, The Port-Royal Logic

Introduction My first scholarly article was about the work of A. C. Graham.1 Unfortunately, I never met him, but my copies of his books became so worn from overuse that I had to replace them.2 My second, now equally worn, copy of Later Mohist Logic, Ethics, and Science opens to a statement that inspires my work: “A consistent nominalism has to extend its principle to the particular utterances of the name itself; I pronounce the sound ‘stone’ over X and afterwards convey that Y is like X by pronouncing a similar sound.”3 This claim has two important implications. First, in early Chinese texts the feature that makes one name (míng 名) the same as another is not an abstraction. Second, what makes one míng the same as another is a matter of pronouncing similar sounds. In this essay, I explore these two implications by analyzing illustrations of failures to name in early Chinese texts.

137

138 | Jane Geaney As Graham asserts, a míng consists in a sound pronounced over a thing. I have developed this claim by adding that míng and its frequent c­ omplement, shí (實 “action-thing”), constitute an aural-visual polar balance.4 In light of the patterns of aural/visual pairing in early Chinese texts, the fact that míng are audible while shí are visible suggests a yin/yang-type relationship, wherein a míng is not a shí and a shí is not a míng, but they unite.5 This model departs in several ways from standard conceptions of míng/shí implied in translations like “names and objects,” “words and things,” and “language and reality.” Polarity is not necessarily characteristic of the relationships of names to objects, words to things, or language to reality. The nature of the relations within these pairs is open to interpretation. For instance, while names could be nonobjects, they might be a subset of objects. We might think of words as things, but we also might think of (ideal) words and (material) things as mutually exclusive. With regard to language and reality, we might conceive of each side as excluding the other, or we might posit that one is a portion of the other. Thus, the extent to which my claim deviates from traditional interpretations of míng/shí depends on how we conceptualize the relations within these pairs. Models of language also establish paradigms of the things to be named. My proposed model differs from the standard translation in its prototypical “named-things.” In ancient Greek texts, from which the standard translation equivalents ultimately derive, nomos is used to mean both “names” and “nouns.” Accordingly, the prototype of what is named is a substance. In early Chinese texts, a prototypical shí might be a fruit, with associated intimations of becoming full and useful. Early Chinese texts also use shí to mean “dividend” and “capacity.”6 The unlikelihood of being both a substance and a quantity to be divided justifies considering the possibility that the standard translation equivalents might limit our understanding of uses of the term shí. This essay extends my prior arguments by focusing on passages that discuss the utterances of blind and lame people in early Chinese texts, from which I infer that these disabilities epitomize the state of being only partly capable of using míng, insofar as a complete use of míng requires visibly acting, not just speaking. I contrast early Chinese texts’ treatments of visual or visible disabilities (blindness and lameness) with those of taste impairment in order to advance my theory that the texts assume that naming (as something audible) should be complemented by action (as something visible). In other words, a míng is not like a word-type or its tokens, the correct use of which might simply require properly linking them together. As Graham says, míng are like the sound “stone” pronounced over X and Y, to which I add that in the view of early Chinese texts, the correct sounds should be matched by visible acts such as picking a stone.

Míng (名) as “Names” Rather than “Words”  | 139

Background: Detached versus Immersed Views of Language To appreciate the first aspect of Graham’s radical idea, it helps to stipulate an explicit contrast between “names” and “words.” In this contrast, names are denotative, while words are connotative. In other words, names “refer,” whereas words have “senses” or word-meanings.7 The point of this contrast is not to say that anything that can be used as a name is necessarily different from something that can be used as a word, but to clarify why there are two terms at all. As terms whose function is identification, names are inherently grounded in material contexts. By contrast, for example, a word like “not” can be rooted in relations among words, rather than spatiotemporal things or events. We can also differentiate names from words by examining the idea of a dictionary.8 Insofar as names simply “refer,” they are rarely included in dictionaries because they have no word-meanings to be listed there. Word types and their word-meanings are the kind of things included in dictionaries. Names are incorporated into dictionaries only if they are used as words. For instance, dictionaries do not contain “Smith,” the name. The “smith” that appears in dictionaries reflects a usage like “a person who works with metal.” From this perspective, we can see why Graham implies that early Chinese texts do not use míng to mean something like “word” as a type-token concept. Because it is an abstraction, that sort of concept does not conform to the nominalism he posits for Early China.9 Graham’s consistent nominalism implies that “stone” has the metaphysical status of a name, not a word type. The second significant implication of Graham’s statement concerns thinking of míng as sounds, not written “sinographs” or an abstraction that might appear in either medium. The absence of theorizing about the nature of abstract linguistic elements in Early China (which, of course, is not to say all abstract theorizing is absent) is consistent with the presence of a different kind of attention to language, and much can be inferred from the speech situations in which early Chinese texts present míng. I focus here on the fact that, of all the possible ways to illustrate not knowing a míng, texts from Early China often tell stories about physical disabilities, most notably, blind people speaking about color. Thus, to explore early Chinese concepts of míng, I look to something we seldom take into account when thinking about language: the context of utterances and the bodies that make them. In considering such bodies, I pay scant attention to the texts’ own purposes. Instead, I start with the texts’ overt normative goals, only to bracket them in order to foreground the semantic and epistemological implications in marginal illustrations.10 In this way, I argue that, while the trope of blind people’s lack of color knowledge might be familiar from the dominant

140 | Jane Geaney Western philosophical tradition, in early Chinese texts, the implications for the underlying conception of language are different.11 This analysis of early Chinese ideas about míng draws upon scholarship that contrasts “detached” from “immersed” approaches to language.12 Detached approaches are those that involve, for example, conceptions of words having determinate meanings in isolation from their use in the context of an utterance. Detachment of that sort is implicit in metaphors about words having or transmitting word-meanings. That is, a statement like “words have core meanings” implicitly posits word-meanings as abstractions belonging to abstractions (word types). This common usage is typical of formalist theories of language (such as those of Ferdinand de Saussure and Noam Chomsky), which downplay the importance of fleeting individual moments of language use. By contrast, immersed theories emphasize how material contexts and acts of use produce the meanings of utterances (and their components).13 They identify the meaning of utterances by analysis of the situational factors related to it. In the former case, meaning tends to be found in sentences, propositions, and words; in the latter, it lies in contextualized actions (utterances). As an immersed approach to language, early Chinese texts use the term míng in ways that are different from technical uses of “word,” while resembling ordinary uses of “name.” To clarify the nature of uses of míng in early Chinese texts, as previously noted, I use “name” to mean something that designates, and in particular, something that designates things or actions rather than linguistic abstractions. By “word,” I mean a unit whose identity is determined by some form of relation (possibly function or size) to other units in a language. This technical-sounding definition of “word” is currently part of colloquial usage for many speakers. A less technical use of “word” would be something like, “May I have a word with you?”14 As I use the term “name” here, a name is one way to talk about bits of speech—one that emphasizes not intra-speech function or unithood (as “word” might), but the designation of something external to speech.15 Early Chinese texts use míng to mean “personal name,” in addition to using it to mean “fame” (as in “making a name”). In many ways, the uses of míng in early Chinese texts correspond to colloquial English uses of the term “name.” Míng function as labels of things or actions. In passages that call attention to míng labeling, the items that are labeled by míng are neither other bits of language nor bits of thought. Paradigmatically, míng point “outside” rather than toward other míng.16 That is, the primary functions that early Chinese texts ascribe to míng do not involve either what one might call “intra-míng” relations or relations to conceptual items or thoughts.17 Míng select entities or actions. This includes things like “Yao’s dutifulness,” which we might think of as a concept but can be understood as a description of his actions. Míng also

Míng (名) as “Names” Rather than “Words”  | 141 appear to be more like names than words insofar as míng might not be the “building blocks” of speech. Passages that supply examples of míng never choose such words as modal particles, auxiliary words, or “logical words”— like “not” (fēi 非) or “thus” (gù 故).18

Prefatory Remarks The passages I discuss here come from a variety of early Chinese sources dating to different time periods. Some of the passages are from the synoptic Mòzǐ (fourth century BCE), one is from the “Later Mohist Canons” (third century BCE?), one is from the Huáinánzǐ (second century BCE), and one is from the Yántiělùn (first century BCE). It is likely that the later passages rework the stories from the earlier ones, but it is not an easy matter to determine the dates of the passages.19 In any case, they feature a common thread that they refashion and reuse. I take the thread that unites these different uses to justify discussing them as being, in some sense, the same sort of passages, worthy of analysis in light of one another. The recycling of the passages suggests that they are not isolated cases, but rather they represent a familiar way of approaching a particular problem. The passages sometimes use the term yán 言 (speech) rather than míng, but several passages make it clear that naming is still at issue. For instance, the section from the Mòzǐ “Guì Yì” (discussed later) makes the fact that it is about naming explicit by mentioning a distinction between naming and picking (qǔ 取). Moreover, the passage from the Mòzǐ’s “Fēi Gōng Xià” chapter (discussed later) explains that the example of black and white is meant to illustrate the use of a particular name in relation to the actions (shí 實) to which it refers. There are challenges to establishing that early Chinese texts treat the term míng like “name” rather than like “word,” not the least of which is the fact that the distinction itself is obscure. The most obvious evidence is the abundance of uses of míng to mean a sound, as well as the almost nonexistent uses to mean a written word, which I have discussed in other publications.20 Here I will consider the rare passages where we can draw out something like the notion of a difference between a name and a word. These passages, which concern disabled people naming, suggest the unusual idea that it is possible to know how to use míng independent of knowing the things to which they refer. They evoke a distinction between míng and the physical actions that demonstrate the presence of knowledge of míng. In this way, the passages do something rather extraordinary for early Chinese texts. While it might not strike contemporary readers as unusual to talk about knowing the meaning of words independently of knowing how to act on them,

142 | Jane Geaney that kind of claim almost never occurs in relation to míng in early Chinese texts. Of course, early Chinese texts frequently contrast míng to action, but they rarely approach the subject of míng by attending to the correctness of a míng used independently of action. I will argue that these passages where the term míng is used similarly to “word” constitute additional evidence for my argument for two reasons: they highlight the referential use of míng and they do so specifically by characterizing míng as an audible phenomenon that should be paired with visible behavior. Some of these passages I examine here extract míng from the context of their use by contrasting the correctness of míng as uttered to the actions that should go with míng. They imply that certain acts of naming or speaking are correct even though they necessarily fail to refer.21 For the blind people who are asked to pick colors by name, a crucial piece of the context (the visible element) is lacking. In this way, the example of blind speakers draws attention to the possibility of thinking of the relationship of names (or naming utterances) to one another. In other words, because these uses of míng inevitably fail to connect to their referents, they end up being correct only insofar as they connect to one another. To that extent, we might say that the passages briefly entertain something like the idea of treating knowing how míng relate to one another as a form of knowing míng. They leave this knowledge unnamed, but we might call it quasi-knowledge, undeveloped knowledge, or “correct but inactive knowledge” of míng. The important point is that the texts implicitly acknowledge that blind speakers’ correct color naming is founded in knowing míng in relation to one another, and yet they do not pursue this way of thinking about míng.22

Blind Naming Passages Early Chinese texts criticize people who are corrupt or inefficient by means of blindness analogies—one of several tropes of physical disability. The contexts are not explicit, but passages that use blind people naming color as their sole illustrative example seem to concern problems like advocating ethical behavior, but not acting on it. Readers might be sympathetic to people who struggle to discern the difference between ethical and unethical behavior. Indeed, we might even feel grateful when people manage to say the right things about something like duty, even if they do nothing about it. But these tropes argue for the opposite: that is, at least blind people have an excuse. In other words, these disability analogies serve to condemn failures to act when one has the ability to do so. The assumption is that a person with a moral faculty can differentiate dutiful from undutiful behavior as easily as a sighted person can discern black from white.23

Míng (名) as “Names” Rather than “Words”  | 143 Like a ruler who speaks too lightly of duty, the targets of this rhetoric are people who are fully able to see what names refer to and perfectly capable of acting on them. Rhetorically, the disability has the force of making the able-bodied target of criticism look even worse. The point is that, despite their abilities, these people act like blind musicians who have been asked to name colors. Here is one example: If you ask blind musicians, “What is plain white like?” they will say, “It is like unbleached silk.” If you ask them, “What is black like?” they will say, “It is like dark.” But if you take white and black to show them, then they cannot locate (chǔ 處) them. That by which people look at white and black is the eye, and that by which they speak of black and white is the mouth. The blind musicians have the means to speak of white and black, but not the means to know (zhī 知) white and black. Thus, in speaking of white and black, they are the same as other people, but in separating/discerning (bié 別)24 white and black, they are different from other people. 問瞽師曰：「白素何如？」曰：「縞然。」曰：「黑何若？」曰：「黮然。 」援白黑而示之，則不處焉。人之視白黑以目，言白黑以口 ，瞽師有以言 白黑，無以知白黑，故言白黑與人同，其別白黑與人異。

(Huáinánzǐ 9 “Zhǔ Shù Xùn” 《主術訓》) The blind people in these illustrations are capable of using names, but incapable of producing the actions that should accompany those sounds, because their eyes cannot see what action to take. Thus, the passages contrast the blind people’s ability to speak about colors with their inability to identify the referents.25 In these illustrations, what the mouth does is speak, while the eye is supposed to guide the hand in the act of something that the texts call locating (chǔ 處), dividing (fēn 分), choosing (qǔ 取), distinguishing (biàn 辯), or separating (bié 別). Speakers must act in some visible way in relation to the referent about which they speak: whether by pointing, picking, embodying or something else. Hence, for example, a ruler’s failure to act on his talk of duty might be a matter of failing to do something comparable to separating. He should use his eyes to locate (chǔ 處), divide (fēn 分), choose (qǔ 取), distinguish (biàn 辯), or separate (bié 別) dutiful behavior and use his body to act on it.26 That action might involve something like rewarding dutiful behavior or even behaving dutifully himself. Thus, the blind analogies might, for example, make the case that a ruler is faced with the task of showing that he can discriminate dutiful from undutiful behavior, just as the

144 | Jane Geaney blind people are faced with the task of showing that they can discriminate black from white colors. The ruler, however, has the advantage of being able to see; hence, there is no excuse for his failures. The interesting feature of these passages is not limited, however, to the way they illustrate people failing to act. In fact, they specifically target the combination of being unable to act while simultaneously being able to talk about things. Without denying the correctness of blind people’s color naming—indeed precisely in light of that correctness—the passages suggest that it has no value: This is like blind people naming “white” and “black” names like other people, but not being able to divide (fēn 分) the things (wù 物). How can this be called discernment (bié 別)?27 此譬猶盲者之與人同命白黑之名，而不能分其物也，則豈謂有別哉。

(Mòzǐ 5.3 “Fēi Gōng Xià” 19 《非攻下第十九》) Thus, blind people engage in a two-sided process. On the one hand, they name or speak, and on the other, they should divide (fēn 分) the things to which those names point. Only success in both constitutes successful “separation” or discernment (bié 別). Hence, the function of the blind people in such passages is not merely to illustrate an inability to act. It specifically illustrates the failure to act on the correct knowledge of names. In this way, the analogy criticizes people who do not live by what they say: that is, those who behave as if they are unable to put into practice their knowledge of names. It is worth noting that, in one sense, there is nothing exceptional about these stories: the analogies to blind people speaking about color are consonant with a general keen interest in early Chinese texts in the idea of strategically employing speech and names. As the Lúnyǔ puts it, “If a jūnzǐ (gentleman) names it, it should necessarily be spoken. If a jūnzǐ speaks it, it should necessarily be done” (Lúnyǔ 13/3).28 Or, as the Mòzǐ says, “If the speech is sufficient to promote action, make it your standard. If the speech is not sufficient to promote action, do not make it your standard” (Mòzǐ “Guì Yì”). We see this practical interest when the Huáinánzǐ’s discussion of a blind person’s failure concludes that people need to exhibit things in action (xíng 行): It is dutiful to be filial toward one’s parents when inside, and loyal to one’s ruler when outside. Foolish or wise, worthy or not, everyone knows that. But rare are the people who exhibit loyal and filial actions or know the means to make them emerge.

Míng (名) as “Names” Rather than “Words”  | 145 入孝於親，出忠於君，無愚智賢不肖皆知其為義也，使陳忠孝行而知所出 者鮮矣。

(Huáinánzǐ 9 “Zhǔ Shù Xùn” 《主術訓》) In a similar way, the passage from the Mòzǐ “Fēi Gōng Xià” chapter (cited earlier) proceeds to make a point about the necessity for action: Hence, the ancient wise people, in acting as the standard for the world, certainly compliantly considered their duty and only then acted on it. 是故古之知者之為天下度也，必順慮其義而後為之行。

(Mòzǐ 5.3 “Fēi Gōng Xià” 19《非攻下第十九》) So, too, the Mòzǐ “Guì Yì” passage implies that the blind people pass the test of using names correctly, but they are not able to perform an action—the act of choosing. If you mix white and black and cause a blind person to choose, they are not able to know. Therefore, when I say that the blind person does not know white and black, this is not because of their naming, but because of their choosing (qǔ 取). 兼白黑，使瞽取焉，不能知也。故我曰瞽不知白（墨）〔黑〕者，非以其 名也，以其取也。

(Mòzǐ 12.1 “Guì Yì” 47《貴義第四十七》) This might imply that a blind person does know the names “white” and “black” and yet does not know how to use the names for referring. Arguably, in the Yántiělùn there is a different variation on this theme of blind people’s inability to act on their color naming. It compares the speech of the Ru (“Confucians”) to that of blind people, because they both sit and talk, without getting up and acting on what they say. The blind can speak of “white” and “black,” but they do not have the eyes to separate/discern (bié 別) them. The mouths of the Ru can speak about “order” and “chaos,” but they do not have the ability to act on them. Now, when seated talk is not acted on, then shepherd boys are joined with the strength of Wu Huo, and

146 | Jane Geaney the long-beards possess the potencies of Yao and Shun. Therefore, if they were caused to speak about the near at hand, how could the Ru dither about order and chaos and how could the blind dither about white and black? If you do not speak things aloud, then the shame of not embodying (gōng 躬) them will not catch up with you. Thus, cases where the lowly speak of the elevated and where there is an ability to speak but no ability to act: these are the sorts of things of which the gentleman would be ashamed. 盲者口能言白黑，而無目以別之。儒者口能言治亂，無能以行之。夫坐言不 行，則牧童兼烏獲之力，（逢須）〔蓬頭〕苞堯、舜之德。故使言而近， 則儒者何患於治亂，而盲人何患於白黑哉？言之不出，恥（窮）〔躬〕之 不逮。故卑而言高，能言而不能行者，君子恥之矣。

(Yántiělùn 7 “Néng Yán” 40《能言第四十》) In other words, it is inadequate to sit comfortably and chat about things that one cannot put into action upon standing up. The Ru in the Yántiělùn passage responds by defending himself against this accusation with the claim that he has abilities to act, as well as to speak.29 The blind, the Ru, and those who are effectively lame (the young shepherds and old long-beards) all have something in common: they can speak, but they cannot act on the things they speak about. Thus, these passages use the example of disabled people’s naming in order to criticize the failure to act on it. This emphasis on putting speech into action, or “walking the talk,” is characteristic of the kind of interest in language often found in early Chinese texts.

Criteria for the Correctness of a Name To some extent then, these stories are typical, but their choice of a blind person as a trope is not random. There are good reasons why the texts often make this point using examples of the color-naming of blind people. Those reasons reflect the blindness of the person doing the naming and the nature of the item that the eyes identify. Early Chinese texts presume that blind people are not only capable, but even proficient, in their sense of hearing.30 Hence, they will understand what they are asked to do. Moreover, color is strictly perceived by the eyes. Texts from Early China typically characterize the things that the eyes see as colors (sè 色), shapes (xíng 形), patterns (lǐ 理), bodies (tǐ 體), and walking/action (xíng 行).31 Among them, only colors cannot also be perceived by the sense of touch. Thus, the unknown item is a color, which rules out assistance from other sensory faculties.

Míng (名) as “Names” Rather than “Words”  | 147 Setting aside for the moment the ultimate function of these analogies, we can see that the passages credit the blind people with being capable of naming color correctly. That is, it is safe to assume that the blind people’s responses represent some level of correctness, because the Mòzǐ adds that the clear-sighted would not speak differently: Now, blind people say, “That which is light32 is white. That which is dark is black.” Even the clear-sighted (míngmùzhě 明目 者)33 would have no basis to alter that. But [if] white and black are mixed together and [you] make the blind person select (qǔ 取) them, they are not able to know (zhī 知) [which is which]. 今瞽曰：「鉅者，白也。黔者，（墨）〔黑〕也。」雖明目者無以易之。 兼白黑，使瞽取焉，不能知也。

(Mòzǐ 12.1 “Guì Yì” 47《貴義第四十七》) The passage affirms the correctness of the sounds produced by the blind people, but at the same time, the choice of using blind people to speak about color makes it inevitable that the correctness of their utterances goes no further than equating sounds. What are the grounds for assuming the correctness of these sounds at all? The passages propose two things that render these utterances correct. First, the blind people’s utterances employ names in the same way as others do. The Mòzǐ “Fēi Gōng Xià” says, “This is like blind people naming ‘white’ and ‘black’ names like other people.” And the Huáinánzǐ says, “Thus in speaking of white and black, they are the same as other people, but in separating/discerning (bié 別) them, they are different from other people.” Second, the clear-sighted (míngmùzhě 明目者) would endorse this naming. They would not change how the blind people use names in these utterances. Hence, the correctness lies in two things: the fact that others commonly use similar names for white and black colors, and the fact that the clear-sighted would not alter the utterances. As for social norms, they contribute to what makes these utterances correct, but the nature and extent of their contribution varies. Because the blind speakers cannot see dark colors, for instance, their utterances are entirely reliant on social norms about the habits associated with the míng “black” and “dark.” That is, assuming these examples concern blindness from birth, the blind people’s utterances only draw on the awareness that people commonly use the sound “black” in the same way as “dark.” By contrast, if a sighted person had made the utterance, social norms would only contribute one element. In addition, a sighted person would see dark colors. Thus, while

148 | Jane Geaney sighted people rely on social norms to know which míng pick out the colors that they personally see, the blind person relies on social norms exclusively without ever using míng to select colors. In other words, I take it that the capacity to say something like “Some hair is black” does not demonstrate a blind person’s ability to connect míng to colors. This kind of assertion is simply an extension of linking míng to each other according to social norms. On my reading, these “blind utterances” are examples of equating míng in intra-míng relationships, because a blind speaker is incapable of making the míng refer to colors. The second component of what makes míng correct is not available to the blind people. The specific reference not only to sighted people (the “others”), but to clear-sighted (míngmùzhě 明目者) people in the “Guì Yì” passages implies that, in addition to mastering the community’s habits for equating míng, correct naming requires visual ability. If people who are clear-sighted are experts at judging whether the blind person’s utterance needs to be altered, then assessing the utterance must in some way involve seeing dark colors rather than just knowing whether the utterance reproduces correct social norms for associating similar kinds of naming. As noted earlier, if the utterance came from sighted people, a background of seeing things would contribute to knowing which ones to call “dark.” More importantly, the passage presents clear-sighted people as knowing even better than ordinary sighted people which things to call “dark.” With their exceptional visual skills, these clear-sighted people function as experts who ratify color utterances. But the suggestion that clear-sighted people are singularly important judges of the blind person’s utterance is puzzling. It seems to deny the nature of the correctness of the blind person’s utterance even as it affirms it. If the correctness of the blind person’s utterance is merely a matter of duplicating social norms for associating míng to one another (i.e., “Things that are dark are black” is correct simply because that is how “we” use these terms), then the clear-sighted should not be singled out to judge the utterance. After all, the ability to emulate the way the linguistic community associates two míng is separable from the ability to observe things clearly. Nevertheless, the passage invokes the clear-sighted as judges, which means the ability to observe dark things must be relevant to assessing the correctness of a blind person’s utterance. Hence, even though the clear-sighted have no basis on which to alter the blind person’s utterance, their mere presence as judges already signals that there is something wrong with it.

Words and Names Returning now to the distinction I described earlier between words and names, the legitimacy or value of these blind “míng equations” will differ

Míng (名) as “Names” Rather than “Words”  | 149 depending on whether we interpret míng to be words, on the one hand, or names on the other. That is, it is possible to say either that the blind utterances “equate words” or that they “equate names.” Now, if this were a case of equating words (without going into what is entailed in equating one word with another), we could assert that the ability to equate one word with another in a correct way amounts to effectively knowing the word. In the case of equating names, however, it is less obviously true. That is, on the stipulated definition by which names refer externally, correctly equating one name to one another might not necessarily seem sufficient to count as a significant use of names. To explain my point, for simplicity, I will focus on the “Guì Yì” utterance, which says, “That which is dark is black.” The goal of the “Guì Yì” passage is, again, to make a normative point about fulfilling one’s speech in action, and it makes this point by recourse to an illustration involving a blind person. By means of a stylistic parallelism, the “Guì Yì” passage offers an analogy between the utterance “That which is dark is black” and the use of the míng “humane.” It says these two things: 1. “That which is light is white, and that which is dark is black.” Even those who are clear-sighted (míngmùzhě 明目者) would have no basis to alter/change (yì 易) that. 「鉅 者，白也。黔者，（墨）〔黑〕也。」雖明目者無以易之。

2. Now, the gentlemen of the world’s naming “humane,” even Yu and Tang would have no basis to alter/change (yì 易) that. 今天下之君子之名仁也，雖禹湯無以易之。

(Mòzǐ 12.1 “Guì Yì” 47《貴義第四十七》) On the model of naming things “humane,” we can take the blind person’s utterance to be talking about naming things “black.” In effect, it is saying, “That which is dark is [named] ‘black,’ ” just as it might say, “Benevolent [acts] are [named] ‘humane.’ ”34 This is where the difference between “words” and míng becomes relevant. In the case of words, we might expect that knowing how to equate one word (“humane”) with other similar words (i.e., “benevolent”) in and of itself would seem to be enough to constitute knowing the word. But if, on the other hand, “humane” is understood as a name, things are a bit different. There seems to be something inadequate about only knowing how to equate the name “humane” with other names like “benevolent” that label similar actions. This also appears to apply to merely knowing that “black” names the same colors as “dark.” If we think of the míng in question as a name, not having knowledge of the color to

150 | Jane Geaney which it refers seems more like a deficiency. Thus, the fact that the passages proceed to assess the utterances in terms of their uselessness, instead of in any way foregrounding the correct affirmations about the relationships of míng to one another, helps illustrate that míng resemble names more than words.

Why Blind People, After All? We could dismiss the invocation of the clear-sighted people in the “Guì Yì” passage as a by-product of the Mòzǐ getting carried away with its own rhetoric. In favor of ignoring the reference to the clear-sighted, we might remember that the passage goes on to use the same pattern to say that the ancient sages would not change the way “gentlemen” use the name “humaneness.” Hence, the comment about the clear-sighted might simply set the stage for that assertion. On the same lines, insofar as early Chinese texts often play with rhetorical extremes (blind versus clear-sighted, sages versus “gentlemen”), the comment might also be accounted for as a mere result of stylistic compulsion. Nevertheless, a more interesting reading is worth considering. The presence of the clear-sighted judges and the way the passages brush past the correct blind “míng equations” bring us back to the choice of blind people in the first place. The premise of these examples is that blind people speaking about color is a good illustration of the problem of simultaneously knowing míng, in some sense, and yet not being able to put them to use. This would be the case only if it is customary to think of míng as occurring in concrete circumstances. In other words, insofar as these passages focus on the differences between blind and clear-sighted people naming colors, they reinforce the impression that early Chinese texts treat “language” as situated and treat the meanings of utterances as dependent on their interaction with the nonlinguistic environment.35 If it is plausible that early Chinese texts present language as something that interacts with the nonlinguistic environment, then the kind of bodies that use language becomes important. Considering the bodies that employ the language highlights some differences between míng and “word,” because, if the goal was to exemplify being in a state of knowing what words mean without knowing how to act on them, one might choose any sort of impairment. This does not happen in early Chinese texts. If míng were thought of as words, such partial knowledge of míng could be illustrated by having deaf people be asked to write about music and then fail a test of distinguishing sounds. Or taste-impaired people (more on this later) could be asked to speak about flavors and then show themselves to be unable to discern different tastes. It is by no means uncommon for early Chinese texts to employ other physical disabilities as pedagogical illustra-

Míng (名) as “Names” Rather than “Words”  | 151 tions. However, blindness and lameness are the only disabilities used to demonstrate this particular idea: knowing how to equate names but at the same time not knowing how to use them in action.

A cross-cultural perspective To pinpoint the specificity of this use of blind and lame analogies, let us take a quick and unsystematic comparative glance at disabled-person illustrations in the dominant Western philosophical tradition. They often involve blind people with color. As in my opening epigraph from The Port-Royal Logic, the illustrations sometimes concern the ability to use words, but even then, their focus is on the question of whether having knowledge without experience can be characterized as having certainty. In these examples, the disabilities are not limited to blindness. For instance, David Hume’s use of the blind/ color trope extends the examples to include the deaf. He writes, A blind man can form no notion of colours; a deaf man of sounds. Restore either of them that sense, in which he was deficient; by opening this new inlet for his sensations, you also open a new inlet for ideas; and he finds no difficulty in conceiving these objects.36 Here, deafness serves as well as blindness, because the tropes address a much-debated question about whether linguistic knowledge without sensory knowledge can be certain. Examples of blind people’s knowledge of color statements also sometimes function as illustrations of the uncertainty of faith claims, as in this example from the work of Søren Kierkegaard: If you yourself have never been in love, you do not know whether anyone has been loved in this world, although you do know how many have affirmed that they have been loved. . . . But whether they actually have loved, you cannot know; and if you yourself have loved, then you know you have loved. The blind person cannot know color differences, he must be content that others have assured him that they do exist and that they are thus and so. (emphasis added)37 In other words, without the experience of faith or love, the only certainty you can have about such statements is your experience of having heard them. In these cases where language is contrasted to experience, there is no particular motivation for choosing disabilities related to vision as the examples. Thus, these passages that compare language to experience serve

152 | Jane Geaney as a reminder that, while the disability illustrations might seem similar, putting language into action is not quite the same thing.

Lameness and the motivation for using visual disabilities The fact that early Chinese texts use lame people, in addition to blind people, to illustrate one particular kind of point increases the likelihood that the emphasis on visibility is highly motivated. In the context of Early China, lame and blind people both represent a disability pertaining to vision: the ability to act on certain kinds of naming utterances. Hence, the addition of the lame reinforces the significance of vision in these stories. We have already seen that the Yántiělùn makes a case against what we might call “inactive name knowledge” by using figuratively lame people (old long-beards and young shepherds talking about exploits beyond their abilities), but it also uses literally lame people for the same purpose. The context is an accusation that the Ru display the vice of talking about things on which they cannot act. [For the Ru], during the long period of over a thousand years since the foundation of the House of Zhou, there has been only Wen, Wu, Cheng, and Kang, to whom they would refer whenever they speak. They take up the unattainable and praise it, just like lame people who are able to speak about great distances but cannot walk them.38 自周室以來，千有餘歲，獨有文、武、成、康，如言必參一焉，取所不能 及而稱之，猶躄者能言遠不能行也。

(Yántiělùn 2 “Lún Rú” 11《論儒第十一》) Here, the trope of lameness plays a role similar to that of blindness in the other passages. The accusation is that the lame people have the ability to speak about walking great distances, and yet they cannot act on that ability. They are able to talk but not literally “walk” their talk. Thus, in both cases, the speech is correct, but a constitutional inability to put it into action precludes it from having merit. This is relevant because action is visible. The failure to walk is a visible failure, no less so than the blind people’s failure to use their hands to choose the right colors. According to the conceptions of sense perception in Early China, human action (xíng 行) or embodiment pertains to the field of vision.39 This is subtly apparent when a passage from the “Fēi Gōng Xià” chapter of the Mòzǐ pairs names with shí (實) in the sense of “actions.” Quoting more fully this time, the passage says,

Míng (名) as “Names” Rather than “Words”  | 153 Now the world agrees on dutifulness as being the method of the sage kings. Now the world’s feudal lords are alike in joining together in averting, attacking, and felling [each other]. Thus, this is having the name for praiseworthy dutifulness, but not inspecting the shí [the actions that are named dutiful]. This is like blind people naming “white” and “black” names like other people, but not being able to divide (fēn 分) the things (wù 物). How can this be called discernment (bié 別)? 今天下之所同義者，聖王之法也。今天下之諸侯將猶多皆免攻伐并兼，則 是有譽義之名，而不察其實也。此譬猶盲之與人同命白黑之名，而不能分 其物也，則豈謂有別哉?

(Mòzǐ 5.3 “Fēi Gōng Xià” 19《非攻下第十九》) The míng here is dutifulness, but the shí (實) that feudal lords display are acts of violence instead of duty.40 In other words, the passage says that, upon hearing about a reputation for being dutiful, it is advisable to look to see if the actions fulfill that name. In short, it is important to match the sounds you hear with what you see, like two parts of a tally. Thus, the reason these passages can be illustrated with examples of lame people as well as blind people is that both disabilities involve a problem pertaining to the visible sphere, where one can see people put names into action by choosing colors and “walking the talk.”

Taste impairment and the motivation for using visual disabilities It is instructive to contrast these cases to the way early Chinese texts achieve different rhetorical goals with other types of disabilities. For instance, the texts use the trope of taste impairment to argue for a different kind of inconsistency: adhering to one’s values in minor cases while abandoning them in major ones. In contrast to the examples discussed earlier, this ethical inconsistency involves a failure in knowing how to speak about míng as well as failure to act on them. These cases employ the disabilities to support an argument that people who recognize wrongness in a single instance should also recognize wrongness on a large scale. The point is that if one does not know the latter, then one cannot be credited with knowing at all. The passages mention blindness along with taste impairments, but blindness represents something different in these examples insofar as it is only one of the impairments used to illustrate a complete absence of knowledge. The passages feature a test of knowledge and then frame the results as a total failure.

154 | Jane Geaney Now if there is someone who, on seeing a little bit of black, says “Black,” but on seeing lot of black, says “White,” then we would certainly take this person to be someone who does not know the distinction (biàn 辯) between black and white. If there is someone who, on tasting a little bitterness, says “Bitter,” but on tasting a lot says “Sweet,” then we would certainly take this person to be someone who does not know the distinction (biàn 辯) between sweet and bitter. 今有人於此，少見黑曰黑，多見黑曰白，則〔必〕以此人〔為〕不知白黑之 辯矣。少嘗苦曰苦，多嘗苦曰甘，則必以此人為不知甘苦之辯矣。

(Mòzǐ 5.1 “Fēi Gōng Shàng” 17《非攻上第十七》) The second example is similar:41 Now, what if there is a person who, when being shown a little bit of black calls it black, but when being shown a lot of black, calls it white? He will certainly have to say, “My eyes are disordered and I do not know the separation (bié 別) between black and white.” Now, what if there is a person who, when he is able to taste a little sweetness, calls it sweet, but when he tastes a lot of sweetness, calls it bitter. He will certainly have to say, “My mouth is disordered and I do not know the tastes of sweet and bitter.” 今有人於此，少而示之黑謂之黑，多示之黑 謂白，必曰吾目亂，不知黑白 之別。今有人於此，能少嘗之甘謂甘，多嘗〔之甘〕謂苦，必曰吾口亂， 不知其甘苦之味。

(Mòzǐ 7.3 “Tiān Zhì Xià” 28《天志下第二十八》) In contrast to the passages about knowing how to use names with other names, but not being able to put them into action, these passages elide the distinction between using names in relation to each other. In this way, they do not grant their disabled subjects any knowledge.42 The passages present the failure to correctly name the larger quantity as completely invalidating what might have seemed like the correct assignment of names to the smaller quantity. In this kind of test, the disabled people do not get credit for knowing names in relation to one another any more than they get credit for knowing how to pick their referents. These examples might allow us to say more about what it means to use míng correctly but without acting on them. There could be good rea-

Míng (名) as “Names” Rather than “Words”  | 155 son why certain illustrations might include or exclude taste impairments. The success or failure of an “act” of discerning a flavor occurs inside the mouth where witnesses cannot observe it, which makes it hard to confirm the knowledge of flavor. That is, suppose I were asked what “sweet” is, and I passed the test by saying “sugary.” Then suppose I was also given a piece of fruit and identified it as sweet rather than bitter. If “sweet” was a word, as distinct from a míng, my answer would suffice. If it was a míng, however, it seems that more would be needed. While my answer might be deemed correct, it still would not demonstrate that I actually had the ability to taste that particular fruit, rather than just the ability to use the name that my society associates with it. To determine whether people’s answers were rooted solely in the understanding of names (as distinct from also having an ability to taste), we would have to do two things: avoid telling the taste testers what they were eating, and blindfold them. This idea of a blindfold, however, reintroduces the trope of blindness, thereby reaffirming that knowledge of míng can only be demonstrated in the absence of sight. Taste impairment is not visible enough for witnesses to discern the difference between, on the one hand, being able to use míng in relation to things and, on the other hand, only being able to use míng in relation to each other. The fact that the passages do not use taste impairment as an illustration of this difference confirms that a míng is not like a word, insofar as knowing míng requires visible pairing. Thus, the trope of taste impairment has limited use in this context, but that limitation is not characteristic of the tropes of lameness and blindness. Hence, the use of taste impairment to demonstrate total knowledge failure underscores, by contrast, the uniqueness of the situations represented by the blind and lame people who are credited with some degree of knowledge in naming. It is no coincidence that when early Chinese texts illustrate an inconsistency between míng and action, they choose a blind or lame person; whereas when they illustrate inconsistent behavior in minor and major circumstances, their illustrations might include taste impairment.

The Legacy of Blind Naming in the Mò Biàn The necessity of pairing names with something visible persists in the more abstract and decontextualized discussions of naming in later Mohist texts, even though the idea of disabled bodies drops out of the picture. The Mò Biàn arguably develops this pattern while transforming it. Instead of blind people naming colors, Canon B70 presents a situation in which sighted people hear the name of a color but cannot see the color because it is inside a room. The Canon statement characterizes this as “knowing both,” and the Explanation that follows describes why.43 Although they cannot see the

156 | Jane Geaney color in the room, they can see a color in front of them, and they hear that the color in the room resembles the one in front of them, which they know because they can see it. The Explanation that follows the Canon emphasizes visual clarity by saying, “Names are things that, by means of something clear (míng 明), correct something unknown” (夫名以所明正所不智). Therefore, in Canon B70’s Explanation, the overheard name, by means of the visible color, settles the case regarding the unknown color that is not visible. If this is indeed a development from the earlier Mohist blind-naming scenarios, then the later Mohists no longer even take into account the idea of knowing colors solely by associations among names (blind people’s color knowledge). Instead, B70 replaces the blind people with sighted people naming colors in visually compromised situations. The later Mohist texts apparently contend that sighted people hearing the names of colors that they cannot currently see does constitute a kind of knowledge, but even if they consider this to be “knowing names,” such knowledge of names would still have a foundation in seeing. Perhaps we should not assume this is “knowing names,” however, because the Explanation to Canon B70 describes this kind of knowing as “knowledge by explanation” (shuōzhì 說智). “Knowing by explanation” also seems to require connecting míng to their referents. This kind of knowing is discussed in another canon, A80, which appears to describe three means of knowing.44 The Canon’s Explanation contrasts knowing by explanation with knowing in person and knowing by hearsay. No part of Canon A80 implies, however, that there is a means of knowing that exclusively involves míng. Instead, the Explanation stipulates that míng must be understood in terms of relations to referents by saying, “Míng are things by which [something] is called,” to which it adds, “That which is called is shí.”45 Moreover, the Explanation says that “joining” is a matter of a míng matching their referents: “Míng and shí matching is joining” (名實耦合也). Regarding “doing,” it says, “Aims being enacted is doing” (志行為也). The second half of the Canon’s Explanation is more obscure, but nothing about it hints that míng can be known independently of shí. Hence, if this is the later Mohists’ development of the blind-naming scenarios, whether it is called “knowing names” or “knowing by explanation,” knowledge of míng still presumes that míng must match their referents. My analysis of passages that use physical disabilities to argue for “walking the talk” allows us to glimpse some of the features of naming that early Chinese texts take for granted and rarely articulate explicitly. Comparing the illustrations supplies a context for understanding why they discount the kind of naming knowledge that they attribute to blind and lame people. It is not, as in my examples from the dominant Western philosophical tradition, that this knowledge has no basis in experience. It is not even merely the norma-

Míng (名) as “Names” Rather than “Words”  | 157 tive point that their knowledge of the names cannot be acted on (although it is that). In terms of the implicit semantic and epistemological ideas in these passages, the problem with blind and lame naming lies elsewhere. The sorts of things that do not appear in these passages are as telling as those that do. For instance, in early Chinese texts, people who are deaf and mute are never tested on their knowledge of míng, and indeed never could be, because míng are bits of speech. Moreover, as we have seen, the difference between knowing flavors and merely knowing flavor names cannot be discerned without taking sight out of the equation. Thus, to make the direct contrast that passages are making, the illustrations must use people who can make sounds but not the visible actions that correspond to them. Hence, the examples are people who are either lame or blind. The insistence on transforming speech into action reflects a way of thinking about the human body’s means of knowing. When the blind and lame people produce naming utterances, those are sounds. What is missing from their naming knowledge is not just action, but action understood specifically as something visible that complements the sounds. Paradigmatic knowledge of names involves pairing the aural and visual sensory functions: for example, looking to confirm what is heard, using one’s eyes to select the things to which names refer, or using one’s body to make what is said into an action visible to the eyes. The fact that míng need to be visibly acted upon explains why the passages dismiss mere knowledge of míng in their relations to one another. A “word,” as distinct from a “name” in the sense previously described, has identity and meaning that appears complete without needing to be situated in any concrete, responsive, temporal context of utterance. But a míng, even when used correctly in relation to other míng, does not conform to the paradigm of míng matching something visible. It is therefore deficient. Thus, we cannot fully understand the approach to naming in Early China unless we recognize the kinds of things that the texts take to be míng, the kinds of things they take to be the referents of míng, and their general understanding of embodiment. The texts presume that a crucial feature of reliable knowledge is the complementary relationship between sound and sight. What we learn about conceptions of language in Early China from these passages is that a significant element of naming knowledge is missing when no visible form accompanies the sounds that constitute naming.

