Cultural Realism: Strategic Culture and Grand Strategy in Chinese History 0691002398, 9780691002392

Citation preview

CULTURAL REALISM

PRINCETON STUDIES IN INTERNATIONAL HISTORY AND POLITICS

Series Editors Jack L. Snyder and Richard H. Ullman RECENT TITLES:

A Constructed Peace: The Making of the European Settlement, 1945—1963 by Marc Trachtenberg Regional Orders at Century's Dawn: Global and Domestic Influences on Grand Strategy by Etel Solingen From Wealth to Power: The Unusual Origins of America's World Role by Fareed Zakaria Changing Course: Ideas, Politics, and the Soviet Withdrawal from Afghanistan by Sarah E. Mendelson Disarming Strangers: Nuclear Diplomacy with North Korea by Leon V. Sigal Imagining War: French and British Military Doctrine between the Wars by Elizabeth Kier Roosevelt and the Munich Crisis: A Study of Political Decision-Making by Barbara Rearden Farnham Useful Adversaries: Grand Strategy, Domestic Mobilization, and Sino-American Conflict, 1947-1958 by Thomas J. Christensen Satellites and Commissars: Strategy and Conflict in the Politics of Soviet-Bloc Trade by Randall W. Stone Does Conquest Pay? The Exploitation of Occupied Industrial Societies by Peter Liberman Cultural Realism: Strategic Culture and Grand Strategy in Chinese History by Alastair Iain Johnston The Korean War: An International History by William Stueck Cooperation among Democracies: The European Influence on U.S. Foreign Policy by Thomas Risse-Kappen The Sovereign State and Its Competitors: An Analysis of Systems Change by Hendrik Spruyt America's Mission: The United States and the Worldwide Struggle for Democracy in the Twentieth Century by Tony Smith Who Adjusts? Domestic Sources of Foreign Economic Policy during the Interwar Years by Beth A. Simmons

CULTURAL REALISM STRATEGIC CULTURE AND GRAND STRATEGY IN CHINESE HISTORY

ALASTAIR IAIN JOHNSTON

PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY

Copyright © 1995 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, Chichester, West Sussex All Rights Reserved Library of Congress Cataloging-in-Publication Data Johnston, Alastair Iain. Cultural realism: strategic culture and grand strategy in Chinese history / by Alastair I. Johnston p. cm. Includes bibliographical references and index. ISBN 0-691-02996-2 ISBN 0-691-00239-8 (pbk.) 1. Chins*—History—Ming dynasty, 1368-1644. 2. National security—China. 3. China— Military policy. I. Title. DS753.J64 1995 951'.02—dc20 95-3105 CIP This book has been composed in Monotype Times New Roman Princeton University Press books are printed on acid-free paper and meet the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources Third printing, and first paperback printing, 1998 http://pup.princeton.edu Printed in the United States of America 10 9 8 7 6 5 4 3

CONTENTS LIST OF FIGURES AND TABLES PREFACE

vii

ix

CHAPTER ONE Strategic Culture: A Critique 1 The International Security Field and Strategic Culture 4 Strategic Culture and China 22 Conclusion 27 CHAPTER TWO Some Questions of Methodology 32 Definitions of Strategic Culture 33 Objects of Analysis 39 Methods of Analysis 49 Empirical Analysis 52 CHAPTER THREE

Chinese Strategic Culture and the Parabellum Paradigm Righteous War 69 On Violence 71 On "Not Fighting and Subduing the Enemy " 99 Conclusion 106

61

CHAPTER FOUR

Chinese Strategic Culture and Grand Strategic Preferences 109 A Typology of Grand Strategies 109 Central Paradigms and Grand Strategic Preferences 117 Conclusion 143 CHAPTER FIVE

A Return to Theory

155

The Strategy of Symbols and Symbolic Strategy 156 Some Hypotheses about Ming Strategic Decision-Making

170

VI

CONTENTS

CHAPTER SIX

The Parabellum Paradigm and the Ming Security Problematique 175 Bingshu in the Ming Dynasty 176 Ming Security Problems Along the Northern Border 183 The Parabellum Paradigm and Alternative Grand Strategies 186 On Violence 190 Conclusion 212 CHAPTER SEVEN

Chinese Strategic Culture and Ming Grand Strategic Choice 216 Patterns in Ming Grand Strategic Preferences 217 Strategic Preferences and Coerciveness in Ming Security Policy 231 Conclusion 242 CHAPTER EIGHT

Conclusion 248 APPENDIX A

Coding Procedures 267 APPENDIX B

Terms Used to Describe Legitimate Actions Directed at an Adversary 270 Terms Used to Describe Outcomes of Actions against an Adversary 273 APPENDIX C

Map of Northern Border Areas in the Ming Period 274 REFERENCES INDEX

293

275

LIST OF FIGURES AND TABLES

FIGURES

3.1

Composite Cognitive Map of the Wu Zi Bing Fa

73

3.2

Composite Cognitive Map of the Si Ma Fa

75

3.3

Composite Cognitive Map of the Wei Liao Zi

79

3.4

Composite Cognitive Map of the Huang Shi Gong San hue

83

3.5

Composite Cognitive Map of the Tai Gong Liu Tao

87

3.6

Composite Cognitive Map of the Tang Tai Zong Li Wei Gong WenDui

91

3.7

Composite Cognitive Map of the Sun Zi Bing Fa

94

4.1

Interrelated Levels of Strategy

115

4.2

The Central Paradigm

149

6.1

Fan Ji (1428)

193

6.2

YuQian(1450)

194

6.3

YuQian(1450)

195

6.4

Yu Qian (1452)

196

6.5

Wang Shu (1484)

197

6.6

Wang Shouren (1500)

198

6.7

Yang Yiqing (1506)

200

6.8

Yang Yiqing (1506)

201

6.9

Wang Qiong (1516)

202

6.10

Wang Qiong (1516)

203

6.11

Wang Qiong (1520)

204

6.12

ZengXian(1547)

205

6.13

Weng Wanda (1547)

207

6.14

Weng Wanda (1551)

208

6.15

Yang Jisheng (1552)

209

6.16

Wang Chonggu (1571)

210

Vlll

7.1 7.2

FIGURES AND TABLES

Cumulative Frequencies of Ming and Mongol Conflict Initiation, by Five-Year Periods

237

Yearly Percentage of Conflicts Initiated by the Ming and Mongols, Averaged over Five Years

239

TABLES

4.1

Summary of Predominant Elements in Security Strategy

144

4.2

Congruence between Grand Strategy and Preference Rankings

145

4.3

Congruence between Grand Strategy and Preference Rankings

146

4.4

Congruence between Grand Strategy and Preference Rankings

147

4.5

Congruence between Grand Strategy and Preference Rankings

148

7.1

Predominant Elements in Security: Ming Memorials on the Mongol Threat

232

Correlation between Mean Yearly Percentage of Conflicts Initiated by the Ming, Mean Yearly Percentage of Conflicts Initiated by Mongols, and Mean Yearly Frequency of Rebellion

240

Regression between Mean Yearly Percentage of Conflicts Initiated by the Ming, Mean Yearly Percentage of Conflicts Initiated by Mongols, and Mean Yearly Frequency of Rebellion

240

Association between Ming Initiation, Threat, and Capabilities

241

7.2

7.3

7.4

PREFACE

T

HE CENTRAL QUESTION in this book is twofold: To what extent is there a substantively consistent and temporally persistent Chinese strategic culture, and to what extent has this strategic culture influenced China's use of military force against external "threats" historically? The empirical focus is on the Ming dynasty (1368-1644). By strategic culture I mean ranked grand strategic preferences derived from central paradigmatic assumptions about the nature of conflict and the enemy, and collectively shared by decision makers. This book, then, is about ideas and their relationship to behavior. Despite the centrality of this relationship to political behavior, it is an exceedingly difficult causal connection to show empirically. Perhaps for this reason, the dominant neorealist or realist school in international-relations theory has avoided the problem by trying to construct a theory of state behavior that privileges the incentives and constraints created by particular configurations of power in the international system structure.1 The notion of strategic culture, in principle at least, poses a significant challenge to structural realist claims about the sources and characteristics of state behavior by rooting strategic choice in deeply historical, formative ideational legacies. I do not claim to provide any dramatic conceptual or methodological breakthroughs. This book is merely an effort to see whether strategic culture as one set of ideational inputs into behavior is a useful analytic tool with which to explore the broader relationship between ideas and behavior in international relations and strategic studies. To this end, however, I do try to avoid the pitfalls that critics of political culture argued were inherent in this focus on ideational sources of behavior, namely tautological definitions of culture, a mechanically deterministic conceptualization of the relationship between culture and behavior, and flawed research designs that do not even attempt to distinguish between the influences of "culture" and "structure." I spend a fair amount of time trying to develop a notion of strategic culture that is falsifiable, whose formation and development can be traced empirically, and whose effects on strategic choice can be, in principle, weighed against the effects of other nonideational influences. I argue that strategic culture consists of two basic elements: (a) a central paradigm that supplies answers to three basic, related questions about the nature of conflict in human affairs, the nature of the enemy, and the efficacy of violence; 1 The claim that neorealist theory addresses international politics and not foreign policy seems somewhat disingenuous. Neorealism does make predictions about what individual states do or ought to do to pursue their interests within the constraints of polarity and particular distributions of power. Small, weak states, for instance, combine to balance against rising hegemonic power by pursuing foreign policies aimed at alliance building and intra-alliance negotiation.

X

PREFACE

and (b) a ranked set of strategic preferences logically derived from these central assumptions. These central heuristics and accompanying strategic preferences must be congruent across relevant objects of analysis (e.g., strategic texts representative of a formative period in the development of strategic thought and practice) for a single strategic culture to exist in any particular society. Moreover, one must show, in the historical period of interest, that these heuristics and preferences are consistent across a significant portion of decision makers socialized in this strategic culture before drawing any conclusions about the effect of strategic culture on behavior. Then patterns in strategic behavior must be shown to be consistent with those predicted by the strategic preferences held by these decision makers. The results of this three-step research process in the Chinese case are a rather complex set of conclusions. In essence I argue, contrary to my initial expectations, that there is evidence of two Chinese strategic cultures, one a symbolic or idealized set of assumptions and ranked preferences, and one an operational set that had a nontrivial effect on strategic choice in the Ming period. The symbolic set, for the most part, is disconnected from the programmatic decision rules governing strategy, and appears mostly in an habitual discourse designed, in part, to justify behavior in culturally acceptable terms. The operational set reflects what I call a, parabellum or hard realpolitik strategic culture that, in essence, argues that the best way of dealing with security threats is to eliminate them through the use of force. This preference is tempered by an explicit sensitivity to one's relative capacity to do this. In other words, at its simplest, the operational strategic culture predisposes those socialized in it to act more coercively against an enemy as relative capabilities become more favorable. This is consistent with what Vasquez calls an "opportunity model" of realpolitik behavior, where "states need no special motivation to threaten or use force; rather they are always predisposed to do so, unless restrained by contextual variables" (Vasquez 1993: 115). What predisposes states (at least the Chinese state, since this is not a cross-national study) to act this way, however, are not anarchical structures generating realpolitik self-help impulses, but rather the parabellum strategic culture that, I argue, persists from very early formative periods in Chinese strategic thought and practice up through the Ming dynasty across different strategic structures and objective contexts. In other words, I treat realpolitik decision axioms as cultural in that their content over time is largely independent of interstate structures, and is collectively learned and transmitted through socialization in the parabellum approach to conflict. This finding suggests that while strategic culture exists and plays a role in strategic choice, there may be fewer cross-national differences in strategic culture than many of the users of the concept expect. However, precisely because realpolitik behavior may be strategic-culturally rooted, these systemwide strategic predispositions may also be more mutable and susceptible to purposeful change than realists may expect.

PREFACE

XI

Broadly speaking, I had a number of goals in mind when I started this project. One was simply to try to understand and demystify Chinese strategic thinking. Most of my previous academic work had focused on contemporary Chinese security affairs, arms control, nuclear doctrine, and the like. It was clear to me from reading in this area that much of the Western and Chinese literature accepted that contemporary Chinese strategic thought was to an important degree influenced by ancient traditions in philosophy and statecraft, and that these traditions were unique, or at least substantially different, from what are portrayed as Western traditions. Yet to this point very little systematic work had been done on isolating exactly what elements of this tradition had persisted and how. The notion of "strategic culture," which appeared in the internationalrelations and international-security literature in the 1980s, offered at least the beginnings of a conceptualization of this process. When I started this project I had no particular reason to think that standard views of Chinese strategic thought—embodied for instance in ConfucianMencian denigration of the role of violence in state security, or Sun Zi's alleged emphasis on non- or minimally violent routes to strategic victory—might be inaccurate, misleading, or plainly wrong. I was fully prepared to find that, indeed, Chinese strategists through the ages did think differently from those in the West about questions of war and peace, and that the differences should show up in behavioral patterns substantially different from what, say, a standard structural realpolitik approach might suggest about strategic choice. But as I delved into the classic texts in Chinese strategic thought, every now and then looking at indicators of the frequency with which military violence was used throughout Chinese history, I was struck by two things. One was the prevalence of assumptions and decision axioms that in fact placed a high degree of value on the use of pure violence to resolve security conflicts. The other was the existence in these texts—and even more so in recent scholarship on these texts—of a Confucian-Mencian discourse as well, but one that was more obviously divorced from the strategic preferences that emanated from the classics on strategy. One of the subthemes in this book is what the substance and role of this symbolic discourse is in the context of a predominant emphasis on hard realpolitik or parabellum strategic preferences. The problem with this finding is that it makes testing for the effects of a strategic culture on Ming security policy rather difficult, since the parabellum strategic-culture model does not make significantly different predictions about behavior from that of a simple structural realpolitik model. Grappling with this methodological problem is also, as it turns out, a subtheme I did not anticipate. Another even broader goal, or perhaps motivation, reflects my interest and training in both China studies and international-relations theory. I have long been interested in examining some of the dominant assumptions about the role of culture or cultural variables (in this case a subset, strategic culture) in political behavior used in area studies and international-relations theory. The assumption,

Xll

PREFACE

often unspoken, behind area studies, and China studies in particular, has been that unique cultural variables will generate different, and more accurate, predictions about the behavior of a polity than if culture were left out. Put another way, Chinese political behavior cannot be understood without reference to historical and cultural precedent. Almost invariably, when issues in the China field are put this way, the conclusions stress China's uniqueness or differentness. This assumption applies to both domestic politics and external relations. The dominant paradigm in international-relations theory—neorealism or structural realism and its subvariants—reaches a very different conclusion; namely, that culture does not add any significant explanatory power to ahistorical and acultural models of state behavior. This argument—plus the positivist methodological bias in international-relations theory—leads scholarship to proceed from the assumption of sameness across cases. I do not claim to resolve this conflict here. (Indeed, those on both sides may find my conclusions somewhat disturbing—there is a Chinese strategic culture, but its principal components are not self-evidently unique.) But one of my interests has been to examine both sets of assumptions, using the tools and methods of both fields to examine the role of strategic culture in Chinese behavior. In part I have in mind an even grander goal—a "vision thing" to quote a former American president: to make the discussion of China's external behavior across history, its patterns of conflict resolution, its strategic doctrines, etc., as common in international-relations literature as discussions of the German, French, American, British, and Russian cases. I see this study as one small step in that direction. There are innumerable people and institutions I should thank for emotional, intellectual, and financial support. To begin with I thank Felicity Lufkin for her love, and for reminding me that there is life beyond the academy, particularly when Jackie Chan movies are on at the Brattle. I want to thank my parents, Antony and Margot, and the rest of the Johnstons for being a very, very fine family. Heartfelt thanks go to the members of my most excellent dissertation committee: Kenneth Lieberthal, Michel Oksenberg, Robert Axelrod, Albert Feuerwerker, and Paul Forage. My gratitude goes as well to my graduatestudent colleagues, and others who helped me gather material, sift through ideas, and who provided intellectual support and friendship: Abigail Jahiel, Arthur Waldron, Bruce Dickson, Charles Glaser, Chen Peixiong, Cheryl Shanks, Dale Copeland, David Shambaugh, Ding Zhaoqiang, Edward O'Dowd, Gao Dianfang, Gerald Sorokin, Harold Jacobsen, J. David Singer, Jack Snyder, James Tong, Jason Hu, Jeff Ritter, John Gaddis, Karl Eikenberry, Kate Campbell, Kent Jennings, Kidder Smith, Li Ling, Liang Liu Chun-hua, Liu Di, Liu Qing, Ma Dazheng, Martha Feldman, Nico Howson, Pan Jiabin, Patrick Regan, Paul Godwin, Peter Katzenstein, Ren Li, Samuel Barnes, Shen Mingming, Thomas Christensen, Wang Jianwei, Wang Zhenhui, Wei Rulin, Wei Xiuyun, William Zimmerman, Wu Rusong, Wu Zhemin, Xu Xin, Yang

PREFACE

Xlll

Kai, and Yu Zemin. I am also indebted to institutional, intellectual, and/or financial support from: The Center for Chinese Studies, The Horace H. Rackham School of Graduate Studies, and the Department of Political Science at the University of Michigan; the Social Sciences and Humanities Research Council of Canada; the Social Sciences Research Council/MacArthur Fellowships in International Peace and Security; the Institute of International Relations, Taiwan; the Peking University History Department, Beijing; the Institute for the Study of World Politics; the Olin Institute for Strategic Studies; the Fairbank Center for East Asian Research and the Department of Government at Harvard University.

CULTURAL REALISM

Chapter One STRATEGIC CULTURE: A CRITIQUE

T

HE SOVIET MILITARY exhibited a preference for the preemptive, offensive use of force that was deeply rooted in a Russian history of external insecurity and internal autocracy. The United States has exhibited a tendency towards a sporadic and reluctant though messianic and crusading use of force that is deeply rooted in a fundamental belief in the aberrance of warfare in human relations and the moralism of the early republic. China has exhibited a tendency for the controlled, politically driven defensive and minimalist use offeree that is deeply rooted in the statecraft of ancient strategists and a worldview of relatively complacent superiority. Obviously these are sweeping generalizations that could (and should) be challenged thoroughly on the basis of counterevidence. But all of these generalizations have been proposed in more or less similar fashion by respected students of Soviet, American, and Chinese strategic behavior. Such assessments can all be subsumed under the broad analytic category of "strategic culture." While the term remains remarkably undefined, those who use it explicitly or implicitly tend to mean that there are consistent and persistent historical patterns in the way particular states (or state elites) think about the use of force for political ends. That is, different states have different predominant sets of strategic preferences that are rooted in the "early" or "formative" military experiences of the state or its predecessor, and are influenced to some degree by the philosophical, political, cultural, and cognitive characteristics of the state and state elites as these develop through time. Ahistorical or "objective" variables such as technology, capabilities, levels of threat, and organizational structures are all of secondary importance: it is the interpretative lens of strategic culture that gives meaning to these variables. Strategic choices, therefore, are less responsive to changes in the objective strategic environment, since the weight of historical experiences and historically rooted strategic preferences tends to constrain the effects of environmental variables and to mute responses to environmental change. As a result, if strategic culture does change, it does so slowly, lagging behind changes in "objective" conditions. This does not imply that the strategic-culture approach rejects rationality. Indeed, strategic culture is compatible with concepts of limited rationality, under which strategic culture-based heuristics are employed to simplify reality; process rationality, under which strategic culture determines the most rational process of choosing between options; and adaptive or learned rationality, under which historical choices, analogies, metaphors, and precedents are

2

CHAPTER ONE

invoked to guide choice.1 The strategic-culture approach does seem potentially incompatible with the contingent nature of strategy in game rationality or classical rational choice-expected utility models. Whereas strategies in games focus on making the "best" choice depending on expectations about what other players do (Schelling 1980: 3), strategic culture, as it has been used to date, implies that a state's strategic behavior is less responsive to others' choices.2 There is, instead, a historically imposed inertia on choice that makes strategy less responsive to specific contingencies. Thus, in the view of some American analysts of Soviet strategic culture, the Soviets did not adopt American MADbased deterrence doctrines, as U.S. policymakers had once predicted, since Soviet strategic culture-based preferences historically preceded the emergence of the nuclear era and the American nuclear doctrine. Rather than rejecting rationality per se as a factor in strategic choice, the strategic-culture approach rejects an ahistorical, acultural realist framework for analyzing strategic choices.3 The latter would seem to discount the accumulated weight of the past in favor of a forward-looking calculation of expected utility. The dominant neorealist paradigm assumes that states are functionally undifferentiated units that seek to optimize their utility (Waltz 1979, Keohane 1986, Buzan et al. 1993). Utility is usually defined as power, often as capabilities and resources. Hence states will act to expand their capabilities as long as the resources or opportunities to do so exist. Strategic choices will be optimizing ones, constrained only, or largely, by ahistorical and acultural variables such as geography, capability, threat, and a tendency of states to refrain from behaviors that clearly threaten their immediate survival (Bueno de Mesquita 1981: 29-30, 64).4 This is not to imply that analysts of strategic choice ignore constraints imposed by "nonobjective" variables, including the past. The literature on misperception, deception, and limited-information bargaining, for instance, usefully relaxes assumptions of ahistorical and acultural rationality. But there is a tendency in this literature to avoid looking at deep historical and cultural roots of (mis)perception and limited information. Instead, the limitations on rational 1

All terms are taken from March 1978: 590-92. There is a burgeoning literature, however, that points out that in multiple-equilibria games (e.g., coordination games, iterated prisoners' dilemma games, etc.) cultural norms may be vital for coordinating players' expectations around certain equilibria, not to mention for defining preferences and perceived payoffs. See, for instance, Johnson 1991. 3 Adelman and Shih's recent work (1993) on the Chinese use of force since the mid nineteenth century does explicitly argue that China's symbolic use of force cannot be explained with reference to Western concepts of rationality. Their argument, however, confuses rational choice with successful outcomes and misses the rationally instrumental use of strategic symbols. See also Twining 1989. The authors make the mistake of assuming that rational choice embodies a definition of what constitutes utility (see Elster 1986: 4-5). 4 See chapter 2, n. 32, below. 2

STRATEGIC CULTURE: A CRITIQUE

3

choice are attributed either to eccentric characteristics of particular leaders or, conversely, to universal characteristics of decision makers and decision situations (Jervis 1976, Snyder and Diesing 1977, Lebow 1981). Many also acknowledge organizational constraints on strategic choice, under which organizational missions and decision routines are in some sense historically derived (though usually from not too distant antecedents). But often this organizational perspective is an auxiliary theory used to explain gaps in an essentially neorealist framework. In general, then, neorealist theory has little theoretical room for the strategiccultural dimension of strategic choice. The range of strategic choices presented to and decided upon by decision elites can be explained mostly by so-called structural variables such as the nature of power distributions between states in the here and now, where these distributions are given their particular cast by geography, technology, and military capabilities. The decision by elites to choose one from their range of options is framed generally by a calculation of national interest that is, for the most part, universal, or perhaps determined by organizational interests, but certainly not culturally specific. Any particular set of elites placed in a similar situation ought to make a similar choice.5 Most of the proponents of the strategic-culture approach, however, would disagree fundamentally with this conclusion. Rather, they might counter, particular sets of elites socialized in different strategic cultures will make different choices when placed in similar situations. Thus the strategic-culture approach challenges ahistorical and acultural explanations of strategic choice by rooting strategic preferences deep in history and culture, and not predominantly in system structure, or the distribution of state capabilities. Since strategic cultures are unit-level attributes and vary across states or societies, we should expect similar strategic realities to be interpreted differently. So the problem for culturalists is to explain similarities across cultures in strategic behavior when cultures vary. Conversely, the problem for structural realists is to explain cases of differences in strategic behavior when structural conditions are more or less constant. While there is no a priori reason for predictions about strategic choice derived from strategic culture to be different from predictions 5 A sophisticated example of the neorealist analysis of strategic choice is Barry Posen's study of the sources of military doctrine between the two world wars. When history enters into his explanatory equation, it apparently affects how a state perceives the costs and benefits of expansion, and thus affects how the state responds to opportunities to optimize its influence. Generally, however, this historical influence comes in the form of a lesson learned by most states: namely, that a hegemonic power will be opposed by other states, and that hence it is prudent to uphold the status quo, or at least to curb one's expansionism such that it does not provoke the development of an antihegemonic alliance. Because of this lesson, the security dilemma does not always bring unceasing violence and massive destruction to the international system (Posen 1984: 68-69). Posen does not explore this historical learning much. Moreover, this lesson is more or less cross-cultural.

4

CHAPTER ONE

derived from ahistorical structural approaches (any differences depend on the content of a strategic culture), there is no a priori reason for them to be the same either. The possibility of different predictions about state behavior underscores the potential analytic value of understanding the concept of strategic culture. What, then, does the literature on strategic culture suggest about the sources of a state's strategic choices, and about the ways in which states use force for political ends? What follows is, first, a discussion of the concept of strategic culture in the international-security literature, followed by a look at the literature on Chinese strategic culture. This leads to a final assessment of the conceptual and methodological problems in the current work on strategic culture. It is important to spend some time on these questions because, given the underdeveloped state of the concept, any useful rethinking and testing of strategic culture depends heavily on understanding the flaws in present conceptualizations.

T H E INTERNATIONAL SECURITY FIELD AND STRATEGIC CULTURE

Traditionally, the question of culture has not attracted much attention in international security studies and international-relations theory. As Joseph Nye and Sean Lynn-Jones concluded in a report on the state of the field, strategic studies has been dominated for too long by American ethnocentrism and a concomitant neglect of "national styles of strategy" (Nye and Lynn-Jones 1988: 14-15)—a criticism echoing that of a number of scholars who used strategic culture or a variant in their own research (Booth 1979, Ermarth 1981:52, Lord 1985: 269-270).6 Nonetheless, in the last ten to fifteen years, there has been a growing interest in strategic culture (and culture in general) as explicative of the security behavior of states, so much so that the American Academy of Sciences held a workshop on strategic culture—the first of its kind—in May 1990 in an effort to refine a definition of strategic culture and to discuss the potential contributions of strategic culture to the analysis of states' strategic behavior. The workshop brought together participants from a number of disciplines—social historians, anthropologists, area specialists, and mainstream neorealist strategic analysts. Since then there have been several conferences related broadly to cultural or ideational sources of strategic choice,7 a number of books that explicitly incorporate strategic culture into their analysis (Jacobsen 6

On the relative neglect of culture in international-relations theory see Walker 1990: 3-20. One was a workshop on strategic culture and China held in March 1992 and another a series of conferences on norms and national security held in February 1993, and January and October 1994. Both were funded by the SSRC/MacArthur Program in International Peace and Security. Indeed, an important purpose of the SSRC/ MacArthur Fellowships in International 7

STRATEGIC CULTURE: A CRITIQUE

5

1990, Zhang 1992, Kupchan 1994), as well as a growing number of scholarly articles. The research so far on strategic culture—or, more broadly, cultural and ideational influences on strategic choice—can be divided into three generations.8 The first generation, which emerged in the early 1980s, was composed mostly of security-policy analysts and Soviet area specialists, rather than internationalrelations theorists. It focused mostly on trying to explain why the Soviets and the Americans apparently thought differently about strategy in the nuclear age. It attributed these differences to variations in deeply rooted historical experiences, political culture, and geography, among other variables. In other words, the independent and dependent variables were diffuse, sweepingly broad, and the arguments highly determinist. The second generation, appearing in the mid 1980s, focused as well on the superpowers, but argued essentially from a Gramscian perspective. This generation recognized the possibility of a disjuncture between a symbolic strategic-cultural discourse and operational doctrines, the former being used to reinforce the hegemony of strategic elites and their authority to determine the latter. The third generation, which is just beginning to emerge, has been more conceptually and methodologically rigorous. It has narrowed the focus of the dependent variables in order to set up more reliable and valid empirical tests for the effects of strategic culture, and has looked at a wider range of cases. This generation has tended to study the role of strategic and organizational cultural norms in strategic choice in an effort to explain choices that do not fit with the dominant neorealist explanations.

The First Generation The direct inspiration for the initial analyses of strategic culture came from Jack Snyder's short 1977 RAND study on Soviet limited nuclear war doctrine, in which he coined the phrase "strategic culture." Ironically, Snyder's definition and use of strategic culture was far more restricted conceptually than those who took up the term, and he has since tried hard to disassociate himself from the first generation of literature (Snyder 1990). In his study, Snyder referred to strategic culture as "the body of attitudes and beliefs that guides and circumscribes thought on strategic questions, influences the way strategic issues are formulated, and sets the vocabulary and the perceptual parameters of strategic debate" (Snyder 1977: 9). Strategic culture Peace and Security in the last few years has been to fund projects in which "culture" is a key independent variable. 8 These generations are roughly temporal in sequence, though there is some overlap. Some of the literature that belongs conceptually to the first generation appeared after the emergence of the second generation.

6

CHAPTER ONE

in the nuclear age was the "sum total of ideals, conditional emotional responses, and patterns of habitual behavior that members of the national strategic community have acquired through instruction or imitation and share with each other with regard to nuclear strategy" (ibid.). Unlike operational code analysis, which tended to offer strategic axioms as hard and fast rules for behavior, Snyder suggested that strategic culture contained more general "cognitive propensities." Snyder, in contrast to other writers in the first generation, did not see strategic culture as rigidly constraining strategic choices, but rather as coloring debate. Nor did he view a strategic culture as a monolithic concept. Within a strategic culture there could exist subcultures or substrains of strategic preferences (ibid., 14). Unlike other authors Snyder, did not view strategic culture as being rooted in deeply historical-cultural antecedents and formative experiences. Rather, it sprang from a mixture of recent historical experiences, ideology, high politics, organizational interests, and geography (ibid., 8, cf. Snyder 1990: 4). Snyder's was a sensitive and reasonable discussion of the reasons why the Soviets appeared to have had a "disposition to contemplate" the use of nuclear options—a disposition puzzling to those American strategists who believed that the language and concepts of the nuclear age ought to have had more or less universal meaning. Snyder also recognized, in his discussion of operational codes, the danger of mechanically deterministic explanations of Soviet strategic choices stemming from simplistic descriptions of the sociocultural and cognitive characteristics of Bolshevism. But Snyder's own definition of strategic culture seemed to have trouble completely escaping this determinist trap. He referred at one point, for instance, to strategic culture as including ideas, emotional responses, and "patterns of habitual behavior" shared by members of strategic elites. As I will discuss below in more detail, this type of definition is problematic. I raise the issue now because it touches on the central empirical problems in work on strategic culture: how to measure its effects on behavior, and how to determine its explanatory value vis-a-vis noncultural explanations. By subsuming attitudes and behavior, including semiconsciously held assumptions and axioms, within the same concept, Snyder's definition seemed to imply that there is, or can be, a one-to-one correspondence between attitudes and behavior. This relationship is rejected by many definitions of culture in other disciplines, and even seems to be ruled out by Snyder's own critique of operational-code analysis.9 How does one identify a strategic culture when attitudes and behaviors diverge? By indiscriminately linking the two in this way, there is a tendency towards mechanically deterministic explanations of strategic choices that are vulnerable to easily discoverable counterevidence. 9

On the importance of conceptualizing culture separately from behavior see Schuman and Johnson 1976, Elkins and Simeon 1979, Smircich 1983: 347-351, and Barnes 1986: 19.

STRATEGIC CULTURE: A CRITIQUE

7

The first generation of scholarship seems to have ignored Snyder's fairly limited claims about the sources and effects of strategic culture; at the same time, it inherited and magnified the problem of mechanical determinism. This was evident, for instance, in the work by Colin Gray, Carnes Lord, and David Jones on U.S. and Soviet strategic culture. Gray, who borrowed directly from Snyder's conceptual work, defined American strategic culture as "modes of thought and action with respect to force" that derive "from perception of the national historical experience, aspirations for self-characterization (e.g. as an American, what am I, how should I feel, think, and behave?), and from all the many distinctively American experiences (of geography, political philosophy, of civic culture, and 'way of life') that characterize an American citizen" (Gray 1981: 22).10 Thus strategic culture is distinctive and unique, given the disinctiveness and uniqueness of these inputs. Moreover, these inputs, in contrast with Snyder's list, had much older antecedents, deeper historical and cultural roots extending as far back as the Seven Years' War. By 1945, the net result was a set of "dominant national beliefs" with respect to strategic choices. These beliefs included: a belief in the essential goodness of one's cause, hence a reluctance to wage wars for goals "that are controversial in terms of enduring American ideas of justice" (ibid., 26), hence too a view that wars were aberrations from a natural rational order; a basic optimism about triumph and inevitable victory, hence expectations for total victory in war; a sense of omnipotence derived from a history of successfully fought wars; and a sense of unlimited resources that should be applied with overwhelming effect against an enemy, hence a reluctance to expend many human lives in pursuit of victory (ibid., 27-29). Some of these characteristics were carried across the threshold of the nuclear revolution to produce a peculiarly American approach to nuclear strategy, argued Gray. This approach stressed, among other things: that nuclear wars could not be won because the high level of human casualties would erase any meaningful concept of political or military victory; an optimism about the American technological capacity to provide an effective nuclear deterrent in the face of eventual Soviet advantages in numbers and yields of nuclear weapons; and an optimism that arms-control dialogue could teach the Soviets to think and speak the American nuclear language, thus leading to converging and hence stabilizing views on deterrence and conflict management. Gray concluded that this relatively homogeneous American strategic culture differed fundamentally from that of the Soviet Union, and that Americans were in general incapable of thinking strategically. In other words, unlike the Soviets, American leaders were incapable of thinking about, planning for, fighting, and winning a nuclear war. As an influential advocate of counterforce war-fighting doctrines, Gray's 10

Gray's ideas are developed somewhat further in Gray 1986: 33-63 and Gray 1990: 45-48.

8

CHAPTER ONE

use of strategic culture had an obvious policy agenda.11 Putting this aside, however, his definition presented analytical problems both similar to and different from Snyder's work. First, by subsuming behavior in a definition of strategic culture ("modes of action"), he implies that strategic thought leads consistently to one type of behavior. How does one evaluate a strategic culture where thought and action seem inconsistent? Alternatively, is it always the case that one type of behavior reveals one set of distinctive patterns of strategic thinking? Gray's definition led him to the sweepingly simplistic conclusion that there is one U.S. strategic culture that is incapable of conceptualizing a war-fighting, war-winning nuclear doctrine, and that this feature distinguishes it from the one Soviet strategic culture. Like most mechanistically deterministic cultural arguments, Gray's conclusions avoided ample counterevidence; namely, the evidence that in operational terms planners in the Strategic Air Command always considered counterforce war-fighting war-winning options (Kaplan 1983, Pringle and Arkin 1983, Herken 1985, Sagan 1987).12 This relates to a second problem: the implication that the public is a repository of strategic culture, or that the strategic decision makers and the public at large share a strategic culture. This may or may not be the case. Even granted that the American public has a predisposition to think about nuclear war as unwinnable, and thus is unwilling to conceive of ever rationally fighting one, and even granted that a sizable portion of U.S. political elites and members of the strategic community share, or are constrained by, these predispositions, it is not necessarily the case that those in operational planning share this strategic culture. In other words, there may indeed be a close connection between strategic culture and behavior, but the strategic subculture upon which Gray focused may be less relevant for explaining how the U.S. would in fact behave in a nuclear crisis.13 Following closely in Gray's footsteps was Carnes Lord's study of American strategic culture. Definitionally, Lord seemed at first to avoid the determinist trap by arguing that strategic culture—the "fundamental assumptions governing the constitution of military forces and the ends they are intended to serve"— establishes "a basic framework for, if they do not determine in detail the na-

11 The interest in historically rooted, culturally unique styles of American strategy was not limited to conservatives and hawks. See also Shy 1971, Rappoport 1978: 63-64, and Chace and Carr 1988: 318-321. 12 He also misses evidence that Soviet strategists did adopt some American concepts in arms control and strategic stability. See Adler 1992. 13 In a later work on strategy, Gray comes close to implying a disjuncture between elite and mass strategic cultures. He suggests that strategic culture persists because it is perceived by "those charged with policy navigation for that society" to have had positive effects. This statement opens the door to the argument that strategic culture may be used instrumentally by elites to perpetuate basic strategic policies that serve their own interests. See Gray 1990: 45-46.

STRATEGIC CULTURE: A CRITIQUE

9

ture of, military forces and military operations" (Lord 1985: 271). This basic framework appears to color debate over strategic options by injecting sometimes "unconscious and unsystematic" historically rooted concepts on the nature of war, the necessary preparations for war, and the methods of war (ibid., 289-290n.5). Yet, at the same time, Lord undercut this flexible, delimiting notion of strategic culture by allowing the concept to include both "traditional practices (i.e., behavior) and habits of thought by which military force is organized and employed by a society in the service of its political goals" (ibid., 271). Like Gray, Lord fell into the same mechanically deterministic trap: he was unable to explain disjunctures between practice and thought, and was inclined to conclude that one set of practices reflects one set of thoughts and vice versa. The result, however, fit nicely into his policy agenda as well. American practice and American thought—influenced by geography, international system structure, political ideology, military history, civil-military relations, organizational interests, and technology—has displayed a consistent tendency towards astrategic thinking. "Liberal democracies are deeply disinclined to prepare adequately for war, or to foster the institutions and types of men capable of waging it" (ibid., 273). When war does break out, however, liberal democracies like the United States are "inclined toward waging war a outrance, with insufficient consideration for the political goals which should define strategic objectives and give meaning to victory" (ibid.). Geography, political culture, and military history have given the U.S. a strategic culture that is "fundamentally defensive" at the level of strategy, and that unfortunately, in Lord's view, eschews counterforce war-fighting and war-winning nuclear strategies. This stands in diametrical opposition to Soviet strategic culture (ibid., 272, 277). When the literature turned to Soviet strategic culture it accepted the amorphous definition of strategic culture offered by Gray and others in the first generation. In an essay on Russian cultural influences on Soviet strategy in a volume devoted to the application of strategic culture to the analysis of U.S. and Soviet strategic power, David R. Jones argued that there were three levels to a state's strategic culture: its basic elements born from geographical, ethnocultural, and historical variables; the socioeconomic characteristics of society, and its political structure; and, most narrowly, the nature of contemporary interaction between military and political institutions. Together these variables interact to create a state's strategic culture (Jones 1990: 37). This strategic culture does not just delimit strategic options. Rather, for Jones strategic culture pervades all levels of strategy from grand strategy down to tactics: "Such a 'culture' presumably affects the whole range of a nation's broad security and more narrow military policies, beginning with the basic goals of its diplomacy and ending with the 'style' or 'whole series of proclivities' displayed by its armed forces in peace and war" (ibid., 35). In a sense this conceptualization attributes to a state's formative historical

10

CHAPTER ONE

experience an even more rigid hold over strategy and tactics than perhaps even Gray would admit. Like others in the first generation, however, Jones' determinism runs into problems of logic and evidence. In the first place, there is no room in his definition for multiple or competing cultural inheritances. Like Richard Pipes (1977, 1980), Jones assumed that Russian history and culture embodies only one set of strategic beliefs, assumptions, and preferences, which manifests itself in one set of consistent strategic choices across different technological and threat contexts. But then the strategic culture argument becomes tautological: Soviet military doctrine was the inevitable product of Russian/Soviet strategic culture because the strategic culture is rooted in a historical-cultural milieu that embodied similar perspectives and approaches to violence. Since only one set of historical-cultural perspectives exists, then Soviet strategic culture must necessarily have reflected these. There was no room for competing strategic-culture tendencies or strategic choices. There was no room for competing explanations of strategic choice. There was no room for the possibility that different strategic choices can emerge from similar strategic-cultural backgrounds. Strategic culture becomes unfalsifiable. Second, accepting the possibility that Russian/Soviet military leaders consciously put their own thinking in the context of deep historical traditions, it does not automatically follow that Soviet strategic behavior reflected the influences of premodern Russia. It is possible that these leaders misinterpreted these traditions—as, for example, today's Sun Zi specialists in China and Taiwan exaggerate the role of nonviolence in traditional Chinese military thought. As the literature on cultural symbols and political power suggests, traditions are constantly redefined and reinterpreted by successive generations of elites with a political interest in highlighting or downplaying particular traditions. These redefinitions, however, are invariably cast as true and accurate recreations of the past. Third, simply because different societies have developed under different historical and geographical conditions, this does not mean that any strategic culture produced in these conditions is necessarily unique. It is logically possible for different circumstances to produce similar strategic cultures because two different societies may face similar strategic problems. Since strategic conditions and experiences are also sources of this amorphous notion of strategic culture, it is possible that the strategic predispositions derived from particular geographical or historical conditions could be inconsistent with those derived from particular strategic problems. This opens up the possibility that the diversity within any particular society's geographical, political, cultural, and strategic experience will produce multiple strategic cultures, a possibility excluded by the narrow determinism of the first-generation literature. Why expect that such a diverse range of variables should produce compatible elements of a strategic culture, given that, considered on their own, each variable might make different predictions about strategic behavior? Britain as an island

STRATEGIC CULTURE: A CRITIQUE

11

nation was at different times an expansionist imperialist power and a defensive balancer. Britain as an emerging democracy in the late nineteenth century was a defensive balancer in Europe, while the U.S., as a liberal democracy in the latter half of the twentieth century adopted an offensive strategy of global anticommunist containment. China as a large bureaucratic authoritarian empire exposed to external threats from nomadic groups from the north and west has historically eschewed, according to some, expansionist, offensive doctrines. Russia, also a bureaucratic, authoritarian empire, adopted, according to some, an expansionist, offensive approach to dealing with threats. Clearly, by themselves geography, political culture, or particular historical experiences do not lead to consistent approaches to questions of war and peace. Indiscriminately combined, they create overdetermined, hence analytically useless, explanations of strategic behavior. Finally, Jones and others of the first generation treated strategic culture, in essence, as a constant. To the extent that there are variations in strategic preferences and behavior, strategic culture alone is not a particularly convincing explanation of choice. This is not a big problem for many of the authors because they are unwilling to admit that there is much diversity in any particular state's strategic preferences. Yet at one point in his essay Jones admits that in the nuclear era the Soviets have interpreted Clausewitz's dictum about war as a continuation of politics to justify the use of nuclear weapons during war and to reject nuclear war as incompatible with rational political ends (Jones 1990: 43). How does a single, constant, deeply rooted strategic culture produce such a dramatic variation in strategic nuclear preferences? The first generation of literature on strategic culture, then, falls short of showing that historically and culturally based approaches to strategy do affect strategic choice in any appreciable way. Unfortunately, the work of this generation has tended to dominate the debate over the relationship between culture, strategic culture, and strategic choice.14 In addition to problems in specific studies, the serious flaws in the strategic-culture approach developed in 14

Moreover, members of this generation had considerable influence in developing the justifications for refining counterforce war-fighting nuclear doctrines during the Reagan era. Soviet strategic culture was a logical source of the aggressive strategic behavior that characterized the "evil empire." It was not sufficient for strategic stability to target Soviet cities for retaliation, because Soviet strategic culture did not value populations as much as political and military power. Thus, targeting and preparing to knock out the centers of Soviet power first might just deter the Soviets from launching a first strike. U.S. strategic culture, however, put the U.S. at a disadvantage when facing down the Soviet threat, and explained why remedial efforts to break out of the constraints of an astrategic strategic tradition were so difficult yet so necessary. Despite the shoddy construction and often obvious political use of the strategicculture concept by proponents of nuclear war-fighting, the Reagan era discourse on nuclear strategy was based on the view that history and culture propelled both superpowers along vastly different strategic trajectories that, nonetheless, were highly likely to lead to dramatic confrontation somewhere down the road. See Scheer 1982, Herken 1985, and Kull 1988.

12

CHAPTER ONE

this literature are conceptual, methodological, and empirical. For the moment, let me address the former two issues, since it is against these problems that in the next section I reconstruct and juxtapose a more useful conception of strategic culture. One set of problems is definitional. For the first generation, the elements of strategic culture are quite diverse, yet its effects are extremely constraining. Technology, geography, organizational culture and traditions, historical strategic practices, political culture, national character, political psychology, ideology, even international system structure, among other variables, are all considered to be relevant inputs into this amorphous strategic culture. In this sense, the first-generation arguments are at once both overdetermined (strategic culture as an independent variable is the product of a vast range of other independent variables) and underdetermined (a single strategic culture explains all or most strategic choices). Yet, arguably, these variables are different classes of inputs and come from different levels of analysis. They could stand, and indeed have stood, by themselves as separate explanations of strategic choice (i.e., geography in the form of geopolitical theories; organizational culture in the form of organizational interest and interest-group models of choice; technology as a variable in neorealist models of capability-based choice). There is, to be sure, nothing in principle (other than parsimony) that forbids combining these variables into progressively more complex explanatory models. But if the aggregated variables contain so many separate independent inputs, at least two problems arise. The first is how to weigh the effects of these various inputs. This is important because different weights presumably explain the appearance of different types of strategic culture and strategic choices. There is no effort in the first-generation literature to attach weights to these inputs. Second, and more serious, the aggregate concept becomes unfalsifiable. If strategic culture is the product of most relevant explanatory variables, then there is little conceptual space for a nonstrategic cultural explanation of strategic choice. There is nothing that is not strategic culture. In a related vein, the literature does not indicate when there is enough variation within a state's (or its elite's) strategic propensities such that an integrated strategic culture no longer exists. If it can be shown, for instance, that a state has at least two very contrasting sets of strategic propensities (minimum deterrence versus counterforce war-fighting doctrines, for example) then it obviously makes less sense to talk about a single strategic culture. Yet the literature makes little allowance for the possibility that there is either no strategic culture or that there is more than one strategic culture in a society. The first-generation analysts might respond that they are not mechanical determinists. That is, strategic culture does not predetermine strategic preferences and choices, but only produces tendencies, or modal behaviors. Indeed some backtrack by arguing that strategic culture tends to lead to particular

STRATEGIC CULTURE: A CRITIQUE

13

strategic behaviors, or that strategy is in part a product of culture (Booth 1990: 124, 126). If this is the case, however, the literature is conceptually and methodologically ill-equipped to isolate which part of strategy comes from strategic culture. To make this less determinist claim, the first-generation literature would have to provide a theory explaining why particular tendencies or modes of strategic behavior are prominent at particular times. Thus they need to offer a theoretical explanation of why certain historical tendencies in strategy persist and others do not. But then much of the explanatory power for strategic choice would reside in this intervening variable, not in strategic culture itself. Moreover, to admit that within a particular state there are competing strategic tendencies—as a less determinist claim would admit—is also to admit the possibility that a similar range of competing strategic tendencies may exist in other states. If this is the case, then unique historical, geographical, and experiential conditions in any particular society count for very little since these cannot explain why similar ranges of strategic choices are present in other "unique" societies, much less why particular strategic tendencies become dominant at particular times within any one society. If those in the first generation wanted to create a falsifiable strategic-culture concept in order to isolate the relative influence of strategic culture on strategic choice, this would be impossible given their "everything but the kitchen sink" treatment of strategic culture. Such an approach makes tests of a strategic culture-based model of choice difficult, since there are no nonstrategic cultural variables against which to set up such tests. Certainly the literature does not set these up. In fact, there is no effort to rule out alternative explanations for strategic choice, in part because of the amorphous definitions of strategic culture, and in part because the literature as a rule lacks any notion of research design. Yet the underlying assumption of the strategic-culture literature is precisely that strategic culture-influenced behavior is different from behavior uninfluenced by strategic culture, or influenced by different strategic cultures. That is, the proponents of the strategic-culture model of choice would claim to provide different explanations from other, acultural models; but their writings do not actually test these differences because their definition of strategic culture makes such tests virtually impossible. A second set of problems concerns the relationship between strategic culture and behavior. Given the all-encompassing nature of strategic culture as an independent variable, there is no possibility in most of the first-generation literature for there to be a break or disjuncture between strategic culture and behavior. Instead, the literature tends to suggest one of two possible relationships between "attitudes and behaviors." The first is that strategic culture includes historical behavior as a key input. That is, behavior at time t-1 is a key explanatory variable for behavior at time t. This is hardly a novel proposition. It would be an interesting proposition if this behavior were deeply historical, say, several generations or more removed from behavior at time t, but the

14

CHAPTER ONE

literature is confused as to whether strategic culture derives from deeply historical antecedents or recent ones. As Mary McCauley notes in a critique of work on political culture, "we must know just which period 'history' refers to—when it begins and ends—and how we can assess the influence of that period with the influence of the more immediate past" (McCauley 1984: 22). Instead, the literature on strategic culture skips over this problem, and assumes a relatively unbroken chain between historical strategic behavior and contemporary strategic choice. The second type of attitude-behavior relationship is that strategic culture (excluding behavior) has a measurable effect on strategic choice—that is, that strategic culture exists "out there," a monolithic, independent, and observable constraint or weight on all actors' behaviors. There is little or no appreciation of the potential instrumentality of strategic culture, where strategic culture is used to clothe strategic choices in culturally acceptable language, and hence to justify the competence of decision makers, deflect criticism, suppress potential dissent, and limit access to the decision process. This is unfortunate since, with the absence of a concept of instrumentality, the literature then implies that if a link between attitude and behavior is not found, then strategic culture does not exist. Yet as cultural analyses of religion, ideology, and organizations suggest, it is quite possible to find coherent, integrated, consistent sets of assumptions, ideas, axioms, etc., that have only a tenuous connection to observable behavioral choices (Glenn et al. 1970, Broms and Gahmberg 1983, Laitin 1988:591). A third set of problems concerns the derivation of an observable strategic culture. There is little explication of, let alone agreement on, this process. What are the sources for representations of strategic culture, and over what historical periods? Should we examine the writings of strategic thinkers, doctrinal statements, force structures, the attitudes of individuals in the military, or popular representations of military themes? From which time periods should these "texts" be taken? Why are certain historical periods considered formative and authoritative sources of strategic culture and others not? How is strategic culture transmitted through time? Does it change appreciably through its transmission? In general, none of these questions is explicitly asked or answered in the first-generation literature. In part, this probably relates to confusion over whether strategic culture should be treated as a subcomponent of a society's culture in a broad sense, or as a separate entity, rooted in early and formative strategic thought, and hence at least once removed from broader cultural antecedents and characteristics. If strategic culture is the former, then the objects of analysis—the strategic cultural artifacts—should include nonmilitary texts as well. If the latter, then the objects of analysis are strictly military or strategic artifacts. This choice also has a bearing on hypotheses about how strategic culture is transmitted over time. There is little in the literature to suggest how strategic culture is transmitted, whether through broader cultural processes

STRATEGIC CULTURE: A CRITIQUE

15

(i.e., education, mass media, political socialization, etc.) or narrower strategiccultural processes (i.e., military education, the construction of military histories, the search for military precedents in the strategic decision process, etc.).

The Second Generation The second-generation literature begins from the premise that there is potentially a vast difference between what leaders think and say they do and the deeper motives for doing what they in fact do. While much of the second generation draws more or less explicitly on Gramscian notions of strategic culture as a tool of political hegemony in the realm of war and peace, not all of it does. There is also some provocative historical research that has focused on the disjuncture between rhetoric and explanation in strategic choice. Reginald Stuart's work on the American war myth, for instance, traced the development of a myth about how Americans think about and behave in wars— a myth rooted in republican nationalism of the late eighteenth and early nineteenth centuries. This myth implies—to use a term Stuart does not, but that fits—that U.S. strategic culture stresses the aberrance of war, but also promotes a crusading anti-Clausewitzian totalistic and moralistic approach to fighting wars (Stuart 1982, chap. 7). This myth is perpetuated in most of the contemporary scholarship on U.S. foreign policy and military history. Yet, Stuart argued, it is still a myth, disconnected from the realities of early American thought on war; there is a recognizably Clausewitzian, realpolitik strain in American military thinking historically. This tradition viewed war as an inevitable feature of interstate relations, and recognized the need to use force to expand militarily, if necessary, for national interests, and not simply for great causes. Both of these approaches to war, Stuart argued, affected why and how the United States went to war, why and how it made its strategic choices. He concluded, finally, that the "many components of the American character suggest that the American view of war is really a collection, and historically therefore a series of collections, of views" (ibid., 193). Stuart's work hints at a couple of extremely useful caveats for the study of strategic culture, caveats missing in much of the work of the first generation. First, a state may have a strategic culture, but it may remain at the level of myth. That is, strategic culture may consist of images, stories, and symbols that in their totality serve to legitimize dominant conventions about a state's past, present, and future strategic behavior (Dittmer 1977: 579, Pettigrew 1979: 576).15 But this leaves the relationship between strategic culture and strategic 15

Well before organizational-culture theorists or critics of strategic studies posited a disjuncture between "espoused theories" and "theories-in-use," Singer (1968: 145) pointed out the contrasts between a group's "official or articulated ideology" and its "operative ideology." Ideology, he argues, is one component of culture, and the one most likely to constrain or delimit

16

CHAPTER ONE

behavior even more ambiguous. Is the relationship one of instrumentality? That is, is the strategic-culture myth used consciously to mask or justify essentially nonstrategic culture-derived behavior? Or is strategic culture used semiconsciously as the language of debate, with little or no awareness of the disjuncture between this language and actual strategic choices? Second, there may well be more than one strategic culture—Stuart's "collection of views"—that do delimit choice in some observable fashion, but they do so at different times for different reasons. But this possibility raises at least two additional methodological problems: How does one determine the boundaries of distinctive strategic cultures within one society? That is, what criteria should be used to clarify the clusters of views, assumptions, and beliefs that form separate strategic cultures? And what are the intervening variables that determine when one strategic culture or another emerges temporarily as the dominant one? The conclusion that there is a potential disjuncture between strategic-culture myths and behavior seems to inform the work of writers on the political left, perhaps because culture as an ideology has been a central focus of neoMarxist, Gramscian analyses of the state. Bradley Klein, for instance, treated military strategy as a cultural practice, a reflection of a hegemonistic political order. Strategic culture . . . involves widely available orientations to violence and to ways in which the state can legitimately use violence against putative enemies. In this sense popular representations of violence and of the "enemy" against whom violence is to be legitimately deployed become significant artefacts insofar as they construct as plausible a distinct range of identities and render others unavailable or implausible. To study strategic culture is to study the cultural hegemony of organized state violence. (Klein 1988: 136) Nowhere in this definition were there any hidden assumptions about the effect of strategic culture on strategic choices themselves. Indeed, in reference to American nuclear policy, Klein pointed to the important gap between strategic culture-derived declaratory strategic policy and operational policy. The latter, he suggested, is the nuclear strategy of an America as a strategic

the responses of the group to its environment. There are parallels to this in Knud Krakau's deeply historical explanation for the American national style. Americans have historically assumed war to be an aberration, but have tended to fight wars in a crusading style. Unlike Gray, though like Singer and Stuart, Krakau admitted that one element of this national style is "modes of rationalization and rhetoric" that may mask the absence of a unique national style in behavior. He implied as much when he commented that as the U.S. became increasingly integrated into regional and global affairs there appeared a "growing distance between traditional action [e.g., European, realpolitik behavior] and continued revolutionary rhetoric or verbal strategy used to describe and rationalize this action" (Krakau 1984: 271).

STRATEGIC CULTURE: A CRITIQUE

17

hegemonic power, stressing war-fighting in defense ofAmerican interests (ibid., 139-140). Declaratory strategy is used instrumentally by political elites to fashion a culturally and linguistically acceptable justification for operational strategy, and to silence or mislead potential political challenges to the latter. Elsewhere Klein pushed this argument further, claiming that military strategy is engaged in the construction and circulation of particular visions of political life . . . these visions entail the violent domestication of forces that are presented as external, alien and in need of taming. To conventional wisdom, such a claim must surely stretch beyond credibility the boundaries of plausible interpretation. For the prevailing view . . . is that military strategy is about a real world that has an existence "out there" to which responsible leaders must fashion a response. From this standpoint, to interpret strategy as a textual production is to deny the objective integrity of the enterprise. (Klein 1989: 99-100) While for Klein strategic culture was instrumental, it did not come out of thin air, nor out of the pockets of political and military elites. He implied that strategic culture is a product of historical experience. Since these experiences are different across states, different states exhibit different strategic cultures (Klein 1988: 139). But since there is a radical delinkage between strategic culture and behavior (or operational doctrine), and since behavior is the reflection of the state interests or organizational interests of a hegemonic group, strategic choice is constrained by these interests. It is therefore possible that states speak different strategic-culture languages—as hawkish critics of U.S. MAD concepts stressed about the USSR—but that states' body language (operational doctrine and behavior) are essentially similar. A variation on this theme of instrumentality was provided by Robin Luckham in his study of "armament culture." In contrast to other writers on strategic culture, Luckham argued that armament culture is not a product of national historical development, but of global historical processes. He defined armament culture as arms fetishism that establishes an essential causal relationship between modern weapons, military superiority over enemies, and security. This fetishism seeps into all aspects of culture and mentality. Human consciousness embraces the weapons-security link and accepts that it is powerless in the face of the "war system" to alter it, and thus accepts as self-evident truths the necessity of arms racing, nuclear deterrence, etc. This arms culture serves the interests of its producers—strategists, statesmen, soldiers, and arms manufacturers. Culture is, in this sense, instrumental, but it is cross-national. It is not unique to ethnocultural systems, but it is unique to a level of global industrialization, militarization, and capitalization (Luckham 1984: 1-2). There is, however, a problem with the second-generation literature on strategic culture, arising from the relationship between the symbolic discourse— the strategic culture—and behavior. It is not clear from the literature whether

18

CHAPTER ONE

we should expect strategic discourses to influence behavior. Instrumentality implies that dominant elites can escape from, or rise above, strategic-cultural constraints that they themselves manipulate. Yet recent scholarship on leadership would suggest there is a dialectical relationship between strategic culture and operational behavior. Elites, too, are socialized in the strategic culture they produce, and thus over time are constrained by the symbolic or "textual" myths that they or their predecessors created (Hollander 1985, Snyder 1991: 41-42, 49). This raises the possibility that elites can be captured by the symbolic discourses they manipulate. Thus one should expect cross-national differences in behavior to the extent that these discourses vary cross-nationally. Indeed, the second-generation literature seems of two minds on whether to expect cross-national differences in operational strategy. On the one hand, to the extent that the symbolic discourse delegitimizes certain strategic options by placing them outside of the boundaries of acceptable debate, the strategic possibilities available to states vary across strategic cultures. Thus there is a possibility that behavior may vary. On the other hand, there is a strong implication that strategic elites around the world ought to share similar strategic preferences. These are essentially militaristic or realpolitik since different national strategic discourses tend to create similarly stark visions of a threatening external world (Bjork 1992, Campbell 1992). In an effort to create ingroup solidarity for the purposes of political hegemony, elites foster images of danger. These images tend to correlate with responses to danger that stress zero-sum notions of conflict, the efficacy of force, and the justness of cause. The second-generation literature cannot solve this problem, in part because most of it has not looked at comparative cases to trace empirically how certain discourses and symbolic languages have actually narrowed debate, to determine whether this narrowing differs across cases, and whether the choices and options that remain are different across cases.16

The Third Generation The third generation, which emerged in the 1990s, tends to be both more rigorous and more eclectic in its conceptualization of ideational independent variables and more narrowly focused on particular strategic decisions as dependent variables. Some use military culture, some political military culture, and others organizational culture, but all take the realist edifice as the target, and 16 Interestingly, the second generation comes to similar conclusions as the first generation about exceptionalism and moralism in U.S. foreign policy and strategic discourse (Luke 1989, Bjork 1992, Campbell 1992). The difference is that for the critical analysts this exceptionalism translates into an aggressive capitalist imperialism sold at home as benign, enlightened defensiveness, whereas for the first-generation writers it translates into defensiveness at best and strategic paralysis at worst.

STRATEGIC CULTURE: A CRITIQUE

19

focus on cases where structural definitions of interest cannot explain a particular strategic choice. These definitions, for the most part, explicitly exclude behavior as an element, thereby avoiding the tautological traps of the first generation. Other than this, the definitions do not vary dramatically from those found in political culture, organizational culture, or the first-generation work on strategic culture. The sources of these cultural values are, unlike the first generation, less deeply rooted in history, and more clearly the product of recent practice and experience. Kier (1992), for instance, argues that doctrinal change in the French and British armies between the two world wars cannot be explained by changes in external structures or balances of power nor by putative military organizational preferences for offensive doctrines. Rather, the roots are to be found in changes in the organizational cultures of both militaries. These cultures consisted of three levels of ideas, beliefs, and norms. The first and broadest level concerned values about the nature of international politics and warfare. The second concerned the relationship between the military, the state, and society, and the third concerned internal organization, technology, and officer-soldier relations. Similarly, Legro looks to military organizational cultures—defined in a similar fashion as "patterns of assumptions, ideas, and beliefs that proscribe how a group should adapt to its external environment and manage its internal structure" (Legro 1995)—to explain different levels of restraint in the use of submarine attacks on merchant shipping, the bombing of civilians, and the use of poison gas in the Second World War. Realist explanations, Legro contends, would suggest that there should have been no variations in restraint on the use of these military tools, indeed that there should have been no restraint whatsoever. Whether state interests are defined in terms of survival or the maximization of power, a realist explanation would suggest there should be few selfconscious endogenous limits on the tools of warfare.17 The third-generation literature also tends to be more eclectic in methodology and empirical focus. Some have been influenced by reflectivist or constructivist, broadly postmodern work on norms (Price and Tannenwald 1994). Others exude a Gramscian flavor where strategic culture is the ideational mass product of symbolic manipulation by decision elites aimed, initially, at creating mass support for particular strategic choices (Kupchan 1994). Others hew to a more traditionally positivist methodology, uncomfortable with the scientifically indeterminate findings from strictly intepretivist approaches (Legro 1995). Despite this eclecticism, however, there is a tendency across 17 A similar theme is taken up by Price and Tannenwald 1994 in explaining taboos on the use of nuclear weapons in the Cold War period. Instead of invoking organizational culture as an explanation, however, they suggest that the rise of a norm of nonuse reinforced perceptions in the United States about the military and political disutility of use.

20

CHAPTER ONE

approaches to take ideas and habits of mind seriously as empirically observable constraints on choice. As for determining the relationship between cultural or ideational elements and behavior, the third generation exhibits a couple of strengths. First, it avoids the determinism of the first generation. In part, as noted, this is a function of this generation's care in leaving behavior out of the independent variables. In part it is because some in this generation conceptualize culture in such a way as to allow it to vary. Kier (1992), for instance, views political-military culture as a product of changing domestic political contexts, hence it varies as domestic politics varies. She also examines her cases cross-sectionally and longitudinally, thus introducing variation in strategic cultures across time and societies. Likewise, Legro also allows for variation in both cultural and noncultural variables because, for him, culture is rooted in recent experience, and not in deeply historical experience. Second, this generation is committed explicitly to competitive-theory testing, pitting alternative explanations against each other. Legro, for instance, tests a realist model against institutionalist and organizational-culture explanations of restraint in war. Kier pits realism, structural-functionalist organizational models, and the concept of military culture against each other. This strength highlights the methodological weakness of the first generation. There are, however, a few problems in this emerging literature that are worth examining. First, precisely because of the carefully delimited empirical foci of much of the work, it is hard to compare the research payoffs with the broader concerns of the first and second generation. That is, by taking culture to mean (relatively) quickly changing values, attitudes, and norms rooted in recent past experiences, this literature has less to say about the influences of broader and more deeply historical differences in (strategic) culture across societies, a question that propelled the initial work on strategic culture. Second, the careful focus on strategic choices that are not explained well by realism brings with it some drawbacks. Realism can be, in fact, quite indeterminate about predictions of strategic choice.18 It is therefore hard to set up conclusive tests pitting realist predictions about strategic choice against ideational, cultural, or normative models without providing very strict and clear definitions of interests in the realism model. A third problem concerns the work that posits organizational culture as the key independent variable in strategic choice. The third generation shares the first generation's belief that cultural variables indeed have an observable, measurable effect on behavior. Yet this research misses a key strain in organizational-culture literature that posits a vast disjuncture between symbolic (cultural) strategy and operational doctrine.19 This possibility has not been pur18 19

See chapter 2, n. 31, below. See chapter 5.

STRATEGIC CULTURE: A CRITIQUE

21

sued in third-generation work. Some of the literature safely avoids the question because the dependent variable is behavior—such as the nonuse of nuclear weapons, for instance—and not foreign-policy statements or statements of strategic doctrine. But in some cases, military doctrine is a dependent variable, and this raises the problem that declared and operational doctrines may be different. Fourth, by positing short-run variations in culture (i.e., culture is not a constant) much of the third generation's use of culture, it seems to me, moves close to belief-systems analysis. Why does culture change so quickly if it supposedly reflects conscious, semiconscious, and unconscious orientations to the world? Is there no possibility of cultural inertia persisting across longer periods of time and fairly dramatic variations in domestic and international structures?20 What, in the end, is the difference between this use of culture as an independent variable and the notion of unstable belief systems? Fifth, the definition of culture used by the third generation is a fairly traditional one—namely, one that presents decision makers with a limited range of options or in which culture acts as a lens to alter the attractiveness of different choices. This definition therefore requires an intervening noncultural variable or a theory of politics or even another cultural variable to explain why particular choices are finally made. In other words, if organizational culture creates organizational preferences that in the process of policymaking delimit options available to decision makers, where do preferences among these limited options come from? Are they culturally derived from the strategic worldviews, socialization, and experience of civilian leaders? Another problem follows from this delimiting rather than more deterministic definition of culture: if culture is neither a reflection of an individual's beliefs nor a mere aggregation of beliefs captured by modal points in a distribution of beliefs (as previous national character studies suggested), then any one individual will not be completely socialized in that culture. That is, any one individual will not share all the cultural traits of any other. Yet individuals make decisions, and in times of security crises usually a small number of identifiable individuals make key strategic decisions. If these individuals do not wholly reflect the values of a military or strategic culture, then this creates some slippage between those values on the one hand and behavior on the other, since the relationship is mediated by individuals who are not completely rep20

At times it is unclear how inert culture is for the third generation. Both Legro and Kier imply that some organizational cultures can persist after structural or personnel changes, but they also suggest that the time frame for change can be quite short, even deliberately so. If organizational culture is unstable, there can indeed be variation in the independent variable. But if this is the case, then variation in strategic choice depends on variation in the independent cultural variable. Culture then becomes an intervening variable, further diluting its explanatory power. If culture is so unstable, is "culture" the term we want to use to conceptualize these assumptions and values?

22

CHAPTER ONE

resentative of that culture. If this is the case, the power of culture as an independent variable diminishes somewhat, becoming a contextual variable. We have, then, three groups of research on strategic culture, each with its own sets of conceptual and methodological problems. I am less willing to pass judgment at this point on the ultimate value of the eclecticism of the third generation, given that this literature has not really had enough time to generate a sufficient body of research findings. But I am willing to suggest that the first generation—whose conceptualizations and research dominate the literature on strategic culture at the moment—has generally failed to push the concept of strategic culture forward. We come, then, to an interesting point. The literature on strategic culture and strategic culture-like concepts seems to be of two minds about whether historically rooted ideational variables limit the strategic choices of decision elites. These may or may not constrain choice. There are different research implications depending on which conclusion one supports. The former— whether of the first or third generations—implies that research ought to focus on how to extract strategic-cultural influences on behavior from the effects of other variables, particularly those favored by realist approaches. The latter— the second generation primarily—implies that research ought to focus on the uses of strategic culture to obscure or mask strategic choices made essentially in the interests of expanding the domestic and external control of dominant groups. In the first case, the strategic-culture approach does seem to offer a different explanation for strategic choices than that provided by ahistorical structuralist explanations. In the second, the strategic-culture approach challenges the notion that strategic choices are based essentially on a rational calculation of a national interest that exists as an expression of the state as the primary actor in the international system.

STRATEGIC CULTURE AND C H I N A

In contrast to international-security studies, within the China field there seems to be little controversy about the proposition that "deep" history and culture are critical sources of strategic behavior. Indeed, most students of Chinese strategic thought and practice could be placed safely in a strategic-culture school of analysis, though few use the term explicitly. Moreover, most would fit comfortably in the first-generation literature. Most would argue that Chinese strategic culture uniquely stresses nonviolent political or diplomatic means to deal with adversaries, or—when force is absolutely necessary—the controlled, defensive use of violence. This has given Chinese strategic behavior a distinctive minimally violent character.21 21

Atypical example of this interpretation is Adelman and Shih 1993.

STRATEGIC CULTURE: A CRITIQUE

23

In comparison to the international-security literature, however, the conceptual literature on strategic culture-like notions in the China field is even more impoverished. There are a handful of authors who use the term or one like it. Edward Boylan, for instance, claims that China has exhibited a distinctive "cultural style" in war, rooted in the strategic thought of Sun Zi, the famous strategist from the Warring States period (475-221 B.C.). He defines cultural style as "perspectives, outlooks, ways of formulating strategies, methods of attack, etc., which occurred often in national history and should, presumably, occur again if the nation were to go to war" (Boylan 1982: 341). These characteristics are not necessarily impervious to situational change, nor are they necessarily unique to that state. Rather "what one expects to be unique is the total collection of qualities" (ibid., 342). Here Boylan seems to hint that strategic culture consists of distinctive clusters of strategic characteristics, rather than distinctive characteristics per se. Chong-pin Lin, in his detailed study of Chinese nuclear strategy, tries to make the case that there is a persistent "strategic tradition" rooted in the statecraft and strategic philosophy of ancient China, which has influenced contemporary China's nuclear thinking. He variously terms these traditions "military culture" and "strategic culture," but he leaves these relatively undefined, preferring instead to draw from some of the first-generation work on strategic culture (Lin 1988: 6, 7, 12). Lin also laments the dearth of similar studies of Chinese military strategy that explicitly set out from a cultural paradigm.22 One of the few efforts to explicitly incorporate a strategic-culture perspective, writes Lin, is Gerald Segal's work. And, indeed, in a short article on SinoSoviet detente in the 1980s, Segal argues that much of the improvement in relations was due to changes in perceptions of mutual threat. These perceptions were based on the two states' respective "defense cultures."23 He defines defense culture as a "prism" through which "problems of current policy are filtered and refracted." In part, this prism is a creation of "perceptions of historical traditions of defense policy"—that is, historical perceptions of threat, methods of preserving security, and patterns in the structure and deployment of military forces (Segal 1985a: 180). Actually, Segal is inconsistent about the effects of defense culture on strategic decision making. Elsewhere in a monograph on China's use of force after 1949, he argues that the few patterns that do exist in Chinese strategic behavior are typically realpolitik responses. At the level of grand strategy, he concludes, China has pursued a mix of goals and methods from defense to deterrence to 22 This critique, it seems to me, misses the implicit dominance of the strategic-culture approach in much of the literature on China, most of which assumes a strong element of uniqueness in Chinese approaches to war and peace, as Lin himself admits (Lin 1988: 6). 23 It is not clear from his analysis, however, how a slowly changing defense culture produces two different perceptions of threat in a short period of time.

24

CHAPTER ONE

outright expansionism. Moreover, the success of Chinese techniques of limited war has been overblown by Western and Chinese analysts alike (Segal 1985b). Yet in another article, Segal holds to a golden mean, arguing that analysts should try to avoid both excessive ethnocentrism (i.e., that distinctive historical-cultural influences are irrelevant) and excessive "ethnic-chic" (i.e., that historical-cultural influences are predominant) (Segal 1983-1984: 15). There is, then, little scholarly work in the English literature on China that explicitly incorporates, let alone develops, the concept of strategic culture, even though the literature generally accepts that there are persistent and consistent culturally based principles unique to Chinese strategic thought and practice. Within the Chinese-language literature on Chinese strategic thought, the idea that deeply rooted national styles or cultural impulses affect strategy is certainly not an alien one either.24 Ding Zhaoqiang, a Taiwanese student of strategy, notes, "One can positively state that a state's military thinking, particularly its military strategic thought, is generally influenced by its traditional culture" (Ding 1984: 184). In a lengthy study of deception in warfare, two mainland China scholars, Li Bingyan and Sun Jing, argue that a culture's "military outlook" {junshi guannian JjUJfJl!^) is a critical though understudied element in national strategy (Li and Sun 1989: 20). Some Chinese scholars have used the term "military culture" {junshi wenhua ^ ^ ^ f t ) to describe a consistent thread of strategic thought and practice that was historically developed and inherited by the Han people (Wu 1989:21, 25). This concept of military culture seems to be broader than some Western concepts of strategic culture; it includes styles of military preparations or "weapons culture" (wuqi wenhua jSiUjIZ&lft), military systems or "organizational culture" (bingzhi wenhua J=£ft!l3t/fc), and military thought or "thought culture" {siwei wenhua j g j f ~$£it) (Interview 1990a, Interview 1990b). While the former two cultures are certainly not excluded from some Western definitions of strategic culture, the last one seems closer to notions of strategic culture in the international-security literature. Other Chinese analysts have suggested that what the Western literature terms strategic culture is in fact best seen as the predominant "strategic value system" (zhanlue jiazhi guan 1^B§ ffifjIiS) at a particular point in history. This value system provides a society and its military with definitions of interests, and thus also places limits on the methods and scope of war (Interview 1990b). The origins of this value system 24 Indeed, one of the earliest analyses of national styles of warfare is in Wu Zi's Art of War (ca. fourth century B.C.). In the chapter on "Estimating the Enemy" (liao di $4j§fc) the author presents a lengthy analysis of the sociological and psychological characteristics of the six potential enemy states, their leaders and people, and their concomitant styles of warfare. On the basis of this analysis he proposes different strategic responses to these different characteristics. See Liu 1955a.

STRATEGIC CULTURE: A CRITIQUE

25

lie deep in the formative economic and cultural institutions of a society. The strategic value systems also seem to reflect culturally rooted "thought processes" or "cognitive processes" (si weifangshi JgJU^f ^0 that affect strategic choices. One analyst suggests, for instance, that cyclical or wholistic tendencies in Chinese thought created the notion of a natural limit to the use of force for expansive purposes; once this limit was reached there was a natural decline in morale, capabilities, resources, and qi (Md- Hence to avoid strategic decline, the use of force should go up to, but not beyond, this limit (Interview 1990b, Xu 1986: 25). In general, then, in the small amount of research on strategic culture and China there is a fair amount of conceptual overlap with the first-generation Western literature on strategic culture. Strategic culture (or military culture or defense culture or cultural style) is rooted in early formative military thought and practice. It limits the kinds of strategic choices considered by decision makers by affecting how the strategic environment is perceived, and thus how they respond to it. Unlike the second-generation literature, there is no sense in the Chinese literature of a radical disjuncture between strategic culture and strategic choice where the former may play a mythical role, obscuring, cloaking, or justifying the latter. Nor is there any appreciation that there might exist multiple strategic cultures, each with differing effects on behavior. Indeed, there is a great deal of unanimity about what precisely the characteristics of Chinese strategic culture are and how these are manifested in strategic behavior. These characteristics include, but are not limited to, the following: (1) a theoretical and practical preference for strategic defense—earthworks, walls, garrisons, static positional defense, accompanied by diplomatic intrigue and alliance building rather than the invasion, subjugation, or extermination of the adversary; (2) a preference for limited war, or the restrained application of force for clearly enunciated political ends; and (3) an apparently low estimation of the efficacy of violence, as embodied for instance in Sun Zi's oft-cited phrase, "not fighting and subduing the enemy is the supreme level of skill."25 Moreover, there is also a generally accepted view that these characteristics have changed little from Sun Zi through Mao Zedong, though there is not much explanation of why the evolution of China's alleged strategic culture has been so exceptionally slow. The tendency in the literature is to focus almost exclusively on Sun Zi, compare him with Mao, and assume that an unbroken strategic-cultural chain links the two. One consequence of these interpretations of traditional strategic thought is a tendency in the literature to juxtapose Chinese and Western strategic cultures, and to conclude that the West stresses the application of technology, firepower, and offensive wars of annihilation while the Chinese have a preference for strategem, minimal violence, and defensive wars of maneuver or at25

These characteristics are examined more closely in chapters 3 and 4.

26

CHAPTER ONE

trition. Whether or not this is a fair contrast, it is a common one. Boorman and Boorman, for instance, argue that what they call a Napoleonic model of conflict posits "momentum equals mass times velocity, victory equals maximum momentum concentrated at a decisive point." Chinese strategy, on the other hand, stresses indirection and the manipulation of the enemy's perceptions of the structure of the conflict (Boorman and Boorman 1967: 152, Lin 1988: 3233, Wei and Liu 1985: 435-36, Li and Sun 1989: 24, Chen 1990: 24). Unfortunately, like much of the work on Western strategic cultures, the literature on China fails to offer a convincing understanding of the content and influence of any Chinese strategic culture. Briefly put, there is an almost monolithic acceptance of the view that the Chinese strategic tradition is uniquely antimilitarist, yet there is little explication of the process of analyzing this tradition. There is also a view that this strategic tradition has had a tight, deterministic effect on strategic choice from ancient times up to the present. Traditional scholarship allows no room for the possibility of either multiple strategic cultures or disjunctures between strategic culture and strategic choice. There is little or no effort to set up strategic culture-derived propositions about strategic choice and test these against alternative explanations. In large measure the problems in this literature stem from a heavy reliance on Sun Zi's Art of War as the textual basis of Chinese strategic culture. And the reading of Sun Zi focuses heavily on concepts like "not fighting and subduing the enemy," "the best policy is to 'attack' the enemy's strategy . . . the worst is to attack the enemy's cities," among other axioms. Since these notions are generally assumed to be the core of Sun Zi, hence central to Chinese strategic culture, it is no wonder that there is a pronounced stress on the alleged "disesteem of violence" in the literature on China (Wang 1976: 77, Wu 1989: 1726, Gao 1990: 1-8, Boylan 1982: 342-46, Fairbank 1974: 5,11). The fact that Mao Zedong borrowed from Sun Zi is seen as a confirmation of the persistence and consistency of Chinese strategic culture. There is very little literature on works on strategy other than Sun Zi, despite, as will be discussed in later chapters, the rich and influential tradition of strategic thought beyond Sun Zi's Art of War. And there is little discussion of the variations in interpretations of Sun Zi across the numerous annotations and explanations of the text from Han times to the present.26 When authors do look at other sources of Chinese strategic culture, the interpretation and analysis of these sources is heavily influenced by this standard view of Sun Zi (see Lin 1988). In other words, other texts end up confirming or buttressing the alleged preference for minimal violence embodied in Sun Zi's concept of "not fighting and subduing the enemy." In general, there is 26 For an indication of just how rich China's tradition of strategic thought is, see the lists of texts and annotations in Xu 1988 and Liu 1990d. The former contains 3,380 texts, the latter 4,221 from pre-Qin (pre-250 B.C.) to the end of the Qing dynasty (1911).

STRATEGIC CULTURE: A CRITIQUE

27

simply very little literature that analyzes Chinese strategic thought from a strategic perspective—namely, asking what this strategic thought tells a hypothetical strategist about what to do and how to make choices.27 There is, finally, no apparent recognition of the possibility that strategic culture and strategic choice are unrelated or disjointed, or that strategic culture plays an instrumental or ritualistic role in the strategic decision process. Fairbank wisely added the caveat in his summary of China's strategic traditions that there might be a difference between the normative ideal of minimal violence and actual strategic practice. But there is generally a yawning gap between strategic culture-related literature and historical research on China's use of force. While work on Chinese strategic thought persists in the notion that this thought stressed defensiveness, limited war, the disesteem of violence, and so on, the historical behavior of Chinese rulers and military figures presents a far more mixed picture. Take, for instance, the frequency of wars of extermination (mie $*£) in the history of the Zhou dynasty, recorded in the Spring and Autumn Annals; Han Wu Di's campaigns in the former Han period; expansionism into Turkestan and Tibet during the Tang dynasty; the Ming Yong Le Emperor's annexation of Annam (1407-1427), among other examples of strategic offensiveness. In a massive compilation of internal and external historical wars from the Western Zhou (ca. 1100 B.C.) to the end of the Qing dynasty (1911) scholars at the Chinese Academy of Military Sciences uncovered a total of 3,790 recorded wars. In the Ming alone there was an average of 1.12 external wars per year through the entire 270 plus years of the dynasty (Chinese Military History Group 1986). There has been little systematic research tracing the alleged constraining effects of Chinese strategic culture on the decision processes of various dynastic rulers and military figures. How does one explain the frequency of violence in Chinese historical strategic behavior while at the same time maintaining that traditional Chinese strategic thought is uniquely antimilitaristic?

