Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The Japan-South Korea history issue: the emotional outcome of trusting acts and met/unmet alignment security expectations
(USC Thesis Other)
The Japan-South Korea history issue: the emotional outcome of trusting acts and met/unmet alignment security expectations
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE UNIVERSITY OF SOUTHERN CALIFORNIA
THE JAPAN-SOUTH KOREA HISTORY ISSUE: THE EMOTIONAL OUTCOME OF
TRUSTING ACTS AND MET/UNMET ALIGNMENT SECURITY EXPECTATIONS
A DISSERTATION SUBMITTED TO THE FACULTY OF THE USC GRADUATE SCHOOL
IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
DEPARTMENT OF POLITICS AND INTERNATIONAL RELATIONS
BY
DAVID A. WALKER
LOS ANGELES, CA
DECEMBER 2013
Copyright 2013 by David A. Walker
All rights reserved
To Alison
- iii -
CONTENTS
Chapter 1
Why Historical Animosity in Japan-South Korean Relations Matters
I. The Puzzle------------------------------------------------------------------------------- 1
II. History in Context--------------------------------------------------------------------- 10
A. IR Theory and the Japan-Korea History Issue----------------------------------------- 10
B. Modeling the Japan-Korea History Issue----------------------------------------------- 18
CHA’S QUASI-ALLIANCE MODEL------------------------------------------------------- 18
YOON’S NET THREAT THEORY MODEL----------------------------------------------- 20
LIND’S APOLOGY POLITICS MODEL--------------------------------------------------- 23
KOO’S COMPLEX INTERDEPENDENCE MODEL--------------------------------------- 27
CONCLUSION REGARDING THE FOUR MODELS-------------------------------------- 32
III. Research Strategy-------------------------------------------------------------------- 33
A. Dependent Variable: The History Issue------------------------------------------------ 33
Chapter 2
Introduction to Theory: The Role of Expectations and Emotions
I. Outline of Theoretical Chapters----------------------------------------------------- 57
II. Expectations and Emotions are Everywhere-------------------------------------- 60
A. Emotions and Expectations in IR Theory---------------------------------------------- 61
NEOREALISM------------------------------------------------------------------------------ 61
HUMAN REALISM AND CLASSICAL SCHOLARSHIP---------------------------------- 63
CLASSICAL LIBERALISM AND NEOLIBERALISM------------------------------------- 68
CONSTRUCTIVISM------------------------------------------------------------------------ 78
B. Emotions and Expectations in Models of Japan-Korea Relations--------------------- 86
CHA’S QUASI-ALLIANCE MODEL------------------------------------------------------- 86
YOON’S NET THREAT THEORY MODEL----------------------------------------------- 94
LIND’S APOLOGY POLITICS MODEL--------------------------------------------------- 99
KOO’S COMPLEX INTERDEPENDENCE MODEL--------------------------------------- 104
SUMMARY OF MODELS ON JAPAN-KOREA RELATIONS--------------------------- 107
III. Security Expectations in Japan-Korea Relations ------------------------------ 109
Chapter 3
Theoretical Model #1: Delimited Balance of Threat Model
I. Overview-------------------------------------------------------------------------------- 119
II. Neorealist Alignment Politics------------------------------------------------------- 121
A. Delimiting Walt’s Balance of Threat Theory------------------------------------------ 121
THE FLOW OF EMOTIONS IN DELIMITED BALANCE OF THREAT THEORY------ 130
- iv -
B. Definitions Based on Snyder’s Alliance Politics--------------------------------------- 133
C. Causal Mechanism--------------------------------------------------------------------- 140
FEAR AND INSECURITY ALMOST NEVER LEAD TO POLITICIZATION OF
IDEAS---------------------------------------------------------------------------------------------------- 141
MATERIALLY COERCED LOYALTY IN NEOREALIST ALIGNMENTS-------------- 142
IDEOLOGICALLY COERCED LOYALTY IN NEOREALIST ALIGNMENTS---------- 148
THE POLITICIZATION AND DEPOLITICIZATION OF IDEAS IN NEOREALISM---- 149
MATERIAL COOPERATION IN NEOREALIST ALIGNMENTS------------------------ 157
WHY NEOREALIST FEAR DOES NOT IMPROVE IDEATIONAL RELATIONS
DIRECTLY--------------------------------------------------------------------------------- 158
Chapter 4
Theoretical Model #2: The Common Sense Social Emotions Model
I. Overview: Putting Social Emotions before Material Emotions--------------- 161
II. Social Emotions in Material Theories and Models----------------------------- 166
A. Neorealist and Non-Neorealist Emotions--------------------------------------------- 167
B. Real Neorealism and Social Emotions: The Example of Neorealist Anger--------- 172
C. Social Emotions in the Five Material Models----------------------------------------- 177
CHA’S QUASI ALLIANCE MODEL----------------------------------------------------- 177
YOON’S NET THREAT THEORY------------------------------------------------------- 179
LIND’S APOLOGY POLITICS MODEL------------------------------------------------- 181
KOO’S COMPLEX INTERDEPENDENCE MODEL------------------------------------- 183
THE DELIMITED BALANCE OF THREAT MODEL------------------------------------ 185
III. Common Sense Social Emotions: Minimal Theory and Fairly Robust
Causal Mechanism---------------------------------------------------------------------- 190
A. The Interests--------------------------------------------------------------------------- 191
B. The Emotions-------------------------------------------------------------------------- 200
THE RATIONALE BEHIND WEAPONIZED IDEAS------------------------------------ 200
CHA’S AND SNYDER’S WEAPONIZED IDEAS--------------------------------------- 202
COOPERATION & CONFLICT: THE ROLE OF MATERIAL FEAR AND
IDEATIONAL ANGER & VENGEFULNESS-------------------------------------------- 204
IDEATIONAL COOPERATION: THE ROLE OF IDEATIONAL FEAR AND
GRATITUDE------------------------------------------------------------------------------ 207
Chapter 5
Theoretical Model #3: The Collective Self-Esteem Model
I. Overview------------------------------------------------------------------------------- 213
COOPERATION & CONFLICT----------------------------------------------------------- 214
THE SIGNIFICANCE OF NEW EXPECTATIONS--------------------------------------- 217
II. Static Versus Temporally Rich Security Expectations-------------------------- 219
A. Surprise and Uncertainty-------------------------------------------------------------- 219
- v -
B. The Logical Consequences of Expectational Staticity on Abandonment Fear------- 222
SNYDER’S TEMPORAL DILEMMA----------------------------------------------------- 223
CHA’S AND YOON’S STATIC & ASSUMED EXPECTATIONS------------------------ 232
C. Expectations and Actor Preferences--------------------------------------------------- 239
THE ROLE OF PRIOR EXPECTATIONS------------------------------------------------- 240
THE MEANING OF CHANGING EXPECTATIONS FOR MATERIALISM------------- 242
III. The Collective Self-Esteem Model------------------------------------------------ 245
A. The Interests and Emotions------------------------------------------------------------ 246
COLLECTIVE SELF-ESTEEM AND IN-GROUP PREFERENCE------------------------ 248
IDEATIONAL FEAR---------------------------------------------------------------------- 253
THE TRUSTING ACT I-------------------------------------------------------------------- 255
THE TRUSTING ACT II------------------------------------------------------------------- 260
THE SHORT-TERM OUTCOME OF HUMILATION------------------------------------- 262
THE LONG-TERM OUTCOME OF HUMILATION-------------------------------------- 264
Chapter 6
Statistical Analysis
I. Overview------------------------------------------------------------------------------- 265
II. Econometrics------------------------------------------------------------------------- 265
A. North Korea Policy Variable --------------------------------------------------------- 265
KOREA CLAUSE------------------------------------------------------------------------- 266
REPATRIATION OF KOREANS IN JAPAN TO NORTH KOREA---------------------- 267
TRADE WITH NORTH KOREA---------------------------------------------------------- 268
HUMANITARIAN FOOD AID TO NORTH KOREA------------------------------------- 270
NORTH KOREA POLICY VARIABLE--------------------------------------------------- 271
REGRESSION ANALYSIS OF NORTH KOREA POLICY IV AND COMPONENTS--- 272
B. North Korea Policy Variable in Neorealism------------------------------------------ 274
C. North Korea Policy Variable in Neoliberalism---------------------------------------- 275
III. Conclusion--------------------------------------------------------------------------- 278
Chapter 7
Conclusion------------------------------------------------------------------------------- 279
Appendices------------------------------------------------------------------------------ 283
Bibliography------------------------------------------------------------------------------ 289
- vi -
ILLUSTRATIONS
Diagram 2.1 Summary of Emotions in the Three Alignment Models after One State Hedges
Diagram 3.1 Walt’s Neorealism, Delimited and with No Alignment Politics
Diagram 3.2 Flow of System-Level Emotions to the Transnational Level
Diagram 3.3 Flow of Transnational-Level Emotions to the System Level
Diagram 3.4 The Intersection of Conflicting Security Interests and Ideological Politicization in
Neorealism: Jilted State Exhibits Loyalty
Diagram 3.5 Transnational Processes and Ideological Problems in Neorealist Alignments
Diagram 3.6 Transnational Politicization of Ideas within and between Aligned States
Diagram 3.7 Complementary Security Interest and Improved Ideological Relations in
Neorealist Alignments
Diagram 4.1 Security Interests Trigger Alignment Conflict via Ideological Social Emotions in
a Non-Emergency Material Security Environment
Diagram 4.2 Security Interests Trigger Alignment Cooperation via Ideological Social Emotions
in a Non-Emergency Material Environment
Diagram 4.3 An Immediate, Existential Material Threat to One State Leads to Cooperative
Material and Ideological Relations within an Alignment
Diagram 4.4 An Immediate, Existential Material Threat to Both Security Partners Leads to
Cooperative Material and Ideological Relations within the Alignment
Diagram 5.1 Security Interests Trigger Alignment Cooperation in a Non-Emergency Material
Environment
Diagram 5.2 Security Interests Trigger Alignment Conflict in a Non-Emergency Material
Environment
Diagram 5.3 Weak Expectations and Abandonment Fear in Temporally Static versus
Longitudinal Alignments: the Case of Japan and South Korea
Diagram 5.4 Expectational Change and Abandonment Fear in an Ongoing Alignment: the Case
of Japan and South Korea
Diagram 5.5 Passive or Active Emotions in the Three Alignment Models (Normal Crises)
Diagram 5.6 Security Interests Produce Ideational and Material Cooperation in the Collective
Self-Esteem Model: In-Group Dynamics, Short-Term Emphasis
Diagram 5.7 Security Interests Trigger Alignment Cooperation: Material and Ideational Fear
Emphases
Figure 1.1 Japanese Imperialism in Chosun Ilbo Headlines: 1960-2010
Figure 1.2 Dokdo, and Korean Residents of Japan
Figure 1.3 Japanese Textbooks
Figure 1.4 Comfort Women, and Yasukuni Shrine
Figure 1.5 Key Cooperative Moments
Figure 1.6 Directly Observable Conflict
Figure 1.7 Conflict, Indirectly Measured
Figure 1.8 Cooperation and Conflict
Figure 1.9 Korean Residents of Japan
Figure 1.10 Japanese Textbooks
Figure 1.11 Dokdo
Figure 1.12 Yasukuni Shrine
- vii -
Figure 1.13 Comfort Women
Figure 1.14 Directly Observed Conflict, Alternative Measure
Figure 2.1 DPRK and ROK GDP: 1970-2010
Figure 5.1 Fear of Abandonment in Temporally Static Alignments, 1
Figure 5.2 Fear of Abandonment in Temporally Static Alignments, 2
Figure 5.3 Fear of Abandonment in Temporally Rich Alignments
Table 1.1 Primary DV, DV1, DV2, & DV3
Table 1.2 Primary DV, DV4, & DV5
Table 1.3 Raw Chosun Ilbo Headlines & DV6
Table 1.4 Deflated Chosun Ilbo Healines & DV 6
Table 1.5 Chosun Ilbo Headlines and DV 1
Table 1.6 Chosun Ilbo Headlines & DV 7
Table 1.7 DV2, DV8, & DV 9
Table 1.8 Primary DV, DV 10, DV 4, & DV 11
Table 4.1 Four Types of Crises and their Outcomes in Two-Way Alignments
Table 6.1 Korea Clause
Table 6.2 Nationality of Japanese-Koreans
Table 6.3 Percent Change in Total Trade Over Year Prior
Table 6.4 Food Aid to North Korea
Table 6.5 Regression: North Korea Policy Components
Table 6.6 Regression: North Korea Policy IV
Table 6.7 Regression: Neorealism
Table 6.8 Regression: Neoliberalism
- viii -
ACKNOWLEDGMENTS
This dissertation is the product of many years of interaction with a substantial number of
people, both inside and outside of academia. At USC, Dr. David Kang served patiently as my
dissertation chair. His feedback drove me to revisit my research methodology, and brought me
to embark on a quantitative exploration of Japan-Korea relations that I otherwise never would
have considered. As dissertation committee members, Dr. Jacques Hymans and Dr. Peter
Carnevale provided thoughtful critiques and provided new insights. Dr. Dan Lynch played an
important role not only as an informal member of the committee, but also as a mentor. Dr.
Jarrod Hayes, once my classmate at USC, generously read and marked up an early draft of my
dissertation. Dr. Yoo-Il “Joseph” Bae, another USC POIR alumni, helped translate exceptionally
difficult words and terms. My dissertation would have suffered for the worse, if not for the
involvement of these incredible scholars. Naturally, whatever flaws my project does possess are
due to my shortcomings alone.
I did not develop interests in East Asia in a vacuum. While working on my BA and MA
degrees at UMBC, Dr. Warren Cohen quickly converted me into fan of studying East Asian
diplomatic history. Dr. Constantine Vaporis, who remains an important mentor as well as a
friend, was similarly critical in drawing my attention toward Japan, which surely has the most
interesting pre-modern history of any state.
Whereas scholars helped nurture my fondness for international politics and Japanese
history, “average” Koreans drew me to learn more about their incredible country. Mr. Song
Chang-Min and Ms. Jeon Ji-Ji became life-long friends, seemingly from the very first seconds
they invited my wife and me to join them on their waterproof picnic mat in a sprinkler-soaked
- ix -
park in Daegu. The vacations we took with this amazing couple along South Korea’s southern
and eastern coasts, as well as in Jeju Island, remain most-cherished memories.
As a USC graduate student, I benefited from the amazing competence of Ms. Veri
Chavarin, simply the best and most friendly Graduate Advisor possible. Ms. Joy Kim, the
Curator of the Korean Heritage Library, as well as the librarians at the Von KleinSmid Center
Library for International and Public Affairs, likewise proved inexhaustible and eager to tackle
any task. Mr. Thomas Handel, a neighbor and family friend from Maryland, saved me many
hours by generously sharing his Excel expertise with me over email.
My parents deserve special mention. They provided more support than any child could
reasonably expect. Without their open-ended and unconditional encouragement, understanding,
and love I never would have finished this project. My children, whose rate of learning about the
world remains a source of wonder, have been an incredible and necessary fount of meaning and
joy. And humor: Liam, I still chuckle whenever I recall how you hoped I could use real
weapons at my dissertation defense. Aria, I wish you all the luck in the world with your dream
of becoming a princess who does dentistry on the side.
Finally, but most importantly: No matter how many times I write and delete them, there
are no sentences adequate to the task of properly acknowledging Alison. How else to say it?
She is on every page and within every word. Without Alison, nothing.
- x -
ABSTRACT
Expectations and emotions matter in international relations. In this dissertation I argue
Japan’s and South Korea’s respective alignment security expectations play a critical role in
exacerbating and ameliorating the “history issue” marring their bilateral relationship. More
specifically, Japan and the ROK each expect that the Other will uphold the security interests of
the Self when it comes to establishing North Korea policy. During moments in which their
North Korea policies are aligned, the quasi-allies tend to enjoy warm ideological relations.
When their Pyongyang policies diverge, however, ideological affairs within the Japan-Korea
alignment suffer.
The goal of my research is two-fold. First, I aim to quantify the influence on the history
issue of various phenomena such as North Korea policy, elections, trade dependence, and the
strength of the US commitment to East Asia. Of the nine independent variables I analyze in
various theoretically informed configurations, only two return statistically significant results:
ROK National Assembly elections and North Korea policy convergence/divergence. Each
substantially influences the level of historical animosity.
My second objective is to explain how Seoul’s and Tokyo’s North Korea policies lead to
historical animosity. While I deduce several theoretical models to explain causality, I find only
the third one—the collective self-esteem model—adequately explains all facets of the
relationship between alignment security expectations and ideological cooperation and friction.
Specifically, prior ideological divide means Japan and South Korea must undertake a
considerable trusting act in order to successfully coordinate alignment security policy. Warm
ideological relations emerge through the ceremonies and promulgations associated with the
trusting act.
- xi -
Regrettably, the security expectations fostered via the trusting act are unrealistic, and
with seeming inevitability Japan’s and South Korea’s Pyongyang policies diverge. Ideological
relations invariably deteriorate, because of the humiliation suffered by those vested materially
and emotionally in the trusting act. In order to diminish the humiliation and preserve their
individual and collective self-esteems, these men and women disparage and denigrate the Other.
While disparagement and denigration preserve the moral superiority and trustworthiness of the
Self over the Other, these acts also directly cause bilateral ideological friction and a deterioration
of the history issue.
- 1 -
Chapter 1
Why Historical Animosity in Japan-South Korean Relations
Matters
I. The Puzzle
Japan and South Korea
1
share one of the most enigmatic relationships in the international
system. Despite compelling long-term incentives to cooperate, problems related to Japan’s
colonization of Korea in the first half of the 20
th
century emerge often, dampening hopes the two
states can ever escape their past and develop more normal ties. At its most politicized moments,
the “history issue” even seemingly holds the potential to upend the overall relationship. In recent
years this possibility was most obvious in 2006 when, despite the need for close cooperation in
responding to North Korea’s ongoing nuclear program, South Korean President Roh Moo-Hyun
ordered approximately 20 gunboats to forcibly turn back, if necessary, a Japanese scientific
vessel charged with carrying out a survey of the ocean floor around the disputed
Dokdo/Takeshima
2
Islands. While cooler heads fortunately averted an even more serious crisis,
the regularly scheduled Japan-Korea bilateral summit meetings that historical controversies had
first derailed in 2005 remained canceled until the inauguration of a new Korean administration in
early 2008. Unfortunately Lee Myung-Bak, the new ROK
3
President, felt compelled to call off
1
In order to promote uniformity and, more importantly, impartiality, I alphabetize proper nouns when referring both
to Japan and South Korea and contested geographical entities.
2
To maintain impartiality I always refer to disputed geographical features by both their Japanese and Korean names.
3
I call South Korea the ROK (Republic of Korea), Korea, the South, and South Korea, among other terms.
Similarly, when discussing North Korea, I utilize the DPRK (Democratic People’s Republic of Korea), the North,
and North Korea. I never refer to North Korea simply as Korea, though I would if the DPRK were a primary focus
of my research.
- 2 -
the second “shuttle diplomacy” summit of 2008, following a renewed Japanese claim to the
islands.
History is not merely reliably troublesome, however: By some measures historical
animosity has actually worsened over the past 50 years. South Korea’s decision to cancel or
avoid summits from 2005 to 2008 represented the first ongoing boycott of highest-level, bilateral
meetings since the history-plagued normalization talks of the 1950s and early to mid 1960s.
4
Similarly, Korea’s explicit show of force and accompanying threat of force in the lead-up to
Japan’s planned survey of Dokdo/Takeshima waters marked a high point of conflict not equaled
or surpassed since small-scale armed clashes took place in 1953 and 1954.
5
That the events of
2005 to 2008 happened during a period when maximum coordination by these key American
allies was needed to promote a best outcome to North Korea’s nuclear challenge is truly
remarkable. In all, it suggests that even very serious security matters can be made secondary
when historical animosity spirals out of control.
The general trend of deepening historic animosity and the peak friction of the mid 2000s
are particularly puzzling when contrasted against the preceding brief period of exceptionally
warm ties. For it was only at their summit in 1998 that Japanese Prime Minister Obuchi Keizo
and ROK President Kim Dae-Jung made astonishing concessions about the past, including both
the formal giving and official acceptance of the most sincere, bilateral, state-level apology to
date, and South Korea’s decision to open its markets to Japanese cultural imports. Amidst the
optimism engendered by the summit and in the commitments honored over the following two
4
“Highest-level” refers to the maximum level of formal, bilateral talks technically possible at a given time. By
definition, these meetings include only negotiations on normalization before 1965, and formal summits between
heads of state afterward.
5
Min-Kyo Koo rates the intensity of Dokdo/Takeshima conflict from 1953 to 2005. Note, however, that he employs
a slightly different scale than do I, as will be detailed later in this chapter. His overview of the island conflict is
quite good. See Min Kyo Koo, “Scramble for the Rocks: The Disputes over the Dokdo/Takeshima, Senkaku/Diaoyu,
and Paracel and Spratly Islands” (Ph.D. diss., University of California, Berkeley, 2005), 62-64, 88-139.
- 3 -
years or so, the two states seemed destined for a future in which the past would not play so
destructive a role. And yet it did.
Five distinct historical phenomena have been particularly salient over the past half
century: Dokdo/Takeshima, Koreans in Japan, Japanese textbooks, comfort women, and formal
visits to Yasukuni Shrine by sitting Japanese Prime Ministers. Of these, contestation over the
islands and concern about discriminatory treatment of the hundreds of thousands of Koreans
living permanently in Japan were evident in the early postwar period.
6
Problems regarding
textbooks revisions did not emerge until later, however, when rumors about new guidelines for
Japanese history textbooks began circulating in 1982.
7
Similarly, the plight of the comfort
women—women compelled to work as sex slaves for the Japanese military—remained a societal
secret in both Japan and Korea until brave Korean “grandmothers” began to speak out about their
experiences in fairly large numbers in the early 1990s.
8
Finally, visits to Yasukuni Shrine, where
Japan enshrines its war dead—including, since 1978, Class A war criminals of the colonial era—
became a very serious problem in bilateral ties only in the 2000s.
9
6
Koo notes that US negotiators may deliberately have left out Dokdo/Takeshima in order to improve Japan’s claims
and, therefore, defensive position in the event Communist forces emerged victorious in the Korean War. See Koo
“Scramble for the Rocks,” 91. For the Japanese and Korean politicking leading up to the 1951 Peace Treaty and for
early problems related to Koreans in Japan see Sung-Hwa Cheong, The Politics of Anti-Japanese Sentiment in Korea:
Japanese-South Korean Relations Under American Occupation, 1945-1952 (Westport, CT: Greenwood Press, 1991),
35-45, 57-75.
7
The episode is confusing: Claudia Schneider calls it a “canard.” The Japanese Education Ministry apparently did
not even intend to publish new guidelines that year. Nonetheless, the rumors triggered extensive discussion within
and between states in East Asia. Of great note, Koreans came to realize the extent to which the Japanese viewed
themselves as victims of the war, and not primarily as its perpetrators. Koreans also saw the extent to which
coverage of Japan’s past aggression in the Japanese press was “evasive.” See Claudia Schneider, “The Japanese
History Textbook Controversy in East Asian Perspective,” The Annals of the American Academy of Political and
Social Science 617 (May 2008), 109.
8
Most estimates of the number of women drafted by the Japanese military for sexual services range from 50,000 to
200,000. Korean women represented the majority, however. Comfort women may have tried to break their silence
sooner: Joo Hwang-Kum ( 주황금) contends her requests to the Korean government for compensation in 1966 were
suppressed. See Bang-Soon L. Yoon, “Imperial Japan’s Comfort Women from Korea: History & Politics of
Silence-Breaking,” The Journal of Northeast Asian History 7, no. 1 (Summer 2010): 7-9, 20.
9
Nakasone Yasuhiro’s visit to Yasukuni marked the first formally announced official visit of a Japanese Prime
Minister to the shrine. While China angrily denounced Nakasone’s visit, the South Korean government was
curiously quiet. The ROK protested forcefully only upon Prime Minister Koizumi Junichiro’s first formal visit in
- 4 -
Other facets also color the history issue. These include problems related to Koreans left
behind by the Japanese in Sakhalin, which USSR/Russia has controlled since 1945; Korean
survivors of the atomic bombings of Hiroshima and Nagasaki; and Korean lepers corralled into
terrible leper colonies in Japan. Troubling questions have also existed about the return of
cultural artifacts and the repatriation of the remains of Koreans who died while working in, or
soldiering for, Japan; the correct name for the East Sea/Sea of Japan; the possibility that postwar
Japan has been a neocolonial, mercantilist state seeking to dominate Korea’s economy; whether
Japan has apologized adequately and the extent to which new transgressions or discoveries about
the past merit renewed apologies; and, correspondingly, whether and how the Japanese should
offer more reparations. All of the above issues can become interrelated, of course. Visits to
Yasukuni Shrine can draw Korean attention to Japanese textbooks that gloss over the specifics of
the comfort women, “misname” the East Sea/Sea of Japan, and lay claim to Dokdo/Takeshima.
In turn doubts about the adequacy of past apologies and reparations can arise.
As it did from 1998 to 2008, historic animosity has varied greatly in its intensity since at
least 1960. Figure 1.1 suggests the extent to which the past can become politicized on any given
year. Comprised of Chosun Ilbo newspaper headlines containing the term “Japanese
Imperialism” ( 일제), it reflects with pretty good accuracy some of the peak years of conflict over
history.
10
2001. See Daiki Shibuichi, “The Yasukuni Shrine Dispute and the Politics of Identity in Japan: Why All the Fuss?”
Asian Survey 45, no. 2 (March/April 2005): 210-11.
10
For illustrative purposes only, I use “Japanese Imperialism” because it is a fairly issue-neutral term. In contrast to
a specific phenomenon like “comfort women,” it should provide more of an overall sense of the history issue.
Chosun Ilbo Archive, http://srchdb1.chosun.com/pdf/i_archive/ (accessed May 5, 2011).
- 5 -
Though history-driven passions obviously fluctuate, evidence indicates that some specific
issues have gotten markedly better and others more problematic. Figure 1.2 shows a
considerable decline in the number of Chosun Ilbo headlines pertaining to “Korean residents of
Japan” ( 재일 교포). In contrast, headlines containing the Korean word for “Dokdo/Takeshima”
( 독도) have increased manifold. The dramatic rise of the latter is somewhat surprising, given the
island dispute’s widely acknowledged status as one of the most problematic issues of the 15 or so
years leading up to the establishment of relations in 1965.
11
11
Koo, “Scramble for the Rocks,” 98-99; Brian Bridges, Japan and Korea in the 1990s: From Antagonism to
Adjustment (Cambridge, Great Britain: Edward Elgar Publishing Ltd, 1993), 11. Of course Dokdo/Takeshima was a
prime confounding component in the complicated question of territorial waters. Taken as a whole, the issue of
territorial waters was a chief reason—at least ostensibly—for the ongoing deadlock in the normalization talks. See
Tae-Ryong Yoon, “Fragile Cooperation: Net Threat Theory and Japan-Korea-U.S. Relations” (Ph.D. diss.,
Columbia University, 2006), 154, 279-280.
- 6 -
In comparison to the general idea of Japanese imperialism and the specific issues of
Korean residents of Japan and Dokdo/Takeshima, Chosun Ilbo headlines suggest that problems
related to revisions of “Japanese textbooks” (일본 교과서) have been extraordinarily focused in
their intensity. This can be seen in Figure 1.3.
- 7 -
Like Japanese textbooks, the phenomena of “comfort women” ( 정신대 and 위안부) and
“Yasukuni Shrine” ( 야스꾸니 and 야스쿠니) visits are both relatively recent phenomenon. Unlike
textbooks, however, the Chosun Ilbo headline counts shown in Figure 1.4 suggest their impact is
diffused more across time.
As outlined above, the volatile and episodic saliency of historic animosity is a defining
characteristic of Seoul-Tokyo relations. Notably, at times the past barely seems to matter. On
other occasions, however, it impedes cooperation on even the most important concerns. In this
dissertation I seek to explain why Japan-Korea relations suffer these wild gyrations. Not only do
I intend to clarify why history is sometimes good or bad, or somewhere in between, but why its
importance has not diminished over the course of the normalized relationship.
To answer these questions, I investigate a broad array of causal variables suggested by
International Relations (IR) theory. Investigating from 1960 to 2008, and using quantitative
methodology, I produce and analyze three simple regression models that purport to explain
between 21 to 24 percent of the total variance in my measure of the history issue—i.e., the
- 8 -
dependent variable (DV). Two coefficients return statistically significant results: Korean
Assembly elections and Japan’s and South Korea’s policies regarding North Korea.
12
Both are
associated with a marked rise in the politicization of the past, though divergence in North Korea
policy has a more robust correlation.
13
My main contribution to the model is an IV representing the overall conflict or
coincidence of Japan’s and South Korea’s respective policies toward North Korea. From a
foundation comprised of Japan’s level of support regarding the Korea Clause,
14
I flesh out this
security interest variable by measuring directly observable data on each state’s trade with, and
aid to, North Korea, as well as via data that clearly delineates diverging or converging North
Korean policies in the years before trade and aid began.
This research is important for a number of reasons. While it will surprise few that
Korean electoral politics at the Assembly level correlate positively with politicization of the past,
my investigation marks a first successful attempt to concretely measure the impact of this
commonly assumed variable in combination with key international factors. A more important
aspect of this study, however, stems from the data on Japanese and South Korean interests
regarding the DPRK. Not only is it remarkable how much one can explain about the nature of
affairs between Seoul and Tokyo by looking only at their North Korean policies, but this
discovery marks an excellent starting point to explore the question of how differing and
converging interests regarding a key security matter result in historic animosity.
12
Within the same Neoliberal model, Korean Assembly elections are statistically significant at a 0.04 level, while
North Korea policy is significant at a 0.001 level.
13
In the Neoliberal model, Beta scores are as follows: Assembly elections (0.31); North Korea policy (0.51).
14
The Korea Clause is comprised of official and unofficial public statements made by pertinent and very high level
Japanese leaders regarding the importance of the ROK to Japan’s security. Chong-Sik Lee gives an excellent
introduction to it. See Chong-sik Lee, Japan and Korea: The Political Dimension (Stanford: Stanford University
Press, 1985), 68-104.
- 9 -
Rejecting the idea that material emotions such as material fear, material insecurity, and
desire for economic well-being can explain how conflicting policies politicize the history issue, I
argue ideational social emotions are responsible. I produce three deductive explanatory models
to explore the material and ideational emotions that matter. By blending key tenets of Social
Psychology and IR theories, I argue that Japan and Korea each develop alignment security
expectations—i.e., each trusts—that the Other in their bilateral relationship will uphold the
security interests of the Self. They enact trust when their security interests regarding Pyongyang
are aligned. At these moments the history issue is unlikely to be problematic. Since neither state
ever enjoys the opportunity to internalize the other’s security interests as its own, however, the
undertaken trust and expectations are unrealistic. They are inevitably frustrated by diverging
interests and policies concerning North Korea. Ideational social emotions emerge as security
expectations are disappointed. The reaction of each state to these social emotions causes, or at
least permits, the re-politicization of the history issue. Elevated levels of historic animosity
result.
15
I round out this chapter by giving in Section II an overview of how IR metatheories
explain the kind of conflict and cooperation represented by the history issue, and how four
scholars who explicitly or implicitly model the relationship account for its puzzling trajectories.
In Section III I give an overview of my dissertation, detail the history issue DV, and quickly
introduce some initial regression findings.
15
The history issue is of course not disabled or enabled solely through this process; it exists for a variety of reasons
and has its own self-perpetuating dynamics. Nonetheless, the extent to which it would otherwise exist is magnified
or diminished by group dynamics.
- 10 -
II. History in Context
A. IR Theory and the Japan-Korea History Issue
The historical animosity between Japan and Korea is ultimately a form of interstate
conflict. Likewise, when Seoul and Tokyo work together to improve the past, interstate
cooperation takes place. Unfortunately, IR theory is not very capable of explaining the kind of
conflict and cooperation evident in the history issue. When a given theory does explain well a
potential root cause of the phenomenon, the task of explaining how animosity comes about, or
when it will be particularly problematic, is beyond its scope. Variable confusion also confounds:
While IR theories often impart a sense that state interests shape feelings about the past, they
seldom can explain in a theoretically coherent manner how the history issue shapes state interests.
These shortcomings connote a certain importance to the case of Japan-Korea relations. For if in
its frequently changing tenor of its affairs scholars can clarify how interests impact history, and
how historical animosity shapes interests, they are likely to have something substantial to
contribute to IR theory itself.
Of the most relevant IR theories, Neorealism has the least to say about Japan and Korea’s
history issue.
16
Neorealism, or Structural Realism, argues that state interests are shaped by the
balance of power among the strongest states in an anarchic international system, and by the
inherent desire of all states to survive. While all varieties of Structural Realism claim states set
their survival strategies according to the international system’s overall environment, two main
schools of Neorealism differentiate the policies states will most likely enact to increase their
survival likelihood. Defensive Realists such as Kenneth Waltz posit states will generally be
16
It is worth noting that Defensive Neorealist Stephen Walt’s Balance of Threat Theory holds some potential for
explaining Japan-Korea relations. I detail his work in Chapter 2, in the context of the use of emotions by scholars
whose modeling utilizes Walt’s work on threat perception. Stephen Walt, The Origins of Alliances (Ithaca, NY:
Cornell University Press, 1987).
- 11 -
content fortifying and preserving that which they have.
17
In contrast, Offensive Realists like
John Mearsheimer contend states will always take advantage of opportunities to expand their
power, since more power is ultimately the only reliable way to increase the odds of survival.
18
Though its strengths are best suited for explaining great power behavior, Structural
Realism applied cautiously to the Japan-Korea middle power relationship does produce a few
basic predictions. It indicates that Seoul and Tokyo are most likely to cooperate on all matters
when each perceives, simultaneously, a threat to its respective survival. In contrast, problems
will tend to fester when both feel secure. Importantly, Japan’s and Korea’s perceptions of safety
will be shaped primarily by the balance and concentration of power in the international system.
Neorealism therefore overwhelmingly predicts cooperation in Japan-ROK affairs. For
both states have been strongly allied with the United States for more than half a century, and
have been pitted historically against the communist nations of East Asia. Despite the end of the
Cold War, they continue to face provocations from heavily armed and prickly North Korea,
which maintains considerable ties to China. Given the rise of China, Seoul and Tokyo have in
their backyards a major regional player and, soon, a global superpower. Under these conditions,
they have very strong incentive to work together to balance against the material capabilities of
their neighboring states and thereby increase the likelihood of their own survival. Neorealists
might therefore make a compelling case for ongoing Japanese-Korean cooperation. But they
would have to borrow from elsewhere to explain historical conflict, as they can offer no
compelling reason why states in such an unsafe security environment would allow it to
deleteriously impact relations.
17
Kenneth Waltz, Theory of International Politics (Reading, Mass: Addison-Wesley, 1979).
18
John Mearsheimer, The Tragedy or Great Power Politics (New York: W.W. Norton & Company, 2001).
- 12 -
In contrast with Structural Realism, Classical Liberalism suffers a two-fold problem
when explaining the history issue. First, its arguments that economic interdependence and
political democratization align interests and prevent conflict mean it cannot explain moments of
cooperation during the early years of normalized relations.
19
At that time, Korean autocrats
controlled government. And trade with Japan, though rising quickly, was very low by modern
standards. Second, conflict in recent decades is beyond Classical Liberalism’s purview. Not
only do both states now enjoy firmly entrenched democratic political systems, but economically
they are interdependent, capitalist, free-market, globalized powerhouses. Given that these
complementary political and economic trends have deepened, especially since the late 1980s,
each passing year should have resulted in more closely aligned Japan-ROK interests, and thus
good bilateral relations.
20
Classical Liberalism would thus be very useful for explaining a
relationship in which initially high levels of hostility about the past diminished steadily and
considerably, especially after Korea democratized in 1987. It cannot account, however, for the
real-world actuality of history’s zigzagging saliency and overall tenacity.
Other theoretical approaches, such as Neoliberal Institutionalism and Constructivism, are
equipped far better to explain why the relationship experiences problems. But they can only
problematically predict when relations will sour. Moreover, they contain no theoretically
satisfactory mechanism to explain how material interest divergence might lead to animosity
regarding non-material matters like history.
19
For good introductions to some of the main tenets of Classical Liberalism, see Bruce M. Russet, Grasping the
Democratic Peace: Principles for a Post Cold War World (Princeton: Princeton University Press, 1993); James Lee
Ray, “Does Democracy Cause Peace?” Annual Review of Political Science 1 (1998): 27-46; John R. Oneal and
Bruce M. Russett, “The Classical Liberals Were Right: Democracy, Interdependence, and Conflict, 1950-1985,”
International Studies Quarterly 41 (1997): 267-284; and Dale C. Copeland, “Economic Interdependence and War:
A Theory of Trade Expectations,” International Security 20, no. 4 (Spring 1996): 5-41.
20
For a good overview of Korea-Japan economic interdependence through the early 1990s, see Bridges, Japan and
Korea in the 1990s, 87-109. For a more recent take see Samuel Kim, The Two Koreas and the Great Powers (New
York: Cambridge University Press, 2006), 203-222.
- 13 -
Both theories highlight that friction in bilateral affairs results whenever disincentives
overpower incentives for cooperation. For Neoliberals, these are almost entirely material: Japan
and Korea cooperate for security and economic reasons.
21
Unfortunately, collaboration is
difficult because there are no adequately powerful institutions at the regional, subregional, or
bilateral levels to “provide information, reduce transaction costs, make commitments more
credible, establish focal points for coordination, and in general facilitate the operation of
reciprocity.”
22
Under the above conditions, the considerable incentives for cooperation are often
overwhelmed by concerns about “cheating” and a general lack of trust. Complex
interdependence may make matters worse, given that domestic nationalist groups and legal and
bureaucratic procedures in both countries have emerged as major challenges to the cooperative
agendas periodically set out by leaders in Seoul and Tokyo.
23
In contrast with Neoliberal theorists, Constructivists view competing national identities
within and between each state as setting the tone of relations. Some identities are
complementary: Japan and Korea are both “trading states” and so leaders and citizens in each
21
See, for example, Kenneth Oye, “Cooperation under Anarchy: Cooperation and Strategies,” World Politics 38, no.
1 (October 1985): 1-24; Robert Axelrod and Robert Keohane, “Achieving Cooperation under Anarchy: Strategies
and Institutions,” World Politics 38, no. 1 (October 1985): 226-254. For a discussion about the relationship between
Neoliberalism and Neorealism, see, John J. Mearsheimer, “The False Promise of International Institutions,”
International Security 19, no. 3 (Winter 1994-1995): 5-49.
22
The quote is from Robert O. Keohane and Lisa L. Martin, “The Promise of Institutional Theory,” International
Security 20, no. 1 (Summer 1995): 42. For an overview of Korea in a regional context, see Charles K. Armstrong et
al., eds., Korea at the Center: Dynamics of Regionalism in Northeast Asia (Armonk, NY: M.E. Sharpe, Inc., 2006).
ASEAN and its offshoot, ASEAN Plus Three, are probably the most important regional institutions in Asia.
Regarding an optimistic account of ASEAN in the Asia Pacific region, see Amitav Acharya, Constructing a Security
Community in Southeast Asia: ASEAN and the Problem of Regional Order (New York: Routledge, 2001); for a
more pessimistic view see Thomas Moore, “China’s International Relations: The Economic Dimension,” in The
International Relations of Northeast Asia, ed. Samuel S. Kim (Lanham, MD: Rowman & Littelfield Publishers, Inc.,
2004), 101-134, but esp. 116-126.
23
Bureaucratic textbook revision procedures and the outcomes of court cases regarding comfort women are good
examples of institutional processes at the domestic level that can damage Japan-ROK relations. For complex
interdependence, see Robert O. Keohane and Joseph S. Nye, Power and Interdependence, 2
nd
ed. (Glenville, IL:
Scott Foresman, 1989).
- 14 -
state view their security more in terms of economic well-being than political-military rivalry.
24
Democratic identities since the late 1980s also encourage cooperation, given that norms of
nonviolence, compromise, equality, and tolerance at the domestic level can influence
international interaction between democratic states.
25
Moreover, Japan’s modern identity as a
state obviously reluctant to use military force aids in promoting warm ties.
26
Greatly negating
these positive factors, however, are nationalistic identities in each state, as well as “diverse and
mutually reinforcing fears, prejudices and historical animosities.”
27
Particularly deleterious are
conflicting historical memories, as well as collective Japanese “aloofness” and Korean
“resentment” about issues stemming from Japan’s colonization of Korea.
28
Both theoretical camps agree that serious security challenges would quickly bring both
states to a cooperative footing. In the absence of a very obvious threat to both states, they predict
general bilateral friction caused either by the absence of strong institutions or the presence of
lingering historical issues, respectively. More than Neoliberals, Constructivists would be
sensitive to how emerging divergences in threat perceptions of other states might lead or
24
As “trading state” applies to East Asia, see Thomas U. Berger, “Power and Purpose in Pacific East Asia: A
Constructivist Interpretation,” in International Relations Theory and the Asia-Pacific, eds. G. John Ikenberry and
Michael Mastanduno (New York: Columbia University Press, 2003), 387-419, but esp. 392-398. Also see Peter J.
Katzenstein, “Taming of Power: German Unification, 1989-1990,” in Past as Prelude, History in the Making of a
New World Order, eds. Meredith Woo-Cumings and Michael Lorriaux (Boulder: Westview, 1993), 59-82. For a
general introduction to Constructivism, see Peter J Katzenstein, ed., The Culture of National Security: Norms and
Identity in World Politics (New York: Columbia University Press, 1996); and Emanuel Adler, “Seizing the Middle
Ground: Constructivism in World Politics,” European Journal of International Relations 3, no. 3 (1997): 319-363.
25
Thomas Risse-Kappen, “Collective Identity in a Democratic Community: The Case of NATO,” in The Culture of
National Security: Norms and Identity in World Politics, ed. Peter J. Katzenstein (New York: Columbia University
Press, 1996), 357-399, esp. 366-367; Emanuel Adler, “Seizing the Middle Ground,” 347.
26
This is true to the extent that a Japan more willing to use military force in world affairs would apparently be more
worrisome to Korea. Regarding Constructivist arguments for Japan’s reluctance to use military force, see Thomas U.
Berger, “Norms, Identity, and National Security in Germany and Japan,” in The Culture of National Security: Norms
and Identity in World Politics, ed. Peter J. Katzenstein (New York: Columbia University Press, 1996), 317-356.
Katzenstein argues that the Japanese military has emerged as an “agent of nonviolence”; see Peter J. Katzenstein,
Cultural Norms & National Security: Police and Military in Postwar Japan (Ithaca, NY: Cornell University Press,
1996).
27
Berger, “Power and Purpose,” 392-393.
28
Ibid., 403-404; For a more thorough analysis of the society-wide history problem see Alexis Dudden, Troubled
Apologies Among Japan, Korea, and the United States (New York: Columbia University Press, 2008).
- 15 -
contribute to friction during moderate security crises.
29
Nevertheless, material incentives and
compatible identities do exist and, under the right combination of leadership in each country,
bilateral warmth and remarkable cooperation is possible. Intense conflict is always close by,
however, as multiple kinds of domestic phenomenon can emerge and overwhelm the identities
and material incentives promoting cooperation at any time. For both theoretical schools,
domestic factors are therefore the most important forces improving or exacerbating relations, and
shifting the tenor from a predicted state of moderate friction.
Neoliberalism and Constructivism are therefore compelled to take a remarkably similar
approach to explain phases of warm cooperation and heightened friction in Japan-ROK affairs.
Namely, they have to utilize a general historical narrative to examine how developments at the
domestic level unfold and become major factors causing amicability or hostility. But this is
already the method used by almost every scholar who examines ties between Seoul and Tokyo.
And, indeed, the domestic historic approach, combined with basic insights from Neorealism,
Neoliberalism, and/or Constructivism about cooperation or friction-inducing factors at the
international level, can result in a compelling narration of events. Unfortunately, the narrative
approach is not at all rewarding in terms of discerning generalizable causal phenomenon. A
handful of examples from some of the literature that seeks to explain changes in the temperament
of Japan-Korea affairs, but which is not theoretically rigid or argumentatively focused, highlight
the weakness of the approach.
30
29
Changes of threat perception that take place without changes in the material balance of power, offensive capability,
or geography cannot be explained by Neorealism. It does not illuminate, for instance, diverging Japanese and South
Korean perceptions of the North Korean threat in the early and mid 2000s, since both states could be targets for
North Korean nuclear and missile attacks. For an entertaining discussion about how Realism cannot explain threat
perception, see Jeffrey W. Legro and Andrew Moravcsik, “Is Anybody Still a Realist,” International Security 24, no.
2 (Autumn 1999): 5-55.
30
Later in this section, I discuss scholarly works that explain Japan-Korea relations in a far more theoretically
consistent manner.
- 16 -
In examining why Korea and Japan were able to normalize their relationship in 1965 but
not before, Chong-Sik Lee gives primary causal power to the combination of mutual economic
incentives, the political rise in Korea of Japanophile Park Chung-Hee, and intense US pressure.
31
Brian Bridges argues that the sudden political transition from Park to the more “Korean product,”
Chun Doo-Hwan, caused bilateral friction in 1980. Textbook revision issues in Japan contributed
to problems, and relations were exceptionally strained until newly elected Japanese Prime
Minister Nakasone Yasuhiro exhibited considerable leadership by reaching out to Chun in
1983.
32
For Lam Peng Er, relations reached a post-normalization high in 1998 because of the
leadership, unique character, and background of Korean President Kim Dae-Jung. Not only were
the Japanese “assured by Kim’s image of being a man of principle,” but they viewed Korea
increasingly favorably because of its solidified democracy and high level of economic
development. Japanese concerns about a rising China, and Kim’s need for support from Japan
for his North Korean “Sunshine Policy,” also led to cooperation and warmth.
33
Samuel Kim
argues that relations began to fall apart in 2001, however, in response to “the lack of sustained
leadership.” Thus Japanese textbook revisions, a Japanese Prime Ministerial visit to Yasukuni
Shrine, and a fisheries dispute were allowed to sour bilateral affairs. Fortunately, in response to
US pressure and the need to coordinate after the US launched its War on Terror, Japanese Prime
Minister Koizumi Junichiro undertook a constructive leadership role that improved relations.
34
Byung-Chul Koh agrees that sustained leadership by both sides prevented history-related
31
Lee, Japan and Korea, 43-67.
32
Chun not only arranged for Japan’s favorite Korean democracy proponent, Kim Dae-Jung, to be sentenced to
death (later commuted), but abrasively demanded a $6 billion development loan, tying it explicitly to Korea’s role in
defending Japan and the inadequacies of economic aid given as part of the 1965 normalization treaty. Bridges
credits Nakasone with setting a positive tone for relations through the 1980s and beyond. See Bridges, Japan and
Korea in the 1990s, 14-17. Lee’s more thorough accounting of these events is almost identical to Bridges’. See Lee,
Japan and Korea, 105-164.
33
Lam Peng Er, “The Apology Issue: Japan’s Different Approaches Toward China and South Korea,” American
Asian Review 20 (Fall 2002): 32, 44-46.
34
Kim, The Two Koreas and the Great Powers, 190-191, 202-203.
- 17 -
problems from seriously impacting relations from 2001-2004. By 2005, however, domestic
forces in Japan triggered tremendous bilateral friction over claims to Dokdo/Takeshima and
textbook revisions.
35
As suggested above, the historical narrative approach clarifies that Seoul and Tokyo are
most likely to cooperate bilaterally under the following conditions: 1) a leader emerges who is
willing to reach out to the other side; 2) there are complementary economic and/or security
reasons to overlook differences; and 3) the US is applying considerable pressure. Conversely,
tension is most likely when 1) various domestic-level forces or processes politicize historical
issues related to Japan’s colonization of Korea; and 2) there is a Korean leader who does not
have deep ties to Japan.
However interesting factually, in terms of theoretical generalizability the above
conclusions ask more questions than they answer. Most importantly, they lead to questions
about how these phenomena are related: Are all of the conditions causal, independent variables?
If so, which are the most important? If not, which are dependent variables, and how do they
relate to the independent variables? Are some variables permissive or intervening, rather than
directly causal? And what feedback effect, if any, do dependent variables have on independent
variables?
These are not questions that the historical narrative—and thus Neoliberalism and
Constructivism—can easily answer regarding Japan-Korea ties. They fail because each shift in
temperament within bilateral affairs is marked by phenomena too particularistic to be applied
across cases. Moreover, when similarities do exist, such as the cooperative periods in 1965 and
1998, not only are important causal elements missing that were deemed essential before—e.g.,
35
Byung Chul Koh, Between Discord and Cooperation: Japan and the Two Koreas (Seoul: Yonsei Univeristy Press,
2007), 466-476.
- 18 -
US pressure in 1965—but the narrative approach cannot weigh, within a theoretically coherent
framework, the relative importance of causal similarities.
B. Modeling the Japan-Korea History Issue
Precisely because Neoliberalism and Constructivism, as well as Neorealism and Classical
Liberalism, cannot explain Seoul-Tokyo affairs satisfactorily, scholars have attempted to develop
theoretically coherent models to clarify it. Unfortunately, these models have been outdated by
events, or are too confusing, predetermined, or underwhelming to explain the bilateral
relationship satisfactorily.
CHA’S QUASI-ALLIANCE MODEL
Victor Cha’s asymmetrical quasi-alliance model remains the dominant explanation for
Japan-Korea ties.
36
Regarding the Cold War era, Cha examines affairs between the two states
within the context of their respective security relationships with the United States. Within this
“quasi alliance,” Japan and Korea suffer from asymmetrical abandonment and entrapment fears,
as the ROK worries that more powerful Japan will not honor its commitments, and Japan fears
that it will become embroiled in a conflict on the Korean Peninsula. The imbalance contributes
to friction between the two states, as Korea gets upset when Japan hedges over its security
commitments.
37
Asymmetrical relationships between quasi allies like Japan and Korea are heavily
influenced by their common ally, in this case the United States. Whenever America shows a lack
of commitment or resolve in East Asia, the quasi alliance model predicts Seoul and Tokyo will
undergo a period of cooperation. Mutual support emerges because Japanese and Korean leaders
realize their collaboration lowers the perceived costs to the US of remaining involved in East
36
Victor D. Cha, Alignment Despite Antagonism: The US-Korea-Japan Security Triangle (Stanford: Stanford
University Press, 1999).
37
Ibid., 36-43.
- 19 -
Asia. Conversely, whenever the US is strongly and obviously committed in the area, frictions
caused by the asymmetrical entrapment and abandonment fears discussed above rise to the
surface.
38
Though Cha ably explains all four periods of cooperation or friction in Korea-Japan ties
from 1969 through the 1980s, his model has limited value outside of those years. He can only
awkwardly explain, for example, how concord on the controversial 1965 Normalization Treaty
was achieved.
39
Given the incredibly high level of US regional involvement at the time, Cha’s
model should lead readers to expect conflict, not normalized relations.
40
More problematic,
however, is the collapse of the abandonment-entrapment mechanism following the end of the
Cold War, as Japan worried more than South Korea about the rise of China and North Korea’s
nuclear and missile programs, and Seoul confidently took the lead in implementing a conciliatory
North Korean policy.
The fact that historical animosity continues to plague the relationship in the absence of a
key element of Cha’s model suggests strongly that abandonment-entrapment may have been an
unnecessary component to begin with. It always did sit a little uneasily as a dominant
mechanism, a fact reflected by Cha’s constant need to explain abandonment and entrapment in
terms of each state’s respective threat perceptions. From 1972 to 1974, for instance, US détente
with the Soviet Union did not result in warm ties between Korea and Japan as the straight model
predicts. Japanese security consolidation vis-à-vis the Communist states instead led to security
38
Ibid., 46-57.
39
Yoon makes this same point. See Tae Ryong Yoon, “Fragile Cooperation: Net Threat Theory and Japan-Korea-
US Relations” (Ph.D. diss., Columbia University, 2006), 40-41.
40
Cha reports considerable American pressure on its allies, regarding signing the treaty. See Cha, Alignment
Despite Antagonism, 29-34.
- 20 -
interest divergence with its Korean quasi-ally.
41
Disagreement with the Koreans in turn led to
the triggering of abandonment and entrapment fears and, therefore, animus.
42
Cha’s argument about détente seems reasonable enough, though it remains
underspecified why Japan and Korea did not desire to lower the perceived cost of US
involvement at a time of retrenchment. Was it really that obvious that it was best to bicker
publicly when the Americans were taking steps that could seriously lessen their military footprint
in East Asia? The only retort must be that Japan followed its “overall” interests, rather than the
interests argued by the model. But this is highly unsatisfactory, because in each of Cha’s four
cases relations between Seoul and Tokyo were good when overall interests were aligned and bad
when they were not. If so, why not just argue that the US plays an important role in shaping the
security interests of Korea and Japan, and allow overall interest coincidence or divergence to
predict the tenor of relations?
YOON’S NET THREAT THEORY MODEL
This is exactly the tact Tae-Ryong Yoon takes in his 2006 dissertation. Grounding his
model in Stephen Walt’s Balance of Threat Theory and Glenn Snyder’s Theory of Alliance
Politics, Yoon argues the overall security interests of Japan and Korea will dictate whether they
permit ongoing bilateral problems to negatively impact the relationship. Bilateral issues include
the disputed Dokdo/Takeshima Islands, which Yoon argues have been socially constructed into
an important Realist concern, and more obviously non-material historic contentions.
43
The
overall security milieu—or “net threat”—is the balance between the threat faced by Seoul or
Tokyo separately, minus the counter-threat. The counter-threat includes the resources each state
can bring to protect only itself, plus the safety granted by the level of commitment exhibited by
41
Seoul also tried to reach out to the Soviets, Chinese, and North Koreans. Unlike Japan it was rebuffed by all three.
42
Ibid., 100-140.
43
Yoon, “Fragile Cooperation,” 28-37.
- 21 -
the US at a given time. Japan and Korea are sensitive to the level of net threat: When it increases,
such as when the US commitment is in question, enemies are acting particularly aggressive, or
the ROK experiences considerable political instability, the quasi allies will actively work to
subdue problematic issues. An increase in net threat therefore results in warm relations.
Conversely, when the net threat decreases or remains unchanged neither government will move
to squelch the things that cause bilateral friction.
44
Whereas Cha provides strong incentives for both cooperative and obstinate behaviors, the
game theoretical aspects of Yoon’s model are more limited to positive inducements. Japan
cooperates with South Korea when net threat is rising, because it wants to prop up the declining
deterrent against North Korea. Realizing they will inevitably get sucked into any conflict on the
peninsula, the Japanese dread “deterrence failure.” Tokyo therefore suffers conflict entrapment
anxiety and acts to bolster deterrent in the face of increasing net threat. Seoul, on the other hand,
worries about abandonment by the Japanese. In the face of rising net threat the Koreans
cooperate over controversial issues, thereby encouraging Japanese involvement. Nonetheless,
but especially before the ROK began its remarkable economic development, Koreans fretted
about becoming overly dependent on, and losing autonomy to, Japan. Therefore, even under
rising net threat, Korea sometimes would be less cooperative than one would think.
45
Yoon examines two eras: American reengagement from 1950-1969, and American
retrenchment from 1969-1979. Holding American involvement more or less constant, he
subdivides each of the eras into smaller time periods based on whether net threat was increasing
or decreasing. The game theoretical mechanisms outlined above, plus other factors like total
44
Ibid., 14, 57-58, 85-87.
45
Ibid., 88-93.
- 22 -
enemy strength, source of threat, or South Korean political instability provides directionality to
the overall net threat level during the various time periods.
Since his model does not contain strong, built-in conflictual incentives, Yoon is forced to
adopt the history issue as both an intervening-amplifying and an intervening-causal independent
variable. These variables lead to friction when net threat is unchanged or falling. They act as
catalyst and amplifier for conflicts of interest over more material matters, such as North Korea or
China policies and Dokdo/Takeshima. Though this move satisfies the requirement that models
of Japan-Korea affairs need to explain linkages between material and ideological matters, it is
not unproblematic.
Arguably the biggest difficulty is that the model is too complex to be effective at
explicating causal mechanisms below the level of net threat. Its emphasis on historical matters
as intervening-amplifying and intervening-causal independent variables makes it difficult to see
how it differs from any of the domestic historic narrative approaches discussed earlier. True,
Yoon’s model would be more attuned to the impact of changes in overall threat levels, perhaps
like a good Constructivist or historic-narrative account. But there is something inherently
unsatisfactory in a predictive model that says, after allowing for overall interests, an historical
friction act directly leads to divergent policies regarding material interests and therefore permits
historical friction to worsen while, simultaneously, general and ongoing historical friction leads
to contradictory action caused by material interest divergence that was amplified by general
historical animosity. This multi-directional tangle is made even more complicated when one
realizes that antagonistic North Korea and China policies are not only intervening independent
variables, but must also surely be grounded in perceptions defining the first-tier independent
- 23 -
variable of overall net threat.
46
One might argue that Yoon successfully mapped a real-world
bilateral relationship. To the extent he succeeded, however, he obscured causality.
LIND’S APOLOGY POLITICS MODEL
Like Yoon, and to a much lesser extent Cha, Jennifer Lind relies heavily on Walt’s
balance of threat theory to ground her model.
47
Whereas Cha and Yoon sought to explain both
moments of cooperation and friction, however, Lind focuses only on the question of why Japan
and Korea never healthfully reconciled following World War Two. She argues the type of
remembrance a war-perpetrating society undertakes will be a major variable shaping whether its
erstwhile enemies perceive it to be an ongoing threat. Perception of threat, in turn, is the main
determinate of state reconciliation.
48
Lind examines Japan’s and Germany’s post-war
remembrances to test her theory. She divides each case longitudinally into early, middle, and
late post-war time periods. What emerges are two different narratives of how remembrance
contributed to low and high threat perceptions.
Though neither Germany nor Japan were apologetic in the early period—indeed each
country’s broader societies preferred to think of themselves as victims of their own wartime
leaders—only Japanese elites actively denied or justified the atrocities their state had committed.
In contrast German elites readily admitted to, without apologizing for, the bad things Germany
had done. So effective was unapologetic German contrition on France’s threat perception that
46
Yoon’s model is most clearly made when diagramed. See Ibid., 14.
47
Lind does not state her reliance on Walt. Nevertheless, as I explain here and in later chapters, her approach is
Neorealist in the same vein as Walt’s Neorealism. Ultimately, for Lind, ideational reconciliation remains unfulfilled
not because there has been inadequate ideational reconciliation in and of itself, but because angry Japanese
nationalism compels Seoul to worry about Japan’s intentions and military power. Inadequate ideological
reconciliation emerges from those material concerns.
48
Jennifer Lind, Sorry States: Apologies in International Relations (Ithaca, NY: Cornell University Press, 2008), 9-
23.
- 24 -
Germany became France’s closest ally by 1960. Unfortunately, states like South Korea
continued to fear and distrust Japan because of its denials and whitewashing.
49
In the middle period both Germany and Japan became apologetic. Germany not only
accepted its role in starting the war, but actively educated its population about German
aggression and atrocities. Furthermore, the German government greatly expanded its war
reparations program and publically brought many ex-NAZI officials to justice. Germany’s
straightforward apologies were therefore well received. Japan’s apologies, however, were
ineffective. They were vague and tepid and evoked angry outcry from Japanese nationalists,
albeit one mostly at the elite level. Therefore, in contrast to Germany’s society-wide
introspection, new textbooks in Japan began to play down its military aggression. Moreover,
instead of investigating war criminals, Japan enshrined even its worst offenders at Yasukuni
Shrine. Magnifying the offense to Korea, China, and other states, top-level ruling party leaders
publicly paid respects to all commemorated there.
50
These trends accelerated in the late post-war period, as German leaders offered even
more profuse apologies and reparations and, along with the German public, came to accept that a
broad swath of German society had taken part in and known about the Holocaust. Japan’s
apologies, though deepened and made more specific, were in this period drowned out
domestically by a broad, society-wide nationalist backlash. Visits to Yasukuni increased,
textbook disputes intensified, Tokyo refused to offer reparations to newly realized victims, and
the Dokdo/Takeshima Islands reemerged as a major stumbling block to good relations.
51
Lind draws several fascinating conclusions from the above cases: First, in the Franco-
German rapprochement of the 1950s, she finds reconciliation can take place without significant
49
Ibid., 29-32, 108-114, 124.
50
Ibid., 47-54, 126-135.
51
Ibid., 60-78, 143-149.
- 25 -
contrition. Apologies are therefore not necessary for reconciliation. Second, Japan’s experience
across all periods, but especially in the third, brings her to surmise that apologies can lead to a
nationalist backlash in the apologizing state. Hostile nationalism in turn fosters increased threat
perception among victims, and unfulfilled reconciliation. Consequently, apologies can be
counterproductive. Third, Lind concludes by studying all of Japan’s time periods that
unapologetic remembrances—i.e., whitewashing history and outright denials—fuel distrust of
intentions and harm reconciliation. This finding is most remarkable regarding Japan’s and
Germany’s early postwar years, since whitewashing and denial were the main apology-related
factors differentiating the two states at the time.
52
Lind’s project, however interesting, contains serious shortcomings. Most stem from
biases that emerge from her case selection, a problem she unsuccessfully tries to dispatch with a
few pages here and there.
53
Unfortunately, France-Germany and Japan-Korea remain cases too
diverse from which to draw claims without controlling for, in most careful detail, the chasm that
differentiates them. One brief example about Germany’s and Japan’s respective occupations
should suffice to highlight the point. As Lind briefly acknowledges, allied occupation policy
was far more forceful regarding getting Germans to recognize the Holocaust and the basic fact of
Nazi aggression than US occupation policy was in making the Japanese accept their atrocities
and role.
54
Importantly, Japanese wartime elites continued to wield considerable influence in the
postwar state, in contrast with their purged Nazi peers.
55
Lind is therefore not actually answering
her intended general question about the impact of apologies per se. In the case of Germany, she
is exploring whether a state that has been thoroughly restructured following defeat can gradually
52
Ibid., 188-189.
53
Ibid., 23-25, 188-189.
54
Ibid., 29-32, 105-107.
55
See especially, Ibid., 32.
- 26 -
come to accept full responsibility for its past aggression. Conversely, regarding Japan, Lind is
examining whether a state can acknowledge the depravity of its wartime behavior if the old elite
maintains access to the levers of power in postwar society. Her findings are ultimately within
the realm of what common sense would predict: Revamped states can deal with the past
gradually, while states with fairly broad continuity cannot.
56
A second, albeit far less serious problem, is that Lind’s model is incapable of explaining
short-term gyrations in the nature of affairs between states involved in interstate apology politics.
True, her apology model holds potential for determining the long-term trends of progressively
improving or steadily declining relations. She can therefore reasonably claim to explain the
progressive Franco-German rapprochement following World War Two. But Lind’s argument
would seem far too predetermined if France and Germany had undergone very bitter and
sporadic episodes of historic animosity while on the overall path of eventual reconciliation. It
would, in that case, seem imperative to try and understand these fluctuations and the logics they
might have imparted to the generally improving tenor of affairs. What is true for a hypothetical
Bonn/Berlin-Paris path to reconciliation is true for Seoul and Tokyo’s. Since Lind cannot
illuminate why Japan and Korea undergo both periods of warmth and particularly marked
hostility, her model’s prediction of steady deterioration of relations feels like it is made at the
expense of the ongoing present. While I accept that Japan and Korea have suffered a general
worsening of the history issue—or at least that it remains far worse than one would otherwise
expect—I therefore strongly reject the notion that Lind’s model explains the relationship
adequately.
56
Especially troubling is the second period of 1965-1989, in which apologies caused nationalist backlash among
Japanese elites. At that time, the elite body was constituted to very large extent by wartime elites and their
immediate protégés.
- 27 -
KOO’S COMPLEX INTERDEPENDENCE MODEL
In contrast to the approaches of the three scholars discussed above, Min-Gyo Koo takes a
very different tact in modeling Japan-Korea relations. Instead of looking at the entire history
issue, Koo limits his dissertation to island disputes.
57
He studies three cases: Dokdo/Takeshima,
Paracel and Spratley, and Diaoyu/Senkaku.
58
More important than these divergences, however,
is the econometrics focus of Koo’s work. In order to carry out econometric regressions he
develops an understandable DV, fleshing it out from varying levels of conflict over particular
islands between a given set of states on any one year. He utilizes a five-point ordinal scale,
scoring it from 0 to 4, with 0 representing practically no conflict whatsoever and 4 indicating
full-scale war.
59
Koo IVs are likewise distinguishable phenomena. For the Japan-Korea case, they include
the level of US defense commitment, joint democracy, joint membership in the WTO or its
precursor, the level of dispute intensity the year prior, and whether the political parties of the
executives in Seoul and Tokyo enjoy simultaneous majority control of legislature.
60
What Koo
is most interested in measuring, however, is the level of bilateral trade dependence. For his
dissertation is primarily aimed at testing Neoliberalism’s claim that trade dependence will
decrease conflict.
61
To calculate it for each state, Koo measures the total two-way trade and
divides by the GDP of Japan and Korea, respectively.
62
He does find some support for the Neoliberal claim that trade dependence dampens
conflict. Regarding relations between Seoul and Tokyo, South Korea’s trade dependence on
57
Koo, “Scramble for the Rocks.”
58
China and Vietnam contest the Paracel and Spratley islands, whereas the Diaoyu/Senkaku dispute is between
China and Japan.
59
Ibid., 16-18, 62. I detail more aspects of Koo’s DV in Section III, while explaining my own DV.
60
Ibid., 54, 57-61, 64.
61
Ibid., 33-40, 54-57.
62
Ibid., 62-63.
- 28 -
Japan lowers conflict at a 0.1 level of statistical significance. Unfortunately, Japan’s trade
dependence on Korea is not statistically significant, even at the low threshold Koo allows. But
its insignificance is not unexpected, because two-way trade represented less than one percent of
Japanese GDP for most of his 1953 to 2005 timeframe. It is worth noting that only one other of
Koo’s IVs enjoyed statistical significance in the Japan-Korea case: The intensity of a dispute the
year prior led to higher rates of conflict on a given year.
63
Despite its modest findings regarding Dokdo/Takeshima, Koo’s overall model holds
potential for uncovering the impact on the island dispute of theoretically important IVs.
Considerable shortcomings are nevertheless apparent. Importantly, Koo suffers a
methodological problem regarding his DV. His threshold levels for coding it are grossly
underspecified regarding the difference between a 0 and a 1 coding, for example. Here, the
problem lies in the unclear space between a 0’s “minimal or no diplomatic/political conflict” and
a 1’s “hostile rhetoric and public recrimination.” Since Koo explains neither how one can
identify hostile rhetoric and public recrimination, nor how many instances of either suffice to
shift a year past the “minimal” conflict threshold, his DV is not as transparent as it purports to
be.