Notes   1. When I was completing my dissertation at the University of Chicago, Chad Hansen (my external advisor at the University of Hong Kong) urged me to try to grasp Graham’s research on the Mò Biàn before formulating my own ideas

158 | Jane Geaney about concepts of language in Early China. The outcome was my first publication, “A Critique of A. C. Graham’s Reconstruction of the Neo-Mohist Canons,” Journal of the American Oriental Society 119 (1999): 1–11.  2. My replacement copies include Disputers of the Tao, Chuang-tzŭ: The Inner Chapters, and Later Mohist Logic, Ethics, and Science.   3. A. C. Graham, Later Mohist Logic, Ethics, and Science (Hong Kong: The Chinese University Press, 1978), 287.   4. The association of míng with sound is evident in its use as “fame” (unfailingly described as being heard), as well as its interchanges with mìng 命 (“ordain”) and míng 鳴 (to refer to the sounds animals make). For the argument that shí 實—like xíng 行 and xíng 形—is one of the paradigmatically visible correlates of speaking and naming, see Geaney, On the Epistemology of the Senses in Early China (Honolulu: University of Hawai’i Press, 2002), 121–127, and Language as Bodily Practice in Early China: A Chinese Grammatology (Albany: State University of New York Press, 2018). I gloss shí 實 as action-thing because it is sometimes used to mean action, and because even when it used to mean “thing,” it is more like a thing-in-process than a substance. For the reasons why uses of shí in early Chinese texts should not be understood as “reality,” see my analysis of shí 實 in On the Epistemology of the Senses in Early China; “Grounding ‘Language’ in the Senses: What the Eyes and Ears Reveal about Ming 名 (Names) in Early Chinese Thought,” PEW 60.2 (2010): 251–293; and chapter two of Language as Bodily Practice in Early China.   5. To say that a míng is a shí is paradoxical. This is evident insofar as, in the rare cases when that occurs, it appears to be done in jest. See, for example, Wèi Liáo Zǐ 24.2 “Bīng Lìng Xià” 《兵令下》.   6. Karine Chemla writes, “The meaning of shí as ‘capacity’ is attested in a passage from the Records on the scrutiny of the crafts (Kaogongji, 考工記, dated to the third century b.c.e), which deals with standard vessels and is quoted in Liu Hui’s commentary on Problem 5.25 [Chemla and Guo Shuchun, 2004, 450–453].” Karine Chemla, “On Mathematical Problems as Historically Determined Artifacts: Reflections Inspired by Sources from Ancient China,” Historia Mathematica 36 (2009): 213–246, 242 n. 54.   7. I am calling this “word-meaning,” because the complexity of the various uses of the term “meaning” often provokes confusion.   8. Gilbert Ryle argues that names denote, rather than connote, and that they are “the words which do not appear in dictionaries.” Against this, Jon Wheatley contends that there is no firm distinction between names and words. He asserts, “If we are going to talk about names in the context of meaning, then the obvious difference which seems to obtain between names and other words is that names do not importantly mean anything while ordinary words do.” Wheatley might be right that the contrast can be overdrawn, but for understanding what is at stake in claiming that a míng in early Chinese texts is a name rather than a word, we have to begin with a sense of what that distinction entails. Gilbert Ryle, “The Theory of Meaning,” in British Philosophy in the Mid-Century, ed. C. A. Mace (London: George Allen & Unwin, 1957), 247. Jon Wheatley, “Names,” Analysis 25 (1965): 84.   9. It is possible to postulate that word-types are only hypothetically existent, but that does not change the likelihood that the concept would not fit with Graham’s idea of consistent nominalism. I discuss sinologists’ views on whether early Chinese

Míng (名) as “Names” Rather than “Words”  | 159 texts treat míng as a word-type in other works. See Geaney, “Grounding ‘Language’ in the Senses,” 279; and Geaney, Language as Bodily Practice in Early China. This characterization of míng as denotative is also consonant with another of Graham’s insights about early Chinese views of language, “The Chinese opposition míng/shí ‘name/object’ is very unlike the Saussurian signifier/signified  .  .  .  there is consequently no tendency for  .  .  .  third entities on the same level as the objects and the sounds of names.” A. C. Graham, Disputers of the Tao (La Salle: Open Court, 1989), 228. 10. My method here is influenced by the work of Annmarie Mol, who argues for the benefit of paying attention to “exemplary situations”—the models and metaphors used in philosophical theories (“I Eat an Apple: On Theorizing Subjectivities,” Subjectivity 22.1 [2008]: 28–37). This method is also informed by Sara Ahmed’s “queer phenomenological” discussion of the philosopher’s writing desk, which brings the familiar into the foreground and thereby demonstrates how philosophy might begin with actual bodies and their necessities. See especially the “Orientation Toward Objects” chapter of Sara Ahmed, Queer Phenomenology: Orientations, Objects, Others (Durham: Duke University Press, 2006). 11. See below in the section, “A cross-cultural perspective,” for more on the dominant Western philosophical tradition’s illustrations. 12. William F. Hanks outlines a similar contrast between different approaches to language in Language and Communicative Practices, Series Critical Essays in Anthropology (Boulder: Westview Press, 1995). 13. Neither mode of thinking about language denies the effects of social norms on linguistic meaning. 14. Incidentally, this use is often filled in early Chinese texts by “one yán 言.” I should add that, when I refer to something as being a “name,” I also mean something distinct from a “noun,” because a noun is a type of word, whereas, in early Chinese usage, “míng” is not a type of word. There is no technical concept of “word” of which a noun might be thought of as a type. It is also important to keep in mind that early Chinese texts do not posit meanings as abstract entities (word-meanings) that are possessed by míng. See Geaney, The Emergence of Word-Meaning, SUNY series in Chinese Philosophy and Culture (forthcoming). 15. Some of these same roles I am attributing to míng are ascribed to a lesser degree to “speech” yán 言, which are usually longer utterances that include míng. See Geaney, The Emergence of Word-Meaning in Early China (forthcoming). 16. My point is not that there can be no míng for things that are not visible. There are míng for all kinds of different entities like sounds and smells. The phrase “names and objects” does not imply that names only apply to medium-sized dry goods. In the same way, “language and reality” does not imply that there are no names for unreal or imaginary entities. See below in the section, “Lameness and the motivation for using visual disabilities,” for a case of the míng of a míng (or similarly naming a name) in the “Fēi Gōng Xià” chapter of the Mòzǐ. 17. As Christoph Harbsmeier points out, “For Xunzi, as for all other early Chinese philosophers of language, the main concern is not in this way essentially tied up with a relation between two terms. Chinese philosophers of language were mainly concerned with the relation between names and objects/things.” Christoph

160 | Jane Geaney Harbsmeier, Language and Logic, vol. 7 of Science and Civilization in China (New York: Cambridge University Press, 1998), 195. 18. This is not to say that a logical word cannot be part of a míng. A míng often includes the possessive zhī (之), as in someone being called “X 之 Y.” My point is merely that in early Chinese texts something like zhī 之 is never foregrounded as an individual example of a míng. 19. The time periods of these passages cannot be specified with precision. The image suggested by Erik Maeder—of loose-leaf binders rearranged with additions and deletions—nicely captures the difficulty of dating any individual passage or phrase. Erik W. Maeder, “Some Observations on the Composition of the ‘Core Chapters’ of the Mòzǐ,” Early China 17 (1992): 28. 20. Geaney, “Grounding ‘Language’ in the Senses,” 251–293. 21. For my purposes, the blind people’s speech in these passages is relevant in a way that the usual use of the “X, Y 也” (“X equates with Y”) formula is not. That is, there is seldom any reason to assume that the elements in an “X, Y 也” formula should be interpreted independently of some specific context, because the usage is usually exegetical. 22. Regarding these Mohist illustrations, Chris Fraser aptly notes, “A natural response here would be to point out that the Mohists’ own view implicitly recognizes that knowledge of words, or ‘names,’ constitutes a separate type of knowledge, distinct from knowledge of how to ‘select’ things. The Mohists ought to categorize knowledge more finely, the response would run, by distinguishing between knowing how to use names and knowing how to distinguish the referents of names.” It is important that, as Fraser says, the texts implicitly recognize that the blind people’s color naming is knowledge. Even if they do not describe it as a separate type of knowledge, they subtly acknowledge it as knowing. Fraser adds that the later Mohist texts do this by means of a fourfold categorization of knowing, which treats names, things, relations of names and things, and knowing how to act as four independent forms of knowing. I disagree with this interpretation of Canon A80 for the reasons described in my discussion of Graham’s translation in note 45. Chris Fraser, “Mohism,” in The Stanford Encyclopedia of Philosophy (Summer 2010 edition), ed. Edward N. Zalta, http:// plato.stanford.edu/archives/sum2010/entries/mohism/. 23. In early Chinese texts, colors are among the standard things that the eyes differentiate. Colors (and shapes) are used independently of items to which one might expect them to be attached. Thus, a color is, in effect, a thing, not a property of a thing. The ability to differentiate black and white amounts to the ability to see, because blindness is the absence of that ability: If eyes cannot determine the color of black from white, then call them blind. If ears cannot separate the sounds of clear from turbid, then call them deaf. If the heartmind cannot assess a position of gaining from losing, then call it crazy. 目不能決黑白之色則謂之盲，耳不能別清濁之聲則謂之聾，心不能審得失之地則謂 之狂。

(Hánfēizǐ 20 “Jiě Lǎo” 《解老》)

Míng (名) as “Names” Rather than “Words”  | 161 Thus, distinguishing black and white is fundamental. Apparently, is also easy. If the ruler is sincere and clear about the speech of his ministers, then separating (bié 別) worthy from the degenerate is [as easy as separating] black from white. 為人主者誠明於臣之所言，則別賢不肖如黑白矣。

(Hánfēizǐ 44 “Shuō Yí” 《說疑》) Black and white are easy to separate not because of the extent to which they contrast, but because they are visible. (For this argument, see Geaney, “Binaries in Early Chinese Texts: Locating Entities on Continuums,” International Communication of Chinese Culture 3.2 [2016]: 275–292.) That is not to say everything visible is easy to see, seeing an autumn hair is a metaphor for a visual task that is not easy. 24. Translations like “separating/discerning” are rather wooden, but I use them deliberately because of my interest in the way parts of speech (words or phrases) repeat themselves in different ways, often without our awareness, which is difficult to convey in more polished-sounding translations. 25. To clarify my use of terminology, as I am using the word “referent,” a person who cannot see color cannot see the “referent” of the name, by which I mean, in these cases, the color (the shí 實, or the wù 物, as in Mòzǐ 5.3 “Fēi Gōng Xià” 19). 26. In early Chinese texts, spatial separation is arguably constitutive of the identity of an action or a thing. I discuss this in “Binaries in Early Chinese Texts.” 27. Chris Fraser rightly noted in his comments on a draft of this piece that the last phrase can also be read as asking how there can be any difference between the blind people and the targets of the passage’s criticism. But I stick to this translation because the use of bie 別 seems to refer back to the dividing (fen 分) of things. 28. I am indebted to Dan Robins for suggesting something like this translation of the line. More typically, bì kě 必可 is taken to mean “can” or “is able to,” but comparisons to other examples of the use of bì kě 必可 in early Chinese texts suggest that here it is not being used to mean, in effect, “it is certainly permitted but optional.” I discuss this line more in chapter five of Language as Bodily Practice in Early China. 29. The passage explicitly evokes the discussion in Xunzi 27 (“Dà Lüe”) about acting on one’s speech. My tentative translation of the Yántiělùn passage is: Those who are able to speak and yet unable to act are the treasure of the state. Those who are able to act but unable to speak are the tools of the state. The gentleman has both. Those who lack one are the shepherd boys and the long-beards. The mouth says it and the body enacts it. How is that like silently holding an official post and going through the motions? 能言而不能行者，國之寶也。能行而不能言者，國之用也。兼此二者，君子也。無 一者，（烏獲）〔牧童〕、（逢須）〔蓬頭〕也。口言之，躬行之，豈若默然載施 其行而已。

(Yántiělùn 7 “Néng Yán” 40《能言第四十》)

162 | Jane Geaney My translation follows commentators who emend shī 施 “bestow” to shī 尸 “corpse.” 30. The assumption seems to be that proficiency in sound compensates for deficiency in sight. For example, the Huáinánzǐ says, Now in the case of blind people, their eyes cannot separate day from night or divide white from black; however, when they grasp the qin and pluck the strings, triply plucking and double pressing, touching and plucking, pulling and releasing, their hands are a blur and they never miss a string. 今夫盲者，目不能別晝夜、分白黑，然而搏琴撫弦，參彈復徽，攫援摽拂，手若薎 蒙，不失一弦。

(Huáinánzǐ 19 “Xiū Wù Xùn” 《脩務訓》) Sarah A. Queen and John S. Major translation modified. John S. Major, Sarah Queen, Andrew Meyer, and Harold D. Roth, trans., The Huainanzi: A Guide to the Theory and Practice of Government in Early Han China, by Liu An, King of Huainan (New York: Columbia University Press, 2010), 778, 682. The Wénzǐ says something similar. A tortoise lacks ears, but its eyes cannot be occluded. Its jing 精 (refined essence) lies in clarity. Blind people lack eyesight, but their hearing cannot be occluded. Their jing 精 lies in keen hearing. （聾）〔鱉〕無耳而目不可以蔽，精於明也；瞽無目而耳不可以蔽，精於聽也。

(Wénzǐ 6 “Shàng Dé” 《上德 》) 31. See appendix to Geaney, Language as Bodily Practice. Incidentally, this means that the texts do not imply that blind people cannot use names to refer. They only imply that they cannot rely solely on vision to use names to refer visible things. 32. Some commentators emend jù 鉅 “great” to ái 皚 “white,” as in snow. I do not translate it as “things that are light,” because that would make it like the property of a thing, rather than a thing itself. It seems that we should think of it as “lights” or “brights,” as one might say of sorted laundry. 33. In early Chinese texts, the use of míng 明 in the context of the eyes generally indicates clear-sightedness, not just sightedness. Míng 明 and cōng 聰 are refined characteristics, typical of a sage. While the Shizi contains a passage wherein those who are called míngmùzhě 明目者 seem ordinary (Shizi 尸子 1 〈卷上〉1．12 處道), in this Mòzǐ passage, the parallel between the míngmùzhě 明目者 and the sages (Yu and Tang) indicates that they are exceptional, like the sages: “Now, the gentlemen of the world’s naming ‘humane,’ even Yu and Tang would have no basis to alter that” (今 天下之君子之名仁也，雖禹湯無以易之。). Thus, the míngmùzhě is to the blind as Yu and

Míng (名) as “Names” Rather than “Words”  | 163 Tang are to the “gentlemen of the world.” Other passages that support this include the following: Seeing the sun and moon does not count as clear eyesight (míngmù 明目). Hearing thunder rumblings does not count as keen hearing (cōngér 聰耳). 見日月不為明目，聞雷霆不為聰耳。

(Súnzi 4 “Xíng Piān” 《形篇》) As the Guǎnzǐ puts it: The eyes are in charge of looking. Looking must accompany seeing. If one sees and inspects, call it míng.  .  .  .  If one sees and does not inspect, then it is not míng, and if it is not míng then there are mistakes. 目司視，視必順見，見察謂之明。。。視不察不明，不察不明則過。

(Guǎnzǐ 4 “Zhòu Hé” 11 《宙合第十一》) 34. The sentence that follows the utterance about dark colors being black comes from the interlocutor, the eponymous Mozi, and informs us that these blind people cannot select black colors, not that they cannot select dark colors. Thus, the text is treating black colors and dark colors as interchangeable. Hence, my choice of “benevolent” and “humane” is supposed to represent two terms with very similar uses (so similar, in fact, that both are often used as translations of rén 仁). 35. I put “language” in quotes because by some definitions what I am describing is merely speech and names. 36. David Hume, An Enquiry Concerning Human Understanding (London: A. Millar, 1777), E. 2.7. Accessed from www.davidhume.org (Peter Millican and Amyas Merivale 2001–2012). 37. Søren Kierkegaard, Christian Discourses: The Crisis and a Crisis in the Life of an Actress, ed. and trans. Howard V. Hong and Edna H. Hong (Princeton: University of Princeton Press, 2009), 237. 38. Translation modified, Esson M. Gale, Discourses on Salt and Iron: A Debate on State Control of Commerce and Industry in Ancient China, Chapter I–XXVIII: Translated from the Chinese of Huan K’uan, with Introduction and Notes (Taibei: Ch’eng-Wen, 1967), 13. 39. See note 4. The following things are associated with vision: “walking” (xíng 行), “action-thing” (shí 實), embodying (tǐ 體), and treading (lǚ 履). 40. This passage draws attention to the visibility of shí 實 by mentioning a need to inspect (chá 察) the shí of the name “dutiful.” While it has broader uses, the term chá 察 is often used to mean specifically inspecting by means of the eyes. See appendix to Geaney, Language as Bodily Practice in Early China. 41. This second passage supplies more of the speech of the disabled people and uses some different terms, but these are not relevant to my point.

164 | Jane Geaney 42. They not only flatly deny the knowledge but also do not mention that the disabled people speak the way others do or say things that experts would not change. 43. The “knowing both” is ambiguous, but we might take it as meaning that both the name and the color are known. 44. The grounds for thinking it is about knowing is that “knowing” (zhī 知) is the first graph in the Canon statement. 45. Canon A80 does not introduce four objects of knowledge, as Graham claims (Later Mohist Logic, 30–32). Graham’s argument is based on something that I view as a misreading of a list of seven terms in the Canon statement of Canon A80, three of which he views as means of knowing—wén 聞 (hearsay), shuō 說 (explanation), and qīn 親 (in person). In light of the Canon’s Explanation, this part of his interpretation is plausible. But without providing any justification, Graham asserts that the other four—míng 名, shí 實, hé 合 (join), and wèi 為 (do)—are “objects” of knowing and even “disciplines,” each inherently separate from the other. Evidence in the Canon’s Explanation, however, undermines the persuasiveness of this. There is no reason to assume míng or shí constitute separable and independent objects of knowledge when the Explanation proceeds to assert that they function together: “Míng and shí matching is joining” (名實耦合也). See Geaney, “A Critique of A. C. Graham’s Reconstruction of the Neo-Mohist Canons.”

8

Unfounded and Unfollowed Mencius’s Portrayal of Yang Zhu and Mo Di

Carine Defoort

Three major figures in almost any historical overview of early Chinese philosophy are Mencius (孟子 ca. 371–289 BCE), the supposed founder of Mohism known as Mozi (墨子 ca. 479–381 BCE), and the shadowy figure Yang Zhu (楊/陽朱 fl. ca. 350 BCE).1 Angus Graham wrote extensively about all three of them. They occur together in Mencius’s well-known complaint about the enormous appeal of Mo Di (翟 Mozi) and Yang Zhu (Yangzi) in his days. Worrying about the chaotic times in which he lived, Mencius declared the heretical ideas of these two masters a major cause of the problem. Motivated by his loyalty to Confucius and exasperated by the success of his rivals, Mencius offered the following characterizations of Yang and Mo: Yangzi chooses “for oneself”; if he could have benefited the world by pulling out one hair, he would not have done it. Mozi “cares for each one”; if he could have benefited the world by wearing himself smooth from the crown to the heels he would have done it. 楊子取為我, 拔一毛而利天下, 不為也. 墨子兼愛, 摩頂放踵利天下, 為之.2

A Western philosopher willing to take early Chinese thought seriously would be perplexed by Mencius’s characterization of his two opponents in terms of the amount of hair that they were (or were not) willing to sacrifice for the benefit of the whole world. She could probably make sense of Mencius’s outrage at Yang Zhu’s refusal to offer even one hair for the benefit of the

165

166 | Carine Defoort whole world, but what is so terrible about working day and night, rubbing bald one’s whole body by carrying things or plowing through the marshes for the benefit of others?3 Mencius’s portrayal of the two “heretics” was largely uncontested until the twentieth century, when a booming interest emerged in early Chinese philosophers. Mo Di was relatively well attested in early sources, and most scholars could make sense of Mencius’s claim that he would have sacrificed all this body hair for the sake of the people. But Yang Zhu was another matter: Who was he? What did he think? Did he really have the success that Mencius complained about?4 Within that context grew the suspicion that Mencius might have wrongly attributed this statement to Yang Zhu, or that he misunderstood or misrepresented it. Despite the extreme sparseness of textual records concerning Yang Zhu, most scholars seem to believe that Mencius recorded a claim that was made by a historical person.5 Others, like Graham, believe that he recorded a claim that might have been attributed to Yang Zhu by others “to tell us what Yang Zhu taught as they needed a name to label the teaching, it being the convention to call each school after a supposed founder.”6 I think that neither was the case: Mencius’s portrayal of his two rivals may to a large extent have sprouted from his imagination.7 I believe furthermore that nobody in his day took notice of it. Mencius stood alone with his view: it was both unfounded and unfollowed. This essay consists of four parts. It starts off with Mencius’s unfounded characterization of the egoist Yang Zhu by highlighting the absence of any testimony in the early corpus of roughly six centuries of any master named Yang Zhu or related to him making any claim about “one hair” in relation to the world. The second part discusses Mencius’s much less contested but equally unfounded portrayal, I believe, of Mo Di willing to rub his whole body bald for the sake of the people in the world. The third part shows how unfollowed Mencius’s statement was, thereby challenging Graham’s claim about generations of Mohists and Confucians despising Yang Zhu.8 The fourth and last part concerns the often quoted but difficult passage in the Liezi that may have been inspired by Mencius, when his star had started to rise.

Unfounded: Mencius’s Portrayal of the Egoist Yang Zhu Mencius’s portrayal of Yang and Mo as, respectively, an extreme egoist and altruist is probably one of the oldest received passages in which an early Chinese master attacks others. And it has been exceptionally influential. The book Mencius does not often explicitly refer to Yang Zhu or Mo Di, but when

Unfounded and Unfollowed  | 167 it does, they are usually presented together as each other’s mirror images, both despicable because of their opposite but equally shocking appreciation of body hair in relation to the world. Mencius’s strongest accusation is that Yang Zhu’s choice “for oneself” (為我) amounts to “a denial of one’s lord” (無君) and that Mozi’s “care for everybody” (兼愛) amounts to “a denial of one’s father” (無父); both denials turn them into “beasts” (禽獸).9 On the basis of the received sources, we can trace the increasing influence of this portrayal from—but not before—the late Western Han dynasty until the Qing.10 When it was first criticized, the defenders of Yang or Mo did not focus on the claims about their hair management, but on what they were (not) willing to sacrifice it for; in other words: whom they really cared for. While the earliest modern defense of Mozi pointed out that care for everybody did not necessarily amount to a failure to respect one’s father, support for Yang Zhu focused on the expression li tianxia (利天下). Gu Jiegang (顧頡剛 1893–1980) was one of the early scholars who argued that Mencius had misrepresented Yang Zhu’s unwillingness to actually “benefit from the world” (li tianxia). Yang Zhu had claimed that he would not pull out one hair even if doing so would give him the benefit of the whole world. Thus, Yangzi was not selfishly refusing to help others; he was able to resist the appeal of possessing the whole realm because he realized that, in the end, this would harm his health.11 In support of his interpretation, Gu referred to a passage in Han Feizi “Xian xue” (顯學) in which someone is described as unwilling to enter an endangered city or a battle because “he would not exchange one hair of his shin for getting the benefit of the world” (不以天下大利易其脛一毛).12 That soldier might have earned a big reward if he was victorious in battle, but he considered nothing worth risking his life for.13 Even though Feng Youlan (馮友蘭 1895–1990) considered this reading possible, the parallelism in Mencius’s portrayal with Mozi’s extreme eagerness to benefit the world nevertheless strongly suggested to him that at least for Mencius the expression must have meant “benefit the world.”14 Indeed, the expression “li (利) the world” is almost always used in the sense that Feng attributes to Mencius. In the 1950s, Graham had even questioned the grammatical acceptability of Gu Jiegang’s alternative. But he became convinced of its possibility, mostly by a Lü shi chunqiu fragment in which li tianxia undoubtedly means “benefit from the world.”15 It describes the need for moral servants to oppose greedy rulers, because “only after virtuous power had declined and the realm was thrown into chaos did the Son of Heaven benefit from the world, the ruler of a state from his state, and the heads of office from that office” (德衰世亂 然後天子利天下, 國君利國, 官長利官) (“Shi jun” 恃君).16 Graham therefore acknowledged the suggestion

168 | Carine Defoort that Yang Zhu never objected to “benefit the world” and that “Gu Jiegang was right in arguing that Mencius misrepresented Yang Zhu. It was not that the Yangist would not lift a finger to help the world; the point was that he would not accept the least injury to the body, even the loss of a hair, for the sake of any external possession, even the throne of the empire.” But Graham also nuanced this view: “Rather than misrepresenting Yang Zhu’s philosophy, Mencius may be exposing what he sees as the doctrine’s selfish implications, which Yang Zhu is trying to hide.”17 I believe, however, that there is negative as well as positive evidence to doubt Mencius’s very claim. A certain Yang Zhu may very well have existed in or before Mencius’s time, either as a historical or textual figure (one to whom a set of ideas was attributed). But there is no evidence that he was unwilling to pull out one hair to benefit the world or to benefit from it.

Negative evidence One reason why Gu Jiegang’s suggestion has not been very influential in the field may be—aside from Mencius’s lingering influence—that it is unconvincing, but not on the grounds pointed out by Feng or Graham. The transitive verb li can indeed mean “to benefit from,” occasionally even with the object tianxia. The weakness of his argument, however, lies in the fact that nowhere in the received corpus do the defenders of health—the presumed followers of Yang Zhu—refer to “one hair” or to “li tianxia” when expressing their unwillingness to risk their health. The passage from Han Feizi adduced by Gu is a critical representation of their thought: Han Fei complains about some rulers’ respect for men who “value life more than anything” so that they would not sacrifice one hair in order to get the whole world. Graham’s defense of Mencius is therefore equally unconvincing: Mencius did not expose the immoral implications of Yang Zhu’s philosophy of life, but unfavorably attributed an extreme position to a certain Yang Zhu that no defender of health had ever promoted, defended, or even considered, at least not in the surviving sources.18 There exists no textual evidence, aside from Mencius’s own testimony and those much later sources quoting or repeating him, that supports this extreme portrayal of Yang Zhu’s thought.

Positive evidence Negative arguments are inherently weak since the nonexistence of something is hard to prove. How about the possibility that Yang Zhu or any other health defender may have expressed, orally or in a now-lost written source, the unwillingness to sacrifice one hair in order to li tianxia? Could it be that

Unfounded and Unfollowed  | 169 Mencius merely represented a then-existing view and exposed its implications, as Graham believes? Since only a few textual vestiges of the Warring States intellectual scene have survived the ages, this possibility cannot be excluded.19 Here is where a tentative and positive argument takes over. A possible reconstruction of Mencius’s portrayal on the basis of the few sources that have reached us suggests that it could have grown out of existing arguments as a hyperbolic response to them. Some pro-health narratives mostly preserved in third century BCE sources—four Lü shi chunqiu (呂氏春 秋) chapters and one late Zhuangzi chapter—may have been records of oral or written persuasion that had been circulating when the book Mencius was written.20 In these records, defenders of health consistently oppose body parts to possessions: they weigh a head (首/頭), a finger (指), a hand (手), a foot (足), an arm (臂), a life (生), or body (身/形) against mere possessions such as a thing (物), profit (利), nobility and wealth (貴富), a piece of clothing ( 衣), a cap (冠), a shoe (履), a pearl (珠), gold (金), land (地), a state (國), or the world (天下).21 In these hypothetical acts of weighing, the body always outweighs the possession. Even the item tianxia, however valuable, consistently occurs on the “light” side. I have argued elsewhere that these persuaders use “the world” as the stock example of the weightiest imaginable possession and yet not heavy enough to sacrifice one’s health for. For most people, including rulers, possession of the world was not a real option but stood for the prototypical “thing” (物) that never outweighed the body. Thus went their argument: if the world is not worth sacrificing an arm or hand for, how much less then should one risk ruining one’s whole body (which is heavier than a part) in the pursuit of mere land, luxury, or status (which is lighter than the world). Hair never enters these arguments, since few people would be unwilling to literally pull out one hair in exchange for the world.22 Two characteristics of these pro-health persuasions can be directly related to Mencius’s portrayal of Yang Zhu: the rhetorical use of “the world” as the weightiest thing, and the opposition of body parts to things. But the critical master may have adapted the argument while adopting it, again, in two important ways. First, inspired by these persuasive metaphors, Mencius responded by adding the prototype of the lightest possible body part, opposing it to the prototype of the heaviest thing. The addition of such a ridiculously light body part as “one hair” makes the view that Mencius attributes to Yang Zhu extreme. And second, the expression li tianxia pops up, which never appeared in any of the health-defending arguments. Combined with the negative evidence, these rhetorical connections support the suspicion that Mencius’s portrayal of Yang Zhu may very well have been the creation of his own genius.

170 | Carine Defoort

Equally Unfounded: Mencius’s Portrayal of the Extreme Altruist Mozi Even if we believe that the well-known Mencius passage testifies to the clash of two rhetorical tropes—“one hair” and “the world”—this still does not explain why a certain Yang Zhu was suddenly portrayed as unwilling to “benefit the world,” a position that no health defender had ever considered, let alone defended. While the unwillingness to pull out literally one hair in exchange for the greatest possible possession would have made him look unreasonable, the unwillingness to benefit the world with this minimal sacrifice turned him into a highly immoral egoist. And this was indeed Mencius’s point: Yang Zhu was not willing to sacrifice the lightest imaginable body part! His accusation was, of course, not a neutral description of facts, nor a matter of mere philosophical disagreement, but closely related to Mencius’s deepest concerns.23 And these were not as different from the Mohist ideals as his inflated attacks suggest.24 Even though he sometimes seemed allergic to the word “benefit” (利), the second Confucian sage also believed that the ruler should be good for his people and bring them welfare.25 This shared view on the ruler’s duty toward the world may have influenced Mencius’s parallel representation of his two rivals. This Mohist connection contains a second basis for the hair metaphor. Considering their veneration of the Great Yu (禹) laboring so hard in rivers and marshes that he lost all the hairs on his calves and on his shins, the lack of body hair in Mencius’s portrayal of Mo Di sounds familiar. And this altruistic model could, in turn, have inspired Mencius’s description of Yang Zhu.26 In that case, the prototype of “one hair” not only functioned as the concern for the lightest possible body part presumably opposed to the “world” by men who were obsessed by their health, but also as a contrast with all the body hair that the altruistic cultural hero had sacrificed for the sake of others. While some scholars believe that the opposition with Mohism inspired Yang Zhu to actually make his extreme claim,27 others have accused Mencius of falsely portraying Yang Zhu in extreme opposition to the Mohist model.28 I take the argument one step further, suggesting that even Mencius’s portrayal of Mo Di may have been largely invented. Even though the association of Emperor Yu with Mohism is legendary, it was not particularly strong in Mencius’s days and certainly not exclusive. For one, it does not occur in the Mozi, which counts among the oldest master texts, especially its core chapters.29 The book never mentions hair—or physical deformation for that matter—in relation to Yu’s exploits,30 nor does it give Yu an exceptional status among the other sage rulers.31 Even though the hard-working cultural hero does occur in a variety of other, mostly later texts,32 for instance in the last Zhuangzi chapter, “Tianxia” (天下),33 Yu is

Unfounded and Unfollowed  | 171 never criticized for it in the Mencius, nor is he said to have sacrificed even the smallest part of his body.34 The stereotypical association of Yu with Mohism was, moreover, far from consistent. Yu’s hard work is often praised without reference to Mohism, as in the Rong cheng shi (容成氏). This text records how Yu’s hard labor caused “his face to become dry and coarse, and his shins to no longer grow hair” (面乾粗, 脛不生之毛).35 Likewise, in Han Feizi “Wu du” 五蠹 and later sources quoting it, he is described in terms of hair sacrifice, but without reference to Mohism (see further discussion). The sacrifice of hair or any other body part is also variously attributed to different sage-kings or cultural models. Zhuangzi “Zai you” (在宥), for instance, states that Yao and Shun—not Yu— “then wore off the hair on their thighs and on their shins in order to nourish the bodies of everyone in the world” (堯舜於是乎股無胈脛無毛, 以養天下之形).36 In his critique of physiognomy (“Fei xiang” 非相), Xunzi points out that all major sages had some physical deficiency.37 And in Kongcongzi, finally, Zisi argues that true sages lacked hair and had all sorts of deformities.38 The appearance of Yu with or without hair does not always herald a Mohist position, not even Mohist influence.39 And finally, even if some connection existed between Mohism and Yu sacrificing the hair on his legs in Mencius’s days, his own portrayal of Mozi still remarkably deviates from this. He describes Mo Di (not Yu) as missing hair all over the body (not just the calves, shins, or thighs) in an expression that has puzzled commentators and translators. Mo Di was willing to “wear himself smooth from the crown to the heels,”40 analogous with his rhetorical portrayal of an extreme egoist who “would not pull out one hair.” While the loss of hair on the legs can easily be understood in relation to the hydraulic endeavors usually attributed to Yu, it is less clear what Mencius had in mind with his unprecedented description of Mo Di becoming bald from top to toe. This portrayal was perhaps as extreme and original as that of Yang Zhu. A full-blown Mohist stereotype against which Mencius portrayed Yang Zhu may not yet have existed.

Unfollowed: Post-Mencian Variations on “Hair” and the “World” Even though the two hair-related portrayals in Mencius’s accusation were clearly related to each other, the specifics of their connection—which influenced which—cannot be easily established. My attempt to relate them in a positive way to ongoing debates was tentative. Turning now to the supposed reactions to Mencius’s portrayal, there is an implicit consensus to take Mencius’s word for it. Franklin Perkins claims that “Mozi himself

172 | Carine Defoort was famous for having worn the hair off of his legs in the service of the people,”41 but there is no source besides Mencius’s own portrayal confirming this. No text expresses any awareness of his claim, neither by denying nor by confirming it. Apparently, Mencius’s unique characterization of his rivals was not known or quoted in his days, unless all the relevant evidence has disappeared. Nor is there any indication in the early sources of Confucians and Mohists deriding Yang Zhu’s egoism, as Graham claimed. The passages discussing hair in relation to the world nevertheless seem connected to each other. Four somewhat overlapping variations of the combined tropes can be distinguished.

Praise for not sacrificing one hair to obtain the world One probably relatively late variation of the hair-world metaphor has been preserved in Han and later sources. Chapter “Chu zhen xun” (俶真訓) in Huainanzi describes perfect humans who put techniques (shu 術) aside and merge with the Way (dao 道). Human perfection lies in the indifference to beauty and luxury, profit and status. Nothing can perturb these men, not even the risk of death. Since life is not enough to motivate them, how could benefit be enough to move them? Since death is not enough to stop them, how could harm be enough to frighten them?  .  .  .  Even if such a great thing as the world could be exchanged for one hair on their shin bone, they would consider none of it in their purposes. 生不足以使之, 利何足以動之. 死不足以禁之, 害何足以恐之  .  .  .  雖以天下 之大, 易骭之一毛, 無所概於志也.42

The Kongcongzi “Kang zhi” (抗志) has Zisi expressing the same idea in very similar wording when advising the lord of Wei to focus on the Way rather than on techniques. Superior men in antiquity were not affected in their sense of purpose “even by the offer to exchange the world for a hair on their shin” (雖以天下易其脛毛).43 Here we encounter a particular constellation of the two tropes: “one hair” and “the world.” The unwillingness to sacrifice exactly one hair somewhat resonates with Mencius’s critical portrayal of Yang Zhu, the shin bone is reminiscent of the paradigmatic good ruler, the option to possess the world resembles the health defenders’ weighing arguments collected in the Lü shi chunqiu, but the duty to benefit the world is not considered here. This particular mix endorses a novel view: nothing is worth sacrificing a hair for; the perfect man is not only oblivious to riches and luxury but also to

Unfounded and Unfollowed  | 173 life and death. The care of one hair therefore does not represent a concern with health or physical wholeness. The unwillingness to exchange it for anything at all characterizes the ideal person who has merged with the Way.

Criticism for not sacrificing one hair to obtain the world A second variant, quoted (wrongly, I believe) by Gu Jiegang in defense of Yang Zhu’s position, occurs in a slightly earlier source, namely Han Feizi “Xian xue” (顯學). The author also thought in terms of possessing the world when he described some high-minded persons’ unwillingness to risk life in battle. Imagine somebody whose principles did not allow him to enter a city in danger, lodge in a military camp, or exchange one hair on his shin for the great benefit of [receiving] the world. For sure, the rulers of the age would respect him for it, value his wisdom, and exalt his behavior, regarding him a man who gives weight to life above anything. 今有人於此, 義不入危城, 不處軍旅, 不以天下大利易其脛一毛. 世主必從而 禮之, 貴其智而高其行, 以為輕物重生之士也.44

Han Fei is very critical of such self-defeating rulers: he points out that subjects will only risk death for him if they value the promised rewards more than life. “If the superior now honors and values those who give weight to life above anything and yet demands that the people go out to die and value self-sacrifice for their superior, this will not work” (今上尊貴輕物重生 之士而索民之出死而重殉上事, 不可得也). The high-minded subject described here indeed closely resembles the health defenders, who “give weight to life above anything” (輕物重生),45 even if that thing is the whole world. But by adding the “one hair” as the prototypical body part that they did not want to sacrifice, Han Fei, like Mencius, unfavorably attributes to them an extreme position that they never defended.

Cynical appraisal of the ideal of abdication A third variant of the hair-world trope has also been preserved in the Han Feizi “Wu du” (五蠹). Here the author does not praise or blame people for rejecting the possession of the world but fully understands them. In the old days, when Yao and Yu lorded over the world, they wanted to get rid of it because it made them suffer unbearably. Both rulers are implicitly portrayed in Mohist terms.46 Yu “led the people by personally holding plow and spade, till his thighs and shins had no more hair. Even the labor of a war prisoner

174 | Carine Defoort was not worse than this” (禹之王天下也, 身執耒臿以為民先, 股無胈 脛不生毛. 雖 臣虜之勞不苦於此矣). Life on the throne was so hard that he gladly abdicated: “Passing on the world was not worth making a big deal of” (傳天下而不足 多也). There was nothing moral about this act of abdication. “The fact that he lightly resigned as Son of Heaven was not because he was high-minded but because his power was scant” (輕辭天子, 非高也, 勢薄也).47 If possession of the world causes suffering, then a reasonable ruler does not want it: this is how the Second Emperor understood Master Han’s (韓子) advice, which he very freely quoted in response to Li Si’s (李斯) warnings. Since Yao lived in poor circumstances and Yu labored so hard that “his shins had no more hair” (脛毋毛), the young Qin emperor felt no urge to emulate them.48 The conversation is also recorded in Li Si’s biography, where Han Fei’s description of Yu’s physical deterioration is further enhanced. And this was certainly not what the emperor wanted.49 In this variant of the hair-world trope, presented by Han Feizi and creatively quoted by the second emperor of Qin, hair and the world are not explicitly offered in exchange. But the lack of hair is a sign of service and suffering, and the world is a possession to either enjoy or reject, but not specifically for the sake of one’s health. A ruler only wants the world if it adds to his status, power, and pleasure. For Han Fei, this insight reduced the moral exemplars to ordinary human beings, but for the second emperor it was an argument to exclusively rely on harsh laws and enjoy life, a policy that would have been abhorred by the health defenders. History would prove them right.