CONCLUSION

Despite problems in the literature in both the international-relations and China fields, it seems worthwhile to pursue the question of strategic culture further. First, the literature on Chinese strategic thought and practice is dominated by a strategic culture-like approach, though the term is rarely used. There are a number of assumptions and conclusions about the nature and consistency of 27 The handful of English-language monographs that do look at formative strategic texts are generally sinological or philosophical in orientation. Rand 1977 for instance, stops his analysis at the Han dynasty. Lewis 1990 stops in the Warring States period. Balmforth 1979 focuses solely on the strategy of Sun Bin (fourth century B.C.).

28

CHAPTER ONE

Chinese strategic culture that have gone unexamined both theoretically and empirically. In this study I hope to help demystify Chinese strategic concepts and to develop and test a model of Chinese strategic culture that will fill in the large gaps in Western analyses of Chinese strategic thought. Second, strategic culture has become an object of increased attention in the international-relations and international-security literature of late, yet the notion is so underdeveloped conceptually and methodologically that its value is, in some senses, unassessable. I will try to assess and develop more rigorous and testable methods for analyzing strategic culture. This will, hopefully, provide a basis for more exacting analyses of strategic culture within states and hence more genuine analysis across states. These procedures could then be used to reexamine American, Soviet/Russian, Middle Eastern, and Asian strategic cultures. Third, strategic culture questions ahistorical and acultural explanations of strategic choice by rooting strategic preferences deep in history and culture, and not predominantly in system structure, or the distribution of state capabilities, as the dominant structural-realist paradigm in the international-relations literature does. Further study of strategic culture takes up the call for further analysis of national styles in strategy. The study of strategic culture also taps into a growing interest among some critics of structural realism in so-called second image, or unit-level and ideational sources of state behavior.28 Strategic culture fundamentally problematizes the concept of a universal, ahistorical unitary state-as-rational-actor inherent in neorealist approaches to strategic analysis, and suggests that both conflict and cooperation in international politics are rooted in historically constructed and socially learned assumptions about the strategic environment and appropriate responses to it. This possibility is at the core of some nasty debates between neorealists, claiming that structural anarchy imposes fundamental sameness on the strategic behavior of actors regardless of historical period, and constructivists who contend that there have been historical state systems in which cooperative norms have neutralized or replaced anarchical structures.29 But this is no mere esoteric academic controversy: it touches on a range of critical policy debates in the West over the degree to which post-Cold War 28 Strictly speaking, ideational sources of strategic choice need not be labeled unit-level variables. As work on the diffusion of international norms (e.g., democracy, sovereignty) suggests, units or states may be socialized by "global ideational processes" whose roots lie outside of the state. See McNeely 1992, Barnett 1993, Wendt and Barnett 1992, Soysal 1993. 29 See in particular the exchanges between Hall and Kratochwil (1993), and Fischer (1993) over the degree to which medieval interstate relations were characterized by an anarchic structure and realpolitik behavior or nonanarchical, normative-based functionally cooperative behavior. As my argument about the China case will suggest, both sides may be missing a third possibility: realpolitik behavior could be normatively or ideationally rooted, independent of structure, even though appearing under anarchical conditions.

STRATEGIC CULTURE: A CRITIQUE

29

international relations will be characterized by old clashes of states operating in an immutable anarchical environment (Mearsheimer 1990), by new clashes of cultures where states are the structural tools of ideational civilizations (Huntington 1993), or by purposive efforts by state and nonstate actors to foster cooperative security norms in spite of the absence of supranational governing institutions (Adler and Crawford 1991). Arguably China is a good place to start for the study of strategic culture. Given certain historical and cultural continuities, and the apparently crucial role of precedent in intellectual and policy life, if there is anywhere one should expect to find a single strategic culture, and to see strategic culture-like influences on behavior, it ought to be in the Sino-Confucian system. In this sense, China may be one of the "most likely" cases in which hypotheses about the existence and effect of strategic culture are true. Thus, if strategic culture were found not to exist, or not to have an effect on strategic choice, this would constitute a fairly powerful indication that strategic culture as an analytic concept is a dead end.30 Positive findings, however, would be less generalizable since as the "most likely" case China therefore would be more likely also to be a deviant case. But positive findings about both the existence and influence of strategic culture in the Chinese case would, nonetheless, constitute evidence that further cross-national studies of strategic culture may reveal the existence of other, different strategic cultures. I chose Ming China (1368-1644) as the empirical focus of the study primarily for methodological reasons. To test for "Chinese" strategic culture, it makes sense to look for a period of history in which decision makers are self-conscious heirs of the philosophical and textual traditions and experiential legacies out of which this strategic culture may come. It also makes sense to look at a period in which the decision elites are insulated—to the extent possible in nonexperimental research designs—from the possible effects of non-Chinese or Western strategic cultures, if indeed these exist.311 also sought a period in Chinese history for which the documentation on decision making was rela30

On "most likely" and "least likely" case selection for low-N research designs see Eckstein 1975: 119-20. 31 I am aware of the implicit and risky equation this decision rule draws between ethnicity and strategic culture. Strategically, peaceful and warlike interaction between Han and nonHan led to considerable diffusion in both directions of technology, personnel, and in all probability strategic ideas. All this I acknowledge. However, this study cannot be a pure experiment where groups are randomly assigned to different treatments. To the extent that a Chinese strategic culture may exist, it is more likely to occur in a purer form in a dynastic period where ethnic Hans, drawing explicitly on early Han traditions of statecraft, rule rather than in, say, the Mongol Yuan dynasty or the Manchu Qing dynasty (itself influenced from the mid nineteenth century on by contacts with Western imperialism). Second, to the extent that there is consistency between early formative strategic theory on the one hand and Ming strategic theory on the other, one can be somewhat more confident that the latter has not been completely contaminated by, say, interaction with Mongol strategic culture.

30

CHAPTER ONE

tively, and I stress relatively, rich. The intersection of these three sets of conditions was the Ming dynasty. The first step, then, in the study of strategic culture is to learn from past mistakes, to construct a more rigorous concept of strategic culture that specifies what the scope and content of strategic culture is and what it is not, the objects of analysis, the historical periods from which these are drawn, and the methods for deriving strategic culture from these objects. Then it is necessary to explicate a research strategy that can credibly measure the effects of strategic culture on the process of making strategic choices. This is the focus of chapter 2. Chapter 3 then goes on to apply these procedures to the analysis of traditional Chinese strategic thought as embodied in the Seven Military Classics (Wu Jing Qi Shu tf£!M~bHf). This chapter attempts to uncover the key assumptions of the central paradigm in traditional Chinese strategic thought by focusing on answers to questions about the role of war in human affairs, the nature of conflict with the enemy, and the efficacy of violence in dealing with the adversary. I argue that the answers to these questions reveal that Chinese strategic thought shares many of the same assumptions as parabellum or hard realpolitik worldviews found in some variants of Western realism. That is, warfare is viewed as a relatively constant element in human interaction, stakes in conflicts with the adversary are viewed in zero-sum terms, and pure violence is highly efficacious for dealing with threats that the enemy is predisposed to make. Chapter 4 looks for a set of grand strategic preference rankings that derive from the assumptions of the parabellum paradigm and that are consistent across the objects of analysis. It then asks whether there is enough consistency to constitute a single strategic culture. My findings suggest that, although the results are not especially strong, there is sufficient evidence to show that the Seven Military Classics do share a preference for offensive strategies over static defensive and accommodationist options. There is also a pervasive acceptance of absolute flexibility as a critical element in strategic choice. This flexibility occurs within the context of an overall preference for offensive violence. Thus it introduces into this preference ranking an explicit sensitivity to changes in relative capacities. The analysis in both chapters 3 and 4 also indicates the presence of a set of central assumptions and strategic preferences that are for the most part different from this parabellum strategic culture. However, this alternative strategic culture, which I call Confucian-Mencian, is idealized and divorced from the argumentation structures and decision axioms found in the classic texts. This, of course, raises questions about the reason for and role of the idealized or symbolic strategic culture. Chapter 5 returns to some theoretical literature in organizational studies, political science, and social psychology to provide some tentative explanations for this disjuncture between the Confucian-Mencian

STRATEGIC CULTURE: A CRITIQUE

31

and parabellum strategic cultures. This chapter also offers some hypotheses about the roles of each in strategic choice. Chapter 6 turns to Ming grand strategy towards the Mongols along the northern border to test for the presence of a parabellum strategic culture in Ming policy making. The objective is not to provide the "best" explanation of Ming strategy; it is instead to assess the importance of strategic culture in any such explanation. This chapter parallels chapter 3 in that it focuses on the answers that Ming strategists and policy makers provided to the three questions at the heart of a strategic culture's central paradigm. The analysis finds, in fact, a great deal of congruence between the answers found in the Seven Military Classics and those provided by Ming strategists. Chapter 7 then goes on to examine whether one finds in Ming strategic thinking a parabellum grand strategic-preference ranking similar to that in the ancient texts on strategy. The analysis suggests that there is indeed some preference for offensive over static defensive and accommodationist strategies, but that this preference is heavily mediated by the notion of absolute flexibility, as it was in the Seven Military Classics. This suggests that in general there is a tendency in policy for offensive behavior in periods when the Ming experienced a relatively large advantage in capabilities, but a shift in less coercive directions as this capacity diminished. This chapter then looks at both anecdotal and aggregate evidence to see if this proposition holds. While the evidence is not overwhelmingly strong, there is sufficient reason to argue that variations in Ming strategy generally follow the pattern predicted by a parabellum strategic culture. The chapter also examines the surprisingly minor role played by the Confucian-Mencian strategic discourse in Ming strategy. The final chapter concludes that while there is evidence that a Chinese strategic culture does exist and influences grand strategic choice, this strategic culture is not self-evidently unique, or different from certain strains in Western realpolitik thought and practice. I also there address some of the problems, caveats, and puzzles these findings raise for the further study of Chinese strategic theory, practices, and strategic behavior in international-relations theory. In particular, the Chinese case suggests that to the extent parabellum is crosscultural and learned we should begin to treat realpolitik behavior as a product of a "cultural realist" norm—a historically and temporally bounded, though impressively powerful, ideational source of state behavior.

Chapter Two SOME QUESTIONS OF METHODOLOGY

T

HE ANALYSIS OF ideationally based cross-national variation in strategic behavior has been fundamentally hampered by empirically untestable conceptualizations of strategic culture. One major reason for this is the inattention paid in much of the literature to basic questions of methodology: how should we conceive of strategic culture such that we can isolate its content and boundaries and derive and test hypotheses about its effect on behavior? In this chapter I will clarify how to analyze Chinese strategic culture and offer explanations for the various methodological choices I have made. After surveying different ideas on the nature of culture, and analogues in the realm of strategy, I will develop a notion of strategic culture that is empirically testable. The basis of this conceptualization is congruence in strategic preference rankings across key objects of analysis or texts in a state's strategic tradition. I then go on to apply this definition of strategic culture to the Chinese case and to develop a research design that, in principle, should be able to isolate the effects of strategic culture—should it exist—on choice and behavior. In this process, I stress two fundamental principles in any research on strategic culture. Both address the crucial problem of falsifiability that dogs existing research. First, one has to determine whether or not strategic culture exists across time and across actors within a society in such a way that it may constitute a dominant variable in decision making. And second, one has to ask whether or not strategic culture influences the behavior of decision makers. Both tests have to be passed before we can assess the analytic value of strategic culture as a separate variable in any explanation of strategic choice. If thefirsttest fails but the second succeeds—that is, if there are multiple and varied strategic preferences—then all we are left with are the idiosyncratic preferences of individual decision makers. This is not the same thing as a collectively shared strategic culture. If the first test succeeds but the second fails, then strategic culture can be relegated to a residual variable, and we can refocus on the other standard variables in international-relations and comparative foreign-policy theory. The analysis of strategic culture is not an easy task. For one thing, as the previous chapter indicated, there are few consistent methodological maps provided by the existing literature. How do we distinguish strategic culture as an independent variable? Where historically are we supposed to look to find the roots of strategic culture? What are its signatures, artifacts, or footprints? How

SOME QUESTIONS OF METHODOLOGY

33

do we know it when we see it? Those who have used the term, or a concept like it, have mostly left the contents and boundaries of strategic culture undefined and unexplored. The analysis of strategic culture is also difficult because it enters the realm of cultural analysis. Here one confronts a conceptual and methodological morass where definitions of culture are as numerous as the researchers of culture, where methods range from quantitative content analysis to the first-hand observation of small subcommunities to the examination of physical artifacts to intuitive, avowedly antimethod postmodern narrative accounts. All of this suggests, however, that at this stage in our understanding of the ideational sources of strategic choice we cannot help but be somewhat arbitrary, though explicit, when trying rigorously to define and test a notion of strategic culture.

DEFINITIONS OF STRATEGIC CULTURE

There is no shortage of definitions of culture. Glenn et al. define culture as "the total knowledge existing within a society, concerned in such a way that each item of knowledge is multiplied by a) the proportion of individuals who hold it, andb) the 'leverage' exercised by each of these individuals" (Glenn et al. 1970: 41). This definition implies that subcultures exist within a society, but that a dominant interpretation of "total knowledge" emerges, depending on the proportion and power of those who share this interpretation. The authors note further that the strength of these subcultures has an effect on collective decisions such that "the greater the operative strength of an idea the likelier it is to become the basis of a decision" (ibid.). They do not suggest, however, what this knowledge concerns, nor which collective decisions it is more likely to affect, nor how these decisions are, in the end, affected. Wildavsky calls culture those "codes enabling individuals to make much out of little. Thus cultures may be conceived of as grand theories, paradigms if you will, programs if you prefer, from whose initial premises many consequences applicable to a wide variety of circumstances may be deduced" (Wildavsky 1985: 95). While Wildavsky is somewhat more specific about the form of culture—codes—he does not address their content and their role in decision making or in behavior. Keesing is more concrete about the content of culture and what it does for those who share it. He suggests cultures are systems of cognition that relate peoples and communities to their ecology or environment in an evolutionary, symbiotic relationship. As the environment evolves, so does culture, which in turn affects the evolution of the environment (Keesing 1974: 75-76, 91). In a similarly adaptive vein, Schein suggests culture is "a pattern of basic assumptions—invented, discovered or developed by a given group as it learns to cope with its problems of external adaptation and internal integration—that has

34

CHAPTER TWO

worked well enough in the past to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think and feel in relation to these problems" (Barnes 1986:4). Geertz adds that these cultural assumptions come in the form of control mechanisms, plans, recipes, rules and instructions, which together constitute a "system of inherited conceptions expressed in symbolic forms by means of which men communicate, perpetuate and develop their knowledge about and attitudes towards life" (Geertz 1973: 89). Other images used by various authors to capture the forms of culture include unconscious or hidden standard operating procedures, scripts (Smircich 1983: 347-51), and "easy behavior.. . routine, largely unexplained options followed by most people most of the time" (Barnes 1986: 15-16). Namenwirth and Weber refer to culture as a "design for living" that only has a "programmatic and conceptual existence" (Namenwirth and Weber 1987: 8). Despite the diversity in the terms used to describe forms of culture, there is a common thread—namely that they all describe a bounded, inductive system of assumptions. These are not necessarily internally consistent, nor rigorously formal knowledge structures. Rather, they are learned, satisficing cognitive mixtures that allow people to interpret "the game being played," and to figure out what others "know, believe and mean," thus ordering the environment in a ragged, but more or less identifiable fashion (Keesing 1974: 86-87). Crucially, these cultural assumptions can be instrumental as well. That is, dominant subcultures can impose them on other groups, manipulate them, or convince other subcultures that these dominant cultural forms are in fact their own—a type of false cultural consciousness. The instrumental use of cultural forms is designed to justify and preempt challenges to the status quo (Green 1988: 19). There remains, however, a frustrating level of vagueness in notions of culture about its relationship to behavioral choice—that is, about what it is that culture does in a behavioral sense. Some of the terminology (i.e., recipes, rules, instructions, standard operating procedures, etc.) implies that culture is a necessary but not sufficient determinant of behavior. There is no one-to-one correspondence between cultural forms and observable decisions. In Elkins and Simeon's words, "culture does not explain particular choices which individuals make. Its explanatory power is primarily restricted to setting the agenda" (Elkins and Simeon 1979:130-31, Smircich 1983: 346). Barnes agrees: "culture introduces biases in perception; it limits vision and the range of choice" (Barnes 1986: 19, Elder and Cobb 1983: 85). Yet this definition verges on relegating culture once again to a contextual variable, and forces us to look to other mediating variables to explain why particular choices are made. In general, there are no glaring contradictions between definitions of culture found in anthropology, sociology, and organization theory on the one hand and those found in political science on the other. The concept of political culture is seen as a subset of culture, for the most part, though one that obviously has a more specifically political content. Thus political culture is viewed as

SOME QUESTIONS OF METHODOLOGY

35

political codes, rules, recipes, standard operating procedures, and routines that impose a rough order on perceptions of the political environment. It is a "cognitive, affective and evaluational" orientation towards political objects such as the self and others as political actors and towards governmental authority. It sets basic patterns in groups' assumptions about their own political efficacy, about the legitimacy of authority, and about the legitimacy of certain political inputs and outputs (Dittmer 1977: 553-54, Barnes 1986: 39-40). According to Elkins and Simeon, political culture provides views on the "nature of the political game being played, on proper modes of conduct, and on goals and strategies." Specifically, political culture provides assumptions about the orderliness of the political universe, the nature of causality, principal goals in political life, the relative value of maximax versus minimax strategies, who belongs to the political community, what type of events, actions, and institutions are political, and the trustworthiness of other political actors (Elkins and Simeon 1979: 132).] Political culture can also be instrumental, used deliberately to eliminate alternative institutional, ideological, or behavioral options from the body politic. As for the relationship between political culture and behavioral choice, Elkins and Simeon contend that political culture is a "shorthand expression for a 'mind set' which has the effect of limiting attention to less than the full range of alternative behaviors, problems and solutions which are logically possible" (ibid., 128). Hence, as in more generic definitions of culture, political culturebased behavior is juxtaposed with rational choice in that for some arbitrary, often less-than-conscious reason certain preferences, choices, and outcomes are excluded from consideration. In sum, despite variations in the terminology and emphasis found in definitions of culture, there are nonetheless a number of more-or-less shared elements. Culture consists of shared decision rules, recipes, standard operating procedures, and decision routines that impose a degree of order on individual and group conceptions of their relationship to their environment, be it social, organizational, or political. Cultural patterns and behavioral patterns are not the same thing. Insofar as culture affects behavior, it does so by presenting limited options and by affecting how members of these cultures learn from interaction with the environment. Culture is therefore learned, evolutionary, and dynamic, though the speed of change is affected by culturally influenced learning rates, or by the weight of history. Multiple cultures can exist within one social entity (i.e., community, organization, state, etc.) but there is a dominant one that is interested in preserving the status quo. Hence culture can be an instrument of control, consciously cultivated and manipulated. Given these elements of culture, what might a useful definition of strategic culture look like? We need a notion of strategic culture that (a) is falsifiable or 1 For a somewhat different list that closely parallels the assumptions in political culture see Elder and Cobb's list of "latent dispositional structures" (1983: 44).

36

CHAPTER TWO

at least distinguishable from nonstrategic-culture variables, (b) captures what strategic culture is supposed to do, namely provide decision makers with a uniquely ordered set of strategic choices from which to derive predictions about behavior, (c) can be uncovered in strategic-cultural objects, and (d) the transference of which across time can be traced. As a first cut, for simplicity's sake it seems best to transfer key elements of culture to the field of strategy or the realm of war and peace. Given the conceptual poverty of the literature on strategic culture, such testing of a definition of the term cannot help but be somewhat arbitrary. I assume, then, that strategic culture, if it exists, is, like culture, an ideational milieu that limits behavioral choices. But, unlike most of the literature on culture, I also assume that from these limits one could derive specific predictions about strategic choice. I am partial, then, to using an initial definition of strategic culture that paraphrases Geertz's definition of religion as a cultural system (Geertz 1973: 90). Strategic culture is an integrated system of symbols (i.e., argumentation structures, languages, analogies, metaphors, etc.) that acts to establish pervasive and long-lasting grand strategic preferences by formulating concepts of the role and efficacy of military force in interstate political affairs, and by clothing these conceptions with such an aura of factuality that the strategic preferences seem uniquely realistic and efficacious.2 The problem is relating strategic culture qua culture as a system of symbols to behavioral choices. The first-generation strategic-culture literature is con2 The term "grand strategy," sometimes termed "national strategy," is defined in the United States as the coordination of all elements of national power (economic, political, and military) to accomplish "national goals," primarily security against external threats (Department of Defense 1987: 350-51). "Military strategy" is generally defined as the coordinated use of military force and violence to accomplish the military goals of national or grand strategy (ibid., 232). However, the lines between the two are often not so clear (Jones 1987: 54-55). Contemporary Chinese definitions of national strategy (guojia anquan zhanlue S ^ ^ ^ : K f f i § ) see it as the highest level of strategy, an expression of the basic characteristics of a society, limited and shaped by the general lines and policies of the ruling entity (Sun and Zhang 1989: 471-72). None of these definitions reveals much detail about what kinds of behaviors constitute grand strategic choices. Posen remedies this somewhat when he defines grand strategy as a

political-military means-ends chain, a state's theory on how it can best "cause" security for itself. . . . A grand strategy must identify likely threats to a state's security and it must devise political, economic, military and other remedies for those threats. Priorities must be established among both threat and remedies because given an anarchical international environment, the number of possible threats is great, and given the inescapable limits of a national economy, resources are scarce. Because resources are scarce, the most appropriate military means should be selected to achieve the political ends in view." (Posen 1984: 13) This definition is useful here since it seems likely that if strategic culture exists, it will appear at a level where definitions of the enemy, of the threat environment, and of fundamental, deeply rooted preferred ways of responding to this environment are salient. My only quarrel with Posen here is that I would not axiomatically subsume strategic ends within a concept of grand strategy means. For further discussion see chapter 4.

SOME QUESTIONS OF METHODOLOGY

37

fused on this point, mostly implying that the linkage should be quite direct: a single strategic culture is the critical determinant of a strategic preference. Sometimes, however, it suggests the linkage should be indirect. Strategic cultural assumptions establish the values and assumptions of decision makers through socialization, particularly through education and training as decisionmaking elites. The strategic culture-based values of these decision makers then provide a limited set of strategic options. A more useful argument would suggest that strategic culture-based values provide a limited set of strategic options, but then rank these as well. Suffice it to say, how strategic culture affects specific choices is an extremely complex problem. But before we can hope to conclude that it does affect this choice, at the very least we have to show that strategic culture first limits in some way the options considered. Hence we need to trace strategic culture from its sources through the socialization process to the values and assumptions held by particular key decision makers. This requires "operationalizing" strategic culture, or at least suggesting what its empirical referents are so as to trace them through these first two stages. Thus, let me suggest that strategic culture as a "system of symbols" is comprised of two parts: the first consists of basic assumptions about the orderliness of the strategic environment—that is, about the role of war in human affairs (i.e., whether it is aberrant or inevitable), about the nature of the adversary and the threat it poses (i.e., zero-sum or positive sum), and about the efficacy of the use of force (i.e., about the ability to control outcomes and eliminate threats and about the conditions under which the use of force is useful). Together these comprise the central paradigm of a strategic culture. Note the parallels with the core assumptions embodied in traditional definitions of political culture. The second part of strategic culture consists of assumptions at a more operational level about what strategic options are the most efficacious for dealing with the threat environment as defined by answers to these three sets of questions. These lower-level assumptions should flow logically from the central paradigm. It is at this second level that strategic culture begins to impact directly on behavioral choices.3 Thus the essential components or empirical referents 3 This parallels somewhat Taber's (1987: 3-4) notion of war learning. A state's preferred choice within a range of strategies is set by specific heuristics about the strategic environment at time t. These in turn are derived from a broader, more deeply rooted, less contingent collection of central heuristics (a paradigm) that outlines the nature of this environment and how force fits into dealing with it. See also Ott's notion of "basic assumptions" which, he argues, are the key building blocks of organizational culture (1989: 44, 55, 61). The above characterization of strategic culture comes admittedly close to Holsti's definition of belief systems (Holsti 1969) and George's notion of operational codes (George 1969). And, indeed, other authors see these concepts as essentially related (Lebow 1981: 193). Arguably, however, there are at least two important differences. First, strategic culture is collective: it refers to collectively held preferences. Operational-code and belief-system analysis, on the other hand, are more clearly

38

CHAPTER TWO

of a strategic culture will appear in the form of a limited, ranked set of grand strategic preferences over actions that are consistent across the objects of analysis and persistent across time.4 They are not, therefore, necessarily responsive to changes in noncultural variables such as technology, threat, or organization.5 I use ranked preferences instead of a simple menu of strategic options for a number of reasons. First, simply in terms of the numbers of, and variations in, the strategies considered by decision elites in different societies, it is likely that there will be a wide enough range within a single state that there will be considerable overlap across states, hence the difficulty in determining separate strategic cultures across states. By looking at preference rankings one, in effect, weighs these within-state preferences. If different societies have different strategic cultures, they ought to put different weights on these choices—that is, rank them differently. This assumption allows testing for consistency in strategic culture within societies and thus for differences between societies (Elkins and Simeon 1979: 133). This approach also provides a concept of strategic culture that is falsifiable. If preference rankings are not consistent across objects of analysis then at that particular time a single strategic culture cannot be said to exist.6 Conversely, a strategic culture exists and persists if preference rankings are consistent across objects of analysis from deeply historical, formative periods up to the period under examination. Second, I use ranked strategic preferences because these should yield more explicit predictions about behavorial choice than a limited, unranked range of preferences. The latter would imply a definition of strategic culture that has a "permissive" influence (Terhune 1970: 18) on choice rather than a determining reductionist approaches, and generally focus on the individual's psychological characteristics. Second, and relatedly, the units of analysis differ: strategic-cultural analysis focuses on collectively produced and shared artifacts rather than on an individual's belief system. 4 Here I adapt Wildavsky's "cultural theory of preference formation" (1987). See also Powell 1994 on the importance of distinguishing between preferences over outcomes (ends) and preferences over actions (means). 5 Admittedly, this last is a tricky assumption. Should not different levels of technology and threat cue different repertoires of cultural reactions? There are a number of possible responses to this question. First I am looking at a level of strategic choice—grand strategy—that while not totally unresponsive to the cumulative impact of technological change should, in principle, not be as responsive as the operational levels of strategy and tactics, since it concerns broader, more enduring philosophical notions about when and where to use force. Second, if it is to have analytic value, or at least if one is to test its analytic power, strategic culture must be assumed to have powerful influence on behavior, be slow to change, and be able to override or substantially alter the influence of exogenous variables that, if acting by themselves in a strategic culture-less situation, have different effects on decision making. (On the stability of cultural traits, see Boyd and Richerson 1985.) Finally, for the sake of methodological simplicity it seems reasonable to me to begin by assuming great stability in the content of a strategic culture. This assumption can be modified if the evidence warrants. 6 Individual objects or texts, of course, may well embody ranked preferences. But if there is no congruence across texts then we cannot talk about shared preferences or a strategic culture.

SOME QUESTIONS OF METHODOLOGY

39

(though probabilistic) effect. If we instead begin by assuming that strategic culture has a more positive, active effect on choice, it makes distinguishing a strategic-culture model of choice from other models easier. Again, it seems sensible to begin by giving strategic culture the benefit of the doubt by expecting a considerable amount of explanatory power and modifying this proposition only in the face of empirical evidence. Third, while preference rankings and culture are not synonymous, they should not be considered completely separate either. Strategic culture as a system of symbols embodies assumptions about what the security problematique is, and therefore about how to best deal with it. In principle, then, strategic culture ought to embody preferred methods for responding to the threat environment. Moreover, preferences established in the past become part of the symbol system that is learned and reproduced, thus becoming part of the strategic-culture milieu at time t.1 This definition meets my criteria for assessing the analytic value of strategic culture: it is falsifiable; it can provide empirical predictions about strategic choice that, depending on the content of a particular strategic culture, can be tested against other models of choice; it has, in principle, empirical referents (e.g., languages and preferences) that can be observed in strategic-culture objects (e.g., texts, documents, doctrines, etc.); and its evolution (even dissolution) over time can be traced, since one can observe whether successive generations of decision makers are socialized in and share the basic precepts of the strategic culture as embodied in key strategic-culture objects.

OBJECTS OF ANALYSIS

As noted in chapter 1, one of the major flaws in the study of strategic culture is the lack of specification about what exactly should be analyzed. What are the artifacts or objects of analysis of a historically embedded, ranked set of grand strategic preferences? In principle the variety of objects of analysis could be formidable. They could include the writings, debates, thoughts, and words of strategists, military leaders, and "national security elites," however defined, or weapons designs and deployments, war plans, images of war and peace portrayed in various media, military ceremonies, even war literature.8 The amount of strategic-cultural artifacts could be overwhelming even when looking at a brief historical period. One way of getting around this problem is to analyze the content of some 7 1 am grateful to Audie Klotz for provoking my thinking on this point at the S SRC/Mac Arthur Fellows Conference, Budapest, May 1990. 8 For a comprehensive listing see Luckham 1984, tables 2 and 3. Kier uses curricula from military schools, training manuals, journals, languages, symbols, taboos, etc. (Kier 1992). Legro 1995 examines planning documents, regulations, military exercises, and memoirs.

40

CHAPTER TWO

most recent texts and some from the distant past, and assume that if there is congruence in preference rankings that strategic culture exists and has persisted up to the present. The longer the time across which this congruence stretches, the mor6 powerful and persistent the strategic culture. Many of those who write from a strategic-culture perspective, especially on China, take this route. But as McCauley usefully warns, "what we cannot assume from the existence of two similar sets of beliefs at different periods of time is that they enjoy an unbroken existence. The 'same' beliefs can sprout from different roots, at different periods" (McCauley 1984: 24). It is important, therefore, that the content analysis of strategic-cultural objects begin at the earliest accessible point in history, where strategic-cultural preference rankings may reasonably be expected to have emerged, or where those who use these strategic traditions imply the mots of their thought lie. From this point one moves systematically forward. Only this way can one determine whether later strategic culture is a direct descendant of a formative strategic culture, a return to earlier patterns, a break from more recent ones, a reflection of a particular subculture, or simply nonexistent. In essence, much of the work on strategic culture is mechanically deterministic because it asks, "Here is a set of strategic beliefs; where do they come from?" The researcher then moves back in time to a point where she or he finds similar beliefs. The alternative proposed here is to ask, "Here are some historical strategic beliefs; where do they go?" This approach, however, almost guarantees the researcher will be overwhelmed with potential objects of analysis or strategic-cultural texts. Fortunately, in the Chinese case one need not be too arbitrary about choosing which objects or texts to analyze: there is an obvious group of appropriate texts called the Seven Military Classics, comprised of the Sun Zi Bing Fa (M^F^ii;) the Wu Zi Bing Fa O ^ ^ p ^ r S ) , the Si Ma Fa (W] ^ & ) , t h e Wei Liao Zi (WiW?)> the Tai Gong Liu Tao Cfc&/\ fg), the Huang Shi Gong San hue (]gf 54>HB&), and the Tang Tai Zong Li Wei Gong Wen Dui (jM^^^M&MWWhile the dating and authorship of most of these texts is still far from settled, the individual texts appeared over a period of about fifteen hundred years of varied intellectual, political, social, and strategic contexts, beginning with Sun Zi Bing Fa in the fifth century B.C. and ending with the Tang Tai Zong Li Wei Gong Wen Dui {Wen Dui for short) written sometime in the tenth century A.D. 1. Sun Zi Bing Fa: This is the earliest extant military-strategic work and supposedly reflects the thinking of Sun Wu, a native of the state of Qi, born in the sixth century B.C. during the Spring and Autumn period. The dating and authorship of the text are both open to debate. The most recent archaeological and textual research suggests the text appeared around 500 B.C. (Yates 1988: 216-19). 9 One of China's top au9 Sawyer 1993: 149-50 cites evidence for three different datings, but in the end leans towards Yates's dating.

SOME QUESTIONS OF METHODOLOGY

41

10

thorities on ancient military texts gives roughly the same period. He argues that the extant text is probably a revision of an original text by Sun Zi's disciples (Xu 1990: 97, see Wang and Xu 1985: 122). This text is, of course, the one classical Chinese military work known in the West as the epitome of traditional Chinese strategic thought, and for good reason. In Chinese scholarship the Sun Zi Bing Fa has been, in general, considered the most profound, comprehensive, transcendent, and direct expression of the general principles of warfare ever to have appeared. In its various versions, the text comprises 6.6% of all known military texts (bingshu jiflf) up to the end of the Ming dynasty.11 It has been variously called a "military classic," a "sacred" work on military affairs, and the "progenitor" of military studies (Xie 1990: 33, Wang and Xu 1985: 114). Its preeminence among the military texts was formalized in the earliest surviving annotation of the text by the famous general of the Three Kingdoms period (A.D. 220-265), Cao Cao (Li 1990). In the Wen Dui the Tang emperor Tai Zong is quoted as saying, "I have read all the works on military affairs. None surpasses Sun Zi" (Liu 1955a: 97). The text's thirteen chapters cover topics such as the general role of force in warfare to strategy, logistics, and intelligence.12 2. Wu Zi Bing Fa: This text is attributed to Wu Qi, born around 430 B.C., a student of one of Confucius's disciples, and a practitioner and theorist of war. He has been identified variously as a Confucian realist, a Legalist, and a member of a separate class of strategists (bingjia J=£liQ.13 The text included in the Seven Military Classics was only a portion of a more complete text. While less well known in the West than Sun Zi, and less frequently reproduced in Chinese history, the historical status of the Wu Zi Bing Fa is nonetheless quite high. The Wei Liao Zi, for instance, cites Wu Zi more frequently than Sun Zi in its discussion of strategy and of the use of force for political ends. Han Fei Zi, a philosopher of the late Warring States period, noted that at his time (mid third century B.c.) discussions on strategy were quite popular among scholars, and many possessed copies of both Sun Zi and Wu Zi (Xu 1990: 101, Sawyer 1993: 191-92). Wu Zi, along with Sun Zi, was studied by military 10

The Spring and Autumn period in Chinese history lasted from 770 to 476 B.C.. The period saw the gradual atrophy of the feudal system set up under the rulership of the Zhou dynasty (eleventh century-221 B.C.). By the Warring States period dynastic power had all but collapsed as the remnant feudal states fell into a period of continuous warfare where more powerful states swallowed up weaker ones. 11 Calculated from Liu 1990d. 12 In 1972 archaeologists unearthed the earliest extant version of Sun Zi, written on bamboo strips and buried in a Western Han tomb. In addition to the central thirteen chapters, there were portions of an additional sixty-nine chapters. These latter chapters, however, had evidently been lost since the Western Han, and were never included as part of subsequent versions of the text. 13 Legalism was a school of thought originating in the Warring States period that stressed the role of a firm and clear system of rewards and punishments (laws) to govern all aspects of political, economic, military, and social life. Legalists tended to stress the importance of economic military power as the basis of state security, in contrast to the Confucian emphasis on correct and moral government.

42

CHAPTER TWO

officials of the Eastern Han dynasty (25-220) (Li 1978: 265, Wang and Xu 1985: 5). While Cao Cao annotated only the Sun Zi text, he evidently accorded a great deal of value to Wu Zi as well, according to his military contemporaries Wang Shen and Zhuge Liang (Li and Wang 1986: 7). Altogther various versions of the Wu Zi BingFa comprise close to four percent of the total military texts produced up to the end of the Ming dynasty. The texts provide extensive discussions on gauging the nature of the enemy, on command and training, on morale, and on how to respond to various strategic contingencies. 3. Si Ma Fa: This text is allegedly written by Si Ma Rangju, a military official of the Qi state. Sources date him variously from the early sixth century to the late third century B.C.. While the text may therefore have appeared after Sun Zi and Wu Zi, much of its description of techniques of warfare appears to predate the former works, and provides useful information on chivalric warfare in the Spring and Autumn period. Indeed, one author speculates that the extant Si Ma Fa may have been a fusion of an earlier Si Ma Fa and Si MaRangju's commentaries (Lan 1987: 193—94, Sawyer 1993: 111). Whatever its origins, the historical evaluation of the text has been consistently high. The Han historian Sima Qian praised it as "vast and profound." The early histories of China, the Shi Ji Gfejfg) and the Han Shu (MIS), used the text as a guide to Western Zhou military structures. The text was said to have had a strong influence on the military thought and practice of Han Xin, the military strategist for Liu Bang, the founder of the Han dynasty (206 B.C.-A.D. 220) (Liu 1955f: 82). It is claimed as well that the powerful Han emperor Han Wu Di chose his military leaders on the basis of their knowledge of the Si Ma Fa (Xu 1990: 107, Lan 1987:192-93). The text is also cited favorably on topics from statecraft to just war in such key historical works as the Shi Ji, the Lu Shi Chun Qiu ( S J S # f t t and Cao Cao's annotation of Sun Zi (Tian 1990: 20-21). The Si Ma Fa provided one of the earliest discussions of the justification of war as an instrument of policy, as well as commentaries on statecraft and internal governance, military command, maneuver, and the conduct of soldiers during warfare. The text is sometimes identified as a Confucian military work since there is a fair amount of commentary on virtuous and moral government as a basis of security, and on righteous war (Li 1991).14 4. Wei Liao Zi: Like the other texts, controversy surrounds the dating and authorship of the Wei Liao Zi. There are, in fact, historical records for two Wei Liao's— one a Legalist who lived in the mid to late fourth century B.C. and served as an official for king Liang Hui, and the other a general for the state of Qin in the last half of the third century B.C. (Yates 1988: 225-32). One view is that the text is possibly a fusion of two texts from both periods, or is a continuation of a text from the earlier period that was finished in the late Warring States period (Xu 1989: 24-25). Regardless of its dating, however, the text reflects a strong, though not exclusively, Legalist influence. Its early chapters spend a considerable amount of time on the 14

See chapter 3 for a discussion of Confucian views on war and peace.

SOME QUESTIONS OF METHODOLOGY

43

economic and military basis of state security (Xu 1990: 116). Many of its later chapters are devoted to the problems of creating and controlling an effective military capability—a reflection of the increasingly complex and "professionalized" nature of warfare in the late Warring States period. One scholar, however, argues that the text's stress on civility and virtue (wen de jfcll), an( * o n moral rectification within the government of a state (nei xiu p3 0 ) as a basis of security, makes the Wei Liao Zi a successor to Wu Zi's Confucian realism (Zeng 1972: 81). 5. Tai Gong Liu Tao: The Liu Tao is surrounded with an aura of mystery. It is structured as a dialogue between the Zhou kings Wen and Wu (circa twelfth century B.C.) and their chief strategist Jiang Tai Gong, but legend has it that the text was passed down to the Han strategist Zhang Liang by a mysterious old man whom he chanced upon on a bridge. Modern scholars give the text at least four datings, from the Spring and Autumn period to the Warring States period to the Qin dynasty (221— 207 B.C.) to the Qin-Han transition. The dominant position leans towards the late Warring States period, in part because the book clearly borrows concepts from Confucian, Daoist, and Legalist schools of strategic thought. These did not contend for prominence until the Warring States period. The military and political terminology used (such as references to cavalry) seems to accord with this dating, though no one knows who actually wrote the text (Kong 1987: 6-7, Xu 1990: 118, Li 1990). The Liu Tao is recognized as the purest expression of the strategem school (quan mou fl fHI) of military texts.15 In addition to standard discussions of period weaponry, training, tactics, and military geography, the text devotes a fair amount of space to politicomilitary strategems aimed at undermining the political and military cohesion of the enemy, similar to some of the strategems in Sun Zi. The Liu Tao evidently had an important influence on military thinking in the Han, Three Kingdoms, and Tang periods (Kong 1987: 30, Wu 1988a: 59-60). 6. Huang Shi Gong San Lue: The dating and authorship of the San Lue are similarly obscure. Legend has it that it was also written by Jiang Tai Gong, yet the text does not appear as a separate titled work until the Western Han (207 B.C.-A.D. 24). Chinese scholarship currently holds that the work is a theoretical synthesis of the military thought and practice of Zhang Liang, one of the military leaders who overthrew the Qin dynasty in 207 B.C., but that it was disseminated under the name of Huang Shi Gong (a.k.a. Jiang Tai Gong) (Xu 1987: 3, Xu 1990: 125). This would place the text later in the Western Han period. One scholar suggests, however, that the text may have emerged around the same time as the Liu Tao and may have been written by members of the "strategems school," based on the former's frequent citation of a lost text, the Jun Chan (51 HI). This title is very close to the name of one of the works attributed to Jiang Tai Gong, the Mi Chan ( f $ H ) (Li 1990). 15 A late Qing introduction to the military thought of Jiang Tai Gong noted, "Tai Gong's strategems and plans were varied. Thus later generations said, 'In military affairs [the exploitation of] opportunity, extraordinary strategems and plans all descended from Tai Gong'" (Wang 1955: 17). The Tang emperor Gao Zu Tai Zong ordered "martial temples" built to honor Tai Gong—the military equivalent of Confucian temples honoring Confucius (Wei 1987: 18—19).

44

CHAPTER TWO

The text provides a fairly elevated and abstract discussion of techniques of political control and military victory. It stresses, among other notions, righteous and magnanimous rule, the employment of capable officials, and self-abnegation and selfrestraint by leaders, among other internal bases of external security. It also comments on intelligence, geography, and military organization. Like the other texts in the Seven Military Classics, the San hue was considered an important work on political and military strategy in its own right, and was included along with the Wei Liao Zi in the Tang-dynasty compilation on techniques of rulership and statecraft, the Zhong Shu Zhi Yao ^.WfeW: (Xu 1987: 26). Song dynasty (960-1279) scholars gave the text a very positive evaluation. One concluded that of the seven major military texts the San hue most closely approached a true understanding of the Way; it had great practical value in leading rulers to meritorious achievements, and in providing for the security of the state (citedin Xu 1985a: 57). 7. Tang Tai Zong Li Wei Gong Wen Dui: This text is constructed as a dialogue between the Tang emperor Tai Zong (599-694) and his chief military advisor Li Jing (571-649). Again, controversy surrounds the text's authorship and dating. Some historical commentaries claim the text is a Song forgery (Boodberg 1930: iii-ix), while others claim it indeed records Li Jing's exchanges with Tang Tai Zong on questions of strategy and tactics. The best that can be said is that the text probably reflects Li Jing's military thought. The work itself may have appeared after the Tang, before the end of the Five Dynasties period (906-960), or even as late as early Northern Song, since the Wen Dui is not mentioned in the Song war encyclopedia, the Wu Jing Zong Yao (5£ $§&§[§?), which appeared in the mid eleventh century (Lan 1987: 214-16, Chinese Military History Group 1988a: 204, Xu 1990: 131— 32). Li Jing's position in the pantheon of Chinese strategists and generals is quite high. His biography in the Jiu Tang Shu ( K B iir) notes that of all strategists only Li Jing was really able to understand Sun Zi and Wu Zi's military thinking (Xu 1990: 131). The Wen Dui was the only post-Han text included in the Seven Military Classics, an indication in the view of an American student of the text that the "Lz Wei Kung Wen Tui has remained to the present day the last monument of Chinese original thought on the problem of war" (Boodberg 1930: xxii). The text, like the others in the set, covers a wide variety of topics ranging from military command, training, tactics, battle formation, border defense, and deterrence. It also provides one of the more extensive and sophisticated discussions of military dialectics of any ancient Chinese military work. While obviously each of these texts tends to stress different topics, together they constitute as extensive a discussion of statecraft, grand strategy, operational strategy, tactics, and military organization as any other single work among China's rich store of military texts. Indeed, different versions of these seven texts comprise 15.5% of the recorded military titles to have appeared in ancient China up to the end of the Ming dynasty. 16 16

Calculated from Liu 1990d.

SOME QUESTIONS OF METHODOLOGY

45

Besides the proportion of all military texts comprised by the Seven Military Classics, there are other reasons for focusing on these texts. First, they mix elements from Confucian-Mencian, Legalist, and Daoist traditions in Chinese statecraft.17 Among Confucian-Mencian elements, for example, there is a fairly consistent notion of righteous war across texts, and general agreement on the critical importance of upright and virtuous officials in civilian and military administration. 18 As for Legalist elements, the texts generally also stress the importance of economic power and military force in the attainment and consolidation of state security. Finally, a number of texts, including socalled realist texts such as the Sun Zi Bing Fa, incorporate Daoist imagery in comments on the minimalist use of force for politico-military ends (Xu 1985a: 56-57, Wu 1988a: 60, Liu 1989: 3, Hatori 1989: 3). Given that these texts, then, do not seem to lie far outside the major Chinese philosophical traditions, any evidence of strategic culture in them ought not to be divorced from broader cultural and intellectual traditions. In other words, by looking at these military classics I doubt I will miss elements of alternative strategic cultures found in nonmilitary texts.19 These are not narrowly military texts that focus solely on what to do once war has broken out. If they were, we would naturally expect them to concentrate, for instance, on intrawar strategy and tactics. Rather, they are texts that deal with everything from questions of prewar statecraft to battle formations. Indeed, there is in some cases very little difference between the content of these military classics and other texts on statecraft that have been traditionally categorized as philosophical classics. 17

One American scholar (Rand 1977) has usefully disaggregated six different philosophical approaches to war and peace up to the Han. This is a valuable corrective to the tendency by some Chinese scholars to lump works on strategy into a separate "militarist" category, distinct from the Confucian, Daoist, and Legalist traditions (See Zeng 1972, Li 1978). The main problem with this classification, however, is methodological: Rand offers few criteria for his decisions to classify a particular text as representative of a particular subcategory, and seems to miss the synthesis found in many of the military texts. 18 Confucius (551-479 B.C.) was a philosopher whose teachings (and their interpretations) became the basis of the predominant orthodoxy in political and moral thought in China. Mencius (3907—305? B.C.) was one of the more influential interpreters of Confucius's thought. At the risk of oversimplifying a complex body of thought, Confucianism stressed that external security rested on a ruler's ability to provide for the material and moral needs of his people through virtuous personal conduct and enlightened policies. This way, the people of the realm would be content with their lot, and potential enemies would submit willingly in order to partake of the ruler's magnanimity. Moral education was sufficient to transform potential enemies (including non-Chinese peoples) into willing and submissive allies. Confucius did not oppose military preparations, though he downplayed their role in the security of the state. Mencius, in particular, pushed Confucian ideas in a more extreme direction, arguing that a virtuous ruler had no need to use military force because he would have no enemies. 19 1 have not ignored the nonmilitary classics on questions of war and peace. I have read and compared a number of key pre-Han texts that deal in places with issues of statecraft and strategy, including Confucian texts, texts associated with Mencius, Huai Nan Zi, Shang Yang, Lao Zi, Guan Zhong, Wen Zi, and the Lu Shi Chun Qiu, among others.

46

CHAPTER TWO

A second general reason for looking at these seven texts pertains to their compilation as a body of strategic work in 1083 as the Seven Military Classics. As early as 1040 the Song court began work on compiling strategic advice from ancient military texts. In the wake of an upsurge in violence along the northern borders in the first half of the century new attention was focused on the strategic requirements of border security. In 1040 the Song Renzong emperor ordered his officials to cull ancient texts on strategy—including the Sun Zi Bing Fa, the Wu Zi Bing Fa, the Liu Tao, the Si Ma Fa, and the Wei Liao Zi—for information on organization, strategy, and tactics that might be of practical value. The result was an encyclopedic work, the Wu Jing Zong Yao (5£$§ f§Ic), completed in 1044, that was clearly influenced by a number of military classics (Boodberg 1930: i-ii, Wang and Xu 1985: 114, Chinese Military History Group 1988a: 207, Xu 1990: 356-61).20 In 1072, in order to provide more institutionalized training for military officials, the Song set up the first military school to train candidates (wuju ren jSJIfi A ) for military examinations. The required study materials included the Sun Zi Bing Fa, the Wu Zi Bing Fa, and the Liu Tao. In 1080, in the further institutionalization of the ancient texts, the Song Shenzong emperor ordered the chief tutor of the Imperial Academy, one Zhu Fu, and the first recipient of the boshi (fifdb) degree in military affairs, He Qufei, to compile a text for military study comprised of the seven classics. These became the basic study materials of the Song military-examination system, and remained as such through the Ming and Qing dynasties (Brunnert and Hagelstrom 1978: 271, 313, Xu 1985: 90, Wang and Xu 1985: 117, 13637, Chinese Military History Group 1987: 337-38, Zhang 1988: 44). The central role of these texts in military education from the Song dynasty on has an important bearing on a critical problem raised, but not dealt with, in most of the literature on strategic culture—namely, the transmission of strategic culture from its formative period across time. It is clear that whatever consistent strategic precepts were embodied in the military classics would have been transmitted, at the very least, through the military-education system (as well as through the education of the emperor and his key advisers). During the Ming the reformalization of a military-examination system began in 1387 when the first emperor, Ming Taizu, ordered the creation of a separate military education system for officals at the provincial level, after the Song system had atrophied during the Yuan dynasty (1271-1368).21 Ming Taizu also ordered 20

Many of the historical examples of key strategic principles were drawn from Sun Zi. Even so, the formal implementation of this system seems also to have atrophied from time to time. In 1442 the court ordered once again that military officers should study the military classics more diligently. Training in military studies was to include—in addition to the standard nonmilitary classics—the Seven Military Classics, as well as works on how to gauge the enemy, and extraordinary methods for achieving victory (Cheng Jingzong in HMJSWBa 4: 7 7 81). A similar order was issued in 1465 (Yang 1960: 877-78). Still, the frequency of exams was inconsistent. During the Cheng Hua reign (1465-1488), the examinations were held sometimes every three years, sometimes every six. Often senior Ming government and military officials 21

SOME QUESTIONS OF METHODOLOGY

47

that all hereditary military officials and their children at and below the level of imperial princes, imperial sons-in-law, earls and provincial military governors were required to study the Seven Military Classics (Li 1983:415-21, Xu 1985c: 263-64, Xu 1985b: 92). One result of this edict was the compilation of perhaps the best known and one of the more authoritative annotated versions of the text by the Ming scholar Liu Yin (Wang and Xu 1985: 118). There existed as well during the Ming a question-and-answer version of the Seven Military Classics that enabled students to memorize succinct answers to questions about the content of the texts (Xu 1990: 58). It is clear, too, that those who studied the Seven Military Classics were not limited to professional military officials. Many high officials in the Ming, including various emperors, were keen students of the military classics. We know, for instance, that Yu Qian, the Minister of War from 1449 to 1457, wrote a text entitled/,/ DaiJing Wu Yao Lue [Summary of Historical Military Classics] (jfg iXK^M:^)Wang Shouren, alias Wang Yangming, an important Ming NeoConfucian philosopher and a key official in the Cheng De period (1506-1522), wrote six military texts and a commentary on the Seven Military Classics. Gao Gong and Zhang Juzheng, top officials in the court during the 1570s, both wrote books on military affairs based on their study of the classics.22 In addition to the role of the Seven Military Classics in the socialization of many military officials and top court officials, the work provided the textual and intellectual basis for much of the extensive writing on military affairs in the Ming period. The massive late-Ming war manual, the Wu Bei Zhi (3£$f ^C), for instance, frequently cited the military classics, as did the highly regarded early Wan Li period treatise, the Cao Lu Jing Lue (jpL3t,fgBI§).23 The famous mid-sixteenth-century general/official Qi Jiguang also cited the seven texts extensively in his works on military affairs. Indeed, in one text Qi proposed a reformed military-education system composed of four parts—the first being a study of military and historical texts including, among others, the Seven Military Classics (Zou 1987: 106-7). In fact, the writing of military texts seems to have been somewhat of a growth industry during the Ming dynasty, despite blamed the poor performance of the Ming military on military officers' poor grasp of the Seven Military Classics. See, for instance, Ma Wensheng's memorial on the need to study both the Seven Military Classics and the Song military text, the Wu Jing Zong Yao (Ma Wensheng in HMJSWBa 5:520-21, Zhu Jian in HMJSWBa 4:243-52, Li Shimian in HMJSWBa 3:648-51, cf. Ye Sheng in HMJSWBa 5:341). 22 Scholar/generals such as Xu Da, Zhu Jian, Ma Wensheng, Ye Sheng, Yang Yiqing, Wang Chonggu, among others were also keen students of the Seven Military Classics. See chapters 7 and 8 for a discussion of the strategic thinking of some of these officials. My thanks to General Ding Zhaoqiang (ret.) in Taiwan for providing a list of Ming officials known to have studied the military classics. See also the sources listed in Liu 1990d. 23 There are in the Cao Lu Jing Lue at least sixteen references to Sun Zi, five to Wu Zi, four to the Wei Liao Zi, three to the Liu Tao, two to the Wen Dui, one to the San Lue, and one to the Seven Military Classics as a body of work.

48

CHAPTER TWO

the intellectual prominence in the mid and late periods of the dynasty of allegedly antimilitary neo-Confucian intellectual trends. According to one recent source, 1,164 military texts appeared during the Ming dynasty, of which 217 (18.6%) were individual or collected versions of the texts in the Seven Military Classics (Liu 1990d). In sum, given the importance of the Seven Military Classics in the production of military thought in China, and in the socialization of scholars-officials and military officers, these texts can serve as valid objects of analysis for the rigorous study of Chinese strategic culture. There are, however, some potential problems in the selection of the Seven Military Classics. First, the fact that the majority of these texts originated during the Warring States period—a time of incessant conflict among more or less technologically equal, culturally similar, "sovereign" states—rather than later periods where a vast empire was involved in ongoing wars against non-Han "barbarians" raises the question of whether the strategic thinking in these classics may differ from works produced after China's unification in the third century B.C. I am not too concerned about this potential problem. For one thing, the Wen Dui consciously applies the precepts of Warring States texts to problems of imperial defense against northern barbarians. For another, as noted below, the Seven Military Classics's prominent role in military thinking through the end of the imperial era indicates that later thinkers continued to see the relevance of these texts despite transitions in political structure and security problematique. Indeed, even in the Warring States period texts preserved the notion (often fictional) of a political center, namely the Zhou court. The strategic advice often concerned how to ensure that those in "all directions" submitted to the authority of the center, a security problem not dissimilar to that formulated by strategists after China's unification. Finally, I have also been careful to examine key annotations of these texts—particularly the Song dynasty Shi Shi Qi Shu Jiang Yi (Bfg.R'bilf IS III) and the Ming dynasty Wu Jing Qi Shu Zhi Jie (J^IM'fc Hf til @?)—as a control of sorts on how, if at all, the lessons of the Warring States works were adapted to different strategic contexts. A second general critique might be that strategic texts are objects of analysis produced by individuals, and thus are not properly cultural products. Since culture is collectively shared and since individuals are often incompletely enculturated, individual texts on strategy may not capture the core elements of a dominant strategic culture.24 This problem may be of lesser concern here, since most of the works seem to be compilations of the teachings of a strategist with collective additions by various followers, students, transmitters, and annotators, not solely attributable to an individual. A final problem might be whether the content of the Seven Military Classics 24

This point is made in Namenwirth and Weber 1987: 13-14. See also Vertzberger's (1986: 231) caution that responses to historical and cultural stimuli may be "individual specific."

SOME QUESTIONS OF METHODOLOGY

49

in fact addresses grand strategic questions. Do these works explicitly or implicitly discuss the pros and cons of different types of grand strategies? Or are most largely concerned with military strategy, tactics, and training? The short answer is that in some cases the discussion focuses explicitly on issues related to grand strategy—the role of force in the pursuit of state aims, typologies of warfare, the preferred means for achieving security, among other topics.25 In some cases the texts introduce concepts that have grand strategic, strategic, and tactical implications. They remain at a level of abstraction that can incorporate a range of choices running along a vertical axis from tactics to grand strategy and a horizontal axis from nonviolent to violent behaviors (Wang 1976: 77, Chinese Military History Group 1988a: 43). Finally, in some places the military texts do deal only with choices below the grand strategic level. But here, too, these choices appear to derive from, or can be traced back to, separate, definable, though implied, grand strategic choices.

M E T H O D S OF ANALYSIS

How, then, should one go about analyzing the Seven Military Classics? How to extract the central elements of strategic cultures, if they indeed exist? My approach to content analysis is fairly eclectic. It seems sensible to assume that different methods might tap into different levels of meanings in the texts and act as cross-checks of sorts on the meanings uncovered by each method respectively. Multiple methods can triangulate the central meanings in the texts, ascertaining if they are consistent on all levels of meaning. In some instances one need not go any further than to pick out explicit statements on strategic preferences, but such references are fairly rare. Sometimes such statements could be subject to less obvious interpretations, and sometimes, as will be argued later, they are not always consistent with other causal arguments in the texts. Thus I also use two other forms of content analysis, namely a modified form of cognitive mapping and symbolic analysis. My reasoning is simple: I am interested in what the Seven Military Classics appear to be telling a strategist to do, how to rank choices, and thus how to make choices. As Dessler argues in a general discussion of foreign-policy choice, causal judgments are a key step as decision makers reason about how certain types of behavior will help secure basic foreign-policy goals. Hence, a critical unit of analysis is the "policy argument," or the "network of statements that a) defines policy goals and standards and b) recommends the adoption of a particular policy option or criticizes the recommended adoption of another, on the basis of projected eventtrends linked to the specific implementation of specific policy options" (Dessler 1986: 18-19, Goldstein andKeohane 1993: 13-14). 25 Wei Rulin and Liu Zhongping, for instance, call Sun Zi's third chapter a chapter on "national strategy" (Wei and Liu 1985: 92).

50

CHAPTER TWO

Cognitive mapping is one technique for uncovering these sorts of "policy arguments," or linkages between certain causal axioms and their estimated behavioral effects. This method seems particularly well suited for uncovering grand strategic axioms. "A cognitive map is designed to capture the structure of causal assertions of a person with respect to a particular policy domain, and generate the consequences that follow from this structure" (Axelrod 1976: 58). This is a crucial step in ranking the value of alternative strategic choices. Cognitive mapping enables the researcher to trace the relationship between different types of proposed strategic actions and results that are considered to have both positive and negative value. This technique, then, ought to help clarify strategic preferences when they are obscured by great amounts of text, or by false, perfunctory, causally irrelevant options. One is looking, in effect, for the "text's own logic," as opposed to the surface logics employed by its author(s).26 Comparing cognitive maps across texts can determine the degree of congruence in estimations of the efficacy of different strategic choices. The ancient Chinese military texts are, of course, often of unknown authorship, and some appear to have been incompletely transmitted over time.27 Mapping partial texts as though they were whole documents may create causeeffect relationships that are not supposed to exist. Conversely, one can miss cause-effect relationships that have been left out by gaps in the texts. I have therefore analyzed both larger interconnected maps as well as submaps of smaller, but probably relatively complete, sections of texts.28 As for symbolic analysis, much of the literature on cultural analysis suggests that symbols are the vehicles through which cultural forms (i.e., shared rules, axioms, preferences, etc.) become engaged, activated, or manifested empirically, such that culture can be communicated, learned, or contested. As Elder and Cobb summarize, "A symbol is any object used by human beings to index meanings that are not inherent in, nor discernable from, the object itself. 26 See Gregory 1989: xvii. This is a methodological goal of the postbehavioralists. On this score, the use of a replicable technique like cognitive mapping to uncover the "deep structures" of a text suggests there need not be a priori conflict between certain postbehavioralist epistemologies and certain positivist research methods in the social sciences. On the potential overlap between positivist social science and postmodernist approaches see Vaillancourt 1987: 90. 27 Yates (1988: 219), for instance, argues that the transmitted version of Sun Zi BingFa may be more fragmented than scholars have heretofore supposed. 28 To this end I coded all of the Seven Military Classics for cause-effect statements relating to the use of force for the purpose of state security—that is, causes that led directly or indirectly to the military/political defeat of the enemy and to the protection, preservation, or prosperity of the states (N=2225). I then entered these statements into a simple database computer program so that I could pull up particular cause or effect statements using key words. Causes and effects could therefore be aggregated or disaggregated in the process of analysis. The effect concepts were aggregated into four categories: (1) the weakening, diminution, disadvantaging of the enemy short of military defeat; (2) victory over, or the military defeat of, the enemy; (3) state security; and (4) unspecified notions of utility. The cause concepts were then analyzed to ascertain what sorts of strategies and behaviors led to these outcomes. See appendix A on coding.

SOME QUESTIONS OF METHODOLOGY

51

Literally anything can be a symbol: A word or a phrase, a gesture or an event, a person, a place, or a thing. An object becomes a symbol when people endow it with meaning, value or significance" (Elder and Cobb 1983: 28, Dittmer 1977: 577,579, cf. Pettigrew 1979: 574, Wuthnow et al. 1984: 37, Namenwirth and Weber 1987:21). From a symbolic perspective, then, strategic culture may be reflected by symbols about the role of force in human affairs, about the efficacy of certain strategies, and hence about what sorts of strategies are better than others. The possibility that certain symbols can cue certain repertoires or packages of behavioral options, or reveal "clear-cut modes of actions" (Dittmer 1977) seems to accord with more recent thinking in social psychology on cognitive processing and social cuing. Decision makers are constrained by specific "intuitively derived heuristics" or "mental aids" (Powell et al. 1987: 204, Bennett 1981: 76). These heuristics consist in large measure of symbols. An analysis of symbols in strategic texts, then, may reveal a great deal about how strategic axioms in a text might be interpreted behaviorally—that is, what sorts of strategicpreference rankings flow from these axioms.29 Of course, symbolic analysis should be used with caution. As Barnes warns, "It is not easy to relate political symbolism to political outcomes even with intensive knowledge of the culture itself. . . . It is difficult to convert insights into arguments that are intersubjectively convincing, rather than merely plausible" (Barnes 1986: 56). And as Kertzer adds, there may be changes in the interpretations of meanings over time even while the symbols themselves remain consistent (Kertzer 1988: 67-69). Avoiding these potential pitfalls in an absolute sense is a daunting methodological task. But a couple of precautionary steps can be taken, namely: (1) being inclusive when choosing symbols for analysis; and (2) trying to corroborate the implications of certain symbols through cognitive mapping and through the contextual analysis of the texts and their annotations. There is some reason in the Chinese case, however, to expect a high degree of consistency in both the symbols and their meanings over time. Chinese intellectual history and traditional scholarship shows marked consistency, indeed tyranny, in the repetitive use of analogies and metaphors. As Lo Jung-pang puts it, When confronted with problems, anyone mightfirstof all do well to resort to past experience in search of ready made formulas. This pragmatic attitude occurs par29

In using symbolic analysis I examined a variety of units of analysis: (1) frequently used idioms and phrases that are axiomatically accepted as valid descriptions of a strategic context and that can indirectly contain strategic preferences; (2) key words that appear to embody certain behavioral axioms, or that are used to describe legitimate actions directed at an adversary; and (3) analogies and metaphors that function as shorthand definitions of a strategic environment and that correlate to certain responses to this environment. (See Stenelo 1981: 8—14, MacCormac 1985: 23-24, Kolodner and Simpson 1986, Vertzberger 1986: 225, Kull 1988: 48, Khong 1992: 10.)

52

CHAPTER TWO

ticularly to the Chinese mind, which is steeped in literary tradition and which places great stock on the guidance of history. In the conduct of foreign affairs, as in social intercourse, there are maxims and precedents that were so constantly quoted that they became cliches and, like political slogans, exerted an influence in the shaping of policy and the making of decisions. (Lo 1969: 51)

EMPIRICAL ANALYSIS

The utility of strategic culture as an analytic concept disappears rapidly without an effort to test for its effects on strategic behavior. One of the problems that has plagued cultural analysis, however, has been precisely the difficulty in determining the "attitude-behavior" relationship. Often the problem arises when the researcher is using very general attitudes to predict specific acts (Terhune 1970: 45, Schuman and Johnson 1976: 162, 170, 173, Brown 1984: 260, Gaenslen 1986: 82). Often, too, the problem shows up when the link between group or societal values, a specific individual's attitudes, and that individual's behavior is left unspecified. Thus, at the very least one must trace the presence of strategic culture-based preference rankings from the objects of analysis through to the attitudes of specific decision makers, and by doing so outline likely modes of transmission. The influence of strategic culture-derived preference rankings on cause-effect assumptions held by decision makers before the moment of decision can show much about where strategic preferences come from. From there one should look at the relationship between these preferences and actual decisions made over time and across strategic contexts. In principle, then, a rigorous approach to the attitude-behavior problem in the Chinese case involves three steps. The first is to test for the presence of and congruence between the strategic preference rankings across the Seven Military Classics. The second is to test for the presence of and congruence between preference rankings found in a sample of documents taken from the decision process using the content-analysis techniques outlined above. These documents should be taken from different times and across different strategic contexts. The assumption here is that if strategic culture is to have a traceable behavioral effect, it must at least have an effect on the action-oriented perceptions of key decision makers. The third step is to test for effects of decision makers' preference rankings on politico-military behavior. Here there are two related methodological issues: (1) conceptualizing the relationship between strategic culture and behavior; and (2) case selection. Regarding the first consideration, the research problem is to control for the effects of culturally exogenous variables that, by implication, may provide competitive explanations of behavioral choices. This is not a clear-cut process. There are at least six ways of conceptualizing the relation-

SOME QUESTIONS OF METHODOLOGY

53

ships between strategic culture and other exogenous independent variables, which I address from least to most complex.30 In the first case, no exogenous variables need be taken into account. Strategic culture directly explains strategic choice for one of two reasons: first, it provides a single strategy option. This is implied in much of the first-generation literature. There is no intervening effect from exogenous variables because these have all been subsumed into the concept of strategic culture itself. As I suggested earlier, this conceptualization is unconvincing, deterministic, and unfalsifiable. Second, strategic culture provides a preference ranking of choices that is insensitive to change in structural variables such that its predictions are distinct, say, from a determinate version of structural realism. In a second conceptualization, strategic culture provides a limited range of choices or tendencies, and an intervening variable determines which strain kicks in and when. This selection process could be a function of leadership change, elite transformation, bureaucratic politics, technology cycles, internal debates, external crisis, among other variables, that might create conditions for certain strains to appear as the dominant strategic culture. This is a less deterministic view of the culture-behavior link than the first conceptualization. The researcher, however, has to identify the universe of logically possible options and suggest that some were ignored due to the delimiting effects of culture. In the third instance, strategic culture could be conceived of as a consistent set of ranked preferences that persist across time and across strategic contexts. Decision makers, however, are sensitive to structural or exogenous conditions (i.e., relative capabilities) in a culturally unique way. Thus changes in exogenous conditions interact with a constant strategic culture to produce variation in the composite independent variable. This interaction may (though need not) yield different predictions from a purely structural model of choice, or from a different interactive strategic-culture model. Fourth, strategic culture could be manifested in a form other than a preference ranking. It could mediate or moderate the effects of the independent variable, for example, by defining policy process (Sampson 1987). This conceptualization of strategic culture as process overlaps with much of the organizationalculture literature (Legro 1995). One difficulty with this model is that it seems less applicable to highly personalized political systems. Another is that one needs an additional theory of politics to indicate why particular organizations and their cultures become dominant in the policy process. If policy is somehow a compromise between organizations, then choice will reflect a hybrid of strategic cultures. A variant of this model would be a three-stage process where domestic political variables influence a state's strategic culture, which in turn 30 1 drew initial inspiration for this conceptualization from John Mearsheimer's remarks at the American Academy of Arts and Science Workshop on Strategic Culture, May 1990.

54

CHAPTER TWO

interacts with the external environment to produce behavioral choices (Kier 1992). Fifth, strategic culture could be an instrument, justifying policy, and excluding alternatives to debate in the domestic game, while decisions are made on an entirely different basis (e.g., the political and economic interests of strategic elites cast as the "national interest"). This conceptualization is consistent with the second-generation work on strategic culture. It is not clear, however, from this literature whether one should expect some interactive effects between strategic culture used instrumentally and those variables that affect strategic choice directly. Finally, one could conceive of a dynamic model, where strategic culture as an instrument at time t evolves into a determinant of behavior at t+1 through two processes: a process of "blowback" or "echo" where elites internalize the strategic-culture symbols they initially manipulate (Van Evera 1984, Snyder 1991), or a process where elites are constrained by strategic culture-derived mass preferences (Kupchan 1994). The problem with this conceptualization is that it does not tell us what is going on at t—1. Why at time t are elites unaffected by strategic-culture symbols used in the domestic game at t—11 Where does the process start? Also, what does strategic culture look like in states where attentive or mass publics have no meaningful role in strategic choice (e.g., Stalin's Russia or Mao's China)? Some of these conceptualizations have more merit than others. The key issue is how to measure the separate effects of a constant or slow-to-change variable such as strategic culture on an outcome that is supposed to vary like strategic choice. The first approach—embodied by most of the first-generation literature—is clearly inadequate since neither the independent nor dependent variable varies within any particular case. The fifth conceptualization posits, in effect, that strategic culture has little or no effect on behavior, and thus is not interested in testing strategic culture against other explanatory variables. The sixth conceptualization limits the presence of strategic culture to a narrower range of regimes and is therefore of limited value. The second through fourth approaches, however, allow one to consider strategic culture as a constant that, in interaction with noncultural variables, varies the independent inputs into strategic choice. These approaches therefore allow one to test the influence of strategic culture against purely noncultural models. Moreover, in contrast to the sixth approach, these three allow for a considerably expanded set of cases. Thus, any tests of alternative explanations are likely to hold more validity. I take the third approach because it combines strategic culture (a unitlevel, ideational variable) with changes in relative capabilities (a system-level structural variable) on the assumption that if behavior is more consistent with this combined independent variable than with behaviors predicted by a purely structural model, we are in a better position to attribute this difference to strategic culture.

SOME QUESTIONS OF METHODOLOGY

55

As for case selection, some might argue that in principle the research design for this question ought to be comparative across states, where ahistorical, acultural variables are similar for both states, but where variation in strategic cultures is maximized.31 This book is not a cross-national study, however. As a first cut at strategic culture a comparative design may be premature. Given my definition of strategic culture, the crucial first question is the consistency and persistence of strategic culture. This is a within-system problem. Before any cross-national or cross-cultural comparisons can be of value, the issue of consistency within a state or society needs to be resolved. This is also a practical question: the familiarity with the strategic history of a state needed to carry out this research design is quite daunting. Comparative designs are, in the long run, essential, but realistically can probably wait until country-specific studies have determined whether or not a strategic culture exists. This does not mean that a more-or-less rigorous design cannot be set up for comparisons within one state across time. There are in fact a couple of designs that could be used. The first is an interrupted time-series design where behavior before and after changes in strategic culture would be observed in a time period in which exogenous variables such as technology or threat did not change. If there were significant post-test variations in behavior, then one could infer strategic-culture effects more confidently. For example, one might test for the effects of strategic culture across the Song-Yuan or Ming-Qing transitions; in both cases a "foreign" dynasty superseded a Han Chinese dynasty. This design faces a couple of problems, however. One is a practical one: the researcher would also have to become familiar with Mongolian and Manchurian strategic cultures. A second is the possibility that, say, in the Ming-Qing transition the Manchurian elites were already highly "sinified," and thus did not bring with them a different strategic culture. An alternative design—and my initial choice for this study—is to look at one fairly lengthy historical period— i.e., the Ming period—in which competing models of strategic behavior can be tested against a strategic culture-derived model, assuming of course that such a model would yield different predictions from the others. This has the advantage of, in effect, controlling those variables that are being tested by the other models. Thus I had intended to test the behavioral propositions from a strategic-culture model against those from other relatively powerful, but simple structural models found in the internationalrelations literature. 31 Alternatively, as Przeworski and Tuene suggest (1970), one could use a "most different systems" (MDS) design, choosing states where variation in ahistorical and acultural variables is maximized, but where strategic cultures are similar. In principle, MDS designs allow researchers to use a larger number of cases than "most similar systems" designs. But the starting assumption for most of the literature on strategic culture is precisely that strategic cultures are distinctive because states are not likely to share deeply historical, formative paths of strategic experience, hence one is not likely to find many states with shared strategic cultures.

56

CHAPTER TWO

The most logical competing model might be called a realpolitik-dynastic cycle model. This assumes that states' decision makers generally share an undifferentiated interest in expanding the capabilities of the state. States will expand as long as their resources allow, since greater relative capabilities increase the relative strength of the state, and increases in the relative strength of the state increase the probability of success.32 Since preferences are constant— 32 There is controversy within the realist tradition surrounding this model of state behavior. Some argue that to assume states are inherently expansionist power-maximizers—given the opportunity—flies in the face of the evidence of considerable stability in the international system. Rather, states are first and foremost interested in state survival. The problem is that the realist theoretical edifice is itself vague or inconsistent, with Mearsheimer in the power-maximizer school and Waltz in the survival-first school (see Mearsheimer 1990, Waltz 1979, Zakaria 1992: 190-96). Given that in neorealism state preferences over ends have this vast range (Waltz 1979: 117-18), the range of optimal preferences over means can vary dramatically as well. In the nuclear age, for instance, if mere survival is preferred, then a minimum or finite-deterrent capability should suffice, and once this is achieved, a state should be generally unconcerned about relative gains in power capabilities. Capabilities beyond this level, or strategic behaviors (e.g., alliances, arms racing) that do not add to this finite deterrence can safely be eschewed. If "universal domination," to use Waltz's words, is the goal, then the desirable level of arms or power depends entirely on what others have, and indeed should ideally vastly exceed that of others. For this state, relative gains and associated realpolitik behaviors (e.g., arms racing, alliance building, etc.) characterize strategic choices. Once achieving a sufficient level of superiority, a state should act to impose its will on other states through coercive diplomacy or war. Note the vast difference in types of expected behaviors even within the neorealist theoretical construct. Which assumption about state preferences is more consistent with realism? Perhaps both, depending on the type of state system. The argument that states are survivalists or selfpreservers first and power maximizers second is an effort to explain why most major powers in the European state system have been generally unwilling to challenge the status quo. This is a lesson that has been learned over time (see Posen 1984: 68-69). The implication is that absent this lesson—either because it was never learned, or as Snyder (1991) argues because the interests of domestic coalitions override it—then states will tend to try to expand. Others argue that states have means to achieve security other than war, expansion, and the attempted elimination of all threats. One such option is alliances (Snyder 1991: 22-25). All this may be true for states operating in the standard vision of an anarchical interstate system such as existed in, say, nineteenth century Europe. But one might also argue that these sorts of restraints on power maximization are generally absent for empires. That is, empires may be precisely that form of state in which the lessons of survival versus power maximization are not learned, since the states or political entities that might teach this lesson—through anti-imperial defensive alliances for instance—are weak. Moreover, by virtue of being an empire the state has in fact learned that expansion pays, a lesson reinforced by domestic interests (including the legitimacy of the imperial leadership). Alliances are not an alternative to the use of force to eliminate threats since the capabilities gap between the empire and any potential alliance partners may be so great as to make little difference for the security of the former. With the absence of these two types of restraint on opportunist expansionism, a power-maximizing model of imperial security policy may not be an unreasonable one. Within the realist construct it seems legitimate, then, to assume that the Ming empire was a power-maximizing state. This approach is consistent with Doyle's metrocentric structural theory of imperial expansion. The theory posits that empires act to increase relative power vis-a-vis adversaries in order to increase security, subdue all threats, and preempt imperial decline. Given the opportunity to do so, empires will attempt to expand (Doyle 1986: 27-28, 123-24).

SOME QUESTIONS OF METHODOLOGY

57

expansion ranks highest—an increase in the probability of success will increase the expected utility of expansion. This increase in the expected .utility of expansion may also introduce a degree of optimism in decision makers' estimation of their ability to advantageously alter the adversary's behavior (Bueno de Mesquita 1981: 29, 30, 64). The realpolitik-dynastic cycle model would therefore predict that as the empire consolidates and mobilizes resources in the earlier stages of the dynastic cycle, it will adopt increasingly expansionist, coercive strategies (i.e., extended campaigns beyond the frontiers, preventive colonization, formal annexation of new territory etc.). The empire will become more, not less, belligerent (Vasquez 1986: 321, Vasquez 1987: 367-68). As the dynastic cycle peaks, the empire is overextended financially and militarily; bureaucratic conservatism slows the empire's ability to mobilize more resources; military technology and technique has diffused to the "barbarians"; internal unrest increases; the military and political leadership atrophies; and the empire begins to collapse inwardly. As decline sets in, the state turns to less offensively coercive, more static defensive strategies, and from there to more accommodating strategies— peace treaties, bribes, territorial concessions, etc. In thefinalperiod of imminent collapse, one might expect to see an increasing reliance on military means— static defense of contracted frontiers—in a last-ditch fight for survival. According to this model, then, the empire's expansion is constrained by resources, not by strategic-cultural baggage. The realpolitik-dynastic model comes closest to embodying the key elements of a structural-realist conceptualization of strategic choice, and given the latter's hegemony in the international-relations field, is the most appropriate alternative to a strategic-culture model. A strategic-culture model based on preference rankings found in the secondary literature (preference rankings that reflected a Confucian-Mencian equation linking moral state government and external security) would suggest a more-or-less constant trend in strategic choices during the dynastic cycle. Even as the empire mobilized resources, strategic culture would dictate policies that manifest the magnanimity of the ruler, his "awesomeness" (wei g£) and "virtue" (de flO.33 The empire would rely on a constant but low coercive mix of policies such as accommodation, detente, trade, intermarriage, treaty making (broadly termed he qin ^PiM) and static defense even though the capacity to pursue expanionist, offensive grand strategies had increased. A variation of this would suggest—if strategic culture's influence is seen as being mediated by changing levels of threat—that levels of coerciveness should decline somewhat as the power of the empire increases. During the initial stage of consolidation, when the external threat is quite high, leaders would focus on the establishment of 33 Indeed weide was a term traditionally used to describe a successful ruler, one whose majesty, power, and moral rectitude underlay efficient and incorruptible government, economic welfare, social contentment, and hence external peace.

58

CHAPTER TWO

secure frontiers, and thus rely on essentially coercive and violent strategies. But as the empire mobilized resources and the relative distribution of capabilities between,the empire and its adversaries changed in the former's favor, the leadership would rely increasingly on less coercive mixes of policies. Over time, as the empire's capacity to mobilize resources atrophies and the relative degree of threat increases, the regime once again turns to more coercive mixes of strategy. Gills neatly summarizes the relationship between standard interpretations of Chinese strategic tradition on the one hand and actual behavior on the other: The East Asian (predominantly Chinese) approach to security, "is derivative of the views of Mencius in that it downplayed the role of force and upgraded the role of value and moral example. In periods of dynastic consolidation the frequency of wars seems to be less than in periods of dynastic ascendence" (Gills 1993: 195). To summarize these two models, each shares a dependent variable—namely, change in mixes of grand strategic policies along a spectrum of coerciveness. Each makes different predictions about the basic trends in these policy packages. For the strategic-culture model the primary independent variable is culturally derived strategy-preference rankings.34 For the realpolitik-dynastic cycle model, the primary independent variable is the dynastic cycle as it affects changes in the resources and capabilities of the empire. The advantage of this design is that the strategic-culture model would be tested explicitly against a nonstrategic-culture ahistorical model. Since each makes different predictions about Ming strategy choices, any confirmation of the strategic-culture model would be more valid, since a critical alternative explanation of behavior will have been ruled out. Moreover, the models represent different levels of analysis—a cognitive and societal unit-level (strategic culture) and a structural/systemic level (realpolitik). There is an ongoing debate in the international-relations literature over which level to privilege in explanations of behavior, with the unit level generally given shorter shrift, at least until recently. To the extent that the results of this test indicate which level of analysis offers more useful explanations, the Chinese case could make a constructive contribution to this theoretical debate. Unfortunately, the content of the Seven Military Classics renders the results of this design problematic: contrary to the implications of standard interpretations, Chinese "strategic culture" does not embody a set of preference rankings that would make unambiguously different predictions from those of the 34

Technically, the strategic-culture variable does not vary in this design. That is, there is no change in strategic-preference rankings over time. What causes variation in the effect of strategic culture is its interaction with structural conditions, namely the changes in relative balance of capabilities vis-a-vis the adversary. The dynastic cycle, then, is an intervening variable. This accords with the implications in the literature that strategic culture predisposes states to respond differently to similar structural conditions.

