1 Jervis, , ‘Realism, Game Theory, and Cooperation’, World Politics, 40 (1988), pp. 317–349, p. 318CrossRefGoogle Scholar .

2 Butterfield, Herbert, History and Human Relations (London: Collins, 1951)Google Scholar ; Herz, John, Political Realism and Political Idealism: A Study in Theories and Realities (Chicago: University of Chicago Press, 1951)Google Scholar ; Jervis, Robert, Perception and Misperception in International Politics (Princeton, N.J.: Princeton University Press, 1976)Google Scholar , chap. 3; and idem, ‘Cooperation under the security dilemma,’ World Politics, 30 (1978), pp. 167–214.

3 Glaser, Charles L., ‘The Security Dilemma Revisited’, World Politics, 50 (1997), pp. 171–201, 172CrossRefGoogle Scholar .

4 Keohane, Robert O. and Martin, Lisa L., ‘The Promise of Institutionalist Theory’, International Security, 20 (1995), pp. 39–51Google Scholar ; Cederman, Lars-Erik, ‘Back to Kant: Reinterpreting the Democratic Peace as a Macrohistorical Learning Process’, American Political Science Review, 95 (2001), pp. 15–31Google Scholar , and Wendt, Alexander, ‘Constructing International Politics’, International Security, 20 (1995), pp. 71–81Google Scholar .

5 Horowitz, Donald L., Ethnic Groups in Conflict (Berkeley, CA: University of California Press, 1985)Google Scholar ; Rothschild, Joseph, Ethnopolitics: A Conceptual Framework (New York: Columbia University Press, 1981)Google Scholar .

6 Kaufmann, Chaim D., ‘Possible and Impossible Solutions to Ethnic Civil Wars’, International Security, 20 (1996), pp. 139–175CrossRefGoogle Scholar .

7 I restrict my attention to those works that have employed the security dilemma as a major analytical tool. Chaim Kaufmann employed the security dilemma to advance partition as a preferred solution to ethnic conflict, but his discussion of the security dilemma has been brief and largely recited Posen's elaboration. See, Kaufmann, ‘Possible and Impossible Solutions’, pp. 147–50. William Rose claimed to have developed some hypotheses about the security dilemma and ethnic conflict but he did not really deploy the security dilemma as an analytical tool. Rose, , ‘The Security Dilemma and Ethnic Conflict: some new hypotheses’, Security Studies, 9 (2002), pp. 1–51Google Scholar . Instead, his analytical tool is offense-defense theory because he mistakenly took offense-defense theory as an integral part of the security dilemma theory. I address the problems of offense-defense theory and its relationship with the security dilemma in detail in Tang, Shiping, ‘Offense-Defense Theory: Toward a Definitive Understanding’, Chinese Journal of International Politics, 3 (2010), pp. 213–260Google Scholar . Melander (1999, 2009) also relied on the security dilemma but has essentially focused on ‘the need to preempt’, which I discuss in the section on Paul Roe below.

8 Tang, Shiping, ‘Fear in International Politics: Two Positions’, International Studies Review, 10 (2008), pp. 451–470CrossRefGoogle Scholar . We shall differentiate the security dilemma from security dilemma theory (or model). The security dilemma is a concept for labelling a particular phenomenon. Security dilemma theory is the body of knowledge that seeks to understand the underlying causes, regulation, and implications of the security dilemma.

9 Here, it is critical to differentiate intentions from motives. Motives are states' (immediate) interests or preferences over goals or outcomes. Intentions are states' preferences over strategies/behaviours. Hence, while all states want power, malign states seek power by intentionally threatening others, whereas benign states do not. On the differences between preferences over strategies and preferences over goals, see Powell, Robert, ‘Anarchy in International Relations Theory: The Neorealist-Neoliberal Debate’, International Organization, 48 (1994), pp. 313–344CrossRefGoogle Scholar , esp. 318–21.

