Skip to main content
Has "the Clash of Civilizations" Found Empirical Support?
Oct 1, 2006

Samuel Huntington first published his article “The Clash of Civilizations?” in 1993. Later, in 1996 he published “The Clash of Civilizations and the Remaking of World Order,” giving a more detailed and broader explanation of his theory and providing more cases to support his argument. Debates among not only political scientists, but also scholars of different social studies followed both studies.

In his studies, (1993, 1996) Huntington suggests alternative explanations to understanding conflicts and disputes in the post-Cold War period. To him, “culture and cultural identities, which at the broadest level are civilizational identities, are shaping patterns of cohesion, disintegration, and conflict in the post-Cold War” (Huntington 1996, 20).

The purpose of this paper is to test Huntington’s argument (1996) by addressing the following question: “Does a civilizational difference cause militarized interstate dispute?” The significance of the answer to this question is obvious: if Huntington’s claim is empirically supported, then dominant theories of international politics, which do not base conflicts on civilizational differences, have to be revised. And, therefore, Huntington’s theory will be useful in understanding the emerging world-order and explaining militarized disputes and conflicts that have occurred in the post-Cold War era.

Literature review

There were many responses and objections to Huntington’s argument that resulted in an extensive debate. Many of them claimed that the civilization theory is unable to explain the causes of conflicts and international order. In spite of the great number of responses and studies, few of these have been supported by quantitative research.

One of the important studies that evaluated Huntington’s argument and which deserves close attention was conducted by Russett, Oneal, and Cox in 2000 (in Triangulating Peace 2001, 239-305). The main question to be answered was “Does a difference in civilization increase the likelihood that a pair of states will become involved in a militarized interstate dispute?” (Russett et al 2001, 248). To perform their study, they used the University of Michigan’s Correlates of War (COW) database that provided Militarized Interstate Disputes (MID). In their view, limiting their analysis to MID that occurred during the 1950-92 period was appropriate, because “Huntington addressed events in these years when he first presented his argument in 1993” (249). Variables of liberal and realist theories were also added to their analysis in order to observe which theory had had higher explanatory and predictive power. Results of logistic regression indicated that variables of liberal and realist theories had much more influence on the probability of conflict than on civilization theory. Belonging to different civilizations increases the risk of dispute between two states only by 12 percent with .30 level of statistical significance. (255).

Huntington (2000, 609) replied to Russett et al, suggesting that they retest their argument as it was dealing with the post-Cold War period. Russett et al (2000, 611), in turn, argued that Huntington (1993) had used many disputes and conflicts that had not occurred in the post-Cold War era to justify and support his argument. That was the reason why they used MID of the 1950-92 period.

Another study to test Huntington’s argument was conducted by Jonathan Fox in 2001. He tried to examine whether conflicts between civilizations, particularly Western and Islamic conflicts, had increased with the end of the Cold War (Fox 2001, 459). He used the Minorities at Risk (MAR) dataset for his quantitative research and came up with interesting results. From a global perspective, the ratio of non-civilizational conflicts to civilizational conflicts had not changed significantly after the end of the Cold War. From an Islamic civilization perspective, there had been little change in the amount of conflicts among Islamic and other civilizations. However, from the perspective of Western civilizations, the proportion of conflicts between Western and Islamic civilizations had significantly increased since the end of the Cold War. Fox concludes that the findings illustrate that it is important which perspective one looks at conflicts from. If one looks from the Western perspective, they will find an increased amount of conflicts between Western and Islamic civilizations, which supports Huntington’s argument. But from the global and Islamic civilization perspective, little evidence can be found for a “clash of civilizations” in the post-Cold War period (459).

Although Fox’s study deserves attention, it does not examine conflicts among the other seven civilizations, but only focuses on the Western versus Islamic civilizations. Furthermore, his units of analysis are conflicts within states, not interstate. Interstate disputes are important in analyzing Huntington’s argument, because a nation as a whole has specific civilizational characteristics. According to Huntington’s map (1996, 26-27) of civilizations, each nation belongs to one of the nine civilizations and conflicts mostly arise among these. Conflicts that occur within states are only one minor part of Huntington’s argument.

