The Clash of Civilizations
The Clash of Civilizations
Samuel P. Huntington argued after the Cold War that the clash of superpowers would be replaced by the clash of civilizations in the post-Cold War period. His argument resonates with many, who see the rise of China or Muslim terrorism as evidence that Western civilization clashes with Chinese and Islamic ones, respectively. Opponents argue that the clash of civilization argument is not useful to understand world politics. It embraces a flawed conception of civilization, which treats civilization as fixed, and it also dangerous because it risks becoming a self-fulfilling prophecy. Which side is right?
in this assignement you will support Huntington point using the three article posted extensively ....others sources can be used but not that much those three sources presented have to be use extensively though.... you have to do an independent research and will put together a paper (maximum 10 pages long, 12 Times New Roman, double spaced), which represents supporting the author huthington argument in a clear and persuasive manner.Other reading can be used and should be presented but not as extensively as the one attached
The structure of the paper:
a) Introduction (one paragraph)
b) Main arguments, including theoretical concepts and supporting facts
c) Discussion of one main counterargument to the position which an argument agaisnt the presented samuel huthington argument up about (one paragraph)
d) Conclusion (one paragraph)
International Studies Quarterly ~2001! 45, 317338.
Clear and Present Strangers:
The Clash of Civilizations
and International Conflict
Errol A. Henderson
Wayne State University
and
Richard Tucker
Vanderbilt University
Huntingtons ~1993a, 1993b, 1996! clash of civilizations thesis suggests
that states belonging to different civilizations are more likely to become
involved in conflict with one another. To evaluate the empirical accuracy of Huntingtons claims, we examined the relationship between
civilization membership and interstate war between 1816 and 1992. We
find that civilization membership was not significantly associated with
the onset of interstate war during the Cold War era ~1946 1988!, which
is consistent with one aspect of Huntingtons thesis; however, we also
find that for the preCold War period ~1816 1945! states of similar
civilizations were more likely to fight each other than were those of
different civilizations, which contradicts Huntingtons thesis. Most importantly, our analysis reveals that during the postCold War era ~1989 1992!,
the period in which Huntington contends that the clash of civilizations
should be most apparent, civilization membership was not significantly
associated with the probability of interstate war. All told, our findings
challenge Huntingtons claims and seriously undermine the policy recommendations that devolve from his clash of civilizations thesis.
A growing literature has emerged on the impact of cultural factors on world
politics ~e.g., Chay, 1990; Mazrui, 1990; Ryan, 1990; Midlarsky, 1992; Huntington,
1993a, 1993b, 1996; Gurr, 1994; Gurr and Harff, 1994; Carment and James, 1997;
Henderson, 1995, 1997, 1998!. This renewed focus on culture has become so
pronounced that ethnicity has become, for some, the new master explanatory
variable in world politics ~Holsti, 1997:8!. Other studies evince an almost Dr.
Strangelove-like quality in their apocalyptic vision of seething cultural cauldrons, leading one author to conclude that animosity among ethnic groups is
beginning to rival the spread of nuclear weapons as the most serious threat to
Authors note: Earlier versions of this article were presented to the International Studies Association, Washington, DC ~February 1999! and the American Political Science Association, Atlanta, GA ~September 1999!. The
authors thank Don Babai, Andrew Enterline, Cynthia Ervin, Kristian Gleditsch, Pat James, Volker Krause, Anthony
Loh, James Lee Ray, J. David Singer, Michael Tomz, and John Vasquez for comments on earlier drafts.
2001 International Studies Association.
Published by Blackwell Publishers, 350 Main Street, Malden, MA 02148, USA, and 108 Cowley Road, Oxford OX4 1JF, UK.
318
Clear and Present Strangers
peace that the world faces ~Maynes, 1993:5!.1 In this context, cultural realists
provide a theoretical rationale for explicating the putative processes at work in
cultural conflagrations occurring throughout the globe.
The clearest and most controversial articulation of cultural realism is found in
Huntingtons ~1993a, 1993b, 1996! clash of civilizations thesis, which posits that
conflict is more likely to occur between states of different civilizations ~see
Henderson, 1997:657; Schulman, 1998:304 305!.2 Huntingtons thesis has become
the centerpiece of scholarly discourse on the impact of cultural factors on international conflict, as evidenced by its having been translated into several languages while becoming one of the most cited articles to appear in Foreign Affairs
in the past decade. In addition, it has become increasingly influential in policy
circles, leading Roman Herzog ~1999!, the former president of the FRG, and four
other renowned scholars to pen a recent volume entitled Preventing the Clash of
Civilizations, which is aimed at providing strategies to prevent the scenario of a
clash of civilizations from becoming reality.
In this study we examine Huntingtons thesis on the relationship between
civilization membership and the likelihood of interstate war. In the first section
we review Huntingtons basic arguments regarding intercivilizational conflict
and derive several propositions from Huntingtons clash of civilizations thesis.
The second and third sections of our study describe the research design used to
evaluate these propositions and contain the findings from our data analysis. In
the final two sections we present a discussion of the findings and suggest their
implications for subsequent research and policy.
The Clash of Civilizations Thesis
The Main Premise of Huntingtons Thesis
The central contention of Huntingtons ~1993a, 1993b, 1996! clash of civilizations ~hereafter, CoC! thesis is that states belonging to different civilizations are
more likely to fight each other, while those belonging to the same civilization are
less likely to fight each other. According to Huntington ~1996!, the CoC has
emerged in the postCold War era as a result of several factors, including: ~1! the
increased interaction among peoples of different civilizations; ~2! the deWesternization and indigenization of elites in non-Western states; ~3! increased
economic regionalization, which heightens civilization consciousness; and ~4! a
global resurgence of religious identity, which is replacing diminishing local and
state-based identities. In addition, demographic and economic changes have
shifted the balance of power among civilizations as the capabilities of nonWestern statesespecially Asian and Islamic statesare rising to challenge Western hegemony. The interaction of these factors has resulted in the increased
salience of civilization membership in global politics. Since civilizational characteristics are basic and essential, civilizational differences are increasingly likely to
generate conflict. The result is that cultural factors have replaced ideological
ones as the major source of conflict in world politics.
1
A growing number of scholars now challenge the view that even prototypal ethnic conflicts such as those in
the former Yugoslavia, Rwanda, and Liberia are ethnic conflicts at all ~see Banac, 1992:143; Pfaff, 1993; Gagnon,
1995:334; Destexhe, 1994; Schlichte, 1994; Bowen, 1997!. Henderson ~1997:650653! maintains that ethnic conflict is often a misleading moniker for a variety of conflicts driven by a diversity of issues.
2
Cultural realism is a derivative of realism that focuses on cultural factors as determinants of a states foreign
policy; moreover, it posits that, ceteris paribus, states belonging to different cultural backgrounds are more likely
to become involved in conflict than those that share a common cultural background. This view is contrasted with
cultural idealism ~see Henderson, 1997!. This conceptualization of cultural realism should not be confused with
Johnstons ~1995!, which derives from his analysis of strategic culture.