10 Due to the enormous complexity of prescribing sound policies to containing ethnic conflict, however, I shall take on this task in another work. For an earlier and solid attempt, see Byman, Daniel, Keeping the Peace: Lasting Solutions to Ethnic Conflicts (Baltimore: John Hopkins University Press, 2002)Google Scholar .

11 Tang, Shiping, ‘The Security Dilemma: A Conceptual Analysis’, Security Studies, 15 (2009), pp. 587–623CrossRefGoogle Scholar .

12 Booth and Wheeler (2008) is a more recent major treatise on the concept. Wheeler, Ken Booth and Nicholas, The Security Dilemma: Fear, Cooperation, and Trust in World Politics (New York: Palgrave Macmillan, 2008)Google Scholar . Unfortunately, their understanding of the concept too suffers from serious misunderstandings. For a detailed discussion, see Tang, ‘The Security Dilemma’.

13 Tang, ‘The Security Dilemma’.

14 In the context of ethnic conflict, anarchy can be de facto anarchy (that is, collapse of central authority). Anarchy should be understood as not the direct cause of the security dilemma, but rather as a necessary and permissive condition for the security dilemma to arise. See also Collins, Alan, The Security Dilemma of Southeast Asia (London: Macmillan 2000), p. 12CrossRefGoogle Scholar .

15 Although uncertainty about others' intentions must logically precede fear and the two are ontologically different, they are so closely intertwined that I list them together in Fig. 1. For a useful discussion, see Tang, ‘Fear in International Politics’.

16 I have followed the tradition of restricting the discussion on security dilemma to dyadic cases. Adding a third party that can be a direct part of a potential conflict to the picture will make the problem more complex.

17 The existence of security dilemma only means that some measures of self-help will be self-defeating. In other words, some measures of self-help are not self-defeating: they do increase a state's security.

18 For instance, deterrence policies can also and, often do, produce unintended and self-defeating consequences. See Jervis, Perception and Misperception, pp. 81, 90.

19 For instance, the early part of the Cold War (circa 1944–1948) was a classic spiral (that is, the relationship between the Soviet Union and the US steadily deteriorated), but this part of the Cold War was not, nor caused by, a security dilemma because the Soviet Union under Stalin was a malign state.

20 For a more detailed discussion on this key question, see Tang, Shiping, A Theory of Security Strategy for Our Time (New York: Palgrave Macmillan, 2010)CrossRefGoogle Scholar , chap. 3 and the references cited there.

21 Collins, The Security Dilemma, p. 24.

22 This, of course, immediately begs the question how to read others' present intentions and anticipate others' future. This problem is beyond the scope of this article, and I shall merely point out that there are essentially two ways for reading another state's present intentions: 1) observing its behaviour toward other states, and 2) reassurance, that is signalling one's benign intentions and then gauging the other state's intentions by reading into its reaction toward one's signals of benign intentions. I develop a theory of reassurance as a theory of cooperation-building via intention-reading in Tang, A Theory of Security Strategy, chap. 5. The question of anticipating others' future intentions is to be addressed separately in another work.

23 For a more detailed discussion of this continuum, see Tang, ‘The Security Dilemma’. This dynamic can be understood as a change in preference over strategies – from defensive to offensive, even though the preference over goals (that is, security or power) remain the same. Here, it is critical to reject the notion of ‘security-driven’ or ‘defensive’ expansions. Once we differentiate preferences over strategies from preferences over goals, it becomes clear that expansions are signatures of malign intentions and the label of ‘security-driven’ or ‘defensive’ expansions should not retain the connotation of having benign intentions that is now conferred by those adjectives. Accepting the notion of ‘security-driven’ or ‘defensive’ expansions is equal to accepting the notion that there are no fundamental difference between malign states and benign states. For a more detailed discussion, see Tang, A Theory of Security Strategy, esp. chap. 1 and 3.