SOME QUESTIONS OF METHODOLOGY

59

realpolitik-dynastic cycle model. Chinese strategic culture essentially shares a realpolitik-preference ranking. Moreover, Chinese strategic thought is highly contingent; the effect of an a priori ranking of strategic preferences is mediated by a powerful notion of absolute flexibility. The operational decision rule is not to minimize violence in pursuit of security, as suggested in the secondary literature, but rather, to do what is necessary in any particular contingency to destroy the enemy militarily and in some cases politically. This translates into a pattern of strategic behavior in which more coercive strategies are chosen in periods of relative power advantage. In essence, as will be argued, Chinese strategic decision making reflects an essentially realpolitik calculus of force and opportunity. This characterization of Chinese strategic culture and behavior is precisely opposite that offered by Gills and others. Yet superimposed on this deep operational structure is a set of strategic preferences based on the minimization of violence and on the expression of wei and de. This suggests the possibility, raised by the second-generation literature, that an idealized or mythological strategic-culture canon exists that has less to do with behavioral choice and more to do with how this choice is framed or justified in culturally acceptable language. Specifically, while the deep structure of Chinese strategic thought revolves around absolute flexibility and a high level of acceptable violence, there is a persistent myth or idealized vision of "not fighting and subduing the enemy" and all its strategic corollaries. At both this idealized level and at the operational levels there are discemibly consistent preference rankings, hence discernible strategic cultures. A critical empirical question then becomes what role this idealized strategic culture plays in the decision process. Does it impinge upon any effect the operational strategic culture might have on decision making and strategic choice? Empirically, this study focuses on both a detailed textual analysis of the strategic arguments behind Ming policies towards the Mongols and on aggregate changes in Ming strategy across the entire 270 plus years of the dynasty. My reasoning is that any arguments that are confirmed both by aggregate data and richer anecdotal information are more powerful than arguments confirmed by one or the other. The richer data on policy argument comes primarily from four periods: (1) the reign of the Yong Le emperor (1403-1425), in which he launched five offensive campaigns against the Mongols between 1414 and 1424; (2) the last half of the fifteenth century, when after the disastrous defeat at Tu Mu in 1449 the Ming were thrown onto the strategic defense for the first time; (3) the mid 1500s when intensive debates erupted over the merits of an offensive strategy to recapture the He Tao region west of Beijing; and (4) the 1570s when, under the leadership of the grand secretary Zhang Juzheng, the Ming pursued a policy of detente of sorts with the Mongols. The periods cover both sides of the dynastic cycle; they cover changes in the relative military capability of the empire; they cover periods when the Ming were on the political and military offense, and when they were not. But by focusing on conflict with

60

CHAPTER TWO

the Mongols in the north and northwest, I hold constant as best as can be hoped the nature of the threat and geography. As Engels said, the proof of the pudding is in the eating. What do these procedures reveal about grand strategic preferences in Chinese history? The first item on the agenda concerns assumptions found in the Seven Military Classics about the role of war in human affairs, about the nature of the enemy, and about the efficacy of force. These general assumptions provide the basis for more specific assumptions about the efficacy of different grand strategies.

Chapter Three CHINESE STRATEGIC CULTURE AND THE PARABELLUMPARADIGM

T

HE FUNDAMENTAL ELEMENTS of a strategic culture deal with, inter alia, three broad interrelated questions about the role of war in human affairs, the nature of the adversary, and the efficacy of military force and applied violence. Together answers to these three questions form the central security paradigm of a particular strategic culture.1 From this paradigm should flow logically connected strategic preferences about how to deal with threats to security. In this chapter I examine the Seven Military Classics for answers to these three questions using the analytic techniques outlined in the previous chapter. The short of it is, the cognitive maps and symbolic analysis indicate that, for the most part, the texts accept that warfare and conflict are relatively constant features of interstate affairs, that conflict with an enemy tends towards zerosum stakes, and consequently that violence is a highly efficacious means for dealing with conflict. Together these three sets of assumptions create a parabellum or hard realpolitik view of security whereby the sine qua non of state security is sufficient military capabilities and, preferably, the military defeat of the adversary. This parabellum paradigm at the core of the military classics stands in contrast with the standard image of Chinese strategic thought found in much of the secondary literature. Lebow has argued that acceptance of the inevitability of war has led, in a number of cases in international politics, to preferences for preventive war or preemptive strategies. If war is inevitable, the logic goes, it makes sense to act before the enemy inevitably does (Lebow 1981: 254-63). Zero-sum conceptualizations of conflict are often based on worst-case assumptions about the intentions of the adversary, and in worst-case assumptions the probability of an event is at, or close to, one (Kanwisher 1989: 656). Thus the perceived probability of an enemy attack is extremely high. Given this assumption it is better to strike the enemy first. The "cult of the offensive" in European military doctrines prior to World War I was rooted, arguably, in both short-term estimations of inevitable conflict between the two major alliances, and in more 1

See Taber's notion of a dominant paradigm as a collection of heuristics "that a nation uses to guide the selection of problem solving strategies for some specifiable period of time. Individual leaders may or may not be aware of the influence of a dominant paradigm on strategy choice" (Taber 1987:4).

62

CHAPTER THREE

persistent social Darwinian, zero-sum visions of international conflict (Snyder 1984: 17, 39, 199). In principle, then, this zero-sum view of the enemy ought to be associated with a belief that war is both an inevitable characteristic of human affairs and central to the security problematique of the state. The root cause of conflict is the aggressive disposition of the adversary. The security of the state, therefore, rests to the extent possible on the elimination of external threats first. This requires, ultimately, the use of violence since the enemy is not predisposed in the long run to tolerate one's own survival. Military force, if not the first resort, is indispensable to the security of the state. Moreover, since state survival is at stake, there should be no a priori moral or political limits on the application of force, regardless how it is justified. Conversely, a general belief that war is an aberrant or at least preventable event in human affairs ought to be associated with a nonzero-sum view of the adversary. If conflicts are in the main negotiable, then presumably the adversary has a price short of one's own capitulation. If this is the case, then ceteris paribus, highly coercive, violent strategies—which can entail severe economic and political costs—would seem to be strategies of last resort, since tradeoffs and "log rolling" opportunities appear more cost effective in managing security threats. Security is multidimensional, in part a function of the behavior of the adversary, and in part of one's own state's internal cohesion and socioeconomic well-being. In reality, the linkages between the answers to these three central questions are not always so clear or consistent. Students of Chinese strategic thought have generally not focused on these linkages in any systematic fashion. But there is, arguably, some consensus in the literature about the content of these answers. And generally the literature accepts that the second set of answers more closely reflects the central paradigm in Chinese strategic thought historically. In particular, the secondary literature is soaked with the notion that in Chinese strategic thought war was a "last resort," (Fairbank 1974:7, cf. Cleary 1989: 20). In the standard formulation found in the Chinese classics, "war and weaponry is an inauspicious tool, and should only be used in unavoidable

circumstances" (bingzhe xiong qiye, bu deyi eryong zhi J^^^^-tiL^^F^S

MflB5l)- This reluctance to resort to force rested, it is argued, on a low estimation of the efficacy of violence. This "disesteem" of violence was inherent in the Confucian moral order (Fairbank 1974: 25). "The resort to warfare (wu) was an admission of bankruptcy in the pursuit of wen [civil, nonmilitary action]. . .. Herein lies the pacifist bias of the Chinese tradition.. .. This stress on gaining victory without fighting is not a Utopian fantasy, but part of the larger view that seeks to maintain the established order without the use of violence" (ibid., 6-7, 11). According to Western analysts, traditional and contemporary Chinese strategic thought allegedly deemphasizes violence. Different authors put it different

THE PARABELLUM PARADIGM

63

ways: a stress on psychological warfare over the use of weaponry and firepower (Pillsbury 1980/1981); a stress on "gaining victory while keeping as much intact as possible both socially and materially rather than destroying who or whatever stands in the way" (Cleary 1988: 19); an "anti-militarist bias" (Wei and Liu 1985: 437); an emphasis on humans or mind over weaponry and the use of nonviolent strategems and deception to overcome "bare strength" (Tan Eng Bok 1984: 4, Boylan 1982: 345); an "aversion to violence in war" (Lin 1988: 31-33); the absence of a notion of ideologically based total war (Boodberg 1930: xii-xiv); the minimization of war and the corollary that wars that were not minimized were ethically unjust (Cleary 1989: 20); a "systematic denial of belligerence" (Lewis 1990: 103);2 a view that war is an "aberration" (Adelman and Shih 1993: 31); and "strategy by strategem" as opposed, presumably, to strategy involving the application of pure violence (Boorman and Boorman 1967: 152, Wang 1976: 77) among other similar characterizations .3 Scholars trace the roots of this minimal-violence heuristic to a number of disparate sources. One is Sun Zi's notion of "not fighting and subduing the enemy" (bu zhan er qu ren zhi bing ^ l ^ M i S A5lJ3c)- These scholars interpret this to mean primarily the use of nonviolent or nonmilitary actions to defeat an adversary by preventing it from achieving its politico-military goals (Wang 1976: 77, Wu 1989:19, Fu 1990: 68, Gao 1990: 2,3, Liu 1990b: 20, Yu 1990:27). Others suggest that this heuristic comes from Lao Zi and his doctrine of using "softness to overcome hardness" (yi rou ke gang JJilP: j£fK|!l) (Armed Forces University, 1984: 12-13, Li and Sun 1989: 185-89). Still others contend that the deprecation of violence is rooted in the ConfucianMencian emphasis on the ruler's cultivation of virtue and good government as the basis for the security and prosperity of the state. That is, external security rests on internal rectification, on the ruler's employing capable officials, on reducing the economic burdens on his subjects, and on creating conditions such that people will be content with their place in the socioeconomic-political order. Carried to its utmost degree, such circumstances will cause the adversary to submit willingly to the ruler's authority, thus dissolving security threats. As the Wen Zi ($Crf) summarized, "As for the ruler's being righteous, [it means] controlling the kingdoms and clans [i.e., the state], ordering affairs inside the 2

Lewis is referring to the emergent strategies in the Warring States period as representing a radical break from concepts of heroic martiality in the Spring and Autumn period warrior culture. Many of the main works on Chinese strategic thought (including most of the books in the Seven Military Classics) spring from the Warring States period. 3 Challenges to these characterizations are generally rare, but tend to come from those who have done detailed empirical research on Chinese strategic behavior that goes beyond the analysis of ideas. See, for instance, Sariti's argument that even in the Northern Song period, a period renowned in later historiography for its defensive grand strategy and alleged aversion to expansion, "None, not even the most outspoken critic of government policy, rejected war or the military as legitimate instruments through which to settle foreign policy problems" (Sariti 1973: 5).

64

CHAPTER THREE

boundaries, implementing benevolence and righteousness, spreading virtue, displaying kindness, establishing correct rules, blocking the way of treachery, [such that] the multitudinous officials are close and united [loyal], the masses are peaceful and united, superiors and inferiors are of one mind, the many officials unite their efforts, the feudal lords submit to [the ruler's] awesomeness, and surrounding states embrace his virtue" (Qiu 1985: 254, cf. Zeng 1972: 72-73, Li 1978: 124-25, Xu 1983: 126, Lewis 1990: 66). Such was the way the sage kings and legendary rulers of the golden ages glorified by Confucians handled the security problematique. This construct was summarized by a host of idioms passed down through history: "emphasize civility, deemphasize martiality; stress virtue and downplay physical strength" (zhong wen qing wu, zhong de bu zhong li fi^ClMSiSIB-^fi^J); "if one has virtue, one cannot be matched [e.g., by an enemy] (you de bu ke you di ^'jM^f RlWiSSO; "display virtue and do not flaunt the military instrument" (guan de bu yao bing H! tS^^fSiSr)- It is even argued that Sun Zi was a proponent of these doctrines when he placed the dao (Jl) as thefirstelement decision makers should consider in security affairs. In this instance some have interpreted dao in Confucian terms to mean "the policies of the ruler being enlightened, the masses supporting the ruler, and superiors and inferiors being united in heart and mind" (Yu 1990: 22). When the use of force becomes "unavoidable," then, the literature suggests, Chinese strategic propensities lean to the defensive and limited use of force. Offensive wars of annihilation, it is argued, were rarely used historically. Chinese political and military leaders generally eschewed exterminating states, occupying territory, or killing the people of an enemy state (Xu 1983: 126). A number of reasons are given for this. One is that China's rulers and strategists were generally pessimistic about the ability to eliminate external threats. The barbarian problem, for instance, had existed since ancient times and could only be managed, not resolved (Rand 1977, chap. 1, Tao 1983: 74, 78, Wu 1978a: 188, Xu 1987: 318). Loewe argues that in their efforts to control the Xiongnu barbarians on their northern and northwestern borders, rulers of the Han dynasty, with the exception of Han Wu Di, resorted largely to a static defense posture. Offensive campaigns, when they were launched, were punitive and aimed mostly at deterrence and pacification rather than annihilation. The strategy was classically defensive: "to deny an enemy access to Chinese cities and farms" (Loewe 1974: 106, Wallacher 1975: 343-44). Cleary suggests a moral source of defensiveness. In the Daoist and Confucian ethical systems, he argues, "[war] should be taken only as a last resort, and only in a just cause. This generally means defensive war, but can also mean punitive war to stop the strong from bullying the weak" (Cleary 1989: 20). Another scholar argues that defensiveness can be traced back to Mo Zi, a fifthcentury B.C. military philosopher and practioner of siege defense, who propounded a doctrine of "nonoffense" (fei gong #X£) (Lin 1988: 19-20, 57).

THE PARABELLUM PARADIGM

65

Two of Taiwan's most prominent scholars of ancient Chinese military thought, retired Guomindang generals Wei Rulin and Liu Zhongping, also find cultural roots for this alleged defensive use of force. Chinese culture, they contend, stressed self-protection by means of walls, earthworks, and strategic strong points rather than invasion, a predisposition rooted in the security imperatives of a sedentary agricultural society. Wei and Liu also point to the alleged use of culture rather than territory to define "us and them" in ancient China as militating against pressure to expand boundaries aggressively. As Confucius said, "If the Yi and Di barbarians invade the central states [i.e., the core cultural area], then sinify them" (yi dijin ru Zhongguo ze Zhongguo zhi 51 ^ M A 11 ^ SWi^ft i@/^L). If the adoption of Sino-Confucian cultural norms was sufficient to place others inside the Chinese realm, then the active territorial and political expansion of the realm was unnecessary. If this were the case, then the solution to barbarian threats to the realm's security was enculturation, not extermination. And enculturation was, fundamentally, a question of example and inducement, not force (Wei and Liu 1985: 88, 302-3, Xu 1983: 75, 125, 134, 159-60, Liu 1990a: 375, Interview 1991a).4 Interestingly, this enculturation doctrine implies a nonzero-sum view of the adversary. If through externally directed policies of "education through civility" (wenjiao >C$SO and expressions of magnanimity the barbarian can be sinified and pacified, then, in principle, there is no fundamentally irresolvable conflict of interests. As various idioms suggest, the enemy is a threat to the realm not dispositionally but rather situationally.5 Another alleged element of Chinese strategic thought and behavior is a stress on limited wars. A key characteristic of limited wars is the controlled calibration of violence to well-defined, limited political goals (Brodie 1973). These limits can also include well-defined spatial and temporal restrictions on violence. Together these conditions separate limited war from unlimited war, where any and all means are used to obtain the unconditional surrender or destruction of the enemy as soon as possible. Limited wars, therefore, are distinguished by tight controls on the pressures to escalate conflict. Much of the secondary literature accepts a similar characterization of Chinese strategic thought and practice. Bobrow (1964), the Boormans (1967, 1969, 1972), Whiting (1975), Chan (1978), Godwin (1984), Tan Eng Bok (1984), and Lin (1988) all, in various ways, point to the Chinese exercise of military force as a strategic tradition capable of maintaining a rational balance between 4 Some P.R.C. scholars come to a similar conclusion about the defensive preferences of agricultural societies (Interview 1990b, Liu 1990c: 1-2). 5 For example, "Those within the four seas are of one family" (si hai neiyijia 23}gp3 — | ^ ) ; "within the four seas all are brothers (si hai zhi neijie xiong H9$§;£. p3 W 51); "there are no discords [space, fracture, difference] between Han and barbarian" (Hua Yi wujian ^£~^kjB£f$); "In their essense Han and barbarian are one family" (Hua Yi benyi jia I j i 9 l ^ — W . ) - See Liu 1990a, and Yang 1990.

66

CHAPTER THREE

political ends and military means. China has been relatively successful, it is argued, in using limited amounts of force, in coordination with diplomatic tools, to pursue clearly defined political aims. These authors also contend that grand strategic behavior in the post-1949 period has been characterized by defense at or only slightly beyond immediate borders, a relative rather than zero-sum concept of victory, pacification and deterrence rather than the extermination of the enemy, a nonzero-sum concept of conflict that reduces escalation pressures, and a strict hierarchy of political goals, among other traits. This ability to control the level and breadth of violence stems, in principle, from a strict hierarchy of goals during conflict. Very often, pressures to escalate war beyond the originally perceived limits come from an inability to rank the utility of these goals and to "quit while one is ahead"—that is, to try to cease hostilities once certain goals are met while leaving others partially or wholly unmet. If these goals are not clearly ranked, then setbacks, say in lowerpriority political objectives, have a tendency to transform these goals into higher-priority ones, which therefore more urgently require a favorable outcome. This problem is probably especially acute when conflict is viewed as zero-sum, where the prospect of partially unmet goals is viewed as an absolute loss. One solution to this escalation dynamic is to hold, consciously or not, to a long-term vision of the value of different strategic outcomes. The value of future goals is sustained in the face of short-term setbacks, making these more palatable. This reduces the pressure to escalate violence in the present. In the Chinese case, some argue that a distant horizon is a key cultural characteristic of Confucian economic ethics, namely a willingness to accept material sacrifices in the interests of long-term welfare (Lin and Ho 1989: 4). Others see this quality as central to "strategic patience" in Chinese military thought and practice. As one study put it, Chinese strategy has shown an ability "to maintain an objective over an unusually long period of time without seeing frustration or impotence resulting from long delays" (Boorman and Boorman 1967: 150, Boorman 1969: 23, Wu 1990: 93 ). On the basis of the secondary literature, then, we have one set of answers for China to the three basic assumptions, or central paradigm, of strategic culture: war is inauspicious and to be avoided; the enemy is not necessarily demonized—it can be enculturated and pacified, though not exterminated or annihilated; violence is a last resort. When violence is used by the state to deal with external security threats, it is generally applied defensively and is limited in nature. This reflects a view that violence is not particularly efficacious in eliminating threats or producing security. Rather, the more reliable basis for state security is the moral and administrative quality of internal rulership. What, then, do the military classics have to say about the content of this central paradigm? Most of the texts repeat in various forms the statement that weapons are inauspicious instruments and are only to be used under unavoidable circumstances. The San Lue, for instance, notes, "As for the military, it is

THE PARABELLUM PARADIGM

67

not an auspicious instrument; it is the way of heaven to despise it" (Liu [SKQS]: 726.124, Shi 1222: 33.10b). The Wei Liao Zi states, "As for the military, it is an inauspicious instrument; as for conflict and contention, it runs counter to virtue" (Liu 1955c: 81). But rather than concluding from this that war can be avoided, the military classics generally view war as an inevitable product of conflict between human interests, and indeed of the human character. The first line in Sun Zi's The Art of War summarizes the ascribed role of war as a constant social phenomenon: "Military affairs are a vital issue of the state. It is the realm of death and life; it is the way of survival or ruin. It cannot but be investigated" (Wu 1983: 29, Rand 1977: 7).6 The sources of war, according to the military texts, were variegated but ever present. This conclusion is reflected in the sources and typologies of war found in the Wu Zi Bing Fa. Wu Zi names five sources of war: a struggle for fame or reputation; a struggle for material benefit; an accumulation of resentments and hatreds; internal chaos and rebellion; and famine or poor internal economic conditions. These conditions produce the occasion of war, but these occasions can then be divided into five types of war: righteous war; wars of coercion and aggression; willful wars; violent or wanton wars; and wars of rebellion against the status quo, or correct moral order (Li and Wang 1986: 58).7 Given the multiple sources of conflict and the different forms of war, weapons and warfare could not be done away with. Indeed, while constant warring threatens to drain the resources of the state and lead to its destruction, to be unprepared for war is also dangerous and foolish public policy. As the Si Ma Fa notes, "Thus, although the state is powerful, if it is bellicose it necessarily perishes; although the realm is at peace, if [the state] forgets how to war, then it necessarily faces danger" (Liu 1955b: 5).8 To eschew war in the face of 6 Sun Bin, the author of another Warring States military classic, discovered in the early 1970s, writes, "Military victory is the reason for states surviving and perishing, for future generations continuing and being cut off. Military defeat is the reason for having land annexed and the temples of grain and land being threatened. Thus one cannot but investigate military affairs" (Xu 1987: 317). This conclusion is paralleled in other works on statecraft. Zhuang Zi notes, for instance, the source of military affairs was contention over things. "The masses of people take that which is not necessary and turn it into something that is necessary, thus there is much warfare" (Qiu 1985: 235). This is echoed in a Tang dynasty text on military aspects of the Dao De Jing. The author notes in his preface, "People take that which is unnecessary and turn it into something which is necessary, thus this is the reason why war and invasion flourishes more day by day. . . . As for contention and conflict, it is the source of the military instrument, the root of calamity and chaos" (Wang 812: 4a-b). The pre-Han Lu Shi Chun Qiu concludes, "As for martial affairs, one cannot get rid of them.. .. The source from which the military instrument comes is ancient" (Lin 1986: 174). 7 The Wen Zi, another Warring States period work on statecraft, citing Lao Zi, offers a somewhat similar typology: righteous war, wars of response to aggression, wars of indignation and anger, covetous wars arising from a desire to acquire territory and wealth, and wars of arrogance and aggressiveness. Cited in Qiu 1985: 242. 8 See also Sun Zi on the dangers of protracted war (Wu 1983: 41), Wu Zi on five types of

68

CHAPTER THREE

willful, wanton aggression and rebellion was both bad policy and an abdication of moral responsibility. As Wu Zi argues, "If one meets an enemy and does not advance, one cannot be said to have reached a state of righteousness" (Li and Wang 1986: 57). Sun Bin, born about 100 years after the death of his purported ancestor Sun Zi, wrote that "If in war one does not win, then as a result territory is sliced off and the 'altars and temples' [i.e., the state] are endangered" (Sun 1985: 49). Logically, then, one should have the capabilities to win a war in order to protect the state. Thus, the admonition to use force only under unavoidable circumstances implies that one can legitimately resort to force when the actions of the enemy are such that no efficacious alternative is available. That is, the actions of others compel one to resort to force. This is not the same thing as saying that force should be resorted to infrequently, since the frequency is wholly determined by the behavior of the opponent. Given the assumption of a fair degree of constancy in the conflictual behavior of others, the military texts therefore imply that the legitimate resort to force is also not likely to be infrequent. From this perspective, then, the implication of "using force under unavoidable circumstances" is not so much that war as a matter of policy is a last resort to be used when all other means are exhausted, but that war as a matter of reality in the face of a broad range of sources and types of conflict is a relatively constant social practice. Moreover, the linguistic construct "using force under unavoidable circumstances" shifts the responsibility for warlike behavior onto the enemy. One is thus exonerated from responsibility for creating conflictual conditions, and one's use of force is therefore never illegitimate. Indeed, in the military texts, the only conditions under which force can be legitimately employed is to fight a "righteous war" (yi zhan §1®) against those who have created the conditions for war. "Using force under unavoidable circumstances" does not limit one's use of violence to retaliation for unrighteous wars initiated by others, but allows one to initiate war in the name of righteousness. As the Si Ma Fa puts it explicitly, "If one kills [the enemy] in order to bring peace to his people, then killing is permissable. If one attacks the enemy's state in order to love and protect his people, then attacking is permissable. If one uses war to prevent war, although it is still war, it is permissible" (Liu 1955b: 3-4).9 This interpretation of "using force under unavoidable circumstances" turns the complete phrase, "the military is an inauspicious instrument that should victory, and Wei Liao Zi (Liu 1955c: 19.1, 25.1). 9 See also Shi Zimei's annotation of the San hue where he cites Si Ma Fa on this score (Shi 1222: 33.1 Ob). There is a parallel argument in the Lu Shi Chun Qiu: "If the military is righteous [i.e., wielded for righteous purposes], then attack and invasion are permissible, coming to someone's rescue and defense is also permissible. If the military is unrighteous, then attack and invasion are wrong, and coming to someone's rescue and defense is wrong as well" (Lin 1986: 181).

THE PARABELLUM

PARADIGM

69

only be used under unavoidable circumstances," into a seeming contradiction. How can war be both an inauspicious, immoral instrument and an instrument of righteousness? The resolution lies in the fact that the Seven Military Classics see the nature of the military (bing J i ) in relative terms. The military is not an inauspicious instrument in an absolute sense, only in a relative one. Some uses are more inauspicious and immoral than others. Some uses are more righteous than others. This implies that the military, weapons, or war are themselves neutral tools of policy. What gives them moral content is who uses them for what purpose.10 What makes the use of force legitimate as a tool of state policy, then, is its use for the purposes of upholding righteousness.

RIGHTEOUS WAR

The notion of righteous war pervades the Seven Military Classics as well as other classics on statecraft.11 Generally the righteous use of force means "sending forth armor and weapons in order to punish the unrighteous," "punishing the rebellious and suppressing the unrighteous" (Liu 1955c: 32.2, Lin 1986: 188), or "punishing the violent and cruel, and suppressing the rebellious" (Liu [SKQS]: 726.124). As for what concretely constituted cruel, violent, rebellious, or unrighteous behavior, the military texts offered several examples, most of which are roughly patterned off the Zhou Rites's "Nine Punishments."12 These include a ruler: bullying weaker states (fan ^B, qin fj|); bullying his own people; being violent internally (bao JH) and insulting other states externally; neglecting agriculture (bu ren ^fiZ);u killing one's own people (sha |j£) and disregarding the regularities of heaven, earth, and humankind (i.e., demonstrating a lack of propriety—bu yi ^ § | ) ; behaving like a wild animal (luan j§L); and rebellion against the established political and social order (pan fS) (Liu 1955b: 9-12, Zeng 1986: 12, Lin 1986: 188). Against such behaviors, righteous actions included cutting off a portion of the violator's land (xue fflj); attacking and suppressing him (fa fjc); sending him into exile (tan ijj[); exterminating him (mie $$,); invading his state to compel his submission (qin {§); executing him (zhu |$); 10

Interestingly, this finds resonance in the Maoist view that who uses war and weapons for what purposes determines whether or not the use of force is justified (Kim 1980: 230). 11 The exception is Sun Zi Bingfa where righteousness is mentioned only once, in the final chapter on the use of spies. 12 The Zhou Rites text outlined the correct organization and procedures of government and rulership, including rules regulating the relationship between the Zhou dynasty court and feudal states (e.g., the Nine Punishments). There is some debate when exactly the text appeared, though most place it in the Warring States period. The text is considered a Confucian classic because it embodied Confucian ideals of statecraft. 13 Literally, neglecting benevolence.

70

CHAPTER THREE

destroying him (can ^$); and isolating him from other states (Liu 1955b: 1315, Zeng 1986: 112). In short, in the face of unrighteous behavior one must negate the enemy with force (// jj) (Shi 1222: 34: 17a). Under the banner of righteousness, the destruction (po 5$) of the enemy was both necessary and desirable (ibid.: 31:9b). Righteous wars would restore the correct order of things, people, and policies. In short, righteous war was a conservative, restorative notion (Liu 1955b: 42.2).14 This brief survey leads to a point worth underscoring, namely, that within the context of righteous war, the ends clearly justify the means. Once the ends of war are deemed righteous, then any and all means become righteous by themselves. Righteousness does not constrain one's options, but rather opens them up. Indeed, as the range of choices in the list of righteous actions indicates, actions that can be unrighteous (i.e., invasion and extermination) are righteous when infused with moral intent.15 Righteousness (yi H ) in some sense means controlling or ordering things (zhi N/n) as they should be. If this ordering is done benevolently (ren JZ), then the process is called just (zheng IE). If benevolence is inadequate to the task, then the ruler must use force. But the act of controlling or correctly ordering things by definition remains righteous, so the ruler can resort to force and still be righteous. Thus it would seem that any action that falls under the rubric of force (quan ^ | ) is, by definition, a righteous act if used for righteous purposes (Liu 1955b: chap. I).16 14

The concept of righteous war was certainly not unique to the military texts, nor was the prominent role played by legitimate violence. A number of other classic texts from different schools of thought all contained similar notions. The Guan Zi, commonly classified as a Legalist text, notes that the "best use of the military instrument is not to acquire territory, but to punish peoples' [unrighteous] rulers" (Qiu 1985: 334). The righteous use of the military instrument, moreover, was necessarily violent (ibid., 397). Similarly, the Wen Zi, a Warring States text with a Confucian and Daoist flavor, argues that the military instrument is not for acquiring land or others' material wealth, but for suppressing disorder and eliminating threats to people, including those in other states. Mo Zi, the proponent of the doctrine of nonoffense, in fact justifies offensive war by relabeling it "righteous" on the basis of a utilitarian calculation of who benefits. If the war benefits the greatest number of people, if it protects the weak as well, then it is "just" (ibid., 27-28). Even Mencius, the architect of Confucian antimilitarism, does not reject the use of force if it is for just causes. Indeed, he takes the term for attacking (zheng f]E)—a behavior to which he objects on moral grounds—and names it just (zheng IE) if done by a benevolent and virtuous ruler. In other words, the empirical act he condemns becomes justifiable if done for righteous purpose. 15 A similar conclusion was presented to me by two experts at the National Defense University in Beijing (Interview 1990a), and by a retired Guomindang general (Interview 1989b). There are parallels with Machiavelli's view of morality and war: "Morality in Machiavelli's usage is entirely instrumental: it is part of a prince's arsenal to be used to greater or lesser effect. In no sense does it restrain state behavior—nor should one expect it to do so because the state in Machiavelli's treatment is beyond such restraint" (Smith 1986: 10). 16 The same argument is made in a Ming dynasty text on strategy. The author argued that although military force is an inauspicious instrument, when used in conformity with the principles of heaven it was, by definition, a benevolent and righteous action. The author criticized dogmatic

THE PARABELLUM

PARADIGM

71

There are, then, no a priori moral limits on the means and methods of righteous war. Thus, the conclusion in much of the secondary literature that righteous war embodies admonitions against the invasion or the extermination of the adversary (Ding 1984:236) is not entirely accurate. Suggesting, as Cleary does (1989: 20), that righteous war justifies only limited punitive expeditions confuses the action with the language by which it is justified.17 The only a priori element of restraint that the military texts seem to place on behavior taken under the rubric of righteous war concerns how to treat the enemy's people after their defeat. The victors were to be benevolent in their treatment of the enemy, and were to refrain from hunting, destroying land and crops, stealing tools and livestock, wrecking temples and houses, and harassing women and children (Liu 1955b, 1955c, Anonymous 1986: 173). But this restraint, interestingly, seems to have been primarily politically motivated—rather than morally so—since this enabled the righteous forces to capture the hearts and minds of the vanquished, thus making control easier and more legitimate, and thus also defusing resentments and feelings of revenge born from defeat (Anonymous 1986:173). The faster these political goals could be accomplished, the sooner the righteous forces could withdraw.18

O N VIOLENCE

This discussion of righteous war provides clues to the content of the other two basic elements in a strategic culture's central paradigm, namely the nature of the adversary and the role of violence in state security.19 The righteous-war Confucians for neglecting this principle and thus putting the interests of the state in jeopardy (Zhao 1569: preface, l a - l b , 5a-5b). See also the annotation of the first chapter of the .S1/ Ma Fa in the Qing edition of the Seven Military Classics (Zhu 1989: 122). Shi Zimei used this principle to argue that the classic Confucian notion of "cultivating the Way"—typically identified as an accommodationist concept of security—could also justify the use of force. "Using perfect benevolence to punitively attack those who are not benevolent, and using perfect righteousness to punitively attack the unrighteous, this is 'cultivating the Way'" (Shi 1222: 486a). 17 It is interesting to note the linguistic constructs used to justify China's invasion of Vietnam in 1979. It was termed a "defensive counterattack" designed to "teach a lesson." The first term denotes that China was the aggrieved party and its response was taken "under unavoidable circumstances." The second term implies the response was merely punitive, aimed at restoring a correct order in the political, military, and geographical relations between China, Vietnam, and Kampuchea. The language is consistent with the notion of righteous war. 18 The temporary presence of the righteous army in the defeated state is a prevalent theme in the comments on righteous war. The Wei Liao Zi, for instance, states that once an enemy's army has been defeated, its general killed, its cities destroyed, and some unspecified amount of land acquired, the victorious army should then return home (Liu 1955c: 14). Withdrawal does not rule out annexation of portions of a state's territory, and in fact the San Lue is premised on the creation of a "virtuous and benevolent" puppet ruler. 19 The discussion here moves back and forth across a number of levels of strategy in order to

72

CHAPTER THREE

doctrine embodies a zero-sum view of the adversary, since it places unrighteous violators of the moral-political order beyond the pale. The enemy is irredeemably an enemy, one who cannot be won over but must be destroyed. The concept of enemy (di fSft) itself connotes unrighteousness. An enemy is qualitatively different from a rebel or a troublemaker who can be won over by bribes or even kindness. The San hue seems to imply as much when it comments that to deal with rebels one must try to get them to return, both literally and figuratively, to the political and social fold (pan zhe huan zhi f&^M^.), but enemies must be exterminated (di zhe can zhi jfft#?|5l) (Liu [SKQS]: 726.103). As Shi Zimei's annotation of the Liu Tao puts it, within the realm there cannot be two centers of authority. In conflicts deemed righteous, there can only be one victor and one vanquished (Shi 1222: 36.7b).20 The demonization of an unrighteous adversary—and the corollary that any and all means of eliminating this enemy are legitimate—is described well by modern-day attribution theory. The adversary's behavior is seen as dispositional. It is by nature predisposed to acting unrighteously, hence nothing short of destroying the enemy or preventing it from acting according to its nature alters the basic threat. One's own behavior, on the other hand, is circumstantial, a reaction to a dangerous situation created by the adversary, hence one's response is taken "under unavoidable circumstances." This construct creates "an image of the situation for personal and public consumption that releases the subject from moral inhibitions and allows the subject to deal with the threat or opportunity without restraint. The threat will be killed with pride and the opportunities taken without embarassment" (Herrmann 1988: 183,185). The use of force under these conditions is framed as defensive and of complete necessity (Hart 1978: 150, Eiser and van der Pligt 1988: 45-66, cf. Mack 1990: 6061, Stein 1990: 82). This leads to a related conclusion. In the military texts the use of the military instrument (bing J=£) is considered highly efficacious, indeed necessary, for dealing with security threats. This is not to say that war or violence is the sole basis of state security, but there is a strong expectation that violence inheres in human social processes, and that preparations for, and the use of, violence are vital for self-preservation. Thus the status of violence in these texts is higher than implied in the secondary Western and Chinese literature. It is also higher than implied in the Confucian-Mencian notion of internal rectification as the basis for external security that infuses much of this secondary provide some insights into the status of violence in traditional Chinese military thought as it applies to different levels of strategic behavior. The next chapter takes these insights and applies them to the analysis of grand strategy. 20 The Lu Shi Chun Qiu describes this zero-sum relationship succinctly: "As for [my] victory in military affairs, it means the enemy's loss. . . . If the enemy obtains life from me, then I obtain death from the enemy. If the enemy obtains death from me, I obtain life from the enemy" (Lin 1986:211,217).

military strategem

THE PARABELLUM weakening, diminution of the enemy

PARADIGM

73

defeat, submission of the enemy>

political, diplomatic strategem

military preparations, the application of force

internal rectification, quality of internal rule

military morale, mobilization of human and material resources

i

security of the state

direct positive causal relationship implied positive causal relationship

Fig. 3.1. Composite Cognitive Map of the Wu Zi Bing Fa. literature. To get at this question perhaps the best route is to go through the Seven Military Classics book by book, examining the texts for the causal relationships, metaphors, analogies, and languages that may reveal assumptions about the relationship between violence and security. Wu Zi Bing Fa It is quite obvious that in the Wu Zi Bing Fa the sources of state security are varied and include both martial {wu ^ ) and civil or nonmilitary {wen ~$Q elements. One causal path in the text's cognitive map—a route that is consistent with Confucian-Mencian notions of security—moves from the cultivation of the four virtues ("the way," righteousness, propriety, and benevolence) within the state directly to the survival and flourishing of the state (152.2a.6).21 Another route links enlightened internal policies to "victory in the battle" (154.1a.5).22 However, the unstated linking concept here appears from the context to be 21

A direct cause of a particular effect refers to a cause that connects directly to an effect, and does not move through an intermediary concept. An indirect cause is at least once removed from the effect. The three numerals in the brackets given to a particular cause-effect relationship refer respectively to the page number in the version of the text 1 coded, the line or section number on that page, and the assigned number of the cause-effect relationship found in that line or section. In instances where there are a great number of cause-effect statements tracing a similar causal path, I only supply a few of the representative samples. 22 The phrase used in the text could be translated as "victory already being won" (zhan yi sheng yi fJccLll^l), meaning being victorious before having to fight by virtue of attracting such overwhelming support that one faces no enemies. This is consistent with the Confucian-

74

CHAPTER THREE

"morale." That is, soldiers will fight more courageously—hence military capabilities will be more efficacious—if they are convinced both of the benefits of defending a righteous ruler and of the moral correctness of their fight. Thus this "internal rectification" causal concept generally does not circumvent "martiality" on its way to state security. Indeed, in two places Wu Zi traces a positive relationship between a cause that stresses both internal rectification and martiality and state security. The argument is, in effect, that the ruler should cultivate virtue internally while applying martiality externally (151.5.6; 154.5.1).23 Elsewhere Wu Zi contends that the cultivation of internal virtue to the neglect of the military instrument leads to the extermination of the state (151.5.1). Military preparations, Wu Zi argues, are the most valuable means by which the state can be made secure (Li and Wang 1986: 70). The indispensability of the military instrument is reflected as well in the causal link between the use of military power, the weakening or diminution of the enemy, and the security of the state. The primary causes of the enemy's military and political decline or diminution almost all involve the application of force. This can be broken down into a couple of subcauses: (1) striking at the enemy swiftly and preemptively or otherwise attacking it directly (e.g., 155.2.8,155.2.9,157.1a.l, 157. la.2); and (2) the use of military ruse (155.2.7). There are two cause-effect relationships that do not explicitly involve the use of force, but that apparently involve the use of political-diplomatic methods to manipulate the emotions and motives of the enemy. In one case, the cause concept implies offering some sort of advantage to encourage the enemy to act in a particular way (155.2.5); the other implies instilling fear and hesitation in the enemy (155.2.14b). These causes then both lead to the military weakening of the adversary. However, both cause concepts are vague and could easily mean the use of military means (i.e., feigning retreat and lying in ambush, or displaying impressive force) to instill in the enemy a false sense of advantage on the one hand or a^ false sense of disadvantage on the other. The political and military diminution of the enemy is itself a cause, though an implied one, of another effect concept: that of fighting few wars yet winning them all. Infrequent but decisive victories are, in turn, linked positively to state security (152.1b.9). As for the defeat of the enemy, the key causal inputs are the possession of weapons and well-trained soldiers to use them Mencian notion of security. One contemporary annotation, however, interprets this phrase as meaning that the ruler will then be able to achieve victory in war (Li and Wang 1986: 64). My interpretation is closer to the latter because it is more consistent with the indispensability of martiality for the security of the state and for the defeat of the adversary. 23 This conclusion is also reflected in one of the few analogies used in the text. Wu Zi cites two ancient leaders, one from the Cheng Sang tribe and one from the Yu Hu tribe. The first stressed the cultivation of virtue and neglected military preparations; his state was extinguished. The second was bellicose and concentrated solely on developing his military capabilities, and consequently he "lost" his state (Griffiths 1982: 151).

military strategem

THE PARABELLUM -^weakening, diminution of the enemy v x x

PARADIGM

75

defeat, submission of the enemy

political, diplomatic strategem

military preparations, the application of force

internal rectification, • quality of internal rule

military morale, mobilization of human and material resources

security of the state

direct positive causal relationship implied positive causal relationship

Fig. 3.2. Composite Cognitive Map of the Si Ma Fa. (151.4.1), rigorous and effective internal military organization (159.1.8, 162.1.4), and the use of military strategems to envelope and then attack the enemy (164.1b.2). There is no cause leading to the defeat of the enemy that does not involve the use of military force. All of this creates a picture of state security that rests primarily on the military defeat of the enemy through the development, organization, training, and application of military capabilities. While the Wu Zi Bing Fa does not attach probabilities to the cause-effect relationships, the text leaves the role of internal rectification in state security highly underspecified, and in any case it is not differentiated from martiality as an independent source of security.24 Si Ma Fa Like the Wu ZiBing Fa, the causal paths in the Si Ma Fa at first glance seem to indicate that state security is caused by two parallel processes, one that is essentially civil in nature and the other martial. One cause-effect relationship, 24

Using utility as an effect concept again finds the primary cause is the application of martiality. In fact, the largest proportion of inputs into utility involve the preemptive use of force when the enemy is caught in various disadvantageous circumstances (156.1a.l, 156.1a.7, 157.2a.l-157.2a.16). There are a handful of cause-effect relationships where the source of utility is the quality of the ruler's leadership (e.g. his expression of "the dao" of propriety and benevolence etc.). But in context these refer to the qualities of "sage rulers," historical figures who in the military texts are often seen as ideal types, and of less immediate relevance to the practical problems of security.

76

CHAPTER THREE

for instance, establishes a positive link between the ruler's ability to obtain the people's support and the survival of the state (4.4.1). Another indicates a positive relationship between correct ritual in rulership and the steadfastness of the state's defense (19.4.1). Others link the ruler's benevolence, righteousness, and trustworthiness to the control of disorder (42.1.1, 42.1.2, 42.1.5). All of these linkages, however, are open to some interpretation. The first two suggest that the quality of the leader's rule—its conformity to the correct order of heaven and earth—results in high morale in the populace, and hence an increased willingness to sacrifice for the military defense of the state.25 In the last set of cause-effect relationships, "disorder" could refer to general chaos and instability externally, within the state, or within the army. The context of these relationships suggests it is the last of these, since they appear in the midst of an explicit discussion of how to create and maintain high morale and discipline in the military. Thus the implied next step in achieving state security is the successful application of military power, where benevolence-derived morale and military order are key internal inputs into military success.26 This is consistent with another explicit direct cause of state security, namely a state's military capability. The Si Ma Fa uses a simple deterrence calculus to argue that a state's peacetime preparations for war are a source of security (5.4.2). While the text warns that a state must also avoid using its military capabilities in a reckless, belligerent fashion if its security is to be assured (5.4.4), this is a call for restraint in the use of force; it negates the value of the military instrument for state security. The sources of this military capacity lie in mastering the timing of heaven (tian shi ^cB^f), obtaining abundant military supplies through pillaging enemy stores, estimating the relative military advantages and disadvantages of terrain, using first-class weapons, and motivating soldiers to exert themselves in war (30.4.1-30.4.5).27 None of these inputs, strictly speaking, is subsumed under the category of internal rectification, where the hostile intentions of potential enemies dissolve in the face of righteous and benevolent rulership. Another input into state security is, of course, victory over the enemy. What leads to the enemy's submission? The implied cause is the diminution or weakening of the enemy. As in the Wu ZiBingFa, the weakening of the enemy is a result of the application of military force, not the application of nonmilitary strategems or more general policies of internal rectification. Rather, the number, diversity, and quality of weapons (38.3.3, 31.3.1), numerical superiority 25 This is the implication of Shi Zimei's annotation (1222: 17.3b), Liu Yin's annotation (1955b: 4), as well as the modern annotations by Tian (1990: 50) and Liu (1986: 36). 26 Another cause-effect relationship makes this linkage explicit: the preparedness of the military in training and armaments leads to the state's ability to defend itself over the long term (46.3.3). 27 "Timing of heaven" {tian shi 3^0$) refers to the limits placed on operations by anything from climatic conditions to agricultural cycles. Mastering the constraints imposed by tian shi means being able to discern the precise advantages of time and conditions for the preemptive application of force.

THE PARABELLUM

PARADIGM

77

in soldiers (53.4.1), and the quality of military strategem (54.1.1) are positively related to the diminution of the enemy's military power short of its defeat. The direct causes, again, lean heavily to the wielding of the military instrument. These include superior technology and weaponry (22.3.6, 46.3.4), numerical superiority when operating in enemy territory (47.2.4, 53.4.1), and superior morale resulting from the proper combination of rewards and punishments and from the general's "awesomeness" (wei$0 (53.3.7,46.2.2,24.1.2,24.3.2, 43.1.1, 48.3.2). There is no direct causal route, say, from nonviolent political/ diplomatic strategems to military victory. Nor, again, is there any hint that internal-rectification policies alone can dissipate external threats or cause the enemy to submit bloodlessly.28 An analysis of the cognitive map, then, seems to indicate the centrality of the military instrument in the pursuit of state security. Does this mean that civil-related concepts are inconsequential? Not entirely. The first chapter on "Benevolence as the Root," in particular, refers in a general way to the ruler's righteousness and benevolence as a source of the people's "delighting" in their ruler, thus submitting willingly to his control (4.3.1, 4.3.2). This submission also adds to the ruler's awesomeness. But the next section of the text strongly implies that the value of this support lies in the willingness of the masses to then defend the state against adversaries (4.4.1, 4.4.2). An analysis of the metaphors and analogies used in the text adds a degree of complexity to this issue of civil- (wen) versus military- (wu) related sources of security. Two sets of metaphors in particular address the relationship between these two elements. One refers to wen as being like the right-hand side of things (you ;jzf) and wu as the being the left-hand side (zuo ;££) (Liu 1955b: 26). This suggests an unequal relationship since in classical Chinese "right" commonly refers to a place or position of esteem, while "left" refers to the inferior or secondary position. Elsewhere, in a similar vein, benevolence, a we/z-related concept, is described by the term for "root" (ben ^), while martiality is refered to as the "branch" (mo 7^). The text implies that in statecraft it is best to rely on the root first and on the branch secondarily. This would seem to be similar to the standard Confucian-Mencian admonition that war should be a last resort. 28

A large proportion of the relationships in the text are between a cause and an implied utility in warfare. Here the causes of utility can be broken down into five types: military intelligence (37.1.1, 55.1.1-55.1.5); morale (e.g., 41.2.1-41.2.6); strategem (e.g. 55.1.2, 55.1.4, 55.2.4, 55.2.5); motivating capable officers and soldiers by means of kindness (37.2.3); and using military force to attack the enemy's vulnerable or weakened forces preemptively and with flexibility (e.g., 30.4.1, 37.1.2, 37.2.2, 53.1.1-53.1.5, 54.5.1, 54.5.3, 55.1.1, 55.2.4). The vast proportion of these causes increases the efficacy of the military instrument. The civil or wenrelated causes do not stand out as separate routes to utility. The kindness cause, for instance, refers to motivating soldiers before battle. In the case of strategem the context implies that one should use military force, say, to provoke the enemy, feign weakness, or induce a reaction from which one can draw conclusions about the enemy's strengths and weaknesses, capability, and intentions. These conclusions then help decide when, where, and how to apply military force.

78

CHAPTER THREE

Yet analogies elsewhere in the text appear to put this "civility over martiality" relationship in a bounded historical context. While the Si Ma Fa recognizes that the "first kings" and ancient sage rulers relied on benevolent and righteous rule and on moral education to elicit the willing submission of all those in the realm, the text traces a linear, secular evolution in grand strategy from this internal rectification-based approach down to the use of violence. In a discussion of the legendary rulers Xia Hou Shi ( i J g f t ) , Yin Tang (|£i§), and the Zhou kings Wen ( J i l ^ i E ) and Wu (JUx^Hi), the text argues that Xia Hou Shi used a yielding approach, cultivated his virtue, and ensured the security of the realm; Yin Tang relied on righteousness primarily; and the Zhou kings used military strength (// ~jj) (Liu 1955b: 20-21). There is no strong sense that a return to the golden age of Xia Hou Shi or Yin Tang was historically possible. Elsewhere, the lessons the text draws from the statecraft of the ancients mostly have to do with the use of force, namely using superior capabilities to attack tired or weary enemy forces or forces that appear intimidated or fearful (ibid., 53). In short, as in the Wu Zi Bing Fa, the causal arguments made and the metaphorical or analogical lessons drawn in the Si Ma Fa do not seem to confirm the standard Confucian-Mencian interpretation of the role of violence in achieving state security. Rather, they suggest a strong faith in the efficacy of applied military force to weaken the enemy, achieve victory, and preserve the political status quo. Wei Liao Zi Like the previous two texts, the Wei Liao Zi appears to trace the security of the state back to both civil and martial causes. Consistent with its reputation as a Legalist-influenced book on strategy, the Wei Liao Zi states that the well-being and order of the state depend on agricultural development (5.1.1, 19.3.1) and a large and well-regulated population (5.1.2). Both these elements, it is argued, give the state and its ruler such awesomeness that the military instrument never need be applied. "[If the state has] wealth and order, then the people do not 'remove the skids' [on war chariots], and armor is not brought out [of storage]" (Liu 1955c: 5). Note that in this Legalist version of the notion of internal rectification it is wealth (fu H ) and order (zhi tp), not benevolence (ren {H) and righteousness (yi i i ) that cause state security. But for the Wei Liao Zi the implied link is "awesomeness" (wei g£), meaning, in part, the presence of an overwhelming ability to exercise military power if necessary. Wealth and order are prerequisites for a large and powerful military organization that should pose, at the very least, a latent threat to potential adversaries. The security of the state is also caused directly by a ruler's employing capable officials—a construct that stresses the wen-based nature of security. Here too, however, the context suggests this could be a reference to both capable civilian

military strategem

THE PARABELLUM weakening, diminution of the enemy

PARADIGM

79

defeat, submission of the enemy

political, diplomatic strategem

military preparations, the application of force

internal rectification, -* quality of internal rule

military morale, mobilization of human and material resources

security of the state

direct positive causal relationship implied positive causal relationship

Fig. 3.3. Composite Cognitive Map of the WeiLiao Zi. or military officials. The Wei Liao Zi argues that with such officials the ruler is able to capture and hold territory and control populations and resources belonging to the adversary. Without these capable officials, the text warns, the state will inevitably face military defeat and see its general killed (Liu 1987: 42).29 There are, finally, clearly military-related direct causes of security, namely an excellent general. A general's excellence is primarily defined by his command and organizational abilities and his effectiveness in maneuvering his army to attack at precisely the most opportune moment. "The flourishing or perishing [of the state], and its [facing] peace or danger" resides "in the head of the drumstick" (Liu 1955c: 34), meaning it resides in the general's command abilities.30 Unlike the Wu ZiBingFa or the Si Ma Fa, the WeiLiao Zi outlines a clearer, 29

It is not clear whether the officials are the victorious state's officials and officers or those of the vanquished state. See chapter 4 for a fuller discussion of the problems of consolidating victory. 30 The general's drumstick was used to signal advance, hence it is a symbol of his command. Another direct cause of security that may be military-related is an effective defense structure and strategy (31.1.6). The phrase "perishing consists of nothing being preserved/defended" (wangzaiyu wu suo shou tffi 1 ?Mffi^) could also be interpreted as preserving and upholding military orders and discipline, given that the next line refers to the need to have effective rules and commands if the state is to avoid danger. The first interpretation is found in Xu Yong 1989: 93, the second in Liu Zhongping 1987: 103. Shi Zimei annotates it as "not preserving sufficient stores" (supplies, food, etc.), which is also consistent with the Legalist stress on the material bases of power (Shi 1222: 36.6a).

80

CHAPTER THREE

more unadulterated link between the application of military force and the enemy's diminution short of military defeat. There are two basic causes at work. One is the general. Here the Wei Liao Zi uses a number of metaphors to describe the general's use of military force in overwhelming the adversary. He is likened to a great wall pressing in on the adversary (7.1.6), or like a great cloud enveloping the enemy (7.1.8). The general's ability to maintain the military cohesion of chariot forces, special operations forces (qi bing 1?f J=£), and the masses of foot soldiers is also related positively to the army's ability to damage and weaken the adversary (9.2.4). The second set of direct causes here relates, in various forms, simply to the application of military force (81.3.3, 16.1.2): its speed and awesomeness (7.1.3);31 its ability to cut off the enemy's escape routes, occupy its towns, attack its strategic strongpoints (25.3.3-25.3.5); its preemptive application of both regular (zheng IE) and special-operations (qi iff) forces (54.1.1, 54.1.2); and, when on the defensive, the accumulation of sufficient soldiers and weapons (27.1.3) as well as the implemention of a scorched-earth policy outside of one's own besieged cities (27.1.4). There is no mention of nonviolent strategems (nor military strategem and ruse for that matter) as a direct cause of the enemy's diminution. Clearly the enemy's diminution is itself a cause of military victory, and hence the direct causes noted above all count as indirect causes of military victory. But there are many other direct causes of military victory found in the text. And here we come across one of the clearest statements of a nonviolent route to victory in the Seven Military Classics. We also therefore come up against a complex analytical and strategic question, namely how to gauge the relative centrality or status of nonviolent routes to state security in traditional Chinese strategic thought. First, though, the less vexing sources of military victory. The range of causes is, as in the Si Ma Fa, fairly extensive. These include whether the ruler has correctly estimated his population and resources, whether the state has allocated sufficient resources to defend the realm, and whether a system is in place for the efficient mobilization of these human and material resources (4.1.2, 4.1.3, L38.3). Another cause relates to the commander's ability to train (21.2.7) and command forces (24.3.1, 22.1.2, 81.4.1, 81.4.2), and his ability to motivate and inspire soldiers to fight (L34.2, 81.4.5, 81.4.6). The quality of weapons is also positively related to the ability to defeat the adversary and capture his cities (2.2.3), and to the efficacy of the military instrument in general (e.g., 38.2.1, 38.2.2).32 31

Here the text likens the military in action to a mountain and a flame. Here the text uses two metaphors to describe a military that has both highly effective weapons and martial courage. Such forces are likened to a bird striking its prey (fa zhi ru niao ji -§|;^I$[] JSjifi:), and to having the ability to leap over an endless abyss (rufu qian ren zhi qi $D )- The imagery is of inexorable, efficacious force. 32

THE PARABELLUM

PARADIGM

81

There are no inherent contradictions among these various sources of victory. In one important chapter, however, the Wei Liao Zi outlines three alternative causes of victory. The first is called "victory by means of the dao" (dao sheng MM), where the ruler or his general so understands the essence of military affairs, and has so effectively estimated the enemy's capabilities and intentions that nonviolent (but unspecified) measures dissipate the enemy's qi (HO, or essence and life force, such that its army scatters before the victorious side has deployed its forces (14.1.1).33 The second type of victory is called "victory by means of awesomeness" (wei sheng MLW), where the system of military rules and regulations is highly developed, the system of rewards and punishments is clearly delineated, and weapons are in a high state of preparedness. As a result, the soldiers are determined to fight for the state and its ruler. The effect is to intimidate the enemy and compel it to submit, or at least to abandon its aggressive plans (14.1.2). The third form of victory is termed "victory by means of military strength" (// sheng j~J |]§). This involves the application of violence such that the enemy's army is destroyed, its general killed, its cities besieged and captured, its population scattered, and portions of its territory captured (14.1.3). The text does not explicitly rank these types of victories from most to least morally and militarily acceptable.34 If the Wei Liao Zi were consistent with Confucian-Mencian notions of security, then "victory by means of the dao" would be clearly superior to either "victory by means of awesomeness" or "victory by means of military force." While "victory by means of the dao" and "victory by means of awesomeness" share a similar result—the nonviolent submission of the enemy—arguably victory by means of awesomeness and victory by means of military force are of the same type, since both rely on wielding and displaying the military instrument. Thus "victory by means of 33

The causal mechanism at work here is unclear in the text, and neither Shi Zimei's nor Liu Yin's annotations throws any light on how this qi is dissipated in a nonviolent manner. Nor is Rand (1977: 87) very illuminating here. One contemporary Chinese commentator sees it as the use of strategem to undermine the enemy, but is equally vague about its content and application. He merely says the ruler "thinks of a way" to dissipate enemy qi (Xu 1989: 73). Taiwanese commentary is similar (Liu 1987: 48). At first glance, the mechanism might be awesomeness, or the intimidation of the enemy. But this is the key element in the second type of victory discussed below. "Victory by means of the dao" may mean that the enemy realizes its "form" (xing 7&), or capabilities and intentions, is so well understood and its strategy and movements have been so completely laid bare that it abandons its aggressive plans and submits voluntarily. 34 Shi Zimei's commentary does provide a ranking. "Military force is not as good as awesomeness, and awesomeness is not as good as victory by means of the dao" (Shi 1222: 24.1a). However, he also implies that while this moral order exists, these three types of victory also reflect a linear historical development from "victory by means of the dao"—characteristic of ancient sage kings—to "victory by means of awesomeness"—characteristic of the hegemons of the Spring and Autumn period—to "victory by means of military force"—characteristic of the Warring States period. This appears to be a pragmatic recognition of a historical evolution away from this moral ranking.

82

CHAPTER THREE

the dao" would rank higher as a desirable strategic axiom in the text. This would tend to offset the prevalence of military-related causal concepts in the process leading to the defeat of the enemy and to the attainment of state security. On the other hand, an argument can be made that "victory by means of awesomeness" and "victory by means of military force" are more central to the Wei Liao Z/'s strategic calculus. For one thing, none of the other direct causes of the enemy's submission relate to "victory by means of the dao." In fact, the remaining causes can be divided safely into those related to victory by means of awesomeness and victory by means of military force. For instance, with regard to the latter the text positively links increases in military capabilities to the pursuit and capture of the enemy (21.2.2). The use of military forces to block the enemy's source of supplies and route of retreat is also causally related to the defeat of the enemy (55.2.4), as is attacking the enemy with regular forces, followed by "clutching" him with special-operations forces (83.1.1). Perhaps more so than military force, victory by means of awesomeness figures prominently in the text as a cause of the adversary's defeat. There are at least six instances where the causal concept is awesomeness. These relate to the general's cultivation of his own awesomeness (L229.1, 21.2.6) after his receiving the mandate to launch a righteous war against transgressors of the moral political order (15.2.2, 56.4.2, 15.2.5). Another important facet of awesomeness is the cultivation of high morale among soldiers, such that the enemy is intimidated, particularly when one is conducting operations within its territory (15.2.3, cf. Liu 1987: 50-51). The end result of the effective cultivation of awesomeness is the submission of the enemy without resort to direct military engagement. Indeed, in one chapter the Wei Liao Zi argues this explicitly (Liu 1955c: 56). This nonviolent submission is clearly preferred in the Wei Liao Zi, but it is the result of military strength (i.e., a function of will and capabilities).35 The outcome does not rest directly on the quality of political rulership, nor on nonmilitary political and diplomatic strategems, which seem to lie at the heart of the notion of victory by means of the dao. Rather, the mediating concept between the military instrument on the one hand and nonviolent victory on the other is military awesomeness. The mediating concept between the military instrument and violent victory is the application of superior military forces. The metaphors used in the text also seem to underscore the centrality of awesomeness and applied military power in the achievement of security. The ideal general is likened to a great wall pressing in on the enemy (ibid., 7). Heavy forces (i.e., well-armed and trained, numerically superior forces) are likened to a mountain, a forest, and a great river—immovable, overwhelming obstacles in the way of the enemy (ibid.). Wielding the military instrument is 35 As the text notes elsewhere, the strengthening of the military instrument can result in the unification of the realm "without staining the blades with blood" (bing bu xue ren (Liu 1955c: 37).

THE PARABELLUM weakening, diminution of the enemy **-*..^

military strategem

PARADIGM

83

defeat, submission of the enemy

political, diplomatic strategem

internal rectification quality of rule

military preparations, • the application of force

\

security of the state

military morale, mobilization of human and material resources

direct positive causal relationship implied positive causal relationship

Fig. 3.4. Composite Cognitive Map of the Huang Shi Gong San Lue. compared to holding a crossbow in constant readiness to pull the trigger (ibid.). This is an image of latent force that would fly forth with inexorable speed and accuracy should the enemy not submit.36 Applied force is characterized as a whirlwind (ibid.), as a bird of prey attacking (ibid., 34), as thunder and lightning striking before the enemy can cover its ears or close its eyes (ibid., 37), as a river bursting its banks (ibid., 41, 56), where the army attacks with the speed and determination of one trying to save a drowning person (ibid., 22). All these metaphors imply a basic belief in the efficacy of the preemptive use of overwhelming violence against the enemy, if latent force is insufficient to handle the security problem.

Huang Shi Gong San Lue Unlike the other military classics, the cognitive map of the San Lue appears to place far less emphasis on the military-operational causes of the enemy's submission and the security of the state. Rather, the text appears to outline in a very clear manner the nonmilitary routes to security—namely the causal rela36 Interestingly, one of the idioms used in Chinese to characterize awesomeness is "to draw the bow but not to shoot" (yin er bufa 51 ffjj ^WO- This fits with the crossbow metaphor in the Wei Liao Zi as a description of a simple deterrence calculus. Indeed, the character for awesomeness or latent force (wei i£) is the first character in the compound term for "deterrence" (wei she S f i r ) , which itself is quite old. As early as a Qin dynasty text, the Lu Shi Chun Qiu, one finds a simple statement relating awesomeness to the enemy's immobilization from fear. "To employ [humankind's] inauspicious nature [i.e., martial ferocity] one must certainly [display] awesomeness. Because of awesomeness one terrorizes [the enemy]" (Lin 1986: 199).

84

CHAPTER THREE

tionship between internal rectification and external security. This perhaps reflects the fact that of all the Seven Military Classics the San Lue devotes most of its space to statecraft, problems of government, and ruler-minister relations, while giving far less space to questions of strategy, operations, and military organization. If the Tai Gong Liu Tao stands at one pole as a manual on strategy, tactics, logistics, and battle formations, then the San Lue stands at the other pole as a more elevated philosophical, often obscure essay on "political strategy" (Xu 1986: 20). This is not to say the text neglects military sources of security—rather it traces two clearly separate routes, as well as various interacting causal paths, to state security. This makes it difficult to extract a dominant path and to weigh the relative effects of violence and nonviolence on "good" strategic outcomes. The text traces the security of the state back to four major sets of causes. By far the most common, however, is the quality of rulership within the state. Here the text stresses a number of elements: the ruler's ability to find talented officials and prevent the rise of cabals in the court (726.100.8, 726.111.6-8); the implementation of internal policies that do not impose economic burdens on the people but that help them to become contented with their lot (726.111.5, 726.119.15, 726.122.4, 726.124.1); the ruler's own satisfaction with the territory and authority he already possesses (726.120.7); and the ruler's authority over the military branch of government (726.124.7, 726.124.8, 726.124.9). This last causal concept refers to the dangers inherent in a state controlled by those whose spirit and talents are expressed through the exercise of military force. Should they control political power, the text suggests, the state will end up in an endless, exhausting series of conflicts. This last point touches on another set of causes of security, namely the ruler's (or general's) ability to balance "softness" (rou f]5) and "weakness" (ruo f§) with "hardness" (gang HI) and "strength" (qiang 3j|j). The exact meaning of these terms is not entirely obvious. In one passage "softness" implies qualities and behaviors that express virtue, while "hardness" refers to behaviors that are contrary to virtue, or are unrighteous, or involve the use of violence. Hence, at one point the text seems to argue that virtue can overcome nonvirtuous, violent behavior (rou nengzhigang l ^ f b f l M ) . This would appear consistent with a classically Mencian view of state security, where the quality of internal rule ultimately dissolves external threats. Liu Yin's annotation upholds this interpretation: those who are "soft" or virtuous will attract support from others; those who are hard and unrighteous will alienate potential supporters, thus helping to build an overwhelming coalition against them (Liu [SKQS]: 726.100). Later, however, the text seems to imply a somewhat different sense of softness and hardness, where these refer mainly to nonviolent and violent means, respectively, for dealing with security threats. Here the San Lue suggests the ruler or general should possess both sets of qualities, though with a nuanced

THE PARABELLUM

PARADIGM

85

preference for softness. Being excessively soft can encourage aggressive neighbors to slice off portions of the state, but being excessively hard can lead to the extermination of the state, presumably through endless warfare (ibid., 726.102). Liu Yin's annotation suggests that softness can also refer to nonviolent strategems while hardness refers to violence. He uses the famous story of Sun Bin's defeat of Pang Juan at the battle of Maling in 341 B.C. to illustrate his point. Sun Bin is said to have duped Pang Juan into thinking Sun's weakening army was retreating by lighting progressively fewer campfires at night over a three-day period. Pang Juan thus gave chase, but fell into Sun Bin's ambush. This coordinated use of softness and hardness implies a somewhat different relationship between nonviolent and violent sources of security, a more balanced view that is less consistent with the classically Confucian-Mencian conceptualization. Yet there are also a couple of causes of state security that fit comfortably into the category of applied violence. One is the hegemonic use of power (quart H ) , including military force, to control the realm (726.115.2). Another is simply the general's achieving military victory (726.108.4). Victory over the enemy is, in turn, an effect of several causes. These include the application of violence, specifically preemptively attacking the enemy before its defense preparations are completed (726.109.15, 726.109.19); applying the "way" of "softness" in warfare, meaning conforming to the nature of the enemy and devising counterstrategies that cut to the heart of the enemy's weaknesses (726.101.15, 726.103.35); military organizational cohesion, namely the clarity and correctness of military administration, rewards, and punishments (726.106.12); and a vague notion of internal rectification, where the ruler cultivates and cherishes the people (726.110.6, 726.110.8). This last cause, however, can, as in other texts, refer to cherishing the people who forrrTthe army. In other words, the notion implies both providing military training for the people and cultivating their morale, in effect cultivating both their capabilities and their willingness to fight for their ruler. Indeed, this interpretation fits the predominant causes of military victory in the text. These include the general's sharing of hardships with his troops (726.100.4, 726.105.13), his bestowing favors on the soldiers (726.110.7, 726.110.5), and his preserving the unity of mind and purpose in the army (726.110.9). Most of the causes of the enemy's submission, then, are products of the relationship between the ruler or general and his troops, and the effect of this relationship on morale, fighting ability, and military awesomeness. There is very little direct mention of the application of violence per se as a cause of victory. This imbalance changes somewhat when one turns to measures that weaken the enemy, short of defeat. Here two of the causal statements are explicit about the efficacy of applied violence. Specifically, the text relates capturing the enemy's capable military officers (726.105.1) and attacking the enemy with speed (726.109.15) to the enemy's weakening. Another cause can be safely

86

CHAPTER THREE

categorized as nonviolent strategy, whereby the enemy is tricked diplomatically and militarily (726.103.36). A final cause refers more generally to the level of military discipline within the army (726.106.15). We see in the San Lue's cause-effect statements, then, a fairly ambiguous role for violence in the military successes and security of the state. Whereas in the Wu ZiBingFa and the Si Ma Fa, for instance, the effects of internal rectification can be linked to state security through its impact on morale within the military—that is, on the efficacy of the military instrument—in the San hue that linkage is not so obvious. The explicit role of violence in the defeat of the enemy is also downplayed in this text. Rather, the focus seems to be primarily on military morale, itself a function of the affective relationship between the general and his troops. Unfortunately, the metaphors and analogies used in the text do not clear up the ambiguity about the role of violence. On the one hand, in a lengthy discussion of the internal political sources of state insecurity, the text lambasts corrupt officials and their cliques, comparing them to intertwining brambles and vines. These usurp the ruler's authority, deceive him, provide bad advice and thus threaten the survival of the state (Liu [SKQS] 726.112,726.113). This reinforces the conclusion that the text takes a Confucian-Mencian view of the quality of government as central to the existence of the state. On the other hand, one passage refers to the general metaphorically as the "life" or "fate" of the state (Jiangzhe guo zhi mingye T ^ ^ H S L np iM), concluding with the causal statement that if the general is able to achieve victory, then the state will be secure and stable (ibid., 726.108). The text is clear, however, about the legitimacy of violence for righteous causes. Endorsing a notion of righteous war common to the other military classics, the San hue notes near the end that while the wise king does not revel in the application of military force, he must use it nevertheless to punish the violent and suppress disorder (ibid., 726.124). The metaphor used to describe righteous war leaves no doubt about the efficacy of overwhelming military force. "The sage king does not delight in using the military instrument. It is used to punish the violent and suppress disorder. Using righteousness to punish the unrighteous is like a great river bursting [its banks] and snuffing out a flame" (ibid., 726.124).37 The metaphor's sense of overproportionality provides an interesting contrast to the stress at the start of the text on softness overcoming hardness; but it is consistent with the metaphors used in the Wei Liao Zi and in the Sun Zi Bing Fa to describe the efficacious use of the military instrument. 37 An earlier passage put it similarly, "To exterminate rebels and retaliate against the enemy, this is the bursting forth (jue $£) of righteousness" (Liu [SKQS] 726.120). The term/we is used in the passage above to mean the bursting of a river's banks, (juejiang he

THE PARABELLUM military strategem

PARADIGM

8 7

weakening, diminution of the enemy defeat, submission of the enemy

political, diplomatic strategem

military preparations, the application of force

internal rectification, quality of internal rule

military morale, mobilization of human and material resources

security of the state

direct positive causal relationship implied positive causal relationship

Fig. 3.5. Composite Cognitive Map of the Tai Gong Liu Tao.

Tai Gong Liu Tao More so than the other military classics, the extant Liu Tao provides a comprehensive discussion of different levels of statecraft, public policy, and what would be termed "national security" today. Its first section is devoted to the broadest level of politics: how a ruler should order society, recruit his officials, and implement policies broadly aimed at achieving the pacification, prosperity, and security of the state. The remaining five sections deal more specifically with military-security issues, moving from macro questions of grand strategy to increasingly detailed questions about tactical and battlefield contingencies, military training, and organization. I make this observation only to put the Liu Tao's discussion of the sources of state security into context. While at a very general level the text tends to focus on internal rectification as the basis of state security, the context suggests that, like the other military classics, the military basis of security is of greater concern than one would conclude from all the statements connecting internal rectification to external security. One finds again, as in the other texts, a strong faith in a primed and usable military capability as an indispensable source of security. Of undeniable importance as a cause of security, according to the Liu Tao, is the personal character and talents of the ruler. A ruler's wisdom is seen as a source of social order in the state (9.2.4). His material self-abnegation, and hence his unwillingness to impose heavy burdens on the masses, is crucial for

88

CHAPTER THREE

the happiness and material well-being of his people (11.3.1, 11.2.2, 13.1.5). His stress on righteousness over desire and on benevolence in his government is a source of the state's greatness and prosperity (17.2.2,35.2.1,35.2.3,38.2.2). It is also, more broadly, the cause of the willful submission to his rule of those in the realm (8.1.1, 8.2.3, 47.1.3, 47.1.4). The ruler's ability to cultivate the loyalty of feudal lords is also positively related to the defense of state territory (20.2.1). All this contributes to his legitimacy—his ability to avoid "losing his masses" (shi qizhong ^igljcjc)—and thus to the security of the state (21.1.4). Directly related to the quality of the ruler is the quality of public policy. Minimum levels of levies and taxes prevent the people from "appearing starved and cold" (11.3.4,13.1.4). The ruler's providing "great sufficiency" within the realm is directly related to the protection of the realm (47.1.1). The presence of righteous, upright officials quietly but capably implementing policies leads to the flourishing of the state (13.1.6, 14.1.2, 14.1.3, 35.2.1, 38.3.2, 33.3.3, 43.3.3). The text also outlines some externally directed sources of security. The ruler must also "manage" (yu f^P) the four borders correctly if he is to be able to defend the state. Yu connotes a sense of control, whereby what is controlled is driven like one would drive a chariot or ride a horse. This does not imply that the most efficacious methods of control are only coercive ones, but it does not exclude coercion either. Indeed, the text explicitly argues that using physical force to suppress those who rebel against internal and external order—the unrighteous—leads to the submission and harmony of the realm (22.2.8). In one passage, the text outlines a direct positive relationship between the extension of the ruler's coercive power (quan S ) and the allegiance of the realm. This distinctly martial-based source of state security is found as well in statements concerning the role of the general. The Liu Tao notes that the ruler's ability to employ a capable and upright general is a key element in the protection of the state (59.3.3), and is positively related to the flourishing of the state (80.2.2, 80.2.4). This conclusion parallels the other military classics' views about the general's role as the arbiter of the state's survival. On this score the text also concludes—as the Sun Zi Bing Fa does in its first line—that warfare is directly related to the survival or ruin of the state (59.3.1,59.3.2). The implication is that warfare is a vital activity of the state, or at least enough of a constant in the state's environment that the ability to wage war cannot be neglected, given its implications for the fundamental existence of the state. As Shi Zimei annotates these passages, "Whether in warfare there is success or failure [determines] whether the state survives or perishes" (Shi 1222: 36.7a, 36.27b-28a). Thus, in the Liu Tao the military instrument and military victory are causes of state security; internal rectification is not the only source. How, then, does one apply the military instrument so as to achieve victory, the defeat of the enemy, and the management of external threats? Here the direct role of violence

THE PARABELLUM

PARADIGM

89

is especially evident. The text notes that attacking with speed and momentum results in the destruction of the adversary. And indeed the metaphors used in these cause-effect statements imply that the use of force should be preemptive. The general acts with lightning- or thunder-like speed, striking before the enemy has time to "cover its eyes and ears," that is, before its military preparations are complete (71.1.2, 71.1.1). The army itself moves forward like a rushing torrent of water or like a furious wind, overwhelming anything caught in its path (75.1.3, 75.1.4). The forces attack with the speed, direction, and efficacy of an arrowhead, with the inexorableness of a released trigger of a crossbow (77.1.1, 77.1.2). The army attacks suddenly where the enemy does not anticipate, thus creating military victory (33.1.2).38 A couple of cause-effect statements in the text outline a somewhat more detailed path to military victory. One suggests that a victorious army uses military strategem to create advantageous conditions for the application of military force, thus destroying the enemy's army and capturing his general (77.2A, 77.2.2). Another argues that when deploying forces in distant areas deep within enemy territory, one should attack with explosive speed, then feign retreat, presumably luring the enemy out of defensive positions. Then one can capture the enemy's walled cities and force the submission of its towns (78.1.1). Another direct cause of the enemy's defeat has to do with morale within the army— more specifically, the general's use of righteous war rhetoric to whip up morale encourages the soldiers on to victory (80.1.5).39 As with some of the other texts, it is clear that for the Liu Tao another cause of the enemy's defeat is its physical weakening and mental confusion (43.5.4, 43.2.4). This cause is, in turn, an effect of a mix of military and political strategies. It is at this point that nonmilitary strategem plays a role, though an indirect one, in the ultimate suppression of external threats. On the military side, the text offers a general admonition that the ruler's reluctance to exercise military power is positively related to the strengthening of the enemy state (28.3.3). This statement comes in the context of two other cause-effect relationships that suggest that potential threats to the state are more or less a constant feature of the environment, both internal and external to the state, and that therefore the ruler must act decisively and preemptively to suppress or eradicate such dangers.40 In context, then, the first cause-effect 38

The phrase is "jiji qi buyi $el|i:j£^TJr." This is similar to Sun Zi's admonition that it is best to "appear [with military force] where the enemy does not anticipate it" (chu qi huyi 39 This recognition of the instrumental value of declaring one's own forces righteous is not uncommon in traditional military texts. Sun Bin's comments that "[when] troops are few [outnumbered] but militarily strong, [this means] they have righteousness" (Sun Bin 1985: 49). 40 Specifically, the text argues, "Thus if the ruler is not furious when he ought to be furious, then treacherous officials will flourish. If he does not kill when he ought to kill, then great thieves [enemies] will spring forth. If military power is not exercised, then the enemy state will be strong" (Liu 1955d: 28).

90

CHAPTER THREE

statement implies a direct causal relationship between wielding the military instrument and weakening or controlling the enemy. The text also expounds on the role of military strategem in this process of weakening the enemy. Here the Liu Tao stresses the presentation of false appearances as key to maneuvering the enemy into a militarily disadvantageous position. The army should appear outwardly chaotic, but retain internal order (32.3.1, 77.3.1); it should appear weak when in fact it is strong (73.2.1); its soldiers should appear hungry and poorly supplied while remaining well fed (32.3.2); the army should appear outwardly incompetent while remaining alert and highly trained (32.3.3, 33.3.4); plans and strategems must remain concealed (32.3.4, 32.3.6); elite troops should be waiting in ambush (32.2.8). All of this is aimed at forcing the enemy to commit itself to a particular strategy, thereby exposing its military and strategic weaknesses. This creates an advantageous position from which to attack with military force. Part of the process of defeating the enemy, however, also involves undermining its political cohesion. Here the Liu Tao outlines a number of political strategems all designed to confuse, manipulate, and destabilize the adversary's leader. Bribing, or otherwise cultivating ties with, the enemy's officials leads to growing dissension within the enemy court (44.2.1, 44.2.2, 49.4.4, 49.4.1, 49.2.7) and the appearance of cabals and factions (46.1.1). Bribing the enemy ruler's advisers allows one to obtain deep intelligence (43.4.1), and encourages these officials to neglect their duties (44.3.1). Courting those in whom the enemy ruler places his trust leads to divisions within his mind, and suspicions about divided loyalties among those around him (43.3.1). This, in turn, contributes to the decline of the enemy's cohesion and will (43.3.2). Using a specious manner and pleasing words to communicate with the adversary leads to the atrophy of its will to fight (43.5.3). Bribing the enemy and developing in it a sense of trust makes it easier to manipulate the ruler (45.1.1,49.4.6). Similarly, adopting a fawning demeanor towards the enemy also helps develop its trust, thus creating opportunities for manipulation (45.3.1, 45.3.3). Gifts of riches, women, and music accentuate the moral perversion of enemy officials, and help to divert their attentions away from problems of internal governance and from one's own intentions (45.5.1, 43.5.2, 46.2.2, 46.2.3). In sum, all these political strategems lead to the degradation of the enemy ruler's ability to assess the nature of the conflict. They undermine his ability to gauge the nature of one's own intentions and capabilities, and thus he is unable to judge which responses are the most appropriate. But these political ploys do not by themselves lead directly to the defeat of the enemy, or to the eradication of the threat it poses. Rather, these all set up advantageous political conditions that, along with advantageous military conditions created by military strategem, allow for the more efficacious application of military force. Indeed, in one critical section of the Liu Tao—a chapter on "nonmilitary attack" {wen fa j^CfJc)—the text lists twelve key political strategems but closes these off by stating that only once they are all prepared, when conditions are

military strategem

W

THE PARABELLUM weakening, dimunition of the enemy V ^ ^

political, diplomatic strategem

internal rectification, quality of internal rule

military preparations, the application of force

PARADIGM

91

defeat, submission of the enemy

•

security of the state

military morale, mobilization of human and material resources

direct positive causal relationship implied positive causal relationship

Fig. 3.6. Composite Cognitive Map of the Tang Tai ZongLi Wei Gong Wen Dui. ripe, can one complete one's military tasks (shi erjie bei nai cheng wu shi

Tang Tai Zong Li Wei Gong Wen Dui The Wen Dui is essentially a multi-leveled, extended discussion between the Tang emperor Tai Zong and his chief strategist Li Jing on national security issues. These range from broad questions on strategies to deal with barbarian threats, to the dialectics of "regular" and "extraordinary" military operations, to battlefield formations and tactics, to assessments of the military philosophies of other military texts. Compared to the other military classics, the Wen Dui pays very little attention to the political, economic, or moral sources of state security. But it does share an essential faith in the importance of the military instrument for achieving state security. This faith is tempered by the caution, such as one finds in the Si Ma Fa and Sun Zi Bing Fa, that excessive reliance on warfare, particularly protracted, geographically extended warfare, is harmful to the well-being of the masses and the security of the state. The text cites with approval the statement in the Si Ma Fa that preparing for war, even in times of peace, leads to the security of the state (135.1.2); under41

The only metaphors used in the text to describe deceptive behavior and strategem, whether political or military, do imply that this deception is a prelude to the violent destruction of the enemy. In one passage the text notes that just as birds of prey fly low and fold in their wings before striking, and wild animals press back their ears and lie prostrate in ambush before pouncing, so the sagacious general takes on a doltish, feeble, hence vulnerable appearance before striking first (Liu 1955d: 38).