Errol A. Henderson and Richard Tucker
319
For Huntington ~1996:43!, a civilization is the highest cultural grouping of people and the broadest level of cultural identity people have short of that which distinguishes humans from other species. He maintains that a civilization is a culture
writ large ~p. 41! and is the biggest we within which we feel culturally at home
as distinguished from all the other thems out there ~p. 43!. Civilizations are quite
diverse in composition and may involve a large number of people, such as Chinese civilization, or a very small number of people such as the Anglophone Caribbean ~p. 43!. The central defining characteristic of a civilization is its religion
~p. 47!, hence, the major civilizations in human history have been closely identified with the worlds great religions ~p. 42!. These civilizations include the Sinic,
Japanese, Hindu, Islamic, Orthodox, Western, Latin American, ~apparently! Buddhist, and possibly African civilizations ~pp. 4748!.
Since shared religion is the single most important indicator of a civilization,
Huntington maintains that intercivilizational clashes are usually conflicts between
peoples of different religions ~p. 253!. That religious difference should be the
fulcrum on which the CoC thesis rests stems from Huntingtons view that religion is possibly the most profound difference that can exist between people.
Therefore, he contends that warfare between states of different civilizations is
greatly enhanced by beliefs in different gods ~p. 254!. This leads to Huntingtons most concise statement of the CoC: @c# ivilizations are the ultimate human
tribes, and the clash of civilizations is tribal conflict on a global scale ~p. 207!.
The Theoretical Logic of the CoC Thesis
In its basic concern with the impact of culture on interstate conflict, Huntingtons CoC thesis is not wholly divorced from the arguments that are derived from
the major paradigms of world politics regarding this relationship. Such arguments have largely centered on the failure to recognize the basic right of national
~largely conceived as cultural! self-determination and freedom from alien rule as
a precipitant of international conflict. Morgenthau ~1985 @1948#:350!, for example, insisted that the wars of the nineteenth century grew out of either different
interpretations of the national principle or the refusal to accept it at all ~p. 350!.
Scholars across the realist-idealist continuum, from Morgenthau ~1985 @1948#:180!
to Wright ~1942:1002!, have discussed the role of cultural factors in exacerbating
international conflict. However, these scholars do not ascribe to cultural factors
the significance that Huntington does. More recently, scholars have given greater
attention to the impact of cultural factors in the decision-making process of
disputants in interstate conflicts. For example, building on the arguments of
Mansbach and Vasquez ~1981:186 231! and Vasquez ~1993:76 82!, Henderson
~1998! claims that where cultural factors become increasingly salient in disputes,
concrete stakes such as territory become loaded with transcendent values that
tend to make territorial issues indivisible and subsequent related disputes less
susceptible to compromise. For example, a territorial dispute may be resolved by
simply partitioning the disputed territory; however, it is quite another thing
altogether to partition Jerusalem to the satisfaction of the disputants. Huntington echoes this view:
Differences in material interest can be negotiated and often settled by compromise in a way cultural issues cannot. Hindus and Muslims are unlikely to resolve
the issue of whether a temple or a mosque should be built at Ayodhya by
building both, or neither, or a syncretic building that is both a mosque and a
temple. Nor can what might seem to be a straightforward territorial question
between Albanian Muslims and Orthodox Serbs concerning Kosovo or between
Jews and Arabs concerning Jerusalem be easily settled, since each place has deep
historical, cultural, and emotional meaning to both peoples. Cultural questions
like these involve a yes or no, zero-sum choice. ~1996:129 130!
320
Clear and Present Strangers
Implicit in Huntingtons argument is the conviction that disputes between
states of different civilizations are more likely to become culturally loaded in
such a way as to exacerbate intercivilizational conflict levels. Where Huntingtons
primordialist analysis departs from more instrumentalist views is his location of
these conflict-enhancing factors in the identities of the disputants themselves,
rather than in the nature of the issues in the dispute. The logic of this assumption is simple for Huntington: if difference in cultural identity is conflict laden,
and identity is largely exclusive and immutable, then ones identity in a dispute
is more important than ones position in a dispute; therefore, disputes related to
identity are less likely to be amenable to compromise. Huntington ~1996:289
290! implies that culturally dissimilar decisionmakers are likely to employ a
civilizational realpolitik ~i.e., a cultural realist! strategy in their intercivilizational disputes, which, in his view, increases the probability of escalation to a
fault-line war.3 Moreover, since civilization membership is rooted primarily in
religious identity, this further reduces the likelihood of compromise, and, in
fact, intensifies the conflict proneness of intercivilizational disputes.
Huntington ~1993a! suggests that compromise in identity conflicts is difficult
among religiously dissimilar disputants because of the often unreflective and
intransigent nature of territorial claims based on religious beliefs. By comparison, although a cultural characteristic such as spoken language is cumulative in
that one can learn to speak several languages, this is not the case with religion.
The exclusivity of religion is far more complete than that entailed by race or
ethnicity since @a# person can be half-French and half-Arab and simultaneously
even a citizen of two countries; however, @i# t is more difficult to be halfCatholic and half-Muslim ~p. 27!. The exclusivity of religious-based differences
derives from the fact that they reflect divergent views on fundamental issues of
human existence, such as the nature of life and the fundamental ordering principles of self, family, and society, as well as issues of eschatology and salvation. As
noted elsewhere, @u# nderwritten by God, political mobilization rooted in religious identification often takes on the aspect of a secular religion that routinely
requires human sacrifice ~Henderson, 1997:663664!.
Propositions
The most explicit proposition that one can derive from Huntingtons work is that
in the postCold War era, the most prevalent form of global conflict will occur
at the fault lines of the major civilizations.4 Huntington stated it unequivocally:
It is my hypothesis that the fundamental source of conflict in this new world
@the postCold War era# will not be primarily ideological or primarily economic.
The great divisions among humankind and the dominating source of conflict
will be cultural. Nation states will remain the most powerful actors in world
affairs, but the principal conflicts of global politics will occur between nations
and groups of different civilizations. The clash of civilizations will dominate
global politics. The fault lines between civilizations will be the battle lines of the
future. ~1993a:22!
3
While empirical evidence ~e.g., Leng, 1983! suggests that in repeated disputes with the same adversary the use
of coercive bargaining strategies by decisionmakers increases the likelihood of escalation to war, at present, no
evidence has been found to support Huntingtons civilizational realpolitik claims.
4
Although the present study is focused on interstate conflict, there is also an intrastate dimension of Huntingtons thesis, which posits that the CoC is operative within states where groups representing different civilizations
are more likely to fight each other. This intrastate dimension is not systematically examined in this article, although
previous empirical studies have examined aspects of it ~e.g., Gurr, 1994!.