24 Posen, Barry, ‘The Security Dilemma and Ethnic Conflict’, Survival, 35 (1993), pp. 27–47CrossRefGoogle Scholar , esp. pp. 27–35. I address the offense-defense balance (ODB) as a key component of offense-defense theory in Tang, ‘Offense-Defense Theory’. Briefly, I show that ODB is a theoretical hoax that holds little value.

25 Posen, ‘The Security Dilemma and Ethnic Conflict’, p. 37.

26 Kaufman, Stuart J., ‘The Irresistible Force and the Imperceptible Object: The Yogoslva Breakup and Western Policy’, Security Studies, 4 (1994–1995), pp. 281–319Google Scholar ; ‘An “International” Theory of Inter-ethnic War’, Review of International Studies, 22 (1996), pp. 149–71; ‘Spiraling to Ethnic War’, International Security, 21 (1996), pp. 108–38. Kaufman's more recent work essentially retained his earlier understanding of the security dilemma, although he now seems to downgrade its importance. See Kaufman, , Modern Hatred: the Symbolic Politics of Ethnic War (Ithaca, N. Y.: Cornell University Press, 2001), pp. 9–10Google Scholar , p. 12; idem, ‘Symbolic Politics or Rational Choice? Testing Theories of Extreme Ethnic Violence’, International Security, 30 (2006), pp. 45–86.

27 Kaufman, ‘The Irresistible Force’, pp. 282, 285. Emphasis added.

28 Ibid., p. 293.

29 Kaufman, ‘The Irresistible Force’, pp. 293. Snyder, Jack, ‘Perceptions of the Security Dilemma in 1914’, in Psychology and Deterrence, edited by Jervis, Robert, Lebow, Richard Ned, and Stein, Janice (Baltimore: John Hopkins University Press, 1985), p. 155Google Scholar .

30 Kaufman, ‘An “International” Theory’, pp. 150–1. Hostility can mean both an emotion state and behaviour (a hostile act). Kaufman uses hostility to denote an emotional state. ‘People do not engage in ethnic violence unless they are hostile, that is, unless they actively want to harm each other.’ (Kaufman, ‘Spiraling’, p. 111; emphasis added) Apparently, such a state equals to malign intentions.

31 Kaufman, ‘Spiraling’, pp. 109, 111.

32 These verbs were literally littered in Kaufman's two papers. See, for example, Kaufman, ‘An “International” Theory’, pp. 150, 154–7, 161; idem, ‘Spiraling’, pp. 107, 109, 111–2, p. 117. A fitting verb for most these statements should be ‘exacerbate’, ‘aggravate’, or ‘harden’. ‘Provoke’ can mean both ‘cause/incite’ and ‘aggravate/inflame/arouse’, but it is evident that ‘provoke’ is equivalent to ‘create’ or ‘cause’ for Kaufman.

33 Kaufman, ‘An “International” Theory’, pp. 158, 170. Emphasis added.

34 Ibid., p. 150, emphasis added.

35 Kaufman, ‘Spiraling’, p. 112; emphasis in original. See also idem, ‘An “International” Theory’, p. 156; Modern Hatred, p. 12, p. 34.

36 Kaufman, Modern Hatred, pp. 19–22, 63.

37 Kaufman, ‘Spiraling’, pp. 117, 124; emphasis added.

38 Tang, ‘The Security Dilemma’ (2009). Collins too pointed out that the intellectual foundation of Kaufman's application was Jack Snyder and that Kaufman misunderstood Snyder somewhat, without getting into the details. Collins, The Security Dilemma, p. 25, n. 67.

39 Snyder, ‘Perceptions of the Security Dilemma’, pp. 155–6, 160. The imperialist's (security) dilemma depicts a situation between a malign state (the imperialist state) and a benign realism state. As such, Snyder is mistaken to label it as a security dilemma. Deadlock is equivalent to the situation between two malign states. For a more detailed discussion, see Tang (2009). Snyder uses security-seeking states to denote benign states. See the discussion below.