92

CHAPTER THREE

standing the way of warfare leads to peace, the expansion of the ruler's awesomeness, and the fulfillment of the people's needs (Z152.1.1—Z151.1.3).42 But the text also warns that a belligerent though powerful state will end up sacrificing its security to incessant war, and thus will invariably perish (135.1.1). The text also explicitly borrows Sun Zi's admonition that the application of force in distant places over long supply lines impoverishes the people (126.2.2), and hence threatens the prosperity, stability, and security of the state. It is clear, then, that successful warfare plays an important role in achieving and preserving state security. Security threats can be, and are, dealt with by the use of force. As for what specifically leads to the defeat of the adversary in a conflict, the text outlines four general paths. Very broadly, none of the four paths is unrelated to the application of military force. One route is through the use of strategem, which, by confusing and misleading the enemy, leads directly to its defeat (63.3.4, 132.2.1). While the text does not specifically identify these strategems as exclusively military or political in nature, the first of the two cause-effect statements is found in a quote from a passage in the first chapter of Sun Zi Bing Fa, where it lists twelve strategems that create political and military conditions advantageous to the use of military force. A second path to victory involves grasping the strategic initiative from the enemy (135.1.1). In context this does not refer simply to maintaining an offensive posture, attacking preemptively, or forcing the enemy permanently onto the defensive. Rather it means controlling the enemy's perceptions of the strategic environment. Sometimes this may involve appearing weak and on the defensive so as to entice the enemy to attack when in fact one is prepared to attack. It may involve frightening the enemy into retreat when in fact one's military strength is insufficient to defeat it outright (Liu 1955e: 133-34). A third source of victory is the correct application of regular {zheng IE) and extraordinary, irregular, or special (qi isf) forces and operations. This comes from understanding the constant dialectical transformations of zheng and qi and the enemy's strengths and weaknesses (62.1.1, 65.2.1, 98.2.3).43 Metaphorically, the use of zheng forces is likened to a mountain; they are immovable, solid, rooted, and overwhelming. The use of qi is likened to lightning that strikes quickly and preemptively with pinpoint accuracy (Liu 1955e: 118). Analogically, the text cites numerous historical cases where the use of both zheng and qi military forces resulted in the defeat of the enemy (ibid., 60, 61, 63, 64). A fourth route is through the use of overwhelming military force to intimidate 42 This refers to understanding "heaven" (climatic conditions), "earth" (the advantages and disadvantages of various types of terrain), the "interior" (the mind and morale of oneself and one's enemies), and the "exterior" (the enemy's military strengths and weaknesses) (Zeng 1986: 151). 43 That is, one uses zheng forces against the enemy's strengths, and qi forces to attack the enemy's decisive points of weakness.

THE PARABELLUM

PARADIGM

93

the enemy into submission (Z185.3). It is hard to tell whether "awesomeness" here refers to some limited demonstration or threat of latent power, or its application in battle. But from the context it is clear that the basis of awesomeness is military power or martiality. It is also evident that martial-based awesomeness is a tool to be directed outwards toward the enemy, while civil or we«-related behavior is directed inwards towards one's own people in order to obtain their willing submission and support. This suggests that external security is rooted primarily in martiality. Finally, victory is also a function of the military training and battlefield discipline of the army. The army's orderliness and its ability to maneuver effectively on command are positively related to victory (64.3.4,64.3.1,124.1.4, Z203.2).44 As for the weakening of the enemy short of military defeat, again the text borrows heavily from Sun Zi. The cause-effect relationships reveal two basic paths. One is a path of pure violence, whereby the application of zheng and qi forces in alternating fashion results in a constantly weak and confused enemy (99.1.1, 99.1.2). Excelling at both the offensive and defensive use of forces also leads to the enemy's confusion about what should be defended and what should be attacked (132.3.3, 132.3.4). A second path relies on the use of military strategem and ruse de guerre to entice or direct the enemy into a disadvantageous military position. Again, the text borrows statements from Sun Zi about taking the initiative and offering the enemy (unspecified) advantages that expose its vulnerabilities (63.3.3, 98.1.1, Z244.2.1). Feigning military "insufficiency" will entice the enemy to attack points that should have been avoided (133.3.3). Conversely, feigning "sufficiency" will cause the enemy to fall back on the defensive (133.3.4). Sun Zi Bing Fa I have left Sun Zi for last because the text presents special problems in any discussion of the role of violence in traditional Chinese military thought. More so than other texts, secondary commentaries on Sun Zi Bing Fa stress the centrality of "not fighting and subduing the enemy" and other nonviolent paths to state security. Given, too, Sun Zi's prominence in the pantheon of military writers, it is not surprising that the dominant interpretations of Chinese strategic 44

At one point in the text, Tai Zong lists a hierarchy of choices and outcomes in military strategy. The highest form of strategy, he quotes from Sun Zi, is "not to fight but to subdue the enemy." The middle level is to fight and win one hundred battles. The lowest is to rely on high walls and deep moats to meet the enemy from a purely defensive position (Liu 1955e: 152). While this would seem to indicate a strategic preference for the nonviolent submission of the enemy, the notion is not part of any clear cause-effect path. There is no indication that this statement negates the pragmatic importance of violence in the attainment of victory. More of this will be said later in the discussion of Sun Zi and violence.

9 4 CHAPTER THREE military strategem £• \

wealkening diminution ofth eenen ly '.., i

\ political, diplomatic strategem

internal rectification, quality of internal rule

N

'** defeat, submission of

the enemy

J V

\

military preparations,' **" the application of force

\

\

/

/

^

security of the state

military morale, mobilization of human and material resources

direct positive causal relationship implied positive causal relationship

Fig. 3.7. Composite Cognitive Map of the Sun Zi BingFa.

culture rest heavily on this standard view of Sun Zi's strategic preferences. It is important, then, to take a closer look at the status and nature of violence in Sun Zi's thought. This involves, in particular, a closer look at the concept of "not fighting and subduing the enemy" and its strategic corollaries, namely "the best strategy is to attack the enemy's strategy, next best is to attack his alliances, next best is to attack his army, and the worst is to besiege his cities"

(shang bingfa mou qi cifajiao qi cifa bing qi ci gong cheng iiJ^fJcI

But first, in the interests of analytical consistency, a look at the cognitive maps and cause-effect relationships in the Sun Zi Bing Fa. When the standard secondary commentaries on Sun Zi note the importance of non- or minimal violence in his strategic thought, this does not mean the text makes a ConfucianMencian argument that internal rectification is the primary basis of security. Among the Seven Military Classics, the Sun Zi Bing Fa is unique in that it is not a text on general problems of statecraft and governance. Rather it is a treatise on war that moves back and forth from the nature of military power to grand strategy, to diplomatic and military strategem, to tactics, logistics, intelligence, and geography. It does not generally address how a ruler should choose officials or how to exhibit his benevolence and magnanimity in governing the realm. Nor, in fact, is there any discussion of righteous war and other justifications for the use of the military instrument. One finds, then, comparatively little comment on the causes of state security in a broad sense.45 45

Kidder Smith argues that from this perspective Sun Zi is the "least culturally specific, the least 'Chinese' of the Warring States'writings" (Smith 1990: 8), though most analysts take it to be precisely the opposite.

THE PARABELLUM PARADIGM

95

But right from the start one does find a fundamental acceptance of the centrality of military power in achieving security. The text begins with the famous maxim that military affairs or warfare affect the survival or ruin of the state.46 This phrasing suggests a recognition that warfare is a common enough human experience that preparations for warfare, as well as the ability to conduct war successfully, are critical determinants of the state's survival. As Shi Zimei comments in terms similar to his annotation of the Liu Tao, "[Whether one is] victorious or defeated [affects whether] the people live or die; [whether] in warfare [outcomes] are as they ought to be or not [affects whether] the state survives or perishes" (Shi 1222: Lib). Given this intimate relationship between warfare and state security, Sun Zi goes on to stress, warfare should most certainly be studied and researched. Indeed, one of the direct causes of state security is the general's understanding of the "way" of warfare (76.21.2). More precisely, the general who truly understands the nature of war is the "arbiter of the people's fate" (min zhisi ming J§ j£ WJ ffp), and "the master [i.e., determining element] of the security of the state" {guo an weizhizhu |Jil5cJfe5l3i) (Wu 1983: 42). Elsewhere the text again underscores the importance of the general and his army by drawing a positive causal connection between the general's assistance in the protection of the state and the strength of the state (81.16.1).47 This does not amount to support for an aggressive, expansionist use of military forces in the interests of state security, however. Sun Zi is, above all, prudent. He is keenly aware of the drain on state resources and internal morale that the extended projection of force can entail. In one of the few general comments on the relationship between the quality of internal rule and state security, the text argues that an enlightened ruler is cautious and prudent when resorting to war. Thus the state is kept secure (143.19.3). This theme of caution in the use of force runs through the handful of other causes of state security. Eschewing protracted campaigns ensures that the state does not exhaust its resources and impoverish its people in times of war (73.4.1, 73.7.1, 74.11.1, 74.13.1). In this way domestic unrest is avoided (144.1.1). These are admonitions against belligerence in the pursuit of state security. They are grounded in an appreciation of the logistical limits on protracted warfare, but they do not add up to a rejection or downplaying of warfare or violence on a priori moral, political, or military grounds. Indeed, as in the other texts, the correct handling of security threats rests in large measure on the defeat of the enemy. In the Sun Zi Bing Fa the causes of 46 Griffiths (1982: 63) and Pan Jiabin (Wu 1990: 263) translate the sentence literally that warfare is "the road to survival or ruin." They take the character dao ($$t) and render it "road." This is problematic since dao could also be translated as "way," as in method or process. This latter interpretation would underscore warfare as a consciously wielded tool or policy that, depending on how it is used, determines the fate of the state. 47 One finds a similar argument in the works of the famous Ming general Qi Jiguang. In one passage in his text, the Lian Bing Shi Ji OH J^-jflB) "As for the way of being a general, [it affects] the peace or danger [security] of the lands within the frontiers." Cited in Fan 1990: 75.

96

CHAPTER THREE

victory are varied. The first is, essentially, intelligence (149.23.1). Victory comes in part from correctly estimating the political, climatic, geographical, and organizational conditions each side faces (65.11.1, 84.33.1, 129.25.1)—"knowing the enemy and knowing oneself—as well as from intelligence about the enemy's plans and strategies (144.3.1). Another element could be called destructive and constructive morale within the military. The former involves stirring up soldiers' anger and hatred of the enemy (75.16.1); the latter refers to the correct use of rewards and punishments to unite and command one's own forces (123.48.1, 83.29.1, 62.23.2). Much is made in the secondary literature of the centrality of deception or ambiguity (gui dao UGIS)—the so-called indirect approach—in Sun Zi's work. Sun Zi's text is exalted as the quintessential historical work on strategy on this score, and is usually contrasted with Clausewitz, who dismissed strategems, cunning, and deceit as being too costly logistically and of "little strategic value" (Clausewitz 1976: 202). To be sure, Sun Zi does elevate the strategic value of deception and strategem beyond the level in Clausewitz, but there is nothing in his cause-effect relationships to suggest that deception and strategem— whether political/diplomatic or military—are distinct direct routes to the defeat of the enemy, and separate from the application of violence. In one instance the text suggests that responding to the enemy's nature and deceiving him, in combination with the concentration of military forces for attack, leads to the defeat of the adversary (139.56.1). In no other instance does strategem or deception lead directly to the enemy's submission. Rather, in several places Sun Zi argues that it is the application of violence that directly causes victory. Attacking the enemy where it is unprepared, or where it does not provide adequate defenses leads to certain victory (69.26.1, 96.7.1). Putting oneself in an indefeatable military position from which to wait until the enemy reveals weaknesses and vulnerabilities prevents one from losing the opportunity to defeat the enemy (87.13.1, 85.5.1, 87.13.3). By first attaining a position of strategic superiority and then attacking, one is assured of victory (67.13.4). Attacking preemptively with speed and inexorable force is also positively related to victory (85.7.2, 88.20.1). Here on the question of the efficacy of force Sun Zi employs a couple of vivid metaphors that seem to underscore an acceptance of massive overproportionality in the use of violence. In one case he describes the general's attack as though he were descending from the "nine levels of heaven" (85.7.2).48 "Heaven" {Han X ) connotes, in addition to great height and momentum, weather-related characteristics, namely thunder and lightning. We have seen other texts use this imagery to describe preemption, a sudden and massive military strike at an adversary before it can "block the ears" and "close the 48

The number nine in traditional China was an abstract number designating an extreme— e.g., the highest point in heaven.

THE PARABELLUM PARADIGM

97

eyes." In another case (88.20.1), the causal link between an attacking army and victory is likened to the release of pent-up water into a bottomless abyss, and to two weights balanced against each other, the victor weighing 480 times the vanquished.49 Elsewhere—while not causally connected to victory per se— Sun Zi likens the application of actualized military power (shi §§) against the enemy's weak points to throwing a grindstone against an egg, to torrential waters tossing boulders out of the way, and to a boulder rolling with inexorable force and direction down from the top of a mountain (Wu 1983: 65). It is, in the end, this act of attacking the enemy with massive force that directly creates the possibility of winning (ke sheng gong ye Rl!$£$C"ttL) (85.5.2). It is only in the process of getting to the point where one is attacking from a position of superiority that strategem plays a role. That is, political and military strategem and deception are elements in weakening, or attriting, the adversary short of its outright defeat and submission. For instance, in one passage in his first chapter Sun Zi begins with his well-known statement that "as for wielding the military instrument, it is the way of deception" (Wu 1983: 29). He then goes on immediately to list twelve examples of strategems that could be used against the adversary. One can, for example, feign weakness (66.16.1); or appear to be closer to or farther away from the enemy than one actually is (66.19.1); or offer the enemy some sort of advantage to entice it to act (66.20.1, 75.17.1); or vex and confound the enemy's general when he is easily angered (67.22.1); or encourage the enemy's arrogance by appearing lowly and subservient (67.23.1); or draw close to the enemy (or his advisers) and create splits, divisions, and doubts within the enemy camp (69.26.1). Some of these strategems appear purely political and diplomatic, some involve military actions, and some are ambiguous. But the important point is that the passage is closed off by two statements that could be interpreted plausibly as admonitions to use force, namely "attack where the enemy is unprepared [to defend], and appear [with military forces] where the enemy does not anticipate it." The effect of these twelve strategems, then, is that the enemy becomes unprepared and unaware. Once these vulnerabilities appear, then force is applied to inflict decisive defeat. The primary purpose of strategem is to control the movements and actions of the enemy and to create political, military, and psychological boundaries of action within which the enemy must operate. This creates weaknesses, openings, and opportunities to be exploited with military force. Thus, "causing the enemy to have form" (xing zhi Jf^^) 50 leads to the enemy's manipulation (93.20.1). Offering advantages, whether political or military, causes the enemy 49 This connection is made by a number of historical commentaries on Sun Zi. See Du Mu's and Du You's commentaries in the Song dynasty collection of commentaries, SBSYJZSZ1978:1, 40a-b, 50 Meaning manipulating the enemy's intentions and capabilities in such a way as to create predicted, patterned behaviors that are favorable to the attacker.

98

CHAPTER THREE

to move in a manner that plays to one's own advantage (93.20.3, 96.3.1, 140.61.1). The converse, of course, is that strategem also aims to obscure and mask one's own movements, such that one has no form (wu xing $$&), or discernible behavioral patterns. In this way the enemy is unable to anticipate when or where force will be used against it (98.14.1, 134.29.2, 137.48.1, 137.45.1). In addition to strategem and deception, the use offeree itself shows up as a source of the enemy's vulnerability. Defensive and offensive skills (97.8.1, 97.8.2); an ability to concentrate forces while attacking and dividing enemy forces (98.13.2); capturing the enemy's strategic strong points (134.28.1); attacking the enemy's weak points (97.10.1); attacking territory or targets that the enemy feels compelled to expend considerable energy and resources to save (97.11.1); and engaging one's numerically superior, elite troops against the enemy's inferior and weaker troops (127.15.1) all lead to the enemy's weakening short of outright capitulation. There is, however, one critical and troubling cause-effect relationship that does not appear to fit this motion from deception to applied force to victory. At one point Sun Zi outlines a direct causal connection between a general's excelling at warfare and his ability to compel the enemy to submit without resorting to war (79.10.1), and to capture enemy cities without having to attack and besiege them (79.10.2). This relationship echoes Sun Zi's often cited statement—found in the same chapter—that to "fight and win one hundred battles is not the supreme skill, but to subdue the enemy without fighting is." From this general concept follows several corollaries—namely, that in any conflict situation the superior strategy is to undermine the enemy's strategy and his alliances. These two strategies are usually interpreted as the use of nonviolent strategem to undermine the enemy's political cohesion and military strategy, and the use of diplomatic means to isolate the adversary. Only under unavoidable circumstances (that is, circumstances dictated by the adversary's actions, not by one's own initiative) should one choose to attack the enemy's army and besiege his cities, or use pure violence to achieve political and military goals. Given the conceptual link between subduing the enemy without fighting and attacking the enemy's strategy and alliances, on the one hand, and the appositioning of attacking the enemy's strategy and alliances with attacking the enemy's army and cities, on the other, it is not surprising that in most of the secondary literature subduing the enemy without fighting is interpreted quite literally as relying on nonviolent and nonmilitary means. This construct not only undergirds the argument that nonviolence is a central concept in Sun Zi's thinking, but also appears to establish an a priori strategic preference ranking where, in abstract terms, violent strategies are less preferred to nonviolent ones. We arrive, then, at a problem. Given what seem to be obvious causal linkages in the cognitive map between violence on the one hand and victory and state security on the other, how do we reconcile these with the

THE PARABELLUM

PARADIGM

99

apparent centrality of a preference ranking where nonviolent strategem is an efficacious cause of these two outcomes as well? This problem requires some elaboration.

O N " N O T F I G H T I N G AND SUBDUING THE E N E M Y "

As noted, there is general agreement among Western and Chinese scholars today that the concept of "not fighting and subduing the enemy" is one of the core notions in Sun Zi's strategic thought, and by extension, in Chinese strategic thought as a whole. There is also general agreement that u not fighting and subduing the enemy" leads to a preference for attacking the enemy's strategy and alliances. This is usually interpreted to mean the use of nonviolent strategem. Finally, there is also agreement that the notion of "not fighting and subduing the enemy" infuses many of the other classic texts. One finds in the Liu Tao, for instance, the passage "One who excels in war does not rely on sending forth the army . . . the superior way of warfare is not to engage the enemy" (Liu 1955d: 72). The Wei Liao Zi argues that if a state is correctly ordered— that is, farmers are content in their fields, merchants content in their shops, warriors and officials content and effective in carrying out their duties—this will translate into military awesomeness, thus allowing for the pacification of the realm without "the blades [of war] being stained with blood" (bing bu xue ren jS^lftlJJ) (Liu 1955c: 32). Liu Yin suggests that the passage in the Si Ma Fa—"use benevolence as the way of victory"—means winning over the enemy through the extension of benevolence. This was the ancient "way of not killing" (bu sha zhi dao ^f^^M) (Liu 1955b: 19, 50).51 Concepts, phrases, and passages such as these have sustained a view in much of the secondary scholarship that Chinese strategic culture downplays the efficacy of violence, that it exhibits a marked "disesteem of violence." This essential characteristic is often juxtaposed explicitly with an alleged Western or Clausewitzian tradition where war is seen as "an act of violence pushed to the utmost bounds" (Lin 1988: 33). The starkness of this alleged contrast is evident in a work by Wei Rulin and Liu Zhongping, two of Taiwan's top experts 51

Other strategy texts contain similar concepts. See the Tai Bai Yin Jing (>£ fi f#|M) (Chinese Collected Military Texts Editorial Committee 1987b: 453), and the Chronicles of the Three Kingdoms (Wu 1988b: 212). There is a Confucian-Mencian spin on Sun Zi's notion of not righting and subduing the enemy. That does not address the role of nonviolent political strategem in the defeat of the enemy. Rather, it argues that by completely capturing the hearts and minds of the people through good government the ruler can ensure the nonviolent submission of potential enemies. (See the annotation by the powerful late Ming official Zhang Juzheng in Wei 1988: 101.) The suppression of threats with physical coercion does not resolve the basic security problem; it still leaves people with resentments that only proper rule can alleviate (Shi 1222: 3.1b). This is not a narrow argument about the efficacy of deception over force.

100

CHAPTER THREE

on the history of Chinese military thought. In a summary chapter on the special characteristics of Chinese strategic thought, they provide a table that contrasts Chinese and Western concepts along several dimensions. While the Western nomadic/commercial culture traditionally stresses expansion and acquisition through war, China's agricultural, sedentary culture has traditionally emphasized defensiveness, strategem, and the search for peace. Whereas the West stresses humankind's struggle for mastery over nature, Chinese culture stresses the harmony between the two, and hence noncontention. Western culture emphasizes achievement, advantage, and the use of physical force to compel the submission of others, while Chinese culture stresses benevolence, righteousness, and the use of virtue to extract their voluntary submission. In the West a typical symbol of attitudes towards conflict is boxing and contests of martial strength. In China it is taiji quan and the use of softness to overcome hardness. Finally, Western culture stresses the primacy of military warfare, while Chinese traditions stress the primacy of political warfare (Wei and Liu 1985: 436, Xu 1983: 126, 134). This facile comparison is, perhaps, too easy a target. But it reflects in extreme a commonly held stereotype of the differences between Chinese and Western strategic cultures. The cognitive maps and symbolic analyses seriously undermine these standard interpretations, but this seems to contradict the nature and centrality of "not fighting and subduing the enemy" in Sun Zi and many of the other texts. There are three sets of assumptions, however, that underlie the conventional wisdom, and each deserves closer scrutiny. The first is that "not fighting and subduing the enemy" is central to Sun Zi's strategic thought. The second is that "not fighting" indeed means not fighting. And the third set are the assumptions that "attacking the enemy's strategy and alliances" (fa mou, fajiao {JclStfe5C) only involve nonviolent methods. As should be evident, I argue that all these assumptions are not unproblematic. Generally speaking, it is ambiguous whether "not fighting and subduing the enemy" is the essence of a nonviolent or minimally violent core in Sun Zi, and my analysis suggests that even this axiom and its strategic corollaries embody elements of applied violence. Concerning the first assumption, what indeed is the conceptual status of "not fighting and subduing the enemy" in the Sun Zi Bing Fa and the other classics? If we assume that this concept and its corollaries constitute, in essence, a nonviolent strategic-preference ranking and decision calculus, and if we assume that Sun Zi is a handbook of sorts on strategy, a source of broad axiomatic guidelines for strategic decision making, then what role does "not fighting and subduing the enemy" play as a decision rule in strategic choice? There is some evidence that the military texts as a whole view "not fighting and subduing the enemy" as an idealized form of strategy, and relegate it to a golden age of sage kings and legendary rulers. Thus it has less relevance to the pragmatic, operational problems to which the texts devote much of their space. As noted earlier,

THE PARABELLUM

PARADIGM

101

the Si Ma Fa outlines a linear evolution in strategy from the nonuse of force to the use of violence in a discussion of the legendary rulers Xia Hou Shi, Yin Tang, and the sage kings Wen and Wu of the Zhou dynasty. The San hue also describes a linear evolution in the way in which the security problematique was addressed down through the ages. The five legendary emperors Shao Hao, Zhuan Xu, Gao Xin, Tang Yao, and Yu Shun pacified the realm through personal example, the employment of able, gracious ministers, and the enculturation of the people. The three sage kings Yu, Tang, and Zhou Wen governed by means of the dao such that everyone was content with their position in life, rules and rites were used to preserve this order, the people were thus united, and though military forces existed there was no need to employ them. The hegemons of the Zhou dynasty, however, ruled by political and military power, and used rewards and punishments to preserve social order (Liu [SKQS]: 726.114-726.115). In perhaps the most detailed and thoughtful historical annotation of the Sun Zi Bing Fa, Shi Zimei also implies that nonviolent strategem is essentially historically bounded and idealized. He suggests that the ability to not fight and subdue the enemy essentially belonged to the ancient sages Yao and Shun, and to the Zhou kings Wen and Wu. (Shi differs with the Si Ma Fa on who could and could not subdue the enemy without fighting, but the allegorical principle remains the same.) "The best of the best is certainly not warring. Only Shun could do this" (Shi 1222: 3.2a). Since Shun's time, according to Shi, there has been no one capable of implementing this strategic principle. Famous generals have all had to use violence and to fight protractedly in order to defeat an enemy. This, he concludes, has been a "constant principle" (ibid., 3.5a).52 At a more mundane level, the notion that "not fighting and subduing the enemy" is the core of Sun Zi loses sight of what the rest of the text tells a strategist to do: the last eight of the thirteen chapters explore the principles of mobile warfare—attacking, defending, and invading other states under maximal 52

The fourth century B.C. Legalist scholar-official, Shang Yang, also argued for a linear view of strategic history when he concluded that strategy was historically bounded. The idealized, nonviolent methods used by the legendary Shen Nong to rule were not appropriate by the time of the Yellow Emperor, since conditions had changed. Power relations between individuals and groups had appeared, and often these could only be resolved through punishment and war (He 1988: 142). Similarly, Li Quan, the author of the Tang text, Tai Bai Yin Jing, presents a linear account of the shift from the reliance on virtue and benevolence for ordering the world to the use of coercive power. While he considers the latter an inferior method of governance—it ensures no enemies, but does not ensure permanent order—it was not condemned per se, but was viewed as a valid method within a particular context (Chinese Collected Military Texts Editorial Committee 1987b: 445-46). One of the few contemporary authors who implicitly downgrades the centrality of "not fighting and subduing the enemy" argues that Sun Zi's thought itself was bounded by historical strategic circumstances of the Warring States period. Xu Zihong suggests that Sun Zi reflected the consciousness of the period, which was belligerent, glorified martiality, and exhibited a preference for active offensive strategies; his thought was best suited to the pursuit of hegemonic power (Xu 1990: 104).

102

CHAPTER THREE

geographic, logistic, and tactical conditions. Chapters 4 and 5 present more abstract discussions on how to control the strategic relationship with the adversary, prior to the sudden and decisive application of violence. This violence is described metaphorically in terms that underscore overwhelming, out of proportion, inexorable force leading to the enemy's destruction. It is really only in the first and third chapters that one finds strategic axioms that might be construed in the standard way as stressing "not fighting and subduing the enemy." But, the cognitive map indicates, these are not necessarily central to the text, nor to traditional Chinese strategic thought in general. If the military classics as a whole viewed "not fighting and subduing the enemy" as an idealized and historically bounded decision axiom, what then were the operative strategic decision rules in these texts? I will deal with this in detail in the next chapter, but the more dominant decision rule is "absolute flexibility," embodied in the notion of quan bian OfHH).53 That is, given that constant change is the key characteristic of conflict situations, a strategist must be prepared to adapt to dangers and opportunities as they suddenly appear. The strategist cannot be restricted, constrained by, or wedded to self-imposed a priori political, military, or moral limits on strategic choices. Whereas "not fighting and subduing the enemy" as a decision rule implies an a priori strategicpreference ranking in which nonviolent methods are preferred, the notion of quan bian lifts this restriction, since the nature of conflict requires an ability to transcend fixed responses to particular contingencies. Quan bian in effect states that when facing a contingency, choose any and all actions that will achieve one's goals. One could argue, then, that the essence of strategic choice in the military texts is not "not fighting and subduing the enemy" but "respond flexibly to the enemy and thus create conditions for victory" (yin di erzhi sheng 0j§£

mmmv4

Turning to the second assumption, the argument about the idealized status of "not fighting and subduing the enemy" accepts the notion that the concept and its strategic corollaries are nonviolent and nonmilitary in nature, and that these embody a strategic-preference ranking that places nonviolent strategies at the top. Arguably, however, this premise can also be questioned. Here there are a couple of interpretations worth noting. Does "not fighting and subduing the enemy" mean "not fighting" as usually interpreted in English and modern Chinese? One interpretation is that "buzhan" (/f§jt%), or "not fighting," can be read in classical Chinese as "wei zhan\^. !£), a vaguer, more relative negation meaning "having not yet fought" (Interview 1991b, Yang 1986: 34). Under this interpretation the phrase implies some53 Or "sui quan ying bian" (WWLMW), meaning responding to changing circumstances according to a weighing of all relevant factors in a strategic situation. 54 This notion of absolute flexibility is nested within an overall expectation that violent solutions to security problems are more efficacious. Thus quan bian undermines an a priori nonviolent preference ranking, but not necessarily a violent one.

THE PARABELLUM

PARADIGM

103

thing along the lines of "first set up maximally advantageous conditions for the defeat of the enemy before fighting, and only then fight and defeat the enemy." That is, fighting is still required, violence must still be applied, but it is done so minimally (in relative, not absolute terms) since the objective conditions for the enemy's defeat already exist. This accords with another comment in the Sun Zi Bing Fa, "first achieve victory and only then go to war" {xian sheng er hou zhan 7fc|]§Tfff in !£)> where the character for victory (sheng $&) has the more nuanced meaning of a superior position from which victory is assured (Interview 1991b).55 This interpretation throws a somewhat different light on the strategic corollaries "attack the enemy's strategy," "attack the enemy's alliances," "attack the enemy's army," and "besiege the enemy's cities." Conventionally, these are viewed as a ranked set of preferences moving from best to worst choices. If, however, "not fighting and subduing the enemy" means "first weakening the enemy and then fighting," then the operational strategies used to achieve this outcome can be seen not as a ranked set of preferences, but rather as one of two things: (1) a menu of strategic choices; or (2) a temporal sequence of strategies.56 The view that these corollaries constitute a menu of choices implies that under different military, economic, political, diplomatic, and psychological conditions, a decision maker chooses different strategies. This interpretation accords with the notion of quan bian. The nonviolent strategic preferences are therefore highly contingent, even to the point of being irrelevant, rather than a priori. Nonviolent means are not intrinsically better than violent strategies, but only under certain specific, limited conditions. If these corollaries are seen instead as a temporal sequence of strategies, 55 This interpretation is similar to the notion found in the Wen Zi: "One who excels in the use of the military instrument first weakens the enemy and only then fights" (xian ruo di er hou zhan 5fci!8fcrFD JalK) (Qiu 1985: 248). For a recent interpretation of this phrase as "first fighting and then subduing the enemy," see Wang Sanxin's study of Mao's military thought. '"Not fighting' should be flexibly explicated. One cannot accept that 'not fighting' means not 'firing a shot' Sometimes one must first fight and then compel the enemy to submit" (Wang 1988: 139). 56 This latter possibility was admitted in two interviews I conducted with researchers from the Academy of Military Sciences in Beijing (Interview 1990c, Interview 1991b). The problem with these two interpretations, admittedly, is with the linguistic construct of the four strategies. The phrase begins with the term shang bing ( _ h ^ ) , where shang usually translates as "highest," "top" or "best." The last strategy is prefaced with the term qi xia C R T ) where xia means "lowest," "bottom" or "worst." This would certainly imply a preference ranking. One can at times give these terms a temporal flavor, interpreting shang as "before" and xia as "after," which would be consistent with the second interpretation. The entire phrase would more obviously constitute a menu of choice if each strategy were prefaced with the term huo (jgK), meaning "sometimes" or "or." In this sense, the first interpretation is the weaker of the two. On the other hand, as I indicate in chapter 7, some Ming strategists used the corollaries as only one element in a long list of strategic options available to deal with Mongol threats along the northern border. This suggests that the corollaries were used as a menu of sorts. In other words, Chinese strategists themselves have been willing to decontextualize these corollaries and ignore the strict linguistic constructions of the phrases.

104

CHAPTER THREE

then war becomes a process of defeating the enemy. First one attacks the enemy's strategy, then one attacks its alliances. Once these moves weaken the enemy and create conditions for its defeat, one then attacks its army and besieges its cities. This interpretation is consistent with the relationship Sun Zi sets up in the first chapter between deception and nonviolent strategem on the one hand and the application of military force on the other. His twelve nonviolent strategies—some of which could be categorized as attacking the enemy's strategy and his alliances—act as inputs into the successful application of force in places where the enemy is not prepared.57 This interpretation is also consistent with most of the cognitive maps in the other military texts; these indicate that nonviolent strategem is not a separate route to victory. If "not fighting and subduing the enemy" in its conventional form were indeed the central decision rule in the military texts, one would expect to find direct causal paths from this concept and its strategic corollaries to desirable ends, such as the defeat of the enemy and the security of the state. Yet, in general, nonviolent strategems (along with violent ones) lead only to the weakening, confusion, or diminution of the enemy. In almost all the texts the defeat of the enemy at this point requires the application of violence. In sum, "attacking the enemy's strategy, etc." does not necessarily represent an a priori preference ranking that places nonviolence at the top. Rather, it can be viewed as an idealized preference ranking, a contingent one, a menu of choices where violence is not necessarily the least-preferred route to security, or a temporal sequence of actions. Turning finally to the third assumption, the question of whether "not fighting and subduing the enemy" and its corollaries represent a menu of strategic choices or a temporal sequence sets out from the common assumption that "attacking the enemy's strategy and alliances" is indeed nonviolent, and that "attacking the enemy's army and cities" is indeed violent. This premise, too, can be reexamined. Most scholars define "attacking the enemy's strategy" as using spies, sowing disunity, and weakening the enemy's will in order to undermine the enemy's strategy. Most define "attacking the enemy's alliances" as using diplomatic means to isolate the enemy and cut it off from allies. One confounding problem is that Sun Zi does not explicitly elaborate the meaning of "attacking the enemy's strategy" and "attacking the enemy's alliances." This leaves the notions open to interpretation. And in fact, historically there were differences from presentday intepretations. Both Cao Cao's and Du You's annotations imply that "attacking the enemy's strategy" means essentially the preemptive use of violence against the enemy before it can implement its plans to attack (SBSYJZSZ 1978: 1.25a). Another historical annotation suggests that "attacking the enemy's strategy" means once the enemy's strategy is revealed, one must respond to it, devise a counterstrategy and then attack (note the contingent nature of this 57 It is logically possible to hold that these four strategies constitute a menu of choice that from time to time should be implemented sequentially. The opposite construct is more difficult to hold.

THE PARABELLUM

PARADIGM

105

interpretation). While the text is ambiguous, it does not at least limit this counterattack to nonviolent means (ibid., 1.25b). This raises an important point, namely that a key element in undermining the enemy's strategy involves causing it to be revealed so that one can then devise an effective counterstrategy. The question is whether this process is necessarily nonviolent. Sun Zi makes a somewhat abstract argument in chapter 6 that force can be used to provoke a reaction from which to gauge the enemy's strengths and weaknesses. The chapter is in part a discussion of how to evaluate and exploit enemy weaknesses. In a summary of the process, the text notes that by applying military pressure, thus compelling the enemy to react, one can understand the principles behind its responses. And by butting against or clashing with the enemy (jiao ^=J), one can then understand its strengths and weaknesses. In sum, the process of undermining an enemy's strategy need not be limited to nonviolent methods. As for the notion of "attacking the enemy's alliances," once again historical annotations and commentaries offer a more complex formulation than the standard contemporary ones. A Tang commentator and author of texts on strategy and statecraft, Li Quan, argues that it means using force to attack the enemy's potential alliance partners before an alliance can consolidate (ibid., 1.26a). Shi Zimei suggests that it means using force preemptively to attack the enemy's existing allies before they can act in coordination to come to the main enemy's rescue (Shi 1222:3.2a). Sun Zi does not really explicate the process of "attacking the enemy's alliances." But in chapter 11 the text does offer a hint of what the strategy entails. One passage seems to indicate that the application of "awesomeness" (wei Js£) or overwhelmingly threatening, though latent, force is necessary to prevent the enemy's alliances from consolidating. The question is what is the nature of wei? Is it, in effect, a policy of flaunting force and intimidating the enemy without the use of violence? Or is it the application of overwhelming military power (wei li $£3]), as one annotator suggests (SBSYJZSZ 1978: 2.28b)? There is nothing in the notion of wei that excludes the use of force or violence. Indeed, in traditional Chinese strategic thought the selective, careful application of limited amounts of force for demonstrative purposes was often viewed as a necessary step in building up, consolidating, and bringing wei into play. Again, any conclusions about the elements of violence in "attacking the enemy's strategy" and "attacking the enemy's alliances" are necessarily tentative. But what this discussion suggests is that these strategies do not simply constitute a preference for nonviolent strategies. Rather, at each level of strategy there is a potential element of applied violence. From this perspective, then, "attacking the enemy's strategy" and "besieging the enemy's cities" are not polar opposites, but embody different proportions of violence.58 58

This conclusion can hold whether these four strategies are ranked as preferences, are a menu of choices, or are arrayed as a temporal sequence.

106

CHAPTER THREE

CONCLUSION

This chapter has focused on three interrelated questions the answers to which, a model of strategic culture would suggest, constitute a central paradigm from which should flow logically connected grand strategy preferences about how to manage external conflict. These questions concern the role of war in human affairs, the nature of the adversary, and the role and efficacy of violence or military force. The central paradigm can, at its extremes, take one of two ideal forms. One is characterized by assumptions that war is aberrant and preventable; that conflict is variable sum and that the enemy has a price; and that highly coercive strategies are generally the least efficacious and are last resorts (or at least not first resorts). An alternative paradigm assumes war is inevitable or extremely frequent; that war is rooted in an enemy predisposed to challenge one's own interests; and that this threat can best be handled through the application of superior force. I have suggested that most students of Chinese strategic thought would identify the first paradigm as fitting more closely with traditional Chinese strategic beliefs. While I do not deny that elements of this central paradigm are found in traditional Chinese strategic thought, there are also elements in this tradition that fit as closely, if not more so, with the second paradigm. In general, within the Seven Military Classics the key elements of the first paradigm are unrelated, or are not directly connected causally, to the security of the state. War is considered a relatively constant characteristic of the human condition. Whether one resorts to war or not depends on the adversary. It is the enemy's disposition that determines whether one faces a security threat. This disposition to war is, by definition, unrighteous, since the moral order requires one not to threaten the security of another. One's own resort to force, therefore, is not only legitimate and necessary, it is not bounded by any a priori moral limits, since the enemy is beyond the pale. The utter defeat of the enemy requires the application of superior force. An analysis of the cause-effect statements in the military classics suggests that violence is highly efficacious and that nonviolent strategies in general do not lead directly to the submission of the enemy, but that they are at most a prelude to the application of overwhelming military power. The role of force is described well by the metaphors used in the texts to illustrate the effect of a general and his army on the adversary. In five of the seven texts, when nonmilitary or civil (wen) sources of security are discussed, they affect security indirectly by first having a positive effect on morale in the army and on the awesomeness of the state and its ruler. High morale and awesomeness, in turn, have a positive effect on the ability to defend the state against external threats. Only in the San Lue and the Liu Tao does one find civil causes of state security (i.e., causes related to internal rectification) that seem unrelated causally to martial-based inputs. All seven texts indicate that preparations for war, knowledge of the "way" of warfare, and the

THE PARABELLUM

PARADIGM

107

resort to war are positively related to the achievement and preservation of state security. None of the texts argues that the application of political and military strategem only leads directly to state security. In all the texts the use of violence is identified as a direct cause of victory over the enemy or the enemy's capitulation. In only two texts—the San Lue and the Wei Liao Zi—is there a direct, positive relationship between internal rectification, or the quality of internal rule, and the capitulation of the adversary. But in the Wei Liao Zi the implied link is between internal rectification and the state's ability to mobilize resources for war. In the San Lue the implied link is between morale and the willingness of the people and army to sacrifice for the state. In only two texts are there cause-effect relationships that might suggest that nonviolent diplomatic actions can lead directly to the defeat of the adversary. But in the Wei Liao Zi the statement comes in the context of a discussion of "victory by means of the dao" As I suggested, this form of victory was an ideal one, and not of central concern in the text. In the Wen Dui it is ambiguous whether the reference is to nonviolent diplomatic strategem or to military deception. From the context, it appears the text is referring to Sun Zi's passage on deception, which lists twelve political and military strategems that are, in effect, preludes to the application of violence. What then can we conclude about the central paradigm in Chinese strategic culture? What is striking about the answers provided by the cognitive maps and symbolic analysis to the three questions raised at the beginning of this chapter is how closely these parallel the parabellum paradigm within Western traditions of realpolitik behavior. Parabellum stands for the well-known concept, si pacem parabellum ("if you want peace, prepare for war"). Linguistically, the phrase has its parallels in the Chinese terms, "thinking about danger and threat while residing in conditions of peace" (Ju an si wei jlf ^ c ^ f a ) and "with [sufficient] military preparations there will be no calamities" (you bei wu huan W f S t t ^ ) - 5 9 Parabellum, to borrow Vasquez's terms, is a "set of practices," rooted in a particular view of the world that has to be "culturally acceptable" to show up behaviorally (Vasquez 1987: 369, 372).60 The parabellum paradigm has a number of elements. It proceeds from an assumption that threats to A's peace and security come from B's capability and inten59 Both idioms are found in modern Chinese, but their origins appear to be the Confucian classic, the Chun Qiu Zuo Zhuan. "The Book [of History] says, 'When residing in peace, think of danger. If one thinks of danger, then one will make military preparations. With military preparations, there will be no calamity" (Legge 1935: 451). The argument, made in somewhat different terms, is found in the even older text, the Yi Jing (or Zhou Yi)\ "the ruler is secure but does not forget about threats/dangers; he survives, but does not forget about perishing; he orders [the realm] but does not forget about chaos. Thus he is secure and the state and clans can be protected" (cited in Wu 1988b: 95). 60 This is to distinguish parabellum realpolitik as a historically bounded set of practices from realpolitik or realism as an intellectual "theory" of state behavior.

1 08

CHAPTER THREE

tion to use force in a way that could overwhelm A. The environment is highly conflictual due to the "fact" that the adversary is constantly prepared to use force if given the opportunity. From A's perspective, then, the adversary is dispositionally given to challenge A's security. The relationship is, therefore, perceived to be zero sum. These threats from B are best held in abeyance by preparations for war, demonstrations of resolve, and bargaining strategies that are highly coercive, aimed at demonstrating the capacity and willingness to use force, thus denying the enemy the chance to act dispositionally. The assumption is that the more capabilities one has both relatively and absolutely, the more secure one is. The use of violence and variations in the means of violence are always justifiable, since it is the enemy's constant threat to resort to force that challenges the moral order. Moreover, a solution to this security problematique ideally involves the elimination of the adversary, since resolution short of this does not change the enemy's disposition, and hence does not eliminate the threat. As Vasquez writes, realpolitik as a set of strategic practices is characterized by "perceptions of insecurity (the security dilemma), struggles for power, the use of Machievellian strategems, the presence of coercion, attempts to balance power, and the use of war to settle disputes" (Vasquez 1983: 216, Vasquez 1987: 369, and Vasquez 1993: 114).61 The parabellum paradigm appears to pervade the causal relationships and argument structures in the Seven Military Classics. But a perplexing problem is how or if Chinese strategic thought resolves the apparent tension between this paradigm and the more benign Confucian-Mencian one. The immediate question, however, is what sort of logically derived grand strategic-preference ranking emerges from this central paradigm. 61 See also Leng 1983: 381. Not surprisingly, the parabellum "mindset" is associated with the values and beliefs inculcated into military organizations. See, for example, the answers to questions on what matters in war (question 32), the role of morality in determining foreign policy (question 33), and the moral limits on military options (question 4) in Brunk et al. 1990: 105, 107. The parabellum paradigm also lies at the heart of "myths of expansion and empire." As Snyder argues, a powerful myth created to justify and perhaps motivate imperial expansion is a "faith in threats"—that is, a faith in the efficacy of military force to compel the submission of the external threats (Snyder 1991: 5). I am also struck by the similarities between Chinese parabellum and Stuart's description of the dominant American war myth that evolved in the early years of the republic. This myth was characterized by a "belief in the inevitability of war, in the defensive-mindedness of Americans, and in the need for force as an instrument of policy. Arguments for preparedness flowed naturally from this" (Stuart 1982: 156). See also Johnson's analysis of American preference for "strategies of annihilation," in particular the relationship between beliefs in the morality of the war cause, the demonization of the adversary, and justifications of no-holds-barred uses of military force (Johnson 1981: 259-62). On the inherent nature of conflict in the realist worldview see Smith 1986: 4-16 and Wight 1991.

Chapter Four CHINESE STRATEGIC CULTURE AND GRAND STRATEGIC PREFERENCES

T

HE NEXT STEP in the process of determining the existence and influence of strategic culture is to reexamine the strategic texts to see whether any strategic-preference ranking exists, and whether this ranking is consistent with the central paradigm. I begin with a discussion of what grand strategies one should expect to find ranked or unranked in the texts, and suggest a working typology of strategies that strategic culture, in principle, ought to rank in some order. I then go on to look at each of the Seven Military Classics for evidence of a grand strategic-preference ranking consistent with the parabellum paradigm. The rankings are compared in order to gauge the degree of congruence across texts. As I outlined in chapter 2, the degree of congruence is the basis of any conclusion about the existence of a strategic culture. The results are both reassuring and less so. On the one hand, there is evidence that the texts embody a preference for offensive, preemptive uses of force against the enemy, as the parabellum paradigm would suggest. On the other, within the broad preference for more coercive grand strategies, there is some ambiguity whether this translates into an across-the-board preference for political and territorial expansion through offensive uses of force.

A T Y P O L O G Y OF G R A N D STRATEGIES

If the influence of strategic culture on behavior begins with its effect on how decision makers rank grand strategic preferences, we have to know first what grand strategic preferences exist. Some of the literature implies that strategic culture ought first and foremost to place a priori limits on the range of strategic options considered. That is, what distinguishes one strategic culture from another are differences in the types of grand strategies up for consideration. This argument fits with Laitin's contention that culture does not provide preferred ways of behaving, but rather shared "points of concern." The positions taken on these points of concern may be different, but the range of points is uniquely limited to a particular culture. He argues, for instance, that what distinguishes American political culture from, say, that of Neo-Confucianism in ancient China is that the fundamental questions of politics are different. In the former, political debate derives ultimately from contention over the balance of rights and

1 10

CHAPTER FOUR

demands the individual and the state can legitimately claim from the other. In the latter, the rights and demands of individuals simply do not exist as a point of concern or contention (Laitin 1988: 589-90). This is not the place to debate the merits of Laitin's conceptualization of political culture, nor his characterization of American versus Neo-Confucian points of contention. But there are a couple of objections to this approach in the realm of strategy. First, defining strategic culture as a culturally distinct, a priori, limited but unranked range of strategic choices makes it difficult to derive culturally distinct sets of predictions about what choices will be made over time. Without these distinct sets of predictions one faces considerable difficulties testing a strategic-culture model of choice against noncultural models. Second, and more important here, it is not at all clear that in the realm of grand strategy that the potential range of options is great enough for there to be significant differences across societies in the kinds of strategies considered by decision makers. It is, of course, an empirical question whether, to the extent that grand strategy addresses timeless and universal problems in the management of threats to political hegemony and territorial control, one should expect little cross-cultural variation in the number and type of strategic options. Certainly when Ming decision makers identified their range of options or strategic practices these included all of the same general behaviors available to British, French, German, and Japanese decision makers in the nineteenth and twentieth centuries (see Kupchan 1994). If strategic culture is to have an effect on choice, it is more likely through a distinctive ranking of universal options than the a priori exclusion of certain options. Just how these options should be categorized, however, is none too clear in the international relations and security literature. Much of this literature categorizes military doctrine (i.e., deterrence, defensive, offensive) (Snyder 1984, Posen 1984), or operational strategy (i.e., attrition, blitzkrieg, limited aims) (Mearsheimer 1983). But there is generally far less specificity about the categories of grand strategies available to decision makers.1 In part this is probably because there is comparatively less literature on the question of grand strategy. In part, too, it reflects the multiple spheres or arenas of state activity that are encompassed by grand strategy. Any typologies of grand strategy can, in principle, touch on economic, political, and/or military-strategic actions, •A recent volume edited by Paul Kennedy, Grand Strategies in War and Peace, provides a number of very useful historical case studies of different states' grand strategies (Kennedy 1991). Nowhere, however, is there any sustained discussion of types of grand strategies. Evangelista's essay in Rosecrance and Stein's edited volume on grand strategy provides a list of different kinds of strategies, but there is no detailed explication of the content and boundaries of these (Evangelista 1993: 163). Desch discusses four types of strategic interests of major powers, but does not really suggest how particular grand strategic means relate to these particular ends (1993: 10-11).

GRAND STRATEGIC PREFERENCES

111

each of which can have their own categorizations. Finally, the absence of generally accepted typologies reflects the ambiguity at this level of strategy about whether ends or means are the primary determinants of the content of grand strategies. Categories of political goals or ends may require different labels than categories of grand strategy. There is, however, a small literature that does attempt to provide typologies of grand strategy. Luttwak (1987: 182, 192) suggests two kinds of grand strategy, both of which subsume political goals and grand strategic means under one concept: one type seeks to maximize power through military and territorial expansion; the other aims to preserve the political and territorial status quo and to increase internal levels of wealth, usually through diplomacy. Luttwak therefore implicitly accepts a simple dichotomy between expansionist or revisionist ends and means on the one hand, and status quo ends and means on the other. The problem with Luttwak's typology is that it excludes a range of grand strategic means that could, in principle, be used to pursue either grand strategic end. A (self-perceived) status-quo power could use highly coercive, expansionist means in a conservative, restorationist effort to eliminate rising threats to its hegemony.2 A revisionist state could, in temporarily expedient circumstances, resort to diplomatic negotiation, alliance building, and/or political subversion, all the while attempting to alter fundamentally the relative distribution of power within the international system or subsystem. Neither of these possibilities fits into Luttwak's conceptualization.3 Thus, his typology mixes political ends and grand strategic means, or at least leaves them indistinct. The result is that certain other combinations of ends and means are excluded from the typology. By leaving out logically and empirically probable alternative mixes of ends and means, Luttwak's categorization is not exhaustive, a critical element of any typology.4 An alternative typology comes from Kupchan's work on the overexpansion of empires. He identifies three basic strategies: compellence, deterrence, and accommodation. "Compellence," he writes, "is a coercive strategy aimed at forcing the adversary through punishment or the threat of punishment to acquiesce to specific demands." Deterrence is "a strategy of dissuasion aimed at 2 The United States in the post-World War II period seems to fit in this category of an expansionist, self-perceived status-quo power. 3 Elsewhere Luttwak offers a somewhat different typology, though it too runs into similar problems. In a book on the grand strategy of the Roman Empire, Luttwak outlines three systems of security that integrate military, economic, and diplomatic policies for certain political ends: one is "hegemonic expansionism," the second is "territorial security" and the last he terms "sheer survival." Preferences for each system reflect separate worldviews or self-images (Luttwak 1976:3-4). 4 It also violates another criterion for typologies—namely, the exclusivity of categories. It is not altogether clear from the description what the difference is in terms of either ends or means between "territorial security" and "sheer survival" as systems of state security.

1 12

CHAPTER FOUR

preventing the adversary through punishment or the threat of punishment from undertaking specific actions." Accommodation, Kupchan suggests, "is a conciliatory strategy aimed at moderating the adversary's behavior through concession and compromise" (Kupchan 1994: 67-68). These, he implies, are the universe of grand strategies available to decision makers. Kupchan's categories are more satisfying than Luttwak's if only because they include a wider range of plausible grand strategic choices, and because he does not try to limit specific choices to specific state goals. But his typology still begs a couple of questions. First, how useful is the distinction between deterrence and compellence? Compellence, for instance, can be used at time t to reinforce the credibility of deterrence threats at time t+1. On what basis can one a priori draw a distinction between a state that is acquiescing to specific demands and one that decides not to take a specific action? Acquiescing to a demand, after all, entails a decision not to act in some other way. Second, and more important, how does one distinguish between defensive uses of force and punishment for the purposes of deterrence, for example? Indeed, where do defensive strategies fit in, since deterrence and compellence can both involve threats to use force offensively inside the enemy's territory? As it stands, Kupchan's categories leave out a range of defensive and offensive uses of force designed to deny the adversary its strategic goals. His categorization appears to leave indistinct the difference between three plausible types of grand strategy: (1) highly coercive, expansionist grand strategies aimed at the elimination of the adversary and the annexation of its territory or the acquisition of its resources; (2) highly coercive, expansionist grand strategies aimed at preserving the political and territorial status quo;5 and (3) less coercive, defensive, or deterring grand strategies aimed at preserving the political and territorial status quo by denying the enemy the ability to challenge it. How, then, might we categorize grand strategy in such a way that (a) the distinction between types, while perhaps blurred at the margins, is generally clear and exclusive, and (b) all or most plausible politico-military behaviors are included such that the typology is generally exhaustive? For the moment let me propose three types of grand strategy. 1. Accommodationist: This strategy relies primarily on diplomacy, political trading, economic incentives, bandwagoning, and balancing alliance behavior, among other low-coercion policies. Security is achieved primarily through informal and formal alliance building, or uni-, bi-, or multilateral concessions. Accommodationist grand strategies imply that the ends of policy, while not necessarily well defined, exclude the physical and political elimination of the adversary and the annexation of its territory. 5 This type of grand strategy would be typified by extended punitive military campaigns, followed by withdrawal to preexisting boundaries. It may or may not involve efforts to overthrow the adversary's political leadership and install a friendlier, more compliant regime.

GRAND STRATEGIC PREFERENCES

1 13

2. Defensive: This grand strategy is more coercive in nature than an accommodationist strategy. It relies primarily on static defense along an external boundary. The use of force is not designed to annex territory or to destroy the political leadership or structures of the enemy state. Security is supplied primarily through the internal mobilization of resources for military purposes rather than through alliance building. Defensive grand strategies imply that the ends of policy are not, at that moment, expansionist or annexationist. This category captures the notion of deterrence through denial or limited punishment. 3. Offensive/Expansionist: This strategy is highly coercive, relying primarily on the offensive, preventive, preemptive, or predominantly punitive uses of military force beyond immediate borders.6 The strategic goal behind the use of military force is total military victory and the political destruction of the adversary, including annexation of at least some territory. As for the political ends of an expansionist grand strategy, these are not necessarily revisionist or imperialist in nature. As noted above, strategic expansionism can also be motivated by efforts to preserve or return to a political or even territorial status quo.7 In fact, this category of grand strategy makes no assumptions about the political ends of the state, though clearly, by implication, if a state did have revisionist or expansionist political goals it would presumably prefer this type of grand strategy over a defensive or accommodationist one. Regardless of political aims, however, there is a common denominator: namely, at a minimum, the elimination of the adversary's military capabilities and, at a maximum, the destruction of the adversary's political capacity as a means of achieving security. These three grand strategies are ideal types. In reality the boundaries, particularly between the latter two, can blur and states pursue mixes and matches. While a combination of accommodationist and expansionist strategies is improbable 8 —a mix of accommodationist and defensive, and defensive and expansionist/offensive strategies is plausible. It is important to note that these grand strategies refer primarily to the means of security, not to the political 6 A defensive grand strategy implies that at a minimum the state unilaterally accepts the legitimacy of its territorial boundaries. Whether the adversary does is relevant only insofar as the issue is a source of conflict. An expansionist or offensive grand strategy implies only that temporary territorial boundaries exist; it does not require that either the state or its adversary accept these as legitimate. 7 For instance, a state with an expansionist or offensive grand strategy may seek to preserve its recognized territorial boundaries by reducing bordering states to politically malleable buffers. It seeks an expanded sphere of influence in order to preserve the territorial and political status quo. There was a time when some experts on the Soviet Union made this argument to explain the "defensive expansionism" of the Soviet Union. Others have made a similar argument about American policy in the Western Hemisphere and Asia. See Nijman 1991: 65, 69. 8 This is so if one assumes that states act as if they are unitary actors. Obviously, if this assumption is relaxed and one accepts interest-group, bureaucratic, or factional models of state behavior, then it is quite possible for a state to alternate between the two strategies, and perhaps even pursue them simultaneously.

1 14

CHAPTER FOUR

ends towards which they might be applied. Hence my assumption about the incompatibility of pursuing accommodationist and expansionist strategies. Admittedly the exclusion of explicit ends from this categorization might be somewhat controversial.9 Some might argue that grand strategy, involving the coordinated application of economic, military, and diplomatic tools, by definition subsumes assumptions about political ends since, in the absence of these ends, there would be no guidance or justification for coordination in the first place. States with expansionist aims are likely to adopt expansionist grand strategies, while states concerned primarily with preservation of their political and territorial status quo, or with mere survival in the face of overwhelmingly powerful adversaries, will adopt accommodationist or defensive grand strategies. Critics of a strategic-cultural approach, then, might argue that before one turns to cultural explanations of grand strategic choice, one should look first at a state's preferred political ends. And this choice of ends, from a simple structural realist perspective, is conditioned by a state's position in the system's power hierarchy. A more relaxed structural explanation would suggest that, to the extent that formative cultural or historical experiences did influence grand strategic choice, they would do so only at the margins, after a state's political goals established a preference for particular grand strategic means. I do not think it is all that helpful, however, to subsume political ends and grand strategic means within the same concept. As I hinted above, it seems to me that there is no clear a priori relationship between types of political goals and types of grand strategic means. Envision, for instance, a state's security problematique as a hierarchy of integrated levels of behavior, moving from the basic political goals of the state down through grand strategic means, to military doctrines used to implement grand strategy, down to operational strategies available to put military doctrines into effect (see Fig. 4.1).10 The arrows indicate plausible relationships between the various categories of ends and means. The important point is that both status-quo and revisionist states can, 9 Much of the literature on grand strategy subsumes both notions within the concept of grand strategy. See, for example, the essays by Rice and Kennedy in Kennedy 1991: 147, 168. 10 Here I use goal types that are commonly found in the international-relations literature (see Vasquez 1983: 63). My own view is that this distinction, as it is presently constituted in the literature, is not too helpful for either describing or explaining most state behavior, since there are few clear theoretical criteria offered for determining whether a state is status quo or revisionist or indeed what the status quo between states is at any particular moment. More fundamentally this categorization assumes that a state, or state elites, know what its goals are, and that other states concur with this characterization. Much imperialism—of the classical and the neo-kind— has been motivated, and certainly justified, by efforts to preserve a certain self-defined vision of a state's domestic legitimacy and international status. Morgenthau recognized this problem in defining state goals, but his use of the terms status quo and imperialism (which he substitutes for revisionism) to describe behavior, as opposed to motivation, still does not solve the problems tagged above. For a definition of grand strategy that does include means and ends, see Desch 1993: 3.

Political Goals

Grand Strategy

GRAND STRATEGIC PREFERENCES Status Quo Revisionist

accommodationist

defensive

Military Doctrine

115

expansionist

offensive

Strategy

limited aims

blitzkrieg

(a) temporary strategies adopted for expediency (b) conservative imperialism (c) punitive expeditions (d) strategy designed to capture territory vital for strategic defense (e) compellence designed to reinforce future deterrence credibility

Fig. 4.1. Interrelated Levels of Strategy. under particular circumstances, share any one of the three grand strategies. Status-quo powers are not restricted to accommodationist or defensive grand strategies.11 Nor are revisionist states restricted to expansionist grand strategies, though ultimately it is accepted that at some point these should be pursued in order to expand the boundaries of effective political and territorial control. Thus one cannot necessarily start with a state's political goals and deduce logically connected grand strategic means, nor can one necessarily start with a state's grand strategic means and infer its political ends as revealed preferences over outcomes; ends, in and of themselves, are not necessarily good predictors of grand strategic means. This being the case, it seems safer conceptually to separate the two. A second reason for leaving political ends out of a concept of grand strategy is that by doing so one avoids the tricky problem of determining what exactly a state's goals are. Categories of political ends are highly susceptible to nor11

As an example, Hattendorf's (1991: 19-20) analysis of British grand strategy against France during the War of Spanish Succession, 1702—1713, indicates that Britain employed an offensive grand strategy comprised in part of a military strategy of offensive war of strategic encirclement and attrition in order to preserve its existing political influence in Europe, its trade routes, and its domestic security, all of which had been threatened by the growth of French power.

1 16

CHAPTER FOUR

mative judgments made by strategic elites and scholars alike about what the systemic or subsystemic status quo is, what constitutes revision of this status quo, and the degree to which a particular state hopes to preserve or change this status quo. What one state calls expansionism or imperialism another calls the legitimate defense of national or community interests.12 Categories of grand strategic means, however, are more obviously descriptive (though the problem of normative bias is not entirely absent). One can, in principle, more easily observe whether a state relies on static defense and occasional punitive uses of force short of trying to achieve total victory over an adversary, or whether the state adopts highly offensive means, expands its range of political and military control, and tries to destroy the adversary politically and militarily. Having settled on a working typology of grand strategies, there is one more issue to consider: how well do these categories of grand strategies translate into the Chinese context? There is in the small literature on traditional Chinese strategic and foreign policy a fair variety in the range of strategies allegedly used by successive dynastic leaders to deal with external threats. Taken together various authors have suggested as many as seven different types of grand strategy: (1) formal alliance building, and alliance subversion; (2) he qin (fPiJl) or alliance through marriage;13 (3) ji mi (fUSI), or "loose rein" policies, including less formal hierarchical trade and tributary relations and other material inducements; (4) static defense, or a reliance on walls and strategic garrisons; (5) punitive expeditions, or the use of forward bases to launch military expeditions beyond the frontiers in order to intimidate, punish, or otherwise weaken the adversary;14 (6) preventive colonization, or the gradual expansion of a state's security perimeter;15 (7) conquest, or the application of large-scale offensive military force resulting in the military occupation or the formal annexation of enemy territory (Lattimore 1962: 112, Wu 1978a, 1978b, Wang 1983, Tao 1983: 78-79).16 Some scholars have attempted to collapse these into more manageable categories. Yang Lien-sheng distills four basic grand strategic options pursued at various times throughout Chinese history: (1) appeasement through material reward; (2) punitive expeditions only slightly beyond the gates; (3) deep penetration of "barbarian" territory; and (4) static defense, manifested most clearly in the Great Wall (Yang 1968 :28-30).17 Waldron, in his important study of the 12 Witness, for example, the similarities in language and application of the Johnson Doctrine and the Brezhnev Doctrine. See Franck and Weisband 1971. 13 See Rand 1977: 212-18, Barfield 1981: 52, 56, and Jagchid and Symons 1989: 141-64. 14 Sometimes referred to as "pacification" {zhao an S ? c ) . 15 Sometimes referred to as can shi ( R ^ ) , o r nibbling away at territory as a silkworm eats mulberry leaves. 16 Sometimes referred to as mie ($$)> or "extermination." 17 It is interesting to note the degree to which these choices are similar to those outlined by Clausewitz in On War. He lists four types of strategies: "passively awaiting the enemy's attacks"; "temporary occupation or invasion"; "projects with an immediate political purpose"

GRAND STRATEGIC PREFERENCES

1 17

origins of the Great Wall, whittles these choices down to three—namely, diplomatic accommodation, static defense mixed with occasional punitive expeditions, and offensive conquest (Waldron 1990: 41-42).18 Paul Forage's work on Northern Song policy towards the Tangut identifies similar categories of grand strategic choice.19 Waldron's and Forage's aggregation is quite helpful. Each category subsumes a number of strategies found on other lists, but this aggregation does not substantially lose any relevant information about the options considered by decision makers. As Waldron's work indicates, to the extent that grand strategic options were debated in Ming policy circles, these were generally not disaggregated into fine-grained policy categories. Put more concretely, alliance building, he qin, trade and tributary relations, and the ji mi policy, for instance, were not viewed as mutually exclusive approaches to handling the barbarian problem (ibid., 42-46, Zhang 1990: 443-44). In sum, the working typology of grand stategies outlined above seems like a reasonable one that captures the range of grand strategic options extant in China.

CENTRAL PARADIGMS AND G R A N D STRATEGIC PREFERENCES

How are these three grand strategies ordered and ranked in the Seven Military Classics and are they logically derivative of the parabellum paradigm, which seems to lie at the heart of Chinese strategic thinking? How does this differ from the order derived from the idealized, more benign Confucian-Mencian paradigm? Perhaps the conventional Confucian-Mencian paradigm is the place to begin. According to much of the secondary literature, Chinese decision makers historically preferred to defuse security threats internally through moral government, and externally through diplomatic maneuvering, bribes, formal and informal alliances, he qin-type policies, trade, tributary relations, appeasement, and other policies within what could be broadly be labelled accommodationist grand strategies.20 Next best was to rely on somewhat more coercive, but primarily (i.e., punitive strikes); and "destruction of the enemy's forces and the conquest of his territory." The only category not shared by Clausewitz is material appeasement of an external threat. See Clausewitz 1976: 94. 18 Diplomatic accommodation can subsume a broadly defined notion of he qin. That means, in essence, the defusion of threats by means of negotiation, trade, and the maintenance of reasonably peaceful political relations. On this broader meaning of he qin see Zhang 1990: 449. On he qin see also Rand 1977: 212-18. 19 Personal conversation, Feb. 1992. 20 The term for appeasement (sui jing $£3|f) does not carry the pejorative connotations in classical Chinese as it does in both modern Chinese or post-World War II English. It means to cause others to be at peace and stable by creating appropriate economic and political conditions or the wherewithal to be content with the status quo (Interview 1989b).

1 18

CHAPTER FOUR

defensive grand strategies that could entail everything from static defense operations (wall-building, networks of garrisons, the so-called tun Han (?£, EH) system used in various forms by different dynasties21), to occasional punitive expeditions designed to defuse threats temporarily, either by denial or punishment. The least-preferred way of handling security threats was through highly coercive offensive or expansionist grand strategies that entailed extended military campaigning, the occupation or formal annexation of enemy territory, and the political and military destruction of the enemy. And indeed, a more methodical look at key elements of the idealized Confucian-Mencian central paradigm leads to a similar conclusion about preference rankings. There are certainly numerous places in traditional texts on statecraft where one finds explicit statements apparently embodying a similar ranking. The Wen Zi, for instance, outlines a hierarchy of means where the most secure state is one where the ruler is supremely righteous. He properly orders the state, implements benevolent and righteous policies, and displays his virtue and magnanimity. His officials are all united and loyal, the masses interact harmoniously, and there is unity of will between superiors and inferiors. Thus surrounding feudal lords all submit to this morally based awesomeness. The next best approach is one where the state is materially wealthy and abundantly populated, the ruler is capable and his generals skilled, and the state can support a powerful military force. This gives the state an intimidating image such that when its forces are mobilized the adversary flees before actual hostilities have broken out. The least-preferred mode of security is one that rests on the bloody defeat of the adversary in battle (Qiu 1985: 254).22 Similarly, the Warring States text, the Yi Zhou Shu (SUHU), states, "The best [method of overcoming] is to have [your opponent] submit by showing him respect; next best is to take him by sheer desire; next best is to take him by subjugation; next best is to overcome him in combat; and worst is to move [against him] with the ruler relying [only] on his physical strength" (cited in 21 This was a system of military farms populated by farmer-soldiers. The farms were supposed to be self-sufficient, providing food and supplies to the army in peacetime and supplying soldiers in wartime. See Wang 1965. 22 The Huai Nanzi expounds a similar preference ranking. See Rand 1977: 222. A text written by He Qufei, first recipient of a boshi degree in military affairs during the Northern Song, suggests that a ruler has three methods at his disposal. The first is to use virtue to obtain the willing submission of the adversary; the next is to use physical strength or military power to suppress the enemy; and the third is to use strategem to capture the enemy (the implication is to capture the enemy general without violence). He concludes, "In general using military force is not as good as victory by means of strategem; to use strategem is not as good as complete [submission] by means of virtue" (Gong 1989: 13). Shi Zimei suggests a reason for this order in his commentary on Sun Zi, "To use force to compel the submission of the adversary is not as good as using virtue to compel his submission. [Although] one uses force to compel submission, the adversary still considers you an enemy." Force does not fundamentally resolve the conflicts at stake (Shi 1222: 3.1b).

GRAND STRATEGIC PREFERENCES

119

Rand 1977: 45). Elsewhere the text notes, "One who excels at government does not go on the offensive; one who excels at going on the offensive does not invade other states; one who excels at invading other states does not campaign; one who excels at campaigning does not form up in battle; one who excels in battle formation does not engage in battle; one who excels at battle does not fight; one who excels at fighting does not lose" (cited in Wu 1990: 90). Zhuge Liang, the famous general from the Three Kingdoms period (221277), provides a variation on this theme in his work, the Xin Shu OC?iIr): "Anciently, those who excelled at governing did not send forth armies. Those who excelled at sending forth armies did not form up for battle. Those who excelled at battle formation did not actually war. Those who excelled at warring did not lose. Those who excelled at losing did not perish" (Li 1957: II, 4: 11). In both citations, excelling at the antecedent activity avoided the need to rely increasingly on military means to solve security problems. The root cause of this avoidance process was proper government. Other texts are somewhat more explicit about the preferred means short of force to achieve security. The Zhan Guo Ce (sjjfc US f§), a book of stylized stories of interstate struggle from the Warring States period, suggests a state should "Use humble words and rich gifts to serve her [the adversary state].. . . If that does not suffice, cede territories to tempt her. . . . If that does not suffice then raise troops and attack her" (cited in Rand 1977: 261). The 72K' Bai Yin Jing (>fc £j |^|M), a Tang-period text, concurs: "First use civility and virtue to cherish them. If one cherishes them but they do not submit, then use ornaments, jade and silk to entice them; if one entices them but they they do not come [to submit], only then should you give the mandate [to punish] to your general, train soldiers and horses, hone armor and sharpen weapons, and attack where [the enemy is] undefended and emerge where he is unaware" (Liu and Su 1988,2:463-64). While these two quotes do not indicate whether at the violent end of the spectrum defensive/punitive strategies are preferred to offensive/ expansionist ones, they are clear that accommodationist strategies, in this case he qin-typQ policies, are preferred to the former two. Moreover there is no indication in these texts that this preference is conditional or expedient; it is apparently not a function of strategic or military weakness. The apparent preference for accommodationist strategies is also expressed more indirectly within Daoist notions of using "weakness to overcome strength" and "softness to overcome hardness." Here weakness and softness implies concessions or deception directed at potential adversaries in order to dissipate their aggressive intentions.23 This parallels the notion of "victory by means of 23 There is also a mixed Mencian-Daoist position on the best route to security. According to a Tang text on Daoist military thought, the best strategy when facing a reckless, violent adversary is to retreat or yield and to "propagate to him using civil words, and posture in front of him with shields and banners. The enemy will certainly ascertain our righteousness and retreat. Thus one is naturally without enemies" (Wang 812: 69.6b-7a).

120

CHAPTER FOUR

the dao" found in the Wei Liao Zi. This notion of victory, one will recall, involved dissipating the enemy's qi such that it abandoned its belligerent plans. While the text was unclear as to how one did this, the notion of yielding, conceding, and accommodating embodied in the concept of softness gives some clues to the concrete elements of "victory by means of the dao." Hsiao suggests that in the Daoist tradition the basis of security lay in big and small states being mutually yielding and deferent to each other's needs. The large states would accommodate or condescend to the small, and the small would serve the large (Hsiao 1979: 287). Both the Confucian-Mencian and Daoist traditions, then, seem to embody a preference for accommodationist grand strategies, particularly the extension of material benefits or status to potential adversaries. But this only addresses the preference for nonviolent, he qin-typQ approaches to security over the application of force. When force must be used, one can also find an apparent preference for static defensive strategies over offensive ones. With Chinese writings on statecraft, there are a number of moral and strategic justifications given for this defensive orientation. One moral argument derives from a rejection of the chaos unrestrained and selfish uses of force invariably create. Zhuge Liang, for instance, drawing apparently on the classifications in the Zhou Rites, argued that of the many strategic uses of force, warfare aimed at acquiring other states' land or material resources was illegitimate, as was the use of force by large states to bully smaller states. Rather, there were only two legitimate forms of force: one was righteous war, used to punish the violent and save weak states; the other was defensive war (literally "responding warfare" or ying zhan HHPc) against a state that attacked or planned to attack first (cited in Wu 1988b: 8-9).24 Another moral argument is made by Mo Zi, the advocate of the doctrine of nonoffense. His contention that brotherly love (Jian 0* Ht§?) should impel all moral action led him to reject offensive warfare because it was the strategic equivalent of theft and murder (Yates 1980: 557-58). This conclusion also led Moists to a more pragmatic causal argument against offensive strategies— namely, what goes around comes around. Expansive, offensive uses of force will sow the seeds of long-standing conflicts of interest that will eventually bring harm to the initiator (Wei and Liu 1985: 50). Instead, should "brotherly love" be insufficient to ameliorate conflict, the state should concentrate on purely defensive preparations, namely defense against sieges. Indeed the Moists became reknown for their expertise in defensive military technologies (Yates 1980, Yates 1988). Effective siege defense, along with diplomatic initiatives to acquire allies, provided a sufficient basis for security. "Mo Zi said: My walls and moats are prepared, defensive weapons at the ready; grain stores are sufficient; there is mutual love between those above and below; and I have 24

Note that offensive uses of force are not ruled out if done under the banner of righteousness.

GRAND STRATEGIC PREFERENCES

121

obtained the help of the lords of the neighbouring states. For these reasons my state is sustained" (cited in Wei and Liu 1985: 57, Li 1978: 151). This strategic calculus, argue Wei and Liu, constituted a notion of "armed pacificism" (1985: 53).25 Afinalmoral argument against offensive or expansionist uses of force comes from cyclical or holistic concepts of nature derived from the Chinese notion of yin yang, which contends that there is a natural limit to the offensive use of force; once this limit is reached there is a natural decline in military and psychic strength and resources. The use of force can go up to but not beyond this limit. To violate this threshold means both to violate the natural order, and more pragmatically, to create the conditions for strategic decline. As the Daoist classic the Jing Fa (&MS;) argues, "Those who lose their form . . . within and who wrongly launch expeditions outside [their borders] will be destroyed. Those who go against precept and overflow upward without knowing where to stop will perish" (cited in Lewis 1990: 234-35, Interview 1990c). The strategic arguments in favor of defensive grand strategies or against offensive ones come in a number of forms. One is the argument, implied by the Guan Zi, an early Han text within the Legalist tradition, that the aggressive use of force along and beyond the borders was counterproductive insofar as it unsettled the state's neighbors and created animosities. "The first kings did not use the warlike and fierce [i.e., aggressive force] along the borders. Thus the borders were peaceful. If the borders are peaceful, then neighboring states are friendly. If the [result is that] neighboring states are friendly, then this is how one's actions should be [i.e., proper] (Qiu 1985: 340). Another strategic argument stressed that defensive strategies played to the military strengths of the Han people. An official in the Northern Wei period (386-535), Gao Lu, argued, for example, that a grand strategy based on static defense—wall building between strategic garrisons—fit with the Chinese comparative advantage in static warfare, blunted barbarian advantages in cavalrybased mobile warfare, provided better forward reconnaissance and early warning of barbarian raids, and reduced the costs of border defense (Waldron 1990: 44-45). A third strategic argument stems from the linkage, found in some writing on statecraft, between defense and what is essentially a minimax strategy. Minimax—the minimization of maximum losses—is a quintessentially conservative, risk-averse strategy that eschews high-payoff victories (i.e., the extermination of the adversary or the acquisition of its political, economic, or territorial assets) because these are high-risk victories. The purpose of force is to avoid maximum loss rather than achieve total victory. In strategic terms a preference for 25 While the Moists as a cohesive school of thought on statecraft did not last past the Han dynasty, Yates (1980) contends that some Moist military techniques and concepts were absorbed and passed on by other schools of thought.

1 22

CHAPTER FOUR

minimax strategies can be translated into a preference for deterrence through the capacity to deny the enemy military victory. This capacity usually rests on effective defensive uses of force. This relationship between minimax and defensive grand strategies is made explicitly by He Qufei, the first recipient of a degree in military affairs during the Northern Song. "As for plans to resist and throw the enemy army back, [these rely on] the so-called military forces which 'save one from losing' (Jiu baizhi shi fJd&^Bifj). • • • Forces which 'save one from losing' are advantageous for strong defense" (Gong 1989: 7, emphasis mine).26 The diverse moral and strategic arguments presented would suggest, then, that the strategic-preference ranking derived from the Confucian-Mencian, non-parabellum central paradigm does indeed show up in many of the texts on strategy and statecraft. That is, there is some reason to conclude—as much of the secondary literature does or implies it does—that accommodationist strategies are preferred to defensive strategies that are preferred to offensive ones. Given the working definition of strategic culture outlined in chapter 2, there then might be some basis for concluding that China does indeed have a strategic culture—a strategic-preference ranking that is consistent across a number of key texts—and it is apparently one that does not correspond too closely to Western realpolitik traditions. But does this particular strategic-preference ranking show up in the Seven Military Classics! Is there evidence that the Seven Military Classics share a preference for accommodationist or he qin-typQ strategies over defensive ones, and defensive ones over offensive ones? As I suggested in the previous chapter, one is likely to find references to this benign preference ranking in the Seven Military Classics insofar as these derive from the idealized visions of nonviolent security in these texts. But given the underlying causal and metaphorical arguments that constitute & parabellum decision calculus, one is also likely to find at the operational level a preference for violent strategies over nonviolent ones. As for a more fine-tuned preference between types of violent grand strategies, this is a harder question to answer, but one that I will address later. Despite grounds for pessimism about finding the Confucian-Mencian preference ranking in the Seven Military Classics, it is worth first taking a look at how these texts treat the moral and strategic arguments raised in favor of this particular ranking. By doing so we can at the same time examine the extent to which the seven texts individually and collectively embody a different prefer26 An argument for deterrence based on solid defensive/denial capabilities is found as well in a much earlier text, the Lu Shi Chun Qiu. "Concerning the enemy's coming [to attack] it is because he seeks benefits. Presently he comes but receives death, moreover he retreats in order to benefit [avoid death]. If enemies all take retreating as being to their advantage, then there is no 'clash of blades'" (ren buyujie JJ^J&lfic) (Lin 1986: 217). This is a defensive version of "not fighting and subduing the enemy," as opposed to the offensive flaunting of overwhelming, intimidating force.