Errol A. Henderson and Richard Tucker
321
The ascendance of cultural factors in world politics results largely from the
view that @s# tates define threats in terms of the intentions of other states, and
those intentions and how they are perceived are powerfully shaped by cultural
considerations ~p. 34!. Huntington insists that decisionmakers are much more
likely to see threats coming from states whose societies have different cultures
and hence which they do not understand and feel they cannot trust ~p. 34!.
Therefore, states are more likely to cooperate with and ally themselves with
states with similar or common culture and are more often in conflict with
countries of different culture ~p. 34!. As a response to the changing locus of
power in the system, the linkages among states that previously were a function of
basic security concerns ~Walt, 1987!, balance of power considerations ~Waltz,
1979!, or ideological orientation ~Siverson and Emmons, 1991; Simon and Gartzke,
1996; Thompson and Tucker, 1997a, 1997b! have come to reflect the shared
cultural orientations of the states. Hence, for Huntington ~1993a:31!, the Velvet
Curtain of culture has replaced the Iron Curtain of ideology and the core
states of the major civilizations are supplanting the two Cold War superpowers as
the principal poles of interaction and repulsion for other countries ~Huntington, 1996:155!. The result is that the wars of ideology that shaped the Cold War
landscape have given way to wars of civilizations in the postCold War era.
Huntington labels these wars between civilizations fault-line wars.
Further, in this new world, according to Huntington ~p. 125!, cultural identity
is the central factor shaping a countrys associations and antagonisms and, therefore, states increasingly define their interests in civilization terms ~p. 34!. In the
postCold War era, the question, Which side are you on? has been replaced by
the much more fundamental one, Who are you? and @e# very state has to have
an answer. For Huntington, @t# hat answer, its cultural identity, defines the states
place in world politics, its friends, and its enemies. Therefore, Huntingtons primary CoC thesis contends that:
Proposition 1.1:
During the postCold War era, intercivilizational difference is
associated with an increased likelihood of interstate war.
Although the postCold War focus of Huntingtons ~1993a:2225; 1993b:187;
1996:20! argument is clear, the CoC thesis actually purports to account for the
incidence of interstate conflict over a much longer time span. This is implicit
in Huntingtons ~1996:39, 255! assertion that during the Cold War period the
CoC was suppressed by the superpower rivalry. This suggests that during the
Cold War the impact of the superpower stand-off effectively mitigated the CoC
because alignments were guided largely by ideology ~p. 34!a view supported
by empirical analyses of alliance aggregation during the Cold War era ~see
Siverson and Emmons, 1991; Simon and Gartzke, 1996; Thompson and Tucker,
1997a, 1997b!. One may draw two conclusions from the putative relationship
between cultural factors and war, which Huntington suggests was operative during the Cold War era. On the one hand, one might maintain that the greater
salience of ideological factors during the Cold War era should have vitiated any
significant association between cultural factors and interstate war. Therefore,
the relationship between civilization membership and war during the Cold War
era should have been insignificant. On the other hand, one might contend that
the absence of intercivilizational conflict, coupled with intracivilizational cooperation ~which is assumed to be the general state of affairs, see below!, should have
resulted in a negative relationship between civilization membership and interstate
war in the Cold War era. Since either of these outcomes is consistent with Huntingtons thesis we propose the following:
322
Clear and Present Strangers
Proposition 1.2:
During the Cold War era, intercivilizational difference is negatively or insignificantly associated with the likelihood of interstate war.
In addition, the alleged suppression of the CoC during the Cold War assumes
that it was operative prior to that era. This implies a much longer temporal
pedigree for the CoC. It follows from Huntingtons premises that the relationship between civilization membership and conflict that is manifest in the post
Cold War era is, to some degree, an extension of the relationship between joint
civilization membership and conflict throughout history. However, one does not
have to rely solely on the implications of Huntingtons statements to realize that
his CoC thesis is assumed to be operative over a much longer time period; he is
also quite explicit with regard to his claims of a longue dure for the CoC. This
is evident in Huntingtons ~1993a:25! claim that @o# ver the centuries . . . differences
among civilizations have generated the most prolonged and the most violent
conflicts. He also notes that @c# ommon membership in a civilization reduces
the probability of violence in situations where it might otherwise occur and
where intracivilizational conflicts occur, he insists that these conflicts are likely
to be less intense . . . than conflicts between civilizations ~p. 38!. Huntington
~1996:126! adds that @p# eople rally to those with similar ancestry, religion, language, values, and institutions and distance themselves from those with different
ones. Moreover, after delineating the various issues that have been the sources
of conflict between humans throughout history, Huntington ~1996:208! avers
that @w# hen states from different civilizations are involved, however, cultural
differences sharpen the conflict.
Emblematic of the more protracted nature of the CoC, for Huntington, is his
view of the historic conflict between Christian and Islamic societies. Drawing on
Richardsons ~1960! data, he observes that 50 percent of wars involving pairs of
states of different religions between 1820 and 1929 were wars between Muslims
and Christians ~Huntington, 1996:210!. He insists that the causes of the prolonged conflict between Islamic and Western civilizations lie not in transitory
phenomena such as twelfth-century Christian passion or twentieth-century Muslim fundamentalism, but rather, @t# hey flow from the nature of the two religions and the civilizations based on them ~p. 210!. He forecasts that @s# o long
as Islam remains Islam . . . and the West remains the West . . . this fundamental
conflict between two great civilizations and ways of life will continue to define
their relations in the future even as it has defined them for the past fourteen
centuries ~p. 212!. Therefore, beyond his assessments of the factors influencing
postCold War and Cold War conflicts, Huntingtons CoC thesis suggests that,
ceteris paribus, culturally dissimilar states, throughout history, have been more
likely to fight each other than culturally similar states.5
Proposition 1.3:
During the preCold War era, intercivilizational difference is
positively associated with an increased likelihood of interstate
war.
All told, Huntingtons CoC thesis makes three clear, empirical assumptions:
~1! in the postCold War era, states of different civilizations have been more
likely to fight each other; ~2! during the Cold War era, the superpowers constrained intercivilizational conflict; and ~3! throughout history states of different
civilizations have been more likely to fight each other than have those that share
a common civilization. These empirical claims beg systematic examination, and
5
The only temporal exception to this generalization is the Cold War period ~discussed above!.
Errol A. Henderson and Richard Tucker
323
Huntington, as well as others, has provided some empirical analyses of these
relationships. We turn to a discussion of several of these studies in the next section.
Previous Empirical Tests of Huntingtons Thesis
Huntington ~pp. 257258! conducts some cursory analyses of the relationship
between civilization membership and involvement in ethnopolitical conflicts.