By benefiting the world, one deserves to benefit from it A fourth and last variant consists of stories in which sacrificing one’s body hair for the sake of the world is associated with the right to possess it. This view combines the two possible meanings of li tianxia—benefit (from) the world—even though this expression does not occur here. A saying in the Shizi makes the point: having described the ancient sage rulers as laboring for the people of the world—including “Yu whose shins did no longer grow hair” (禹脛不生毛)—it continues: “Hence, they enjoyed the riches of the world and were ennobled as Sons of Heaven” (故富有天下, 貴為天子矣).50 A similar idea is expressed indirectly by Sima Xiangru who was sent as an envoy to Shu.51 The possession of the world was for the person who benefited it without expecting anything in return, a common idea that was expressed by various masters, often without reference to Yu or to the sacrifice of hair.52 To conclude, the sample of political-philosophical arguments discussing hair in relation to the world is small and hardly echoes Mencius’s extreme

Unfounded and Unfollowed  | 175 portrayal; even his choice of vocabulary does not resonate, except perhaps the “one hair” trope. While hair in general is usually thought of as lacking from the legs of exemplary rulers, “one hair” more specifically stands for the smallest body part that one would not exchange for the whole world. All the authors think of the “world” in terms of a possession to be enjoyed or rejected; only some see it (like Mencius) as standing for the people who deserve to be benefited. The most striking conclusion, however, is the variety of views that are expressed with the hair-world tropes. Almost every passage makes a different point, not only from Mencius but also from each other, even the two that are collected in the same book (Han Feizi) or that explicitly quote one another (the second emperor’s use of master Han Fei’s words). Of course, we do not expect a persuader to quote a master simply in order to repeat him but rather to make a point of his own. And in these passages, the point never amounts to Confucian and Mohist moralists deriding Yang Zhu for his egoism.

One Superficial Follower: “Yang Zhu” in the Liezi Early sources thus suggest that most authors may not have known Mencius’s critical portrayal, or perhaps they did not consider it worth quoting, discussing, defending, criticizing or even repeating, except for a glimpse of it in one passage53 roughly six centuries later. Even though Mencius is not mentioned, the implicit connection with his portrayal has given this textual fragment a certain status throughout Chinese history.54 It occurs in a chapter explicitly named after Yang Zhu in the Liezi, a book that can for the most part be dated around CE 300. The hedonism promoted in this chapter is generally considered as late as the book itself, but some scholars believe that parts of it date from the Warring States period, including the well-known passage that echoes Mencius’s claim about Yang and Mo.55 Graham’s view was more complex: this fragment was indeed a direct testimony of the historical Yang Zhu’s view (fourth century BCE), but had been distorted by a Mohist author in a now-lost source (ca. third century BCE), and had then ended up in the hedonist Liezi chapter (ca. CE 300).56 The “surface” translation of the first fragment, according to Graham, goes along these lines: Yang Zhu said: “Since Bocheng Zigao would not benefit others at the cost of one hair, he renounced his state and retired to plow the fields. Since the Great Yu did not keep even his own

176 | Carine Defoort person for his own benefit, he worked to drain the flood until his whole body was limping and emaciated. If men from antiquity could have benefited the world by the loss of one hair, they would not have given it; if the world had been given to them alone, they would not have taken it. If nobody would lose one hair, and nobody would benefit the world, the world would be well ordered.” 楊朱曰: 伯成子高不以一毫利物, 舍國而隱耕. 大禹不以一身自利, 一體偏 枯. 古之人損一毫利天下不與也. 悉天下奉一身不取也. 人人不損一毫, 人 人不利天下, 天下治矣.57

The original pro-health message, which “is still visible behind the passage” and “which makes much more sense,” depends on the alternative translation of the verb lì (利) as “benefit from”: since “Bocheng Zigao would not benefit from things (利物) at the cost of one body hair” he “renounced his state and retired to plow the fields.” More generally, “if men from antiquity could have enjoyed the benefit of the world (利天下) by sacrificing one hair, they would not have given it.” Therefore, “if nobody sacrifices one hair, and nobody enjoys the benefit of possessing the world (利天下), the world would be well ordered.” Somewhat like Mencius, who, in Graham’s eyes, exposed Yang Zhu’s egoism, this unknown Mohist author “distorted” the original message.58 The Mohist voice only becomes explicit in the second part, when Mozi’s disciple, Qin Guli, confronts Yang Zhu implicitly with the Mencian criticism but in terms that are less ambiguous than the expression li tianxia: Qin Guli asked Yang Zhu: “To discard one hair of your body in order to save the whole age, would you do that?” Yang Zhu said: “The age is certainly not something that one hair could save.” “But suppose that it could, would you then do it?” Yang Zhu did not answer him. 禽子問楊朱曰: “去子體之一毛以濟一世, 汝為之乎?” 楊子曰: “世固非一毛 之所濟.” 禽子曰: “假濟, 為之乎?” 楊子弗應.

Yang Zhu’s initial answer as well as his ultimate failure to respond show that he may never have thought in terms of saving or benefiting the world. Here a disciple, Mengsun Yang, tries to defend his master by immediately

Unfounded and Unfollowed  | 177 replacing hair by skin and flesh, and by reformulating the question in terms of gaining things rather than benefiting others. “You don’t understand my Master’s mind. Let me explain it. If you could win ten thousand pieces of gold by injuring your skin and flesh, would you do it?” “I would.” Mengsun Yang continued: “If you could gain a state by cutting off one of your joints, would you do it?” Qinzi was silent for a while. Mengsun Yang said: “It is clear that one hair is a trifle compared with skin and flesh, and so are skin and flesh compared to one joint. But an accumulation of single specimens of hair amounts to skin and flesh; an accumulation of skin and flesh amounts to one joint. One hair is inherently one thing among the myriad parts of one body. How can one treat it lightly?” 孟孫陽曰: “子不達夫子之心. 吾請言之. 有侵若肌膚獲萬金者, 若為之乎?” 曰: “為之.” 孟孫陽曰: “有斷若一節得一國, 子為之乎?” 禽子默然有閒. 孟 孫陽曰: “一毛微於肌膚, 肌膚微於一節,省矣. 然則積一毛以成肌膚, 積肌膚 以成一節. 一毛固一體萬分中之一物. 柰何輕之乎?”59

The beginning of the disciple’s argument resembles the original weighing arguments used by the defenders of health in the Lü shi chunqiu. Like them, Mengsun Yang invites his opponent to make consecutive choices between body parts and things, until Qinzi seems to hesitate about sacrificing a joint of his body for the possession of a state. As a response to Qin Guli’s attack, Mengsun Yang’s defense of Yang Zhu could have ended here. But it does not: what follows is a strange argument that somewhat resembles the Later Mohist adaptations of the weighing metaphors in that different body parts are now compared to each other in an almost technical manner.60 Graham considered the first part of the whole fragment “a garbled account of the doctrines of the historical Yang Zhu from a much older source, probably a document of the rival school of Mozi.”61 It is indeed rather garbled, and Graham’s extremely complicated interpretation hardly solves the problems.62 It is, moreover, burdened by his expectation of a clear division between schools of thought, of philosophical consistency over time, and of the historicity of a Yang Zhu unwilling to pull out one hair. The scanty references to Yang Zhu combined with other fragments that have survived—many of which have been quoted in this essay—do not provide much evidence for these assumptions. The Liezi passage is a complex variation on the hair trope, perhaps influenced by Mencius’s slowly rising star in the later Han dynasty and the emerging sense of lineages.63

178 | Carine Defoort

Conclusion Current views about Yang Zhu and, to a lesser extent, Mo Di have relied heavily on one statement in the Mencius. I have argued that this statement was probably largely “unfounded” or invented by him. Extant sources strongly suggest that no Yang Zhu—neither as a historical nor as a textual figure—ever refused to pull out one hair for the sake of the world, nor was there a Mo Di rubbing the hair from his whole body to benefit the world. Mencius’s double portrayal may have been two hyperboles created as a rhetorical response to some ongoing debates about, respectively, healthdefending arguments and world-benefiting ideals. This essay has provided negative evidence (the lack of textual confirmation) and tentative positive evidence (rhetorical responses to ongoing debates) to support this claim. Although this evidence can never be conclusive, I think it is stronger than the opposite view that we have inherited from the imperial past. Aside from being unfounded, Mencius was also “unfollowed” in his portrayal of Yang and Mo. The relatively few references to “hair” and “the one hair” trope in post-Mencian sources, the variety in their messages, and their apparent disinterest in his nowadays well-known portrayal all suggest that pre-imperial and early imperial sources were much less impressed by Mencius than we are. It is disconcerting to see how difficult we find it to shed his age-long legacy and how contemporary references to the early Chinese heritage as a worthy source of contemporary ethical reflection are by this very mission hindered to give their full attention to the textual evidence.

Notes   1. When talking about Mencius, Mozi, Yang Zhu, or any other “Chinese philosopher,” I refer to the personas to whom the sayings are attributed, who are not necessarily the quoted speakers or authors. The least well known of the three, Yang Zhu, was also named Yangzi 楊/陽子 or Yang Sheng 陽生. For possible names of this figure, see also Attilio Andreini, Il pensiero di Yang Zhu (IV secolo a.C.) attraverso un esame delle fonti cinesi classiche (Trieste: Edizione università di Trieste, 2000), 11–18. Andreini’s monograph is to my knowledge the most exhaustive study of Yang Zhu in Western scholarship.   2. Mencius 7A26. For the translation of the expression 摩頂放踵 (discussed later), I have followed Angus Graham, “The Right to Selfishness: Yangism, Later Mohism, Chuang Tzu,” in Individualism and Holism: Studies in Confucian and Taoist Values, ed. Donald Munro (Ann Arbor: University of Michigan, 1985), 73–84, esp. 73. All my references to primary sources, except when fragments are commonly recognized by a number (e.g., Lunyu 1.2; Mencius 3A9; Lü shi chunqiu 21/4.3), are from D. C. Lau,

Unfounded and Unfollowed  | 179 ICS Ancient Chinese Texts Concordance Series (Hong Kong: Commercial Press, 1995– ). The chapter number is given first, followed by a colon and then the page number and line number separated by a slash. For convenience, I often refer to an existing translation (“see also”) without necessarily strictly following it.   3. We could bring up indications of the cultural importance of hair as an integral part of the body in early China: for example, the filial duty to preserve one’s body totally intact, the shaving punishment as a type of corporal mutilation, the idea of self-sacrifice by cutting one’s hair, the association of cut hair (and tatoos) with barbary, the aesthetic appreciation of various types of hair (beards, whiskers, eyebrows), and the correlation of personhood and mental sanity with a pretty hairdo. See, for example, John Emerson, “Yang Chu’s Discovery of the Body,” Philosophy East and West 46.4 (1996): 533–566, esp. 548–550.   4. There is little evidence in early sources of his existence and ideas, let alone of the success that Mencius attributes to him. Aside from references to Yang Zhu in the Mencius and Liezi (discussed later), there is one in the Huainanzi (13: 123/21) and in the Lü shi chunqiu (17.7: 107/4). No figure, school, or book is mentioned in Xunzi “Fei shier zi,” Zhuangzi “Tianxia,” or Shiji “Taishigong zixu.” See, for example, Qian Mu 錢穆, Xian Qin zhuzi xinian 先秦諸子繫年 (Taibei: Dongda, 1886 [1935]), 245–246; and Graham, “The Right to Selfishness,” 73.   5. Huang Chun-chieh, for instance, notes that “not much was recorded about Yang Zhu,” but he does not therefore question Mencius’s portrayal. He rather speculates that this was perhaps “because his [Yang Zhu’s] naked individual egoism was too straightforward, or his pure individualism is shunned by the Chinese people.” See Huang Chun-chieh, Mencian Hermeneutics: A History of Interpretations in China (New Brunswick/London: Transaction, 2001), 73. See also Fung Yu-lan, A History of Chinese Philosophy, trans. Derk Bodde (Princeton: Princeton University Press, 1973 [1937]), vol. 1: 135, 137. James Legge had also stated that “doubtless it was owing to Mencius’s opposition that the foul and dangerous current was stayed.” James Legge, The Chinese Classics, with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes, 5 vols., The Works of Mencius (Taipei: SMC, 1991 [1861]), 2: 99.   6. Graham, “The Right to Selfishness,” 73. While Graham says this of “Mencius and the rest of them,” the remainder of his argument shows that he does not suggest that Mencius invented this claim of Yang Zhu.   7. It is not my intention to join the polemics about Yang Zhu’s existence, his doctrine, his dates, provenance, school, and so on. For an overview of Yang Zhu scholarship, see, for example, Graham, “The Right to Selfishness,” 74–75; and Cheng Yifan 程一凡, “Shui shi Yang Zhu?—Ting Shi huaci de 谁是杨朱—听史华慈的,” in Shi Huaci yu Zhongguo 史华慈与中国, ed. Xu Jilin 许纪霖 and Zhu Zhenghui 朱政惠 (Changchun: Jilin Chubanshe, 2008), 214–241, esp. 237–239.   8. “For moralists such as the Confucians and Mohists, to refuse a throne would not be a proof of high-minded indifference to personal gain, but a selfish rejection of the opportunity to benefit the people. They therefore derided Yang Zhu as a man who would not sacrifice a hair even to benefit the whole world.” A. C., Graham, The Book of Lieh-tzu: A Classic of Tao (New York: Columbia University Press, 1990

180 | Carine Defoort [1960]), 135–136. In my quotes from secondary sources, I have consistently adapted the transliteration to pinyin.   9. Mencius 3B9. Other fragments are Mencius 7A26 (quoted earlier), Mencius 7B26, and Mencius, 3A5. In the Ming dynasty, some scholars began to question Mencius’s harsh criticism, but they initially defended Mozi more than Yang Zhu. See Carine Defoort, “The Modern Formation of Early Mohism: Sun Yirang’s Exposing and Correcting the Mozi,” T’oung Pao 101.1–3 (2015): 208–238, esp. 227–233. 10. Mencius’s criticism of Yang and Mo is first repeated in Han sources such as Fayan 法言, Fengsu tongyi 風俗通義, and Lunheng 論衡, but they do not mention his claims about hair. For a brief history of Mencius’s portrayal of his rivals and the oldest criticism, see Defoort, “The Modern Formation of Early Mohism,” 221–224. 11. Gu Jiegang 顧頡剛, “Cong Lü shi chunqiu tuice Laozi zhi cheng shu niandai” 從呂氏春秋推測老子之成書年代, in Gushi bian 古史辯, ed. Gu Jiegang et al. (Taibei: Landeng wenhua, 1987 [1926–1941]), 4: 462–520, esp. 493–494. 12. Han Feizi 50: 151/9. Han Fei’s complaint quoted later, makes clear that “the benefit of the world” is something that this person would receive. For the various meanings of li (benefit), see also Carine Defoort, “The Profit That Does Not Profit: Paradoxes with lì in Early Chinese Texts,” Asia Major 21.1 (2008): 153–181. 13. For scholars siding with Gu Jiegang, see, for example, D. C. Lau, Mencius (London: Penguin Books, 1984 [1770]), 30; Shun Kwong-loi, Mencius and Early Chinese Thought (Stanford: Stanford University Press, 1997), 45; and Aloysius Chang, cited in Emerson, “Yang Chu’s Discovery of the Body,” 533. 14. Fung, A History of Chinese Philosophy, vol. 1, 134n2. 15. Graham, “The Right to Selfishness,” 76, 81–82n25. For his original scepsis, see Angus Graham, “The Dialogue between Yang Ju 楊朱 and Chyntzyy 禽子,” Bulletin of the School of Oriental and African Studies 22.1/3 (1959): 291–299, esp. 295. 16. See Lü shi chunqiu, 20/1.3: 128/23–31. See also John Knoblock and Jeffrey Riegel, trans., The Annals of Lü Buwei (Stanford: Stanford University Press, 2000), 512–513. 17. Graham, “The Right to Selfishness,” 75–76, emphasis added. 18. See also Cheng Yifan, “Shui shi Yang Zhu?,” 235. 19. See, for example, Shun Kwong-loi, Mencius and Early Chinese Thought, 36. 20. Scholars who identify the ideas expressed in Mencius with the dates attributed to the figure Mencius tend to consider these stories of a later date, sometimes even a later evolution of “Yangsism.” I will not assume any fixed chronology since the book Mencius and these stories (oral or written) are all of uncertain dates. For the Mencius, see, for example, E. Bruce Brooks and A. Taeko Brooks, “The Nature and Historical Context of the Mencius,” in Contexts and Interpretations, ed. Alan Chan (Honolulu: University of Hawai’i Press, 2002), 242–281; and Michael Hunter, “Did Mencius Know the Analects?,” T’oung Pao 100 (2014): 33–79. 21. These weighing metaphor stories are gathered in Lü shi chunqiu, “Ben sheng (Taking life as basic)” 本生, “Zhongji (Valuing the self)” 重己, “Gui sheng (Honoring life)” 貴生, “Shen wei (Being attentive to aims)” 審為, with much overlap with Zhuangzi, “Rang wang (Abdicating the throne)” 讓王. 22. See Carine Defoort, “Heavy and Light Body Parts: The Weighing Metaphor in Early Chinese Dialogues,” Early China 38 (2015): 55–77.

Unfounded and Unfollowed  | 181 23. See, for example, Benjamin Schwartz, The World of Thought in Ancient China (Cambridge: Harvard University Press, 1985), 259. 24. For the similarity between Mencius’s and Mozi’ thought, see, for example, Gu Jiegang, “Cong Lü shi chunqiu tuice Laozi zhi cheng shu niandai,” 494; and Hsiao Kung-chuan, A History of Chinese Political Thought, trans. Frederick Mote (Taibei: Caves Books, 1980), 231–332. 25. See Jörg Schumacher, Über den Begriff des Nützlichen bei Mengzi (Bern: Peter Lang, 1993), 22, 112–114. 26. Emerson, “Yang Chu’s Discovery of the Body,” 549, adds “from his leg” in his interpretation of Yang Zhu’s supposed unwillingness to pull one hair. 27. See Schwarz, The World of Thought, 178–179. 28. Arthur Waley described “the polemic references to Yang Zhu in Mencius  .  .  .  mere parody.” See Arthur Waley, trans., The Way and Its Power. A Study of the Tao Te Ching and Its Place in Chinese Thought (New York: Grove Press, 1958), 34n1. See also Emerson, “Yang Chu’s Discovery of the Body,” 549; Andreini, Il pensiero di Yang Zhu, 57; Attilio Andreini, “The Yang Mo 楊墨 Dualism and the Rhetorical Construction of Heterodoxy,” Asiatische Studien 68.3 (2014): 1–59, esp. 27. 29. Schumacher, Über den Begriff, 150–152, speculates that Mencius might have been inspired by the portrayal of Yu by the Later Mohists. The core chapters of the Mozi (chaps. 8–37) are usually dated from the early fourth until third century BCE. 30. Nor is Yang Zhu’s unwillingness to pull out one hair ever criticized, even though I believe that some late Mozi chapters might have been influenced by the health defenders’ weighing metaphor. See Mozi 42: 83/6–9, Mozi 44: 92/20, and Mozi 47: 103/23–26. These chapters are generally dated around the late fourth and third century BCE. For their adaptation of the weighing metaphor, see Defoort, “Heavy and Light Body Parts,” 75–77. The conventional explanation why the defense of health is not mentioned in the book Mozi is that the presumed historical Yang Zhu is traditionally dated after the presumed historical figure Mo Di. 31. A core chapter “Jian ai, zhong” (兼愛中) describes Yu as hard-working to benefit the people in the whole known world but not by sacrificing any body part for it, not even hair. See Mozi 15: 26/17–20 and Knoblock and Riegel, Mozi, 153–154. This chapter is generally dated around 350 BCE. See, for example, John Knoblock and Jeffrey Riegel, Mozi: A Study and Translation of the Ethical and Political Writings (Berkeley: University of California Press, 2013), 34. 32. See Shizi 2: 42: 16/7; Shiji, “Qin shi Huang benji” 秦始皇本紀”; Shiji “Li Si liezhuan” 李斯列傳, and Shiji, “Sima Xiangru liezhuan” 司馬相如列傳; Huainanzi, “Yao Lüe” 要略; Hanshu, “Sima Xiangru liezhuan” 司馬相如列傳; Kongcongzi, “Ju Wei” 居 衛; Liezi, “Yang Zhu”; and possibly the oldest among them: Rong Cheng shi 容成氏, a manuscript from the Shanghai Museum collection purchased from the black market and tentatively dated around 300 BCE. Yuri Pines dates this manuscript in the fourth or early third century BCE. See Yuri Pines, “Political Mythology and Dynastic Legitimacy in the Rong Cheng shi Manuscript,” Bulletin of the School of Oriental and African Studies 73.3 (2010): 503–529, esp. 504. 33. Zhuangzi 33: 98/19–29 states that Mozi declares Yu his hero for dedicating his life to the harsh labor of stemming the floods, cutting channels, and opening

182 | Carine Defoort hydraulic communications in the world. Mohists emulated Yu and demanded the same physical sacrifices of their followers. 34. In the Mencius, Emperor Yu is always praised. So are his hydraulic endeavors for the sake of the people. They are never related to the loss of hair or any other physical damage. The argument is rather that Yu followed nature. See, for example, Mencius 3A4, 3B9, 4B26, 4B29, and 6B11. 35. I am unsure here about the translation of 之 (or the variant 趾), but I have followed Pines, “Political Mythology and Dynastic Legitimacy,” 511–512. 36. Zhuangzi 11: 27/3. 37. “As for Yi Yin’s looks, his face had neither beard nor eyebrows, Yu was lame, and Tang was limping. Yao and Shun had irregular pupils” (伊尹之狀, 面無須麋; 禹跳湯 偏; 堯舜參牟子). Xunzi 5: 17/22–24. See also John Knoblock, trans., Xunzi: A Translation and Study of the Complete Works (Stanford: Stanford University Press, 1988), 1: 204. 38. Zisi concludes that, compared to virtue, body hair is a trivial matter: “A person’s worth and sagacity lies in his virtue. How could it be a matter of looks?” (人之賢聖在德, 豈在貌乎). Even his own “late Master had no beard or eyebrows” (先君生 無鬚眉). Kongcongzi 2.4: 14/1–2. See also Yoav Ariel, K’ung-ts’ung-tzu: The K’ung Family Masters’ Anthology (Princeton: Princeton University Press, 1989), 107–108. 39. To consider all these Mohist influence suggests that there initially was a clear and systematic association of these ideas with Mohism. I have challenged this by tracing the occurrence of “care for all” (兼愛 jian ai) in Warring States and Han texts. See Carine Defoort, “Do the Ten Mohist Theses Represent Mozi’s Thought? Reading the Masters with a Focus on Mottos,” Bulletin of the School of Oriental and African Studies 77.2 (June 2014): 337–370, esp. 367. 40. Yang Bojun is puzzled by the expression 摩頂放踵, disagrees with the commentary of Zhao Qi explaining 放 as 至 (till), speculates that this was perhaps a saying in Mencius’s days, and leaves the translation vague. See Yang Bojun 楊伯峻, Mengzi yizhu 孟子譯註 (Hong Kong: Zhonghua shuju, 1997), 313n2. I have followed Graham, “The Right to Selfishness,” 73. Lau, Mencius, 188, translates “shaving his head and showing his heels.” 41. Franklin Perkins, “The Mozi and the Daodejing,” Journal of Chinese Philosophy 41 (2014): 18–32, esp. 19. Qian Mu, Xian Qin zhuzi xinian, 246, pointed out Mencius’s lack of followers. 42. Huainanzi 2: 12/11–12; see also Major, The Huainanzi, 90–91. 43. Kongcongzi 10: 20/18–19; see also Ariel, K’ung-ts’ung-tzu, 125. 44. Han Feizi 50: 151/9–10; see also W. K. Liao, trans., The Complete Works of Han Fei Tzu, 2 vols. (London: Arthur Probsthain, 1959), 2: 301–302. 45. This expression is indeed commonly used by the health persuaders: see, for example, Lü shi chunqiu “Ben sheng” (1.2.5), “Gui sheng” (2.2.4), “Shen wei” (21.4.4), and in the parallel persuasions in Zhuangzi, “Rang wang.” My numbering of Lü shi chunqiu fragments follows Knoblock and Riegel, The Annals of Lü Buwei. 46. Mozi is not mentioned, but Yao’s description resonates with that of Sima Tan’s description of the Mohists in Shiji “Taishigong zixu” 太史公自序. 47. Han Feizi 49: 145/28–146/1. See also Liao, The Complete Works of Han Fei Tzu, 2: 277–278. A similar point is made, but without any reference to hair, by Rober Zhi in Zhuangzi 29: 90/14–18.

Unfounded and Unfollowed  | 183 48. Adding that “in general, the reason for one to be noble enough to possess the world is that he can give full rein to his wishes and let his desires run wild” (凡所為貴有天下者, 得肆意極欲) Shiji 6: 271. Quotes from dynastic histories are to the Zhonghua shuju edition of 1992. See also William Nienhauser, The Grand Scribe’s Records (Bloomington: Indiana University Press, 1994), 1: 159–160. The same view is recorded in Liezi 7: 42/6–7 quoted earlier. 49. “His thighs and shins had no more hair; his hands and feet were callused and his face blackened. He then died far away from home and was buried on Mount Kuaiji. The labor of a war prisoner was not tougher than this” (股無胈, 脛無毛; 手足胼 胝, 面目黎黑. 遂以死于外, 葬於會稽. 臣虜之勞不烈於此矣). The emperor responded: “When a worthy man possesses the world, he exclusively uses it to suit himself and that is it. This is how he is honored by possessing the world. The one who is called a worthy man must be able to secure the world and govern the ten thousand people. Now if he personally is not able to benefit from it, how could he then govern the world?” (彼賢人之有天下也, 專用天下適己而已矣. 此所以貴於有天下也. 夫所謂賢人者, 必能安天下而治萬 民. 今身且不能利, 將惡能治天下哉?) Pandering to the emperor’s whims, Li Si hastened to express his full agreement by quoting Shen Buhai: “To possess the world and yet not throw off all restraints is called ‘shackling oneself with the world’ ” (有天下而不 恣睢, 命之曰以天下為桎梏). Yao and Yu wearied their bodies while serving the people and therefore deserve to be called “shackled” (or perhaps 至梏 “extremely emaciated,” which resonates with 桎梏; the term gu 梏 is often used in the stereotypical description of Yu’s deformity). Li Si therefore wonders “what is worth honoring in this?” (何足貴哉?). Shiji 87: 2553–2555; see also William Nienhauser, The Grand Scribe’s Records (Bloomington: Indiana University Press, 2006), 7: 347–349. For Shen Buhai, see Herrlee Creel, trans., Shen Pu-hai. A Chinese Political Philosopher of the Fourth Century B.C. (Chicago: University of Chicago Press, 1974), 381. 50. Shizi 2.59: 18/3–4; see also Paul Fischer, Shizi: China’s First Syncretist (New York: Columbia University Press, 2012), 123. 51. In a letter defending Emperor Wu’s policy, he compares the ruler with the Great Yu during the great flood, when people were moving around and found no security: “[Yu] of the Xia worried about this; he blocked the flood, opened up the Yangzi Stream, conducted Yellow River, divided and deepened the beds to avert disasters. He diverted it all eastward to the ocean and the world was forever at peace.  .  .  .  Because his heart was troubled with concern, he personally labored for it. His body was calloused and without down, his skin did no longer grow hair. Thus, his merits were visible for eternity, and his reputation continues till today. (夏后氏戚 之; 乃堙鴻水, 決江疏河, 漉沈贍菑. 東歸之於海, 而天下永寧.  .  .  .  心煩於慮而身親其勞, 躬胝無胈, 膚不生毛. 故休烈顯乎無窮，聲稱浹乎于茲). Shiji 117: 3050; see also Burton Watson, trans., Records of the Grand Historian of China, 2 vols. (New York: Columbia University Press, 1961), 327. See also in Hanshu 57B: 2585. Yu deserved the throne by hard labor and concern for the people. As does, by implication, Emperor Wu. 52. See, for example, Xunzi 10: 47/5–6, playing on the ambiguity of the verb li 利: “One who benefits them and does not benefit from them, who cares for them but does not use them, will get the world” (利而不利也, 愛而不用也者 取天下矣). 53. Some scholars consider it two consecutive passages instead of one. Men Qiming called them the two “for oneself” (wei wo 為我) fragments and believed

184 | Carine Defoort that they preceded Mencius, who had quoted them. See Men Qiming 門啟明, “Yang Zhu pian he Yangzi zhi bijiao yanjiu” 楊朱篇和楊子之比較研究, in Gushi bian 古史辯, 4: 592–610, esp. 596. 54. From Han (but mostly Song) till Qing, he was criticized along with Mozi as a “heretic”; from the twentieth century onward he was appreciated as a “philosopher.” Graham explained the “importance” of this fragment by its being “the longest of the few direct accounts of Yang Zhu’s teaching, and the only record of a dispute between his school and that of Mozi.” See Graham, “The Dialogue between Yang Ju 楊朱 and Chyntzyy禽子,” 299. 55. Aside from Men Qiming mentioned earlier, and Graham, see also Waley, The Way and Its Power, 34n1. He believed that this fragment was a “late hearsay account in which the original teachings of Yang Zhu are mixed up with the hedonistic doctrine which grew up in one branch of this school.” 56. Graham, The Book of Lieh-tzu, 136. In an earlier paper, he had speculated that this lost source had to be “earlier than the final victory of Confucianism and Taoism over the ‘Hundred Schools’ during the Former Han dynasty,” perhaps indirectly through the now-lost Shizi. See Graham, “The Dialogue between Yang Ju 楊 朱 and Chyntzyy禽子,” 298. 57. Liezi 7: 41/18–20; see also Graham, The Book of Lieh-tzu, 148–149. 58. See Graham, The Book of Lieh-tzu, 149n1. 59. Liezi 7: 41/23–25; see also Graham, The Book of Lieh-tzu, 149. 60. In Mozi “Jingshuo shang” 經說上, Mozi “Choosing the Greater” 大取, and Mozi “Valuing Justice” 貴義. See Defoort, “Heavy and Light Body Parts,” 75–76. 61. Graham, The Book of Lieh-tzu, 149n1. 62. Graham’s alternative translation is not complete and indeed poses problems. His attribution of the fragment to a Mohist author wavers between “evidently” (136) and “probably” (149). He does not convincingly date the Mohist source, nor does he explain what constitutes the distortion by the Mohist. As for the attribution of this fragment to a Mohist author, Graham gave four reasons in an earlier paper, but none is very conclusive: (1) Mozi’s student, Qin Guli, is referred to as a master, namely Qinzi (禽子), while Yang Zhu is also called by his name; (2) Qin Guli twice has the last word in the second fragment; (3) in the first fragment, Yang Zhu first refers to the Mohist model emperor Yu and then to Bocheng Zigao; (4) the similarity with a fragment about choices in Mozi 47. See Graham, “The Dialogue between Yang Ju 楊 朱 and Chyntzyy 禽子,” 297. My brief response to these arguments is: (1) Yang Zhu is twice called Yangzi 楊子 and calling him by his name does not necessarily indicate a lack of status; (2) in fact, they agree to disagree; (3) reference to Emperor Yu and to Bocheng Zigao (his critic in a Zhuangzi story) is not a sign of Mohist authorship; (4) the fragment in Mozi 47 is probably an adaptation of the health-defending stories in four Lü shi chunqiu chapters. 63. The “Yang Zhu” passage seems to reflect or take part in the construction of lineages, which is not surprising in a post-Han context: on the Mohist side, the Great Yu is mentioned as authority, Mozi as master, and Qin Guli as disciple and master (Qinzi). On the other side, there is Bocheng Zigao mentioned as authority, Yang Zhu as master, Mengsun Yang as disciple, and Guan Yin with Lao Dan as authorities.

9

Reconstructing A. C. Graham’s Reading of Mencius on xing 性 A Coda to “The Background of the Mencian Theory of Human Nature” (1967)

Roger T. Ames

Sharing a Hobby Horse One of the topics on which Angus Graham is still most cited in the scholarly literature is his interpretation of the Mencian notion of “human nature” (xing 性). Graham first wrote a seminal essay in 1967 that is referenced in the title of this essay and then returned to ride this issue time and again in his published oeuvre. In this essay I propose to follow his thinking on this topic as it evolved across his academic career, and as it inspired a debate among many other scholars, including a prolonged and delightful exchange I had in book chapters and journal articles with the wonderful and fondly remembered Irene Bloom.1 In this essay I will once again mount this Graham hobby horse and take it for one more ride around the track.

Setting the Problem Historically, Mencius, in his concerted attempt to formulate the conditions required to achieve the consummate human experience, has not been well served by his commentators. Xunzi with his own agenda is perhaps the first philosopher remembered by history who gives Mencius’s notion of xing

185

186  |  Roger T. Ames (性)—conventionally translated as “human nature”—a skewed, naturalistic reading that is not only not Mencius’s own, but an understanding of xing Mencius flatly rejects. Following Confucius in arguing for the assiduous effort it requires to achieve moral competency, Xunzi caricatures Mencius’s position as being otherwise: Mencius says that “it is because human beings can learn that their xing is good.” I would argue that this is not so. Mencius’s understanding of the human xing being itself wanting, he was unable to register the distinction between xing and deliberate activity. In general terms, xing is what is given by nature; it can neither be learned nor acquired. Moral dispositions such as achieving propriety in our roles and relations (li 禮) and seeking what is most appropriate in any situation (yi 義) are the products of the sages, and hence are something that can be learned and applied, acquired and mastered. What cannot be learned and cannot be acquired but simply is inherent in persons is what is called xing.2 Xunzi assumes here that his own understanding of xing as being an unlearned “given” is what xing actually means. And hence according to Xunzi’s logic, when Mencius says “it is because human beings can learn that their xing is good,” Mencius is claiming that being good is both unlearned and learned at the same time because we all know that xing is unlearned. Of course, this interpretation of xing that Xunzi is imputing to Mencius—what you are born with is xing—is explicitly rejected by Mencius as a reductio ad absurdum: “It is what you are born with (sheng 生) that is meant by xing (性),” said Master Gao. “Is saying that ‘what you are born with is what is meant by xing’ the same thing as saying that ‘white’ is what is meant by ‘white’?” asked Mencius.3 “Indeed it is,” replied Master Gao. “Then is the whiteness of white feathers the same as the whiteness of snow, and the whiteness of snow the same as the whiteness of jade?” “Yes, indeed,” replied Master Gao. “And then it follows that the xing of a dog is the same as the xing of an ox, and the xing of an ox is the same as the xing of a person?”4

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 187 Despite Mencius’s rejection of this tautological naturalism—that is, humans are good because they are good—it is just such an impoverished reading of Mencius—xing is an unlearned given that makes us good—that not only persists but in fact prevails among commentators even today. The expression xingshan (性善) is a claim that in the contemporary interpretive literature on Mencius—both Western and Chinese alike—has almost ubiquitously been understood as meaning that “human nature is good.” And as Michael Sandel, in his Liberalism and the Limits of Justice, pointedly observes at least for Western commentators: “To speak of human nature is often to suggest a classical teleological conception, associated with the notion of a universal human essence, invariant in all times and places.”5 Indeed, the uncritical assumption is that for Mencius xing references a universal, inborn, fixed, self-sufficient endowment that is defining of all human beings and that programs us naturally as human beings to be good in what we do. Lee Yearley in his comparison of Mencius with Aquinas refines this notion of a given human nature by making a distinction between an ontological or discovery model of xing that he would reject as a familiar yet mistaken interpretation of both Mencius and Aquinas, and a development or biological model that he would embrace as the right way of reading them. Of the former, Yearley says: In a discovery model . . . human nature exists as a permanent set of dispositions that are obscured but that can be contacted or discovered. People do not cultivate inchoate capacities. Rather they discover a hidden ontological reality that defines them. The discovery model reflects, then, ontological rather than biological notions. An ontological reality, the true self, always is present no matter what specific humans, particular instances of it, are or do.6 Yearley instead advocates for the latter developmental or biological understanding of the Mencian xing that he describes as follows: What can be called a biological framework informs Mencius’s ideas on human nature and its characteristic successes and failures. . . . To speak of the nature of something within such a framework is to refer to some innate constitution that manifests itself in patterns of growth and culminates in specifiable forms.7 While the developmental model of xing is more compelling than the discovery model, it is still strongly Aristotelian in its highly teleological understanding of human potential as the innate defining “capacities humans

188  |  Roger T. Ames possess”8 that are manifested and actualized in determining who we will become. What recommends this developmental model to Yearley and others is Mencius’s frequent appeal to a horticultural and husbanding analogy in which, for example, seeds of barley if uninjured and nurtured, will achieve the characteristic form of barley.9 Indeed, in the Mencius and other canonical Confucian texts, the familiar appeal to the horticultural and husbanding metaphors is often construed as reinforcing the idea that plants and animals, in growing to become what they essentially are, are simply actualizing the potential inherent in their “seed” or “root.” Given our own interpretive context in which substance ontology is common sense, teleology is our default assumption. But what in fact makes these horticulture and husbanding analogies appropriate for capturing the growth of relationally constituted “human becomings” is the acute dependence of farming and the raising of animals upon a contrived environment and upon concentrated human effort. Without sustained, radical intervention, most seeds, far from becoming what they “are,” become anything and everything else. Without the benefit of intensive intervention and cultivation on behalf of what we think they will “naturally” become, most acorns become squirrels, most corn becomes cows, most eggs become omelets, and most apples become compost. Indeed, only one in a million acorns becomes an oak tree. The “root” or “seed” of anything and what it will become is as much a function of intense cultivation together with the contingencies of circumstances as it is of the initial conditions from which it “begins.” We might look to the term “culture” and the metaphors in which it is embedded within our own narrative to register the assumptions we bring with us in making horticulture and husbanding an apposite metaphor for the production of culture. Indeed, it is because we default to these same assumptions that the essentialist interpretation of Mencius on xing persists as the prevalent understanding. In his Keywords: A Vocabulary of Culture and Society (1976), Raymond Williams famously describes “culture” as one of the two or three most complicated terms in English.10 He attributes this in part to the relative recency with which the meaning of “culture” has been metaphorically extended from its original sense of the physical processes of nurturing and cultivation—that is, the practices of horticulture and husbandry—to point toward a characteristic mode of material, intellectual, spiritual, and aesthetic development. Just as our common sense dictates, we see these practices as teleologically driven in bringing to fruition characteristic forms inherent in the object of cultivation where human intervention serves as both a source of discipline and control, and as an external facilitation. The assumption is that the plant or animal flourishes if it is unimpeded and properly nourished.

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 189 It was only in the eighteenth century that “culture” was first used consistently to denote the entire “way of life” of a people, and only in the late nineteenth and early twentieth centuries that it was identified with specific civilization-distinguishing patterns of practices and values. In this latter case, it was used in the context of theories of progressive “social evolution” as something that sets apart and divides societies, making one more advanced than another. One contemporary legacy of this emerging sense of contest is the contemporary media’s frequent characterization of multicultural tensions in education as “culture wars.” As in premodern European languages, there was no single term in premodern Chinese and Japanese that had a conceptual reach comparable to that of contemporary uses of the word “culture.” But the term that came to be used throughout the geographical region to appropriate and translate this emerging modern concept differs markedly in its metaphorical implications from those assumed with the English word “culture.” While Chinese and Japanese abound with words that, like “culture,” are rooted in instrumental physical processes of cultivation and nourishing (e.g., xiu 修, yang 養, xu/chu 畜), these terms are bypassed as points of metaphorical departure in favor of wenhua (文化)—a compound word that combines the characters for “the inscribing and embellishing of the literary/civil/artistic traditions” (wen) and “transforming” (hua). Whereas metaphorically rooting “culture” in practices of plant and animal domestication invites seeing cultural norms as having a transcendent disciplinary force with respect to that which is being “cultured,” wen was understood (with significant political implications) as collaborating with nature and elaborating upon it rather than as regulating its spontaneous growth. As is demonstrated by its use in texts dating to the Han dynasty (202 BCE–220 CE), the term wenhua is an ancient one, and the idea is even earlier. Wenhua is a Japanese kanji term derived from classical Chinese that is first attested in the court bibliographer Liu Xiang’s (劉向 77–6 BCE) Shuoyuan (說 苑): “It is only when civilizing efforts do not bring them up to standard that punishments are to be imposed.”11 And, by at least the fifth century, Chinese literary theorists such as Liu Xie (劉勰 465?–522?) identified wen explicitly with the self-arising (ziran 自然) and ceaselessly creative (shengsheng buxi 生 生不息) dynamics of the natural world (dao 道), affirming that nature and nurture were not opposed, but rather constituted a coevolving, contrapuntal process understood to be at the heart of realizing a symbiotic, natural and societal, harmony. Moreover, in sharp contrast with the contemporary use of “wars” as a metaphor for cultural tensions, from ancient times wen was contrasted explicitly with the destructive and dehumanizing use of military force (wu 武). Wen instead denotes the expansively civil dimension of the human experience that emerges when the life of a community is guided

190  |  Roger T. Ames by an aesthetically and critically enriching counterpoint between persistent canonical texts and those continuing commentaries that have been crafted in each generation as a response to current conditions. In sum, the conceptual genealogy of wenhua implies that culture emerges through an intrinsic relationship between persistence and change (biantong 變通)—a relationship between tradition and transformation in which cultural conservation and prospective change, far from being opposed, are complementary and mutually enhancing. Indeed, it is this complementary, contrapuntal dynamic that has immediate relevance for A. C. Graham’s revisionist reading of Mencius on xing.