64
Relatedly, Koo does not specify the factual basis for the coding of every single year of
his DV. Whereas he carefully illuminates via qualitative narration the events leading to scores
above 0, his reasoning behind the categorization of 0-scored years is usually not evident.
65
Absent explanation regarding these years and the opacity of his DV’s threshold levels are
awkward because there are quite a few “uneventful” years in which island-related events actually
took place. In 1981, for instance, a Japanese patrol vessel entered Dokdo/Takeshima waters,
63
Ibid., 65-68.
64
Ibid., 16-18, 62.
65
Ibid., 88-139.
- 29 -
leading to a discussion of it in the Korean Assembly and a formal remonstration.
66
Japan
similarly protested news reports in 1982 that a Korean fisherman had successfully registered his
vessel as domiciled in Dokdo.
67
By 1989, Japanese officials were making it a priority to bring
up the sovereignty of Dokdo/Takeshima at regularly scheduled, formal, bilateral meetings, a fact
pointedly mentioned in Japan’s Diplomatic Blue Book for several years beginning in 1990.
68
So
regular were problems related to the island dispute, in fact, that a very quick search of the 18
consecutively zero-scored years beginning in 1978 suggests that Dokdo/Takeshima was only
truly a non-issue in 1985, 1987, 1994, and 1995.
69
Koo’s DV suffer similarly from the concept of a “return to the status quo.” Whenever
conflict in a given year returns to the state it was in before becoming problematic Koo scores it a
0.
70
While ostensibly reasonable, the coding policy is actually somewhat troublesome. It has,
for example, a broad definitional problem in that a true return to a “status quo” situation is
impossible—no two years will ever really be the same. The phenomenon of status quo is,
therefore, necessarily broad and vulnerable to considerable interpretation bias. It is nevertheless
obvious that scholars utilizing it in their coding schemes can obviate a good amount of
disagreement by explicating and reasoning the best criteria for measuring a return to status quo
in a given situation, and then evidencing it exactingly.
66
“Assembly Queries Foreign, Defense Ministers,” Seoul Domestic Service, September 2, 1981, translated by
FBISDRAP, September 3, 1981.
67
“ROK Action on Takeshima Formally Protested,” Kyodo, August 18, 1983, reported by FBISDRAP, August 18,
1983.
68
See, for example, Ministry of Foreign Affairs of Japan, “Diplomatic Blue Book 1990: Japan’s Diplomatic Affairs,”
under “Item 2. The Korean Peninsula,” http://www.mofa.go.jp/policy/other/bluebook/1990/1990-contents.htm
(accessed September 9, 2011).
69
In addition to checking Japan’s Diplomatic Blue Book issues, I searched FBIS reports for the terms “Tokto” and
“Takeshima.” Foreign Broadcast Information Service Daily Reports, in NewsBank, http://www.newsbank.com/
(accessed September 9, 2011); Ministry of Foreign Affairs of Japan, “Diplomatic Blue Book,”
http://www.mofa.go.jp/policy/other/bluebook/ (accessed September 9, 2011).
70
Koo, “Scramble for the Rocks,” 16-18.
- 30 -
Regrettably, Koo is not vigilant enough on these points, as best illustrated by his
coverage of the Dokdo/Takeshima dispute in 1978, a year in which disagreement over the islands
was considerable.
71
At its most harried moment, Korean diplomats filed a formal complaint
following a May penetration of the island’s waters by Japanese fishing vessels, and Japanese
Foreign Minister Sonoda Sunao was compelled to present an emergency report to the Japanese
Diet.
72
Moreover, the conflict seems to have spilled over into other areas of the relationship, as
suggested by reports claiming Japan was withholding a previously negotiated ¥4 trillion loan to
Korea because of it.
73
In all, 1978 seems to meet safely the threshold for a category one coding. Due to his
finding that conflict returned to the status quo of 1976, however, Koo marks it a zero.
74
He
explains his decision in two sentences: In the first he states “by early 1978 the sovereignty issue
quickly subsided without inflaming the nationalist sentiments any further on both sides.” And he
notes in the second that Japan and Korea reached “an unofficial tacit agreement” allowing
occasional Japanese fishing within Dokdo/Takeshima waters.
75
Regrettably, neither claim is
without complication. Concerning his first statement, Koo does not evidence it at all, or clarify
his measurement assumptions. Readers have no idea, therefore, why they should discount
something like ROK Defense Minister Roh Chae-Hyon’s October testimony to the National
Assembly on Korean preparedness to protect Dokdo/Takeshima, or Korea’s September
enforcement decree regarding its formally legislated, 12-mile, maritime territorial declaration of
71
Ibid., 111-115.
72
“Sonoda Rejects ROK Protest of Fishing Near Takeshima,” Kyodo, May 9, 1978, reported by FBISDRAP, May
10, 1978.
73
“Japan Denies Loan Linked to Tok-To Dispute,” Haptong, June 20, 1978, reported by FBISDRAP, June 20, 1978.
74
Koo, “Scramble for the Rocks,” 62, 111.
75
Ibid., 111.
- 31 -
the year prior.
76
These laws implicitly declared all the waters in a 12-mile radius around
Dokdo/Takeshima to be Korean territory, a fact that evinced alarm from Japanese newspapers.
77
Koo’s sole reliance on the work of economist Hideki Kajimura to support his claim of a tacit
agreement is likewise disappointing.
78
For Kajimura’s statement about a tacit agreement is itself
apparently based only on one Japanese newspaper article, and officials in Seoul denied firmly
they had made a deal with Japan.
79
Moreover, in their October testimony to the National
Assembly, ROK Home Ministry officials reported that all of the 101 Japanese vessels caught
fishing around Dokdo/Takeshima had been forcibly expelled since the 12 mile territorial limit
first came into effect in April 1978.
80
The opacity of Koo’s decision making regarding his DV means his conclusions should be
accepted only with the greatest skepticism. While his econometrics approach fosters the
76
“Defense Minister Cites Readiness to Protect Tok-To,” Haptong, October 3, 1978, reported by FBISDRAP,
October 3, 1978. “Japanese Papers Report on New Sea Limit, Tok-To Island,” Haptong, September 20, 1978,
reported by FBISDRAP, September 21, 1978.
77
Neither Korea’s 1977 Territorial Sea and Contiguous Zone Act, which went into effect on April 30, 1978, nor the
related 1978 Enforcement Decree of Territorial Sea and Contiguous Zone Act mentions Dokdo/Takeshima. Their
omission is expected because UN rules for drawing baselines permit them only “in localities where the coastline is
deeply indented and cut into” or “if there is a fringe of islands along the coast in its immediate vicinity.”
Accordingly, the ROK delineates in its zone act and enforcement decree practically none of its remarkably smooth
and clear-cut eastern edge, and was not able to physically draw lines around remote islands far off the coast.
Omitted islands included the uncontested and obviously Korean Ulleungdo, as well as the considerably contested
Dokdo/Takeshima. Nonetheless, Korea’s Zone Act and Enforcement Decree effected a 12 mile zone of Korean
territory around the entire east coast and the uncontested and contested islands in the East Sea/Sea of Japan. See the
United Nations Convention on the Law of the Sea as described and analyzed by the US Department of State’s
Bureau of Oceans and International Environmental and Scientific Affairs, “Limits in the Seas, no. 121, Straight
Baseline and Territorial Sea Claims: South Korea” (Washington, DC: US Department of State, September 1998) 3-4,
http://www.state.gov/documents/organization/57685.pdf (accessed September 13, 2011). The State Department
document also contains an English translation of the zone act and the enforcement decree in its annexes. See ibid.,
11-20. See also Hideki Kajimura, “The Question of Takeshima/Tokdo,” Korea Observer 28, no. 3 (Autumn 1997):
467; Seo-Hang Lee, “Korea’s Claims to Maritime Jurisdiction,” Korean Journal of Comparative Law 18 (1990): 62-
89. For Japanese newspapers see “Papers Report on New Sea Limit, Tok-To Island,” Haptong, September 20, 1978,
reported by FBISDRAP, September 21, 1978.
78
Koo, “Scramble for the Rocks,” 111.
79
Kajimura, “The Question of Takeshima/Tokdo,” 472. On Korea’s denial of an agreement, see Sungjoo Han,
“South Korea 1978: The Growing Security Dilemma,” Asian Survey 19, no. 1 (January 1979): 46. For a Korean
refutation of the possibility of a tacit agreement see, for example, “Foreign Ministry Denies Reported Agreement on
Tok-To Island,” Haptong, November17, 1978, reported by FBISDRAP, November 17, 1978.
80
Another five Japanese vessels were caught within Dokdo/Takeshima waters, but were not expelled because,
apparently, they were not fishing when intercepted. See “Japanese Boats Violated Tok-To Waters 106 times,”
Haptong, October 11, 1978, reported by FBISDRAP, October 11, 1978.
- 32 -
impression of an impartiality exceeding that of Cha’s, Yoon’s, and Lind’s qualitative
methodologies, Koo actually makes result-changing decisions with much less clarity.
Reluctance to accept his judgment is necessarily compounded by the limited statistical
significance of his key finding regarding trade dependence. Changing Koo’s DV by even the
smallest amount could make trade dependence inarguably statistically insignificant, and might
even call into question his entire project.
Despite my harsh conclusion about Koo’s results, however, I ultimately believe his work
is quite important. He appears to be the first and only scholar to examine econometrically the
historical animosity between Japan and Korea. His effort seems like a positive step forward,
given the replicable nature and obvious impartiality of well done quantitative work. Moreover,
the many fluctuations of the history issue make it seem ideally suited for econometric analysis.
Econometrics should help determine correlated and statistically significant IVs, while weeding
out those that are neither. Precisely because of these reasons, I utilize Koo’s methodology as the
starting point for my own quantitative work, as I outline beginning in Section III.
CONCLUSION REGARDING THE FOUR MODELS
Each of the four Japan-Korea relations models discussed above enjoys strengths and
suffers weaknesses. Cha’s quasi-alliance model correctly emphasizes how coincident and
conflicting security interests can impact the tenor of affairs, but has been outdated by the nearly
complete disappearance of South Korea’s abandonment fears and Japan’s worries about
entrapment. Yoon’s net threat theory is right to draw attention to the role of overall interests,
though it suffers from incredible complexity. Lind’s apology politics model, while too
predetermined in its prediction of the past’s increased or decreased politicization, reasonably
points out the potential dangers of apologies, especially if offered by states in which a sizeable
- 33 -
number of important constituents oppose them for deeply personal reasons. Finally, Koo’s
methodology represents a groundbreaking treatment of an aspect of the history issue, even if one
must view his conclusions with the great skepticism.
III. Research Strategy
In writing this dissertation I face a dual challenge. First, I need to uncover which of the
commonly assumed IVs actually impact causally the history issue DV. Particularly important
are any IVs related to Japan’s and Korea’s respective material security interests. If I fail to show
a strong connection between security interest convergence and divergence and the DV, then I
cannot begin to argue that social-psychological phenomena like expectations and social emotions
related to fluctuations of security interests are important determinates of historical animosity. I
utilize quantitative methodology to evidence this first kind of data.
My second task is to demonstrate that social-psychological phenomena matter. This is no
small feat. Indeed the experience of scholars discussed above suggests I will have already
accomplished a lot if I am successful in crafting a good DV and reasonable IVs concerning
security interests and, via regression analysis, show the level of correlation between them. I use
deductive and comparative logic to argue for the existence and impact of social psychological
elements.
A. Dependent Variable: The History Issue
In this section I detail my history issue DV and quickly introduce the North Korea Policy
IV. I delay detailing my IVs until after I develop my three theoretical models in chapters 3, 4,
and 5.
Koo’s work on the island dispute provides the foundation for my own DV. His central
methodological idea was to adopt as the starting point for his DV Charles Gochman and Zeev
- 34 -
Maoz’s Correlates of War (COW) classification system for the highest level of conflict in
Militarized Interstate Disputes (MIDs) between two states in a given dyad-year.
81
Since the
COW dataset begins coding at the threat-of-force level, however, it cannot accommodate the
lesser forms of interstate political conflict related to historical animosity.
82
In order to include
conflict of a non-military nature, Koo was therefore compelled to significantly adjust the COW
methodology. Changes include the addition of a 0 coding for “minimal or no” conflict, and a 1
coding for “moderate to high levels of diplomatic and political conflict.” Koo blends more
extreme forms of diplomatic pressure, like sanctions and restrictions on bilateral ties, with threats
of force to help define a 2 category. “Mild clashes and isolated incidents between police or
armed forces” comprise a 3. Finally, Koo declares conditions a 4 when there are more
substantial military clashes or full-scale war.
83
Koo does not adopt Gochman and Maoz’s
criteria for the end of a conflict. Instead of coding a conflict over after a six-month lapse of
codable events, Koo sets a return to the status quo as the litmus.
84
I have already discussed the opacity inherent in Koo’s coding methodology, both in terms
of unclear differences between categories and regarding the uncertainty surrounding a return to
the status quo. But his DV also suffers another kind of problem that makes adopting it
unchanged problematic: It is too complicated. Its complexity is the main reason why one cannot
know whether Koo has coded each year correctly, not only in terms of replicability by other
scholars but also in accordance with his own criteria.
81
A MID is “a set of interactions between or among states involving threats to use military force, displays of
military force, or actual uses of military force.” It “must be explicit, overt, nonaccidental, and government
sanctioned.” Charles S. Gochman and Zeev Maoz, “Militarized Interstate Disputes, 1816-1976: Procedures,
Patterns, and Insights,” Journal of Conflict Resolution 28, no. 4 (December 1984): 587.
82
Gochman and Maoz find three kinds of military acts: Threat, display, and use of force. Use of force technically
includes war, though in practice they evaluate war as a fourth kind of military phenomenon. Gochman and Maoz,
“Militarized Interstate Disputes,” 588-589, 600, 607.
83
Koo, “Scramble for the Rocks,” 51-52, note 2.
84
Ibid., 16-18; Gochman and Maoz, “Militarized Interstate Disputes,” 589.
- 35 -
The difficulty emerges from the fact that Koo must evaluate just about every bit of
conflict related to a given island dispute if he wishes to operationalize it numerically and
convincingly. Given the tremendous amount of information that exists on each dispute, however,
this is not a practical research strategy, especially in terms of the 50 years Koo examines that he
codes a 2 or less.
85
Yet until scholars have exhaustively dissected each of those years for events
below the MID level, it is impossible to have faith that a previously unexplored, non-military
conflictual act will not be discovered. Of course any newly uncovered act could shift higher the
“final” categorization of any particular year. And, even if all the information were known, there
would still be difficult categorization questions. Regarding Koo’s contention of a return to the
status quo in 1978, for instance, archives may open that support the notion of a secret deal
allowing for occasional Japanese fishing in Dokdo/Takeshima waters. But even that information
would not eliminate the heated contention and frequent public recrimination that emerged after
rumors of a secret deal surfaced. What would be more important in deciding a return to the
status quo: The secret agreement, or the new and extensive society-wide conflict about its
contemporarily alleged existence?
Indiscriminate borrowing of Koo’s DV would require that I handle far more information
than he in terms of the Japan-Korea case. Whereas Koo’s DV covered only the island dispute
portion of the history issue, mine must represent all of its major elements. With far more events
per year, I would face considerably more difficultly in differentiating between levels of political
conflict below the MID level. Ultimately, the more encompassing nature of my exploration of
the history issue means I need to adapt rather than adopt Koo’s DV. I especially need to simplify
it below the level of major, easily countable events.
85
The physical clashes that demarcate conflict levels 3 and 4 are more obvious than the elements that define 0
through 2 codings, and therefore do not contribute to the complexity problem.
- 36 -
I begin building my DV by incorporating only the most significant, directly observable
phenomena. Exactly like the COW project, I measure the highest level of conflict within a given
year when measuring conflict. In line with the COW, I declare a particular conflict to be
continuous until there is a six month absence of codable events. Unlike the COW, however, I
code all years a 0 unless a codable event actually takes place. I hold to this standard even if a
recordable history act last occurred on December 31
st
of the year prior. Another important
difference between my methodology and the COW and Koo systems is that I code both key
moments of conflict and cooperation. My scale runs from negative (cooperation) 3 to positive
(conflict) 3. I therefore utilize a seven point ordinal scale.
To measure cooperation, I record the first occurrence of several kinds of important
conciliatory events. In addition to being a first, these events either must generally be considered
very helpful in improving the history issue, or must mark an obvious improvement in attitudes
about the past. These firsts are limited to general apologies, agreements, and visitations, as well
as Japanese admissions of responsibility, or important agreements, regarding the five main
components of the history issue. For first apologies, I include only those apologies exclusively
targeting a South Korean audience. Apologies meeting these requirements are the first formal
apology by a Japanese Cabinet member (1965), the first apology by a sitting Japanese Prime
Minister (1984), the first vague apology by the Japanese Emperor (1984), the first specific
apology by the Japanese Emperor (1990), and South Korea’s first agreement to settle the history
issue by accepting formally and with a strong sense of finality a Japanese Prime Minister’s
apology (1998).
Comprising general agreements are the secret Kim-Ohira memo (1962) in which Tokyo
consented to give Seoul financial restitution, the bilateral normalization treaty and related
- 37 -
agreements (1965), and Korea’s decision to end the import ban on Japanese cultural products
(1998). There are three key visitations and one particularly remarkable visitation agreement: the
first visit to postwar Korea by a sitting Japanese Prime Minister (1967), the first state-level visit
to Korea by a sitting Japanese Prime Minister (1983), the first state-level visit to Japan by a
sitting Korean President (1984),
86
and the first mutual agreement to arrange for a visit to Korea
by the Japanese Emperor (1999). Regarding the five primary components of the history issue, I
include Japan’s decision to provide permanent residency and related special benefits to third
generation Korean-Japanese (1990), and Japan’s first formal admission that it used compulsion
in its wartime recruitment of Comfort Women (1993).
The coding for the cooperation laid out above can be seen in Figure 1.5. I treat
cooperative moments additively: If one cooperative event occurs per year, I score the year a -1,
two events per year equal a -2, and so on. Cooperation is therefore potentially a continuous
variable, since it is possible that any number of key first events might take place on any given
year. In practice, however, it is exceptionally unlikely that more than two or three cooperative
phenomenon meeting the appropriate threshold level can take place in one year. Consequently,
my cooperation variable operates very much like an ordinal categorical variable.
86
Park Chung-Hee’s 1961 visit to Japan does not qualify as a first visit, because Syngman Rhee visited Tokyo as
early as 1950. For details of Rhee’s visit as well as the political context, see Cheong, The Politics of Anti-Japanese
Sentiment in Korea, 80-84.
- 38 -
Contrasting with cooperation, I operationalize conflict in two ways: by direct and indirect
measurement. First, I directly measure the highest level of conflict regarding the history issue on
a particular year.
87
In order to avoid the kind of exceptionally tricky coding problems that
weakened Koo’s DV, however, I do not code below the level of easily differentiated conflict. I
therefore begin coding the directly observable aspects of conflict at a 2. Level 2 conditions are
met when historical animosity derailed, stalemated, caused cancellation, or otherwise contributed
to the absence of the pre-1965 normalization talks, including any secret talks. Similarly, I also
code as a 2 those years in which the history issue led to the cancellation or postponement, for six
months or longer, of prearranged Presidential/Prime Ministerial summit meetings.
88
Importantly,
a cancelled or delayed summit need not exceed the 6 month mark if it was part of an ongoing
string of cancelled executive talks.
89
Conditions meet the threshold for a 3 coding when either
87
See Appendix 3, years 1960-1965 and 2005-2008.
88
I include the criteria of a six month delay or cancellation of summit meetings in order to weed out summits that
have been delayed for other reasons, including domestic elections and deliberate use of summit postponement as a
short-term negotiating strategy. My decision to utilize six months as the starting point for a serious event is based
on the importance attached to the timeframe by the COW.
89
As noted earlier conflicts are ongoing until there is a break in coding for six months or longer.
- 39 -
Japan or Korea issues a threat of force or carries out a show of force.
90
Figure 1.6 shows these
directly observable phenomena.
91
Second, I indirectly measure conflict by examining the five main components of the
history issue as enumerated in Chosun Ilbo headlines from 1960 to 2008. Selecting only the
peak years of conflict regarding Dokdo/Takeshima, Koreans in Japan,
92
Japanese Textbooks,
Comfort Women, and Yasukuni Shrine results in up to two of these phenomena falling on
particular years. I show this indirect measure of the history issue in Figure 1.7. Importantly,
90
I limit my DV to 3, because conflict never rises above this point from 1960 to 2008. The scale could easily be
extended, however, in the event of small-scale clashes between military or police forces (4), larger but isolated
physical clashes (5), and war (6).
91
Coding decisions regarding canceled meetings are detailed in Appendix 3. Appendix 3 also pertains to an
alternative DV—DV 7—suggested to me by Jacques Hymans and included later in this chapter. Regarding the
directly observable conflict data in Figure 1.6 and included in Appendix 3, only the years before 1965 and after 2004
are pertinent. I create DV 7 and other DV variants in response to Hymans’ excellent criticism of my DV. He
pointed out to me the possibility that years in which no highest-level possible meetings were considered might be
more conflictual than years in which they were planned but cancelled or otherwise derailed for six months or more.
Consequently, I explore in-depth below the possibility my coding of directly observable conflict is incorrect. I note
for now that my research suggests strongly my coding is correct. The alternative possibility does not hold, probably
because annual meetings were not normalized and expected until the 1980s, after which there were annual meetings
in Japan or Korea until conflict over the past peaked again in the mid-2000s.
92
I only measure Koreans in Japan beginning in 1970, the year pro-Seoul Korean-Japanese began to outnumber
those who were pro-Pyongyang. Before 1970, but especially before the 1965 normalization treaty, the issue of
Koreans in Japan was a regime legitimization issue—i.e., a key security issue—rather than part of the history issue.
I detail and evidence this contention in Chapter 6 while discussing my IVs.
- 40 -
though I treat these history issues additively here for illustrative purposes, when I combine them
with the directly observable elements of historical conflict, I score them only as a level 1 conflict.
By adopting this tactic, I claim ordinal categorical status for the conflict component of my DV,
since these years of strife are worse than zero scored years, but less conflictual than years
categorized as two or higher.
The final step in creating my DV is to create a history-issue amalgam out of cooperation
and directly and indirectly observed conflict. This process is relatively straightforward, as
mismatch between cooperation and conflict occurs only in 1962 and 1965. Fortunately, bilateral
cooperation in the form of the 1962 secret restitution memo and the 1965 normalization treaty
and apology was far more remarkable than the conflict over Dokdo/Takeshima, the primary issue
of contention in those years.
93
Figure 1.8 shows the amalgamated cooperation and conflict
data—i.e., what I call the Primary DV for this project.
93
Though choosing to favor cooperation is the obvious choice regarding mismatched cooperation and conflict
within these particular years in the Japan-Korea relationship, the inconsistency of the directly and indirectly created
data does raise questions regarding whether my methodology can be applied to other cases. The answer is probably
that it can, though I foresee the need to make decisions that will be more complicated than that required here. One
choice that would practically always be justifiable would be selection of the directly observed data over the
- 41 -
The quality of the history issue DV is remarkable for a number of reasons. In
comparison with the models of Japan-Korea relations discussed earlier, it is more obviously
impartial and replicable. True, scholars wishing to recreate it would need to pick from the
headline data the most conflictual years for each of the five main history issue components, and
differences could emerge in the process. Fortunately, there are a number of reasons why
selection is not a major concern.
First, I choose conflictual years from the headline data very conservatively. While I do
not adopt rigid scientific selection criteria, I also strive to prevent artistry. To that end I use four
guiding rules: 1) A conflictual year should represent a considerable peak relative to either the
year before or after it;
94
2) the conflict level of a peak year must nearly double the intensity of
the lowest level of conflict within the preceding two years;
95
3) the selected year cannot be much
indirectly observed, since the former should be more accurate than the latter. At any rate, having to choose between
inconsistent instances of cooperation and conflict is something about which all scholars studying complex, real-
world cases must make choices. No matter one’s methodology, scholars cannot escape the fact that cooperation and
conflict are quite often ongoing and overlapping.
94
I used a 75 percent increase regarding either the year before or after as a rule of thumb.
95
A fifth rule might be necessary if one wishes to apply my methodology to other cases. Guiding rules one and two
should be overridden in the event there is an obvious, ongoing peak lasting more than two years. Nonetheless, since
- 42 -
lower than both peaks immediately before and after it;
96
and 4) a peak year cannot be so low
relative to any selected years before it that it risks being unremarkable. These rules should
promote replicability and impartiality. They should also mitigate phenomena like headline
inflation, which occurred as the number of pages in the Chosun Ilbo tended to increase over time.
The years I select via these guiding rules are marked by diamond shapes in figures 1.9 to 1.13.
97
volatility and issue complexity should be prerequisites for similar cases, it seems unlikely that ongoing static peaks
will be common within any given conflictual issue. There are none among the five core history issues in Japan-
Korea affairs.
96
If a line were drawn across peaks within a given history issue, the line must not form a deep ‘V’ shape. If peak
years are very far in time, the absolute trough of the middle year should not be considerable.
97
The same data, without the peak years marked, is available on figures 1.2 to 1.4 in a larger and therefore easier to
see size.
- 43 -
The second reason why peak year selection is not a highly critical matter is the general
robustness of the relationship between the IV reflecting Japan’s and South Korea’s North Korea
policies and the Primary DV.
98
As seen in Table 1.1,
99
not only does the North Korea Policy IV
enjoy statistical significance and robust correlation when regressed against the Primary DV, but
it also does well when regressed against DVs comprised solely of directly observed conflict, and
directly observed conflict and cooperation together.
100
Only when regressed against directly
observed cooperation are the results statistically insignificant at the 0.05 level.
101
In other words,
Chosun Ilbo headlines are remarkable more for supplying missing data than for providing data
essential to create good results. Given the robust foundation upon which indirectly observed
conflict rests, it seems very unlikely scholars examining the headline data could select years in
such a way that my overall project would become threatened. Moreover, though I do not detail it,
98
Because I analyze multiple DVs, I frequently refer to the main DV as the Primary DV.
99
I provide regression tables in this section for illustrative purposes only. The two IVs I run as coefficients here are
selected based not on any particular theoretical foundation, but because they are the only IVs that regularly return
statistically significant results in the theoretically grounded analyses I carry out in Chapter 6.
100
For ease of comparisons across tables, I label the DVs Primary DV, DV1, DV2, and so forth. A DV of the same
name is always comprised of the same inputs.
101
Remarkably, the North Korea policy variable remains statistically significant if I permit Koo’s threshold of 0.10
P-value.
- 44 -
the regression output for the North Korean policy IV actually improves a little when I do attempt
to select years without applying a particular methodology.
102
Table 1.1 Primary DV, DV 1, DV 2, & DV 3
Primary
DV
DV 1: Directly
Observable Conflict
DV 2: Directly Observable
Cooperation
DV 3: Directly Observable
Cooperation and Conflict
Beta, North Korea Policy Coef. .481474 .4767999 .2369204 .475711
P>|t|, North Korea Policy Coef. 0.000 0.001 .100 0.000
Beta, Korean Assembly Elections
Coef.
.2322803 .1034524 .168323 .1724081
P>|t|, Korean Assembly Elections
Coef.
0.068 0.425 0.239 0.181
Adjusted R2 0.2604 0.2074 0.0467 0.2278
Prob >F 0.0004 0.0018 0.1251 0.0010
The Chosun Ilbo headlines are nevertheless important. For by combining them with the
directly observable cooperation data, it becomes possible to create a cooperation-conflict DV
that not only spans the entire 1960 to 2008 time period under consideration, but does so without
including any directly observable conflict data. Doing so is worthwhile, because there is a
chance the directly observable conflict portion of the Primary DV is terminally flawed.
Specifically, since directly observable conflict is comprised to a large extent by regularly
scheduled, highest-level possible meetings that have been canceled, derailed, or stalemated for
six months or more, there exists the possibility that a more conflictual coding criterion is being
overlooked in the DV. One can alternatively argue, for instance, that years in which no meetings
were planned in the first place are actually more conflictual than years in which they were
scheduled but then cancelled.
103
Creating an alternative cooperation-conflict DV, one excluding
the directly observable conflict data, permits a side-by-side evaluation of it—DV 4—and the
Primary DV. The regression results are shown in Table 1.2 below. I also include the output
produced by DV 5, which is comprised solely of the newspaper headline data.
102
Though I avoided using specific rules while selecting years, I remained cognizant that headline inflation might be
a general problem across time. Therefore, I tended to be more cautious about years in which results were declining
in comparison with years prior.
103
I thank Jacques Hymans for pointing out this argument.
- 45 -
Table 1.2 Primary DV, DV 4, & DV 5
Primary
DV
DV 4: Directly Observable Cooperation
and Indirectly Observable Conflict (via
Chosun Ilbo Headlines)
DV 5: Indirectly Observable
Conflict (via Chosun Ilbo
Headlines)
Beta, North Korea Policy Coef. .481474 .3364226 .2760401
P>|t|, North Korea Policy Coef. 0.000 0.015 0.050
Beta, Korean Assembly Elections Coef. .2322803 .2645866 .2293049
P>|t|, Korean Assembly Elections Coef. 0.068 0.053 0.102
Adjusted R2 0.2604 0.1522 0.0941
Prob >F 0.0004 0.0084 0.0387
As seen in the regression outputs, the difference between the Primary DV, DV 4, and DV
5 is that of robustness and confidence, rather than something actually conflictual. Not only is
directionality of correlation the same regarding both IVs implemented, but the North Korea
Policy IV is statistically significant across all three variants of the DVs. In sum, these results
suggest it is highly unlikely that a coding error as critical as the above-described one regarding
directly observable conflict has been committed. If the Primary DV contained such a serious
flaw, there is little reason to believe DV 4 and DV 5 would both comport so closely with it.
The Chosun Ilbo headlines provide a second avenue through which to evidence that a
directly observable conflict coding error did not take place. Namely, the raw newspaper
headlines regarding Dokdo/Takeshima, Koreans in Japan, Japanese textbooks, Comfort Women,
and Yasukuni Shrine can each be utilized as separate IVs and regressed against a version of the
DV stripped of newspaper data. With the indirectly observable conflict removed, such a DV—
DV 6 below—would be comprised of what remains of the Primary DV: Directly observable
conflict and directly observable cooperation. The objective is of course to check for the overall
accuracy of the directly and indirectly observable data, by using each as a measurement of the
other. Though the unfiltered headline data certainly contains a lot of “noise,” those history issue
components that are statistically significant should correlate positively, and strongly, with DV 6.
I carry out this exercise in Table 1.3. In addition to running the standard regression of
data comprising the 1960 to 2008 period, I delimit the timeframe to 1970 to 2008. Such a step is
- 46 -
necessary, because no data exists before 1970 regarding the issue of Koreans in Japan.
104
Thus, I
regress against DV 6 the other four history issue components from 1960 to 2008, and all five
history issue components from 1970 to 2008. Note the Chosun Ilbo headline data in Table 1.3 is
“raw,” meaning I do not deflate it by overall headline wordiness, as I do in Table 1.4 below. I
discuss the results of Table 1.3 after illustrating Table 1.4.
Table 1.3 Raw Chosun Ilbo Headlines & DV 6
DV 6 (1960-2008): Directly Observable
Cooperation and Directly Observable
Conflict
DV 6 (1970-2008): Directly Observable
Cooperation and Directly Observable
Conflict
Beta, Dokdo/Takeshima Coef. .2160626 .3605993
P>|t|, Dokdo/Takeshima Coef. 0.211 0.036
Beta, Koreans in Japan Coef. NA .1288866
P>|t|, Koreans in Japan Coef. NA 0.384
Beta, Japanese Textbooks Coef. -.0179091 .0429417
P>|t|, Japanese Textbooks Coef. 0.893 0.744
Beta, Comfort Women Coef. -.1169081 -.0201034
P>|t|, Comfort Women Coef. 0.398 0.890
Beta, Yasukuni Shrine Coef. .3386235 .4307747
P>|t|, Yasukuni Shrine Coef. 0.047 0.012
Adjusted R2 0.1691 0.3753
Prob >F 0.0156 0.0008
In Table 1.4, I carry out the exact same exercise, except for one difference: I “deflate”
the Chosun Ilbo headlines by dividing each year of headline data by a proxy ratio of the overall
number of words comprising the headlines in a given year. Doing so helps to accommodate both
the fluctuating number of headlines and the changing wordiness of headline styles across time.
In order to estimate the number of words comprising headlines, I searched Chosun Ilbo headlines
for the following frequently appearing grammar forms and one noun: 가, 이, 을, 를, 의, 으로, 는,
에, 면서, and 명.
105
104
As I explain in Chapter 6, the issue of Koreans in Japan was a North Korea policy security issue until 1970; it
was not a history issue component until then. Thus, headlines regarding Koreans in Japan cannot be considered part
of the IV from 1960 to 1969.
105
After assembling each category’s yearly count, I summed the count of each category on a per-year basis. I then
created the wordiness ratio by dividing each year by the sum of all categories in 2008.
- 47 -
Table 1.4 Deflated Chosun Ilbo Headlines & DV 6
DV 6 (1960-2008): Directly Observable
Cooperation and Directly Observable
Conflict
DV 6 (1970-2008): Directly Observable
Cooperation and Directly Observable
Conflict
Beta, Dokdo/Takeshima Coef. .0458862 .2775104
P>|t|, Dokdo/Takeshima Coef. 0.759 0.077
Beta, Koreans in Japan Coef. NA .0860807
P>|t|, Koreans in Japan Coef. NA 0.576
Beta, Japanese Textbooks Coef. -.0204734 .0556091
P>|t|, Japanese Textbooks Coef. 0.884 0.696
Beta, Comfort Women Coef. -.1011785 -.0238602
P>|t|, Comfort Women Coef. 0.468 0.873
Beta, Yasukuni Shrine Coef. .3823397 .4701492
P>|t|, Yasukuni Shrine Coef. 0.013 0.004
Adjusted R2 0.0921 0.2770
Prob >F 0.0826 0.0068
In combination, Table 1.3 and Table 1.4 provide very good evidence that the
hypothesized coding error regarding directly observed conflict has not taken place. Importantly,
in every instance in which a given history issue coefficient is statistically significant at a 0.10
level or better, the directionality of correlation comports with the idea that the history issue—
directly observed—will be worse when the history issue—indirectly observed—is bad.
Particularly interesting in this regard are Chosun Ilbo headlines pertaining to Yasukuni
Shrine. For whether the Yasukuni Shrine headlines are analyzed raw or deflated, the correlation
with directly observed historical conflict and cooperation is always positive, considerable, and
statistically significant at a 0.05 level or better. Indeed, a one unit increase in Yasukuni Shrine
headlines correlates with a minimum of a 0.34 unit increase in the history issue directly observed,
to a maximum of a 0.47 unit increase. If a coding error as critical as the one described earlier
had been made—i.e., if I had missed the directly observed conflict represented by meetings not
even being planned from 1960 to 2008—it is all but inconceivable that the indirectly observed
Yasukuni Shrine conflict data would comport to the directly observed conflict and cooperation
so well.
In Table 1.5 below, I carry out another set of analyses pertaining to the coding of directly
observable conflict. I regress raw and deflated history issue headline IVs against DV 1, which is
- 48 -
comprised solely of directly observable conflict. In contrast with DV 6 as used above, DV 1 is
missing directly observable cooperation. I run this analysis to eliminate the possibility that
directly observable cooperation is overwhelmingly responsible for the correlations uncovered in
Table 1.3 and Table 1.4. The results, which are slightly better overall than those returned when
both directly observable cooperation and conflict are considered together, as was the case with
DV 6, confirm the accuracy of my coding for directly observable conflict. Most importantly, in
seven of eight instances Dokdo/Takeshima and Yasukuni Shrine headline data return positively
correlated, statistically significant, robust results. Scholars perceiving an error in the coding
scheme I adopt for directly observable conflict would need to explain why the headlines for
Dokdo/Takeshima and Yasukuni Shrine consistently comport closely with it.
Table 1.5 Chosun Ilbo Headlines & DV 1
Raw Deflated
DV 1 (1960-2008):
Directly Observable
Conflict
DV 1 (1970-2008):
Directly Observable
Conflict
DV 1 (1960-2008):
Directly Observable
Conflict
DV 1 (1970-2008):
Directly Observable
Conflict
Beta, Dokdo/Takeshima Coef. .3441966 .4867388 .1979874 .3741043
P>|t|, Dokdo/Takeshima Coef. 0.035 0.000 0.178 0.009
Beta, Koreans in Japan Coef. NA .0046215 NA -.0669692
P>|t|, Koreans in Japan Coef. NA 0.967 NA 0.622
Beta, Japanese Textbooks Coef. -.0588878 .0059234 -.0546412 .0168713
P>|t|, Japanese Textbooks Coef. 0.634 0.952 0.688 0.893
Beta, Comfort Women Coef. -.165195 -.0696849 -.1363194 -.0792537
P>|t|, Comfort Women Coef. 0.201 0.525 0.317 0.547
Beta, Yasukuni Shrine Coef. .3225614 .4565712 .3266736 .4589804
P>|t|, Yasukuni Shrine Coef. 0.042 0.001 0.028 0.002
Adjusted R2 0.2822 0.6434 0.1396 0.4364
Prob >F 0.0009 0.0000 0.0305 0.0002
A final method for investigating the possibility I erred when measuring directly
observable conflict is to actually create the DV suggested by the criticism about it, as explained
above, and then analyze it via Chosun Ilbo headline data. Specifically, I can score years in
which no highest-level possible meetings took place as being one step more conflictual than
years in which regularly scheduled meetings were canceled, derailed, or stalemated for six
months or longer. Importantly, since it is difficult to ascertain whether an actual meeting took
place on a given year, I code years as having no conflict if a Japanese Prime Minister travelled to
- 49 -
Korea, or a Korean President travelled to Japan. The only exception to this rule is enacted when
one of two conflictual events occur on a given year: Years in which highest-level possible
meetings and talks are cancelled, derailed, or stalemated for six months or more are coded a 1,
and years in which there is a show of force or threat of force are coded a 3. It follows that years
in which no cross-border visits take place are coded a 2. Secret agreement in 1962 and the 1965
normalization agreement continue to trump any conflictual events that may have taken place on
those years. This DV is illustrated in Figure 1.14, and detailed in Appendix 3.
106
As revealed in Table 1.6 below, the Chosun Ilbo headlines evidence that coding directly
observable conflict via the alternative method is incorrect. Most problematically, headlines
pertaining to comfort women correlate robustly and negatively with directly observed conflict,
and do so from 1960 to 2008 at 0.02 level of statistical significance or better. The overall
correlation remains the same, whether the headlines are analyzed raw or deflated. Scholars who
argue the alternative coding method for directly observable conflict is superior to the method
106
Data is from Victor D. Cha, “Bridging the Gap: The Strategic Context of the 1965 Korea-Japan Normalization
Treaty,” Korean Studies 20 (1996): 123-160; Kenji Hiramatsu, “Leadup to the Signing of the Japan-DPRK
Pyongyang Declaration,” Gaiko Forum (Winter 2003): 10; Yoon, “Fragile Cooperation.”
- 50 -
utilized while creating the Primary DV and DV 1 will need to contend with the fact that a one
unit increase in headlines about the comfort women correlates with a 0.37 to 0.41 per unit
decrease in directly observable conflict across the 1960 to 2008 time period. That is, the
researchers would have to explain why one of the five components of the history issue is sharply
discordant with Japan-Korea conflict. Though I decline to detail the full results, zeroing out the
coding for Koreans in Japan from 1960 to 1969, and regressing all five history issue components
from 1960 to 2008 similarly produces statistically significant, robust, negative correlation
between the comfort women headline data and DV 7.
107
Table 1.6 Chosun Ilbo Headlines & DV 7
Raw Deflated
DV 7 (1960-2008):
Directly Observable
Conflict, Alternative
DV 7 (1970-2008):
Directly Observable
Conflict, Alternative
DV 7 (1960-2008):
Directly Observable
Conflict, Alternative
DV 7 (1970-2008):
Directly Observable
Conflict, Alternative
Beta, Dokdo/Takeshima Coef. -.0014706 .0821697 -.0595517 .0956625
P>|t|, Dokdo/Takeshima Coef. 0.993 0.671 0.694 0.573
Beta, Koreans in Japan Coef. NA .2983997 NA .3092302
P>|t|, Koreans in Japan Coef. NA 0.088 NA 0.075
Beta, Japanese Textbooks Coef. .0652227 .1113179 .0618757 .1176281
P>|t|, Japanese Textbooks Coef. 0.636 0.468 0.662 0.456
Beta, Comfort Women Coef. -.4145421 -.3091464 -.3702104 -.2735027
P>|t|, Comfort Women Coef. 0.006 0.073 0.011 0.103
Beta, Yasukuni Shrine Coef. .1465691 .1793132 .0218012 .0514387
P>|t|, Yasukuni Shrine Coef. 0.398 0.347 0.885 0.761
Adjusted R2 0.1078 0.1543 0.0718 0.1187
Prob >F 0.0601 0.0593 0.1225 0.1010
If my coding scheme for directly observed conflict is justifiable via the above statistical
analyses, the weakest remaining aspect of the Primary DV remains its directly observable
cooperation component. In particular, either my selection of first instances of cooperation might
be too liberal, or I may have accidentally missed some key cooperative firsts that other scholars
would have coded. Consequently, it seems prudent to investigate whether the lesser ones are
107
For comfort women headlines raw and deflated, the P-value is 0.014 and 0.027, and the Beta scores are
-.3689257 and -.324748, respectively. None of the other IVs are statistically significant in either regression, even at
a 0.10 level.
- 51 -
playing too decisive a role, and to create a more conscribed directly observable cooperation DV
component.
In order to scrutinize directly observable cooperation, I temporarily remove the less
obvious cooperative firsts from the DV. I eliminate the first informal visit to Korea by a sitting
Prime Minister (1967), the first vague apology by the Emperor (1984), Japan’s agreements
regarding the rights of third generation Korean-Japanese (1990), the Emperor’s first direct
expression of regret (1990), Japan’s first admission of guilt regarding the comfort women (1993),
and the first mutual agreement to arrange for the Emperor’s visit to Korea (1999). What is left—
the secret financial restitution memo (1962), the normalization treaty and related cabinet-level
apology (1965), the first state-level visit to Korea by a sitting Japanese Prime Minister (1983),
the first state-level visit to Japan by a sitting Korean President (1984), the first apology to Korea
by a sitting Prime Minister (1984), the first formal Korean acceptance of Japan’s apology (1998),
and Korea’s decision to allow Japanese cultural imports (1998)—results in a far more
conservative cooperation component. Nevertheless, as shown in Table 1.7 below, being more
selective about directly observed cooperation actually strengthens the correlation between the
North Korean Policy IV and the DV, as seen in the DV 2 and DV 8 regression results. This
outcome is not surprising: Presuming the North Korean Policy IV accurately reflects divergence
and convergence of Seoul’s and Tokyo’s Pyongyang policies, one expects correlations to
strengthen as criteria for measuring history issue-related cooperation and conflict are tightened.
Table 1.7 DV 2, DV 8, & DV 9
DV 2: Directly
Observable
Cooperation
DV 8: Directly Observable
Cooperation, Conservative
Elements Only
DV 9: Directly Observable
Cooperation, Liberal Elements
Only
Beta, North Korea Policy Coef. .2369204 .2868375 .0331292
P>|t|, North Korea Policy Coef. .100 0.046 0.822
Beta, Korean Assembly Elections Coef. .168323 .123987 .129018
P>|t|, Korean Assembly Elections Coef. 0.239 0.380 0.382
Adjusted R2 0.0467 0.0602 -0.0247
Prob >F 0.1251 0.0901 0.6593
- 52 -
It bears mentioning that conservatively selected, directly observable cooperation DV
components also strengthen the correlation and statistical significance of the North Korea Policy
IV, when combined with either directly observable conflict or indirectly observable conflict.
The tightened relationships are viewable in Table 1.8 below, when one compares the results
between the Primary DV and DV 10, as well as the outputs of DV 4 and DV 11.
Table 1.8 Primary DV, DV 10, DV 4 & DV 11
Primary
DV
DV 10: Directly
Observable Cooperation,
Conservative Elements
Only, and Directly
Observable Conflict
DV 4: Directly Observable
Cooperation and Indirectly
Observable Conflict (via Chosun
Ilbo Headlines)
DV 11: Directly Observable
Cooperation, Conservative
Elements Only, and Indirectly
Observable Conflict (via
Chosun Ilbo Headlines)
Beta, North Korea
Policy Coef.
.481474 .5240101 .3364226 .3831534
P>|t|, North Korea
Policy Coef.
0.000 0.000 0.015 0.005
Beta, Korean Assembly
Elections Coef.
.2322803 .1450151 .2645866 .2532795
P>|t|, Korean Assembly
Elections Coef.
0.068 0.246 0.053 0.059
Adjusted R2 0.2604 0.2688 0.1522 0.1816
Prob >F 0.0004 0.0003 0.0084 0.0037
The above examination of variations of the Primary DV informs that the Primary DV is a
reasonable estimation of the history-derived conflict suffered by Japan and the ROK. In those
areas where is it potentially weak, investigation reveals strength. Most critically, the direct and
indirect measures of conflict data comprising the Primary DV are compatible with each other,
whether combined with directly observable cooperation components or treated separately. And
the directly observable conflict component of the Primary DV is compatible with the unfiltered
Chosun Ilbo headline data, whereas the alternative method for coding directly observable
conflict is not. If I have made a mistake in my data coding, it most likely resides in the North
Korean Policy IV and not the Primary DV. I detail the North Korean Policy IV in Chapter 6.
The overall quality of the DV is not unexpected, because it was highly valid internally.
In other words, out of 49 possible years in which cooperation and conflict could have been in
direct disagreement, they were opposed only in 1962 and 1965, when conflict indicated by the
- 53 -
headline data clashed with directly observable cooperation. In each case I chose conflict over
cooperation, a position I defend in two ways.
First, the cooperation evident in those years was far more remarkable than the conflict.
To put it simply, 1962 and 1965 represent the first substantive Japan-Korea agreements inked
since Japan regained its sovereignty in 1952. They are therefore extremely significant. In
contrast, the politicization of Dokdo/Takeshima in 1962 and 1965 was nowhere near as
remarkable. Most obviously, the islands did not prevent agreement on key matters at the secret
negotiations in 1962 or the publically announced meetings in 1965. There was similarly neither
a threat or show of force, nor small-scale clashes as seen in the early 1950s. There was nothing,
in other words, that might require a second-guessing of the decision to prioritize cooperation
over conflict on the history issue.
Second, I choose cooperation over friction because the directly observed data is almost
certainly a more accurate representation of overall relations than the more indirect headlines.
Whereas the headline data should indicate the overall information a large percentage of the
public is receiving and is willing to accept, directly observed government-to-government
cooperation happens only when the governments are cooperating enough, according to their own
ideas of what is enough, on the key history issues. Counterfactually, for example, no matter
Korea’s other incentives to cooperate, it is impossible to conceive that agreement would have
been made in 1962 and 1965 if the Japanese had landed a small police or military force on
Dokdo/Takeshima just before either set of talks began, even if the public did not yet know of it.
The fact that my coding of cooperation and conflict clashes in only two out of 49 years
under consideration suggests I have found a very good way of measuring the history issue. To
- 54 -
put it another way, despite having 21 conflictual years
108
and 14 cooperative events on 9
cooperative years, these opposed phenomena occur during the same year only twice. And on the
two years they do disagree, the conflictual element is noteworthy for being unremarkable. In
other words, the divergence is much smaller and therefore less significant than if cooperative
events had taken place on a year in which historical animosity proved considerably more
damaging. Such bounded incongruity is especially surprising since the annual unit of analysis
leaves plenty of space, temporally speaking, for both cooperative and conflictual events to take
place “simultaneously.”
Reliance on headline data is of course not ideal, even if advantageous regarding ease of
research and interpretation. It is not unlikely, in all, that I included some years that were not
actual peaks while excluding some that were. Nevertheless, there is good reason to be satisfied
with the headline data. Given the conservative methodology, it seems improbable that peak
years of conflict were chosen in which the history issue was minimally or entirely unproblematic.
More to the point, though I cannot have confidence that I selected all the peak years of conflict, I
can be relatively certain that I selected years more conflictual than those in which there was, in
actuality, minimal or no conflict. Since the Primary DV is ordinal and categorical, so long as I
have selected years that are more conflictual—on average—than years coded 0, and less
conflictual than years coded as level 2 conflict, then I have done well.
I concede that my methodology for DV creation is unusual. My blending of directly
observable conflict and both types of cooperation seems particularly uncommon. Similarly odd
is my use of additive cooperation elements and purely categorical levels of conflict. Nonetheless,
I have accomplished what I set out do: create an ordinal, categorical DV. On average, the
following should be true: Level 3 cooperative years should be more cooperative than level 2
108
There are 21 conflictual years including 1962 and 1965. But, as noted earlier, only 19 are included in the DV.
- 55 -
cooperation; level 2 cooperation should be more cooperative than level 1 cooperation; level 1
cooperation should be more cooperative than the cooperation found in 0 coded years.
Conversely, level 3 conflict years should be more conflictual than level 2 conflict; level 2
conflict more conflictual than level 1 conflict; and level 1 conflict more conflictual than the
conflict seen in 0 coded years.
109
One criticism that can be made about the DV as it applies to the Japan-Korea case, and
therefore about my project in general, is that I am examining a relatively narrow band of
conflict.
110
In particular, if relations are nowadays seemingly self-limited to canceled summit
meetings at worst, why does it matter whether Japan and Korea suffer historic animosity? After
all, surely both states would pull together under a threat undeniable to their mutual national
securities. Although I can foresee a worst case scenario in which ongoing subpar security
coordination between Seoul and Tokyo leaves them—and therefore their American ally—less
than fully prepared to respond to an extremely pressing and somewhat surprising security
emergency, the general sentiment is most likely right: Japan and Korea would work together to
ensure security threats of the highest magnitude are met.
The importance of my study is therefore not highly reliant on questions about the
response capability of the Japan-Korea-US security triangle. Rather, it consequence lies in
pointing out that the history issue is precisely the kind of non-material phenomenon that prevents
Seoul and Tokyo from better internalizing each other’s interests. It is why there is so very little
“We” and “Us” in the relationship. Indeed, Japan and South Korea have not followed the
Franco-German path to more fulfilling rapprochement because diverging interests regarding
109
In Chapter 6 I do accommodate for the possibility that the DV categories are incorrectly sorted. Thus, I create an
alternative DV extending only from -1 (cooperation) to 2 (conflict). The output difference is not substantial,
however.
110
I thank Dan Lynch for pointing out this argument.
- 56 -
North Korea, especially, almost always lead to politicization of the past. Historic animosity in
turn prevents the two quasi-allies from better internalizing each other’s security interests, thereby
setting the stage for conflicting security policies and more historic animosity the next time
interests clash. The purpose of my project is to uncover how security interest convergence and
divergence are related to historic animosity. Though some traditional explanations have merit, I
argue they cannot adequately explain the robust relationship between security interests and the
history issue. I argue social phenomena—alignment security expectations and social emotions—
are needed to explain Japan-Korea affairs.
- 57 -
Chapter 2
Introduction to Theory: The Role of Expectations and Emotions
I. Outline of Theoretical Chapters
Expectations and emotions matter in international relations. In the following three
chapters I create three alignment politics models, each one blatantly dependent on security
expectations and the emotions produced when alignment members satisfy or disappoint those
expectations. The current chapter’s key objective is to highlight how expectations and emotions
are ubiquitous in both IR theory and alignment politics modeling. This task is critical, because I
need to establish that much of what I do is just make explicit what most scholarship keeps
implicit.
I begin this chapter by highlighting many of the expectations and emotions used
explicitly and implicitly by theorists working within the main schools of IR, as well as by some
of the classic philosophers and thinkers who inspired them. After surveying IR theory, I shift my
attention to the work of four scholars who modeled the Japan-Korea bilateral relationship,
examining in particular the nature of their usage of expectations and emotions. Finally I give a
brief overview of Japan-Korea relations, targeting especially the question of which state tended
to suffer disappointed alignment security expectations during a given time period.
111
As emphasized below in Diagram 2.1, each of the three models I create is remarkable
primarily for its emotional ontology. Because Neorealism dominates alignment modeling, I first
111
Japan and Korea are not allies. But, as I argue later, they each expect into existence considerable security
obligations of the other. I define terms like “allies” and “alignment members,” as well as “security expectations”
and “security obligations” in Chapter 3. I explain why security expectations exist where there ought be none in
Chapter 5.
- 58 -
attempt to construct a model as true to Neorealism as possible. I call this first theoretical model
the delimited balance of threat model. Hewing to Neorealism in critical ways, it is most
remarkable for including as determinate the passive social emotions of friendliness and enmity.
Of course these social emotions are decidedly not Neorealist. Though I attempt to utilize them in
a way I perceive to be reasonably befitting the metatheory, their ultimate incompatibility along
with other, parallel ideational shortcomings leads me to break from Neorealism in my second
alignment politics model.
In my second model—the common sense social emotions model—I select emotions more
or less without deductive scientific method. I nevertheless do have reasonable confidence in
them. For not only are the emotions of ideational fear and ideational anger used implicitly or
explicitly by all four of the materialist scholars whose models I detail, but I utilize academic
literature from the field of Social Psychology (SP) to support my argument. Similarly, though
my utilization of ideational gratitude is not found in the four IR modelers’ research, I do explain
its nature in terms of SP scholarship.
I call my third model the collective self-esteem model. Whereas collective self-esteem is
a hidden hand in all alignment politics models I examine, my third model is unique because I
explicitly grant national interest status to the interest of collective self-esteem. Another critical
factor in the model is ideational trust—an important “emotional attitude.” It helps to explain
why security expectations come into being in alignment politics. Trust also gives those security
expectations emotional meaning. Consequently, when trust is proven by events to have been
misplaced, security expectations degrade and critical actors suffer ideational humiliation.
Humiliation causes those feeling it to undertake an aggressive defense to protect their collective
self-esteem. They lash out with pen and tongue at the semi-defecting alignment partner. By
- 59 -
denigrating and diminishing the offender, these people establish to themselves their superiority
over the other. This strategy diminishes the humiliation, because the actions of an “inferior”
state are less likely to be interpreted as being humiliating. Unfortunately, it also triggers intra-
alignment ideational friction.
As I argue in this and the following chapters, a concrete focus on expectations and
emotions reveals serious shortcomings in all of the alignment models I examine, including in at
least one of my own. Most importantly, it uncovers the models’ thorough dependence on the
existence and outcomes of both expectations and social emotions. Such ubiquitous reliance on
ideational phenomena suggests scholars should not begin deducing alignment models from
metatheories incapable of explaining and encompassing them. I find Neorealism, in particular,
especially unsuited to the task, though Neoliberalism also suffers what I consider to be fatal
flaws.
112
112
I note one telling example for now. Namely, neither material metatheory can permit actor preferences to change.
But expectations influence actor preferences and, as my examination of alignment security expectations shows,
Diagram 2.1
Summary of Emotions in the Three Alignment Models After One State Hedges
(Pertains only to normal alignment crises as prevalent in the modern world; not to immediate, existential crises.)
Name of
Model
Emotion(s)
Causing
Material
Cooperation
Emotion(s) Causing
Material Conflict
Emotion(s) Causing
Ideational Cooperation
Emotion(s) Causing
Ideational Conflict
Delimited
Balance of
Threat
Material Fear NA* Ideational Friendliness Ideational Enmity
Common
Sense Social
Emotions
Material Fear
Ideational Anger &
Ideational
Vengefulness
Ideational Gratitude &
Ideational Fear
Ideational Anger &
Ideational
Vengefulness
Collective
Self-Esteem
Material Fear
Ideational Humiliation
& Ideational Desire to
Preserve Collective
Self-Esteem
Ideational Fear,
Ideational Trust, &
Ideational Desire to
Promote Collective
Self-Esteem
Ideational Humiliation
& Ideational Desire to
Preserve Collective
Self-Esteem
* No material conflict in Neorealist alignments, because the state fearing abandonment always exhibits material
loyalty to its hedging partner.
- 60 -
II. Expectations and Emotions are Everywhere
Expectations and emotions are ubiquitous in both IR theory and in models of alignment
politics. Regarding Japan-Korea relations, I argue their interaction structures an important causal
mechanism that explains how converging and diverging security interests and policies lead to
cooperation or friction over the history issue.
Expectations and emotions are thus the cornerstones of the three theoretical models I
develop in the following chapters. Specifically, expectations—or, more accurately, unrealistic
expectations—set the stage for deleterious emotional reaction once the expectations are smashed
by reality. But expectations of security support do more than just lead to conflict. When they
are satisfied or exceeded by an alignment partner, they produce positive emotions and therefore
lead to good ideational relations.
I therefore face a dual task. First, I have to argue that states share expectations and
experience emotions. In particular I must explain why Japan and Korea each tend to expect that
the other will consider the security needs of the alignment their relationship comprises. Many
scholars will no doubt find this contention puzzling, since Japan and Korea have only recently
begun contemplating a security commitment. Second, I need to detail convincingly how
emotional reaction to unmet security expectations leads to ideational friction, and how the
affective response to satisfied security expectations induces ideational harmony. I begin these
tasks below, and keep them in mind as I construct my three models.
Before discussing expectations and emotions in both IR theory and in models of Seoul-
Tokyo affairs, I wish to reemphasize that my project’s DV is primarily intended to address
change all the time. Since nearly every longitudinal study of alignments will encounter changing security
expectations, it will therefore also find changing actor preferences. Hence, a specific focus on actor expectations
reveals changing actor preferences, and the consequent incompatibility of the material theories for alignment politics
modeling.
- 61 -
Japan’s and Korea’s ideational cooperation and conflict. Though some of its components—e.g.,
canceled and concluded summit meetings—are certainly complementary to material relations,
the DV should not be considered something representing material cooperation and conflict. Thus,
while my models permit predictions about the tenor of material relations, my statistical analysis
provides a firm measurement only for the DV of Japan and Korea’s bilateral history issue.
A. Emotions and Expectations in IR Theory
International Relations (IR) literature has remarkably little to say about emotions or
expectations. One suspects it should, however, because emotions always play a key role in
theory,
113
as do expectations. I discuss here how intellectuals across time, and scholars within IR,
have explicitly and/or implicitly utilized emotions and expectations to explain the world around
them. As I show via the below discussion of theory, modern IR theory slowly, and mostly
steadily, has been returning to its human nature-based, emotional and expectational roots, after
undertaking a comparatively inhuman scientific turn in the late 1970s with the advent of
Neorealism. I make this claim very reservedly, however, as I wish to imply neither that the
transformation of the field is anywhere near complete, nor that it is ever likely to become so.
NEOREALISM
Neorealism is dependent on two causal emotions related to state survival: fear and
insecurity.
114
These emotions emerge in response to the terrible realities of existence within an
anarchic international system and, in tandem with expectations about the meaning of power,
incentivize almost all of the significant conflict and cooperation seen in an ideal Neorealist
113
Neta C. Crawford, “The Passion of World Politics: Propositions on Emotion and Emotional Relationships,”
International Security 24, no. 4 (Spring 2000): esp. 116-123.
114
Crawford lists nationalism and its related love and hate components as Realism-sanctioned emotions. I exclude
nationalism and the like because I am discussing Neorealism, or Structural Realism, which I believe cannot
scientifically accommodate them. Ibid., 116-117.
- 62 -
world.
115
Most significantly, states completely unable to experience fear or insecurity would not
balance materially against each other, as concerns about the balance of power in the international
system simply would not exist.
116
Likewise, states incapable of anticipating that their own power
provides security, while the power of others leads to insecurity, would neither try accreting or
consolidating it, nor be prodded by it, when in the hands of others, to experience fear and
insecurity.
While general expectations about power exist in ideal Neorealist worlds, whether
expectations about the likelihood of conflict with specific others will be present is less clear and
probably depends on the type of Neorealism. It is hard to say, for instance, that Defensive
Realism’s ideal will produce expectations that a given state will pursue an aggressive agenda. It
seems most likely that states operating under its logics need not expect anything of others.
117
The tendency of most states to hunker down and be satisfied with what they have already secured
will be counterbalanced by anomalous aggressors. Since states have no capability of reliably
sorting out in a timely and therefore safe manner the aggressors from the status quo states, those
that somehow developed expectations of completely satisfied neighbors will have been the most
likely to get destroyed over time. Once-held expectations should have disappeared along with
their sovereignties. Conversely, there is no reason to suspect that states in a largely status quo
system would ever develop expectations that their neighbors will be particularly aggressive,
115
Waltz, the founder of Neorealism, is exceptionally clear about how the international system socializes states to
care about the balance of power. See Kenneth Waltz, “Reductionist and Systemic Theories,” in Neorealism and its
Critics, ed. Robert O. Keohane (New York: Columbia University Press, 1986), 62-67. Mearsheimer downplays
social process in his brand of Offensive Realism. Nonetheless, he acknowledges the role of socialization: Great
powers are “concerned mainly with figuring out how to survive…” and they “quickly realize” how best to survive.
Moreover, as he concedes, the main difference between offensive and defensive realisms is regards only “how much
power states want.” Mearsheimer, The Tragedy of Great Power Politics, 21-22.
116
Richard Ned Lebow, A Cultural Theory of International Relations (2008; repr., New York: Cambridge
University Press, 2010), 88-93.
117
Waltz notes that things related to expectations—“ideology, form of government, peacefulness, bellicosity, or
whatever”—do not matter; only capability in the form of power does. Kenneth Waltz, “Political Structures,” in
Neorealism and its Critics, ed. Robert O. Keohane (New York: Columbia University Press, 1986), 93.
- 63 -
because these expectations will almost always go unmet. States in pure Defensive Neorealist
systems will be insecure and feel fear throughout all their interactions. But their expectations
regarding others, to the extent they have them, should be limited to expectations about the
unreliability and dangerousness of having expectations.
Offensive Realism is far more likely than its defensive cousin to contain expectations
about others. These expectations will exist in a world defined by this brand of Neorealism for
the very same reason that emotions like fear and insecurity exist. Namely, just as those states
capable of feeling fear and insecurity will be more liable to survive than those that cannot, states
able to learn to expect aggression from all neighbors will also be more likely to survive and
prosper in such a competitive system. States likewise must have developed expectations about
the desirability of offensive strategy. Otherwise they would not be inclined strongly to accrete
power when opportunity arises. In all, expectations of aggressive states and zero-sum
calculations will be the norm. Remarkably, this means states will experience emotions like
surprise and puzzlement whenever other states do not act aggressively at opportune moments.
118
This is clearly the case, because Neorealism permits emotions where there are expectations, and
Offensive Neorealism’s logics allow for a deduction of expectations about aggressive others.
HUMAN REALISM AND CLASSICAL SCHOLARSHIP
Neorealism’s scientific approach to understanding international politics differs greatly
from its forebear, Human Realism. In contrast to the socializing role Neorealism’s anarchy plays
in producing a dangerous world, Human Realism argues that problems emerge from human
nature. Hans Morgenthau is clearest on this point among the post-World War Two Realists,
119
118
Indeed, Mearsheimer is surprised and puzzled that Germany did not attack France in 1905 after Russia had been
defeated by Japan. Mearsheimer, The Tragedy of Great Power Politics, 10.
119
Hans Morgenthau, Politics Among Nations: The Struggle for Power and Peace (New York: Alfred Knopf, 1978),
esp. 4-5.
- 64 -
though the role of human nature is a trend of thought extending from ancient Greek philosophers
like Thucydides, Plato, and Aristotle, through European intellectuals such as Machiavelli,
Hobbes, and Kant.
While all of these great thinkers recognized the general problem of anarchy, or at least of
weak and easily manipulable government at the domestic level, they never emphasized how a
lack or absence of government might socialize states, state-like units, or individuals that initially
were not influenced by human emotions or thought. Instead, anarchy was seen as problematic
precisely because it allowed the worst of human nature to go unchecked. In turn, human nature
within anarchy injected extremely deleterious kill-or-be-killed type logics into the relations of
actors. This was the case even for Hobbes, the scholar who placed the most emphasis on the
catalyzing impact of anarchy.
120
The question at hand is whether theories imparting causal power to human nature can
abide either emotions or expectations, or both. The answer is most certainly yes, since
phenomena like emotions and expectations are intrinsic characteristics of human nature. Among
emotions, a quick look at the classic human nature literature suggests fear and insecurity are the
most abundant. They are evident almost everywhere in the literature and a few examples suffice
to illustrate the point. Most famously, Thucydides argued, “What made [the Peloponnesian War]
inevitable was the growth of Athenian power and the fear which this caused in Sparta.”
121
Machiavelli similarly informed that a prince who wishes to stay in power “must have two kinds
120
Hobbes was nevertheless unique among the classic scholars. He emphasized that only a relatively small number
of men needed to have insatiable appetites and aversions for there to be a war of all against all in the state of nature.
Anarchy catalyzed the impact of these unsatisfiable men, however. Since men of average appetites and aversions
could never tell if they were in the presence of an insatiable man, and since all men are equal enough in capability to
kill each other, even men of average appetites had to war on everyone in order to best guarantee their own survival.
For a quick reading of this line of thought see C.B. Macpherson, “Reading 2.1: The Theory of Human Nature in
Society,” in Reading Political Philosophy: Machiavelli to Mill, ed., Nigel Warburton, Jon Pike, and Derek
Matravers (New York: Routledge, 2000), 100-105.
121
Thucydides, “Introduction,” ed. M. I. Finley, trans. Rex Warner, in History of the Peloponnesian War, rev. ed.
(New York: Penguin Books, 1972), 49.
- 65 -
of fear: one internal as regards his subjects, one external as regards foreign powers.”
122
In post-
war Human Realism, Morgenthau argued that leaders’ fear of changes in the balance of power
could lead to wars.
123
Other emotions and similar elements of human nature are also apparent in the classic
literature. Plato argued the causes of war were found in man’s need to care for his corporeal
body: “Only the body and its desires cause war, civil discord, and battles, for all wars are due to
the desire to acquire wealth, and it is the body and the care of it to which we are enslaved, which
compel us to acquire wealth…”
124
For Aristotle, well-formulated society was needed to contain
man’s passions. Without adequate society humans are “the most savage, the most unrighteous,
and the worst in regard to sexual license and gluttony.”
125
Kant contended that war “appears to
be engrafted on human nature; it passes even for something noble, to which the love of glory
impels men…”
126
Sentiments identical to these are seen in Morgenthau’s explanation for
conflict: “The drives to live, to propagate, and to dominate are common to all men.” Moreover,
“The tendency to dominate, in particular, is an element of all human associations, from the
family through…to the state.”
127
Among the classical researchers Hobbes clarified emotions best. He called them
passions and divided them into appetites and aversions. The most important appetites are the
“desire of Power, of Riches, of Knowledge, and of Honour.” These are reducible to desire for
122
Niccolo Machiavelli, “The Prince,” ed. Max Lerner, in The Prince and the Discourses (New York: Random
House Inc., 1950), 67. In some sections of The Prince fear is a near constant topic. See chapters XV-XIV, for
instance, in ibid., 56-77.