GRAND STRATEGIC PREFERENCES

123

ence ranking—one that, as the parabellum paradigm would suggest, places violent, possibly offensive, strategies before less coercive strategies. As in the previous chapter it is perhaps best to analyze each text separately and then test for congruence between any preference rankings that might emerge. Wu Zi Bing Fa An analysis of the Wu Zi Bing Fa's composite cognitive map suggests that, at the very least, accommodationist grand strategies are not the most preferred choice. There is no direct causal relationship between diplomatic and political strategem and defeat of the adversary or the security of the state (see Fig. 3.1). This does not mean, however, that in all cases Wu Zi advocated aggressive wars of expansion to deal with security threats or problems. Like other works on strategy Wu Zi believed the use of force merely to acquire territory or material benefit was immoral and ought to be opposed (Li and Wang 1986: 58). For Wu Zi the only legitimate use of force was for the purposes of righteous war. Righteous war, as I have suggested, however, does not limit its practitioner to one particular set of strategies, whether at the level of grand strategy, military strategy, or tactics. It certainly did not limit grand strategy to diplomatic actions or to static defensive postures. Indeed, for Wu Zi there was a great deal of utility in the preemptive application of force against an adversary's forces, particularly when these forces were vulnerable strategically, materially, and psychologically. There are over twenty cause-effect statements in the text where such action is linked positively to a general notion of utility (156.la.l-3, 156.la.5-8, 157. la. 1-2, 157.2a.2-14, 157.2a.16). In the process of carrying out righteous war there was also no rule against the invasion and plundering of enemy territory. "As for the Way of attacking the enemy and besieging his cities, once the cities and towns have been destroyed, then enter his buildings, commandeer his wealth and collect his weapons and materials" (Li and Wang 1986: 97). The two historical analogies used to illustrate righteous war both refer to ancient sages invading the moral offender's territory and being welcomed by the oppressed peoples (ibid., 58). There is, however, a strong flavor of contingency and expediency to all of this. The Wu Zi is one of the few of the seven texts to distinguish more precisely between different types of security threats, and hence different types of responses to these threats. Within the overall parabellum framework, the text did allow that some conflicts were possibly resolvable short of resorting to force. But this statement is heavily qualified because it is not clear in the short passage in which Wu Zi appears to advocate nonviolent means to defuse particular types of threats whether these means are separate from or sequential to the use of force. Wu Zi lists five types of wars: wars fought for righteous causes (yi bing § | &) ; wars using massed force to invade other states {qiang bing ?®^); wars

124

CHAPTER FOUR

fought out of anger and rage (gang bing W\&)\ wars fought by those who abandon all propriety and seek material benefits (bao bing |j|j=£); and wars of rebellion against the political and moral order (ni bing J^Ji). Each of these types of wars required different strategies to resolve the underlying sources of conflict. The roots of righteous war were, for instance, the adversary's unrighteous behavior. Thus ultimately one had to use propriety or li Of!) to cause the enemy to submit. It is not clear from the text, however, whether this propriety comes before or after the application of violence. Certainly in other texts, righteous war is used first to defeat the military power and usurp the political authority of the offending elites, after which the victorious side sets up, or restores as the case may be, a moral political elite that will follow the precepts of li. As for wars of aggression or brute force by powerful militaries, Wu Zi recommended adopting a humble, somewhat conciliatory posture in order to defuse or deflect aggression—apparently a policy of appeasement. Again, it is not clear whether this is seen as a resolution of conflict, or merely a temporary expedient to be adopted by the weaker state. As will be discussed, in other texts one finds the argument that analogous accommodationist responses are preludes to the application of superior force once the enemy has been sufficiently mollified and deceived. Wars of rage were apparently to be dealt with in a similarly accommodationist fashion: one must use pleasing words to convince the enemy to desist. As for wars of acquisition, one should resort to deception and strategem. The text does not explicate how diplomatic or political deception alone might buy off or mollify this type of aggressor; Wu Zi may have been referring either to military operations that feigned weakness and vulnerability, thus enticing an aggressor into military traps (as Sun Zi's passage on deception proposes), or to offers of wealth and territory. The term used, zha (ff), implies trickery, so such efforts to buy off the adversary might be part of a longer-term, more complex military strategy. Finally, in the face of wars of rebellion, Wu Zi advocated the flexible use of military force. The passage in the Wu Zi on contingent strategies is quite short and abstract. In a far more detailed, concrete discussion of the six states that at the time posed potential threats to the security of his own state, Wu Zi spent a great deal of time analyzing the national character and military styles and capabilities of these threats. Against each different type of threat he recommended a different approach. Indeed, the terms he used to classify these different states are similar in some cases to the terms he used in the passage describing different types of war. Thus the responses Wu Zi recommended to these concrete situations should provide some clues to the content of some of the strategies outlined in the paragraphs above. Against the easily enraged, wealthy, oppressive, and divided state of Qi (which Wu Zi classifies as gang or "hard," "harsh"), Wu Zi recommended what appears

GRAND STRATEGIC PREFERENCES

125

to be offensive war to divide, hunt down, and destroy Qi forces. How this fits into his general admonition to use pleasing words against a gang state is not clear. The only hint—and this is speculative—is that, given the political oppressiveness of the state and divisions of will within its army, groups inside the Qi state were potentially susceptible to overtures aimed at further exacerbating internal conflicts. This would weaken the Qi politically and militarily, thus allowing for the application of force. The Qin state, on the other hand, was qiang (5S)j o r preferred to wage wars of brute force. Its territory was strategically located and characterized by many defiles and narrow access points. Government was strict, and its people had an unyielding fighting spirit. Against this type of adversary Wu Zi recommended offering inducements and advantages to disperse and divide Qin forces. Then one should hunt down these separate units, set up ambushes, and at just the right moment attack and capture the enemy's commander. Here again it is not clear whether the feigned inducements refer to monetary or military advantages. But it is clear that this is but a step in a sequence in which force is eventually applied. This suggests, in reference to Wu Zi's more abstract discussion on how to deal with qiang enemies, that accommodating strategies are deceptive preludes to violence. As for the state of Chu, it was weak, its territory was large, its government unstable, and its people exhausted. Its military power therefore lacked sustainability. Wu Zi recommended a strategy that mixed preemptive attacks and attrition: one should attack Chu military camps swiftly, and then withdraw, relying on constant harassment of remaining enemy forces in order to undermine their morale and weaken their military capabilities. The people of the state of Yan were simple, honest, and industrious, and they prized bravery and righteousness. But they lacked strategems in warfare. Thus they excelled at defensive war and did not retreat. Against this type of enemy, Wu Zi argued, one must engage and intimidate its forces, invade, and then immediately disengage, striking next at their rear. This will sow doubts and confusion among Yan leaders and instill fear among Yan soldiers. Finally, with regard to the three Jin states (Wei, Zhao, and Han), these countries were at peace and their governments stable. The people were worn out by warfare, having been constantly subject to it, and therefore tended to disregard commanders. Their armies lacked the will to fight to the death. Wu Zi recommended that one should box in their forces, and press in on them. When they attacked one should resist; when they retreated one should pursue, thus tiring them out. I should make a couple of observations here. First, beneath the surface, there are no glaring inconsistencies between the strategies in Wu Zi's abstract discussion of how to handle different threats and his concrete policy recommendations in his discussion of Qi, Qin, Chu, Yan, and the three Jin states. In every instance the application of violence is recommended or, as I have suggested,

1 26

CHAPTER FOUR

implied by context. In those cases where the application of violence is explicit it requires, at some point, offensive war aimed at defeating the enemy's forces or capturing the enemy's general. Diplomatic and political actions or accommodationist approaches are instrumental; they are used to influence the characteristics of the conflict in such a way that the use of force is more efficacious. While there is no indication from the text that the acquisition or annexation of enemy territory is a necessary condition for the resolution of conflict, the notion of righteous war does imply that the political usurpation of the adversary's leadership and its replacement with some new righteous political order is an important part of effective conflict resolution. Second, Wu Zi's strategic advice is highly flexible and contingent. In the process of defeating the adversary and resolving security problems no one strategy is ruled out. This flexibility, however, comes within the context of an overall preference for violent strategies—a preference rooted in causal arguments that suggest security threats cannot be resolved satisfactorily without the enemy's defeat and submission. We can conclude, then, that there is at least a preference for violent grand strategies over accommodationist ones. At the violent end of the strategy spectrum, there is some evidence to suggest as well that more offensively oriented strategies are preferred to defensive measures. Certainly in neither the abstract nor more concrete discussions of different threat environments and strategic contingencies are there arguments in favor of predominantly defensive strategies based on "deep moats and high walls." Si Ma Fa Like the Wu Zi Bing Fa, there are no direct causal linkages made in the Si Ma Fa between nonviolent political, diplomatic actions, or internal rectification policies on the one hand and the defeat of the enemy or the security of the state on the other. This would suggest, therefore, that at the very least accommodationist grand strategies are not the most preferred since these are not particularly effective. This conclusion requires more elaboration, however, since there is one passage in the text that, according to a recent commentator, does imply a strategic-preference ranking that places nonviolent political or diplomatic strategies first. Early on the author lists six ways in which the king and the hegemon can order the realm and respond to threats to this order.27 The first is to use differential distributions of land and wealth to establish a hierarchy among the feudal states. The second is to use rules and regulations to control these states. The 27 The text refers to the ideal political order as it was supposed to have been during the Zhou dynasty. The son of heaven ruled the realm, but order was kept among feudal states by the hegemon, who was the ruler of the chief feudal state and who acted as an intermediary between the king and lesser nobility. The hegemon was also the king's enforcer of order should a feudal state stray from the correct path, and led the righteous use of force on behalf of the king.

GRAND STRATEGIC PREFERENCES

127

third is to use rites or li and trustworthiness to unite with them. The fourth is to employ officials with strength of character to put the minds of the feudal lords at rest. The fifth is to use clever strategists to devise political means to keep the lords united. And the sixth is to use military force to make the feudal lords submit (Liu 1955b: 12-13). These six strategies could be structured one of three ways. First, they could be a ranking of preferred options. The two least-preferred means of securing the realm—one involving the application of violence—are both the most divorced from the Confucian-Mencian ideal of good government. This would imply that the preference is first for accommodationist measures, or measures associated with traditional notions of internal rectification, and only if these are ineffective should the ruler resort to force.28 Second, they could be a list of policies to be pursued simultaneously for the most effective ordering of the realm. On the one hand, the unity and willing submission of the states within the realm would be ensured through enlightened government. On the other, the ever-present threat of military force or punishment would deter challenges to this order. Finally, the six strategies could constitute a menu of choices, the selection from which would be contingent on the nature of the security problem. There would be no a priori preference for any particular option. We can get an indication of which interpretation most closely reflects the strategic axioms in the text by translating these three alternatives into the simple wen-wu (civil and martial) or ben-mo (root and branch) dichotomies. Options one through five can be classified as wen- or Z>e«-related strategies. Option six is clearly what the text would classify as wu- or mo-related. And indeed the text suggests at one point that the best strategy is to use ben, while the next best is to rely on mo options (ibid., 50). This would suggest, then, that the six options are structured roughly in a preference ranking, moving from most preferred to least preferred. How is this apparent preference for nonviolent strategies reconciled with causal arguments in the text that suggest only an indirect nonviolent route to security? The resolution of this apparent tension is found in the notion of quan bian dHH), or the doctrine of absolute flexibility. In fact the text itself undermines this wen over wu ranking when it argues that circumstances should be the sole determinant of when to use ben and when to use mo. "[The use of] the root or the branch is solely a question of quan. This is [the essence of] war" (ibid.).29 This flexibility is reflected in the opening paragraphs of the text, cited earlier, that if attacking and killing the enemy brings peace to the realm and eliminates those who offended the moral order, then it is justifiable behavior. Thus, while 28

This is Tian's interpretation (1989: 42). In modern and classical Chinese quan can refer to power and authority. In classical Chinese it also means literally to weigh or balance; in strategy this translates into the weighing of options. More often than not, in the military texts it refers to the application of coercive force by legitimate authority after a careful weighing of all relevant factors in a conflict situation. 29

1 28

CHAPTER FOUR

at some idealized level there seems to be a preference for nonviolent strategies, the flexibility axiom embodied in the notion of quart Man—a notion at the heart of the pragmatic strategic advice of the text—undercuts this ranking. Does the flexibility axiom preclude us from drawing conclusions about alternative preference rankings, say ones that place offensive or defensive strategies on top of accommodationist ones? Unfortunately the text provides little concrete indication of the threshold beyond which a threat must be met with the use of force instead of accommodation. If this threshold were low—approaching a zero-sum conception of conflict—then it follows that the state would be compelled to use violent means relatively frequently. If any transgression of the moral order, regardless of its severity, is seen as an existential threat to the broader security of the realm, then the preferred means of dealing with this threat would be at the violent end of the strategy spectrum. If there are no gradients of threat, then in any particular threat situation the preferred strategies would lie at the violent end of the spectrum of choice.30 One needs to find evidence, then, that indeed there are no gradients of threat, that the threshold that triggers the use of force is so low that any and all transgressions of the order of the realm justify a violent response. There is one passage in the text that arguably provides this evidence. This occurs in a discussion of the process of the launching of a righteous war. From the description there are few grey areas between states that adhere to the moral order and those that offend it.31 The process began with an imperial inspection tour of the feudal states to determine which states were deserving of reward and which of punishment. The latter were states (or state rulers) that bullied and harassed other states, were internally cruel and barbarous, neglected agriculture and generally rebelled against the established political and social order.32 Once the ruler determined which states belonged in which group, the chief minister of the empire gathered loyal feudal lords together and declared that such and such a state had violated the dao. All other states were ordered to contribute military forces to an expedition to suppress the offending state(s). This was the formal mandate by which any particular use of force was deemed righteous. There were no clear gradients of offenders, nor once a state was deemed to have violated the dao was there any range of options other than the 30 Conversely, if this threshold were high—in other words, if there were many shades of transgression that lay below this threshold—then at least some significant proportion of threats could be met without resort to force. While this would not establish a clear preference for nonviolent strategies—since this would depend on where any particular transgression lay in relation to the threshold of threat—it would at least establish that violent strategies were not generally preferred. 31 The text situates itself historically sometime in the Zhou dynasty, when a nominally ruling king presided over a realm comprised of smaller feudal states. Interstate violence was justified almost invariably by all sides as an effort to punish the unrighteous and restore the moral order. 32 See chapter 3 for a list of nine types of unrighteous behavior, based on the Zhou Rites.

GRAND STRATEGIC PREFERENCES

129

33

use of military force. Righteous war was militarily offensive and, once defeated, the feudal lord and his ruling clique could meet a number of different ends from exile to death to the forced secession of a portion of territory (ibid., 13-15). On the basis of this passage, then, we can tentatively conclude that the preferred means for handling political and military threats to the realm were violent ones, since all threats fell under the rubric of unrighteous behavior. Certainly this is consistent with the causal relationships in the text. But what of a more specified preference between offensive and defensive grand strategies? The evidence is, again, fairly indirect. The behavioral choices in the text embody elements of both types of grand strategies. On the one hand, the political and military destruction of the transgressing state, as well as the annexation or redistribution of portions of its territory, were valid steps in solving the security problem. These all required the offensive use of military forces inside enemy territory. This would suggest a preference for an offensive and expansionist grand strategy. On the other, punishing an adversary did not necessarily require the permanent occupation of the enemy state. Indeed, at one point the text argues against the deep penetration of enemy territory on strategic grounds. If one uses large numbers of heavily armed forces deep in enemy territory they may be easily attacked, or they may be hard to control, plundering and destroying and thus failing to win the hearts and minds of the locals. If light forces are sent deep into enemy territory they are likely to be defeated by better prepared, more desperate defenders. However, heavy forces sent just inside enemy territory will be successful, since they are easier to control and enemy forces are more dispersed. This kind of operation is designed primarily to punish the adversary (ibid., 47). Some commentators even suggest that the Si Ma Fa implies offensive wars of total victory are immoral. In one passage the text admonishes not to pursue enemy forces more than one hundred steps. This decision rule, Liu Yin comments, demonstrates compassion, benevolence, and self-control towards one's own soldiers as well as enemy forces (ibid., 7). All of this would indicate a preference for somewhat less coercive punitive grand strategies, though not for static defensive ones per se. It is clear, however, that these admonitions in favor of spatially and militarily restrained, punitive uses of force are highly conditional, and in the end do not constitute a hard and fast preference over offensive strategies. For one thing it is hard to see how this restraint fits with some of the punishments the Zhou 33 In the list of nine transgressions there is one that appears to call for a response short of offensive military options. If the offending lord violates orders from the Son of Heaven or insults the imperial government's policies, then he is to be isolated, his relations with other states cut off. The passage does not explicitly refer to the use of military force to carry out this punishment, but presumably some force would be deployed to enforce this embargo.

130

CHAPTER FOUR

Rites call for—namely, the invasion of the enemy's state, and the removal of the enemy ruler. More importantly, the caution against deep penetration of enemy territory is wholly qualified by the military risks involved, not the moral implications. The text states, for instance, that if one's chariots are sturdy and fast and weapons sharp or effective, then light forces can strike at strategic points well within enemy territory (ibid., 48). As for the admonition not to pursue a retreating enemy too far, here too the reasoning is based on military caution: "By not [pursuing the adversary] far, it is hard to be entrapped. By not catching up to [and defeating a retreating enemy] then it is hard to fall into an ambush" (ibid., 19). Presumably, then, the limits on pursuit are flexible, dependent only on intelligence about the enemy's activities. What does all this add up to? Like the Wu Zi Bing Fa, the Si Ma Fa argues the only legitimate use of military force is for righteous war. Within this context the use of force is governed by a flexibility axiom: depending on strategic circumstances, any and all means can be used to punish the unrighteous. While the text does not explicitly discount or reject the value of we«-related means, the nature of "threat" implies that the most efficacious route to security is through the application of violence. As for whether the preferred form of violence is defensive or offensive, here the text is somewhat ambiguous. It would seem, however, that preferring defensive strategies over offensive ones is conditional or contingent. That is, should strategic conditions allow it, the offensive/expansionist application of force is likely to be more efficacious, given the range of punishments that can and should be meted out to transgressor states. Wei Liao Zi The cause-effect relationships in the Wei Liao Zi clearly rule out any direct positive association between political or diplomatic strategies on the one hand and the submission of the enemy or the security of the state on the other. The only suggestion that nonviolent strategies are more efficacious in dealing with adversaries comes in the text's reference to "victory by means of the dao" As I argued, however, this type of victory is an idealized one; it is mentioned, then essentially dropped, as the text goes on to expound on the problems of achieving victory through the wielding of military forces. Moreover, relying on the dao for victory is not the same as relying on internal rectification,34 or using diplomatic alliances, economic or territorial trade-offs, among other tools of accommodationist strategies. At no point in the texts is there any explicit 34

In fact, Liu Yin takes the Wei Liao Zi to task for neglecting this Confucian-Mencian element. "Now [victory by] the dao is not limited to thoroughly investigating and understanding martiality, and on estimating the enemy alone. Rather, one must first cultivate virtue, implement correct policies, look after the people, employ the capable, that is all" (Liu 1955c: 14).

GRAND STRATEGIC PREFERENCES

131

advocacy of accommodation. In fact, there is one passage that can be interpreted as a criticism of states that rely on he gin-typQ strategies both to resist an adversary or to buy the military assistance of other states (Liu 1987: 37). It is safe to conclude, then, that this text does not embody a grand strategicpreference ranking that places accommodationist strategies at the top. Rather, there is an obvious preference for violent strategies, and within this a fairly clear preference for offensive or expansionist ones over defensive options. The notion of "victory by means of physical force," for instance, involves the invasion of enemy territory, the destruction of enemy forces, and the capture of its cities and towns.35 To be sure, this is followed by withdrawal of one's own forces, but not without the annexation of at least some enemy territory. Moreover, withdrawal does not necessarily mean a return to the political status quo ante. One passage in the text could be interpreted to mean that if a state wants to control the land and population of another state it must also get the support of the enemy state's capable officials, presumably making use of them after disposing of the enemy ruler. If political consolidation over the vanquished states is not carried out, then the military threat remains. "If you cannot acquire their capable [officials] and [yet] wish to control the realm, then [the enemy will] necessarily swallow up your army and kill your general" (ibid., 42).36 It is clear elsewhere in the text that the military defeat of the enemy should be followed by the replacement of the enemy's political leadership and the implementation of liberal policies within the vanquished state designed to "relieve the people of their hardships"(ibid., 258).37 The premium placed on the invasion of the enemy, the annexation of some of its territory, and the usurpation of the enemy ruler's political power, all suggest a preference for offensive grand strategies. But what of the notion of "victory by means of awesomeness"? While this form of victory relies on the intimidating display of military force, if successful it leads to the enemy's willing submission short of outright military defeat. Does this type of victory reflect any preference between defensive or offensive strategies? The notion does imply deterrence through an impressive capacity 35 See as well the cause-effect statements on the efficacy of offensive, preemptive warfare (25.3.3, 25.3.4, 25.3.5, 54.1.1, 54.1.2). 36 An alternative reading is simply that if one wants to control another state one must first have capable officials in government within one's own state. The difference turns on whether the possessive pronoun qi (jR) refers to the subject or the object (the enemy) in the sentence. In most cases the pronoun refers to "the other," in this case the enemy. Moreover, this latter interpretation puts too much of a Mencian spin on the text, which does not jibe with the clear stress on military sources of security. For a similar reading see Xu 1989: 66. The Mencian reading is found in Liu 1987: 42. 37 A similar recommendation of restraint aimed at winning the support of the enemy's people during military occupation is found in the Si Ma Fa. See also Anonymous 1986: 123. The content and purpose of these rules were similar to Mao Zedong's "Three Great Disciplines and Eight Points of Attention."

132

CHAPTER FOUR

to defend or deter through denial—a concept that I earlier subsumed within the category of defensive grand strategy. And indeed, there is one passage that suggests that "deep moats and high ramparts" is one way of diminishing a threat without resorting to military clashes (ibid., 44). But other components of strategic awesomeness suggest that intimidating shows of force need not be limited to static, defensive, or deterrence posturing. Awesomeness comes in part from the general's having received a formal mandate to launch a righteous war, which we have seen almost invariably means the invasion of enemy territory. Awesomeness rests as well on the intimidating presence of righteous soldiers within enemy territory.38 Awesomeness can also result from forming up in battle array and declaring the enemy's crime, as a prelude to carrying out righteous punishment (ibid., 44, 168). Thus, "victory by means of awesomeness" does not restrict strategic choice to defensive options; the rubric is broad enough to include actions that entail the invasion of enemy territory. Which form of "awesomeness" is most appropriate or efficacious—defensive, offensive, invasive, etc.—is flexibly determined; it requires above all correctly estimating the enemy's strengths and weaknesses (ibid., 44). Of course, strategies that might be termed defensive are not ignored in the Wei Liao Zi. A state's having deep moats and high ramparts around its towns, and having well-prepared defensive technologies are both positively related causally to the ability of the state to protect itself from threats (15.2.4, 18.2.5, 28.1.1,28.1.6,28.1.3). Nor does the text ignore the possibility that one's forces may be thrown on the defensive, and therefore should be trained to fight defensively behind artificial and natural obstacles (27.1.2, 34.3.2, 27.2.1). But this type of warfare is not the most desirable form; generally the defense of towns is not viewed as a direct cause of the enemy's submission. Indeed, it is best to avoid having to defend towns in the first place by retaining the capacity to war against the enemy (Liu 1987: 54). And warring outside of one's own territory is more efficacious than warring within or at the gates. At best, defense is viewed as a temporary posture adopted until conditions are such that the enemy weakens militarily and psychologically, thus allowing one's own forces to go on the offensive (19.2.1). The determination ofjust when these conditions are ripe is dependent on quan bian, or the flexible calculation and exploitation of opportunity. Once so determined, force should be applied preemptively. "If one applies quan bian techniques before the enemy does, the enemy will be unable to use his power to engage. By applying military force before the enemy does, the enemy is thus without power to engage. Thus in military affairs one prizes preemption" (ibid., 54). In sum, like the Wu Zi Bing Fa (and perhaps the Si Ma Fa) the Wei Liao Zi 38 This is termed "the doctrine of crossing the bank," meaning crossing the riverbank or boundary into enemy territory with highly efficacious forces. There are two references to this doctrine as a source of bloodless victory. See Liu 1987: 44, 168 and Liu 1955c: 15, 56.

GRAND STRATEGIC PREFERENCES

133

exhibits a grand strategic-preference ranking favoring offensive strategies over defensive ones and defensive ones over accommodationist options. Huang Shi Gong San Lue As noted above, the San Lue places more emphasis on the nonviolent sources of state security than do most of the other texts in the Seven Military Classics. There is a clear, direct causal relationship, for instance, between internal rectification and the security of the state. Diplomatic and political strategem are also positively related to the diminution or weakening of the adversary. On the other hand, the text also clearly outlines causal paths from the violent, invasive application of force in the name of righteous war to the submission of the enemy and the security of the state. From these multiple sources of security, it is not entirely clear whether there is a grand strategic-preference ranking in the text. On the one hand, the San Lue appears to reject expansive notions of security. Central to the text is an extensive discussion of the concept of yielding, which apparently draws from the Daoist notions of "using softness to overcome hardness" and "using weakness to overcome strength."39 The yielding axiom can translate into a number of different types of political and diplomatic behaviors. In the political-strategic realm, yielding or softness can include a Confucian-Mencian notion of internal rectification. This, in turn, translates into an admonition against the sole reliance on military means—defensive or offensive—to achieve security. "Stressing the expansion of territory [will lead to internal] barrenness. Stressing the expansion of virtue [will lead to] flourishing and strength," argues the text. A moral leader does not covet another's territory or wealth, but rather is content with what he already possesses. This will lead to the state's being at peace. Coveting what others possess will lead to the destruction of the state and long-term disaster, presumably through endless warfare (Liu [SKQS]: 726.120). The ruler should instead concentrate on problems of internal order, and set aside concerns about distant affairs. Commentaries generally interpret this to be an argument for a defensive posture vis-a-vis external adversaries, an admonition to concentrate on the cultivation of virtuous rule, coupled with efforts to establish and maintain friendly relations with those far away.40 These sorts of policies will lead to the adversary's self-submission. At worst, if the adversary resists, then virtuous, 39

Shi Zimei explicitly draws an intellectual connection between the San Lue and Lao Zi (Shi 1222: 31.2a-2b). 40 In his commentary in this passage, Shi Zimei criticizes Tang Tai Zong for his invasion of Korea because it eventually brought harm to later generations. Instead, he suggests, being "content with what one possesses" means withdrawing from forward defenses and closing up access points. Extended campaigns designed to establish political and territorial control over the

134

CHAPTER FOUR

moral leadership—softness to the aggressor's hardness—will attract the assistance of others in the realm. Hardness or aggressiveness will only cause others to rise up against a state (ibid., 726.100).41 Yielding also broadly subsumes diplomatic measures designed to deflate or defuse threats either by temporarily acting submissively or by extending a helping hand to a potential adversary (ibid., 726.103-726.104). Taken together, then, the nature of yielding in the San Lue suggests a preference for a mix of accommodationist and defensive grand strategies over offensive uses of force. On the other hand, the text, like all the other strategic classics, embraces a concept of righteous war that does not exclude the invasion of enemy territory and the overthrow of its political leadership. There are some cause-effect statements that point to the utility of attacking the enemy preemptively, occupying its territory, seizing its cities and capturing its officials (726.109.15,726.109.19, 726.104.1-726.104.3). Once the enemy has been defeated its government should be reformed, a new ruler installed, and new upright officials (apparently native to the vanquished state) employed. Then the victorious forces should be withdrawn, and there should be no need to attack the (former) adversary again (Liu [SKQS]: 726.104, Xu 1986: 43). In general, this process is similar to the one outlined in the WeiLiao Zi. One difference is that the San Lue apparently does not advocate the continued occupation of enemy territory or the appropriation of enemy wealth as part of a political solution in the aftermath of a righteous war.42 Thus, although under the rubric of righteous war the San Lue shares a preference for the military and political destruction of the adversary, it rejects territorial annexation as a solution to security threats. Morever, it is clear from the emphasis on yielding strategies in all their forms, as well as from the relatively small amount of space devoted to problems of righteous war, that the San Lue views the threshold of threat beyond which righteous war is necessary as being relatively high, in barbarians were counterproductive; the barbarians had different customs, lived in unfamiliar terrain, and hence it would be difficult and costly to control and defend such gains (Shi 1222: 33.4a). See also Xu 1986: 104. Liu Yin uses a common idiom to elucidate the meaning of "putting aside that which is far away, to plan for that which is close": yuanjiao jin gong (3SI5J 3fi$0. The phrase means to "establish good relations with those far away and to attack those close by" (Liu [SKQS]: 726.120). 41 Shi Zimei comments, "An enlightened king concentrates on expressing virtue, thus the four barbarians submit to his rule. Thus by propagating virtue one can then make those distant submit. What need is there to rely on expanding territory?" (Shi 1222: 33.4b). See also Xu 1986:41 and Wei 1986:40. 42 In a discussion of how to respond to the enemy under different strategic conditions, the text only reads, "On obtaining, don't take it as your own." Liu Yin annotates this to mean that the victor should redistribute occupied territory and acquired wealth. He does not suggest to whom these resources should be distributed (Liu [SKQS] 726.104). Wei Rulin suggests the meaning here is that the resources should be redistributed to aid the people of the vanquished state (1986: 50).

GRAND STRATEGIC PREFERENCES

135

comparison to, say, the Si Ma Fa. In other words, by placing a fair degree of faith in the efficacy of soft strategies like internal rectification and diplomacy, the text argues implicitly that there are a range of threats that need not be met with righteous violence. Within this range, the preferred strategic means would be a mix of accommodationist and defensive approaches. This would suggest, then, that the San Lue does not reflect a consistently hard zero-sum view of conflict, nor indeed, that it fits as well within the parabellum paradigm as the other texts.43 As expected, this conclusion needs to be qualified somewhat. While the yielding axiom suggests a grand strategic-preference ranking that places accommodationist and defensive strategies first, there is another facet that loosens this preference, namely a deeply embedded notion of strategic flexibility. A yielding or soft approach to security also includes responding to the constantly changing nature of conflict, conforming to the nature of the enemy, and waiting for the enemy to move first. This way the enemy reveals its intentions and capabilities, thus maximizing the efficiency of any countermoves. This axiom is linked causally to the ability to control the realm (726.101.1, 726.101.2), to the pacification of the "nine barbarians" (726.101.6, 726.101.11), and to the extension of the ruler's awesomeness (726.101.4, 726.101.7). Although the term is not used in this text, clearly the strategic principle at work here parallels the concept of quart bian. Even in this the least realpolitik of the military classics, preferences for nonviolent solutions to security problems are moderated somewhat by the requirements of strategic flexibility. This implies that no methods can a priori be excluded from the strategic options available to decision makers, and that any strategic-preference ranking is conditional. The difficult question is whether quart bian makes this ranking so conditional as to be essentially irrelevant as a guide to strategic choice. Tai Gong Liu Tao Like the San Lue, the Liu Tao draws a direct, positive causal relationship between internal rectification and external security. But other elements of an accommodationist strategy, such as he qin-typz diplomatic actions, are conspiciously missing as direct causes of the submission of the enemy or the security of the state. There are only two roles for political and diplomatic strategies. One is, as elaborated in the chapter on nonmartial attack (wen fa ~$C {%), to weaken the adversary politically and diplomatically, thus paving the way for the application of military forces. In this case, the strategems range from gifts of riches, fawning subservience and friendship, bribery, political 43

According to one source, the San Lue was said to have influenced the Han emperor Guang Wu's (6 B.C.-A.D. 57) opposition to offensive war against the Xiongnu barbarians (Wang 1985: 292).

136

CHAPTER FOUR

subversion, among other techniques that weaken, divide, and expose the enemy. The other role for diplomacy is to buy the support of other states to assist in defeating the adversary when one's own military forces are inferior to those of the enemy. In this case, one should treat the officials of potential allies with respect, heap gifts upon them, and address them with lowly, subservient words (Xu 1986: 187).This, too, is a prelude to the use of force against a superior enemy, however. In short, the text does not present any strong arguments that would lead us to conclude that accommodationist grand strategies were clearly preferred to more coercive, violent approaches to security. As for any demonstrated preference for defensive grand strategies, here again there is precious little evidence that the Liu Tao advocates static defense. There are only a few instances where the text discusses static defensive postures (e.g., "deep moats and high ramparts"). Mostly, these refer either to the defense of staging areas for attacks inside enemy territory, or defenses and obstacles around the army's encampment when inside enemy territory. In only one case is there a discussion of how to defend deep within one's own territory against a marauding adversary (ibid., 174-75). In general, defensive preparations are not linked causally to the defeat of the adversary or the security of the state. Rather, they are linked to obtaining advantage over the enemy in a stalemate (32.3.7, Xu 1986: 154), and to the ability to fight protractedly (79.2.2). This leaves punitive and expansionist uses of force. Once again we are confronted with a frustrating degree of ambiguity in the text. There can be little doubt that the Liu Tao places a high premium on the preemptive, offensive use of force within enemy territory. This preference shows up in a number of forms: causal paths that lead to the weakening or destruction of the enemy (i.e., 33.1.2, 73.1.3, 77.1.1, 77.1.2, 78.1.1); an extensive discussion of the problems of operating militarily deep in enemy territory;44 and metaphors that underscore the advantages of preemption.45 But there is little indication whether the employment of offensive force is for the strategic purposes of punishment followed by withdrawal to preexisting boundaries or for the destruction of political leadership and the annexation of territory. In contrast to other texts, there is no discussion of permanently seizing enemy territory or of overthrowing the adversary's political leaders and replacing them with more righteous (i.e., compliant) officials (Liu 1955d: 22). Thus, while we can safely conclude that the Liu Tao demonstrates a preference for violent grand strategies over nonviolent, accommodationist and static defensive ones, the discussions of preferred modes of 44 There are altogether fifteen sections that specifically address hypothetical strategic contingencies while operating deep inside the enemy state. 45 Some of these are used directly to describe the best use of force (see chapter 3). Others are more general admonitions to act preemptively while political and strategic problems are still manageable (i.e., blocking a bubbling well before it turns into a river, extinguishing a flame before it becomes a blaze, and nipping buds before they become a tree that requires the use of an axe) (Liu 1955d: 20).

GRAND STRATEGIC PREFERENCES

137

violence are not specific enough to reveal a preference between more actively defensive strategies on the one hand and unambiguously offensive ones on the other. The only indication that perhaps the text leans towards offensive strategies is language that implies that the strategic goal of military force is to destroy or exterminate the enemy (45.3.4, 75.1.1-75.1.4, 77.1.1, 77.1.2, 77.2.1, 77.2.2). But this is only one of a number of criteria for offensive grand strategies. The only certainty is a preference for the offensive use of force at the operational level, though whether this is for punitive purposes or expansionist ones is unclear. Within this broad preference for violent solutions to security problems, however, there is, as with most of the other texts, an emphasis on the need for strategic flexibility. The ability to conform to the nature of the enemy, exploit opportunity, and adopt the most efficacious strategies are all positively related to the defeat and destruction of the enemy (43.2.1,43.2.1 a, 71.2.2, 77.4.1). As one metaphor puts it, one must be prepared to feign acceptance of the enemy's will and respond to his actions as though you were brothers. This way you can control the enemy and, when the opportunity arises, destroy him (Liu 1955d: 45). Or, as the text remarks elsewhere, "The directional power [of the military] comes from responding to the movements of the enemy; the flexible exploitation of change is born from the interaction of the two army's battle formations" (ibid., 71). Tang Tai Zong Li Wei Gong Wen Dui The Wen Dui is an extended dialogue on grand strategy, operations, tactics, and military training between an emperor and his strategist. Significantly, the whole discussion is prompted by the failure of diplomacy. It opens with a question from the emperor Tang Tai Zong: "The Gao Li have repeatedly invaded the Xin Luo. I dispatched an envoy to go order [a cessation of war]. But the Gao Li did not acknowledge the edict. I want to punish them [by military means]. How should this be done?"46 The starting assumption, then, is that force in some form or another has to be used, an assumption that is confirmed by the composite cognitive map described in Fig. 3.6. Indeed, at one point in the text Tang Tai Zong signals his agreement with what the ancients once said, "Using wen [i.e., benevolence] one can draw close to the people [of one's own state]; using wu [martiality] one can intimidate the enemy" (Zeng 1986: 185).47 46 The Gao Li and the Xin Luo were two of three main tribes situated in the Korean peninsula, an area within the system of Tang tributary relations. In Tang Tai Zong's time, the Xin Luo accepted, while the Gao Li rejected, a tribute relationship with China. For a fuller explanation of the strategic situation facing Tang Tai Zong in Korea see Zeng 1986: 33-38. 47 This recalls the famous Three Kingdoms strategist Zhuge Liang's comment, "As for the Rong and Di [barbarians] it is hard to use principle to transform them; it is easy to use martiality to force them to submit" (Jiang 1989: 138). The Wen Dui appears somewhat contradictory on

138

CHAPTER FOUR

The question is whether this use of martiality should be predominantly defensive or offensive. There is some evidence that the Wen Dui advocates a mix of static defensive and punitive offensive strategies, while eschewing costly efforts to exterminate the enemy militarily and politically, or to acquire new territory. The text cites Sun Zi's axiom that the deployment of the army over great distances leads to the impoverishment of the people (126.2.2). This does not mean the text opposes the offensive use of forces beyond the gates. At an abstract level, the text accepts that it is best to be the host or master (zhu ]£) in military affairs. One should take advantage of familiarity with one's own territory, make best use of short lines of communication and supply, and entice the enemy, or guest (ke § ) , to operate under strategic conditions over which it has little control. But being the host does not necessarily restrict one to operating defensively at, or within, one's borders. One can also play the host within enemy territory by plundering the enemy for supplies and by retaining strategic initiative (Liu 1955f: 126). This way the admonition against protracted punitive campaigning is a relative restriction.48 At a pragmatic level, the advice Li Jing gives to the Tang emperor is to launch a punitive campaign against the Gao Li in the east, relying primarily on regular forces.49 In the west Li advocates a more or less static defensive posture, linking signal posts and protective commanderies in a form of zonal defense. He suggests that the Tang make better use of non-Han soldiers in its employ by deploying these forces—skilled in mobile operations—in forward positions along the boundaries. Most Han forces should be deployed in internal areas well within the borders, thus saving food and provisions, presumably by this issue. At one point, in a discussion on the integration of barbarian and Han troops in the Tang border defenses, the text appears to accept the benign Mencian view of barbarians, namely that Heaven had not originally differentiated between Han and the barbarians when they were created. The differences in culture and habits were products of geography, largely. If the Han used kindness and trust, then the barbarians could become Han; enculturation would eliminate any differences. As I have argued, this assumption can underpin nonzero-sum assumptions about conflict that can justify accommodationist approaches to security problems. In the Wen Dui, however, this argument is not taken very far, and does not affect Li Jing's strategic advice (Liu 1955d: 100). 48 Host and guest are thus relative terms. One can be at the same time a grand strategic host, a strategic guest, and a tactical host, or any combination of these. The determining factor is which side controls initiative at which level. This is different from asking which side is on the move, since the parameters could have been established by the other side. 49 The text uses the term "regular forces" (zheng bing I E ^ ) in a number of ways. Regular forces can refer to those that are numerically predominant in an army, those that are used first to engage the enemy head on, those that retain strict formation in battle, are awesome and intimidating, or are used for extended campaigning. In contrast, "irregular" forces {qi bing I s f ^ ) are variously those forces used in decisive moments in war to strike at the enemy's weak points, forces used to attack from the side or back, forces that strike second, proportionately inferior forces, cavalry, etc. Qi forces are used situationally, that is, to exploit strategic opportunities created by the use of zheng forces.

GRAND STRATEGIC PREFERENCES

139

reducing the length of supply lines, and placing soldiers in areas that are arable, hence somewhat self-sufficient. If there is an emergency on the border, the Han soldiers could be moved forward temporarily (Liu 1955f: 100). The punitive use of force in the east and static defenses in the west were not fixed strategies, however. The Wen Dui records, for instance, that Li Jing used a punitive strategy earlier against the western barbarians (ibid., 61). The text implies that a balance between punitive campaigning and static defense is the most efficacious combination. It argues against purely static or passive defensive postures, yet also implies that extensive punitive campaigning is costly and risky.50 Indeed, the text warns against strategies aimed at wholly exterminating the adversary because victory can become a function of fortune more than a result of carefully applied force. The wiser strategy is to "not seek great victories" and thus also avoid "great defeats" (Zeng 1986: 240). This requires first and foremost putting oneself in an indefeatable position (i.e., static defense), and then waiting only for the opportunity to go on the offensive against the adversary (Liu 1955f: 133-34, Zeng 1986: 240). Until this opportunity appears one should "defend against losing" (Zeng 1986: 241). This phrase captures nicely the minimax strategic assumptions behind Li Jing's advice. The decision axiom here is to minimize maximum losses. Put another way, defending against losing is an admonition against offensive/expansionist efforts to destroy the enemy militarily and politically. It is at the same time an admonition to rely on a mix of static defenses and punitive expeditions that eschew the pursuit of great victories. In sum, the argument structures and strategic advice presented in the Wen Dui all suggest that accommodationist grand strategies are essentially irrelevant to the security problems presented in the text. Rather, the text exhibits a preference for violent grand strategies, and within this context, a mix of static defensive and punitive strategies over offensive-expansionist ones. But once again this preference is colored by strategic flexibility born from a deep sense of the constantly changing nature of conflictual interaction with the adversary. The dialectics of conflict prevent too strict an adherence to any a priori ranking of strategic choices. Sun Zi Bing Fa There is in Sun Zi's text very little discussion of he qin—type diplomatic activities. There are only a couple of instances in which diplomatic actions are mentioned; in one case that role is severely conditioned, and in the other it is ambiguous as to whether in fact Sun Zi is referring to the active use of diplomacy. 50

"In military affairs one prizes controlling the enemy; one should not wish [only] to resist him" (Liu 1955f: 127).

140

CHAPTER FOUR

In the first instance, in his chapter on "Nine Types of Terrain," Sun Zi argues that only in "focal terrain" {qu di ffffife) should a state resort to alliance building and accommodation as a tool of security policy. "Focal terrain" refers to territory or a state that is at the hub of primary communications and transportation routes. It is thus strategically vital and the object of attention from many states. Thus if one governs or simply occupies focal terrain, the potential number of enemies is overwhelming. Under these conditions, it is best, argues Sun Zi, to establish close and friendly relations with other states (Wu 1983: 110), though there is no indication that diplomacy alone is sufficient to resolve conflict. Indeed, accommodation is, at most, a prelude to the swift and overwhelming use of force. This relationship is revealed best in the mix of two metaphors used to describe the process of creating and then exploiting opportunities for victory. One should be like a seductive maiden, writes Sun Zi, until the enemy opens the door (when he has exposed his weaknesses); then one should become like a fast rabbit and rush in. Thus the enemy is unable to block the exploitation of this opportunity (Li 1991: 73). A second instance is found in Sun Zi's famous statement, discussed in the previous chapter, that "the best strategy is to attack the enemy's strategy, the next best is to attack his alliances, the next best is to attack the enemy's army, and the worst is to besiege his cities." Almost all contemporary commentators interpret the second part of the phrase to mean using diplomacy to undermine the adversary's alliances, but elsewhere the text suggests that awesomeness (wei jg£) is one tool to prevent the enemy's alliances from coalescing. And demonstrating awesomeness can involve anything from flaunting superior force in order to intimidate the enemy to the limited use of force to strengthen credibility. Whichever interpretation of wei holds, the application of awesomeness is not in the same class of strategic behaviors as those associated with accommodationist diplomacy. That diplomacy plays such a small role in Sun Zi's strategic thinking should not be surprising; as the composite cognitive map indicates (Fig. 3.7), diplomatic actions and nonviolent political strategem are only indirectly causally connected to the defeat of the adversary or the achievement of state security. The resolution of the security problematique ultimately takes place within the context of applied violence. Within this context, however, is there a discernable preference ranking between defensive and offensive strategies? Wei Rulin argues that Sun Zi's concept of "first making oneself undefeatable in order to wait for the enemy to become defeatable" is an inherently defensive posture that eschews the invasion of other states (Wei 1988: 116). There are, in addition, a number of passages that warn against protracted campaigning within other states because these are negatively related to state security; they deplete military strength, drain resources, and undermine the livelihood and morale of the masses (73.4.1, 73.7.1, 74.11.1, 74.13.1). There are also strategic reasons for avoiding extended military operations. Armies that rush forward, lightly

GRAND STRATEGIC PREFERENCES

141

armed, over long distances become strategic guests (ke ^ ) , and possess all the disadvantages therefrom: they have extended lines of communication; their forces may not all arrive in one place at the same time; and some will be insufficiently rested (Li 1991: 40). The cause-effect statements also indicate that being skilled in defense is the basis of being in an indefeatable position (85.5.1); this in turn means the enemy cannot discern one's weaknesses and thus is confused as to where to attack (97.7.2). Defensive preparations are also positively related to general utility, particularly when the enemy is militarily superior (67.21.1, 80.16.1). As we have come to expect, however, Sun Zi's comments on defense should be heavily qualified. First, it is fairly evident that Sun Zi opposes static defense. He admonishes against defensive postures that use "bound horses" and "buried chariots" as linear defense lines. His reasoning, suggests the Tang commentator Du Mu, is that such a posture, while appearing solid, is insufficient to support or increase the valor, fighting spirit, and unity of the army (Wei 1988: 264, 266). Second, for Sun Zi defense and offense, whether at the grand strategic or operational levels of strategy, are dialectically linked. "First putting oneself in an indefeatable position" means waiting for the enemy to reveal its intentions and capabilities, while one's own strategic posture of apparent immobility and obscurity conceals one's own intentions and capabilities. Once the enemy reveals its disposition by committing to an action (possibly after it is forced or enticed into such a move by military or political strategem), one can assess its strengths and weaknesses. At this point one shifts to an offensive posture, striking at the enemy's "empty" (xu fSi) points and disarming it Defense is thus a preparatory stage for offense: it is a stage that outwardly appears passive, but in fact one retains the initiative by establishing the parameters within which the enemy must operate (Shi 1222: 4.1a, Guo 1987: 30, Wei 1988: 115, 117). This relationship is causally linked in the text to the defeat of the adversary (85.5.1, 87.13.3,139.56.1). There is nothing in the text to suggest that the shift to the offense need be restricted to operations within one's own territory (Yu 1990: 13).51 Third, it is also quite obvious that Sun Zi does not oppose the use of military force inside enemy territory. The advice against protracted campaigns is relative, conditioned by logistical problems rather than moral absolutes. The logic of the advice suggests that should these logistical problems be solved, then the admonitions against extended operations would cease to hold. Sun Zi himself 51

Interestingly, Clausewitz sees the dialectic in much the same way. He suggests that the advantage goes to the side that first takes the defense, since it is thus in a position to wait and see. The defense retains the initiative since the enemy is compelled to act first, thus revealing its intentions, and allowing the defender to respond maximally by going on the offensive and hitting the enemy's "center of gravity." It is absurd, writes Clausewitz, to consider defense the "final purpose" of war (Clausewitz 1976: 358, 370-71, Aron 1985: 147-48).

142

CHAPTER FOUR

provides the solution when he recommends "relying on the enemy for supplies" (yin liangyu di g $ i f|jf£) and plundering to support one's forces when deep in enemy territory (zhong di ze lue JSJfelO flO- These sorts of actions are positively related to victory (76.20.1), to sustaining the supplies and physical strength of the army (134.31.1), and to general notions of utility (76.18.176.18.3, 131.13.2, 131.20.1).52 The emphasis on the offensive uses of force, however, does not automatically translate into a preference for offensive/expansionist grand strategies. There is no discussion of righteous war, for instance, as one finds in the other texts of the Seven Military Classics. Indeed, Sun Zi uses the term "righteous" only once to describe the desired quality of a leader who employs spies. The term is never used to describe a concept ofjust war. Often in the other texts the process of conducting righteous war is linked inextricably to the military and political destruction of the adversary and to the annexation of portions of its territory. There are only hints of this in the Sun Zi Bing Fa. In one passage Sun Zi implies that after victory one should plunder the enemy's towns and divide up its people, expand one's own territory and divide up the wealth (Li 1991: 40-41, Zhang 1987b: 202). But the text does not speak of replacing the enemy's political leadership or employing a new cadre of officials from the vanquished state, as the Si Ma Fa does, for instance. The text generally only addresses the invasion and the military destruction of the adversary as an important step in the achievement of state security.53 Is there enough in the discussion so far to draw any conclusions about a grand strategic-preference ranking? We can safely conclude that violent grand 52 Cleary's contention (1989: 14) that traditional Chinese strategic thinking, including Sun Zi's, rejects invasive and aggressive warfare on ethical grounds misses both the pragmatic flavor of Sun Zi's argument as well as its conditionality. This leads Cleary somehow to ignore the numerous descriptions of the legitimate use of military force deep inside enemy states. 53 1 touched on this earlier in chapter 3 in the discussion of the metaphoric language in the text, language that stresses preemptive, overproportional uses of violence. See also the causeeffect statement 106.13.3 where "invading and plundering the enemy like fire" is positively associated with a general notion of utility. The acceptability of the destruction of the enemy state raises an important question of interpretation in the secondary literature. At the start of Sun Zi's third chapter he uses the following phrase\fanyong bingzhifa quart guo wei shangpo guo ci zhi (Kf$^^&i=kMM±.1(&M:#l£-)The text then leads into the passage about "not fighting and subduing the enemy." Most commentaries, including Liu Yin, interpret this to mean that the best strategy is to "take a state intact; to ruin it is inferior to this" (Griffith 1982: 77, Cleary 1988: 66), or some close variant. This is taken as evidence that Sun Zi opposed offensive, invasive, and destructive warfare. I would argue that left alone this interpretation sets up a contradiction with Sun Zi's recommendation to plunder the enemy's countryside and divide up the enemy's wealth. It also contradicts cause-effect states in the text where "invading and plundering the enemy like fire" is positively associated with a general notion of utility. This contradiction only exists because the term quan guo or "intact state" ( ^ @) is taken to mean the enemy's state. In fact, there is nothing in the original Chinese to indicate that the subject of "intact state" is the enemy; it could just as easily refer to one's own state. In the earliest extant annotation of the text, by the Three Kingdoms' strategist Cao Cao, the willing

GRAND STRATEGIC PREFERENCES

143

strategies are preferred to nonviolent or accommodationist ones. We can also conclude that the Sun Zi Bing Fa rejects static defense as the primary means of defeating the adversary and securing the state. As with the Liu Tao, the absence of an explicit notion of righteous war, coupled with the conditional admonitions against extended campaigning within enemy territory, suggest an inherent caution against expansionist grand strategies, though there is a more-or-less clear preference for offensive uses of force.

CONCLUSION

There is little evidence to suggest that the Seven Military Classics embody a grand strategic-preference ranking similar to that found in conventional interpretations of traditional Chinese strategic thought. There is no shared preference for accommodationist or he qin-type grand strategies over either defensive/ punitive or offensive/expansionist ones. This should not be surprising since, as the previous chapter has shown, the texts reflect essentially a parabellum conceptualization of the security problematique. Consistent with this paradigm most of the texts show a preference for violent grand strategies over accommodationist ones. The exception is the San Lue, and this is the one work that least fits the parabellum paradigm. Table 4.1 summarizes the elements of grand strategy that figure prominently submission of the enemy state comes only after one's own forces have invaded enemy territory, occupied it for a while, besieged its cities, and cut off its communications with other states (Wei 1988: 98). There is no indication that quart guo means taking the enemy state intact without resort to at least some level of violence. At least two twentieth-century Chinese annotations and a recent English-language edition of the text suggest that quan guo refers specifically to one's own state, in other words using force while minimizing the costs to one's own state (Zhang 1987b: 111, Wei 1988: 94-95, Ames 1993: 111, 284). This is consistent with the interpretation in a memorial on border security written by a Ming official, Ni Qiu. In the memorial Ni writes that the forces of the emperor should achieve victory by means of "completeness." This, he implies, is the same as Sun Zi's principle of "avoiding becoming exhausted" (HMJSWBa 6:304-5). This interpretation is also consistent with other passages in Sun Zi where restraint in the use of force (i.e., avoiding extended campaigns) is born from concerns about the potentially ruinous economic and political effects for one's own state. The notion that quan guo refers to the enemy state comes from the distinctly Confucian-Mencian spin that most commentators put on this passage. Taking the enemy state intact, without the use of violence, is evidence that a ruler and his strategists cherish the enemy's people, and wish to punish the enemy's ruler while winning over the hearts and minds of his people (See Shi 1222: 3.1a, Guo Ying's commentary in the Qing compilation of commentaries in Zhu 1989: 17, Zhang Yu's commentary in SBSYJZSZ 1978: 1.24). The problem is that the Sun Zi text is one of the least Confucian-Mencian of the seven classics, bereft of any substantial discussion of internal rectification or the righteous use of force. I do not mean to suggest that this ends the debate; I mean to point out only that the view that Sun Zi stresses nonviolence, or noninvasive or nonoffensive strategy rests on contestable evidence and cannot explain the deep structure and language of the strategic arguments of the text.

144

CHAPTER FOUR

TABLE 4.1

Summary of Predominant Elements in Security Strategy military political annexation destruction destruction

accommodation

static defense

invasion

WuZi

yes (a)

no

yes

yes

yes

no

Si Ma Fa

yes (b)

no

yes

yes

yes

yes

Wei Liao Zi

no(c)

yes (d)

yes

yes

yes

yes

San Lue

yes

yes

yes

yes

yes

no

Liu Tao

yes (a)

no

yes

yes

no

no

Wen Dui

no

yes

yes

no (e)

no

no

yes (a)

yes (d)

yes

yes

no

yes(f)

Text

SunZi

Notes (a) in limited number of contingencies as strategem (b) after victory over the adversary (c) idealized, contained within discussion of "victory by means of the Dao" (d) when militarily inferior and as a prelude to offense (e) eschews "great victories" in order to avoid "great losses" (f) brief mention of division of territorial spoils

in the process of defeating the enemy and achieving state security and indicates which elements are found in which text. As is evident, accommodationist strategies play at best a secondary role in the security process. The texts that do mention the role of diplomacy, internal rectification, and other nonviolent political strategies view these behaviors either as preludes to the application of violence, as actions that can consolidate the political gains of warfare, or as part of an idealized strategy that has little relation to the central advice of the text. In four of the seven texts static defense plays a key role in the defeat of the adversary and the achievement of state security. But in at least two of those four cases, static defense is a stage in the process of applying violence against the enemy within enemy territory. All the texts consider the invasion of an enemy state to be a legitimate step in the pursuit of state security, and six of the seven texts explicitly accept the destruction of the enemy's military capability as an important part of this process. This conclusion is corroborated by the language used in the texts to describe legitimate actions against aifenemy and legitimate outcomes of these actions (see appendix B). Among the terms used are at least seventy verbal expressions that describe offensive, invasive, and destructive actions: these include terms such as "attack," "invade," "campaign against," "punish," "penetrate," "execute," "control," "threaten," "destroy,"

GRAND STRATEGIC PREFERENCES

145

TABLE 4.2

Congruence between Grand Strategy and Preference Rankings Strategy accommodation

defense

offense

WuZi

3

2

1

Si Ma Fa

3

2

1

Wei Liao Zi

3

2

1

San Lue

1

2

3

Liu Tao

3

1

2

Wen Dui

3

1

2

Sun Zi

3

1

2

Text

Notes Numbers signify the ranking of each strategy from most preferred (1) to least preferred (3). Kendall's W = .3878 Sign. = .085

"suppress," and "dismember," among others. There are also at least fifteen terms used to describe legitimate outcomes of these actions against the enemy including the "extermination," "execution," "killing," and "destruction" of the adversary. There is less agreement among the texts about the validity of the political destruction and reform of the adversary's state. In most instances the legitimacy of this outcome is associated with a clear and prominent notion of righteous war. There is still less congruence on the question of territorial annexation. Only three of the seven texts appear to include this strategic action as a part of the process of defeating the enemy and achieving state security. Two of these three cases also accept the validity of political destruction. In all there is a high degree of agreement on the role of offensive, invasive uses of force. The question is whether this adds up to overall congruence in the grand strategic-preference rankings in the texts. In other words, is there evidence for the existence of a strategic culture as I have defined it? Given the absence of unanimity on the issues of territorial annexation and the political destruction of the enemy, and given differences on the role and importance of static defense, one might suspect that there is not enough agreement on all the key elements of each grand strategy to constitute a basic congruence in preference rankings.

146

CHAPTER FOUR TABLE 4.3

Congruence between Grand Strategy and Preference Rankings Strategy Text

accommodation

defense

offense

WuZi

3

2

1

Si Ma Fa*

3

1

2

Wei Liao Zi

3

2

1

San Lue

1

2

3

Liu Tao

3

1

2

Wen Dui

3

1

2

SunZi

3

1

2

Notes Numbers signify the ranking of each strategy from most preferred (1) to least preferred (3). Kendall's W = .4285 Sign. = .051 * Alternative ranking

Statistically, however, the degree of congruence is moderate but significant both statistically and substantively (see Tables 4.2-4.5).54 There is, then, enough variation in the preferences for defensive/punitive and offensive/expansionist grand strategies to conclude that evidence for a strategic culture is not as strong as one might think. The differences over whether security requires the utter political and military destruction of the enemy and territorial annexation or whether offensive, punitive force aimed primarily at the military destruction or weakening of the enemy is sufficient suggests two conclusions about Chinese strategic culture. The first is that this apparent difference between preferences for different degrees of destruction indicates some variation in the hardness of the parabellum paradigm across texts (see Fig. 54

As Table 4.2 indicates, a Kendall's W test for congruence between multiple rankings shows a coefficient of congruence (W) of only .39, which is not significant at the .05 level, though it is significant at the .1 level. Thus the statistical significance of congruence among the texts on strategic-preference rankings is somewhat problematic. In Table 4.3 I use an alternative ranking for the Si Ma Fa since either ranking seems plausible. The coefficient of congruence (W) rises somewhat to .43. This time the coefficient is significant basically at the .05 level. In Table 4.4 I use the ranking for the Si Ma Fa from Table 4.2 and a different, though plausible, ranking for the Liu Tao. In Table 4.5 I use the ranking for the Si Ma Fa from Table 4.3 and the alternative ranking for the Liu Tao. In both cases the coefficient of congruence (W) is .39 and is not significant at the .05 level, but remains significant at the .1 level.

GRAND STRATEGIC PREFERENCES

147

TABLE 4.4

Congruence between Grand Strategy and Preference Rankings Strategy accommodation

defense

offense

WuZi

3

2

1

Si Ma Fa

3

2

1

Wei Liao Zi

3

2

1

San Lue

1

2

3

Liu Tao*

3

2

1

Wen Dui

3

1

2

Sun Zi

3

1

2

Text

Notes Numbers signify the ranking of each strategy from most preferred (1) to least preferred (3). Kendall's W = .3878 Sign. = .085 * Alternative ranking

4.2). Specifically it indicates that there is some variation along the dimensions of the three elements that constitute the central paradigm of a strategic culture. Any shift away from the high extreme on any of these dimensions would soften the parabellum nature of the assumptions behind strategic preferences. The variation in this case is along the "efficacy of violence" dimension. There is little disagreement in the military classics about the ubiquitousness of conflict or about the zero-sum nature of security threats. But in some texts the efficacy of violence is qualified by logistical constraints on the use of force against the enemy. Although the texts may view the political and military destruction of the enemy as essential for the resolution of conflict, some are more obviously concerned about the logistical and mobilization problems involved, and thus are less clear about issues such as the political destruction of the enemy or territorial aggrandizement. To the extent that soft and hard parabellum assumptions found in the strategic texts are reflected in and compete within the strategic decision-making process, one would expect debate to revolve largely around whether strategy should be punitive or expansionist and, within these parameters, around the degree to which logistical issues are considered central or soluble.55 55 This suggests that a reconceptualized notion of strategic culture should combine Wildavsky's notion of culture as preferences, and Laitin's idea of culture as shared points of concern that are

148

CHAPTER FOUR TABLE 4.5

Congruence between Grand Strategy and Preference Rankings Strategy accommodation

defense

offense

WuZi

3

2

1

Si Ma Fa*

3

1

2

Wei Liao Zi

3

2

1

San Lue

1

2

3

Liu Tao*

3

2

1

Wen Dui

3

1

2

Sun Zi

3

1

2

Text

Notes Numbers signify the ranking of each strategy from most preferred (1) to least preferred (3). Kendall's W = .3878 Sign. = .085 * Alternative ranking

However, if we loosen the distinction between punishment and extermination, then the Seven Military Classics show a greater measure of congruence than otherwise. That is, except perhaps for the San Lue, the texts share a preference for offensive uses of military force over diplomatic, accommodationist strategies.56 To the extent that theparabellum paradigm is reflected in strategic decision-making processes, accommodationist and static defense postures of security ought not to be considered acceptable long-term alternatives to the offensive use of violence. There is, however, a caveat. And that is whether the notion of quan Man undermines the possibility of a generalized preference for offensively oriented (punitive and expansionist) grand strategies. As I suggested earlier, most of the texts implicitly or explicitly embody a flexibility axiom. The Si Ma Fa debated. Within a shared metapreference for violent grand strategies there is room for differences over whether violence should be limited and punitive, or unlimited and expansionist. 56 Xin Jiaxuan, a southern Song writer, asked in his collection of essays, "which is more advantageous, sending forth troops to attack the enemy, or sitting and waiting for the enemy's attack? Which pays off, righting on enemy territory or retreating and fighting on one's own territory? Generally, when war can't be avoided, nothing is better than sending forth your army first in order to fight on enemy territory. This is certainly the best strategy under heaven, and a strategist's superior plan" (cited in Wu 1988b: 386-87).

GRAND S T R A T E G I C P R E F E R E N C E S

149

A

High

Frequency of conflict in human affairs

Low

High

Zero-sum nature of conflict

A = parabellum (hard realpolitik) B = idealpolitik

Fig. 4.2. The Central Paradigm. summarizes it nicely when it states simply, "As for war, it is [a question of] 'expedient assessment'" (Liu 1955b: 33).57 The notion of absolute flexibility also pervades other texts outside of the Seven Military Classics. Flexibility is essential because the constant process of change in conflict situations requires a constant awareness of the appearance and disappearance of opportunity. The Lu Shi Chun Qiu remarks, "If one is knowledgeable about strategems then one understands the transformations of time [opportunity]. If one understands the transformations of time, then one understands the interchangeability of weakness and strength, flourishing and declining, and one understands how to evaluate first and last, distance and closeness, when to advance and when to hold up" (Lin 1986: 208). Gui Guzi, a Warring States philosopher, wrote, "In affairs one stresses controlling the enemy and not being controlled by the enemy. As for controlling the enemy, [it is a question of] mastering the way of versatility" (Li and Li 1990: 50). He Qufei argued in his work on strategy, "Thus success in all actr * s under heaven invariably arises out of the calculation of change, and lack of success invariably arises from inflexibility" (Gong 1989: 90). The Ming strategist He Liangchen devoted a whole chapter to "Adapting to Circumstances" in which he argued, "Therefore, one who excels in military affairs adapts to the enemy and exploits change, adapts to the enemy and adopts differing policies, adapts to terrain and creates momentum, adapts to [the enemy's] situation and creates 57 The phrase is "fan zhan quanye" ( / L U S i f e ) . Here I use Lewis's (1990: 118) translation oV'quan bian" The Chinese dictionary, the Ci Hai, defines quart bian as "adapt to opportunities and respond to change" (suijiyingbian

1 50

CHAPTER FOUR

his [own] form" (Chen 1984: 199). In another Ming strategic text, the Ton Bi Fu Tan (JxitJpf §£), one finds a similar argument about the crucial role of flexibility: As for the way of employing the military instrument, there should be no constant pattern in being "host" or "guest"; there should be no constant form in either attacking or defending; there should be no constant rules for either dispersing or uniting forces; there should be no constant time period when one is in motion or at rest; there should be no constant directional momentum, either when extending or retracting [one's forces]. By constantly changing when one appears and disappears the enemy will be unable to discern [your intentions and capabilities]. This is what is known as critical ingenuity in military affairs." (Academy of Military Science Editorial Group 1984: 87) The flexibility axiom, therefore, seems to preclude predetermined strategic choices. The Chronicles of the Three Kingdoms contains the simple statement, "[Given] change and transformation in military affairs, there is certainly not one method [in strategy]" (Wu 1988b: 321). The Song military encyclopedia, the Wu Jing Zong Yao (S^IMIHIc), summarized the pragmatism of other strategic thinkers and concluded that after gauging the enemy's situation, "if it is possible [i.e., advantageous] to go to war against the enemy, then go to war. If it is possible to establish peace, then establish peace. If one sees benefit, then advance and attack. If one discerns difficulties [disadvantages] then retreat and defend. Sometimes it is best to penetrate deeply, sometimes it is best to penetrate only slightly" (Zeng 1043: chuan 2.9).58 Does, then, the notion of quan Man undermine any a priori preference rank58

Oddly, the centrality of quan bian and absolute flexibility in traditional Chinese strategic thought is recognized by some of the secondary literature, particularly Chinese scholarship. At the same time that this scholarship recognizes and endorses the importance of this, however, it also persists in the view that over all there is a preference for non- or minimally violent approaches. For modern Chinese students of traditional strategic thought the flexibility axiom translates into a rejection of a priori limits on the means and methods of defeating an adversary. According to Li and Sun, although the strategic goal may be connected to benevolence and righteousness, one cannot adopt benevolent and righteous means for dealing with the enemy (Li and Sun 1989: 31). Yang Bing'an writes, "Towards the enemy one must implement deceptive strategems and extraordinary plans, and cannot speak of benevolence, righteousness or morality" (Yang 1986: 10). Zeng Guodan reaches a similar conclusion: "It is obvious that the goals of warfare generally arise from righteousness. But the methods of warfare are only for achieving results; one cannot therefore speak of righteousness" (Zeng 1972: 93). See also Xu 1986: 24-25, Guo 1987: 30, Wu 1989: 54, Ma 1990b: 8, Liu 1990b: 150). As I was told by a retired Guomindang general and by scholars at the Academy of Military Sciences in Beijing, in warfare quan bian takes precedence over righteousness (Interview 1989b, Interview 1991b). Lewis is one of the few Western scholars to acknowledge the importance of quan bian, but he seems to confine it to a tactical principle. Boorman and Boorman (1967: 149) and Pye (1988; 82-83) also recognize the role of flexibility in Chinese strategic behavior, but they do not trace this back to quan bian.

GRAND STRATEGIC PREFERENCES

15 1

ing, including one that places offensive violence on top? In most instances the flexibility axiom translates into a recognition that under certain expedient circumstances (usually conditions of military or political weakness and inferiority) nonviolent behaviors are acceptable means for creating conditions for the more efficacious application of military force. The texts urge strategists, whether at the grand strategic level or at the level of tactical operations, to exercise patience, respond to the enemy's actions, determine or weigh (quan) the constantly changing nature of the conflict (bian) with the enemy, and then at the opportune moment strike violently and decisively at the enemy's exposed weaknesses (cf. Shi 1222: 35.20b and 31.9a). Nonmilitary or nonviolent strategies, therefore, are instrumental, serving the purposes of violent, usually offensive and preemptive uses of force. On the surface this conceptualization of flexibly responding to the enemy implies an initial passivity and a predilection for striking second. In fact, however, the texts argue that adapting to the enemy's nature is a means of retaining the initiative. To the extent this is symbiotically linked to the lightning exploitation of opportunities to destroy the enemy, waiting and seeing and only then moving against the enemy are steps in the same preemptive process. As the Sun Zi BingFa notes succinctly, "move after the enemy, and arrive [at point of attack] before him" (Liu 1955b: 32, 54, Liu 1955c: 24, Xu 1986: 194, Wu 1989: 54-58, Xie 1990: 42-43, Li and Li 1990: 49-57).59 This implies two things. The first is that quan bian is not an axiom about flexibility within the context of nonviolent or minimially violent preferences. Rather, it is an axiom about flexibility within the context of conflict, a context within which there is a high expectation of violence due to the nature of the parabellum assumptions. Within these expectations of violence, a key decision rule is to be sensitive to changing opportunities to defeat the enemy (and, by implication, opportunities for the enemy to defeat you). So one fights when conditions are optimal—meaning when gains are maximized and losses are minimized. When and how to fight are determined by quan bian, but there is still an expectation that one will most likely have to fight. The second implication, then, is that the determining factor in weighing strategic choices is whether one is capable of defeating the adversary. This capability, according to the texts, is multifaceted: domestic cohesion, economic and military resources, military training and morale, the skills of the general, among other elements, are all important considerations. But the nature of quan bian—weighing changes in the interactive relationship with the enemy (Shi 1222: 31.6a)—suggests that any composite assessment of these factors is a 59 A work on strategy written by another Warring States strategist, Sun Bin, puts it nicely: "Silently waiting, this is also movement" (Xin 1985: 127). On the importance of preemption in other strategic texts see the Ming period Cao Lu Jing hue (Anonymous 1986: 233, 284) and Bai Zhan Qi hue (Zhang 1987a: 113) and the Tang text by Li Quan, the Tai Bai Yin Jing (Li 1957: 2, 5:58).

152

CHAPTER FOUR

relative one; the absolute levels of any of these indices are irrelevant. What counts is change relative to the enemy's measurement along these dimensions.60 This is, in effect, a structural view of strategic choice. In other words, as the relative balance of these factors shifts in one's favor, the efficacy of violence increases as does the value of strategies that lie at the violent, offensive end of the spectrum. In this respect, quan Man is an explicit admonition to strategists to act rationally in the expected-utility sense of rationality. Put another way, quan Man is an axiom that urges decision makers to calculate the optimal way of setting up conditions for the defeat of the enemy. Since the texts tend to argue that as a practical matter the defeat of the enemy is not likely to happen without the application of violence, the calculation of expected utility therefore does not refer to the expected utility of violent versus nonviolent paths to the enemy's defeat, but instead refers to the expected utility of nonviolent paths to optimal conditions for the application of violence versus that of the direct application of violence. The notion of quan Man, linked to the causal structures of the Seven Military Classics, eliminates any possibility of an a priori grand strategic-preference ranking in which accommodationist choices rank higher than defensive and offensive choices. But it does not preclude the possibility of a grand strategicpreference ranking in which offensive uses of force (whether punitive or expansionist) are loosely ranked higher than static defensive and accommodationist choices. We come back, then, to the initial question in this concluding section: Do the parabellum paradigm and this looser preference ranking together constitute a single Chinese strategic culture, or along with an apparently different preference ranking coming out of the Confucian-Mencian central paradigm, do they constitute a strategic subculture, and if so, which is dominant and which is subordinate? The answer to the first two questions rests on a decision as to whether the looser preference ranking that emerges from the Seven Military Classics fits the criteria for a strategic culture outlined in chapter 2. Certainly there is a high degree of congruence across the objects of analysis about the preference for offensive, preemptive uses of force. And this preference is consistent with the core assumptions in the central paradigm. But there are also apparent differences over alternative subchoices within the broader strategic preference (i.e., between punitive and expansionist uses of force). To the extent, however, that one can still derive from this looser preference ranking more-or-less determinant predictions about whether to expect strategic culture-influenced decision makers to prefer using force offensively as opposed to pursuing static 60

On this score, the Chinese case confirms Vasquez's hypothesis that realpolitik worldviews will tend to stress relative capabilities. "A social construction of reality that is based on principles of power politics is going to place more of an emphasis on capability because of the greater role given to warfare as a means of resolving intractable issues" (Vasquez 1993: 118).

GRAND STRATEGIC PREFERENCES

153

defensive or accommodationist strategies, then let us for the moment accept that the parabellum paradigm and the looser preference ranking do constitute a strategic culture. Does this mean, then, that there are two Chinese strategic cultures or subcultures? Certainly the Confucian-Mencian paradigm does embody a preference at least for accommodationist strategies over violent ones, and there is some evidence to suggest that within the violent end of the spectrum there is a preference for defensive over offensive choices. In the Seven Military Classics, however, this paradigm and concomitant preferences are idealized and disconnected causally from, or only indirectly related to, the defeat of the adversary and the security of the state. The language of this strategic culture, it seems, is almost perfunctory or symbolic, since it appears to have little effect on or relationship to the parabellum decision rules and axioms in the texts. Arguably, then, there are two strategic subcultures. Which one is dominant in the decision-making process is in principle an empirical question that could be answered by testing these subcultures in alternative models of strategic choice, but let's assume for the moment that my procedures and analysis are wrong, and that I have simply missed the operative role of the axioms associated with the so-called idealized strategic culture. One test of whether the deeply held decision axioms are consistent with the Confucian-Mencian or the parabellum strategic culture would be to ask questions about specific strategic contingencies (i.e., different levels of threat, different balances of capabilities, etc.) and how decision makers ought to respond to these. If the answer is in all contingencies the same—namely, to choose the least coercive options consistent with the Confucian-Mencian assumptions—then the former strategic culture is arguably the operative one. But if the answers stress options other than accommodationist or minimally violent ones, then arguably the deeply held, reflexive decision rules do not come from the Confucian-Mencian paradigm.61 Chapters 6 and 7 on Ming grand strategy take up this test in more detail, but the short answer is that Ming decision makers and strategists did not advocate the same strategies for different strategic contingencies, and for the most part rejected accommodation and minimalist violence as a basic solution to their security problematique. When Ming decision makers did consider these options they were, in effect, a subset within a range of strategies, subordinate to and less preferred than more offensively oriented choices, but considered efficacious in the short term under specific circumstances. Thus the relationship between the two strategic cultures is not one between alternative or alternating influences on strategic choice. Rather, they may be situated in a much more complex relationship to one another. Put simply, the deep structure of the Seven Military Classics reflects the parabellum strategic 61

For a brief discussion of this method for eliciting semiconsciously held, operative decision axioms and beliefs, see Ott 1989: 41-44.

154

CHAPTER FOUR

culture. But overlying this deep structure, clothing it, is a different set of languages, logics, and decision rules that are consistent with the ConfucianMencian strategic culture. Exploring this relationship empirically is not as straightforward as merely setting up and testing alternative models of choice. It requires first a return to theory in order to help explain this relationship, followed by a look at the ramifications for empirically testing the effects of strategic culture on decision making.

Chapter Five A RETURN TO THEORY

T

HERE ARE AT LEAST two different sets of clear strategic preferences in the Seven Military Classics that derive from two different central paradigms. These constitute two strategic cultures. One, derived from a broad Confucian-Mencian central paradigm, places nonviolent, accommodationist grand strategies before violent defensive or offensive ones in a ranking of strategic choices. This preference ranking is associated in the texts with language that reflects the Confucian-Mencian stress on benevolence, righteousness, virtue—or good government—as a basis of security. This language casts military force as "inauspicious," to be used only under unavoidable circumstances. It stresses "softness" overcoming "hardness" (in this case derivative as well from Daoist traditions) and the submission of the enemy without resorting to force. The other, derived from a parabellum paradigm, generally places violent, offensive strategies before static defense and accommodationist strategies; whether offense for punitive or expansionist purposes is preferred depends on the softness or hardness of the parabellum axioms in a particular text. This paradigm characterizes the external environment as dangerous, adversaries as dispositionally threatening, and conflict as zero sum. Under these conditions, the application of violence is ultimately necessary to deal with threats. But these two sets of preferences do not stand in the texts as two separate but equal strategic cultures. Rather, the Confucian-Mencian language represents an idealized level of discourse. For one thing, none of the texts (with the exception of the San hue) devotes very much space to any detailed explication of the concrete application of Confucian-Mencian concepts of security. For another, a number of the texts, along with some historical commentaries and annotations, relegate these vague strategic axioms to indistinct golden ages of sage kings and legendary rulers, and suggest the historical and strategic irrelevance of these axioms. Finally, as the cognitive maps, argument structures, and metaphorical or analogical language of the texts indicate, critical security concepts in the Confucian-Mencian paradigm are causally disconnected from or only indirectly related to desirable outcomes such as the defeat of the adversary or the security of the state. That is, the Confucian-Mencian paradigm is disconnected from the operational advice, axioms, and decision rules that derive from the parabellum paradigm. But the Confucian-Mencian paradigm is still there. What, then, is the relationship between these two different strategic cultures? What purposes does

1 56

CHAPTER FIVE

the idealized paradigm serve in these texts, and—if this hierarchy of strategic cultures is reflected in decision processes—what purpose does it serve in strategic choice? What in principle does the presence of idealized and operational sets of strategic preferences do for empirically testing the effects of strategic culture on decision making?

T H E STRATEGY OF SYMBOLS AND SYMBOLIC STRATEGY

That traditional Chinese strategic thought consists of an idealized set of languages, logics, and axioms as well as an operational level should not be surprising in light of a growing body of literature in social psychology, anthropology, organizational culture, linguistics, and political science on the role of symbols in human behavior. Some of the propositions and conclusions of this disparate literature are more relevant to the Chinese case than others, but I present the overlapping elements of these literatures because all have implications for strategic decision making across cultures. Essentially, the literature suggests that symbols can be used for at least three major related purposes directed inter alia at the self, the group, and relationships between groups. Regarding the self-directed use of symbols, studies in organizational culture suggest that elites use symbols or symbolic strategies to create or reinforce a sense of competence where decisions—regardless of the process or outcome of these choices—are symbolically successful because they are tied to a legitimating legacy. In this case, strategic plans devised by organizational leaders and managers consist largely of symbolic language and decisions that are primarily for the purposes of autocommunication. Unlike the conventional conceptualization of strategy as a rational process that "matches internal resources to environmental opportunities and threats" (Green 1988: 22), autocommunicative strategies are not specifically designed to be implemented or to be used in the organization's interaction with the outside, but rather are designed to reinforce the sense of competence and hence legitimacy held by decision makers. When designing plans that accord with preconceived notions of what competent decisions look like, how these plans are implemented is secondary. Of course, people are still making decisions that are implemented, but these decisions are likely to be arrived at less consciously, often by different levels of decision makers, and are likely to conform to entirely different sets of decision axioms. Autocommunicative strategic plans are mantras or espoused theories. They are mirrors that indicate to the decision makers and their organizations, "This is what you should look like" (Broms and Gahmberg 1983: 490). This image of correct decision making may or may not have any connection to the "theories-in-use" that actually determine the operational behavior

A R E T U R N TO T H E O R Y

157

of the group (ibid., Westerlund and Sjostrand 1979: 42-45,120-22, Gahmberg 1987: 398). In the realm of security studies Luke (1989) implies that deterrence theory serves a similar autocommunicative purpose. "As myth deterrence theory presents an idealized ahistorical story of how strategic actors supposedly do behave, creating for decision makers a representation of how they should behave in managing national security" (214). The more decision makers and strategic thinkers accept and conform to this myth, the more their plans take on an air of authority and competence. Klein argues, in parallel with the autocommunication literature, that there exists a gulf between "declaratory" strategic doctrine and "action" policy. Declaratory doctrine explains what strategic-decision makers want their decisions to look like. In the American case, strategic doctrine is thus presented in terms of its defensive, deterring, or retaliatory purposes. Action policy (nuclear targetting plans, single integrated operational plans, etc.), on the other hand, as a "theory-in-use" has largely stressed offensive war-fighting and preemptive nuclear warfare (Klein 1988: 138, Pringle and Arkin 1983, Sagan 1987). Another use of symbols by elites is outwardly directed at nonelites with the same organization or state. In this case, elites create an "official language" of discourse (Bourdieu 1991: 41-65), or "managerial ideology" (Weiss 1986: 35-47), or "administrative ideology" (Urban 1982: 5), that excludes alternative strategies, undermines challenges to their authority and legitimacy, mobilizes support and otherwise reinforces their hegemony in the decision process (Eisenberg and Riley 1988: 136-39, Edelman 1977: 11; Laitin 1988: 591). Symbols or symbolic language act as a source of authority. Those who wield these symbols are recognized as competent and legitimate authorities, and hence others are more likely to accept the correctness of strategic decisions regardless of their nature or consequence. Symbolic language in a strategic discourse, then, serves the same function as other symbols of authority such as a uniform, religious clothing, and job titles. As social psychologists have discovered, people are more apt to accept the decisions of putative authorities if these are clothed in authoritative symbols, even if these decisions appear incompatible with some of the interests of those submitting (see Cialdini 1984: 208-10,215-23). Hierarchy is reinforced, power structures are legitimated, and potential challenges to authority are preempted or coopted either because these alternatives are symbolically beyond the pale or because the potential challenger must adopt the delimiting discourse of the official language in order to participate in the group (Conrad 1983: 186-92, Wuthnow et al., 1984: 37). Thus the power of symbols comes in part from their preexisting authority in the eyes of both their producers and consumers (Bourdieu 1991:75-76). Credible participants in a strategic discourse are those who adopt the symbolic language of that discourse. Like debating rules in parliaments, strategic symbols

158

CHAPTER FIVE

authorize or prevent participation in the discourse. This does not necessarily mean that all debate is thus eliminated within these restrictions, but it does mean that those with alternative visions and strategies that cannot be framed comfortably in terms of this symbolic discourse are excluded (ibid., 58). For instance, although there were differences within the American strategic studies and policy communities during the Cold War between proponents of nuclear war-fighting and strategic defense on the one hand and proponents of assured destruction—based notions of deterrence on the other (Jervis 1984, Williams 1992: 68), both groups were debating ways to enhance the stability of nuclear deterrence. Both sides accepted that nuclear weapons were, depending on deployment posture, efficacious tools for dealing with a threatening Soviet Union. Within the symbolic boundaries of these assumptions there was very little opportunity for proponents of minimum deterrence, unilateral disarmament, or social defense and others who fundamentally rejected the premises of this strategic debate to enter into it or challenge it.1 This argument about the hegemonic roles of strategic discourses is not the monopoly of extremist postmodernists. Stephen Walt, a scholar who could hardly be considered outside the realist mainstream of strategic studies, has also argued that strategic elites have bureaucratic, political, and personal interests in delimiting strategic debates, just as those outside of the debates who want to participate have self-interested reasons to conform to the official discourse. Far from being a "free exchange of ideas," therefore, debates on strategy are heavily influenced by the political and organizational interests of the participants. In other words, much of the strategic community has a greater interest in defending their positions than in pursuing truth. . . . Potential critics are also deterred by the fear of being ostracized. We should not underestimate these social and political pressures; few of us would welcome being entirely excluded from the "respectable" strategic studies community. If one values being part of an active policy community (which can bring invitations to conferences, increased access to important journals, offers to perform consulting work, and greater access to research funding), then one's willingness to criticize prevailing wisdom probably declines. (Walt 1987: 147-48) Jack Snyder is another who has made the argument that strategic ideologies— 'Luckham suggests that the net effect of this restricted debate was the creation of an ideology (or "armament culture") that justified the strategic hegemony of security intellectuals, defense policymakers, arms manufacturers, and all those who accepted this link between the Soviet threat, nuclear weapons, and U.S. security. This ideology was designed "to neutralize the premonition of danger and to convert this fear into a source of power and profit for those who control the institutions under which we live." This was done by setting up an equivalence between security, defense, and armament, such that "we will never be safe until we have established overwhelming military superiority over all possible enemies" (Luckham 1984: 4, Klein 1988: 134, 139, Luke 1989: 221-23).