Angus Graham and the Narrative Interpretation of xing In this essay, “we” (because I include my collaborators David Hall and Henry Rosemont as embracing this same effort) want to celebrate the sustained attempt of our teacher, colleague, and friend, Angus Graham, to save Mencius from the familiar essentialist misreadings of xing—both the discovery and the developmental models—by pointing the commentarial tradition in the direction of a possible third position. Graham offers us what we would call a “narrative” understanding of xing in which person and world evolve together in a dynamic, contrapuntal relationship. The identities of persons are certainly grounded in the native beginnings of family, community, and environing relationships that need to be both nurtured and protected from loss or injury, but such identities only emerge in the process of these relationships achieving thick resolution as they are cultivated, grown, and consummated over their lifetimes. Their potential far from being a given, in fact, emerges contrapuntally in the always transactional events that in sum constitute lives lived in the world. That is, the “potential” for becoming human is not simply the “beginnings”—something inborn “within” the persons exclusive of context and family relations. In the first place, in this natural cosmology, there are no such deracinated persons or such a metaphysical understanding of how anything begins. Persons do not exist inside their skins; they exist only in their associations. And since persons in their nested, narratives-within-narratives are constituted by these evolving, eventful relations, the “potential” of persons and their achieved identities in fact emerge pari passu from the specific, contingent transactions of their lives. Thus, the best sense we can make of “potential” here is: Potential, rather than being wholly antecedent as a set of given conditions, is also prospective and evolves with ever-changing circumstances; rather than being generic or universal, it is always unique to the career of this specific relational person; and rather than existing simply as an inherent and defining endowment, it can only be known post hoc after the unfolding of the particular narrative.

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 191 What is at stake here is Mencius’s answer to perhaps our most basic and important philosophical question: How should we understand what it means to become fully human? How do we explain birth, life, and growth of the human “being”?—by reduplicative causal accounts (the infant is a ready-made adult), by teleological accounts (the infant is simply preliminary to the existing ideal), or as a human “becoming” that appeals to a contextual, narrative account available to us through a phenomenology of deliberate personal action? How do we define what it means to be a human “being”?—by speculative assumptions about innate, isolatable causes that locate persons outside of the roles and relations in which they live their lives, or alternatively, as having “become” human by taking account of the initial, native conditions and context within which persons are inextricably embedded, and then by assaying the full aggregation of consequent action as their life stories unfold? In our own world in which individualism has become an ideology in the sense that it has a monopoly on intellectual conscious without any robust alternatives, we must ask whether our own seemingly default commonsense assumptions about individual human “beings” are consistent with the Mencian project as it was situated and developed within the natural qi cosmology that serves this tradition as the context for such personal growth.

Graham’s Initial Developmental Understanding of the Mencian xing (性) But Graham did not come to the narrative understanding of xing easily. Early on he had endorsed the developmental model of Mencius on xing by associating it quite specifically with an Aristotelian understanding of a given human potential and its actualization. In discussing the Mencian notion of xing, Graham allowed that we can best understand Mencius by recognizing the fact that . . . Aristotle was pursuing a similar line of thought with the much greater logical rigour of the Greeks. Indeed it is difficult to write of hsing [xing] in English without resorting to the Aristotelian terminology of “potentiality” and “actualization,” and the former word has several times slipped into the present study.12 Graham, appealing to the Zhongyong for textual proof to support this essentialist reading of Mencius, argued that xing denotes the self-sufficient and self-completing nature of things: Each thing has its nature, and “becomes complete” (成 ch’eng [cheng]/*DIENG) by fulfilling the capacities of its nature. In man

192  |  Roger T. Ames this state of maturity, by which we act wholeheartedly according to our nature and become in the full sense men, is 誠 ch’eng [cheng]/*DIENG “wholeness, integrity,” defined in Chung Yung [Zhongyong] by “Integrity is self-completion” (誠者自成也).13 Of course, were Graham to have cited this same Zhongyong passage in its entirety, he would have had to allow that this text, far from corroborating his claim, is making precisely the opposite point: that is, cheng (誠) cannot mean merely “self-completion.” Indeed, this same Zhongyong passage goes on to categorically contradict the opening, patently rhetorical phrase that Graham has appealed to here in making his case for an essentialist reading of xing: But cheng 誠 is NOT simply the self-completing of one’s own person; it is what consummates all things. Consummating oneself is achieving virtuosity in one’s roles and relations (ren 仁); consummating all things is living wisely (zhi 知). This is a virtuosity achieved by xing (性) and is what integrates the more internal with the more external. Thus wherever and whenever this virtuosity is invoked, it is fitting.14 If we want to translate cheng (誠) as “integrity” as Graham chooses to do here, we might observe that in Aristotle’s substance ontology, the integrity of persons—the eidos that is immanent in matter and directs the entire teleological structure of individual existents—guarantees self-actualization through the realization of a given set of capacities and their potential to produce a human “being” as a final cause. But in the Chinese process cosmology that serves the Zhongyong as its interpretive context, such “integrity,” far from being a self-sufficient given or a final cause, would also have to entail full integration of persons with their irretractable social and natural environments. That is, the possibilities and potential of any person would emerge within the transactions that unfold pari passu in the narrative of persons and their worlds becoming-one-together. The Zhongyong in this passage clearly disputes any suggestion that persons can be “self-completing” in some exclusive sense and quite specifically defines the “virtuosity of xing” itself as what coordinates the external and the internal and allows all things to find their fit and to coalesce productively with each other as we live wisely in the world. Far from defining cheng (誠) as “self-completion,” the text argues that such personal cultivation is invariably a collaborative attempt to optimize the possibilities available at the vital intersection between purposeful persons and their environing conditions, where cheng gives human beings the capacity to both shape

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 193 their world and to respond efficaciously to the pressures of this world as it shapes them. The Zhongyong gives a clear account of such collaboration between person and world: Those of the greatest virtuosity (dade 大德) are certain to gain status, emoluments, reputation, and longevity. For the generosity of nature (tian 天) in giving birth to and nurturing things is certain to be in response to the quality of the things themselves.15

Graham’s Evolving Understanding of xing But Graham in his later writings becomes dissatisfied with and indeed repudiates his earlier essentialist reading of xing. Reflecting on the evolutionary turn that brought him to a different, more situated understanding of xing, he offers us this novel, revisionist interpretation: Waley says that “hsing [xing] (nature) meant in ordinary parlance the qualities a thing has to start with,”16 and I have myself in previous publications translated the definition 生之謂性 by “Inborn is what is meant by ‘nature.’ ” Yet early Chinese thinkers who discuss hsing [xing] seldom seem to be thinking of fixed qualities going back to a thing’s origin  .  .  .  rather they are concerned with developments which are spontaneous but realize their own potentialities only if uninjured and adequately nourished. Mencius in particular seems never to be looking back towards birth, always forward to the maturation of a continuing growth.17 Graham here clearly wants to distinguish himself from any discovery or ontological model of xing and introduces language that might seem to favor the developmental alternative. But to understand his interpretation more clearly, such a prospective, processual, and developmental interpretation of the maturation of an “uninjured and adequately nourished” xing needs to be located within Graham’s understanding of early Chinese process cosmology more broadly. For Graham, proper attention to the interdependence and irreducibly contextual nature of all things will enable us to disambiguate some of the central philosophical vocabulary of classical Chinese philosophy such as tian (天) and dao (道) by identifying equivocations that emerge when we elide cosmological presuppositions indigenous to the classical Chinese worldview with our more familiar classical Greek ontological assumptions. Graham cautions us about such frequent equivocations, using as his specific example

194  |  Roger T. Ames our tendency to treat xing as referencing some “transcendent” and hence independent existent: In the Chinese cosmos all things are interdependent, without transcendent principles by which to explain them or a transcendent origin from which they derive.  .  .  . A novelty in this position which greatly impresses me is that it exposes a preconception of Western interpreters that such concepts as Tian “Heaven” and Dao “Way” must have the transcendence of our own ultimate principles; it is hard for us to grasp that even the Way is interdependent with man. . . . the translation of hsing [xing] by “nature” predisposes us to mistake it for a transcendent origin, which in Mencian doctrine would also be a transcendent end.18 Graham, rejecting the relevance for Mencius of an essentializing Greek idealism (eidos), and the radical teleology that follows from it, would locate his notion of xing within the generic features of an early Chinese process or “event” ontology in which putative “things” and their contexts are interdependent and thus inseparable. What it means to become human, far from being an antecedent given that takes us back to our origins, is in fact provisional and emergent within the context of an evolving cosmic order. It is such a worldview that we and our collaborators following Marcel Granet, Joseph Needham, and Graham himself have argued for at length as the most appropriate interpretive context for understanding classical Confucianism.19 For Graham xing as a dynamic, gerundive concept references a developmental process that is spontaneous and realizes its own potentialities when it is nourished and unimpeded. But given that Graham’s understanding of early Confucian cosmology assumes a doctrine of internal relations constitutive of putative “things” rather than external relations that merely conjoin discrete and independent things, his interpretation of “its own potentialities” would make them historicist, particularist, and genealogical. Graham in his clarification of the nature of “relations” that is relevant to this Chinese cosmology introduces precisely this distinction: As for “relationships,” relation is no doubt an indispensable concept in exposition of Chinese thought, which generally impresses a Western as more concerned with the relations between things than with their qualities; but the concern is with concrete patterns rather than relations abstracted from them . . .20 Again, Graham has already cautioned us of a fundamental difference in defining the terms of art available for theorizing classical Chinese cosmol-

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 195 ogy that “accords with one’s general impression when groping towards an understanding of early Chinese concepts, that often they tend to be more dynamic than their closest Western equivalents, and that English translation freezes them into immobility.”21 It is only recently over the past century that we have called into question the assumption that concepts can serve us as a univocal currency to guarantee the cogency of our arguments, and with the insights of contemporaries such as Wittgenstein and more recently Lakoff and Johnson, have come to see concepts as irreducibly dependent upon context. Graham remarks on this transition: “We are losing the faith, except in logic and mathematics, that a concept can be established by precise definitions which free the word from the analogies which guide its ordinary usage.”22 Such a difference in our assumptions about the nature of concepts themselves as we look for cultural equivalencies might be interpreted as arising from a contrast between a substance ontology that naturalizes form and stasis, and thus favors the more stable, decontextualizing noun on the one hand, and a dynamic process cosmology that assumes the inseparability of a rhythmic, contrapuntal forming and functioning (tiyong 體用) and thus favors the contextualizing gerund (or verbal noun) on the other. We have seen that Graham describes xing as “spontaneous process with the direction continually modified by the effects on it of deliberate action rather than of its origin or goal.”23 In damping the assumed retrospective limits of a given teleology and invoking the more important role of deliberate, prospective action, he offers us an understanding of xing as a vital and purposeful conatus that in its growth is self-directing and yet collaborative, as in its narrative it shapes and is shaped by its environments. In Graham’s own words: Still riding my own hobby horse, might one distinguish xin “heart” and xing “nature” as the centres of awareness and spontaneity respectively?  .  .  .  With the exercise of the heart for thinking, spontaneous inclination shifts away from the direction in which it is pulled by the mere action of the perceived thing on the senses. . . . In unawareness you know only present inclination; it is by full exertion of the heart to understand the things which act on you that you come to know what Heaven has ordained as the spontaneous preference of your nature in full awareness of them.24 Such a sense of radically situated awareness and creativity is lost when we understand xing (性) as referencing only “natural endowment at birth” rather than including also the life and growth of these native conditions within their contextualizing relationships, and we understand sheng (生) as

196  |  Roger T. Ames simply “birth” rather than as “birth, growth, life, and vitality.” Given that xin is an integral aspect of xing (性 as 心 + 生), xing would seem to be the vital and purposeful narrative of life, growth, maturation, and vitality of those native conditions captured in the inchoate xin as these preliminary constitutive relations in family and community captured in the “four beginnings” (siduan 四端) formula of renyilizhi (仁義禮智) are thickened to alter and be altered in their evolving encounter with their environing conditions.

Familiar Misreadings of Mencius on xing One familiar misunderstanding of Mencius on xing occurs when on the analogy of default metaphysical conceptions of “human nature” we decontextualize and detemporalize xin, and following either the discovery or the developmental model of xing, treat such generalized features captured in the expression “four beginnings” (siduan) as innate, internal, fixed, independent, and isolatable conditions. But xin, far from being the inborn and exclusive nature defining of discrete individuals, is the concrete pattern of constitutive relations that serve as a general description of the native beginnings of unique yet radically situated human “becomings” available to the process of further nurturance, growth, and articulation. Attributing a notion of an essential “human nature” as a principle of individuation to Mencius is a misreading of xin on at least two counts. First, it is certainly counterintuitive within the context of a holistic qi (氣) cosmology that begins from the primacy of process and vital relationality, and in which everything is vulnerable to transformation. Xin (心) in fact references the initial conditions and tendencies of a human “becoming” captured in the aspectual language of renyilizhi that locates us in our constitutive roles and relations, and in those relations, incline us to efficacious growth (shan 善). “Acting with wisdom” (zhi 智), for example, is not first “having” wisdom and then applying it to a situation but is rather a quality of acting that arises gradually within the efficacy of one’s actions themselves. Similarly, xin is no more an essential and inborn “given” than is the aspectual ren (仁) that, as first an incipient relational disposition, in the fullness of time evolves to become an acquired consummatory virtuosity in one’s roles and relations. Both are fairly described as native inclinations that in the fullness of time, through the continuing articulation of what were initially only tentative dispositions, evolve to become an achieved quality in the deliberate growth of the roles and relations that constitute us. This nonanalytic renyilizhi language of “the four beginnings” is “aspectual” in that each term provides a particular perspective on the same relationally constituted, vital phenomenon and anticipates how this inchoate,

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 197 socially embedded person, located within a grammar of familial and cultural conditions, will be inclined to grow and mature as a source of communal and ultimately cosmic meaning. And as we have argued, since nothing is static in a process cosmology, this nurtured and unimpeded xin as the seat of a person’s purpose and resolution continues to expand its influence as integral to the growth and maturation of xing. But perhaps even more important philosophically, the understanding of persons as discrete and isolatable individuals that is being imputed to Mencius has been repudiated as erroneous thinking within the continuing, internal critique of the Western philosophical narrative. For A. N. Whitehead, for example, the notion of the discrete individual is a prime and powerful example of what he calls the fallacy of simple location: that is, the familiar and yet fallacious claim that isolating, decontextualizing, and analyzing things as simple particulars is the best way to understand the content of our experience. For Whitehead, the notion of the discrete individual that has deep roots in the Western narrative and is assumed in much of liberal theorizing is a specific, persistent example of this deformation professionelle.25 Such theorizing is fatally reductionistic and suffers from what Whitehead has further described as “the fallacy of misplaced concreteness,” where we equate the abstract and independent “thing” with what is real while at the same time being insufficiently cognizant of the genuine connectedness of things as transactional events in our ordinary experience. Misplaced concreteness occurs when we take the antecedent ideal to be what is really real at the expense of the processual transitivity of things.26 A second, corollary misinterpretation of Mencius on xing arises with the frequent eliding of “heartmind” (xin) and xing, and the treatment of them as equivalents in referencing the same given “human nature” antecedent to the process of growth. If the Mencius parses xin as a general description of the native human condition (and it certainly does), and if xing is understood as denoting “human nature” without further qualification (which for commentators it most often is), then this equivocation seems unavoidable. Zhang Dainian, for example, suggests that “since the sage is also a man, the xing that the sage has is the xing that all people have.”27 Since xin in Mencius can serve as a general description of native human conditions available for cultivation, Zhang certainly has a warrant to assert that the sage’s xin in this sense is a generic feature shared by all people. In fact, the Mencius describes the xin as an organ (guan 官) on a plane with the organs of sight and hearing.28 But Zhang’s assumption here in eliding xin and xing seems to be that xing like xin is a given, latent potential shared by all persons by definition, and that the sage is simply the one who successfully actualizes it.

198  |  Roger T. Ames Tang Junyi argues that in our understanding of xing we must give appropriate value to the fact that xing (性) is quite literally a combination of xin (心) (initially heartmind as native conditions available for growth) and sheng (生) (the spontaneous and vital birth, growth, and life of these native conditions within a specific narrative as this same xin achieves its maturation). Importantly, we need to introduce two clarifications here. First, if we register fully the “birth, growth, and life” that comes with sheng, the possibilities entailed by xing do not lie solely within things themselves; rather, the xing of human beings in particular, while certainly deliberate and resolute when properly directed by xin, is a defused and creative collaboration with its various environments. And secondly, while xin certainly references the native human condition, it is also the locus of the dynamic process of the articulation and maturation of persons that enables them to act deliberately in this growth. Tang Junyi underscores the vital, collaborative, and emergent nature of this process: Within Chinese natural cosmology what is held in general is not some first principle. The root pattern or coherence (genben zhi li 根本之理) of anything is its “life force” (shengli 生理), and this life force is its xing (性). The xing is expressed in the quality of its interactions with other things and events. The xing or “life force” then entailing spontaneity and transformation has nothing to do with necessity.  .  .  .  The emergence of any particular phenomenon is a function of the interaction between its prior conditions and other things and events as external influences. So how something interacts with other things and events and the form of this interaction is not determined by the thing in itself. . . . Thus the xing of anything in itself is inclusive of this process of transformability in response to whatever it encounters.29 As we can see, while xin and xing have a different and distinct reference at the beginning of the process of growth and must not be elided, they gradually emerge as terms denoting the same result. This point is made explicitly in the Mencius when it insists that “those who make the most of their ‘heartminding’ (xin) realize their xing.”30 That is, the parsing of xin as both the initial, native conditions and the resolute and deliberate xin that has been achieved by a person through the process of growth and articulation has the same referent as this person’s xing, and hence there is no equivocation. A third misreading of Mencius on xing is prompted by the frequent analogy he appeals to between our desire to nourish the “body” and our similar impulse to nurture our “heartmind” (xin). Influenced by our own

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 199 persisting commonsense, we are inclined uncritically to assume the outcome of personal cultivation to be an intellectual and spiritual ascent that elevates us above our mundane physical experience. But such dualistic thinking is anathema to Mencius. And sinologists generally, responding to the realization that our familiar separation of the cognitive and the affective is not relevant for Mencius, attempt to overcome this persistent dualism by having standardized the translation of xin as some variant of “heartmind.” But this is just a beginning. In addition to resisting this cognition/affect dualism, Mencius’s notion of xin precludes the relevance of other familiar dualisms as well: that is, mind-body, inner-outer, subject-object, and agent-action as well. With this acknowledgment, we might be doing greater justice to this term if we read xin gerundively (if ungrammatically) as a “vital, focused bodyheartminding within our field of experience.” In any case, we will need to reconceive xin in this more capacious direction if we are to rescue it from our suppressed dualistic assumptions.

Using Mencius to Restate Graham’s Insights into the Notion of xing But where are we? Despite Graham’s important insights into this alternative, narrative understanding of xing, we continue to have a lingering concern. Although Graham abandoned his earlier essentialist reading and has provided a relatively full account of his more capacious interpretation of xing in his Disputers of the Tao, these same insights continue to be either misunderstood or ignored. At the very least, we need only note that in the interpretive literature Graham’s revisionist reading has failed to trump the persisting understanding of Mencius’s xing as a default, inborn, and morally good “human nature.” One significant factor in this persistent misreading is the seemingly invincible commitment to a default individualism in our contemporary culture to the extent that we can only struggle to think otherwise. It has only been by locating Graham’s position within an interpretive context by appeal to excerpted fragments drawn from his later publications broadly that we have been able to bring his dynamic, irreducibly contextual, narrative interpretation of xing into clearer focus for ourselves and hopefully for our readers. At the same time, in our own recent work we have also sought to register the importance of the interpretive context in trying to argue for a Confucian role ethics that begins from the fact of associated living and, on that basis, contests individualism with a radically different relational understanding of what it means to become a person. We will now turn to

200  |  Roger T. Ames a reading of the Mencius that will hopefully show an alignment between Graham’s understanding of xing and our concept of role ethics.

The Mencius on the in medias res of Native Conditions In further elaborating on Mencius here, we will find that a close reading of the Mencius confirms Graham’s argument that the project of becoming consummately human, far from simply actualizing ready-made human “beings,” “begins” and “ends” in medias res through the deliberate personal cultivation of our roles and relations. That is, human becomings begin “in the middle” within narratives and end the same way. Further, the Mencian project of vital human “becomings” seems to demand not only motivation and real purposeful effort on our part, but also imagination and a creative responsiveness to the constraints set by our ever-changing circumstances. Graham, in registering “what Heaven has ordained as the spontaneous preference of your nature” is referencing our incipient “four beginnings” (siduan 四端)—that is, the native, constitutive conditions Mencius generalizes about the human experience and that he identifies with the inchoate heartmind (xin 心). In the Mencius we read: The heartmind in feeling pity at suffering has the first beginnings of consummate conduct in one’s roles and relations; the heartmind in feeling shame at crudeness has the first beginnings of appropriate conduct in one’s roles and relations; the heartmind in feeling a sense of modesty and deference has the first beginnings of achieving propriety in one’s roles and relations; the heartmind in feeling a sense of approval and disapproval has the first beginnings of acting wisely in one’s roles and relations. Persons having these four beginnings (siduan) is like their having their four limbs. . . . Now acknowledging that these four beginnings are defining of me, the process of realizing the development and fruition of them is like a fire beginning to burn or a spring of water beginning to gush forth. Persons who are able to bring them fully to fruition can vouchsafe everyone within the four seas; persons unable to do so cannot even serve their own parents.31 Commenting on this passage, Graham observes: It is essential to Mencius’ case that although moral education is indispensable it is, like the feeding of the body, the nourishing

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 201 of a spontaneous process. The process once launched accelerates, like fire catching, because we discover the pleasure of it.  .  .  . A man becomes bad, not because the incipient impulses are missing from his constitution, but because he neglects and starves them.32 Graham’s appeal to the terms “constitution” and “incipient impulses” are important because Mencius is here referencing the native inclinations of the heartmind of the newborn infant: “Mencius said: ‘Great persons are those who do not lose the heartmind of the newborn babe.’ ”33 Mencius’s description of incipient “beginnings” not only introduces the radically embedded newborn as the locus of growth but further defines the “constitution” of this infant in terms of these same organic relations. That is, the newborn emerges in medias res as an inchoate narrative within the larger narrative of family and community. The general, concrete pattern of family and communal relations of an infant born into a mature culture can be captured both descriptively and prescriptively by an appeal to the impulses of these constitutive familial bonds. For this infant, the four beginnings are thick native conditions that are irreducibly biological, familial, and social. They are thus by nature resolutely relational and inclusive, and a source of moral growth (shan) as they become more robust. Said simply, the newborn as a concrete fact is constituted by the gerundive relations that define it as it follows the promptings of these relations and grows them into an emerging personal identity; the infant as a discrete and separate “individual” is no more than a retrospective abstraction from this same dynamic manifold of relations. It is the nature of the radial, constitutive relations themselves and their dynamic inclination for growth and extension that locates the newborn within an inchoate habitude and disposes it to positive moral growth.34

The Mencius on the Role of Assiduous Personal Cultivation Contra Xunzi’s caricature of Mencius as offering a naturalistic interpretation of xing, the impulse that guides us in the direction of efficacious moral conduct is incipient and thus, as we have seen, is certainly at risk of being lost. But becoming “great persons” is so much more than simply not losing it. In the Mencius there is in fact an inordinate emphasis on the role of assiduous personal cultivation in consolidating the virtuosic habits that are expressed in and are defining of an exemplary life. This process extends from our raw animality with the slight human advantage of the incipient “four beginnings” as its starting point, to the possibility of attaining a full-blown human sagacity as its apex:

202  |  Roger T. Ames Mencius said: “What distinguishes people from the brutes is ever so slight, and where the common run of people are apt to lose this propensity, exemplary persons preserve it. Shun was wise to the way of all things and had real insight into human roles and relationships. He acted upon his moral habits to be consummatory and appropriate in his conduct rather than merely doing what is deemed consummatory and appropriate.”35 First, a vital distinction here is the difference between persons who are merely able to follow conventional values in acting in a way deemed virtuosic by the community and the sage who through an assiduous personal regimen is able to act from a cultivated and aggregated habitude of virtuosity. Secondly, we need to be careful in how we parse this journey from animality to sagehood. Commenting on the inseparability of physical and moral growth, Graham observes: Moral inclinations belong to nature in the same way as the physical growth of the body. They germinate spontaneously without having to be learned or worked for, they can be nourished, injured, starved, they develop if properly tended but their growth cannot be forced.36 The familiar and repeated analogy between the “four beginnings” and the “four limbs” or “body” is important in at least two senses. First, the growth in our relations and in our bodies is of something profoundly but not exclusively physical (rather than metaphysical) and is prompted by what we desire and aspire to. And secondly, the qualitative virtuosity that emerges in the radial growth of our relations that has been inspired by the promptings of the inchoate heartmind produces a transformed moral physicality—what Richard Shusterman calls a “somaesthetic.”37 Stated simply, the distinction between human animality and sagacity is not between the physical and the intellectual aspects that are, in fact, no more than two ways of viewing and evaluating the same phenomenon. Virtuosic conduct is necessarily embodied and is transformative of our entire person as expressed through a quality of experience that is at once physically and intellectually virtuosic: Mencius said, “. . . What exemplary persons cultivate as their xing—their inclination to act consummately in their roles and relations (ren), to act appropriately (yi), to achieve propriety (li), and to act wisely in those same roles and relations (zhi)—is rooted in their heartmind (xin), and the complexion that develops in this endeavor glows radiantly on their faces, is reflected in their

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 203 carriage, and extends throughout their extremities. Without their bodies having to say anything at all, everyone is keenly aware of this personal growth.”38

Mencius and the Holography of xin as xing The focus-field notion of person assumed in this daode cosmology stands in stark contrast to the metaphysical realist distinction between an inner, private domain and a shared outer world. It begins from the doctrine of internal, constitutive relations and requires a fundamentally different understanding of persons in which their particular identities and the unsummed totality—their foregrounded focus and its field—are two holographic and thus mutually entailing ways of perceiving the same phenomenon. Just as each played note in a symphony has implicated within it the entire performance, so each focal event has implicated within it its entire field. Indeed, to grasp this holographic understanding of xin the following, oft-cited passage from the Mencius that calls into question the distinction between an inner self and an outer world might require a more literal reading than it usually receives: “Mengzi said, ‘The myriad happenings of the world are all implicated here in me.’ ”39 To make sense of this Mencian claim we need an alternative to our commonsense understanding of the inner self and the outer world as two separate domains. We must clarify the background cosmological assumptions about the processive nature and the radical contextuality of the human experience, and about the perceived relationship between particular persons and their experienced world. Most obviously, as previously noted it is a commonplace that xin does the work of both cognizing and feeling; we have both felt thoughts and cognitively informed feelings. Similarly, there is no strict dichotomy between intellection and sensation, between body and mind, between structure and function, between thinking and doing, between center and context, between nature and culture. These aspectual distinctions are nonanalytic and mutually entailing; they do not serve to separate and isolate different components within our “bodyheartminding” experience nor fragment the activities that are defining of it. If we use the language of focus and field to give an account of the virtuosity achieved in “bodyheartminding,” it is first and foremost the dynamic focus of a specific systemic center of thinking and feeling (de 德) that extends out radially as a physiological, psychological, and sociological experience to the furthest reaches of the unbounded cosmos as its contextualizing field (dao 道). Indeed, xin is only derivatively and abstractly taken

204  |  Roger T. Ames to be the physical organ that then becomes metonymic for a full complex of interactions and events, where this one focal aspect is isolated as a symbol for the holistic and eventful functions—both physical and psychic—that constitute a continuing human life. We might recall another related passage in the Mencius cited earlier that describes and advocates for the symbiotic and mutually entailing phases of making the most of our “bodyheartminding,” an effort that in turn enables us to grow our initial conditions in family and community fully and to make our own distinctive contribution to the realization of our cosmic context: Those who make the most of their “bodyheartminding” (xin) realize their xing. And in realizing their xing they are realizing their natural and cultural context (tian). Consolidating their bodyheartminding and nourishing their xing is the way to pay service to nature and to culture (tian).40 The familiar dualistic separation of inner and outer domains in its appeal to a doctrine of external relations brings with it a familiar psychological exercise called “introspection.” Such introspection is usually conceived as turning from a normal outward orientation to a reflective examination of one’s own internal mental states and feelings. Inspired by this Mencian understanding of “bodyheartminding,” however, we might want to challenge this definition of what takes place when we look inward by inventing an alternative term—“intraspection.” Such a neologism would signal the fact that the process of “looking into our own bodyheartminding” is at the same time a looking outward into the quality of the coalescence this “bodyheartminding” is achieving in its shared narrative with its contextualizing world. Indeed, such “intraspection” as a looking “into” the productive connectivity of our bodyheartminding with the “outer” world is both inner and outer at the same time. The point is that bodyheartminding is holographic, and indeed, since “everything is here in me,” in “making the most of our bodyheartminding,” we are literally bringing the entire cosmos into more meaningful focus and resolution from our own unique perspectives. And in so doing, we are “realizing our natural and culture context.” Citing this Mencius passage more fully: Mengzi said, “Is there any enjoyment greater than, with the myriad happenings of the world all implicated here in me, to turn personally inward and to thus find resolution with these happenings. Is there any way of seeking to become consummate in my person more immediate than making every effort

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 205 to act empathetically by extending myself into the perspectives entertained by others.”41 Again in this passage we see that becoming consummate as a human being is a holographic process, where the inner resolution of our connectivity with all of the happenings in the world and the outer reach and influence we are to have on other things are coterminous and mutually entailing. There is a symbiosis between consolidating our relations within our unique focus and thereby bringing these relations into meaningful and clear resolution (cheng 誠) on the one hand, and extending the field of relevance of this focal identity outward through deferring to and producing meaning in our expanding circle of personal relations (shu 恕) on the other. Sincerity and resolve as achieved in our relations “within” are manifested as consummate conduct in our relations “without.” In this way, not only are the myriad happenings of the world implicated here in me, but more importantly, they are within the focus of my person made optimally meaningful by my capacity to bring full resolution to my connectivity with them in my own person.

Mencius and the Reserving of xing for That Which Is Distinctively Human Mencius again chooses to reserve the use of the term xing for what is most exclusively and distinctively human. But the distinction he wants to make between xing (性) and “propensities” (ming 命) in the following passage cannot be reduced to dichotomies between the intellectual and the sensual, because not only is physicality a ubiquitous aspect of human activity, but just as with the intellectual and emotional it is reflective of the assiduous cultivation needed to become consummately human. The mouth’s penchant for taste, the eye’s for color, the ear’s for sound, the nose’s for smell, and the body’s for comfort—these are xing, and yet because our propensities (ming) also have a role in these capacities, exemplary persons are not given to referring to them as xing. A penchant for consummate conduct in the father-son relation, for appropriateness in the ruler-subject relation, for ritual propriety in the guest-host relation, for wisdom in the relations of superior persons and others, and for sagacity in the way-making of tian, are certainly propensities (ming) that we have, yet because xing also has a role in these capacities, exemplary persons are not given to referring to them as propensities.42

206  |  Roger T. Ames Ming as “propensities” is used here as elsewhere in the Mencius to designate those circumstances in the life experience over which we have less rather than more influence and control. Xing by contrast references outcomes that carry the strongest human signature: our capacity to purposely and deliberately nurture our familial and communal roles, and in so doing, to elevate the human experience into an elegant and exemplary narrative.

The Mencius and a Virtuosity That Is Deliberate and Resolute The habituated growth and achieved virtuosity that was initially prompted by our native conditions, when properly disciplined and directed, suffuses our entire persons to achieve a quality of resolution that is deliberate, and that serves as the ground of our personal identities: Mencius said: “.  .  .  I haven’t felt agitation in my heartmind from the age of forty on”  .  .  .  One’s purposes (zhi 志) are what guide one’s qi, and one’s qi is what fills one’s body. Wherever one’s purposes take one, one’s qi will follow behind. Thus it is said: Be firm in your purposes and do not distress your qi.”43 We have seen that xin 心 (native conditions available for growth) and sheng 生 (the birth, growth, life, and vitality of these native conditions) functioning in combination both describe and prescribe the vital narrative in which we become human (xing 性). Similarly, etymologically the combination of xin 心 (native conditions available for growth) and zhi as either 之 or 止 (both meaning “going to”) are combined to denote the purposefulness and resolution (zhi 志) that must attend this vital process of achieving virtuosity in one’s conduct.

We might reflect on what seem to be certain suppressed premises in the cosmological and normative language of “virtuosity”—dao (道) and de (德)—that make it a specifically human achievement. It is significant that the character for dao found on the bronzes is not simply a road—“the Way” as it is conventionally translated—but is reflexive in including within it a clear

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 207 graphic representation of walking-and-seeing persons journeying resolutely ahead on their way forward, including both where they have come from and where they are headed. Simply put, dao is not the way; it is our waymaking as we move straight ahead. With an appropriate awareness of the fact that the etymology of “world” as w(e)oruld (Old English) is derived from a Germanic compound of wer (man) and eld (age), and thus means, quite explicitly, “the age or life of man,” this specific reference to the human sojourn might allow for an alternative translation of dao as “world-making” in a human-centered cosmogony.

On the oracle bones, de that is conventionally rendered nominally as “virtue” or “excellence” appears as and is similar to the graph for dao in again depicting persons walking deliberately forward with eyes focused on the road ahead.44 As we might anticipate in this holographic cosmology, there is an immediate graphic resonace between dao as a resolutely and specifically human understanding of forging our way forward in the world and its dyadic correlate de that denotes human flourishing within the context of our world of experience. Indeed, on the bronzes, the heartmind (xin 心) signific is added as an additional element in the de character, underscoring the intellectual, the affective, and the physical dimensions integral to the life experience as we cultivate and accrue the virtuosity needed to extend our way forward to optimum effect in the world.45

When we parse this notion of “straight ahead” and consider the normative qualities of action that would follow from it, we have a glossary of terms: true, direct, candid, authentic, rightly positioned, immediate, timely, undiluted, deliberate, resolute, intense, and so on. Such disciplined conduct in sum is indeed the meaning of “virtuosity”: that is, being resolute in

208  |  Roger T. Ames ­ ptimizing those resources and their creative possibilities that allow us to o find the most efficient, productive, and expedient way forward. This same Mencius 2A2 passage then continues: “Is there any way of getting past agitation in one’s heartmind?  .  .  .  May I dare to ask after your success in this respect?” Mencius replied: “I understand discourse, and I am adept at nourishing my flood-like qi.” “May I ask what you mean by ‘flood-like qi’?” Mencius replied: “It is difficult to express. It is the most extensive and the most intensive quality of qi. If one nurtures it faithfully and without respite, it will fill up all between the heavens and the earth. As a quality of qi, it is a piece with achieving optimal appropriateness in one’s conduct (yi 義) and moving ahead on the proper way (dao 道). Without this qi, one lacks proper nourishment. It is what is born of a cumulative habit of optimally appropriate conduct and is not something that can be had through random acts of appropriateness.”46 Again, restated in the cosmological language of focus and field, Mencius attributes his success in achieving resolution in his conduct to “understanding discourse” (zhiyan 知言) and being “adept at nourishing his flood-like qi.” The expression “understanding discourse” might well be an allusion to the Analects wherein Confucius allows that “someone who does not understand discourse (bu zhiyan 不知言) has no way of understanding others.”47 It denotes a quality of discoursing that is the substance of consolidating relations—the quality of communicating that inspires a communing community. So when properly nurtured and cultivated, Mencius’s “flood-like qi” achieves its greatest “extensive magnitude” (至大 most vast) and its most “intensive resolution” (至剛 most firm, unyielding, indomitable, constant, enduring). In the language of the text as it continues, to nourish one’s qi most successfully is to achieve the greatest degree of meaningful resolution (yi 義) within the most extensive field of qi (dao 道). In this manner, a sustained excellence in one’s conduct (daode 道德) is attained through acquiring the greatest degree of potency and effectiveness (de 德) in relation to the most far-ranging elements of one’s environments (dao 道).

In Summary Bringing Graham’s revisionist reading of Mencius on xing and our own elaborations on the relationally constituted person into summary form, we

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 209 might begin by observing that xin (心) references the native conditions and tendencies of a human “becoming” as a manifold of incipient bonds that locate us within the narratives of family and community (renyilizhi 仁義禮 智) and incline us to growth and efficacy in our constitutive roles and relations (shan 善). These native conditions are not only a generic description of the pattern of relations of persons as they are born into the continuing narrative of a mature culture, they are also normative in the sense that if we cultivate ourselves to live and grow according to these native inclinations, we can anticipate flourishing as human beings. However, although we are so inclined, it is also possible through inattention and a lack of effort to lose this capacity for normative growth. Since Mencius’s notion of xin defies the familiar dichotomies of cognition/affect, body and mind, agent and action, inner and outer, subject and object, rather than translating it as “heartmind” that only challenges the first of these dualisms, to do its complexity justice, we might have to render it awkwardly but more capaciously as “vital, resolute, and focused bodyheartminding within our field of experience.” That is, xin as fundamentally relational and vital rather than discrete denotes a particular dynamic focus within an extended field of relations and is thus holographic, with “the myriad things all implicated here in me.” Xing (性) is a process of cultivation that includes xin as the ground for the habitual dispositions one is able to grow from the native conditions. When this growth is directed by a purposeful and resolute xin and is fully informed by both feelings and circumstances (qing 情), it has the capacity for creative redirection in service to human flourishing. This process of growth is made possible through a cultivation of qi (氣) within the discursive context of our roles and relations (zhiyan 知言). The cultivation of qi requires us to be resolute (zhigang 至剛 and cheng 誠) in our purposes (zhi 志), where such purposes are themselves driven by our need to achieve optimal appropriateness (yi 義) as the ground for our contribution to the reach and influence (zhida 至大) of the proper way (dao 道). In the absence of those biological and metaphysical assumptions that lie behind the notion of the discrete individual, it is the cultivating of unity and resolve in the deliberate purposes of physically and socially defused persons as these now-focused intentions are manifested in our embodied action that becomes the signature of personal uniqueness and coherent identity. There is thus an important and repeated distinction made between simply acting according to established norms (li 禮) and achieving the resolution and focus that consolidates these norms as one’s personal identity. An achieved propriety in the roles and relations of our narrative identities (li 禮) entails an analogous growth of our lived bodies (ti 體) that are also expressive of this virtuosity. Indeed, just as virtuosity in the roles that we live comes to distinguish us in our families and communities, so the

210  |  Roger T. Ames internalization of this virtuosity is manifested in the glow and complexion of our human physicality. It is only in giving Mencius such a reading that we can appreciate the extent to which he is indeed an extension of Confucius, who also demands assiduous effort in the project of becoming consummately human. And it is the narrative understanding of xing that we can tease out of Graham’s oeuvre that provides an alternative conception of person sufficiently robust to challenge the ideology of liberal individualism.