40 Tang, ‘The Security Dilemma’.

41 Kaufman, ‘Spiraling’, pp. 112, 124.

42 Kaufman, ‘An “International” Theory’, p. 152. Evidently, Kaufman here interpreted Snyder's two labels literally: ‘perceptual’ means false (thus less dangerous), while ‘structural’ means real (thus more dangerous).

43 Ibid., p. 162; see also p. 158.

44 Ibid., pp. 154–5.

45 Kaufman, ‘Spiraling’, p. 112.

46 A major source of Roe's errors has been that he relies heavily on several misleading dichotomies for labelling two types of states – benign states and malign states. These misleading dichotomies include status quo vs. revisionist and security-seeking vs. power-seeking, among others. For a more detailed discussion, see Tang, A Theory of Security Strategy, chap. 1. Interestingly, although in 2001 Roe shifted his attention from states' goals to their strategies (he called them ‘security requirements’) – a move that should force him to focus on actors' intentions, he actually ended up in making many more errors because he has dropped lack of malign intentions from the definition of security dilemma altogether then. See the discussion below.

47 Roe, , ‘The Intrastate Security Dilemma: Ethnic Conflict as a “Tragedy”?’, Journal of Peace Research, 36 (1999), pp. 183–202CrossRefGoogle Scholar ; idem, ‘Former Yugoslavia: The Security Dilemma that never was’, European Journal of International Relations, 6 (2000), pp. 373–93, esp. pp. 375–80.

48 Roe, ‘The Intrastate Security Dilemma’, p. 186; ‘Former Yugoslavia’, pp. 378–9. Unfortunately, like many others, Roe does not differentiate the security dilemma from a spiral (model).

49 Roe, ‘The Intrastate Security Dilemma’, p. 200; ‘Former Yugoslavia’, p. 388. Melander, Erik, Anarchy Within: The Security Dilemma between Ethnic Groups in Emerging Anarchy (Department of Peace and Conflict Research, Uppsala University, 1999)Google Scholar , Report no. 52.

50 Roe, ‘Former Yugoslavia’, pp. 377–8, 380.

51 Wheeler, Nicholas and Booth, Ken, ‘The Security Dilemma’, in Baylis, John and Wheeler, Nicholas (eds), Dilemma of World Politics (Oxford: Oxford University Press, 1992), p. 30Google Scholar .

52 Roe, ‘Former Yugoslavia’, p. 375.

53 Kydd, Andrew, ‘Game Theory and the Spiral Model’, World Politics, 49 (1997), pp. 371–400CrossRefGoogle Scholar ; Tang, ‘Offence-Defence Theory’.

54 Roe, ‘Former Yugoslavia’, p. 376; see also Roe, ‘“Actors” Responsibility in Tight, , Regular, or Loose Security dilemmas’, Security Dialogue, 32 (2001), pp. 103–116Google Scholar , at p. 105; and idem, ‘Which Security Dilemma? Mitigating Ethnic Conflict: The Case of Croatia’, Security Studies, 13 (2004), pp. 280–313, at p. 283.

55 For an in-depth discussion on the role of worst-case assumption over intentions in IR theory, see Tang (2008).

56 Roe, ‘Former Yugoslavia’, p. 376; ‘“Actors” Responsibility’, p. 106; ‘Which Security Dilemma’, pp. 283–4. See also, Melander, Anarchy Within, p. 21.

57 Tang, ‘The Security Dilemma’.

58 Roe, ‘“Actors” Responsibility’, pp. 111–2. In his two later works, Roe basically employed the theoretical arguments developed here. Undoubtedly, his stretching of the security dilemma was encouraged by a similar endeavour from Jack Snyder and Robert Jervis. Snyder, Jack and Jervis, Robert, ‘Civil War and the Security Dilemma’, in Walter, Barbara F. and Snyder, Jack (eds), Civil Wars, Insecurity, and Intervention (New York: Columbia University Press, 1999), pp. 15–37Google Scholar , esp. pp. 19–20.