However, his findings are unpersuasive since his analysis is devoid of relevant
controls. It focuses primarily on intrastate rather than interstate conflict, and it
is temporally restricted to the 199394 period. It also appears that Huntingtons
utilization of Richardsons ~1960! findings to support his contentions regarding
the enduring conflict between Islam and Western Christendom ignores the more
complex relationships uncovered by Richardson. First, it is important to remember that Richardsons ~1960! study, which provides the earliest systematic treatment of the impact of cultural similarity on international conflict, focuses on
disparate types of conflicts ~most of which are not interstate wars! and is restricted
to the period from 1820 to 1929. Second, Richardson ~1960:223230! found that,
in the main, common religion did not have a dampening effect on the incidence
of war ~nor did common language!. Although shared Confucianism appeared to
be associated with a decreased likelihood of war, no such relationship obtained
for Islam or Christianity ~pp. 238239!. While there appeared to be a relationship between religious dissimilarity and conflict in the case of Christianity versus
Islam ~p. 245!, Richardson also found that Christianity more than Islam was
subject to internecine conflict ~pp. 235239!. Therefore, Richardsons ~1960!
findings do not provide the unequivocal support for either the CoC thesis, in
general, or Huntingtons conception of Islams Bloody Borders, specifically. In
fact, a casual review of the relationship between religious similarity and interstate war using Correlates of War Project data reveals that of the 30 dyadic
interstate wars between states of different civilizations from 1816 to 1992, only 11
~37%! were between Islamic and Christian states, and of those only 4 ~13% of the
total! were between Islamic and Western states, the others occurring between
Islamic and Orthodox Christian states. These relationships seem to challenge the
view that a disproportionate amount of intercivilizational wars has occurred
between Islam and Western Christendom. Moreover, they encourage a more
informed systematic examination of the CoC thesis.
In addition, Hendersons ~1997, 1998! empirical analysis of the relationship
between cultural similarity and interstate war found that while common religion
is associated with a decreased likelihood of interstate war, both shared ethnicity
and shared language are associated with an increased likelihood of interstate
war. Although the findings with regard to the impact of religious dissimilarity on
interstate war may provide support for one aspect of Huntingtons thesis, nevertheless, the overall relationship between cultural similarity and the likelihood
of war in these studies seems to militate against the sweeping generalizations of
the CoC thesis. More tellingly, Gurrs ~1994! analysis of ethnopolitical conflicts,
which utilizes the same Minority at Risk data that Huntington ~1996:257258!
relies on to support his CoC claims, largely gainsays Huntingtons contentions.
In fact, Gurrs ~1994:347! empirical analysis indicates that the primary culprit in
ethnopolitical conflicts among groups of different civilizations in the postWorld
War II era is the transformation of the political system of the states in which they
are located. Specifically, he demonstrates that the presence of a political transition within countries undergoing state formation, revolution, or democratization
is the most important factor in intercivilizational disputes within states. The end
of the Cold War simply increased the number of such transitions as new states
were added to the system. What is more, Gurr ~p. 347! insists that although
324
Clear and Present Strangers
@c# ommunal conflicts across fault lines between civilizations and religious traditions are more intense than others, nevertheless, they have not increased in
relative frequency or severity since the end of the Cold War.
However, it is important to note that none of these studies that are critical of
Huntingtons thesis has explicitly tested the CoC thesis. For example, Hendersons ~1997, 1998! studies focus on the religious, ethnic, and linguistic characteristics of states rather than on the broad civilizational characteristics emphasized
in the CoC thesis. Also, the temporal domain of those studies ends with observations in 1989, therefore they do not examine the postCold War relationships
that are important to the CoC thesis. In addition, Gurrs ~1994! analysis focuses
primarily on intrastate rather than interstate conflict. Two other studies ~Chiozza, 1999; Russett et al., 2000! provide more straightforward tests of Huntingtons
thesis explicitly; however, they both focus on international conflict short of war.6
To be sure, Huntington ~1996:262! himself acknowledges that the evidence
required to support his claims is beyond that which he provides in his primarily
anecdotal conceptualization of the CoC thesis. He also states that his study is
not intended to be a work of social science ~p. 13!, although this does not
prevent him from making forecasts and developing testable propositions from
his self-described civilization paradigm ~pp. 1314, 35, 37!, which he argues is
superior to universalist as well as realist and other state-centered theoretical
frameworks ~pp. 34 39!. Even a cursory review of Huntingtons claims reveals
that despite his assertions to the contrary, he is clearly promulgating testable
propositions that are appropriate subjects of social-scientific analysis.7 Therefore, our objective in the present study is to provide a rigorous examination of
Huntingtons claims using a research design that explicitly tests his CoC thesis
over several spatial-temporal domains while employing relevant controls that
impact on the presumed relationship between civilization membership and interstate war. In the next section we delineate our research design that we utilize to
evaluate the propositions that devolve from the CoC thesis.
Research Design
Outcome Variable
The outcome variable in this study is the presence0absence of an interstate war
for the state dyad. If the dyad experienced an interstate war, it is coded as 1; if
it did not, it is coded as 0. The outcome variable does not include wars
involving nonstate actors, such as colonies, but only those recognized states
according to COW criteria ~Small and Singer, 1982!.8 The primary predictor
6
Specifically, Chiozza ~1999! focuses on international crises and Russett et al. ~2000! focus on militarized
interstate disputes.
7
Huntington ~1996:37! not only clearly formulates propositionseven predictionsderived from the CoC
thesis, he even conducts empirical tests of those propositions derived from his civilizational paradigm ~see above!.
Further, he states that not only do paradigms inform foreign policy, but @p# aradigms also generate predictions, and
a crucial test of a paradigms validity and usefulness is the extent to which the predictions derived from it turn out
to be more accurate than those from alternative paradigms. He then attempts to show how the predictions that are
derived from his CoC thesis are more accurate than those from a statist perspective drawing on, inter alia, the
relationship between Russia and the Ukraine as a case in point. Therefore, his comments to the contrary notwithstanding, it is clear that even Huntington seems to agree that one can derive testable propositions from the CoC thesis.
8
The COW project defines an interstate war as an armed conflict involving the regular armed forces of two
recognized states in which a participant incurs a minimum of 1,000 battle deaths. For some, this coding rule
arbitrarily excludes what are clear clashes of civilizations between imperialist powers and their subjects; however,
since there is a dearth of data on the relative capabilities, regime-type, etc., of the nonstate actors, then their
inclusion would militate against the type of analysis we wish to conduct. Further, Huntingtons cultural realist
thesis, in its international ~as opposed to domestic! variant, focuses on the behavior of states more than the
interaction of state and nonstate actors. For an analysis of the correlates of extrastate wars, see Henderson, 2000.
Errol A. Henderson and Richard Tucker
325
variable is civilization membership, with additional controls for geographic proximity, regime-type, and power parity.
Predictor Variable
Civilization membership follows the criteria outlined in Huntington ~1996:4548,
Map 1.3!. The civilization categories include Sinic, Japanese, Hindu, Islamic,
Orthodox, Western, Latin American, African, and Buddhist civilizations. States
that do not fall within these categories are classified as Other. Ten codings are
used for civilization type: the nine specific civilizations, and an Other category.