Notes   1. See Roger T. Ames, “The Mencian Conception of Ren xing 人性: Does It Mean ‘Human Nature’?,” in Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus C. Graham, ed. Henry Rosemont Jr., 143–175 (La Salle: Open Court, 1991), and “Mencius and a Process Notion of Human Nature,” in Mencius: Contexts and Interpretations, ed. Alan K. L. Chan, 72–90 (Honolulu: University of Hawai’i Press, 2002); and Irene Bloom, “Mencian Arguments on Human Nature (Jen-hsing),” Philosophy East and West 44.1 (1994): 19–53, “Nature and Biological Nature in Mencius,” Philosophy East and West 47.1 (1997): 21–23, and “Biology and Culture in the Mencian View of Human Nature” in Chan, Mencius, 91–102.  2. Xunzi 23: 孟子曰：「人之學者，其性善。」 曰：是不然。是不及知人之性，而不察乎人 之性偽之分者也。凡性者，天之就也，不可學，不可事。禮義者，聖人之所生也，人之所學而能，所 事而成者也。不可 學，不可事，而在人者，謂之性 。Although Xunzi is making a distinction

between what we are born with and what is deliberately acquired, we must resist translating Xunzi’s naturalism into an essentialistic metaphysical claim that would align him with the dualistic realism of a Plato or an Aristotle in which something that is “inherent in persons” (zairen 在人) separates persons from their world.   3. As D. C. Lau points out, because sheng 生 and xing 性 are used interchangeably in these early texts and were close in pronunciation, Mencius here imputes a tautology to Master Gao: “A is A,” allowing Mencius to draw the analogy with the claim that “white is white” and to then take the discussion on to assert that Master Gao’s position is a reductio. See D. C. Lau, Mencius (Hong Kong: The Chinese University Press, 1984), 225.   4. 6A3: 告子曰：「生之謂性。」孟子曰：「生之謂性也，猶白之謂白與？」曰：「然。 」 「白羽之白也，猶白雪之白；白雪之白，猶白玉之白與？」曰：「然。」 「然則犬之性，猶牛 之性；牛之性，猶人之性與？」

  5. Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982; 2nd ed. 1998), 50.   6. Lee Yearley, Mencius and Aquinas: Theories of Virtue and Conceptions of Courage (Albany: State University of New York Press, 1990), 60.   7. Ibid., 58–59.   8. Ibid., 60.   9. Ibid., 59.

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 211 10. Raymond Williams, Keywords: A Vocabulary of Culture and Society (New York: Oxford University Press, 1976). 11. 文化不改，然後加誅。 12. A. C. Graham, Studies in Chinese Philosophy and Philosophical Literature. This collection of essays was first published by the Institute of East Asian Philosophies, National University of Singapore, in 1986, and reprinted by State University of New York Press in 1990, 55. But importantly, this specific essay, “The Background of the Mencian Theory of Human Nature,” was published early on in Tsing Hua Journal of Chinese Studies 6.1–2 (1967): 215–271. 13. Ibid. 14. 誠者非自成己而已也，所以成物也。成己，仁也；成物，知也。性之德也，合外內之 道，故時措之宜也。

15. Zhongyong 17: 。 。 。大德必得其位，必得其祿，必得其名，必得其壽。故天之生

物，必因其材而篤焉。

16. Arthur Waley, Three Ways of Thought in Ancient China (Stanford: Stanford University Press, 1939), 205. 17. Henry Rosemont Jr., ed., Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus C. Graham (La Salle: Open Court, 1990), 287. 18. Ibid. 19. See, for example, David L. Hall and Roger T. Ames, Thinking from the Han: Self, Truth, and Transcendence in Chinese and Western Culture (Albany: State University of New York, 1998), 23–78, and Roger T. Ames and Henry Rosemont Jr., The Analects of Confucius: A Philosophical Translation (New York: Ballantine, 1998), 20–45. See also Roger T. Ames, Confucian Role Ethics: A Vocabulary (Honolulu and Hong Kong: University of Hawai’i Press and Chinese University Press, joint publication, 2011), chap. 2. 20. Rosemont, Chinese Texts, 288–289. 21. A. C. Graham, Studies in Chinese Philosophy, 8. 22. A. C. Graham, Disputers of the Tao: Philosophical Argument in Ancient China (La Salle: Open Court, 1989), 120. 23. Rosemont, Chinese Texts, 288. 24. Ibid., 290–291. 25. A. N. Whitehead, Process and Reality: An Essay in Cosmology, ed. David Ray Griffin and Donald W. Sherburne, corrected edition (New York: Free Press, 1978), 137, observes: “This presupposition of individual independence is what I have elsewhere called, the ‘fallacy of simple location.’ ” Charles Hartshorne, A History of Philosophical Systems (New York: Philosophical Library, 1950), 443, elaborates: “as Whitehead has most clearly seen—individuals generally are not simply outside each other (the fallacy of ‘simple location’) but in each other, and God’s inclusion of all things is merely the extreme or super-case of the social relativity or mutual immanence of individuals.” 26. See Whitehead, Process and Reality, 10. 27. Zhang Dainian 張岱年, 中國哲學大綱 (An outline of Chinese philosophy) (Beijing: Chinese Academy of Social Sciences Press, 1982), 250–253. 28. Mencius 6A15. 29. Tang Junyi 唐 君 毅, Complete Works 唐君毅全集 (Taipei: Xuesheng shuju, 1991), 4: 98–100: 中國自然宇宙論中，共相非第一義之理。物之存在的根本之理為生理，此生理即

212  |  Roger T. Ames 物之性。物之性表現於與他物感通之德量。性或生理，乃自由原則，生化原則，而非必然原則。。 。 。蓋任一事象之生起，必由以前之物與其他之交感，以爲其外緣。而一物與他物之如何交感或交感之 形式，則非由任一物之本身所決定。。 。 。因而一物之性之本身，即包含一隨所感而變化之性。 30. Mencius 7A1: 盡其心者知其性也。 31. Mencius 2A6: 惻隱之心，仁之端也；羞惡之心，義之端也；辭讓之心，禮之端也；是 非之心，智之端也。人之有是四端也，猶其有四體也。 . . . 凡有四端於我者，知皆擴而充之矣，若 火之始然，泉之始達。苟能充之，足以保四海；苟不充之，不足以事父母。

32. Graham. Disputers, 126, 129. 33. Mencius 4B12: 孟子曰：「大人者，不失其赤子之心者也。」 34. In Just Babies (New York: Random House, 2014), Paul Bloom argues that humans rather than being blank slates are hardwired with a sense of morality. The problem with this understanding of morality is that it locates it exclusively in the babies themselves rather than in the relationally informed narrative of the baby—a new narrative within a larger continuing narrative. 35. Mencius 4B19: 孟子曰：人之所以異於禽獸者希。庶民去之，君子存之。舜明於庶物， 察於人倫，由仁義行，非行仁義也。See also 6A8. 36. Graham, Disputers, 125. 37. See, for example, Richard Shusterman, Body Consciousness: A Philosophy of Mindfulness and Somaesthetics (Cambridge: Cambridge University Press, 2008). 38. Mencius 7A21: 孟子曰：「。 。 。 君子所性，仁義禮智根於心。其生色也，睟然見 於面，盎於背，施於四體，四體不言而喻。」 39. Mencius 7A4: 孟子曰：萬物皆備於我矣。 40. Mencius 7A1: 盡其心者知其性也。知其性則知天矣。存其心養其性所以事天也。It is because it is important to appreciate the performative implications of zhi 知—con-

ventionally translated as “knowing”—that we have rendered it here as “realizing” something in the sense of “making it real” rather than simply “knowing” something cognitively. 41. Mencius 7A4: 孟子曰：萬物皆備於我矣。反身而誠，樂莫大焉。強恕而行，求仁莫 近焉。

42. Mencius 7B24: 口之於味也，目之於色也，耳之於聲也，鼻之於臭也，四肢之於安佚 也，性也，有命焉，君子不謂性也。仁之於父子也，義之於君臣也，禮之於賓主也，智之於賢者也， 聖人之於天道也，命也，有性焉，君子不謂命也。 43. Mencius 2A2: 孟子曰：「。。。我四十不動心。” 。 。 。 夫志，氣之帥也；氣，體 之充也。夫志至焉，氣次焉。故曰：『持其志，無暴其氣。』」 44. See Kwan Tze-wan’s “Multi-function Character Database” 殷墟文字甲編 2304,

http://humanum.arts.cuhk.edu.hk/Lexis/lexi-mf/. 45. Ibid., 2837. Reinforcing this understanding of de as a kind of “conatus”—the deliberate effort of striving forward and excelling in the life experience—is the fact that in several of the recently recovered archaeological texts, the character de 德 is written using the graphic variant, 悳. This graphic alternative has a heartmind radical xin 心 placed underneath the character zhi 直 that itself again means “moving straight ahead.” There is clearly a cognate relationship between these two graphic variants for de—德 and 惪—on the one hand, and the character zhi 直 meaning “moving straight ahead,” on the other. In the early texts the latter occasionally appears as a loan character for the former, and in the archaic language they are markedly similar in pronunciation.

Reconstructing A. C. Graham’s Reading of Mencius on xing 性  | 213 46. Mencius 2A2: 曰： “不動心有道乎？” 。。。“敢問夫子惡乎長？” 曰：“我知言，我 善養吾浩然之氣。” “敢問何謂浩然之氣？” 曰：“難言也。其為氣也，至大至剛，以直養而無害， 則塞于天地之閒。其為氣也，配義與道；無是，餒也。是集義所生者，非義襲而取之也。” 47. Analects 20.3: 不知言，無以知人也。

10

Reason and Spontaneity Reconsidered

Lisa Raphals

Prefatory Note In his work on the Zhuangzi and elsewhere, Angus Graham was consistently occupied by two key themes and concerns. One is the importance of skill knowledge, and this essay was originally intended as a treatment of issues pertaining to skill knowledge in excavated texts that became available since Graham’s death. Another is the importance of informed, spontaneous performance. This is of course a preoccupation in Graham’s treatment of the Zhuangzi. But it receives more sustained attention as central to the argument of Reason and Spontaneity (1985). I remember that in conversations within the last few years of his life, Angus had expressed pleasure that “anyone had read it,” so it seems fitting to center this essay on the arguments of that book, one of his few essays into formal philosophy. This is not to say that anyone would expect—or want—the arguments and presentation of that book—including an interlude in the form of a long poem—to echo the language or style of contemporary analytic philosophy. In Reason and Spontaneity, Angus Graham argued that humans are agents who choose our ends and purposes but, in doing so, encounter Hume’s understanding that no normative statement about values can be derived logically from declarative statements about facts.1 As Graham puts it: “I am not an instinctive being like an animal” (more on this later). “I have to choose, and on the [Humean] position we are here considering, all imperatives are ungrounded.”2 Graham argues that rationalists and moralists have been unwilling to acknowledge that much of what they value arises from:

215

216 | Lisa Raphals the vast area of human behaviour which shares the spontaneity of physical events. Physical events are caused, human action is willed; causes determine effects, the will is free. To the extent that activities are spontaneous it appears that they belong to the realm of the caused (which in the case of biological process is obvious enough), and that he is a free agent only to the extent that he learns to direct them.3 Graham also emphasizes that he is not suggesting that we become more spontaneous, but rather that when we reason about means, ends, and principles, we need not be troubled about “that little puzzle about passing from ‘is’ to ‘ought.’ ” Once we recognize that our ultimate goals are spontaneous, the only necessary first principle becomes “Be aware.”4 He also rejected Kant’s solution, which was to ground ends and values in reason.5 It is important to note how Graham uses the key term “spontaneity,” because in English the term has two distinct meanings: something that is self-caused and something that is random and uncaused. Graham ascribes “spontaneous” behavior to characters in the Zhuangzi in the former sense. Characters who butcher or carve or swim “spontaneously” do so by virtue of a cultivated disposition that makes their actions effortless yet efficacious. Graham argues that our ends are grounded not in reason but in inclinations: “We find ourselves compelled in practice to start from inclination as from perception, questioning inclinations like perceptions only when they conflict, without reason having authorized the initial step.” But without perception and inclination, reason has nothing to engage with; when it does have them, it can criticize and guide.6 He does not define this term, but he seems to take inclinations as generated by psychological states, but what is particularly important about them is that they are spontaneous.7 As Yukio Kachi points out, Graham advances both a proposal and an empirical thesis. The proposal is a general theory of value that grounds all values in the imperative to “be aware.”8 The empirical thesis addresses causal connections between awareness and motivation in a wide variety of contexts. Graham describes this proposal as a “quasi-syllogism,” which goes as follows:9 In awareness of everything relevant to the issue (= everything which spontaneously moves me one way or the other), I find myself moved towards X, overlooking something relevant I find myself moved towards Y. Be aware. Therefore, let yourself be moved towards X (= choose X as end).10

Reason and Spontaneity Reconsidered  | 217 The quasi-syllogism is central to two of Graham’s most important books: the philosophical study Reason and Spontaneity (1985) and, indirectly, Disputers of the Tao (1988). In Reason and Spontaneity, he states that the book’s line of thinking arose in connection with his work on and translation of the Zhuangzi.11 Graham rejects both Kantian rationalism and romantic irrationalism in favor of a notion of self in which awareness integrates reason and spontaneity: Like the animals, I am an organism which spontaneously senses, analogizes to the already experienced, and tends towards or away. Unlike them, I am self-conscious, can detach myself from spontaneous process in order to analyze and criticize perceptions, analogies and reactions, choose ends from my spontaneously emerging goals, choose means to my ends. In becoming selfconscious I require an imperative by which to choose between spontaneous tendencies as they veer with changing awareness, but only one, “Be Aware.”12 This view is in part explicitly derived from the Zhuangzi, but the reasons are important. Graham, as elsewhere, identifies the Zhuangzi with what he calls anti-rationalism, as distinct from irrationalism. Irrationalism refuses to submit spontaneity to the test of awareness; anti-rationalism recognizes the need to “be aware” but rejects reason as the basis for awareness or the link between awareness and action.13 Graham’s argument is both philosophical and empirical, and I want to respond to both in the light of subsequent arguments in philosophy and subsequent research in several areas of psychology and biology. In the first section I address resonances between several strands of research on autonomy and Graham’s account of inclination informed by awareness (rather than unmitigated reason) as the basis for agency and choices among ends. I also argue that this account does not require Graham’s “anti-rationalism.” Much scholarly ink has been spent on his account of the Zhuangzi in particular as anti-rationalist. Rather than trying to engage in those debates (on which I have a view), I instead point out that they are not necessary to his view of agency. In the second section, I turn to his empirical argument and show how it is supported by recent research on the biology of agency. Finally, in the third section I take up one point where I think his argument may miss the mark: in his rigid distinction between humans and animals.

Graham’s Awareness and Non-Kantian Autonomy One problem with Graham’s account is the assumption—by no means unique to him—that the only or the primary account of agency in the “Western”

218 | Lisa Raphals tradition is Kantian. Three other approaches resonate far more with his dictum to “Be aware” than does the Kantian rational agent he takes as his philosophical opponent. A powerful alternative to Kantian personal autonomy is Millian agency, which combines an account of individual autonomy with a naturalistic account of action.14 On Mill’s view, individuals choose to implement their desires, but they also own or take charge of them, a state Mill describes as “having a character.” Someone whose desires and impulses are not her own has no more character than a steam engine.15 Character requires a person to own or prioritize certain desires over others, but these priorities must arise from natural causes. John Skorupsi argues that resisting strong desires for good reasons is the paradigm of an autonomous act; and autonomy is the capacity to respond to good reasons. Autonomy, as he puts it, consists of recognizing and responding to a reason.16 This is interestingly close to Graham’s “Be aware.”17 But other strands in contemporary ethics also suggest non-Kantian approaches to the problems of choice and autonomy. Jonathan Schneewind suggests five sources for renewed interest in autonomy since about 1970: new ideas on free will and philosophy of action; medical ethics and bioethics; feminism, debates on liberalism within political thought, and neo-Kantian ethics.18 While these developments all occurred within Graham’s lifetime, they were not central to his interests (philosophical or sinological), and much of their growth has come in the years since his death. But these approaches also suggest accounts to fact and value that prioritize awareness. An example is Harry Frankfurt’s account of “second-order desires” in his famous 1971 essay “Freedom of the Will and the Concept of a Person.” Frankfurt notes that humans reflect on our desires and form “second-order” desires based on that self-conscious reflection. The ability to form secondorder desires is what distinguishes humans from animals and underlies free will, which consists in being able to choose which first-order desires to act on. In later writings Frankfurt augments this account with notions of wholeheartedness.19 Again, Frankfurt’s second-order desires bear some interesting resemblances to “Be aware.” Frankfurt also made a now-classic argument that what defines humanity is not rationality but freedom of the will. He argues that the usage of “person” as an entity with both mental and physical properties also applies to some animals, as well as imaginable non-humans such as extraterrestrials. But neither animals nor extraterrestrials—who have both psychological and material properties—are persons as the term is normally used.20 His point is not to elucidate a dividing line between human and nonhuman species (a point to which I will return later) but rather to identify the attributes that are most fundamental to human persons, attributes that we could in principle share with nonhumans:

Reason and Spontaneity Reconsidered  | 219 What interests us most in the human condition would not interest us less if it were also a feature of the condition of other creatures as well. Our concept of ourselves as persons is not to be understood, therefore, as a concept of attributes that are necessarily species-specific. It is conceptually possible that members of novel or even of familiar nonhuman species should be persons.21 He distinguishes the rational agent from the “wanton,” who may be rational, but who has no second-order desires: What distinguishes the rational wanton from other rational agents is that he is not concerned with the desirability of his desires themselves. He ignores the question of what his will is to be. Not only does he pursue whatever course of action he is most strongly inclined to pursue, but he does not care which of his inclinations is the strongest.22 But a non-wanton, such as an unwilling addict, cares which of his firstorder desires gains the upper hand.23 He identifies the capacity to form second-order volitions with the ability to have or lack free will, a capacity he considers essential to persons and a distinguishing mark of the human condition.24 Freedom of will is thus different from freedom to do what one wants, a point on which Graham would agree!

The Biology of Choice and Agency Graham emphasizes that much human behavior shares the spontaneity of physical events, and he notes that spontaneous actions seem to belong more to the realm of the caused than to freedom of the will.25 Recent research in several sciences has clarified some of the ways in which spontaneous inclinations may be said to be caused. David Hume famously argued that the self is a bundle of momentary impressions strung together by the imagination. On his view, the self is a (useful) narrative fiction. This view continues in contemporary “narrative” theories of the self.26 Contemporary neuroscience suggests that an ensemble of neurological processes make up the experience of the self. They are distributed across several regions of the brain, with the result that there is no self-contained neurological “self.”27 On this model of the self, spontaneous action plays an important part in several ways. First, important aspects of consciousness precede, and are not accessible to, reflective thought. As Shaun Gallagher puts

220 | Lisa Raphals it, some structures of consciousness are “prenoetic”: hidden from immediate phenomenological experience—things that “happen before we know it.”28 They also tend to be inaccessible to reflective consciousness. Gallagher asks how consciousness and cognitive processes—including perception, memory, and imagination—are structured prenoetically by virtue of being embodied.

Prereflective awareness: phenomenal body image and prenoetic body schema Central to Gallagher’s account of the embodied mind is a distinction between “body image” and “body schema,” which he describes as two different but closely related systems: A body image consists of a system of perceptions, attitudes, and beliefs pertaining to one’s own body. In contrast, a body schema is a system of sensory-motor capacities that function without awareness or the necessity of perceptual monitoring. This conceptual distinction between body image and body schema is related respectively to the difference between having a perception of (or belief about) something and having a capacity to move (or an ability to do something).29 As Gallagher explains it, body image involves more than perception; it can include mental representations, beliefs, and attitudes insofar as they concern one’s own body. By contrast, body schema involves motor capacities, abilities, and habits that enable (and constrain) movement and posture. But body schema also applies to objects of perception and intention beyond one’s own body. The difference is like the difference between perception of movement and actual movement. The body schema operates below the level of self-referential intention. It involves “tacit performances” that are almost automatic: “in this sense the body-in-action tends to efface itself in most of its purposive activities.”30 But intentional, goal-directed activity can also shape movements controlled by the body schema. Thus a body schema is not a form of consciousness, but it can support (or undermine) the intentional activities of the body image.31 This prenoetic performance helps to structure consciousness but does not explicitly show itself. It affects and structures the style and organization of our relations with our environment, including habitual postures and movements. As Gallagher puts it, “the carpenter’s hammer becomes an operative extension of the carpenter’s hand.”32 In other words, it also potentially informs spontaneous and skilled performance. The interest of this distinction for Graham’s work is that both spontaneous inclination or action and the

Reason and Spontaneity Reconsidered  | 221 skill knowledge he makes so much of significantly involve what we might call the extended action of body schemas, beyond the workaday monitoring of the body to far more complex activities.

Somatic markers Other research gives similar accounts of choices that are in some sense physically “caused.” For example, according to Antonio Damasio’s somatic marker hypothesis, the brain associates physiological signals (somatic markers) and the emotions generated from them with past actions and outcomes, with the result that somatic markers bias decisions toward some behaviors and away from others.33 Other research suggests that affective reactions are often faster and more basic than cognitive evaluations, and that anticipatory emotions may be as important as cognitive evaluations in making risky decisions.34 THE SELF IN THE BRAIN: SELF-REFERENTIAL PROCESSING

Several neuroscientists have argued for the existence of a physical self, variously described as a sensorimotor “proto-self,” distinguished from several other “selves” by the stimuli to which it responds and the domain in which it acts.35 Georg Northoff notes that this “self” resembles what William James (1890) called the physical self, and interacts with several other “selves” with distinct domains of activity. These include what has been called a “minimal self” or “core or mental self” (and resembles James’s account of a mental self) and what has been called an “autobiographical” or “narrative self” (with some resemblance to James’s spiritual self).36 But the identification of these distinct domains of selfhood in the brain leaves unanswered the question of what links them together in what we commonly recognize as a self or person. It has been suggested that this sense of self is created in the brain through “self-related processing” (SRP).37 This kind of processing operates on prereflective stimuli associated with a strong sense of selfhood.38 SRP operates through a central integrative neural system made up of cortical midline structures (CMS), understood both anatomically and functionally.39 CMS seem to be involved in self-referential processing across several domains, including language, spatial perception, memory, emotion, facial recognition of oneself and others, and perception of agency and the ownership of one’s movements.40 These structures are probably not unique to humans, and may be homologous across mammalian species. That issue is addressed in the following discussion. •

222 | Lisa Raphals Recent research from a range of disciplines, including neuroscience, psychology, cognitive science, phenomenology, and philosophy of mind, suggests the physical basis of emotion, reason, and decision-making (rather than the nature of the identity of the “self” who thinks, decides, etc.) This possibility gives Graham’s dictum to “Be aware” an expanded meaning. While prenoetic processes truly seem beyond the range of self-reflective activity, basing our decisions on maximal awareness includes awareness of somatic states and inclinations. But this research exacerbates another problem Graham tackles: how “caused” action can be free. Research by Benjamin Libet suggests that unconscious cerebral processes initiate apparently voluntary acts (such as choosing to flick one’s wrist) before the onset of any conscious intention to act. This sort of scenario might seem to undermine free will, but as Gallagher points out, free choice is not about tiny time intervals in the firing of neurons. Free will concerns intentions and purposive actions, however much somatic states and prenoetic knowledge are an important part of our thought processes and decisions.41

Animal awareness I now turn to a different problem in Graham’s account of spontaneity and awareness: the clear boundary he draws between his “aware” self and “instinctive” animals. Both accounts in the Zhuangzi—his preferred texts for the kind of agency he is advocating—and recent biological evidence mostly after his time suggest that this distinction could be reconsidered and nuanced. ANIMAL AWARENESS IN THE ZHUANGZI

Despite his extensive work on the Zhuangzi, Graham never explores how or why the Zhuangzi attributes awareness to animals (and plants), but several bear mention. The Zhuangzi describes “destiny” (ming 命) in the biological senses of life span (sheng ming 生命) and “years allotted by heaven” (tian nian 天年). The text is striking in its insistence that ming in this sense is not limited to humans.42 Understanding ming as life span nuances a continuum between human and animal in the Zhuangzi. By juxtaposing the allotments of ming and the “natural” life spans allotted by Heaven (tian nian), we see a continuum in the “fates” in living things. This account of ming locates our human decisions within a natural continuum of living things, mirroring the Zhuangzi’s attitudes toward human roles in the cosmos. It suggests an appreciation of what in modern terms we would call the shared biological heritage between humans and animals.

Reason and Spontaneity Reconsidered  | 223 Second, the Zhuangzi recommends animals as models because of their freedom from destructive emotions. Animals do not fret over changes in their environment, and are not upset by the illusory shifts of human emotions. In Zhuangzi 21, Lao Dan advises Confucius: Grass-eating animals are not upset by a change of pasture; water creatures are not upset by a change of stream. They go along with minor change, provided they do not lose the great constancies. [Be like this] and happiness, anger, grief, and pleasure can never enter your breast.43 草食之獸不疾易藪，水生之蟲不疾易水 行小變而不失其大常也，喜怒哀樂 不入於跄次。

On this account, animals do not understand or care about their ming, but they respond naturally to change and are not vexed by the illusion of happiness. Here, the Zhuangzi describes felicity as a quality not limited to humans and even seems to recommend the equanimity of animals. This state of felicity accords with dao and with ming and makes it possible to live out one’s allotted life span. Elsewhere, the Zhuangzi suggests that all living things have a natural life span, determined in part by the norms for particular species. For example, the morning mushroom lives a day; the long-lived trees of southern Chu live for centuries.44 Each individual—animal as well as human—has a ming, but it is subject to circumstance, and there is no guarantee that any individual (animal or human) will survive to complete its ming. But despite animals and even plants who reflect upon their own actions, the text does—as Graham seems to believe—distinguish between the agency of humans and other living things. Animals are caught in traps because of their nature, not because of individual decisions or mistakes. And even animals that live out their allotted ming do not control or deliberately create the characteristics that “save” them.45 In this sense, the Zhuangzi maintains an ontological difference between humans and other living things. Our life spans are determined by combinations of accident and individual circumstance and choice, not by class membership. Only humans make deliberate choices that optimize their ming. HIEROCLES ON ANIMAL SELF-PERCEPTION

A comparable claim appears in the “Elements of Ethics” of the second-century (CE) Stoic philosopher Hierocles, in a papyrus discovered at Hermopolis in 1901.46 Hierocles argues that what motivates all animals is “self-ownership,”

224 | Lisa Raphals a reflexive version of oikeiōsis: “appropriation” or “ownership” of oneself. This disposition manifests in animals’ universal instinct for self-preservation. However, as Hierocles argues, self-preservation requires self-awareness, since an animal must perceive itself before it perceives anything else: One must know that an animal immediately, as soon as it is born, perceives itself [aisthanetai heautou].47 Animals perceive their own parts [merōn tōn idiōn aisthanetai]. Thus, winged creatures, on the one hand, are aware of the readiness and aptness of their wings for flying, and, on the other hand, every land animal is aware both that it has its own members and of their use; and we ourselves are aware of our eyes and ears and other parts.48 Every hegemonic faculty [hēgemonikē] begins with itself. In this way a cohesive structure [hexis], which binds together what pertains to it, is first binding of itself.49 The concept of self-perception (heautou aisthanesthai, antilēpsis, or sunaisthēsis) seems to be a Stoic invention.50 Hierocles’s animal self-perception is not grounded in experience or learning; it is pre- or nonconceptual.51 What then is it? Writing in 1986, James Brunschwig and Anthony Long identified it with what neurologists call proprioception, a kind of “muscular sensation” that allows an animal to monitor and adjust the state and position of its limbs and other moving parts.52 Long argues that the Stoics were interested in the principles that make animals function as well-organized wholes, enabling them to coordinate movement and maintain appropriate physical orientation of themselves and their bodily parts. Further, the interaction of exteroception and proprioception produces the self-image (phantasia) that animals use in self-perception and self-concern.53 My point is not to identify Hierocles’s animal “self-perception” with either Stoic or post-Cartesian self-consciousness, but rather to suggest interesting parallels between the minimalist Zhuangzi account of animal felicity, the more substantial Stoic account of oikeiōsis, and biological evidence about animal brains and bodies. Subsequent research may allow us to refine this picture. EMPIRICAL EVIDENCE FOR ANIMAL AWARENESS AND AGENCY

Several findings from evolutionary biology allow us to nuance this picture. In the nineteenth century, Charles Darwin argued that the differences in

Reason and Spontaneity Reconsidered  | 225 the mental lives of animals is one of degree, not of kind.54 Now, there is some empirical evidence for the existence of a neurological “core self” across species. This evidence is of different kinds. Bernhard Baars argues that the homologues of the human brain structures that govern cognition and conscious perception also occur in animals, and that evidence from animal anatomy and physiology suggests that “consciousness of one kind or another may be biologically fundamental and phylogenetically ancient.”55 Jaak Panksepp argues for a “Spinozan-type dual-aspect monism” in which affective consciousness arises from complex neural networks that control instinctual emotional actions.56 The problem, as Panksepp puts it, is that certain ontological positions assume that consciousness is based upon the human rationality and command of language.57 If we understand consciousness in this way (as Graham appears to do), it is easy to conclude that animals lack consciousness. But Panksepp and others present substantial experimental evidence for internal affective states in animal minds. For example, human emotions depend on subcortical brain systems that are shared with other mammals and are controlled by similar regions of the brain.58 The point for Graham’s argument is that, on Panksepp’s dual-aspect monism, raw emotional feelings do not require processing or interpretation by any higher cognitive apparatus. Rather, they reflect the neurodynamics of emotional operating systems and their associated brain mechanisms.59 Moving the argument for animal consciousness and agency one step further, some cognitive scientists argue that many animal species possess the core ability of “self-related processing,” which coordinates internal processes such as emotions, motivations, and homeostasis with external sensory stimuli in relation to goal-directed activities. Mammals have the capacity to relate bodily states, intrinsic brain states, and environmental stimuli to life-supporting goal orientations. It has been suggested that self-related processing operates through a central integrative neural system made up of subcortical-cortical midline structures (SCMS) that are homologous across mammalian species.60 Finally, studies of animal group behavior suggest that the superior awareness of a few individuals in a collective can alter the behavior of collective groups. Recent research has begun to reveal the principles of collective decision-making in animal groups and the complex relationship between individuals and group-level properties in the collective behavior of organisms such as swarming ants, schooling fish, flocking birds, and so forth. In such groups, alignment among individuals (the tendency to move in the same direction as immediate neighbors) makes it possible to transmit information about a change in direction as a rapid wave, extending over a great distance. This behavior makes it possible to amplify local fluctuations in order to react to threats such as predators, since the turning movement

226 | Lisa Raphals of the group creates a larger “sensorium” than individual perception ever could. Thus, one individual detecting a predator and changing direction can rapidly amplify into a propagating wave of turning, so that many individuals or even a whole group turn away from a threat. Nor is it dependent on the specific leadership of any one individual, nor does it require deliberate signaling.61 The interest of this phenomenon for the present discussion is that the spontaneous “turning” behavior of animals, based on “awareness” of danger and spontaneously acting to protect both the individual and the group meets at least some of the requirements of Graham’s quasi-syllogism. I don’t wish to push this point beyond where it will go, or claim that such behavior is equivalent to the kind of awareness he is recommending in Reason and Spontaneity. But rather I want to make the more modest point that we can view awareness as a continuum. In summary, a range of biological evidence over the past fifteen or twenty years significantly extends our account of animal consciousness beyond proprioception. Evidence for a neurological “core self,” for self-related processing in the brain, and for collective decision-making by animal groups suggest far more continuity between animals and humans than had been previously supposed. Here, perhaps, Graham missed the mark. But these developments make his core account of spontaneity and awareness all the more suggestive and prescient.

Notes   1. A. C. Graham, Reason and Spontaneity: A New Solution to the Problem of Fact and Value (London: Curzon Press, 1985), which is in turn informed by two earlier works: the book The Problem of Value (1961) and “Taoist Spontaneity and the Dichotomy of ‘Is’ and ‘Ought,” in Experimental Essays on Chuang-tzu, ed. V. Mair, 3–23 (Honolulu: University of Hawai’i Press, 1983).   2. Graham, Reason and Spontaneity, 2.   3. Ibid., 7.   4. Ibid., 9.   5. Ibid., 10.   6. Ibid., 10.   7. Ibid., 2–3, 7–9.  8. Yukio Kachi, “Reason and Spontaneity by A. C. Graham,” Philosophy East and West 40.3 (1990): 389.  9. For more on the quasi-syllogism, see Harold Rosemont Jr., “Remarks on the Quasi-syllogism,” Philosophy East and West 42.1 (1992): 31–35. 10. Graham, Reason and Spontaneity, 7.

Reason and Spontaneity Reconsidered  | 227 11. Ibid., 184. 12. Ibid., 151. 13. Kachi, “Reason and Spontaneity by A. C. Graham,” 396. 14. As Onora O’Neill puts it, contemporary admiration for individual autonomy owes more to Mill than to Kant because the former attempts to provide a naturalistic account of individual autonomy. O. O’Neill, Autonomy and Trust in Bioethics (Cambridge: Cambridge University Press, 2002), 29–34 15. J. S. Mill, On Liberty, in Utilitarianism and on Liberty, ed. M. Warnock, 88–180 (Oxford: Blackwell, 2003 [1863]), 135. 16. J. Skorupski, John Stuart Mill (New York: Routledge, 1989), 33. 17. It may seem counterintuitive to introduce Mill into any argument on China and autonomy because of his very negative view of China as the antithesis of the “character” that he recommends. For this view, see Mill, On Liberty, 142, 144–145, cf. 161, 165, 174. Nonetheless, this negative view is separate from the merits of his account of agency and autonomy. 18. J. B. Schneewind, “Autonomy after Kant,” in Kant on Moral Autonomy, ed. O. Sensen, 146–168 (Cambridge: Cambridge University Press, 2010). 19. H. G. Frankfurt, The Importance of What We Care About: Philosophical Essays (Cambridge: Cambridge University Press, 1988). 20. H. G. Frankfurt, “Freedom of the Will and the Concept of a Person,” Journal of Philosophy 68.1 (1971): 5. 21. Ibid., 6. 22. Ibid., 11. 23. J. D. Velleman, “The Way of the Wanton,” in Practical Identity and Narrative Agency, ed. C. Mackenzie and K. Atkins, 169–192 (New York and London: Routledge, 2008). 24. Frankfurt, “Freedom of the Will and the Concept of a Person,” 14. 25. Graham, Reason and Spontaneity, 7, discussed earlier. 26. D. Hume, A Treatise of Human Nature, ed. L. A. Selby-Bigge, 3 vols. (Oxford: Clarendon, 1960 [1888]), 1.4.6:252. 27. S. Gallagher, “Philosophical Conceptions of the Self: Implications for Cognitive Science,” Trends in Cognitive Science 4.1 (2000): 14–21. 28. S. Gallagher, How the Body Shapes the Mind (Oxford: Clarendon Press, 2005), 2. 29. Ibid., 24. 30. Ibid., 24. 31. Ibid., 26. 32. Ibid., 32, 35. 33. A. Damasio, Somatic Markers and the Guidance of Behavior (New York: Oxford University Press, 1991), 217–299; A. Damasio, Descartes’ Error: Emotion, Reason, and the Human Brain (New York: G. P. Putnam’s Sons, 1994). 34. G. Loewenstein, E. Weber, C. Hsee, and N. Welch, “Risk as Feelings,” Psychological Bulletin 127.2 (2001): 267–286. 35. For the proto-self, see A. Damasio, The Feeling of What Happens: Body and Emotion in the Making of Consciousness (New York: Harcourt Brace, 1999), and J. Panksepp, Affective Neuroscience: The Foundations of Human and Animal Emotions (New

228 | Lisa Raphals York and Oxford: Oxford University Press, 1989) and “The Periconscious Substrates of Consciousness: Affective States and the Evolutionary Origins of the Self,” Journal of Consciousness Studies 5.5–6 (1989): 566–582. 36. See G. Northoff et al., “Self-Referential Processing in Our Brain: A MetaAnalysis of Imaging Studies on the Self,” NeuroImage 31 (2006): 440. For “minimal” and “narrative” self, see Gallagher, “Philosophical Conceptions of the Self,” and H. L. Gallagher and C. D. Frith, “Functional Imaging of ‘Theory of Mind,’ ” Trends in Cognitive Science 7.2 (2003): 77–83. For core or mental self and autobiographical self, see Damasio, The Feeling of What Happens. 37. G. Northoff and F. Bermpohl, “Cortical Midline Structures and the Self,” Trends in Cognitive Sciences 8 (2004): 102–107; Northoff et al., “Self-Referential Processing in Our Brain”; A. D’Argembeau, P. Ruby, et al., “Distinct Regions of the Medial Prefrontal Cortex Are Associated with Self-Referential Processing and Perspective Taking,” Journal of Cognitive Neuroscience 19.6 (2007): 935–944. 38. S. Gallagher and D. Zahavi, “Phenomenological Approaches to SelfConsciousness,” The Stanford Encyclopedia of Philosophy, ed. E. N. Zalta, https://plato. stanford.edu/entries/self-consciousness-phenomenological/; D. Legrand, “How Not to Find the Neural Signature of Self-Consciousness,” Consciousness and Cognition 12.4 (2003): 544–546, and “Being a Body,” Trends in Cognitive Science 9.9 (2005): 413–414. 39. Cortical midline structures include the medial orbital prefrontal cortex (MOFC), the ventromedial prefrontal cortex (VMPFC), the sub/pre- and supragenual anterior cingulate cortex (PACC, SACC), the dorsomedial prefrontal cortex (DMPFC), the medial parietal cortex (MPC), the posterior cingulate cortex (PCC), and the retrosplenial cortex (RSC). See Northoff and Bermpohl, “Cortical Midline Structures and the Self,” and Northoff et al., “Self-Referential Processing in Our Brain,” 441–442. 40. Northoff et al., “Self-Referential Processing in Our Brain,” 446–449. 41. B. Libet, “Unconscious Cerebral Initiative and the Role of Conscious Will in Voluntary Action,” Behavioral and Brain Sciences 8 (1985): 529–566; cf. Gallagher, How the Body Shapes the Mind, 238. 42. For a detailed account of this evidence, see L. A. Raphals, “The Zhuangzi on ming: Perspectives and Implications,” in What Is Philosophy: China, ed. R. Gassmann and R. Weber (Leiden: Brill, 2014). 43. Zhuangzi 21, 714; cf. Graham 1981, 131. 44. Zhuangzi 1, 39. 45. For example, oxen with white foreheads, pigs with upturned noses, and humans with piles cannot be used as sacrificial victims (4, 177), but they do not choose these features. 46. First published in H. von Arnim, Hierokles: Ethische elementarlehre (Pap. 9780) (Berlin: Klassikertexte, 1906), Heft iv, just after his Stoicorum Veterum Fragmenta. This absence may have contributed to his inaccessibility; cf. A. A. Long, “Hierocles on oikeiōsis and Self-Perception,” Stoic Studies (Berkeley and Los Angeles: University of California Press, 1996), 251–253. 47. Hierocles I.35–40, text and translation by G. Bastiannini and A. A. Long, “Hierocles’ Elementa Moralia,” in Hierocles the Stoic: Elements of Ethics, Fragments, and Excerpts, ed. I. Ramelli, trans. D. Konstan (Atlanta: Society of Biblical Literature, 2009), 4–5; cf. von Arnim, Hierokles, and A. A. Long, “Representations of the Self

Reason and Spontaneity Reconsidered  | 229 in Stoicism,” in Psychology: Companions to Ancient Thought, ed. S. Everson, 102–120 (Cambridge: Cambridge University Press, 1991), 1, 107. 48. Hierocles I.50–55, trans. Bastiannini and Long, “Hierocles’ Elementa Moralia,” 4–5. 49. Hierocles VI.10–15, trans. Bastiannini and Long, “Hierocles’ Elementa Moralia,” 16–17. 50. Aristotle (De sensu 7 448a26) uses the phrase autou aisthanesthai, but clearly of a human being. For claims that oikeiōsis is a Stoic invention, see C. O. Brink, “Οἰϰείωσις and Οἰϰειότης: Theophrastus and Zeno on Nature in Moral Theory,” Phronesis 2 (1956): 123–145, and Long, “Hierocles on oikeiōsis and Self-Perception,” 250–254. For claims for a peripatetic origin, see H. von Arnim, “Arius Didymus’ Abriss der peripatetischen Ethik,” Sitzungsberichte der Academie Wien 204.3 (1926). Brink, “Οἰϰείωσις and Οἰϰειότης,” presents a detailed history of the issues and evidence. 51. Long, “Hierocles on oikeiōsis and Self-Perception,” 256. 52. J. Brunschwig, “The Cradle Argument in Epicureanism and Stoicism,” in The Norms of Nature: Studies in Hellenistic ethics, ed. M. Schofield and G. Striker, 113–145 (Cambridge: Cambridge University Press, 1986), 137; Long, “Hierocles on oikeiōsis and Self-Perception,” 258. This term was used by the neurologist Charles Sherrington to distinguish between exteroceptive, interoceptive, and proprioceptive perception. These referred to the sensation of external stimuli (such as vision, hearing, etc.), internal sensations, and “muscular sensations” concerned with the mechanics of locomotion. See C. Sherrington, The Integrative Action of the Nervous System (New Haven: Yale University Press, 1906), 116, 131–135 (reflexes), 308, 316–320 and especially 336–345 and 347–349. 53. Long, “Hierocles on oikeiōsis and Self-Perception,” 258–260. 54. C. Darwin, The Expression of Emotions in Man and Animals (London: John Murray, 1872; rpt. New York: Oxford University Press, 3rd ed., 1998), 127. 55. B. J. Baars, “Subjective Experience Is Probably Not Limited to Humans: The Evidence from Neurobiology and Behavior,” Consciousness and Cognition 14 (2005): 7. We can infer human subjective experiences from behavioral and brain evidence, and similar evidence exists for other mammals and perhaps other nonmammalian animals. But biological evidence suggests that subjectivity may be conserved in species with humanlike brains and behavior. 56. Panksepp, Affective Neuroscience, “The Periconscious Substrates of Consciousness,” and “Affective Consciousness” (Panksepp’s research in this field is too extensive to quote in full). For discussion of animal cognition, see J. Parvizi and A. Damasio, “Consciousness and the Brainstem,” Consciousness and Cognition 14 (2005): 135–159; A. K. Seth et al., “Criteria for Consciousness in Humans and Other Mammals,” Consciousness and Cognition 14 (2005): 119–139; and S. R. Taylor, W. Parker, R. W. Mitchell, and M. L. Boccia, Self-Awareness in Animals and Humans: Developmental Perspectives (Cambridge: Cambridge University Press, 2006). 57. Panksepp, “Affective Consciousness,” 39. 58. These core emotional networks shared by all mammals include (using Panksepp’s capitalization convention) FEAR, SEEKING, anger-RAGE, sexuality-LUST, nurturance-CARE, separation distress-PANIC, and joyful PLAY. Panksepp, Affective Neuroscience and “Affective Consciousness.”