Many other scholars have qualitatively criticized Huntington’s clash of civilizations theory. Ali Mazrui (1997) finds three major problems in Huntington’s argument. First, there is a “factual fallacy.” “It may not be factually true that the main lines of conflict of the future, following the Cold War, will be lines of clash of civilizations. It could be states or economic blocs” (Mazrui in Rashid 97, 27). Second, there could also be a “conceptual fallacy” in Huntington’s argument. Mazrui suggests a “clash of races” as an alternative to the clash of civilizations (27). Thirdly, there is a “temporal fallacy,” which means that inter-civilizational conflicts are not phenomena of the emerging world, but rather that they have been the most important issue of the world for at least the past four to five hundred years; within these conflicts we can include the Crusades, the “trans-Atlantic slave-trade, and European colonization of much of the world” (29).

In response to his critics (Foreign Affairs 1996, 67), Huntington stated that none of these scholars had come up with a better alternative to explain the conflicts of the new era. To him, the clash of civilizations remains the most powerful theory to understand disputes in the post-Cold War period.

In this context, this paper is of a crucial importance because 1) it will focus on militarized disputes of the post-Cold War period; 2) the disputes are among the states and civilizations they belong to; 3) and the study will conduct a quantitative method to analyze the clash of civilizations. Thus, this paper takes into account the weaknesses of previous studies and establishes a strong critique to Huntington’s argument.

Research design

This analysis uses the Correlates of War (COW) project’s Militarized Interstate Disputes (MID). The COW provides MID starting from 1816 to 2001. As Huntington (2000, 609) claims, the clash of civilizations theory is helpful in understanding conflicts that occurred after the end of the Cold War, specifically after 1989 (Huntington 1996, 21; Huntington 2000, 609). The COW project had MID reaching up to 2001, so I chose those MID which occurred from 1989 to 2001. Although the number of disputes in this period seems to be small to test any theory, it should be quite enough to evaluate Huntington’s argument, as post-Cold War disputes were the basis of his theory. I had 620 militarized interstate disputes in total. Each dispute occurred between at least two states; one state in side A, and another on side B. The sides are coded as 0 and 1. I had information on which pair of states had disputes. Some disputes had more than two sides. I rearranged these coalitions as if they had disputes separately as dyads. For instance, if there were two states on side A and two states on side B in a dispute, than there would be four dyads involved in the same dispute.

States in these disputes belong to one of the nine civilizations, which are Western, coded as 1, Latin American (2), African (3), Islamic (4), Sinic (5), Hindu (6), Orthodox (7), Buddhist (8), and Japanese (9). Huntington (1993, 1996) does not explicitly specify the number of civilizations. In his article (1993) he mentions about “seven or eight major civilizations.” To him, Buddhism “has not been the basis of a major civilization.” (Huntington 1996, 48). Also Japan has not founded any civilization as there are no nations that share the same religion, language, or cultural values with Japanese culture. However, map 1.3 (The World of Civilizations: Post-1990) in his book (1996, 26-27) shows that there are nine civilizations in the world, including Buddhist and Japanese. I used nine civilizations as shown in the map. Russett et al (2001, 251) used eight, excluding Japanese. I wonder how they dealt with disputes that had Japan on one of the sides.

Having nine coded civilizations and many more states in the dataset, I coded all states as any of the nine civilizations they belonged to, according to map 1.3. (Huntington 1996, 26-27). I had difficulties, as did Russett et al (2001, 251), in classifying some states, because the map was not clear and Huntington did not explicitly mention them in the text. For example Nigeria, according to the map is separated into two civilizations, Islamic and African. Since religion is the most important factor that distinguishes each civilization, I have used the CIA’s World Factbook to find out the proportion of religion in Nigeria, and classified it as Islamic. India is shown as a state of both Islamic and Hindu civilizations, but I have classified it as Hindu. Sri Lanka was classified as Buddhist in relevance with Russett et al (2001). The Philippines, in my classification, go into Sinic civilization, but Russett et al (2001) classify it as Western. The most interesting problem that occurred in classification was to which civilization Israel belongs. The question is significant, as Israel has had many disputes with its neighboring states. Huntington thinks Israel is an Islamic civilization according to the map. It is ironic that Israel is considered as an Islamic civilization, first because it has nothing to do with the religion of Islam, and second because such an approach decreases empirical support for Huntington’s argument. Although it does not make sense, I classified Israel as Islamic, strictly adhering to the map.