123
He argued this in terms of the root cause of World War One, for instance. Morgenthau, Politics Among Nations,
207.
124
Plato, Phaedo, ed. Stephen M Cahn, in Classics of Western Philosophy, 6
th
ed. (Indianapolis, IN: Hackett
Publishing, 2002), 54.
125
Aristotle, The Politics,” ed. Trevor J. Saunders, trans. T. A. Sinclair (1981; repr., London: Penguin Books, 1992),
59-61.
126
Immanuel Kant, “Perpetual Peace,” ed. and trans. Lewis White Beck, in Kant: On History (Upper Saddle River,
NJ: Prentice Hall Inc., 2001), 110-111.
127
Morgenthau, Politics Among Nations, 37.
- 66 -
power, Hobbes argued, because they “are but several sorts of Power.” Paramount among
aversions are “Feare of Death” and wounds, though aversions of losing power and any of its
subcomponents are also obviously key.
128
Importantly, the “principall causes of quarrel” are
“First, Competition; Secondly, Diffidence; Thirdly, Glory.”
129
Expectations are less apparent than emotions in the classic literature. Interestingly, they
are most evident in Hobbes. This is surprising, because the deductive technique he employed
imparts the sense that Leviathan is more scientific than that of the other classic texts, and
expectations are not something one readily expects in scientific studies. Nevertheless, there are
at least three main kinds of expectations in his tome. The most interesting are expectations
associated with man’s escape from the state of nature. Remarkably, Hobbesian Man is naturally
endowed with the capacity to anticipate the Leviathan’s ability to enforce its sovereignty at the
domestic level.
130
If he does not expect the sovereign to perform effectively and immediately
after the social contract is signed, Hobbesian Man will obviously never accept it and the state of
nature will continue.
131
Hobbes’s second kind of expectation is very much in line with Offensive Neorealism’s
expectations—i.e., a learned behavior that aggression is the best way to counter insecurity and
fear. Hobbes is clear on this point, saying: “And from this diffidence of one another, there is no
128
Hobbes writes extensively about a considerable range of passions and aversions. Thomas Hobbes, Leviathan, ed.
C.B. Macpherson (New York: Penguin, 1968), 118-130, 139, 160-168, 183-188.
129
Diffidence means distrust or suspicion in this context: “[Competition], maketh men invade for Gain; [diffidence],
for Safety; and [glory], for Reputation.” See Ibid., 184-185.
130
Hobbesian Man’s expectations about the Leviathan cannot be learned behavior. If learned, man could never have
left the state of nature on the first occasion. The ability to have expectations about the Leviathan must be part of
mankind’s inborn nature.
131
For a nice discussion of expectations in Hobbes’s Leviathan see Susan Sreedhar, Hobbes on Resistance: Defying
the Leviathan (Cambridge: Cambridge University Press, 2010), 40-47. Sreedhar argues that man does not surrender
to the social contract his right to self-defense. If true, then the Leviathan would not have to be completely effective
at protecting people from each other; effective enough would suffice to permit the escape from the state of nature. I
do not consider Sreedhar’s argument at odds with mine, though accepting her reading of Hobbes requires that I alter
my argument a bit. To restate: If Hobbesian Man expects nothing from the Leviathan—if he places no trust in its
ability to be effective at all—then the social contract will never be worth more than words on a page, if it can even
get to the point at which words are written down.
- 67 -
way for any man to secure himself, so reasonable, as Anticipation.”
132
Perhaps even more so
than fear and insecurity, anticipation of physical violence by others is paramount in Hobbes’s
theory; it provides the impetus for all men to strike each other first, thereby causing the war of all
against all. Without it, Hobbes could not have produced the antecedent of Offensive Neorealism.
Hobbes’s third set of expectations pertains to individual satiation. Hobbes argued that
most people have upper limits regarding how much power they desire. Indeed, only madmen
either do not have limits or have theirs set absurdly high.
133
This means most men have
expectations about what is an appropriate amount of power, given their own, unique
circumstances.
134
Unfortunately, madmen do exist. And because of them, when men are outside
of society—i.e., when they are in the state of nature—all men by necessity become insatiable and
uncompromising power-seekers in order to save themselves.
135
While only Hobbes emphasizes how expectations allow man to escape the state of nature,
expectations that cause fear and insecurity and make anarchy so damaging to peace are
practically ubiquitous in the human nature literature. These are the expectations that reify
violence. In Thucydides, for example, expectations of other city-state’s hostile intentions
produced the “necessary law of nature” in which one must “rule whatever one can.”
136
Machiavelli informed rulers they must “learn how not to be good” and to “do evil” if they
wished to maintain their power in a world in which there are very many bad men and no ultimate
132
Hobbes, Leviathan, 184.
133
“And to have stronger, and more vehement Passions for any thing, that in ordinarily seen in others, is that which
men call Madnesse.” Ibid., 139.
134
Machiavelli has a darker view of man, claiming “when men are no longer obliged to fight from necessity, they
fight from ambition, which passion is so powerful in the hearts of men that it never leaves them, no matter to what
height they may rise.” Niccolo Machiavelli, “The Discourses,” ed. Max Lerner, in The Prince and the Discourses
(New York: Random House Inc., 1950), 208.
135
“If others, that otherwise would be glad to be at ease within modest bounds, should not by invasion increase their
power, they would be able, long time, by standing only on their defense, to subsist.” Hobbes, Leviathan, 185.
136
Thucydides, “The Melian Dialogue,” ed. M. I. Finley, trans. Rex Warner, in History of the Peloponnesian War,
rev. ed. (New York: Penguin Books, 1972), 402-404.
- 68 -
recourse against them beyond fear and violence.
137
In more recent scholarship, Morgenthau
noted that states care about interest defined in terms of power and, therefore, the balance of
power. The “universally valid,” “objective category” of interests defined as power means that
state expectations about it are also universal. For if there were no expectations about the
meaning of power in the international system, states would not care about it beyond that
predicted by chance.
Other kinds of expectations are less common. Nonetheless there are clearly quite a few.
Kant’s declaration that man’s quest for glory leads to war means men have expectations about
being glorified. Machiavelli likewise noted that men are motivated and satiated by desire for
honor, implying they have expectations about it.
138
Returning to Hobbes, he finds that men act
in order to attain “Fame from new Conquest…ease and sensuall pleasure…admiration, or being
flattered for excellence in some art, or other ability of the mind.”
139
Men incapable of expecting
glorification, honor, fame, pleasure, admiration, flattering, or any other impetus for action would
never start wars or otherwise set out in hope of attaining these rewards.
CLASSICAL LIBERALISM AND NEOLIBERALISM
Like Neorealism and interpretations based on human nature, liberal schools of IR such as
Classical Liberalism and Neoliberalism employ emotions and expectations. Classical
Liberalism’s best explanation for the democratic peace is, for example, blatantly built upon
expectations while being almost equally, if less obviously, reliant on emotions. To illuminate,
consider Bruce Russett’s well known argument that normative factors prevent democratic states
from warring against each other. Russett contends that, as an ongoing “presumption of amity”
137
See, for example, Niccolo Machiavelli, “The Prince,” ed. Max Lerner, in The Prince and the Discourses (New
York: Random House Inc., 1950), 56, 64-65, 71-72.
138
Honor is a common theme in “The Discourses,” though not in “The Prince.” For one example see Machiavelli,
“The Discourses,” 163.
139
Hobbes, Leviathan, 161.
- 69 -
between democracies is solidified, “wars against other democratic states are neither expected nor
considered legitimate…” Expectations that war between democracies cannot and will not
happen create the political space in which democratic “norms of regulated political competition,
compromise solutions to political conflicts, and peaceful transfer of power are externalized by
democracies…” In a virtuous cycle of nonviolence, these internationalized domestic norms
reinforce expectations that war is not possible between democracies.
140
Though not obvious at first look, Russett’s norm-based argument is heavily reliant on
emotions. The dependence emerges from the enforcement role emotions serve in relation to
norms. Namely, as argued by Jonathan Mercer, powerful emotions such as “embarrassment,
guilt, or shame” will be felt by deliberate norm transgressors.
141
Similarly, observers party to the
norms will feel “anger or indignation” in response to breaches of norms. Reactions like these are
an important hidden-hand explaining why actors are constrained by well-internalized norms.
Mercer even goes so far as to claim that “someone who had no emotion would not be influenced
by norms.”
142
Emotions and expectations also comprise much of Neoliberalism. To make better sense
of these phenomena within Neoliberalism, I explore how they might operate at three of
Neoliberalism’s “levels.” These levels are 1) the systemic level, in which Neoliberalism relies
on the logics of Neorealism; 2) the international regime level, in which Neorealism’s systemic
140
Russet, Grasping the Democratic Peace, 30-38, 88-92.
141
Jonathan Mercer, “Approaching Emotion” (paper presented at the International Studies Association conference in
San Diego, CA, April 25, 1996), 11. Though unpublished, Mercer’s paper clearly marked a watershed in the
exploration of emotion in IR. It is cited frequently in the literature that followed. I would like to thank Dr. Mercer
for emailing to me a draft of the paper.
142
Ibid., 11. While I agree with Mercer that emotions enforce norms, I can foresee that thoroughly internalized
norms would not depend on emotions, except in the broadest sense. In Canadian-US relations, for example, the US
does not ever consider invading and conquering Canada. Emotional responses are therefore never activated. They
would become paramount only when a handful of Americans got to the point when they actually began to espouse,
with real intent, an invasion of Canada.
- 70 -
logics are mitigated;
143
and 3) the transnational level, where domestic phenomena influence
international relations.
One set of emotions and expectations within Neoliberalism comes from Neoliberalism’s
general acceptance of Neorealism’s contentions about anarchy and balance of power.
144
States in
ideal Neoliberal worlds will therefore be capable of feeling material fear and insecurity at
Neoliberalism’s first level, just as states in Neorealism do. If suffering the logics of a Neoliberal
theoretical foundation based on Offensive Realism, states will also be capable of holding clear
expectations about their neighbors, and would necessarily act offensively to attain power were it
not constrained by regimes at Neoliberalism’s second level.
Emotions and expectations also exist in Neoliberal systems at the second level. Unlike
those at the Neorealist-like first level, these stem from the system-level logics of Neoliberalism
that differ from Neorealism. Most significant among these differences is the recognition by
Neoliberals that cooperation unrelated to the current distribution of power can take place
internationally.
145
Cooperation is most likely when international regimes exist and promote legal
143
Keohane defines regimes as “institutions with explicit rules, agreed upon by governments, that pertain to
particular sets of issues in international relations.” Robert O Keohane, “Multilateralism: An Agenda for Research,”
International Journal 45, no. 4 (Autumn 1990): 733.
144
See, for example, Keohane’s discussion of building on the basic foundation of Neorealism in Robert Keohane,
“Theory of World Politics: Structural Realism and Beyond,” in Neorealism and its Critics, ed. Robert O Keohane
(New York: Columbia University Press, 1986), 192-197. Keohane and Nye admit “the result of our synthetic
analysis in Power and Interdependence, and of subsequent work such as Keohane's After Hegemony, has been to
broaden neorealism and provide it with new concepts rather than to articulate a coherent alternative theoretical
framework for the study of world politics. See Robert O Keohane and Joesph S Nye, Jr., “Review: Power and
Interdependence Revisited,” International Organization 41, no. 4 (Autumn 1987): 733; Robert O Keohane and
Joseph S Nye, Power and Interdependence, 3
rd
ed. (New York: Longman, 2001). Wendt argues that “Neoliberals
are caught in a Realist trap.” See Alexander Wendt, Social Theory of International Politics (Cambridge: Cambridge
University Press, 1999), 136-137.
145
Most regimes reflect their contemporary international power configurations when first established. But they
often outlive their formative distribution of power, because they are useful to states. Moreover, new regimes are
difficult to create, and powerful states will often find old regimes more preferable to risking regime breakdown or
the absence of a regime. Regimes also enjoy unexpected longevity, because each individual regime is often seen as
part of the whole system of international regimes. See Robert O Keohane, After Hegemony: Cooperation and
Discord in the World Political Economy, Princeton Classic ed. (Princeton, NJ: Princeton University Press, 2005),
97-106.
- 71 -
liability, lower transaction costs, and reduce uncertainty.
146
Expectations are the lynchpin of all
voluntary regimes. Indeed, by Stephen Krasner’s definition of regimes, states cannot agree to
form or participate in one unless their “expectations converge in a given area of international
relations.”
147
It follows that too much divergence of expectations among states regarding what a
particular regime is about and what it is capable of accomplishing would result in its demise.
Similarly, states incapable of developing expectations would not be able to learn how to
cooperate via iterated prisoner’s dilemma games. Not only would they fail to understand the
meaning of emerging patterns of cooperation, but they could not recognize how their own
defections regularly lead to short term punishment and worse outcomes over the medium and
long terms. Surely they could never comprehend “reputation” and “credibility,” two related
phenomena important to the Neoliberal understanding of cooperation.
148
Yet, as the Neoliberal
project has clarified, these kinds of learning take place.
149
At the absolute minimum,
expectations will always exist in economics: As argued by Robert Axelrod and Keohane, “in
economic relations actors have to expect that their relationships will continue over an indefinite
period of time,” because it is very unlikely that one party will be able to eliminate the other
economically.
150
Emotions also exist at the level of Neoliberalism’s second level. By far the most
important is the emotion of desire for economic well-being, which fosters all of the important
146
Keohane, After Hegemony, 88-96.
147
Krasner defines regimes as “sets of implicit or explicit principles, norms, rules, and decision-making procedures
around which actor’s expectations converge in a given area of international relations.” Stephen D Krasner,
“Structural Causes and Regime Consequences: Regimes as Intervening Variables,” in International Regimes, ed.,
Stephen D Krasner (1983; repr., Ithaca, NY: Cornell University Press, 1984), 2.
148
For reputation, see Keohane, After Hegemony, 105-106. For credibility, see Robert O Keohane and Joseph S Nye,
Jr., “Power and Interdependence in the Information Age,” Foreign Affairs 77, no. 5 (September/October 1998): 90.
149
See for example, Robert Axelrod and Robert O Keohane, “Achieving Cooperation under Anarchy: Strategies and
Institutions,” World Politics 38, no. 1 (October 1985): 226-254; Keohane, After Hegemony, 65-84. For an alternate
take, see Robert Powell, “Absolute and Relative Gains in International Relations Theory,” American Political
Science Review 85, no. 4 (December 1991): 1303-1320. Powell argues that learning is most likely to take place if
the cost of war is high.
150
Axelrod and Keohane, “Achieving Cooperation under Anarchy,” 232.
- 72 -
non-Neorealist conflict and cooperation in Neoliberalism.
151
While economic self-interest is
generally thought of as rational and timeless, Mercer points to the work of economist Albert
Hirschman to argue it was not always thought so dispassionate. Change came only when the rise
of capitalism fostered a “semantic shift” in which the desire for economic success came to be
seen not as a passion but as “a narrow understanding of interest.”
152
The desire for economic well-being emerges in the space Neoliberals perceive to exist
between the security and economic spheres. Neoliberal perceptions of the disconnect between
the two realms are purported to be somewhere between a “neat dividing line”
153
and “no clean
analytical line.”
154
Regardless of the division, state interests transform considerably in it. Most
important in terms of IR theory discussion to date, the space permits states to begin caring about
absolute gains instead of just relative gains. In terms of emotionality, the more states come to
view economics as being unrelated to military power, the more they learn to desire economic
growth for its own sake. Thus is the desire for economic well-being fostered.
Beyond its inclusion of desire for economic well-being, it remains doubtful whether
Neoliberalism can accommodate other non-Neorealist emotions at the second level. The
problem is that Neoliberalism is constrained by an emotionality-focused version of
Mearsheimer’s and Wendt’s “Neorealist trap.”
155
Specifically, as Neoliberals come to view
immaterial emotions as causally effective, material desires lose their explanatory power. The
151
Desire for economic well-being differs from Neorealism’s desires. In Neorealism, economic growth is desired
only because it mitigates material fear and insecurity.
152
Mercer notes that famed philosopher and economist Adam Smith was part of this process. Smith “squeezed
human emotions down to one—the drive for economic well being. [He] substituted passion and vice with interest
and advantage.” See Mercer, “Approaching Emotion,” 3-4. For Hirschman, see Albert O Hirschman, The Passions
and the Interests: Political Argument for Capitalism before its Triumph (Princeton, NJ: Princeton University Press,
1977.) For a similar explication of wealth in ancient Greece, see Lebow, A Cultural Theory of International
Relations, 169.
153
Mearsheimer, “The False Promise of International Institutions,” 20.
154
The line is blurred because “information” can sometimes transform even the logics of international security into
something more akin to economics. Keohane and Martin, “The Promise of Institutional Theory,” 42-44.
155
The term is Wendt’s, but he notes Mearsheimer is aware of the trap, too. Wendt, Social Theory of International
Politics, 136-137; Mearsheimer, “The False Promise of International Institutions,” 20.
- 73 -
threshold is thin, and is crossed once theory accepts that ideational emotions can compel action
at odds with the imperative of material interest. The line has been crossed, for example, when a
state acts angrily and vengefully to recoup lost economic well-being from the source of a
perceived economic injustice, instead of from a less costly source, like an uninvolved third state.
As I detail later, once immaterial emotions can overpower material emotions and interests
within a given theory, all scholars must shift their investigation toward the question of how
emotions contribute to state perceptions about the meaning of material power. To restate the
point more illustratively, the Neorealist and Neoliberal meanings of material power collapse, as
it becomes clear that material power means something remarkably different to State Z, when it is
wielded by states X and Y, where X is angry and vengeful toward Z, while Y is not. In effect,
ideational emotionality transforms materialist scholars into Constructivists, a fact that holds
unless the separation of material security and material economics is somehow complete within
Neoliberalism’s second theoretical level.
Regarding scholarship which does theorize coherently a clear dividing line between
material security and economics, it becomes possible to imagine other emotions working
alongside desire for economic well-being at Neoliberalism’s second level. Much as was the case
with Classical Liberalism, these emotions are produced whenever expectations go unmet.
Regime defections that disappoint expectations might produce feelings of regret, shame, and
disappointment, as well as emotions related to anger and vengefulness. Punishment meted out
by jilted states will be a product of emotions, most broadly defined, as even the most materially
grounded states will be unable to react to defections if they cannot feel.
156
The defecting state
156
At minimum, emotions related to desire to uphold self-interest would be responsible for punitive action, or for a
decision to ignore the defection.
- 74 -
meanwhile may similarly evince shame or embarrassment, or perhaps even pride, at its
defection.
157
It is import to note, however, that any reactionary ideological emotions felt by defecting
states will not necessarily play a crucial role in enforcing ongoing regime participation, as do
similar emotions in Classical Liberalism’s Democratic Peace Theory. That is because
Neoliberalism’s expectations and norms are much “thinner” than those upholding a prohibition
on war between democracies.
158
Indeed, Neoliberalism’s ongoing cost-benefit rationalization for
international cooperation via regimes does not leave enough room to deduce comfortably that
non-material factors like ideological emotions are likely to play an important enforcement role
regarding self-corrective behavior.
159
Even when discussing matters likely to be laden with
social emotion, Neoliberal scholars downplay non-material explanations. Witness Keohane’s
discussion of reputation, in which he emphasizes that states will care about their reputations
because “the costs of acquiring a bad reputation as a result of rule-violations are imposed
specifically on the transgressor.” While he understands that “personal honor and self-respect”
might play a role, his focus is squarely on the likelihood of being materially punished in iterated
games.
160
Neoliberalism is potentially most laden with expectations and emotions at the third
level—the transnational level—where all kinds of entities operating domestically can impact
interstate politics. As clarified by Keohane and Nye, international politicking is usually marked
157
It strikes me that pride tied to nationalism would in most cases be a phenomenon of the third, or transnational,
level. But something akin to pride at a smart decision to serve state interests by leaving a regime upheld by weak
norms could easily take place at Neoliberalism’s second level.
158
Neoliberalism’s norms are thin in comparison to Democratic Peace norms, which are identical to Constructivist
norms. See Emanuel Adler and Michael Barnett, “Security Communities in Theoretical Perspective,” in Security
Communities, eds. Emanuel Adler and Michael Barnett (Cambridge, Cambridge University Press: 1998), 9-15; Peter
J Katzenstein, “Introduction: Alternative Perspectives on National Security,” in The Culture of National Security:
Norms and Identity in World Politics, ed. Peter J Katzenstein (New York: Columbia University Press, 1996), 20-22.
159
Self-corrective behavior by potential defectors contrasts with the emotions felt by jilted states.
160
Keohane, After Hegemony, 105-106.
- 75 -
by a “complex interdependence” in which multiple channels connect societies, there is an
absence of hierarchy among issues, and military force is not used.
161
The impact on agenda
setting of formal and informal organizations, and even individuals, operating at the domestic
level is responsible for much of the “noise” that prevents states from pursuing Neorealism’s
predicted and narrow interests internationally.
Keohane and Nye are mostly concerned about the influence on international policy of
businesses, labor organizations, bureaucracies, and legislative bodies. Significantly, considering
these entities means the level of analysis has shifted away from the apparently rational and
singularly focused state toward the more obviously emotional, expectant, and diverse humans
who run and comprise them. It becomes essential to consider, at minimum, the impact of leaders
and key members within many institutions, including the following: business; politically-
oriented organizations; local and federal bureaucracies; and the judiciary, legislature, and
executive administrations at all levels of society.
162
Naturally, the more influence individuals
exert within their organizations, the more important emotions and expectations become in any
complete explanation of international relations. By theoretically justifying causality in
international politics at a level where individuals are very likely to play important roles,
Neoliberals like Keohane and Nye are empowering expectations and emotions to almost the
same extent as did the ancient Greek and early European intellectuals discussed above.
163
161
Keohane and Nye, Power and Interdependence, 20-30.
162
Some might exclude the national executive office from this list. But even in this relatively small, state-focused
group, individuals will have unique world views and interests that conflict with the interests emphasized by
Neorealism. A President might prioritize reelection over the national interest, for instance. See the political
considerations of top leaders in Allison’s 3
rd
model, the bureaucratic politics model, in Graham T Allison,
“Conceptual Models and the Cuban Missile Crisis,” American Political Science Review 63, no. 3 (September 1969):
707-715.
163
It is generally understood that individuals at the domestic level will often be impacted by the logics of the
organization to which they belong, or for which they work. As the famous adage puts it, where one stands depends
on where one sits. I do not mean to imply that the logic behind organizational thinking is incorrect. Nevertheless, it
is undeniable that the lower one sets the level of analysis, the more important the analysis of individual humans
- 76 -
At ideal Neoliberalism’s third level, human nature is arguably most empowered within
the issue of nationalism. Indeed, even the quickest examination of interstate relations will show
that nationalism among domestic actors emerges quite often as a prime motivator for
international policies and events, and sometimes leads to international policies far removed from
Neorealism’s predicted interests.
164
Nevertheless, while Keohane and Nye mention nationalism
as an important force in terms of American-Canadian affairs, they spend practically no effort
discussing it in their theoretical discussion.
165
Perhaps their oversight occurred because it is
fairly obvious how nationalism works in complex interdependence.
Presuming the prior existence of nationalist tendencies within a given state, nationalism
may surge there when nationalists individually and/or in groups feel their nation has somehow
been slighted internationally. Nationalism is, in other words, an expectation-driven phenomenon.
Nationalists incapable of holding expectations about how their state should be accorded cannot
perceive transgressions against their state’s dignity. They likewise cannot evoke emotional
responses to disrespecting acts they are incapable of recognizing. In an important sense,
nationalism is an international norm that is not shared between two or more parties. Thinking of
nationalism as a one-sided norm clarifies that emotions are produced from violations of the norm,
and that most emotional reaction will take place in the one state party to the norm. When the
nationalist norm is disappointed via international actions and incidents, or ostensibly domestic
becomes. For organizational and bureaucratic models in IR see Ibid., 689-719. For an excellent Constructivist
interpretation of the Cuban Missile Crisis that includes a considerable emotions and expectations component, see
Jutta Weldes, Constructing National Interests: The United States and the Cuban Missile Crisis (Minneapolis, MN:
University of Minnesota Press, 1999). For a general introduction to putting human decision makers at the “center”
of theory, see Richard C Snyder, HW Bruck, and Burton Sapin, Foreign Policy Decision-Making: Revisited (New
York: Palgrave Macmillan, 2002).
164
Japan and Korea are a case in point, since their competing nationalisms can make cooperation difficult even
under pressing security conditions related to third parties.
165
Curiously, Keohane and Nye conclude that Canadian nationalism was a force the Canadian state learned to
control for its own myopic interests. In the Japan-Korea case, nationalism is often at odds with state interests
rationally defined. Keohane and Nye, Power and Interdependence, 145-190, esp. 185.
- 77 -
events in foreign states, nationalists are likely to feel emotions like embarrassment, shame, anger,
and/or indignation.
166
For Keohane and Nye, nationalism is primarily a phenomenon that impacts and is
impacted by commercial matters. Canadian nationalists, for example, were agitated when
American auto unions attempted to preserve jobs in the US by convincing negotiators to demand
Canadian-US pay parity. Similarly, the voyage of the US tanker Manhattan through Canadian
Straits in the Arctic triggered an outcry from nationalists who feared both a diminishment of
Canadian sovereignty and American commercial domination. Finally, US balance of payments
guidelines bothered nationalists after a Canadian politician politicized the issue domestically.
167
While the interplay between economics and nationalism certainly explains some
outcomes in Japan-Korea relations, Keohane and Nye’s discussion of process is too narrow to
capture an essential aspect of their nationalisms. Importantly, it misses the fact that nationalists
in the two states partake in what is effectively a back-and-forth conversation, albeit one ongoing
since at least the 1950s. They communicate via the practically unlimited number of information
channels that connect Japanese and Korean societies. These channels of course include official
and unofficial government communications at the executive, bureaucratic, and legislative levels,
as well as business, organizational, and personal contacts. Arguably more significant than these
usually deliberate and direct two-way contacts between societies are channels that spread
information in a way not necessarily intended to foster intersocietal conversation. Among the
most important are traditional news services that primarily report to subscribers in their home
country. Similarly, the foreign press; nationalist group announcements and actions; the
blogosphere, web-based social networks and chat rooms; cell phone text messages; and video,
166
An outburst of nationalism is evidence that norms related to nationalism have been violated. See Mercer,
“Approaching Emotion,” 11-12.
167
The cases discussed here are mentioned in Keohane and Nye, Power and Interdependence, 167, 176.
- 78 -
telephone, and in-person conversations spread news within, and therefore between, states. Given
the ease with which information spreads via the above channels, it is no surprise that states often
have tremendous difficulty controlling nationalist sentiment among their respective publics.
168
Perceived sleights committed by nationalistic individuals or organizations at any level of
one society have the potential to produce nationalistic responses within the other state. And
reactions can induce a feedback effect in which upward spiraling nationalism impacts policy.
Even unintentional insults can foment deliberately disrespectful responses. Politicians and
bureaucrats often play an important role in kick-starting a new round of the conversation, or
continuing it by responding angrily to the initial or prior insult. They might act provocatively
because they, too, are nationalistic or because they are using nationalism to gain or consolidate
advantage over political rivals. Whether leaders are levelheaded or not, surging nationalism can
quickly constrain their policy options, making bilateral relations very difficult to manage in
Seoul and Tokyo. In the event nationalistic conversation between the two states leads to foreign
policies out of line with each state’s respective material interests as established primarily within
Neoliberalism’s second level, then Keohane and Nye’s general predictions about the impact of
complex interdependence’s multiple channels have been supported.
CONSTRUCTIVISM
As suggested by the discussion of norms in Democratic Peace Theory and nationalism at
Neoliberalism’s third level, Constructivism’s explicit leverage of norms means it is the
theoretical school of IR most dependent on expectations and emotions related to them. For
Constructivism’s norms are sometimes very “thick,” just like the ideas that make inter-
168
China, for example, sometimes has a hard time containing popular nationalism. See Susan L Shirk, China:
Fragile Superpower (New York: Oxford University Press, 2007).
- 79 -
democratic war exceptionally unlikely or that comprise the one-sided nationalist norm.
169
And,
as discussed above, where there are robust norms, expectations are strong and emotions play an
important enforcement role.
Expectations and norms are in fact inextricable. As defined by Peter Katzenstein, norms
are “collective expectations for the proper behavior of actors with a given identity.”
170
Norms
are both constitutive and regulative. They constitute actors by “specify[ing] the actions that will
cause relevant others to recognize and validate a particular identity and respond to it
appropriately.” Norms are regulative because “they operate as standards for the proper
enactment or deployment of a defined identity.”
171
Though norms are causative, they are of
varying strength and often operate at cross-purposes. Because of these and other reasons, “the
presence of a norm does not dictate compliance.”
172
Significantly, states incapable of possessing expectations cannot recognize whether the
behavior of other states comports with a given identity, because they cannot anticipate the kinds
of actions that validate identity in an ideal Constructivist world. States will consequently be
unable to respond to each other appropriately, meaning it will be impossible for them to
reconstitute their identity through interaction.
173
This suggests that states incapable of expecting
will similarly fail to have an interstate identity, except perhaps occasionally and at the most
rudimentary level.
174
169
The strength of norms can vary greatly. Ronald L Jepperson, Alexander Wendt, and Peter J Katzenstein, “Norms,
Identity, and Culture in National Security, in The Culture of National Security: Norms and Identity in World Politics,
ed. Peter J. Katzenstein (New York: Columbia University Press, 1996), 55-56.
170
Katzenstein, “Introduction,” 5.
171
Second and third quote from Jepperson, Wendt, and Katzenstein, “Norms, Identity, and Culture in National
Security,” 54.
172
Ibid., 55-56.
173
Whether each state, individually, would be regulated by norms is beside the point. What is important is that
states would be unable to interpret the actions of other states.
174
A state incapable of expecting could probably sustain an enemy identity vis-à-vis a state that was currently
invading it.
- 80 -
While emotional capability plays a crucial role in empowering norms, discussion of
emotions does not play a prominent role in the large majority of Constructivist scholarship.
Constructivism’s omission is especially surprising, because ideas often have an undeniable
emotional component. Perhaps because of the obvious role of emotions within Constructivism,
there does seem to be an increasing awareness about them among major Constructivist theorists.
Here I compare three seminal Constructivist texts to suggest the trend, though I caution that it is
still very much an emergent phenomenon. I briefly examine Katzenstein’s 1996 edited volume,
The Culture of National Security; Alexander Wendt’s 1999 Social Theory of International
Politics; and Richard Lebow’s 2008 A Cultural Theory of International Politics.
Katzenstein’s volume is remarkable regarding emotion in that no variation of the word
appears in the index or the theoretical discussion. Indeed, the entire book appears to be absent
the term.
175
Emotions nevertheless clearly play an essential role for many scholars writing case
studies for the tome, and I highlight three cases. Dana Eyre and Mark Suchman’s chapter on
status as a motivator for states to acquire weapons is one example.
176
While the two researchers
do not examine whether the status gained by obtaining certain kinds of weaponry is more
important for bolstering regime legitimacy at the domestic level or sovereign legitimacy at the
international level, the general idea of increasing one’s status is reliant on an emotional
aspiration to obtain status. Even if status is desired for very narrow interests—e.g., decreasing
the likelihood of domestic rebellion against the regime—the heightened status will be intended to
operate by satiating the expectations of the target constituency. If the target audience is unable
175
An electronic search of Katzenstein’s entire volume through Amazon’s and Google Books’ respective text search
feature shows that “emotion” and “emotions” are not used.
176
Dana P Eyre and Mark C Suchman, “Status, Norms, and the Proliferation of Conventional Weapons: An
Institutional Theory Approach,” in The Culture of National Security: Norms and Identity in World Politics, ed. Peter
J Katzenstein (New York: Columbia University Press, 1996), 79-113.
- 81 -
to feel, efforts to fulfill its expectations for what a proper regime or state looks like will not be
successful.
Martha Finnemore’s chapter on humanitarian intervention is equally reliant on emotional
capability.
177
Interventions, which she claims are often a product of “shared values and
expectations” about “standards of justice,” would not happen if states were incapable of
experiencing the emotions produced when injustice is discovered.
178
Bereft of emotion, states
could neither perceive justice,
179
nor feel angry or disturbed by the results of whatever substitute
for justice they used to measure humanitarian tragedy. Finally, emotions are implicit in Richard
Price and Nina Tannenwald’s chapter on the “taboo” of using either chemical or nuclear
weapons.
180
Obviously, states incapable of feeling sorry, guilty, or some variant of regret-based
emotion would be completely unconstrained by a sense that it is wrong to use particular kinds of
weapons.
It is important to note the theme of expectations in the above cases. In the first one,
expectations regarding weaponry symbolically appropriate for possession by sovereign states
dictated purchases. The second case was marked by expectations about justice and the
desirability of infringing the sovereign rights of other states to resolve gross humanitarian
177
“Martha Finnemore, “Constructing Norms of Humanitarian Intervention,” in The Culture of National Security:
Norms and Identity in World Politics, ed. Peter J Katzenstein (New York: Columbia University Press, 1996), 153-
185.
178
Ibid., esp. 159; Robert C Solomon, A Passion for Justice: Emotions and the Origins of the Social Contract (New
York: Addison-Wesley, 1990).
179
One might argue that states could perceive of injustice in terms of scientific logics and measurements. An
instantiation of injustice could be said to take place whenever lawbreakers go unpunished via a legally proscribed
mechanism, for instance. Contentions like this become problematic, however, when one realizes that states would
still be compelled to differentiate between illegal acts. Without emotions states would be incapable of rank-ordering
crimes in terms of type, and perhaps intensity, and would therefore be incapable of determining whether a given
crime merits international intervention.
180
Richard Price and Nina Tannenwald, “Norms and Deterrence: The Nuclear and Chemical Weapons Taboos,” in
The Culture of National Security: Norms and Identity in World Politics, ed. Peter J Katzenstein (New York:
Columbia University Press, 1996), 114-152.
- 82 -
injustice. Lastly, strong expectations about which kinds of weapons are legitimate in warfare
proscribed the use of nuclear and chemical weapons.
In comparison with Katzenstein’s edited volume, Wendt’s accounting of international
politics elevates the causal role of emotions considerably. Though variants of the word “emotion”
are still missing from the index of Social Theory of International Politics (STIP), an electronic
search of the text turns up one specific use of it,
181
as well as a short discussion about how
emotions are produced whenever various kinds of human needs are or are not met.
182
Much
more important than this finding, however, is Wendt’s argument that states must satisfy the
collective self-esteem needs of their constituents if they are to survive.
183
While collective self-esteem—or “a group’s need to feel good about itself, for respect or
status” from others—is but one of four state interests in STIP, Wendt’s elevation of it to an
interest equivalent to physical security, autonomy, and economic well-being is truly
remarkable.
184
It represents an emphasis on emotions completely lacking in Katzenstein’s
theoretical account, since the concept of a given state’s collective self-esteem being satisfied or
dissatisfied internationally is blatantly an emotional phenomenon. Including it as an essential
interest means scholars relying on Wendt’s work must weigh equally certain emotional needs of
states alongside the other top-tier interests. Just as important, researchers must also endeavor to
understand the expectations states have regarding the amount of respect or status they require
from significant others. Though Wendt does not highlight the role of expectations regarding any
of the interests, it is obvious that states operating in an ideal world delineated by STIP are
181
An electronic search of Wendt’s entire volume was conducted through Amazon’s and Google Books’ respective
text search feature.
182
Satisfaction is produced when “material” needs are met; and “anxiety, fear, or frustration” when they are not.
Wendt, Social Theory of International Politics, 131-132. An electronic search of Wendt’s entire volume was
conducted through Amazon’s and Google Books’ respective text search feature.
183
Ibid., 237.
184
Note that Wendt suggests an inductive approach to establish how a given state prioritizes its interests. For this
recommendation as well as his introduction to the four interests see Ibid., 235-238.
- 83 -
capable of expecting. In the case of collective self-esteem, for instance, states incapable of
holding expectations about appropriate levels of respect or status would never know when their
collective self-esteem was being violated or satisfied.
If Wendt’s work is remarkable for elevating emotions to the same prominence as material
interests, Lebow’s is extraordinary for its claim that emotionality is more important than
materialism.
185
In A Cultural Theory of International Relations (CTIR), Lebow sees a world in
which the “universal human need[s]” of “spirit” and “appetite” provide the drive, or motivation,
for all action.
186
Spirit causes a desire for self-esteem, which is obtained via honor and standing
as recognized by others.
187
Appetite leads to aspirations for corporeal pleasure and material
things.
188
While Lebow argues that spirit and appetite are not emotions, they are clearly
inextricable from emotionality.
189
Therefore, even if one is talking about a war fought for
material gain, the conflict will be firmly rooted in the emotion of desire for material reward.
Spirit and appetite can be tempered by “reason,” an ability with which humans are
naturally endowed.
190
Through reason, man can “educate” his spirit and appetites.
191
Order
emerges as norms that constrain these otherwise destructive drives develop and solidify.
192
This
process creates “balance,” which is achieved only with great difficulty.
193
Through it,
185
The importance of emotion is reflected by the word’s presence in the index and its use, in various forms, at least
50 times within the text. Word count provided by Amazon’s text search feature. See “Look Inside!” at
Amazon.com.
186
Lebow, A Cultural Theory of International Relations, 62.
187
Ibid., 64.
188
Ibid., 72-76.
189
Ibid., 89.
190
Ibid., 78.
191
Ibid., 78-79.
192
Ibid., 77.
193
Ibid., 88.
- 84 -
individuals and societies can cooperate for mutual good, even if compromise does not satisfy the
immediate interest of all members.
194
The breakdown of order in CTIR is never far off, however. Imbalance can occur rapidly,
usually as a result of “the unrestricted pursuit by actors—individuals, factions or political units—
of their parochial goals.”
195
When these players break the rules for their own advantage, others
begin to fear for their own “ability to satisfy their spirit and/or appetites, and perhaps for their
survival.”
196
As more actors begin breaking rules to remain viable, a deleterious, reverse “norm
cascade” can take place.
197
The result is a fear-based world very much like that described by
Neorealist scholars.
198
But fear-based worlds are not the only possibility before order is established or after
balance has broken down. Worlds in which appetite or spirit dominate are also possible. In
appetite-based systems, desire for economic gain dictates policy preferences. Conflict and
cooperation are equally common and depend almost completely upon the temporary and shifting
interests of units. Appetite-based worlds are therefore very similar to a system predicted by
Neoliberalism.
199
Spirit-based worlds are not represented in mainstream IR theory. When spirit dominates,
claims Lebow, actors are willing to sacrifice themselves to obtain self-esteem in the form of
honor and/or standing.
200
Among likeminded members of a domestic or international
community, conflict in spirit-based worlds is likely to be “frequent, but the ends of warfare and
194
Ibid., 77.
195
Ibid., 89.
196
Ibid., 89.
197
Wohlforth makes the point about the deleterious norm cascade. See William Wohlforth, “A Matter of Honor,”
International Theory 2, no. 3 (2010): 468. For a norm cascade beneficial to order, see Martha Finnemore and
Kathryn Sikkink, “International Norm Dynamics and Political Change,” International Organization 52, no. 4
(Autumn, 1998): 887-917.
198
Lebow, A Cultural Theory of International Relations, 90.
199
Ibid., 75-76.
200
Lebow also lists “autonomy,” which I do not discuss since he does not usually emphasize it. See Ibid., 19.
- 85 -
the means by which it is waged tend to be limited,” somewhat like duels between two individuals
who are part of disagreeing factions. The limited nature of conflict between community
members stems from the fact that one’s honor and standing is to a considerable extent a product
of recognition by a given community. When violence moves beyond that which is acceptable by
the community, honor and standing are much less likely to be gained through it.
201
CTIR therefore predicts four ideal worlds: spirit, appetite, reason, and fear. Lebow
acknowledges, however, that the real world is a mixture of the four, which are best thought of as
“analytical constructs, useful to understand the behavior of societies, but without direct
correspondence in reality.”
202
Given that his purpose is to emphasize the central role of
emotionality in international relations, he focuses his case studies on the task of evidencing how
spirit has led to conflict over the past two millennia. This strategy is appropriate, because spirit
is far less dependent on material concerns than either appetite or fear.
The point of the above summary is, of course, that CTIR is thick with emotions. Though
Lebow does not emphasize the point, his theory is also rich in expectations. Indeed, practically
all action in an ideal spirit-based world is carried out with the expectation of achieving or
solidifying honor or standing. Similarly, units driven by appetite expect material reward, and
those in fear-based worlds strive only to improve their security. If actors cannot anticipate
rewards of honor or standing, material wealth, or security, they will not act determinedly to
obtain those things.
Importantly, and returning to a remark made earlier about Constructivism in general, if
individuals, factions, or states are incapable of expecting, they cannot anticipate or recognize
actions that validate identity in Lebow’s ideal worlds. This means, for instance, units otherwise
201
For community and the above quote, see Ibid., 71
202
Ibid., 93-94.
- 86 -
part of a spirit-based community would be unable to ascertain through interaction whether they
adhere to the same set of norms. As a result, each perceives and treats the other as an outsider.
Failing to reconstitute their mutual identity through the interaction, their spirit-based identity will
dissipate and eventually disappear.
203
Without expectations, spirit and appetite would arguably
remain as motives or drives, but “worlds,” if they could instantiate at all, would certainly not
exist for long.
B. Emotions and Expectations in Models of Japan-Korea Relations
As argued above, IR scholars always employ or depend on emotions and expectations in
their theories. Here I continue Chapter 1’s examination of the four models of Seoul-Tokyo
affairs, and highlight their sometimes explicit—but usually implicit—utilization of emotions and
expectations. This is important because my theoretical models rely on both phenomena, and I
wish to establish that much of what I do is simply make explicit that which the other scholars
either keep implicit or otherwise do not consider carefully enough. As I show, in all models
deliberate exploration of emotions and expectations is necessary for understanding the link
between the most important causal variables and the level of historical conflict or cooperation in
Japan-Korea relations. Focusing on emotions and expectations also clarifies whether there are
logical inconsistencies within a given model, and where the scholars are making unsubstantiated
assumptions.
CHA’S QUASI-ALLIANCE MODEL
Victor Cha’s quasi-alliance model makes explicit use of expectations and the emotions of
fear and insecurity, as well as implicit utilization of a generic and unspecified emotionality
component. Using fear and insecurity is an appropriate and adept move for Cha, a Neorealist,
203
The possibilities for cooperation and conflict would change. Regarding conflict, Lebow argues it would become
more brutal without the constraining norms found in honor societies. See Ibid., 71-72.
- 87 -
because they are the two emotions sanctioned by Neorealism. Applying the unstated set of
emotions is without theoretical merit, however.
Building on the alliance politics work of Glen Snyder, Cha argues explicitly that
expectations play an essential role in creating fear and insecurity via the
abandonment/entrapment mechanism I described in Chapter 1.
204
Though I detail Snyder’s work
Chapter 3, it is worth quickly introducing his thinking about expectations here. For Snyder,
security expectations are the foundation of any alliance, if formal, or alignment, if informal.
205
He defines the aligned aspect of both types of relationships as “expectations of states about
whether they will be supported or opposed by other states in future interactions.”
206
Uncertain
expectations are an important source of fear in alignment politics. Specifically, abandonment
fears emerge from “the uncertain expectation of another state’s support against an adversary.”
207
Cha’s method of utilizing expectations is interesting and worth exploring. Most
remarkably, it seems to deemphasize temporal causality. Thus, for Cha, the uncertainty of
expectations feels more or less unrelated to the more certain expectations that preceded it:
Strong expectations equate with no fear; uncertain expectations produce fear. There is very little
sense of the temporal linearity between once having enjoyed strong expectations of support and
then suffering the realization that those expectations were wrong.
208
But the most intense alignment fears are unlikely to come from a static uncertainty of
expectations. If X always has been, and remains, all but certain that Y will not help it in a
confrontation against Z, then X’s insecurity will not increase by much, if any, upon the continued
204
Cha, Alignment Despite Antagonism, 38-43.
205
Glenn Snyder, Alliance Politics (Ithaca, NY: Cornell University Press, 1997), 6, 9, 11, 21, 35-36.
206
Ibid., 6.
207
Cha, Alignment Despite Antagonism, 40.
208
For example: When Cha notes that “uncertain expectation of another state’s support against an adversary
generates abandonment fears,” he does not mention the role of prior expectations. It is as if they do not matter.
Ibid., 38-40.
- 88 -
noninvolvement of Y. Y’s ongoing noninvolvement represents a more or less net-neutral fear
event for X, because X’s nearly complete uncertainty regarding Y’s security support never
allowed X to enjoy a substantial, preemptive decline in its fear of conflict with Z. Only if X had
already expected Y to be an important source of security in the confrontation with Z would
intense fear be produced by Y’s obvious shift to neutrality. Under relatively normal
circumstances, the most severe abandonment fears consequently emerge from expectational
transition, from certain to uncertain, rather than simply from uncertainty itself.
The above transition argument is not just semantics; it represents a core difference
between my thinking and Cha’s. I therefore elaborate my argument about expectations
considerably as I build my models later in this research. What is important to note for now is
that Cha’s de-emphasis of the transition from expectation to uncertainty allows him to ignore the
emotions Social Psychology predicts will be produced as important expectations between groups
are disappointed. Instead of discussing anger and vengefulness following feelings of betrayal,
for example, it permits him to focus on material fear and insecurity as the emotions leading to
bad Japan-Korea relations, while leaving all other emotions merely tacit and greatly under-
theorized.
As noted above, Cha implicitly utilizes a third, albeit generic and unspecified,
emotionality component to explain how Seoul and Tokyo become entangled in a conflict spiral.
Using this emotion element is without merit in Neorealism. Assuming its presence and role
represents a major shortcoming of Cha’s model.
The connoted set of emotions is best shown deductively. Before carrying out the exercise,
however, it is critical to note that Cha never details the precise mechanism ultimately producing
the predicted friction, beyond stating that a “disparity in strategies leaves [Seoul] unsatisfied”
- 89 -
whenever Korea fears abandonment by Japan while, simultaneously, Japan fears entrapment by
Korea.
209
This emotionally opaque and incomplete process can be outlined as: 1) Korea attempts
to assuage its fear of abandonment by showing a stronger commitment to hedging Japan, so as to
make Tokyo value Seoul more. 2) Continuing to fear entrapment because of their asymmetrical
capabilities, however, the Japanese perceive in Korea’s expressions of loyalty more room to
semi-defect. Tokyo therefore hedges again to the extent it thinks it can. 3a) If Seoul is not yet
unsatisfied, the cycle continues, with Korea moving closer to Japan and Japan correspondingly
pulling back once more. 3b) If the ROK feels unsatisfied, however, friction emerges. Being the
aggrieved party in the alignment, Korea initiates the friction.
While seemingly straightforward, the above argument is actually problematic within
Cha’s Neorealist model. To understand why, consider Seoul’s two possible insecurities within
the quasi-alliance model: abandonment by Japan and entrapment by Japan. As Cha understands,
Seoul’s “only rational play” when fearing abandonment is to enact “a more cooperative strategy”
toward Japan.
210
Importantly, unless some new emotion or idea appears, there is no reason why
the ROK would not just keep moving closer to Japan every time the Japanese pull away,
presuming Korea’s abandonment fears remain primary. In fact, if Cha’s deductive logic is taken
literally—i.e., if fear and insecurity are the only first-tier causal emotions—Korea should
progressively move toward Japan until the Koreans fear becoming entrapped in Japan’s conflicts
more than they worry about being abandoned by their quasi-ally. Intriguingly, once that
threshold is crossed, Korea will pull back from Japan, leading Tokyo to move toward Korea until
Japan shifts past the threshold comprising its balance of interests. This back-and-forth will
continue more or less indefinitely, presuming external factors stay the same, with shifts
209
Ibid., 46.
210
Ibid., 46.
- 90 -
occurring whenever the state fearing abandonment crosses its balance threshold. In short,
conflict should never appear if Japan and Korea experience only the emotions of fear and
insecurity, as the state with incentive to create trouble within the alignment—the state fearing
abandonment—is the very state that will always appease its quasi-ally. To put it another way, in
the quasi-alliance model there is no point at which the Koreans become dissatisfied while
simultaneously having incentive to create conflict with Japan.
Cha might be arguing implicitly that the Koreans become “fed up” and/or they begin to
fear the Japanese will merely pull away again following another cooperative move by Seoul. But
“fed up” is obviously an emotion, or set of emotions. How this ideological phenomenon would
override Korea’s fear of abandonment is unclear, especially since making another concession to
Japan has no cost other than increasing the likelihood of being entrapped in Japan’s conflicts.
Likewise, while the Koreans may fear that Japan will pull back one more time, that insecurity is
remarkable for being just a once-removed fear of abandonment. Why the Koreans would
privilege this costless and less certain pre-cooperative fear of abandonment over the immediate
and more definite fear of abandonment they are experiencing is completely unclear. Moreover,
being but another form of fear of abandonment, pre-concession fear should result in a greater
show of alignment loyalty, up until the point at which entrapment insecurity becomes more
pressing. With emotionality conscribed to fear and insecurity, rational states fearing
abandonment more than entrapment will never enact fractious behavior.
Though it is already obvious that material fear and insecurity are inadequate to the task of
explaining conflict in Cha’s model, it is worth detailing further how abandonment and
entrapment fears, and general insecurity, are supposed to lead to conflict under firm America
commitment to East Asia. Presuming the impossible—i.e., that Japan’s most recent hedge has
- 91 -
led Seoul to express loyalty to Tokyo for the final time, but the Koreans have not yet crossed
their alignment-entrapment threshold: 1) Japan hedges yet again, because it still fears entrapment
by an otherwise confident and reassured South Korea.
211
2) Insecure and fearing abandonment
to the point of feeling unsatisfied by Japan, Korea now acts out against it, thereby sending a price
signal to Tokyo that the current level of abandonment will be costly. The Koreans leverage the
history issue, because they perceive it to be an effective method of imposing cost on Japan.
212
Therefore, 3) because the Japanese realize hedging will be costly, Tokyo suffers renewed
entrapment fears. Consequently Japan, too, lashes out at its quasi-ally, thereby launching a
retaliatory tit-for-tat conflict cycle.
213
As can be seen, ongoing problems can take place in an ideal quasi-alliance model only if
it is reasonable to propose that Korea’s high price signal—its acting out in step 2—represents
such a compelling cost signal to Japan that the Japanese in step 3 actually fear it might force
them to reverse their hedge and support Korea against their will. Of course it is impossible to
believe that the increased price of non-compliance signaled by Seoul’s angry but nonviolent
outburst would raise even a modicum of security concern in Tokyo, presuming one is working
within Neorealism. Yet without Japan’s fears and insecurity in step 3, there would be no
retaliatory cycle and enmity would dead end. Without it, I would have to rewrite it as: 3) Japan
has no need to retaliate against Korea, because it simply does not care about Korea’s impotent
response and distancing itself from Korea was what it always wanted.
214
211
Ibid., 45-46.
212
Korea leverages the history issue, because it tries “to elicit security cooperation from [Japan] by exercising
leverage over other issue areas.” During the era in which Cha is primarily interested, Korea had minimal leverage
with Japan in the material realms of security and economics, and so it had to assail Tokyo ideationally. Ibid., 43.
213
Cha does not make this point. Nevertheless, it is essential if he wishes to produce a conflict spiral, as I explain
immediately below in the text.
214
Cha might be tempted to claim that Japan engages Korea in hostile conversation over the past because it
increases the emotional distance between the two states, therefore helping Tokyo accomplish its goal of hedging its
security position vis-à-vis Seoul. But since Korea’s angry lashing out is always product of its fears of Japanese
- 92 -
Only when I assume that Cha is utilizing a generic emotionality component to explain
Japan’s response to Korea’s coercive use of history does the model continue to make sense. Cha
seems to be leveraging some combination of embarrassment, anger, and vengeance, as well as
emotions related to nationalism. If so, I can change the deduction to be: 3) Embarrassed by
Korea’s nonviolent lashing out at Japan, Japanese nationalists at all levels of society become
angry and vengeful. They make defamatory statements about Korea and/or act provocatively
about the past, thereby triggering a retaliatory conflict cycle. Whatever the case—and there are
quite a few complementary emotions that could be utilized—Cha clearly is reaching far outside
of the Neorealism that underpins his model to explain enmity in Seoul-Tokyo affairs. As it is
written and deduced, however, Cha’s model cannot explain ongoing conflict between Seoul and
Tokyo when the US is highly involved in East Asia.
While Cha obviously encounters difficulty clarifying ideological conflict among the
quasi-allies, at first blush his model’s explanation for Japanese-Korean cooperation in the face of
wavering US security commitment seems more plausible. Indeed, Cha’s primary contention that
American retrenchment leads to fear and insecurity in Seoul and Tokyo, and these emotions
impel the Japanese and Koreans to cooperate deliberately on material and ideological relations,
will make ring true to many IR scholars.
215
After all, common sense informs that when survival
is at stake, all matters with potential to prevent necessary security cooperation must be resolved
or at least significantly diminished. Unfortunately, despite being intuitive, the claim is
troublesome within Neorealist theory. For, while Cha’s use of Neorealism to explain how
insecurity leads to material collaboration is without obvious fault, his assumption that material
abandonment, Seoul has already gotten the message. Joining Korea in a history conversation therefore has a high-
risk low-reward profile for Japan: Dialogue about the past risks poisoning a relationship that is already empirically
distanced enough to satisfy Japan’s security interests. In Cha’s rational choice model, Tokyo’s leaders should
understand this and, therefore, will ignore Korea’s impotent protests rather than engage them.
215
Ibid., 45-57.
- 93 -
security concerns should lead to cooperation on ideological tribulations like the history issue
remains unexplained and puzzling.
216
The greatest challenge is the contradiction therein—the presence of a material-ideational
paradox. Simply put, regarding Neorealism, since non-Neorealist ideas like the history issue are
not powerful enough to impinge upon the calculations of states as they estimate their security
interests, then how can obviously non-Neorealist ideological conflict prevent robust security
collaboration? After all, if disagreements over ideas can prevent adequate security response to
the point that states feel they must actively ameliorate ideological issues, then immaterial
phenomena obviously can impact security interests.
217
And if immaterial phenomena can
influence material interests so powerfully—i.e., if they can dissatisfy states to the point at which
they will not cooperate on material security issues in order to assuage their material fears—then
clearly Neorealism is not adequate to the task of encompassing an explanatory model like the
quasi-alliance model. At minimum, scholars attempting to use Neorealism to explain ideational
outcomes should carefully delineate which ideas matter and when, and discuss as precisely as
possible how material and ideational phenomena relate to each other.
The weakness of causal opacity between the material and ideational realms is not Cha’s
alone and, in fact, he enjoys considerable cover within the field of IR by couching his model to
some extent within Walt’s Neorealist Balance of Threat Theory, a treatise permitting broad
blending of things material and ideational, and which does so without providing any substantial
guidelines for handling or clarifying ideas and their interactions with material phenomena.
218
I
briefly discussed Walt’s work in Chapter 1, and I return to touch on it again below before
216
Cha never explains how material and ideational matters relate to each other, beyond the general assertion that
security imperatives can lead to better ideological relations. Ibid., 43-50.
217
The contention that Japan and Korea cooperate over the history issue so as to influence America’s interpretation
of its material interests is at the core of Cha’s argument. Ibid., 43-57.
218
Ibid., 48.
- 94 -
detailing it in-depth in Chapter 3. For now the point to make is that Cha’s contention about how
ideological relations improve between states in response to their fears and insecurities empowers
ideas to a level on par with material concerns. Yet, despite this implicit acknowledgment that
non-Neorealist ideas can be first-tier causal variables, Cha is seemingly automatic in relegating
them to dependent variable status or, at best, to assumed and unexplained independent variable
standing. Regarding his model’s causal level, he admits specifically only to material fear and
insecurity.
219
YOON’S NET THREAT THEORY MODEL
Like Cha, Tae-Ryong Yoon utilizes fear and insecurity to explain Japan-Korea relations.
But net threat theory also makes explicit the use of the emotion-laden history issue as a causal
variable. Importantly, reliance on the power of the past means Yoon leverages equally the
emotions and expectations related to nationalism.
It is worth illuminating these expectations. Primarily, Yoon implicitly accepts that
Japanese and Korean nationalists have expectations regarding how their states should be
accorded internationally. While the exact parameters of these expectations remain unclear, they
are obviously present. For if the Japanese or Koreans had no expectations whatsoever regarding
representations of their state by others, they would be incapable of reacting emotionally to the
slights of others. Similarly, it is significant that these expectations for respect were completely
irrational during the 1950-1979 period Yoon examines, because no agreement existed claiming
the past had been settled.
220
This implies that Yoon’s rational choice account is reliant on the
219
Ibid., 36-58.
220
One might argue that such an agreement was made in the 1998 Kim-Obuchi joint statement. Prior to 1998
Japanese apologies were arguably too vague to count, and Korean leaders had not yet formally agreed to forego
bringing up the past at the governmental level.
- 95 -
existence of irrational nationalistic expectations.
221
The point is not that these expectations did
not exist, however; surely most observers would agree they have been and continue to be present.
Rather, the point is that they are extant despite being completely irrational at the international
level.
Yoon’s use of fear and insecurity is also remarkable. Recall that under rising net threat
Japan fears getting drawn into conflict on the peninsula and therefore works to ensure adequate
deterrent. Simultaneously, Korea fears abandonment by the Japanese and cooperates on the
history issue to encourage Japanese involvement. The resulting collaboration leads to good
material and ideational relations.
Using these emotions in this manner makes sense within Yoon’s larger theory. Crucially,
Yoon avoids a great bit of Cha’s logical contradiction about emotions within Neorealism,
because Yoon, as I show below, actually grants some ideological phenomena first-tier causal
power.
222
But contending that Korean abandonment and Japanese entrapment fears compel good
material and immaterial relations introduces an interesting but resolvable problem for Yoon.
Namely, since abandonment and entrapment fears lead to improved relations, it becomes difficult
logically to leverage fear and insecurity as emotions that trigger the history issue under falling or
static net threat.
223
This problem might explain why Yoon does not emphasize fear to explain
heightened nationalism, which can technically occur among the public whether net threat is
rising, falling, or staying the same.
224
221
Cha’s implicit reliance on nationalism, as outlined above, means he is equally reliant on these same irrational
expectations.
222
See the below discussion of betrayal in net threat theory.
223
Yoon is clear on the point that “the cause of frequent frictions in Japan-ROK bilateral relations rests in other
factors than the asymmetric bilateral structure of the [abandonment/entrapment] fears.” Yoon, “Fragile
Cooperation,” 94.
224
Of course leaders will have incentive to constrain nationalism when net threat is rising. But that does not mean
public nationalism cannot become aggravated at that time.
- 96 -
To illustrate the difficulty, consider Tokyo’s diplomatic overtures to Pyongyang at a time
of falling or static net threat. Despite relatively benign net threat environments, these advances
have often provoked considerable consternation in Seoul. But Yoon cannot easily argue South
Korea’s consequent abandonment fears explain any historical animosity surfacing concomitantly.
He is constrained because his primary mechanism calls for South Korea to placate Japan at a
time when Seoul fears abandonment, and it is unclear why the same kind of insecurity should
produce different reactions within his model.
Yoon solves his logical conundrum by introducing an expectation-related emotion to
explain why historic animosity tends to bubble up when net threat is falling or staying the same:
Betrayal. That is, for example, when Japan reaches out to North Korea without South Korea’s
blessing, Koreans in the south feel betrayed. And betrayal leads to enmity.
225
With the idea of, and emotions related to, betrayal playing the role a key role, troubles
about the past emerge due to the following causal process: 1) Net threat falls. 2a) Suddenly
unconcerned about deterrence failure on the peninsula, Tokyo’s interests diverge from Seoul’s.
2b) Lack of concern means Japanese leaders do not try to constrain historical animosity if the
history issue reemerges. 3) Divergent interests lead Japan to enact a security policy that
disappoints the ROK’s expectations of security loyalty. 4) South Korean’s feel betrayed and
come to view Japan with enmity. 5) Leaders in Seoul cease tamping down the fallout of
historical friction acts when they occur. 6) The history issue flairs up as Japanese and Korean
nationalists inevitably rekindle the history issue and engage in an ongoing tit-for-tat conflict
spiral.
Conversely, problems regarding history ameliorate because: 1) Net threat rises. 2aa)
Tokyo becomes worried about deterrence failure on the peninsula and its interests therefore align
225
Ibid., 98-99.
- 97 -
with Seoul’s. 2ab) Koreans worry about Japanese abandonment. 2ba) Concern about deterrence
means Japanese leaders actively try to minimize the history issue. 2bb) Fear of abandonment
compels South Korean leaders to attempt to improve the history issue. 3) Good relations emerge
because governments in Seoul and Tokyo collaborate successfully on the history issue.
Yoon’s mechanism leading to heightened historical friction is remarkably more passive
than Cha’s. Whereas, Cha argues that fearful Koreans leverage the history issue to impose costs
on a hedging Japan, Yoon’s Koreans simply step back and allow the past to become problematic.
Conversely, when interests are aligned, both Cha and Yoon call for Japanese and Korean leaders
to actively ameliorate historic animosity. But only Cha asserts that the Koreans cease utilizing
history to impose costs on Japan.
Cha’s active mechanism is not without major fault, as I detailed above. Nevertheless, his
active mechanism feels necessary, given the many, often rapid, shifts in the tenor of Seoul-
Tokyo affairs. Without an active mechanism, Yoon must rely on the ongoing constant of historic
animosity combined with random
226
domestic and international events to elucidate sharp
politicizations of the history issue. But if random events and ongoing nationalism are so volatile
a combination, there is no reason to think the leadership in Seoul and Tokyo could contain as
well as they do the consequences of random historical friction acts in a rising net threat
environment.
Furthermore, since net threat occasionally begins rising at a time when historical
animosity is already extremely agitated, then it is not just a matter of leaders in Seoul and Tokyo
intervening earlier than otherwise to stop problems from getting out of control, as Yoon
226
Phenomena are random to the extent they are unrelated to rising or falling net threat. Note that Yoon’s
theoretical discussion on sources of heightened nationalism is exceptionally limited, and usually does not extend
beyond arguing that ongoing nationalism combined with historical friction acts leads to problems. In addition to his
use of betrayal, he acknowledges specifically that domestic politics can intensify contestations over the past. Ibid.,
94, 98.
- 98 -
claims.
227
Temporal randomness of historical friction acts vis-à-vis rising net threat means the
leadership will instead sometimes have to intercede at moments of thoroughly unruly nationalism.
It is hard to imagine how these elite interventions counter domestic nationalistic emotions so
effectively at these difficult times.
All in all, the elites’ remarkable ability to contain historic animosity at the popular level
suggests 1) problems about the past must, in actuality, not be very severe; or 2) the history issue
itself must often be agitated and assuaged directly by rising and falling net threat. Yoon
certainly does not agree with the former. Concerning the latter, however, he lacks the requisite
active mechanism. Indeed, his main argument seems to be that “historical animosity plays an
important role of amplifying conflicting interests,” rather than conflicting interests intensifying
the history issue via an active, causal mechanism.
228
Despite this shortcoming, Yoon’s injection into his model of the emotion of betrayal is
fascinating. For “betrayal”—more so than any other term—implies very strong expectations of
security loyalty. I agree that feelings of betrayal are present. Unlike Yoon, however, I argue that
betrayal produces more than just comparatively passive emotions like enmity among the
leadership. As I detail later, it evokes feelings among all those sensitized to it and demands
action to assuage these feelings. In my model, betrayal therefore aggravates historical animosity
directly, often at many levels of society.
Yoon’s discussion of betrayal is also interesting in that he does not mention its emotional
corollary: satisfied loyalty. If betrayal produces deleterious emotions, loyalty surely evokes
positive ones. As I contend in my theoretical chapters, satisfied loyalty expectations induce
emotions, among adequately aware members of society, that directly help pacify the history issue.
227
Ibid., 14.
228
Yoon does argue, however, that the history issue would eventually disappear without conflicts of interest.
Compare pp. 94 and 96 in Ibid.
- 99 -
Another area where Yoon is weak—and Cha shares this flaw—concerns expectations.
Simply stated, it is almost completely unclear why Cha and Yoon assert that security
expectations exist in Seoul-Tokyo relations. After all, given the enmity of the general bilateral
relationship, the on-and-off-again nature of security ties, and the lack of a formal bilateral
security commitment, why would the ROK—an apparently highly rational state—ever develop
expectations of Japanese loyalty to Korea’s security interests? South Korean expectations were
obviously irrational to begin with and became more so with every consecutive jilting by Japan. I
explore this issue in Chapter 5. For now I note both Cha and Yoon should have spent more than
two or three paragraphs explaining the presence of these highly irrational expectations in their
rational choice models.
229
LIND’S APOLOGY POLITICS MODEL
Jennifer Lind’s apology politics model makes explicit use of the Neorealist emotions of
fear and insecurity, and couples them with explicit emotions and implicit expectations related to
the history issue. Lind model comports strongly with Stephen Walt’s balance of threat theory.
230
Though ostensibly a Defensive Neorealist, Walt’s theory departs considerably from Neorealism.
Instead of spotlighting how states balance against power, he contends they balance against threat,
as defined by aggregate power, geographic proximity, offensive power, and aggressive
intentions.
231
Walt’s last point about intentions particularly differentiates his work from the
structural basis of Neorealism, as intentions and structure are unrelated phenomenon.
232
As I detailed in Chapter 1, Lind examines the impact of apology politics on both France-
German and Japan-Korea relations. In postwar Europe, Germany’s “unapologetic contrition”
229
Ibid., 98-99; Cha, Alignment Despite Antagonism, 39-40.
230
Walt, The Origins of Alliances.
231
Ibid., 21-26.
232
Legro and Moravcsik entertainingly examine Walt’s and others’ Realisms. For their take on Walt, see Legro and
Moravcsik, “Is Anybody Still a Realist,” 36-38.
- 100 -
tempered French perceptions of a German threat, and paved the way for a gradual and deep
reconciliation. In contrast, Japan effected “unapologetic remembrances” after its defeat. These
whitewashings and denials were followed by apologies. Though these acts of contrition were
often vague and seemingly halfhearted, they nevertheless produced backlash among Japanese
nationalists. Therefore, in contrast with France’s perception of a benign—indeed, very
friendly—Germany, Korea continued to fear Japan’s intentions and potential military power.
233
As a result, Seoul-Tokyo reconciliation remains unfulfilled. Lind’s focus is therefore set
squarely on Walt’s argument about perceptions of threat.
Lind’s explicit and implicit use of expectations and emotions in the Japan-Korea case is
generally well thought out, particularly concerning the Japanese. Though she does not draw
attention to it, her model is reliant on the expectations of Japanese nationalists. Like all
nationalists, they have expectations about the way their state should be accorded internationally.