Each state is classified by its civilization type; a dummy variable ~Mixed Civilization! is coded 1 if the states in the dyad represent different civilizations; 0 if
they share a common civilization. The civilization coding for each state can be
found in the Appendix.
It is important to note that there are several inconsistencies in Huntingtons
designation of major civilizations and the states that fall within them. For example,
he is ambivalent as to whether there is an African civilization although he begrudgingly includes it in his list of major civilizations ~p. 47!. At times he suggests that
Buddhism is a civilization ~pp. 109, 137138, 176, 219, 257!, while in other places
he suggests that it is not ~pp. 45, 47!, and in still others he is ambivalent about it
~pp. 4748!. In the end, Huntington appears to include Buddhism as a major
civilization since he includes it in Map 1.3, which depicts the worlds civilizations.
Therefore, we code it as such.9 It also is not clear where some of the Caribbean
states fall in the civilizational spectrum; that is, are they Western or Latin American? For example, while Huntington includes Cuba and the Dominican Republic
in the Latin American civilization, Haiti appears to be neither Latin American nor
Western; therefore, we place it in the Other category. There are also several
countries such as Guyana, Kenya, Nigeria, the Philippines, Surinam, and others
listed on Map 1.3 that are torn between civilizations and we coded these as
Other. In the case of Comoros and Mauritius, which are primarily Islamic
and Hindu states, respectively, we follow Huntingtons classification and code
them as members of the African civilization. In addition, Huntington excludes
Judaism from his list of civilizations. At first blush this might appear to have
major repercussions for our analysis since it will force us to include Israel in
the Other category and risk losing the conflicts between Israel and several of
its Islamic neighbors. In practice, though, this coding should not skew our
results because conflicts between Islamic states and Israel will be captured as
conflicts between Islamic and Other states.10
To be sure, one may take issue with Huntingtons classificatory scheme, and
in several cases we do; however, in order to assess the CoC thesis properly it
is important that we stay as close to Huntingtons typology as is reasonable.
Without doing so we run the risk of failing to test Huntingtons thesis explicitly.
9
In its latest incarnation, Huntingtons ~1999:18! discussion with regard to Buddhism is just as inconsistent as
that found in his earlier work. For example, he states: What can legitimately be described as a Buddhist civilization,
however, does exist in Sri Lanka, Burma, Thailand, Laos, Cambodia; and Tibet, Mongolia, and Bhutan. He then
adds that the virtual extinction of Buddhism in India and its incorporation into existing cultures in other major
countries means that it has not been the basis of a major civilization.
10
0n the other hand, since Huntington does not include Judaism as a civilization, then one might contend that
the Middle East Wars cannot, therefore, represent clashes of civilizations in the strict sense. Nevertheless, the
removal of cases that involve Israelas well as the removal of the category Other as a wholefrom our analyses
does not have a significant impact on our findings. Also, since Huntington seems to view civilization membership
as relatively time invariantalthough he intimates that during the nineteenth century the U.S. was not as firmly
situated in the West as it is nowwe assume that states belong to the same civilization during the entire 1816 1992
period.
326
Clear and Present Strangers
Therefore, although our codings of states in their respective civilizations do not
necessarily conform to our definitions, they closely conform to Huntingtons.11
Control Variables
Our study is not designed to evaluate the full range of relevant counterarguments to Huntingtons thesis, some of which focus explicitly on the CoC thesis
~e.g., Ajami, 1993; Bartley, 1993; Mahbubani, 1993; Rubenstein and Crocker,
1994; Mottahedeh, 1995; Walt, 1997; Midlarsky, 1998; Rosecrance, 1998; Senghaus, 1998!. Others address the larger relationship between cultural similarity
and conflict ~e.g., Gleditsch and Singer, 1975; Henderson, 1998!. Instead, we
focus on the empirically basedrather than the more policy oriented
counterarguments to the CoC thesis. Therefore, the control variables we utilize
in this study are derived from theoretical counterarguments to the CoC thesis
which insist that in the presence of relevant non-civilization-based factors, the
impact of civilization membership on international conflict will be vitiated. By
refuting the empirical basis of Huntingtons thesis, we would also undermine the
policy recommendations that derive from it. On the other hand, if the impact of
civilization membership is significant in light of these non-civilization-based
control variables, we would have gone far in corroborating the CoC thesis.
Further, by substantiating Huntingtons empirical claims one would buttress his
policy prescriptions. The three non-civilization-based factors that we include as
controls are geographical proximity, regime-type, and relative capability.
Geographic Proximity. We assume that the impact of civilizational difference on
war will be manifest only among proximate civilizations. It is not surprising that
states in noncontiguous civilizations ~e.g., Latin American and Sinic states! are
less likely to fight since they rarely have an opportunity to interact. The major
alternative empirical thesis on the relationship between cultural similarity and
the likelihood of interstate war centers on the role of geographic proximity as a
mitigating factor in intercultural conflicts. Richardson ~1960! concluded that
territory was an important mitigating factor in the deadly quarrels of culturally
dissimilar disputants. This view is supported by Holsti ~1991!, who concluded that
concerns over territory have generated more wars than any other issue for the
period 16481989. Diehl ~1985! and Vasquez ~1993! showed that 59 of the 67
~88%! interstate wars between 1816 and 1980 involved neighbors; and Bremer
~1992! has found that contiguity has a greater impact on the likelihood of war
within state dyads than alliance membership, regime-type, levels of development,
extent of militarization, power status, and differentials in relative capabilities.
Singer ~1981! reminds us that neighbors are more likely to be both culturally
similar and prone to conflict. Gleditsch and Singers ~1975! study of the impact
of distance on interstate war concludes that proximity more than culture is a
predictor to war onset. They reason that since cultural cousins are more often
neighbors, and neighbors are more likely to fight, distance rather than culture is
the significant variable in interstate conflict. But, even controlling for contiguity,
Henderson ~1997, 1998! has found that religious, ethnic, and linguistic factors
are significantly associated with the likelihood of war onset. However, since
11
One may quarrel with Huntingtons classificatory schema on a number of levels and with respect to a number
of cases. Beyond the issue of whether it is accurate to conceptualize civilizations at all, in the way that Huntington
does, clearly several cases are contentious. For example, although there is a majority Christian population in the Philippines, Huntington classifies this as a hybrid state with Sinic, Christian, and Muslim characteristics. Kazakhstan is heavily Orthodox and Muslim, but Huntington codes it as an Orthodox state. Israel is coded as a Western civilization by
Russett et al. ~2000! but this classification is inconsistent with Huntingtons ~1996! thesis, which, although at times is
ambivalent as to whether Judaism represents a major civilization ~contrast p. 48 with p. 317!, is quite clear that Israel
is not a Western state ~see pp. 71, 90, 186!.