A RETURN TO THEORY

159

analogous in many respects to the postmodernists' notion of strategy as symbol—are wielded instrumentally by elites. Strategy is "bound up with the social order, the political balance of power within it, its legitimation, and the justification of policies favored by particular social groups" (Snyder 1991: 31, Kupchan 1994). In Snyder's view, historically oligarchies of economic and military elites have used these beliefs to justify self-serving policies of imperial overexpansion. Pro-expansion groups monopolized specialized knowledge of the threat environment and the conditions for security, thus justifying strategic choices and preempting challenges to these decisions by those who could not make a similar claim to knowledge (Snyder 1991: 35-36). One of the myths in the expansionist ideology is the "faith in threats" myth, namely a faith in the efficacy of military force to compel the submission of external threats (ibid., 5). From this myth follow arguments about the necessity of using military force against potential enemies, and the advantages of preemptive expansion. Other myths of empire embody zero-sum interpretations of the game being played with the adversary and faith in the advantage of the offensive, notions that I have argued inhere in & parabellum view of security (ibid., 3-5,78-79). 2 Symbols are also critical tools for political entrepreneurs, be they politicians, corporate leaders, or revolutionaries, if they are to mobilize potential constituents. The leadership credibility of these entrepreneurs depends on their "visible commitment to an ethos" (Johnson 1991: 100-104), a set of symbols that even if inchoate must preexist. Whether over time the entrepreneurs try to change or maintain this ethos they must at least buy into it first in order to convince followers of the legitimacy of their claim to authority. As Peng Pai, the first communist leader of a peasant movement in China discovered, his initial efforts to mobilize peasants in the early 1920s fell on deaf ears when he exhorted them while wearing the clothes and speaking the language of an urban intellectual. Only after he changed these and used puppet shows, a phonograph, and magic tricks to entertain did he begin to attract positive attention from the farmers (Hofheinz 1977:146-50). Indeed, as the literature on leadership suggests, leaders must first manipulate positively held symbols to build up capital among followers. This allows her/him to deviate later from the demands, interests, and traditional patterns of behavior of the organization or group (Hollander 1985). 2 Note the similarities in Snyder's myths of empire and Luckham's notion of an armament culture. I should point out, however, that there is a big difference between Snyder's argument that these parabellum myths are merely justifications for the decisions based on the economic and political interests of pro-empire groups and my argument that parabellum axioms are exhibited in the argumentation structures of the Seven Military Classics. While these parabellum axioms may be based on myths about the nature of the threat environment, they are, it seems to me, internalized decision heuristics. Indeed, in the case of the Seven Military Classics the justification myths are the idealized images of security through internal rectification, not the parabellum axioms.

160

CHAPTER FIVE

In the security field, Regan (1992) has found that powerful symbols of militarization and patriotism, whether consciously manipulated by decision elites or not, are positively related to the militarization of economies and societies. "The findings for the United States do point to a fairly strong relationship between the creation and maintenance of symbols in society, the maintenance of the perception of threat in the media, public attitudes toward the military, and the extent to which the United States can be considered militarized" (175). Given this relationship, decision elites that benefit from the militarization of domestic policy or external behavior have a self-interest in perpetuating symbols that mobilize support for, or at least acquiesence to, these policies. A third purpose behind strategic symbols is also outwardly directed, but it has a bearing on the relationship between a group, organization, or state and those outside of it. This has to do with the creation and perpetuation of a sense of in-group solidarity directed, consciously or not, at would-be adversaries. Bormann argues that the creation of a political community first requires the formation of a "rhetorical community" bound together by shared myths and languages that underscore the uniqueness of the community (Bormann 1983: 100-106, Pettigrew 1979: 576, Eisenberg andRiley 1988: 136-39, Green 1988: 11). Since uniqueness, like power, is relative, the process of defining a sense of community by definition also establishes who does not belong in the community, who does not share its purposes, and who is a potential threat to it. The more the language of group discourse creates a perceptual distance between the in-group and the other, the more legitimate are any actions, particularly coercive ones, directed at the adversary (Campbell and Levine 1968: 552). In Edelman's words, "One of the most frequent forms of political categorization is the definition of some large group of people as so serious a threat that their physical existence, their most characteristic ways of thought and feeling, or both, must be exterminated or ruthlessly suppressed" (Edelman 1977: 32). Ingroup solidarity requires the dehumanization of the out-group, and the more it is dehumanized the easier and more necessary it becomes to relax the boundaries on the types of actions taken against the enemy. Thus, the content of an idealized or symbolic strategic discourse is less important than its existence, since it is its mere existence, and its claim to uniqueness, that creates the distance between the group and the adversary.3 At the same time, symbolic language can rationalize group behaviors that are otherwise inconsistent with the self-professed preferences of the group or 3

This suggests, then, that any society or state that shares a parabellum strategic culture will also tend to share an idealized strategic culture that, regardless of its concrete content, is designed to accentuate "us-them" differences, and therefore justify parabellum behavior. Similarly, any group with an idealized strategic culture that in effect underscores the group's uniqueness will also tend to exhibit parabellum behaviors, since these are perceived to be the most efficacious in dealing with zero-sum conflicts created by the idealized strategic culture.

A RETURN TO THEORY

16 1

"disguise the motives for social action" (Urban 1982: 8, Eiser and van der Pligt 1988: 16, Eisenberg and Riley 1988: 136-39). This can be done simply by renaming objectionable behaviors with symbolic terms that are culturally acceptable to the group. Empirically identical behaviors become differentiated by their labels, and the authoritativeness of these labels allows groups to act in ways that they might otherwise oppose (see Luckham 1984: 15). Myths used to describe both the behavior of the group and that of the adversary can "explain and reconcile contradictions between professed values and actual behavior" (Pettigrew 1979: 576), thus providing both a psychological resolution of dissonance and public justification for behavior.4 Kull found in his interviews of American security intellectuals, for instance, that Manichean images of the U.S.-Soviet relationship were used—sometimes even consciously—to mask or deny the reality of mutual vulnerability and mutual threat. This denial was necessary in the view of some of his informants because to accept that the structure of the superpower relationship created mutual threat would be to deny the moral correctness of the American responses to the Soviet threat and thus would remove the moral justification for American strategic war plans (Kull 1988: 308-16). Alternatively, as attribution theory might suggest, if the discrepancy between proposed behavior and self-professed values has to be acknowledged openly, it can be rationalized by an appeal to the alleged inevitability of the circumstances. The group has no choice but to act, because the disposition of the adversary is such that all reasonable responses except one have been eliminated. This being the case, the choice is not really a choice, and hence cannot be subject to sanction for contravening the alleged values of the group. These justifications are products of symbolic portrayals of the nature of the group and the adversary or, through incantation in strategy, are symbols in and of themselves. Recall the ubiquitous phrase in Chinese strategic texts: "war is an inauspicious instrument, and is only used under unavoidable circumstances." In sum, based on this brief review of a wide-ranging literature, we should not be too perplexed by the apparent disjuncture between an idealized set of grand strategic preferences and the operational set in the Seven Military Classics or in strategic decision processes that may reflect the argument structures of these texts. The literature suggests that this disjuncture could serve a number of purposes, from reassuring strategists and decision makers of their own 4 In this respect the incantation of symbolic language to justify behaviors takes on a ritualistic air. According to Kelman and Baron, this is one important way of dealing with cognitive dissonance: "A person may be able to avoid a sharp confrontation between two incompatible elements and thus maintain the inconsistent relationship between them by engaging in certain ritual behaviors that hide the inconsistency without removing it. Such mechanisms are particularly likely to come into play when there is a conflict between what a person knows he ought to do and what he is able or prepared to do" (1968: 673).

162

CHAPTER FIVE

competence and authority, to establishing this authority vis-a-vis the broader community, to justifying strategies against an adversary in culturally and psychologically acceptable language. It is clear, however, that not all of these explanations are equally applicable to the Seven Military Classics. It is also clear that any explanations for this disjuncture are tentative, given that the ethnographic and controlled experimental research that undergirds much of the literature cited above obviously cannot be reproduced in this case. (Moreover, it seems reasonable that those explanations that more plausibly apply to the Seven Military Classics as texts may or may not apply to the Ming decision process.) But before examining which explanation(s) seems best to account for the tension between China's idealized and operational strategic cultures, I want to look at some of the implications of the literature on symbols for decision making and strategic choice because these have a bearing on the implications of the idealized-operational disjuncture for decision making in the Ming dynasty. The short of it is that within the literature there appear to be two very different conclusions about the effects of symbolic strategy on choice. One view is that symbolic strategies restrict a group's strategic options and thus delimit behavior; the other is that there is no a priori reason to expect such a restriction. I would argue that in the Chinese case the latter is a more accurate description of the relationship between the idealized strategic culture and behavior. Concerning the first conclusion, the literature suggests that restrictions on strategic choice will come from three different sources. One source is the organizational or political interests of strategic elites. As Walt, JSnyder, Klein, Luke, and others have argued, military organizations (including the intellectual security community in the U.S.) have a self-interested preference for security strategies that stress the employment of military means (though not necessarily the use of violence). To the extent that military organizations, or decision makers who accept the paradigmatic assumptions of militaries, dominate the strategic decision-making process, the boundaries of strategic debate will be set by their language, logics, and conceptual categories. A second way in which options are narrowed is through a less self-aware acceptance of limited debate. This process, termed "blowback" or "echo" in security literature, suggests that strategic symbols, initially wielded instrumentally by elites against alternative strategic visions, over time are accepted and internalized by these same elites or successive generations of elites (Jervis 1989: 191-92, Snyder 1991: 41-42). 5 The strategic symbols provide fast and 5

Actually Snyder is inconsistent about the implications of "blowback." On the one hand, he rejects a cognitive-process explanation of imperial overexpansion because he believes that strategic ideologies and myths are not cognitive heuristics deeply ingrained in the decision processes of pro-expansion interests, but rather are self-consciously wielded by these groups at time t to justify choices that serve their interests (Snyder 1991: 10). On the other hand, he

A RETURN TO THEORY

163

easy decision heuristics or cues about the nature of the threat environment and the best ways of dealing with this environment. The net effect is to limit the search for alternative means to a particular end. Sometimes these strategic symbols-as-heuristics interact with heuristics used in "everyday reasoning" to reinforce the narrowing of options. Kanwisher notes, for instance, how a human tendency to overestimate the probability that sequential events are causally related reinforced the myth held by American policy makers that the '"fair of one country to communism will increase the probability of subsequent 'losses' to communism" (Kanwisher 1989: 663-64). Related to this is the use of historical precedent as a shorthand guide to identifying the nature of a strategic problem and, hence, the most appropriate or efficacious means of dealing with it. To the extent that historical analogies govern decision making, argues Vertzberger, "all alternatives not having historical precedent are eliminated without being considered further" (Vertzberger 1986: 229, Vertzberger 1990: 269). Both the identification of and solution to security problems are skewed by selective readings of history, but the bias is not so much self-consciously manipulated to justify decisions as it is the result of generational socialization.6 Thus, according to some scholars, the Munich syndrome and the Yalta axioms came to dominate Cold War interpretations of Soviet intentions and appropriate strategic responses (Yergin 1977, Khong 1992).7 Similarly, Whiting has argued that Chinese policy towards Japan

concedes that over time subsequent generations of decision makers may internalize these justifications, thus transforming them into cognitive heuristics at time t+n (ibid., 2, 10). But why does this cognitive transformation begin at time t and not earlier? Why would elites at time t not have already internalized justifications used at time t-nl Moreover if elites at time t+n internalize these justifications and transform them into decision heuristics, then the argument that overexpansion is the result of the oligarchic pursuit of economic and organizational interests becomes problematic. 6 Moreover, this bias is perpetuated, and the constraints persist, because elites want to appear consistent in strategic choice, in part for domestic political reasons, and in part out of concern for external credibility in the eyes of allies and enemies alike (Anderson 1981: 740, 741). 7 Snyder's critique that historical analogies, like the Munich syndrome, are rationalizations rather than sources of decisions (Snyder 1991: 278-79) does not undermine the argument that analogies can be used to invalidate challenges to decisions, and hence narrow the range of debate. If the analogies are part of the official discourse, then people who reject their premises or offer counteranalogies are less likely to be accepted into the discourse. The Munich case is interesting, too, because it is an analogy that reflects a parabellum view of conflict: conflict with the adversary is zero sum (a "fact" that Chamberlain misperceived), the disposition of the enemy is such that conflict is irresolvable through diplomacy and bargaining (hence the ineffectiveness of Chamberlain's diplomacy), and at a minimum efficacious responses include the development of matching, and better still, overwhelming military capabilities (thus the need to rearm and organize an anti-German alliance). Snyder may well be correct in a narrow sense that the Munich analogy did not cause U.S. decision makers to adoptparabeHum-based grand strategies to deal with the Soviet Union, but it seems plausible that this sort of analogy fell on ground made fertile by the preexistence of parabellum axioms and myths. From this perspective,

164

CHAPTER FIVE

continues to be constrained by the images of Japan's invasion of China in the 1930s (Whiting 1989). A third way in which options are constrained is an unintended result of the deliberate manipulation of strategic symbols by decision elites. According to Kupchan, as these symbols and associated means for dealing with the external environment become rooted in mass attitudes (what he calls strategic culture), efforts by elites to change strategic course become constrained by public opinion and by the requirements of domestic political legitimacy. Thus one reason why elites perpetuate policies of imperial expansion even in the face of rising costs is that "political legitimacy becomes wedded to a specific notion of empire." If for strategic reasons elites try to back away from this notion, they become entrapped by the mass-based strategic culture they helped to create in the first place (Kupchan 1994: 90,487). Elites tend to choose political survival over strategic reality.8 Snyder agrees that rhetorical entrapment can prevent elites from reversing the myths of empire, since this would threaten their political legitimacy at home (Snyder 1991: 41-42). In contrast, the autocommunication literature implies that there are no a priori reasons to expect organizational or group symbols, myths, and symbolic strategies to have any effect on the behavior of the group in its interaction with others. As long as idealized strategies and plans are directed at reinforcing self-perceptions of competence and authority held by decision elites there are no particular reasons why the behavior of the organization cannot be generated by other processes such as inertia, standard operating procedures, or even rational choices made by operational managers (see Broms and Gahmberg 1983: 489). These other processes may or may not narrow options and bias choices, but if they do it is not a result of the idealized level of strategy. Some of the literature on the use of symbols to consolidate in-group identity also implies that strategic discourse can accentuate us-them differences, dehumanize the other and thus lift a priori moral and political boundaries on appropriate ways of dealing with the out-group. The social-psychology literature on cognitive dissonance likewise suggests that far from narrowing the range of strategic choices in an effort to reconcile these with professed preferences, strategic symbols, myths, and rituals may be used alternatively to justify or obscure these differences. In the case of the Seven Military Classics, my analysis suggests that the idealized Confucian-Mencian strategic culture should not impose any a priori limits on strategic choice at the operational parabellum level. Moreover, to the extent that the concept of righteous war functions at both the idealized level then, Munich was not so much a rationalization as a convenient, powerful, and effective heuristic in the strategic decision-making process. 8 The problem here is equating public opinion with strategic culture. How, then, is strategic culture falsifiable?

A R E T U R N TO T H E O R Y

165

(i.e., force is used reluctantly only to restore the moral order, and as an expression of benevolence to oppressed peoples) and at the operational level (the enemy is beyond the moral pale), then the idealized strategic culture lifts any moral or political limits on the application of violence. This, in conjunction with the flexibility axiom (quart biari) found in the operational strategic culture, allows for a wide range of strategic behaviors aimed at increasing the efficacy of the military instrument. If options are narrowed it is due to the underlying parabellum set of strategic preferences that ranks static defensive and accommodationist strategies last. These strategies are not eliminated from the range of options considered, however; they are only discounted as independent routes to security. Under certain circumstances determined by quart bian these options are viewed as viable means for creating advantageous conditions for the eventual application of force. Thus within the parabellum paradigm there are no endogenous restrictions on the range of strategic options; there are only exogenous limitations (such as the balance of capabilities between the two sides) that interact with parabellum strategic preferences to produce behavioral choices. The question then, is, why there is this disjuncture between an idealized strategic culture and the parabellum strategic culture in the Seven Military Classics, and why the former may have little effect on strategic behavior in Ming China. Clearly, some of the theoretical explanations are more useful than others, though convincing tests for any of them are hard to set up. For one thing, the classics are texts, not the strategic plans of organizations or corporations; they are not documents taken from an actual decision-making process. So the disjuncture between the symbolic Confucian-Mencian strategy and the parabellum axioms in the Seven Military Classics may be explained differently than any disjuncture that may exist in the decision processes in the Ming period, for instance. For another, it is hard to gauge the motives and intentions, conscious or otherwise, of the authors of these texts, since authorship is disputed in almost all of the texts, and in some cases that authorship may have been collective, taking place over a number of years. On the other hand, because the authorship may be collective, explanations that account for the symbolic and operational discourses in the texts may be somewhat more generalizable. I do not intend to offer an extended, definitive discussion of the historical, political, military, and personal context of each of the seven texts. This is beyond the scope of this work, and moreover is probably impossible given the ambiguities about dating and authorship of almost all of the texts. What follows is a rather speculative discussion that links these theoretical explanations to a broader discussion of historical context in order to assess more or less plausible reasons for the division. The autocommunication hypothesis is very hard to test in this instance. This notion is usually applied to decision makers who seek to affirm their own competence and sense of purposive control over the decision process. The

166

CHAPTER FIVE

concept is both psychological and social in that this sense of competence has effects on individual decision makers as well as on the group's perceptions of the individual as leader/manager. It is hard to see how it might apply to the author(s) of a text, however. It is possible, of course, that writers of texts on strategy in ancient China used particular languages and symbols to reinforce their own sense of competence, just as the mastery of deterrence theory might create a sense of competence and community identification for a strategic analyst in the United States. But it is awfully hard to show this was the case for obscure writers who lived and died centuries ago. The "official language" hypothesis suggests the Confucian-Mencian discourse would have been used to validate the authority of the strategist. In the Warring States period, for instance, there was a great deal of competition among different strategists and itinerant specialists on statecraft to gain the attention and patronage of the rulers of different states of that period. These strategistsfor-hire had an interest in presenting their ideas as the most efficacious and presumably also hoped to discount the strategic advice of their competitors. Different schools attacked stereotypes of other schools. Legalist strategists, such as the author of the Wei Liao Zi, attacked the mysticism of the yin yang school.9 Confucians charged strategists in the tradition of Sun Zi and Wu Zi with overstressing the allegedly immoral practices of deception, deceit, and guile (Lewis 1990: 129-30). Throughout this period, however, there was no sustained critique of the use of force per se. Even the Confucian-Mencian advocates, as I have noted, clothed their support for war in the language of righteous punishment. But the act of labeling a competing school of strategy with terms that implied incompatibility with this moral order, with the nature of humans, or with the nature of cosmic forces may have been motivated by this competitive atmosphere. It is interesting in this context that the military classics, many of which are thought to have emerged in this period, all contain language that might appeal to some or all of the different schools of thought. Wu Zi incorporated Confucian language, but referred as well to Legalist notions. The Si Ma Fa contains evident references to Confucian code language such as "benevolence as the root" of order, but also advocated stealth, deception, and surprise attacks in warfare. The San Lue drew on Daoist imagery, particularly its stress on softness overcoming hardness, but incorporated Confucian language as well. The WeiLiao Zi drew on Legalist traditions in its discussions of how societies should be organized for war, yet also used Confucian imagery in discussions of how to mobilize and motivate the populace. Perhaps, then, the rhetorical eclecticism of the military texts was a result in part of the need to survive the competitive rhetorical environment of the Warring States period. But it is important to note that much of this rhetoric tapped into 9

The yin yang school stressed divination and proper ritual as the key to winning in battle. See Hsiao 1979: 61-65, Yates 1988: 233-37, and Lewis 1990: 143.

A RETURN TO THEORY

167

a proto-orthodoxy in the language of statecraft and strategy, despite the competition between "one hundred schools." This emergent orthodoxy appropriated the symbolic legacy of the Zhou dynasty, and Confucius was the key figure in this process. Indeed, the Confucians became the strongest claimants to this legacy, but even so it was never challenged in a sustained fashion by any of the other competing schools. The legacy included certain values, institutions, and rhetoric associated with an idealized vision of the early Zhou under the first kings Wen and Wu, and even included certain values associated with even more primeval legendary sage kings such as Yao and Shun. In the realm of strategy and statecraft this legacy included such notions as righteous punishment, internal rectification, the expression of benevolence, and magnanimity by political rulers, the importance of capable and upright officials who administered policies that did not impose oppressive burdens on the people, among other concepts. The rulers of the various states in the Warring States period also tried to lay claim to these legitimating symbols. It is not unreasonable to speculate that strategists drew on this symbolic and rhetorical legacy as well; this was required both for their competition with others and for their own sense of legitimacy.10 Indeed, two of the Seven Military Classics—the San Lue and the Liu Tao—are linked by legend to the first kings of the Zhou dynasty through their putative author, Tai Gong. Tai Gong was allegedly a strategic advisor to both the founder of the Zhou, King Wen, and his son King Wu. The Si Ma Fa draws explicitly on the Zhou legacy in the regulation of interstate relations in its discussion of the Zhou Rites and the Nine Punishments (Liu and Yao 1988: 95). Other texts are also linked to the emergent Confucian proto-orthodoxy through personal ties. Wu Zi, for instance, was the student of Zeng Shen, who in turn was the student of one of Confucius's own students, Zi Xia (Li 1990).n And almost all the texts would in various places preface a statement on statecraft and strategy with reference to ancient precedents established by sage kings. Usually the precedent referred to actions or values that accorded with Confucian-Mencian strategic precepts such as internal rectification as the basis of security or the use of force only for righteous punishment.12 10 Sun Zi is a possible exception to this process. Unlike the probable authors of the other military classics, he was a contemporary of Confucius, and therefore did not draw on the emerging Confucian rhetorical proto-orthodoxy. His is the one text with almost no references to the standard Confucian code words. "I am not arguing that the parabellum paradigm existed in Chinese thought and practice in some form well before being clothed in the myths of the Zhou legacy and the ConfucianMencian paradigm's rhetoric. Indeed, there is evidence that the two emerged roughly in tandem. Lewis has done an excellent job of analyzing the evolution of warfare from highly ritualized individualistic combat in the early Zhou to the larger scale, professional warfare of the Spring and Autumn period and after (Lewis 1990). Contemporary Chinese scholars argue that the concept of quart Man emerged in the seventh century B.C. (Liu and Yao 1988). 12 Not all of these references contradicted the parabellum axioms in the texts. As I have

168

CHAPTER FIVE

The important legitimating role of these Confucian-appropriated symbols persisted even after the Qin dynasty succeeded in unifying China in the third century B.C. Despite the Legalist roots of much of the Qin statecraft, and despite the execution of certain Confucian scholars by Qin Shi Huang Di, recent scholarship suggests that Confucian code language was still crucial for legitimating some statecraft practices (Rand 1977: 201, Bodde 1986: 75-76). After the Qin, Confucianism, in all its variations, deviations, and permutations, emerged more clearly as the rhetorical orthodoxy in statecraft and strategy. Given the Confucian critique of certain strategic practices, of wanton violence, and of reliance on military force as the basis for security, it is highly unlikely that any works on strategy or military affairs could have emerged that did not incorporate certain key Confucian codes and symbols. This is not to say that the authors of the Seven Military Classics deliberately sanitized their language to fit the Confucian-Mencian rhetoric. This is impossible to know. As the literature on symbolic strategy suggests, the process by which operational predispositions are harmonized with the professed political, social, and moral values of the group, organization, or state need not be a wholly conscious one, though it is an imperative one. It is probably true that most societies have a need to make external strategic behavior a seemingly natural or legitimate extension of the state's internal activities. Symbolic strategy seemingly reduces inconsistencies between the civil nature of the state and the martial nature of the state's external behavior.13 Whether societies are traditional empires, authoritarian dictatorships, or liberal democracies there is an interesting tension between military activity and the goal of constructing an image of an ordered, legitimate society. On the one hand, military institutions and activities often seem not to fit very comfortably with self-professed values of the state. On the other, a coercive apparatus directed outwardly can reinforce the supposed contrast between external disorder and internal order. This can reinforce an in-group sense of siege, thus both justifying the internal order and external behavior directed at the out-group that threatens this internal order. Symbolic strategy provides a way of integrating the military behavior into the legitimate activities of the state as best as may be possible given its uniquely violent and destructive nature. In the Confucian-Mencian paradigm martiality was generally viewed as a argued in chapter 3, in some instances in the Seven Military Classics a reference to the nonviolent resolution of conflict in Yao's or Shun's time implied the historical irrelevance of these ancient sages to the problems of strategy. On the role of sage-king myths in the emergence of a "correct" view of violence in the late Zhou period, see ibid., 167-74. 13 As postmodern and neoliberal institutionalist critics of realist theory argue, there is much less legitimate order in domestic societies (a great deal of which is coerced) and much more noncoercive, self-interested order and institutionalization in international societies than standard realist theory would have us believe (Keohane and Nye 1977, Klein 1988: 134, Dalby 1992: 101).

A R E T U R N TO T H E O R Y

169

separate realm of state activity, governed by its own rules and norms of behavior (Rand 1977: 34-41, Lewis 1990: 126-27). Reconciling these with the very different rules and norms of internal civil order required symbolically downplaying martiality and justifying it only in terms of these civil norms. Even the most benign code language in the Confucian-Mencian paradigm could indirectly justify the application of violence against out-groups. The notions that those "all under heaven are brothers," or "those within the four seas are of one clan" would not seem at first to reinforce a sense of in-group uniqueness. Yet these notions were often tied to a corollary, found in the Wen Dui for instance, that barbarians may be cut from the same cloth as the Han, but through differences in geography, economic activity, socialization, and such developed in a culturally inferior direction. Their claim to have come from the same primordial clan only meant that they could be reeducated or sinified. While this conclusion could justify an accommodationist strategy of enculturation, it has the secondary effect of accentuating differences between the in-group and outgroup and of reinforcing the in-group's claim to unique superiority. It is not a huge leap to the conclusion that should the out-group resist enculturation, or threaten to "barbarianize" the in-group (the Han), coercion becomes justifiable. There is, finally, the cognitive-dissonance explanation for the disjuncture between an idealized strategic culture and the operational parabellum axioms. In the Chinese case Lucian Pye comes close to making the argument that the Confucian-Mencian paradigm bridged the vast gulf between the ideal and the real by clothing the latter in a discourse infused with images of the former. For the Chinese a basic function of ideologies has been to make it possible for them to avoid a direct confrontation with the realities of their politics. Political reality in any society must reflect disorder, potential or actual conflict and the expression of open or latent aggression. These are anxiety-provoking subjects for the Chinese. Thus the tradition developed that the elaboration of politics could be respectable only if it was enveloped in moral discourse, and any explicit discussion of the realities of politics without the wrappings of ideology was vulgar. Furthermore, the extent to which the Chinese have had to rely upon ideology has been indicative of the degree to which they have tended to repress conflict; and more importantly they have related the need for authority to the need to control all manifestations of human aggression. From this exaggerated need for the comforts of ideology we can see a hint of the concern about aggression. . . . We argue that the Confucian tradition, both structurally and ideologically, created forms of authority that gained strength by denying the legitimacy of sentiments of aggression. (Pye 1968:31-35)

In a more recent work Pye argues that Chinese pragmatism can explain the absolute flexibility in Chinese political-strategic decision making, but that this pragmatism, for legitimation reasons, has to be wrapped in the mantle of ethical and moral superiority (Pye 1988: 103).

1 70

CHAPTER FIVE

Pye's arguments are unfortunately cast in Sinocentric terms. The process of managing dissonance between idealized behavior and operational decision rules is certainly not unique to the Chinese, and it is not clear on what basis the need for the comforts of ideology are exaggerated by the Chinese. The management of dissonance through ideology inheres in any process by which alleged superior values held by an in-group conflict with the types of strategies required to deal with a threatening out-group. The use of Confucian precepts to mask the advocacy of the extermination of the adversary is no different functionally from the use of democracy and the rule of law to advocate the overthrow of antiAmerican governments, or the invocation of fraternal socialist community values to invade an errant Soviet client state. But Pye's general point may well apply in the case of the Seven Military Classics and other strategic texts. Again, however, any conclusions about the management of dissonance cannot be tested rigorously on the authors of these texts. We have, then, three broad explanations for the disjuncture between an idealized strategic culture andparabellum axioms in the Seven Military Classics. None of these explanations can be confirmed or disconfirmed with much certainty, given the paucity of the kinds of information needed, but the latter two seem more plausible within the limitations of the evidence.The dissonance explanation, for instance, might provide some footprint that could indicate its presence. Specifically, one might look for a large and growing disjuncture between the status of the idealized strategic discourse in the policy process on the one hand and actual behavior on the other. That is, as actions diverge from the standard of the idealized strategic culture, the group is increasingly compelled to frame this behavior in the discourse of the ideal. So ever shriller affirmations of the symbolic discourse would confirm the dissonance explanation. Unfortunately this would also be evidence for the in-group identity explanation. I have also suggested that whatever the explanation, the symbolic strategic culture ought not to have much effect on behavior, because it is essentially disconnected from operational advice. Moreover, there are plausible theoretical grounds to expect little impact on narrowing the range of options available to decision makers. Indeed, all three explanations might lead to the same conclusion about the noneffects of the symbolic discourse.

SOME HYPOTHESES ABOUT M I N G STRATEGIC D E C I S I O N - M A K I N G

We come finally to the problem of linking the findings about Chinese strategic culture to grand strategic behavior. Put another way, how much of the variance in state behavior is accounted for by strategic culture? Or can we safely relegate historically and culturally rooted images, axioms, and preferences to a residual category as the dominant structural-realist paradigm does? It makes

A RETURN TO THEORY

171

little sense to leave any discussion of strategic culture at the textual or ideational level without asking how this level impacts on the strategic preferences and behavioral choices of decision makers. Otherwise the study of strategic culture ends up treating only half of the "attitude-behavior" problem. Ultimately the dependent variable in the social sciences in general, and international relations theory specifically, is behavior. There is no sense, then, in analyzing images, perceptions, worldviews, doctrines, norms, and other ideational variables unless this is part of a broader research program that links these to the behavior of individuals, groups, organizations, states, and systems. This does not mean we should expect to find a clear attitude-behavior relationship. As a number of social scientists have warned over the years, the relationship is notoriously hard to nail down empirically in a convincing manner. Attitudes may obscure or clarify behavior or remain only tenuously related, but none of these would be a trivial finding. Negative findings are just as important for cumulative knowledge as the confirmation of hypotheses, since they can rule out theoretical dead ends. What hypotheses about the grand strategic choices of the Ming emerge from the content of Chinese strategic culture? The answer is complicated by a critical question: namely, whether one should assume there are any interactive effects between the idealized or symbolic strategic culture and the parabellum set of preferences. But as I have argued, there is little indication that the symbolic strategy seeps into decision axioms in the deep structure of most of the Seven Military Classics. Thus, if these texts do embody a decision structure that informs behavioral choice then there is no particular reason to assume grand strategic choice in the Ming will be constrained by the symbolic level of strategy, either through the effects of blowback, as Snyder suggests, or through the constraints of strategic culture-influenced public opinion, as Kupchan argues (or literati opinion in the Chinese case).14 Moreover, to assume from the start that symbolic strategy has a delimiting effect on parabellurn-dGrivQd decisions pushes any hypotheses into the minefield of counterfactuals. How would one gauge the degree to which symbolic strategy dampened levels of coerciveness in strategic choice, since one could not know with much certainty how much more coercive strategic choices might have been had there been no interaction? It seems best to keep the hypothesis testing as simple as possible at the moment, and to alter any assumed relationship between the two levels of strategic culture on the basis of the results of the empirical tests. Given this assumption, one should expect to see the persistence of a Confucian-Mencian strategic discourse during the Ming period regardless of the grand strategic choices made by the dynasty. Since the official language of strategy was established during the Warring States period and did not change appreciably 14 This does not mean that decision makers are insensitive to challenges to their decisions framed in terms of symbolic strategic precepts. There is a difference between decisions taken because of symbolic challenges and decisions justified in the face of symbolic challenges.

1 72

CHAPTER FIVE

over time we should expect that strategic texts, debates, and decisions in the Ming exhibited a similar symbolic discourse.15 Strategic decisions that leaned towards the accommodationist end of the spectrum could be framed in terms of expressions of benevolence and magnanimity. When strategic choices leaned towards the offensive or expansionist end, these could be justified by appeals to the notion of righteous war and expressions of concern for the peoples living under the unrighteous yoke of certain barbarian rulers. Or these choices could simply be named expressions of the emperor's benevolent awesomeness. The point is that one should expect to see different grand strategic choices in different periods framed at the symbolic level by the same ConfucianMencian discourse. In other words, symbolic strategic culture should have no effect on strategic choice. At the operational level, however, if Ming decision makers were influenced by the parabellum strategic culture, we should expect to find a general preference for offensive strategies over static defensive and accommodationist choices. Choices at this operational level should be tempered by quan bian. If decision makers follow this axiom they will qualify any preference for offensive uses of force with a calculation of its expected efficacy under different strategic conditions. This calculation should take into account the relative balance of military capabilities (including internal economic resources), morale, will, and the advantages and disadvantages of terrain. One should expect to find, then, a direct association between changes in levels of coerciveness or offensiveness and changes in the capacity to implement such strategic policies. Accommodationist and static defensive policies should be temporary expedients associated with a low or declining capacity to implement offensive strategies and pursued only because the regime has no viable offensive option. As the capacity to act offensively against the threat drops one should expect to see a preference first for static defense and then accommodation. A critical question is whether we should expect this changing capacity to act offensively to be objective or subjective. If there is a disjuncture between objective and subjective assessments of relative capacity to act offensively, the indicators and evidence used to assess the accuracy of predictions under one situation may be useless for assessing the other. To be more concrete, objective indicators (i.e., changing balance of military capabilities, changing capacity of the state to mobilize economic resources, levels of internal cohesion, etc.) may be inaccurate representations of how Ming decision makers assessed their relative capacity to use offensive violence against the Mongols. If this 15 The existence, collation, and institutionalization of the texts that comprise the Seven Military Classics is evidence that the official language of strategy did not change up to the Ming period. It is clear from many of the other key works on strategy from the Song and Ming periods that outside of the seven classics there was no appreciable change in the symbolic discourse in strategic affairs. There was no open rejection of or attack on the Confucian-Mencian orthodoxy.

A R E T U R N TO T H E O R Y

173

were the case then any longitudinal analysis of objective changes in strategic capacity would yield inaccurate predictions about changing levels of coercion. On the other hand, there is no glaring reason to believe there should be a huge discrepancy between objective indicators and subjective assessments of the empire's strategic capacity. If the two are generally in sync, then the question is amenable to statistical analysis, otherwise a more detailed focus on the subjective assessments of strategic opportunities might be more appropriate. My own preference is to do both, since this will help determine the congruence between objective and subjective indicators of the capacity to act offensively against the Mongols. Moreover, as I argued in chapter 2, a richer focus on decision makers' perceptions and assumptions is necessary to establish any connection between the decision axioms in the parabellum strategic culture, the decision axioms of key decision makers, and their strategic choices. A broader longitudinal approach would simply assume a relationship between parabellum axioms in Chinese strategic culture on the one hand and strategic behavior on the other. A longitudinal analysis, however, may catch a general (i.e., probabilistic) relationship between changing strategic capacity and offensive behavior that a handful of case studies may miss. Indeed, as the literature on the attitude-behavior problem warns, general attitudes are poor predictors of specific cases (Eiser and van der Pligt 1988: 26, Schuman and Johnson 1976). A longitudinal analysis might catch general trends in behavior not caught in specific studies of decision makers. To summarize, then, given the dichotomized nature of Chinese strategic culture there are two basic hypotheses about the relationship between strategic culture and strategic choice in the Ming period. The first hypothesis is that we should not expect to find any relationship between the idealized grand strategicpreference ranking and Ming strategic choice. Instead we should see the persistence of an idealized or symbolic discourse in the decision-making process regardless of the strategic behavior of the Ming. The second hypothesis is that given the parabellum axioms in the operational strategic culture, we should expect to find a positive relationship between changes in the relative capacity of the Ming to act offensively against the Mongols and offensive strategic choices. We should not expect to see the adoption of accommodationist or static defense strategies when the regime has an increasing relative capacity to act offensively, as the initial strategic-culture model outlined in chapter 2 would predict. It should be immediately apparent that this latter prediction is not much different from the prediction one might derive from a simple realpolitikdynastic cycle model of choice. But one of the reasons the concept of strategic culture is worth a closer look is that predictions from a strategic-culture model of choice may well be different from those of variants of structural-realist models though the predictions need not be that different either. It seems obvious that the parabellum strategic culture, coupled with the concept of quart bian—

1 74

CHAPTER FIVE

a concept that is virtually an explicit reminder to decision makers to calculate the expected utility of different strategies—makes predictions similar to those of a rough structural-realpolitik model. If a state has the capacity to use force offensively against threats, it will do so; if it does not have this capacity, it will not. This capacity is a function of the relative balance of capabilities (i.e., of the structure of the relationship with the adversary). A structural model assumes that decision makers will act as if they have the ability to estimate the expected utility of various alternatives. But there is a great deal of literature in the areas of bureaucratic politics, bounded rationality, anthropology, and social psychology critiquing this assumption. To the extent that parabellum axioms plus quart Man show up in the decision process, then strategic culture appears to be filling in structural-realpolitik theory. If one could show that structural-realpolitik predictions do not work well unless one finds the presence of a. parabellum strategic culture in a particular society, this might have two very different implications for the soundness of the realist theoretical edifice. On the one hand it would seem to strengthen the argument made by realists and neorealists that their explanations and predictions about state behavior are not culturally bounded or Eurocentric. On the other hand, a finding that realpolitik behavior is rooted in a culturally learned view of the world might also leave open the theoretical possibility that societies could exist that did not exhibit a parabellum strategic culture and thus would not act in a realpolitik manner, and that particular societies that exhibited a parabellum strategic culture at time t are not inexorably bound to exhibit one at time t+1. Put another way, if & parabellum calculus is rooted in strategic culture and not structure, if it is learned and transmitted, then this suggests that a parabellum strategic culture is also historically and culturally bounded. It is not necessarily a valid model of strategic culture for all societies at all times, whereas structural-realist theory suggests that precisely because if is structural, a structural-realpolitik calculus is not historically or culturally bounded. It applies to street gangs as much as it applies to the Warring States in China or the city-states of ancient Greece, anywhere states operate under conditions of anarchy. Since it is therefore theoretically possible that the operational strategic culture varies across societies, it is therefore still of great theoretical interest to determine whether it is necessary to appeal to strategic culture to help explain strategic choice. Obviously this determination would be much easier if the operational strategic culture yielded predictions that were different from a structural-realpolitik model of choice. Unfortunately, this is not the case for Ming China.

Chapter Six THE PARABELLUM PARADIGM AND THE MING SECURITY PROBLEMATIQUE

A RIGOROUS APPROACH to analyzing the relationship between strategic / \ culture and choice involves at least three steps. The first is to test for JL \*the presence of and congruence between grand strategic-preference rankings in strategic cultural objects of analysis (in this case the Seven Military Classics). There is a nontrivial degree of congruence in preference rankings across the texts, but these strategic preferences generally reflect & parabellum view of the nature of conflict and the efficacy of violence for achieving state security. This ranking is tempered, however, by a notion of absolute flexibility— or quan bian—that implies that any grand strategic choices based on this preference ranking will be positively affected by assessments of the likelihood of achieving desirable ends from any particular choice. In other words, ceteris paribus, decision makers ought to prefer the offensive use of force in order to defeat the adversary and achieve state security. But if they determine that this use of force faces temporarily unfavorable strategic conditions (say, a balance of capabilities that reduces the likelihood of success), then other choices such as defense or accommodation will be made in anticipation of an eventual favorable change in the balance of capabilities. The next step in this research process is to test for the presence and status of \hs parabellum strategic culture in the Ming period. Chapter 7 will test for the effects of this strategic culture on Ming strategy towards the Mongols. Put another way, this chapter focuses first on whether or not the parabellum paradigm anchors the strategic thinking of key decision makers in the Ming period. The next chapter examines the relationship between these attitudes and Ming security policy towards the Mongol threat along the northern border. I hope to offer somewhat more concrete conclusions about the role of the symbolic Confucian-Mencian strategic culture in Ming policy towards the Mongols. The evidence indicates that, for the most part, Ming decision makers accepted the key elements ofthe parabellum central paradigm, modified by the concept of absolute flexibility. Thus, while there were debates about which strategies were most likely to solve the security problem, the disagreements revolved around differing assessments of the balance of capabilities between the Ming and Mongols. In other words, as ^parabellum strategic culture would suggest, a great many decision makers preferred to use force, either offensively or in the context of active defense, rather than static defense or accommodation.

1 76

C H A P T E R SIX

Arguments in favor of static defense (e.g., wall building) or accommodation (e.g., opening tribute relations or setting up trading fairs with the Mongols) were often made on the grounds that the Ming were militarily and strategically incapable of doing anything else. By implication, offensive strategies, while in principle the preferred means of dealing decisively with the Mongols, would likely not work under the circumstances. The arguments in favor of nonoffensive strategies were not based on a priori Confucian-Mencian strategic cultural preference rankings or some moral-political aversion to offensive uses of violence, but on contingent strategic arguments. Some simple aggregate data analysis also tends to back this up. Increases in offensively oriented activity tended to be associated with periods when the Mongols were less threatening or Ming capacity to direct its resources to the Mongol threat improved. Moreover, there is evidence to suggest a secular shift from more offensively oriented behavior to more defensive and even accommodationist behavior as the Ming capacity to mobilize resources against the Mongols declined.

BlNGSHUm

THE MlNG DYNASTY

There is one issue to be dealt with before turning to an analysis of Ming security policy, and that is the influence of the Seven Military Classics during the Ming period. Any effort to test for the effects of strategic culture must also address the transmission of this strategic culture across time and across different strategic contexts; the objects of analysis that embody this strategic culture must play a role in the socialization of strategic elites in the strategic culture. This step determines whether the basic assumptions of the primordial strategic culture are accepted by decision makers in later periods or whether these assumptions are reinterpreted. The Seven Military Classics were relatively widely read in the Ming period among scholar officials, military leaders, and emperors. The impact of the Seven Military Classics on strategic thinking can be seen in the strategic texts produced during that time. I have not subjected the Ming texts to the same rigorous content analysis I used on the original classics, in part because despite the proliferation oibingshu (texts on strategy and statecraft), individual Ming texts were not as widely read, nor were they required study in the military education system. For the most part, as well, they were derivative from the Seven Military Classics. I simply wish to indicate that the strategic thought embodied by the Ming texts openly drew on the seven classic texts. Because of the vast number of bingshu titles produced during the Ming period, I will only touch on the content of three of these titles: the Zhen Ji (PI&E), the Cao Lu Jing Lue (#^SI1K§), and the Ton Bi Fu Tan (SISjff §£)• Contemporary scholarship identifies these as the key Ming texts on strategy and warfare (or

THE MING SECURITY PROBLEMATIQUE

177

the art of war, bingfa J^Sc) and most similar in scope to *the Seven Military Classics (Xu 1990).l

Zhen Ji This text appeared sometime in the Jia Qing-Wan Li reign periods (mid to late sixteenth century). It was written by He Liangchen (fof^ |§), a scholar-official with military experience along the northern border (see ibid., 165). Like the seven classics, this text is a wide-ranging discussion of strategy, operations, training, maneuver, and technology. The text frequently cites or paraphrases the earlier classics, including Sun Zi, Wu Zi, and the Wei Liao Zi. There is little difference between the strategic axioms that the Zhen Ji explicates and the parabellum axioms found in the Seven Military Classics. The Zhen Ji does not exhibit any Confucian-Mencian variable-sum view of the adversary, nor any obvious preference for the nonviolent or minimally violent defensive resolution of conflict. Rather, in one list of five internal and external dangers or threats to the state the text notes that the fourth comes from the "four barbarians." The source of this threat, argues the author, is inadequate attention to military affairs, clearly implying that the correct application of violence is the primary means for dealing with this external threat. "The [threat] from the four barbarians arises from offense and defense not being clearly understood, and from the means of controlling [the enemy] not being systematized" (He 1984: 84). Elsewhere the author discounts the role of nonviolent or minimally violent strategem in security affairs in his comparison of three different approaches to the use of the military instrument. One approach, he writes, stresses martiality and the offensive application of force. Another emphasizes heaviness, solidity, and steadfastness, with implications of a defensive preference. The third stresses ingenious strategem and deception. Of the three, the latter is considered the least useful, the most divorced from the pragmatic experience of warfare. The implication is that in the martial realm the argument that one can rely solely on strategem is merely "talking about military affairs" ( 1 ^ ^ ) (ibid., 187).2 The application of violence, moreover, should be preemptive, including well within enemy territory. Indeed, the author cites favorably a discussion in the !

This list excludes works like the Ji Xiao Xin Shu ()£$&$?#) and the Lian Bing Shi Ji ($fc ft If &B) by the famed general of the latter half of the sixteenth century, Qi Jiguang. His bingshu, however, are for the most part primers on training and maneuver, not texts on higher levels of strategy. 2 He Zhongsheng ({Hlf1!1^*), the author of another Ming bingshu, written in 1606, concurs that the notion of not fighting and defeating the enemy is an unrealistic one. "As for the way of using the military instrument, it is difficult to achieve decisive victory without warring" (He 1606: 9.2a).

1 78

C H A P T E R SIX

Wei Liao Zi about a preemptive seizure of enemy territory (ibid., 181-82). Elsewhere the text repeats the metaphors found in the Seven Military Classics to describe the speed, direction, and awesome power of the efficacious use of the military instrument. The effective use of force is likened to wind and rain, a diving bird of prey, thunder and lightning, rolling stones, and water bursting forth to extinguish a flame (ibid., 71, 83-84, 178-80). Like some of the military classics, however, the Zhen Ji does inject an element of restraint in the application of violence—namely, caution in offensive operations and in the pursuit of a retreating enemy. Again, however, this caution is not based on any particular a priori moral restraint, but on the grounds that reckless offense or pursuit exposes one's own forces to traps, ambushes, and logistic foul-ups (ibid., 52). As with the Seven Military Classics, the prominent role of violence is qualified by the guiding principle of absolute flexibility. The Zhen Ji devotes a whole chapter to the concept of responding to changing circumstances (yin shi @ §§). The author neatly summarizes its application to military affairs: "Thus one who excels at the use of the military instrument necessarily responds to the adversary and takes advantage of change" (ibid., 198-99). Concretely, this involves striking at the enemy's weak points, waiting for the enemy's capabilities to decline, and attacking and grabbing what is undefended. Cao Lu Jing Lue The author of this text is unknown, though scholars believe it appeared sometime in the early Wan Li period (late sixteenth century) (Anonymous 1986, Xu 1990: 159-60). Like the Zhen Ji, this text liberally cites the Seven Military Classics. Indeed, a great number of the chapters in the Cao Lu Jing Lue begin by citing one of the classics. This text, more so than the Zhen Ji, exhibits some of the language of the Confucian-Mencian strategic discourse. It stresses, for instance, internal rectification as an element in the successful use of military capabilities. The text cites Wu Zi's dictum, "If within the state there is no harmony, then one should not send forth the army. If there is no unity within the army, then one should not deploy it in formation. If there is no unity in formation, then one should not advance into battle. If not united in battle, then [the army] cannot win decisively" (Anonymous 1986: 86). The text does not argue, however, that internal rectification will lead to the enemy's voluntary submission in order to partake of the benefits of Chinese virtue. The closest it comes to making this argument is in a discussion of how to handle the Qiang (7^) barbarians in northwestern Sichuan and southern Gansu. Here the advice is to use a two-track policy of offering them posts, rewards, and material goods, and improving defensive military capabilities, while at the same time eschewing punitive campaigns so as to avoid driving the Qiang into an alliance with other tribes in the area (ibid., 304). A chapter

THE MING SECURITY PROBLEMATIQUE

1 79

on dealing with the Man (fi) barbarians in the south offers similar recommendations (ibid., 299). Elsewhere the support for internal rectification comes from the imperative to improve the deterrent capabilities of Ming border defenses. In a clear reference to the poor organizational and logistics systems supporting Ming forces along the borders, the text notes "If the people are impoverished and warriors are weak and famished, and if in government and private affairs all face dire straits, then the enemy takes advantage from the outside, and there are upheavals on the inside. When things become like this, there has never been a state which has been peaceful and secure" (ibid., 71). In general, the text argues that force is necessary to deal with external threats. Indeed, the author begins with a statement, which recalls the first line in Sun Zi's Art of War, to the effect that nothing historically has been more important in the decline of states than the neglect of military preparations. He then cites Confucius's comment that not teaching the people how to war was to abandon or neglect them (ibid., 1). Another chapter addresses the efficacy of force in dealing with the external threat posed by the Rong (j£) and Di ($c) barbarians (northern, including Mongolian and other nomadic peoples). Against these peoples the most appropriate approach is to exterminate the threat. Far from adopting a Confucian-Mencian view of non-Chinese and Han as being from one, albeit unequal, family, the author describes the Rong and the Di in rather stark zero-sum terms: "The Rong and Di are ravenous like wolves; they cannot but be killed" (ibid., 167). In practice, however, given China's weaknesses in offensive capabilities (namely, a lack of cavalry and an abundance of logistical problems), the best Chinese forces could do was to adopt an active defense policy along the border. This would combine static defensive measures like wall building, the establishment of a better warning system, the accumulation of rations and supplies, and improvements in military training and discipline with more aggressive, though limited, use of mobile, rapid response troops (qi bing iif J=£) to harass, ambush, and wear down attacking enemy forces (ibid., 126, 295, 297). However, this defensive strategic preference was, in essence, a contingent, capabilities-based argument; according to the author, the Ming were not technologically capable of meeting the northern barbarians in mobile offensive warfare in the steppe lands, and an offensive strategy of punitive campaigning made little sense under those conditions. This was no flat rejection of the offensive use of force. The author believed that were the Ming militarily capable of taking the offensive, this would be preferred to a more-or-less passive defensive posture.3 Elsewhere the text argues that in principle the use of force 3 As the author remarks, "When my side is superior and the enemy inferior, then it is suitable to fight open/field warfare. When my side is inferior and the enemy is superior then to employ open/field warfare is dangerous" (Anonymous 1986: 254-55). He also critiques a purely static

1 80

CHAPTER SIX

does not exclude offensive incursions into enemy territory, and it discusses a number of scenarios involving operations deep in enemy territory. These operations are strategically legitimate if one has adequate supplies, does not fight protractedly, avoids enemy ambushes, maintains strong defenses, and attacks an exposed, strategically valuable weak point. This offensive predilection is reinforced by the argument that force is most effective when applied preemptively. Borrowing from imagery in the Seven Military Classics, the author likens the use of force to a bird of prey striking its victim, or lightning striking before one has time to cover one's eyes. At one point he outlines twenty scenarios in which force should be applied preemptively. The import of the argument is that it is best to attack the enemy before its defensive preparations are complete, before the morale of enemy soldiers firms up, before it has completed its deployments, and before it has acquired advantageous terrain. Not to do so is to lose "circumstantial opportunity" (ibid., 233, 284, 286). As for the role of nonviolent political strategem and deception, here too the author does not stray from the causal arguments in most of the Seven Military Classics. Clearly, deception and cunning play important roles in the security realm (there is a chapter devoted to the topic). "Who says one can neglect deception and trickery? If one speaks only of benevolence and righteousness in warfare, this is to follow in the footsteps of Song Rang Cheng An. How can one thus avoid being defeated?" (ibid., 77).4 However, political and military strategem is linked only indirectly to the defeat or submission of the adversary. Rather, it is used to weaken, confuse, and otherwise control the enemy. Once this has been accomplished then violence should be applied, and the enemy attacked at its most vulnerable point (ibid., 77, 225, 227, 242). Much like the arguments made in the San hue and the Si Ma Fa, the author accepts that once the enemy has been defeated, the righteous side has the right to reorganize the internal affairs of the defeated state (ibid., 123, 272). The Cao Lu JingLue does not advocate unrestrained offensive or preemptive uses of forces. Indeed, like other works on strategy, this notion of quan bian or absolute strategic flexibility is central to the text. In a paraphrase of Sun Zi, the author notes, "As for military affairs, it is [a case of] action when presented defensive strategy like wall building for wasting funds (ibid., 55). Towards coastal pirates he recommends an offensive posture: "Defending against their landing [on shore] is not as good as defending against them at sea. Defending against them in coastal waters is not as good as heading out to sea and defending against them outside of coastal waters" (ibid., 302). 4 Song Rang Cheng An (Song Rang Gong) was a ruler of the state of Song during the Spring and Autumn period. He is known for continuing to hold to the principles of ritualistic warfare of the early Zhou dynasty even as they were being abandoned. He suffered a defeat when he refused to attack the army of the state of Chu when it was most vulnerable as it crossed a river. To have done so, he argued, would have been to offend the principles of chivalry.

THE MING SECURITY PROBLEMATIQUE

18 1

with opportunity. Attack where the enemy is unprepared, and appear where it does not anticipate" (ibid., 84). One must be able to respond to changing circumstances and exploit opportunity. "One who excels in military affairs follows [changes in] the circumstances [or relative circumstantial power of the two sides] and responds to them. Thus our side has no fixed position" (ibid., 182).5 Indeed, one must be capable of directing this change by manipulating the enemy through political and military moves so that the enemy's intentions are exposed. This may require waiting or operating on the defensive for a time. But once the opportunity to decisively attack enemy weaknesses appears, it must be exploited quickly (ibid., 53, 156, 231). Given this principle of absolute flexibility, restraints on force become strategic or logistical, not moral or political. Offensive operations within enemy territory, for instance, are constrained by the need to protect food supplies and supply routes from enemy counterattack (ibid., 73). The author also cites the Si Ma Fa's axiom against recklessly pursuing a retreating enemy. Again, however, this restraint derives from strategic concerns about falling into enemy traps and ambushes. If one can ascertain that the enemy has truly been routed and has no hope of being rescued by outside forces, then it should be pursued quickly and defeated decisively (ibid., 238). The text also adopts the restrictions concerning treatment of the populace in the enemy state found in the Si Ma Fa and the Wei Liao Zi, for instance. Whenever invading an enemy state, one must take care of the old and young, avoid destroying houses and temples, avoid plundering the people's wealth, and use polite words to enjoin the population to submit. This approach allows the victor to display its great virtue, dissipates the enemy's will to fight, and thus hastens the total capitulation of the adversary. This use of benevolence and righteousness as a political tool in a broader offensive policy is a mechanism by which a "guest" (i.e., invading) army can be turned into a "host" in enemy territory (ibid., 167, 173, 238).

Tou Bi Fu Tan This text, which appeared in 1604, was, according to some scholars, probably written by He Shoufa ({RjTffife)who lived in the latter half of the sixteenth century (Cui 1989: 1, Xu 1990: 169). He also annotated a version of the Seven Military Classics (Liu 1990d: 147), and his own book was obviously modeled on Sun Zi's Art of War. In the first chapter He, like Sun Zi, appears to argue that one can defeat the enemy without resort to actual violence. If one can respond successfully to and control changes in enemy behavior, presumably through deception and strategem, then one can "capture enemy cities without 5 "Position" here is a translation of ju (j|j). It implies positional status in a game of conflict, like wei qi or chess.

1 82

C H A P T E R SIX

attacking them, and command the enemy's capitulation without surrounding him." The use of violence is characterized as a last resort, to be taken in "unavoidable circumstances" (Cui 1989: 18). Even so, it becomes clear as the text goes on that this minimalist vision is qualified and weakened by other strategic axioms. In later chapters, for instance, He appears to view the cause-effect chain from nonviolent strategem to defeat of the enemy in much the same way as most of the Seven Military Classics—namely, as an indirect path that passes through the application of pure violence. A discussion of "attacking the enemy's mind" (gong qi xin J& ;Mw[>) comes shortly after a passage where it is clear that this is a prelude to a more effective violent attack on an enemy city (ibid., 78-80). The objective of strategem is, in general, to create opportunities for attack (ibid., 36). This, then, relates to the concept of absolute flexibility, or responding to changing circumstances. The notion allows for the application of violence, offensively if conditions allow, to defeat an adversary. In one passage He discusses how to transform oneself from vulnerable guest army into a more stolid, invulnerable, and controlling host army while remaining deep in enemy territory (ibid., 37). The contingent nature of offensive and defensive strategies is never clearer than in a chapter specifically on weighing the relative advantages of attack and defense. Not surprisingly, given the content of quan Man, the author concludes that both approaches have strengths and weaknesses. Defensive strategies allow one to rest, while offensive ones tend to exhaust the army's strength. On the other hand, offense allows for the concentration of physical power, while defense tends to force its dispersal. The opportune moment for both strategies is not fixed; when to use one or the other depends entirely on changing circumstances (ibid., 77). While this argument translates into a cautious approach to the use of force, it does not limit a state to a narrow range of strategic choices. However, given the importance of military forces qua military forces in the defeat of the enemy and the security of the state (see ibid., 17, 149), and given the absence of any Confucian-Mencian language about the extension of benevolence or the enculturation and voluntary submission of the adversary as a basis of security, it is probably safe to say that this text considers active defense strategies as more efficacious than static defense or accommodation.6 The Ming bingshu themselves deserve a more thorough examination. A sample of three does not do justice to the richness of texts published in this 6 Like many other bingshu, the Tou Bi Fu Tan sees benevolence and righteousness as instruments in the application of military force, not as separate routes to security. By declaring one's own forces righteous and the enemy's unrighteous, one can motivate one's own forces, prevent them from defecting to or sympathizing with the enemy, and place the enemy beyond the moral pale, as is required to justify the application of violence (Cui 1989: 19, 54). The concept of absolute flexibility makes quick work of any Confucian-Mencian notion of good government as the basis of security.

THE MING SECURITY PROBLEMATIQUE

183

period. My purpose, however, is only to show the continuing direct influence of the Seven Military Classics on the production of strategic thought during the Ming dynasty. The methodological objective has been to determine the presence of decision axioms, embodied in strategic-cultural objects of analysis, in the historical period on which the empirical analysis of strategic culture focuses, and to determine whether there have been any changes in interpretation during this transmission. This is a lead into a discussion of the influence of the parabellum paradigm on Ming strategy towards the Mongols. Ming bingshu embody many of the characteristics of the parabellum paradigm found in the Seven Military Classics: a highly conflictual, even zero-sum conceptualization of the relationship with the adversary; a recognition of the efficacy of violence, including offensive or preemptive violence, in the process of defeating the enemy and achieving state security; and, within the context of legitimate violence, a notion of absolute flexibility that, while introducing an element of caution into strategic decision-making, lifts restrictions on acceptable means for dealing with an adversary. There is very little of the ConfucianMencian discourse in these texts, certainly less than one finds in the Si Ma Fa or San hue. Indeed the Ming texts view this discourse largely as a political tool for defining the nature of conflict with the adversary and fostering morale within the populace and the army.

MING SECURITY PROBLEMS ALONG THE NORTHERN BORDER

Before finally turning to an analysis of Ming decision makers' strategic preferences, it would probably be helpful at this point to provide some historical context. Obviously space does not permit a complete recounting of MingMongol relations. There are already a number of good secondary sources on the question (see Lai and Li J954, Serruys 1967,1987, Farmer 1976, Pokotilov 1976, Wu 1978a and b, 1981, Chen 1984, History of Chinese Wars Editorial Committee 1986, Wang 1985, Twitchett and Loewe 1986, Zhang 1986, Waldron 1990, Yang 1990). I will only highlight some basic events and themes in Ming security policy towards the Mongols. For most of the dynastic period the Mongols constituted the primary security problem faced by the Ming rulers and decision makers, outweighing both the uprisings and rebellions of non-Han groups along China's southern and southwestern borders and the violence from Chinese and Japanese pirates operating along the eastern and southeastern coast. Except for the last few decades of the dynasty, when conflict with the Manchus in the northeast took precedence over all other external-security issues, the conflict with the Mongols preoccupied Ming rulers (Lai and Li 1954: 12, Yang 1990: 289).7 This is borne out in 7

For representative examples of this basic threat assessment see the memorial from Liu Qiu

1 84

C H A P T E R SIX

some basic statistics. The Ming were involved in 308 external wars throughout the dynastic period (1.12 wars per year on average). Of these, 192 or about 62% of all external wars were with the Mongols (0.69 wars per year on average). The Ming fought wars with the Mongols in about half of the years in the dynastic period (n=152 or 55%).8 Not surprisingly then, most of Ming military and financial resources were devoted to the military security of the border with various Mongol tribes, primarily the Oirat, the Da Dan, and the Uriyanghkha, that lived outside the borders running from Gansu in the west to Shanhaiguan in the east. During the first two major regencies—that of the Hong Wu (1368-1399) and Yong Le (1403-1423) emperors—after the Mongols had been pushed out of the north China plain in the 1360s, the primary concern was to cripple a still-potent Mongol military capability, divide and conquer the numerous Mongol tribes that occupied the areas along the northern border, and push the Ming's boundary of control as far north as possible, even into the inner Mongolian steppe lands. These goals were essentially accomplished by the end of the Yong Le period, largely through a series of extended offensive campaigns deep into Mongol territory in the 1370s through 1380s and in the 1410s through 1420s. But two developments in particular led to a revived and particularly acute Mongol threat by the mid-1400s. The first was the retraction of the Ming forward security line that had been established by the end of the fourteenth century. The Yong Le emperor himself had turned control of the Da Ning commandery northeast of Beijing over to the Uriyanghkha in 1403, at that time a loyal Mongol tribe (see appendix C). This in turn had led to the isolation of another forward garrison to the west, the Kai Ping garrison. Due to problems with supplying this garrison in the face of Mongol attacks, the Ming abandoned it as well in 1430, thus pulling Ming security lines back three hundred kilometers. Another garrison at Dong Sheng on the northeast edge of the Yellow River loop west of Beijing had been abandoned in the 1430s because of the vulnerability of its extended supply lines. Thus by the mid-1400s the forward security line had contracted sufficiently that Beijing, as well as other parts of the north China plain—the "stomach" of China—were far more vulnerable to Mongol military might. Whereas Da Ning, Kai Ping, and Dong Sheng had previously constituted a third and outer fence around the capital and central China, now against the campaign in Lu Quan along the border with Burma in the 1440s in HMJSWBa 4:94-96, and Ye Sheng's memorial on defense against Mongol bandits in 1450 in MCZY: 45. 8 These statistics are derived from the annual lists of wars and rebellions in Chinese Military History Group 1986. There are a number of problems with this list, one of which is that the compilers do not provide references for their data. It is clear, however, that they count only large-scale conflicts with the Mongols because the annual figures are too low to include what other historical sources indicate were at times more frequent (i.e., monthly, small-scale raiding of the northern border by the Mongols). See Pokotilov 1976: 139.

THE MING SECURITY PROBLEMATIQUE

1 85

the Da Tong and Xuan Fu commanderies, previously the second fence, became the outer line of security. The second development was the revival of Mongol military power in the mid 1400s under the leadership of Esen (Ye Xian), who had unified a number of Mongol tribes from the northwest to the northeast through force, threats, or promises. The Mongols under Esen began raiding and invading Ming boundaries, culminating in a massive offensive along the length of the border in July 1449. The Ming response was a counteroffensive led personally by the emperor. The campaign met disaster when Ming forces were utterly defeated at Tu Mu, west of the capital. The emperor was captured and Mongol forces reached the walls of Beijing. The Tu Mu defeat marked the beginning of a shift in Ming security policy to a strategy that stressed border fortifications, wall building, and linear static defense (Xu 1983: 649-50). From this point on, despite fluctuations in both the degree of unity among the Mongols and in the degree of coherence and activism in Ming security policy, the Ming military forces were essentially in a holding pattern along the border. Mongol raiding was countered with more defensive measures. On a number of occasions over the next one hundred years many Ming decision makers agreed that these static defense measures were by themselves inadequate in the long term for preventing or ameliorating a more or less constant pattern of Mongol raiding and invasion. Indeed, even proponents of wall building agreed that a more active, offensive policy aimed, in particular, at driving the Mongols out of their bases and staging areas in the He Tao region within the Yellow River loop was a more fundamental answer to the security problem.9 The issue was whether or not the Ming had the ability to do so, whether it could muster the financial resources to sustain massive offensives as well as shore up border fortifications. Debates on offensive strategies flared in the 1470s, at the turn of the century, and again in the 1540s, and usually occurred after a period of severe, sustained Mongol raids. As Tong (1985) has amply shown, the Ming capacity in the sixteenth century to mobilize these resources was steadily declining, given other external threats (the coastal pirates in mid century), the rising incidence of internal social unrest and collective violence, official corruption and a bleeding treasury, among other indicators of a downturn in the dynastic cycle. By the end of the century the Mongol threat had died down somewhat. This was not due, however, to any improvement in Ming capacity to deal militarily with the Mongols. Rather, it was due primarily to Mongol weakness from internecine strife. By the 1580s wars with 9 Before the late 1400s Mongol tribes had moved periodically into the He Tao to nomadize during springs and summers. By the 1490s the Mongol presence had become more stable and long-term, and by the first decade of the 1500s the Mongols had established a permanent presence in the He Tao from which they launched devastating raids east, south, and west. On the He Tao as a strategic problem, see Yi 1968, Chen 1984, Waldron 1990.

1 86

CHAPTER SIX

the Mongols became increasingly infrequent, with only eighteen major conflicts in the last sixty years of the dynasty (or an average of 0.3 per year). Only a third of Ming external wars in this last period were with the Mongols, a great deal below the average for the entire dynastic period. It is within this broad strategic context, then, that Ming officials debated and discussed how best to apply the economic, political, and military resources of the dynasty to defeat the Mongols. As will be evident, the proposed solutions reflected, for the most part, a parabellum approach to security problems, an approach generally consistent with the decision axioms embodied in the Seven Military Classics.

T H E PARABELLUM

PARADIGM AND ALTERNATIVE G R A N D STRATEGIES

ThQ parabellum paradigm at the core of operational Chinese strategic culture accepts that conflict and war are constant characteristics of human affairs, views the adversary as disposed to challenge one's own interests in such a manner that conflict becomes zero sum, and places a great deal of faith in the efficacy of military preparations and the application of violence to resolve these highstakes conflicts. It is clear from an analysis of 120 memorials on border policy submitted to the imperial court by a wide range of Ming officials in disparate time periods that these three elements or sets of assumptions undergirded much of the thinking about how to deal with the Mongols.10 At a very general level one finds among Ming strategists and officials a view that war and the military instrument were, at their simplest, characteristics of human society. The first emperor of the Ming, Hong Wu, stated that, "Those who run states have never discarded warfare" (Wu 1978a: 188). Zhao Benxue, the author of a mid-sixteenth century annotation of Sun Zi's Art of War accepted, as well, the necessity of military capabilities, in terms similar to what we in the West might call social Darwinian. "Because between heaven and earth there are people, therefore there is conflict. Because there is conflict, therefore there is chaos. Because chaos cannot be ordered, therefore there is warfare" (Zhao 1569: preface 1). In a more specific reference to the northern barbarian problem, the Hong Wu emperor noted in 1389 that the calamity created by the Rong and Di had deep roots. Previous dynasties had also been harassed and invaded by the northern barbarians, and this was a source of great shame for the Han (cited in Yang 1988: 6). Ming memorialists echoed this assessment, though with greater exasperation as the Mongol threat grew after the Tu Mu defeat. Ma Wensheng, a 10 Memorials on border affairs generally functioned as policy papers submitted by officials to the emperor for approval, comment, or decision. They were the primary means of communication between officials and the emperor. On the memorial system, see Wu 1968.

THE MING SECURITY PROBLEMATIQUE

1 87

minister of war in the late fifteenth century, had argued in a couple of memorials that the Mongols had long been a threat to China. "From ancient times [we] have been unable to exterminate [them], and they have been China's constant calamity" (Ma Wensheng HMJSWBa 5:523, 5:493-97, Li Chengxun HMJSWBb 2:901). Wang Shu, the minister of war in the Nanjing capital,11 memorialized in 1484 that although at that particular moment the conflict with the Mongols appeared to have abated somewhat, there was no guarantee that this state of affairs would continue for long, nor that the Mongols would not step up their attacks on Ming border defenses (Wang Shu MCZY: 74). In other words, even when there were lulls in Ming-Mongol conflict, pessimistic assumptions pervaded assessments of the long-term nature of the conflict with the northern nomads. This view of the constancy of conflict is not surprising given the zero-sum nature of Ming perceptions of the Mongol threat. The discussions of strategy are striking in their lack of Confucian-Mencian references to the Han and barbarian being of one family or similar in essence. Instead the Mongols were almost uniformally characterized as subhuman "dogs and sheep," "fierce and wild" opportunists with insatiable appetites for the wealth of Han China. As the Hong Wu emperor stated, the border tribes "are not of our race; their hearts and minds are different" (cited in Yang 1990: 284). In a text on statecraft that he is said to have compiled and annotated personally, the Yong Le emperor cited a Han dynasty assessment, "As for the Rong and the Di, in all respects they have a different essence [from the Han].12 There is no difference between them and birds and beasts" (Ming Chengzu, 1410: 524).13 The memorialists were equally direct and racialist, and more or less explicitly rejected Confucian arguments about the potential for enculturating the enemy. Li Xian, who submitted memorials in the decade prior to the Tu Mu defeat on Mongol officials and leaders in the capital who had previously surrendered to the Ming, wrote that the Han must be loved like "newborn babies," while the Mongols should be "rejected as animals" (cited in Serruys 1987: 240-41). The Yi14 and Di barbarians were only "people on the outside, but wild animals at heart" (Li Xian HMJSWBa 4:313, 309, Liu Dingzhi HMJSWBa 4:692-93). In context, u From the early 1400s on, the Ming operated two capitals, the original one in Nanjing and the real seat of power in Beijing. See Farmer 1976. 12 Rong and Di are ancient and derogatory names for the nomadic peoples along China's northern and northwestern borders. In the Han period the terms were used to describe the Xiongnu tribes, in the Tang the Tu Que, and in the Ming the Mongols. 13 The Yong Le emperor tended to issue contradictory statements about the nature of barbarians. Yang Shaoyou quotes him also as saying "as for the goodness of people's essence, there are no differences between the Man and Yi [barbarians] and the Han." As discussed below, emperors, as protectors of the Confucian legacy, were at times caught between appealing to the ConfucianMencian language and abandoning it in their discourses on security. 14 Another term for the northern nomadic tribes.

1 88

C H A P T E R SIX

rejecting the Mongols meant expelling them from the capital and scattering them in settlements inside China in order to break up any cabals; they were of suspect loyalty and consumed food and wealth that could be turned over to Chinese. Yu Qian, at the time the minister of war, remarked in memorials immediately after the Tu Mu defeat that the Mongols were "fierce and wild" {changjue MM), treacherous and deceitful by nature. In a phrase repeated in other memorials throughout the Ming period, Yu termed the relationship between China and the Mongols one of irreconcilable conflict and hatred, literally an enmity that prevented the two from living under the same heaven together (Zhongguo yu hi you bu gong dai tian zhi chou cf319 J£|fj5 W^F^-N Hc^^lfA) (Yu Qian HMJSWBa 4:125-26, Yu Qian MCZY: 24-25).15 A few decades later Qiu Jun, a senior scholar in the Hanlin Academy, argued in a memorial on what to do with Mongols who had submitted and now resided in the capital, that the Han and the barbarians were different by nature and that the latter could by no means be considered similar to the Chinese people. Despite being treated liberally by the Ming, rewarded with posts and salaries, even the submissive Mongols could not really change their nature, and posed a constant internal threat to China whenever it faced conflict with the Mongols along the border (Qiu Jun HMJSWBa 6:83-87).16 Even those who cautioned against an aggressively offensive strategy towards the Mongols accepted these zero-sum characterizations of the nature of the Mongols. Calling the Mongols of his time especially fierce and wild in comparison to those in earlier years of the dynasty, Yang Yiqing, at the time supreme commander of the Yansui, Ningxia, and Gansu military regions, argued in one memorial on border defense that they did not simply pose a threat to the security of the border areas, but to the very stability of the state (Yang Yiqing HMJSWBb2:1085).17 15 The phrase also appears, for example, in the famous memorial from Zeng Xian in 1547 proposing a massive offensive campaign to clear the Mongols out of the He Tao region (MCZY: 428), and in a memorial by Yang Jisheng arguing against negotiating peace with the Mongols in 1552 (ibid., 442). 16 Waldron identifies Qiu as an advocate of static defensive strategies (1990: 112). I7 Weng Wanda, a vigorous critic of an ambitious plan to go on the offensive to recapture the He Tao region in the mid-sixteenth century, concurred with this basic assessment of the nature of the Sino-Mongol conflict, with one qualification. The Mongols, in his view, were by nature dogs and sheep, but like humans they could reason and understand costs and benefits. Thus there was some basis for negotiation over the Mongol request in 1550 for tribute relations with the Ming, as long as the Ming were militarily prepared to deter or respond to any trickery (Weng Wanda HMJSWBb 3:2349). This was the most charitable characterization of the Mongols in the 120 memorials I reviewed. The racialist terminology and the view that the Mongols were unenculturable call into question generalizations about the Confucian influence on imperial China's relations with the outside. Samuel Kim's comment that traditional China's conceptualization of barbarians was "devoid of any racial or nationalistic imperative, as it merely conveyed to the native-born that the people so-designated stood outside the pale of Chinese cultural and

THE MING SECURITY PROBLEMATIQUE

1 89

These racialist characterizations of the Mongols tended to translate into zerosum assessments of the strategic relationship between Han and barbarian. According to many Ming memorialists the Mongols' primary strategic goal was to get access to China's material wealth, if not through plundering and raiding, then through trade and gifts. This goal, however, was not a limited one. In the view of many memorialists, limited raiding—which would have constituted for the Ming a minor, if irritating border problem—was not enough to satisfy Mongol desire for the Han people's material goods. Trading or imperial gifts bestowed on tribute delegations would only lead to increasingly strident demands, backed up by the threat of renewed raiding and invasion. Indeed, according to some, the Mongols tended to use trade and tribute delegations to spy on Ming defenses and to assess their strengths and weaknesses as a prelude to more raiding.18 Thus in the words of a number of memorialists from different periods, the Mongols' desire for "benefit" was "insatiable" (wu ya zhi qiu $£ M £ 3 < ) (Yu Qian MCZY: 34, Ye Sheng MCZY: 46, Yang Yiqing HMJSWBb 2:1109, Yang Jisheng MCZY: 443, Yin 1532: 2).19 Despite the belief, then, that the Mongols' goals were economic, there was little conceptual room for transforming the Ming-Mongol conflict into a variable-sum one where trade or tribute relations could be bargained for border security. Although most memorialists did not see the Mongol intentions as immediately political or directly aimed at overthrowing the Ming state, some argued that continuous Mongol raiding would threaten the peace and security of inner China, and thus lead to social instability and rebellion (Yang Yiqing MCZY: 197). These indirect political consequences reinforced the view that the conflict was indeed zero sum, with the essential security and stability of the state at stake. The logic of this assessment of the Mongols' intentions pushed Ming strategists to conclude that strategically and territorially the two sides were locked in a zero-sum conflict. As Yu Qian wrote in a memorial in 1450, "If we retreat a foot, then the bandits [Mongols] advance a foot. If we lose an inch, the bandits gain an inch" (Yu Qian HMJSWBa 4:152, Yu Qian MCZY: 31-32). Wang Shu argued similarly in 1484 that there was a direct relationship between the Ming sending forth its military forces and the Mongols retreating from along the border (Wang Shu MCZY: 74). Sixty years later, Zeng Xian applied the same zero-sum characterization in assessing the strategic importance of linguistic refinement" (Kim 1979: 24) does not hold for Ming characterizations of the Mongols. 18 One is reminded of John Foster Dulles's view of the Soviet Union: conflictual behavior was proof of the aggressive intentions in the Kremlin; cooperative behavior was a trick to subvert Western vigilance, hence also proof of the aggressive intentions in the Kremlin. 19 Waldron identifies Ye Sheng and Yang Yiqing as opponents of an offensive strategy towards the Mongols and as supporters of wall building and other more purely defensive measures (1990: 101, 119).