What makes Lind’s approach unique is her reasonable assertion that Japanese nationalists
anticipate that their own leaders will uphold a positive and unflinching state image vis-à-vis
other states. Formal apologies and admissions of wrongdoing or, indeed, just apologies that
imply immoral behavior in the past, disappoint the expectations of these men and women within
and without government.
Transgression of nationalist expectations induces many emotions, including surprise,
puzzlement, or embarrassment, as well as anger, enmity, and vengeful feelings. Lind’s
assumption that angry Japanese nationalists desire to fight back combatively, and therefore lash
out at Korea directly or indirectly, seems plausible, since they are likely to understand and regret
233
Lind clarifies that Korea does not fear Japan’s contemporary power at any given time. Institutional and structural
factors like Japan’s constitution, the US alliance, and the rise of China impose adequate constraints on Japan’s
military capability. What the ROK worries about is Japan’s future power. See, for example, Lind, Sorry States, 81-
84.
- 101 -
that Korea was not just a bystander in the instance of any particular Korea-directed apology or
concession. Moreover, the fact that an historical friction act
234
directly infuriated the nationalists
means Lind can argue logically that nationalists will be inclined to focus on the history issue.
Unlike Yoon, she is not missing an active causal mechanism leading to historical animosity. In
contrast with Cha, Lind does not require a missing assumption of a rational state utilizing the
history issue because it is valuable for sending a signal that defection will be costly.
Where Lind’s model weakens is in her underutilization of Korean expectations. Since
Japanese nationalists have expectations about the representation of Japan in Lind’s model, their
Korean peers certainly have similar expectations concerning Korea’s depiction. Moreover, like
the Japanese, Korean nationalists will often anger and fight back provocatively against perceived
diminishments of their state. This means Japanese and Korean nationalists will be involved in
the kind of ongoing bilateral conversation I described while explaining the nature of their
nationalisms under complex interdependence. Nevertheless, a pure version of the apology
politics model is comparatively unidirectional: 1) Japan apologizes; 2) there is a nationalist
backlash in Japan; 3) Koreans perceive that Japan has bad intentions; and 4) reconciliation
remains unfulfilled as Seoul worries about Japanese power.
235
Lind thereby tends to portray
Korea very much like a participator in a video chat who can see, hear, and internalize the
conversation, but can only minimally respond. This position is untenable in Lind’s expectations
and emotions reliant model, unless the Japanese are somehow unable to receive adequate
information from Korea.
If Korean nationalists inside and outside of government are given the same expectations,
emotions, and agency as nationalists in Japan, Lind’s model must change considerably. Her
234
In Lind’s model, Japanese apologies clearly represent historical friction acts for Japanese nationalists.
235
For Lind’s three mechanism’s linking the history issue to security see Ibid., 11-13.
- 102 -
unidirectional model becomes: 1) Japan apologizes; 2) there is a nationalist backlash in Japan; 3a)
Koreans perceive that Japan has bad intentions while, simultaneously, 3b) Korean nationalists
lash out at Japan directly and indirectly, because they resent the slights made during the backlash;
4a) reconciliation remains unfulfilled because Koreans fear Japanese power,
236
while 4b)
ideological relations remain very tense as nationalists at all levels of society in both states remain
engaged in an ongoing, zero-sum conversation about the status of their states.
The real challenge for Lind’s model appears, however, when one asks why Japanese
leaders apologize. Curiously, Lind never properly addresses this obviously important question.
She does note that “scholars increasingly cite [apologies’] importance in international
reconciliation,” and that “the world criticizes Japan for failing to come to terms with its past.”
237
Moreover, she occasionally contextualizes Tokyo’s apologies to the ROK. On the instance of
the first occasion, for example, she remarks it was “part of diplomatic normalization” in 1965.
238
Similarly, Emperor Akihito’s 1990 apology was designed “to compensate for the vagueness of
[Emperor] Hirohito’s 1984 statement.”
239
Given Lind’s concordance with Walt’s balance of threat theory, it seems reasonable to
presume that Japanese leaders apologize because they wish to promote Korea’s security
cooperation by ameliorating its perception of a Japan threat. Intriguingly, this line of thinking
does not make much sense in the Franco-German case, however. Indeed, France perceived “no
friend but Germany” by 1960, some five years before the German “awakening.”
240
Germany
was apologizing, in other words, to a state that viewed it as very friendly. Insights provided by
236
Whether Tokyo perceives Seoul as having bad intentions is moot; Korea is not powerful enough to represent a
serious threat to Japan in the foreseeable future.
237
Ibid., 2.
238
Ibid., 47.
239
Ibid., 61.
240
Ibid., 124, 126-131.
- 103 -
Ruth Benedict’s 1946 landmark book, The Chrysanthemum and the Sword, are interesting in this
context. Her argument that there are two kinds of societies—guilt and shame based—suggests
Germany apologizes to assuage its guilt, while shame makes Japan unwilling to face the past.
241
Especially when combined, Walt’s theory and Benedict’s concept of shame societies implies
Japan will normally apologize only when it perceives that the history issue is preventing essential
cooperation in crucial material matters. In most cases, in others words, the history issue will
have to be very bad before Japan will apologize.
But if Japanese leaders tend to apologize on occasions when historical animosity is
already seriously straining Seoul-Tokyo relations, then the unidirectional nature of Lind’s model
faces a considerable challenge. It suggests Japanese and Koreans are usually engaged in
nationalistic conversation even before nationalists in Japan have an apology against which to
push back. Were this the case—were nationalists battling before apologies—then only with the
most careful analysis can observers begin to discern whether nationalistic anger following an
apology represents a backlash against the apology, an ongoing nationalistic conversation, or a
combination of both. Similar inquiry is merited to address whether apologies impact the actual
level of mobilized hostile nationalism, or mostly just the direction of the debate.
242
Lind is regrettably not careful enough on these points. Indeed, while discussing
competing explanations in the Japan-Korea case, she primarily asks if Korea’s distrust and
relatively high threat perception regarding Japan are real, and whether their reality matter. She
does not critically explore alternatives for the apology-backlash portion of her model.
243
This
241
Ruth Benedict, The Chrysanthemum and the Sword: Patterns of Japanese Culture (1946; repr., New York:
Houghton Mifflin, 2005).
242
It might be the case, for example, that Japanese refusal to apologize during or following a period of exceptionally
tense ideational relations would provoke a powerful Korean material or ideological backlash against Japan. If so,
missing Japanese apologies would still lead Koreans to view Japan’s long-term material intentions with concern, just
as in Lind’s model.
243
Lind, Sorry States, 39-40, 54-55, 78-80, 92-100, 186-189.
- 104 -
missing inquiry is awkward, because the key contribution of her scholarship rests in its claim that
apologies lead to relationship-damaging backlash. The idea that Seoul’s threat perceptions
contribute to reconciliation problems is comparatively uninteresting.
KOO’S COMPLEX INTERDEPENDENCE MODEL
Of the four models of Japan-Korea relations detailed here, Min-Gyo Koo’s makes the
least use of emotions and expectations. Their relative absence is directly related to the fact that
he exerts practically no effort to understand the causal mechanism through which increased trade
dependence ameliorates conflict. Instead, Koo adopts wholeheartedly the general argument of
Neoliberalism’s complex interdependence theory that trade dependence reduces conflict.
244
In his qualitative case study of the Dokdo/Takeshima dispute, Koo disappointingly does
not systematically evaluate the sub-processes of complex interdependence that scholars predict
will decrease conflict. This feels odd, because he does show how problems regarding the islands
emerge via the pathways predicted by complex interdependence. Koo argues convincingly, for
example, that the islands became politicized or were aggravated by policies implemented by
prefectural authorities and federal bureaucrats, the imperatives of national-level domestic politics,
the worldwide trend of establishing 200-nautical-miles maritime boundaries, the United Nations
Convention on the Law of the Sea (UNCLOS), and angry mass nationalism within both Japan
and Korea.
245
In contrast to these comparatively bottom-up processes, Koo’s accounting of conflict de-
escalation is top-down. In terms of trade, for instance, he offers scant evidence of suffering,
cross-border businesses lobbying governments to make concessions on Dokdo/Takeshima or,
likewise, of friendship societies working together to promote understanding during tense
244
Koo, “Scramble for the Rocks,” 18-19, 33-40, 251.
245
Ibid., 93-95, 108-110, 120, 130-133.
- 105 -
moments. What emerges instead is a sense that conflict de-escalation is born of an almost
Neorealist concern about the health of the economy by national leaders. Thus, for instance, Koo
presents the history-related concessions made by Korea to secure Japanese loans during the
desperate 1997 East Asian financial crisis as evidence of dependence.
246
While most scholars
surely would agree that financial interdependence encouraged cooperation during the emergency,
Koo’s overview of high politics gives one the sense that he is supporting his qualitative cases
with his quantitative work, rather than the other way around.
Because Koo chooses not to illuminate the processes through which trade dependence
improves relations, discussing his implied use of emotions and expectations is not as rewarding
as it was regarding Cha’s, Yoon’s, and Lind’s models. Nevertheless, it should still be stated that
trade dependence models are reliant on the emotions of desire to obtain or maintain material
wealth. They are also dependent on expectations that gains are possible, and that mutual trade
cooperation is the best way to fulfill material desires over the long-term. Importantly, these
expectations and emotions must be shared by a considerable portion of society in both Japan and
Korea, given the efficacy with which leaders in Seoul and Tokyo are able to temper widespread
nationalistic sentiment whenever conflict over Dokdo/Takeshima seriously threatens the
economy. As relations near the breaking point, Japanese and Koreans who prioritize economic
well-being over collective self-esteem must somehow overpower those preferring the opposite.
Likewise, within individuals emotionally conflicted about the two interests, decisions to uphold
economics are obviously being made with great consistency.
A deeper investigation of expectations and emotions does reveal a potential problem
within the overall idea that trade dependence necessarily leads to a tendency for improved
relations between Japan and Korea. The trouble emerges from Koo’s frequent use of the one-
246
Ibid., 127-129.
- 106 -
sided nationalist norm to explain conflict over Dokdo/Takeshima. Implementing the
expectations and emotions associated with this norm to explain conflict means Koo ideally
should have explored how nationalists in both Japan and Korea perceive their respective state’s
trade dependence with the other, and whether these in turn ever exacerbate conflict.
Nationalists in Korea, for example, feared being economically dependent on Japan,
especially until the 1997 financial crisis.
247
Their concerns suggest dependence could have
intensified or even kick-started certain episodes of the island dispute, or at least that the
pacifying effects of economic interaction were an outcome of debate and deliberation among
competing groups and were not more-or-less automatic. Similarly, Japanese nationalists with
superiority complexes vis-à-vis Korea very well might be displeased with Japan’s growing
economic dependence on South Korea, particularly given Japan’s comparative economic
stagnation since the early 1990s. If so, sentiments about Korea being an unwelcome economic
upstart and challenger could pose problems for those within Japan whose material interests lead
them to view these ties more positively. While Koo’s quantitative data inform that economic
interests trump nationalism and therefore promote conflict de-escalation over the entire 1953-
2005 period, case studies that focused more closely on process might show circumstances
differed in certain episodes. Detailed research might even suggest that the past could become
more problematic over the short and medium terms due to Japan’s increasing, yet still modest,
trade dependence on Korea. This would happen in the event Japanese anger over dependence on
Korea temporarily overwhelms the rationality compelling economic cooperation.
247
For example see Judith Cherry, “Killing Five Birds with One Stone: Inward Foreign Direct Investment in Post-
Crisis Korea,” Pacific Affairs 79, no.1 (Spring, 2006): 9-27. Cherry’s argument about the change in Korean
perceptions regarding inward FDI following the 1997 financial crisis does not discuss Japan or any other states
specifically. I nevertheless apply it to Japan, because of its historic outsized role in Korea’s inward FDI.
- 107 -
SUMMARY OF MODELS ON JAPAN-KOREA RELATIONS
All in all, among the four models detailed in this section Koo’s is the least weakened
from critical examination informed by an understanding of expectations and emotions.
Importantly, his comparative success is due to insulation provided by exceptionally minimal
specificity of causal process, rather than because he has generated a well-designed, precise
predictive model. His strength emerges from weakness, in other words.
Whereas Koo does not seriously attempt to uncover how conflict increases or decreases,
the other three scholars do. As shown above, examining their model’s processes with an
expectations/emotions deductive lens reveals serious flaws. It uncovers that Cha’s rational
choice model is reliant on South Korea’s surprising tendency to develop irrational expectations
about Japan’s commitment to Korea’s security. The lens also shows that Cha’s
abandonment/entrapment mechanism is missing a link crucial for explaining why Japan gets
drawn into a tit-for-tat conflict spiral whenever Korea attacks it via the history issue. More
precisely, Cha implicitly requires Japanese within and without government to be motivated by
nationalistic expectations and emotions—to fight back via the history issue whenever Korea lobs
history bombs.
Equally problematic, by arguing that material fears prod leaders in Seoul and Tokyo to
deliberately improve the history issue, Cha elevates ideas to the highest tier of causality. These
assumptions about ideas and emotions are awkward because they suggests these very same
expectational and emotional phenomena should be at work elsewhere in the model, and very well
might be independent variables working in tandem with, or even supplanting, fear and insecurity.
This logical incongruity consequently challenges the very idea that a model based on Neorealism
can explain satisfactorily the cycles of historical animosity in Seoul-Tokyo affairs. Similarly but
- 108 -
less crucially, the examination reveals that Cha requires Koreans to lash out via the history issue
because they believe it will effectively impose costs on Japan. How they would rationalize this
within the dynamics of a world explained by Neorealism remains unclear.
The expectations/emotions lens also reveals much about Yoon’s net threat theory. In line
with Cha, Yoon’s explicit use of the betrayal emotion requires Koreans to have unrealistic
expectations that Japan will be loyal to Seoul’s security interests. Korean leaders experience
enmity toward Japan whenever Tokyo betrays its implicit commitment. Nevertheless, net threat
theory’s exploration of betrayal remains greatly undertheorized. Most problematically, Yoon
never explains why expectations of security loyalty exist in South Korea. Yoon moreover does
not consider betrayal’s corollary: Satisfied loyalty. Net threat theory therefore misses
completely one process through which Korea’s satisfied expectations regarding Tokyo’s security
loyalty encourage leaders to tamp down society-wide historical animosity and bring about better
bilateral relations. Finally, the lens also reveals that Yoon relies primarily on a passive
mechanism to explain shifts in the tenor of historical animosity. While his model does permit
leaders in Japan and Korea to depoliticize the history issue when they experience fear, they are
given no active role in aggravating historical animosity. Since they cannot actively make it
worse, they cannot reverse the actions with which they might have otherwise accelerated the
historical issue. With such a passive mechanism, I lack confidence that Yoon can explain the
relationship between security interests and the history issue.
Finally, exploring expectations/emotions within Lind’s work shows her apology politics
model is incorrectly unidirectional. By highlighting the expectations Lind implicitly gives to
Japanese nationalists, it is easy to argue that Koreans should be given the same expectations and,
consequently, the same emotions and agency as the Japanese. This acknowledgement greatly
- 109 -
weakens Lind’s model by suggesting that apologies are usually given within the context of an
intense bilateral nationalistic conversation, and that a large percentage—indeed, potentially all—
of the Japanese nationalistic backlash following Tokyo’s apologies is just part of the ongoing
international dialogue.
III. Security Expectations in Japan-Korea Relations
As I detail in Chapter 3 while developing my first alignment model, Japan and South
Korea are members of an informal bilateral security “alliance.” More specifically, they are
aligned states, rather than allies. Despite lacking a formal security agreement, each holds
security expectations of the other, just as do all states in alignments and alliances. North Korea
has been the primary object of Seoul’s and Tokyo’s strongest security expectations. Simply
stated, each expects the other to promulgate North Korea policies conforming to the security
interests of the self.
In the following paragraphs, I briefly overview the post-1960 Seoul-Tokyo security
relationship, with the specific objective of explaining which alignment member was the primary
transgressor of security expectations during a given period of time. I argue there have been two
distinct periods within the decades I examine throughout this dissertation. During the first,
which runs from at least 1960 through to the early or mid 1990s, Japan was the primary
expectations transgressor. Its periodic endeavors to reach out to North Korea departed
considerably from Seoul’s much more conservative Pyongyang policy, thereby disappointing the
ROK’s security expectations and inducing severe emotionality in it. By the late 1990s, however,
South Korea took the lead in crafting North Korea policy and consequently ran afoul of Japan’s
security expectations. With its security expectations disappointed, Tokyo became the
emotionally unsatisfied alignment member.
- 110 -
To argue the above, I join Lind and Yoon, and to some extent, Cha, in utilizing Walt’s
balance of threat theory. As noted earlier, Walt is widely considered to be a Defensive
Neorealist. He argues that aggregate power, geographic proximity, offensive power, and
aggressive intentions define state interests.
248
Applying balance of threat theory’s four interests
to Japan-Korea relations reveals a number of findings regarding their respective security interests,
which in turn help provide a basic understanding of their security expectations.
249
First, Walt’s theory suggests Japan and South Korea are not a threat to each other,
especially over the short term but also over the intermediate term. The ROK is too weak to
threaten Japan, and has most of its military power dedicated to preventing a North Korean
invasion. South Korea moreover does not have a nuclear weapons program with which to awe
its eastern neighbor. On the other side, Tokyo controls a much larger aggregate power base than
does Seoul and does not have a comparatively preoccupied military. But Japan’s power is
nevertheless mitigated considerably by factors such as the East Sea/Sea of Japan, security
dependence on the US, the generally defensive nature of Japan’s military, and constitutional and
social constraints on its military forces.
Whereas Lind’s hyper-focused account of the ROK’s threat perception of Japan led her to
conclude that South Koreans fear Japanese power, casting balance of threat theory more widely
suggests Seoul’s insecurities regarding a revitalized Japanese empire are, at most, of secondary
importance. A broader consideration of Walt’s theory informs, instead, that the focus of any
examination should be North Korea, and how changes in Seoul’s and Tokyo’s respective
perceptions of the security threat it represents impact the Japan-Korea relationship. After all, in
248
Walt, The Origins of Alliances, 21-26.
249
Snyder explains that security interests are a means through which one can estimate security expectations within
alignments. I discuss Sndyer’s research in-depth in Chapter 3. For quick overview, see Snyder, Alliance Politics,
30-31.
- 111 -
comparison with Japan’s greatly constrained power, North Korea is contiguous to South Korea,
has maintained a very large military along their lengthy shared border, possesses chemical
weapons of mass destruction, has made a concerted effort to develop and test nuclear weapons,
has often been militarily provocative,
250
has clearly established that the mission of its military is
to unite Korea by force, and has never shared an ally common to Japan or Korea to keep its worst
tendencies in check.
251
Japan, too, has reason to be wary of North Korea. Nevertheless applying Walt’s four
lenses suggests Tokyo’s fears of North Korean military power should generally have been much
lower than Seoul’s during much of the era under consideration. For, unlike the ROK, the East
Sea/Sea of Japan separates North Korea and the island nation. Moreover, the vast majority of
Pyongyang’s military capability is confined to land. Those North Korean military vessels or
airplanes that might otherwise threaten Japan are hopelessly outclassed by Japan’s modern naval
and air forces, to say nothing of American capabilities within and around Japan. In fact, only
with the advent of Pyongyang’s advanced missile capability and nuclear weapons program in the
1990s did North Korean power become a direct threat to Japan’s security. Before then its threat
was merely indirect. For instance, while Tokyo would have realized great insecurity following a
collapse or military defeat of South Korea, the Japanese had no reason to worry about a major
attack on Japan by North Korean forces.
252
Renewed war on the peninsula during the Cold War
likewise risked the possibility of drawing in the superpowers, greatly increasing the risks to
250
Dick Nanto, North Korea: Chronology of Provocations, 1950-2003, (Washington D.C.: Congressional Research
Service, March 2000), 3-11, from http://www.fas.org/man/crs/RL30004.pdf, accessed May 2, 2005.
251
US Department of Defense, Annual Report on the Military Situation on the Korean Penninsula (Washington
D.C.: Dept of Defense, Sept 2000), from http://www.csis.org/burke/mb/asia_neac_dod_korea.pdf, accessed May 1,
2005.
252
Yoon is very good on this point. Recall how net threat theory emphasizes Japan’s fears of being drawn into
conflict on the peninsula. That is why Tokyo does its part to bolster deterrent there under rising net threat. See
Yoon, “Fragile Cooperation,” esp. 85-90.
- 112 -
America-allied Japan. But even this terrible possibility was fully one step removed compared to
the direct threat North Korea posed to South Korea’s population, economy, and sovereignty.
In all, the last three paragraphs inform that Seoul-Tokyo relations suffered a natural
tendency to have divergent security policies regarding North Korea, especially until Pyongyang’s
weapons programs began directly threatening Japan in the 1990s. Through to that point,
comparatively low levels of Japanese fear and insecurity afforded Tokyo considerable leeway in
its dealings with the DPRK. Its northern-friendly strategy emerged from within that space.
Whenever Tokyo’s security strategy contrasted blatantly with the ROK’s, it risked disappointing
Seoul’s security expectations. If realized, these dashed expectations produced an emotional
reaction in Korea causing or permitting historical animosity to blossom between Japan and South
Korea.
Of course Pyongyang was not the only contributor to security concerns in Seoul and
Tokyo. Balance of threat theory suggests Cha is correct to consider variations in US
involvement and Yoon is right to look at the overall net threat faced by Japan and Korea.
Fluctuations in these phenomena sometimes produced conflicting or coincident perceptions of
threat in Seoul and Tokyo. Milieu threat perception divergence was arguably clearest during the
US-USSR détente of the early 1970s, for example. At that time, Seoul-Tokyo interests
contrasted considerably as South Korea achieved practically none of the success Japan enjoyed
in improving its relations with China, North Korea, and the Soviet Union. Tokyo’s realization
that large scale conflict was unlikely therefore contrasted sharply with Seoul’s view. As Japan
pulled away from South Korea and began to favor the communist states more than before, the
ROK’s security expectations regarding the Seoul-Tokyo alignment were dashed. The emotional
- 113 -
consequence generated incredible bilateral animosity, including over historical matters.
253
Resolution materialized only in the aftermath of collapsing détente and US President Jimmy
Carter’s plan to withdraw all US troops from South Korea. These challenges quickly aligned
Seoul’s and Tokyo’s respective interests and policies, thereby satisfying both the ROK’s and
Japan’s security expectations.
254
The emerging picture is thus of a very complex bilateral security relationship.
Contrasting with Seoul’s often intense fears, Tokyo’s more benign perception of the North
Korean threat produces less insecurity, and therefore more scope for Japanese rapprochement
with Pyongyang. This overall tendency should have held, at least until the DPRK’s weapons
programs gained momentum in the 1990s. Nevertheless, the potent influence of the US as well
as the overall regional level of threat on the bilateral Japan-ROK relationship means Seoul and
Tokyo affairs are very unstable. More than just occasionally, these phenomena compellingly
pushed and pulled Japanese and Korean interests together and apart, leading to repeated periods
satisfied and disappointed bilateral security expectations.
Even this complicated description is incomplete, however. For adopting Walt’s theory
makes it imperative to consider how threat perception changes over time. I have already noted
Japan’s increased threat perception in response to the DPRK’s weapons programs.
255
But South
Korea’s insecurity regarding North Korea also shifted, albeit toward a much lower overall
perception of threat. It is unclear exactly when the ROK’s view began transforming, or by
precisely how much. Though aided greatly by the comparatively rapid ending of the Cold War,
the process was no doubt gradual and a product of mutually reinforcing material and ideological
253
Cha is best on this episode. Cha, Alignment Despite Antagonism, 100-140.
254
Ibid., 141-168.
255
Green argues North Korea played an important role in transforming Japan into a “reluctant realist.” Michael
Green, Japan ’s Reluctant Realism: Foreign Policy Challenges in an Era of Uncertain Power (2001; repr., New
York: Palgrave, 2003).
- 114 -
factors. On the material side, as shown in Figure 2.1, South Korea’s amazing economic growth,
combined with North Korea’s comparative and absolute economic stasis, resulted in the ROK’s
unequivocal economic dominance of the peninsula.
256
With superior financial resources, Seoul
transformed its military into a modern fighting force while Pyongyang’s remained relatively
antiquated.
257
By the 2000s, if not before, there seemed little doubt that South Korea could
defeat a North Korean blitzkrieg attack before DPRK troops captured Seoul, even without help
from the US.
258
A RAND report clarifies the ideological change in the ROK. It argues there was a
“gradual creation of a link between the struggle for democracy and the desire for unification.”
256
UN, “National Accounts Main Aggregate Database,” under “GDP and its Breakdown at Constant 2005 prices in
US Dollars,” http://unstats.un.org/unsd/snaama/dnlList.asp (accessed January 28, 2012). I use this data for
illustrative purposes only; data on North Korea is notoriously questionable. For an excellent overview of the
challenges of using data about the DPRK, see Mika Marumoto, “Project Report: Democratic People’s Republic of
Korea Economic Statistics Project (April-December 2008),” Presented to Korea Development Institute School of
Public Policy and Management and the DPRK Economic Forum, U.S.-Korea Institute at Johns Hopkins University-
School of Advanced International Studies (March 2009), http://uskoreainstitute.org/2009/03/02/project-report-dprk-
economic-statistics-project-march-2009/ (accessed January 28, 2012).
257
Anthony H Cordesman and Robert Hammond, “The Military Balance in Asia: 1990-2011: A Quantitative
Analysis” (Center for Strategic and International Studies, 2011), http://csis.org/publication/military-balance-asia-
1990-2011 (accessed January 28, 2012).
258
See, for example, Jae-Jung Suh, “Assessing the Military Balance in Korea,” Asian Perspective 29, no. 4 (2004):
63-88.
- 115 -
“Many came to see achieving democracy in South Korea not only as a critical need in its own
right but as the essential first step toward promoting peace and unification on the peninsula more
broadly. For ‘progressives’ in particular, the struggle to achieve democracy in South Korean
society became synonymous with the struggle to promote reconciliation between the two
Koreas.”
259
There was also a strong connection between the political power of the military,
which tended to be very conservative, and the prioritization of security. In a significant way, as
the military lost control over the levers of power, the more possible it was for the leadership in
Seoul to view the North as something other than just a direct threat to be feared. These
complementary processes surely explain an important part of the reason why Kim Young-Sam,
the first South Korean President without a major military career behind him, agreed in 1994 to
have a summit meeting with Kim Il-Sung in Pyongyang. The northern Kim’s death just days
after the summit agreement had been reached marked the end of this progress until Kim Dae-
Jung, another civilian President, ascended the presidency in Seoul.
Though Japan did not experience a concomitant ideological shift and reduced threat
perception from it, it did consolidate its economic advantage over North Korea. But this
improvement did not permit the Japanese to bolster their superior military position over the
North Koreans, because Pyongyang’s missile and nuclear programs sidestepped Japan’s
geographical and material defenses. The No-dong missile exemplifies this trend. First detected
by intelligence in 1990, it quickly transformed Tokyo and other major Japanese cities into
259
Norman Levin and Yong-Sup Han, Sunshine in Korea: The South Korean Debate Over Policies Toward North
Korea (Santa Monica, CA: Rand, 2002), http://www.rand.org/pubs/monograph_reports/MR1555.html (accessed
January 30, 2012), 18.
- 116 -
targets.
260
By 2008, North Korea reportedly deployed about 200 of these chemical weapons-
capable missiles.
261
Whereas the No-dong and the DPRK’s even more advanced missile technology severely
eroded Japan’s impregnability to North Korean attack in the 1990s and 2000s, they merely added
a little extra danger to the incredibly high level of threat already faced by South Korea. For
North Korea had been successfully producing Hwasong-5 SCUD-type missiles since
approximately 1987.
262
With a range of 300km, these CWMD-capable missiles comfortably
target all of the greater Seoul metropolitan area. By 2000, Pyongyang deployed around 500
SCUD-type missiles, including the Hwasong-6, a missile with a range that threatens all of South
Korea.
263
But the primary threat to South Korean civilians almost certainly comes from the
North’s long-range artillery. In contrast to single-use missiles, around 500 artillery systems can
strike all of Seoul repeatedly, until destroyed with difficulty by US and South Korean air and
artillery counterstrikes. Additional shorter range artillery can target fairly large population
centers north of Seoul, like Paju and Ilsan. South Korea also risks exceptionally high military
casualties in a war with the DPRK. North Korea garrisons about 700,000 troops, 8,000 artillery
systems, and 2,000 tanks within 100 miles of the Demilitarized Zone.
264
Though inferior in
quality to ROK and US military forces, the DPRK’s exceptionally large deployment means war
will be very costly for South Korea’s military.
260
Joseph S Bermudez, Jr., “Occasional Paper No. Two: A History of Ballistic Missile Development in the DPRK,”
Center for Nonproliferation Studies, Monterey Institute of International Studies (1999),
http://www.fas.org/nuke/guide/dprk/missile/nd-1.htm, (accessed January 31, 2012), 21.
261
Daniel A Pinkston, “The North Korean Ballistic Missile Program,” Strategic Studies Institute, US Army War
College (February 2008), http://www.strategicstudiesinstitute.army.mil/pdffiles/pub842.pdf (accessed January 31,
2012), 16.
262
Ibid.,”16.
263
Ibid., 47.
264
US Department of Defense, Annual Report on the Military Situation on the Korean Penninsula, 5-7.
- 117 -
The point is that North Korea’s 200 or so intermediate range missiles and potential future
nuclear weapons capability are relatively unimportant when contrasted with the comparatively
immediate and very substantial North Korean conventional threat faced by South Korea’s
military and civilian population. Importantly, they do not cancel out much of the material
dominance Seoul increasingly brought to bear against Pyongyang. What alarmed Japan
therefore produced little more than a shrug in South Korea, especially after the very “progressive”
and famed democracy proponent Kim Dae-Jung assumed the ROK presidency in 1998. For Kim
was the embodiment of the ideological shift in South Korean society regarding both democracy
and the perception of North Korea’s threat. The incredible political power of the Presidency in
South Korea all but ensured Kim would succeed in transforming his progressive beliefs into
national policy.
Though Japan’s and the ROK’s respective security degradation and consolidation were
gradual and incomplete by 1998, the year nevertheless marked an important turning point in how
both states perceived the North Korean threat.
265
Not only did Kim Dae-Jung ascend the Korean
presidency in February and quickly begin clarifying how South Korea would shortly embark on a
northern-friendly “sunshine policy,” but Tokyo’s insecurities were crystallized by North Korea’s
Taepo-dong missile test in August. Though carrying just a satellite, this ballistic missile shocked
the Japanese because it flew directly over their main island of Honshu. Moreover the rocket’s
three-stage capability surpassed intelligence estimates and implied Pyongyang could now launch
larger and more destructive payloads.
266
Coupled with the DPRK’s stalled but still extant
nuclear technology, the new kind of missile was viewed very menacingly by Tokyo.
267
265
For an excellent summary of Japan’s experience during this time see Green, Japan ’s Reluctant Realism, 116-130.
266
Intriguingly, the Taepo-dong represented very little new danger to Japan. Its impact was therefore mostly
psychological. The No-dong missile could already target Tokyo and other major Japanese cities, and was apparently
- 118 -
By 1998, therefore, it can be said that Seoul’s and Tokyo’ perceptions of the North
Korean threat had reversed, with South Korea’s quickly falling below Japan’s. Kim Dae-Jung’s
sunshine policy toward the DPRK accelerated the trend, as Pyongyang responded to it by acting
less provocatively toward Seoul. As foreshadowed by the 1998 Taepo-dong test, North Korea
concomitantly became more aggressive toward Japan.
268
Especially in contrast with Seoul’s
ground-breaking achievements with Pyongyang over the next decade, Tokyo suffered failure and
experienced heightened fear and insecurity.
designed to carry nuclear warheads. For missile capabilities and the 1998 missile test see Bermudez, “Occasional
Paper No. Two,” 16, 21, 27, 29-30.
267
For Japanese reaction see Green, Japan’s Reluctant Realism, 124-128. The 1994 Agreed Framework ostensibly
ended North Korea’s attempts to develop nuclear technology that could be weaponized. Of course the technology to
date was not unlearned. For the agreement see the Korean Peninsula Energy Development Organization, “Agreed
Framework between the United States of America and the Democratic People’s Republic of Korea,” Geneva,
October 21, 1994, http://www.kedo.org/pdfs/AgreedFramework.pdf (accessed February 1, 2012).
268
For a good overview of North Korean provocations through early 2003, see Nanto, North Korea: Chronology of
Provocations.
- 119 -
Chapter 3
Theoretical Model #1: Delimited Balance of Threat Model
I. Overview
In this chapter, I endeavor to create my first theoretical model, one useful for explaining
the wild swings of historical animosity evident in Japan-Korea relations via security expectations
and the emotions produced when those expectations are satisfied or disappointed. Grounded as
much as possible within Neorealism, I generate it with the shortcomings of the four models
examined in Chapter 2 in mind. I particularly strive to produce a causal mechanism completely
linking security interest convergence and divergence to fluctuations in the intensity of the history
issue.
As discussed in Chapter 2, the four models I examined all suffer incomplete causal
mechanisms. Cha cannot explain why the Japanese engage Koreans in a zero-sum conversation
over the history issue, when what Tokyo desires is merely to distance itself from Seoul.
Relatedly, given his model’s reliance on Neorealism, Cha’s contentions that South Korea utilizes
the history issue to impose cost on Japan, and that Seoul and Tokyo are compelled by material
interests to improve the history issue, remain out of place theoretically. Regarding net threat
theory, Yoon does not comprehend that satisfied expectations regarding security loyalty produce
emotions that contribute to ameliorating the history issue. Moreover, Yoon’s inelegant model
lacks an active-enough causal mechanism to satisfactorily explain the often rapid and marked
shifts in the tenor Japan-Korea relations.
In contrast to Cha’s and Yoon’s scholarship, Lind’s model is complete in the sense that
no steps are missing between apology, backlash, perceived threat, and unfulfilled reconciliation.
- 120 -
Regrettably, her model attains elegance by ignoring the causal process leading to apologies. To
put it another way, Lind is missing the most important part of the mechanism: Unfulfilled
reconciliation, often in the form of a heated, ongoing, bilateral, nationalistic dialogue, leads to
apologies in the first place. Lastly, Koo’s effort to explain conflict de-escalation over
Dokdo/Takeshima is missing practically every aspect of the causal process. Simply stated,
beyond that predicted by top Neoliberal scholars he has no idea how trade dependence leads to
better relations.
Like Yoon and Lind, and to some extent Cha, I ground my model in Walt’s Balance of
Threat Theory. But using Walt’s Defensive Neorealism as the starting point for modeling is not
unproblematic because, as I explain shortly below, its ontological expansiveness permits the
inclusion of practically any international relations phenomena. Consequently I begin by
delimiting Walt’s theory so as to constrain it in ways more compatible with Neorealism.
Most worthy of note, I delimit its emotional ontology to just five primary emotions. The
emotions of material fear and material insecurity are active, meaning they compel direct action
by the actors who matter most. Three ideational emotions—enmity, friendliness, and
ambivalence—are passive in their influence. None of the primary actors undertake a specific
action when experiencing just these passive emotions. A sixth emotion—the desire to mitigate
material fear and material insecurity—is actually always in play. But I almost always ignore it
for reasons of streamlining, and therefore must ask the reader to understand that its presence is
implied whenever material fear and material insecurity are being considered.
In order to flesh out my delimited balance of threat model, I follow the example of Cha
and Yoon by incorporating into it the groundbreaking alliance politics research of Glenn Snyder.
Snyder’s work, which is completely compatible with Walt’s, is especially useful for helping to
- 121 -
establish the nature of things like security expectations, obligations, and loyalty. In fact, as will
readily be apparent, my dissertation hinges on Snyder’s emphasis on security expectations in all
alliances and alignments, and the predicted emotional reaction to expectations when they are
satisfied or disappointed.
It is worth stating upfront that I do not succeed in my quest to build a Neorealist
alignment politics model. It remains problematic in at least two ways. First, its Neorealism
necessarily conscribes to indirectness the relationship between material interests and ideational
cooperation and friction. Consequently, I am highly skeptical that the delimited balance of threat
model can explain the sharp gyrations in the tenor of Japan-Korea relations. Second, as I detail
in Chapter 4, the model simply is not Neorealist. That is, despite succeeding in crafting a model
that is more true to Neorealism than the four material comparison models, I still fail. Having
been unsuccessful, I argue Neorealism is especially poorly suited for the task of alignment
politics modeling. I therefore depart from it beginning with my second model.
II. Neorealist Alignment Politics
General application of Neorealist theory informs that Japan’s and Korea’s respective
security interests are likely to be key drivers of the history issue. I begin by joining Lind and
Yoon, and to a lesser extent, Cha, in adopting Walt’s balance of threat theory to explain how
convergence and divergence of Seoul’s and Tokyo’s Pyongyang policies impact the
politicization of the past. Like Cha and Yoon, I rely heavily on Snyder’s much more precise
alliance politics research to flesh out Walt’s basic theoretical insights.
A. Delimiting Walt’s Balance of Threat Theory
As noted earlier, Walt is widely considered to be a Defensive Neorealist. He argues that
aggregate power, geographic proximity, offensive power, and aggressive intentions comprise
- 122 -
state interests.
269
Adopting balance of threat theory is not straightforward, however, as Walt’s
fourth interest of aggressive intentions is notoriously slippery.
270
A matter of identity, aggressive
intentions can be stretched to include practically any international security phenomenon
impacting state perception of others. Not only can it encompass concrete phenomena like
declarations of war, troops amassed on a state’s border, recent aggression, and formal warnings
or threats, but there is no inviolable reason why it could not explain something as intangible as
Wendt’s ideal Hobbesian, Lockean, and Kantian systems.
271
Indeed, I could couch all three of
my theoretical models within Walt’s theory, if I so desired, since social psychological factors
impact perceptions of others and because I believe that Walt’s first three interests matter,
especially when refracted through his fourth.
Because I intend my first theoretical model to be as true to Neorealism as possible,
especially regarding emotions and expectations, I delimit Walt’s aggressive intentions interest as
follows:
1) States can evaluate others’ intentions as part of the process through which they estimate the likelihood of
conflict or cooperation over the short, medium, and long terms. States comprehend that the accuracy of
these calculations diminishes across time.
2) Perceptions of threat can change over time. States understand it is possible, even if difficult, to increase
or decrease the likelihood of conflict with others by proactively taking steps to change security perceptions.
3) While threats and opportunities can be perceived by people at all levels of society, the views of the
foreign policy executive usually matter most.
272
Nevertheless, the executive can be constrained and
269
Walt, The Origins of Alliances, 21-26.
270
For the “indeterminacy” of Walt’s theory, see Legro and Moravcsik, “Is Anybody Still a Realist,” 36-38.
271
Seen in this light, Wendt’s ideal worlds are simply very intricate statements about the kinds of intentions states
have. Wendt, Social Theory of International Politics, 246-312.
272
As I explain shortly below, I borrow Lake’s foreign policy executive, which he defines as “the high-ranking
bureaucrats and elected executive officials charged with the overall conduct of defense and foreign affairs…”
Though not immutably so, the executive acts “in the interest for society,” while others act “in the interests of
society.” Of course the executive need not be elected. More importantly, my society is much broader than that of
Lake, who is primarily concerned with trade and those involved in it. See David A Lake, “The State and American
- 123 -
compelled by societal perceptions and demands.
273
To put the point another way, any individual can react
to phenomena informing material security interests, but only to the extent he or she is politically aware.
Likewise, all people can respond to their perceptions of state interests, but their influence extends only as
far as their political relevance at a given moment.
4) States and societies suffer extremely stunted emotionality regarding all matters pertaining to material
security. They do, however, experience fear and insecurity, and desire things like economic growth,
military power, and alliances to the extent they are perceived to directly or indirectly mitigate fear and
insecurity. Aside from fear and insecurity, the predominant emotions states and societies can feel regarding
international material security are enmity, friendliness, and ambivalence. These three emotions emerge
both from events at the transnational level
274
and from materially important occurrences at the systemic
level. While individuals and groups can experience broad and rich emotionality at the transnational level,
real-world emotions at that level are significant in terms of material security only in so far as they
contribute to enmity, friendliness, or ambivalence. Thus, emotionally complicated transnational ideological
issues like the Japan-Korea history issue can influence security interests via the three emotions, though
these comprise but one part of one interest with which states must deal.
5) Relative to material fear and security, ideational enmity, friendliness, and ambivalence are passive
emotions. They are nevertheless important for two reasons. First, they are key determinates of the
perceptions states have regarding others’ intentions, and therefore influence overall levels of fear and
insecurity. Enmity, friendliness, and ambivalence therefore impact material security interests at the system
level. Second, these emotions also influence whether state leaders and pertinent members of society are
inclined to act cooperatively or obstinately on all sorts of issues otherwise unrelated to material interests.
Enmity makes people less cooperative on ideological matters, for example; friendliness induces more
cooperation. Emergent cooperation or obstinacy does not take place to the extent called for by more
action-demanding emotions, like anger or vengefulness. Those who experience enmity are, instead, merely
Trade Strategy in the Pre-Hegemonic Era,” in The State and American Foreign Policy, ed. John G Ikenberry, David
A Lake, and Michael Mastanduno (Ithaca, NY: Cornell University Press, 1988), 36-38.
273
For an interesting take on how social forces related to the Democratic Peace constrain leaders, see Jarrod Hayes,
Constructing National Security: US Relations with China and India (New York: Cambridge University Press, 2013).
274
As I clarify later, I categorize domestic-level processes and institutions as being transnational, so long as they
either are influenced by international phenomenon or exert international impact. Consequently, even an ostensibly
domestic election is transnational if, for example, politicians and voters are influenced by international events or the
election itself impacts international politics.
- 124 -
less likely to help resolve problems that arise, while those perceiving friendliness are less inclined to make
matters worse. Importantly, all proactive cooperation or obstinacy regarding ideological matters occurs
because of matters unrelated to international security. In this sense, as I detail below, all deliberate action
regarding ideas occurs randomly relative to material security interests.
6) While states can experience friendliness, they cannot internalize each other’s interests. Likewise, though
enmity exists, it is always the product of more-or-less immediate material or ideational phenomena. In
other words, the kind of enmity and “friendship” present in the third degree internalization of Wendt’s
Hobbesian and Kantian cultures of anarchies does not exist in Neorealist worlds.
275
7) States and societies enjoy expectational ability. Without it, they could not take part in alliances and
alignments. Likewise rivalry and neutrality would be impossible. Security expectations equate with Walt’s
aggressive intentions interest. The primary inputs of expectations are the emotions of fear and insecurity,
and enmity, friendliness, and ambivalence though, as I explain below, the non-emotional characteristics of
the systemic level can, on occasion, become determinate “as if” there were no security expectations. If
emotions power expectations, expectations are also important determinates of emotions. For example,
satisfied expectations of support within an alignment are significant because they ameliorate fear and
insecurity, and produce feelings of friendliness. Disappointed security expectations within an alignment
lead to higher levels of fear and insecurity, as well as enmity.
The above delimitation of Walt’s theory ascribes to the emotions of enmity, friendliness,
and ambivalence, and to security expectations, an important function in bridging the system and
transnational levels. By “system,” I am specifically referring to Walt’s first three interests:
aggregate power, geographic proximity, and offensive power, as they are made significant by the
275
Wendt deduces three cultures of anarchy—Hobbesian, Lockean, and Kantian—each of which can be internalized
up to three degrees. Third degree Hobbesian and Kantian anarchies produce states that are enemies or “friends,”
respectively. Importantly, their views of each other are compelled neither by the first-degree’s coercion or force,
nor are they shaped by the second-degree’s calculations of self-interest or cost/gain. Perceptions of others at the
third-degree of cultural internalization are instead a product of legitimacy. That is, in third-degree Hobbesian
cultures, states are enemies because they have come to measure their own worth in terms of their efforts to harm
other states. A failure to fight the enemy because of, say, economic desires or fear of some force external to the
system—perhaps global warming—would be tantamount to ongoing self-defeat defeat and failure. Likewise, states
in third-degree Kantian systems would perceive a direct loss to themselves if one of their “friends” suffered defeat
or catastrophe. The loss experienced is a product of the internalization of the other’s interests as one’s own, of
coming to view the Other as part of the Self. For cultures of anarchy and their internalization, see Wendt, Social
Theory of International Politics, 246-312, but esp. 272-278 and 305-307.
- 125 -
logics of anarchy Kenneth Waltz famously perceives to be structuring the international
system.
276
“Transnational,” in contrast, indicates all other political phenomena, so long as 1)
they impact international politics or are influenced by international occurrences; and 2) they are
neither the bridging emotions of enmity, friendliness, or ambivalence, as fully formed within the
bridging level, nor security expectations. Though many of its elements are yet to be defined and
explained fully, I illustrate Walt’s delimited theory in Diagram 3.1.
276
Waltz, Theory of International Politics.
- 126 -
Four points regarding the bridging phenomena in Diagram 3.1 require immediate
commentary. First, security expectations comprise the aggressive intentions interest. They
result from the interests and emotions of the system level and the three bridging emotions of
Diagram 3.1
Walt’s Neorealism, Delimited and with No Alignment Politics
Transnational Level
Actors: Including but not limited to politicians, bureaucrats, non-governmental
organizations, businesses, international organizations, and individuals, but only to the extent
they influence, or are influenced by, ideological phenomena in international politics.
Emotions: Broad and rich, as present in the real world. Can be influenced indirectly by the
bridging emotions.
Bridging Level
Emotions: Enmity, friendliness, and ambivalence, as informed by systemic phenomena
refracted by security expectations, and via transnational emotions.
Security Expectations—i.e., the aggressive intentions interest: Formed by the interests and
emotions of the systemic level, and the bridging emotions.
Actors: The foreign policy executive and pertinent others (FPE & PO). They are aware of
the material interests and emotions of the systemic level, as well as the emotions and security
expectations bridging the systemic and transnational levels. They are actors only to the
extent they are influenced by these logics at the transnational level, or can influence the
interests and emotions of the systemic level.
System Level
Actors: States
Emotions: Limited to fear and insecurity, and desire for material things to mitigate them.
Interests: Comprised by aggregate power, geographic proximity, and offensive power within
Waltzian anarchy.
- 127 -
enmity, friendliness, and ambivalence. Security expectations are perceived into existence by the
foreign policy executive and pertinent others (FPE & PO). Borrowing from David Lake, in
democratic states the FPE is comprised of “the high-ranking bureaucrats and elected executive
officials charged with the overall conduct of defense and foreign affairs…”
277
The PO, in
contrast, is comprised by groups and unaffiliated individuals that are not formally part of the FPE
but, like the FPE, are aware of their state’s security expectations to the point they react
emotionally, in a predictable manner, when these expectations are satisfied or disappointed by
material occurrences. Moreover, all members of the FPE & PO must either possess some
capability, even if miniscule, of influencing material security interests at the systemic level, or be
actors at the transnational level.
278
They may also be both.
My emphasis on the larger FPE & PO minimizes the significance of Lake’s point that the
FPE generally acts “in the interest for society,” while others act “in the interests of society.”
279
It
is diminished because my focus is not on how the FPE’s and PO’s differing interests directly
influence outcomes regarding material security. Indeed, while there is always potential for upset,
in the majority of cases the FPE is likely to dominate the PO in terms of determining interests.
Rather, what is of primarily concern for building alignment models is the joint emotional
reaction of both the FPE & PO to satisfied and disappointed security expectations. Crucially,
unless their respective interests regarding other states are radically different, their emotional
reactions to transgressed and well-met security expectations will be roughly similar. These
emotions—in the form of feelings of enmity, friendliness, and ambivalence—influence
277
Of course authoritarian states do not elect the FPE by popular vote. See Lake, “The State and American Trade
Strategy in the Pre-Hegemonic Era,” 36-38.
278
The point is definitional: If a member of the FPE & PO neither influences material security interests, nor
occupies a spot within the transnational level, then it simply cannot be a bridging actor.
279
Ibid., 38.
- 128 -
ideational phenomena at the transnational level and perceptions of material interests at the
systemic level.
Second, the FPE & PO exist solely neither at the systemic level, nor at the transnational
level. These actors are not, therefore, perfectly substitutable for the state at the system level.
They are likewise not purely transnational actors with no international security interests. The
FPE & PO are pushed and pulled instead via the logics of both levels, usually in combination.
Hence, even compelling material changes within the system level will almost always be viewed
by these actors through the lens of their security expectations, as informed both by past and
present systemic materialism and transnational ideas.
Third, in cases best illuminated by Walt’s theory, security expectations—i.e., the
aggressive intentions interest—will be a product of complementary emotional reactions to
systemic and transnational phenomena. As I detail below, the emotions will tend to correspond,
because the aspects of the bridging emotions produced by satisfied or disappointed material
expectations impact the tenor of ideological transnational matters indirectly while, at the same
time, the tenor of ideological transnational matters influences the formation of material security
expectations. In short, in longer term relationships, especially, one can expect a good measure of
circular reinforcement between security interests on the one hand and transnational ideational
phenomena on the other.
Nevertheless, security expectations can also form at those times when emotional
reactions to material and ideological inputs contrast greatly. For instance, in the event a state is
weak and/or distant to a very powerful one, system logics may induce expectations of reasonably
good material-security relations, despite the existence of high levels of transnationally derived
- 129 -
enmity.
280
Conversely, a powerful neighbor may induce expectations of conflict, despite
considerable enthusiasm about the otherwise threatening state.
281
One question worth pondering is what happens when a state experiences complete, or
nearly complete, ambivalence about another state, as might occur when material and ideological
inputs of emotion and, consequently, security expectations conflict? In these no-doubt unusual
cases, security expectations are likely to be established “as if” there were no such thing as
bridging emotions. Were this situation to occur, Walt’s reputation as a Defensive Neorealist
suggests one would be wise to adopt Waltz’s Defensive Neorealism as a means of understanding
the extent of the expectations likely to be present. Though no doubt a simplification, my earlier
argument about Waltz providing little room for expectations about the intentions of others might
hold true.
282
Fourth, and finally, practically all actors of the bridging level—the FPE & PO—are also
actors within the transnational level at any given moment. To a lesser extent the obverse is also
true, as many transnational actors comprise part of the bridging actors. Essentially, therefore, the
actors are individuals and groups who wear different “hats” when considering different issues;
they are like electrons occupying two different spaces at the same time. This is the case with the
highest ranking foreign policy executive, who must not only consider how to respond to the
expectations-informed material demands of the system level, but must also strive to ensure
his/her ability to govern domestically.
283
At the opposite end of the power spectrum, an
unimportant member of an unimportant Korean high school organization dedicated to raising
280
Consider the case of Saudi Arabia, regarding which many Americans felt obvious enmity following the 9/11
terrorist attacks. Systemic considerations produced expectations of good relations despite concerns about Saudi
responsibility for the attacks.
281
Consider the German Anschluss of Austria in 1938.
282
See my discussion of expectations in Neorealism in Chapter 2.
283
As is the case with all phenomena, domestic politics are transnational if they impact, or are influenced by, the
bridging emotions, and are neither system level interests and emotions, nor the bridging emotions and security
expectations. If, for example, a domestic election is impacted by foreign affairs, the election is transnational.
- 130 -
awareness about an ideological issue like the comfort women can be said to become an actor at
the bridging level when she begins worrying about whether Japan will develop nuclear weapons.
THE FLOW OF EMOTIONS IN DELIMITED BALANCE OF THREAT THEORY
In the delimited version of Walt’s Neorealism, emotions flow from the system level to
transnational level, and vice versa. As seen below in diagrams 3.2 and 3.3 there are constraints
on their movement, however. To better highlight the transmission of emotions, I temporarily
remove expectations from balance of threat theory in both diagrams and in their immediate
explications. Diagram 3.2 shows the direct transfer of the emotions of fear and insecurity, from
the system to the bridging and transnational levels.
284
In aggregate, all members of the FPE &
PO aware of the appropriate phenomena at the systemic level experience fear and insecurity
regarding their state’s material interests. Consequently, they act to mitigate their fear and
insecurity, to the extent they can, if they perceive their participation is required to carry out the
counteraction. Their collective action explains why systemic phenomena often elicit predictable
responses by states. Fear and insecurity derived from the system level also impact transnational
affairs, albeit indirectly, via the bridging emotion of enmity. Bridging actors who are also
members of a given state’s transnational society respond to these passive emotions by becoming
less likely to help resolve ideological problems vis-à-vis the state toward which there is enmity.
Since these transnationally sourced ideas involve broad and rich emotions, and because they are
more likely to be experienced under conditions of enmity at the bridging level, it can be said that
systemic fear and insecurity influences affect at the transnational level.
284
I exclude discussion of ambivalence, because it is a neutral emotion.
- 131 -
The movement of emotions from the transnational level to the systemic level originates
from the logics of transnational phenomena and is unrelated, in a direct causal sense, to
considerations of systemic material interests. It usually begins when an idea involving at least
two states is created and politicized, or is re-politicized, by a transnational actor operating for his,
her, or its own interests within the transnational level.
285
For example, a politician in State X
attempting to consolidate his or her position over domestic rivals might deliberately inflame an
idea pertaining to State Y. As the emotionally driven issue gains traction within X, it generates
widespread enmity toward Y. Should enmity due to this particular idea overpower the other
inputs of the aggressive intentions interest,
286
leaders and pertinent members of society in X
285
Ideas can become politicized accidentally, too. Hence, “usually” in the sentence.
286
As explained earlier, the inputs are the emotions of enmity, friendliness, and ambiguity, as they are produced by
fear and insecurity at the system level and most real-world emotions at the transnational level.
Diagram 3.2
Flow of System-Level Emotions to the Transnational Level
State X’s material fear and insecurity decreases
due to State Y, at the system level.
X experiences
more
friendliness
toward Y.
Emotions related to problematic ideological
issues are less likely than normal to arise,
because leaders and pertinent members of society
within X become less inclined to aggravate
ideological problems regarding Y, should they
arise at the transnational level.
State X’s material fear and insecurity increases due
to State Y, at the system level.
X
experiences
more enmity
toward Y.
Emotions related to problematic ideological
issues are more likely than normal to arise,
because leaders and pertinent members of society
within X become less likely to help resolve
ideological problems regarding Y, should they
arise at the transnational level.
FPE & PO experiences
more fear and insecurity
regarding material
interests.
FPE & PO experiences
less fear and insecurity
regarding material
interests.
- 132 -
come to perceive an increase in the likelihood of material conflict and a decrease in the
likelihood of material cooperation vis-à-vis Y. If so, the consequences are increased levels of
fear and insecurity at the system level. The process through which the varied and deep emotions
of the transnational level are bottlenecked into the three bridging emotions is shown in Diagram
3.3.
Much remains to be clarified regarding my delimitation of Walt’s theory and application
of it to alignments, in general, as well as to the Japan-Korea history issue in particular. I
especially must explain expectations and detail the actors and logics comprising the transnational
level. I also need to argue what systemically induced fear and insecurity can and cannot do at
the transnational level. I turn to these tasks immediately below, as part of and following my
discussion of Glenn Snyder’s groundbreaking alliance politics work.
Diagram 3.3
Flow of Transnational-Level Emotions to the System Level
Transnational phenomena unrelated to security
interests politicize a contentious idea within the
X-Y relationship.
X and/or Y experience enmity toward the other.
The FPE & PO in X and/or Y perceive an
increase in the likelihood of material conflict
and a decrease in the likelihood of material
cooperation vis-à-vis the other.
Material insecurity and fear increases at the
systemic level.
Transnational phenomena unrelated to
security interests depoliticize a contentious
idea within the X-Y relationship.
X and/or Y experience friendliness toward
the other.
The FPE & PO in X and/or Y perceive a
decrease in the likelihood of material
conflict and an increase in the likelihood of
material cooperation vis-à-vis the other.
Material insecurity and fear decreases at the
systemic level.
- 133 -
B. Definitions Based on Snyder’s Alliance Politics
Before explicating how divergent and convergent security policies impact the history
issue under Neorealism, it is essential to highlight a few key terms. I adapt the definitions below
from Snyder’s research on alliance politics. I say “adapt,” because Snyder’s work suffers some
irregularity about both the expectational and emotional capabilities of states, a matter I detail
later. Consequently, I must delimit Snyder’s definitions to better match Neorealism, the theory
to which he comports most closely while writing about the alliance security dilemma, his most
important contribution to the study of alliance politics. Since Snyder is somewhat inconsistent in
his Neorealism, I frame the aspects of his modeling I adapt within my understanding of Walt’s
Neorealism, as demarcated above. Though, as stated here, the results of this adaption process
may seem predestined to controversy, in practice my application of Snyder’s work in my first
theoretical model is fairly straightforward.
The alliance security dilemma is built upon Snyder’s deduction that one alignment
partner’s fear of abandonment is inversely related to another’s fear of entrapment. How
dependent one partner is on the other impacts the dilemma greatly. For example, highly
dependent states tend to fear abandonment, while more powerful alignment members usually
fear entrapment. Tensions emerge from the former’s efforts to ensure the latter’s involvement,
while the latter endeavors to temper its commitment to the alignment.
287
As I detail considerably in Chapter 5, Snyder’s alignment security dilemma is remarkable
for being the most Neorealism-friendly part of his fascinating research project. Consequently,
scholars like Cha and Yoon who adapt Snyder’s work tend to utilize the abandonment-
entrapment mechanism, and the security expectations underpinning it, in an overly Neorealist
fashion. This propensity is unfortunate, because it encourages scholars to overlook the dynamic
287
See especially, Glenn H. Snyder, Alliance Politics (Ithaca, NY: Cornell University Press, 1997), 180-192.
- 134 -
nature of security expectations as they change across time. It also prevents them from
adequately considering the social emotions likely to be accompanying such a deeply social
process.
Alliances and alignments are essentially the same thing, according to Snyder.
288
Both
usually require two or more states to develop mutual expectations of support on any number of
security matters.
289
Alliances and alignments can be set against a specific threat or, more
expansively, may be with a state against all potential security challenges. Alliances are, by
definition, underwritten by formal public or secret agreements between two or more states.
Nevertheless, Snyder notes that the significance of alliances “lies not in the formal contract but
in the expectations that are supported or created…among all interested bystanders.” While the
formal nature of alliances facilitates, ceteris paribus, stronger expectations of security support
than informal alignments, alliances do not always enjoy expectational advantage over alignments.
Some alignments are marked by very strong expectations, and some allies suffer incredible doubt
about the reliability of their partner(s). In sum, alliances are a subset of alignments, and both
exist only to the extent they are underpinned by expectations of support.
290
Because alliances are
a type of alignment, I use terms related to ‘alignment’ to refer both to aligned states and allied
states. Likewise states can be ‘partners,’ despite having no formal security agreement. Only
when I use words founded on ‘ally’ do I refer solely to alignments upheld by formal
agreements.
291
288
Walt fully agrees on this point. Walt, The Origins of Alliances, 12.
289
Even one-way alliances and alignments are possible. See Snyder, Alliance Politics, 12.
290
Snyder, Alliance Politics, 6-16. Quote is from Ibid., 9. A formal contract is a key indicator of the ‘degree of
commitment’ of aligned states. Nevertheless, it is not the only determinate of expectations. For a quick overview
see Ibid., 188-189.
291
Terms like ‘quasi-allies’ represent an obvious exception to this rule.
- 135 -
Expectations of security support on one or more security issues must be present for
states to be considered aligned. Even very long-term alliances cease to exist once all
expectations of security backing are shed, though the pieces of paper formalizing the erstwhile
alliances may not yet be nullified. Conversely, the weakest alignment can be said to come into
being despite being instantiated by extremely feeble expectations. Snyder intriguingly speculates
that state intentions do not matter in alignments. It might be said, therefore, that alliances and
alignments can exist despite the accurate knowledge of all members that they will never uphold
their own obligations to others. Indeed, what matters most when talking about the dynamics of
alignment politics is whether states expect their security partners to uphold their stated or implied
obligations to the alignment, and the strength of those expectations.
292
According to Snyder, allied and aligned states approximate their expectations of their
partners in at least four ways.
293
Of these, three “incentives” shed some light on a partner’s
future behavior. The partner’s degree of commitment—i.e., how clearly, formally, publically,
repeatedly, and unchangingly it demarcates its role in, and views of, the alignment—is the first
incentive.
294
Second is the alignment members’ relative dependence on each other. It helps to
establish security expectations, because it illuminates a partner’s general interests within the
alliance security dilemma. Snyder’s third incentive is interest, by which he means each partner’s
respective view of threats and opportunities presented by others, and regarding which alignment
members must “bargain.”
295
292
Ibid., 6-16, esp. footnote ‘*’ on 6.
293
Ibid., 186-189.
294
Written and spoken commitments are important in material models, because states obviously abrogating their
responsibilities to others come to be seen as unreliable. States viewed as untrustworthy bear material costs for their
caprice or dishonesty in iterated games with others. See Keohane, After Hegemony, 105-106.
295
Snyder, Alliance Politics, 186-192.
- 136 -
Though the three incentives inform expectations to some extent, they “yield only
uncertain judgments” about whether an aligned state will uphold its commitments. As a result,
Snyder argues states most heavily rely on a fourth factor: past behavior. Whereas the incentive
of degree of commitment comprises what the partner has said it will do and how it will do it, past
behavior refers to what it actually has done in the recent and distant past.
296
Obviously, partners
with lengthy histories of following through well and dependably on their alignment obligations
inspire stronger expectations than alignment members with either minimal or bad reputations.
The three incentives and the past behavior variable influence expectations in any possible
combination, and can be complementary or work at cross purposes. For example, even states
with a poor history of working together on security matters, and with no formal alliance or
fungible security dependence on other matters, might develop tremendously high mutual
expectations if they suddenly find themselves exceptionally dependent on each other, due to an
obvious threat from an implacably aggressive and comparatively powerful state. Conversely, a
state recently reassured in the strongest terms by its long-term, faithful ally should enjoy strong
expectations of support, even if their shared interest in the security challenge at hand is not
obviously convergent, and the faithful partner is apparently only minimally dependent on the
alignment.
Importantly, Snyder’s determinates of alignment expectations are completely compatible
with both Walt’s original theory and its delimited version. Indeed, each of Snyder’s three
incentives and his past behavior variable comprise an aspect of balance of threat theory. Degree
of commitment is often a critical determinate of the bridging emotions and security expectations.
Relative dependence is a material security interest understood via aggregate power, offensive
296
There is some overlap, as degree of commitment is clearly a part of past behavior. For quote and past behavior
see Ibid., esp. 189.
- 137 -
power, and geographic proximity. Issue interest is highly predictable via understanding of
systemic interests and security expectations.
297
Finally, the history variable influences both
security expectations and the bridging emotions—i.e., the aggressive interest variable.
Alignment obligations are created by expectations in Neorealism. Though true, it is not
obvious that expectations determine security obligations so powerfully.
298
Would it not be the
case that an obligation can be said to exist if state Y truthfully declares its intention to support
state X on some security matter, even if Y correctly understands that every pertinent observer to
Y’s claim, including X, completely believes Y’s declaration false? Though seemingly
reasonable, the rhetorical logic is incorrect because, under Neorealism, states like Y lack the
emotionality necessary to feel compelled to honor their obligations for ideological reasons. Only
if Y fears X’s response, or fears more generally that it will endure insufferable reputational
damage among key observers, will Y desire to honor its stated obligation, for the obligation’s
own sake.
299
At the very least, therefore, if Y is to perceive itself as being obliged, Y must
expect that X and/or important others expect Y to uphold its commitment to X. If neither X nor
Y fears the outcome of an unmet obligation, then none exists in Neorealism, no matter past or
present statements to the contrary.
300
Conversely, if Y fears the consequences of not satisfying a
particular security expectation of a significant other, then it is obligated by it.
301
297
Consider two obvious reasons why the US restored Kuwaiti independence in the first Iraq War. America’s
aggregate power and offensive power made it an essential participant and the primary leader of the UN-sanctioned
war effort.
298
Indeed, Snyder is somewhat unclear about obligations, as I explain in the next paragraph.
299
Y might decide to honor its stated alliance commitment out of its own fears of the specific threat. If so, its
decision results from interests paralleling the obligation, not from the obligation itself. Ibid., 350-351. Likewise, Y
can opt to follow through on its disbelieved, stated commitment in order to convince its partners and all observers
that it is reliable. But this is not an obligation, because Y is acting to create future expectations, not to fulfill current
ones.
300
Snyder argues there is a “moral” element upholding obligations. That may be the case in the real world, but it
cannot be an important process in a Neorealist explication, because states in Neorealism lack the emotional
capability required to feel compelled by ideological factors. Snyder is inconsistent on morality and other ideological
phenomena, a matter I detail later. For now I note that Snyder seems to have downplayed by too much the
- 138 -
Two kinds of security obligations exist: stated, or formal obligations; and implied, or
informal ones. Snyder never carefully delineates the two types. Indeed, he is frequently
inconsistent about the meaning he attaches to the term ‘obligation’ itself. At times, Snyder
writes as if security obligations are commitments one must make consciously and even
formally.
302
In this sense a state is obliged only when it promises an alignment partner that it
will defend its interests in some particular way, and in exchange for a specific and reciprocal
promise from the partner. Yet Snyder understands that obligations can exist even if one does not
make an actual commitment. Military demonstrations and statements of diplomatic support can
unintentionally create military obligations, for instance.
303
Furthest from conscious and formal
commitment, Snyder contends in one of his final analyses that a recipient’s expectations—even
if completely false—can create powerful obligations for its alignment partners.
304
Consequently, I treat obligations as stated if they meet Snyder’s first criteria. Implied
obligations are commitments expected by the recipient that are not formally made by the giver.
Implied obligations need not be entered into consciously, or even be understood by the giver to
exist. Because stated obligations coincide with formal declarations demarcating the commitment,
ceteris paribus, they induce stronger expectations than do implied ones. Importantly, however,
both stated and implied obligations are actualized via the process of expectation formation, as
Neorealism-compatible argument that alignments change strategic interests as states line up with and against each
other, and as alignments become more institutionalized and/or intertwined materially. Ibid., 350-359.
301
If one follows this logic to its furthest extent, Y is obligated by the expectations of others even if it was either
initially unaware of their expectations or never intended to satisfy them. This logic holds in Neorealism, so long as
Y fears the consequences of failing to meet these expectations.
302
See, for example, Ibid., 35-36.
303
Ibid., 15-16, 359.
304
Snyder contends that statesmen who understand this dynamic work very hard to ensure that false expectations
never take hold among their alignment partners. See Ibid., 358-359.
- 139 -
outlined earlier. And they are present only to the extent they are expected into existence by the
recipient of the obligation.
305
Security loyalty is something all states in alignments expect to receive from their
security partners. When alignment members fully honor their obligations as expected by
recipient states, they are completely loyal.
306
If they completely fail to fulfill their obligations
when action is required, they are totally disloyal. Between these two poles falls a mixture of
imperfect loyalty and incomplete disloyalty.