Errol A. Henderson and Richard Tucker
327
previous studies have not specifically tested the CoC thesis, the alleged mitigating effect posed by distance on relationships between civilization membership
and interstate conflict remains plausible. In our analysis we utilize a distance
variable that is constructed using a capital-to-capital measure to capture the
proximity ~in kilometers! of states in the dyad; however, if two states are contiguous by land, the distance is zero regardless of the distance between their respective capitals ~see Bennett and Stam, 1998!. Distance between national capitals is
calculated using the Great Circle distance formula as delineated in Fitzpatrick
and Modlin, 1986:xi.12
Regime-Type. In addition, the conflict-dampening impact of joint democracy
appears to be evident across cultural spheres ~Russett, 1993; Ray, 1995; Maoz,
1997!. It follows that regardless of civilization membership, democracies of diverse
civilizations will be less likely to fight one another, a claim that is at odds with the
CoC thesis. The findings of the relationship between cultural factors and the
democratic peace seem to support this viewpoint ~Henderson, 1998!. In order to
examine the mitigating impact of joint democracy on intercivilizational conflict,
we include a joint democracy variable ~ Joint Democracy!, which is computed
from the values of the individual states democracy and autocracy scores using
the Polity III data set ~ Jaggers and Gurr, 1996!. It is coded 1 if both states in a
dyad score greater than 6 on individual state scores; otherwise, it is coded 0.
Our regime-type indicator is assigned a one-year time lag.
Capabilities. Lastly, realists contend that relative power capabilities are more
salient than cultural ~or civilizational! factors in a states decision to go to war.
Findings on the relationship between the relative power capabilities of states
support the contention that power parity between states is associated with an
increased likelihood of war, while asymmetries of power between states are associated with an increased likelihood of peace ~Weede, 1976; Bremer, 1992!. We
construct our power parity variable in two steps. First, we draw our asymmetry
scores for each dyad using Beck and Tuckers ~1998! measure of asymmetry,
where a total score of 1 on the asymmetry scale indicates a completely asymmetrical dyad and a value of 0 represents a perfectly symmetrical dyad. Second,
we subtract the asymmetry score from 1 and the resultant score is our power
parity score. Since our power parity variable takes the value of 1
the asymmetry score, then the greater the value of this variable ~i.e., the more equal the
two states in terms of their relative capabilities!, the greater the likelihood of war
~Organski and Kugler, 1980; Kugler and Lemke, 1996!. The power parity variable
~Power Parity! is also assigned a one-year time lag.
Methodology
The spatial-temporal domain of the analysis encompasses all state dyads observed
annually for the period 1816 1992 ~Tucker, 1998!. Since our outcome variable
the onset of international conflictis a binary choice ~0 or 1!, estimations are
made using logit regression ~Aldrich and Nelson, 1984!. The baseline model takes
the following form: Pr ~Wari,t ! 1 0~ 1 e Zi !; where Pr ~Wari,t ! is the probability
that the outcome variable ~the onset of interstate war! equals 1; and Z i is the sum
12
Distance between national capitals is calculated as: Cos ~ D ! ~Sin ~ L1 !! Sin ~ L2 !! ~Cos ~ L2 ! Cos ~ DiffLo !!;
where L1 is the latitude of the capital of state A; L2 is the latitude of the capital of state B; DiffLo is the difference
in longitude between the respective capitals; D is the arc distance ~in degrees! between the respective capitals.
Within the formula, northern latitudes are specified as positive values and southern latitudes are negative; eastern
longitudes are specified as positive values and western longitudes are negative. Given the arc distance D, distance
in miles is computed by multiplying D by the average number of miles per degree, 69.16 ~see Bennett and Stam,
1998!.
328
Clear and Present Strangers
of the product of the coefficient values ~ bk ! across all observations of the predictor variables ~ Xi,t !, that is: a b1 Mixed Civilization b2 Distance b3 Joint
Democracy b4 Power Parity. We report both bivariate and multivariate findings
on the relationship between civilization membership and interstate war ~with the
multivariate regression including our control variables! for each time period.
Beck and Tucker ~1997! and Beck, Katz, and Tucker ~1998! stress the importance
of accounting for spatial and temporal dependence in binary time-series crosssection analysis and suggest a variety of ways to detect, and correct, for such
dependencies. The most straightforward method they propose is the addition of
a set of dummy variables, constructed from an indicator of the length of time
since the last outcome ~in this case, the length of time since the previous war
onset! to the original specification. Another correction employs the use of natural cubic splines in the logit estimation. The following analyses correct for
specifications that exhibit temporal dependence by including the time counter
variable ~Peace Years! and its accompanying spline variables.13 In addition, we
calculate robust standard errors, as discussed in Beck and Tucker, 1997, to
control for potential problems of spatial dependence. Having discussed the basic
research design, in the next section we evaluate the propositions outlined above.
Findings
PostCold War
One of Huntingtons primary claims is that the CoC is operative in the postCold
War era, and therefore civilization difference should be associated with an increased
likelihood of interstate war in this period. An examination of the results found
in Column I, Table 1, however, suggests that states of different civilizations are
actually less likely than states of the same civilization to fight one another. This
finding is inconsistent with Proposition 1.1 and appears to turn Huntingtons
thesis on its head. This finding suggests that we are not witnessing the emergence of a new era of clashing civilizations in world politics; however, simple
bivariate regressions cannot provide definitive support for or against the relationship between a particular factor and the onset of international violence. A
multivariate analysis will allow us to determine if the relationship between civilization membership and conflict holds when controlling for potentially mitigating factors ~specifically, those discussed above!.
When we control for geographic distance, joint democracy, and power parity,
there is no statistically significant relationship between civilization membership
and the likelihood of interstate war as revealed in the multivariate findings in
Column II of Table 1. These findings reveal that states belonging to different
civilizations are neither more nor less likely to fight one another during the
postCold War era. On the other hand, both joint democracy and geographic
distance clearly reduce the likelihood of interstate conflict ~p , 0.001! and these
two findings corroborate two of the most consistent findings in international
relations research. We found no statistically significant relationship between power
parity and war onset during this period. Altogether the results in Table 1 indicate
that when we control for other important factors, the effect of civilization membership during the postCold War era is not statistically significant.
A counterargument to these conclusions is that our data analysis only examines the relationships between 1989 and 1992 and therefore fails to take into
13
Beck, Katz, and Tucker ~1998! suggest that the choice of either approach ~temporal dummies or splines! has
almost no consequences for the estimation of the parameter estimates. Tucker ~1999! contains a set of Stata
functions used here to produce the Peace Years variable, and the accompanying temporal dummies and splines
described in Beck, Katz, and Tucker ~1998!.
Errol A. Henderson and Richard Tucker
329
Table 1. Logit Analysis of the Clash of Civilizations, 1989 1992
Column I
Column II
Mixed Civilization
2.11***
~.53!