230 | Lisa Raphals 59. Panksepp, “Affective Consciousness,” 64. 60. Northoff and Bermpohl, “Cortical Midline Structures and the Self”; Northoff et al., “Self-Referential Processing in Our Brain”; G. Northoff and J. Panksepp, “The Trans-Species Concept of Self and the Subcortical-Cortical Midline System,” Trends in Cognitive Sciences 12.7 (2008): 259–264. 61. See I. D. Couzin, “Collective Cognition in Animal Groups,” Trends in Cognitive Sciences 13.1 (2008): 36–43. For supporting studies, see I. D. Couzin, “Collective Minds,” Nature 445 (2007): 715, 715; I. D. Couzin and J. Krause, “Self-Organization and Collective Behavior in Vertebrates,” Advances in the Study of Behavior 32 (2003): 1–75; L. Conradt and T. J. Roper, “Consensus Decision-Making in Animals,” Trends in Ecology and Evolution 20 (2005): 449–456; and D. J. T. Sumpter, “The Principles of Collective Animal Behavior,” Philosophical Transactions of the Royal Society B: Biological Sciences 361 (2006): 5–22. It has also been proposed that there are important commonalities between neuronal processes and collective animal behavior. For turning behavior, see I. D. Couzin et al., “Effective Leadership and Decision-Making in Animal Groups on the Move,” Nature 433 (2005): 513–516.

11

Spontaneity and Marriage

Paul Kjellberg

Between Chou [Zhou] and the butterfly there was necessarily a dividing: just this is what is meant by the transformation of things. —A. C. Graham, Chuang-tzŭ: The Inner Chapters

Introduction: The Setting A few years after my wife and I were married, an old friend came to visit with his girlfriend. Later in the afternoon when we were alone, he asked how I liked being married and I said very much. Then he said, somewhat predictably, that he had thought about it and believed that he loved his girlfriend but that he wasn’t sure he’d feel the same way in the future. His remark struck me as simultaneously familiar and strange. I remembered having been the sort of person who would say the same thing. At the same time, I felt that I had changed and was now a different type of creature. “Marriage isn’t like that,” I thought but didn’t say. “It’s not a prediction but a vow. I know I will always love my wife because I have promised to. There is no guesswork involved.” This putative exchange stuck in my mind because, though I meant it, it seemed inconsistent with other things I thought I believed. Specifically, it sounded like a repudiation of the value I had always placed on openness to change and the unknown. This ideal of life as a process of ongoing transformation was inspired by the Daoist philosopher Zhuangzi, or Zhuang Zhou (“Chou” in the epigram to this chapter), who awoke from a dream about a butterfly unable to determine if he was a person having dreamt a

231

232 | Paul Kjellberg butterfly or a butterfly dreaming a person. Zhuangzi’s modern student and our teacher, Angus Graham, explained this as a reliance on spontaneity to support and to guide us beyond the limits of our former experience. On the face of it, openness to change is the opposite of commitment to a vow. As I thought about the discussion with my friend, marriage seemed incompatible with spontaneity. The following essay is my effort to sort this conflict out.

The Background: Graham’s Anti-Rationalism In Reason and Spontaneity: A New Solution to the Problem of Fact and Value, Graham lays out a model of moral reasoning in which all choices are ultimately decided on the basis of spontaneous inclinations. Though Graham draws the solution from his studies in Chinese philosophy, he presents the problem in terms of the history of Western thought: Throughout the two and a half millennia of the history of Western philosophy it has been supposed that, once the key is found, it will be possible to establish rationally the ends which man should pursue. But when it became clear that Kant failed to establish a logically a priori imperative, after Hume’s separation of “is” and “ought” had discredited in advance any ethic based on theological, physiological, sociological or other factual generalizations about actual goals, it became impossible to sustain this hope. Reason can find the means to already accepted ends, and can reconcile or decide between ends, but the ends themselves spring from somewhere beyond its range.1 Western philosophy, according to Graham, has traditionally assumed that ultimate values must be established in one of two ways, based either in experience or on pure a priori reflection. After Hume invalidated the first and Kant failed at the second, the modern West found itself in a crisis due to its inability to establish any absolute ground for moral values. It is worth reviewing the arguments here both because, though familiar to some, they are increasingly shrouded in the mists of time and because Graham’s understanding of the problem is important for an appreciation of his proposed solution. Hume’s argument is not technically an argument so much as an observation: that arguments beginning with accounts of what is end with conclusions about what ought to be, without ever explaining how they make the transition.2 Hume does not deny that we should make this transition; indeed, he does so himself regularly. His observation simply made it impossible for people afterward to pretend that they had a rational

Spontaneity and Marriage  | 233 foundation for doing so. That is, no account of the facts, however complete or accurate, by itself determines what people should do on the basis of those facts. Staring into the sun will blind you. But whether you choose to stare into the sun is still a different question, depending, for instance, on whether or not you are Oedipus. Even your desire to be blinded or not does not settle the issue since the fact that you want something does not necessarily mean you should have it. Facts and values are two different things, according to Hume, and no accumulation of the first will get you to the second. The far-reaching implications of Hume’s observation are demonstrated by the many names his distinction goes under in our modern lexicon: between is and ought, fact and value, description and prescription. Kant accepted Hume’s distinction and was prompted by it to explain the possibility of moral obligation (“ought”) without basing it on any particular fact (“is”). He described his approach as “a priori” in the sense that it would have to come before—that is, be logically prior to—any particular set of facts. He does this by basing his argument not on any assertions about the way the world is but on the way we would have to think of it in order to conceive of it as a world in which morality exists. He concludes that morality only makes sense if we think of people as capable not just of doing what they desire, but of doing the right thing, what anyone ought to do under similar circumstances. In one famous formulation, he says, “Act only on that maxim [principle] through which you can at the same time will that it should become a universal law.”3 He describes this imperative as “categorical,” meaning “across the board,” in the sense that it underlies all other particular moral judgments. This requirement to act universally was to him a purely rational moral value not dependent on any particular facts. Not everyone follows Kant’s argument or agrees with his conclusion. Even those who do still have to acknowledge, as Kant himself admits,4 that he has not proven that morality in this form actually exists, only described what it would have to be in order to exist. This is what Graham means by describing Kant’s attempt as a failure. Given the presumption that the foundation for moral judgment must take one or the other of these forms—based either on facts or on reason alone—the discovery that neither one of them worked left Western philosophy in a crisis. The project of digging a rational foundation for morality had in effect run out of dirt. Graham distinguishes three different positions in the fallout of this crisis that he calls “rationalism,” “irrationalism,” and “anti-rationalism.”5 Rationalism is the belief that ultimate values can be established either by reasoning about facts or on the basis of purely rational reflection. This was the belief, so long cherished by the Western tradition, that Graham argues is no longer tenable in the modern world. Irrationalism is one response to reason’s failure to establish value in facts, either a priori or empirically.

234 | Paul Kjellberg According to the irrationalist, not only are valuations not determined by logic or by facts about the world, they are entirely independent of them. Values are not found but created; things are important because we care about them and not vice versa. The result is a subjectivism in which “intensity is put before awareness.”6 Graham associates irrationalism with romanticism, though the equation of it with existentialism might be equally appropriate. The Western world is caught on the horns of this dilemma, according to Graham, either committed to the discredited task of establishing values in facts or else jettisoning facts altogether and plunging into moral subjectivism and solipsism. Graham begins his search for a third alternative with a thought experiment, eschewing principles and focusing instead exclusively on facts: “Let us start by imagining a manner of life in which  .  .  .  I appeal only to what is, never to what ought to be.”7 The basis for judgment on this scheme would not be some previously held belief, whether based on reason, experience, or passionate intensity, but rather one’s spontaneous responses to the circumstances as they present themselves. Of course, beliefs, reasons, experience, and passion play a role in all this; they are part of what one is responding to. But the choice cannot be reduced to them. The determining factor is something else, which Graham calls “spontaneity.” Graham defines “spontaneity” as what takes place in the absence of deliberate decision-making: “We describe as ‘spontaneous’ all activity which is not the result of a considered choice.”8 Spontaneity is thus opposed to considered choice or deliberation: “Without resorting to principles or moral decisions, one is simply moved in one direction or the other in greater or lesser awareness.”9 Facts by themselves do not determine our responses to those facts, as Hume observed. Indeed, even our feelings do not determine whether or not we will follow those feelings. What is left over can only be a spontaneous response.10 At this point, it might seem as though Graham’s spontaneous response is totally arbitrary, but this is importantly not the case. As a response to circumstances, the value of spontaneity will depend on awareness of those circumstances: As long as I stay confined to this moral nihilism, the single unquestioned imperative will be to act intelligently.  .  .  .  let us pick “Face Facts” [as] a roughly formulated injunction which carries with it its own punishments for disobedience, without any need to appeal to God, philosophers, or the police.11 Elsewhere he formulates this principle as “Be aware!” Graham refers to this third alternative as “anti-rationalism.” Reason is concerned exclusively with matters of fact. Values, according to Graham, are not themselves facts and

Spontaneity and Marriage  | 235 so are by nature beyond reason’s compass. But although they are not facts themselves, they are responses to facts. Consequently, they can be objectively assessed, since, of two differing responses, the one is preferable that is based on the most accurate and complete representation of the facts as they stand. “ ‘Be aware’ will prefer the intelligent to the stupid reaction.”12 Anti-rationalism, on Graham’s reading, recognizes the limitations of reason while still acknowledging the important role it has to play. Anti-rationalism provides for some other source of values without slipping into the abyss of utter irrationalism. Admittedly, the number of facts is potentially infinite. We cannot be aware of all of them. Graham clarifies that one must be aware of all the relevant facts and defines as relevant “everything which would spontaneously move me one way or the other.”13 This includes not only how one happens to feel from one’s current viewpoint but how one is likely to feel in the future and also how one might feel from alternative viewpoints. “ ‘Be aware’ therefore prescribes the same neutrality for personal as for spatial and temporal viewpoints; it refuses a privileged status to ‘I’ as to ‘here’ and ‘now.’ ”14 Failure to consider the perspectives of other people is simple ignorance, willful or otherwise. Seeing an object clearly, for instance, involves not simply seeing it from one’s own perspective but understanding how it would look from other perspectives, which is how we know that distant objects are not in fact smaller. So, too, understanding a situation requires exploring it from all possible angles: “moving to the viewpoints of other persons is not in itself a moral act, any more than temporal viewpoint-shifting, so that a reduced capacity for either is not a moral but a cognitive defect.”15 What he is describing is not “empathy” because he is not counseling the adoption of another person’s perspective. He is not even dividing perspectives into “one’s own” and “other people’s.” They are all simply part of getting a complete understanding of the situation. The inclusion of the injunction to “Be aware!” means that there is more to Graham’s spontaneity than simply the absence of choice or deliberation cited earlier. Many things can hamper awareness beyond simple ignorance of the facts, such as the submission to habit. Lack of imagination can blind one to relevant viewpoints and possibilities. Complex and subtle decisions call for focused attention, as in performing brain surgery or Bach. On the one hand, this may tempt people to block out all the other “noise” and so violate the injunction to “Be aware!” But it does not require them to do so; in fact, being a successful surgeon or musician may require sensitivity to the unexpected inputs we have described. Graham’s spontaneity is not a rejection of rational processes so much as a refusal to be limited by them. In response to the criticism that such a spontaneous approach to life is somehow shallow, Graham gives a lengthy example to illustrate this decision-making procedure in practice:

236 | Paul Kjellberg To some temperaments it may seem that in crediting myself only with a capacity to choose between reactions which themselves are not chosen but caused, I would be representing myself as less than human. But what could be worthier of my human intelligence than to discriminate between and interrelate the parts of a painting or musical composition with the concentration of a mathematician solving a problem, yet respond to them with the immediacy of a dog pricking up its ears? Suppose that I am sitting by a bowl of fruit; my hand hovers over a pear, then a peach catches my eye; their distinctive flavours revive in memory and pull against each other, then my hand moves over and picks out a peach. Let us assume nothing but a causal connexion between the imagined tastes and the motion of the hand. I responded like a monkey to what I saw and smelled, but in full awareness of the two flavors, in obedience to “Be aware.” Could I in fact have chosen in a manner more worthy of my dignity as a rational agent? It will be said perhaps that I should have combined some principle of conduct with propositions about the flavours. But I do not even have a vocabulary to describe the distinctive tang of a pear or peach. The best I could do would be to say “I like peaches better,” but quite apart to the logical objection to deriving “Choose the peach” from this psychological statement, reliance on a generalization about my preferences could get me into a habit which would dim my awareness of the tastes, until I fail to notice that I no longer like peaches as much as I did, or that at this moment I hanker after a pear, so that the abortive try at rationalization would make my choice less intelligent.16 There is nothing shallow or irrational about the spontaneous decision described here. It seems like a perfectly legitimate approach, arguably the only available one. It may be objected that this example represents a judgment of taste, not a moral decision. This defect can be remedied without change to Graham’s argument. Perhaps I know that you love peaches too and that this is the last one; that you are starving; or that this is the last peach on earth, carefully preserved in a bowl of fruit in the United Nations Seed Bank. The same procedure would still be in play, just on a larger scale: duly considering all the different perspectives—mine, yours, future generations’—and then responding. If we add into the mix principles—do unto others as you would be done by, life should be preserved at all costs, future generations have no greater a right to the finer things in life than we do and someone’s got to eat that peach—what else could we do but be aware of them and

Spontaneity and Marriage  | 237 respond accordingly? (The poor blind mathematician solving the peach-pear problem would be lost if his seeing-eye dog did not prick up its ears.) As Graham says, what more could we ask of a rational agent than that she choose intelligently, and what more could we mean by “intelligence” in this context than awareness? Certainly, it would be wrong to be unaware or superficially aware. But beyond being aware of the alternatives and responding accordingly, the procedure resists systematization; any system we try to apply would have to be based on this foundation anyway. Failure on Graham’s model is primarily a matter not of moral depravity but of ignorance. Ignorance may be due to many factors: a simple lack of information, the refusal or failure to consider the information available, or the inability to deal with so much of it. Graham describes this last possibility: “The danger is of a breakdown, in the disorientation of excess of information, the compass needle going wild as multiple pulls drag it hither and thither, and then a self-defensive shutting off of information and lapse into irrationalism.”17 We often have multiple spontaneous responses simultaneously to different aspects of the situation. The resulting tension tempts us to turn our eyes from facts that confuse us. But while understandable, this is simply foolish. Note that “irrationalism” here is not illogic but unawareness, a failure or refusal to “Face facts.” Indeed, illogic need not be a problem in principle but only as it contributes to misinformation about the facts. Kant spoke about the danger of “misology,” hatred of logic and its demand that we universalize our actions. For Graham the danger is rather hatred of knowledge because it prompts us in so many different ways at once. Thus, there is a role for reason within spontaneity but it is a diminished one, auxiliary rather than determinative. Reason in the form of science, for instance, has proven tremendously successful in revealing facts. But the revelation of facts does not necessarily tell us what we should do with those facts; knowing how to build a bomb does not mean knowing how to use it. Reason also has the shortcoming of dealing with propositional knowledge, which, though powerful, inevitably falls short of the richness of actual experience: Why have we chosen to say “aware” when philosophers generally say “know”? Philosophy has an incorrigible bias toward thinking of knowledge in terms of the verbally formulable, the proposition.  .  .  . Awareness, though aided by propositional knowledge, is primarily of the concrete situation.18 Reason reveals facts, but so does art in its own way, drawing attention to perspectives and aspects of a situation we might otherwise not have appreciated. Giving reason a monopoly on truth paradoxically limits awareness,

238 | Paul Kjellberg thus encouraging stupid reactions. By contrast, limiting reason lets other kinds of knowledge compete for our attention and maximizes intelligence.19 Spontaneity as Graham understands it had no place in the traditional rationalist model. Simply spontaneous choices, ones for which no explanation could be given, must either be confined to nonmoral matters of taste or else considered irrational and immoral. By placing spontaneity not to the side of reason and morality but underneath them, he both finds a home for this valued part of our lives and simultaneously fills a void by providing the foundation that rationalism was missing. Though he calls the result “anti-rationalism,” it can be better understood as a grounded rationality, one that is located within our nonrational, prerational, or extrarational relations to the world.

The Quarrel: Spontaneity versus Marriage Now it is time to return from our long detour back to the original question, whether Graham’s version of spontaneity, which he calls “anti-rationalism,” is consistent with marriage as it is traditionally understood. On the surface, the two seem to conflict. Spontaneous responses are unpredictable. They vary with changes in the facts, and one relevant set of facts are people’s feelings. There would be no reason for the spontaneous person to stray once his or her feelings change. Marriage, by contrast, is based on a vow. But what reason is there to keep a vow if your feelings have changed? Kant argues that the institution of vows only makes sense on the presumption that people will keep them,20 otherwise they are only rhetorical flourishes like the promises of a mischievous child. But while the spontaneous person may choose to take this logic into account, he or she need not be bound by it if it is outweighed by other factors to the contrary. This is not to say that marriage needs to be a loveless commitment. Part of abiding by the vow may mean doing those things that encourage spontaneous feelings, like celebrating anniversaries, giving gifts, and deliberately seeing things from the other’s perspective. But the fact that these activities promote spontaneity suggests that they are prior to it, deliberate, and hence not spontaneous themselves. Ideally, one buys flowers spontaneously. Still, one would be wise to make a note on one’s calendar and to do so deliberately whether spontaneously moved or not. Let us call this position “guided spontaneity.”21 But this does not appear to be the spontaneity that Graham has in mind. Spontaneous responses still play a role in guided spontaneity, as reason plays a role in anti-rationalism, but they are not determinative. In guided spontaneity, spontaneous responses would still be determined by what they are guided by, that is, the vow. It is the vow that guides spontaneity, not vice versa.

Spontaneity and Marriage  | 239 Not only is marriage not completely spontaneous, it is not even clear we would want it to be. I do not remain faithful to my wife because, having considered an affair with every woman who steps in the room—everything that might spontaneously move me one way or the other22—I find myself inclined to be faithful. She would not want me performing that calculus even if it ruled in her favor, nor I her. Spontaneity and marriage seem to have run against irreconcilable differences.

Mediation: Levels of Spontaneity Before trying to arbitrate this dispute, there are several small questions raised by our outline of Graham’s position that need to be addressed. For instance, Is spontaneity monolithic? Graham normally refers to spontaneous reactions in the singular. For instance, “ ‘Be aware’ will prefer the intelligent to the stupid reaction,” suggests that there is only one of each.23 And yet it seems as though we often have a range of responses to a single situation. We might be simultaneously attracted, repelled, and unmoved on different levels. Spontaneity is an ocean, with different currents running at different depths. If we have conflicting spontaneous responses, how do we choose between them? Can they all necessarily be cashed out in terms of more and less aware and summed like vectors? We will have to wait to see if this becomes a problem. Similarly, one might ask, Are facts monolithic? It is notoriously difficult to distinguish the facts themselves from our characterizations of them. Even within a circumscribed body of facts there is a larger if not infinite number of interpretations that may demand consideration. In response to the question, “What are you doing?” the answers “I am writing a philosophy paper” and “Nothing, really,” might both be honest and accurate responses. Graham acknowledges this: Traditionally it has been assumed that thinking, whether about value or about fact, has to be fully detached from the spontaneous, which engages with it only as emotion biasing judgment. We have argued that for questions of value reason cannot achieve this independence. That the same is true of questions of fact is a conclusion that lately has become increasingly difficult to resist.24 That is, the traditional rationalist believed spontaneity had no role in the deduction either of values or facts. Feelings and emotions were only so much static. But Graham argues this traditional view is no longer tenable. That reason alone cannot deduce values has been shown. And it is becoming clear that reason alone cannot determine the facts, either. Some interpretation is

240 | Paul Kjellberg required in order to establish the facts. This takes place spontaneously. Feelings and emotions are part of the way we interact with the world. Rather than eliminating them, the point is to make sure that they are based on the facts. But this poses a problem: How can I respond spontaneously to the facts when the facts change with my spontaneous response?25 This puzzle becomes a problem because our responses are often based more on our characterizations of the situations than on the situations themselves. For instance, it may be a significant moment when I stop thinking of the woman next to me as my wife and begin to look at her as just another person with no greater or lesser claim over me than anyone else. In this case, we might almost say that the marriage has ended at that point, so that the interpretation determines the fact. Since my characterization of the facts may depend on my responses, how can I base my responses on the facts alone? A third set of questions pertains to Graham’s “roughly formulated injunction” to prefer greater over lesser awareness, which “carries with it its own punishments for disobedience.”26 The point may seem obvious but that need not exempt it from scrutiny. Perhaps a better way to get at this question is to ask: Where does the roughly formulated injunction come from? That is to say, is it a spontaneous preference or a principle derived in some other way? Would Kant’s impartial observer have purely rational grounds for preferring intelligent spontaneity over the stupid response?27 Phrased this way, however, the answer to our third question is straightforward: consistency demands that we interpret the injunction as spontaneous. Graham is not telling us we should be more spontaneous but informing us that we have been fundamentally spontaneous all along. Admittedly, his rhetoric on the point is ambiguous. For instance, in the preface, he says that “even our most deeply considered choices of ends are often (might it be always?) choices between goals which are themselves spontaneous.”28 Yet his tentative tone here is misleading. The answer to the parenthetical question is Yes; the point of his argument is that there is no “might” about it. In the subsequent paragraph he says: Although conscious of being less rational than the eighteenth and nineteenth centuries supposed, we are not yet accustomed to thinking of ourselves primarily as spontaneous beings, whose intellect, even if we happen to be philosophers, is in the conduct of life doing nothing else but choosing between the directions in which we are spontaneously pulled.29 But he is still talking here about the way we are accustomed to think of ourselves as being, not about the way we are. In the body of the work he expresses himself more straightforwardly:

Spontaneity and Marriage  | 241 A stream of spontaneous desire and aversion continue to pour into the centre of me, and I never cease to choose new ends from among its goals. Indeed, as far as my self-interest is concerned, on what principle could I choose as an end sufficient in itself something which I do not spontaneously want?  .  .  .  We are not advising the rational man to be any more spontaneous than he already is  .  .  .  We wish also to persuade him that once he takes fully into account that his ultimate goals are spontaneous (are, not ought to be; what they ought to be is intelligent as well as spontaneous), he will require no other principle than “Be aware” for choosing between them.30 If Graham were telling us to be spontaneous, he would encounter logical difficulties. He would be appealing either to our spontaneous feelings, in which case he would argue circularly, or to some principle, in which case he would be inconsistent. His argument makes the most sense interpreted in this way, not as telling us to be spontaneous but as informing us that we already are. Indeed, even the formulation of his point as an injunction is potentially misleading since it tempts readers to read the book in the imperative mood, as advocating spontaneity. And yet, if we look at the imperative itself, it is not telling people to be spontaneous, which the argument asserts they cannot help but be anyway; rather, the imperative tells them to be aware. This point about the injunction therefore tells us something important about how to read Graham’s book: he is not advocating for spontaneity versus deliberation so much as helping us to negotiate the struggle between spontaneous drives on different levels. In particular, he is negotiating a struggle between misinformed drives, which presume a rationalistic model of the self, with well-informed drives understanding the self as spontaneous. He is not telling us to be spontaneous, which we were already, like it or not. Rather he is helping us to be intelligent, helping us to “Be aware!” The rationalistic model of ourselves was misinformed and thus led, not away from spontaneity, but into clumsy spontaneity. By giving us a better picture of the facts, he enables us to be spontaneous better. For example, if we think of ourselves as rationalistic, we will assume that we should act on principle and therefore search for the principles on which we should act to the exclusion of all other things, such as feelings, changes in our situation, and so on. If we think of ourselves as spontaneous beings, “doing nothing else but choosing between the directions in which we are spontaneously pulled,”31 we will take these other factors into account and presumably make better choices. Or if, as rationalists, we strive for consistency, and if we think of ourselves as peach people, we will stubbornly

242 | Paul Kjellberg take the peach despite the fact that today we would prefer a pear. Graham is not telling us to do anything differently since, by his account, we have made our choices spontaneously all along. Rather, by illuminating the way that we choose, he means to help us do what we have done all along better, that is, with greater awareness, more intelligently. In addition to clarifying how to read Graham’s book, this point about the injunction also answers our earlier questions. We asked whether spontaneity was monolithic and whether acknowledging levels of spontaneity poses a problem for Graham’s theory. The consistent answer to both questions is No. There can be spontaneous responses at various levels, like ocean currents at different depths or weather patterns at different altitudes. How do we choose between two spontaneous responses? Spontaneously, inclining naturally in one direction or another as we are more and better informed—informed about the facts, about other perspectives, and most importantly about ourselves. We respond moment-to-moment, the way we drive. And we do so based on the information available to us. Some of this information is moment-to-moment, such as road and traffic conditions. Other information is comparatively static, such as our knowledge (or lack of it) of the workings of our engine. But the importance of that static knowledge is the role it plays in moment-to-moment judgments. Our second question was whether facts are monolithic. Graham seemed to acknowledge they are not, that the same situation can be characterized in multiple ways and hence yield different sets of facts. But he did not see that as a problem. Though he doesn’t offer his reasoning, we can fill it in for him. Yes, it is true that the world is ambiguous, often allowing of multiple descriptions. And yet, given this fact, what alternative is there but to become as aware as possible before responding? Here again, we have different spontaneous interpretations and we chose between them spontaneously. That much is inevitable. Graham’s point is to help us choose intelligently by informing us what we are about.

Reconciliation: Spontaneously Guided Spontaneity This more subtle reading of Graham’s project holds the key to reconciling spontaneity and marriage. We spoke earlier of the apparent incompatibility between according with a vow and openness to an ongoing process of change. Arguably, in accordance with the vow, one might be wise to cultivate spontaneous feelings through rituals such as anniversaries as well as less formal practices. One might make a point of always leaving the last peach, for instance, knowing that she likes them. We called this “guided spontane-

Spontaneity and Marriage  | 243 ity.” But to the extent that what guides spontaneity is not itself spontaneous, this would still not be what Graham has in mind. But our new reading of Graham may give us a tool to solve this problem. Given the fact that spontaneity is not monolithic, what guides a spontaneous drive—indeed, even what guides deliberation—need not be something unspontaneous but may simply be another spontaneous drive. We are composed of numerous spontaneous drives on numerous levels. These drives may be in conflict, and hence clumsy, or in harmony, hence intelligent. We need not think of the order imposed on these drives as an external force, like the coach on a pee-wee football team. The order may arise spontaneously from the process of life, like learning to dance. The function of weddings vows, anniversary gifts, and valentines—not to mention apologies and acts of contrition—would be to bring our superficial impulses into harmony with deeper feelings. They are, so to speak, expedient means: You cannot by mere exertion of the will hit on a new thought, love your wife, have faith in the cause, or repent the crime. Someone may ardently desire to create, love, believe, repent, but if his hope is fulfilled it will be because of a spontaneous process of maturation or crisis of conversion, which reason and will can coax but not force.32 The fact that these activities promote spontaneity need not entail that they are not spontaneous themselves, but rather spontaneous on a different, deeper, level. Earlier we said that guided spontaneity is not completely spontaneous. But what we have here is spontaneously guided spontaneity, which is not only spontaneous but, with luck, intelligently so. While guided spontaneity was not necessarily compatible with Graham’s position, spontaneously guided spontaneity is not only consistent with it but provides the most consistent reading of it. Understood in this way, it is finally possible to be spontaneous and married at the same time.

The Problem Resurfaces: Lingering Doubts All seems well again in our little home. The apparent tension between spontaneity and marriage has been resolved for practical purposes. But there are still theoretical problems that may reemerge. It forces us to think about marriage in a different way. On this model, marriage would still be a prediction. My putative statement to my friend, that I know I will love

244 | Paul Kjellberg my wife because I have promised to, will strictly speaking not be untrue. It must be reinterpreted as a prediction, though hopefully one that tends toward self-fulfillment, like the Little Engine chanting, “I know I can!” Weddings, rather than the union of two in one flesh, must be understood as akin to rain dances or pep rallies, as invocations of spontaneous feelings. Though this may seem like a terminological distinction, it makes a difference, for example, in the case of divorce. There will always be divorce however one approaches marriage: spontaneously, deliberately, or some combination of the two. Even the most well-intentioned and industrious relationships will sometimes not work out. But on the spontaneous model there will not necessarily be any grounds for regret: things simply changed. You may not have wanted them to change. You may have worked hard to keep them from doing so. But they did. It is just bad luck, something to be sorry about, perhaps, but no more a cause for regret than an unexpected change in the weather. The spontaneous person simply moves on. From the traditionalist’s perspective, however, divorce would be grounds for regret. Indeed, from this perspective it is not clear that the spontaneous person could even get married, at least not without a good deal of clarification for all parties involved, for example, that “I do” means “I hope I do.” You might say that spontaneous people can live together but can’t get married in the traditional sense. The wedding would function like italics, marking a change in emphasis but with no real meaning beyond that. Another way to think of this is as a forced redefinition of what we mean by “fidelity.” I speak here primarily of marital fidelity, though it has implications for religious thought, as well. If at our core we are, as Graham describes, streams of spontaneous desire and aversion pulling now in one way and then in another, it is not clear that we can really make a promise. For faith to be possible in either sense, religious or matrimonial, we have to be the sort of creatures for whom it would be meaningful to make a covenant in the first place, which we would not be if we are ultimately spontaneous as described. This last example suggests that, despite practical similarities between marriage and spontaneously guided spontaneity, there are underneath them very different conceptions of the self. It may be useful to distinguish between natural and social models of the self. The former is spontaneous, the latter deliberate. The spontaneous self is subject to ongoing change, while the social self strives for constancy. In this sense, the social self is fragile—it can fail—while the natural self simply becomes something new. The social self is temporary, while the natural self endures. The social self tries, while the natural self goes with the flow. The natural self just is, while the social self is something—a boss, a spouse, a congregant.

Spontaneity and Marriage  | 245 We might speak of them as different identities, though we should do so with a grain of salt. The natural self has no identity, strictly speaking, which is what allows it to be open to ongoing change. The social self, by contrast, takes a name. The natural self, we might say, identifies with the process; the social self with a shape the process takes. While similar for practical purposes, marriage and spontaneity, even spontaneously guided spontaneity, presuppose different notions of the self, different identities. Faced with these two models of the self, one might want to ask: Which one is right? Which one is the real self? But I doubt these questions can be answered in any useful way. Arguably, different conceptions of the self are appropriate in different circumstances. There is a lot to be said for the spontaneous self as we navigate our way through life’s peaches and pears. On the other hand, it is not clear that we would want soldiers on watch or police officers on duty—not to mention parents—to be constantly weighing their spontaneous inclinations to see in which direction they are pulled. Rather than asking which one is right, I want to ask a different question: What are the implications of identifying with one or the other? What difference does it make which way you think of yourself?

The Story of Zhuangzi’s Wife To explore this question in the spirit of Angus Graham, let us consider the story of the death of Zhuangzi’s wife. As is frequently the case, the differences between these conceptions shows up most clearly against the background of mortality. When Chuang-tzu’s [Zhuangzi’s] wife died, Hui Shih [Hui Shi] came to condole. As for Chuang-tzu, he was squatting with his knees out, drumming on a pot and singing. “When you have lived with someone,” said Hui Shih, “and brought up children, and grown old together, to refuse to bewail her death would be bad enough, but to drum on a pot and sing—could there be anything more shameful?” “Not so. When she first died, do you suppose that I was able not to feel the loss? I peered back into her beginnings; there was a time before there was a life. Not only was there no life, there was a time before there was a shape. Not only was there no shape, there was a time before there was energy. Mingled together in the amorphous, something altered, and there was the energy; by the alteration in the energy there was the shape,

246 | Paul Kjellberg by alteration of the shape there was the life. Now, once more altered, she has gone over to death. This is to be companion with spring and autumn, summer and winter, in the procession of the four seasons. When someone was about to lie down and sleep in the greatest of mansions, I with my sobbing knew no better than to bewail her. The thought came to me that I was being uncomprehending towards destiny, so I stopped.”33 There are many interpretations of this passage that raise interesting questions. For instance, why did Zhuangzi grieve in the first place? Was it the force of convention to which he momentarily succumbed and then subsequently freed himself, thus returning to spontaneity? Or was the grief itself a spontaneous but temporary response, which in time yielded naturally to a greater equanimity? Either way, commentators generally agree that Zhuangzi reconciles himself to his wife’s death by seeing her life as part of a larger process.34 It is interesting how Zhuangzi orchestrates this response. It is by viewing his wife as part of a spontaneous process that Zhuangzi redefines his own identity, either returning to spontaneity temporarily lost or connecting to a deeper spontaneity than he had known before. Graham describes this identification with the cosmic process: “In losing selfhood I shall remain what at bottom I always was, identical with everything conscious or unconscious in the universe.”35 The natural self identifies with the process; the social self with a shape the process takes. Zhuangzi relinquishes the latter to gain the former. And indeed, this is the attitude Zhuangzi expresses in the face of his own death. When his disciples want to bury him, he says, “Above ground I’ll be eaten by the crows and kites; below ground, I’ll be eaten by ants and molecrickets. You rob one of them to give to the other; how come you like them so much better?”36 Reconciliation with death is effected by surrendering attachment to one’s social identity to connect with the natural world. This is the only reference in the text to Zhuangzi’s wife. One wonders what kind of person shared a life with him, what role they played in each other’s lives, and what influence they had on one another. Graham and others have argued that Zhuangzi underwent a conversion experience.37 Given the text’s emphasis on ongoing transformation, it is unlikely that this was an isolated occurrence. And if Zhuangzi’s experience was developmental, if he grew and changed throughout his life, it seems likely that we would not have had our Zhuangzi without her. She must have been an interesting person. It is possible that she took a different approach to her death than he did. It is possible that she opted to identify with her social self, the one she shared with her husband and family. That is, rather than identifying with the process of change that would go on and become something new, she

Spontaneity and Marriage  | 247 staked her claim in the wife, mother, and friend that was about to pass. If so, death would have been hard for her, a cause for sorrow. If she did this, it would not have been because she didn’t realize what was at stake: a painful identity versus a tranquil identitylessness. She chose to spend her life with Zhuangzi, after all, so we can assume she understood as well as he did what was happening. She just chose to identify with her social rather than her natural self. She wasn’t interested in reconciliation with death. She loved their life together and would never willingly relinquish it. In a variation of the old cowboy motif, she died with her ring on. If so, what do we make of Zhuangzi’s wife? Was she simply responding in awareness, just differently from Zhuangzi? Or was she doing something else? Was this a spontaneous development or an act of will? But I fear we have come to the limits of our definitions here. Instead I would simply point out that these two definitions of the self, natural and social, have different costs and presumably also different rewards. Perhaps it is not a question of which self we are, but which self we wish to be.

Conclusion Returning to our original question, my friend said he thought he loved his girlfriend but was not sure he would feel that way in the future. I remembered having been the sort of person who would have said that but felt as though I had become a different kind of creature in the meantime. Whether or not any actual change occurred, it was certainly true that I had come to think of myself in a different way: I identified with a particular form rather than the process. Doing so created new rewards and also new risks. It was possible for me to fail now in a way it had not been before. The future horizon now included possible catastrophes where before there had only been transformations. Does thinking of ourselves in different ways make us into different kinds of creatures? That arguably goes beyond the scope of this essay and certainly of this author. But whether or not they make any difference in fact, it is clear that the commitments we make and what we mean by them make a difference to us, or at least to me now.

Notes   1. Angus Charles Graham, Reason and Spontaneity: A New Solution to the Problem of Fact and Value (London: Curzon Press, 1985), 156.   2. David Hume, A Treatise of Human Nature, ed. L. A. Selby-Bigge and P. H. Nidditch (Oxford: Oxford University Press, 1978), 469–470.

248 | Paul Kjellberg   3. Immanuel Kant, Groundwork of the Metaphysic of Morals, trans. H. J. Paton (New York: Harper Torchbooks, 1964), 88.   4. Ibid., 74.   5. Graham, Reason and Spontaneity, 159.   6. Ibid., 160–161, cf. 185, 188.   7. Ibid., 1.   8. Ibid., 4. This is not the only possible definition of “spontaneity.” Philip Ivanhoe identifies two definitions of spontaneity as 1. arising from natural impulse or 2. free from external compulsion (“The Values of Spontaneity,” in Taking Confucian Ethics Seriously: Contemporary Theories and Applications, ed. Philip J. Ivanhoe, Yu Kam-por, and Julia Tao [Albany: State University of New York Press, 2010], 183–184). The second of these is compatible with Kantian autonomy and is not what Graham has in mind here.   9. Graham, Reason and Spontaneity, 6. 10. Edward Slingerland examines spontaneity in Effortless Action: Wu-wei as Conceptual Metaphor and Spiritual Ideal in Early China (New York: Oxford University Press, 2003) and Trying Not to Try: The Art and Science of Spontaneity (New York: Crown, 2014). These studies are wide-ranging, covering not only the traditional Confucianism, Daoism, and Buddhism but modern neuroscience and psychology as well, and so would take us far afield in this essay. But they are recommended to interested readers. 11. Graham, Reason and Spontaneity, 1. 12. Ibid., 8. 13. Ibid., 7. 14. Ibid., 16. 15. Ibid., 29. 16. Ibid., 10–11. 17. Ibid., 52. 18. Ibid., 38. 19. Graham describes this position, which limits rather than rejects reason, as “anti-rationalism” (ibid., 159). 20. Kant, Groundwork, 90. 21. This is similar to Ivanhoe’s second definition of spontaneity, cited earlier. It is distinct from Graham’s position for the reasons given here. 22. Graham, Reason and Spontaneity, 7. 23. Ibid., 8. 24. Ibid., 52. 25. Though we might be tempted to think of this puzzle as distinctively “postmodern,” it is anticipated in pre-Qin China. The later Mohists tried to base their ethics on an analysis of the world based on similarities and differences between things to arrive at a correct system of names based on, so to speak, atomic facts (Angus G ­ raham, Disputers of the Tao [La Salle: Open Court, 1989], 183–186). Zhuangzi’s response is characteristically simple and to-the-point: “If you look at them from the viewpoint of their differences, from liver to gall is as far as from Ch’u [Chu] to Yueh [Yue]; if you look at them from the viewpoint of their sameness, the myriad things are all one” (Graham, Chuang-tzŭ, 76–77; emphasis added). We cannot base our interpretations

Spontaneity and Marriage  | 249 on similarities and differences between things, because whether things are similar or different is itself a matter of interpretation. 26. Graham, Reason and Spontaneity, 1. 27. Kant, Groundwork, 61. 28. Graham, Reason and Spontaneity, vii. 29. Ibid., vii. 30. Ibid., 9. 31. Ibid., vii. 32. Ibid., 13. 33. Graham, Chuang-tzŭ, 123–124. 34. David Nivison says, “His wife’s death was part of the process that included her birth and her life” (“Hsun Tzu and Chuang,” in Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus Graham, ed. Henry Rosemont Jr. [La Salle: Open Court, 1991], 138). As Roger Ames puts it, “death is equated with the process of transformation itself” (“Death as Transformation in Classical Daoism,” in Death and Philosophy, ed. Jeff Malpas and Robert C. Solomon [New York: Routledge, 1998], 65). According to David Wong, “Zhuangzi becomes reconciled to her death because he identifies with the whole” (“The Meaning of Detachment in Daoism, Buddhism, and Stoicism,” Dao: A Journal of Comparative Philosophy 5.2 [2006]: 214). Lee Yearley describes an “intraworldly mysticism,” in which “each changing state is embraced and then surrendered” (“The Perfected Person in the Radical Chuang-tzu,” in Experimental Essays on Chuang-tzu, ed. Victor Mair [Honolulu: University of Hawai’i Press, 1983], 131 and 135). Amy Olberding resolves the tension between Zhuangzi’s initial grief and subsequent calm not by rejecting the local orientation but by locating it within a global one (“Sorrow and the Sage: Grief in the Zhuangzi,” Dao: A Journal of Comparative Philosophy 6.4 [2007]: 339–359). 35 Graham, Chuang-tzŭ, 23. 36. Ibid., 125. 37. Henri Maspero, Le Taoïsme et les Religions Chinoises (Paris: Gallimard, 1971), 215. Graham, Chuang-tzŭ, 117–118. David Nivison in Chinese Texts and Philosophical Contexts, 132–133. Philip Ivanhoe, “Zhuangzi’s Conversion Experience,” Journal of Chinese Religion 19.1 (January 1991): 13–25.