307
It is essential to note that the phrase
‘expectations of security loyalty’ is no different from terms like expectations of ‘security support’
or expectations that states will ‘honor their commitments.’ An alignment member only obligated
informally to supporting its partner in a specific UN Security Council resolution, for example,
will be perceived by the partner as being disloyal if it does not perform the action expected of it.
Whether its disloyalty matters usually depends upon the combination of the depth of the
expectations held by the recipient, and the importance of the particular security issue. Ultimately,
asserting that states expect loyalty from their security partners does not mean they expect
alignment members to uphold most, or even all, of their security interests. That said, there is no
reason why obligations cannot be extremely expansive in unusual cases, or when two or more
states embark on a very broad and deep, long-term alliance.
308
305
Though there is no doubt some causal circularity between expectation formation and the existence of stated
obligations, in most instances expectations are formed before the obligation is stated. Brian Rathbun likewise
argues that trust in institutions predates acts of international cooperation. I detail further my views on this causal
phenomenon, as well as Rathbun’s work, in my third theoretical model. See Brian C Rathbun, Trust in International
Cooperation: International Security Institutions, Domestic Politics and American Multilateralism (New York:
Cambridge University Press, 2012).
306
Observing states have expectations, too, though these might be different from the expectations of the recipient
state.
307
I am treating treasonous behavior as if it falls outside of the loyalty-disloyalty poles.
308
Snyder does not define security loyalty. But he consistently uses the term as I describe in this paragraph. See,
for example, Snyder, Alliance Politics, 11, 23, 36, 46, 47, 88, 184, 198.
- 140 -
C. Causal Mechanism
Truncating Walt’s aggressive intentions interest and recombining it with his other three
interests produces an extremely passive causal mechanism linking Japan-Korea security conflict
and historical animosity. When divergent interests lead to conflicting security policies, leaders
in Seoul and Tokyo experience enmity, which is a product of their disappointed security
expectations. Enmity impacts relations indirectly, as it makes the FPE & PO more unlikely to
take part in resolving historical flare-ups, which emerge from temporally random transnational
historical friction acts.
309
Enmity also increases fear and insecurity, because it impacts
perceptions about the likelihood of bilateral cooperation and conflict on important security
matters. Consequently, Japan and/or South Korea are likely to experience increased levels of
fear and insecurity regarding other states, like North Korea or China, and perhaps even each
other. Members of the FPE & PO in Seoul and Tokyo almost never respond to these action-
demanding emotions by lashing out at each other with the history issue, however. As I explain
shortly below, they refrain because no powerful incentive compels them.
Security interest and policy convergences mitigate historical animosity in a similarly
indirect manner. Namely, confluence results in satisfied alignment expectations. Feelings of
bilateral friendliness result, meaning the FPE & PO in Japan and Korea become less inclined to
aggravate the history issue should it become politicized by matters unrelated to security. My
indirect explanation for ideological cooperation stands in sharp contrast with those Neorealist
scholars who assert atheoretically that Japan’s and Korea’s coincident and heightened fears and
insecurities provide Seoul and Tokyo compelling reasons to minimize all kinds of
disagreements—including history-related problems—in order to maximize their ability to
cooperate more effectively on security matters. As I explain later, such claims are problematic,
309
They are temporally random only in the sense they have nothing to do with material security interests.
- 141 -
because they elevate ideas to a level equal to material security concerns, a fact representing an
impossibility in Neorealism.
FEAR AND INSECURITY ALMOST NEVER LEAD TO POLITICIZATION OF IDEAS
Why, in the delimited balance of threat model—i.e., in Neorealism—do fear and
insecurity not lead to deliberate and punitive use of the history issue? The short answer is that
states in Neorealism lack the emotions making ideological attacks a preferred option in the real
world. Neorealism’s stunted emotionality and rational materialism produce logics through which
states seldom lash out, even materially, when fearing abandonment by allied or aligned security
partners. Indeed, Neorealist rationales explain Snyder’s central finding that the “standard
response” of states experiencing abandonment fears “is to move closer to the ally in some way so
as to increase its perception of one’s loyalty.”
310
To understand why coercion will seldom be used against defecting partners, consider the
following scenario: State Y hedges its alignment with state X regarding state Z, which X fears.
X consequently experiences fear of abandonment by Y and, therefore, greater insecurity
regarding Z. At this point, it seems intuitive and not unreasonable to assert that security
imperatives lead X to impose costs of some kind on Y, because of Y’s role in increasing X’s
insecurity and/or because X wants to compel Y to uphold its alignment obligations. It is actually
awkward to make this claim, however, unless the analyst declares that states have advanced
emotional capability. With emotionality conscribed along the Neorealist lines discussed above,
X is very unlikely to impose material costs on Y. The likelihood of X coercing Y’s loyalty via
an ideological phenomenon like the history issue is even more remote.
An emotionally stunted, yet abandonment-fearing X will usually refrain from imposing
material costs on hedging Y for three reasons. First, X cannot persecute Y out of some
310
Snyder, Alliance Politics, 183-184.
- 142 -
perception of injustice related to disappointed alignment expectations. Neither fear and
insecurity, nor enmity, friendliness, and ambivalence are adequate emotional determinates of just
and unjust action; other emotions are needed at the causal level for injustice to be perceived and
acted upon.
311
Second, and parallel with the first point, X is incapable of imposing costs out of
anger, spite, or vengefulness, because it simply does not experience these emotions adequately.
MATERIALLY COERCED LOYALTY IN NEOREALIST ALIGNMENTS
However unlikely, X is capable of deciding to coerce Y’s behavior via a third
rationalization. Responding to its own fears of abandonment, X can leverage material costs to
warn Y that it risks abandonment if it remains an unsatisfactory alignment partner. In this
situation, X imposes economic or security costs on Y to induce in it fear and insecurity. If Y
experiences fear and insecurity beyond the levels of emotionality causing it to defect from X in
the first place, Y might return to the old alignment status quo.
Though not impossible, the above material compulsion scenario will normally not take
place. One mitigating factor emerges from Neorealism’s stunted emotionality. As argued above,
X is constrained emotionally and cannot pass judgment on Y’s defection. Similarly, X is
incapable of being attached to Y’s alignment affectively. X will therefore only coerce Y’s
alignment recommitment if doing so is the best method for improving X’s security materially. If
it is not, X will pursue other options such as expressing loyalty toward Y, courting new partners,
developing new military capabilities, or even improving relations with Z.
312
X is unlikely to compel Y because, as Snyder clarifies, intimidation is dangerous. Y’s
prior defection has already indicated that the balance of its material interests requires it to
311
Robert C Solomon, “Sympathy and Vengeance: The Role of the Emotions in Justice,” in Emotions: Essays on
Emotion Theory, ed. Stephanie van Gooze, Nanne Van de Poll, and Joe Sergeant (Hillsdale, NJ: Lawrence Erlbaum,
1994), 291-311; Solomon, A Passion for Justice.
312
Improving relations with Z might also express loyalty to Y, depending on the circumstances. For ways to reduce
dependence see Snyder, Alliance Politics, 167.
- 143 -
distance itself somewhat from its alignment with X, at least as far as Z is concerned.
Unfortunately for X, successfully compelling Y to return to the old status quo will diminish the
value Y places on the alignment to the same extent that Y appreciated its semi-defection.
Coercion consequently risks pushing Y passed the threshold beyond which it views its alignment
with X as undesirous. Once that point is reached, X’s threats to abandon Y would end up leading
Y to assuage its own fears of abandonment by courting Z more aggressively, establishing new
security relationships, or by enhancing its military power, among other options. Only if Y does
not cross the threshold might it assuage its own fears of abandonment by expressing loyalty to
X.
313
Importantly, the entire experience diminishes the bilateral relationship more than just
materially. For hedging and coercion generate enmity, thereby impacting each state’s estimate
of the likelihood of bilateral security cooperation and conflict moving forward. X’s enmity
toward Y emerges when Y’s hedging disappoints X’s alignment expectations. Y’s enmity
toward X is produced when X attempts to coerce it materially, thereby disappointing Y’s
alignment expectations.
314
The upshot is that X must view as still desirable the relationship it
now perceives as diminished, if it is to coerce Y’s compliance.
315
In parallel, a materially
coerced Y must evaluate the ideologically devalued relationship as being above its “stay”
threshold, if it is to express loyalty by returning to the old status quo.
316
In contrast with materially punishing Y to coerce its loyalty, X’s option of exhibiting
loyalty to Y is generally a much better option. True, moving closer to Y requires X to express
loyalty despite its perceptions of a materially and ideologically diminished alignment. But
313
Snyder, Alliance Politics, 184-185.
314
This holds so long as Y expects X to be loyal to Y’s endeavors to improve relations with Z.
315
X must also feel this way if it were to express loyalty to Y.
316
The term is Snyder’s. Ibid., 184-185.
- 144 -
showing loyalty has great value, too. Namely, it helps establish X’s reputation for loyalty with Y
and others. Rather than feeling enmity and discerning a materially and ideologically devalued
alignment, as would have happened in the event X coerced Y’s loyalty, Y comes to view its
partner via the lens of friendliness. This emotion compels Y to perceive that X will be more
cooperative and less conflictual moving forward, a fact that may explain Snyder’s conjecture that
X’s exhibition of loyalty might grant it greater leeway in future, interest-specific bargaining with
Y.
317
In sum, whereas coercion is a risky way for X to assuage its abandonment fears, moving
closer to Y is a much safer strategy for feeling secure, at least as far as the bilateral X-Y
alignment is concerned.
While X should normally feel compelled to express loyalty to Y by the advantages listed
above, Snyder’s work informs that X’s preferred course of action will also be shaped by its
interests in the alliance security dilemma. Ceteris paribus, X will be less likely to move closer to
Y if it fears getting entrapped in Y’s conflicts with others. Conversely, if X is not particularly
concerned about entrapment in Y’s affairs, it will be more likely to express loyalty to Y.
318
It is worth enumerating some of the reasons why X’s desire to assuage its fears of
abandonment may lead it to compel Y materially, instead of enticing it with an expression of
loyalty. First is the entrapment-fear scenario described just above. Second, X might choose to
intimidate Y if they share a security relationship in which X enjoys considerable material
317
I infer Snyder’s supposition from his argument about commitments being “interdependent,” and his claim that a
state may feel increased pressure to support its ally, if the state had defected from the ally in a prior bargaining
session. See Ibid., 183-185.
318
In Snyder’s alliance security dilemma, avoiding entrapment in an alignment partner’s conflicts is the fundamental
reason why states resist drawing closer in alignments. Ibid., 180-183. This dynamic is important to understanding
the Japan-ROK case. The primary historical antagonist of the time periods I examine—i.e., the state, not
coincidentally, whose security expectations were disappointed by its partner’s efforts to establish better relations
with North Korea—had little reason to worry about being entrapped in the other’s conflicts at that time. Snyder’s
alliance security dilemma consequently suggests that Seoul had tremendous material incentive to exhibit loyalty to
Tokyo whenever Japan courted North Korea, especially through the mid-1990s. Likewise, Tokyo should have
shown loyalty to Seoul after 2000, as South Korea initiated the sunshine policy to warm relations with Pyongyang.
Because each time period’s respective, expectationally disappointed state did not normally express loyalty, I argue
later that Neorealism’s materialism and stunted emotionality cannot explain the Japan-Korea history issue.
- 145 -
leverage. Alliances or alignments in which Y is highly dependent on X—e.g., a deeply
institutionalized military alliance, an economic relationship in which X enjoys tremendous
advantage, or a security partnership marked by X’s overwhelming military dominance—
certainly meet this criteria. Third, desperation could lead X to warn Y that courting Z risks
material loss and abandonment. In particular, X might be inclined to go this route if it had no or
limited means of building up its military power, finding new allies, or solidifying its alignments
with others. Likewise, X would have to perceive that better relations between Y and Z
represented an unacceptable security risk. And, fourth, chance can lead to coercive enforcement
of alignment loyalty.
319
Importantly, the above logics should not regularly lead X to enact forceful behavior, even
if it is relatively unconcerned about the risks of pushing Y past its stay threshold. In fact, if X’s
entrapment fears are key, X should neither frequently nor repeatedly suffer Y’s hedging. Y’s
semi-defection should be rare because, as Snyder clarifies, the more powerful alignment member
is the partner prone to experiencing entrapment fears, not the weaker one.
320
But if jilted X is
more powerful than hedging Y, Y will naturally be disinclined to semi-defect from X in the first
place. Indeed, being comparatively weaker than X, Y’s primary concern in their bilateral
alignment politics should be to avoid abandonment by X.
321
In other words, while weaker states
can and do semi-defect from their stronger partners, Snyder does not grant Y strong incentive to
semi-defect from X.
In fact, no matter which partner is dominant militarily, entrapment fears are unlikely to
bring X to a coercive position. The chance of entrapment-inflicted coercion is remote, because
Y’s attempts to better relations with Z, to the extent they are successful, should decrease Y’s risk
319
Chance in Neorealism subsumes materially irrational decision-making.
320
Ibid., 166-168.
321
Ibid., 180-186.
- 146 -
of conflict with Z. The entrapment downside for X of expressing loyalty to Y, and therefore
being more tightly bound to its alignment with Y, is consequently diminished. With its
entrapment fears assuaged, X should prefer exhibiting kindness to Y.
322
Since it is awkward to argue that states fear entrapment when their security partners try to
establish better relations with an adversary, it makes sense to limit discussion to the role of
abandonment fears. If jilted X suffers from strong fears of abandonment due to hedging Y’s
move toward Z, however, X is likely to be the weaker state in the X-Y alignment. Moreover,
X’s abandonment fears will correlate positively with its security dependence on Y.
323
Ironically,
the more X fears abandonment, the less power of coercion it will have over Y. Conversely, the
less X cares about Y’s move toward Z, the more capability it will have to punish Y materially. It
is noteworthy that a similar dynamic works from Y’s perspective: The more material leverage X
enjoys over Y, the less likely Y will be to semi-defect from it in the first place. These logics
decrease considerably the likelihood of jilted states coercing hedging partners from positions of
strength.
Desperation seems to have more potential to explain how, in Neorealism, abandonment
fears lead the jilted to lash out materially at their hedging alignment partners. It seems
reasonable to assert that X may attempt to coerce Y’s loyalty if X perceives it has no other
options to counter its insecurity. As noted earlier, X should be especially prone to compel Y
when X perceives Y’s improved relationship with Z to be an unacceptable security risk.
Moreover, since the jilted state is normally the weaker and more dependent state, it makes sense
322
One counterargument to my logic might be that X could fear entrapment in Y’s conflicts with states other than Z.
That outcome is especially possible in multipolar systems, and should be considered inductively, on a case-by-case
basis. It is highly unlikely in bipolar systems, however, since Z should have security ties to the competing faction.
323
Ibid., 166-168.
- 147 -
logically to assume that X will sometimes suffer severe-enough abandonment fears to feel
desperate.
Desperation is no incontrovertible argument for coercion, however. It remains
problematic, for instance, that the more desperate jilted X is, the weaker it is likely to be relative
to hedging Y. When in particularly dire straits, X is therefore unlikely to have material power
adequate to the task of instilling bilateral abandonment fears in Y. Because its desperation is
negatively correlated with its influence to compel, X should normally only attempt coercion
when better relations between Y and Z are completely and absolutely unacceptable. In this event,
however, compelling Y’s loyalty becomes especially dangerous. As discussed earlier, it risks
pushing Y further from its alignment with X. It may even push Y closer to Z, a possibility that
would put X in an even more untenable situation. Facing such terribly risky and unacceptable
outcomes, when stripped of adequate emotionality by Neorealism, X’s decision to accept or
reject Y’s move toward Z is likely to correlate well with chance, the fourth reason why X might
coerce Y. Of course chance is random and will wash out X’s agency in any longitudinal
behavioral study. Consequently, chance is much less important and interesting than the first
three factors that might lead X to compel Y materially.
The above discussion does not mean jilted X will never coerce hedging Y materially
when operating under strict Neorealist logics. The arguments I lay out are in no way absolute
enough to eliminate X’s agency, especially when particularistic circumstances arise. Indeed, my
intention was merely to invoke, and explain and support, Snyder’s finding that states fearing
abandonment will most often assuage their concerns by treating their hedging security partners
with kindness, not to argue that other outcomes are impossible.
- 148 -
IDEOLOGICALLY COERCED LOYALTY IN NEOREALIST ALIGNMENTS
Crucially for the purposes of my overall argument, if states experiencing abandonment
fears in Neorealist systems are unlikely to compel their hedging alignment partners materially,
they are even less likely to attempt to coerce them with ideological matters like the Japan-Korea
history issue. Not only do the logics mitigating the desirability of material coercion apply more
or less equally to ideological coercion, but it simply makes no sense, prima facie, for states
operating under Neorealist logics to deliberately leverage ideological force in hopes of
compelling material response.
Nevertheless, and returning to the discussion of states X and Y above, if X were
determined to utilize an idea to compel better behavior in Y, how might it happen? The answer
is X would have to rationalize that an idea can generate security fears equal to or exceeding the
materially-derived insecurities that led Y to semi-defect from X in the first place. X might hope
to do this in one of two ways. First, it might utilize an ideological matter to instill fear in Y
directly. If X can politicize an idea to the point that it pains Y tremendously, Y’s fears of X’s
ongoing idea barrage could prompt it to end its materially-derived hedge. Of course such an idea
is preposterous in Neorealism, since states lack the emotional depth necessary to realize
immaterially derived fears at anywhere near the potency of the material insecurities they suffer.
Second, X might use an idea to instill fear in Y indirectly. In this instance, X leverages
an ideological matter as both a signal of its dissatisfaction with Y and its intent to punish Y
materially. While this rationale appears much more plausible in Neorealism than the first one, it
remains terribly problematic. No doubt the biggest concern is that Neorealist theory cannot
imbue ideological matters like the history issue with adequate social significance in the realm of
- 149 -
material security.
324
Ideological issue signals are consequently no more useful in resolving
alignment security crises than simple statements of dissatisfaction accompanied by firm
warnings of material intentions. In fact, leveraging contested ideas to convey intentions is much
riskier than straight-forward demarches, because the former leaves considerably more room for
message misinterpretation than the latter. The possibility of states in Neorealist systems ever
signaling their material intentions indirectly, via an ideological issue—to say nothing of the
chance they might develop a preference for this strategy—is obviously practically nonexistent.
THE POLITICIZATION AND DEPOLITICIZATION OF IDEAS IN NEOREALISM
As briefly outlined earlier in this section, in the delimited balance of threat model, the
relationship between international security matters and the politicization and depoliticization of
ideas is indirect. It occurs through enmity and friendliness, respectively. The relationship
between security interests and the aggravation of ideological phenomena is, in fact, extremely
indirect. Indeed, relative to Neorealist security concerns, states initiate idea contestation
completely randomly. Ideological conflict occurs at random because ideas become politicized
somewhere other than the level in which states bargain over international security matters, and
security interests do not impel the deterioration of ideological matters. As I explain below, the
ideological action takes place at the transnational level, and security interests are merely
permissive of the results of transnational ideological contestation.
Diagram 3.4 shows how interests in Neorealism contribute to idea conflict. As Snyder
predicts, X’s fears of being abandoned by Y lead X to move closer to Y. Y’s security
expectations are satisfied by X’s actions, and the FPE & PO in Y consequently experiences
324
As argued earlier, the FPE & PO can only feel the emotions of fear and insecurity, and enmity, friendliness, and
ambivalence. Any individual considering a material security issue is by definition a member of this group.
Consequently, the entire audience for an ideological signal of intent regarding the possible termination of a material
alignment cannot perceive its real-world significance.
- 150 -
friendliness toward X and becomes less likely to aggravate any ideological contestation
emerging from transnational politics. In contrast with satisfied Y, X’s FPE & PO experiences
enmity in response to its disappointed security expectations, and therefore becomes less likely to
help alleviate any ideological problems that might develop concomitantly at the transnational
level. Enmity also means that members of X’s FPE & PO remain just as inclined as ever to
utilize ideational conflict for their own transnational purposes, such as gaining domestic political
advantage. X therefore contrasts with Y, as security induced friendliness makes Y’s FPE & PO
less like to aggravate ideas for domestic gain.
325
Nevertheless, Y’s propensity for inaction is not
accompanied by a desire to better ideological ties, because friendliness is not an active emotion.
Should transnational processes produce ideational conflict, Y will be no more inclined than
normal to tamp it down internally. The upshot, when there is a semi-defection by one security
partner followed by an expression of loyalty by the other, is that ideational conflict will tend to
emerge from within the jilted state, and will usually be worse than a “hypothetical” average.
326
325
Some foreign policy leaders are also key domestic players. Friendliness therefore makes these foreign-domestic
leaders less likely to exploit ideational conflict for their own domestic, political ends.
326
What I am calling the “hypothetical” average can theoretically be measured in those hypothetical moments when
neither the shared security milieu, nor the security relationship itself, exacts emotional impact on a nevertheless
extant alignment.
- 151 -
As noted in Diagram 3.4, ideological matters become politicized only at the transnational
level. While members of the FPE & PO who experience enmity from disappointed material
expectations do become more permissive of bilateral ideological conflict with hedging alignment
partners, it is important to note there is no security-related force to compel the initiation of
ideological conflict within the transnational level. In Neorealism, therefore, a given set of
alignment members must already suffer a propensity for ideas to become troublesome bilaterally
or multilaterally. If states in alignments lack a history of ideological discord, the absence of a
Diagram 3.4
The Intersection of Conflicting Security Interests and Ideological Politicization
in Neorealism: Jilted State Exhibits Loyalty
X suffers disappointed
security expectations.
States X & Y suffer conflicting security interests due to differing
perceptions of threats and/or opportunities in their shared security
milieu.
X feels enmity
toward Y.
X suffers
abandonment
fears.
Y hedges/semi-defects.
Y experiences satisfied
security expectation.
X becomes less likely to
help resolve ideological
problems, as they emerge
from the transnational level.
X exhibits security
loyalty toward Y.
Y becomes less likely to
aggravate ideological
problems, as they
emerge from the
transnational level.
Ideological issues are more
likely than average to become
problematic, with contestation
usually emerging from X.
- 152 -
problematizing mechanism in Neorealism means there is no reason to think security interest
conflict will correlate significantly and robustly with idea contestation.
327
It is worth summarizing a few of the many ways in which ongoing, interstate idea
conflicts can become politicized transnationally.
328
One is through domestic politics, as key
politicians at all levels of domestic society seek to improve or consolidate their political position
over rivals by capitalizing on volatile ideological matters vis-à-vis other states.
329
Similarly,
important bureaucratic players and institutional processes may deliberately or accidentally
politicize international issues.
330
Finally, actors such as nationalists, businesses, and
international organizations can intentionally or inadvertently enflame international ideological
conflict.
331
Diagram 3.5 shows how these processes coincide with enmity, after it is produced
following disappointed security expectations within an ongoing alignment.
332
327
This means, of course, Neorealism lacks a mechanism to explain how ideological conflicts emerge between states
in the first place.
328
Review my discussion of transnationalism in Neoliberalism in Chapter 2, as well as the discussions of complex
interdependence in Keohane and Nye, Power and Interdependence.
329
Waltz’s work on the ‘second level’ represents a classic statement on the impact of domestic politics on foreign
policy. Kenneth N Waltz, Man, the State, and War: A Theoretical Analysis (2001; repr., New York: Columbia
University Press, 1954). See also the political considerations of top leaders in Allison’s 3
rd
model, the bureaucratic
politics model. Allison, “Conceptual Models and the Cuban Missile Crisis,” 707-715.
330
See Allison’s 2
nd
model, the organizational process model, as well as his third model, the bureaucratic politics
model. Ibid., 698-715.
331
Keohane and Nye, Power and Interdependence, 20-32, 145-190.
332
Though the diagram shows enmity emerging in X, it could emerge in Y. But X’s enmity and Y’s friendliness
mean ideological problems are slightly more likely to appear within X.
- 153 -
Diagram 3.5 clarifies three main points regarding ideological conflict within Neorealism.
First, conflictual security policies do not cause or impel the politicization of ideas. Rather,
security interest divergence merely makes the FPE and PO less likely to tamp down transnational
quarrels over ideas. Second, if ideational conflict does not emerge at the transnational level,
ideas about immaterial things do not become politicized. In Neorealism, whenever a jilted state
Diagram 3.5
Transnational Processes and Ideological Problems in Neorealist Alignments
X experiences enmity toward Y in response to disappointed security expectations.
Ideological
conflict between
X and Y
emerges at the
transnational
level for reasons
completely
unrelated to
immediate
security
concerns.
X & Y do not
suffer ideological
conflict.
X & Y suffer
ideological
conflict.
The FPE & PO in X becomes less likely to help
resolve ideological problems.
Ideological
conflict between
X and Y does
not emerge at
the transnational
level for reasons
completely
unrelated to
immediate
security
concerns.
Transnational Causes of Ideological
Discord, Unrelated to Security Interests
Domestic Politics
Politicians seek to improve or consolidate
their political position over rivals, within and
without their respective coalitions by
capitalizing on volatile ideological matters
vis-à-vis other states.
Bureaucracies & Institutions
Bureaucratic competition or
institutional processes at the
domestic level politicize
international ideological issues.
Other Transnational Actors
Businesses, non-governmental organizations
(including nationalist groups), and
international organizations intentionally or
inadvertently enflame international
ideological conflict.
- 154 -
exhibits loyalty toward its hedging partner, intersocietal ideational relations remain unchanged
unless some phenomenon unrelated to security politicizes a new or once-contested idea. And
third, actors within the transnational level interact with each other within a given state. The
politicization of a particular idea can become more intense as it is absorbed and reacted upon by
transnational groups. Though not shown in the diagram, the politicization of an idea can spread
into related ideational issue areas through this process. Importantly, both states are connected
transnationally, a matter clarified in Diagram 3.6. Contestation over an idea can therefore
worsen and shift to somewhat similar matters as it moves back and forth between states as well.
Should the ideological contest result in an ongoing feedback cycle between the two states,
ideological relations between the two states stand a good chance of remaining poor until their
shared security milieu incentivizes betters relations.
Despite the fact that states operating under Neorealist logics necessarily suffer conscribed
emotionality when it comes to material security, groups and individuals within states enjoy broad
Diagram 3.6
Transnational Politicization of Ideas within and between Aligned States
State X State Y
Ideological
conflict
between X
and Y
emerges at the
transnational
level for
reasons
completely
unrelated to
immediate
security
concerns.
Information
Flows
between
States
Domestic Politics
Bureaucracies
& Institutions
Other Transnational
Actors
Domestic Politics
Bureaucracies
& Institutions
Other Transnational
Actors
- 155 -
emotional capabilities and can experience and understand many of the ideas of the real world.
Societal emotionality remains markedly delimited in two important ways, however: First, as
argued earlier in this section, societies in Neorealism cannot feel most emotions when it comes to
their material security. All emotions related to international material interests are sharply
delimited to the active emotions of fear and insecurity, and the desire to mitigate them, and the
passive emotions of enmity, ambivalence, and friendliness. Second, security interests remain
somewhat insulated from the effects of transnational ideational conflict and cooperation. While
transnational ideas can and do contribute to feelings of enmity, ambivalence, and friendliness,
and therefore provide states with an estimation of the amount of security cooperation or conflict
they can expect vis-à-vis a particular other, these passive emotions are but one input in the
security calculus. More important in calculating what the future holds for alignment members
are Snyder’s three incentives and the totality of the past behavior variable, as they impact
security expectations and the emotions of enmity, ambivalence, and friendliness.
Ultimately, these two caveats mean the transnational and security realms are not only
very different from each other but, by and large, are also remarkably non-interactive phenomena.
That this should be the case in Neorealism is not at all surprising though, as I emphasize below,
scholars utilizing Neorealism to examine the Japan-Korea case routinely make the atheoretical
assumption that societies have broad emotionality when it comes to international security matters,
even while states do not. Equally awkward theoretically is the frequent supposition that the
security and transnational realms are directly interactive.
The relationship between convergent security interests and the depoliticization of
ideological phenomena is as indirect as the process problematizing them. As shown in Diagram
3.7, depoliticization occurs due to the passive emotion of friendliness. These sentiments emerge
- 156 -
as alignment members’ individual expectations of security loyalty are satisfied by their joint,
cooperative responses to the security milieu’s challenge. As members of the FPE & PO
experience friendliness toward alignment members, they perceive an increase in the likelihood of
security cooperation and a decrease in the probability of security conflict within the alignment
moving forward. Consequently, they become less likely than normal to aggravate any ideological
problems that may emerge at the transnational level.
The deduction that complementary fears of the shared security milieu promote
cooperation within a Neorealist alignment is not surprising. Two processes remain unclear,
however. It is uncertain how emergent milieu fears lead to security collaboration within
alignments. And it remains unexplained why heightened fear of the security environment does
not lead more directly to better ideational relations among alignment members.
Diagram 3.7
Complementary Security Interest and Improved Ideological Relations in
Neorealist Alignments
X & Y enjoy aligned security interests because of shared fears of security milieu.
Milieu fears cause X & Y to exhibit security loyalty.
X & Y become less likely to aggravate ideological matters, as they emerge transnationally.
Ideological matters tend to become or remain depoliticized.
The FPE & PO in X & Y enjoy satisfied security expectations.
- 157 -
MATERIAL COOPERATION IN NEOREALIST ALIGNMENTS
There are at least three causal pathways leading to robust, material security cooperation
within alignments. All lead the normally hesitant, entrapment-fearing state to embrace its
alignment. First, a security threat can strengthen an alignment when it brings states fearing
entrapment by an alignment partner to come to fear abandonment by it instead. Such an
occasion occurs, for instance, when states begin fearing that an alignment partner will, without
strong support, bandwagon with a threatening state.
333
Should collaboration-inhibiting
entrapment fears suddenly become overwhelmed by bandwagon-abandonment fears, the
alignment will be much more likely to pull together.
334
The second route to cooperation occurs when the nature of entrapment fears change. As
argued persuasively by Yoon, the odds of cooperation rise greatly if entrapment-fearing states
come to worry about deterrence failure.
335
An alignment member realizes this variant of
entrapment fear when it comprehends that clashes involving its security partner are more likely
to occur from lack of deterrence of the enemy than because of too much support of the partner
within the alignment, and that new or rekindled conflict will invariably require its own
participation. States fearing deterrence failure have very strong incentive to avert conflict by
bolstering commitment to the alignment.
336
333
Walt argues that states are most likely to balance, rather than bandwagon. Nevertheless, he claims states do
bandwagon, especially when a given state 1) is much weaker than the most threatening state; 2) has no allies
available; 3) perceives that the most threatening state appears to be winning an extant war; and 4) perceives that
appeasing the most threatening state is a viable option. Walt, The Origins of Alliance, 28-32, 172-178.
334
Abandonment is the “most undesirable side effect” of hedging within an alignment. Snyder, Alliance Politics,
195.
335
See my discussion of Yoon in Chapter 1. See also Yoon, “Fragile Cooperation,” 88-90.
336
Yoon’s argument that Japan suffers deterrence failure-type, conflict-entrapment fears is not especially at odds
with Snyder’s. Snyder understands fully there is an “adversary game,” and that it occurs at the same time as the
“alliance game.” One of the “bads” of the adversary game is that withheld support within an alignment can
encourage the adversary to become “more intransigent and aggressive,” and consequently lead to conflict. What
makes Yoon’s argument unusual is his not-unreasonable assertion that Tokyo simply does not fear abandonment by
Seoul. Hence, from the Japanese perspective, Tokyo is involved in an adversary game with its opponents, but not in
an alliance game with Seoul. For Yoon, see Ibid., 88-90. For Synder, see Snyder, Alliance Politics, 192-198.
- 158 -
Third, the odds of cooperation between alignment members increase when milieu fears
render impotent the entrapment fear dynamic of Snyder’s security dilemma. Cha’s quasi-
alliance model is a good example of this mechanism at work. Recall one of the central
arguments of Cha’s model: Japan and Korea pull together when they both experience fear of
abandonment by the US, in order to reduce the costs of ongoing American involvement in
Asia.
337
Although somewhat complicated and obscured by Cha’s examination of Seoul’s and
Tokyo’s interests via the lens of their respective alliances with the US, Cha is essentially
contending that some interest external to the Japan-Korea security relationship more or less
removes fear of entrapment from the bilateral alignment. Importantly, Tokyo’s entrapment fears
regarding its ties with Seoul have not been replaced by fears of being abandoned by the ROK, as
occurs in the first scenario above. And the nature of Japan’s entrapment fears regarding Korea
have not changed, as they did in the second one. American retrenchment instead causes Tokyo
to worry about the direct threat to the home islands, and this insecurity overpowers Japanese
fears of entrapment on the Korean Peninsula.
338
Concern about survival, for its own sake, leads
Japan to cooperate with the ROK to lower the costs to the US of remaining involved in East Asia.
WHY NEOREALIST FEAR DOES NOT IMPROVE IDEATIONAL RELATIONS DIRECTLY
The fact that pressing security threats can promote active cooperation in the material
realm will suggest to many scholars that imperiled sovereignty should also produce ideological
cooperation, and that efforts to improve immaterial relations will be vigorous and deliberate.
After all, common sense informs that when survival is at stake all phenomena with potential to
337
See my discussion of Cha in Chapter 1. See also Cha, Alignment Despite Antagonism, 50-57.
338
Cha understands that Japan also fears deterrence failure—i.e., it suffers an entrapment fear—regarding the
Korean Peninsula whenever the US retrenches. Nevertheless, even at these moments it is fair to classify Cha’s work
as being external to the Japan-Korea alignment security dilemma, because Cha’s emphasis is clearly on Tokyo’s
insecurity regarding the Japanese mainland itself, not on getting entrapped in conflict on the peninsula. Ibid., esp.
56.
- 159 -
prevent necessary security cooperation must be resolved or at least significantly diminished. Cha
argues along these lines when he claims Tokyo’s and Seoul’s fears compel leaders in both
capitals to strive intentionally to ameliorate the history issue. Yoon contends similarly that
Japanese fears of deterrence failure and Korean fears of abandonment have the same effect.
Koo’s trade dependence model makes similar claims about urgent economic imperative
producing deliberate efforts to improve the Dokdo-Takeshima dispute.
339
Unfortunately,
Neorealist and Neorealist-type claims of this type are certain to be woefully incomplete,
unsatisfactory, and/or atheoretical.
The biggest problem, as briefly noted in earlier discussion of Cha’s model, is the paradox
contained within the very thought of Neorealism’s material imperatives incentivizing ideological
cooperation among rational actors. This material-ideational contradiction emerges from “real”
Neorealism’s postulate that ideas are not powerful enough to impinge upon the calculations of
states as they estimate their security interests.
340
Given that proposition, how can ideological
conflict prevent robust security collaboration? After all, if disagreements over ideas can prevent
adequate security response to the point that states feel they must actively ameliorate ideological
issues, then completely immaterial phenomena are obviously impacting material security
interests powerfully. Remarkably, the causal emotions at work are fear and insecurity, as states
come realize that ideas have the power to prevent adequate security response. If ideas can
induce fear and insecurity regarding material capability so compellingly, however, then
Neorealism has failed terminally.
341
Conversely, if ideas cannot impact material cooperation,
339
Recall, especially, Koo’s argument about the 1997 financial crisis. See Koo, “Scramble for the Rocks,” 127-129.
340
Legro and Moravcsik, “Is Anybody Still a Realist,” esp. 18.
341
If ideas shape security interests, then Neorealism’s contribution to scholarship cannot extend beyond unnecessary
reminders, in the present-day, that states care about power. Nevertheless, Realists and Neorealists played an
obviously important and necessary role in emphasizing the role of power in the past. And they may yet again. For
the classic manifestation of this phenomenon, see E.H. Carr, The Twenty Year ’s Crisis, 1919-1939: An Introduction
to the Study of International Relations (1981; repr., New York: Palgrave, 2001).
- 160 -
then Cha’s, Yoon’s, and Koo’s rational actors have no security-related incentive to limit
ideational conflict.
342
This inherent incongruity is precisely the reason why my delimitation of
Walt’s Neorealism introduces an FPE & PO incapable of fearing ideological phenomena directly
when considering material security interests. Were the delimited balance of threat model to
acknowledge ideas as being equal in potency to material interests, I could no longer claim my
first model is Neorealist. Importantly, if my model were not Neorealist, I could no longer make
sweeping claims ontologically delimiting active emotionality to fear and insecurity.
342
Koo’s might at the domestic/transnational level, because he is a Neoliberal. By failing to show these processes at
work adequately, however, he is writing as if he were a Neorealist.
- 161 -
Chapter 4
Theoretical Model #2: The Common Sense Social Emotions Model
I. Overview: Putting Social Emotions before Material Emotions
In building my first theoretical model in Chapter 3, I attempted to take Neorealism
seriously. Despite my best efforts to stay within the metatheory’s confines while still having
something meaningful to say about Japan-Korea alignment politics, I invariably made
theoretically incoherent claims. I detail these inconsistencies below, giving particular attention
to atheoretically utilized social emotions. I argue they imply strongly that Neorealism represents
a remarkably poor starting point for modeling alignment politics. In Chapter 5, after examining
the theoretical implications of dynamic expectations on stable actor preferences, I argue
Neoliberalism in all its guises is similarly insufficient for alignment politics modeling.
After evidencing some of the ways in which Neorealism is inadequate in my delimited
balance of threat model, as well as in the work of Cha, Yoon, and Lind and, to some extent, Koo,
I build a second theoretical model. As will be readily apparent, it is most remarkable for putting
ideational social emotions at the heart of alignment politics. In other words, I essentially
abandon Neorealism while developing my second model. Importantly, I do not seriously attempt
to ground my work in this chapter within another coherent IR metatheory. Instead I take what I
consider to be a “common sense” approach to choose the social emotions that matter.
Consequently, I refer to my second model as the common sense social emotions model. While
the decision to work without an explicit theoretical framework may puzzle some scholars and
feel incorrect, my strategy is primarily differentiated from the work of the models I examine
- 162 -
most carefully only in the fact that I make my departure from metatheory deliberately and
professedly, not for the fact of the departure itself.
Leaving formal theory is moreover somewhat advantageous, in that a more or less
atheoretical accounting of social emotions generates a causal process remarkably different from
the group dynamics model I produce in my third theoretical model in Chapter 5. To the extent
my group dynamics deductions and process are convincing, the weaker the “common sense”
arguments of my second model become. In the event my group dynamics work is unconvincing,
however, then the work in the current or prior chapters may prove a better starting point for
additional scholarship.
Before exploring the theoretical incoherence of the use of social emotions by the material
models, it is worth outlining the most basic causal process of the common sense social emotions
model. As shown in Diagram 4.1, the starting point of inter-alignment conflict is remarkable for
being the same as that of my first theoretical model. Namely, conflicting milieu security
interests lead one alignment partner to hedge its commitments to, or semi-defect from, the
alignment.
What changes after X’s alignment expectations are disappointed is X’s emotional
reaction. Instead of feeling fear of abandonment and the passive emotion of enmity, X primarily
experiences the active emotions of anger and vengefulness. These emotions cause the FPE & PO
in X to lash out at Y deliberately. Of course various constraints exist: X’s material
interdependence with Y might prevent it from attacking Y in the economic or security realms, for
example. Similarly, as hinted in the title of the diagram, conflictual outcomes are likely only
when X’s abandonment fear is below an emergency threshold. If material insecurities are so
high as to constitute an existential crisis, X will exhibit material loyalty toward Y. Under certain
- 163 -
conditions which I outline later, existential crises also lead to exhibitions of ideological loyalty.
Importantly, the existential crisis threshold is seldomly met or crossed within alignments in the
modern world. In practically all cases, therefore, immaterial social emotions determine conflict
in alignment politics.
As outlined in Diagram 4.2, the cooperative mechanism in the common sense social
emotions model is a little less straightforward. Convergent milieu security interests lead both
alignment members to honor their obligations to each other. Both X and Y consequently
experience satisfied security expectations, a feeling producing the fairly passive emotion of
gratitude. Gratitude in turn leads the FPE & PO in both states to become less likely to aggravate
material and ideological problems as they emerge from the transnational level. Importantly, the
alignment partners’ coinciding material fear and insecurity can promote deliberate ideological
Diagram 4.1
Security Interests Trigger Alignment Conflict via Ideological Social Emotions
in a Non-Emergency Material Security Environment
Material and/or Ideological Conflict is/are deliberately instigated by X,
depending on its capabilities and interests.
X suffers disappointed security expectations.
States X & Y suffer conflicting security interests due to differing perceptions of
threats and/or opportunities in their shared security milieu.
The FPE & PO in X feels primarily
angry and vengeful toward Y.
Y hedges/semi-defects.
- 164 -
cooperation, in addition to the active material cooperation noted above. The FPE & PO does so,
however, only to the extent ideological issues are perceived to threaten essential material
cooperation. Thus, states can fear ideas in the common sense social emotions model.
Remarkably, true security emergencies in which all alignment members feel existential material
fear and insecurity never produce deliberate ideological collaboration.
343
They do not
incentivize it, because conflictual ideational phenomena are simply not powerful enough to
prevent material cooperation among aligned states in extreme security scenarios. Shared, severe
crises consequently convey on their own zero incentive for states to handle ideational issues
actively.
343
Passive ideological cooperation can occur, primarily because gratitude will be experienced as security
expectations are satisfied.
- 165 -
Passive versus active emotions represent the key explanatory difference in my first two
theoretical models. This divergence is critical, as it often results in two very different
relationships between the generic IV of material security interests and the generic DV of
ideational phenomena. The delimited balance of threat model accommodates only an indirect
relationship between the two, implying the history issue will not be immediately and powerfully
affected by changes in milieu interests. The common sense social emotions model, in contrast,
suggests there will be a remarkably close relationship between the two variables, as changes in
the IV will often lead to deliberate action on the DV. Purposeful action will nearly always be at
Diagram 4.2
Security Interests Trigger Alignment Cooperation via Ideological Social
Emotions in a Non-Emergency Material Environment
X & Y feel gratitude toward each other.
X & Y experience satisfied security
expectations.
States X & Y enjoy convergent security interests due to similar perceptions of
threats and/or opportunities in their shared security milieu.
X & Y honor their material alignment obligations.
If the FPE & POs in both X & Y fear ideological
problems will prevent material cooperation, they
actively strive to ameliorate and prevent ideological
conflict.
The FPE & PO in X & Y become less likely to aggravate
ideological problems as they emerge from the transnational level.
Ideological problems resolve and/or become unproblematic. The pace of change depends on whether
ideological problems are perceived to threaten essential material cooperation.
- 166 -
the root of conflict, and will similarly mark cooperation whenever the FPE & PO fear that
ideological conflict can upend the desired material cooperation.
The most critical similarity linking both models, and indeed connecting all the Japan-
Korea models examined up to this point, is their inclusion of security expectations. None of
these models bother to explain why expectations are present. Instead, each treats them as given,
even though the logics of the theories and/or models usually suggest such expectations should be
absent or, at most, very weak. This fact represents a major weakness in all of the models
considered through the current chapter, and is a matter I address in Chapter 5 while constructing
my third theoretical model.
II. Social Emotions in Material Theories and Models
None of the Japan-Korea models examined most carefully in this paper are true to “real”
Neorealism. The point is fairly obvious, as each work of scholarship utilizes social emotions that
are simply not Neorealist. Often less apparent, the models use these non-Neorealist emotions to
explain powerful, causal linkages connecting the material and ideational realms. Doing so
further distances the respective work of Cha, Yoon, Lind, and Koo from its ostensible material
underpinning.
In this section I summarize how far each scholar has diverged from the emotional and
causal constraints of Neorealism. The evaluation is merited, because the four scholars make bold
claims about the emotions that matter—about emotional ontology. Yet, in particular but critical
instances, they base their assumptions about the affect in play on nothing more than the fact their
research is grounded within material IR theories. Revealing the extent to which each model is
non-materialist consequently evidences that these ontological assumptions are without merit. It
also makes apparent the equal possibility that non-Neorealist emotions are just as likely to be
- 167 -
both legitimate and causal as the material ones the scholars presuppose into their arguments.
Though I believe I can rightfully claim my delimited balance of threat model is “more Neorealist”
than the others, it too suffers shortcomings. With the ideal of real Neorealism in mind, I
compare and contrast my first theoretical model with the other four material models of Japan-
Korea relations.
A. Neorealist and Non-Neorealist Emotions
Scholars utilizing Neorealism to model alignment politics face substantial constraints on
the emotions to which they may accede. They are limited because emotions other than fear and
insecurity, and the desire to mitigate them, end up irreparably damaging Neorealism’s defining
focus on material interests as made significant within anarchy by ubiquitous and unchanging
expectations about the meaning of power.
344
While Neorealism could not exist without any one
of its three components–namely anarchy, expectations about power, and the emotions of fear and
insecurity—shared and unchanging expectations about the meaning of power form the obvious
core of the theory. For if power ever comes to mean different things to those considering it,
there is no reason why some, and perhaps even all, states cannot come to feel unafraid and secure
no matter the dispersion of power and existence of anarchy in the international system.
Neorealism, in short, would cease to be, except in those spaces and times in which states
experience the Neorealist meaning of power. Such a situation would mark the transformation of
Neorealism into a Wendt-type project in which the scholar’s primary task is to uncover the
meaning of power within the international system, or at least among a particular grouping of
states.
345
344
See my discussion of Neorealism in Chapter 2. For another take on three assumptions of “real” Neorealism, see
Legro and Moravcsik, “Is Anybody Still a Realist,” 12-18.
345
This is essentially the point Wendt makes when he differentiates the meaning of power by categorizing it
according to Hobbesian, Lockean, and Kantian norms. See Wendt, Social Theory of International Politics.
- 168 -
The requirement of holding constant expectations about the meaning of power is the
wellspring of Neorealism’s delimitation of emotionality to materially-centered fear and
insecurity. Neorealists are constrained in their consideration of emotionality because many
ideological phenomena end up transforming expectations about the meaning of power.
Regarding emotions, the constraint on Neorealism is two-fold. First, Neorealist states cannot
experience fear or insecurity from ideas; only material things can trigger these emotions.
Scholars adhering to Neorealism must never ascribe to ideational phenomena the power to instill
fear and insecurity, because states capable of categorizing a given idea as “bad” for their security
will necessarily also have the ability of considering other ideas as being “not bad.” As a given
state comes to view the power of the states around it via the lens of the ideas they are perceived
to follow, its expectations about the meaning of material power change accordingly. Having lost
the expectational leg of its tripodal underpinning, real Neorealism quickly degrades.
The second constraint on the use of emotions in Neorealism is simply a complete
prohibition of all emotions that change the meaning of power to something other than the
meaning of power produced by fear and insecurity under anarchy. This requirement applies
equally to emotions scholars might otherwise apply directly as independent variables, as well as
to affects produced at the dependent variable level if, by implication of theoretical coherence,
there are follow-up consequences for the meaning of power.
Consider the ramifications on his or her theory of a Neorealist scholar finding that anger
can lead to conflict. There are multiple ways to explain why the researcher’s argument is
problematic for Neorealism. One is to note that anger cannot exist outside of society. Indeed,
one must be angry at someone, or at least at a temporarily mentally-anthropomorphized
- 169 -
something, for having carried out an “offense” of some kind.
346
But an Other cannot commit an
anger-triggering offense unless there exist norms to be broken. Importantly, since only societies
can create norms, a state experiencing anger necessarily views the offender as being both part of
its international society, and as having broken one or more of the society’s norms. In short, by
concluding that anger produces conflict, the Neorealist researcher is inadvertently arguing for the
existence and importance of an international system of norms, a phenomenon which would
influence the meaning of power in anarchy.
Another route to understanding why Neorealists need to remain highly cautious about
utilizing emotions is to explore the fact that many emotions cannot be experienced by those
operating in an emotional vacuum. In other words, emotional prerequisites must often be
satisfied before a given actor can be declared capable of feeling a particular affect. Humiliation,
for example, is tightly linked to “perceived unfairness”—a sense that the humiliation is not
deserved.
347
In this regard it is practically identical to anger, which requires its experient to
make a “judgment that the frustration, interruption, power reversal, or harm is illegitimate—that
the situation is contrary to what ought to be.”
348
Importantly, realizing and acting upon these
“ought violations”
349
implies the existence of multiple, related emotions. Before an actor can
pass judgment, for instance, s/he must be able to recognize justice and injustice, ideological
phenomena which are themselves “a complex set of passions” including “compassion and caring,”
346
Solomon, A Passion for Justice, 253, 255-256. One can also be mad at oneself, though this does not have any
obvious implications for conflict with others.
347
See discussion of research in Bernhard Leidner, Hammad Sheikh, and Jeremy Ginges, “Affective Dimensions of
Intergroup Humiliation,” Plos One 7, no. 9 (September 2012): 2.
348
Phillip Shaver et al., “Emotion Knowledge: Further Exploration of a Prototype Approach,” Journal of
Personality and Social Psychology 52, No. 6 (1987): 1077.
349
Michael F Mascolo and Sharon Griffin, “Alternative Trajectories in the Development of Anger,” in What
Develops in Emotional Development? (New York: Plenum, 1998), 219-249.
- 170 -
as well as “envy, jealousy, indignation, anger, and resentment.”
350
In other words, when a
scholar makes even the quickest claim about the importance of an emotion like anger or
humiliation, s/he is also deliberately or accidentally accepting the equal significance of justice
and injustice, and the emotions and ideas associated with them. Even implicit acknowledgment
that states can perceive justice and injustice is of course highly damaging to Neorealism, because
those sensitive to these ideological phenomena will shape their policies according to their
perception and interpretation of an objective right and wrong. Since an internationally shared
sense of right and wrong would ultimately help shape the meaning of power, wherever justice
exists the scholar’s task is to uncover what it means within its society of adherents, and the
extent to which it is upheld in interstate relations.
351
The existence of ideological phenomena like anger and humiliation, and justice and
injustice has other follow-on implications. A state cannot punish its transgressors to assuage its
anger or humiliation, for example, if it cannot also experience an emotion like vengefulness. But
vengeance is not “endless and wholly destructive.”
352
Predicated on perceived injustice,
vengefulness calls for retaliatory acts to cancel out the injustice, not to completely destroy the
offending party. While an act of vengeance may of course get out of hand and become much
more destructive than the injustice it seeks to balance, the experient’s purpose of restoring justice
is remarkable. For vengefulness’s “conception of its own satisfaction and limits” means states
seeking vengeance are bound by the norms in their minds.
353
These norms challenge Neorealism
because, if even vengeful states are constrained by conceptions of international justice, then
350
The quotes are from Solomon, A Passion for Justice, 243. See also Ibid., 242-296. Intriguingly, if the anger
realized is a product of viewing an Other’s experience, then the angry party has “take[n] the experience
personally”—i.e., it has “identif[ied] in some sense with the offended person or party.” Neorealism of course cannot
abide this kind of “we-feeling.” Ibid., 253.
351
To foreshadow a point made below in my discussion of expectations, the strength of social emotions felt by a
state should correlate positively with its perception of the strength of the international system of norms.
352
Ibid., 44.
353
Quote is from Ibid., 44. See also Ibid., 256-258, 272-74.
- 171 -
surely states responding to changes in the balance of power are as well. The emotion of
vengeance therefore has broad implications regarding the severity of anarchy, as it means that
societal and international norms related to justice and injustice conscribe the use of power to
“reasonable” levels. Moreover, to the extent states recognize their neighbors as being vengeful
and not vengeful, the affect attributes different meanings to material power in a second sense.
If it is true that materially oriented theorists and modelers must be very careful when
considering emotions like anger, humiliation, and vengefulness, they should also exercise great
caution when taking into account emotions related to nationalism. For it is obvious that
nationalism will too often be incapable of inducing states to action if they experience neither the
aforementioned emotions, nor affects similar to them. Nationalists, for example, will never start
a war to recover lost national honor if they cannot feel some variant of one or all of the emotions
discussed above. If these kinds of emotions are a precondition for the activation of nationalism
as a causal factor, however, then Neorealists must either build their theories and models without
nationalism, or exercise great prudence when acknowledging it. Those Neorealist theorists and
modelers who include nationalism haphazardly might very well be undermining the coherence of
the theory with which they are working.
There is another reason why Neorealists must be careful regarding nationalism. The
problem is that nationalism is, at its very root, a social phenomenon. It is closely tied to a kind
of national culture—i.e., “a shared set of [national] symbols and memories” that helps demarcate
an Us and the Them, and comprises an important part of the historical, national narrative.
354
Distinguishing between Us and Them would not be problematic for Neorealism, if every
nationalistic state perceived the meaning of every other states’ power equally. But nationalist
354
Barry R Posen, “Nationalism, the Mass Army, and Military Power,” International Security 18, no. 2 (Autumn
1993): 85.
- 172 -
narratives are too often set against particular Others.
355
Nationalism, in short, is frequently a
power-differentiating social force. Consequently, while it may explain how modern states
developed mass armies and the kind of military power about which Neorealists must theorize,
356
nationalism also often undercuts Neorealism’s core assumption about a static and ubiquitous
meaning of power.
B. Real Neorealism and Social Emotions: The Example of Neorealist Anger
It is worth summarizing two ways in which emotions like anger, humiliation,
vengefulness, and the emotions associated with nationalism differ from materially-derived fear
and insecurity. First, as laid out above, the prior set of emotions suggests the existence of an
international society in which norms exist and are important. Material fear and insecurity, in
contrast, arguably connote nothing about societal norms beyond Neorealism’s shared
expectations about the meaning of power and the existence of anarchy. This tremendous
variance is not surprising, as emotions related to nationalism, as well as anger, humiliation, and
vengefulness are social emotions—i.e., they are triggered via one’s active participation in, and
experiences within, a society.
357
Fear and insecurity are comparatively asocial: Although
humans can certainly feel them in social settings, society is not required.
358
People need not
anthropomorphize the bear that is attacking them, the water in which they are drowning, or the
material power another state wields in order to be made fearful or insecure about them. Likewise,
355
See, for example, Gretchen Schrock-Jacobson, “The Violent Consequences of the Nation: Nationalism and the
Initiation of Interstate War,” Journal of Conflict Resolution 56, no. 5 (2012): 825-852; and Stephen van Evera,
“Hypotheses on Nationalism and War,” International Security 18, no. 4 (Spring, 1994): 5-39.
356
Posen “Nationalism, the Mass Army, and Military Power.”
357
While scholars such as Turner and Stets include anger on their short list of hard-wired emotions, Hareli and
Parkinson’s review of the literature on emotions finds that all the work on anger they examined included a
“consideration of other’s involvement or agency.” In contrast, none of the research on fear examined the role of
others. See Schlomo Harelli and Brian Parkinson, “What’s Social about Social Emotions,” Journal for the Theory
of Social Behavior 38, no. 2 (2008): 133; Jonathan H Turner and Jan E Stets, “Sociological Theories of Human
Emotions,” Annual Review of Sociology 32 (2006): 46. See also, Brian Parkinson, “Emotions are Social,” British
Journal of Psychology 87 (1996): 663-683.
358
Harelli and Parkinson, “What’s Social about Social Emotions,” 137-38,140, 147.
- 173 -
while they might not like the fact they are being eaten, are drowning, or are being threatened
militarily, the fear and insecurity people experience at these moments does not require the
perception of an ought violation. Sensitively to questions of justice and injustice, and ideological
forces related to them, are therefore not critical in these and similar instances.
The comparably asocial nature of fear and insecurity hints at the second way in which the
two sets of emotions differ. Simply stated, fear and insecurity seem more hard-wired into the
brain than the more social emotions discussed above.
359
The comparatively more primordial
nature of fear and insecurity is no doubt why these two emotions work so well in Neorealism. If
fear and insecurity were not so easily felt, or especially if they required societal norms to be
experienced, then Neorealism would have failed immediately and outright regarding its claims
about the existence of an anarchy and a meaning of power that are quintessentially Hobbesian, or
pre-social. The theory would have collapsed, because a fear and an insecurity requiring social
norms would have implied the existence of some kind of social order and, therefore, a non-
Neorealist meaning of power.
If Neorealism topples whenever fear and insecurity require social context, then something
similar surely occurs when Neorealists attempt to explain outcomes by reaching for emotions
like anger, humiliation, and vengefulness, or calling upon affects related to nationalism. It is
therefore worth contemplating the Neorealist case for including these emotions. Doing so helps
clarify what social emotions ought to look like when considered within Neorealism, and provides
a reasonable comparison point for viewing how far a particular Neorealist or Neorealist-type
scholar has strayed from a more or less objective Neorealist standard.
359
Anger is also a primordial, or “basic,” emotion. But it differs from fear and insecurity in that it requires a social
appraisal. Ibid., 137-38,140, 144, 147.
- 174 -
As argued above, in Neorealism emotions cannot transform the meaning of power.
Should they influence power in this way, the scholar’s primary task necessarily morphs into one
in which the meaning of power must first be considered before Neorealist principles can be
assumed dominant and applied across cases. Consequently, emotions must not imply the
existence of an effectual international society. Similarly, emotions that do determine material
interests must be “material” themselves, just as fear and insecurity are material.
Rather than present Neorealism’s case for every non-Neorealist emotion discussed above,
it makes sense in terms of streamlining and opportunity for depth to illustrate just one: Anger.
Anger is arguably the best third emotion for incorporation in Neorealism, because it is clearly a
“basic” emotion; like fear and insecurity, it is primordial.
Anger is nevertheless very different from fear and security. I have already discussed how
anger—unlike fear and insecurity—requires its experient to perceive an ought violation and,
therefore, that those who are angry must also be emotionally and normatively capable of
distinguishing between justice and injustice. Moreover, I explained that people feeling angry
must also experience an emotion like vengefulness if they are to carry out aggressive
counteractions in response to their anger. Since Neorealism is very unlikely to survive the
ontological challenge represented by this kind of social depiction of anger, to make the case for
Neorealism I therefore need to argue that anger can also work more like material fear and
insecurity. Namely, I must claim it can be felt without the presence of an ought violation and
relieved without need of an additional action-permitting emotion.
By my reasoning, including anger in Neorealism requires six deductions. First, a given
state will anger only when its material capability is harmed or reduced by other states.
Importantly, it does not matter whether the actions deleteriously impacting its power are
- 175 -
deliberate, because the angry state cannot distinguish the intentions of states as they strive to
consolidate or gain material power. This inability also mandates that states cannot anger too far
in advance of being harmed materially. Second, just as Neorealist states desire those things
which improve their material security, they must also possess a natural desire to assuage their
anger. Like fear and insecurity, it must pain them in some way. Framing anger as something
which states strive to eliminate obviates most of the problematic implications of arguing they
require some other emotion, like vengeance, with attendant and theory-complicating
requirements for a sense of justice and ought violations.
Third, Neorealist anger can only be assuaged via action if the response produces one or
both of two outcomes. It must reduce the material power of the party responsible for the
experient’s initial loss of power, and/or lead the responsible party to replace the material
capability it harmed or otherwise took away from the angry state. It is remarkably different from
fear and insecurity in this regard. For, while states can often mitigate the fear and insecurity they
suffer in response to losing material power by reacquiring it from anywhere,
360
states must target
the actual source of their anger if they wish to mitigate their emotionality aggressively. Fourth,
unlike Neorealist fear and insecurity, anger dissipates over time. States therefore remain angry at
their transgressors over the long-term only if the latter harms or reduces the former’s material
power with adequate regularity.
Fifth, states will not enact policies to reduce their anger, if doing so runs counter to their
material interests. In other words, states would rather be angry while enjoying good material
prospects than be anger-assuaged, yet materially more insecure. Sixth and finally, states are just
as bad at determining others’ level of anger than they are at measuring others’ fear and insecurity.
360
This might not be the case in some bi-polar systems, however.
- 176 -
This point is important because—at least in anger-inclusive Defensive Neorealism
361
—the ability
to estimate others’ fear, insecurity, and anger reasonably well implies states would categorize
each other as being satisfied or dissatisfied emotionally. Since emotional dissatisfaction is
ultimately what drives states to accrue power in Defensive Neorealism, states enjoying the ability
to measure accurately the emotionality of others would come to impart new meaning to power, in
the form of categorizing states as being satisfied or dissatisfied with their level of power.
The six deductions produce a unique type of theoretical anger. Ontologically, this
Neorealist anger becomes a “material” emotion, like fear and insecurity. Just as states in
Neorealism cannot experience fear and insecurity from ideas, anger can no longer flow from
ideas. Just as fears and insecurities cannot produce or impel ideas that in turn impact the
meaning of power, Neorealist anger can no longer bring about or promote ideational phenomena
if, in their follow-on effects, they come to influence the meaning of power. Intriguingly, states
will only act on their anger if doing so does not counter their material interests regarding the
balance of power. With one notable exception, therefore, anger-targeted states will suffer
counter aggression or enjoy being ignored, no matter the existence or intensity of the experient’s
anger. The only occasion in which Neorealist anger plays a distinct role occurs when balancing
interests call for a still-angry state to accrue more power immediately, but distinguish neither the
desirability of targets nor the best method of accreting power. Under these conditions, states that
are the source of anger will fall victim to it.
Tightly bound to materialism, Neorealist anger is radically different from the kind of
anger and other non-material, social emotions implicitly or explicitly used by the four Japan-
Korea scholars examined most carefully in this study. The contrast is remarkable since all four
361
As noted in Chapter 2’s discussion, Offensive Neorealism differs from its defensive cousin in that states are
never satisfied with their level of power. In Offensive Neorealism, therefore, hegemony-striving states are always
more or less equally fearful.
- 177 -
of these scholars ground their alignment models in Neorealism or its material cousin,
Neoliberalism. Admittedly, the Neorealist anger I deduce also differs from my use of the social
emotions of enmity, ambivalence, and friendship in the delimited balance of threat model—i.e.,
in the model I placed as firmly as possible in Neorealism. My failure to create a satisfactory
material model—one completely bereft of social emotions—is the major reason I move far
beyond Neorealism in this project.
C. Social Emotions in the Five Material Models
Neorealist Anger provides a useful measurement point to help gauge the Neorealism of
social emotions present in a material given model. Here I focus my discussion on particular
social emotions utilized in the four models of Japan-Korea relations, as well as in my own
delimited balance of threat model. As noted above, the purpose of this exercise is to point out
the extent to which a given model is not compatible with Neorealism or material Neoliberalism,
a fact which greatly weakens any argument in which material emotions are presupposed without
thorough and convincing rationale, and which consequently equalizes my common sense use of
emotions in my second model.
CHA’S QUASI ALLIANCE MODEL
Cha’s quasi-alliance model seemingly presupposes from Neorealism that fear and
insecurity are the only active emotions influencing the history issue and other forms of
cooperation and conflict. I say “seemingly,” because Cha attempts to distance himself from
“Realism,” by arguing it cannot explain the “reality” of Japan-Korea relations.
362
Nevertheless,
in terms of theoretical compatibility Cha clarifies that his divergence from Stephen Walt’s
Defensive Neorealism is primarily marked by Walt’s sole focus on the perception of the enemy’s
362
Cha, Alignment Despite Antagonism, esp. 17-19.
- 178 -
threat, versus Cha’s equal inclusion of “promises” of the “ally.”
363
Remarkably, both scholars
emphasize as active emotions only fear and insecurity, even though they are deeply dependent
implicitly on a whole panoply of unrecognized emotions to activate fear and insecurity.
364
While Cha’s research is emotionally compatible with Walt’s, the quasi-alliance model’s
main inspiration is obviously Glen Snyder’s work on alliance politics. As I emphasize below,
Snyder is primarily a Neorealist, despite his occasional but not-inconsiderable divergence from
real Neorealism. What makes Cha’s model Neorealist, is the fact that practically all of its
contribution to scholarship is grounded in Snyder’s abandonment and entrapment mechanism,
which happens to be the portion of Snyder’s alliance politics research most true to real
Neorealism. More specifically, as I show below, Snyder’s discussion of abandonment and
entrapment is the part of his work that strips out practically all of the non-Neorealist phenomena
that Snyder knows, and not-infrequently argues, is critical. What Cha produces is consequently,
at least in ostensible terms, very Neorealist.
Despite his Neorealism, the emotions Cha implicitly utilizes diverge greatly from the
standard established by Neorealist anger. As argued in Chapter 2, the most substantial
incompatibility emerges in Cha’s necessary reliance on one of two phenomena: He either
requires the Japanese to experience material fear and insecurity from the history issue, or he is
363
Ibid., 48-49.
364
I discussed this point regarding Cha in Chapter 2, and I return to it briefly in this section. The implicit use of
active emotions other than fear and insecurity is sometimes harder to see in Walt’s theory, because of his theoretical
impreciseness and his method of writing. For example, Walt notes that much of the West treated the USSR with
disdain until Germany became powerful in the run-up to WWII. While that position seems entirely true to real
Neorealism, surely balance of threat theory acknowledges Germany’s non-Neorealist emotional dissatisfaction—its
obvious nationalism, anger, and vengefulness, for example—as a key reason why Britain, France, and the US came
to fear Germany’s growing power in the first place. Thus, Walt has presented something grounded inherently in
non-Neorealist active emotions as being completely Neorealist. Importantly, Walt is occasionally blatant in his use
of non-Neorealist, active emotions. For instance, weak regimes pursue “popular” foreign policy goals in order to
build their legitimacy. This fact means foreign policy goals appealing to mass emotionality—i.e., making the
masses proud, happy, honored, etc.—actively shape international politics. Walt, The Origins of Alliances, 37-39.
- 179 -
reliant on the existence of a compelling non-Neorealist emotionality component.
365
Both are
terribly incompatible with Neorealist anger, because Neorealist anger cannot be experienced
from ideas; and neither does it require emotions related to nationalism, nor embarrassment, anger,
and vengefulness.
Cha’s cooperative mechanism similarly drifts too far from Neorealism, when it calls for
Japan and Korea to collaborate deliberately on the history issue in order to entice the US by
lowering the costs of involvement in East Asia. By making this argument Cha has fallen into the
material-ideational paradox, wherein Seoul and Tokyo must fear that troubles over an idea—the
history issue—can influence US material interests, even while the model’s primary causal
mechanisms are deduced from Neorealism. As hinted in this criticism, the quasi-alliance model
also suffers from the direct linkage between the material and ideational realms. For, just as
Neorealist anger can neither be triggered nor satiated by ideas, Neorealist fear and insecurity
cannot directly impact, or be directly influenced by, something as immaterial as the history issue.
Remarkably, without the above non-Neorealist qualities, the quasi-alliance model ceases
to function. In particular, it no longer explains Seoul-Tokyo conflict when the US is eager to
participate in East Asia. And it likewise cannot explain Japan-Korea cooperation intended to
entice American involvement. Importantly, if Cha utilizes such powerful non-Neorealist
phenomena to account for Japan-Korea relations, what basis is there for scholars accept his
deduction from prior Neorealist research that fear and insecurity of material things lies at the
heart of ideological cooperation and friction among the quasi-alliances?