To answer my research question I had to use civilization dyads and observe if these dyads are prone to clash or not and how frequently they have disputes. Thus, I arranged these nine civilizations in dyadic form and came up with 45 civilization dyads, extracting each one of the same dyads as 1,2 v. 2,1. A dyad is labeled 0 if the civilizations in it are the same and 1 if the civilizations are different. After arranging civilization dyads, I plugged each dispute into one of the dyads and totaled it in regard to the 45 civilization dyads. Then I arranged the number of disputes in each dyad according to the proportion to the total number of disputes.

Huntington’s main hypothesis is that civilizations have mattered and have had positive impact on disputes in the post-Cold War period. I have my doubts about his argument and hypothesize that civilizations do not matter in interstate.

Data analysis

After running my Probit model, I have found that difference in civilizations had negative impact on interstate disputes with relatively small statistical significance (10). The findings of this paper contradict what Huntington (1993, 1996) suggests as his explanation for the causes of interstate disputes in the post-Cold War era. In other words, civilization differences do not cause interstate conflicts.

There have been a total of 620 militarized interstate disputes during the 1989-2001 period, according to the COW dataset (see Table 1). Some of civilization dyads did not have any disputes, whereas other dyads conflicted several times. Huntington’s argument that different civilizations clash does not find support with the results of this study. We can see from Table 1 that 19 dyads that consist of different civilizations have not had any militarized dispute since the end the Cold War. How can Huntington explain this phenomenon on the basis of his clash of civilizations argument? Can interstate disputes be explained by civilizational differences? The answers to these questions are obviously negative. Different variables of realist and liberal theories may explain the findings, but that is not within the scope of this study.


One of the important limitations is the absence of a clear list of civilizations and states in these civilizations. If Huntington had provided these data then there would be no dissimilarities among the scholars who tried to test his argument. Some studies used eight civilizations while others used nine. Some states, like Israel and the Philippines were placed in different civilizations in different studies. This may influence the results of these studies, as these states have had many militarized disputes.


This study shows that Huntington’s argument is insufficient to explain conflicts in the new world order. As he argues, “culture and cultural identities, which at the broadest level are civilizational identities, are shaping patterns of cohesion, disintegration, and conflict in the post-Cold War” (Huntington 1996, 20). However, he did not carry out a quantitative analysis of the argument. He provided only case studies and offered qualitative analysis as support. Many studies, including this one, have been conducted in response to his clash of civilizations argument, with the conflicts of the post-Cold War period being analyzed quantitatively. None of these studies has found evidence to support Huntington’s argument. The finding of this study illustrates that a difference in civilization is not sufficient on its own to understand what causes conflicts these days nor does it have a positive impact on militarized disputes.


  • Foreign Affairs. 1996. Samuel P. Huntington’s the Clash of Civilizations? The Debate. New York: W.W. Norton & Company,
  • Fox, Jonathan. 2001. Two Civilizations and Ethnic Conflict: Islam and the West. Journal of Peace Research 38(4): 459-472.
  • Huntington, Samuel P. 1993. The Clash of Civilizations? Foreign Affairs 72(3): 22-49
  • If Not Civilizations, What? Foreign Affairs 72(4): 186-94, 1993.
  • The Clash of Civilizations and the Remaking of World Order. New York: Simon and Schuster. 1996.
  • Try Again: A Reply to Russett, Oneal & Cox. Journal of Peace Research 37(5): 609-610, 2000.
  • Rashid, Salim. 1997. The Clash of Civilizations? Asian Responses. New York: Oxford University Press.
  • Russett, Bruce M. and John R. Oneal. 2000. A Response to Huntington. Journal of Peace Research 37(5): 611-612.
  • Triangulating Peace: Democracy, Inter dependence, and International Organizations. New York: W.W. Norton & Company, 2001.