12

Rationalism and Anti-Rationalism in Later Mohism and the Zhua¯ngzıˇ

Chris Fraser

Introduction Zhuāng Zhōu: “I’m flattered you think some of the jottings in that pile of notebooks they named after me basically get the Way right. Personally, I couldn’t say whether I know which way is the Way or not. I just try to find my way along as smoothly as I can, without quite knowing how I manage it. “I’m puzzled why you call me an ‘anti-rationalist,’ though.1 I pointed out that any judgments we act on assume certain preconditions, which depend on the perspective we take and often shift over time. We’d be wise to treat our judgments as provisional, because the basis for them isn’t fixed or absolute.”2 A. C. Graham: “Of course. You urged us to attend to the situation and respond, rather than analyzing, following rules, or reasoning from general principles.”3 Zhōu: “I did suggest it’s best to stay flexible and not overthink things. But why does that make me an ‘anti-rationalist’? How I could be ‘anti’ something I never heard of, which nobody in my day ever talked about? ‘Rationalism’ holds that reason is the chief source of knowledge. It contrasts with empiricism, which holds that experience is the main source of knowledge. Whether reason is a special source of knowledge just wasn’t an issue for us back then.” Graham: “Maybe I’m using the word ‘rationalism’ differently from you. I’m referring to the ideal of ‘bringing all knowledge within the scope of reason,’4 paired with the conviction that ‘analytic reason’ or ‘the posing of

251

252 | Chris Fraser alternatives’ gives us knowledge of objective reality.5 You rejected attempts to identify the Way or guide action by biàn (distinction-drawing 辯) and clearly distinguishing shì-fēi (this/not 是非), right?” Zhōu: “Indirectly, sure. The outcome of biàn is always conditional, provisional, and incomplete. It might be acceptable in a particular context, for particular purposes, but it has no privileged or absolute status. But what’s the connection to rationalism?” Graham: “You forbid us from thinking about what to do, instead of simply responding to the situation spontaneously.”6 Zhōu: “Why would I forbid people from thinking about what to do? Some of the stories I collected suggest we perform best by getting out of our own way, putting worries aside, and not self-consciously directing our actions with the heart or mind.7 The idea is to act from a blank, focused state that lets our skills and creative knack for problem-solving engage with the situation.8 But the same stories also depict people thinking about how to act.9 Nothing mysterious here—anyone who plays a sport or a musical instrument knows what I’m talking about. While you’re learning or preparing, you may need to do some thinking and visualizing. Once it comes to the actual performance, you just focus, loosen up, and act. “But how does any of this make me an ‘anti-rationalist’? It doesn’t really concern whether reason is a source of knowledge or whether knowledge falls within the scope of reason.” Graham: “Sure it does. You reject biàn as a source of authoritative answers to normative questions, and you reject it as a decision procedure or action-guidance procedure for concrete contexts. So you reject reason as a basis for knowledge of the Way and how to follow it.” Zhōu: “Biàn is a process of judging what is or isn’t the same kind of thing, usually by reference to models, exemplars, or precedents.10 It comprises mainly analogical reasoning and judgment, often based on perceptual comparisons of similarity. A commitment to biàn as a source of knowledge is no closer to rationalism than to empiricism. I question the authority of biàn because there are a plurality of potentially justifiable ways to distinguish things as similar or different, and I reject it as a decision procedure because often the most effective way to act requires novel, creative, context-sensitive responses not captured by preconceived models or familiar distinctions. That’s not rejecting ‘rationalism.’ ” Graham: “Biàn employs reason to form judgments. So rejecting biàn is rejecting reason. You reject biàn as authoritative, so you qualify as an anti-rationalist.” Zhōu: “Biàn can include pieces of reasoning, of course. But confidence in biàn doesn’t commit you to rationalism. There’s a big difference between engaging in reasoning—which everyone does—and holding that reason is the

Rationalism and Anti-Rationalism in Later Mohism  | 253 fundamental source of knowledge—a view only some philosophers endorse. To hold that view, you probably need to have an explicit concept of reason, to be concerned with sources of knowledge, and to think there’s something distinctive about reason that makes it a good foundation for knowledge. Rationalism emerges from a particular sort of discursive context. Did early Chinese philosophy provide such a context?”

Huì Shı¯ and Rationalism Graham: “It did! Your friend Huìzǐ was a rationalist. He used analytic thinking to argue that ‘all division leads to contradiction and therefore everything is one.’ ”11 Zhōu: “That makes him a rationalist?” Graham: “He reaches that conclusion by reasoning. He starts from the assumption that divisions exist, derives a contradiction, and so concludes that actually they don’t exist and the cosmos is a unity. He uses reason to justify a view of reality that’s very different from the world we observe.” Zhōu: “Your interpretation seems a bit hasty. He says things like ‘Just as things are alive, they are dying’ and ‘The south has no limit yet has a limit.’12 You take him to be offering a reductio ad absurdum against divisions between things, but the reductio isn’t in the text. He might be implying only that distinctions can be drawn in a plurality of ways, some mutually incompatible. The implied argument isn’t that distinctions generate contradictions, so distinctions are mistaken and monism is correct. It’s that there are many scales or perspectives from which to deem things ‘the same’ or ‘different.’ From one of them, we can deem ‘the myriad things all the same’ or ‘heaven and earth one unit.’13 Maybe he considered this the perspective of the cosmos itself, or maybe he thought it just one more perspective, as defensible as any other.14 “Even if we accept your interpretation, it’s hard to see how Huìzǐ’s views amount to rationalism. He doesn’t identify reason as a special source of knowledge. Certainly, I didn’t see myself as engaged with him in a ‘controversy over the place of reason.’ ”15

Knowledge and Change in Later Mohism Graham: “Well, I mentioned Huìzǐ because you were friends. The clearest example of rationalism in Chinese philosophy is the Later Mohists. ‘They share with the Greeks the faith that all their questions can be settled by reason—in their own terms, by biàn.’ ”16

254 | Chris Fraser Zhōu: “All their questions, including questions like whether it’s raining right now? By reason alone?” Graham: “I should have said ‘fundamental questions.’ The point is that they aimed to establish Mohist teachings ‘on impregnable foundations,’17 grounding them in logical necessity and so giving them ‘a certainty  .  .  .  invulnerable to time.’ ”18 Zhōu: “Why would they be so concerned with necessity and certainty? Earlier Mohists thought it enough to argue that their Way was followed by Heaven itself and that it conformed to the ‘Three Models’—the precedent of the sage-kings, the stuff of people’s ears and eyes, and beneficial consequences when applied in government policy and the penal code. A focus on timeless, necessary truths of reason seems out of place in Chinese philosophy.19 Doesn’t Chinese thought focus on a conception of the Way as immanent in a dynamic natural and social world?” Graham: “Yáng Zhū’s ethical egoism sparked a metaphysical crisis that convinced the Later Mohists that justificatory appeals to Heaven were no longer cogent, because a criminal could argue it was Heaven’s intent that he fulfill his naturally endowed criminal nature.20 They sought new foundations for their ethics independent of the authority of Heaven. The search for such foundations raised worries about how temporal change undermined knowledge.”21 Zhōu: “But arguably the early Mohist appeal to Heaven was more of an expression or illustration of their ethics than a foundational justification.22 Heaven merely articulates their conviction that the right norms are those that impartially promote the benefit of all. It’s not obvious they’d perceive a Yangist doctrine of fulfilling our individual nature as threatening, rather than simply misguided.” Graham: “They explicitly mention this threat. They say, ‘Expounding for a criminal Heaven’s being right while his nature is criminal is singing Heaven’s being wrong.’ ”23 Zhōu: “That’s a single obscure line from a text you called ‘an assemblage of mutilated scraps.’24 There’s no context and no indication this point was so unsettling that it prompted a revolutionary turn in Mohist philosophy.” Graham: “Still, they never mention Heaven again.” Zhōu: “We can’t really infer much from that. In any case, what’s puzzling here is your claim that the relation between knowledge and change became ‘the deepest and most troubling of problems’ for the Mohists.25 I’d have said their deepest problem was how to explain and justify the basis for distinguishing kinds one way rather than another.”26 Graham: “They lived in an age of rapid social transformation in which ancient authority was no longer an adequate guide to conduct.”27 Zhōu: “But the Mohists as a philosophical movement just don’t seem worried by change. Consider their ‘Three Models.’ They never claim to

Rationalism and Anti-Rationalism in Later Mohism  | 255 follow ancient ways, only to ‘root’ their teachings in the sage-kings’ precedent. The other models—what people observe and what in practice benefits everyone—are explicitly empirical or pragmatic, not rationalist. The Mohists accept the observable, changing world as real. They welcome reform and innovation, provided it promotes the benefit of all.” Graham: “In Canon B10, they identify ‘having passed’ as a source of doubt.”28 Zhōu: “Sure, but it’s only one of four sources of doubt, along with accidental circumstances, inconclusive evidence, and causal overdetermination. Nothing they say calls attention to transience as especially troubling. A striking aspect of their epistemology is how unworried they are by any difficulties in attaining knowledge.” Graham: “Their explanation of perceptual knowledge shows a concern with time. Canon A5 indicates that perceptual knowledge requires the ability to describe a thing after having passed it. Canon B46 indicates that knowing ‘differs from perceiving in that it continues after perception is past.’ ”29 Zhōu: “Their point is just that knowledge obtained from perception can persist without continued contact with its cause. They credit us with perceptual knowledge if, having ‘contacted’ a thing in some way (A5), such as by our sense organs (B46), we can describe it, even if we’re no longer in ‘contact’ with it. In B46, they explain how perceptual knowledge can endure even when we no longer perceive the object. Suppose I know something by seeing it, using my eyes, by firelight. Just as we wouldn’t say it’s the fire that sees, the Mohists assert it’s not the eyes that see either—it’s I myself who does. So even when my eyes are no longer in contact with the object, I can still have perceptual knowledge of it, presumably by memory. “Far from being anxious about temporal change, the Mohists seem confident that perceptual knowledge normally endures over time.” Graham: “I’m impressed. How do you know so much about the Mohists?” Zhōu: “Many scholars back then were exchanging ideas about the kinds of topics that interested the Mohists. As you yourself showed,30 the author of the Zhuāngzǐ ‘Essay on Seeing Things as Equal’ was familiar with Later Mohist concepts and terminology.”

Transience and Permanence in Later Mohism Graham: “Still, I think the Mohists were concerned with transience, because it grounds their pivotal distinction between the temporarily ‘staying’ (zhǐ 止), in Canon A50, and the unending, the ‘necessary’ (bì 必), in A51.31 The necessary is the eternal, which is not subject to temporal change and is beyond doubt, according to Canon A83. It contrasts with ‘staying,’ which endures

256 | Chris Fraser for a period and then ends. Their conception of the necessary as eternal provides the basis for the rationalist justification of their ethics.” Zhōu: “I don’t see a contrast between the transient and the eternal here. Canon A50 explains that all ‘staying’ has ‘duration,’ and A51 explains that what’s ‘necessary’ is ‘not ending.’ ‘Staying’ and the ‘unending’ seem intertwined, not antithetical. The unending obviously has ‘duration,’ and some of what has ‘duration’ could be unending.” Graham: “If you look at its ‘explanation,’ A50 is about remaining something x and non-y, such as ‘ox’ and ‘non-horse,’ over a period of time. ‘Staying’ x comes to an end, as when an animal begins as a horse and ends as a non-horse.”32 Zhōu: “ ‘Horse’ and ‘non-horse’? That’s x and non-x, not x and non-y.” Graham: “Yes, that’s a case in which something ‘begins as x and ends as non-x,’ so it’s no longer x and non-y. That’s why it’s about transience.” Zhōu: “I don’t understand. The text doesn’t mention beginning as one thing and changing into something else. It seems to contrast a typical example of ‘staying’ with one of ‘not-staying.’ ‘Staying’ is illustrated by how two logically consistent terms, such as ‘ox’ and ‘non-horse,’ both fit a thing. ‘Not-staying’ is illustrated by how two logically contradictory terms, such as ‘horse’ and ‘non-horse,’ can never both ‘stay’ in the same thing. The text calls this ‘not-staying that endures.’ “Since ‘staying’ requires duration, nothing instantaneous can count as ‘staying.’ So the text notes that even in the ‘ox’ and ‘non-horse’ example, if we consider just a single instant, ‘ox’ and ‘non-horse’ don’t ‘stay.’ But otherwise they do ‘stay.’ There’s no implication as to whether this ‘staying’ ends or not. It could last indefinitely.” Graham: “To explain ‘not-staying that endures,’ the text gives the analogy of a person crossing a bridge. The person will reach the other side. Similarly, duration comes to an end.” Zhōu: “That analogy is just a parenthetical example to illustrate the contrast between something that has duration and something instantaneous, like an arrow flying by a pillar. So I wouldn’t read too much into it.” Graham: “We know ‘staying’ ends, because it contrasts with the ‘necessary,’ which is ‘not-ending.’ ” Zhōu: “This doesn’t seem a strict contrast. Isn’t ‘not-ending’ a subset of ‘staying’? Among the things that ‘stay,’ some might eventually ‘end’ and some not.” Graham: “In argumentation, ‘staying’ can refer to settling a term’s reference, such as by fixing its criteria of application.33 So it pertains to relations between ‘names’—words or terms—and objects. But ‘the necessary’ refers only to logical connections between names or causal ones between objects. It doesn’t refer to relations between names and objects. So there’s an important contrast.”

Rationalism and Anti-Rationalism in Later Mohism  | 257 Zhōu: “You’re right that the Mohists use ‘necessary’ or ‘must’ (bì) in talking about logical relations, such as excluded middle. A51 suggests that shì-fēi (this/not) are ‘necessary’ (bì), probably insofar as anything must be either shì or fēi. A74 indicates that of two contradictory terms, such as ‘ox’ and ‘non-ox,’ only one can fit a thing and the other ‘necessarily’ doesn’t fit. But A78 also uses ‘necessary’ of name-object relations. It states that any object ‘necessarily’ takes the completely general term ‘thing.’ In the case of ‘kind’ terms, such as ‘horse,’ having used ‘horse’ for an exemplar of the kind, we ‘necessarily’ use the same name of any similar animal.” Graham: “Actually, that part of A78 isn’t about relations between names and objects.” Zhōu: “It isn’t? The canon distinguishes three kinds of ‘names’—‘allreaching,’ ‘kind,’ and ‘private’—by what they refer to. ‘All-reaching’ names refer to everything, ‘kind’ names to all similar things that constitute a kind, and ‘private’ names to a single individual.” Graham: “You’re misreading the part about kind names, because you overlooked a quotation. The point is that using a kind name amounts to using the quoted phrase ‘like the object.’34 The text is referring to the necessary relation between using a kind name and saying ‘like the object’ of things of that kind, not the transient relation between names and objects.” Zhōu: “Quotation? The text reads, ‘As to that which is [yě zhě 也者] like the object, one necessarily uses this name.’35 Isn’t the point that for all objects that are relevantly similar, such as all horses, we must use the same name, such as ‘horse’? The ‘Canons’ usually indicate quotation by yuē (‘say’ 曰). There’s no yuē here.” Graham: “Yě zhě (也者) indicates quotation. Every time it’s used in the ‘explanations’ after a single word, that word is a quotation from the corresponding ‘canon.’ ”36 Zhōu: “If it’s just a single word each time, how do we know it’s a quotation? The purpose of the ‘explanations’ is to explicate the corresponding ‘canons,’ so of course they repeat key terms from them. Yě zhě is a widely used topicalization device in classical texts, especially when discussing abstract concepts or resuming a topic mentioned earlier. The yě nominalizes its complement, which the zhě topicalizes. Yě zhě probably functions the same way here. If so, then A78 does refer to some name-object relations as ‘necessary.’ Of course, it isn’t necessary that we call horses ‘horse.’ We could use another name. But once we dub them ‘horse,’ it’s ‘necessary’ that we call all animals of that kind ‘horse.’ “The ‘explanations’ of ‘staying’ and ‘necessary’ are obscure and lack context, so any interpretation is tentative. But I don’t see a sharp contrast between transient name-object relations and eternal name-name relations. At least there’s no reason to think this contrast is central to Later Mohist thought.”37

258 | Chris Fraser

A Fourfold Division of Knowledge? Graham: “There’s important contextual material you’re missing. Once we see that those two canons are about transience versus eternal necessity, we can see that the entire structure of the ‘Canons’ is organized around them. They are pivotal to the Mohist account of knowledge—a ‘vision of universal knowledge organized in four disciplines.’38 Two of these disciplines concern knowledge that is logically or causally necessary and eternal, in the sense identified in A51. Two concern knowledge that is transient, depending on the temporary ‘staying’ of names in objects, as indicated in A50.” Zhōu: “Four disciplines? The ‘Canons’ look at best only partly organized. There are obvious series on knowledge, ethics, biàn, geometry, optics, mechanics, and other topics. But there’s no conspicuous overall pattern, and many canons don’t seem to fit into any discernable scheme. The very first two, for instance, are unrelated to each other and to the series of four that follows. Canon A1 introduces two technical terms—‘minor cause’ and ‘major cause’—that never appear again.”39 Graham: “Canon A80 is the key. It identifies four areas of knowledge, and on close inspection the canons divide into just these four areas.” Zhōu: “A80? The key to the overall organization is buried in one canon in the middle of the corpus?” Graham: “You have to read carefully.” Zhōu: “Is this organizational framework announced anywhere? In a table of contents or a preface, maybe?” Graham: “No.” Zhōu: “Are the different sections divided off? Are there headings indicating where they begin?” Graham: “No.” Zhōu: “Suppose I keep a research notebook. I write a note identifying different areas of knowledge. I write other notes about various philosophical and scientific topics. When I get interested in something, I write a sequence of notes about it. Wouldn’t my notes tend to form groups corresponding roughly to the areas of knowledge I identified, without any deliberate organization?” Graham: “I suppose.” Zhōu: “So even if some groups of canons fall into the four areas identified in A80, this observation could be explained without assuming they were deliberately organized into four disciplines. The clustering could occur just because the writers wrote sequences of canons on whatever topics caught their interest.” Graham: “Maybe, as a matter of mathematical probability. Here, though, we have writers aiming to give a systematic treatment of knowledge. So they’d probably use the four areas as an organizing principle.”

Rationalism and Anti-Rationalism in Later Mohism  | 259 Zhōu: “Why do you think their aim is a systematic treatment of knowledge?” Graham: “Only such a treatment can meet the challenge of establishing Mohist doctrines on an indubitable, necessary foundation and resolve the threat to knowledge posed by temporal change.” Zhōu: “That’s just what I’ve been questioning! I doubt they seek an indubitable foundation or worry about temporal change.” Graham: “Well, it’s a holistic approach. The overall interpretation is justified by how plausibly it explains the parts. The interpretation of the parts is justified partly by how plausibly they fit together into a coherent overall interpretation.” Zhōu: “Then we need to consider how your hypothesis about the organizational framework helps explain the content of individual canons. What exactly is the organizing scheme?” Graham: “The ‘Canons’ as a whole divides into a series of ‘definitions’ and one of ‘propositions.’ Both then divide into four disciplines, corresponding to the four objects of knowledge listed in A80. These are ‘description’ or ‘discourse,’ the study of how to relate names to objects; ethics, the study of how to act; the sciences, the study of objects; and ‘disputation’ or ‘argumentation,’ the study of relations between names.40 The first two disciplines, ‘discourse’ and ethics, produce knowledge that is transient, enduring only as long as the names we use or the desires we affirm ‘stay’ in objects. The second two produce knowledge that is necessary and eternal—causally necessary, in the case of the sciences, and logically necessary, in the case of ‘argumentation.’ ” Zhōu: “Which canons fall under which disciplines?” Graham: “Altogether, the canons run from A1 to A98 and B1 to B82. The ‘propositions’ begin at A88. ‘Discourse’ includes A1–6 from the ‘definitions’ and A88–B12 from the ‘propositions.’ Ethics includes A7–39. The sciences—geometry, optics, mechanics, economics—include A52–69 and B17–31. ‘Argumentation’ includes A70–75 and B32–82.” Zhōu: “Hang on. I’m trying to keep track of which canons go where. There’s a jump here from A39 to A52, and another one from A75 to A88.” Graham: “Right. A40–51 are a different group. Those form a ‘bridging’ section that treats knowledge and change. Since the relation of knowledge to change is pivotal to the whole system, there’s a special section in the middle for it.” Zhōu: “So there are actually five divisions, not four?” Graham: “Plus the appendix.” Zhōu: “The appendix?” Graham: “Yes, A76–87 are an appendix to the ‘definitions.’ They list different uses of some key words.”

260 | Chris Fraser Zhōu: “So there are actually six divisions, not four?” Graham: “Six among the ‘definitions,’ yes, and four among the ‘propositions.’ ” Zhōu: “Four, one for each of the disciplines?” Graham: “Not exactly. There’s one for the sciences, one for ‘discourse,’ and one for ‘argumentation.’ There’s also one for the ‘bridging’ section.” Zhōu: “None for ethics?” Graham: “No, they didn’t need one, because they’d already written about ethics elsewhere. A collection of fragments about ethics is preserved in the ‘Greater Selection’ (Dàqǔ 大取).” Zhōu: “So your hypothesis is that the four areas of knowledge provide the underlying organizing principle. But you identify six, not four, thematic groupings among the ‘definitions,’ and two of these don’t correspond to anything in the ‘propositions.’ ” Graham: “That’s right.” Zhōu: “And by your count, twenty-four of the eighty-seven ‘definitions’ actually don’t fit into the four-way scheme.” Graham: “I didn’t say all the canons fit neatly into the four disciplines. I claimed that they form groups that run parallel through the ‘definitions’ and ‘propositions,’ and the four areas of knowledge provide an organizing principle that explains this.”41 Zhōu: “But since there are no ‘propositions’ for ethics, they don’t really run parallel, do they?” Graham: “Here’s an example. In the ‘definitions,’ A40–51 treat space, time, movement, and change, while in the ‘propositions,’ B13–16 treat space and time. Then, in the ‘definitions,’ the next group, A52–69, treats geometry, while in the ‘propositions,’ the next group, B17–31, treats optics, mechanics, and economics. See the parallels?” Zhōu: “These could certainly count as parallels if other parts of the scheme also fall into place. If not, though, these might be just unsurprising instances of the writers exploring some interrelated topics and then returning to them again later. Also, the main parallel here is the two treatments of space and time in your ‘bridging’ sections, which don’t fall into any of the four disciplines.” Graham: “Another supporting correlation is that each of the four disciplines aligns with a distinct source of doubt, as presented in B10—the ‘accidental’ for ‘discourse,’ the ‘undemanding’ for ‘ethics,’ ‘having passed’ for knowledge and change, and the ‘coinciding’ for the sciences.”42 Zhōu: “Does B10 indicate that the sources of doubt apply to just those disciplines?” Graham: “No.”

Rationalism and Anti-Rationalism in Later Mohism  | 261 Zhōu: “Couldn’t there be multiple sources of doubt for judgments in different areas? The passing of time could create doubt about whether something we said in ‘discourse’ remained correct, for example. That something happened accidentally could undermine a finding in the sciences. The coinciding of several factors could raise doubt about whether someone acted ethically or just from self-interest.” Graham: “Maybe. But there are four sources of doubt and four disciplines, so this seems a reasonable correlation.” Zhōu: “But the correlation isn’t exact. You have four disciplines but sources of doubt for only three. What produces doubt in ‘argumentation’?” Graham: “ ‘Argumentation’ concerns the necessary and eternal and so isn’t subject to doubt.” Zhōu: “What exactly is ‘argumentation,’ and how is it different from ‘discourse’?” Graham: “ ‘Argumentation’ is what the Mohists called biàn.43 It refers to arguing over alternatives to decide which is shì (this) and which fēi (not).” Zhōu: “How does ‘argumentation’ yield knowledge that’s eternal and indubitable? A major theme of the ‘Lesser Selection’ (Xiǎoqǔ 小取) is that, since the methods of biàn are analogical and depend on semantics as well as logic, they’re highly fallible, can’t always be applied, and must be examined carefully.” Graham: “The ‘Lesser Selection’ isn’t about ‘argumentation’ in this sense. It’s about ‘discourse.’ It concerns transient name-object relations, not eternal relations between names.’ Zhōu: “The opening words of the ‘Lesser Selection’ expressly announce that it’s about biàn. It begins, ‘Now as to biàn, we use it to clarify the divisions between shì and fēi.’ ” Graham: “As I reconstruct it, those aren’t the opening words. In any case, biàn in that context refers to ‘discourse,’ while biàn in the ‘Canons’ refers to ‘argumentation.’ These are distinct disciplines.” Zhōu: “This text makes good sense as it stands, doesn’t it? It doesn’t seem to need ‘reconstruction.’ But as to your interpretation, if these are distinct disciplines, why do the Mohists call them both biàn?” Graham: “Originally they used the word narrowly, but later they adapted it to include three of the four disciplines.”44 Zhōu: “But earlier Mohist writings, such as ‘Condemning Fate,’ also use biàn in a broad sense. So I suggest the Mohists, early and late, probably used biàn roughly the same way all along. In Canon A74, they say that biàn is ‘contending over “other,”’ as when one side calls a thing ‘ox,’ the other calls it ‘non-ox,’ and exactly one of the contradictory terms fits. In B35, they say it’s when one side calls a thing shì and the other calls it fēi. These passages

262 | Chris Fraser indicate that biàn in the ‘Canons’ concerns name-object relations. In your treatment of ‘discourse,’ or name-object relations, you cite A70–71 to show the Mohists explain whether a name fits an object by comparing the object to a model (fǎ 法) of the kind of thing denoted by the name.45 But A70–71 come from a series you label ‘argumentation,’ where they are part of an account of biàn. So when we examine the content of canons that treat biàn, the proposed distinction between ‘discourse’ and ‘argumentation’ collapses.46 Graham: “The two overlap, of course. But, thematically, one section of the ‘Canons’ tends to treat name-object relations, the other logical argumentation based on relations between names.” Zhōu: “These thematic generalizations seem questionable. The ‘definitions’ that supposedly treat ‘discourse,’ or name-object relations, actually treat ‘cause’ (A1), ‘unit’ (A2), and four terms related to thought and knowledge (A3–6). Of course, knowledge for the Mohists involves description, but these ‘definitions’ aren’t directly about name-object relations. Among the ‘propositions’ about ‘discourse,’ A93–B2 could just as well be about ‘argumentation,’ since they mention arguing with an opponent and apply concepts introduced in the ‘definitions’ about argumentation, such as comparison with models. “A31–32 are about name-object relations, but they’re in the section you designate ‘ethics,’ not ‘discourse.’ That section also includes canons about ‘life,’ ‘sleep,’ ‘dreaming,’ and time (A22–24, 33). The ‘bridging’ section on ‘knowledge and change’ is mainly about space, time, and movement and doesn’t actually mention knowledge. “By far the largest group is B32–82, which are supposed to be ‘propositions’ in the discipline of ‘argumentation’ based on necessary, logical relations between names. Instead, they look like substantive discussions of a medley of topics. Some fall under the sciences (B43, B47, B56). Some are about knowledge (B41, B46, B50, B70). Many don’t seem to fit into the four-way framework.” Graham: “They’re thematically diverse. But these diverse issues are generally treated as resolvable purely by considering names, without observation of objects. So they fall under ‘argumentation.’ ”47 Zhōu: “That’s true of some of them, such as B35. It applies the principle of excluded middle to argue that in biàn, one of the two opposing claims must ‘win.’ But others concern name-object relations and the status of objects, not just knowledge of names. Some seem to present substantive theories, either about things, how we know them, or how language relates to them. Examples include B43 on the ‘five phases,’ B46 on perceptual knowledge, B47 on fire being hot, B66 on finding the right criteria for distinguishing different kinds, and B70 on knowledge by explanation. I don’t see how these canons all concern knowledge of necessary logical relations between names.”48

Rationalism and Anti-Rationalism in Later Mohism  | 263 Graham: “The thematic cohesion within the divisions may not be obvious to the modern reader, who instinctively applies classifications from Western philosophy and science. This just shows that we need to read carefully, applying the Mohists’ own fourfold classification from Canon A80.” Zhōu: “But the Mohists themselves might take A80 to refer to four dimensions or aspects of knowledge, not four distinct disciplines or fields. Nothing in A80 itself indicates they regard knowledge as organized into four areas, two dealing with the transient, two with the eternal.” Graham: “That’s possible. But my proposal coheres with their classification and their distinction between the temporarily ‘staying’ and the unending ‘necessary.’ ” Zhōu: “Your interpretation may be internally coherent, but it doesn’t explain what we find in the texts very well. I doubted whether A50–51 are about the transient versus the necessary. You claimed that the fourfold organizational scheme would clarify things. But I don’t see clear thematic differences between the canons you assign to ‘discourse,’ the sciences, and ‘argumentation.’ ‘Discourse’ and ‘argumentation’ overlap so much in terminology and content that it’s highly improbable the writers distinguished between one discipline devoted to transient name-object relations and another yielding eternal, necessary knowledge of names. The hypothesis about the fourfold organization doesn’t support your interpretation of A50–51, it rests on it.”49 Graham: “You need to see the overall patterns and the parallels to grasp the organizing principles.” Zhōu: “Well, I just don’t see them as you do.”

The a priori Zhōu: “Let’s trace our way back to where we started. How does this elaborate interpretation show that the Later Mohists are ‘rationalists’?” Graham: “Through their discipline of ‘argumentation,’ which investigates logically necessary relations between names, they develop a body of knowledge grounded in reason. They give an a priori account of their ethics that places it in the realm of the eternal, necessary, and indubitable.” Zhōu: “What’s ‘a priori’?” Graham: “It refers to knowledge that’s justified independently of experience. It rests only on reasoning or rational reflection, along with understanding of the relevant language, concepts, or relations.” Zhōu: “Besides the four objects of knowledge, A80 identifies three sources of knowledge—hearsay, explanation, and personal observation. Knowledge by explanation (shuō 說) could conceivably include knowledge

264 | Chris Fraser obtained by reasoning. But according to B70, knowledge by explanation also includes knowing a hidden object’s color because someone explains it’s the same color as an object we can see. That’s not a priori knowledge. So in A80, where they list various aspects of knowledge, the Mohists don’t seem to recognize a priori knowledge.” Graham: “Yes, but elsewhere they refer to knowing something in advance of observing it. How can we know something before observing it? By reasoning from its definition. So I propose this refers to a priori knowledge.” Zhōu: “A priori knowledge isn’t knowledge of something in advance of observing it. It’s knowledge for which observation is irrelevant—like knowing that anything green is colored.” Graham: “The Mohists might conceptualize it differently.” Zhōu: “What are their examples?” Graham: “They give two. First, suppose there’s a circular object on the other side of a wall. According to Canon A93, when we jump over the wall, the circle ‘stays’ as we expect it, because we can ‘know beforehand’ (xiān zhī 先知) things that ‘follow from or exclude each other.’ I suggest the example refers to a priori knowledge about the circle obtained from its definition.”50 Zhōu: “Isn’t this scenario like the one in B70? Both suppose there’s an object we can’t see. In B70, we’re told it’s the same color as a white object we can see. In A93, we’re informed it’s a circle. According to B70, if we know the color of the unseen object is like that of the white object we do see, then we know the color of the hidden object, because ‘necessarily’ it’s white. B70 ties this point to Mohist philosophy of language. Using a name, such as ‘white,’ of something listeners don’t know informs them the thing is like other objects they do know that are denoted by that name. Similarly, in A93, if the unseen object is called a ‘circle,’ and we know the word ‘circle,’ then we know before seeing it what it’s like. This knowledge isn’t independent of experience. It’s from being told the object is a circle.” Graham: “My idea is that the example refers to inferring from the name ‘circle’ that the object will be round or have other features of a circle. That would be a priori knowledge.” Zhōu: “Then why mention jumping a wall? Why not just state that, since it’s a circle, we know it’s round?” Graham: “That’s what I think the example implies. The text mentions ‘following from or excluding each other.’ Being round follows from being a circle.” Zhōu: “ ‘Following from or excluding each other’ could plausibly refer to a relation of logical consequence, which could be grounds for a priori knowledge. But it could also refer to inductive or analogical inference. The Mohists’ own explanation of how we know what something called ‘circle’ will look like seems grounded in an implicit analogical inference.”

Rationalism and Anti-Rationalism in Later Mohism  | 265 Graham: “Consider the second example, then, from B57. If we deem a pillar round, then when we see it, there’s no change from our thought of it. We ‘know beforehand’ a mental image of the round pillar. We don’t know what its color or material will be, but we know what its shape will be like, from our understanding of what a pillar is and what round is.”51 Zhōu: “So if we know the words ‘pillar’ and ‘round,’ we can visualize the shape of something called a ‘round pillar’ before we see it. Isn’t this just knowledge of what ‘pillar’ and ‘round’ denote? Of course, in this scenario we could have a priori knowledge that, for instance, a round pillar isn’t square. But a remark about being able to visualize an instance of a familiar kind of object before seeing it seems a weak basis for crediting the Mohists with a conception of knowledge independent of experience.” Graham: “Consider the example of a circle. They define ‘same length’ (A53) and ‘center’ (A54) and from these define ‘circular’ (A58) as ‘having the same lengths from a single center.’ From these definitions, one could know by reason alone what a circle is before seeing one. So I suggest this amounts to a priori knowledge.” Zhōu: “Reasoning from definitions could indeed yield a priori knowledge. But does this sort of reasoning play any role in Mohist theories? Their account of how we identify a circle seems to be that we check whether candidate circles match a relevant model, such as a concrete exemplar of a circle, a compass, or a mental image of a circle.” Graham: “That would be for transient name-object relations, falling under ‘discourse,’ whereas a priori knowledge pertains to eternal knowledge of names, falling under ‘argumentation.’ ” Zhōu: “But the example of identifying a circle by reference to a compass or exemplar is from A70, which comes under ‘argumentation’ in your fourfold division. So if your organizational scheme is correct, the Mohists’ own treatment of how to identify a circle in ‘argumentation’ doesn’t apply the conception of a priori knowledge you attribute to them.”

A priori Grounds for Ethics? Graham: “Still, they do apply it to ground their ethics. They hold that the names of moral concepts have ‘essentials’ (qíng 情), or features expressed in their definitions, which determine whether the name fits a thing. A sage who understands these ‘essentials’ will desire and dislike certain things on behalf of all people a priori. By considering their ‘essentials,’ we can learn from the sage that these things are necessary, and thus they form an a priori basis for normative ethics.”52 Zhōu: “The canons say all that?”

266 | Chris Fraser Graham: “No, that comes from some fragments I pieced together from the ‘Greater Selection.’ But this interpretation emerges once you see how the Mohists are concerned with necessary, eternal knowledge, use ‘beforehand’ (xiān) as a technical term for a priori knowledge, and use qíng as a technical term for the essential features that define a name.” Zhōu: “So far, you haven’t convinced me they’re concerned with eternal knowledge or have a concept of a priori knowledge.” Graham: “You’re still struggling to see the big picture.” Zhōu: “We’ve talked about naming and argumentation, but you haven’t said anything about qíng before.” Graham: “Throughout pre-Han philosophy, qíng has a precise meaning similar to the Aristotelian notion of ‘essence.’ The qíng of x is all that’s conveyed in its definition, without which it would not be a genuine x, conceived of as something behind its form (xíng 形) and looks (mào 貌).53 You should know this—I learned it from you! You once said the sage lacks the qíng of a person, and Huìzǐ asked in that case how you could still call the sage a ‘person.’54 His question implies that normally qíng are the essential features that determine whether something takes a certain name.” Zhōu: “That’s odd. My interpretation of that conversation is the reverse of yours. Do you remember how I answered Huìzǐ?” Graham: “You said, ‘The Way gives him the looks, Heaven gives him the form, how could we not call him a “person”?’ ” Zhōu: “Exactly. I was invoking prevailing theories of naming. For the Mohists, ‘form’ and ‘looks’ are the basis for naming things like people. The ‘Greater Selection’ expressly mentions naming on the basis of ‘form’ and ‘looks.’55 The sage has the form and looks of a person, so the name ‘person’ fits him. Qíng has no role at all in Later Mohist philosophy of language. It doesn’t refer to the ‘essentials’ captured by a definition, because it is simply not part of their theory. The word qíng never occurs in any Mohist discussion of naming.” Graham: “How do you explain the passage about the sage having a priori knowledge of the ‘essentials,’ from which we learn what is necessary?” Zhōu: “You agree that qíng can refer to the facts of a situation.56 So that passage says, ‘As to all things the sage desires or dislikes in advance on behalf of people, people necessarily on the basis of their qíng—their situation or constitution—obtain them.’ The sage understands general features of people’s situation and so, in advance of considering our particular circumstances, he has certain desires on everyone’s behalf that we can feasibly satisfy. These are probably desires for the basic goods of Mohist ethics. The passage contrasts these with desires arising from people’s contingent circumstances, which may or may not be feasible to fulfill.

Rationalism and Anti-Rationalism in Later Mohism  | 267 “I wouldn’t invest much in this—or any—interpretation, though, as the passage is isolated and corrupt.” Graham: “Well, your interpretation may still work for my purposes. The sage, the wisest of persons, starts from these desires and dislikes on behalf of all, which are given either ‘a priori’ or ‘in advance,’ as you please. Working from this starting point, the Mohists establish the basic tenets of their ethics by reason alone. ‘Benefit’ and ‘harm’ are defined in terms of ‘desire’ and ‘dislike,’ and ethical terms such as ‘benevolence,’ ‘right,’ and ‘filiality’ are defined in terms of benefit and harm. The result is a systematic series of interlocking definitions that establishes a priori that the benevolent and the right are what will be desired on behalf of all by the sage—a rationalist foundation for Mohist ethics.”57 Zhōu: “How do they get from the sage’s desires and dislikes to ‘benefit’ and ‘harm’?” Graham: “They define ‘benefit’ as ‘what one is pleased to get’ (A26) and ‘harm’ as ‘what one dislikes getting’ (A27).” Zhōu: “So actually they explain ‘benefit’ in terms of pleasure, not desire.” Graham: “It amounts to the same thing, but since you can’t desire something you already have, they use ‘pleased’ instead of ‘desires.’ ” Zhōu: “But ‘desires’ and ‘pleased’ aren’t the same thing at all. You can desire something yet find you’re displeased to get it, and you can not desire something yet find you’re pleased to get it.” Graham: “Still, the Mohists can move from here to define the other key ethical concepts. For example, they define ‘benevolence’ (rén 仁) as ‘concern for units’ and ‘right’ (yì 義) as ‘to benefit.’ ” Zhōu: “How does equating ‘right’ with ‘benefit’ follow from an a priori derivation? That looks like just a blunt statement of the Mohists’ core ethical conviction. Do they argue that whatever the sage desires is right, the sage desires benefit for all, and therefore right is benefit?” Graham: “No, not exactly.” Zhōu: “What about benevolence? How do they move from benefit and harm to ‘concern for units’? Through a definition of ‘concern’ (ài 愛) and one for ‘unit’ (tǐ 體)?” Graham: “Right. They define ‘unit’ as ‘a division from a whole.’ ” Zhōu: “How about ‘whole’ and ‘concern’?” Graham: “We can make a good guess at the definition of ‘concern.’ It would be something like ‘desiring benefit and disliking harm to the person, on the person’s own behalf.’ ”58 Zhōu: “A guess?” Graham: “The ‘Canons’ don’t include those definitions. They omit words that appear in the titles of the ten core Mohist doctrines. For instance, ‘worthy’

268 | Chris Fraser (xián 賢), ‘aggression’ (gōng 攻), and ‘fate’ (mìng 命) are also missing. Probably there was another text, now lost, that defined these important terms.”59 Zhōu: “Is there any evidence such a text existed?” Graham: “No, but these terms are conspicuously absent from the ‘Canons.’ Since the Mohists aimed to produce a systematic account of ethical knowledge, they wouldn’t have omitted them. They must have been treated in another text.” Zhōu: “Wait a moment. The content of the ‘Canons’ turns out not to corroborate your interpretation of their purpose, but instead of revising your interpretation, you postulate a lost text that supports it? That’s a textbook example of an ad hoc rescue.” Graham: “Isn’t it strange that there are no canons on ‘fate,’ ‘concern,’ and so forth?” Zhōu: “We don’t know enough about the ‘Canons’ to draw conclusions about what’s strange or not. You assume the Mohists must have treated these topics because they were engaged in a systematic project with a particular aim. But maybe there was no systematic project or overarching aim. Maybe the ‘Canons’ are a haphazard collection of notes. That’s what they look like.” Graham: “Well, as I said years ago, ‘my explanations will not always convince others as easily as myself.’ ”60

Are the Mohists Rationalists? Zhōu: “Going back to my original question, suppose we accept your theory that the ‘Canons’ present a tightly wrought system of interdefined ethical concepts. Would the result really be a brand of ‘rationalism’? After all, the texts don’t explicitly mention a concept of reason or of knowledge grounded in reason. On your interpretation, their ethics is grounded in the sage’s desires on behalf of everyone. Nothing suggests these desires arise from reason. The reasoning part concerns the relation between concepts such as ‘benefit,’ ‘concern,’ and ‘benevolent,’ not the sage’s foundational attitudes of desire and dislike.” Graham: “My idea was that the sage’s desires and dislikes are independent of experience. He has them ‘in advance’ of any contingent circumstances.” Zhōu: “That doesn’t entail they’re based on reason. They could simply express basic, bedrock values.” Graham: “But the a priori nature of the system provides a rationalist argument that the sage—the most knowledgeable of persons—desires the benevolent and the right on behalf of all.61 This argument gives an authoritative justification for Mohist ethics by grounding it in necessary, eternal relations between concepts.”