YOON’S NET THREAT THEORY
Yoon’s net threat theory is also vulnerable to this line of inquiry. Yoon’s most
fundamental departure from Neorealism arguably occurs when he contends Neorealist interests
365
Likewise, Cha requires the Koreans to rationalize material-like efficacy for the history issue.
- 180 -
incentivize deliberate cooperation on the history issue. Of course this represents an instantiation
of the material-ideational paradox, since material interests cannot produce deliberate action
regarding ideas, if ideas cannot influence material interests. The analysis accompanying
Neorealist anger similarly suggests this kind of direct connection between the material and
ideational realm is impossible in real Neorealism.
As noted in earlier discussion, Yoon’s work is most interesting for its assertion that
betrayal and enmity—both blatantly non-Neorealist emotions—induce conflictual relations. But
if net threat theory diverges so sharply from real Neorealism, perhaps it is best just to note that
Yoon’s model is simply not Neorealist. If not Neorealist, there still remains plenty of room to
critique it on its own terms.
Though such an approach might be reasonable, net threat theory does merit being
analyzed as a Neorealist model for two reasons. First, real Neorealism appears to be the only
reason why the emotions that matter in Yoon’s cooperative mechanism are limited to fear and
insecurity. Having deduced this ontology from Walt’s and Snyder’s theories, Yoon offers no
discussion about why non-Neorealist emotions produced by satisfied expectations of loyalty are
not extant and more compelling. In fact, there is no evidence to suggest he considers any other
possibility, even though he obviously believes non-Neorealist emotions like betrayal and enmity
are more important than fear and insecurity in other parts of net threat theory. It is this
apparently automatic return to Neorealist emotions that marks net threat theory as being
Neorealist, despite its inclusion of ideological affects like betrayal and enmity, as well as the
material-ideational paradox.
Second, though betrayal and enmity are far from being Neorealist, Yoon utilizes them in
a way not necessarily incompatible with the metatheory. In particular, betrayal and enmity exert
- 181 -
no direct impact on ideational phenomena like the history issue. Instead, enmity’s influence is
indirect: It basically deprives leaders in Seoul of their incentive to ameliorate problems about
the past as they emerge for ideational reasons. Importantly, it does not compel the leadership to
strive to make the history issue worse. Though the argument is not perfectly logical—namely
Korea should still be expressing loyalty to Japan in material terms and, regarding other parts of
Yoon’s work, it remains doubtful whether the historical animosity should be “amplifying
conflicting interests” in real Neorealism—there is no clear contradiction in his finding that the
past can end up becoming politicized for reasons having nothing to do with security interests
when the Japanese hedge their alignment commitments.
366
LIND’S APOLOGY POLITICS MODEL
Lind’s apology politics model is in some ways the least Neorealist and material of the
four models. Indeed, the only reason for readily perceiving its Neorealism is Lind’s tight
accordance with Walt’s Neorealism. In reality, her argument that apology-induced nationalistic
backlash in Japan makes the Korean people fear Japan’s long-term intentions is intriguingly
much more akin to recent, “bottom-up” Constructivist work by Ted Hopf than it is to anything
permitted by real Neorealism.
367
Nevertheless, Lind appears to make one tremendous, albeit implicit, Neorealist
assumption, and it informs the actuality of her Neorealism. In particular, careful analysis reveals
she is presupposing Japan’s material fear and insecurity as the only factor compelling Japanese
apologies about the past. As the apology politics model is written, it is clear Tokyo’s leadership
does not apologize because the history issue is contemporaneously problematic bilaterally with
366
Yoon, “Fragile Cooperation,” esp. 94.
367
Ted Hopf, Reconstructing the Cold War: The Early Years, 1945-1958 (New York: Oxford University Press,
2012), Ted Hopf, Social Construction of International Politics: Identities & Foreign Policies, Moscow, 1955 &
1999 (Ithaca, NY: Cornell University Press, 2002).
- 182 -
Korea. That possibility is unworkable logically within the model. If it were the case, the
Koreans and Japanese would already be suffering conflict over the past, and Seoul would
therefore be sensitive to Japan’s unreformed intentions before Japanese nationalists could
respond angrily to Tokyo’s apologies.
368
In other words, the apology-backlash portion of Lind’s
model would become largely spurious. Consequently, it is obvious that Lind has presupposed
from Neorealism why the Japanese apologize. Specifically, for the apology-backlash model to
remain relevant, Tokyo must be apologizing because it is suffering material fear and
insecurity.
369
Unfortunately, if Neorealism’s twin emotions are incentivizing Japan’s ideation-related
behavior, then Lind has fallen into the material-ideational trap of deducing from Neorealism that
only materially derived fear and insecurity matter, even while they simultaneously compel
historical apology, an obviously immaterial phenomena. Of course materialist logic cannot
oblige ideational change, unless ideas are capable of influencing the way things material are
perceived. In that event, however, Lind should not be deducing from Neorealism implicitly that
only material fear and insecurity are inducing Japanese apologies.
Lind’s considerable reliance on real Neorealism to explain Tokyo’s apologetic behavior
has follow-on effects on her model. Neorealist anger suggests Lind should have been much
more precise and careful in her application of social emotions like anger and distrust, which she
uses to directly and powerfully link the ideational and material realms. Particularly remarkable
in this regard is her general point that distrust caused by an ideological phenomenon is the single
368
Under these conditions, Lind would need to argue that a refusal of the Japanese to apologize regarding something
already politicized would not lead to backlash from the Korean side. In other words she would have to argue against
the possibility of an apology-refusal backlash model.
369
One alternative possibility for other cases would be the argument that the apologizing-backlash state apologizes
quickly, before a newly uncovered revelation about the past becomes controversial vis-à-vis pertinent others.
Japan’s obvious reticence about apologizing means this argument cannot be applied to the Korea-Japan case, of
course.
- 183 -
most important determinant of Seoul’s perceptions about Tokyo’s long-term material intentions.
Moreover, given Neorealism’s material imperatives, Lind should not have been so eager to
emphasize Korea’s backlash-induced fears about Japan’s long-term material intentions at the
expense of Korea’s more immediate and frequently changing milieu fears. My earlier discussion
on expectations, as well as my delimitation of Walt’s theory, informs that Neorealism is nothing
if not presently focused. Yet Lind largely, and awkwardly, ignores the very thing that a material
model should be attempting to examine in the relations of two aligned states: the impact of
contemporarily changing material interests on bilateral perceptions of threat and opportunity.
The apology-politics model is, in final analysis, so incompatible with Neorealism that one
wonders why Lind opted to focus so strongly on threat perception in a general balance of threat
framework.
KOO’S COMPLEX INTERDEPENDENCE MODEL
Koo’s complex interdependence model is grounded formally not in Neorealism, but in
Neoliberalism. More specifically, as foreshadowed by the model’s name, it is built within
Neoliberalism’s third level—what I call the transnational level—where complex interdependence
is rooted. The primary emotions compelling cooperation are material fear and insecurity, and the
desire for economic well-being. While these material interests comprise the principal force
leading to cooperation, they sometimes work in tandem with other self-interested emotions. For
example, leaders may or may not care a lot about the economic well-being of their states. But,
presuming they aspire to maintain political power for one reason or another and they believe
their selectorate desires economic well-being, they will act as if they desire it.
Whereas cooperation is linked to material desires in Koo’s complex interdependence
model, conflict emerges whenever emotions related to nationalism surge. This affective set is
- 184 -
comprised of many different configurations of various social emotions, potentially including love
of country and hatred of an Other, as well as embarrassment, anger, and vengefulness.
If the social emotions are more or less clear, the emotional parameters of the actors are
less obvious in Koo’s model. It remains uncertain, for example, whether the individuals and
groups pushing for cooperative behavior at the domestic level are primed to feel only
cooperation-inducing emotions, or whether they can also feel nationalistic ones. Likewise, it is
unknown whether the nationalistic individuals and groups can experience the cooperation-
inducing, material emotions. Presuming all of the actors feel both sets of emotions, the amount
each individual and group prefers one set of emotions over the other is indeterminate. In short,
serious questions exist about whether the primary battle for control of state policy takes place
within people and groups, or between individuals and groups. This missing inquiry matters
because one’s understanding about the locus of the conflict dictates one’s research methodology.
Remarkably, Koo’s model does not appear to suffer any of the material-ideational
contradictions of the three Neorealist models discussed above. By grounding his scholarship in
Neoliberalism’s third level, Koo eliminates the imperative of deducing the meaning of power
from the systemic level. He need not do it, because the state in Neoliberalism’s third level
“represent[s] some subset of domestic society, on the basis of whose interests state officials
define state preferences and act purposively in world politics.”
370
Given that states are reducible
to the set preferences of whichever individual or group is contemporarily dominating policy, and
since preferences at the “bottom” can be material and/or ideational in any permanent
configuration, nothing constrains this aspect of Neoliberalism’s ontology at the third level.
Consequently, even though there is some reason to suspect that Koo is more or less deducing
370
Andrew Moravcsik, “Taking Preferences Seriously: A Liberal Theory of International Politics,” International
Organization 51, no. 4 (Autumn 1997): 518.
- 185 -
from Neoliberalism’s second level his cooperation mechanism
371
—a fact which would
immediately introduce material-ideational logical inconsistencies—the complex interdependence
model emerges largely unscathed from an examination of its social emotions via the lens of
Neorealist anger. It is worth recalling, however, that Koo’s conflictual mechanism is grounded
in emotionality related to nationalism, something inherently ideational and utterly dependent on
the unrealistic expectations of the one-sided nationalist norm. As I argue in Chapter 5, the
complex interdependence model’s main theoretical shortcoming lies in its utilization of these
unrealistic expectations, because expectations change and preferences change in response to
transformed expectations. In short, expectations suggest material Neoliberalism in any of its
configurations is inadequate for modeling alignment politics.
THE DELIMITED BALANCE OF THREAT MODEL
My delimited balance of threat model ostensibly utilizes only one set of social emotions.
As described earlier it is comprised of enmity, ambivalence, and friendliness. Though they are
merely passive emotions and despite their relative unimportance next to Snyder’s three
incentives of alignment behavior and past behavior variable,
372
they still exert not-inconsiderable
influence on perceptions of cooperation and conflict within alignments and between states. In
other words, no matter how passive their effect, these social emotions influence expectations
about the meaning of material power. Unfortunately, as noted in the explication of Neorealist
anger, a fluid meaning of power irreparably challenges the Neorealist foundation of my model.
Despite this shortcoming, I argue shortly below that my first theoretical model is more
true to real Neorealism than the other material models. Before evidencing this claim, however, it
is worth detailing another manner in which my model is remarkable for being non-Neorealist.
371
See my discussion of Koo in Chapter 2.
372
As outlined earlier, and mentioned again below, the three incentives are degree of commitment, relative
dependence, and interest.
- 186 -
Namely, it breaks from Neorealism by relying on Snyder’s three incentives of alignment
behavior and past behavior variable. For, despite the apparent Neorealism of Snyder’s
abandonment and entrapment discussion, not one of these elements is actually Neorealist.
Snyder’s most material of the three incentives is probably the relative dependence of
allies, which Snyder defines as “the amount of harm allies can do to each other by deserting the
partnership or failing to live up to expectations of support.”
373
The difference of material
capabilities between allies is a key aspect of relative dependence. Particularly potent is the fact
that weak states abandoned by powerful allies suffer far more than strong states abandoned by
weak allies.
374
Snyder declares military dependence to be “a function of the degree of threat [a state]
faces from its adversary, the extent to which [its] ally can contribute to deterrence and defense
against the threat, and the availability and cost of alternative means of meeting the threat.”
375
This definition’s ostensibly sole material foundation collapses when one notes the “degree of
threat” perceived by weak allies will on some occasions be dictated in large part by obviously
non-material factors. For instance, a weak state might not fear abandonment by a powerful ally
in the event the primary regional security challenge emerges from a state with which the weak
state shares ethnic, cultural, or ideological affinity.
376
Likewise, a strong state’s power
advantage within an alliance may be incredibly diminished if it shares these same characteristics
with a weak ally, because they make abandoning it more difficult. In either of these situations,
the weak state is no longer dependent on its powerful partner to the extent dictated by material
considerations.
373
Snyder, Alliance Politics, 30-31.
374
Ibid., 166-168, 180-192.
375
Ibid., 31.
376
Snyder understands these non-material factors can play a powerful role. See Ibid., 7, 129, 134, 143.
- 187 -
Snyder’s degree of commitment incentive is also influenced powerfully by ideological
factors. Degree of commitment is comprised of both written or spoken words and deeds, as well
as “strategic or intrinsic” interests. Written and spoken commitments are obliging because they
cause a state to lose political “prestige and reputation” if it fails to honor its pledge.
Commitments also “engage moral and legal values that were previously absent or weak in the
relationship.” In short, ideological phenomena are essential to explaining why promises matter.
They are also imperative in commitments due to strategic or intrinsic interests. Though Snyder
probably intends these interests to be materially derived, his acceptance elsewhere of the
importance of things like ethnic, cultural, and ideological affinities means strategic interests will
often be determined by these nonmaterial matters. Moreover, his understanding that these
affinities impact the meaning of material power inadvertently challenges his assertion that power
will always contain some sort of intrinsic meaning.
377
Snyder’s third incentive is issue interest, by which he means “the parties’ interest in the
specific issue about which they are bargaining.” Differences in issue interest can arise due to
wholly material factors, such as when one ally abuts an adversary, while the other ally enjoys
considerable protective distance. In such an instance the ally neighboring the adversary may
have stronger opinions—i.e., more issue interest—than its partner regarding how to handle the
opponent. Nonmaterial factors can be more important in many cases, however. For example, an
ally sharing ethnic, cultural, and ideological affinities with a prospective alliance member might
be more enthusiastic about its application than a partner possessing none of them.
378
Like
dependence and degree of commitment, an ally’s level of interest in an issue is therefore often
established or influenced by immaterial factors.
377
For commitment and quotes see Ibid., 168-170. For non-material factors, see Ibid., 7, 129, 134, 143.
378
For issue interests see Ibid., 170-172.
- 188 -
Intriguingly, of the four the ways a state can attempt to estimate an ally’s future behavior,
Snyder’s fourth one—the past behavior component—is not only likely to be the most
accurate,
379
but is also completely immaterial. It is never material, even when analysis of an
ally’s past behavior suggests the ally only honors its alliance obligations when materially
advantageous to itself. A judgment like this remains unrelated to materialism, because it is
merely a measurement of how often the ally decides to prioritize its material interests over its
alliance commitments. Discovering an ally always acts for material reasons therefore uncovers
only its unswerving ideological preference for material advantage, rather than some kind of
absolute, theoretically predetermined inability to develop another policy preference.
What careful examination of Snyder’s three incentives informs is that Neorealism’s
materialism cannot satisfactorily explain, on its own, a single one of the determinants of behavior
in alliance politics. Even more noteworthy is that material factors have nothing to do with the
past behavior component, on most occasions the best method with which a state can gauge its
ally’s reliability, and which consequently raises or lowers sharply the state’s abandonment fears.
Reliant as it is on Snyder’s research, my delimited balance of threat model is therefore even
more irreconcilable with Neorealism than suggested by my earlier discussion about the social
emotions of enmity, friendship, and ambivalence, and their influence on the meaning of power.
For better or worse, given their equal reliance on Snyder’s research, analysis of Snyder’s
immaterialism reveals similar theoretical incoherence in Cha’s and Yoon’s scholarship.
Despite these material-ideological inconsistencies, the delimited balance of threat model
is more Neorealist than the other material models in one significant way. Namely, the history
issue DV is influenced only indirectly by the ostensibly material interests. As seen in Diagram
3.4 in Chapter 3, enmity does not compel the FPE & PO to politicize the history issue. Instead,
379
Ibid., 189.
- 189 -
ideological problems are given impetus only within the transnational level, a phenomenon
outlined in diagrams 3.5 and 3.6. This means my deduction about enmity’s indirect influence is
identical to Yoon’s conflict process. It contrasts sharply from the conflictual mechanism in
Cha’s model, however. In his quasi-alliance model, Korea’s fear of abandonment, combined
with its eventual dissatisfaction in the subsequent alignment bargaining, directly rekindles
historic animosity at the state level. Likewise in Lind’s apology-politics model, Japan’s material
fear and insecurity compel the Japanese to apologize, an act which ignites the history issue via
the apology backlash carried out by Japanese nationalists. This backlash causes the Koreans to
view Japanese material intentions suspiciously, which in turn makes the apologies ineffective.
Lastly, Koo’s complex interdependence model diverges completely from Neorealism, as various
material and ideological incentives at the transnational level politicize the past directly.
As reflected in Diagram 3.7, the delimited balance of threat model’s cooperative
mechanism is as indirect as its conflictual process. In it, material security interests lead no
individual or group to actively depoliticize the history issue. In contrast, Cha, Yoon, and Koo all
argue these material fears compel leaders, especially, to ameliorate conflict over the past.
Finally Lind, as noted in Chapter 1, does not seriously consider how material interests improve
the history issue. Her omission is awkward, particularly regarding the Franco-German ties,
because the bipolar nature of the emerging Cold War predated the emergence of German
apologies and a stand-alone, if divided, German state. Moreover, French and German material
interests were more or less aligned and stable across the entire time period Lind examines. The
interest stability these security partners enjoyed is particularly remarkable when contrasted with
the interest instability suffered by the Japan-Korea alignment. Why a materially-focused scholar
- 190 -
would understate the influence of material interests external to the two alignment pairs is
undetermined.
III. Common Sense Social Emotions: Minimal Theory and Fairly Robust Causal
Mechanism
In this section I detail my second alignment politics model. It is critical to note from the
outset that I am not striving to work within in a deductive framework, from a narrow metatheory
like Neorealism. Indeed, I do not attempt accomplish that particular task even in my more
theoretically tethered third model. I instead select and apply emotions with “common sense”
judgment, more or less like all of the scholars whose work I detail carefully throughout this study.
The primary difference between my second model and their research is: 1) My common sense
approach is clearly stated; and 2) I introduce research on emotions to bolster my argument.
In considerable solidarity with some of arguments of the other scholars, I argue the non-
Neorealist social emotions of anger and vengefulness drive much of the ideological conflict
within the Japan-Korea alignment. My cooperative process differs significantly from their
scholarship, however, as I contend the odds of cooperation are increased by the passive,
ideational social emotion of gratitude. But, under the right material and ideational conditions,
material interests can also directly compel ideological cooperation. This cooperative mechanism
engages whenever the FPE & PO fears that conflictual ideas might prevent adequate security
coordination within an alignment. The outcome is a theoretically vague model in which there is
a fairly robust causal relationship flowing from the IV of security interests to the DV of the
Japan-Korea history issue.
- 191 -
A. The Interests
My second theoretical model represents an extension of my first model, the delimited
balance of threat model. Consequently, unless specified, I carry forward the interests and
definitions outlined in Chapter 3. Other than my decision to work without deducing from a firm
theoretical framework, the primary divergence pertains to the delimited balance of threat model’s
stunted emotionality. Specifically, I abandon the notion that material fear and insecurity within
alignments are dominant. Likewise, while I continue to believe enmity, friendliness, and
ambivalence are useful tools for considering many emotions as they funnel “upward” toward the
system level and “downward” to the transnational level, I no longer consider these passive
emotions to be the only ideational social emotions directly influenced by, and influencing,
material security. Largely muscling past the passive emotions to produce conflict within
alignments are active affects comprised of anger and vengefulness. Cooperation in contrast can
be produced passively, via the emotion of gratitude; or actively, in response to fear and
insecurity. Importantly, there is no reason why anger, vengefulness, and gratitude must comprise
the only new social emotions introduced my second model. Only in the interest of parsimony do
I focus on them in the common sense social emotions model.
Material fear and insecurity still play powerful roles in alignments. As noted earlier,
perceptions of material threats and opportunities within state security milieus continue to guide
both alignment formation and decisions to uphold or hedge alignment obligations. But fear of
ideas can also leads to ideological cooperation within alignments. Specifically, when the FPE &
PO fears ideological discord will prevent the material cooperation it deems necessary, it works
actively to ameliorate the ideational conflict. Thus, states fear both material and ideological
phenomena in my second model.
- 192 -
While states remain highly attuned to the emotions of material fear and insecurity, I
differentiate between extreme and non-extreme levels of material fear and insecurity. States
experience extreme material insecurity when there is an immediate, existential threat to their
sovereignties or their people’s way of life. Like Wendt’s “hotel fire” scenario, nothing beyond
surviving the material threat at hand matters.
380
In an emergency of this magnitude, the FPE &
PO of a given state may feel deeply angry and vengeful over an alignment member’s semi-
defection from their alignment. But, if acting rationally, these men and women will present
nothing but kindness to their hedging partner, in order to increase its perceived value of the
alignment. There seems to be no reason to doubt that material expressions of loyalty will
practically always take place under this kind of extreme situation. Importantly, a jilted state will
also express ideological solidarity to its hedging security partners, if it anticipates that
ideological barriers are influencing how its worth is being evaluated. In short, when existentially
insecure, a jilted state will express material and/or ideational loyalty to its partners, so as to make
them value more the alignment. And it will ignore any conflictual social emotions it feels. The
fact social emotions experienced by the jilted state will not matter in such dire circumstances is
reflected in Diagram 4.3.
380
Wendt, Social Theory of International Politics,” esp. 121-122, 159, 233.
- 193 -
Extreme and immediate existential threats to both X and Y simultaneously also lead to
good relations, presuming the threat compels both to rely on their alignment for safety. Better
ideological outcomes are shaped only indirectly in this instance, however, via two mechanisms.
The most important is the passive emotion of gratitude. It induces the FPE & PO in both states
to be less likely to aggravate ideational problems emerging from the transnational level. The
second mechanism is also passive. Namely, the intense material security threat at hand focuses
minds; it causes the FPE & PO to deprioritize—to forget about—comparatively minor things like
Diagram 4.3
An Immediate, Existential Material Threat to One State Leads to Cooperative
Material and Ideological Relations within an Alignment
Aligned states X & Y suffer conflicting security interests due to differing perceptions of the threat
represented by Z. X perceives an immediate existential threat to its sovereignty. Y, in contrast, does
not fear Z very much.
Y hedges its alignment obligations to X, because Y worries about becoming
entrapped in a conflict with Z, a state it does not fear very much.
X experiences fear of abandonment by Y.
Provided X’s material and/or ideational expressions of loyalty to Y have induced Y to value
more its alignment with X, and consequently to uphold its alignment obligations adequately,
X and Y will enjoy good bilateral relations.
X suffers disappointed security expectations
and becomes angry and vengeful toward Y.
X’s fear of abandonment by Y in
an existentially challenging
security environment leads it to
ignore its feelings of anger and
vengefulness toward Y.
X expresses loyalty toward Y, in order to entice Y’s
material support. If ideological issues are perceived to
be a stumbling block to good alignment relations, X
makes ideological concessions in addition to material
ones.
- 194 -
ideational squabbles with friendly states. Why does the FPE & PO in one or both states not
undertake deliberate steps to tamp down immaterial issues, when both are under extreme security
threat? They do not because they lack incentive. Since the alignment partners desperately need
each other, there is no chance their ideological problems will prevent mutual material support.
Consequently, nothing compels purposeful action to improve ideational issues. How
indirectness leads to good relations is illustrated in Diagram 4.4.
Fortunately, as Wendt notes, there “are very few ‘hotel fires’ in international politics.”
381
Social emotions are consequently capable of exerting considerable influence in most episodes of
381
Quote is from Ibid., 129 but see also 159, 233, 256, 237.
Diagram 4.4
An Immediate, Existential Material Threat to Both Security Partners Leads to
Cooperative Material and Ideological Relations within the Alignment
X & Y enjoy aligned security interests because of the extreme and immediate material threat to their
sovereignties and/or way of life.
X & Y honor their alignment obligations to each other, meaning they are enjoying good material relations.
X & Y experience satisfied security expectations.
X & Y ignore bilateral ideational
phenomena, as they are focused on
material survivability.
X & Y feel gratitude toward each other, and therefore
become less likely to aggravate ideational conflict if
it emerges transnationally.
X & Y enjoy good material and ideational relations.
- 195 -
alignment bargaining. Regarding Japan-ROK relations, surely no episode occurring after the
Korean War qualifies as a true hotel fire. While I believe America’s ongoing troop presence in
both Japan and South Korea provides substantial prima facie evidence for my claim that there
have been no immediately pressing existential challenges to either quasi-ally, I note that
existential challenges are often predicted to generate completely different outcomes than are
normal crises. A basic test therefore exists to see whether the quasi-allies perceive their
alignment crises to be existential.
Consider the two likely outcomes in an otherwise identical episode of Seoul-Tokyo
bargaining, in which Korea semi-defects from Japan. In an existential crisis experienced by
Japan but not Korea, Japan’s material fear and insecurity would compel it to exhibit material and
ideational loyalty to the ROK. In a normal crisis, however, Tokyo’s anger and vengefulness
would cause it to lash out at Korea in either, or both, the material and ideational realms. Not
insignificantly, the latter conflictual outcome matches the general predictions Cha and Yoon
make regarding what happens whenever diverging material interests lead one of the quasi-allies
to hedge its alignment obligations. In short, though Cha, especially, tends to exaggerate Korea’s
material fear and insecurity,
382
both Cha and Yoon consistently evidence there have been no
security events in modern Japan-Korea relations so extreme as to deaden social emotions. That
really is no surprise, however: As emphasized elsewhere in my research, both scholars employ
ideational social emotions liberally.
382
Cha’s argument that Korea experiences sharp abandonment fear when Japan hedges in response to America’s
galvanized presence is awkward, for example. Korean fears over Japan’s semi-defection should be considerably
minimized by the US’s bolstered commitment, considering the power differential between the US and Japan. Cha is
especially weak in this regard when he discusses the 1980s. See Cha, Alignment Despite Antagonism, 169-198.
Snyder’s argument suggests the US presence would greatly diminish Korea’s dependence on Japan. Hence, the
ROK’s abandonment fears would be very low. See Snyder, Alliance Politics, 167.
- 196 -
Diagram 4.1, located at the beginning of the current chapter, illustrates the process
leading to material and/or ideational conflict in an alignment, when one partner semi-defects
during a non-emergency crisis. At the root of conflict were the active emotions of anger and
vengefulness, as experienced by the jilted state. It is important to note these emotions will
almost always be ideational, despite being activated by material fear and insecurity. A quick
review of Neorealist anger reveals their shared immaterial ontology. Neorealist anger, as argued
above, directs action against the source of one’s anger only when material fear and insecurity are
neutral to the best source of recovering the lost material power. Consequently, in the unlikely
event material fear and insecurity directly incentivize the coercion of a hedging partner’s loyalty,
Neorealist anger exerts zero influence.
383
And when the twin material emotions point to other
methods of recovering the lost material support—e.g., self-help, finding new security partners,
revitalizing an extant alignment with a third state—Neorealist anger is overridden. In short, the
possibility Neorealist anger might play a role is extremely small. Anger, when activated, will
practically always be ideational. The vengefulness accompanying it will therefore also be
immaterial.
Whereas Neorealist anger requires jilted states to coerce or assail hedging partners
materially, regular anger and vengefulness can produce responses in both the material and
ideational realms. In order to predict and explain outcomes, one must examine the extent of the
material and ideational leverage jilted states have over their hedging partners. While I decline to
make firm and precise propositions about the phenomena that matter when anger and
vengefulness bring about an aggressive response within an alignment, it makes sense to declare
that things like relative military advantage, economic interdependence, and the nature of
383
As I discussed in Chapter 3, material fear and insecurity will seldom lead to coercion of a hedging security
partner.
- 197 -
ideological relations must be considered. I demonstrate this general approach with three distinct
illustrations, including the modern Japan-Korea case.
First, at one end of the material-idea advantage spectrum sits the example laid out in the
Melian Dialogue, in which the people on the island of Melos refused to live up to the security
obligations expected into existence by Athens. Because Athens suffered inadequate ideological
leverage over the Melians, the angry and vengeful Athenians could only attack their reluctant
security partner materially.
384
Importantly, Athens’ material response was remarkably
unconstrained: Its incredible material advantage over the Melians and in the region around Melos
meant the Athenians could pursue whichever retaliatory policy they wished, provided adequate
preparation time. Likewise, whatever minimal economic interdependence there was between
Athens and Melos was obviously too insubstantial to shape Athenian choices. Lastly, no modern
human rights norms inhibited Athens’ final response. The Athenian decision to kill or enslave
all of the Melians is accordingly in the range of imaginable outcomes.
385
Second, as best clarified by Janice Bially Mattern’s research on the 1956 Suez Crisis, the
UK-US alliance exists near the other end of material-ideational spectrum. When British security
interests led the UK to invade Egypt in coordination with French and Israeli forces, both British
and American security expectations were simultaneously disappointed. The resulting anger-
fueled vengeance was carried out primary ideationally, with the Americans publicly accusing the
British of having a “character that was distinctly bellicose and unfamiliar,” while the British
384
It is worth noting the Athenians did degrade and disparage the Melians in the dialogue, the main purpose of
which was to use ideas to convince the Melians to enslave themselves voluntarily. In 20
th
century Cold War terms,
the representatives from Athens argued for the idea that it is “better to be red than dead.”
385
Thucydides, History of the Peloponnesian War, 400-408. The decision to kill, rather than enslave, all the men of
military age evidences the anger and vengefulness of the Athenians.
- 198 -
loudly berated the Americans for their lack of loyalty. In short, they strived in their anger and
vengefulness to harm the other via disparagement and delegitimization.
386
“We-ness” norms, though greatly destabilized during the crisis, seem to explain why the
US “never even discussed” using its vastly superior military options to pressure the UK.
387
The
US did apply minor economic sanctions to help coerce its ally’s loyalty, however. American and
British diplomats and statespersons nevertheless relied primarily on language to compel the
leadership in both states to return to the Anglo-American special relationship status quo. They
utilized what Bially Mattern calls “representational force.” By utilizing distinct words and
phrasing, these men and women solidified the wavering norms of the special relationship, and
greatly damaged the emergent identity that had underpinned the conflict between the UK and US
in the first place. In particular, it became absurd once again for people to think the two states
were competing with each other for domination of the West.
388
Third, the Japanese-Korean relationship is situated somewhere between the Athens-
Melos and UK-US alignments. A product of the modern world, it is no doubt much closer to the
latter than the former. In some sense, therefore, Seoul’s and Tokyo’s tendency to pursue anger-
fueled vengeance in the ideational realm is easy to envision. Nevertheless, as best evidenced by
President Roh’s show and threat of military force in the 2006 Dokdo/Takeshima crisis, ideational
constraints on the use of military force are clearly far less robust in the Japan-ROK alignment
than those present in the UK-US alliance. With bilateral norms about the appropriate use of
military force comparatively limited in Seoul-Tokyo relations, it becomes apparent that ideas
paralleling the UK-US special relationship are very unlikely to be the primary determinates
386
Janice Bially Mattern, Ordering International Politics: Identity, Crisis, and Representational Force (New York:
Routledge, 2008), 79-81.
387
Ibid., 61.
388
For Bially Mattern’s summary of her theory, see Ibid., esp. 13-20.
- 199 -
Japan-Korea relations. It therefore becomes critical to consider the extent to which material
interests impose fairly rigid limitations on the type of vengeance the quasi-allies can carry out
against each other.
Simply stated, material incentives make it very difficult for an angry and vengeful Japan
or Korea to damage each other militarily or economically. And with retaliation nearly always
falling outside of security and economics, things ideational become the default target. America’s
military presence in, and alliances with, Japan and the ROK respectively are critical in
explaining military constraints. The US military’s overwhelming importance to Seoul’s and
Tokyo’s securities makes it all but inconceivable that either Japan or Korea would harm each
other militarily, no matter their emotional pique. This is the case even for Japan, the more
powerful member of the Japan-Korea alignment. Thus, for example, the Japanese may have felt
angry and vengeful toward the Koreans following Seoul’s semi-defection from Tokyo during the
Sunshine Policy decade. But Japan’s military dependence on the US, coupled with America’s
alliance with Korea, meant there were no good options for the Japanese to pressure Seoul
militarily. Vengeful action against Korea in the security realm would have harmed the security
interests Washington expects Tokyo to uphold, a fact almost certain to bring about problematic
Japan-US relations.
Several economic phenomena help explain why conflict is also unlikely in the economic
sphere. First, as Koo argues, Japan and Korea have become economically interdependent.
389
Angry retaliation directed at an economic concern would accordingly impart economic harm to
both states, no matter which one acted vengefully first. Second, the establishment and
maturation of rule-setting and enforcing international institutions such as the WTO further
decreases the likelihood of serious economic conflict between Seoul and Tokyo. And third,
389
Koo, “Scramble for the Rocks.”
- 200 -
economics and security are often fungible issues. Thus, Japan’s security relationship with the
US prevented Tokyo from punitively leveraging its considerable economic advantage over Seoul,
because Washington’s overall Cold War strategy by the mid-1960s called for Japan to take the
lead in bolstering the Korean economy, thereby permitting a more robust and stable “anti-
communist arc” across much of East Asia.
390
Japanese failure to invest in, or trade with, Korea
consequently threatened to disappoint the security obligations the US expected Japan to fulfill.
391
B. The Emotions
THE RATIONALE BEHIND WEAPONIZED IDEAS
But why bother carrying out angry and vengeful attacks in the ideational realm in the first
place, no matter one’s material limitations? More specifically, with what rationality does the
FPE & PO desire to lash out deliberately at hedging security partners with pen and tongue? One
possibility, which applies primarily to the current chapter, requires the FPE & PO to believe
ideas can be harmful in international relations, and that they are therefore weapons well-suited
for acts of vengeance. As I discuss below, the FPE & PO accordingly uses pain-inflicting ideas
as weapons to restore a sense of justice. A second explanation, which I pursue in Chapter 5,
departs from the idea of vengeance but remains somewhat linked to anger. Namely, the FPE &
PO says and writes nasty things about disappointing others in order to feel better about itself—to
protect its collective self-esteem following a humiliating security betrayal. The third
rationalization I consider is a domestic politics blending of the first two. In particular, certain
members of the FPE & PO target semi-defecting security partners for disparaging treatment, with
390
Indeed, Washington’s desire for Japan to bolster South Korea’s economy was a major reason for US
normalization pressure in the mid-1960s. Cha, Alignment Despite Antagonism, 29.
391
Japan’s tendency to view Korea in terms of the Japan-US alliance was particularly marked during the 1960s and
1970s, which was precisely when the Japanese enjoyed their greatest economic leverage over the Koreans. Byung-
Joon Ahn, “Japanese Policy Toward Korea,” in Japan ’s Foreign Policy After the Cold War: Coping with Change,
ed. Gerald L Curtis (New York: M. E. Sharpe, 1993), 267.
- 201 -
the primary goal of solidifying their own political position domestically. It resembles the first
argument in that the FPE & PO must perceive transnationally derived ideas as having a kind of
material-equating power, at least at home,
392
and it is parallels the second rationale to the extent
the FPE & PO believe the electorate/selectorate will respond positively to other-disparaging
words that make them feel better about themselves.
The thrust of the current chapter requires members of the FPE & PO to experience ideas
powerfully, at least in the targeted state and under certain conditions. As outlined earlier, ideas
can remain salient, and compelling, within an alignment at any time but one: When all alignment
members simultaneously suffer an immediate, materially-founded existential crisis. Under such
pressing circumstances, the kind of ideational phenomena plaguing alignments like the bilateral
Japan-Korea quasi-alliance simply become irrelevant within relationships.
But the FPE & PO does more than just equate ideas with power. In order for the
common sense social emotions model to be effective, the FPE & PO must regularly weaponize
ideas for the purpose of harming semi-defecting states. Feeling angry and vengeful, the FPE &
PO in the jilted state strives to damage its hedging partner in some way. One possibility target is
the reluctant alignment member’s collective self-esteem.
393
The initial disparagement and
belittlement battle accompanying the UK-US Suez Crisis represented a collective self-esteem
attack, for instance.
The angry and vengeful state might wish to damage the material interests of the offending
party instead of, or in concert with, harming its collective self-esteem. When material interests
are the target, states initiate idea barrages with the purpose of causing the reluctant security
392
This is true at least to the extent politicians perceive they can utilize angry, alignment-targeting ideational attacks
as a means to provide “cover” for economic policies that are not in the material interests of large chunks of the
electorate or selectorate.
393
Collective self-esteem is “a group’s need to feel good about itself, for respect or status.” Wendt, Social Theory of
International Politics, 236.
- 202 -
partner to do something against its material interests. While the latter possibility could represent
an instance in which ideas vengefully leveraged end up compelling the semi-defecting partner’s
alignment loyalty—i.e., coercing it to return to its prior alignment commitments—vengeful states
do not specifically try to reverse or repair the harm they themselves have suffered.
394
Rather,
vengeance compels the jilted to seek justice via the restoration of the “the balance of
suffering.”
395
Restoring the alignment is therefore unlikely to be the specific goal of
vengefulness, unless the vengeful state perceives coerced loyalty to be the best way for making
the semi-defecting state suffer.
CHA’S AND SNYDER’S WEAPONIZED IDEAS
While at first blush it may seem odd to contend the FPE & PO leverages weaponized
ideas to harm hedging security partners, similar claims pepper IR alignment literature. Among
the four Japan-Korea models studied most carefully in this research, Cha’s was by far the most
remarkable in this sense.
396
As argued in chapters 2 and 3, and also above, the quasi-alliance
model requires Korea to rationalize material efficacy for the history issue. If not, why else
would these rational and materially-minded men and women utilize the past to compel Tokyo’s
security loyalty? Similarly, Cha’s Japanese must be pained materially by Seoul’s idea barrage.
Otherwise they would not continue to fear entrapment, and they would accordingly never engage
in an ideational conflict spiral with the ROK. In effect, the quasi-alliance model empowers ideas
394
Nico H Frijda, “The Lex Talionis: On Vengeance,” in Emotions: Essays on Emotion Theory, ed. Stephanie H M
van Goozen, Nanne E van de Poll, and Joseph A Sergeant (Hillsdale, NJ: Lawrence Erlbaum Ass., 1994), 266.
395
Ibid., 272.
396
As argued previously, Yoon, Lind, and Koo all realize that ideas enjoy material-equaling power. But their
respective models do not require weaponized ideas to make sense, as does the quasi-alliance model.
- 203 -
even more than the common sense social emotions model, because Cha’s Koreans weaponize the
history issue for the specific purpose of compelling Japan’s alignment loyalty.
397
Coercing reluctant alignment partners with weaponized ideas precedes Cha’s research,
however. Consider the work of Glenn Snyder, upon whose abandonment-entrapment
mechanism the quasi-alliance model is built. Like Cha, Snyder understands that certain ideas, if
sharply stated, can compel alignment members to act against their material interests. Snyder’s
most extraordinary emotionality claim in this regard is that states can effectively leverage the
emotions of anger and contempt to force a hesitant ally to uphold its alignment obligations.
Indeed, this ability comprises a portion of one of the seven actions a state can take to assuage its
fears of abandonment. They are: 1) ally with an additional state; 2) increase military
preparedness; 3) acquire additional resources through military action; 4) conciliate the adversary;
5) reduce the ally’s incentive to defect and increase its desire to reciprocate by evidencing
loyalty to it; 6) threaten to defect in order to pressure the ally to reaffirm its commitment; and 7)
effect some kind of sanction against the hedging ally to impose costs on disloyalty.
398
Incredibly,
“probably the most potent sanction” is attacking one’s ally with the emotions of “anger and
contempt”
399
The idea that assailing an ally with anger and contempt can be efficacious in a generally
Neorealist model is astounding. Snyder himself seems quite confused about it. Early in his book,
for example, he claims that “norms are weak because the costs [i.e., the sanctions in response to
397
As argued before, I perceive Cha’s central argument to be that Korea tries to “elicit security cooperation [from
Japan] by exercising leverage over other issue areas.” The only issue over which Korea enjoys leverage apparently
adequate to the task is the history issue. Cha, Alignment Despite Antagonism, 43.
398
I am combining Snyder’s four ways of reducing dependence on an ally with his two methods for avoiding
entrapment and his discussion of sanctions in response to disappointed norms. See Snyder, Alliance Politics, 167,
183-184, 363-364.
399
Ibid., 363-364.
- 204 -
norm violations] are small compared to the value of ‘interests.’”
400
Yet he illustrates the real-
world power of anger and contempt attacks by asserting that Austria’s reproaches in response to
German hedging wore down the Kaiser and helped bring about Germany’s firm alliance
commitment in the run-up to World War One. Paradoxically, therefore, “weak” sanctions
targeting the emotions enforcing norms of alliance loyalty contributed considerably to the cause
of the materially catastrophic, history-altering First World War.
401
COOPERATION & CONFLICT: THE ROLE OF MATERIAL FEAR AND IDEATIONAL
ANGER & VENGEFULNESS
Snyder’s intuition about emotions is truly remarkable. For, in contending the “standard
response” of a state suffering material abandonment fear is to express material loyalty to its
hedging partners, and in linking ideational anger to attacks on semi-defecting alignment
members, Snyder’s arguments parallel the findings of more recent Social Psychology (SP)
research.
402
Importantly, the SP scholarship supports Snyder’s judgment that fear should lead to
expressions of kindness, while anger should lead to aggression.
403
Within the SP scholarship, fear is generally associated with passivity due to uncertainty
and “a sense that factors beyond one’s control shape outcomes.” In contrast, anger tends to
produce action, because it “arises from appraisals of certainty and individual control.”
404
Moreover, fear and anger usually lead to contrary risk assessments: fear exaggerates risk
400
Ibid., 17.
401
Ibid., 363-364.
402
Snyder, Alliance Politics, 183, 363-364.
403
See, for example, Jennifer S Lerner et al., “Effects of Fear and Anger of Perceived Risks of Terrorism: A
National Field Experiment,” Psychological Science 14, no. 2 (March 2003): 144-150; Jennifer S Lerner and Dacher
Keltner, “Fear, Anger, and Risk,” Journal of Personality and Social Psychology 81, no. 1 (2001): 146-159; Jennifer
S Lerner and Dacher Keltner, “Beyond Valence: Toward a Model of Emotion-Specific Influences on Judgment and
Choice,” Cognition and Emotion 14, no. 4 (2000): 473-493.
404
Ibid., 480.
- 205 -
perception; anger mitigates it.
405
In other words, states experiencing fear are likely suffering an
exaggerated perception of risk, increased uncertainty, and sense they are not in control. Under
these conditions, and considering the risks inherent in attacking one’s hedging alignment
partners materially or ideologically,
406
no wonder states primarily fearing abandonment prefer
being nice to their reluctant alignment partners. And given that primarily angry states are
probably enjoying greater certainty and control, while perceiving less risk, material and/or
ideational aggression in response to disappointed alignment security expectations is also highly
logical.
The dynamics of fear explain why states in the common sense social emotions model do
not attempt to coerce alignment loyalty with weaponized ideas, as they do in Cha’s model.
When material fear is the primary emotion, as it is in existential security crises, jilted states
exhibit material loyalty and kindness.
407
While these incredibly insecure states almost certainly
feel ideational anger and vengefulness concomitantly, the severity of their fearfulness causes
them to ignore the twin social emotions as they strive only to survive. In contrast, when fear is
low enough, as it is under normal—i.e., non-existential—crises, anger and vengefulness become
activated. With material insecurity low, however, there is no longer great incentive to coerce
alignment loyalty. Primarily angry and vengeful states accordingly focus more on the task of
simply restoring the balance of suffering. They predominantly act on their vengefulness via their
anger, in other words.
Returning to the Seoul-Tokyo case, understanding the kind of action one can expect when
either fear or anger are predominant helps resolve the puzzle of why the Koreans and Japanese
lash out at each other whenever their respective bilateral security expectations are disappointed.
405
Ibid., 485.
406
See my discussion in Chapter 3 on the danger inherent in coercing alignment loyalty.
407
They may show ideational loyalty, as I discuss shortly below.
- 206 -
Instead of wondering why material fear and insecurity do not compel the jilted state to exhibit
kindness to its semi-defecting partner, understanding the directionality and tenor these material
emotions should normally impart in alignment affairs causes one to recognize that angry and
vengeful states are primarily angry and vengeful, not fearful. And states that act kindly and
otherwise express loyalty when being abandoned are primarily fearful and insecure materially;
they are not experiencing some social emotion like happiness or satisfaction.
Before continuing onward to causes of ideational cooperation in alignments, it is worth
emphasizing three points. First, there are four kinds of alignment crises in two-way alignments.
There are existential and normal crises, and each of these can be internally divided in two,
depending on whether the two alignment members’ material milieu interests are convergent or
divergent. As shown below in Table 4.1, alignment outcomes depend primarily on which one of
the four crises the security partners experience. Importantly, jilted states in the modern world
rarely experience existential material fear and insecurity when alignment partners hedge their
obligations. Insecurity in these instances is much lower than it may have been in the past,
because the world is simply nothing like the deadly place it used to be, at least for sovereign
states at the international level.
408
Second, material fear and insecurity lead, on their own, only to material cooperation. By
themselves, the twin Neorealist emotions cannot incentivize ideational cooperation, no matter
how existentially threatening a given security milieu. As I explain shortly below, the FPE & PO
must also fear ideas, if material interests are to promote ideational cooperation. Third, whereas
material fear and insecurity incentivizes only material cooperation, ideational anger and
vengefulness can bring about both material and ideational conflict. As discussed earlier, the type
408
Hedley Bull, The Anarchical Society (New York: Columbia University Press, 1977).
- 207 -
of quarrel these social emotions provoke depends on the jilted state’s interests and capabilities
relative to the state or states at which it is ideationally angry and vengeful.
IDEATIONAL COOPERATION: THE ROLE OF IDEATIONAL FEAR AND GRATITUDE
Though anger and vengefulness can produce material and/or ideational conflict,
ideological cooperation requires either ideational fear or gratitude.
409
These ideological
409
As argued earlier, in existential crises suffered by all members of an alignment simultaneously, ideational
relations might improve due to distraction of the FPE & PO. In this event, pressing material interests tend to cause
the FPE & PO to forget about and/or ignore ideational squabbles. Consequently, ideational troubles should
Table 4.1
Four Types of Crises and their Outcomes in Two-Way Alignments
Type of Crisis Existential Existential Normal Normal
State or States
Perceiving
Crisis
X & Y X X & Y X
Material
Interests of
State X &
State Y
Align Diverge Align Diverge
Material
Action by Y
and Causal
Emotion
Deliberate
Loyalty,
Material Fear.
Deliberate Hedge,
Material Fear.
Deliberate Loyalty,
Material Fear.
Deliberate Hedge, Material
Fear.
Material
Action by X
and Causal
Emotion
Deliberate
Loyalty,
Material Fear.
Deliberate Loyalty,
Material Fear.
Deliberate Loyalty,
Material Fear.
Deliberate Conflict, Anger
& Vengefulness.*
Ideational
Action by Y
and Causal
Emotion
Indirect
Cooperation,
Gratitude.
NA
Indirect Cooperation,
Gratitude. And might
exhibit Deliberate
Loyalty due to
Ideational Fear.
NA
Ideational
Action by X
and Causal
Emotion
Indirect
Cooperation,
Gratitude.
Might Exhibit
Deliberate Loyalty,
Ideational Fear.
Indirect Cooperation,
Gratitude. And might
exhibit Deliberate
Loyalty due to
Ideational Fear.
Deliberate Conflict, Anger
& Vengefulness.*
Notes on
Ineffective
Emotions NA
X feels Anger and
Vengefulness. But
they are overridden by
Existential Material
Fear.
NA
X will often experience a
relatively low level of Fear
of Abandonment. But it is
overridden by the
ideational emotions.
* The type of conflict X initiates depends on its interests and capabilities.
- 208 -
emotions exert their influence in different ways, however. Ideational fear propels the FPE & PO
to tamp down ideational friction deliberately. Gratitude, in contrast works indirectly. It merely
makes the FPE & PO less likely to aggravate fractious ideas as they emerge from the
transnational level.
Remarkably, without concomitantly experienced ideational fear, material fear and
insecurity cannot incentivize purposeful ideational cooperation. No matter the extent to which a
crisis is existential or normal, the FPE & PO in one or all aligned states must fear, specifically,
the capability of ideological issues to prevent desired material cooperation. The FPE & PO must
fear ideas, in other words. If these men and women cannot fear ideas, they will never
deliberately strive to tamp down ideational conflict for material security reasons, because no
incentive will compel them.
While ideational fear obliges purposeful handling of contentious ideas, gratitude’s
influence is much more indirect. Triggered when alignment security expectations are satisfied
materially, gratitude reduces the likelihood the FPE & PO will aggravate intra-alignment
ideological problems as they are initiated or propelled forward by transnational actors. Gratitude
is therefore a fairly passive affect in international relations. It is passive in the sense it does not
compel a specific action: No one strives in his or her gratitude to improve ideational relations
within alignments deliberately. But it is somewhat akin to an active emotion, because it does
counter the motivations of individuals to politicize the past for personal gain.
Gratitude’s influence occurs through one of two mechanisms. First, it can balance out
directly the desire of certain members of the FPE & PO to use ideological alignment troubles as
a means of achieving personal objectives. And second, it can deaden their incentive indirectly,
ameliorate due to lack of interest by a major proportion of the population. Because this phenomenon should occur
primarily in existential crises, I deemphasize it in my research.
- 209 -
because the broader FPE & PO’s gratitude makes these politically aware individuals less
receptive to the arguments and claims of those striving to use ideological discord for political
leverage. Thus, a prime minister or legislator who might otherwise seek to bolster his/her
reelection odds by politicizing further an alignment-related ideological issue might be influenced
by his or her gratitude and become less likely take advantage of it. Or s/he will simply realize
the tactic is less likely to succeed under the current environment of gratitude among the FPE &
PO.
Gratitude is an expectations-related, pro-social emotion. It can be triggered when one
person or group receives a benefit from another.
410
Unlike the relationship between expectations
and material fear,
411
however, expectations and gratitude are not positively correlated. Instead,
gratitude declines as expectations of receiving a benefit rise.
412
The influence of gratitude in
alignments is consequently not as straightforward as one might initially anticipate.
Research has revealed an intriguing explanation why expectations are negatively
associated with gratitude. Namely, certainty of expectations regarding receiving a benefit evokes
indebtedness largely in place of gratitude.
413
While similar at first blush, these emotions are
actually quite different. Gratitude is much more pro-social in its influence, meaning it induces
the desire to “help, as well as refrain from harming,” those who have provided a benefit.
414
Indebtedness, in contrast, is far less pro-social. Not only does it promote less desire to help the
benefactor, but it does not seem to induce the desire to eliminate antisocial behavior.
415
To some
410
For other definitions of gratitude see the concise discussion in Jo-An Tsang, “Gratitude and Prosocial Behaviour:
An Experimental Test of Gratitude,” Cognition and Emotion 20, no. 1 (2006): 139.
411
I explicate the relationship of expectations and fear in Chapter 5.
412
See the discussion in Michael E McCullough, Robert A Emmons, Shelley D Kilpatrick, and David B Larsen, “Is
Gratitude a Moral Affect,” Psychological Bulletin 127, no. 2 (2001): 255.
413
Philip C Watkins, Jason Scheer, Melinda Ovnicek, and Russell Kolts, “The Debt of Gratitude: Dissociating
Gratitude and Indebtedness,” Cognition and Emotion 20, no. 2 (2006): esp. 233-234.
414
Tsang, “Gratitude and Prosocial Behaviour, 139.
415
Watkins et al., “The Debt of Gratitude,” esp 233-234.
- 210 -
extent, therefore, my claim that high expectations produce little gratitude might be shorthand for
saying there should be comparatively higher levels of indebtedness. Only in the interest of
simplification do I focus primarily on gratitude.
One outcome of the negative correlation between expectations of security support and
gratitude, is the level of gratitude in alignments is often very weak. In particularly strong
alignments like the contemporary UK-US alliance, mutual material support will accordingly
produce little gratitude. In remarkably weak alignments, in contrast, alignment loyalty will
induce strong feelings of gratitude, especially in the event security support exceeds the security
expectations of the benefit-receiving state. This phenomenon no doubt explains why two
researchers define gratitude as “an emotional response to a gift.”
416
The Japan-Korea relationship enjoys neither incredibly powerful, nor remarkably weak
alignment expectations. Nevertheless, as evidenced by the expectations dependent findings of
powerful abandonment fear by Cha; considerable but assuageable abandonment and conflict
deterrence failure fears by Yoon; and my own research on the significant influence of anger-
fueled vengefulness and gratitude, alignment expectations are obviously considerable in Japan-
Korea relations. Consequently, the gratitude experienced by the FPE & PO will on many
occasions be remarkably weak in its influence. Gratitude’s relative impotence in Seoul-Tokyo
affairs explains why its influence on the history issue is more passive than the effects of anger
and vengefulness. Only when alignment expectations are actually exceeded by the amount of
security support carried out will gratitude compel deliberate ideational cooperation. Since states
seem not to set their expectations low relative to the support they tend to end up receiving,
gratitude should usually be passive in alignment politics.
416
Quoted from Tsang, “Gratitude and Prosocial Behaviour,” 139.
- 211 -
Importantly, gratitude’s passive effect depends greatly on the ontological nature of
security expectations. As I explain in the next chapter, there are two ways of considering
security expectations: temporally static or temporally longitudinal. In the former, expectations
are set at the beginning of an alignment and do not change until the alignment collapses.
Alignment cooperation and conflict depends greatly on whether the level of intra-alignment
support satisfies or dissatisfies these stable security expectations at any given moment. Both of
my first two theoretical models, as well as all the alignment modeling literature I examined most
carefully, are remarkable for primarily incorporating temporally static and substantively stable
expectations.
In contrast, the collective self-esteem model I detail in Chapter 5 utilizes longitudinal
expectations. Their actualization still occurs at the moment of initial alignment, via what I call a
“trusting act.” But they are no longer stable. Instead, when alignment support fails to satisfy
prior expectations, members of the FPE & PO realize the security partnership is not as strong as
it once was. Expectationally disappointed, the FPE & PO’s alignment expectations degrade, a
process which causes emotionality damaging to the security relationship. Conversely, trusting
acts are the primary process through which expectations are raised beyond the level of those
prior. Being driven by material interests to undergo a trusting act, the FPE & PO brings about
both stronger expectations and alignment cooperation.
There is a point to conclude regarding expectational ontology and ideational gratitude.
Under temporal stasis, alignment expectations are more or less set until the partnership
terminates, and they are seldom exceeded by the level of alignment loyalty. Hence the passivity
of gratitude under staticity. In contrast, the depth of alignment expectations in longitudinal
ontology can shift upwards or downwards as frequently as material considerations permit. In
- 212 -
some security relationships, therefore, the level of security aid received often surpasses prior
expectations. If so, the influence of the gratitude these alignments experience will be much more
active relative to gratitude produced under temporal staticity. The temporal ontology of security
expectations is therefore critical to understand, particularly in alignments undergoing repeated
cycles of security-driven cooperation and conflict. Expectational ontology is essential to
understand in Japan-Korea relations, in other words.
- 213 -
Chapter 5
Theoretical Model #3: The Collective Self-Esteem Model
I. Overview
While constructing my first two theoretical models I underexplored an important question:
Why do security expectations exist in alignments? As argued in Chapter 3 while discussing
critical aspects of Snyder’s work on alliance politics, expectations of security loyalty underpin
every alignment on earth. The issue is definitional, since alignments cannot exist without
expectations of security support. Thus, in Snyder’s research, expectations explain both the sense
of security states enjoy once in an alignment, and the fear of abandonment suffered as those
expectations are disappointed.
417
Precise, predictive models like Cha’s and Yoon’s, as well as
my own models, operate by exactly the same mechanism. This similarity is no surprise, given
the great indebtedness these models owe to Snyder’s path-breaking research.
In this chapter, while developing my third model—the collective self-esteem model—I
address the question of why expectations exist in alignments. Though Snyder is not incorrect in
arguing that states form expectations via estimations of the three alignment incentives and the
past behavior variable,
418
I contend another phenomenon better encapsulates expectation
formation: Trust, contextualized by key aspects of group dynamics. More specifically, the FPE
& PO undertakes a materially and emotionally costly trusting act in order to establish
alignments—i.e., in order to establish expectations. By doing so these men and women give
417
Expectations also underwrite key aspects of the fear of entrapment. This is particularly the case when the state
fearing entrapment experiences insecurity at the thought of its partner acting more boldly and confrontationally with
its enemies, precisely because the partner enjoys expectations of security support from the entrapment-fearing state.
418
As a reminder, the three incentives are degree of commitment, relative dependence, and issue interest.
- 214 -
meaning to the security expectations they create, and therefore to the obligations they expect into
existence.
COOPERATION & CONFLICT
Material cooperation in the collective self-esteem model is a direct product of convergent
perceptions within the alignment of material threats and opportunities in the shared security
milieu. But given the broad emotionality societies enjoy at the transnational level, especially, it
often proves very difficult for the FPE & PO to carry out the material commitment necessary for
producing an adequate alignment response to the emergent security situation. As a consequence,
particularly if there is no immediate existential threat, the FPE & PO in one or both states must
undertake a public trusting act in order to solidify the alignment—i.e., in order to bring along
wavering members of the FPE & PO, as well as much of the broader populace. In effect, those
wanting closer material relations must create the ideational inertia
419
to bridge some of the
material gap created by the ideational differences between the alignment members.
420
Material
fear therefore generates the desire for material cooperation, while ideational fear—i.e., concern
that ideational difference will interfere with material imperative—produces the desire for
ideational cooperation. These processes are shown in Diagram 5.1. It is worth noting quickly
that trusting acts in Japan-Korea relations have been most remarkable for including the giving
and acceptance of history-related apologies.
419
“Intertia” is used here as in the field of Physics. Namely, the FPE & PO must create new ideational “movement,”
which continues until interrupted by something.
420
The ideational gap between states correlates with the difference in their material interests. For example, states
with no ideational divergence whatsoever are already perfectly aligned. Provided they have not yet merged to form
a larger state, they need not carry out a trusting act.
- 215 -
As was the case in both my first and second alignment models, friction emerges in the
collective self-esteem model after divergent perceptions of material threats and opportunities in
the security milieu lead one alignment member to semi-defect from its partner. Instead of
experiencing enmity or anger-fueled vengefulness from its disappointed security expectations,
however, the jilted state suffers humiliation, a powerful ideational social emotion.
Humiliated, the disregarded state endures damage to its collective self-esteem.
421
Friction emerges in the alignment as the jilted state enacts offensive strategies to preserve its
collective self-esteem. Primarily, the humiliated FPE & PO publicly denigrates and disparages
its defecting partner, thereby establishing in its mind its state’s superiority of the other and
therefore preserving its collective self-esteem from the humiliation. Unfortunately, the FPE &
PO’s aggressive ideational defense represents an initiation of ideological conflict within the
421
Collective self-esteem is “a group’s need to feel good about itself, for respect or status.” Wendt, Social Theory of
International Politics, 236.
Diagram 5.1
Security Interests Trigger Alignment Cooperation in a Non-Emergency Material
Environment
X & Y cooperate materially.
The FPE & PO in both X & Y carry out a public,
trusting act in order to overcome ideational
differences.
Aligned states X & Y enjoy convergent security interests due to similar
perceptions of a new threat and/or opportunity in their shared security milieu.
The FPE & PO in both X & Y desire to deepen their material alignment due to
material fear, but fear being hampered by intra-alignment ideational differences.
X & Y cooperate ideationally.
- 216 -
alignment. Material conflict is also possible. There is nevertheless no element of anger and
vengefulness in the collective self-esteem model. Material friction occurs only if the FPE & PO
instinctually understands it can preserve its collective self-esteem by damaging the semi-
defecting state materially. And, of course, material incentives are highly likely to stand in the
way of an aggressive defense in the material realm. Consequently, when states operate via the
logics of the collective self-esteem model, alignment conflict should be most notable for its
ideological nature. Regarding Japan-Korea relations, I argue this is why the history issue not
only remains problematic, but by some measures has actually gotten worse since 1965.
Diagram 5.2
Security Interests Trigger Alignment Conflict in a Non-Emergency Material
Environment
*X will only act in the material realm, if its FPE & PO perceives material action to be an effective way of
preserving X’s collective self-esteem. If so, ideational and material conflict might take place.
X & Y suffer material conflict, with X
being the initiator.*
X primarily suffers ideational humiliation, which
threatens its collective self-esteem.
Aligned states X & Y suffer divergent security interests due to conflicting
perceptions of a new threat and/or opportunity in their shared security milieu.
Y hedges/semi-defects from the alignment, due to material fear/insecurity.
The FPE & PO in X protect X’s collective self-
esteem by publicly denigrating and disparaging Y.
X & Y suffer ideational conflict with,
X being the initiator.
- 217 -
THE SIGNIFICANCE OF NEW EXPECTATIONS
In contrast to my first and second models, the collective self-esteem model is most
remarkable for explaining how security expectations come into being. For reasons I establish
below, new expectations create new actor preferences. But changing actor preferences are not
something the twin material theories can abide. This is the case even for Neoliberalism at what I
call its transnational, third level. Consequently, while I attempted to remain as true to
Neorealism as possible in the delimited balance of threat model, and operated in an opaque
theoretical environment while assembling the common sense social emotions model, I piece
together the collective self-esteem model within Constructivism. For, of the three most
prominent IR metatheories, only Constructivism can tolerate changing actor preferences.
422
In establishing one of the mechanisms through which actor preferences transform, the
collective self-esteem model helps to clarify how states begin internalizing each other’s security
interests—how they begin to develop “we-feeling.” Specifically, states bridge ideational
differences via trusting acts in order to accomplish short-term material objectives. Through this
process states diminish the kind of ideational differences that hinder collective identity formation.
In the event material conditions incentivize more and more material cooperation, the FPE & PO
in the states concerned must initiate trusting acts in order to achieve it. The trusting acts
undertaken by these men and women gradually reduce the potency of the ideological phenomena
differentiating their alignment’s societies. Under the right conditions, special friendships—i.e.,
states sharing a considerable degree of collective identity—develop.
When an alignment’s security milieu is not so cooperative, however, its influence on
alignment identity is deleterious. States like Japan and South Korea, whose bilateral alignment is
regularly pushed together and pulled apart by the material interests I described in Chapter 3,
422
Stable versus changing actor preferences is the key factor differentiating ideological and material theories.
- 218 -
suffer increasingly more divergent identities, or at least identities more divergent than they
would be otherwise. Ideational conflict becomes progressively worse, due to the logics of the
cycles of materially-driven ideational cooperation and conflict. Most importantly, via a
mechanism I explain below, these cycles sensitize aligned partners’ societies to the ideational
issues differentiating their states. This outcome greatly damages whatever collective identity a
given alignment may have developed while material interests incentivized cooperation. The end
result, after multiple rounds of trusting acts and societal humiliation, is an alignment fairly bereft
of we-feeling, one driven almost solely by the states’ respective material interests of the moment.
Before considering the above processes more carefully, I turn one last time toward
expectations. I particularly strive to illuminate how expectations in the real world tend to be
temporally rich, rather than temporally static, and why the difference matters. I argue current
alignment expectations usually inform very little about the level of emotionality alignment
members experience, unless one enjoys some understanding about the extent of expectations
prior to a given moment. Yet Neorealist scholars often treat alignment expectations as if they are
primarily static, a phenomenon with no past. I contend temporally static treatment of security
expectations explains why the scholars ignore or downplay the importance of ideational social
emotions in alignment relations. Highlighting the longitudinal nature of most alignment
expectations in the modern world in-and-of-itself draws one to ask questions about whether an
array of emotions much broader than material fear and material insecurity must be in play
following the satisfaction or disappointment of this critical social phenomenon.
- 219 -
II. Static Versus Temporally Rich Security Expectations
A. Surprise and Uncertainty
In all International Relations theories and predictive models predicated on the causal
power of emotions, the significance of affect is equaled only by the importance of
expectations.
423
For, within IR, there are surely very few exceptions to the rule that emotional
reaction requires expectations.
424
It is possible to understand the veracity of this statement by
exploring two emotions commonly experienced in the international relations of states, and
regarding which one might reason broad emotions can emerge without expectations: surprise and
uncertainty.
One definition within Webster’s online dictionary defines “surprise” as, “to strike with
wonder or amazement especially because unexpected.”
425
There are, ostensibly, two reasonable
ways to interpret the word “unexpected.” The first is the notion that unexpected equates roughly
with having no expectations. The second, in contrast, requires expectations to be present for the
unexpected to occur. I contend the latter definition is the superior way of considering the
phenomenon of the unexpected.
423
My discussion of emotion as a phenomenon that occurs in response to expectations in IR theory obviates the need
to talk about the existential roots of emotions. Discussion is unnecessary, because I start with the argument that
emotions are extant; I accept them as given. Consequently, I do not need to argue the biological and evolutionary
roots of emotional capability. I need only to deduce why and when members of the FPE & PO will experience the
emotions they have been biologically and evolutionarily primed to feel. My specific answer is members of the FPE
& PO feel emotions in response to their expectations of others, and of themselves in relation to others. This
argument makes sense given the wide acknowledgement that “the most common cause of emotion is some kind of
social event,” and since practically all events in international relations are social. See Hareli and Parkinson, “What’s
Social about Social Emotions?” 131. For a brief but good summary of Sociology research on roots of emotions, see
Turner and Stets, “Sociological Theories of Human Emotions,” esp. 46-48. See also Parkinson, “Emotions are
Social.”
424
Exceptions surely include emotions produced through physical processes of the body. As is commonly known,
chemical imbalances in the brain can produce feelings of depression and anxiety that are unrelated to expectations,
for example. Hence certain analyses of leaders may reasonably focus on emotions related to brain function as being
causal before expectations are formed.
425
“Surprise,” verb def. no. 3, http://www.merriam-webster.com/dictionary/surprise, accessed November 5, 2012.
- 220 -
The difference between the two positions might appear subtle, but the contrasting
outcomes they permit impact theory greatly. Consider, for example, the fact that those who can
experience surprise when they have zero expectation can also feel fear and insecurity, and any
number of other emotions, in reaction to their surprise. Assuming that unexpected equates with
no expectation consequently implies that claims linking extant expectations to emotional
outcomes might often be spurious.
Though seemingly reasonable, arguments equating unexpected with zero expectations are
based upon a false equivalence. Namely, saying something is unexpected does not, in actuality,
indicate the existence of a vacuum in which there is no expectation. Rather, it points to the
existence of incorrect expectations. In other words, the unexpected takes place precisely because
other expectations are held. Emotions like surprise emerge when those expectations are proven
incorrect by reality, or at least via perception of reality.
Consider the case of a man being led into a bar by his wife for the surprise birthday party
she planned, and to which all of his best friends are secretly in attendance. Presuming he is not
unconscious and is mentally competent, the husband will have expectations about the barroom,
no matter whether he has been there before.