59 Roe, ‘Which Security Dilemma’, p. 288, fn. 34. See also Roe, ‘“Actors” Responsibility’, p. 110; Glaser, ‘The Security Dilemma’, pp. 190–1; Schweller, Randall L., ‘Neorealism's status quo Bias: What Security Dilemma?’, Security Studies, 5 (1996), pp. 91–121CrossRefGoogle Scholar , esp. pp. 117–9.

60 For instance, Hitler wanted to win WWII, but ended up in losing it.

61 Boulding, Kenneth E., ‘National Images and International Systems’, Journal of Conflict Resolution, 3 (1959), pp. 120–131CrossRefGoogle Scholar , esp. 130–1.

62 Roe, ‘Which Security Dilemma’, p. 288, fn. 34. Roe's inclination to rely on (illusory) incompatibility was already apparent in his earlier works (for example, Roe, ‘The Intrastate Security Dilemma’, pp. 187–8; ‘Former Yugoslavia’, pp. 379–80, fn. 5), but he did not develop his thoughts back then.

63 Admittedly, this is partially due to the lack of a good theory of reading intentions then.

64 Jervis, ‘Perception and Misperception’, pp. 75–6.

65 Boulding, ‘National Images’, p. 130; emphasis added.

66 On this critical point, see also Wolfers, Arnold, ‘“National Security” as an Ambiguous Symbol’, Political Science Quarterly, 67 (1952), pp. 481–502CrossRefGoogle Scholar .

67 Of course, if the illusory incompatibility is due to one or both sides' misperception that their security interests are incompatible yet neither side harbours malign intentions toward each other, the situation is a classic security dilemma.

68 Tang, ‘The Security Dilemma’, pp. 605–7.

69 Roe, ‘Actors’ Responsibility’, pp. 106–110; idem, ‘Which Security Dilemma’, pp. 300–11.

70 Ibid., p. 106.

71 See, for example, ‘Which Security Dilemma’, p. 302.

72 Niebuhr, Reinhold, Moral Man and Immoral Society: A Study in Ethics and Politics (New York: Charles Scribner's Sons, 1960 [1932]), p. 42Google Scholar .

73 For a more detailed discussion, see Tang, A Theory of Security Strategy, chap. 1.

74 Roe, ‘“Actors” Responsibility’, pp. 106–11; idem, ‘Which Security Dilemma’, pp. 284–91.

75 Roe had the three types of security dilemma in mind earlier, under a different terminology (that is, resolvable short of war, difficult to resolve short of war; and irresolvable short of war). Roe even claimed that his three types of security dilemma can be linked to Waltz's three level of analysis. He was wise enough to drop this misleading linking with Waltz later. See, Roe, ‘Former Yugoslavia’, pp. 388–9, 391, fn. 18.

76 Snyder, ‘Perception of Security Dilemma’, p. 155.

77 Roe, ‘Which Security Dilemma’, p. 287; see also idem, ‘“Actors” Responsibility’, p. 109.

78 Roe, ‘“Actors” Responsibility’, p. 109; see also Roe, ‘Which Security Dilemma’, p. 288.

79 Ibid., p. 110.

80 Haden, Robert, ‘Constitutional Nationalism in the Former Yugoslav Republics’, Slavic Review, 51 (1992), pp. 654–673CrossRefGoogle Scholar ; Haakan Wiberg, ‘Divided Nations and Divided States as a Security Problem: The Case of Yugoslavia’, Copenhagen Peace Research Institute (COPRI), Working paper, no. 14; Roe, ‘Which Security Dilemma’, pp. 297–301. To argue that Tudjman was an aggressor does not mean that Milosevic was not an aggressor. It is entirely possible that Tudjman was made into an aggressor by Milosevic's rhetoric and behaviour, but this does not nullify the point that Tudjman was an aggressor.

81 ‘Which Security Dilemma’, pp. 297–8.