Rationalism and Anti-Rationalism in Later Mohism  | 269 Zhōu: “I don’t see that. Earlier Mohists claimed the sage-kings cared about and promoted the benefit of all. On your reconstruction, the Later Mohists claim that a sage would desire and dislike certain things for all, on the basis of general knowledge of the human condition. Both claims appeal to the exemplary attitudes of ideal moral agents as fundamental models by which to support Mohist views. How does the later approach provide a more convincing justification than the earlier one? Both remain vulnerable to critics who dispute the Mohist view of benevolence and right or the status or attitudes of the sages.” Graham: “You don’t see any advantage to their project of classifying knowledge, identifying a realm of necessary, indubitable knowledge, and building an a priori ethical system? This is among the greatest achievements of Chinese philosophy.” Zhōu: “The hypothesis that they’re engaged in such a project just doesn’t correspond well to what we see in the texts. Their supposed ‘rationalism’ seems a projection of your own philosophical concerns, not something there in their writings.” Graham: “But surely there’s something distinctive about Later Mohist thought. I propose it’s their rationalism.” Zhōu: “They’re meticulous and methodical. They develop interesting, plausible theories to explain knowledge, language, and argumentation by appeal to analogical pattern recognition. They’re optimistic that biàn can clarify and resolve all sorts of epistemic, semantic, ethical, and political issues. They’re confident the world is organized into regular, coherent patterns by which to distinguish things into kinds and identify a beneficial, sustainable Way. They identify and apply logical norms such as noncontradiction and excluded middle, but are mainly interested in semantics and informal logic. Their theories and methods are a distinctive, impressive achievement, but they don’t add up to a species of ‘rationalism.’ ”

Anti-Rationalism or Expertise? Graham: “But it’s their rationalist theorizing that inspired your own incisive anti-rationalism.”62 Zhōu: “That’s what perplexes me. You call me an ‘anti-rationalist’ and suggest I mounted an ‘assault on reason.’63 I don’t see that I did.” Graham: “But you reject Mohist ethics, don’t you?” Zhōu: “Mainly because of their dogmatic confidence that a single, determinate value—the benefit of all—is sufficient to model the Way. I ridicule their dogmatism, narrowness, and epistemic optimism. None of this has much to do with ‘reason’ or ‘rationalism.’ ”

270 | Chris Fraser Graham: “Still, you advocate the antithesis of rationalism, so you’re an anti-rationalist.” Zhōu: “What’s the antithesis of rationalism?” Graham: “Spontaneity. You hold that we can coincide with the Way only by attending to the total situation and responding by means of an inexpressible knack, without analyzing, posing alternatives, reasoning from first principles, or following rules.”64 Zhōu: “How is that ‘anti-rational’? Suppose while driving home I see a group of children playing soccer. The ball rolls into the road and one of them darts after it. I immediately brake to avoid hitting her. I don’t consider alternatives, invoke rules, or deliberate from first principles. Still, in this scenario braking is rational, isn’t it? When we’re talking about action, ‘rational’ refers to our conduct being justified by or consistent with relevant considerations—reasons for or against doing one thing or another. Obviously, there are good reasons to avoid hitting the child.” Graham: “Yes, but your action isn’t guided by analyzing, posing alternatives, or reasoning. It’s a spontaneous response.” Zhōu: “Of course, I don’t engage in an explicit process of reasoning. But let’s not confuse explicit reasoning with being rational or conforming to reason. My braking is still rational. As I see it, what you’re calling ‘antirationalism’ doesn’t really address ‘rationalism,’ the doctrine that reason is the basic source of knowledge. Nor does it reject ‘rationality,’ understood as a normative relation between our actions and reasons for or against them. It seems to be the view that explicit reasoning or deliberation grounded in inflexible rules is an ineffective process for guiding action. But that view needn’t entail rejecting the norms of rationality or denying that reasoned deliberation can help us find our way in difficult situations.” Graham: “Still, reason can’t provide the wisdom needed to act adeptly. The fitting course—the Way—meanders, shifting direction in varying conditions. We miss it if we analyze, follow rules, or rigidly adhere to explicitly formulated codes.65 Isn’t that your view?” Zhōu: “As we encounter it, the Way often twists and shifts, sure. Flexibly adapting to what we find generally serves us better than stubbornly following a fixed, predetermined model.” Graham: “Right, so I credit you with the view that we can coincide with the Way by ceasing to draw distinctions and just using our spontaneous aptitude, our ‘potency’ (dé 德), to respond appropriately to particular situations.”66 Zhōu: “But doesn’t all action involve drawing distinctions? When I brake for the child, I distinguish between the child and an open road. When Cook Ding carves up an ox, he distinguishes between the meat and his fin-

Rationalism and Anti-Rationalism in Later Mohism  | 271 gers. When the whitewater swimmer navigates the rapids, he distinguishes between water and boulders.” Graham: “It’s all right to make fluid distinctions that vary with circumstances. Thinking goes wrong when we make rigid distinctions that mislead us into judging things are permanently what we temporarily find it convenient to name them.67 That’s why you forbid us from thinking about what we ought to do. To avoid being misled by rigid distinctions, we should simply respond with spontaneous action or spontaneous approval or disapproval.”68 Zhōu: “Hold on, that doesn’t follow. Your main point is that adept action issues from an intelligent, fluid responsiveness. But thinking can be fluid and responsive, while spontaneous reactions can be biased, inflexible, and blind to the facts. Consider people who are spontaneously racist, sexist, or homophobic.” Graham: “Well, I don’t mean to saddle you with the view that we should simply do whatever comes spontaneously. The good is the spontaneous reaction in fullest awareness of how things are patterned.”69 Zhōu: “So the stance you attribute to me really isn’t a rejection of rationality. Nor does it endorse impulsive, spontaneous responses. The crux is actually what you call ‘awareness of how things are patterned.’ ” Graham: “Yes, awareness is the capacity to take things into account in choices. Changes in awareness produce inclinations that are spontaneous, in that they are not chosen but caused by our awareness.”70 Zhōu: “That sounds like a contradiction. Fuller awareness increases our capacity to notice factors relevant to making choices, which causes us to form inclinations that are not choices after all? Also, if the inclinations are caused by awareness of factors outside us, then are they really ‘spontaneous’? ‘Spontaneous’ usually refers to how thought and action issue from the agent, rather than being caused by external factors.” Graham: “Look at your own example—it illustrates what I call a spontaneous reaction in awareness of things. By braking to avoid the child, you spontaneously reflect the objective situation, your motions deriving not from you as an individual but from Heaven—forces of nature beyond our will—working through you.”71 Zhōu: “What you’re calling ‘spontaneity’ seems to be just immediate, noninferential responses. These might appear to be independent of reason, intellect, or intention, directly triggered by factors outside the agent. But in a case like braking to avoid hitting someone, they’re in fact the outcome of practiced, expert agency. Much thought and training might have been involved in acquiring the capacity for such responses. In the driving example, I’m capable of this sort of automatic, reflexive response only because I’m an experienced driver who is paying attention to what he’s doing, is familiar

272 | Chris Fraser with the possibility of kids stepping into the road, and values those kids’ lives. My response isn’t a mere ‘reflection’ of the situation, the product of external forces working through me. ‘Spontaneous’ doesn’t seem an informative description of it, either. Really, it’s expertise or virtuosity, as displayed in intelligent, adaptive agency.” Graham: “I guess in the end that’s basically what I’m trying to get at.” Zhōu: “Well, if the view you’re attributing to me is just that following the Way lies primarily in an uncodifiable, adaptive virtuosity, rather than in identifying and applying explicit, fixed models, I can accept that.” Graham: “Hm.  .  .  .  You’ve got me thinking. Maybe reason versus spontaneity is another of the dichotomies we need to leave behind to find the Way.”72 Zhōu: “Maybe it is! As you said, let’s set aside rigid, permanent distinctions in favor of open, fluid ones. Aren’t rationalism versus anti-rationalism and reason versus spontaneity precisely the sort of artificial distinctions we should give up?” Graham: “You know, I once said that understanding Chinese philosophy doesn’t depend on swallowing my or anyone else’s line of thought. What it does depend on is philosophizing for oneself.”73 Zhōu: “Thank you for inspiring us to do just that.”74

Notes   1. A. C. Graham, Disputers of the Tao (La Salle: Open Court, 1989), 7, 176.   2. Chris Fraser, “Skepticism and Value in the Zhuangzi,” International Philosophical Quarterly 49.4 (2009): 439–457.   3. Graham, Disputers of the Tao, 186.   4. Ibid., 7.   5. Ibid., 194–195.   6. Ibid., 189.   7. William Hung, ed., A Concordance to Zhuāngzǐ, Harvard-Yenching Institute Sinological Index Series, supplement no. 20 (Cambridge: Harvard University Press, 1956), 4/23ff., 4/43–44.   8. Chris Fraser, “Heart-Fasting, Forgetting, and Using the Heart Like a Mirror: Applied Emptiness in the Zhuangzi,” in Nothingness in Asian Philosophy, ed. J. Liu and D. Berger (New York: Routledge, 2014), 197–212.   9. Zhuāngzǐ, 4/29–30, 4/44–53, 4/56–64. 10. Chris Fraser, “Distinctions, Judgment, and Reasoning in Classical Chinese Thought,” History and Philosophy of Logic 34.1 (2013): 1–24. 11. Graham, Disputers of the Tao, 176. 12. Zhuāngzǐ, 33/71–73. 13. Zhuāngzǐ, 33/72, 33/74.

Rationalism and Anti-Rationalism in Later Mohism  | 273 14. See Chris Fraser, “School of Names,” in The Stanford Encyclopedia of Philosophy (Winter 2012 edition), ed. E. Zalta, http://plato.stanford.edu/archives/win2012/ entries/school-names/, section 5.1. 15. Graham, Disputers of the Tao, 183. 16. Ibid., 142. 17. Ibid., 142. 18. A. C. Graham, Later Mohist Logic, Ethics, and Science (Hong Kong: The Chinese University Press, 1978), 22. 19. Jane Geaney, “A Critique of A. C. Graham’s Reconstruction of the ‘NeoMohist Canons,’ ” Journal of the American Oriental Society 119.1 (1999): 1. 20. Graham, Later Mohist, 15–23. 21. Graham, Disputers, 141. 22. Chris Fraser, The Philosophy of Mozi (New York: Columbia University Press, 2016), chap. 4. 23. Graham, Later Mohist, 245. 24. Ibid., 100. 25. Ibid., 33; Graham, Disputers of the Tao, 141. 26. Chris Fraser, “Mohist Canons,” in The Stanford Encyclopedia of Philosophy (Fall 2013 edition), ed. E. Zalta, http://plato.stanford.edu/archives/fall2013/entries/ mohist-canons/, sections 6.1 and 7.1. 27. Graham, Later Mohist, 33. 28. Ibid., 33; Disputers of the Tao, 142. 29. Graham, Later Mohist, 33; Graham, Disputers of the Tao, 140. 30. A. C. Graham, “Chuang-tzu’s Essay on Seeing Things as Equal,” History of Religions 9.2–3 (1969–1970): 137–159. 31. Graham, Later Mohist, 299; Graham, Disputers of the Tao, 142–143. 32. Graham, Later Mohist, 299; Graham, Disputers of the Tao, 143. 33. See Canons A78, A96, A97, and B1. 34. Graham, Later Mohist, 38, 141. 35. 若實也者必以是名也. 36. Graham, Later Mohist, 140. 37. See Geaney, “A Critique,” 4–8. 38. Graham, Disputers of the Tao, 137. 39. Christoph Harbsmeier, Review of Later Mohist Logic, Ethics, and Science, Bulletin of the School of Oriental and African Studies 43.3 (1980): 617–619. 40. Graham, Later Mohist, 30, 230; Graham, Disputers of the Tao, 139. 41. Graham, Disputers of the Tao, 138. 42. Ibid., 139. 43. Ibid., 167. 44. Graham, Later Mohist, 31, 476. 45. Graham, Disputers of the Tao, 148. 46. See Geaney, “A Critique,” 3 and 7–10, and Chris Fraser, “Introduction: Later Mohist Logic, Ethics, and Science after 25 Years,” in A. C. Graham, Later Mohist Logic, Ethics, and Science (Hong Kong: The Chinese University Press, 2003; rpt. ed.), xxviii–xxix. 47. Graham, Later Mohist, 399.

274 | Chris Fraser 48. See Geaney, “A Critique”; Harbsmeier, Review, 618; and Fraser, “Introduction,” xxvi–xxvii. 49. See Geaney, “A Critique,” 11. 50. Graham, Disputers of the Tao, 144. 51. Graham, Later Mohist, 429. 52. Ibid., 248; Graham, Disputers of the Tao, 144–145. 53. Graham, Later Mohist, 179. 54. Ibid., 180–181. The reference is to Zhuāngzǐ 5/55–60. 55. Graham, Later Mohist, 472. 56. Ibid., 179. 57. Graham, Disputers of the Tao, 146. 58. Ibid., 145. 59. Graham, Later Mohist, 236; Graham, Disputers of the Tao, 138. 60. Graham, Later Mohist, xiii. 61. Graham, Disputers of the Tao, 146. 62. Ibid., 7. 63. Ibid., 176. 64. Ibid., 186. 65. Ibid., 188. 66. Ibid., 188. 67. Ibid., 190. 68. Ibid., 189. 69. Ibid., 209. 70. Ibid., 383. 71. Ibid., 190–191. 72. Ibid., 190. 73. Ibid., ix. 74. I am grateful to Roger Ames and Carine Defoort for comments that inspired a completely new version of the final section.

About the Contributors Roger T. Ames is Humanities Chair Professor at Peking University, a Berggruen Fellow, and Professor of Philosophy Emeritus at the University of Hawai’i. He is former editor of Philosophy East & West and founding editor of China Review International. Ames has authored several interpretative studies of Chinese philosophy and culture: Thinking Through Confucius (1987), Anticipating China (1995), Thinking from the Han (1998), and Democracy of the Dead (1999) (all with D. L. Hall), and most recently Confucian Role Ethics: A Vocabulary (2011). His publications also include translations of Chinese classics: Sun-tzu: The Art of Warfare (1993); Sun Pin: The Art of Warfare (1996) (with D. C. Lau); the Confucian Analects (1998) and the Classic of Family Reverence: The Xiaojing (2009) (both with H. Rosemont), Focusing the Familiar: The Zhongyong (2001), and The Daodejing (with D. L. Hall) (2003). Carine Defoort is Professor in Sinology at the KU Leuven (University of Leuven), Belgium. She studied sinology and philosophy at KU Leuven, National Taiwan University, and the University of Hawai’i. Her fields of interest are, primarily, early Chinese thought and, secondarily, the modern interpretation of that thought in the twentieth century. She is the editor of Contemporary Chinese Thought (1997) and coeditor of The Mozi as an Evolving Text: Different Voices in Early Chinese Thought. (2013). She authored The Pheasant Cap Master (Heguanzi): A Rhetorical Reading (1997) and various articles on the legitimacy of Chinese philosophy, regicide, benefit, the Shizi, the Zhuangzi, and the Mozi. Chris Fraser is Associate Professor in the Department of Philosophy at the University of Hong Kong. He holds a BA from Yale University, a research master’s degree from National Taiwan University, and a PhD from the University of Hong Kong, where he won the Li Ka Shing dissertation prize in 1999. His research focuses on early Chinese philosophy, particularly philosophy of mind, epistemology, action theory, and the various ways in which these fields intersect with ethics. His publications include The Philosophy of

275

276 | About the Contributors the Mozi (2016), The Essential Mozi (forthcoming), and several dozen journal articles and book chapters. Robert H. Gassmann is Professor Emeritus of Sinology at Zurich University (Switzerland). He is honorary member of the Swiss Asia Society, over which he presided from 1990 to 1996 and from 2002 to 2004. From 1988 to 2005 he was chief editor of the quarterly Asiatische Studien/Etudes Asiatiques (Asian Studies) and of the monograph series published by the society. His fields of interest were language, history, and thought of Early China. Gassmann authored several studies and articles in all these fields: “Understanding Ancient Chinese Society: Approaches to Rén and Mín” (2000), Antikchinesisches Kalenderwesen (2002; reconstruction of the calendar of the Chunqiu period); Antikchinesisch (textbook and grammar of ancient Chinese with several imprints beginning in 1997, the latest with Wolfgang Behr); Ver­ wandtschaft und Gesellschaft im alten China (2006; on kinship and society in Early China), “Coming to Terms with dé 德: The Deconstruction of ‘virtue’ and a Lesson in Scientific Morality” (2011), Menzius (2016; biography, text study, and translation). Jane Geaney is Professor of Religious Studies at the University of Richmond. She is the author of several books on concepts of language and the body in Early China: On the Epistemology of the Senses in Early Chinese Thought (2002), Language as Bodily Practice in Early China: A Chinese Grammatology (2018), and The Emergence of Word-Meaning in Early China: Normative Models for Words (forthcoming). She investigates intersections of the body and language in early Chinese texts by scrutinizing terminology for terms related to “word,” “name,” “language,” “word-meaning,” “intention,” and “reality.” Her research has implications for understanding early Chinese epistemology, ethics, and concepts of the self. Paul Kjellberg had the privilege of working with Angus Graham, Roger Ames, and Carine Defoort in Hawai’i in 1989 and is now a Professor of Philosophy at Whittier College in Los Angeles, where his teaching and research interests include Daoism, Buddhism, and Quakerism. He is the author of “Skepticism, Truth, and the Good Life in Zhuangzi and Sextus Empiricus” (1994), coeditor of Essays on Skepticism, Relativism, and Ethics in the Zhuangzi (1996), and translator of Zhuangzi in Readings in Classical Chinese Philosophy (2000). Esther S. Klein is a Lecturer at the University of Sydney (School of Languages and Cultures) in the Department of Chinese Studies. Her fields of interest include traditional Chinese historical narrative, intertextuality and

About the Contributors  | 277 the history of texts, and ideas of truth and authenticity in Early China. Her book, The Father of Chinese History, explores traditional Chinese views of Sima Qian’s authorial roles and is forthcoming from Brill. She has also published on the Zhuangzi, Lunheng, and “Hengxian.” Xiaogan Liu is currently Chair Professor of the Normal University of Beijing and the founding and honorary director of the Research Centre for Chinese Philosophy and Culture, the Chinese University of Hong Kong. He also taught and conducted research at Peking University, Harvard, Princeton, Singapore, and other institutions. He is the editor and contributor of Dao Companion to Daoist Philosophy and author of Zhuangzi’s Philosophy and Its Development, Laozi Gujin (The Laozi from Ancient to Modern), Quanshi yu Dingxiang (Hermeneutics and Orientations), among others. He has received prizes and awards for excellence in teaching and research at universities in Beijing, Singapore, and Hong Kong. Michael Nylan is Professor of History, University of California, Berkeley, and works mainly on Early China and on contemporary views of the distant past. Her earlier work focused on the reception of the Five Classics and on Han history, with her latest book Chang’an 26 BCE: An Augustan Age in Rome (2014). Her current work includes a book entitled The Chinese Pleasure Book (forthcoming), a complete translation of the Shangshu, according to Han dynasty commentaries (in progress, with Kai Vogelsang and Hu Ruyue), and various essays on modern identity politics and contemporary humanism. Lisa Raphals studies the cultures of Early China and Classical Greece, with interests in comparative philosophy, religion, and history of science. She is Professor of Chinese, Classics, and Comparative Literature, University of California Riverside (cooperating faculty Philosophy, Religious Studies) and on the faculty of the Tri-Campus Program in Classics. She is the author of Knowing Words: Wisdom and Cunning in the Classical Traditions of China and Greece (1992), Sharing the Light: Representations of Women and Virtue in Early China (1998), Divination and Prediction in Early China and Ancient Greece (2013). Representative scholarly articles include: “Skeptical Strategies in the Zhuangzi and Theaetetus” (1994), “Debates about Fate in Early China” (2014), “Sunzi versus Xunzi: Two Views of Deception and Indirection” (2016), and “Body and Mind in Early China and Greece” (2017). Henry Rosemont Jr., is George B. and Willma Reeves Distinguished Professor of the Liberal Arts Emeritus at St. Mary’s College of Maryland and Visiting Scholar of Religious Studies at Brown University. He was Fulbright Senior Lecturer at Fudan University in Shanghai and has also taught at Johns

278 | About the Contributors Hopkins SAIS. With Roger Ames, he has translated the Analects and the Chinese Classic of Family Reverence, and with Daniel Cook, Leibniz: Writings on China. He has edited five volumes, including Graham’s Festschrift, Chinese Texts and Philosophical Contexts, and written five others, the most recent of which is Against Individualism. Harold D. Roth is Professor of Religious Studies and founding director of the contemplative studies program at Brown University. He is a specialist in Chinese philosophy and textual analysis, the classical Daoist tradition, the comparative study of contemplative experiences, and a pioneer of the academic field of contemplative studies. He has published six books and more than fifty scholarly articles in these areas, including The Textual History of the Huai-nan Tzu (1992), Original Tao (1999), and A Companion to Angus C. Graham’s Chuang Tzu: The Inner Chapters (2003). He worked with John Major, Sarah Queen, and Andrew Meyer to translate the Han dynasty Daoist compendium the Huainanzi (2010).

Index a priori, 8, 232–233; in Later Mohism, 263–269 abdicate, abdication, 173–174 action, 2, 6–7, 55–60, 63, 83, 138, 140–146, 149–155, 157, 191, 195–196, 207–209, 216–225, 237, 252, 270–271; effortless, 51, 57, 62, 63, 65, 67, 74 n.26, 216, 248 n.10 agency, agent, 2, 8, 217–219, 221–225; in Zhuangzi, 269, 171–272 Ames, Roger T., 5–7, 100–105, 107–110, 249 Analects (Lunyu), the, 3, 20, 44, 101, 107, 144, 208 animals, 8, 52, 62, 116–117, 121, 129, 188–189, 215, 217–218, 222–226, 256–257; animality, 201–202 Annals of Master Lü, see Lüshi chunqiu anti-rationalism, 8, 217, 232–235, 238, 251–252, 269–270, 272. See also rationalism archaeology, 3, 14, 16, 29, 31–33, 39, 44–45 argumentation, 20, 29, 31, 43–44; in Later Mohism see bian Aristotle, 80, 105, 191–192 assumptions, assume, 2–8, 11–12, 14, 16–17, 19–21, 28–33, 38, 41, 43, 45, 80, 82, 101, 105, 107, 112, 114, 124, 129, 138, 142, 147, 156, 177, 186–189, 191–199, 203, 209, 217, 225, 232, 236, 239, 241, 247, 251, 253, 258, 268 attention, 4, 51, 54–58, 61, 63, 65–69, 71, 235, 237–238, 271

attunement, 4, 49–51, 57–58, 62–64, 67–69, 71–72 author, authorship, authorial voice, 3–4, 11–22, 32, 36–37, 40, 42–45, 58, 60, 64, 75–76 n.34, 93 n.4 autonomy, 83, 101, 217, 218 awareness, 3, 5, 7–8, 50–51, 54–58, 63, 66–67, 82–83, 85, 195, 216–218, 220, 222, 224–226, 234–242, 247, 271 Ban, Gu, 14, 16 Baxter, William H., 19 Be Aware, see awareness Be Fully Present, 88–89 benefit (li), 111, 165–183, 254–255, 267–269 bian (distinction-drawing, argumentation, discourse), 143, 154–155, 252–253, 258, 261–263, 269 binomes, binomial expressions, 17, 115 biology, 8, 52, 187, 201, 209, 216–217, 219, 222–226 Bocheng, Zigao, 175–176, 184 n.62, n.63 body, blindness of, 137–153, 155–157, 160–163; deafness of, 150–151; image of, 220; hands or eyes picking, choosing, selecting, dividing or locating, 138, 141–145, 147–148, 152– 154, 162, 163; lameness of, 138, 146, 151–53, 155–157; putting into practice, 143–144, 146, 149, 153, 156, 161, 163; schema of, 220; sense of touch, 146, 152; sitting instead of standing and acting, 145–146; mouth, 143, 145, 154,

279

280 | Index body (continued) 155, 161; taste-impairment of, 138, 150, 153–155, 157 Book of Rites, see Liji Branner, David, 19 Brunschwig, James, 224, 229 Bruya, Brian, 55 Butcher (Ding), 88, 216 butterfly anecdote, 85, 231–232 Cahn, Rael, 56, 74 n.27 care for everybody (jian ai), 165, 167, 173, 269 categorical imperative, 8, 233 causation, causality, 146–147, 216–217, 221, 222, 236, 243–244, 255–256, 258, 262, 271 change, 39, 56–58, 82–83, 86, 89, 126, 149–150, 190, 223, 225, 231–233, 238, 240–247; in Later Mohism, 253–255, 259–260, 262, 271 Chappell, Timothy, 85 Cheng brothers, 28 choice, 8, 82, 101, 103–110, 177, 217–219, 221–223, 232, 234–236, 238, 240–242, 271; choiceless awareness, 56–57 Chomsky, Noam, 5, 140 Chuang-tzu, see Zhuangzi Chunqiu Zuozhuan, 6, 112–113, 115, 121, 125–127 cognition, 51–53, 55, 57, 59–60, 62–65, 69–71, 73, 199, 209, 225 comparativism, 100–101, 103 conceptual scheme, 5, 99, 102–106 Confucianism, Confucians, 7, 38–40, 52, 62, 80, 101, 103, 145, 166, 170, 172, 175, 188, 194, 199 Confucius, 3, 30, 33–40, 43–44, 60, 63–64, 66–69, 84–85, 101, 104, 165, 186, 208, 210, 223 consciousness, conscious, self-conscious, 4, 51–72, 191, 217–220, 222, 224–226, 246, 252 contemplation, 51–52, 88 contemplative phenomenology, 50–51, 57, 59–61, 63, 65–66, 70–72

Contemplative Studies, 4, 50–52 corruption (textual), 2, 11, 13, 17, 20, 41 cortical midline structures, 221, 225 Csíkszentmihályi, Mihály, 55 culture (wenhua), 50, 55, 79, 104–110, 188–190, 199, 201, 203–204, 209 Damasio, Antonio, 221 Darwin, Charles, 224–225 Davidson, Donald, 104–106 death, dead, 60, 62, 81, 88–89, 127, 129, 172–173, 245, 247 Defoort, Carine, 6 Disputers of the Tao, 2, 7, 28, 30, 99, 108, 199, 217 doing nothing, see wuwei doubting antiquity, 29, 33, 39 Duke Mu of Lu Asks Zisi (Lu Mugong wen Zisizi 魯穆公問子思), 33 dyads, dyadic pairs, 115–118, 125, 127–131, 207 eisegetics, 102, 108 embodiment, embody, 50, 55, 70–71, 143, 146, 152, 157, 202, 209, 220 emotions, 51, 56, 66, 70, 83, 91–92, 205, 221–223, 225, 239–240 empiricism, empirical, 8, 53, 104, 216–217, 224–225, 233, 251–252, 255 essentialism, 101–102, 188, 190–193, 199 Essentials of Nature and Destiny (xingming zhi qing 性命之情), 62, 63 ethics, 100, 103, 199–200, 218; Later Mohist, 254, 256, 258–260, 262–263, 265–269 existentialism, 17, 234 experience, 4–5, 50–51; first-person, 51–53, 55, 57; no-person, 4, 53, 55, 57, 60–61, 63–64, 67, 70–71; objective, 52–53, 70; pure, 4, 53, 55, 57, 63, 71. second-person, 52–53, 57, 65; subjective, 50–53, 69–70; third-person, 52, 55, 57, 69, 70–71 extraterrestrials, 218 fasting of the mind, 60, 66, 69, 72

Index  | 281 Feng, Youlan, 14–15, 167–168 Fingarette, Herbert, 104 flow, 4, 51, 53, 55, 57–59, 61, 63, 67, 69–71, 73 focus-field, 203–205, 208–209 for oneself (wei wo), 165, 167, 183 forget, 60, 66, 72, 88, 92 Frankfurt, Harry, 218, 227 Fraser, Chris, 7–9, 14, 17–18, 21 freedom, 59, 63, 70, 80, 86, 101, 218–219, 223 Fuyang (tomb), 22, 44 Gallagher, Shaun, 219–220, 222 Gassmann, Robert, 5–6 Geaney, Jane, 5–6 genuine, 60, 64, 90, 104, 197, 226 Giles, Herbert A., 30–32 God, 102–103, 234 Graham Memorial Lectures, 9 Graham, Angus, 99–110, 137–139, 157, 159–160, 164, 165–169, 172, 175–180, 182, 184, 231ff.; biography 1–2, 11–13, 16, 18, 19–20, 27, 28, 30, 35, 36, 37, 38, 42, 43, 45; on Chinese cosmology, 193–195; on contemplative states, 73; on mysticism, 49, 50; textual scholarship on Zhuangzi, 57–58, 75, n.33 grammar, 2, 5, 103, 105–107 Gu, Jiegang, 167–168, 173 Guanyin, 36, 67 Guanzi, 34–35, 163 Guo, Xiang, 15–16, 22, 41, 81–82 Hall, David L., 50, 100, 104, 107–108, 190 Han dynasty, 6, 14, 16, 21–22, 30–35, 37, 39, 43, 85, 167, 172, 177, 189 Han Feizi (Hanfeizi), 43, 160–161, 167–168, 171, 173–175, 180, 182 Han, Yu, 15 Hanshi waizhuan, 39–40 Hanshu, “Treatise on Literature and the Arts” (Yiwenzhi), 16, 32 health, 167–170, 172–174, 176–178 heaven(s), see tian

Heguanzi (Ho-kuan-tzu), 2–3, 33 Hierocles, 223–224 holography, 203–205, 207, 209 Hu, Shi, 29 Huainan, 16, 22, 58, 76 Huainanzi, 14, 30, 58, 80, 141, 143–145, 147, 162 Hui, Shi (Huizi), 9, 59, 90–91, 245, 253, 266 human condition, 5, 80, 82, 83–84, 88, 197–198, 219, 262 human nature (xing), 3, 7, 101–103, 109, 185–213; developmental (biological model), 187–188, 191, 193; discovery (ontological) model, 187–188, 193; distinctively human, as what is, 205–206; narrative interpretation, 190–191 Hume, David, 8, 151, 215, 219, 227, 232–233 information processing, 56–57 Inner Chapters, see Zhuangzi intersubjectivity, 52–53 irrationalism, 8, 217, 233–235, 237 James, William, 4, 53, 55–56, 221 Kachi, Yukio, 216 Kant, Immanuel (Kantian), 2, 8, 216–218, 232, 237–238, 240 Kaufmann, Walter, 16–17 Kierkegaard, Søren, 151 Kjellberg, Paul, 7–8 Klein, Esther, 3–4, 93 knowledge, as requiring experience, 151, 156; knowledge, in Later Mohism, 254–255, 258–260, 263–266 Kongcongzi, 171–172, 181–182 Kongzi jiayu, 39–40 Kongzi, see Confucius Lakoff, George, 195 language, views on, 5–6, 91, 99–102, 104–108, 138–141, 146, 150–152, 157, 221, 264, 266, 269

282 | Index Lao Dan, 35, 36, 37, 60, 63–64, 68 Laozi, 28–31, 33–40, 44–45, 63, 68, 91 Lau, D. C., 4, 34, 37–38 Learning, Task-Irrelevant Perceptual (TIPL), 53–55 Lewis, Mark Edward, 13–14 Li, Si, 174, 183 n.48 Li, Xueqin, 44 Liang, Qichao, 29 Libet, Benjamin, 222 Liezi (Lieh-tzu), 2, 28, 99, 166, 175, 177 Liji, 39–40 Liu, Xiang, 16, 189 Liu, Xiaogan, 3–4, 11, 14, 17–18, 20–21 Liu, Xie, 189 Liu, Xin, 16 Liutao, 33 Long, Anthony, 224 Lunyu, see Analects Lüshi chunqiu (Lü shi chunqiu), 14, 18, 21–22, 39–40, 43, 167, 169, 172, 177 marriage, 8, 231–232, 238–245 McCraw, David, 18–19, 76 n.34 meaning, 6, 85, 104, 111–114, 117–118, 122–123, 126–128, 132, 139–140, 150, 157, 174, 216, 233, 276 meditation, 56–57, 65–67 Mencius (Mengzi), 6–7, 165–182, 184; on human nature, 185–213; misreadings of, 196–199; personal cultivation, on, 201–203, 206–207; Xunzi’s critique of, 186–187 Mengsun, Yang, 176–177, 184 Mill, J. S., 218, 227 mind, mirror-like, 67 ming (decree, destiny, propensities), 101–102, 128, 158 n.4, 205, 222–223, 268 ming (name, naming), 6, 124, 137, 157, 266 Moeller, Hans-Georg, 85 Mohism, 9, 256, 165–166, 170–173, 175–178; Later, 2–3, 6, 99, 111, 113, 137, 141, 155–156, 177, 253–269

Mozi (Mo Di), 6, 28, 141, 144–145, 147–154, 165–167, 170–171, 175–178 Munro, Donald, 27 necessity, 83, 148, 198, 220; in Later Mohism, 254–258 neurophenomenology, 57, 75 n.32 neuroscience, 51–52, 56, 219–226 Nishida, Kitaro, 53, 57 Northoff, Georg, 221 Nylan, Michael, 4–5 oikeiosis (self-ownership), 224 Olberding, Amy, 249 one hair (yimao), 165–173, 175–178, 181 ought (is/ought), 3, 8, 216, 232–234, 271 Panksepp, Jaak, 229, 230 parallels, parallelism, 43, 55, 106, 149, 167, 170, 224; in Later Mohism, 260, 263; textual, 14, 18, 21–22 Peng bird, 59, 85–86 perception, 54–57, 65–68, 152, 216–217, 220–221, 224–226 physical wholeness, 173 Polanyi, Michael, 54 Polich, John, 56 power, 62–65, 68–69, 71, 79, 89, 118, 130, 167, 174 Problem of Value, The, 2 propositional knowledge, 55, 237 proprioception, 224, 226 Qian, Mu, 4, 29–30, 33–34 Qin dynasty, 34, 36–37, 44; second emperor of the, 174–175 Qin, Guli, Qinzi, 176–177, 184 qing (facts, essentials), 265–266 Qiwulun, 15, 49–50, 68–69, 86 quasi-syllogism, 7, 108, 216–217, 226 Rand, Christopher, 15–16 Raphals, Lisa, 7–8 rationalism, 8, 215, 217, 233–239, 241, 251–253, 263, 267–272

Index  | 283 rationality, 8–9, 218, 225, 238, 270–271 Reason and Spontaneity, 2, 7, 79, 82, 99, 215, 217, 226, 227, 232 reason, 4, 8–9, 50, 82, 105, 131, 216–218, 222, 232–239, 243, 251–253, 263, 268, 270, 272 relativism, 70, 84, 99, 104 Robber Zhi, 22 Romanticism, romantic, 8, 217, 234 Rong cheng shi, 171, 181 n.32 Rosemont, Henry, Jr., 5, 28, 104, 190 Roth, Hal, 4, 13, 22, 41, 162 Ryle, Gilbert, 54 sage (shengren), 5, 59–62, 64–65, 67–70, 82–87, 89, 91, 150, 153, 170–171, 174, 186, 197, 202, 254–255, 265–269 sample argumentation, 20, 43–44 Sandel, Michael, 187 Sapir-Whorf theory, 99, 104–105, 108 Saussure, Ferdinand de, 140 Schneewind, Jonathan, 218 second-order desires, 218–219 Seitz, Aaron, 53–54 self, 8, 51–53, 55–61, 65, 69–72, 187, 203, 217–219, 221, 223–226 self-related processing (SRP), 221, 225–226 Sherington, Charles, 229 shi (thing, matter), 124, 129, 138, 141, 152, 153, 156, 158, 161, 163, 164 shi-fei (this/not), 61, 69, 252, 257, 261 Shiji, 22, 39, 85, 174 Shizi, 174 Shun (emperor), 146, 171, 182 n.37, 202 Shuoyuan, 39–40 Sima, Qian, 14, 16, 21–22, 30, 34–35, 37–38, 40 Sima, Rangqie, 35 Sima, Tan, 35, 174 Sima, Xiangru, 174, 181 simplicity, the simple, 62–65 somatic markers, 221–222 spontaneity, spontaneous, 2–5, 7–9, 50–51, 53, 55–57, 59, 61–63, 65, 67,

70–71, 73, 80, 82–84, 89, 193–195, 198, 200–202, 215–217, 219–220, 222, 226, 232–233, 234–247, 252, 270–272 states (of mind), 49, 51, 53, 55–57, 63, 65–68, 73 stillness, still, 63–65, 67–68, 72 Stoics, 223–224 Sun Bin Arts of War (Sun Bin bingfa), 33 Sunzi’s Arts of War (Sunzi bingfa), 33 tacit knowledge, 4, 54–55, 57, 62–63, 65, tacit performance, 220 Taishi Dan, 34, 36 Tang, Junyi, 102, 198 Thompson, Evan, 52, 57 tian (heaven, heavens), 60, 65, 84, 101–102, 120, 124, 193, 204–205, 222 tianxia (world), 165–169, 174, 176 token, 116, 118–120, 122, 125–127, 132, type-token distinction, 138–139 tranquility, 51, 64 transcend, transcendence, 50, 53, 82–83, 86, 88, 102–104, 189, 194 translation, 2–3, 12–13, 28, 49, 81, 93, 99–104, 107, 1112–113, 118, 132, 138, 175–176, 194–196, 199, 207, 217 trope, 63, 92, 139, 142, 146, 151–153, 155, 170, 172–175, 177–178 Tu, Wei-ming, 27 Twain, Mark, 37 unhewn (pu), 62, 63, 65 Utmost Persons (zhiren), 59, 64, 67 Varela, Francisco, 51, 74 n.11, 75 n.32 Velleman, J. D., 227 vitality, vital, 5, 68, 79–93, 192, 195–196, 198–200, 202, 206, 209 vow, promise, 8, 231, 238, 242–244 Waley, Arthur, 34, 193 Wang, Shumin, 15, 93 wanton, 219 Warring States, 14, 16, 21, 35, 169, 175 Watanabe, Takeo, 53

284

|

Index

Way (dao), 50, 60–66, 68–71, 84, 90, 172–173, 194, 206, 251–252, 254, 266, 269–270, 272 weighing (metaphor), 6, 82, 91, 169, 172–173, 177 Weiliaozi, 33 wenhua, see FXOWXUH Wenzi, 33 :KLWHKHDG$ 1  :LOOLDPV 5D\PRQG  Wittgenstein, Ludwig, 195 :RQJ 'DYLG  :RRGZRUNHU 4LQJ ² :RROI 9LUJLQLD ² ZRUGV DV GLVWLQFW IURP QDPHV ² 149–151, 158 ZRUOG VHH tianxia wu WKLQJ DJJUHJDWH DVVHPEODJH VHW  6, 111–128, 130 wuwei (nonaction), 36, 63, 81 xing ZDON DFWLRQ DJHQW     146, 152 xing, see human nature ;XQ]L     ²  Yang, Zhu, 6–7, 9, 165–173, 175–178, 254 Yanzi, 33