426
The surprise he experiences as he recognizes his
friends and the nature of the event will be felt only because he expected a typical barroom and
evening out with his wife. If, however, just before entering the bar, the man imagines that his
friends might be just inside the door—i.e., he develops some expectation about seeing his
friends—his surprise will be diminished by the extent he expects to see them. Similarly, should
he intuit with complete certainty that his friends were waiting for him, he would have to feign
surprise because he would not actually feel any. He would be surprised in this latter case,
426
If he has never patronized the establishment, knowledge garnered directly from similar experiences, or indirectly
via the shared experiences of others, will inform his expectations. Direct or indirect exposure to media likewise can
influence what he anticipates.
- 221 -
however, if his wife mistakenly told her husband’s friends to go to the wrong bar, and all the
man saw upon opening the bar door was a typical, anonymously populated room. Seen in this
light, for clarity one might rewrite Webster’s definition as, “to strike with wonder or amazement
because one’s expectations are proven false by events.”
The case is no different when dealing with states in international politics. If a state is
surprised by a sudden shift in allegiance of an ally, it can only be because the state had
expectations counter to the betrayal. Likewise, if the state expected complete abandonment from
its “ally,” surprise on this matter could only take place if the betrayal never takes place or is
otherwise incomplete. Surprise consequently rests on incorrect expectations. So do any
emotions—such as material fear and insecurity—following surprise.
Uncertainty represents the other puzzle to be resolved. If one experiences total
uncertainty does one not lack expectations? Likewise, in international politics, if a state becomes
completely uncertain about its ally’s level of support, is that not the same as saying it has no
expectations regarding whether its ally will help? The answer to both questions is no, and rests
on the deduction that individuals and states are reliant on past experience—either their own,
directly, or from others, indirectly—when they form expectations.
427
Another way of saying this
is that feelings of certainty or uncertainty can be formed only in response to expectations about
what comprises satisfactory levels of knowledge necessary for predicting outcomes confidently.
Consequently, when members of the FPE & PO tell each other they have no idea whether an ally
will be supportive, what they are really saying is they are experiencing the emotion of
uncertainty in response to their expectations about what complete uncertainty is. And when they
communicate certainty about an outcome, they are actually expressing their emotional response
to their expectation of what certainty feels like. These arguments make sense within the context
427
The domestic realm is one obvious source of these expectations. Hopf, Reconstructing the Cold War, 13.
- 222 -
of the overall point that practically all emotion emerges in response to met and unmet
expectations.
Returning to surprise, a state can indeed be shocked when it is completely uncertain
about whether an event will occur. But, like the man who imagined the possibility that his
friends were just beyond the barroom door, its level of surprise will be diminished by the amount
it thought a particular outcome possible. If, for example, leaders of state conclude they are 50
percent certain an ally will act supportively in a given instance in which the ally has only two
options, the level of surprise they experience upon betrayal will be much lower than if they had
been 100 percent confident about the ally’s assistance. The level of surprise experienced by
members of the FPE & PO is consequently related to the certainty of expectations and the
number of outcomes perceived to exist. Importantly, if the ally somehow responds via an
unimagined third option in the above scenario, the surprise experienced by the leaders will be
produced by their incorrect expectation that only two outcomes were plausible.
B. The Logical Consequences of Expectational Staticity on Abandonment Fear
Why, one might reasonably ask at this point, does it matter whether emotions emerge in
response to met and unmet expectations? I emphasize it primarily because those who fail to
understand the role of expectations in producing emotions are much more likely to treat emotions
as if they are a product of temporally static forces, instead of expectations satisfied or dissatisfied
through longitudinal social processes. Scholars who perceive staticity suffer a tendency to
overlook the deeply social nature of interstate security relationships. Staticity, in other words, is
remarkable for privileging biologically-founded emotions such as material fear and insecurity,
and rendering as far less important more obviously social emotions such as ideational anger,
- 223 -
vengefulness, gratitude, and humiliation.
428
Static interpretation consequently privileges
materialist thinking about which emotions are important.
SNYDER’S TEMPORAL DILEMMA
Consider Snyder’s work on the alliance security dilemma, which heavily influences both
Cha’s quasi-alliance model and Yoon’s net threat theory, as well as my own thinking. Though
Snyder obviously understands the social and temporally vital nature of expectational
development, his theoretical discussion of fear of abandonment and fear of entrapment remains
strongly focused on the security interests of a given moment, as if the past were not a significant
component.
429
As I contend below, Snyder’s failure to emphasize adequately the temporally
relational nature of security expectations and material fear and insecurity has influenced those
who have adapted or adopted the alliance security dilemma. Most troubling, in the cases of Cha
and Yoon, I argue it explains to a considerable extent why they overlooked Snyder’s past
behavior variable as a force greatly weakening the strong expectations they require to be present
for Seoul, especially, to experience high levels of abandonment fear.
My claim about Snyder merits some examples. Here are three of his situational
arguments: 1) In the event allies are threatened by different enemies, “then both the likelihood
of abandonment and the cost of entrapment will be high. The allies will simultaneously be
skeptical about each other’s loyalty and anxious lest they be trapped into” rescuing the other at
great cost.
430
2) “Vague or ambiguous alliance agreement, such as an entente, tends to maximize
428
For an introduction to biological causes of emotions, see Turner and Stets, “Sociological Theories of Human
Emotions,” 46-48. For social emotions, see Hareli and Parkinson, “What’s Social about Social Emotions?”
429
Compare Snyder’s portrayal of expectations versus his description of fears of abandonment and entrapment. In
contrast with the temporal depth of the former, the latter tends to be explained solely in terms of the security
concerns of the moment. For expectations, see Snyder, Alliance Politics, 6-11, 188-189. For the alliance security
dilemma, see Ibid., 180-186.
430
Ibid., 187.
- 224 -
fears of abandonment, but it minimizes fears of entrapment.”
431
And, 3) “If the ally’s apparent
commitment declines, the state will become more concerned about abandonment and may move
closer to the ally to discourage its defection.”
432
Significantly, in the first two situations, there is no sense that expectational change
brought about insecurity. Indeed, the arguments as stated seem to apply equally to states that
once enjoyed solid expectations of security support and now do not, as well as to states sharing
no previous alignment—and therefore no prior expectation of security support—yet suddenly
finding themselves allied against enemies and in situations one and/or two above. This equality
is very unlikely, however. In fact Snyder is clear on this point when discussing abandonment
fear, saying “the cost of being abandoned…is the difference between the security experienced
with the alliance and the security that could be obtained in an alternative alliance or no
alliance.”
433
Or, to put it another way, in order to suffer the most severe abandonment fears,
states must first enjoy substantial expectations of material support.
434
For, without security
expectations of their allies, states cannot experience the feeling of being secure in alliances in the
first place.
Reflect on the ramifications of two radical cases: If two states facing situations one or
two above suddenly allied, and from the very beginning one ally was certain it would receive no
support from the other, it would experience zero abandonment fear.
435
Similarly, if the same
suddenly-allied state believed there was only a one percent chance of getting assistance from its
431
Ibid., 188
432
Ibid., 189-190.
433
Ibid., 183.
434
Snyder clarifies why disappointed security expectations lead to fear. They are reducible to his argument that
readjusting security policy is costly. Ibid., 321.
435
With no expectations of support the ostensible allies would actually not be allied, by definition. See Ibid., 6.
- 225 -
brand-new, ostensible ally, then its abandonment fears would be very, very low, because it would
not have put much stock in receiving help from the very first moment of alignment.
Contrast these two examples with a state now suffering diminishing expectations across
time: A state that once expected considerable alliance support, with high confidence, suddenly
discovers its ally is not reliable. Moreover, its security analysts predict on current trends there
will soon be a zero or one percent chance of getting material assistance from it if no
counteraction is taken to bolster the alliance. While the alignment member in the above
paragraph experiences no, or exceptionally little, fear of abandonment, the state currently being
examined suffers tremendous insecurity, as important material support it requires to feel secure is
being quickly withdrawn. Importantly, this incredible difference in levels of fear is uncovered
only through a focus on expectations as they change longitudinally.
In order to further clarify the significance of temporality and expectations it is a useful
exercise to consider another hypothetical circumstance in which two states with no prior security
relationship and, therefore, no prior security expectations suddenly ally. Quick logic informs
they most likely joined forces because both feared material power external to their alliance.
436
Taking the scenario further, though the threat they face is apparent, it is not imminent; there is
time to prepare. But the new allies suffer limited options for increasing their security, as they
have no other states to turn to for support. Consequently, each ally’s only real alternative to
relying on the material power of the other lies in transforming its adequate but latent aggregate
power into deployable military power. Also remarkable is the allies’ exactly equal level of fear
regarding the threat(s) they face. Nevertheless, after considering Snyder’s three incentives and
past behavior variable, the FPE & PO in each allied state separately conclude there is exactly a
436
It does not matter in this discussion whether they each face the same threat, or whether each faces a different
security challenge.
- 226 -
50 percent chance that its materially adequate partner will provide sufficient security support to
prevent disaster. In both states, this calculation takes place instantly, just at the point of
alignment.
Given their expectations of security support, it seems reasonable to declare the states
suffer abandonment fears immediately after allying. This type of situation, therefore, meets the
criteria represented by the overall trend of Snyder’s temporally static theoretical discussion of
state behavior under the alliance security dilemma. It appears also to represent the maximum
amount of abandonment fear possible in the ahistorical account laid out above. Insecurity is at
its zenith, because static situations containing more than 50 percent likelihood of support induce
more confidence that support is forthcoming. And more confidence means less fear of
abandonment. Indeed, abandonment fear gradually falls to zero as the FPE & PO comes to
realize with 100 percent certainty that alliance support is assured. Similarly, in the above static
scenario, perception of a likelihood of support less than 50 percent results in lower abandonment
fear, because the FPE & PO in each state responds to its ally’s unreliability by relying less on its
ally’s questionable intentions. Instead of fearing abandonment, the new allies instantly imagine
plans to increase domestic military capability. Given each state’s adequate latent capability, fear
of abandonment in fact declines gradually to zero, as observers increasingly recognize there is no
chance of obtaining alignment support. The resulting bell curve of insecurity is approximated in
Figure 5.1.
- 227 -
The bell curve is the product of the ability of members of the FPE & PO to substitute
mentally, at the moment of alignment, their state’s capability of augmenting its military power
for abandonment fear. In other words, if an instant ally had no, or inadequate, latent defensive
capability, there would be no bell curve. Extending the scenario, it seems reasonable to assert
that all states lacking the capability of carrying out self-help within an uncertain alignment and
of securing new support from without the security partnership, experience the most extreme
abandonment fears. As their expectations dip to the lowest point—say a one percent certainty of
their ally fulfilling its security obligations—fear of abandonment peaks.
437
At the confident end
of certainty, in contrast, fear of abandonment falls to zero. In other words, fear of abandonment
correlates negatively with the level of certainty about the alignment. I approximate these logics
in Figure 5.2.
437
I use one percent here instead of zero percent, because states with no expectations of security support remain
technically unaligned. Of course, unaligned states cannot suffer abandonment fears in temporally static alignments.
- 228 -
Figure 5.2 is remarkable for being the best approximation of Snyder’s theoretical
discussion of abandonment fear. Considering the difference between the first figure’s bell curve
and the latter figure’s straight, diagonal line, it is clear that Snyder is primarily concerned with
the theoretical logics of static alignments in which states suffer no, or very inadequate, capacity
for self-help beyond the immediately formed alliance.
Intriguingly, though the most intense abandonment fear is demarcated by the identical
score of 1.0 in both of the above figures, there is no doubt that the peak fear of abandonment in
Figure 5.2 is much more intense than that experienced in Figure 5.1. The maximum fear
represented in Figure 5.1 is lower, because those who undergo its logics are not completely
reliant on their alignments for security. In Figure 5.2, in contrast, they are absolutely reliant on
their partner(s). Hence the terrifyingly intense fear of abandonment as expectations of support
collapse to near zero.
Adding temporal richness to the calculus changes the logics behind fear of abandonment
tremendously. Whereas under staticity fear of abandonment is best represented by a bell curve
- 229 -
when self-help and/or other alignments are possible, and a diagonal line when impossible,
accounts considering both new and old certainties of security expectations are approximated only
by a diagonal line. The difference emerges because partners in temporally rich alignments
undergo expectational change across time, and the level of their current fear of abandonment
depends on the difference between current and old expectations.
As illustrated in Figure 5.3, an ally who once expected alignment support with absolute
certainty, but now perceives there will be practically none, experiences the maximum increase in
fear of abandonment. Conversely, an ally who once felt there was basically no chance of
obtaining material help within its alignment, but now enjoys absolute certainty that alliance
obligations will be fulfilled, experiences the greatest possible decrease of abandonment fears.
Importantly, in the event expectations of security support do not change over time, then there is
no difference in the level of abandonment fears experienced. The overall logic of these three
trends holds, no matter whether a state enjoys latent self-help or new alignment options.
- 230 -
There are several points to take away from the three figures above. First, states in static
alliances—i.e., in alignments with no history—do not suffer the most intense abandonment fears
unless they have inadequate capability for self-help beyond the given alignment. This is clear
because, as noted earlier, states extinguish their abandonment fears when possible.
438
Those
states which do enjoy other options in terms of adequate time, and external alignment
alternatives and/or ample latent power, do not excessively fear abandonment, because they make
immediate adjustments at the moment of alignment to compensate for what would have been
their heightened sense of fear.
Second, there is a major difference between the nature of abandonment fear in static
alignments versus in longitudinal ones. When a state in a static alignment enjoys adequate self-
help capability, the adjustments it makes in response to uncertainty—i.e., to compensate for its
would-be fear of abandonment—impart no extra costs beyond that required if it did not join the
alignment in the first place. Extra costs do not exist under staticity, because all calculations are
made exactly at the moment of alignment. In contrast, the otherwise identical state in a
temporally real alignment suffers actual costs of adjusting the security policies it implemented in
response to once-held alignment expectations. And the more integrated an alignment, the higher
these costs will be. Damages can be especially high if, via the logics of balance of power, the
once-satisfying alignment ended up producing enemies among tertiary states that either would
have remained neutral or even become allies.
439
Intriguingly, since perceptions of costs are the wellspring of abandonment fear in both
static and temporal alignments, one can generally claim the potential intensity of fear of
abandonment is higher in temporally rich alignments. One exception to this rule occurs when a
438
Ibid., 180-183, 321.
439
Snyder also lists a couple of immaterial reasons why costs are high in temporally rich alignments. But they are
not compatible with Neorealism, and I leave them out for now. See Ibid., 321.
- 231 -
state in a static alignment suffers inadequate self-help capability and understands it enjoys no
chance of securing more support via new or old allies. Unable to reduce its fear of abandonment,
it must endure it.
Returning to the three situational arguments laid out above, in the third instance Snyder
does exhibit a sense of temporal transition and longitudinal causality. But, even in noting the
apparent decline in commitment and the simultaneous concern it produced, Snyder makes no
mention of changing expectations within the observing state. That is, despite placing
expectations at the definitional center of alliance and alignment, Snyder presents changing
alliance behavior as if it, itself, is producing the observer’s fear. This style of arguing imparts a
sense that fear is produced completely outside of the observer’s head instead of inside it. This is
not the case, however, as the actual process is far more interactive. To summarize: The
observer’s fear is produced by the diminishment of its initial expectations in response to its
perception of the decline in its ally’s willingness to honor old obligations.
Snyder might tend to leave out expectations in his game theoretical discussion of
abandonment and entrapment for two reasons. First, constantly mentioning security expectations
is no doubt tedious and tiresome for both writer and reader. Moreover, since security
expectations are certain to be actualizing the abandonment fear Snyder discusses, it feels
unnecessary to mention them repeatedly. In this sense Snyder’s style of writing about
expectations is very similar to those, like myself, who perceive importance for material fear and
insecurity, but who never emphasize equally the critical emotionality component of the desire to
mitigate material fear and insecurity. Doing so on an ongoing basis would in most cases feel
superfluous and bothersome.
- 232 -
Second, Snyder may have underemphasized expectations because Neorealism coexists
with the phenomenon only with great unease. Downplaying security expectations might
therefore have been practically inevitable. As I argued in Chapter 2, Neorealism seems to
consistently permit only expectations about the meaning of power. Expectations of others within
Neorealism are specific to the type of Neorealist system. I deduced, for example, that states in
ideal Defensive Neorealist worlds are unlikely to be capable of confidently expecting anything of
others. Neorealism as a whole certainly leaves little scope for states to discern with assurance
the future strategies of others, based on their current policies and what they otherwise say they
will do. For instance, in his field-changing Defensive Neorealist treatise, Waltz argues
“outcomes cannot be inferred from intentions and behaviors…because structures select by
rewarding some behaviors and punishing others.”
440
Likewise, one of Offensive Neorealist John
Mearsheimer’s core theoretical assumptions is “states can never be certain about other states’
intentions.”
441
Snyder’s inclusion of alignment expectations within a general Neorealist framework is
accordingly somewhat awkward.
442
The mismatch is not unexpected, however: It was
foreshadowed by the fact that none of Snyder’s determinants of alignment expectations—the
three incentives as well as the more critical past behavior component—are completely material
in nature.
CHA’S AND YOON’S STATIC & ASSUMED EXPECTATIONS
While including the critical influence of security expectations on an ongoing basis may
be tedious, tiresome, and feel unnecessary, as well as be awkward within Neorealism, Snyder’s
440
Waltz, Theory of International Politics, 73-74.
441
Mearsheimer, The Tragedy of Great Power Politics, 31.
442
In earlier alliance politics research Snyder is a self-proclaimed Neorealist. See Glenn H Snyder, “Alliance
Theory: A Neorealist First Cut,” Journal of International Affairs 44, No. 1 (Spring 1990): 103-123.
- 233 -
reasonable inclination to deemphasize expectations in key parts of his research appears to have
contributed toward a disappointing outcome. Namely, Neorealist scholars like Cha and Yoon
who have been eager to adapt Snyder’s abandonment-entrapment mechanism seem not to have
considered seriously the critical role of expectations in producing the emotional reaction their
models require. The scholars therefore never investigate methodically whether the states they
study ought to perceive the kind of alignment security expectations they regularly enjoy or suffer.
In other words, they assume the presence of expectations.
Both Cha and Yoon require Korea, especially, to repeatedly experience fairly powerful
security expectations of Japan. The presence of this ideational phenomenon is clear, because the
emotional reaction—the fear of abandonment, or betrayal and enmity—following the
disappointment of Korea’s alignment security expectations is solid proxy evidence for the
existence and strength of these expectations.
443
To clarify, how could Seoul experience
alignment-altering levels of fear of abandonment, or betrayal and enmity, if it never enjoyed
strong expectations of Japanese security loyalty in the first place? And if the Koreans could not
recognize what it felt like to perceive that Tokyo’s security support would be adequately robust
and forthcoming, how could Japan’s supportive actions ever assuage the ROK’s fear of
abandonment, or soothe its sense of betrayal and thus enmity? In sum, reliance on Snyder’s
abandonment-entrapment mechanism means both Cha and Yoon require powerful expectations
of security loyalty to exist for their respective models to work.
What does a more expansive reading of Snyder’s alliance politics research suggest about
the probability of the existence of Korean expectations? More specifically, if one moves past
Snyder’s Neorealism-friendly discussion of the alliance security dilemma, what do his arguments
443
My argument is an extension of Mercer’s: “One way to test for the presence of norms is to look for emotion; it
seems unlikely that we would find one without the other. If norms play an important role in relations between states,
then we should be able to see emotion when putative norms are violated.” Mercer, “Approaching Emotion,” 11-12.
- 234 -
imply for expectations in Seoul-Tokyo alignment relations? Unfortunately for both Cha and
Yoon, Snyder’s discussion of expectations formation via the three alignment incentives—the
degree of commitment, relative dependence, and issue interest—and the past behavior variable
suggest that Korea is exceptionally unlikely to develop strong alignment security expectations of
Japan.
444
This means that Cha and Yoon both rely on very weak expectations to induce strong
emotional reaction—an impossibility.
One reason Seoul has little reason to expect much of Tokyo stems from the fact that
Japan and South Korea have never formally established their alignment.
445
While formality is
not critical to alignment formation, Snyder clarifies that it does help states develop more robust
security expectations. Second, thanks to geography and history Seoul and Tokyo often have very
different issue interest. Most significantly, the ROK has not been alarmed to the extent of Japan
by North Korea’s missile and nuclear weapons programs, and Japan has not suffered a long-term
direct threat from Pyongyang’s conventional forces.
446
The Seoul-Tokyo alignment has
accordingly tended to suffer from divergent perceptions of the North Korean threat.
Third, Japan and South Korea’s relative dependence—i.e., Japan’s much more substantial
capabilities over the ROK—is problematic. While Snyder’s argument informs that this dynamic
magnifies Seoul’s abandonment fear, such a possibility seems highly probable only in temporally
static alignments. In longitudinal security partnerships with all other things being equal,
however, relative dependence on the Japan-South Korea scale seems more likely to prevent
expectations formation in the first place, precisely because Seoul should rationalize that Tokyo’s
considerable military and economic advantages within their alignment will tend to make it an
444
See my discussion in Chapter 3, especially, and also Snyder, Alliance Politics, 186-189.
445
I limit my discussion here to South Korea’s security expectations, because it is on the losing end in both Cha’s
and Yoon’s models. That is, Seoul is the source of historical animosity in both cases. Hence, its alignment
expectations are more critical when discussing the work of these two scholars.
446
See my overview of interests in Chapter 2, Section 3.
- 235 -
unreliable security partner. If so, the ROK would have a hard time developing the depth of
security expectations required to realize extreme fear of abandonment from Japanese hedging in
the first place.
447
After all, alignment obligations are expected into existence. And one cannot
suffer fears of unfulfilled obligations, if one never truly perceives the obligations in the first
place.
If the three alignment incentives imply Seoul’s alignment expectations of Tokyo should
be minimal, application of Snyder’s past history variable suggests they will be practically
nonexistent. Indeed, the very objective of Cha’s and Yoon’s precise alignment modeling is to
explain how material interests produce the wild swings of cooperation and conflict characterizing
modern Japan-Korea relations. Since the zag following each zig represents moments in which
the Koreans perceive considerable Japanese “disloyalty” following interest divergence, how
could any rational Korean analyst of the alignment believe and espouse anything other than a
finding that Japan could not be trusted to honor its contemporary alignment obligations? This
tendency should have been bolstered considerably by what preceded normalized bilateral
relations in 1965. Specifically, Japan’s history-related whitewashings and denials in the two
decades preceding normalization,
448
the fact of Japan’s colonization of Korea itself, and the not-
forgotten Hideyoshi invasions of Korea in the late 1500s should have weakened any Korean
tendency to expect that Japan would be a reliable security partner.
The above discussion reveals that Cha and Yoon face a regrettable logical paradox: They
are either asking very weak expectations to produce remarkably strong emotions, or are utilizing
strong expectations within the logics of a theoretical framework predicting only minimal
expectations in the case of Seoul-Tokyo relations. To restate the point more clearly, Cha and
447
All information on the three incentives under temporal stasis is from Snyder, Alliance Politics, 186-189.
448
For the best overview of this phenomenon, see Lind, Sorry States, 32-47.
- 236 -
Yoon utilize scholarship contraindicated to the task for which they apply it. Had they recognized
the nature and role of security expectations in longitudinal alignments, they might have realized
the necessity of working far outside of a materialist framework and, therefore, of radically
changing their models.
Given that Snyder’s research predicts overwhelmingly that Korea will suffer the
insecurity of weak security expectations, it makes sense to conclude that both Cha and Yoon
operate within the paradox of weak expectations producing strong emotional outcomes. I
illustrate this process in Diagram 5.3. For comparative purposes I show how their respective
models should operate under the logics of longitudinal temporality, thus revealing that both
scholars are thinking in terms of staticity.
- 237 -
Differing temporal ontologies lead to divergent outcomes. In longitudinal alignments, as
seen on the right side of Diagram 5.3, Korea’s weak expectations of security support influence
the way Seoul perceives its alignment and security milieu. Understanding Tokyo is unlikely to
provide adequate support, the Koreans have no reason to count on it being provided. As a result,
Korean fears of the security milieu decline very little. But never relying on massive Japanese aid
Diagram 5.3
Weak Expectations and Abandonment Fear in Temporally Static versus
Longitudinal Alignments: the Case of Japan and South Korea
*The exact reason is not critical.
**In Yoon’s temporally static model, Korea only cares greatly about its expectations of Japan when net threat is
rising. That is part of the reason why there is no cooperation when net threat is falling.
Cha & Yoon (Actual): Staticity
Cha & Yoon (Hypothetical): Longitudinal
Korea’s expectations of security support are weak,
and are formed at the moment of alignment.*
Japan and Korea align, because they fear a material
threat in their shared security milieu.
Korea’s milieu
fears decrease by
the amount of
security Japan
would ideally
provide—i.e., a
lot.
Korea suffers intense
abandonment fear from
the moment of
alignment, because it
expects Japan will not
support Korea to the
extent Seoul requires to
feel secure.**
Alignment suffers considerable friction (Cha) or
cooperation (Yoon) due to abandonment fear.**
Korea’s expectations of security support are weak, and
are formed at the moment of alignment.*
Japan and Korea align, because they fear a material
threat in their shared security milieu.
Korea’s abandonment fear is all but
nonexistent at the moment of
alignment, because it has not yet
carried out any real-world
adjustments in response to the
alignment. Even after making
adjustments over time, Seoul suffers
only weak abandonment fear,
because the Koreans never plan to
receive from Japan a massive aid
effort they do not expect.
Alignment suffers unsubstantial emotionality due to
abandonment fear.**
Korea’s
milieu
fears
decrease
by the
amount of
security
Seoul
expects
Tokyo to
provide—
i.e., a little.
- 238 -
also means Seoul has little reason to fear being abandoned by Japan. Why, after all, should the
Koreans be especially fearful about not receiving aid they never planned to get?
In Cha’s and Yoon’s comparatively static models, as illustrated on the left side of
Diagram 5.3, weak Korean expectations continue to influence how Seoul views its alignment and
security milieu. But under temporal staticity they produce strong emotions from the very
moment of alignment, because Korea becomes instantly dependent on Japan’s support. That is,
despite lacking the time to adjust its own capabilities to better work with Tokyo’s, Seoul is
somehow immediately and remarkably dependent on Japan.
Instant dependence and the consequent intense emotionality are remarkable in three ways.
First, they suggests either Seoul has no other options for achieving its security goals outside of
reliance on Japan, or the Koreans lack the time to achieve them. As noted earlier, adequate
alternatives mitigate fear of abandonment. Second, emotional dependence requires logically that
Korea believes its support from Japan will be broad and robust. The amount of aid Seoul hopes
to receive is therefore “idealized,” meaning the positive correlation between the strength of
security expectations and expectations regarding the amount, type, and quality of aid
disappears.
449
With expectations delinked, Koreans judge that Japan will provide the kind of aid
Seoul requires, despite Korea’s minimal alignment expectations. And third, the locus of the
ROK’s material insecurities accordingly transfers from the wider milieu to specific questions
about Japan’s alignment loyalty. Suffering weak expectations while enjoying idealized hopes of
security support, Seoul’s powerful emotionality is directed at Japan rather than the milieu threat.
449
Logically, Korea must idealize Japan’s alignment support. Otherwise Seoul could not suffer intense
abandonment fear in a static model comprised of weak expectations. Weak expectations under temporally staticity
mean Korea would continue to fear the milieu threat, but would have little reason to fear abandonment by a security
partner perceived to be unlikely to provide the desired support.
- 239 -
How did Cha and Yoon miss the weak expectations-strong emotions paradox? One key
reason, as argued above, is found in the nature of Snyder’s discussion of the abandonment-
entrapment mechanism. Particularly, Snyder’s marked predilection for underemphasizing the
role of security expectations in contributing to abandonment and entrapment fears, coupled with
his penchant for treating those expectations as static rather than dynamic, greatly increased the
odds that scholars adapting his work on the alignment security dilemma would ignore his
arguments regarding expectation formation. The other most likely reason is simply that strong
security expectations are obviously present in the relationship, at least occasionally. Their
presence and strength is clear, because without them Korea could never experience intense fear
of abandonment, or betrayal and enmity, in its alignment with Japan.
C. Expectations and Actor Preferences
Contrasting with the marked tendency of treating security expectations as temporally
static phenomenon, I emphasize expectations as being longitudinal in the collective self-esteem
model. In other words, I perceive as most critical in alignment politics expectational change
across time. In particular, the longitudinally varying strength of those expectations is what
matters most since, in alignment politics, the strongest emotionality is normally experienced only
when powerful prior expectations are disappointed or satisfied by contemporary phenomena.
Focusing on longitudinal expectational change also informs an arguably even more critical issue
in IR: The question of whether actor preferences are stable or vary across time. I argue below
that expectational change and actor preferences are interlocked. Studying expectational change
as it occurs through social interaction, accordingly informs not only that actor preferences
transform but also clarifies one of the mechanisms through which socializing brings about new
preferences.
- 240 -
THE ROLE OF PRIOR EXPECTATIONS
Importantly, expectations cannot be frustrated without degradation of one’s prior
expectations. When either Japan or Korea perceives the other will not honor some particular
security obligation, its alignment security expectations weaken. This expectational shift occurs
no matter whether the expectations-disappointing state actually ever promulgates the
expectation-damaging policy or policies. Conversely, if either Seoul or Tokyo enacts a strategic
shift proving eventually to be deleterious to the other’s security, emotionality cannot result until
the partner realizes it fully—i.e., until it adjusts its expectations in response to its perception of
the facts.
As illustrated in Diagram 5.4, critical aspects of Cha’s and Yoon’s models can be
portrayed accurately by asserting strong initial alignment expectations and longitudinal ontology.
Regarding Cha, as originally strong Korean expectations are shifted downward, Seoul begins to
fear abandonment acutely. When Yoon’s Koreans likewise realize Japan cannot be counted to
meet Korea’s security needs under rising net threat, the ROK’s old expectations of support
weaken and severe abandonment fears arise.
450
Thus, in contrast with Diagram 5.3, the primary
difference is that Korean expectations start strong as the alignment is actualized, and degrade
only after Japan hedges or semi-defects. These two changes result in drastically different
mechanisms. Critically, whereas the ROK’s abandonment fear was immediate under temporal
staticity, the Koreans do not suffer it until their expectations are disappointed. And while
Seoul’s abandonment fear could only be weak in the hypothetical longitudinal case in the right-
450
Lest Yoon’s increased milieu threat appears to represent an instance in which the ROK’s abandonment fear
emerges only in response to heightened security risk rather than weakening expectations across time, it is worth
noting the hidden hand of changing expectations. Namely, Seoul’s once-strong expectations of alignment support
are now perceived by Koreans to be unrealistic; Japan has become unlikely to live up to the obligations the ROK
once expected of it. In contrast, under temporal staticity and as noted in Diagram 5.3 earlier, the Korean’s suffer
weak security expectations constantly. But they do not care, because net threat is falling.
- 241 -
hand column of the earlier diagram, in the latter its abandonment fear can be experienced
powerfully as expectations deteriorate.
451
451
The emotionality experienced should reflect to a considerable extent the difference in the prior and new set of
alignment security expectations.
Diagram 5.4
Expectational Change and Abandonment Fear in an Ongoing Alignment: the
Case of Japan and South Korea
*As argued earlier, Cha and Yoon cannot explain why Korea initially enjoys strong alignment expectations.
**In Yoon’s model, Seoul only fears abandonment when net threat is rising. When net threat falls, the ROK’s
emotional response is betrayal-fueled enmity.
Cha and Yoon (Actual): Longitudinal
Korea suffers severe emotionality as it downgrades its expectations of security support.
Material interests lead Japan to hedge/semi-defect from its alignment with Korea
(Cha), or at least make it much more likely (Yoon).
Korea suffers heightened abandonment fear and
initiates alignment conflict (Cha) or shows
alignment loyalty (Yoon).**
Korea’s milieu insecurity increases by an amount
related to the degradation of its alignment
expectations.**
Korea’s expectations of security support are strong, and are formed at the moment of alignment.*
Japan and Korea align, because they fear a material threat in their shared security milieu.
Korea’s abandonment fear is all but nonexistent at the moment of
alignment, because it has not yet carried out any real-world
adjustments in response to the alignment. Even after making
adjustments over time, Seoul suffers only very weak abandonment
fear, precisely because its security expectations are considerable.
Korea comes to realize its prior security expectations were unrealistic.
Korea’s milieu fears
decrease by the amount of
security Seoul expects
Tokyo to provide—i.e., a lot.
- 242 -
What Cha and Yoon cannot explain are Korea’s strong expectations, formed as the
alignment comes into being. As argued earlier, considerable security expectations must be
present. Their presence is apparent because, otherwise, neither Cha nor Yoon can explain
Seoul’s powerful emotional responses to Japanese hedging. The existence of strong Korean
expectations—so clearly counter-indicated by Snyder’s discussion of expectation formation and
yet absolutely critical—is the most significant puzzle Cha and Yoon overlook. It is from the fact
of the irrationality of Korean expectations under materialism that one must shift irrevocably to
ideationalism and the full panoply of expectations and emotions it permits.
THE MEANING OF CHANGING EXPECTATIONS FOR MATERIALISM
Grounding alignment politics within materialism is exceptionally difficult and might even
be impossible. Throughout my research, I have demonstrated how the ostensibly materialist
scholarship I examined most carefully is highly remarkable for its broad utilization of ideational
phenomena to explain critical causal mechanisms and outcomes. These potent immaterial forces
range from weaponized ideas capable of compelling states to go against their material interests,
to the notion that ongoing ideational conflict between states is so critical in establishing intra-
alignment bilateral threat perception that one need not seriously examine the influence on threat
perception of contemporarily changing and far more compelling material threats and
opportunities in the alignment’s broader security milieu. Indeed, most significant in the
scholarship is the fact that ideas frequently overpower material interests within alignments.
While materialism itself cannot explain alignments adequately, such a finding cannot
immediately eliminate Neoliberalism as a framework for modeling alignment relations. As I
noted in chapters 2 and 4, Neoliberalism might enjoy the capability of grounding alignment
politics models at what I call its second and third levels. Within its second level, for example,
- 243 -
provided the separation of the security and economic realms remains complete regarding a given
economic issue, one may tentatively speculate that scholars may employ any ideational
phenomenon to explain economic outcomes. If, however, the economic concern is not
completely divorced from state security interests, scholars must delimit their usage of emotions
to material fear and insecurity, and the material desire for economic well-being. For once actors’
perceptions of state security are viewed or refracted through the lens of ideas, scholars must
avoid deducing state security interests directly from material metatheories.
Since any examination of alignment politics is by definition focused on phenomena
directly or indirectly related to state security, grounding alignment research within
Neoliberalism’s second level is just as difficult as deducing the nature of those politics from
Neorealism. Accordingly, scholars determined to investigate alignment relations via
Neoliberalism must work from within its third level, the transnational level. Doing so is
advantageous because state interests are no longer held constant, as they are when scholars
deduce security interests from the material metatheories. Instead, state interests reflect changes
at the “bottom,” as various transnational actors temporarily succeed in influencing state policy.
452
Importantly, any given transnational actor’s interests can be comprised of any ontological
phenomena. Thus, in transnational Neoliberalism, state preferences change in innumerable ways,
so long as there is a transnational entity with those preferences, and the people comprising it
temporarily succeed in influencing state policy on at least one issue.
As is easily imaginable, transnational Neoliberalism is therefore incredibly flexible.
With it scholars seem capable of explaining any event in IR, via any ontology. While solid
evidence is of course preferable, researchers must at least imply that a given transnational actor
with the preferences exhibited by the state contributed to the policy’s promulgation. Precisely
452
Moravcsik, “Taking Preferences Seriously,” 518.
- 244 -
because of transnational Neoliberalism’s incredibly broad scope, I pointed out in Chapter 4 that
Koo’s complex interdependence model is the only scholarship I examined closely remaining free
of obvious ontological paradoxes.
Despite its endless ontological possibilities at the state level, transnational Neoliberalism
faces one absolute constraint: The preferences of the virtually innumerable actors competing to
influence state interests cannot change. In this one fundamental aspect, transnational
Neoliberalism is wholly material. Not surprisingly, constant actor preferences also comprise the
most essential factor differentiating transnational Neoliberalism and Constructivism. Whereas
the ontological configuration of transnational Neoliberalism shifts only in response to the
rotation of actors across issues or time, in Constructivism the preferences of the actors
themselves may also change as they interact with each other and with other social entities.
453
Indeed, no matter whether the unit of analysis is comprised of states operating in the
international system,
454
something resembling the FPE & PO, or some other social combination
of people,
455
Constructivism’s main point is always that actor preferences change.
456
What do alignment security expectations imply about the important question of actor
preferences? Directly stated, changing expectations suggest strongly that actor preferences also
change. And since expectations are transformed via social interaction, preferences are, too.
Intriguingly, understanding the influence on actor preferences of security expectations informs
another potential reason why the materialist scholars I studied tended so strongly to treat
453
Hopf makes this important point quite succinctly. See Hopf, Reconstructing the Cold War, 17-20.
454
Wendt, Social Theory of International Politics.
455
Lebow, A Cultural Theory of International Relations; Katzenstein, The Culture of National Security; Hopf,
Social Construction of International Politics; Hopf, Reconstructing the Cold War.
456
Hopf’s Societal Constructivism entails no fundamental, private actor, because socializing requires more than one
actor, and different kinds of actors interact with each other. Ibid., 17.
- 245 -
alignment expectations as temporally static: Repeatedly shifting expectations are ontologically
mismatched with materialism’s requirement of stable actor preferences.
In the remaining portion of this chapter, while clarifying the primary mechanisms of the
collective self-esteem model, I explain two processes through which expectational change
transforms actor preferences. The first is material: Various members of the FPE & PO bear
political costs in order to bring about the trusting act that actualizes and gives meaning to
alignment security expectations. Since, active participation in the trusting act causes many in the
FPE & PO to become vulnerable materially to the expectations they helped to create, they
develop new preferences regarding both the nature and value of the alignment. The second
mechanism is ideational: Through the trusting act and consequent expectations the FPE & PO
broaden and deepen their state’s collective self-esteem. An important interest that must be
defended in combination with other state interests, the transformed collective self-esteem
represents a considerable shift in actor preferences. Importantly, no swapping of actors is
required to bring about transformed preferences. The preferences of the actors themselves are
instead altered.
III. The Collective Self-Esteem Model
In this section I develop the final aspects of the collective self-esteem model. I ground it
within Constructivism, because the model’s explication of security expectation formation and
degradation informs that actor preferences change longitudinally. And among Neorealism,
Neoliberalism, and Constructivism, only the latter can accommodate transformed preferences.
My utilization of Constructivism is deliberately minimal, however. I do not detail a complete
ontology. My theoretical approach remains firmly deductive and narrow.
- 246 -
The present model differs from my first and second models, and from the modeling
research I examine most carefully in four critical ways: 1) I utilize the concept of the trusting act,
which creates new expectations and rearranges actor preferences; 2) Ontologically longitudinal,
expectations fluctuate over time; 3) I formally introduce the interest of collective self-esteem,
which most of the prior seems to utilize, at least implicitly; and 4) I focus most closely on the
ideational emotions of humiliation and the desire to preserve collective self-esteem. As
illustrated in Diagram 5.2 at the beginning of this chapter, humiliation and the desire to preserve
collective self-esteem drive intra-alignment conflict. The role of material fear and insecurity in
generating cooperation is shown in Diagram 5.1.
A. The Interests and Emotions
The collective self-esteem model represents an extension of my earlier theoretical models.
Unless otherwise specified, I continue to utilize the definitions and interests delineated in
chapters 3 and 4. Remarkably, the current model is the only one in which members of the FPE
& PO always respond to their emotionality with determination. In contrast, while the delimited
balance of threat model accounts for deliberate material cooperation, it contains no mechanism
producing active ideational cooperation or conflict. Most interestingly, its adherence to critical
aspects of Neorealism also means it cannot explain any uptick in material conflict within an
ongoing alignment. The common sense social emotions model is comparatively robust. Its
emotions always accommodate intentional material cooperation and conflict. But while
ideological conflict results directly from the common sense emotions on every occasion,
ideational cooperation should normally occur only indirectly, via the fairly passive emotion of
- 247 -
gratitude.
457
I illustrate these general trends under normal alignment crises, and specify the
appropriate emotions in Diagram 5.5.
I decline to diagram the collective self-esteem model’s predictions for immediate,
existential alignment crises. I skip it because the results under severe crises are practically
identical to those found in the existential crises illustrated in diagrams 4.3 and 4.4, though one
must allow for the sometimes different emotional ontology of the current model and the common
sense social emotions model.
458
On no-doubt exceptionally rare occasions, the collective self-
esteem model will nevertheless produce radically different outcomes. Specifically, since
collective self-esteem is an actual interest, states will sometimes prioritize it over obvious
material interests such as those found in severe, shared security crises. Thus, a given state might
457
Recall from Chapter 4 that active gratitude requires one’s prior security expectations to be surpassed by
alignment loyalty, something unlikely under the temporally static expectations of the common sense social emotions
model.
458
See Chapter 4.
Diagram 5.5
Passive or Active Emotions in the Three Alignment Models (Normal Crises)
Type of
Conflict or
Cooperation
Delimited Balance of Threat
(Causal Emotion(s))
Common Sense Social
Emotions
(Causal Emotion(s))
Collective Self-Esteem
(Causal Emotion(s))
Material
Cooperation
Active
(Material Fear)
Active
(Material Fear)
Active
(Material Fear)
Material
Conflict
NA*
(Material Fear)
Active**
(Ideational Anger and
Ideational Vengefulness)
Active**
(Ideational Humiliation and
Ideational Desire to Preserve
Collective Self-Esteem)
Ideational
Cooperation
Passive
(Ideational Friendliness)
Usually Passive***
(Ideational Gratitude)
Active
(Ideational Trust, Ideational
Fear, and Ideational Desire to
Promote Collective Self-Esteem)
Ideational
Conflict
Passive
(Ideational Enmity)
Active**
(Ideational Anger and
Ideational Vengefulness)
Active
(Ideational Humiliation and
Ideational Desire to Preserve
Collective Self-Esteem)
*Neorealism permits no material conflict in ongoing alignments.
**While ideational conflict is expected, the jilted state’s material interests and capabilities over its semi-
defecting partner determine the likelihood of material conflict.
***It is normally indirect, because of the model’s temporally static expectations.
- 248 -
actually prefer falling to an aggressive power, rather than relying on an alignment member
regarding which its collective self-esteem is extremely sensitive.
459
The FPE & PO might prefer
being “dead” than “red,” in other words. While I do not detail extreme cases like these, just their
very possibility suggests the importance of understanding collective self-esteem.
COLLECTIVE SELF-ESTEEM AND IN-GROUP PREFERENCE
Collective self-esteem is a critical national interest. Wendt defines it as “a group’s need
to feel good about itself, for respect or status.”
460
Along with physical survival, autonomy, and
economic well-being, he argues it must be defended for the state to survive over the long-term.
461
In this discussion, I utilize collective self-esteem to explain ideational cooperation and conflict in
two ways. First, utilizing literature concerning individuals, groups, and self-esteem, I argue
alignments create in-group preference, a phenomenon which reduces ideational conflict in the
short term and should also contribute to intra-alignment interest internalization over longer time
frames. The emotion at work—the desire to promote collective self-esteem—is active in its
influence, and promotes intra-alignment cooperation.
Though I believe this first mechanism does exert some influence in alignment politics, I
perceive a second mechanism related to collective self-esteem as being much more critical in
determining the tenor of alignment relations in the near-term. Specifically, the FPE & PO
suffers humiliation when the trust it undertook while creating new alignment expectations is
proven misplaced. Unfortunately, humiliation damages collective self-esteem. Because the FPE
& PO desires to preserve its collective self-esteem, it determinedly disparages and diminishes the
459
Since expectations mark the presence of an alignment, states can technically be aligned with states from which
they would never accept security support. Consider the case of the Netherlands at the beginning of World War Two.
Formally attempting to remain neutral, it refused aid from the UK and France before the German invasion.
Nevertheless, since the Allies were clear about their willingness to support the Netherlands, if permitted, the
Netherlands no doubt expected aid. Thus, a formally neutral state was definitionally aligned with states from which
it refused aid.
460
Wendt, Social Theory of International Politics, 236.
461
Ibid., 233-238.
- 249 -
semi-defecting alignment partner. Lowering the value of the hedging partner preserves the FPE
& PO’s collective self-esteem. It also initiates intra-alignment conflict.
The idea of collective self-esteem developed from self-esteem research carried out by
adherents of Social Identity Theory (SIT).
462
According to SIT scholars, the “universal desire”
of people to achieve self-esteem leads them to join groups, partly because of the esteem they
convey to members.
463
Because their self-esteem intertwines with the status of the groups they
join, people’s identities become dependent on the success or failure of the group.
464
Consequently, people come to evaluate their group positively.
465
The effect of the resulting
desire to nurture and protect the group’s collective self-esteem is powerful and promotes in-
group favoritism and out-group discrimination.
466
Remarkably, research shows even trivial
differences between groups result in favoritism and discrimination.
467
As argued in Chapter 3, the formation of alignment expectations within one state
actualizes an alignment, a kind of group. While it is surely foolish to claim alignments are as
important as something like “nation,” alignments by definition deal with state security, which is
of course something the FPE & PO cares very much about. Hence, alignments substantiated by
powerful security expectations are very important groups. This logic suggests strongly that
alignment security expectations will correlate positively with the group dynamics discussed
462
For a concise overview by founders of the theory see H Tajfel and JC Turner, “The Social Identity Theory of
Intergroup Behaviour,” in Psychology of Intergroup Relations, 2
nd
ed, eds. S Worchel and WG Austin (Chicago:
Nelson-Hall, 1986), 7-24. For a more thorough overview of key scholarship, see chapters in Henri Tajfel, ed.,
Social Identity and Intergroup Relations (1982; repr., New York: Cambridge University Press, 2010).
463
From summary of literature in Blake E Ashforth and Fred Mael, “Social Identity Theory and the Organization,”
The Academy of Management Review 14, no. 1 (January 1989): 22; “universal desire” quote from Jonathan Mercer,
“Anarchy and Identity,” International Organization 49, no.2 (Spring 1995): 242.
464
For literature overviews on this matter, see Ibid., 22; John C Turner, “Towards a Cognitive Redefinition of the
Social Group,” in Henri Tajfel, ed., Social Identity and Intergroup Relations (1982; repr., New York: Cambridge
University Press, 2010), 15-36.
465
Mercer, “Anarchy and Identity,” 241-243.
466
See discussion in Turner, “Towards a Cognitive Redefinition of the Social Group,” 22-28, 33-36. For a more
recent examination of the literature, see Jeff Spinner-Halev and Elizabeth Theiss-Morse, “National Identity and Self-
Esteem,” Perspectives on Politics 1, no. 3 (Sep. 2003): 515-532.
467
See discussion in Turner, “Towards a Cognitive Redefinition of the Social Group,” 22-23, 34.
- 250 -
above. Most notably, in terms of alignment politics, strong security expectations enjoyed by all
members of an alignment will minimize in-group ideational and material conflict. And when
expectations weaken—i.e., when the alignment weakens—in-group ideational and material
conflict should increase. The weaker a broader we-feeling is in a given alignment, the more
pronounced the influence of this group dynamic should be. It follows that alignment-based in-
group preference exerts the most influence when the alignment expectations instantiating a
security partnership are the only important source of a group’s group identity.
While alignment expectations are important in many ways, SIT suggests they are also
significant just for the fact that they create a group—an often very critical one. Consequently,
security expectations should impart to the FPE & POs of alignments some of the in-group and
out-group dynamics established by SIT scholars. In the case of Japan and Korea, where we-
feeling is obviously very thin, moments when alignment expectations are particularly robust
should produce a noticeable in-group preference—i.e., good relations—at least when compared
to times when alignment expectations are very weak. The divergence occurs, because
deteriorated security expectations weaken the group, and therefore break down quite a bit of the
prior in-group preference.
In-group preference should contribute to material and ideational cooperation among
alignment members, particularly over long time periods as intra-alignment group identification
solidifies. Nevertheless its influence on material security relations, especially, is certainly
dwarfed in the short-term by the material interests of the moment. For, in the collective self-
esteem model, the FPE & PO’s perceptions of material threats and opportunities in the
alignment’s milieu continue to play the critical role in determining material behavior among
most security partners. Consequently, in the short-term, collective self-esteem is most
- 251 -
remarkable for its impact on intra-alignment ideational relations. Diagram 5.6 illustrates how
complementary material fear and insecurity in the alignment’s security milieu brings about short-
term material cooperation and, through group dynamics, produce good ideological relations. As
I clarify below, ideational trust and ideational fear certainly play a more critical role leading to
intra-alignment ideological cooperation. But diagramming the cooperative portion of SIT’s in-
group dynamics is still worthwhile, as it likely explains some of the warm relations.
468
468
The reverse process—divergent interests leading to weakened in-group feeling and influence on ideational
conflict—is straightforward once one understands Diagram 5.6. Consequently, I decline to illustrate it.
Diagram 5.6
Security Interests Produce Ideational and Material Cooperation in the Collective
Self-Esteem Model: In-Group Dynamics, Short-Term Emphasis
*The trusting act is discussed below.
The new or strengthened alignment represents an important,
or more important, group to the FPE & PO in both X & Y.
States X & Y enjoy convergent security interests due to similar perceptions of threats and/or opportunities in
their shared security milieu. Moreover, they perceive that unusually deep security cooperation is required to
achieve their complementary material objectives.
X & Y align or align more deeply—i.e., they each create new security
expectations and, therefore, new alignment obligations.*
The FPE & PO in X & Y develop, or
strengthen, in-group preference.
Intra-alignment ideational friction decreases.
X & Y enjoy good material and ideological alignment relations.
X & Y maintain their new
expectations.
X & Y cooperate materially to the extent
called for by their new, respective
expectations actualized above.
- 252 -
Importantly, collective self-esteem is frequently a hidden hand in alignment politics
research. It is, at minimum, always compatible with the scholarship I examine most closely.
Snyder’s psychology-grounded contention that “commitment to [an] alliance will have increased
the values subjectively associated with that alliance and reduced the attractiveness of alternatives”
easily accommodates a phenomenon like collective self-esteem.
469
Cha’s implied argument that
Japan suffers ongoing entrapment fear stemming from Korea’s politicization of the past, suggests
many Japanese care deeply about their collective self-esteem. Seoul’s deliberate decision to gain
leverage over Tokyo’s material decisions by launching history bombs over the East Sea/Sea of
Japan clarifies the existence of perceptions about the importance of collective self-esteem in the
quasi-alliance model.
470
Yoon’s work is also highly compatible with the notion of collective self-esteem. His
argument that Korean and Japanese leaders tamp down historical troubles as net threat rises
suggests these men and women understand that disagreements over the past are damaging to the
other state’s collective self-esteem. Lind’s argument that Japanese nationalists are pained by
Tokyo’s apologies to the ROK is even more supportive of the importance and existence of
collective self-esteem. Finally, Koo’s contention that mutual material interests compel Japan and
Korea to contain troublesome history-imbued quarrels over Dokdo/Takeshima implies these
ideological disagreements are painful to each state’s collective self-esteem. Failure to
compromise at critical moments might therefore result in an ideational overriding of the material
imperative.
471
While all the above scholarship is at least implicitly compatible with the phenomenon of
collective self-esteem, none of the researchers contend in-group preferences formed through
469
Snyder, Alliance Politics, 321.
470
See my discussion of Cha in Chapter 3, especially.
471
See my discussion of Yoon, Lind, and Koo in Chapter 3, especially.
- 253 -
alignment expectations should reduce ideational conflict. More specifically, all the scholars
seem to agree with the notion that people care about self-esteem, and people in national units
care about their group’s self-esteem. What these academics evidently fail to consider is the
possibility that alignments are instantiations of important groups and, therefore, some element of
group dynamics should influence the tenor of relations.
As noted earlier, I argue a second dynamic concerning collective self-esteem exerts
considerably more influence than the in-group/out-group mechanism outlined above. Briefly,
the FPE & PO experiences considerable humiliation when the trust its members placed in the
alignment is proven by events to have been unwise. Because the humiliation damages their
collective self-esteem, these men and women aggressively defend it. Most importantly, the FPE
& PO disparages and diminishes the offending other, primarily with tongue and pen. The
consequence is intra-alignment ideational friction. Before discussing this second dynamic, I
quickly emphasize ideational fear, an emotion incentivizing deliberate intra-alignment ideational
cooperation.
IDEATIONAL FEAR
Ideational fear—which I discussed in Chapter 4 and therefore treat minimally here—
sometimes provides security partners with considerable inducement to reduce or eliminate their
ideational differences. Its influence almost always requires the concomitant presence of material
fear and insecurity, however, because those fearing the ideational emotion practically always fear
the possibility that ideational problems will prevent desired material cooperation.
Importantly, ideational fear is not friendly to materialism. Its very existence connotes
that ideas can overpower material imperative. Deductive materialists who utilize ideational
fear—even implicitly—are accordingly committing a grave theoretical error. Namely, they are
- 254 -
theorizing within a material-ideational paradox in which fear of ideas can incentivize ideational
cooperation, even while the deductive material framework always places materialism before
ideationalism. The paradox is thus: Actors cannot possibly acquire the incentive to fix
problematic ideas for purely material reasons, because ideas can never overpower material
interests. Nevertheless, as noted in Chapter 4, Snyder and Walt and the four Japan-Korea
alignment politics modelers all agree ideational fear is determinate in alignment politics.
Ideational fear can be particularly powerful in the collective self-esteem model. Its
influence emerges from collective self-esteem’s specified ontological categorization as a critical
national interest. States should therefore choose to sacrifice material interests in favor of
preserving collective self-esteem with some frequency.
472
Members of the FPE & PO who wish
to prioritize material interests over intra-alignment ideational squabbles accordingly suffer real
fear—ideational fear. They fear specifically that ideological divisiveness will prevent the
material alignment coordination they require to feel secure. I illustrate the role of ideational fear
in Diagram 5.7.
472
Wendt advises an inductive approach to establish the prioritization states give to the four national interests. See
Wendt, Social Theory of International Politics, 237-238.
- 255 -
THE TRUSTING ACT I
In the event the FPE & PO in one or more aligned states comes to fear that ideational
problems are threatening the intra-alignment material security cooperation perceived necessary at
that moment, the FPE & PO must successfully carry out a trusting act. Its primary objective is to
establish security expectations adequate to the task of permitting the alignment to carry out the
material coordination desired. After all, if one state does not expect its security partner to fulfill
its alignment obligations, the suspicious state becomes much less likely to expend the resources
necessary to meet its obligations to the latter. What the FPE & PO must do, in other words, is
create adequate intra-alignment trust.
Diagram 5.7
Security Interests Trigger Alignment Cooperation: Material and Ideational Fear
Emphases
*The trusting act is discussed below.
The FPE & PO in both X & Y deliberately strive to overcome
the ideological differences distancing their states. They must
bridge the ideational gap at the root of the material interest gap,
as it existed just before the need to align, or align more closely,
arose. Ideational fear drives their determination.
X & Y cooperate materially to the extent called for by their
new, respective expectations as actualized above, and to the
extent permitted by their improved ideational relationship.
States X & Y enjoy convergent security interests due to similar perceptions of threats and/or opportunities in
their shared security milieu. Moreover, they perceive that unusually deep security cooperation is required to
achieve their complementary material objectives.
X & Y align or align more
deeply—i.e., they each create
new security expectations and,
therefore, new alignment
obligations.*
X & Y enjoy good material and ideational relations.
- 256 -
Trust can be considered an “emotional attitude” that “determines what a trusting person
will believe and how various outcomes are evaluated.”
473
Trust is interesting in the context of
alignment expectations, because the expecting/trusting party perceives that its opposite party is
obligated to fulfill the action expected or trusted.
474
In this sense, alignment expectations and
trust are identical.
Scholars agree trust exists as a continuum, with distrust or lack of knowledge generally
being its opposite. Moreover, trust is usually considered issue specific, in that it is reasonable to
think that Party X is trustworthy on one issue but not another. Therefore, trust exists when X
believes Party Y will not harm X’s interests regarding a specific issue, despite some
uncertainty.
475
In alignments, each state trusts its partners to fulfill their obligations to a specific
aspect of the security of the self. But “aspect of security” can encompass anything from formal,
expansive alliances set against whatever threat might emerge in the future, to informal
alignments regarding one contemporary security issue.
Unfortunately, establishing trust within alignments or among would-be security partners
is difficult. Most critically, security alignments lack the depth of dispute resolution mechanisms
found in advanced international organizations (IOs) like the World Trade Organization (WTO).
States joining an institution as considerable as the WTO can place their trust in its powerful
enforcement mechanisms, rather than in each other.
476
Arguably more accurate, states can trust
473
Brend Lahno, “On the Emotional Character of Trust,” Ethical Theory and Moral Practice 4, no. 2 (June 2001):
171, 174, 182-185.
474
For trust and obligations, see Ibid., 180. For my discussion of alignment expectations in terms of obligations, see
Chapter 3.
475
For quick overviews of the scholarship on trust, see Deborah Welch Larsen, “Distrust: Prudent, If Not Always
Wise,” in Distrust, ed., Russell Hardin (New York: Russell Sage Foundation, 2004), 35-42; Edna Ullmann-Margalit,
“Trust, Distrust, and In Between,” in Distrust, ed., Russell Hardin (New York: Russell Sage Foundation, 2004), 60-
64; Aaron M Hoffman, “A Conceptualization of Trust in International Relations,” European Journal of
International Relations 8, no. 3 (2002): 376-380; and Andrew H. Kydd, Trust and Mistrust in International
Relations (Princeton: Princeton University Press, 2005), 6-12.
476
Keohane, After Hegemony, 85-109; Axelrod and Keohane, “Achieving Cooperation under Anarchy.”
- 257 -
normally untrustworthy members because of recourse to the IO’s machinery. In contrast, most
states wishing to align, or to deepen their alignment, must trust each other directly, with no
substantial recourse beyond their own material and ideational capabilities for punishing
disloyalty. The fact that alignments primarily exist to deal with material security issues
compounds the problem, because misplaced trust in an economic regime should primarily
produce damage in the economic realm. In contrast, trusting a security partner unwisely more
directly damages material security interests.
The trust problem faced by states wishing to deepen their security expectations is more
akin to the situation experienced by states wishing to establish non-military regimes. Brian
Rathbun argues that states wishing to establish an international institution must essentially trust
each other before they begin. Trust preceding the actualization of the institution is required,
because the reciprocity mechanism upon which the Neoliberals rely still depends on one partner
taking the first step. Until there is trust, no state will sacrifice what it values. And if there is no
first step, there can be no reciprocity mechanism.
477
Seen in the light of Rathbun’s research
findings, Krasner’s contention that regimes cannot exist unless “expectations converge in a given
area of international relations” seems prescient.
478
Interpreted liberally, Krasner seems to
understand that a certain amount of trust must be present before an international regime can be
created.
When intra-alignment ideational fear compels a trusting act, the FPE & PO faces another
problem. Namely, the very presence of ideational fear suggests the ideological problems
confronting the alignment are considerable. If they were not, there would be little reason to be
fearful. In the collective self-esteem model, the most severe ideational problems between
477
Rathbun, Trust in International Cooperation, esp. 17-23.
478
Krasner, “Structural Causes and Regime Consequences,” 2.
- 258 -
alignment members threaten the national interest of collective self-esteem. The depth of the
problem facing the FPE & PO is clarified by noting that their task is sometimes to overcome the
imperatives of one national interest—collective self-esteem—in order to accomplish the needs of
a second interest, physical survival.
In order to create the trust necessary for establishing or expanding a given alignment, the
FPE & PO in one or all aligned states must convince four distinct groups of people of its
sincerity. In the case of State X, its most critical audience member is normally State Y’s FPE &
PO. Failure to achieve these people’s trust means they will be much less likely to take the steps
necessary for meeting their state’s alignment obligations. On occasions when X’s FPE & PO is
severely and internally divided, X can be its own target audience. Because the Foreign Policy
Executive should be quite cohesive, however, the FPE & PO problem will normally be rooted in
the Pertinent Others component.
479
As explained in Chapter 3, the PO is a diverse group,
comprised of anyone not formally part of the FPE who is aware of their state’s security
expectations, and either enjoys some influence at the systemic level or is a transnational actor. If
enough of the PO cannot be convinced that a sincere effort is being made to move beyond
ideational divisions, there will be substantial resistance against the trusting act and the
alignment.
480
The final two audiences are comprised of both X’s and Y’s remaining transnational
actors—i.e, those who are not part of the PO. These people are definitionally unaware of their
state’s alignment expectations.
481
Accordingly, the FPE & PO’s aim is not to convince them to
support the alignment itself. The goal is instead to diminish ideological discord at the
479
Lake, “The State and American Trade Strategy in the Pre-Hegemonic Era.”
480
Influential journalists and scholars will strive to undermine the trusting act, for example.
481
Hence they are not part of the PO. See Chapter 3.
- 259 -
transnational level, thereby preventing troublesome ideas from challenging the alignment from
the “bottom.”
The message conveyed in the trusting act by X’s FPE & PO differs according to the
audience. To Y’s FPE & PO and remaining transnational actors, X makes painful ideational
concessions to evidence its sincerity in dealing with the problematic ideas domestically. To
those in X’s PO who are wavering, and to X’s own remaining transnational actors, X’s FPE tries
to express that its commitment is significant enough to bring along Y. In essence, the message is:
“This moment really marks a substantial turning point in the relationship of our two states.” In
the event aligned states suffer a terrible on-again off-again partnership like Japan and Korea’s,
the message changes slightly to: “This time is different. This time the turning point is real.”
Absolutely critically, X’s FPE and members of its PO who supported the alignment at the
point of formulation must believe the cost they bear—their sacrifice—in the trusting act will
make a difference. That is, they must believe their concessions are adequate to the task of
changing behavior among all their audiences, but particularly Y’s FPE & PO. If enough of X’s
FPE & PO do not believe adequate ideational cooperation has occurred, they become much less
likely contribute to maintaining the alignment. Disbelief and inaction in turn greatly increase the
odds of the alignment faltering.
One final point remains to be clarified before discussing the trusting act itself. Namely,
its depth and breadth should correlate positively with the difference between an alignment’s prior
and currently needed security expectations. It is relational, because the prior security
expectations prove inadequate under current material conditions, precisely because the material
gap is a product of intra-alignment identity mismatch. Thus, alignments or would-be alignments
with very little we-feeling will have to undertake a trusting act much more significant than
- 260 -
security partners enjoying an already deeply internalized alliance. In fact, states enjoying deeply
internationalized alignments are likely to have moved past the point of ever needing major
trusting acts. In sum, states must bridge the ideational gap to which the current material interest
gap is rooted if they wish to align, or align more closely. Otherwise they cannot create
successfully the security expectations required to achieve security via alignment.
THE TRUSTING ACT II
Trusting acts vary. When the ideational gap needing to be closed is very large, as it is
when ongoing intra-alignment ideological battles have been unusually damaging to the partners’
collective self-esteems, states must undergo substantial, formal ceremonies to actualize adequate
trust. Ceremonies can range from the formal exchanging and acceptance of apologies, to simple
statements that discord has disappeared or is disappearing. When the ideational gap to be
bridged is small enough, problematic issues can be sidestepped via emphasis on areas where
there is already ideational similarity and cooperation. In this event, the trusting act conveys that
an unrealized intra-alignment ideational closeness exists, and/or that ideational similarities are
much more significant than differences.
In severe cases, the alignment’s FPE & POs are obligated to exhibit ongoing commitment
to the ideas conveyed in the trusting act. Thus, for example, the top leaders in a state’s FPE must
make periodic statements noting the fundamental changes and concessions made, and
downplaying and containing occurrences as they emerge from either partner’s transnational level.
Toward the middle of the FPE & PO, newspaper journalists and editorialists must be relatively
unified in publicizing the importance and sufficiency of concessions made by both partners. At
the very bottom of the FPE & PO, an adequate number of people and groups understanding the
need for closer security partnership must actively support the initial and ongoing effort to show
- 261 -
that things have changed. The goal of the FPE & PO’s endeavors in and following the trusting
act is to prevent or minimize ideational discord as it emerges, or has emerged, transnationally. If
they fail, the odds of maintaining appropriately robust intra-alignment security expectations fall
considerably.
Given the frequency with which materially-driven events necessitate trusting acts in the
Japan-Korea alignment, and considering how their ideational differences are often damaging to
each other’s collective self-esteem, it is no surprise that Seoul and Tokyo frequently initiate
trusting acts. I listed the most important, high-level trusting acts while constructing my DV. As
detailed in Chapter 1, they include five apologies regarding the history issue; agreements
permitting financial restitution, normalized relations, and cultural imports; formal, state-level
visits; and two other major concessions regarding issues critical to disagreement over the past.
Importantly, these trusting acts were intended to close enough of the intra-alignment
ideological gap to permit adequate security cooperation, at least for the short-term. With each
trusting act, the FPE & POs in both Seoul and Tokyo successfully created or renewed security
expectations. Unfortunately, the depth of those expectations always proved inadequate in
comparison with the extensive challenges posed by shifting material circumstances in the
alignment’s security milieu. As Japan’s and Korea’s respective perceptions regarding the
milieu’s threats and opportunities diverged, security policies became mismatched. Conflictual
policies in turn humiliated those members of the FPE & PO who trusted—those who believed
this time was different.
- 262 -
THE SHORT-TERM OUTCOME OF HUMILATION
The “severe emotional pain” following betrayal is grounded in the significant emotional
component of the trusting act.
482
While one IR scholar finds violation of trust resulting in “moral
outrage” and feelings of regret among the trusting party,
483
my focus is ontologically focused on
humiliation.
In the introduction of his outstanding collection of essays on humiliation, William Miller
defines humiliation as “the price we pay for not knowing how others see us.”
484
Sanjay
Palshikar likewise focuses his definition on the Other, arguing it means one has been “rendered
inferior or deficient in some respect by others in a deliberate and destructive way.”
485
Considered in terms of alignment politics, therefore, humiliation is the terrible feeling
experienced when a security partner values one’s security interests and security role less than one
thinks they should be valued, and shows it via deliberate and destructive action. But this
definition is not complete. For humiliation often operates in an opaque Self-Other
environment.
486
Accordingly an actor’s own actions, even if well intended and carried out
honorably, can contribute powerfully to his or her humiliation. In the case of alignment politics,
the FPE & PO of jilted states experience humiliation because the trust they undertook as part of
the trusting act process was proven misplaced by the other state’s actions. The emotionality
experienced can be severe, because humiliation “is perhaps our most powerful socially oriented
emotion of self-assessment.”
487
482
J. David Lewis and Andrew Weigert, “Trust as a Social Reality,” Social Forces 63, no. 4 (June 1984): 971.
483
Larson, “Trust and Missed Opportunities,” 714.
484
William Ian Miller, Humiliation: And Other Essays on Honor, Social Discomfort, and Violence (Ithaca, NY:
Cornell University Press, 1993), x.