82 Ibid., p. 305. Emphasis added.

83 Ibid., p. 297.

84 Ibid., p. 302.

85 Fearon labelled his game theoretical approach as ‘rationalist’. I reject this labelling game theoretical as ‘rational choice’ or ‘rationalist’: doing so is to cede moral high ground to it because it implies its critics are irrational. Moreover, ‘rationalist’ and ‘rationalism’ have other meanings in philosophy. For convenience, I retain the less intimidating label of ‘rational choice’. For an earlier critique of the rational choice approach toward war, see Walt, Stephen, ‘Rigor or Rigor Mortis?: Rational Choice and Security Studies’, International Security, 23 (1999), pp. 5–48CrossRefGoogle Scholar .

86 Kaufman misleadingly put the security dilemma under the ‘hard rationalist approach’, as he tried to downgrade the weight of the security dilemma in his own analytical framework. See, Kaufman, Modern Hatred.

87 Elsewhere, I argue in detail why a genuine security dilemma generally does not lead states to war. Tang, A Theory of Security Strategy, chap. 3.

88 Fearon, James, ‘Rationalist Explanations for War’, International Organization, 49 (1995), pp. 379–414CrossRefGoogle Scholar , at p. 385.

89 For a more detailed discussion, see Shiping Tang, ‘Dimensions of Uncertainty and Their Cognitive Challenges: Toward a Better Framework of Attribution in IR’, unpublished manuscript (2010). Fearon certainly tried hard to emphasise that the commitment problem is different from uncertainty about intentions. Fearon, ‘Rationalist Explanations for War’, pp. 401, 406.

90 Fearon, ‘Rationalist Explanations for War’, p. 401. Emphasis added. By treating motivations as equivalent to intentions, Fearon committed another error. See fn. 9 above.

91 See, for example, Lake, David A. and Rothchild, Donald, ‘Containing Fear: The Origins and Management of Ethnic Conflict’, International Security, 21 (1996), pp. 41–75Google Scholar , esp. pp. 52–3; Toft, Monica Duffy, The Geography of Ethnic Violence: Identity, Interests, and Territory (Princeton, NJ.: Princeton University Press, 2003)Google Scholar ; Walter, Barbara, Committing to Peace: The Success Settlement of Civil Wars (Princeton: Princeton University Press, 2001)Google Scholar .

92 For the original formulation of useful fiction versus miracle maker, see MacDonald, Paul K., ‘Useful Fiction or Miracle Maker: The Competing Epistemological Foundations of Rational Choice Theory’, American Political Science Review, 97 (2003), pp. 551–565CrossRefGoogle Scholar .

93 Fearon, James D. and Laitin, David, ‘Explaining Interethnic Cooperation’, American Political Science Review, 90 (1996), pp. 715–735CrossRefGoogle Scholar ; Walter, Committing to Peace. For counterarguments, see Kaufmann (2006), pp. 188–94.

94 For the sake of convenience, I shall focus on the dynamics of elite-led processes, although the security dilemma can also accommodate the dynamics of mass-led processes as well (for example, Azerbaijan, Georgia).

95 Peaceful separation of two groups also requires some measures to contain the security dilemma.

96 Kaufman, ‘Spiraling’, p. 124.

97 Kaufman, ‘A “International” Theory’, p. 155. If the elite in leadership positions harbour genuine hatred, then there is no security dilemma. Of course, it is difficult to know what a leader (or a member of the elite) really thinks in situ, but this is a totally different issue.

98 Kaufman, ‘Spiraling’, pp. 123–5.

99 Gagnon, V. P., ‘Ethnic Nationalism and International Conflict: The Case of Serbia’, International Security, 19 (1994–1995), pp. 130–166Google Scholar . Kosovo, Independent Commission on, The Kosovo Report (Oxford: Oxford University Press, 2000), p. 132Google Scholar .

100 Tang, ‘The Security Dilemma’.

101 Petersen, Roger, Understanding Ethnic Violence: Fear, Hatred, and Resentment in Twentieth-Century Eastern Europe (Cambridge: Cambridge University Press, 2002)Google Scholar . See also Lebow, Richard Ned, A Cultural Theory of International Relations (Cambridge: Cambridge University Press, 2008)Google Scholar .