190

C H A P T E R SIX

the He Tao region to China. "If the strong enemy is not eradicated [from the He Tao], then the harm to China will increase day by day" (Zeng Xian MCZY: 427).20

O N VIOLENCE

As a number of international relations theorists have suggested, a zero-sum conceptualization of conflict is linked to a preoccupation with relative rather than absolute gains. Under this condition one's own security is directly related to changes in the relative balance of capabilities, rather than some absolute level of one's own capabilities (Powell 1991: 1303, Milner 1992: 470-73). Similarly, Ming strategists saw a direct relationship between the rise and decline of Ming strategic capabilities and changes in the nature of the Mongol threat. As Wang Shouren, at the time an official in the Ministry of Works and later a major intellectual figure in the development of Ming interpretations of Confucianism, put it optimistically in a memorial on border security in 1500, "If I am sufficient, then the enemy becomes increasingly exhausted; if I am flourishing, then the enemy declines; if I am strong and vigorous, the enemy is increasingly bent, weak; if I am rested, then the enemy is increasingly exhausted; if I am strengthened, then the enemy is increasingly empty and weak; if I am sharp, then the enemy is increasingly dulled, and ineffectual "(Wang Shouren MCZY: 167). Of course, the converse was also true. Thus Ming strategists accepted that weakness in the face of this potentially unlimited threat only exacerbated it. As one lengthy analysis of border security from the 1530s put it, "The 'barbarians' are dogs and sheep, and that is all. If you intimidate them with awesomeness [i.e., military power], then they become fearful and flee. If one displays weakness, then they become arrogant and come [to attack]" (Yin 1532: 18). In his volume on statecraft the Yong Le emperor commented, "As for the Yi and Di you cannot use the way of controlling China to govern them. They are like wild animals; seeking to use great forms of government [or great regulations] necessarily leads to great chaos. The first kings understood that things were like this, so they used 'nongovernment' to order them 20 The fact that the zero-sum racialist assessments of the Mongol threat were accepted by advocates of offense, as well as of defensive wall building, and even by some accommodationists, and persisted throughout most of the Ming period appears to weaken Waldron's argument (1990: 109) that these hard-line assessments were used largely in the sixteenth century as politically useful tools for building support for hard-line offensively oriented policies among the literati. Variations in strategic-policy recommendations were not accompanied by differing views of the Mongols or the basic nature of the conflict. Rather, the variations depended on different assessments of what the Ming was physically capable of doing to ameliorate the threat. 21 This seems to refer to the use of methods different from those used for governing the

THE MING SECURITY PROBLEMATIQUE

191

(Ming Chengzu 1410: 527). While the voluntary submission of the opponent was obviously preferred, once the potential adversary had crossed the line and rejected submission—that is, once it had become an enemy—then violence was the only efficacious tool. The Yong Le emperor quoted the Yi Jing, "Use invasion and offensive campaigns to punish those who do not submit" (ibid., 528). In an annotation of a line from Mencius about the use of punitive campaigns to punish inferiors, Yong Le commented "Punitive campaigns are how people are corrected" (ibid., 536). As for Confucius's famous dictum that a ruler's neglect in training his people in the skills of warfare is the same as discarding them, the emperor suggested this meant that the destruction of the state was inevitable if the people were not trained in war (ibid., 535). He then went on to quote a statement that drew a clear relationship between military weakness and the inability to deter threats to state security: "Military affairs cannot be downplayed. If they are downplayed then [the ruler or state] will be without awesomeness. The military instrument cannot be discarded. If the military instrument is discarded then this invites invasion and plundering" (ibid., 537). He returned to this theme in an edict on military preparations in 1420. "Since there has been a realm, no one has not taken military preparations as being most important. If military preparations are complete, then the state is strong and flourishes, and the realm experiences great peace. If military preparations are incomplete, then the state is insignificant and weak" (HMZL 1:446). This simple parabellum equation appeared throughout the memorials submitted on the border question during the Ming period. One author, writing in the mid-fifteenth century, stated plainly, "I have heard of methods for controlling the Di, and indeed none surpasses stressing martiality and the training of soldiers (Li Jian HMJSWBa 4:531). Ye Sheng, at the time an adviser to the commander of the Xuan Fu military region northwest of Beijing, wrote in a memorial shortly after Esen's forces had retreated from Beijing in 1450 that the most basic step in managing the Yi and Di barbarians was to defend the Chinese. Since the latter were favored with benevolent and virtuous policies, it implies that more coercive measures should be used against unenculturable "wild animals." Again, to be fair, the Yong Le emperor quoted other authorities whose statements implied a more orthodox Confucian-Mencian approach to the barbarians. On the same page he cited the great Confucian scholar Zhu Xi to the effect that the sage kings did not rely on awesomeness and strength to control the Yi and Di, but on virtue in government. It is hard to reconcile these differing views of the efficacy of force. And indeed, as I have argued in this study, there is no easy reconciliation of the symbolic and operational strategic cultures. Suffice it to say, the Yong Le's own behavior—personally leading five offensive campaigns against the Mongols—would appear to indicate he more readily accepted the assumptions ofthe parabellum paradigm. According to Hok-lam Chan, Yong Le's text, the ShengXue Xin Fa (J1!||/L>S;) was written "to secure his reputation as a sage ruler," and drew on Song neo-Confucianism to establish an "imperial ideology" (Chan 1988: 218-19). On the efficacy of military force for the security of the state, however, the Yong Le was not setting an example or establishing an imperial orthodoxy. He was simply reflecting the orthodoxy ofChina's parabellum strategic culture.

1 92

C H A P T E R SIX

border, and the key to defending the border was the appointment of able commanders and the training of better military forces (Ye Sheng HMJSWBa 5:23942, Ye Sheng MCZY: 39).22 Yu Zijun, famed for his role in the construction of extensive defensive fortifications and walls in the latter half of the fifteenth century and, in Waldron's view, the "strongest champion of wall-building" as a strategy for dealing with the Mongols (1990: 100), also concluded that military power was the key to Ming security. Defensive fortifications, beacon and early-warning towers, coupled with increased numbers of firearms, he argued in one memorial from the 1470s, would provide sufficient military power in one particularly vulnerable area of the border to prevent a southward invasion by the Mongols. In "one hundred battles and one hundred victories," the unyielding enemy would be coerced into submission (Yu Zijun HMJSWBa 5:347-48).23 In a later memorial he reiterated the general principle: "Only because of military preparations can one suppress conflict, defend against insult [i.e., Mongol invasion], and protect the state and the people. From ancient times no state has not relied on this [military preparations] for security" (ibid., 5: 370).24 Memorialists in the sixteenth century generally argued or assumed no differently. Han Wen, the minister of war in Nanjing briefly in the early 1500s, argued in one criticism of the sorry state of Ming military capabilities that the first emperors had concentrated strength, accumulated resources, trained and encouraged soldiers, and had at the most opportune moment attacked and killed the barbarians, thus pacifying the border. He then recited approvingly the example of Li Mu, a general from the state of Zhao in the late Warring States period, who militarily defeated and destroyed the Xiongnu barbarians. As a result, the Xiongnu did not dare to come close to the borders for ten more years (Han Wen HMJSWBa 6:618, 620-21). 22

He Zhongsheng, the author of a bingshu in the late Ming period, agreed as well that capable commanders were the key to state security. "If one obtains a [capable] general, then the military forces are strong, and the state flourishes. If one does not obtain a [capable] general, then the military forces are weak, and the state perishes" (He 1606: 1.1a). "Recall that Sun Zi's idealized dictum was that "One hundred battles and one hundred victories is not the supreme strategy, but not fighting and subduing the enemy is the best of the best." The area to which Yu Zijun was referring was the border from Da Tong west to Bian Tou Guan, a distance of about two hundred miles. The area was flat, underfortified, and vulnerable to Mongol cavalry. 24 See also the memorials from Ma Wensheng, a military administrator along the northwest border, and minister of war in 1489. Ma draws a direct relationship between military power and security. In one memorial he notes that security along the border "consists of military power being sufficient. If military power is insufficient, then if we go to war [i.e., on the offensive] we cannot win, and if we go on the defensive we cannot be steadfast" (Ma Wensheng HMJSWBa 5:438). In another he linked military power directly to controlling the Mongols: "The security of the realm is related to whether military capabilities are well prepared. If military capabilities are well prepared, then the four Yi will be in fear" (ibid., 5:510). See also ibid., 5:432, 511, 513-14,526-27,533.

military strategem

THE MING SECURITY PROBLEMATIQUE weakening, diminution of the enemy ~-—^^ defeat, submission of the enemy

political strategem

military preparations, application of force

accommodation

mobilization of resources (material, morale)

193

state security

Fig. 6.1. Fan Ji (1428).

What was common to all these memorials, regardless of the strategic policies they supported, were the clear causal paths from military preparations to the defeat or submission of the enemy and then to the security of the state. A more systematic look at the cognitive maps of a selection of memorials confirms this conclusion. The focus here is on a total of sixteen memorials over about 145 years. These were written by Fan Ji, Yu Qian, Wang Shu, Wang Shouren, Yarig Yiqing, Wang Qiong, Zeng Xian, Weng Wanda, Yang Jisheng, and Wang Chonggu, respectively.25

The memorial analyzed here (Fig. 6.1) was written in 1428 on eight issues in Ming statecraft at the time (Fan Ji MCZY: 23). The memorial came a couple of years after the Yong Le emperor's death during the last of his five offensive campaigns against the Mongols. By this time the Mongol threat had abated 25

The coding of these memorials follows the procedures used for the Seven Military Classics. See Appendix A. From these sixteen memorials submitted by nine individuals I have extracted around 430 cause-effect statements relating to the security of the state. This is admittedly not a probability sample of memorialists or memorials. I have relied primarily on Waldron 1990 and Pokotilov 1976 to identify the players in the policy processes at the time. I have chosen figures from different periods in the dynastic cycle, and have included supporters of offensive, defensive, and accommodationist strategies. I have not analyzed all the memorials each wrote on the border, but have chosen a memorial or memorials that most completely cover the author's thinking on the question. After analyzing these memorials and comparing them with the other one hundred or so I have read, I am reasonably convinced that these adequately represent the range of arguments found in debates on strategy towards the Mongols. The two numerals in the brackets assigned to a particular cause-effect statement refer to the page number of the source in which the memorial is found, and the number assigned to a particular statement found on that page.

1 94 CHAPTER SIX military strategem

weakening, diminution of the enemy^ it

^

political strategem

military preparations, application of force

accommodation, internal rectification

mobilization of resources (material, morale)

defeat, submission of the enemy

+

state security

implied causal relationship

Fig. 6.2. YuQian (1450).

somewhat and Ming policy reflected a new focus on the preservation of the status quo along the northern border, internal consolidation, and a preoccupation with the military problems associated with Ming attempts to annex Annam in the south. Here the argument structure shows both indirect and direct relations between military preparations and the security of the state. The direct paths lead from military preparations and/or the application of force directly to the successful defense of the border against Mongol raiding (23.1-23.8, 23.12). This effect concept, however, is also connected by a feedback loop to the weakening or diminution of the enemy. That is, successful defense of the border is positively related to the further exhaustion of the enemy (23.10). But attacking an exhausted enemy is directly related to the defeat of the adversary (23.9), which in turn is causally connected to the defense of the border and the security of the state (23.11). Thus, while the specific military preparations to which Fan Ji refers are primarily static defensive ones—the occupation of strategic passes, coupled with the construction of signal towers and border walls—these measures also appear to be indirect inputs into an improved capacity to attack a weakened enemy. The author advises specifically against extended offensive operations, particularly in pursuit of a defeated enemy (23.12), but he does not exclude (as statement 23.9 implies) the use of active defense measures. There are no nonmilitary routes to security in this section of the memorial (i.e., internal rectification, accommodation, etc.), nor is political or military strategem included as either a direct or indirect cause of security.

THE MING SECURITY PROBLEMATIQVE weakening, diminution of the enemy "^'v^^ ^ " ^ defeat, submission of the enemy

military strategem

political strategem

military preparations, application of force

accommodation, internal rectification

mobilization of resources • (material, morale)

195

state security

(a) refers to formal peaceful relations with the Mongols

Fig. 6.3. YuQian (1450).

Yu Qian Yu was the minister of war immediately after the Tu Mu defeat of 1449. Two of the three memorials examined here were submitted in 1450 and dealt primarily with whether or not to accept Esen's offers of peace after his retreat from the attack on Beijing in 1449-1450 (Yu Qian MCZY: 31, 34). The other was submitted in 1452 and concerned the reorganization of Ming security policy after the immediate crisis precipitated by the Tu Mu incident had passed (ibid., 49). The first memorial presents a fairly simple argument consisting of only a handful of cause-effect relationships on the question of security (see Fig. 6.2). Yu Qian draws a direct causal relationship between military preparations, specifically static defense measures such as the construction of more fortifications and improvements in logistics, and security of the border and the state (31.2, 32.2, 32.3, 32.4). There is also one indirect causal path that links improved border fortifications with security of the Chinese hinterland (32.1). Causal paths linking military capabilities, or the weakening and diminution of the adversary, to the defeat of the adversary and from there to the security of the state are only implied. One other causal factor in the security of the state is the appointment of capable, brave commanders who understand strategy (32.4). There are no causal relationships connecting political or military strategem, or internal rectification and diplomatic accommodation, to the security of the state. These nonviolent inputs come into play in the second memorial but they are related negatively to security (Fig. 6.3). Yu Qian vigorously opposed a negotiated peace settlement with the Mongols that would have allowed the establishment of formal tribute relations and the return of the captured Zheng Tong emperor. He argued that giving in to Mongol demands would not meet their

1 96

CHAPTER SIX

military strategem

weakening, diminution of the enemy defeat, submission of the v x enemy V

political strategem

military preparations, application of force

accommodation internal rectification

mobilization of resources (material, morale)

- •

state security

(a) refers to the virtue and awesomeness of the emperor •

^-implied causal relationship

Fig. 6.4. YuQian( 1452). insatiable appetite for financial and material gain, and displaying weakness would only encourage further aggression. This argument is reflected in the causal relationships in the memorial. Yu traces a direct negative causal path from accommodation with the Mongols to the security of the state (34.3, 34.7, 34.10, 34.14, 35.12). There is an indirect path as well, running from accommodation to a declining capacity of the Ming to use military force against the Mongols, and from there to state insecurity (35.13). Instead, state security is dependent directly on military preparations and the application of violence (including both defensive and offensive measures) (34.4, 34.8, 34.9, 35.13). Military morale is also identified as a source of utility in border security (34.5). Yu argues that by improving the Ming army's capacity to attack and defend, the Mongols will become fearful and retreat (34.1). Military preparations (i.e., training, capable military leadership, sufficient high-quality weaponry, and cultivating military awesomeness) are also directly and positively linked to the outright defeat of the enemy (35.1, 35.4, 35.5, 35.6, 35.8). In the third memorial there are once again a couple of causal paths from military preparations and the application of force to the security of the state (Fig. 6.4). One route is direct: the intimidating power of the state's military awesomeness is linked to pacification of border problems (49.1,49.3). Another is indirect, whereby superior military capabilities (i.e., a capable general, soldiers' ability to maneuver, and the offensive application of force) are causally linked to the defeat of the enemy (49.4, 50.1, 50.2, 50.3, 50.5, 50.10, 50.11),

THE MING SECURITY PROBLEMATIQUE weakening, diminution of the enemy

military strategem

197

defeat, submission of the enemy

political strategem

military preparations, application of force

-(a) accommodation, internal rectification

state security

mobilization of resources (material, morale)

(a) refers to accommodation policy towards Mongols

Fig. 6.5. Wang Shu (1484). which in turn is, by implication, an input into border security.26 The memorial also indicates that superiority in firearms is an input into the military weakening of the enemy (50.4,50.8). Unlike the other two memorials or Fan Ji's comments, Yu Qian here infers a role for strategem as a cause of victory (51.1). But it is unclear from the reference whether he means political or military strategem. Also, Yu draws a direct causal relationship between the virtue and awesomeness of the emperor on the one hand, a variable associated with internal rectification, and deterrence of the enemy and security of the state on the other (50.12).

Wang Shu &ffi) Wang submitted this memorial on maintaining order and protecting the state in 1484 (Wang Shu MCZY: 74) (Fig. 6.5). Although extensive wall building in the 1470s had improved somewhat the Ming capacity to limit Mongol penetration of the border (Mote 1988: 401-2, Waldron 1990: 107-8), this memorial came in the midst of a period of incessant Ming-Mongol conflict in the 1480s (Pokotilov 1976: 79-81), almost all of which was initiated by the latter. Wang, too, argued that the keys to state security were offensive and defensive military 26 In Yu Qian's words, "That which the state relies upon is the military instrument. That which the military depends upon is the general. If one acquires a [capable] general, then of the military forces none will not be first-rate. If among the military none is not first-rate, then the awesomeness of the state will intimidate [the enemy], and the border threat will naturally be pacified" (Yu Qian MCZY: 49).

198

CHAPTER SIX

military strategem

+

weakening, diminution of the enemy defeat, submission of the enemy

political strategem

accommodation, internal rectification

state security

-••mobilization of resources (material, morale)

(a) The relationship is negative for offensive strategies when the Ming are weak, and positive for defensive strategies when the Ming are weak.

Fig. 6.6. Wang Shouren (1500). capabilities. The memorial draws a direct causal link between the qualitative and quantitative superiority of the military and the absence of enemies "in all directions" (74.1, 74.8). Military strength also "causes" one's offensive capabilities to be "invincible" (75.5). The offensive application of military strength is also positively related to the retreat of the enemy (74.7, 75.2). As for the causal inputs into military capabilities, the memorial includes military morale and military finances (75.1). Wang also argues that extended periods of peace lead to the atrophy of military capabilities because the state tends to neglect "thinking of danger when residing in peace." This implies, interestingly, that internal rectification—to the extent it focuses on Confucian-Mencian civility to the neglect of martiality—is negatively related to the security of the state.27

Wang Shouren The great Ming scholar official submitted this memorial on general border policy in 1500 (Fig. 6.6). The previous decade or so had seen a growing, more or less permanent Mongol presence in the He Tao region after a temporary 27 Blaming the atrophy of Ming defenses on long periods of peace was common. This does not mean that all the memorialists advocated high and constant levels of conflict with the Mongols. Quite the opposite, since constant warfare was also blamed for the waste of state resources from corruption, desertion, and the basic costs of defense construction. Rather, the argument reflected reality: when the Mongols did not pose a present danger, fortification construction, training, and discipline tended to slide. See Wang Ao HMJSWBb 2:1146, Yang Yiqing MCZY: 197, Weng Wanda HMJSWBb 3:2364, Li Chengxun MCZY: 374, and Yang Jisheng

THE MING SECURITY PROBLEMATIQUE

1 99

retreat from the area in the 1470s and early 1480s. Although one source suggests that up to 1500 the Mongol raids along the Ming border had declined somewhat due to internecine conflict (Chen 1984: 42-43), others suggest that the 1490s was a period of unrelenting and devastating raids (Pokotilov 1976: 8485). In the decade prior to 1500 the Ming had engaged in at least five major conflicts with the Mongols, in only two of which the Ming had initiated the attack (Chinese Military History Group 1986). In this context, the underlying argument structures in Wang's lengthy memorial, though more complicated, followed much the same pattern as previous memorialists on the question. The memorial identifies three direct causal inputs into the security of the state. One of these, rare for most memorials, is internal rectification, specifically the quality of the officials who run the state and administer the borders (162.1, 164.1).28 Another direct input is military preparations and/or the application of military force. Here, however, the memorial identifies defensive measures, rather than offensive ones, as being positively related to security. Given a strategic context in which the Ming is relatively weak, offensive operations that involve transporting food and supplies over long distances (165.7), and engagements in open/field warfare (167.1) are negatively related to the security and prosperity of the state. Under these conditions, static defensive strategies such as "closing up the gates and consolidating defense" (167.2), extending the system of warning towers and beacons (167.3, 167.6), and training and deploying first-rate soldiers (166.18) are considered either of military utility or lead explicitly to the security of the state. A third direct source of state security and prosperity is the weakening or diminution of the enemy short of its outright defeat and capitulation (165.4). Although the memorial does not draw an explicit direct connection between the submission of the adversary on the one hand and state security on the other, it does imply, of course, that military victory is directly related to this outcome. As for causes of the enemy's defeat and submission, again military force plays a prominent role, both direct and indirect. The memorial directly links the quality of command (163.1, 163.2, 167.10) to the defeat of the adversary and the survival of the state. Other inputs include the use of mobile forces to ambush enemy forces (167.25), enveloping attacks on the enemy (167.28), the use of surprise attack (167.27), attacking the enemy where it is undefended (167.26), the concentration of momentum and physical force against the enemy (167.8), and the general development of military capabilities (166.9). Note here the role of offensive uses of force. Despite the apparent contradiction MCZY: 443. Yang and Weng were on opposite sides of the debate in the late 1540s over whether to go on the offensive to recover the He Tao. It is clear, then, that this assessment cut across differences in grand strategy. 28 Wang also identifies the cultivation of awesomeness as a direct input into state security (166.17), though it is not clear whether this refers to military power or the image of the emperor.

200 CHAPTER SIX military strategem

weakening, diminution of the enemy defeat, submission of the enemy -(a)

political strategem

military preparations, application of force

accommodation

mobilization of resources (material, morale)

- •

state security

(a) refers to offensive use of force when the Ming are weak

Fig. 6.7. Yang Yiqing (1506). with comments about the efficacy of static defense, this is consistent with the contingent nature of the admonitions against offensive force in other causal paths. In principle, offensive actions are causally related to the defeat of the adversary, but in periods of relative weakness static defensive measures are preferred. In addition to the application of violence, not surprisingly, the memorial identifies the diminution, exhaustion, and military vulnerability of the enemy as a cause of its defeat. The relationship is both direct (167.15,167.16,167.18167.21) and indirect, as the enemy's vulnerability creates conditions for the more effective application of violence (167.21-167.23). This variable is then related in a feedback loop to the weakening of the adversary (165.8, 167.12167.14, 166.19). Military morale is also a causal input into the defeat of the enemy (165.9, 166.7). As for variables that could be subsumed under the heading of internal rectification, as I noted, only one leads directly to the security of the state. The others are all inputs, ultimately, into the efficacious application of force. One set of inputs, which is causally related to military capabilities, includes adequate food production and financial levies to supply sufficient military forces (164.3, 164.4, 165.6). Another set relates to the court's declarations of virtue and righteousness as tools for whipping up morale within the military (166.3— 166.5). Morale, in turn, is a causal input into improved military capabilities (166.6, 166.8). Unlike the previous memorials, Wang's also includes military strategem as a causal variable. The use of spies, for instance, is linked directly to the capacity to use military capabilities effectively (167.4), while "attacking the enemy where he is unaware" is linked directly to the defeat of the enemy (167.26).

military strategem

THE MING SECURITY PROBLEMAT1QUE weakening, diminution of the enemy

201

defeat, submission of the enemy

political strategem

accommodation, internal rectification

military preparations, application of force

+

state security

mobilization of resources (material, morale)

Fig. 6.8. Yang Yiqing (1506). This latter causal concept implies the use of deception, but it is not, strictly speaking, a reference to defeating the enemy without resorting to force.

Yang Yiqing (#—tff) Yang was appointed superintendent of border affairs in 1506 in charge of military affairs along the Yansui, Ningxia, and Gansu borders. In the same year he submitted two memorials, one addressing restructuring key aspects of border defenses, the other strategic opportunity. He was writing at a time, according to Pokotilov, when the "entire border of Gansu up to Beijing was exposed to persistent and incessant Mongol attacks . .. [and] the endeavours of the Chinese generals to repulse these attacks failed" (Pokotilov 1976: 98, Chen 1984: 42-43). From 1500 to 1505 the Ming had been involved in five major conflicts with the Mongols, none of which had been initiated by the Chinese side (Chinese Military History Group 1986). Perhaps most worrisome to the Ming was that by the beginning of the sixteenth century the Mongols had established a more or less permanent presence in the He Tao region from which they could launch raids in any direction into China. The cognitive map of Yang's memorial on restructuring border security does not deviate from the general pattern thus far (See Fig. 6.7). State security is an effect of the defeat of the enemy (197.5,198.10). This cause concept is in turn an effect of offensively applied military force (in this case, the recovery of the He Tao region from the Mongols) (198.10), and the military diminution and strategic weakening of the Mongols (197.4). The strategic weakening of the Mongols is also an effect of strong, primarily defensive, military capabilities, including improved border fortifications, better logistics, and well-trained, wellsupplied military forces (197.9,197.10,197.14). The memorial contends that, under conditions of relative weakness, extended offensive operations beyond the borders are negatively related to the strategic weakening of the Mongols.

202 CHAPTER SIX military strategem

weakening, diminution of the enemy defeat, submission of the enemy

political strategem

accommodation and internal rectification

military preparations, + , . application of force

^state security

mobilization of resources (material, morale)

(a) the relationship is positive for offensive strategies and negative for defensive strategies.

Fig. 6.9. Wang Qiong (1516). State security is also "caused" by efficacious military capabilities (197.2), namely a strong static defense capacity (197.3, 197.6, 197.15, 197.16) as well as a capacity to engage in offensive operations (198.7, 198.11). Like Wang Shu's memorial, Yang identifies extended periods of peace as being positively related to the atrophying of Ming military capabilities, in other words, as negatively related, in the end, to Ming security (197.8). It would seem, however, that this particular cognitive map is imbalanced. That is, the direct causal path from military capabilities to state security is a positive one, while the other, indirect path from military capabilities to the diminution and defeat of the enemy to the security of the state is a negative one. Put another way, depending on the causal path followed, an increase in military capabilities or the application of violence leads to both an increase and a decrease in state security. The key to resolving this contradiction lies in Yang's conditionality; this map only remains imbalanced to the extent that the Ming is relatively weak. Under those conditions, offensive operations are negatively related to the diminution of the enemy. When these conditions are removed, offensive strategies contribute to the enemy's weakening, his defeat, and thus to Ming security. Yang's other memorial on border policy (Yang YiqingMCZY: 198) (Fig. 6.8) is somewhat different from the first one in that it does not include a variable associated with accommodation, and it also draws a direct causal linkage between the strategic weakening of the Mongols and Ming security (198.2). In addition this memorial leaves out the conditional or contingent rejection of offense found in the first one. Here he draws a direct causal relationship between limited offense against the Mongols in the He Tao (attacks on enemy camps if they are within two hundred li or one hundred kilometers of the Ming borders)

military strategem

THE MING SECURITY weakening, diminution of the enemy

PROBLEMATIQUE

203

defeat, submission of the enemy political strategem

accommodation and internal rectification

military preparations, application of force

+

state security

mobilization of resources (material, morale)

Fig. 6.10. Wang Qiong (1516).

(199.2-199.4), as well as between more obviously defensive measures (198.3, 198.4) and border security.29 The application of violence is also an indirect input into state security via a path running to strategic weakness of the enemy (199.6), to the military defeat of the enemy (199.7), and finally to state security (199.8).

Wang Qiong (EE3$L) Wang held a number of senior positions in the Ming government, including minister of war from 1515-1520. During this time he was a prolific author of memorials on border security. Once again, this was not a period of much optimism in the Ming court about how to manage the Mongol threat. From 1510 to 1520, in addition to frequent small-scale raiding, the Ming and Mongols engaged in seven major wars. In all of these the Ming responded more or less defensively to the strategic initiatives of the other side (Chinese Military History Group 1986). Wang's first memorial (Fig. 6.9) in 1516 dealt with raiding by the Uriyanghkha Mongols northeast of Beijing (Wang Qiong MCZY: 243). In this memorial military preparations and the use of violence are the sole direct sources of state security. Specifically, offensive operations against the enemy and improved defensive measures are both linked positively to defense of the border (242.3, 243.9, 243.8), as is military strategem, namely attacking the enemy where it does not anticipate it (243.4). Relying solely on static defensive measures or 29 Waldron 1990: 119, 129 rightly argues that Yang did not believe that at the time the Ming forces were capable of going on the offensive to recover the He Tao, and therefore supported more static defensive measures. But it is clear from Yang's causal arguments, and his recognition of the long-term strategic necessity of recovering the He Tao, that he was not an opponent in principle of an offensive strategy, and indeed saw an offensive strategy as essential to resolving the border crisis. For Yang the determining variable was the relative Ming capacity to initiate an offensive strategy and follow through. He did not a priori reject offense.

204

CHAPTER SIX

military strategem

weakening, diminution of the enemy rdefeat,

enemy

political strategem

military preparations, application of force

accommodation and internal rectification

mobilization of resources (material, morale)

+

submission of the

state security

Fig. 6.11. Wang Qiong (1520). recklessly picking minor fights with the enemy are both negatively related to security (243.2, 243.5). Military preparations and capabilities are also positively related to the strategic weakening of the enemy (243.7) and, by implication, to the military defeat of the adversary. Accommodationist policies towards the enemy such as allowing tribute relations, or concluding a peace treaty, are negatively associated with border security (243.1, 243.9). The second memorial from 1516 (Fig. 6.10) is essentially similar to the first (Wang Qiong MCZY: 244). It, too, identifies the combination of offensive and defensive capabilities and actions as causes of border security (243.6, 243.7), but links reckless provocations of the enemy for short-term military gain to decreased security (243.1). Attacking the enemy when and where it does not anticipate it is also directly linked to security (243.5). Effective border defense is positively related to the strategic weakening of the enemy (243.2, 243.4). The only difference between this argument and that of the first memorial is that accommodation with the enemy is linked directly to both the strengthening of the enemy and declining border security (243.2, 254.8). The third memorial (Wang Qiong MCZY: 276) (Fig. 6.11) is consistent with these first two. Military preparations, specifically qualitative and quantitative superiority, and the application of violence are causes of border security. Here Wang stresses the relative value of offensive and active defensive uses of force alone, more so than in the first two memorials. Attacks on the Mongols beyond the border (276.9), and preempting and intercepting Mongol raiding parties (277.1, 277.2, 277.4), are both positively related to border security. In addition, military strategem in the form of night attacks on Mongol bases and camps is directly linked to border security (277.3). The defeat of the adversary is a direct effect of the use of qi bing according to circumstances (276.3). The strategic weakening of the adversary is due variously to the reliance on "awesomeness and martiality" (276.1), and to well-prepared defenses (276.6,276.7).

THE MING SECURITY PROBLEMATIQUE 205 weakening, diminution of the enemy defeat, submission of the enemy

military strategem

military preparations, aDDlication of force

political strategem

accommodation and internal rectification

• • • state security

mobilization of resources (material, morale) (a) refers to internal rectification (b) refers to accommodation, and peaceful relations

Fig. 6.12. ZengXian( 1547).

ZengXian (#ffe) Zeng was the most articulate proponent of an offensive strategy to recover the He Tao region in the mid-sixteenth century. At the time commander of military forces in Yulin, Ningxia, and Guyuan, Zeng wrote his first memorial in 1547, the submission of which sparked an acrimonious debate in the court over strategy towards the Mongols (see Yang 1960: 1220-24, Yi 1968: 197, Pokotilov 1976: 111-12, Waldron 1990: 122-39). By this time the Altan Khan had consolidated Mongol military power under his leadership, and many of these forces were based permanently in the He Tao region. The memorial came after the Ming had rejected repeated requests from the Mongols in the 1540s to open tribute and trade relations. The rejection had led to more frequent and more threatening incursions along the border. It was in this context that Zeng proposed what he believed was a long-term solution to the Mongol threat.30 The memorial is quite lengthy, and the causal relationships fairly complex (Zeng Xian MCZY: 426) (Fig. 6.12). But in essence, Zeng, like the other memorialists, argued that the key to state security was the application of violence, in this case predominantly offensive. There is one direct causal link between internal rectification and the defeat of the adversary (427.1), but this is, in context, clearly a perfunctory use of the Confucian-Mencian discourse. It appears at 30

His plan was to use about sixty thousand footsoldiers, special firearms troops, cavalry, and even shipborne forces to attack Mongol bases in the spring of each year for three successive years. This was the period when Mongol horses, lacking fresh fodder over the winter, were weakest and hence Mongol cavalry least effective. During the fall of each year the Ming would then fall back behind static border fortifications and wear down any counterattacking Mongol cavalry (Zeng Xian MCZY: 429).

206

CHAPTER SIX

the start of the memorial, and is dropped from the argument in the subsequent discussion of concrete security problems. Otherwise related concepts—such as the financial/material wealth of the state (429.10, 429.11, 431.9-431.11)— are causal inputs into military preparations and the application of violence. The sources of state security are threefold. One is the defeat and submission of the adversary, specifically the military elimination of the Mongol from the He Tao region (427.19, 429.1, 429.2). This cause is in turn an effect of the strategic weakening of the enemy (431.5,429.14), defensive military preparations (428.5, 431.2), the application of preemptive offensive violence (428.8, 430.8,430.10,431.1,432.2), the quality of command (431.21), and the quality and quantity of firearms and other weapons (435.1). The strategic weakening of the enemy—in this case preventing the barbarians from encroaching on the northwest border and the He Tao region—is also a direct cause of state security (427.8, 427.12,428.4, 431.7). The other source of state security is simply the application of military force (427.8, 427.16, 428.1, 428.3, 428.9, 429.5, 429.6,429.16,429.17,432.3,434.1). The memorial, however, draws a negative relationship between exclusive reliance on static defense along the Ming borders and state security (427.6, 428.6). Unlike, say Yang Yiqing's conditional rejection of offensive strategies, this is not a contingent rejection of static defense. Regardless of the Ming-Mongol balance of capabilities, static defense is considered disadvantageous, and indeed without an offensive strategy against the Mongols in the He Tao this balance cannot be tipped in the Ming favor. The application of military force also leads to the wearing down of the enemy. To the extent the Mongols are pushed out of the He Tao they will be at a strategic disadvantage when raiding Ming borders in the future. This cause concept can be broken down into smaller categories such as offensives against the Mongols in conjunction with the extension of Ming defense lines further north and west, active defense measures, and superior weaponry, among others (427.3, 427.2, 427.5, 427.7, 427.11, 427.13, 427.17, 427.18, 428.7, 428.10, 430.5, 429.13, 430.3, 430.4, 434.4). There are a couple of feedback loops as well in this memorial. The strategic weakening of the enemy (429.3,430.6), as well as the enemy's defeat (429.12) are positively related to improvements in the Ming capacity to use military force defensively and offensively. There is only one causal relationship drawn between military strategem, namely attacking the enemy where he is undefended, and the defeat of the enemy (430.9). The memorial does not mention political or nonviolent strategem in the process of defeating the enemy or achieving state security.

Weng Wanda ( H H i t ) At the time Zeng Xian submitted his proposal for an offensive strategy against the Mongols, Weng was the governor-general of the Yan Sui military region to

THE MING SECURITY PROBLEMATIQUE military strategem

207

weakening, diminution foi the enemy ^v^x^ defeat, submission of the enemy

political strategem

accommodation

military preparations, application of force

l+

state security

mobilization of resources (material, morale)

(a) refers to offensive use of forces when the Ming is weak ^o.

implied causal relationship

Fig. 6.13. Weng Wanda (1547). the west of Beijing, and perhaps the most articulate critic of Zeng's plans. He was, as well, a supporter in the early 1550s of accommodation with the Mongols after the Altan Khan had defeated Ming forces and raided close to Beijing itself. In 1547 he submitted a memorial in response to Zeng's where he proposed instead that the Ming improve its static defenses in the Xuan Fu and Da Tong commanderies (Weng Wanda HMJSWBb 3:2364) (Fig. 6.13). The memorial is devoted almost entirely to an attack on Zeng's plans. Probably for this reason Weng draws an exclusively negative relationship between the application of offensive military force and state security (2364.15, 2364.16, 2366.8,2366.9). All of these causal relationships have to do with the allegedly enormous financial and material costs of an offensive into the He Tao. Similarly, the memorial argues that an offensive strategy is negatively related to the defeat of the enemy, or put another way, positively related to the defeat of the Ming, again due to the tremendous difficulties in supplying and coordinating such an operation (2364.2, 2364.6, 2364.9, 2364.10, 2364.11, 2365.7, 2365.8). This does not mean, however, that Weng completely rejected the utility of military preparation or military force. Indeed, the memorial establishes a negative causal relationship between extended periods of peace and the weakening of Ming defenses, with concomitant negative effects for China's border security (2364.1). Instead, it draws a positive relationship between more static defensive measures such as improved border fortifications (2366.2, 2366.3) and improved training within the military forces (2366.4) on the one hand and the weakening of the enemy on the other. Military morale is also a cause of the

208

CHAPTER SIX

military strategem

weakening, diminution of the enemy defeat, submission of the enemy

political strategem

military preparations, application of force ~

state security

•(a)

accommodation - (b)

mobilization of resources (material, morale)

(a) Ming attach restrictive conditions to accommodation policy. (b) Ming do not attach restrictive conditions to accommodation policy. implied causal relationship

Fig. 6.14. Weng Wanda (1551). enemy's decline short of outright defeat (2366.5). All of these variables, Weng argues, will allow the Ming to better exploit weaknesses in the Mongols, given the opportunities.31 In another memorial written in 1551, Weng addressed the controversial issue of opening tribute and trade relations with the Mongols shortly after they had attacked right up to the walls of Beijing (Weng Wanda HMJSWBb 3:2366) (Fig. 6.14). The policy was adopted that year and horse fairs were opened in the Xuan Fu and Da Tong commanderies. But they were shut down later that year when the Altan Khan allegedly resumed raiding and plundering along the border. This began another round of intense border conflict for the next five years or so.32 Although he advocated accommodation, Weng's memorial showed a marked degree of ambiguity about the utility of the policy. The cognitive map indicates that accommodation has both positive (2366.2) and negative utility (2366.1) in that it could lead to the Ming appearing militarily weak, or it could buy time and reduce the chances of a retaliatory strike from the Mongols during a period when the Ming were relatively weaker. The solution, according to Weng, is to attach political conditions to accommodation (such as pledges to refrain from raiding for one year, the return of captured Chinese, etc.) so that 31

See also an earlier memorial by Weng in which he is more explicit about the role of defensive force in state security (Weng Wanda HMJSWBb 3:2354; cf. 3:2355). 32 From 1551 to 1555 there were nine major conflicts, only one of which was initiated by the Ming,

military strategem

THE MING SECURITY weakening, diminution of the enemy

PROBLEMATIQUE

209

defeat, submission of the enemy

political strategem

military preparations, application of force

accommodation

mobilization of resources (material, morale)

state security

Fig. 6.15. Yang Jisheng (1552). accommodation would not appear like capitulation. Buying time for improvements in military capabilities will, as the causal relationship between military capabilities and the diminution of the enemy (2366.3) indicates, have a salutary longer-term effect on Ming security (2366.4). The structure of this memorial suggests, then, that accommodation was a conditional strategy useful when the Ming were weak and appeared to have no other choice in reducing the level of threat. Accommodation, therefore, allowed Ming military capabilities to recover. These capabilities, not accommodation per se, were the sources of long-term military security.

Yang Jisheng (i§li§|) Yang was another participant in the debates in the mid-sixteenth century over strategy towards the Mongols. He was appointed vice-director of the Bureau of Equipment in the Ministry of War in 1551, and was a vigorous opponent of the brief policy of accommodation towards the Mongols at that time. His critique came in a memorial submitted in 1552 requesting that the newly opened horse fairs with the Mongols be shut down (Yang Jisheng MCZY: 441) (Fig. 6.15). Like most of the other memorialists, Yang starts from the assumption that the security of the state comes largely from military preparations, military power, and the application of violence. For the most part, this translates into offensively oriented uses of force. Attacking the Mongols in a punitive campaign (442.1-442.4,445.10,445.8) is a direct cause of the security and prosperity of the state, as are active defense measures such as attacking the enemy when it raids (443.17). In addition, the memorial draws a positive causal connection between military awesomeness in the eyes of the Chinese people and the deter-

2 10

CHAPTER SIX

military strategem

weakening, diminution of the enemy defeat, submission of the enemy

political strategem

military preparations, application of force

accommodation, internal rectification

mobilization of resources (material, morale)

state security

(a) refers to use of force when the Ming are weak

Fig. 6.16. Wang Chonggu (1571). rence or suppression of internal unrest and rebellion (443.13,443.15,443.16). Martial awesomeness is also linked to the diminution of the enemy (442.5) as well as to its defeat (442.6). On the other hand, tribute relations and horse fairs—measures that Yang labels a "peace" policy—are related directly and negatively to the security, prosperity, and awesomeness of the state, and to the well-being of its people (441.1,441.2,442.8,442.9,443.1,443.6,444.1,443.19,443.20,444.4,444.8, 445.1, 445.6, 445.9). Accommodation is also related negatively to the defeat of the adversary inasmuch as it allows the enemy to control the Ming rather than the reverse (444.6, 444.7). As for Weng Wanda's argument that accommodation would give the Ming time to build up its own defenses, Yang links accommodation to a loss of Ming military awesomeness (445.4) and the flourishing of enemy military power (445.5). Accommodation is also linked vaguely to military/political advantages for the enemy (445.7).

Wang Chonggu Wang was a key figure in policy towards the Mongols in the 1560s and 1570s. He had command and administrative experience in both campaigns against pirates along the east coast and in the border conflict with the Mongols. In 1567 he was appointed commander of Shaanxi, and was instrumental in the conclusion of a peace settlement with the Altan Khan in 1571. This agreement was due in large measure to the fact that two prominent Grand Councillors, Gao Gong (ri&lft) a n d Zhang Juzheng (3RJg|IE), shared Wang's view that peace with the Mongols, even if temporary, would buy time for the revitalization

THE MING SECURITY PROBLEMATIQUE

2 11

of Ming military power. This had atrophied considerably due to an increase in the frequency of conflict with the Mongols and a surge in coastal piracy and internal rebellion in the two decades since the last short-lived peace (Tong 1985, Chinese Military History Group 1986, Waldron 1990: 184-86). In a 1571 memorial submitted in support of an accommodationist policy he outlined an argument similar in most respects to the one made by Weng Wanda twenty years earlier (Wang Chonggu MCZY: 515) (Fig. 6.16). The memorial seems to establish a negative relationship between the application of violence and China's security. Given Ming weaknesses, efforts to control the Mongols through martiality are causally connected to renewed Mongol attacks on China. That is, unless the Ming accepted a peace settlement, a return to "old" policies that relied on military conflict and the rejection of trade relations would likely lead to Mongol retaliation (516.6, 516.7). Reliance on these "old" policies is also causally linked to increased incentives for the Mongols to continue harassing the Ming (516.5). Accommodation, however, is positively related to the security of the state (515.3), as well as to improved strategic conditions for the Ming (525.2). But, as in Weng Wanda's memorial, the utility of accommodation is conditional; it depends on the Mongols meeting certain demands such as the return of Chinese prisoners, as well as agreement on the Altan Khan's part not to move about within the borders once he submits. These conditions are designed to ensure that China does not appear weak in the face of Mongol requests for formal relations, and to prevent any trickery while the Mongol tribute delegations are inside China. There is also a threat of coercion in all of this. The Chinese threat to declare the death penalty for the Altan Khan is causally linked to his acceptance of the Ming's conditions (515.7). Thus there is an implied feedback loop between the enemy's submission to Chinese conditions for peace and the utility of accommodation. Unlike other memorials, this one also includes a causal role for political strategem. Threats of execution (515.5, 515.6) and efforts to entice other Mongol tribes to submit (516.1) are positively related to deterring the violation of the conditions of accommodation. These are also related to accentuating divisions in the Mongol camp, thus weakening it. The memorial argues that internecine struggles among the Mongols will allow the Ming to rest (516.2), implying that political strategem will give the Ming time to improve its military capabilities vis-a-vis the Mongols. If so, then the memorial sets up an apparently contradictory causal argument where political strategem within a policy of accommodation is desirable in so far as it allows the Ming to improve its military capabilities, but where military capabilities and their use are causally linked to a decline in state security. Once again, though, the arguments against the use of force are conditional. Force is related negatively to security only under present conditions of Ming weakness, not under future conditions of revived Ming power after the successful implementation of a policy of temporary ac-

212

CHAPTER SIX

commodation. Indeed, in another memorial submitted around the same time, Wang makes it clear that in principle the use of force, primarily for active defense purposes, is causally linked to security along the border (Wang Chonggu HMJSWBb 4:3382). The active pursuit of and attack on Mongol forces that have violated Ming borders is connected positively to the deterrence of further attacks (3382.4), as are improved defensive walls, forts, and warning beacons (2382.6-3382.10).

CONCLUSION

I have tried to show that a parabellum paradigm pervades the causal arguments in Ming policy memorials. Ming decision makers viewed war in general, and conflict with the Mongols in particular, as a constant theme in history. The constancy of conflict was due, in their view, to the inherently aggressive disposition of the adversary whose thirst for China's wealth and material power was unquenchable. This assumption implied that conflict with the adversary was essentially zero sum. The resolution of this conflict required superior military capabilities and their application so that the enemy was denied the opportunity to act according to its disposition. The memorialists also accepted, however, that the form of strategy, or variations in the type of force applied, depended on changing circumstances. In addition to the fundamental internalization ofthe parabellum paradigm, then, Ming policy makers also explicitly accepted the concept oiquan Man as a key decision-making axiom. Conflict and war, like other aspects of existence, were processes in constant transformation. The methods of responding to this change were necessarily variegated, and thus could not be a priori limited. In Yu Qian's words, "The methods of employing the military instrument are hard to perceive, just like yin and yang; and hard to comprehend, just like ghosts and spirits. Thus one should stress responding to change at any particular moment. It is difficult to seek [victory] on the basis of a fixed [plan]" (Yu Qian MCZY: 51). Security rested on exploiting enemy weaknesses once the opportunities presented themselves. This might require, as Sun Zi argued, going on the defensive to wait for the enemy to reveal its intentions, strengths, and weaknesses. But it also invariably required the rapid, preemptive, violent exploitation of these weaknesses the moment they appeared, whether on a tactical, strategic, or grand strategic level. The memorials contain numerous recommendations that Ming forces should "attack or defend according to opportunity" (xiangji zhan shou f i ^ f c T F ) or variations on this theme (Yu Qian MCZY: 34, 50,51, Wang Shu MCZY: 75, Yang Yiqing HMJSWBb 2:1102, 1104,1110, Wang Qiong MCZY: 276). Zeng Xian argued in his memorial on the recovery of the He Tao that "opportunity is a function of timing and circumstantial power [or momentum]. . . . If we obtain opportunity and exploit it, then victory resides with us"

THE MING S E C U R I T Y PROBLEMATIQUE

213

(Zeng Xian MCZY: 430). He likened the exploitation of opportunity to defeat the Mongols in the He Tao to the commonly used images of lightning and thunder striking. Opponents of the offensive did not reject the notion of quart bian, but argued only that an advantageous opportunity had not yet arrived. The prevalence of aparabellum calculus in the memorials was accompanied by a virtual absence of Confucian-Mencian language about good government as the basis of security. Occasionally a memorial might refer to the military and warfare as an "inauspicious instrument" to be used "only in unavoidable circumstances" (see Fan Ji MCZY: 24), but for the most part the entire discourse on policy towards the Mongols took place within the context of these "unavoidable circumstances" precisely because of parabellum assumptions about the nature of the enemy. Zeng Xian used this language to defend his proposal for an offensive campaign against criticisms that he was too belligerent. "Although I am stupid, how can I not understand that war is inauspicious and battle is dangerous, and that one must not recklessly act?" (Zeng Xian MCZY: 428). At times some memorialists would refer to the utility of benevolence for enticing the submission of the "four barbarians." Even the hard-line Minister of War Yu Qian wrote, "If the emperor's virtue and magnanimity is displayed afar, the enemy will then definitely not dare to invade and violate the capital" (Yu Qian MCZY: 50). For the most part, however, magnanimity never stood alone as a source of security. Rather, it was linked to a military power designed to intimidate or coerce submission. Yu Qian put it this way: "As for China's controlling the Yi and Di, we should certainly intimidate them with military awesomeness and cherish them with kindness and trustworthiness" (Yu Qian HMJSWBa 4:203, ibid., 4:196, Zeng Xian MCZY: 427). Moreover, these references to Confucian benevolence were found almost invariably near the start of memorials that then, like Yu's, proceeded to argue exclusively for the application of force and violence. Indeed, the only consistent cross-memorial uses of the Confucian-Mencian language came in the normative terms used to describe Ming uses of force. These were often cloaked in the language of righteous war. Yu Zijun, for instance, referred to the historical use of force by Chinese emperors as "punishing the violent, and preventing chaos" (zhu baojin luan HfcH^flL) (Yu Zijun HMJSWBa 5:359). Zeng Xian called the planned campaign to recover the He Tao "heaven's punishment" (tian tao ^ U ) , aimed at "eliminating violence and disorder" (Zeng Xian MCZY: 429). There were, as well, surprisingly few references to idealized notions of minimalist violence associated with Sun Zi. While quotes from The Art of War were not uncommon in the memorials, there was virtually no discussion of the concept of "not fighting and subduing the enemy." Indeed, in one memorial Yu Zijun came close to explicity rejecting the axiom. He suggested that an effective defensive strategy would lead to "one hundred victories in one hundred battles" and the coerced submission of the enemy (Yu Zijun HMJSWBa 5:354).

214

C H A P T E R SIX

In one instance, "not fighting and subduing the enemy" was cited to describe a complex strategy against the Mongols involving feints and deceptions along a static defensive line to keep the enemy guessing about Ming strength and deployments. After a while the enemy will retreat, from indecision mostly, and then the Ming should launch a sneak attack using elite calvalry. The Mongols will then cease raiding. This is what the memorialist called "not fighting." In fact the strategy was a combination of deception and fighting, where deception created the conditions for effective fighting (Guo Deng HMJSWBa 5:195-96). Another memorial, this one by Wang Ao (zESI), refers to Sun Zi's famous axiom about attacking the enemy's strategy and his alliances, but this is used to describe the employment of spies only. This was just one of eight steps for dealing with the Mongols; the others all related to the offensive application of force. In other words, the corollaries of "not fighting and subduing the enemy" were only a small part of an overall violent strategy (Wang Ao HMJSWBb 2:1149). One late Ming text on strategy admitted plainly that "not fighting and achieving victory" was exceedingly difficult in practice (He 1606: 9.2a). The citations from Sun Zi in other memorials clearly dealt with mustering force and applying it in an offensive fashion. One memorial cited Sun Zi's axiom "first put oneself in an indefeatable position and await the enemy to become defeatable." This principle was used to justify a strategy whereby the Ming would first go on the defensive against the Mongols (e.g., by first improving early-warning capabilities and quickening the responsiveness of the army to raids at any particular point along the border). Then once the army accumulated sufficient material power and fighting spirit, it would shift to the offensive, bursting forth like pent-up water overflowing its banks. Note the image of speeding, inexorable destructive force leading to the decisive destruction of the enemy (Wang Shouren MCZY: 166-67, Zeng Xian MCZY: 437). Still others cited the phrase "attack where the enemy is unprepared, appear where he does not anticipate it" to describe the objective of offensive operations (Wang Qiong MCZY: 243).33 In this instance, Sun Zi was used in support of offensive operations against the Mongols. Some memorialists cited Sun Zi only in reference to his warnings about the dangers of extended campaigns against the Mongols (these exhaust the people and the army). The text was used in one of these instances to support arguments for a limited offensive (Zhu Jian HMJSWBa 4:259). In addition, in the memorials there were a number of citations from Sun Zi on the critical role of the general (and by extension "Recall that Sun Zi finished a passage on twelve nonviolent strategems with the phrase, "attack where the enemy is unprepared, appear where he does not anticipate it." While many students of Chinese strategic thought have interpreted the passage as evidence of Sun Zi's stress on nonviolent political strategem, these twelve strategems are preludes to the application of violence. Wang Qiong's interpretation of the last part of the passage seems to fit this latter interpretation more closely. See also Zhu Jian HMJSWBa 4:257.

T H E M I N G S E C U R I T Y PROBLEMATIQUE

2 1 5

military force) in providing security for the state (Liu Daxia HMJSWBa 6:41821). The virtual absence of nonviolent political strategem as a causal concept in the Ming memorials is consistent with this implicit rejection of the notion of not fighting. In the end, then, Ming policymakers placed far more utility in military power as the cause of security than in Confucian-Mencian notions of internal rectification, virtue, or benevolence or idealized concepts of "not fighting and subduing the enemy." Ultimately, the cultivation, preservation, and application of military power was tied to the problem of credibility. By accepting the three basic assumptions of the parabellum paradigm, Ming strategic choice was affected by the real or imagined consequences of appearing weak in the face of the insatiable Mongols. As one official put it in a memorial opposing trade and tribute relations with the Mongols in the 1550s, if China displayed weakness {shi ruo 75 if) by acceding to formal trade relations, it would lose its awesomeness and would appear fearful of the Mongols. This would only encourage their arrogance, and they would then slight or disregard the Ming. If they were to disregard the Ming, they would then begin to "insult" China (i.e., invade, raid, and plunder). This would only prolong the time it would take in the future for the Ming to suppress the Mongols and the chaos they wrought (Chen Shiming HMJSWBb 3:2398). Even the proponents of accommodation in the latter half of the sixteenth century like Weng Wanda and Wang Chonggu acknowledged there was a risk the Mongols would view this strategy as an indication of Ming strategic weakness. Thus they argued for attaching stern, though ultimately hollow, conditions to peace with the enemy. For the proponents, accommodation was a strategy that would buy time until Ming military capabilities could be strengthened. Ultimately, accommodation was an input into an improved capacity to wield force. At a minimum, the fear that weakness would undermine the credibility of future politico-military actions meant that accommodationist strategies were least preferred to violent ones, whether defensive or offensive. At a maximum, Ming parabellum thought dictated that offensive strategies were, in principle, the most preferred, since anything less was a signal of a weakness that would only encourage Mongol aggressiveness. The mediating variable in this preference ranking was, of course, the concept of absolute flexibility; by holding to this notion, Ming strategists recognized that an offensive strategy was not likely to succeed under circumstances of Ming weakness. Once this judgment was made, the next best strategy was a more purely defensive one, tending towards static, linear defense. Only under dire strategic circumstances, conditions of last resort, was accommodation a viable choice.

Chapter Seven CHINESE STRATEGIC CULTURE AND MING GRAND STRATEGIC CHOICE

D

OES A PARABELLUM strategic culture affect Ming policy towards the Mongols? The last chapter indicated the pervasive influence of the parabellum central paradigm on strategic thought in the Ming period, and particularly on Ming debates about security along the northern borders. We must now investigate whether this influence translated into shared grand strategic preferences for dealing with the enemy, and whether these preferences appeared in the grand strategic choices made by the Ming. The short answer is a qualified yes. One finds, as a capability-contingent parabellum model suggests, that Ming decision makers preferred, in principle, more offensive uses of force (including both external extermination campaigns and active defense measures) to static defense and accommodation.1 When the latter two were advocated—as opposed to being preferred in principle—it was generally on the grounds that temporally bounded strategic conditions (i.e., the present balance of capabilities) undermined the efficacy of offensive strategies. In other words, static defense and accommodation were strategies of last resort. Certainly across the memorials there was no consistent rejection of the efficacy of violence in the achievement of security. This argument is very different from preferring defense or accommodation on the basis of a Confucian-Mencian strategic culture.2 One also finds evidence that Ming strategic choice reflected this contingent parabellum calculus. In other words, Ming China tended to adopt more coercive, offensive strategies towards the Mongols when it was more capable of doing so. Ming decision makers resorted to defensive or accommodationist strategies when their ability to mobilize resources for offense was more constrained. 1 Active defense refers to the offensive uses of force against Mongols raiding within Ming territory or within a short distance of border fortifications. It is clear from the memorials that this was viewed as a type of strategic response to the Mongol threat distinct from measures such as "strengthening walls and clearing the fields" and defending against Mongol attacks from static defensive sites. 2 Given the operational hegemony ofthe parabellum paradigm, advocates of static defense and accommodation might use the language of contingent parabellum to justify preferences based on the Confucian-Mencian strategic culture—the reverse of the hypotheses in chapter 6 about the presence of Confucian-Mencian language in the bingshu. This argument does not work, however, because one generally does not find any Confucian-Mencian cause-effect relationships or languages in the memorials which advocate static defense and/or accommodation.

MING GRAND STRATEGIC CHOICE

2 17

PATTERNS IN M I N G G R A N D STRATEGIC PREFERENCES

The central paradigm in the parabellum strategic culture leads logically to a preference for offensive strategies over defensive and accommodationist ones. In the Seven Military Classics this preference ranking is blurred somewhat at the most coercive end because not all the texts argue for extended, expansionist offensive campaigns aimed at the political extermination of the enemy or the annexation of its territory. For the most part, however, the texts did advocate the military destruction of the enemy through offensive uses of force within enemy territory. This was an important precondition for the achievement of state security, and itfloweddirectly from the paradigm's conceptions of conflict, the nature of the enemy, and the efficacy of violence. We should expect, then, that since many Ming decision makers shared the assumptions of the parabellum paradigm, a similar preference for offensive strategies ought to show up in memorials on border security. At first glance this hypothesis seems easily disproved by the evidence that many memorialists advocated static defensive and accommodationist strategies. Yet most of these also accepted the central assumptions of the parabellum paradigm. Either this paradigm was less pervasive than my analysis suggests or there was a cognitive disjuncture between the logic of their arguments and the positions they adopted. As I argue below, however, the many advocates of static defense or accommodation did not reveal their grand strategic preferences but only their calculus of what was strategically possible given their assessment of Ming capabilities. They tended to share with the advocates of offensive strategies a view that, in the long run, less coercive strategies were less likely to improve security along the northern border. What follows is a more detailed examination of the arguments of advocates of each strategy.

Offense The advocates of offensive or more actively coercive strategies were clear about the strategic efficacy of this option. Wang Shu wrote that a policy of going on the attack after careful preparation should be considered the "supreme plan of the state." If the Ming did not send out its forces offensively at an early date, then the enemy's forces would not be forced to retreat any time soon (Wang Shu MCZY: 74-75). Yang Yiqing noted in a memorial that the policy of preemptively attacking the Mongols in the He Tao region if they neared the border was the quickest plan for "pacifying the border and controlling the barbarians" (Yang Yiqing HMJSWBb 2:1104). Elsewhere he stated plainly that the "best policy" was to carry out "heaven's punishment," pushing the Mongols out of the He Tao and establishing the Yellow River loop as the new line of defense (Yang Yiqing MCZY: 198). Not surprisingly, Zeng Xian argued

218

CHAPTER SEVEN

decades later that his plan for an offensive into the He Tao was a plan for "long-term order security" (Zeng Xian MCZY: 429).3 Others implied as much about the efficacy of the offense in their recounting of the success of the first emperors of the dynasty. Han Wen argued, for instance, that the first emperors had concentrated their military strength, accumulated sufficient resources and supplies, trained and encouraged their soldiers, and attacked and killed the barbarians according to opportunity, thus pacifying the border (Han Wen HMJSWBa 6:618). Ma Wensheng made a similarly revealing cause-effect statement in a couple of memorials: he argued that the Yong Le emperor had from time to time sent forces well beyond the borders to attack and punish the Mongols. As a result the Mongols retreated well north, had "fearfully submitted" and did not dare roam south into China. Only after decision makers in the Zheng Tong period neglected this strategy did the Mongol threat reappear (Ma Wensheng HMJSWBa 5:554, 575-76). The purpose of offensive campaigns, he stressed, should focus solely on "attacking and exterminating the barbarian bandits" (ibid., 540-41). Li Dongyang drew a similar causal path between the Hong Wu's offensives against the Mongols and their retreat north (Li Dongyang tfMJSWBa 5:77-78). Whether the memorialists openly declared offensive strategies to be the most efficacious way of dealing with the Mongols or implied this in the use of historical analogies, their conclusions were consistent with the assumptions of the central parabellum paradigm: namely, given the inherently aggressive disposition of the enemy and the efficacy of violence, the best solution was to go on the offensive to eliminate external threats. In some instances, however, short-run strategic conditions made this oneshot elimination of the enemy impossible. Under these circumstances a number of memorialists supported active defensive strategies aimed at wearing down the Mongols or inflicting sufficiently high losses to deter future raiding until Ming forces were capable of launching more extended operations. Yu Qian, for instance, accepted that in the wake of the Tu Mu defeat the best the Ming could achieve militarily, given the inadequacies of supplies primarily, was to attack Mongol forces near or within Ming borders with mobile forces while eschewing extended pursuits. The strategic goal was not simply to beat back or deter the Mongols but to "attack and exterminate" (jiao mie MM) "destroy" (po © ) and "kill" (jiao sha f l j ^ ) them (Yu Qian HMJSWBa 4:120, 131, 133, 141-43, Yu Qian MCZY: 24-25). Presumably, then, if the Ming 3

See also Wang Qiong's advocacy of a preemptive strike against the Uriyanghkha Mongols northeast of Beijing in 1516. While these barbarians did not at that time constitute a grave threat to the Ming, he argued, there was some possibility that they might link up through intermarriage with the northern Mongols and thus quickly become a potential threat very close to the capital. He argued that the Ming should mobilize a large force to "attack and kill them, and sweep them out of their lair" (Wang Qiong HMJSWBb 2:987).

MING GRAND STRATEGIC CHOICE

2 19

capacity to act offensively improved, whether because of improvements in supply, weaponry, the quality of command, or a perceived weakening in the Mongol's capacity to threaten the Ming, then they should go on the offensive. In the mid 1450s when internecine struggles broke out between Mongol leaders Esen and Toghto-bukha (formally the Mongol Khan), Yu argued that this attempt by each side to "swallow" the other was a "heaven-sent opportunity for revenge" that should not be lost. Thus he recommended a limited offensive using several thousand cavalry against the Mongols along the Xuan Fu and Da Tong borders (Yu Qian HMJSWBa 4:163-65, Yu Qian MCZY: 49).4 Wang Shu's arguments in favor of the offense were also rooted in capabilitiesbased reasoning: once the Ming had adequate supplies of horses, food, and fodder, and funds for rewards and perks for the soldiers, it should set a time to send its forces out on the offense, move according to opportunity, and destroy the enemy (Wang Shu MCZY: 75). Wang Shouren suggested in 1500 that to ensure adequate supplies for offensive campaigns, the Ming should improve the tun tian or military farm system, thus reducing the cost and burden of shipping supplies from the Chinese interior to the border areas. This was, in his view, a long-term plan that would allow for more effective offensive use of force (Wang Shouren MCZY: 165). Other advocates of active defense measures argued that Ming forces were too weak and dispersed to pursue Mongol forces in areas very distant from their bases in China. But, as Wu Jie wrote in the early 1450s, the Ming did have sufficient numbers and capabilities to harass Mongol camps, ambush retreating forces, and send out smaller mobile forces to attack Mongol bases beyond the border when the Mongols were themselves raiding inside China (Wu Jie HMJSWBa 3:742-50). Zhu Jian, writing around the same time, concurred that Ming forces were not capable of launching a retaliatory offensive expedition in the wake of the Tu Mu defeat; the problems of coordinating large forces over long distances would expose the Ming to Mongol forces that had just demonstrated their ability to operate over such distances. The Ming had "lost its advantage"; that is, it faced disadvantageous conditions. But it was capable, he argued, of sending out a limited offensive force of cavalry to pursue and attack Mongol forces "according to opportunity" (xiangji zhui 4 At that time Ye Sheng argued along similar lines. In one memorial he advocated that the Ming attack the Mongols' weak points, surprise attack their rear guard, and seek decisive victory by pushing the Mongols back beyond the desert. Significantly, he called this a rare opportunity to defeat the Mongols, implying that given advantageous strategic conditions an offensive policy was preferred. See Ye Sheng HMJSWBa 5:235-39, Ye Sheng MCZY: 45-46. Waldron argues that by the 1470s Ye was an advocate of wall building and static defensive measures largely on the basis of extremely pessimistic assessments of Ming capabilities (Waldron 1990: 101—2). This change in Ye's advice is consistent with a capabilities-based, contingQntparabellum decision rule.

220

CHAPTER SEVEN

jiao #!I|§§JiiJjl!l). He criticized static defensive measures for allowing the Mongols to pillage as they pleased, or slice off territory and thus cut off lines of supply and communication between Ming garrisons along the frontier (Zhu Jian HMJSWBa 4:259-64). This explicit critique of static defense was common among advocates of offensive or active defensive strategies. Wang Yue, the commander of a successful expedition against a Mongol camp in the He Tao in 1473, argued that a policy of static defense along the border required tremendous supply shipments that were easy and attractive targets for the Mongols. Moreover this type of strategy also forced the Ming to disperse forces over a long linear defense line. Since it was hard to predict where the Mongols would attack and where they were headed in the hinterland, the Ming had to second-guess Mongol intentions. This led to a progressive dispersal and thinning of Chinese military power (Wang Yue HMJSWBa 5:714-15). Ma Wensheng was critical of commanders and soldiers in the Da Tong area for retreating within walled fortifications and not actively engaging the enemy outside of these defenses. As a consequence, he argued, the Mongols were able to plunder freely (Ma Wensheng HMJSWBa 5:528). Wang Qiong argued in a memorial on the security of the Beijing area that relying only on defensive forces—forces that specialized in "strengthening walls and consolidating defense"—was dangerous, since if these defenses were lost, one was left with no ability to fight the enemy (Wang Qiong HMJSWBb 2:989). In another memorial in 1516 on defense in the northwest he criticized military officials there for relying primarily on uncoordinated static defense measures. The result, as Ma had argued earlier, was that the Mongols could freely penetrate the Chinese border and retreat with large amounts of booty (ibid., 993-94). Zeng Xian concluded that static defense was too costly; construction or repair of walls and forts took too long and did not meet the immediate threat from the Mongols in the He Tao. Moreover, it left the strategic initiative with the enemy (Zeng Xian MCZY: 428). Active defense, or an offensive defense strategy, then, was a strategy of default. Since, as Wang Yue argued in another memorial on security south of the He Tao, the Ming did not have sufficient strength or resources either to prevent the Mongols from invading or for connecting several hundred li of frontier territory with walls and towers, or for launching several years of extended large-scale external campaigns to wipe the Mongols out, then the least bad policy was to mobilize border forces to attack and harass the Mongols after they came to raid. "Although this [policy] cannot cause [the Mongols] not to invade, it is sufficient to cause [the Mongols] not to dare penetrate deeply" (Wang Yue HMJSWBa 5:705-10). Yang Yiqing argued along similar lines. While he considered a limited offensive against the Mongols in the He Tao to be a superior policy he also attached two preconditions to the strategy: that Mongol camps be relatively close to the Chinese border (i.e., within fifty to one hundred kilometers) so that Ming supplies and reinforcements were less

MING GRAND STRATEGIC CHOICE

22 1

vulnerable to interdiction, and that Mongol military capabilities be relatively weak compared to Ming forces. In other words, an offensive strategy was preferred but implementing it depended on an optimal balance of capabilities and opportunity. As he noted just before attaching these two conditions to an offensive strategy, the Ming should "act according to opportunity" (xiangji er dong ffiUMlfr) (Yang Yiqing HMJSWBb 2:1102-4). In short, in the context of an overall preference for offensive strategies, many memorialists suggested that the Ming should pursue a mix of strategies depending on the quantitative and qualitative nature of the threat. Bai Kui, the minister of war from 1467 to 1475, argued in one memorial that in the face of larger or superior Mongol forces the Ming should "consolidate defenses" and ride out the attack until the Mongols retreated. But when weaker or inferior Mongol forces appear the Ming should concentrate its military power and "exterminate" them. The key rule was "to attack or defend depending on opportunity" (you kou ze xiang ji zhan shou W^IWffilillPGTP) (Bai Kui HMJSWBa 4:512). Yu Zijun, identified in the secondary literature as the quintessential advocate of static defense and wall building in the latter half of the fifteenth century, proposed in one memorial an offensive operation against a Mongol tribe ensconced in the He Tao region. From this base the enemy had been able to spy on the Ming and gauge its strengths and weaknesses. The Ming should thus mobilize its forces and "according to opportunity attack and kill" (xiangjijiao sha %W$$M\%%)tne Mongols and force them to retreat. The decisive factor in whether or not such an action made strategic sense was, again, Ming military capacity. "When food and grass is sufficient then we should plan a great action to search and destroy [the enemy]. When food and grass is insufficient, then we should order [Ming forces] to defend according to the old [strategy]" (Yu Zijun HMJSWBa 5:368, 385-88). Here is a clear case of contingent thinking within the context of parabellum assumptions. Even for Yu Zijun, the architect of the wall-building strategy in the Ming, there was no a priori embrace of static defense strategies. Rather, his argument suggests that, given appropriate strategic conditions and opportunities, it was preferable to exterminate, or badly maul, and push back Mongol forces rather than merely manage the threat through static defense. Consistent with their parabellum assumptions, the advocates of offensive strategies tended to downplay the long-term military capabilities of the Mongols without undermining their assessment of the threat. A number of memorialists from disparate time periods cited in different ways the well-known phrase that the barbarians were not even as numerous as the population in one Chinese district or county.5 Li Xian, a senior grand secretary at the time, concluded from this, in a memorial submitted in 1466, that the only reason the masses of 5

The phrase is taken from a memorial by the Han dynasty official Jia Yi in which he criticized accommodationist policies towards the Xiongnu (Waldron 1990: 41).

222

CHAPTER SEVEN

Chinese had not defeated the Mongols was poor strategy (Li Xian HMJSWBa 4:319). Wang Ao, also citing this phrase, argued that China had sufficient numbers and capabilities to overwhelm the Mongols if these forces were deployed properly to attack the enemy's undisciplined hordes (Wang Ao HMJSWBb 2:1149). Li Dongyang cited this phrase to argue that since the Han dynasty there had never been a time when the northern barbarians had been weaker or more in decline than the present. To mobilize Ming power and drive the Mongols away was thus not really that difficult in his estimation (Li Dongyang HMJSWBa 5:77-78). Zeng Xian also used the phrase to conclude that the Mongols presented ultimately an exterminable threat. Although the Mongols in the He Tao were a real and present danger, they were in the end subject to Ming domination. The problem was Ming passivity, not Mongol strength. Given this fundamental advantage for China, how was it, he asked rhetorically, that the military power of the state could not defend against this threat? (Zeng Xian MCZY: 427). There were throughout the Ming period, then, decision makers who plainly considered offensive strategies to be the most preferred means for dealing with the security problem along the northern border, and who on the basis of their assessment of the Ming capabilities advocated acting on the basis of this preference. This is not surprising given the parabellum assumptions behind this preference and these policy recommendations. But this does not really tell us anything about whether this preference for offensive strategies cut across policy positions, as it should if a contingent parabellum strategic culture pervaded Ming decision making. A more interesting question is whether the advocates of different policy strategies also accepted this preference in principle, and whether their retreat from this in practice was linked to more pessimistic assessments of Ming capabilities rather than, say, a Confucian-Mencian-based preference for less coercive strategies.

Static Defense Arguments in favor of static defense measures tended to be more complex than those supporting offensive strategies. There was to be sure an apparent orthodoxy consistent with the Confucian-Mencian tradition that stressed defensive uses of force and openly eschewed extended offensive operations beyond the borders. This began with the Hong Wu emperor's instructions, "As for the way of defending the border, one must certainly display martial awesomeness, and defend with firmness. If [the enemy] approaches then resist it, if it retreats, then do not pursue. This is the best policy" (cited in Yang 1988: 9). In the Yong Le's compilation of advice on statecraft there were numerous citations against the offensive use of force. As one section put it, When the enemy seeks peaceful relations then treat them well, but don't ally with them. When the enemy invades then strengthen defensive preparations, but don't

MING GRAND STRATEGIC CHOICE

223

stress retaliation. .. . When they are weak they are fearful and submit. When they are strong they invade and rebel. Thus a ruler with the dao awaits them with military preparations, and resists them with constancy. He stresses not relaxing defenses along the border walls, and when the enemy is strong and violent, and plunders and raids, he does not send out his forces to engage in long-distant campaigns. (Ming Chengzu 1410: 524-25, 534) Despite the vast disjuncture between these words and both emperors' record of offensive campaigning against the Mongols, some memorialists appeared to embrace this strategic orthodoxy and argued in an a priori fashion in favor of static defense. Kui Yanliang advised the Hong Wu emperor in the 1380s that in managing the Rong and Di, "defensive preparations come first. Attacking and punishing is second to this. To provoke border quarrels in pursuit of small advantages is the worst [policy]." Interestingly, Kui's memorial was one of the more orthodox Confucian-Mencian ones in tone and language, far more than many later ones. "The Man and the Yi present tribute to the court. If among them are some who don't obey, then one should cultivate civility and virtue in order to attract them and send envoys to instruct them. They will fear our awesomeness and cherish our virtue and none will not submit. Why, then, exhaust military power in distant places?" Yet despite this affirmation of the idealized Confucian-Mencian approach to security, there was a degree of ambiguity in Kui's position. He appeared to differentiate between methods for dealing with the Man and Yi and those for managing the northern Di.6 These he implied were far more rapacious and less amenable to the magnanimity of the Ming empire. Against them he appeared to recommend a two-stage strategy: within the context of a static defense of the borders the Ming should still be prepared to go on the offensive if provoked. "Your sage has considered that we should chose a general, train military forces, divide them into defensive commanderies, and attentively guard [the borders]. As soon as they [the Mongols] raise a pretext for quarrelling, then in one fell swoop we should reduce them to submission" (Kui Yanliang HMJSWBa 3: 192-93).7 Similarly Fan Ji, writing in the late 1420s, also appeared at first glance to prefer static defensive strategies. He suggested that the way to protect Beijing and inner China was to defend strategic points in the outer Da Tong and Xuan Fu areas by expanding the tun tian system as well as walls and signal beacons along the border. He warned against the pursuit of "small advantages" by going afar for other strategic prizes; in other words, it was better to eschew offensive expeditions beyond these boundaries. Instead the strategy should be one of 6 The northern Di (bei di ;|k$0 was a more specific term for the nomadic peoples in the north and northwest, including the Mongols. 7 The term he used implies to destroy and thereby pacify (yiju er tang ping zhi —'

224

CHAPTER SEVEN

"strengthening walls and clearing the fields" (Jian hi qingye ISM'/raS?) to prevent the enemy from obtaining anything of much material or strategic value—a conservative strategy to minimize maximum losses. Like Kui, however, Fan's complete position was somewhat more complex: static defense appeared to be a prelude to a more active defense or offensively defensive posture. After Ming forces had ridden out a Mongol offensive they would strike back, inflict defeat, and only then abandon pursuit (Fan Ji MCZY: 23). The implication of this cause-effect relationship was that static defense created strategic conditions, namely the exhaustion of the enemy, which allowed for the application of offensive force, albeit limited. The offensive use of force, not static defense measures in and of themselves, led to the defeat of the enemy. In general, proponents of static defense started from the assumption that the Ming was strategically incapable of anything else. Guo Deng, a commander along the northern border in the 1450s, argued, for instance, that since China was militarily weak in comparison to the Mongols it could not hope to achieve decisive victory. Given this general assessment of the balance of capabilities he recommended that Ming forces step up efforts to control strategic high points and choke points, improve the early-warning signal system, and build and repair defensive walls. Ming forces should avoid any efforts by the Mongols to engage in direct battle. Instead, the Ming should flaunt its military capabilities; at any particular place along the border it should raise lots of banners and flags and fire off signal cannons to intimidate and unsettle attacking Mongols. After a while, he predicted, Mongol forces would retreat, their morale spent. Then the Ming should send out elite cavalry to surprise and harass the retreating Mongols wherever possible. Since the Mongols prized material wealth, if they left not only bloodied but empty handed they would be reluctant to raid in the future. He termed this strategy one of "not fighting and subduing the enemy," despite the fact that force would be used offensively in a limited fashion (Guo Deng HMJSWBa 5:195-196).8 Liu Dingzhi, a grand secretary during the early Cheng Hua reign period (1465-1488), offered a rather alarmist view of the Mongol threat. "I have observed," he wrote, "that since ancient times the calamity brought by the Yi and Di [barbarians] has never been greater than today." The problem was that the Ming had been weakening militarily for quite some time, and Ming forces were only capable of static defense, or "closing up encampments and strengthening walls" (biyingjian hi |?0ff M i l ) . The Ming military was not capable of sending out forces offensively or even pursuing more active defense strategy (Liu Dingzhi HMJSWBa 4:688). Ye Sheng lamented in one memorial from the early 1470s that although the 8

A similar argument was made by the well-known advocate of wall building Yu Zijun in the 1480s (Yu Zijun HMJSWBa 5:359-60).

MING GRAND STRATEGIC CHOICE

225

northern barbarian threat was an age-old one and the Han had long wished to "exterminate their kind," this was not yet possible. This left China with only one long-term security strategy: "the way of defense" based on "strengthening walls and clearing the fields" (Ye Sheng HMJSWBa 5:317-18). In an earlier memorial he remarked that proposals to send out cavalry to search actively for and destroy Mongol forces seemed to be worth considering, but that in reality, given that Ming troops were too weak and logistics problems too severe at the time, such plans were not practical (ibid., 289). Others viewed the staged relationship between static defense and more offensive measures in a longer-term perspective. Wang Shouren, for example, argued in a memorial submitted in 1500 that Chinese border forces were weak and Mongol power was strong. Under these conditions, he concluded, it made no sense to seek active warfare with the Mongols. This only played to Mongol strengths in mobile warfare. The only sensible policy, then, was static defense, improvements in the signal system and intelligence gathering, stepped-up training of soldiers, and better command and coordination within the military. Over time these measures would lead to the "accumulation of power and the cultivation of sharpness [efficacy]" in the military. Like the damming up of water, it will eventually reach a bursting point such that when it is released its speed and power will be irresistible, leading to the decisive destruction of the enemy.9 In other words, once the balance of capabilities (including intangibles such as morale and command ability) shifted in their favor, the Ming forces should then go on the offensive, attacking the enemy where it was unprepared and unaware. This he termed "victory through completeness" or victory through absolute superiority (Wang Shouren MCZY: 166-67).10 Yang Yiqing argued along similar lines. He saw static defense measures such as wall building as a medium- to long-run means of deterring or resisting Mongol forces until such time that the Ming could accumulate sufficient food, finances, and military supplies to launch major offensives against the Mongols in the He Tao region. He admitted that the more passive first stage of this strategy was not the best plan, but argued that an offensive into the He Tao could not be carried out until the Ming had sufficient capabilities (Yang Yiqing MCZY: 198). 9 He also uses Sun Zi's metaphor about boulders rolling down a mountain to describe the inexorableness of applied violence. 10 The phrase is "sheng yu wan quan' (JK£^P8Ji£:)- It seems to be derived from Sun Zi's concept of quan sheng i^kW)- If this is so, then Wang's interpretation is very different from contemporary interpretations of Sun Zi's phrase. Recall from chapter 4 that many interpret the phrase to mean achieving victory over the enemy while minimizing damage to the adversary-— or "taking the enemy state intact"—thus linking quan sheng conceptually to the notion of "not fighting but subduing the enemy." Wang implies that quan sheng means complete preparation for warfare and the attainment of such overwhelming superiority that when force is applied the destruction of the enemy is complete, with minimal cost to oneself. Wang's interpretation, I believe, is more consistent with the parabellum core in Sun Zi's thought.