485
Sanjay Palshikar, “Understanding Humiliation,” Economic and Political Weekly 40, no. 51 (Dec 2005): 5428.
486
William A Callahan, “National Insecurities: Humiliation, Salvation, and Chinese Nationalism,” Alternatives 29
(2004): 203.
487
Miller, Humiliation, x.
- 263 -
The level of humiliation the FPE & PO in each state feels can reflect different factors.
Less powerful alignment members might experience more humiliation from a partner’s semi-
defection, because smaller states are likely to derive more prestige from group involvement than
larger ones.
488
Conversely, powerful states that feel they should receive deference might
experience more humiliation than one would otherwise anticipate. Moreover, the size and nature
of the organization can matter as well: Humiliation will be magnified in small groups with
informal decision making processes, and muted in large ones with formalized, impersonal
mechanisms for making decisions.
489
The point to make is that hedging and semi-defection from the alignment damages the
collective self-esteem of jilted friends. Damaged collective self-esteem causes impoverished
friends to suffer friction in that realm, because a humiliated FPE & PO will enact offensive
strategies for preserving its collective self-esteem. Primarily, these people will denigrate the
defecting partner to diminish the impact of its act of disregarding. As they emphasize as
negative some aspect of the defecting party, they establish to themselves their moral superiority
and trustworthiness over the other.
490
This strategy diminishes the humiliation, because the
actions of an “inferior” and “untrustworthy” people are much less likely to be interpreted as
being humiliating. In the event the jilted states have maintained a “chosen trauma” as part of
their identity vis-à-vis the defecting state, it will likely be rekindled.
491
Even without a chosen
488
This would especially be the case if states were utilizing their membership in a group to gain “social mobility”—
i.e., trying to “identify with an alternative higher [status] group.” See Leonie Huddy, “Contrasting Theoretical
Approaches to Intergroup Relations,” Political Psychology 25, no. 6 (Dec. 2004): 956. If joining a group results in
more equal status with powerful members, it will result in greater self-esteem. See discussion in Spinner-Halev and
Theiss-Morse, “National Identity and Self-Esteem,” 517-519.
489
Crawford, “The Passion of World Politics,” 132.
490
See the discussion in Spinner-Halev and Theiss-Morse, ““National Identity and Self-Esteem,” 520-521.
491
Vamik D. Volkan, “Transgenerational Transmissions and Chosen Traumas: An Aspect of Large-Group Identity,”
Group Analysis 34, no. 1 (2001): 79-97.
- 264 -
trauma, however, some differentiating factor would become salient, at least temporarily. The
outcome over the short-term is bad intra-alignment ideational relations.
THE LONG-TERM OUTCOME OF HUMILATION
The most interesting consequence of humiliation following unwise trust enactment is its
impact on actor preferences. Specifically, continuing episodes of interest-based alignment semi-
defection and consequent humiliation lead states to prioritize more highly collective self-esteem
as a national interest.
The mechanism leading to changed actor preferences is founded on the intersection of the
trusting act and humiliation. Humiliation, as argued, causes the FPE & PO to aggressively, but
defensively, attack its hedging alignment partner with pen and tongue. Strained ideational
relations centered around the collective self-esteem national interest result. As a consequence,
should the alignment’s material interests once again require close security coordination, the FPE
& POs in both states have a larger ideational gap to bridge if they are to create the trust required
for the actualization of adequate alignment expectations. At the beginning of the second
cooperation-conflict cycle, therefore, the FPE & POs must undertake an even more audacious
trusting act. But greater effort in the trusting act means greater humiliation in the event material
interests diverge once again and produce dissimilar security policies. With every cycle, the intra-
alignment collective self-esteem battle gets worse, as the FPE & PO comes to prioritize more
highly its state’s collective self-esteem. The prioritization of collective self-esteem versus the
other interests marks an important shift in actor preferences. In the final analysis, it suggests that
alignment models must be grounded in ideational metatheories.
- 265 -
Chapter 6
Statistical Analysis
I. Overview
In this chapter I use econometrics to uncover correlations between various Independent
Variables (IVs) and the Japan-ROK history issue Dependent Variable (DV) I described in
Chapter 1: the Primary DV. My chief contribution is an IV representing the convergence,
divergence, and relative neutrality of Seoul’s and Tokyo’s respective North Korea policies. I
evaluate the impact of this variable and its components against the DV both separately and along
with other IVs suggested critical by various theories and models. In particular I create regression
models to reflect Neoliberal and Neorealist metatheories. As evidenced below two IVs return
statistically significant results: The North Korea Policy IV and ROK National Assembly
elections.
II. Econometrics
A. North Korea Policy Variable
The North Korea policy variable measures whether Seoul’s and Tokyo’s respective
Pyongyang policies were convergent, divergent, or comparatively neutral at a given time. It is
comprised of four components: the Korea Clause, the repatriation to the DPRK of Koreans living
in Japan, humanitarian food aid to North Korea, and trade with North Korea. I evaluate the
influence of these IV components on the Primary DV both separately and together.
- 266 -
KOREA CLAUSE
The Korea Clause was a series of statements made by Japanese officials, regarding their
security interests on the Korean Peninsula. In formal and informal statements these officials
clarified whether they valued just South Korean security, or whether they valued security on the
entire Korean Peninsula. South Korea argued firmly that only its security should matter to Japan.
The Korea Clause first became a bilateral issue in the late 1960s, and ceased being
problematic by 1975.
492
In order to promote simplicity and best capture official policy, I code
the Korea Clause via two criteria. First, I include only those statements comprising it when they
accompany Japanese summit meetings with either the US or the ROK. These cooperative (-1)
statements occur whenever Japan specifies only South Korea’s security. In contrast, I code
conflict (1) whenever Japan refers to the entire Korean Peninsula, or otherwise does not mention
the importance of South Korean security in summit-related statements through 1975. Second, I
code Korea Clause statements as cooperation or conflict only when they represent a major
change in the tenor of Japan-Korea affairs. The last summit I code is the Chun-Nakasone
meeting in 1983. I include it because it marked the first time Japan expressed the importance of
Seoul’s security in a formal Japan-ROK summit. I code all other years from 1960 to 2008 as
neutral (0). The Korea Clause was actualized according to the above criteria four times, as seen
in Table 6.1.
492
Dong-jun Lee, “From the Secret ‘Korean Minute’ to the Open ‘Korea Clause’: The United States and Japan and
the Security of the Republic of Korea,” Asian Perspective 36 (2012): 136, fn 23 on 143.
- 267 -
Table 6.1 Korea Clause
1969 Cooperative -1 Nixon-Sato
493
1972 Conflict 1 Nixon-Sato
494
1975 Cooperative -1 Miki-Ford
495
1983 Cooperative -1 Chun-Nakasone
496
REPATRIATION OF KOREANS IN JAPAN TO NORTH KOREA
The issue of repatriating Koreans living in Japan to North Korea comprises the second
component of the North Korea Policy IV.
497
Over considerable South Korean protest, the 1959
Calcutta Accord permitted Japan to repatriate Koreans wishing to live in North Korea. The ROK
considered repatriation a serious security issue. Seoul, which did not normalize relations with
Japan until 1965, feared that repatriation and the DPRK-Japan talks preceding it conveyed
legitimacy to North Korea and portended formal Pyongyang-Tokyo relations.
498
The biggest challenge in operationalizing the repatriation issue emerges from the fact that
the plight of Koreans living in Japan comprises one of the newspaper headline topics
contributing to the history issue DV.
499
Repatriation therefore threatens to be both an IV and a
DV at the same time. Fortunately, it is clear that the general subject of Koreans living in Japan
transformed from being a security issue to being a history issue. The shift occurred by 1970, by
which time the number of Japanese-Koreans registering formally as South Korean outnumbered
those registering as North Korean for the first time. Consequently, I do not input the headline
493
Cha, Alignment Despite Cooperation, 13, 51, 75-78, 203; Lee, Japan and Korea, 71.
494
Cha, Alignment Despite Cooperation, 115-119, 203; Lee, Japan and Korea, 76, 115-116.
495
Cha, Alignment Despite Cooperation. Lee, Japan and Korea, 94; Bridges, Japan and Korea in the 1990s, 54.
496
Ibid., 15, 54.
497
The best overview remains Changsoo Lee, “The Politics of Repatriation,” in Koreans in Japan: Ethnic Conflict
and Accommodation, eds. Changsoo Lee and George de Vos (Los Angeles: University of California Press, 1981),
91-109. For a more recent account, see Tessa Morris-Suzuki, Exodus to North Korea: Shadows from Japan ’s Cold
War (Lanham, MD: Rowman & Littlefield, 2007), 190-192, 195-196, 199, 205, 208, 225, 235-236.
498
Yoon is best on Seoul’s perceptions of the issue. See Yoon, “Fragile Cooperation,” 181-195; Lee, “The Politics
of Repatriation,” 104-105.
499
See Chapter 1.
- 268 -
term “Korean-Japanese” as a DV until 1970. Table 6.2 reports the numbers behind this
legitimacy battle.
Table 6.2 Nationality of Japanese-Koreans
Year ROK DPRK
1950 39,418 495,818
1951 77,433 467,470
1952 95,157 465,543
1953 116,546 454,462
1954 131,437 424,633
1955 138,602 425,620
1966 247,422 366,466
1968 271,624 320,552
1970 314,407 294,082
1971 344,469 257,749
1972 412,744 217,035
ROK Ministry of Foreign Affairs data, as presented by Yoon.
500
The second challenge for operationalization pertains to the selection of years to code as
conflict (1) or no substantial conflict (0). Fortunately, though repatriation took place from 1959-
1972, there were only two years in which a large number of Korean-Japanese were repatriated to
North Korea. In 1960 49,036 were returned, and in 1961 22,801 were repatriated. On any other
specific year, less than 3,500 Koreans were transported to North Korea.
501
Thus the dividing line
is clear. I code 1960 and 1961 as conflictual.
502
All other years are coded as having no
substantial conflictual.
TRADE WITH NORTH KOREA
The third IV component is trade with North Korea from 1962 to 2008. It has two
subcomponents. The first signifies marked change in Japanese trade with North Korea, from
1962 to 1988. I code years in which total Japan-DPRK trade surged by 50 percent or more over
500
Yoon, “Fragile Cooperation,”188.
501
From Lee, “The Politics of Repatriation,” 106-107.
502
Scholars who begin their analyses before 1960 should probably include 1959, because it was the year the
repatriation was promulgated.
- 269 -
the year prior as conflictual (-1) to Japan-ROK relations. Years in which the trade declined by
50 percent I code as cooperative (-1). I code all other years neutral (0). The second
subcomponent represents conflict (1), cooperation (-1), or neutrality (0) in Japan’s and South
Korea’s respective trade policies with North Korea, from 1989 to 2008. As before I am
interested in percent increase or decrease from the year prior. But since there are now two sets
of trading partners, I must consider the difference between them. I utilize the following seven,
firm coding rules:
1) If the overall difference in the change is more than 50 percent, it is conflict.
2) If one increases by 25 percent or more and the other increases by 15 percent or more it is cooperation.
3) If one decreases by 25 percent or more and the other decreases by 15 percent or more it is cooperation.
4) If one increases by 25 percent or more and the other decreases by 15 percent or more it is conflict.
5) If one decreases by 25 percent or more and the other increases by 15 percent or more it is conflict.
Table 6.3 shows the percent increase and decrease on the year prior of both Japan’s and
the ROK’s respective trade with North Korea. Data coded cooperative or conflictual are
connoted by a highlight.
- 270 -
Table 6.3 Percent Change in Total Trade Over Year Prior
Year 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972
DPRK-Japan 0.58 1.13 -0.01 -0.11 0.36 0.45 0.03 0.02 0.02 1.37
DPRK-ROK 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Year 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982
DPRK-Japan 0.24 1.08 -0.32 -0.32 0.15 0.51 0.48 0.29 -0.23 0.08
DPRK-ROK 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Year 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992
DPRK-Japan -0.02 -0.12 0.07 -0.17 0.28 0.23 -0.13 -0.03 0.05 -0.04
DPRK-ROK 0.00 0.00 0.00 0.00 0.00 0.00 0.25 -0.32 7.62 0.55
Year 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
DPRK-Japan -0.03 0.07 0.19 -0.13 -0.08 -0.18 -0.12 0.32 1.81 -0.72
DPRK-ROK 0.07 0.04 0.48 -0.12 0.22 -0.28 0.50 0.27 -0.05 0.59
Year 2003 2004 2005 2006 2007 2008
DPRK-Japan -0.27 -0.05 -0.23 -0.39 -0.92 -0.18
DPRK-ROK 0.13 -0.04 0.51 0.28 0.33 0.01
From the ROK Ministry of Unification,
503
Japanese Ministry of Internal Affairs and Communication,
504
and the IMF
DOT Database.
505
HUMANITARIAN FOOD AID TO NORTH KOREA
The final component of the North Korea policy variable is humanitarian food aid to North
Korea, from 1995-2008. Unlike the repatriation component, food aid had nothing to do with
legitimacy within the Korean Peninsula, a battle which the ROK won irreversibly upon
democratization in 1988. Rather, regarding South Korea it is remarkable for reflecting both the
general trend of the ROK’s declining perception of the North Korean threat, and Seoul’s
determination under the Sunshine Policy to do almost whatever it took to convince Pyongyang
that South Korea wanted better relations. For Japan, in contrast, it reflects Tokyo’s worsening
relationship with the DPRK, and hardening policy.
503
Data and Statistics, Ministry of Unification of Korea, http://www.unikorea.go.kr/ (accessed July 1, 2013).
504
Statistics Japan, Statistics Bureau of the Ministry of Internal Affairs and Communication of Japan,
http://www.stat.go.jp/english/index.htm, (accessed July 1, 2013).
505
International Monetary Fund Direction of Trade Statistics,
http://www.imf.org/external/pubs/cat/longres.cfm?sk=19305.0, (accessed July 1, 2013).
- 271 -
Of the four components, food aid to North Korea is the most difficult to code. Instead of
firm coding rules, I therefore utilize general guidelines. To establish whether food aid policy is
conflictual (1), cooperative (-1), or neutral (0), I consider the overall volume of each state’s aid
to the DPRK, the percent change on the year prior of Seoul’s and Tokyo’s respective aid, and
overall directionality. Each state’s food aid volume per year, and my coding decisions are shown
in Table 6.4. Three years proved difficult to code. My coding conclusion and reasoning for
those years is as follows:
1999: Neutral, because ROK trade dropped precipitously on the year prior, but was above zero.
2001: Neutral, because Japan’s massive increase in aid runs counter to a marked decrease in ROK aid.
Nevertheless the volume of aid from each state remains remarkable.
2005: Neutral, because Japan’s aid halved while ROK aid grew considerably. Nevertheless Japan’s aid is
not negligible.
Table 6.4 Food Aid to North Korea*
Year 1995 1996 1997 1998 1999 2000 2001
ROK 150,000 3,401 62,393 54,126 12,204 351,703 198,000
Japan 378,000 138,574 791 67,000 0 99,999 500,000
Cooperation or Conflict -1 1 1 -1 0 -1 0
Year 2002 2003 2004 2005 2006 2007 2008
ROK 457,800 542,191 406,510 492,743 79,500 431,290 58,605
Japan 0 0 80,803 48,084 0 0 0
Cooperation or Conflict 1 1 -1 0 1 1 1
From the World Food Programme’s Food Information System.
506
*Tons, grain-equivalent.
NORTH KOREA POLICY VARIABLE
In order to construct the North Korean Policy IV, its four components must be combined.
Appendix 1 shows the coding for each component, and the resulting North Korea Policy IV’s
506
The World Food Program Food Information System, http://www.wfp.org/fais/ (accessed July 1, 2013).
- 272 -
coding. Importantly, I do not add together the coding of the four components to produce the
main IV. Thus, a year containing two conflictual instances does not result in a 2 coding, but a 1.
There are no two years in which the coding conflicts. I therefore avoid difficult questions
about how to balance the coding of two contrasting components. The uniformity is not
surprising for three reasons. First, not all of the components overlap, or overlap by very much.
Repatriation could only overlap with the Korea Clause in 1969, for example, because the former
ended in 1969. Similarly, the Korea Clause cannot interfere with food aid policy, because they
do not overlap temporally. Only the trade component presented considerable opportunity for
overlap. Second the fact I only coded major changes meant I maximized the odds of conformity.
Regarding the trade variable, for instance, I would have expected to see less clear signals of
government policy if I had set my primary coding threshold at 25 percent instead of 50 percent.
Finally, and most importantly, one expects these policies to be highly uniform within each state,
precisely because they are all security issues related to Seoul’s and Tokyo’s respective
Pyongyang policies. If I had mistakenly selected a component disconnected from North Korea
security policy, I probably would have seen conflicting data.
REGRESSION ANALYSIS OF NORTH KOREA POLICY IV AND COMPONENTS
In this section I carry out two sets of regression analysis. The first, seen in Table 6.5,
analyzes the four North Korea Policy IV components’ separate influence on the DV, as well as
their combined impact on the history issue DV. The second, in contrast, utilizes the formal
North Korea Policy IV, as assembled in Appendix 1. Table 6.6 shows the results of the latter
regression.
Each regression set contains two DVs. The Primary DV (PDV) represents my project’s
formal DV. Alongside it I utilize a DV I call the Primary DV, Alternative (PDVA). The
- 273 -
PDVA’s purpose is intended to accommodate a potential error in the PDV. Specifically, the
PDV’s range of 3 (conflict) to -3 (cooperation) is built upon a couple of potentially flawed ideas.
First, it assumes that cooperative events are additive. Second, it presumes that derailed, pre-
planned meetings (coded 2) are more conflictual than years in which there are peaks of at least
one of the five history issue components (coded 1), as indirectly observed via Chosun Ilbo
headlines. Both presumptions might be incorrect.
In order to compensate for the above potential errors, I transform all cooperation to -1.
Similarly, I equalize the conflict represented by canceled meetings and newspaper headlines,
setting both to 1. Only South Korea’s show and threat of force regarding Dokdo/Takeshima in
2006 remains remarkable among cooperation and conflict. I therefore code it a 2, because it is
obviously more conflictual than any other event encompassed by my research. The PDVA
therefore runs from 2 (conflict) through to -1 (cooperation), with 0 representing years that are
neither conflictual, nor cooperative. Despite including the PDVA in all of the remaining
regression tables, I focus my discussion in this chapter almost completely on the results
uncovered via the PDV. With one notable exception below, readers will need to peruse the
PDVA’s results on their own.
Regarding the Primary DV, Table 6.5 shows that the four North Korea policy
components comprise a statistically significant model. Together, they predict 23.7 percent of the
total variance in the DV, with a 0.003 level of statistical significance. Nevertheless, as
individual components, only repatriation and trade are statistically significant.
- 274 -
Table 6.5 Regression: North Korea Policy Components
Primary DV Primary DV, Alternative
Adjusted R Squared 0.2369 0.1485
Prob > F 0.0029 0.0250
Beta, Korea Clause .0091765 .0608281
P>|t| 0.944 0.657
Beta, Repatriation .329625 .2383154
P>|t| 0.013 0.082
Beta, Food Aid .1357251 .0965634
P>|t| 0.340 0.519
Beta, Trade .3851873 .3534746
P>|t| 0.010 0.025
Table 6.6 shows that the North Korea Policy IV explains 22.1 percent of the total
variance in the Primary DV, and does so with a statistical significance of 0.0004. The coefficient
ratio reports that a one unit increase conflicting North Korea policy produces a 0.85 unit increase
in conflict over the history issue, and does so with a P value of 0.000.
Table 6.6 Regression: North Korea Policy IV
Primary DV Primary DV, Alternative
Adjusted R Squared 0.2211 0.1771
Prob > F 0.0004 0.0015
North Korea Policy Coef. .8486726 .4902655
P>|t| 0.000 0.002
B. North Korea Policy Variable in Neorealism
In this section I analyze the influence of the North Korea Policy IV within a broader
Neorealist context. Thus, I consider two additional factors. Given Neorealism’s emphasis on the
balance of power in the international system, I utilize the system concentration variable available
via EUGene.
507
Second, based on the logic of Cha’s quasi-alliance model and following Koo’s
lead, I utilize an IV reflecting whether the US exhibits a strong commitment to East Asian
security on a give year. Strong commitment is scored 1, and anything else is scored 0.
508
I
507
Scott D Bennett and Allan Stam, “EUGene: A Conceptual Manual,” International Interactions 26 (2000):179-
204 (version 3.204, accessed Aug 2013; http://eugenesoftware.org).
508
Koo, “Scramble for the Rocks,” 64.
- 275 -
adopt fully Koo’s scoring for the years 1960-2005.
509
Consequently, I only need to code 2006 to
2008. Fortunately, the coding decisions are easy for those three years, because America’s
ongoing War on Terror continued to distract it from East Asia. The coding for these variables is
included in Appendix 2.
Table 6.7 Regression: Neorealism
Primary DV Primary DV, Alternative
Adjusted R Squared 0.2394 0.1493
Prob > F 0.0017 0.0175
Beta, North Korea Policy .4382864 .4061193
P>|t| 0.001 0.005
Beta System Concentration .2148573 .0986802
P>|t| 0.099 0.468
Beta, Strong US Commitment to East Asian Security -.1616515 -.1142982
P>|t| 0.218 0.408
As seen in Table 6.7, the Neorealism model accounts for 24 percent of the total variance
in the history issue, and does so at a 0.0017 level of statistical significance. Remarkably, among
the three components, only the North Korea Policy IV is statistically significant at the 0.05
threshold. A one unit increase in the policy IV produces a 0.44 unit increase in the history issue
at a statistical significance of 0.001. The model cannot support Cha’s contention that US
involvement in East Asia determines the tenor of Seoul-Tokyo relations.
C. North Korea Policy Variable in Neoliberalism
In this section I analyze the influence of the North Korea Policy IV within a broader
Neoliberal context. I consider six additional variables pertaining to either trade or democracy.
Regarding trade I apply the system Koo utilized to measure Japan-South Korea trade
dependence.
510
Using bilateral trade data, I record total Japan-ROK trade from 1960-2008.
511
509
Ibid., 282.
510
Ibid., 62-63.
511
IMF Direction of Trade Statistics.
- 276 -
Then I divide it by the total trade each state conducts with the entire world.
512
The outcome is a
trade dependence IV, or the percentage of Japan’s and South Korea’s respective total trade that is
comprised of their bilateral trade.
In order to measure whether Japan’s domestic politics are influencing the history issue, I
record whether a Japanese Prime Ministerial Cabinet was formed on a given year and, if so, the
number of new cabinets.
513
My aim is to capture both intra-party and inter-party politics. The
former is especially important before the late 2000s, because the LDP’s dominance in Japanese
politics was both considerable and durable. Finally, in order to understand the influence of
Korean politics on the history issue, I create two more IVs by recording the years in which South
Korea held a free and fair Presidential or National Assembly election.
514
Years in which
reasonably democratic elections were held are coded 1; other years are coded 0. Finally, I follow
Koo’s methodology and code whether Japan and Korea were jointly democratic in a given
year.
515
The only coding challenge regarding joint democracy is whether to select 1960 or 1961
as the early, anomalous, joint democracy year. Koo chooses 1961, but I select 1960. Not only
were both elections held that year, but the Polity IV dataset awards only 1960 as being highly
democratic.
516
The Neoliberal component coding and related data are specified in Appendix 2.
512
World Trade Organization, Statistics Database, http://stat.wto.org/Home/WSDBHome.aspx?Language=E
(accessed July 1, 2013).
513
Prime Minister of Japan and His Cabinet, http://www.kantei.go.jp/foreign/cabinet/0031-60_e.html (accessed July
1, 2013).
514
Nils-Christian Bormann and Matt Golder, "Democratic Electoral Systems Around the World, 1946-
2011," Electoral Studies (2013).
515
Koo, “Scramble for the Rocks,” 64.
516
Monty G Marhall and Ted Robert Gurr, “Polity IV Project: Politicl Regime Characteristic and Transitions, 1800-
2012,” http://www.systemicpeace.org/polity/polity4.htm, (accessed July 1, 2013).
- 277 -
Table 6.8 Regression: Neoliberalism
Primary DV Primary DV, Alternative
Adjusted R Squared 0.2127 0.2485
Prob > F 0.0161 0.0075
Beta, North Korea Policy .518592 .5032984
P>|t| 0.001 0.001
Beta, Japanese Cabinet Appointed .0772055 .0792728
P>|t| 0.596 0.578
Beta, Korean National Assembly Election .3095791 .376082
P>|t| 0.039 0.011
Beta, Korean Presidential Election .004209 .0221706
P>|t| 0.978 0.881
Beta, Japan Trade Dependence on South Korea .1277275 .3419463
P>|t| 0.585 0.140
Beta, South Korea Trade Dependence on Japan -.1011398 -.1597763
P>|t| 0.567 0.356
Beta, Japan-Korea Joint Democracy -.3447154 -.6244416
P>|t| 0.196 0.019
As Table 6.8 shows, the Neoliberalism model is remarkable for explaining 21.3 percent
of the total variance on the Primary DV at a statistical significance of 0.02. Among the seven
components of the model, two return statistically significant results. A one unit increase in the
North Korea Policy IV produces a 0.52 unit increase in the history issue, with a statistical
significance of 0.001. A one unit increase in Korean National Assembly election results in a
0.31 unit increase in the history issue, with a statistical significance of 0.04.
While the Neoliberal regression results suggest that the ROK National Assembly’s
electoral politics are contributing to historical animosity, joint democracy’s overall effect might
be to improve the history issue. Regarding this point the Primary DV, Alternative’s results are
interesting. They suggest joint democracy is exerting a considerable soothing effect on the
history issue. Unfortunately, the influence of joint democracy is not statistically significant in
the regression involving the Primary DV.
- 278 -
III. Conclusion
Among the three sets of regressions carried out above, the most robust results appear in
the Neorealism model, with its material security focus. It explained 24 percent of the total
variance in the Primary DV. This is not very surprising, as the North Korea Policy IV itself
attempts to capture the logics of material security calculations, even if they are refracted through
ideas. Moreover, for over 30 of 49 years evaluated—until at least the early 1990s—South Korea
certainly viewed the North Korean threat in terms of Pyongyang’s conventional military
capabilities. And just as the ROK began comprehending that North Korean power was no longer
as threatening as it once was, Japan began to perceive considerable danger from North Korea’s
attitude and its unconventional weapons capability. In other words, for all but a handful of years,
one of the quasi-allies viewed North Korea in ways very compatible with Neorealism.
Neoliberalism also performed well, explaining 21 percent of the total variance in the
history issue. While the North Korea Policy IV was the most robust and statistically significant
of the two statistically significant variables, South Korea’s National Assembly elections also
contributed to the history issue. This is wholly unsurprising, of course. Nevertheless, it is good
to see its influence quantified.
Most remarkably, both the North Korea Policy IV and its components evaluated as a
model return statistically significant findings. Each explains 22 and 24 percent of the total
variance of the DV, respectively. Much more impressive is the per-unit influence (1 to 0.85 ratio)
of the North Korea Policy IV regressed solely against the Primary DV. These findings suggest
the importance of North Korea policy in establishing the tone of Japan-South Korean relations.
- 279 -
Chapter 7
Conclusion
Studying Japan-Korea alignment politics via quantitative analysis within deductive
theoretical frameworks reveals much about Seoul-Tokyo relations, alignment politics, and IR
metatheories. Regarding the quasi-alignment, the econometrics reveal that the history issue is
greatly influenced by the security partners’ respective North Korea policies. Statistics also
inform the extent to which the ROK’s domestic politics at the Assembly level politicize
disagreements about the past. While the quantification of this data is important in and of itself,
its significance extends beyond the numbers.
Regarding the alignment itself, the research suggests Seoul and Tokyo will not be able to
develop a “normal” relationship until North Korea either collapses or begins to behave according
to the norms of the developed world. Since North Korea is obviously a remarkably stable state,
and because trust will be so hard to enact in DPRK-Japan and DPRK-US relations especially,
there appears to be little chance Japan and South Korea can greatly improve the historic
animosity they suffer. This will be the case particularly if North Korea realizes it can damage
the Japan-ROK alignment by threatening one partner while deliberately warming relations with
the other. In the event Pyongyang promulgates such a policy on purpose or accidentally, it will
make coordination within the Japan-ROK-US security triangle very difficult.
Similarly, in the event Japan and South Korea perceive the threat associated with the rise
of China differently, Beijing might be able to use the dynamic to weaken the triangular
alignment. Since neither trade dependence nor joint democracy contribute to conflict or
- 280 -
cooperation in a statistically significant fashion,
517
it is not possible to agree with the general
Neoliberal contention that expanding bilateral economic relations between Japan and the ROK,
and democratic deepening in either Seoul or Tokyo, will mitigate any of the above dynamics.
Indeed, in the case of South Korea, National Assembly elections produce ideational problems.
While there is no good news in the above findings, there is a bit more room to celebrate
the findings of the deductive analyses of alignment politics. Most interestingly, the expectations-
focused discussions suggest that Japan is not solely to blame for its sad relations with the ROK.
Instead, both states are considerably culpable. The joint culpability is found in the expectations
both states develop and lose according to the dictates of their shifting material interests. Mutual
responsibility is clear, because a hedging partner cannot be perceived as lowering one’s own
security if the jilted state perceives the other’s modestly improved security position as
contributing to the jilted state’s own security. Hence, the jilted state shares equal responsibility
with its hedging partner. This finding is good news for two reasons. First, too much of the
Japan-Korea alignment literature puts the onus on Japan for the history problem. My research on
security expectations implies the blame is often misplaced. Second, it means only one state
needs to begin internalizing its security partner’s interest for the process to start working.
More critical for alignment politics scholarship in general are my conclusions about
causal mechanisms. In particular, alignment expectations emerge as a factor second to none.
The question of whether expectations are temporally static or longitudinal, or perhaps a
combination of both, has tremendous implications for alignment modeling. Another highly
significant finding pertains to emotions. Alignment politics researchers all implicitly utilize
social emotions, as they are produced when expectations are disappointed. But many scholars
seem to create their models as if they were only material. The greatest omission in this regard
517
I only refer to data produced by the Primary DV in my conclusion.
- 281 -
pertains to the use of ideational fear. Though it is an ideational social emotion, researchers seem
to believe that material fear alone can incentivize determined coordination to resolve ideational
squabbles. Of course that is impossible: Actors who cannot fear the power of ideas would never
strive to ameliorate them for material reasons. But if socially emotional ideas can actually trump
the imperatives of materialism, scholars should not create models without extensive
consideration of their emotional ontology. One of the main reasons I created three models was
to undergo that ontological process deliberately; I did not want to select social emotions without
lengthy consideration.
Regarding the four Japan-Korea alignment scholars, my research findings are most
suggestive when contrasted with Lind’s apology politics models. Lind’s general conclusion is
that Tokyo must stop apologizing if it wants its relations with Seoul to improve. My collective
self-esteem model, with its longitudinal understanding of security expectations and focus on the
trusting act, suggests that Japan and Korea will not be able to coordinate their security strategies
successfully unless Japan and Korea undergo some kind of trusting act. If apologies are indeed
damaging, the quasi-allies will have to undertake some other painful ideational concession in its
place. Otherwise they stand little chance of meeting their short-term security needs.
That said, the collective self-esteem model’s findings that the combination of the trusting
act, degraded security expectations, humiliation, and the desire to preserve collective self-esteem
are sympathetic to the Lind’s argument about a worsening relationship. Instead of pinning the
deterioration on the apologies, however, my third model argues the primary cause rests on the
prioritization of collective self-esteem as a national interest in response to humiliation.
Of my three models, only the collective self-esteem model can explain why the history
issue is getting worse, or at least why it is a lot worse than common sense suggests it should be.
- 282 -
And this argument—i.e., that collective self-esteem rises as a national interest in response to
mismatched security policies in alignments—is perhaps the most important conclusion in my
research. For by concluding that a national interest has become elevated, I have essentially
argued that actor preferences change. Accordingly the collective self-esteem model challenges
one of the core tenets of materialism.
- 283 -
Appendix 1. The North Korea Policy IV and Its Components
North Korea Policy IV and Its Components
Year Korea Clause Repatriation Food Aid Trade NK Policy IV
1960 0 1 0 0 1
1961 0 1 0 0 1
1962 0 0 0 0 0
1963 0 0 0 1 1
1964 0 0 0 1 1
1965 0 0 0 0 0
1966 0 0 0 0 0
1967 0 0 0 0 0
1968 0 0 0 0 0
1969 -1 0 0 0 -1
1970 0 0 0 0 0
1971 0 0 0 0 0
1972 1 0 0 1 1
1973 0 0 0 0 0
1974 0 0 0 1 1
1975 -1 0 0 0 -1
1976 0 0 0 0 0
1977 0 0 0 0 0
1978 0 0 0 1 1
1979 0 0 0 0 0
1980 0 0 0 0 0
1981 0 0 0 0 0
1982 0 0 0 0 0
1983 -1 0 0 0 -1
1984 0 0 0 0 0
1985 0 0 0 0 0
1986 0 0 0 0 0
1987 0 0 0 0 0
1988 0 0 0 0 0
1989 0 0 0 0 0
1990 0 0 0 0 0
1991 0 0 0 1 1
1992 0 0 0 1 1
1993 0 0 0 0 0
1994 0 0 0 0 0
1995 0 0 -1 -1 -1
1996 0 0 1 0 1
- 284 -
1997 0 0 1 0 1
1998 0 0 -1 -1 -1
1999 0 0 0 1 1
2000 0 0 -1 -1 -1
2001 0 0 0 1 1
2002 0 0 1 1 1
2003 0 0 1 0 1
2004 0 0 -1 0 -1
2005 0 0 0 1 1
2006 0 0 1 1 1
2007 0 0 1 1 1
2008 0 0 1 0 1
- 285 -
Appendix 2. Neorealism IV and Neoliberalism IV Components
Neorealism IV and Neoliberalism IV Components
Year
System
Concentration
Strong US
Involvement
Japan
Cabinet
Appointed
ROK
Assembly
Election
ROK
Presidential
Election
Japan
Trade
Dependence
on ROK
ROK Trade
Dependence
on Japan
Joint
Democracy
1960 0.3 1 2 1 0 0.014 0.316 1
1961 0.3 0 0 0 0 0.015 0.417 0
1962 0.3 0 0 0 0 0.016 0.349 0
1963 0.29 1 1 0 0 0.015 0.289 0
1964 0.29 1 1 0 0 0.010 0.289 0
1965 0.29 1 0 0 0 0.013 0.349 0
1966 0.29 1 0 0 0 0.021 0.421 0
1967 0.29 1 1 0 0 0.023 0.379 0
1968 0.29 1 0 0 0 0.027 0.367 0
1969 0.29 0 0 0 0 0.029 0.368 0
1970 0.28 0 1 0 0 0.027 0.371 0
1971 0.28 0 0 0 0 0.026 0.326 0
1972 0.27 0 2 0 0 0.027 0.344 0
1973 0.27 0 0 0 0 0.040 0.402 0
1974 0.26 1 1 0 0 0.036 0.373 0
1975 0.27 1 0 0 0 0.031 0.291 0
1976 0.26 1 1 0 0 0.036 0.288 0
1977 0.26 0 0 0 0 0.041 0.301 0
1978 0.26 0 1 0 0 0.049 0.314 0
1979 0.26 0 1 0 0 0.045 0.270 0
1980 0.26 0 1 0 0 0.031 0.212 0
1981 0.26 1 0 0 0 0.031 0.191 0
1982 0.26 1 1 0 0 0.030 0.177 0
1983 0.26 1 1 0 0 0.034 0.186 0
1984 0.25 1 0 0 0 0.037 0.191 0
1985 0.25 1 0 0 0 0.037 0.184 0
1986 0.25 1 1 0 0 0.047 0.240 0
1987 0.25 1 1 0 1 0.056 0.244 1
1988 0.25 1 0 1 0 0.060 0.242 1
1989 0.24 1 1 0 0 0.061 0.238 1
1990 0.23 1 1 0 0 0.056 0.217 1
1991 0.22 1 1 0 0 0.059 0.212 1
1992 0.22 1 0 1 1 0.051 0.185 1
1993 0.23 0 1 0 0 0.051 0.186 1
- 286 -
1994 0.23 0 2 0 0 0.056 0.191 1
1995 0.23 0 0 0 0 0.085 0.253 1
1996 0.23 0 2 1 0 0.081 0.219 1
1997 0.23 0 0 0 1 0.073 0.197 1
1998 0.23 1 2 0 0 0.059 0.176 1
1999 0.23 1 0 0 0 0.076 0.210 1
2000 0.24 1 2 1 0 0.083 0.215 1
2001 0.24 0 1 0 0 0.079 0.205 1
2002 0.24 0 0 0 1 0.079 0.189 1
2003 0.24 0 1 0 0 0.083 0.190 1
2004 0.25 0 0 1 0 0.087 0.185 1
2005 0.26 0 1 0 0 0.086 0.175 1
2006 0.26 0 1 0 0 0.063 0.122 1
2007 0.26 0 1 0 1 0.061 0.112 1
2008 NA 0 1 1 0 0.058 0.104 1
- 287 -
Appendix 3. Negotiations, Meetings, and Summits
Negotiations, Meetings, and Summits
Year Note Coding Decision Basis
1960 Negotiation Failure 1 No talks from April 1960 to January 1961.
1961 Ongoing Negotiation Failure 1 Failure of talks on May 16, 1961.
1962 Negotiation Success 0 Kim-Ohira Memorandum.
1963 Negotiation Failure 1 No talks from early 1963 to December 1963.
1964
Negotiation Failure & Ongoing
Negotiation Failure
1 No talks from March 1964 to December 1964.
1965 Normalization Negotiation 0 Normalization Treaty.
1966 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1967 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1968 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1969 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1970 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1971 Highest-Level Possible Meeting 0 Sato in Korea.
1972 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1973 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1974 Highest-Level Possible Meeting 0 Tanaka in Korea.
1975 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1976 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1977 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1978 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1979 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1980 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1981 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1982 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1983 Highest-Level Possible Meeting 0 Nakasone in Korea.
1984 Highest-Level Possible Meeting 0 Chun in Japan.
1985 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1986 Highest-Level Possible Meeting 0 Nakasone in Korea.
1987 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1988 Highest-Level Possible Meeting 0 Takeshita in Korea.
1989 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
1990 Highest-Level Possible Meeting 0 Roh in Japan
1991 Highest-Level Possible Meeting 0 Kaifu in Korea
1992 Highest-Level Possible Meeting 0 Miyazawa in Korea; Roh in Japan
1993 Highest-Level Possible Meeting 0 Hosokawa in Korea
1994 Highest-Level Possible Meeting 0 Kim in Japan; Murayama in Korea
1995 Highest-Level Possible Meeting 0 Kim in Japan.
1996 Highest-Level Possible Meeting 0 Hashimoto in Korea.
- 288 -
1997 Highest-Level Possible Meeting 0 Kim in Japan.
1998 Highest-Level Possible Meeting 0 Obuchi in Korea; Kim in Japan.
1999 Highest-Level Possible Meeting 0 Obuchi in Korea.
2000 Highest-Level Possible Meeting 0 Mori in Korea; Kim in Japan twice.
2001 Highest-Level Possible Meeting 0 Koizumi in Korea.
2002 Highest-Level Possible Meeting 0 Koizumi in Korea; Kim in Japan.
2003 Highest-Level Possible Meeting 0 Koizumi in Korea; Roh in Japan.
2004 Highest-Level Possible Meeting 0 Roh in Japan.
2005 Meeting Failure 1
Koizumi in Korea; Prearranged December Summit
Cancelled
2006 Dokdo/Takeshima 3 Korean show of force/threat of force.
2007 Conflict 2 No Highest-Level Possible Meetings in Japan or Korea.
2008 Meeting Failure 1
Lee in Japan; no second prearranged "shuttle diplomacy"
summit of 2008.
- 289 -
BIBLIOGRAPHY
Acharya, Amitav. Constructing a Security Community in Southeast Asia: ASEAN and the
Problem of Regional Order. New York: Routledge, 2001.
Adler, Emanuel, and Michael Barnett. “Security Communities in Theoretical Perspective.” In
Security Communities, edited by Emanuel Adler and Michael Barnett, 3-28. Cambridge,
Cambridge University Press: 1998.
Adler, Emanuel. “Seizing the Middle Ground: Constructivism in World Politics.” European
Journal of International Relations 3, no. 3 (1997): 319-363.
Ahn, Byung-joon. “Japanese Policy Toward Korea.” In Japan’s Foreign Policy After the Cold
War: Coping with Change, edited by Gerald L Curtis, 263-273. New York: M. E. Sharpe,
1993.
Allison, Graham T. “Conceptual Models and the Cuban Missile Crisis.” American Political
Science Review 63, no. 3 (September 1969): 689-718.
Aristotle. The Politics, edited by Trevor J. Saunders and translated by T. A. Sinclair. 1981.
Reprint, London: Penguin Books, 1992.
Armstrong, Charles K., Gilbert Rozman, Samuel S. Kim, and Stephen Kotkin, eds. Korea at the
Center: Dynamics of Regionalism in Northeast Asia. Armonk, NY: M.E. Sharpe, Inc.,
2006.
Ashforth, Blake E., and Fred Mael. “Social Identity Theory and the Organization.” The
Academy of Management Review 14, no. 1 (January 1989): 20-39.
Axelrod, Robert, and Robert Keohane. “Achieving Cooperation under Anarchy: Strategies and
Institutions.” World Politics 38, no. 1 (October 1985): 226-254.
Benedict, Ruth. The Chrysanthemum and the Sword: Patterns of Japanese Culture. 1946.
Reprint, New York: Houghton Mifflin, 2005.
Bennett, Scott D., and Allan Stam. “EUGene: A Conceptual Manual.” International
Interactions 26 (2000):179-204.
Berger, Thomas U. “Norms, Identity, and National Security in Germany and Japan.” In The
Culture of National Security: Norms and Identity in World Politics. Edited by Peter J.
Katzenstein, 317-356. New York: Columbia University Press, 1996.
Berger, Thomas U. “Power and Purpose in Pacific East Asia: A Constructivist Interpretation.”
In International Relations Theory and the Asia-Pacific, edited by G. John Ikenberry and
Michael Mastanduno, 297-419. New York: Columbia University Press, 2003).
- 290 -
Bermudez, Joseph S. Jr. “Occasional Paper No. Two: A History of Ballistic Missile
Development in the DPRK.” Center for Nonproliferation Studies, Monterey Institute of
International Studies, 1999.
Bially Mattern, Janice. Ordering International Politics: Identity, Crisis, and Representational
Force. New York: Routledge, 2008.
Bormann, Nils-Christian, and Matt Golder. "Democratic Electoral Systems Around the World,
1946-2011." Electoral Studies (2013).
Bridges, Brian. Japan and Korea in the 1990s: From Antagonism to Adjustment. Cambridge,
Great Britain: Edward Elgar Publishing Ltd, 1993.
Bureau of Oceans and International Environmental and Scientific Affairs. “Limits in the Seas,
no. 121, Straight Baseline and Territorial Sea Claims: South Korea.” Washington, DC:
US Department of State, September 1998.
http://www.state.gov/documents/organization/57685.pdf (accessed September 13, 2011).
Callahan, William A. “National Insecurities: Humiliation, Salvation, and Chinese Nationalism.”
Alternatives 29 (2004): 199-218.
Carr, E.H. The Twenty Year ’s Crisis, 1919-1939: An Introduction to the Study of International
Relations. 1981. Reprint, New York: Palgrave, 2001.
Cha, Victor D. Alignment Despite Antagonism: The US-Korea-Japan Security Triangle.
Stanford: Stanford University Press, 1999.
Cheong, Sung-Hwa. The Politics of Anti-Japanese Sentiment in Korea: Japanese-South Korean
Relations Under American Occupation, 1945-1952. Westport, CT: Greenwood Press,
1991.
Cherry, Judith. “Killing Five Birds with One Stone: Inward Foreign Direct Investment in Post-
Crisis Korea.” Pacific Affairs 79, no.1 (Spring, 2006): 9-27.
Chosun Ilbo Archive. http://srchdb1.chosun.com/pdf/i_archive/.
Congressional Research Service. North Korea: Chronology of Provocations, 1950-2003, by
Dick Nanto. Washington D.C., March 2000.
Copeland, Dale C. “Economic Interdependence and War: A Theory of Trade Expectations.”
International Security 20, no. 4 (Spring 1996): 5-41.
Cordesman, Anthony H., and Robert Hammond. “The Military Balance in Asia: 1990-2011: A
Quantitative Analysis.” Center for Strategic and International Studies, 2011.
- 291 -
Crawford, Neta C. “The Passion of World Politics: Propositions on Emotion and Emotional
Relationships.” International Security 24, no. 4 (Spring 2000): 116-156.
Dudden, Alexis. Troubled Apologies Among Japan, Korea, and the United States. New York:
Columbia University Press, 2008.
Er, Lam Peng. “The Apology Issue: Japan’s Different Approaches Toward China and South
Korea.” American Asian Review 20 (Fall 2002): 31-54.
Evera, Stephen Van. “Hypotheses on Nationalism and War.” International Security 18, no. 4
(Spring 1994): 5-39.
Finnemore, Martha, and Kathryn Sikkink. “International Norm Dynamics and Political Change.”
International Organization 52, no. 4 (Autumn, 1998): 887-917.
Foreign Broadcast Information Service Daily Reports. http://www.newsbank.com/.
Frijda, Nico H. “The Lex Talionis: On Vengeance.” In Emotions: Essays on Emotion Theory,
edited by Stephanie H M van Goozen, Nanne E van de Poll, and Joseph A Sergeant, 263-
289. Hillsdale, NJ: Lawrence Erlbaum, 1994.
Gochman, Charles S., and Zeev Maoz. “Militarized Interstate Disputes, 1816-1976: Procedures,
Patterns, and Insights.” Journal of Conflict Resolution 28, no. 4 (December 1984): 585-
616.
Green, Michael. Japan’s Reluctant Realism: Foreign Policy Challenges in an Era of Uncertain
Power. 2001. Reprint, New York: Palgrave, 2003.
Han, Sungjoo. “South Korea 1978: The Growing Security Dilemma.” Asian Survey 19, no. 1
(January 1979): 41-50.
Hardin, Russell, ed. Distrust. New York: Russell Sage Foundation, 2004.
Harelli, Schlomo, and Brian Parkinson. “What’s Social about Social Emotions?” Journal for
the Theory of Social Behavior 38, no. 2 (2008): 131-156.
Hayes, Jarrod. Constructing National Security: US Relations with China and India. New York:
Cambridge University Press, 2013.
Hideki, Kajimura. “The Question of Takeshima/Tokdo.” Korea Observer 28, no. 3 (Autumn
1997): 423-475.
Hirschman, Albert O. The Passions and the Interests: Political Argument for Capitalism before
its Triumph. Princeton, NJ: Princeton University Press, 1977.
Hobbes, Thomas. Leviathan, edited by C.B. Macpherson. New York: Penguin, 1968.
- 292 -
Hoffman, Aaron M. “A Conceptualization of Trust in International Relations.” European
Journal of International Relations 8, no. 3 (2002): 375-401.
Hopf, Ted. Reconstructing the Cold War: The Early Years, 1945-1958. New York: Oxford
University Press, 2012.
Hopf, Ted. Social Construction of International Politics: Identities & Foreign Policies, Moscow,
1955 & 1999. Ithaca, NY: Cornell University Press, 2002.
Huddy, Leonie. “Contrasting Theoretical Approaches to Intergroup Relations.” Political
Psychology 25, no. 6 (Dec. 2004): 947-967.
International Monetary Fund Direction of Trade Statistics.
http://www.imf.org/external/pubs/cat/longres.cfm?sk=19305.0.
Kant, Immanuel. “Perpetual Peace.” In Kant: On History, edited and translated by Lewis White
Beck. Upper Saddle River, NJ: Prentice Hall Inc., 2001.
Katzenstein, Peter J. “Taming of Power: German Unification, 1989-1990.” In Past as Prelude,
History in the Making of a New World Order, edited by Meredith Woo-Cumings and
Michael Lorriaux, 59-82. Boulder: Westview, 1993.
Katzenstein, Peter J. Cultural Norms & National Security: Police and Military in Postwar Japan.
Ithaca, NY: Cornell University Press, 1996.
Katzenstein, Peter J., ed. The Culture of National Security: Norms and Identity in World Politics.
New York: Columbia University Press, 1996.
Keohane, Robert O. “Multilateralism: An Agenda for Research.” International Journal 45, no.
4 (Autumn 1990): 731-764.
Keohane, Robert O. After Hegemony: Cooperation and Discord in the World Political Economy.
Princeton Classic ed. Princeton, NJ: Princeton University Press, 2005.
Keohane, Robert O., and Joesph S Nye, Jr. “Review: Power and Interdependence Revisited.”
International Organization 41, no. 4 (Autumn 1987): 725-753.
Keohane, Robert O., and Joseph S Nye, Jr. “Power and Interdependence in the Information Age.”
Foreign Affairs 77, no. 5 (September/October 1998): 81-94.
Keohane, Robert O., and Joseph S. Nye. Power and Interdependence. 2
nd
ed. Glenville, IL:
Scott Foresman, 1989.
Keohane, Robert O., and Joseph S. Nye. Power and Interdependence. 3
rd
ed. New York:
Longman, 2001.
- 293 -
Keohane, Robert O., and Lisa L. Martin, “The Promise of Institutional Theory,” International
Security 20, no. 1 (Summer 1995): 39-51.
Keohane, Robert. “Theory of World Politics: Structural Realism and Beyond.” In Neorealism
and its Critics, edited by Robert O. Keohane, 158-203. 6
th
ed. New York: Columbia
University Press, 1986.
Kim, Samuel. The Two Koreas and the Great Powers. New York: Cambridge University Press,
2006.
Koh, Byung Chul. Between Discord and Cooperation: Japan and the Two Koreas. Seoul:
Yonsei Univeristy Press, 2007.
Koo, Min Kyo. “Scramble for the Rocks: The Disputes over the Dokdo/Takeshima,
Senkaku/Diaoyu, and Paracel and Spratly Islands.” PhD diss., University of California,
Berkeley, 2005.
Korean Peninsula Energy Development Organization. “Agreed Framework between the United
States of America and the Democratic People’s Republic of Korea.” Geneva, October 21,
1994.
Krasner, Stephen D. “Structural Causes and Regime Consequences: Regimes as Intervening
Variables.” In International Regimes, edited by Stephen D. Krasner, 1-21. 1983.
Reprint, Ithaca, NY: Cornell University Press, 1984.
Kydd, Andrew H. Trust and Mistrust in International Relations. Princeton: Princeton
University Press, 2005.
Lahno, Brend. “On the Emotional Character of Trust.” Ethical Theory and Moral Practice 4, no.
2 (June 2001): 171-189.
Lake, David A. “The State and American Trade Strategy in the Pre-Hegemonic Era.” In The
State and American Foreign Policy, edited by John G. Ikenberry, David A. Lake, and
Michael Mastanduno, 33-58. Ithaca, NY: Cornell University Press, 1988.
Lebow, Richard Ned. A Cultural Theory of International Relations. 2008. Reprint, New York:
Cambridge University Press, 2010.
Lee, Changsoo. “The Politics of Repatriation.” In Koreans in Japan: Ethnic Conflict and
Accommodation, edited by Changsoo Lee and George de Vos, 91-109. Los Angeles:
University of California Press, 1981.
Lee, Chong-Sik. Japan and Korea: The Political Dimension. Stanford: Stanford University
Press, 1985.
- 294 -
Lee, Dong-jun. “From the Secret ‘Korean Minute’ to the Open ‘Korea Clause’: The United
States and Japan and the Security of the Republic of Korea.” Asian Perspective 36
(2012): 123-145.
Lee, Seo-Hang. “Korea’s Claims to Maritime Jurisdiction.” Korean Journal of Comparative
Law 18 (1990): 62-89.
Legro, Jeffrey W., and Andrew Moravcsik. “Is Anybody Still a Realist.” International Security
24, no. 2 (Autumn 1999): 5-55.
Leidner, Bernhard, Hammad Sheikh, and Jeremy Ginges. “Affective Dimensions of Intergroup
Humiliation.” Plos One 7, no. 9 (September 2012): 1-6.
Lerner, Jennifer S., and Dacher Keltner. “Beyond Valence: Toward a Model of Emotion-
Specific Influences on Judgment and Choice.” Cognition and Emotion 14, no. 4 (2000):
473-493.
Lerner, Jennifer S., and Dacher Keltner. “Fear, Anger, and Risk.” Journal of Personality and
Social Psychology 81, no. 1 (2001): 146-159.
Lerner, Jennifer S., Roxana M. Gonzalez, Deborah A. Small, Baruch Fischhoff. “Effects of Fear
and Anger of Perceived Risks of Terrorism: A National Field Experiment.”
Psychological Science 14, no. 2 (March 2003): 144-150
Levin, Norman, and Yong-Sup Han. Sunshine in Korea: The South Korean Debate Over Policies
Toward North Korea. Santa Monica, CA: Rand, 2002.
Lewis J. David, and Andrew Weigert. “Trust as a Social Reality.” Social Forces 63, no. 4 (June
1984): 967-985.
Lind, Jennifer. Sorry States: Apologies in International Relations. Ithaca, NY: Cornell
University Press, 2008.
Machiavelli, Niccolo. “The Discourses.” In The Prince and the Discourses, edited by Max
Lerner, 99-540. New York: Random House, 1950.
Machiavelli, Niccolo. “The Prince.” In The Prince and the Discourses, edited by Max Lerner,
2-98. New York: Random House, 1950.
Macpherson, C.B. “Reading 2.1: The Theory of Human Nature in Society.” In Reading
Political Philosophy: Machiavelli to Mill, edited by Nigel Warburton, Jon Pike, and
Derek Matravers, 100-105. New York: Routledge, 2000.
Marhall, Monty G., and Ted Robert Gurr. “Polity IV Project: Political Regime Characteristic
and Transitions, 1800-2012. http://www.systemicpeace.org/polity/polity4.htm.
- 295 -
Marumoto, Mika. “Project Report: Democratic People’s Republic of Korea Economic Statistics
Project (April-December 2008).” Presented to Korea Development Institute School of
Public Policy and Management and the DPRK Economic Forum, U.S.-Korea Institute at
Johns Hopkins University-School of Advanced International Studies, Washington, D.C.,
March 2009.
Mascolo, Michael F., and Sharon Griffin. “Alternative Trajectories in the Development of
Anger.” In What Develops in Emotional Development? 219-249. New York: Plenum,
1998.
McCullough, Michael E., Robert A. Emmons, Shelley D. Kilpatrick, and David B. Larsen. “Is
Gratitude a Moral Affect?” Psychological Bulletin 127, no. 2 (2001): 249-266.
Mearsheimer, John J. “The False Promise of International Institutions.” International Security
19, no. 3 (Winter 1994-1995): 5-49.
Mearsheimer, John. The Tragedy or Great Power Politics. New York: W.W. Norton &
Company, 2001.
Mercer, Jonathan. “Anarchy and Identity.” International Organization 49, no.2 (Spring 1995):
229-252.
Mercer, Jonathan. “Approaching Emotion.” Paper presented at the International Studies
Association conference in San Diego, CA, April 25, 1996.
Merriam-Webster Online Dictionary. http://www.merriam-webster.com/.
Miller, William Ian. Humiliation: And Other Essays on Honor, Social Discomfort, and Violence.
Ithaca, NY: Cornell University Press, 1993.
Ministry of Foreign Affairs of Japan. Diplomatic Blue Book: Japan ’s Diplomatic Affairs. 1978
through 2011. http://www.mofa.go.jp/policy/other/bluebook/.
Ministry of Internal Affairs and Communication of Japan, Statistics Bureau. Statistics Japan.
http://www.stat.go.jp/english/index.htm.
Ministry of Unification of Korea. Data and Statistics. http://www.unikorea.go.kr/.
Moore, Thomas. “China’s International Relations: The Economic Dimension.” In The
International Relations of Northeast Asia, edited by Samuel S. Kim, 101-134. Lanham,
MD: Rowman & Littelfield Publishers, Inc., 2004.
Moravcsik, Andrew. “Taking Preferences Seriously: A Liberal Theory of International Politics.”
International Organization 51, no. 4 (Autumn 1997): 513-553.
- 296 -
Morgenthau, Hans. Politics Among Nations: The Struggle for Power and Peace. New York:
Alfred Knopf, 1978.
Morris-Suzuki, Tessa. Exodus to North Korea: Shadows from Japan ’s Cold War. Lanham, MD:
Rowman & Littlefield, 2007.
Oneal, John R., and Bruce M. Russett. “The Classical Liberals Were Right: Democracy,
Interdependence, and Conflict, 1950-1985.” International Studies Quarterly 41 (1997):
267-284.
Oye, Kenneth. “Cooperation under Anarchy: Cooperation and Strategies.” World Politics 38,
no. 1 (October 1985): 1-24.
Palshikar, Sanjay. “Understanding Humiliation.” Economic and Political Weekly 40, no. 51
(Dec 2005): 5428-5432.
Parkinson, Brian. “Emotions are Social.” British Journal of Psychology 87 (1996): 663-683.
Pinkston, Daniel A. “The North Korean Ballistic Missile Program.” Strategic Studies Institute,
US Army War College, February 2008.
Plato. “Phaedo.” In Classics of Western Philosophy, edited by Stephen M Cahn, 49-81. 6
th
ed.
Indianapolis, IN: Hackett Publishing, 2002.
Posen, Barry R. “Nationalism, the Mass Army, and Military Power.” International Security 18,
no. 2 (Autumn 1993): 80-124.
Powell, Robert. “Absolute and Relative Gains in International Relations Theory.” American
Political Science Review 85, no. 4 (December 1991): 1303-1320.
Prime Minister of Japan and His Cabinet, http://www.kantei.go.jp/foreign/cabinet/0031-
60_e.html.
Rathbun, Brian C. Trust in International Cooperation: International Security Institutions,
Domestic Politics and American Multilateralism. New York: Cambridge University
Press, 2012.
Ray, James Lee. “Does Democracy Cause Peace?” Annual Review of Political Science 1 (1998):
27-46.
Risse-Kappen, Thomas. “Collective Identity in a Democratic Community: The Case of NATO,”
in The Culture of National Security: Norms and Identity in World Politics. Edited by
Peter J. Katzenstein, 357-399. New York: Columbia University Press, 1996
Russet, Bruce M. Grasping the Democratic Peace: Principles for a Post Cold War World.
Princeton: Princeton University Press, 1993.
- 297 -
Schneider, Claudia. “The Japanese History Textbook Controversy in East Asian Perspective.”
The Annals of the American Academy of Political and Social Science 617 (May 2008):
107-122.
Schrock-Jacobson, Gretchen, “The Violent Consequences of the Nation: Nationalism and the
Initiation of Interstate War.” Journal of Conflict Resolution 56, no. 5 (2012): 825-852.
Shaver, Phillip, Judith Schwartz, Donald Kirson, and Cary O'Connor. “Emotion Knowledge:
Further Exploration of a Prototype Approach.” Journal of Personality and Social
Psychology 52, No. 6 (1987): 1061-1086.
Shibuichi, Daiki. “The Yasukuni Shrine Dispute and the Politics of Identity in Japan: Why All
the Fuss?” Asian Survey 45, no. 2 (March/April 2005): 197-215.
Shirk, Susan L. China: Fragile Superpower. New York: Oxford University Press, 2007.
Snyder, Glenn H. “Alliance Theory: A Neorealist First Cut.” Journal of International Affairs 44,
No. 1 (Spring 1990): 103-123.
Snyder, Glenn H. Alliance Politics. Ithaca, NY: Cornell University Press, 1997.
Snyder, Richard C., H.W. Bruck, and Burton Sapin. Foreign Policy Decision-Making: Revisited.
New York: Palgrave Macmillan, 2002.
Solomon, Robert C. “Sympathy and Vengeance: The Role of the Emotions in Justice.” In
Emotions: Essays on Emotion Theory, edited by Stephanie van Gooze, Nanne Van de
Poll, and Joe Sergeant, 291-311. Hillsdale, NJ: Lawrence Erlbaum, 1994.
Solomon, Robert C. A Passion for Justice: Emotions and the Origins of the Social Contract.
New York: Addison-Wesley, 1990.
Spinner-Halev, Jeff, and Elizabeth Theiss-Morse. “National Identity and Self-Esteem.”
Perspectives on Politics 1, no. 3 (Sep. 2003): 515-532.
Sreedhar, Susan. Hobbes on Resistance: Defying the Leviathan. Cambridge: Cambridge
University Press, 2010.
Suh, Jae-Jung. “Assessing the Military Balance in Korea.” Asian Perspective 29, no. 4 (2004):
63-88.
Tajfel, H., and J.C. Turner. “The Social Identity Theory of Intergroup Behaviour.” In
Psychology of Intergroup Relations, edited by S. Worchel and W.G. Austin, 7-24. 2
nd
ed.
Chicago: Nelson-Hall, 1986.
- 298 -
Tajfel, Henri, ed. Social Identity and Intergroup Relations. 1982. Reprint, New York:
Cambridge University Press, 2010.
The World Food Program Food Information System. http://www.wfp.org/fais/.
Thucydides. History of the Peloponnesian War, edited by M. I. Finley and translated by Rex
Warner. 1954. Revised ed. New York: Penguin Books, 1972.
Tsang, Jo-An. “Gratitude and Prosocial Behaviour: An Experimental Test of Gratitude.”
Cognition and Emotion 20, no. 1 (2006): 138-148.
Turner, Jonathan H., and Jan E. Stets. “Sociological Theories of Human Emotions.” Annual
Review of Sociology 32 (2006): 25-52.
UN National Accounts Main Aggregate Database. http://unstats.un.org/unsd/snaama/dnlList.asp.
US Department of Defense. Annual Report on the Military Situation on the Korean Peninsula.
Washington D.C., September 2000.
Vamik D. Volkan, “Transgenerational Transmissions and Chosen Traumas: An Aspect of Large-
Group Identity,” Group Analysis 34, no. 1 (2001): 79-97.
Walt, Stephen. The Origins of Alliances. Ithaca, NY: Cornell University Press, 1987.
Waltz, Kenneth N. Man, the State, and War:A Theoretical Analysis. 1954. Reprint, New York:
Columbia University Press, 2001.
Waltz, Kenneth. “Political Structures.” In Neorealism and its Critics, edited by Robert O.
Keohane, 47-69. 6
th
ed. New York: Columbia University Press, 1986.
Waltz, Kenneth. “Reductionist and Systemic Theories.” In Neorealism and its Critics, edited by
Robert O. Keohane, 47-69. 6
th
ed. New York: Columbia University Press, 1986.
Waltz, Kenneth. Theory of International Politics. Reading, Mass: Addison-Wesley, 1979.
Watkins, Philip C., Jason Scheer, Melinda Ovnicek, and Russell Kolts. “The Debt of Gratitude:
Dissociating Gratitude and Indebtedness.” Cognition and Emotion 20, no. 2 (2006): 217-
241.
Weldes, Jutta. Constructing National Interests: The United States and the Cuban Missile Crisis.
Minneapolis, MN: University of Minnesota Press, 1999.
Wendt, Alexander. Social Theory of International Politics. Cambridge: Cambridge University
Press, 1999.
Wohlforth, William. “A Matter of Honor.” International Theory 2, no. 3 (2010): 468-474.
- 299 -
World Trade Organization, Statistics Database.
http://stat.wto.org/Home/WSDBHome.aspx?Language=E.
Yoon, Bang-Soon L. “Imperial Japan’s Comfort Women from Korea: History & Politics of
Silence-Breaking.” The Journal of Northeast Asian History 7, no. 1 (Summer 2010): 5-
39.
Yoon, Tae-Ryong. “Fragile Cooperation: Net Threat Theory and Japan-Korea-U.S. Relations.”
PhD diss., Columbia University, 2006.
Abstract (if available)
Abstract
Expectations and emotions matter in international relations. In this dissertation I argue Japan’s and South Korea’s respective alignment security expectations play a critical role in exacerbating and ameliorating the “history issue” marring their bilateral relationship. More specifically, Japan and the ROK each expect that the Other will uphold the security interests of the Self when it comes to establishing North Korea policy. During moments in which their North Korea policies are aligned, the quasi-allies tend to enjoy warm ideological relations. When their Pyongyang policies diverge, however, ideological affairs within the Japan-Korea alignment suffer. ❧ The goal of my research is two-fold. First, I aim to quantify the influence on the history issue of various phenomena such as North Korea policy, elections, trade dependence, and the strength of the US commitment to East Asia. Of the nine independent variables I analyze in various theoretically informed configurations, only two return statistically significant results: ROK National Assembly elections and North Korea policy convergence/divergence. Each substantially influences the level of historical animosity. ❧ My second objective is to explain how Seoul’s and Tokyo’s North Korea policies lead to historical animosity. While I deduce several theoretical models to explain causality, I find only the third one—the collective self-esteem model—adequately explains all facets of the relationship between alignment security expectations and ideological cooperation and friction. Specifically, prior ideological divide means Japan and South Korea must undertake a considerable trusting act in order to successfully coordinate alignment security policy. Warm ideological relations emerge through the ceremonies and promulgations associated with the trusting act. ❧ Regrettably, the security expectations fostered via the trusting act are unrealistic, and with seeming inevitability Japan’s and South Korea’s Pyongyang policies diverge. Ideological relations invariably deteriorate, because of the humiliation suffered by those vested materially and emotionally in the trusting act. In order to diminish the humiliation and preserve their individual and collective self-esteems, these men and women disparage and denigrate the Other. While disparagement and denigration preserve the moral superiority and trustworthiness of the Self over the Other, these acts also directly cause bilateral ideological friction and a deterioration of the history issue.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Status seeking in hierarchy: Korea and Vietnam under Chinese hegemony in early modern Asia
PDF
When passions run high: emotions and the communication of intentions in face-to-face diplomacy
Asset Metadata
Creator
Walker, David A.
(author)
Core Title
The Japan-South Korea history issue: the emotional outcome of trusting acts and met/unmet alignment security expectations
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Politics and International Relations
Publication Date
11/27/2013
Defense Date
08/29/2013
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
alignment politics,alliance politics,historical animosity,history issue,international relations theory,Japan,OAI-PMH Harvest,psychological constructivism,security expectations,social emotions,South Korea,Trust
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Kang, David (
committee chair
), Carnevale, Peter (
committee member
), Hymans, Jacques (
committee member
)
Creator Email
bovoxdave@hotmail.com,dawalker@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-353654
Unique identifier
UC11288006
Identifier
etd-WalkerDavi-2190.pdf (filename),usctheses-c3-353654 (legacy record id)
Legacy Identifier
etd-WalkerDavi-2190.pdf
Dmrecord
353654
Document Type
Dissertation
Rights
Walker, David A.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
alignment politics
alliance politics
historical animosity
history issue
international relations theory
psychological constructivism
security expectations
social emotions