102 Horowitz, Ethnic Groups in Conflict; Kaufman, ‘Spiraling’, pp. 111, 115–6; idem., Modern Hatred, pp. 25–7, 31–2.

103 Kaplan, Robert D., Balkan Ghosts: A Journey through History (New York: Vintage, 1994)Google Scholar ; Kaufman, Modern Hatred, pp. 25, 30–2; idem., ‘Symbolic Politics or Rational Choice? Testing Theories of Extreme Ethnic Violence’, International Security, 30 (2006), pp. 45–86.

104 I thank an anonymous reviewer for this insight.

105 Toft, The Geography of Ethnic Violence; Jervis, ‘Cooperation under the Security Dilemma’.

106 Tang, ‘The Security Dilemma’. On the role of economic interest in driving ethnic conflict, see Woodward, Susan, Balkan Tragedy (Washington, D.C.: Brookings Institution, 1995)Google Scholar .

107 Snyder, Glenn H., ‘The Security Dilemma in Alliance Politics’, World Politics, 36 (1984), pp. 461–495CrossRefGoogle Scholar .

108 Arfi, Badredine, ‘Ethnic Fear: The Social Construction of Insecurity’, Security Studies, 8 (1998), pp. 151–203Google Scholar ; Mitzen, Jennifer, ‘Ontological Security in World Politics: State Identity and the Security Dilemma’, European Journal of International Relations, 12 (2006), pp. 341–370Google Scholar .

109 Wæver, Ole, ‘Securitization and Desecuritization’, in Lipschtz, Ronnie D. (ed.), On Security (New York: Columbia University, 1995), pp. 46–86Google Scholar .

110 Byman, Keeping the Peace, chap. 5.

111 Ibid., chap. 2. Byman's discussion on security dilemma largely followed Posen, Kaufman, and Fearon. Byman also noticed that many applications of security dilemma to ethnic conflict often vary ‘considerably from the classic international relations concept’. (Ibid., p. 229, n. 1).

112 In Rogers Brubaker's label, Tudjman's Croatia and Ba'athist Iraq were ‘nationalizing states’. Brubaker, Rogers, ‘National Minorities, Nationalizing States, and External National Homelands in the New Europe’, Daedalus, 124 (1996), pp. 107–132Google Scholar .

113 Putnam, Robert D., ‘Diplomacy and Domestic Politics: The Logic of Two-Level Games’, International Organization, 42 (1988), pp. 427–460Google Scholar ; Gourevtich, Peter, ‘The Second Image Reversed: The International Sources of Domestic Politics’, International Organization, 32 (1978), pp. 881–912Google Scholar .

114 The scenario in which the mass in both groups is united in malign intention whereas the elite in both groups is united in benign intentions will be extremely rare and difficult to sustain because the mass will eventually demand a malign leadership. As such, such a situation will quickly change into a situation in which both elite and mass are united in malign intentions. War is almost certain in such a scenario.

115 Roe, Paul, Ethnic Violence and the Societal Security Dilemma (London: Routledge, 2005)Google Scholar .

116 Horowitz, Donald, ‘Making Moderation Pay: The Comparative Politics of Ethnic Conflict Management’, in Montville, Joseph V. (ed.), Conflict and Peacemaking in Multiethnic Societies (Lexington, M. A.: Lexington Books, 1990), pp. 451–475Google Scholar .

117 Mueller, John, ‘The Banality of “Ethnic War”’, International Security, 25 (200), pp. 42–70CrossRefGoogle Scholar .

118 De Figueiredo, Rui J. P. Jr. and Weingast, Barry R., ‘Rationality of Fear: Political Opportunism and Ethnic Conflict’, in Walter, Barbara F. and Snyder, Jack (eds), Civil Wars, Insecurity, and Intervention (New York: Columbia University Press, 1999), pp. 261–302Google Scholar .