226

CHAPTER SEVEN

Yang's arguments highlighted a dilemma facing those who memorialized on border questions. On the one hand, many agreed that the more or less permanent presence of the Mongols in the He Tao region by the late fifteenth century constituted a grave weakness in Ming defenses. To ring the southern and eastern portions of the area with walls and fortifications would require enormous one-time outlays in money and labor, and the long-term burden of supplying these forces was not only a drain on state resources but required additional labor and tax levies on the common people. Ming military forces in these areas, complained some memorialists, were entirely too passive, and hid behind existing walls and fortifications to protect themselves. They were unable or unwilling to interdict Mongol vanguard forces, and they had no will to pursue retreating Mongols and to sweep them from their "lairs." The few offensive expeditions sent out to search and destroy Mongol forces were unsuccessful, with corrupt commanders falsely claiming victories in order to acquire rewards and promotions. One strategy, which Ni Qiu called "not a bad policy" (fei bu shan ^ ^ p | l | ) , was to recover the Mongol staging areas, particularly in the He Tao. This would push the Ming defensive line further north and west, taking some pressure off the Ningxia, Yan Sui, Da Tong, and Xuan Fu commanderies, further improving Beijing's security, and leading to the long-term pacification of the border. On the other hand, these memorialists also recognized that an offensive policy faced formidable logistical problems. Defense of the He Tao, for instance, would probably require the deployment of forward garrisons on the northern side of the Yellow River loop in isolated, desert-like conditions where food and military supplies would be exposed to ambush and plundering. In the process of pushing the Mongols out of the area and establishing these forward lines, Ming forces would be vulnerable to harassment and interdiction. Reinforcements might easily be cut off, and there would be no walls or fortifications behind which Ming forces could retreat. Moreover, the Mongols might avoid decisive battle. The net result of unsuccessful extended offensive campaigns might then be, as Sun Zi had warned twothousand years earlier, the exhaustion of the state and the impoverishment of the people (see Ni Qiu HMJSWBa 5:296-307, Qiu Jun HMJSWBa 6:110-14, Wei 1541: 101). In short, the factors that these memorialists considered when debating the merits of a strategically preferred policy towards the Mongols were resource-based or capabilitiesbased ones and had nothing much to do with any a priori Confucian-Mencian preference for nonoffensive security strategies. This, then, was the primary pattern of reasoning behind arguments in favor of static defense strategies. For the most part, static defense was viewed as a means of changing strategic conditions, altering the balance of capabilities, and creating opportunities for a more active and efficacious use of military power. The memorials that advocated static defensive measures tended to embrace a starkly pessimistic view of the Ming ability to go on the offensive

MING GRAND STRATEGIC CHOICE

227

against the Mongols. The pessimism was based on an assessment of the strength and severity of the Mongol threat, or on the short-run costs of offensive strategies, or both. Like the advocates of offensive strategies, proponents of static defense accepted the basic assumptions of the parabellum paradigm, hence they explicitly or implicitly shared the strategic-preference ranking of the latter. It was for capabilities-based reasons that they concluded the preferred means for dealing with the Mongol threat could not be carried out. Had they argued for static defense from a different set of starting assumptions, such as those of the Confucian-Mencian paradigm, or had they rejected offensive strategies and embraced static defense for reasons other than a pessimistic assessment of Ming capabilities, one might conclude that the parabellum strategic culture was not shared by proponents of static defense. And since static defense measures tended to dominate Ming strategy towards the Mongols from the Tu Mu defeat on, one would then have to conclude that the parabellum paradigm had little influence on Ming behavior. But the evidence here suggests otherwise. Of course a still harder test is whether the proponents of accommodation also shared the parabellum assumptions but supported nonviolent, nonmilitary means on the basis of even more fundamentally pessimistic assessments of Ming capabilities; that is, whether, in the view of a major study of border policy written in the mid-sixteenth century, the Ming pursued accommodationist policies only in periods of extreme weakness (Wei 1541: 101).

Accommodation Accommodationist actions were never entirely absent from Ming strategy towards the Mongols. Mongol tribute and diplomatic delegations to Beijing were quite regular through to the 1460s, despite the interruptions from Ming punitive expeditions in the first half of the century and growing Mongol power after the Tu Mu defeat in 1449 (Lai and Li 1954: 53, Serruys 1967: 562-81). Trade fairs along the border were less frequent and often closed quickly by order of the Ming court due to alleged Mongol improprieties such as cheating or spying. But these sorts of interactions did not define Ming strategy towards the Mongols, and for the most part Ming strategists did not view accommodation as an effective alternative to the use of military power or as a long-term solution to the security problems along the northern border. As I noted in chapter 6, many Ming strategists considered accommodation a threat to the credibility of Ming military awesomeness, and an invitation to the Mongols to increase military pressure on China. A number of memorialists also argued that through tribute and trade relations the Mongols were able to acquire materials and technologies that could improve their military capabilities. There was also a suspicion—born from worst-case assessments of the threat—that often the Mongols formed their proposals for tribute or trade relations in such a way that the Ming would have to reject them, thus creating a pretext for more

228

CHAPTER SEVEN

raiding.11 If the proposals were not rejected, then the Mongols would exploit Ming magnanimity and use the opportunity to spy on the Ming military. In general, they argued, more Ming-Mongol interaction would not alleviate conflict in the long run, but instead would increase the number of points of conflict and contention between the two sides (see Lai and Li 1954: 24, Ma Wensheng HMJSWBa 5: 441-44, 519). Nonetheless, a number of memorials did argue for accommodation with the Mongols. However, the cause-effect statements in these memorials indicated that their authors bought into the basic assumptions of the parabellum paradigm: accommodation was viewed as a last resort, as a stopgap measure that would buy the Ming time to improve military capabilities. Accommodation was not a priori preferred to the use of military force, either offensive or defensive. Perhaps the most complete arguments offered in favor of accommodation came from Weng Wanda in the mid-sixteenth century. Interestingly, Weng endorsed the view held by proponents of more offensive strategies. On the one hand, the Mongols were dogs and sheep, and given their wild treacherous nature presented a zero-sum threat (Weng Wanda HMJSWBb 3:2349-50). On the other, in the long run the Mongols presented a threat that could be managed and resolved. Like Zeng Xian, he, too, observed that the Mongols did not equal in population a large district in China. But this long-run optimism was heavily qualified by a short-run pessimism about the Ming ability to deal with this numerically inferior threat. The Mongols were able to concentrate greater numbers of fighters in a shorter time than the Ming. They were warlike by nature, skilled in cavalry and mobile warfare, and were able to cohere and attack better than Chinese forces. The cycle of Mongol power, he argued, had not yet declined, while Ming power was insufficient (ibid., 2356). The present balance of strength was unlike that in the Hong Wu reign when the Ming regularly swept out the Mongols; the Mongols were now stronger and the Ming weaker. It would be like pitting weakness against strength. This was the same as having no strategy at all, he concluded (ibid., 2363). As he summarized in his critique of Zeng Xian's plan for attacking the He Tao, the Ming faced a number of problems: Mongol strength was impressive (he estimated three to four hundred thousand fighters);12 Ming forces were unfamiliar with the terrain outside the borders and had poor intelligence about routes of advance and retreat; supply lines would be overburdened and vul11

Ma Wensheng, for instance, noted that the Mongols used arrogant language in their proposals and that they requested permission to send massive delegations (often three thousand strong) that could easily turn into raiding parties once inside the Chinese border (Ma Wensheng HMJSWBa 6:519). 12 This was an upward estimation of one to two hundred thousand from that in a memorial submitted just prior to his critique of Zeng Xian's plan (Weng Wanda HMJSWBb 3:2356).

MING GRAND STRATEGIC CHOICE

229

nerable to interdiction; the Mongols might simply wait for the Ming to attack and avoid decisive battle, thus preserving their strength as the host army while Ming forces exhausted themselves as the guest. Since the Ming had not been able to defeat the Mongols in battle so far, it would be much more difficult to exterminate the enemy in their base. Finally, the costs of holding the Yellow River as the primary defense line were exorbitant and would require a permanent garrison of three hundred thousand soldiers in addition to new forts, towers, and walls (ibid., 2364-65). Note that Weng's arguments against an offensive strategy implied strongly that relative capabilities determined whether the Ming could pursue different strategies successfully. Thus he agreed in principle with the logic of an offensive strategy, but his strategic assessment led him to conclude conditions were not ripe: "The rationale for recovering [the He Tao] exists, but there is no exploitable opportunity" (ibid., 2363). However, Weng argued, drawing from the notion of constant change and transformation, if the Ming bided their time and improved their border defense and military capabilities, then when a fissure in Mongol power developed it could be fully exploited (ibid., 2366). As he proposed in a memorial a couple of years later during a debate over whether to accede to Mongol demands for trade and tribute relations, accommodation would buy this time. This strategy was, under conditions of Ming weakness, the best of a bad set of options. Weng accepted, for instance, that setting up tribute and trade relations after having so vociferously declared the Mongols to be criminals deserving of punishment would expose the hollowness of the Ming strategic position.13 But to reject the Mongols' peace proposals would be to expose the Ming to a Mongol first strike that the military was incapable of resisting. In this situation, somehow the Ming had to establish a modicum of credibility and initiative. Thus they should attach conditions to any peaceful relations, such as the return of captured Chinese and a Mongol pledge not to invade for one year. Once these commitments were made openly, then the Ming would consent to trade and tribute relations (ibid., 2366).14 13

This memorial was in response to the Altan Khan's request for peace relations after he had penetrated Ming defenses and attacked Beijing in 1550. 14 Similar arguments were made twenty years later by Wang Chonggu during a debate over whether to accede once again to the Altan Khan's demands for peace relations (trade and tribute relations established in 1551 had lasted less than one year before they collapsed) (Waldron 1990: 176). He rejected his critics, arguing that, unlike earlier periods when accommodation was not considered a viable option generally, at the present the Ming were weaker and the Mongols stronger than ever before. If the Ming rejected these demands for tribute and trade relations, then the combined threats from the Altan Khan and lesser Mongol tribes, in conjunction with financial constraints, would only accentuate Ming strategic vulnerability (Wang Chonggu MCZY: 516, Pokotilov 1976: 129-31). This time, with the backing of the powerful grand secretary Zhang Juzheng, combined with a revival of inter-Mongol struggles, trade and tribute relations lasted until Zhang's death in 1582 (Chinese Military History Group 1988b 2:544-46, Waldron 1990: 185-87).

230

CHAPTER SEVEN

There are a couple of observations to be made about accommodationist arguments. First, I do not want to leave the impression that throughout the Ming period one found powerful advocates of accommodationist strategies, and that as a strategy it competed equally for attention with offense or static defense. The policy was rarely adopted as the centerpiece Ming security strategy (see Waldron 1990: 171-89). Indeed, the only sustained period of active accommodation with the Mongols was in the 1570s. Proposals for accommodation were accompanied invariably by demands for increased military capabilities and improved defenses. This two-track approach indicated that accommodation was by itself not considered a preferred means of dealing with the Mongols. Accommodation was a contingent means to an end; it was an option that only made sense, according to its proponents, when the Ming was incapable of effectively using defensive or offensive means to defeat the adversary militarily. Moreover, it was rarely justified by appeals to Confucian-Mencian notions of magnanimity. This relates to the second point: Proponents of accommodation shared the basic assumptions of the parabellum paradigm with the advocates of offensive and static defensive strategies. They accepted that the conflict with the Mongols was essentially zero sum. They often used the same racialist language to characterize the nature of the conflict. They also accepted that despite this threat the application of violence, in principle, could turn the threat into a manageable one. Where they differed was on the question of the temporary Ming capacity to use force in an efficacious way. In other words, had they shared a rough assessment of Ming capabilities the logic of their arguments would have led to shared policy recommendations. Waldron (1990: 124-25, 178-83) argues that the opponents and proponents of accommodation, particularly in the sixteenth century, tended to proceed from very different assumptions. The former argued in culturally exclusivist, emotional, and xenophobic terms, while the latter argued along more "realistic" strategic, capabilities-based lines. The analysis here suggests that the differences were far less stark. In fact, proponents of all three strategies tended to argue along capabilities lines. All accepted that in principle accommodation was the least-preferred option. Since all sides accepted the assumptions of the parabellum paradigm, the differences in policy positions had to be explained by factors other than by basic moral perceptions. The parabellum paradigm suggests that the key variable is change in the relative capacity to act more offensively or coercively.15 15

This raises the question of whether the subjective or objective balance of capabilities is the crucial variable. The fact that there were contemporaneous debates about the relative strengths and weaknesses of the Ming and Mongols suggests that subjective perceptions were at work, unrelated to "objective" conditions. On the other hand, pivotal debates between proponents of offensive versus accommodationist strategies were rare, the most prominent one being the 1547 debate on the recovery of the He Tao. In other words, the advocacy of either accommodation or

MING GRAND STRATEGIC CHOICE

231

STRATEGIC PREFERENCES AND COERCIVENESS IN M I N G SECURITY POLICY

Can we conclude that Ming decision makers and strategists shared a grand strategic-preference ranking similar to that in the Seven Military Classics? Was there enough congruence in preference rankings to conclude that a dominant parabellum strategic culture underlay Ming thinking about security along the northern border?16 The firmest conclusion is that accommodation ranked last, as indicated in the summary of predominant elements in the security strategy often memorialists (Table 7.1). The picture is less clear when it comes to a preference between offensive and active defensive strategies and static defense. None of the memorials supporting static defense also rejected offensive uses of force, whereas a couple of memorials supporting more offensive uses of force did criticize reliance on static defense. And in most cases the logic of the arguments suggests that the defeat of the enemy, hence the security of the state, required more offensive uses of force, though few before the sixteenth century expressed open support for the expansion or annexation of territory. In most cases, recommendations for static or offensive defense were heavily context dependent. In short, there is some basis for arguing that there was a general preference for somewhat more offensive strategies over purely static defensive ones, but that temporary strategic conditions translated this preference into different short-run policy recommendations. If this were the case, then, we should expect to see a significant degree of sensitivity in Ming strategic behavior to changes in China's capacity to act offensively. Specifically, we should expect strategically offensive behavior by the Ming to vary directly with the Ming capacity to act offensively. In other words, given a long-term zero-sum conceptualization of conflict with the adversary, when the capacity of the enemy to threaten the Ming was low, and the capacity of the Ming to mobilize military resources was high, then the Ming would tend to act in a more offensive, coercive manner. The anecdotal evidence seems consistent with this hypothesis. Take for example Yong Le's offensive expeditions against the Mongols in 1410, 1414, 1422, 1423, and 1424. The first was aimed at crushing the Da Dan Mongols' military power, and the next the Oirat tribe. But the decline of Oirat military prowess left the Da Dan tribe, which had initially allied with the Ming against the Oirat, in a position to take over the territory north of Ming boundaries. The nonaccommodation policies tended not so much to be simultaneous among decision makers as much as sequential across different time periods. This is confirmed to some degree by the aggregate data analysis later in this chapter. 16 As opposed to congruence in policy recommendations. My argument is that different policy recommendations can come from similar parabellum assumptions depending on the relative capacity of the state at time t to implement a maximally coercive strategy.

232

CHAPTER SEVEN

TABLE 7.1

Predominant Elements in Security: Ming Memorials on the Mongol Threat Memorialist

military political expansion destruction destruction

accommodation

static defense

offense

no

yes

yes

yes

no

no

no (a)

yes

yes

yes

yes (b)

no

no

yes (c)

yes

yes

yes

no

no (a)

yes (d)

yes (d)

yes

no (7)

no

Yang Yiqing

no

yes

yes(e)

yes

no(?)

yes(f)

Wang Qiong

no

no(g)

yes

yes

yes (h)

?

Zeng Xian

no

no(g)

yes

yes

yes

yes(f)

Weng Wanda

yes (i)

yes

no(d)

yes©

no (7)

yes(f)

Yang Jisheng

no

no

yes

yes

yes (?)

yes

Wang Chonggu

yes

yes

no(d)

yes (j)

no(?)

no

Fan Ji Yu Qian Wang Shu Wang Shouren

Note: In cases where more than one memorial was examined, the entry refers to combined memorials. (a) in one of three memorials. Seems to be a perfunctory reference to the emperor's awesomeness and virtue as a basis of security; rejects formal peace relations (b) refers to the submission of the enemy (c) limited role; the submission of the enemy and the security of the state mostly depends on the offensive use of force (d) depends on capacity to act offensively in a way that will ensure success (e) limited attack on the He Tao region (f) recovery of the He Tao, pushing the Mongols north of the Yellow River loop (g) static defense alone is negatively related to security (h) attack on Mongol bases and camps linked positively to security (i) depends on capacity to use violent strategies in a way that will be effective (j) in the long run necessary to increase security once military capabilities are sufficient (?) indeterminant

expeditions of 1422-1424 were, in essence, efforts to preempt the growth of Da Dan power. These were not small-scale raids on Mongol bases as were the occasional offensives launched in later periods; the first one involved five hundred thousand troops (Lai and Li 1954: 15).17 But at that point in the dynasty

17 There is no good guess about the combined military strength of the Mongols, and there has to date been no systematic effort in China to compile such figures (personal correspondence with Ren Li, Academy of Military Sciences, Beijing, China, July 1992). Ming strategists did not

MING GRAND STRATEGIC CHOICE

233

the military and financial resources at Yong Le's disposal were considerable. While figures for Ming military capabilities are incomplete and sometimes inconsistent, the entire empire had anywhere between 1.5 and 2.5 million soldiers (Wu 1937: 158, Chen 1984: 2). One estimate put Ming military strength in the garrisons along the northern border at the end of the Yong Le period at a total of about 863,000 (Wu 1937).18 Supplies of food and equipment for campaigns were in greater abundance than later periods. Despite a precipitous drop in the food production of military farms (tun tian) in the Yong Le period from about 23,000,000 shi of grain in 1403 to 5,700,000 shi in 1424, production did not bottom out until a decade after the Yong Le period (Wang 1965: 213).19 Evidently this was sufficient to maintain Ming military forces without having to spend central government funds.20 Nor were Ming military resources diverted towards quelling internal unrest, as occurred in the sixteenth century. In the entire Yong Le period there were only eight cases of major internal collective violence, an average of 0.4 incidents per year. This is significantly less than the average for the entire dynasty (about 2.2 cases per year).21 Given these indicators, then, the Ming capacity to act coercively against the Mongols was, relatively speaking, greater during the Yong Le period than later years. And indeed, one saw in this period a series of large-scale offensive operations against the Mongols. During Yong Le's reign the Ming initiated an average of 0.27 major conflicts (n=6) per year with the Mongols, whereas the average over the length of the dynasty was 0.24 conflicts per year (n=65), a difference that is statistically significant at the .01 level.22 The difference is even have solid figures. One finds as well that they tended not to refer to Mongol numbers in their threat assessments until the sixteenth century. Suffice it to say, in the Yong Le period the Mongol tribes could not separately or in concert muster five hundred thousand soldiers for offensive or defensive operations. The Da Dan, for instance, commanded two hundred and forty thousand fighters as of the late 1380s. After they met defeat at the hands of a major Chinese expedition in 1387, about two hundred thousand surrendered (Pokotilov 1976: 12). In the twenty to thirty years subsequent Da Dan strength could not have exceeded this original number. 18 Wu's figures come from the Da MingHui Dian ( ^ 0 5 # ^ - ) , published in the late sixteenth century, which provided detailed information on the official organization and structure of government in the Ming period. The figures may therefore overestimate somewhat actual military strength, though these are likely to be far more accurate than figures for periods after the Tu Mu defeat when desertions and roll padding led to vast discrepancies between paper and actual strength. 19 One shi equals twenty-seven pounds. Thus production dropped from around 310,810 to 77,027 tons. 20 Until the mid-fifteenth century the Ming military was essentially a self-sufficient organization. Most soldiers and officers belonged to hereditary military families and were required to supply their own weapons, uniforms, and other supplies. On military organization see Wang 1965, Taylor 1969, Huang 1970. 21 1 thank James Tong for supplying me with the data for incidents of collective violence that he compiled for his dissertation, Collective Violence in a Premodern Society (1985). 22 1 should add a word here about Ming initiation. The raw data on the number of external wars fought by the Ming come from the annual chronology of wars in Chinese history compiled

234

CHAPTER SEVEN

even greater if one excludes the Hong Wu reign (1368-1399), in which over one-third of Ming initiations took place (n=22). From 1400 to the end of the dynasty the Ming averaged 0.18 initiations per year (n=43). The Ming's behavior in the Yong Le period, then, is consistent with a parabellum decision calculus where a state will act more offensively if it is capable of doing so. A second period worth examining is the few years after the Tu Mu defeat in 1449. Prior to this major disaster, despite a shift away from massive offensives after the Yong Le emperor's death, the Ming maintained the same level of coerciveness as in the previous period. From 1425 to 1449 the Ming initiated seven major conflicts with the Mongols, an average of 0.28 per year. But the frequency of major Mongol attacks on the Ming increased as well, indicative of the former's growing political and military cohesiveness in the 1430s and 1440s (Pokotilov 1976: 36-46). In the same period the Mongols initiated thirteen major conflicts (0.52 per year on average) as opposed to one major conflict in the entire Yong Le period (excluding smaller-scale border raids). To the extent that this increasingly aggressive Mongol behavior constituted, in the eyes of Ming decision makers, an increase in the level of threat, the relative balance of capabilities was beginning to shift to the disadvantage of the Ming. The Tu Mu defeat, however, marked the beginning of a long-term shift away from offensive strategies. In the period from 1450 to 1457, the seven years in which Yu Qian dominated policy making towards the Mongols, the Ming only initiated one major conflict (0.14 per year on average). This is not surprising given the disadvantageous shift in the Ming capacity to implement an offensive strategy. Ming military capabilities had been severely damaged at Tu Mu, with an army of five hundred thousand essentially destroyed in battle (Mote 1974: 264-65). Consequently Mongol forces were able to penetrate Chinese defenses right up to the gates of Beijing. After the Mongol armies withdrew in 1450 in the face of determined resistance from the remaining Ming forces, Yu Qian was preoccupied with rebuilding China's military capabilities (Waldron 1990: 94-95). There was also a growing problem with internal social stability, by scholars at the Chinese Academy of Military Sciences. As a rough indicator of coerciveness in Ming strategic choice, I decided to use the frequency of conflicts in this list that the Ming initiated. The list provides a brief description of each conflict and indicates whether the Ming attacked the Mongols or vice versa. When the description uses terms such as "attacking" {gong XJO, "striking" (ji H ) , "punitively campaigning against" {zheng flE), etc., to describe Ming behavior or otherwise indicate that the conflict was precipitated most immediately by some positive action by the Ming, then I coded the case as an incident of Ming initiation. I recognize the risks in this simple coding rule. Unfortunately, without reconstructing in detail a universe of Ming-Mongol conflicts, the data do not provide enough information to determine whether Ming initiation was a discrete event or part of a broader sequence that may have been initiated by the Mongols. However, to the extent that variations in the frequency of Ming initiations are consistent with anecdotal evidence about changes in strategies towards the Mongols in a particular period, the validity of this indicator is strengthened.

MING GRAND STRATEGIC CHOICE

235

which diverted material and military resources. While incidents of collective violence remained well below the average for the entire dynastic period (n=6, or 0.85 incidents per year), the level was far higher than that during the Yong Le period. In short, strategic conditions made it increasingly difficult to sustain a more offensive strategy towards the Mongols. Since accommodation had essentially been ruled out by Yu Qian, a more defensive strategy seemed to be the most feasible option.23 Given the relationship between a changing capacity to act offensively and changes in Ming strategy towards the Mongols, the parabellum model would also suggest that if accommodationist strategies were adopted these would have to have come in periods of especially severe Ming weakness (or perceived weakness). Once again the anecdotal evidence seems to provide some confirmation of this proposition. As noted, the 1570s was one of the few periods in Ming history in which accommodation with the Mongols was the predominant security strategy, taking the form of formal peace arrangements, tribute relations, and trade fairs (Pokotilov 1976: 123-37, Waldron 1990: 184-87). The justification for this, as exemplified in Wang Chonggu's memorials, was simply that the Ming had no other choice, given their relative military weakness (see Chen 1984: 43). Wang and his powerful supporters in the emperor's court, the Grand Councillors Gao Gong and Zhang Juzheng, were no doubt quite aware of the atrophy of Ming strategic and financial capabilities. Military forces along the northern border had declined quantitatively throughout the sixteenth century, due in part to desertions because of the severity of garrison conditions and in part to the diversion of military resources to deal with a dramatic rise in internal rebellion. In the nine garrisons along the border there were probably less than six hundred thousand deployed—a decline of about thirty percent from the Yong Le period—though many of these forces were undertrained, underequipped, and badly commanded (Wu 1937, Lin 1940, Huo Ji HMJSWBb 3:3446, Tong 1985: 434-36). Military security was also consuming enormous resources. The tun tian system of military farms had essentially collapsed in the previous century, as had the hereditary military family system. In the sixteenth century, the Ming relied increasingly on conscripted forces and corvee labor to build and supply the army (Wu 1937: 18290). As a result the Ming central government was forced to increase military expenditures, so much so that from mid-century through the 1570s funding for the military consumed between sixty and seventy percent of the central government's outlays.24 23

Limited tribute relations were restored with the Mongols (Lai and Li 1954, Serruys 1987: 74), but formal peace arrangements and regular cross-border trade were rejected by the Ming. 24 The figures are incomplete, but in 1564 military expenditures comprised sixty-nine percent of central government outflows of silver. From 1567 to 1569 this had dropped to about

236

CHAPTER SEVEN

The high cost of military expenditures was not surprising given the level of Mongol threat in the decade or so prior to the adoption of an accommodationist policy. One mid-sixteenth-century treatise on border affairs contended that at that time the Altan Khan could muster one hundred thousand troops for major incursions into China. Moreover, it was not only the scale of Mongol forces that posed a growing threat, but also their operational capabilities. In their longtime conflict with the Ming, the author argued, the Mongols had learned the value of taking weapons, blades, axes, armor, and even chariots from the Chinese. They had learned how to fight under fire from Ming firearms and how to deal with Ming fortifications. And they had learned how to coordinate the operations of footsoldiers and cavalry. The Mongols, in short, posed a particular threat by this time because of their ability to learn and change in the face of strategic conditions. They had started out weak, and now they posed a formidable military challenge (Yin 1532: 2-3). This assessment was consistent with overall Mongol behavior in this period. From mid-century on, after the collapse of the short-lived peace relations with the Mongols in 1551, the frequency of raiding on both a minor and major scale picked up considerably. From 1560 to 1570 the Mongols initiated ten major conflicts with the Ming, or about one per year. In contrast, the figure for the entire duration of the dynasty stood at about 0.5 per year. Ming resources were not only being swallowed up in the costs of border security in the north, but were being used to suppress growing domestic social unrest. According to Tong's data, in the ten years prior to the accommodation policy towards the Mongols, the Ming government faced seventy-six serious incidents of collective violence, an average of 7.6 incidents per year and well more than the average for the entire dynasty (and almost twenty times the figure for the Yong Le period). In short, by 1570 the Ming faced a period of military, fiscal, and domestic crisis. The arguments in favor of accommodation at this time explicitly referred to these acutely disadvantageous circumstances as the strategic reason for establishing peaceful relations with the Mongols until such time that Ming financial resources improved and its military capabilities were strengthened. The anecdotal support for the influence of a parabellum calculus on Ming strategy is reinforced by aggregate data: the Ming tended to act more coercively when, in the aggregate, the regime was more capable of doing so. While the indicators of a dynastic cycle are rough, Ming power peaked sometime in the mid- to late fifteenth century. Geographically, the empire reached its largest territorial size by mid-century at 6.53 million square kilometers. By 1500 this area had contracted to 4.71 million square kilometers (Taagepera 1978: 117). sixty-three percent, but jumped in 1570 to seventy-four percent just prior to the decision in 1571 to establish peaceful relations with the Altan Khan. These data come from table 6 in Quan and Li 1973:196.

MING GRAND STRATEGIC CHOICE

237

140 x 120.

O Ming initiation (cumulative) # Mongol initiation (cumulative)

100. 80. N 60. 40-

20.

1350

1400

1450

1500

1550

1600

1650

year

Fig. 7.1. Cumulative Frequencies of Ming and Mongol Conflict Initiation, by Five-Year Periods.* *Each mark along the x axis signifies a five-year period. There are fifty-five five-year periods in the Ming Dynasty. The crossover point in Ming-Mongol cumulative initiations of wars occurred in the twenty-fourth five-year period, 1483-1487.

Fiscally, although the central government remained in the black until the early sixteenth century, the cost of military construction, training, and supply rose steadily in the latter half of the fifteenth century as the military farming and family systems atrophied (Tong 1985: 438-40). Militarily, as we have seen, deployed Ming forces began dropping in size probably beginning in the midfifteenth century. Huang suggests that by the end of the century Ming forces along the northern frontier stood about forty percent of their original strength (1970: 43).25 If we use the frequency of major conflicts initiated by the Ming as an indicator of the level of Ming coerciveness, by the mid 1440s, or within the first quarter of the dynastic period, the Ming had already initiated about fifty percent of the total number of conflicts they would initiate to the end of the dynasty (see fig. 7.1). This works out to an average of about 0.45 per year, as compared to 0.24 per year on average for the entire dynastic period. If we use frequency of major conflicts initiated by the Mongols as an indicator of the level of threat faced by the Ming, we see that up to the mid-fifteenth century 25 This is considerably lower than the figures for the 1570s cited earlier. Judging from incomplete data from a number of sources, it appears that in some garrison areas there was an effort to increase military forces from the mid-1500s on.

238

CHAPTER SEVEN

the Ming faced a comparatively low level of threat.26 Only about eleven percent of Mongol initiations for the entire dynasty occurred in this period. This is an average of 0.19 per year as compared to an average of about 0.5 per year for the entire dynastic period. And if we use the frequency of serious internal collective violence as a very rough indicator of the Ming's ability to mobilize military and financial resources for external security, we find as well that this initial period in which the Ming acted most coercively was also a period in which the level of internal disorder was relatively low.27 According to the cumulative data, by the mid 1440s less than ten percent of the cases of collective violence that plagued the Ming state had occurred, an average of 0.73 per year. This is in contrast to the average of 2.2 per year for the entire dynastic period. These raw data, then, would seem consistent with the proposition that Ming decision makers, viewing the world throughparabellum lenses, chose to act more coercively given the capacity and opportunity to do so.28 We should also, however, subject this proposition to a more rigorous test, one that uses a more useful indicator of Ming coerciveness, namely the percentage—rather than raw frequencies—of conflicts with the Mongols initiated by the Ming. Figure 7.2 indicates that the Ming initiated the majority of conflicts with the Mongols from the founding of the dynasty up to the 1430s and 1440s, roughly the period when the relative power of the Ming grew to its highest level. From then to the end of the dynasty, except for a decade in the 1580s and 1590s, the Mongols clearly took the initiative in major conflicts with the Ming. The statistics bear this out. If Ming decision makers on the whole based their decision to act more coercively on their capability to do so—here defined 26 A better indicator of the Mongol threat would be the changing ratio of Ming and Mongol military strength. Unfortunately, there are no reliable data on raw Ming military capabilities, let alone the Mongols. Aside from occasional references in Ming documents to very approximate (and disputed) figures for, say, the number of Mongols living in the He Tao, or the size of a raid by a particular Mongol leader, Ming intelligence was apparently not reliable enough to construct rough estimates of Mongol strength. On the other hand, in reconstructing Ming threat assessments, the dearth of this sort of detailed data may not be too damaging. Judging from the memorials on the border question the Ming perception of threat levels was mostly a function of the frequency and extent of recent Mongol incursions into Chinese territory, combined with the state of Chinese defenses. Therefore, instead of military force ratios, I used the frequency of conflicts with the Ming initiated by the Mongols as a rough indicator of the level of threat faced by the Ming. 27 Tong (1985: 381) gives a nice summary of why collective violence is a good indicator of changes in the Ming ability to extract resources from the empire. As he shows empirically (see his chapter 6), collective violence is directly related to poor leadership, corruption, fiscal mismanagement, increasing tax burdens, reductions in famine relief, and declining coercive power within the empire. While there are obviously lags and feedback loops between these variables, over a 270-plus year period, rebellion seems to be a valid indicator of the capacity of the regime to mobilize and direct resources both within the state and against external adversaries. 28 These data also seem to confirm that the Tu Mu defeat was an appropriate symbol of and catalyst for the shift to a static defense-oriented security strategy as the Ming entered the extended downside of the dynastic cycle.

MING GRAND STRATEGIC CHOICE

239

1T O Moving average of Ming initiation (% # Moving average of Mongol initiation (

.6-.

.2-.

1350

1400

1450

1500

1550

1600

1650

year

Fig. 7.2. Yearly Percentage of Conflicts Initiated by the Ming and Mongols, Averaged over Five Years (moving average).* *Each mark along the x axis signifies a five-year period. There are fifty-five five-year periods in the Ming dynasty. Each mark along the y axis signifies the percentage of conflicts initiated per year, averaged over five years. The five-year average is then smoothed using a moving average. The percentage of conflicts initiated by the Mongols begins to surpass the Ming percentage in the late 1420s through the late 1430s, the thirteenth and fourteenth five-year periods.

as a relatively low level of threat combined with a relatively high capacity to mobilize resources against it—we should expect to see a negative relationship between coerciveness and Mongol threat and internal disorder.29 The data do indeed indicate this was the case (Table 7.2). There is a reasonably strong and statistically significant negative correlation (r= -.53) between the percentage of conflicts initiated by the Ming each year (averaged over five-year periods) and the lagged percentage initiated by the Mongols. There is as well a significant negative correlation (r= -.36) between Ming coerciveness and the frequency of internal collective violence, though the coefficient is not as high. But what of the combined relationship of these two indicators of capacity 29 Here I lag the indicator of threat by one five-year period on the assumption that the level of Ming coerciveness is a reaction to prior changes in this variable. The indicator for the Mongol threat was also changed from a raw yearly frequency to the mean yearly percentage of conflicts initiated by the Mongols. An increase in the relative aggressiveness of the Mongols at t-1 therefore ought to be associated positively with a decline in Ming aggressiveness at time t. I did not lag the variable capturing constraints on Ming capabilities (internal rebellion). I assumed that decision makers calculate the degree of threat on the basis of prior enemy behavior, while their capacity to respond to this threat is constrained by present resources.

240

CHAPTER SEVEN

TABLE 7.2

Correlation between Mean Yearly Percentage of Conflicts Initiated by the Ming (per five-year period), Mean Yearly Percentage of Conflicts Initiated by Mongols (per fiveyear period) (t-1), and Mean Yearly Frequency of Rebellion (per five-year period) Ming Initiation Mongol Initiation (t-1)

-.53

n=54

Rebellion

-.36

n=55

TABLE 7.3

Regression between Mean Yearly Percentage of Conflicts Initiated by the Ming (per five-year period), Mean Yearly Percentage of Conflicts Initiated by Mongols (per fiveyear period) (t-1), and Mean Yearly Frequency of Rebellion (per five-year period) Coefficient Intercept

Standard Error

t~ Value

.64

MONGINIT.

-.46

.12

-3.68*

REBELLION

-02

.02

-1.34**

Notes Regression equation: y = a - bxj - bx2 + e y = yearly percentage of conflicts initiated by the Ming (MINGINIT.) Xj = yearly percentage of conflicts initiated by Mongols (t— 1) (MONGINIT.) x2 = yearly frequency of rebellion (REBELLION) N = 54 F-test = 11.23 *p = 001 **p=187

and opportunity to the level of Ming coerciveness over time? If we combine them into a simple regression model with Ming coerciveness as the dependent variable, the results do not at least undermine the basic hypothesis (Table 7.3). While the R2 is not overwhelming—about thirty percent of the variation in Ming coerciveness is accounted for by the model—the two independent variables are both in the predicted direction, though the significance level for internal rebellion is approaching a problematic level. The relatively low R2 implies that there are a large number of years in which the dog did not bark, so to speak. In other words, there are quite a few years where conditions allowed the Ming to act more coercively against the Mongols, but where Ming decision makers did not take advantage of them. The question is whether the relationship between capacity and coerciveness in Ming strategy

MING GRAND STRATEGIC CHOICE

24 1

TABLE 7.4

Association between Ming Initiation, Threat, and Capabilities Contingency Table Analysis Years with Ming initiation

Years with no Ming initiation

Mongol initiation (t-1) with no rebellion (t-1)

2 (5)

25 (22.1)

27

Mongol initiation (t-1) with rebellion (t-1)

12(14.8)

69(66.2)

81

No Mongol initiation (t-1) with no rebellion (t-1)

17 (10.8)

42 (48.2)

59

No Mongol initiation (t-1) with rebellion (t-1)

19(19.4)

87(86.6)

106

50

223

273

Notes (Figures in parentheses are expected cases. Observations are years.) DF 3 Chi Square 7.168 Significance .05 Ql Jiguangyanjiu lunji (H£jffi3feW^Sfraft)Beijing, pp. 70-90. Fanner, Edward L. (1976). Early Ming Government: The Evolution of Dual Capitals. Cambridge, Mass. Fischer, Markus (1992). "Feudal Europe, 800-1300: Communal Discourse and Conflictual Practices." International Organization 46:2, pp. 427-66. . (1993). "On Context, Facts, and Norms: Response to Hall and Kratochwil." International Organization 47:3, pp. 493—500. Franck, Thomas M., and Edward Weisband (1971). Word Politics: Verbal Strategy among the Superpowers. New York. Fu Shangkui [ « f £ M ] (1990). "Sun Zi zhanlue sixiang tan yao" {M^W&^MM S ) , in Shehui kexueji kan ( | ± # ? 4 * $ S f U), vol. 2. Gaenslen, Fritz (1986). "Culture and Decision Making in China, Japan, Russia and the United States." World Politics 39:1, pp. 78-103. Gahmberg, Henrik (1987). "Semiotic Tools for the Study of Organizational Culture," in Thomas A. Sebeok and Jean Umiker-Sebeok, ed., The Semiotic Web 1986. Amsterdam, pp. 389-403. Gao Rui [((?!£] (1990). "Sun Zi bu zhan er qu ren zhi bing quan sheng xin tan" ( R ?

^mffOJS A£:ft, ^ E § f S ) in Sun Zixin tan m^fxUl

Beijing.

Geertz, Clifford (1973). The Interpretation of Cultures. New York. George, Alexander (1969). "The Operational Code: A Neglected Approach to the Study of Political Leaders and Decision-making." International Studies Quarterly 13, pp. 190-222. Gills, Barry (1993). "The Hegemonic Transition in East Asia: A Historical Perspective," in Stephen Gill, ed., Gramsci, Historical Materialism and International Relations. Cambridge. Gilpin, Robert (1981). War and Change in World Politics. Cambridge. Glenn, Edmund S., et al. (1970). "A Cognitive Interaction Model to Analyze Culture Conflict in International Relations." Journal of Conflict Resolution 14:1, pp. 35—50. Godwin, Paul (1984). "Soldiers and Statesmen in Conflict: Chinese Defense and Foreign Policies in the 1980s," in Samuel S. Kim, ed., China and the World: Chinese Foreign Policy in the Post-Mao Era. Boulder. Colo., pp. 215-34.

REFERENCES

279

Goldstein, Judith, and Robert O. Keohane (1993). "Ideas and Foreign Policy: An Analytical Framework," in Goldstein and Keohane, ed., Ideas and Foreign Policy: Beliefs, Institutions and Political Change. Ithaca, N.Y., pp. 3-30. Gong Annan [M*£W], ed. and anno. (1989). HeBoshi bei lunyizhu (Mff ± { i H P Y±). Beijing. Gray, Colin (1981). "National Styles in Strategy: The American Example." International Security 6:2, pp. 21-47. —. (1986). Nuclear Strategy and National Style. Lanham, Md. . (1990). War, Peace and Victory: Strategy and Statecraft for the Next Century. New York. Green, Sebastian (1988). "Understanding Corporate Culture and Its Relation to Strategy." International Studies of Management and Organization 18:2, pp. 6-28. Gregory, Donna U. (1989). "Forward," in James Der Derian and Michael J. Shapiro, ed., International/Intertextual Relations: Post Modern Readings in World Politics. Lexington, Mass. Griffiths, Samuel B. (1982). Sun Tzu: The Art of War. London. Guo Huaruo [ f P ^ ^ ] (1989). Guo Huaruojunshi lunwen xuanji (?fli}S;t=f l S ^ ) . Beijing. Guo Zebo [WMlfc] (1987). "Sun Zi fang yu zuo zhan sixiang chu tan" IJcSffilSJ^), in Junshi lishi C S l M S i ) , no. 3, Beijing. Hall, Rodney Bruce, and Friedrich V. Kratochwil (1993). "Medieval Tales: Neorealist 'Science' and the Abuse of History." International Organization 47:3, pp. 479-91. Hart, Thomas G. (1978). "Cognitive Paradigms in the Arms Race: Deterrence, Detente and the 'Fundamental Error' of Attribution." Conflict and Cooperation 3, pp. 147161. Hatori, Chiharu (1989). "Sun Zi bingfa xin tan" ( R i ^ &§f S ) . Paper presented to the first International Symposium on Sun Zi's Art of War, Huimin, China. Hattendorf, John B. (1991). "Alliance, Encirclement and Attrition: British Grand Strategy in the War of the Spanish Succession, 1702-1713," in Paul Kennedy, ed., Grand Strategies in War and Peace. New Haven, Conn., pp. 11—30. He Liangchen [ f q J ^ E ] (1984). Chenjizhu shi {Wfa&WY Beijing. He Lingxu [M^jffi] (1988). ShangJun shujin zhijinyi C B f t l t ^ i i ^ l S ) . Taibei. He Zhongsheng [fqff^^] (1606). Binglue (^ffi§). Ming Chong Zhen ed. 1632. Taiwan National Library. Herken, Gregg (1985). Counsels of War. New York. Herrmann, Richard (1988). "The Empirical Challenge of the Cognitive Revolution: A Strategy for Drawing Inferences about Perceptions." International Studies Quarterly 32, pp. 175-203. History of Chinese Wars Editorial Committee (1986). Zhongguo lidai zhanzheng shi ( ^ H K f ^ K ^ ^ ) . Vol. 14: The Ming Dynasty. Taibei. HMJSWBa [MMWkW~$tM\ Huang Ming jing shi wen bian. Ming Chong Zhen Ping Wu Tang ed. 38 vols. Taibei. HMJSWBb [M 0 ^IS1S5:I§] Huang Ming jing shi wen bian. 6 vols. Hong Kong. HMZL [M ? ^ B 3 ^ P ] Huang Ming zhao ling. Jia Qing 27 nian xiao jie ben, 4 vols. Taibei. Hofheinz, Roy (1977). The Broken Wave: The Chinese Communist Peasant Movement, 1922-1928. Cambridge, Mass.

280

REFERENCES

Hollander, Edwin P. (1985). "Leadership and Power," in Gardner Lindzey and Elliot Aronson, ed., Handbook ofSocial Psychology; vol. 2. New York, pp. 485-537. Holsti, Ole R. (1969). "The Belief System and National Images: A Case Study," in James N. Rosenau, ed., International Politics and Foreign Policy. New York, pp. 543-50. Hsiao, Kung-chuan (1979). A History of Chinese Political Thought. F. W. Mote trans. Princeton. Huang, Ray (1970). "Military Expenditures in Sixteenth Century Ming China." Oriens Extremus 17:1/2, pp. 39-62. . (1981). "The Liao-tung Campaign of 1619." Oriens Extremus 28:1, pp. 30-54. Huang Wenqin [Hr^fe] (1981). "Ming Chengzu shi dai Liaodong de jing hie" ffiBffta JKlft$SK§), in Ming shi yanjiu (BJ*W25), no. 4. Hucker, Charles O. (1978). The Ming Dynasty: Its Origins and Evolving Institutions. Michigan Papers in Chinese Studies, no. 34. Ann Arbor, Mich. Hunt, Michael H. (1984). "Chinese Foreign Relations in Historical Perspective," in Harry Harding, ed., China's Foreign Relations in the 1980s. New Haven, Conn., pp. 1-42. Huntington, Samuel P. (1993). "A Clash of Civilizations?" Foreign Affairs 72:3, pp. 22-49. Interview (1989a). Retired Guomindang general. Taibei. October. . (1989b). Retired Guomindang general. Taibei. October. . (1990a). Researchers at National Defense University. Beijing. November. . (1990b). Researchers at Academy of Military Sciences. Beijing. November. . (1990c). Researchers at Academy of Military Sciences. Beijing. December. . (1991a). Researchers at Academy of Military Sciences. Beijing. March. . (1991b). Researcher at Border History and Geography Research Institute, Chinese Academy of Social Sciences. Beijing. May. (1991c). Researcher, at Border History and Geography Research Institute, Chinese Academy of Social Sciences. Beijing. May. Jacobsen, Carl G., ed. (1990). Strategic Power: USA/USSR. London. Jagchid, Sechin, and Van Jay Symons (1989). Peace, War, and Trade along the Great Wall: Nomadic-Chinese Interaction Through Two Millennia. Bloomington, Ind. Jervis, Robert (1976). Perception andMisperception in International Politics. Princeton. . (1984). The Illogic of American Nuclear Strategy. Ithaca, N.Y. . (1989). The Meaning of the Nuclear Revolution: Statecraft and the Prospect of Armageddon. Ithaca, N.Y. Jiang Weiguo [MMH], ed. (1986). Zhongguo lidai zhanzheng shi ( ^ g U g f t K ^ ^ ) , vol. 14. Taibei. Jiang Yiqing [jl^F pf], ed. and anno. (1989). Kong Ming bingfa Q\M&&). Taibei. Johnson, James D. (1991). Symbols and Strategy: On the Cultural Analysis of Politics. University of Chicago Ph.D. diss. Johnson, James Turner (1981). Just War Tradition and the Restraint of War. Princeton. Johnston, Alastair I. (1992). "Baoli yi fei baoli" (i|;£j^#|jl;fr), in Guoji zhengzhi yanjiu ( H ^ K ^ W c % \ no. 1. Beijing, pp. 61-66. . (1994). "Cultural Realism and Strategy in Maoist China." Paper presented to Third S SRC/MacArthur Conference on Norms and National Security. Stanford, Calif.

REFERENCES

28 1

Jones, Archer (1987). The Art of War in the Western World. Urbana, 111. Jones, David R. (1990). "Soviet Strategic Culture," in Carl G. Jacobsen, ed., Strategic Power: USA/USSR. London, pp. 35-49. Kanwisher, Nancy (1989). "Cognitive Heuristics and American Security Policy." Journal of Conflict Resolution 33:4, pp. 652-75. Kaplan, Fred (1983). Wizards of Armageddon. New York. Keesing, Roger M. (1974). "Theories of Culture," in Bernard Siegal, et al., ed., Annual Review of Anthropology, vol. 3, pp. 73-97. Kelman, Herbert C , and Reuben M. Baron (1968). "Determinants of Modes of Resolving Inconsistency Dilemmas: A Functional Analysis," in Robert Abelson, et al., ed., Theories of Cognitive Consistency: A Sourcebook. Chicago, pp. 670-83. Kennedy, Paul, ed. (1991). Grand Strategies in War and Peace. New Haven, Conn. Keohane, Robert, ed. (1986). Neorealism and Its Critics. New York. Keohane, Robert O., and Joseph S. Nye (1977). Power and Interdependence: World Politics in Transition. Boston. Kertzer, David I. (1988). Ritual, Politics and Power. New Haven, Conn. Khong, Yuen Foong (1992). Analogies at War: Korea, Munich, Dien Bien Phu, and the Vietnam Decisions of 1965. Princeton. Kier, Elizabeth (1992). Changes in Conventional Military Doctrine: The Cultural Roots of Doctrinal Change. Cornell University Ph.D. Diss. Kierman, Frank, and John Fairbank, ed. (1974). Chinese Ways in Warfare. Cambridge, Mass. Kim, Samuel S. (1979). China, the United Nations and World Order. Princeton. . (1980). "Chinese Global Policy: An Assessment," in James Hsiung and Samuel S. Kim, ed., China in the Global Community. New York. Klein, Bradley (1988). "Hegemony and Strategic Culture: American Power Projection and Alliance Defence Politics." Review of International Studies 14, pp. 133-48. *—. (1989). "The Textual Strategies of the Military: Or Have You Read Any Good Defense Manuals Lately," in James Der Deri an and Michael J. Shapiro, ed., International/ Intertextual Relations: Postmodern Readings of World Politics. Lexington, Mass. Klein, Yitzak (1991). "A Theory of Strategic Culture." Comparative Strategy 10:1, pp. 3-24. Kolodner, Janet, and Robert H. Simpson (1986). "Problem Solving and Dynamic Memory," in Kolodner and Christopher Riesbeck, ed., Experience, Memory and Reasoning. Hillsdale, N.J. Kong Deqi [?LH!R] (1987). Liu Tao qian shuo (7\$8Slft). Beijing. Krakau, Knud (1984). "American Foreign Relations: A National Style?" Diplomatic History 8:3, pp. 253-72. Kull, Steven (1988). Minds at War: Nuclear Reality and the Inner Conflicts of Defense Policymakers. New York. Kupchan, Charles (1994). The Vulnerability of Empire. Ithaca, N.Y. Lai Jiadu [ f l ^ S ] and Li Guangbi [$7fcH] (1954). Ming dai dui Wa La de zhan

zhengmiXm'KMmWWy

Shanghai.

Laitin, David (1988). "Political Culture and Political Preferences." American Political Science Review 82: 2, pp. 589-98.

282

REFERENCES

Lan Yongwei [M^cgf] (1987). Chunqiu shiqi de bu bing (#$CB#$§#}*£:£). Taibei. Lattimore, Owen (1962). Studies in Frontier History, Collected Papers, 1928-1958. London. Lebow, Richard Ned (1981). Between Peace and War: The Nature of International Crisis. Baltimore. Legge, James, trans. (1935). The Chinese Classics. Vol. 5: The Ch 'un Ts'ew with the Tso Chuen. Oxford. Legro, Jeffrey (1995). Cooperation Under Fire: Anglo-German Restraint During World War II. Ithaca, N.Y. Leng, Russell J. (1983). "When Will They Ever Learn? Coercive Bargaining in Recurrent Crises." Journal of Conflict Resolution 27:3, pp. 379-419. Levy, Sheldon G. (1992). "Attitudes towards the Conduct of War." Paper prepared for the Fourth World Peace Science Congress, Rotterdam. Lewis, Mark Edward (1990). Sanctioned Violence in Early China. Albany, N.Y. Li Bingyan [^MB] and Sun Jing [Bfcgfc] (1989). Junshi moulue xue (SVifKg). Beijing. Li Bingzhong [^'JR3c] and Li Hongxi [ $ ^ S ] (1990). "Jian lun Sun Zi de zhudong quan sixiang" ( g f f i t S ^ l f t i f t f i i g i i ) , in Sun Zixin taniM^^U). Beijing. Li Jiurui [^%M] (1978). Zhongguo junshi sixiang shi. ( t f l f f S U ) . Taibei. Li Ling [^^|] (1990). Lectures on ancient military texts, Peking University. . (1991). Interview, Beijing. , anno. (1991). "Sun Zi bingfa yi zhu" (M^^&WA) in Luan Baoqun [?£ iSI¥] and Yang Yonglin [|g^k#], ed., Bingjia baojian G f t ^ J t l ! ) . Beijing. Li Quan [ ^ H ] (1957). Tai baiyinjing {X fi ft®, in Li Yuri [^Y4 B ], ed., Zhongguo bing xue daxi ( ^ H ^ ^ ^ ^ ) , vol. 2. Taibei. Li Shuoshi [^WG] and Wang Shijin [ i ^ i ] (1986). Wu Zi qian shuo ( i g ^ S |ft). Beijing. Li Yuri [&%& 0 ] , ed. (1957). Zhongguo bing xue da xi (t£ H ^ l p ; * ; ^ ) . Taibei. Li Zhen [&M] (1983). Zhongguo junshi jiaoyu shi (if S H Vtfc WSt). Taibei. Lin, Chong-pin (1988). China s Nuclear Weapons Strategy: Tradition within Evolution. Lexington, Mass. Lin Pinshi [ # p p 5 1 , ed. and anno. (1986). Lu Shi Chun Qiujin zhujinyi (S.R#1^C 4 > & ^ S ) . Taibei. Lin Shuhui [ # ® 3 i ] (1940). "Ming zhi bei bian bei yu" ($J ^ 4 b ^ f f i S ) , in Shi xue nian bao ( $ ? ^ i ) , vol. 12, pp. 75-99. Lin, T. C. (1936). "Manchuria in the Ming Empire." Nankai Social and Economic Quarterly 8, pp. 1-42. . (1937). "Manchuria Trade and Tribute in the Ming Dynasty." Nankai Social and Economic Quarterly 9, pp. 855-92. Lin, Tzong-biau, and Ho Lok-sang (1989). "Is There a Link among Confucianism, Institutions and Economic Performance?" Paper presented to the Conference on Confucianism and Economic Development in East Asia, May 29-31. Taipei. Liu Di [HUS] (1990a). "Wo guo gudai chuantong zhi bian sixiang chu tan" ($lcBJl5*

iX WMIUTSLSMWW), in Ma Dazheng [If ^ I E ] , ed., Zhongguo gudai bianjiang zhengceyanjiu ( ^ W^^iVAM^WM%)Beijing. Liu Ling [gjfp] (1989). "Sun Zi bingfa yu wu xing shuo"

REFERENCES

283

Paper presented to the First International Symposium on Sun Zi's Art of War. Huimin, China. . (1990b). Baihua Sun Zi bingfa du ben (filSB^rftffitR^). Shenyang. Liu Lumin [glJUJg], and Su Dexiang [ U t i ^ L chief eds. (1988). Zhongguo bingshu ji cheng (t£ [lift U S $c). Beijing. Liu Qing [SIS] (1990c). "Zhong Fa zhanzheng yu wan Qing guofang zhengce de zhuan bian" ( ^ K ^ B f e i t BESi&9l^ff ®). Beijing, unpub. ms. Liu Shenning [gljEf^] (1990d). Zhongguo bingshu zong mu (4 1 H ^ f t & t @). Beijing. Liu Xianting [gU^feS], and Yao Youzhi [MW^S] (1988). "Dong Zhou shiqi de zhanzheng gaikuang" (JKM0#B9 W l ^ t E f f i ) , in Zhongguo junshi kexue (*£ SM9 14*), no. 1, pp. 89-96. Liu Xuetiao [gjjlpifc] (1979). Zhongguo lidai bianjiang da shi nian biao (tip Liu Yin [g!|H], ed. and anno. (1955a). Wu Jing Qi Shu zhijie CK Taibei. . (1955b). Si Ma Fa zhijie (W] JS&fifl?). Taibei. . (1955c). Wei Liao Zi zhijie ( g ^ ^ B W ) . Taibei. . (1955d). Tai Gong Liu Tao zhijie Cfc&ASSllLJS). Taibei. (BX^ . (1955e). Tang TaizongLi Wei Gong Wen Duizhijie # ) . Taibei. -. (SKQS) Huang Shi Gong San Lue zhijie (ilf 5 &.=iN& BiLfl?). (Si Ku Quan Shu [ligj^^fi] version). Taibei. LiuZhongping [glJW2?] (1986). SiMaFajinzhujinyi^W&^&^rW). Taibei. , ed. and anno. (1987). Wei Liao Zijin zhujinyi ( W ' J K ^ ^ & ^ P ) . Taibei. Lo, Jung-pang (1969). "Policy Formulation and Decisionmaking on Issues Respecting Peace and War," in Charles O. Hucker, ed., Chinese Government in Ming Times. New York, pp. 41-72. Loewe, Michael (1974). "The Campaigns of Han Wu-ti," in Frank Kierman and John K. Fairbank, ed., Chinese Ways in Warfare. Cambridge, Mass. Lord, Carnes (1985). "American Strategic Culture." Comparative Strategy 5:3, pp. 269-93. Lovell, John P. (1990). "The United States as Ally and Adversary in East Asia: Reflections on Culture and Foreign Policy," in Jongsuk Chay, ed., Culture and International Relations. New York, pp. 89-102. Luckham, Robin (1984). "Armament Culture." AIternatives 10:1, pp. 1-44. Luke, Timothy W. (1989). "What's Wrong with Deterrence?: A Semiotic Interpretation of National Security Policy," in James Der Derian and Michael J. Shapiro, ed., International/Intertextual Relations: Post Modern Readings in World Politics. Lexington, Mass., pp. 207-29. Luttwak, Edward N. (1976). The Grand Strategy of the Roman Empire from the First Century A.D. to the Third. Baltimore. . (1987). Strategy: The Logic of War and Peace. Cambridge, Mass. Ma Dazheng [>^^:IE], ed. (1990a). Zhongguo gudai bianjiang zhengceyanjiu ( ^ H ^iXmM1SkWM%y Beijing. Ma Dongli [JSlRlt] (1990b). He Boshi Bei Lun zhuyi ({^W±MW&U)Beijing.

284

REFERENCES

Ma Qingping [JSWi¥] (1979). Yu Qian zhi yanjiu ffSi£W?E). China Culture University, M.A. thesis. MacCormac, Earl R. (1985). A Cognitive Theory of Metaphor. Cambridge, Mass. Mack, John E. (1990). "The Enemy System," in Vamik D. Volkan, et a l , ed., The Psychodynamics of International Relationships. Vol. 1: Concepts andTheories. Lexington, Mass. Mao Zedong [35§ljS] (1936). "Zhongguo geming zhanzheng zhanlue wenti" ( ^ H ^^pSi^KffiSr«1ffi), in Mao Zedongxuanji £&WMM$& (1967). Beijing, pp. 154-225. . (1938a). "Lun chi jiu zhan" (HRp&K), in Mao Zedong xuanji ( ^ H l (1967). Beijing, pp. 407-84. -. (1938b). "Zhanzheng yu zhanlue wenti" ( f ^ ! £ K B & F«13), in Mao Zedong xuanji ( = g g ^ g l | i ) (1967). Beijing, pp. 506-21. March, James (1978). "Bounded Rationality, Ambiguity and Engineering of Choice." The BellJournal of Economics 9:2, pp. 587-608. McCauley, Mary (1984). "Political Culture and Communist Politics: One Step Forward, Two Steps Back," in Archie Brown, ed., Political Culture and Communist Studies. London. MccGwire, Michael (1987). Military Objectives in Soviet Foreign Policy. Washington, D.C. McNeely, Connie (1992). "The Determination of Statehood in the United Nations, 1945-1985." Research in Political Sociology, vol. 6. MCZY [Ming Chen Zou Yi $j E # H ] . Mearsheimer, John (1983). Conventional Deterrence. Ithaca, N.Y. . (1990). "Back to the Future: Instability in Europe after the Cold War." International Security 15:1, pp. 5-56. Millinger, James (1973). "Ch'i Chi-kuang—A Ming Military Official as Viewed by His Contemporary Civilian Officials." Oriens Extremus 20:1, pp. 103-17. Milner, Helen (1992). "Cooperation among Nations," World Politics 44:3, pp. 466-96. Ming Chengzu [B^J^H] (1410). Sheng xue xinfa ( $ ¥ ^ f f i ) in Zhongguo zi xue vol. 8, Taibei. mingzhuji cheng Qp WirF^ft^^fiSO, Moravcsik, Andrew (1992). "Liberalism and International Relations Theory." Harvard University Center for International Affairs Working Papers Series, no. 92-6. Morgenthau, Hans (1978). Politics among Nations: The Struggle for Power and Peace. New York. Mote, Frederick (1974). "The T'u-mu Incident of 1449," in Frank Kierman and John K. Fairbank, ed., Chinese Ways in Warfare. Cambridge, Mass, pp. 243-73. . (1988). "The Ch'eng-hua and Hung-chih reigns, 1465-1505," in Frederick W. Mote, Denis Twitchett, and Michael Loewe, ed., The Cambridge History of China. Vol. 7: The Ming Dynasty, 1361-1644, Parti Cambridge, pp. 343-402. Mueller, John (1993). "The Impact of Ideas on Grand Strategy," in Richard Rosecrance and Arthur A. Stein, ed., The Domestic Bases of Grand Strategy. Ithaca, N.Y, pp. 48-62. Namenwirth, J. Zvi, and Robert Philip Weber (1987). Dynamics of Culture. Boston. Nijman, Jan (1991). "The Dynamics of Superpower Spheres of Influence: U.S. and Soviet Military Activities, 1948-1978." International Interactions 17:1, pp. 63-91.

REFERENCES

285

Nye, Joseph, and Sean Lynn-Jones (1988). "International Security Studies: A Report on a Conference on the State of the Field." International Security 12:4, pp. 5-27. Ott, J. Steven (1989). The Organizational Culture Perspective. Pacific Grove, Calif. Pettigrew, Andrew M. (1979). "On Studying Organizational Cultures." Administrative Science Quarterly 24, pp. 570-81. Pillsbury, Michael (1980/1981). "Strategic Acupuncture." Foreign Policy, no. 41, pp. 44-61. Pipes, Richard (1977). "Why the Soviet Union Thinks it Could Fight and Win a Nuclear War." Commentary 64:1, pp. 21-34. . (1980). "Militarism and the Soviet State." Daedalus 109, pp. 1-12. Pokotilov, Dmitri (1976). History of the Eastern Mongols during the Ming Dynasty from 1368-1634. Philadelphia. Posen, Barry R. (1984). The Sources of Military Doctrine: France, Britain, and Germany Between the World Wars. Ithaca, N.Y Powell, Charles A., et al. (1987). "Opening the 'Black Box': Cognitive Processing and Optimal Choice in Foreign Policy Decisionmaking," in Charles Hermann et al., ed., New Directions in the Study of Foreign Policy. Boston, pp. 203—20. Powell, Robert (1991). "Absolute and Relative Gains in International Relations Theory." American Political Science Review 85:4, pp. 1303-20. . (1994). "Anarchy in International Relations Theory: The Neorealist-Neoliberal Debate." International Organization 48:2, pp. 313-44. Price, Richard, and Nina Tannenwald (1994). "Norms and Deterrence: The Nuclear and Chemical Weapons Taboos." Paper presented to Third S SRC/Mac Arthur Conference on Norms and National Security, Stanford, Calif. Pringle, Peter, and William Arkin (1983). SIOP: The Secret U.S. Plan for Nuclear War. New York. Przeworski, Adam, and Henry Tuene (1970). The Logic of Comparative Social Inquiry. New York. Pye, Lucian W. (1968). The Spirit of Chinese Politics: A Psychocultural Study of the Authority Crisis in Political Development. Cambridge, Mass., pp. 31-32. . (1988). The Mandarin and the Cadre: China s Political Cultures. Ann Arbor, Mich.

Qiu Shaohua [fl54>Ig] (1985). Xian Qin zhuzijunshi lunyizhu {^M^^WMWM vi), vol. 1. Beijing. Quan Hansheng [ife9I5H a n d L i Longhua [ $ f | f l ] (1973). "Ming zhong ye hou tai Hang can ru yin Hang de yanjiu" ($} ^Mfa J^1kMAifiMWW55), in Zhongguo wenhuayanjiusuoxuebao ( ^ S ^ 4 f c W ^ # f * ? l i ) 6:1, pp. 169-244. Rand, Christopher C. (1977). The Role ofMilitary Thought in Early Chinese Intellectual History. Harvard University Ph.D. diss. Rappoport, Anatol (1978). "Changing Concepts of War in the United States," in Ken Booth and Moorhead Wright, ed., American Thinking about Peace and War. New York. Regan, Patrick M. (1992). Organizing Societies for War: The Process and its Consequence for State Behavior. University of Michigan Ph.D. diss. Rossabi, Morris, ed. (1983). China among Equals: The Middle Kingdom and Its Neighbors, 10th-14th Centuries. Berkeley.

286

REFERENCES

Russett, Bruce (1993). Grasping the Democratic Peace: Principles for a Post-Cold War World. Princeton. Sagan, Scott (1987). "SIOP-62: The Nuclear War Plan Briefing to President Kennedy." International Security 12:1, pp. 22-51. Sampson, Martin W. (1987). "Cultural Influences on Foreign Policy," in Charles Hermann, et al., ed., New Directions in the Study of Foreign Policy. Boston, pp. 384-405. Sariti, Anthony William (1973). "A Note on Foreign Policy Decisionmaking in the Northern Sung." Sung Research Newsletter, no. 8, pp. 3—11. Sawyer, Ralph D. trans. (1993). The Seven Military Classics of China. Boulder, Colo. SBSYJZSZ (1978). Song ben shiyijia zhu Sun Zi ( T ^ ^ - | — - ^ jftSS-?). Shanghai. Scheer, Robert (1982). With Enough Shovels: Reagan, Bush, and Nuclear War. New York. Schelling, Thomas (1980). The Strategy of Conflict (2d ed.). Cambridge, Mass. Schuman, Howard, and Michael P. Johnson (1976). "Attitudes and Behavior," in Annual Review of Sociology, vol. 2, Palo Alto, Calif. Schweller, Randall L. (1992). "Domestic Structure and Preventive War: Are Democracies More Pacific?" World Politics 44:2, pp. 235-69. Segal, Gerald (1983/1984). "Strategy and Ethnic-chic." International Affairs 60:1. . (1985a). "Defense Culture and Sino-Soviet Relations." Journal of Strategic Studies 8, pp. 180-98. -. (1985b). Defending China. London. Serruys, Henry (1967). Sino-Mongol Relations during the Ming II: The Tribute System and Diplomatic Missions (1400—1600). Brussels. . (1982). "Towers in the Northern Frontier Defenses of the Ming." Ming Studies 14, pp. 9-76. -. (1987). The Mongols and Ming China: Customs and History. London. Seyschab, Carl-A. (1990). "The 36 Strategems: Orthodoxy Against Heterodoxy," in C.-A. Seyschab, A. Sievers, and S. Synkewicz, ed., Society, Culture, and Patterns of Behaviour. Bonn, pp. 97-255. Shek, Richard Hon-Chun (1980). Religion and Society in Late Ming: Sectarianism and Popular Thought in Sixteenth and Seventeenth Century China. University of California, Berkeley, Ph.D. diss. Shi Zimei [Kg-? J i ] (1222). Shi shi qi shujiangyi ( 1 ^ - b f l l ) . Riben wen jiu san nian edition (1863). Taiwan National Museum. Shy, John (1971). "The American Military Experience: History and Learning." Journal of Interdiscipinary History 1, pp. 205-28. Singer, J. David (1968). "Man and World Politics: The Psycho-Cultural Interface." Journal of Social Issues 24:3, pp. 127-56. Smircich, Linda (1983). "Concepts of Culture and Organizational Analysis." Administrative Science Quarterly 28, pp. 339-68. Smith, Kidder (1990). "Making Strategic Choices." Paper presented to the panel on Warfare and Expertise in Ancient China, Association of Asian Studies annual meeting, Chicago. Smith, Michael Joseph (1986). Realist Thoughtfrom Weber to Kissinger. Baton Rouge. Snyder, Glenn, and Paul Diesing (1977). Conflict among Nations: Bargaining, Decision Making and System Structure in International Crises. Princeton.

REFERENCES

287

Snyder, Jack (1977). The Soviet Strategic Culture: Implications for Nuclear Options. Rand R-2154-AF, Santa Monica, Calif. . (1984). The Ideology of the Offensive: Military Decision Making and the Disasters of 1914. Ithaca, N.Y. -. (1990). "The Concept of Strategic Culture: Caveat Emptor," in Carl C. Jacobsen, ed., Strategic Power USA/USSR. London. -. (1991). Myths of Empire: Domestic Politics and International Ambition. Ithaca, N.Y. Soysal, Yasemin (1993). "Limits of Citizenship: Post-National Membership in the Contemporary Nation-State System." Paper presented to first S SRC/Mac Arthur Workshop on Norms and National Security, Ithaca, N.Y. Stein, Howard F. (1990). "The Indispensible Enemy and American-Soviet Relations," in Vamik D. Volkan, Demetrios A. Julius, and Joseph V. Montville, ed., The Psychodynamics of International Relationships, Vol. 1: Concepts and Theories. Lexington, Mass. Stenelo, Lars-Goran (1981). "Predictions and Foreign Policy Heritage." Cooperation and Conflict 16, pp. 3-17. Struve, Lynn A. (1984). The Southern Ming, 1644-1662. New Haven, Conn. Stuart, Reginald C. (1982). War and American Thought: From the Revolution to the Monroe Doctrine. Kent, Ohio. Sun Bin tf&M] (1985). Sun Bin bingfa (i^JK^ffi). Taibei. Sun Xiangming [Bfc|nJ^], and Zhang Huican [WM£H], ed., (1989). Zhanzhengyu zhanlue lilunji cui ( « ^ J f c ¥ i f B £ S M I i # ) . Beijing. Taagepera, Rein (1978). "Size and Duration of Empires: Systematics of Size." Social Science Research 7, pp. 108-27. Taber, Charles (1987). "Modern War Learning: A Markov Model." Paper presented to the Midwestern Political Science Association conference, Chicago. Tan Eng Bok, Georges (1984). "Strategic Doctrine," in Gerald Segal and William Tow, ed., Chinese Defence Policy. London, pp. 3-17. Tang Shunzhi [flf HH£] (1618). Wu Bei ® | § ) . , Si Ku Quan Shu ( H f ^ f ) edition. Shanghai. Tang Xiumei [JfifJI J t ] (1971). Zhang Juzheng de zhengzhi sixiangyu celue yanjiu ( 5 S g i E W i K ^ S f f l l ^ f f i i S W ? S ) . Taiwan National University, M.A. Thesis. Tao Jingshen (1983). "Barbarians or Northerners: Northern Sung Images of the Khitans," in Rossabi (1983). Taylor, Romeyn (1969). "Yuan Origins of the Wei-so System," in Charles O. Hucker, ed., Chinese Government in Ming Times. New York, pp. 73-128. Terhune, Kenneth W. (1970). "From National Character to National Behaviors: A Reformulation." Journal of Conflict Resolution 14:2, pp. 1-61. Tian Xudong [ffljfijR], ed. and anno. (1989). Si Ma Fa qianshuo (W] JSffiSlft). Beijing. Tong, James (1985). Collective Violence in a Premodern Society. University of Michigan Ph.D. diss. Twining, David T. (1989). "Soviet Strategic Culture—The Missing Dimension." Intelligence and National Security 4:1, pp. 169—87. Twitchett, Denis, and Michael Loewe, ed. (1986). The Cambridge History of China. Vol. 1: The Ch 'in and Han Empires, 221 B.C-A.D. 220. Cambridge, pp. 377-462.

288

REFERENCES

Urban, Michael E. (1982). The Ideology of Administration: American and Soviet Cases. Albany, N.Y. Vaillancourt, Pauline M. (1987). "Discourse Analysis, Post-Structuralism, PostModernism, Deconstruction, Semiotics: A New Paradigm for the Social Sciences?" The Polish Sociological Bulletin 4, pp. 89-100. Van Evera, Stephen (1984). Causes of War. University of California, Berkeley, Ph.D. diss. Vasquez, John A. (1983). The Power ofPower Politics: A Critique. New Brunswick, NJ. . (1986). "Capability, Types of War, Peace." Western Political Science Quarterly 39:2, pp. 313-27. -. (1987). "Foreign Policy Learning and War," in Charles Hermann, Charles E. Kegley, and James N. Rosenau, ed., New Directions in the Study ofForeign Policy. Boston, pp. 366-83. -. (1993). The War Puzzle. Cambridge. Vertzberger, Yaacov Y. I. (1986). "Foreign Policy Decisionmakers as Practical-Intuitive Historians: Applied History and Its Shortcomings." International Studies Quarterly 30:2, pp. 223-47. . (1990). The World in Their Minds: Information Processing, Cognition, and Perception in Foreign Policy Decisionmaking. Stanford. Waldron, Arthur (1983). "The Problem of The Great Wall of China." Harvard Journal ofAsiatic Studies 43:2, pp. 643-63. . (1990). The Great Wall of China: From History to Myth. Cambridge. Walker, R.B.J. (1990). "The Concept ofCulture in the Theory of International Relations," in Jongsuk Chay, ed., Culture and International Relations. New York, pp. 3-20. Walker, Richard L. (1953). The Multi-State System ofAncient China. Hamden, Conn. Wallacher, Benjamin E. (1975). "Chinese Ways in Warfare: Review." Harvard Journal ofAsiatic Studies 35, pp. 339-46. Walt, Stephen M. (1987). "The Search for a Science of Strategy: A Review Essay on 'Makers of Modern Strategy'." International Security 12:1, pp. 140-65. Waltz, Kenneth (1979). Theory ofInternational Politics. New York. Wang Gungwu (1983). "The Rhetoric of a Lesser Empire: Early Sung Relations with Its Neighbors," in Rossabi (1983). Wang Jiandong [gEHS] (1976). Sun Zi Bingfa sixiang tixijingjie ( S f f t f f i S i Wang Mingsun [jEE^ISI] (1985). "Zhongguo bei bian zhengce zhi chu qi xing cheng" , in GuojiZhongguo bianjiangxueshu huiyi lun wen Wang Sanxin [3Ezr.jffi] (1988). Mao Zedongjunshi sixiang yanjiu (^JWMM9^M m%\ Beijing. Wang Xianchen [3:015] and XuBaolin [fFiS#] (1985). Zhongguo gudai bing-shu za tan O f f f l S f t J * * * ^ ) . Beijing. Wang Yuquan [3E8fcifc] (1965). Ming dai dejun tun OHfiJ fi^ife). Beijing. WangZhen [ I i ] (812). DaoDeJinglun bingyaoyishu {MWMim^^W^)' Jiu chao ben, Taiwan National Library. Wang Zongxin [SE^rif] (1955). "Wang shi bing xue san shu" (flE-RJ^SHS), in Zhongguo bing xue da xi (t£ ffl^^P^:^), Book 2. Taibei.

REFERENCES

2 8 9

War Theory Research Department (1981). Sun Zi Bingfa xin zhu I Beijing. Wayman, Frank, J. David Singer, and Gary Goertz (1983). "Capabilities, Allocations and Success in Militarized Disputes and Wars, 1816-1976." International Studies Quarterly 27, pp. 497-515. . (1987). "Capabilities Allocations and Conflict Avoidance among Major Powers, 1816-1976." Ann Arbor, Mich., unpub. ms. Wei Huan [8t;J$|] (1541). "Jiu bian kao" (j~Li§L^j), in Bao Cunpeng [ f i S ^ ] > ed., Ming dai de bian fang (B^ft W&K?). Taibei, 1968. Wei Rulin [W&ikM] (1987). "Wo guo wu miao zhi ge yu bian yi zhi yanjiu" ( J i c H S ® i ; p ^ f | ^ S W 5 5 ) , in Wei Rulin, ed., Guofang lunwen xuanji < | | ) . Taibei, pp. 18-26. , ed. and anno. (1986). Huang Shi Gong San Luejin zhujinyi (il

Yi^H). Taibei. . (1988). Sun Zi Bingfa da quan (M^&T&j^lk). Taibei. Wei Rulin [St^Cll], and Liu Zhongping [SW 2 ?] (1985). Zhongguo junshi sixiang shi ( ^ S S W S S S j f i ) - Taibei. Weiss, Richard M. (1986). Managerial Ideology and the Social Control of Deviance in Organizations. New York. Wendt, Alexander (1992). "Anarchy is What States Make of It: The Social Construction of State Politics." International Organization 46:2, pp. 391-425. Wendt, Alexander, and Michael Barnett (1992). "The International System and Third World Militarization," unpub. ms. Westerlund, Gunnar, and Sven-Erik Sjostrand (1979). Organizational Myths. London. White, Stephen (1979). Political Culture and Soviet Politics. London. Whiting, Allen S. (1972). "The Use of Force in Foreign Policy by the People's Republic of China." Annals of the American Academy of Political and Social Science 402,

pp. 55-65.

. (1975). The Chinese Calculus of Deterrence. Ann Arbor, Mich. . (1989). China Eyes Japan. Berkeley. Wight, Martin (1991). International Theory: The Three Traditions, Gabriele Wight and Brian Porter, ed. Leicester. Wildavsky, Aaron (1985). "Change in Political Culture." Politics 20, pp. 95-102. . (1987). "Choosing Preferences by Constructing Institutions: A Cultural Theory of Preference Formation."'American Political Science Review 81 .i, pp. 3-20. Wilkenfeld, Jonathan, Michael Brecher, and Sheila Moser (1988). Crises in the Twentieth Century, vol. 2: Handbook of Foreign Policy Crises. New York. Williams, Michael C. (1992). "Rethinking the 'Logic' of Deterrence." Alternatives 17, pp. 67-93. Woodside, A. B. (1962). "Early Ming Expansionism, 1406-1427: China's Abortive Conquest of Vietnam." Papers on China 17, pp. 1-37. Worthy, Edmund (1983). "Diplomacy for Survival: Domestic and Foreign Relations of the Wu Yueh, 907-978," in Rossabi (1983). Wrightson, Margaret Tucker (1976). "The Documentary Coding Method," in Robert Axelrod, ed., The Structure of Decision: The Cognitive Maps of Political Elites. Princeton, pp. 291-332.

290

REFERENCES

Wu Han [Jg^] (1937). "Ming dai de jun bing" (miX^JM^l in Zhongguo shehui jingji shiji kan Of ffljft#SgS&HT!l) 5:2, pp. 147-99. Wu Jihua (1978a). "Ming dai jian guo dui wai de jiben taiduji juece" ( ^ f t ^ H S t ^ f W S ^ f i f i R ^ B t ) . Journal of Oriental Studies 16, pp. 184-93. - . (1978b). "The Contraction of Forward Defense on the North Chi Frontier during the Ming Dynasty." Papers on Far Eastern History 17, pp. 1-13. -. (1981). "Basic Foreign Policy Attitudes of the Early Ming Dynasty." Ming Studies 12, pp. 65-80. Wu Jiulong [ ^ A f i ] , ed. (1990). Sun Zixiao shi (JS-PK?f). Beijing. Wu Rusong [rj^nffi] (1988a). "Lun Liu Tao de junshi sixiang" (| 38), in Bingjia shi yuan (^^5&?S), vol. 1. Beijing, pp. 58-67. . (1989). Sun Zi Bingfa xin lun ( ^ i F ^ S f f l r o ) . Beijing. . (1990). "Jian lun gudai junshi moulue" (fStm^iX^V-M in Zhongguo junshi kexue ( f H ^ ^ f i f ^ ) , no.l. Beijing. , ed. (1988b). Zhongguo gudai bingfa jing cui lei bian (if H

H $ | ) . Beijing. , ed. and anno. (1983). Sun Zi Bingfa qian shuo \M). Beijing. Wu, Silas (1968). "Transmission of Ming Memorials and the Evolution of the Transmission Network, 1368-1627." T'oungPao 54, pp. 275-87. Wuthnow, Robert, etal. (1984). Cultural Analysis: The Work ofPeter L. Berger, Mary Douglas, Michel Foucault, Ju'rgen Habermas. Boston. Xie Guoliang [Hf BJ3:] (1990). "Lun Sun Zi de shan zhan sixiang" (IRS J^-p^HMBc JS38), in Sun Zixin tan (M~F$tM)' Beijing, pp. 33-48. Xin Jingling [^#MS], ed. and anno. (1985). Sun Bin Bingfa QM%^S). Taibei. Xu Baolin [If ® W (1985a). "San Lue de junshi sixiang" (HB§^j|lljl,S3S) in Man hua Zhongguo gudai junshi (ftfS'f BlSf^Jp-flQ- Beijing, pp. 56-62. . (1985b). "Wu Jing Qi Shu chu yi" C K S ^ l t l g i i ) in Man hua Zhongguo gudai junshi ( S l S f H ^ f t S f l - ) - Beijing, pp. 87-93. . (1987). "Sun Wu he Sun Bin bingfa" (JSiKftlJ^JStrftffiX in Academy of Military Sciences Planning and Organization Department, ed., Junshi sixiang yu junshi lishi (W-W-^M^M^-M^.), vol. 3. Beijing. . (1988). Zhongguo bingshu zhijian lu (cf'(). Beijing. . (1990). Zhongguo bingshu tongjian Of | ). Beijing. , ed. and anno. (1986). San Lue qian shuo (jziffl&SK Beijing. Xu Peigen [£££§$t] (1983). Zhongguo guofang sixiang ( f H KJ ® 39)- Taibei. , ed. and anno. (1986). Tai Gong Liu Tao jin zhujinyi (^^7\*fS^Q:^B?0Taibei. Xu Shu'an [ff Hf 5£r] (1985c). "Ming chao de ke ju xuan guan zhidu" HffMJg;) in Wen xian CStlR), no.3, pp. 256-69. Xu Yong [f^JS] (1989). WeiLiao Zi qian shuo (ftf JS^PStl©- Beijing. Xu Zihong [f^-pSK] (1990). "Lun Sun Zi Bingfa de jiben sixiang yu que xian" ( - ^ ^ ^ ^ j ^ ^ g ^ l S I ^ ^ ) , in Qiu shi (>}