.96
~.67!
Distance
.34 a
~.19!
Joint Democracy
.33***
~.13!
Power Parity
1.01
~.98!
Constant
6.91***
~.33!
4.69***
~.83!
131.23
131.23***
58,274
91.15
33.94***
36,168
2LL
x2
N
Huber0White RSEs are in parentheses; *p , .05 level, **p , .01 level, ***p , .001 level.
a
significant below .08 level; all p-values are estimated using two-tailed tests.
account those that have emerged since then, which may provide support for
Huntingtons thesis. Of course, data limitations render this an empirical issue
that cannot be put to rest directly at this juncture because our findings are
temporally bounded by observations ending in 1992. However, 1992 is also the
last full year from which Huntington could draw his observations on which he
could base his initial CoC contentions, which were first published in the summer
of 1993. Therefore, the time frame for both analyses is generally the same, and
if the CoC thesis is accurate, it should certainly not be falsified over the range of
cases that its progenitor had to draw on in order to promulgate it in the first
place. From where can one derive evidence to support claims that clashes of
civilizations lead to wars if not in the historical record? Clearly, Huntington
argues that the historical record reveals that differences among civilizations
have generated the most prolonged and the most violent conflicts and for him
this historical current is especially ~though not exclusively! evident since the end
of the Cold War. In order to substantiate such a claim, he assesses the salience of
these civilizational factors in world politics. Therefore, if the findings from the
postCold War era are inconsistent with the CoC thesisconsidering that the
period encompasses the full range of empirical observations the author could
draw on in promulgating his thesis in the first placethen by refuting his claims
for the time period 1989 1992 one has seriously challenged a very important
aspect of the CoC thesis.
Cold War
Huntington also argues that the superpower stand-off kept a lid on intercivilizational clashes in the Cold War era. According to this logic, the impact of the
CoC should have been either negative or insignificant. The results in Columns I
and II in Table 2 reveal that this aspect of Huntingtons thesis has some support
since the relationship between mixed civilizations and the probability of war is
statistically insignificant ~although the coefficient is positive!. However, it is not
clear whether Huntington is correct in his view that the superpowers effectively
suppressed intercivilizational conflict since there was hardly an absence of such
conflict during the Cold War era. Rather, it appears that our results simply
reflect the coincidence of both intracivilizational and intercivilizational conflicts
330
Clear and Present Strangers
Table 2. Logit Analysis of the Clash of Civilizations, 1946 1988
Column I
Column II
Mixed Civilization
.33
~.43!
.55
~.41!
Distance
.30***
~.07!
Joint Democracy
3.12***
~1.03!
Power Parity
1.17**
~.46!
Peace Years
1.60***
~.13!
Spline 1 b
.21***
~.02!
Spline 2 b
.15***
~.02!
Spline 3 b
.40***
~.09!
Constant
8.81***
~.39!
2.70***
~.45!
2LL
x2
N
638.02
.58
338,976
1,053.52
857.96***
267,315
Huber0White RSEs are in parentheses; *p , .05 level, **p , .01 level, ***p , .001 level.
a
significant below .08 level; all p-values are estimated using two-tailed tests.
b
Coefficients of Peace Years cubic spline segments
throughout the Cold War era, which served to wash out any significant relationship between civilization membership and war. Conflicts during this period such
as wars in the Middle East, South Asia, Korea, and Vietnam, which involved states
of different civilizations, were contemporaneous with conflicts such as those
between Ethiopia and Somalia, Uganda and Tanzania, China and Vietnam ~twice!,
and Iran and Iraq, which pitted states of similar civilizations against each other.
PreCold War
In addition, Huntingtons CoC thesis implies that intercivilizational difference,
ceteris paribus, is associated with an increased likelihood of interstate war during
the preCold War era. We tested this aspect of his thesis by analyzing data for
the entire 1816 1945 period, from the Congress of Vienna to the end of World
War II. An examination of Column I in Table 3 indicates that during this period,
states of different civilizations were actually less likely to engage in conflict with
each other than were those from similar civilizations, refuting Proposition 1.3 of
the CoC thesis. Moreover, controlling for geographic distance, joint democracy,
and power parity ~Column II! did not cancel out this relationship. Contrary to
the CoC thesis, it appears that in the preCold War era, common civilization
membership is associated with an increased likelihood of interstate war. Thus, in
the only time period in which the relationship between civilization membership
and war is statistically significant below the conventional .05 level threshold, the
general relationship between civilization membership and interstate war is the
opposite of that suggested by Huntington ~1993a, 1993b, 1996!. On the whole, our
Errol A. Henderson and Richard Tucker
331
Table 3. Logit Analysis of the Clash of Civilizations, 1816 1945
Column I
Column II
Mixed Civilization
.60***
~.15!
.59**
~.23!
Distance
.30
~.40!
Joint Democracy
2.37***
~.52!
Power Parity
.17
~.24!
Peace Years
1.50***
~.13!
Spline 1 b
.81***
~.07!
Spline 2 b
.40***
~.03!
Spline 3 b
.50***
~.04!
Constant
5.80***
~.10!
.43***
~.17!
1755.06
15.82***
118,648
2089.57
1,835.20***
97,901
2LL
x2
N
Huber0White RSEs are in parentheses; *p , .05 level, **p , .01 level, ***p , .001 level.
a
significant below .08 level; all p-values are estimated using two-tailed tests.
b
Coefficients of Peace Years cubic spline segments
f indings indicate that Huntingtons CoC thesis is incorrect for two of the three
time periods to which it is applied. Most importantly, it is not supported for the
postCold War era, which is the time period for which it is assumed to be
particularly relevant.
Discussion
In largely refuting Huntingtons ~1993a, 1996! empirical claims, our findings
undermine the salience of the policy prescriptions that he proffers in light of his
CoC thesis. Since our results indicate that civilizational differences are not associated with the fault-line wars of the postCold War era that Huntington anticipates ~1993a:39! and prefigures ~1996:313316!, then his foreign policy prescription
of cultural containment appears ill-advised, at best. That is, viewing the world
through civilizational lenses is more likely to blind decisionmakers to both the
divisions within civilizations ~including their own! and the cross-cutting cleavages
among them, which might provide opportunities for cooperation ~Deutsch and
Singer, 1964!. Moreover, a policy of cultural containment is probably likely to
foment interstate conflict by encouraging more aggressive policies toward states
of different civilizations while creating a false sense of security that members of
the same civilization will have convergent interests.14 Walt ~1997:189! concurs
14
One of the earliest critics of the CoC thesis, Mahbubani ~1993:13! emphasized this point with regard to
Huntingtons speculations with regard to a Confucian-Islamic connection, of which he noted that @t# he real
tragedy of suggesting a Confucian-Islamic connection is that it obscures the fundamentally different nature of the
challenge posed by these forces.
332
Clear and Present Strangers
that the civilizational paradigm that Huntington has offered is not a sound basis
for making foreign policy and that @r# elying upon an overly broad category like
civilization would blind us to the differences within broad cultural groups and
might unwittingly rob policymakers of the flexibility that has always been a
cardinal diplomatic virtue. Therefore, if we treat all states who are part of some
other civilization as intrinsically hostile, we are likely to create enemies that
might otherwise be neutral or friendly.
Much of the problem lies in Huntingtons ~1996! extreme cultural relativism
where civilizations are viewed as largely incommensurable and bound by fundamental traditional values that differentiate them. Huntington ~pp. 69 72! is
struck by the alleged mutual receptivity of democracy and Western Christendom
and the apparent mutual exclusivity of democracy and Islam that he calls the
inhospitable nature of Islamic culture and society to Western liberal concepts
~p. 114!. A strong critique of his views about Islams presumed antidemocratic
bias and inherent propensity to violence is provided by recent systematic evidence that challenges the mutual exclusivity thesis with regard to Islam and
democracy ~Midlarsky, 1998!. Specifically, Midlarskys ~p. 504! study of the relationship between Islam and democracy across 97 states over the period 1973
1987 finds that although Islamic societies appear to be less inclined toward
liberal democracy, @d# emocracy itself and Islam are not mutually exclusive
where democracy is measured using a political rights index such as Gastil, 1988.
Wedding these results to democratic peace findings, which demonstrate that
democracies rarely fight each other ~Russett, 1993; Ray, 1995!and the findings
in this study, which indicate that civilization membership is not a significant
predictor to war involvement but joint democracy isit follows that, were Islamic
states to develop democratic regimes, then the likelihood of warfare between
Islamic states and Western democracies would diminish substantially ~or disappear altogether if the thesis that democracies never fight each other is accurate;
see Rummel, 1979!. Aware of the implications of his findings for clashes between
Islamic and Western states, Midlarsky ~1998:505! asserts that civilizational conflict
between Islamic and Western states is not likely in the foreseeable future, if only
because there are certain compatibilities between democracy and Islam that
deny the mutual exclusivity hypothesis.
Focusing on more rigorous treatments of the relationship between cultural
factors and conflict0cooperation reveals just how much Huntingtons CoC thesis
relies on anecdote and selective ~in! attention to the historical record. For example, throughout his text, Huntington appears to have little problem taking the
mutterings of government leaders at face value where they provide support for
his thesis; however, Walt ~1997:184 185! reminds us that political rhetoric serves
many functions while @t# he real issue is what these leaders ~or their countries!
will actually do, and how much blood and treasure they will devote to civilizational interests. As a case in point, Walt ~1997:185! challenges Huntingtons
~1996:246 252! depiction of the Gulf War as a harbinger of the fault-line wars
that the latter presumes will prevail in the postCold War era. First, Walt reminds
us that the Gulf War did not begin as a dispute between states of different
civilizations but as an intracivilizational dispute between two Islamic states: Iraq
and Kuwait. In this clash of civilizations Western states allied with one Islamic
state, Kuwait, with tacit support from Israel, and repulsed another Islamic state,
Iraq. Undeterred by these peculiarities, Huntington ~1996! attempts to salvage
his thesis by focusing on the sentiments of select individuals and groups from
among the Islamic populations, which he argues favored Iraq. Second, Walt
~1997:185! counters that even if this were true, it merely underscores the fact
that state interests mattered more than loosely felt and politically impotent
loyalties to a particular civilizational entity.
Errol A. Henderson and Richard Tucker
333
Turning to another postCold War clash of civilizations, Walt ~p. 185! points
out that although the war in the former Yugoslavia manifests some of the trappings of a fault-line war, it ultimately fails as an intercivilizational conflict because
the more than 50,000 U.S.-led troops that were deployed to Bosnia in 1996 were
not there to defend Western interests ~represented by Croatia! but Muslims ~as is
the case in the recent NATO intervention in Kosovo, which is aimed at protecting ethnic Albanians who are primarily Muslim!. Further, he contends that while
modest sums of aid for Bosnian Muslims were sent by Islamic countries, the lions
share of support was provided by Western states. Similarly, Russia offered some
rhetorical support to the Serbs, but it backed away from its Orthodox brethren
when Serbian bellicosity made Belgrade an unappealing ally. Even the Western
states failed to line up according to cultural criteria, with Britain and France
being more sympathetic to the Serbs, Germany backing the Croats, and the
United States reserving most of its support for the Muslims. 15
The major critique of the CoC thesis is that it ignores the persistent role of
nationalism in world politics. Clearly, although nationalist conflict may resemble
intercivilizational conflict at times ~e.g., in independence struggles or colonial
wars!, nationalist interests often work at odds with civilizational ones, leading to
intracivilizational wars ~e.g., the wars of Italian and German Unification!. For
example, Walt ~1997:183184! states that @b# eing part of some larger civilization did not convince the Abkhaz, Armenians, Azeris, Chechens, Croats, Eritreans, Georgians, Kurds, Ossetians, Quebecois, Serbs, or Slovaks to abandon the
quest for their own state. Contrary to Huntingtons thesis, the nation, and not
the civilization, appears to be the largest identity group to which people consistently swear fealty in the postCold War era.
Huntington is correct that states continue to be the major actors in world
politics ~and for the foreseeable future!; however, he seems to have ignored the
reason their influence persists: the enduring salience of nationalism in international relations. This neglect of nationalism, for Walt ~p. 187!, is the Achilles
Heal of the civilizational paradigm ~see also Ajami, 1993!. Since civilizations are
more or less ideational constructs rather than political agents, they are devoid of
decision-making power or control over political or economic resources. By comparison, states can mobilize their citizens, collect taxes, issue threats, reward
friends, and wage war; in other words, states can act; moreover, @n# ationalism
is a tremendously powerful force precisely because it marries individual cultural
affinities to an agencythe statethat can actually do something ~Walt, 1997:187!.
To be sure, nationalist struggles may engender intercivilizational conflicts such
as those that Huntington foresees; however, cultural differences, in the future as
in the past, are likely to remain one of several factors including political, military, economic, and demographic ones that give rise to international conflict.
For example, Senghaus ~1998:130! insists that socioeconomic factors are the
primary culprits in the onset of ostensibly intercivilizational conflicts. He also
maintains that cultural variables are, at best, secondary factors in the onset of
wars and probably play a greater role in the ~de! escalation of conflict once it has
begun ~p. 131!. Moreover, scholars have also suggested that the CoC thesis
neglects important conflict-dampening aspects of intercivilizational contact such
as trade ~Rosecrance, 1998!. Earlier Bartley ~1993! argued that the CoC thesis
failed to take into account sufficiently the role of economic development, the
spread of liberal democracy, and the expansion of communications networks as
f